Food, Friends and Foes: Estrogens and Social Behaviour in Mice
by
Amy Elizabeth Clipperton Allen
A Thesis presented to The University of Guelph
In partial fulfilment of requirements for the degree of Doctor of Philosophy in Psychology + Neuroscience
Guelph, Ontario, Canada
© Amy E. Clipperton-Allen, December, 2011 ABSTRACT
FOOD, FRIENDS AND FOES: ESTROGENS AND SOCIAL BEHAVIOUR IN MICE
Amy Elizabeth Clipperton Allen Advisor: University of Guelph, 2011 Professor Elena Choleris
This thesis investigates estrogens' modulation of three aspects of social cognition (aggression and agonistic behaviour, social learning, and social recognition). Sex-typical agonistic behaviour (males: overt attacks, females: more subtle dominance behaviours) was increased in gonadectomized mice by estrogen receptor alpha (ERα) agonist 1,3,5-tris(4- hydroxyphenyl)-4-propyl-1H-pyrazole (PPT), while non-overt agonistic behaviour was increased in male and female gonadally intact mice by ERβ agonist 7-Bromo-2-(4- hydroxyphenyl)-1,3-benzoxazol-5-ol (WAY-200070). Estrogens also affected the social transmission of food preferences (STFP). Acute estrogen and ERβ agonists WAY-200070 and 2,3-bis(4-hydroxyphenyl)propionitrile (DPN) prolonged the preference for the demonstrated food when administered pre-acquisition, likely by affecting motivation or the nature of the social interaction, while acute PPT blocked the STFP. All mice receiving any of the three treatments chronically showed a prolonged demonstrated food preference, suggesting a loss of ER specificity. Individual differences in social recognition may relate to increased oxytocin (OT) and vasopressin (AVP) mRNA, and ERα and ERβ gene activation, in the medial preoptic area, and decreased mRNA for ERs, OT receptor (OTR), AVP and AVP receptors 1a and 1b in the lateral amygdala. Additionally, dorsolateral septum ERs, progesterone receptor, and OTR may relate to social interest without affecting social recognition. Our and others' results suggest that estrogens, OT and AVP are all involved in social behaviours and mediate social recognition, social learning, social interactions, and aggression. ERs differently modulate the two types of social learning investigated here: ERα is critical for social recognition, but impairs social learning, while ERβ is less important in social recognition, and prolongs the demonstrated food preference in the STFP. This may be due to differences in receptor brain distributions or in downstream neurochemical systems that mediate these behaviours. The results of this thesis suggest that estrogens, through the various systems they modulate, have a key role to play in social behaviour. Further investigations of how estrogens effect change in these systems at the molecular and cellular level, as well as the critical brain areas and downstream effectors involved in these complex behaviours, are needed, and could contribute to therapeutic interventions in socially-based, sexually dimorphic disorders, like the autism spectrum disorders, and women receiving hormone replacement therapy for negative peri- or post-menopausal symptoms. ACKNOWLEDGMENTS
This thesis would not have been possible without the help and support of many individuals. I am extremely grateful to my wonderful advisor, Dr. Elena Choleris, who has been both extremely supportive and helped me to move beyond my limitations and do so much more than I would have thought possible. Thank you, Elena. I can’t imagine a better advisor, and I will miss you greatly! I also appreciate the input and advice of Elena and my advisory and examination committees, which have made this document immeasurably better. Additionally, thanks go out to all the undergraduate research assistants over the years; your work was extremely helpful and I wish you all the best in your future endeavours. I would also like to thank the NACS graduate students, without whom I would not have survived this process! Your friendship, support, collaboration and encouragement were invaluable, and I feel very lucky that you were my colleagues, to say nothing of the friendship that developed. In particular, I can’t imagine this experience without my good friend and ‘lab-sister’, Anna Phan. I feel so lucky to have worked and traveled with you, and I will miss you terribly. All my family, both near and far, also deserve thanks. I am grateful for your support throughout my education and for your pleasure in my success. In particular: Mum, Dad and Ben, thank you for being there for me whenever I needed you, and for taking pride in my accomplishments. It means a lot, and it’s wonderful to know that the distance doesn’t matter. Julia, having you in Guelph has been fantastic, and I will miss being able to visit my favourite seamstress and have a cup of tea! Granny, your little presents and notes of encouragement always came when I needed them most. Marou-jan, I appreciate your sympathetic ear, and the wonderful times we always have together. We may be farther apart, and the chats may be on the phone instead of in the driveway, but we’ll always be there for each other. Finally, thank you to Craig, my wonderful husband. You calm me down and help me through every obstacle, and I couldn’t imagine my life without you. I love you. I could not have come this far without the help, support and encouragement of all the above (and likely many more!). I appreciate everything you’ve done, and I hope you realize how important you’ve been, not only in my education, but also in my life. Thank you, and lots of love to you all.
iii TABLE OF CONTENTS
Abstract...... ii
Acknowledgments...... iii
Table of Contents...... iv
List of Abbreviations...... xii
List of Tables ...... xiv
Chapter 1: General Introduction ...... 1
1.1. Social Cognition...... 2
1.2. Estrogens...... 7
1.2.1. Estrogens: What They Are ...... 7
1.2.2. Estrogens: Effects on Non-social Learning...... 11
1.3. Social Interaction and Aggression...... 12
1.3.1. Social Interaction and Aggression: Introduction ...... 12
1.3.2. Social Interaction and Aggression: Effects of Estrogens...... 16
1.4. Social Learning of Food Preferences...... 20
1.4.1. Social Learning of Food Preferences: Introduction...... 20
1.4.2. Social Learning: Effects of Estrogens...... 24
1.5. Social Recognition...... 25
1.5.1. Social Recognition: Introduction...... 25
1.5.2. Social Recognition: Effects of Estrogens...... 31
1.6. Aims...... 34
Part I:
Chapter 2: Effects of an estrogen receptor agonist on agonistic behaviour in intact and gonadectomized male and female mice...... 37
2.1 Abstract...... 38
2.2 Introduction...... 40 iv 2.3 Methods...... 44
2.3.1 Animal Housing...... 44
2.3.2 Experimental Subjects...... 44
2.3.3 Apparatus...... 45
2.3.4 Drug ...... 47
2.3.5 Procedure...... 48
2.3.6 Statistical Analysis...... 52
2.4 Results...... 54
2.4.1 Sex Differences in Gonadally Intact Mice ...... 56
2.4.2 Differences Between Gonadally Intact and Castrated Males...... 56
2.4.3 Differences Between Gonadally Intact and Ovariectomized Females...... 58
2.4.4 Effects of PPT in Gonadally Intact Males...... 58
2.4.6 Estrous Cycle Effects...... 60
2.4.7 Effects of PPT in Castrated Males...... 60
2.4.8 Effects of PPT in Ovariectomized Females ...... 60
2.5 Discussion ...... 62
2.5.1 Sex and Gonadectomy Differences...... 62
2.5.2 PPT Effects in Gonadectomized Mice...... 64
2.5.3 PPT Effects in Gonadally Intact Mice ...... 65
2.5.4 ERα and ERβ in Mouse Agonistic Behaviour...... 66
2.5.5 Vehicle Effects...... 68
2.5.6 Conclusions ...... 69
2.6 Appendix 1: Supplementary Material...... 71
Chapter 3: Agonistic behaviour in males and females: Effects of an estrogen receptor beta agonist in gonadectomized and gonadally intact mice...... 87
3.1 Abstract...... 88 v 3.2 Introduction...... 89
3.3 Method...... 93
3.3.1 Animal Housing...... 93
3.3.2 Experimental Subjects...... 94
3.3.3 Apparatus...... 95
3.3.4 Drug ...... 95
3.3.5 Procedure...... 98
3.3.6 Statistical Analysis...... 102
3.4 Results...... 104
3.4.1 Effects of WAY-200070 in Gonadally Intact Males ...... 104
3.4.2 Effects of WAY-200070 in Gonadally Intact Females...... 104
3.4.3 Effects of WAY-200070 in Castrated Males...... 106
3.4.4 Effects of WAY-200070 in Ovariectomized Females ...... 108
3.4.5 Sex Differences in Uninjected, Gonadally Intact Mice ...... 110
3.4.6 Differences Between Gonadally Intact and Castrated Uninjected Males...... 110
3.4.7 Differences Between Gonadally Intact and Ovariectomized Uninjected Females ...... 112
3.5 Discussion ...... 112
3.5.1 WAY-200070 Effects ...... 113
3.5.2 Sex and Gonadectomy Differences...... 116
3.5.3 Note on Vehicle Effects ...... 118
3.5.4 Conclusions ...... 119
3.6. Appendix 2: Differences in behaviour between sexes and gonadal conditions, and effects of WAY-200070 treatment on behaviour...... 121
Part II:
Chapter 4: Acute effects of estradiol benzoate on the social transmission of food preferences in female mice...... 136
vi 4.1 Abstract...... 137
4.2 Introduction...... 139
4.3 Method...... 142
4.3.1 Subjects ...... 142
4.3.2 Diets ...... 144
4.3.3 Apparatus...... 144
4.3.4 Hormone Treatments...... 145
4.3.5 Procedure...... 145
4.3.5.1 Phase 1 (DEM Feeding and Social Interaction)...... 145
4.3.5.2 Phase 2 (Delay) ...... 147
4.3.5.3 Phase 3 (Choice Test)...... 147
4.3.6 Behavioural Analysis ...... 148
4.3.7 Statistical Analyses ...... 148
4.3.7.1 Food preferences and food consumption...... 148
4.3.7.2 Behavioural analysis...... 149
4.4 Results...... 150
4.4.1 Experiment 1: Pre-acquisition EB Administration ...... 151
4.4.1.1 OBS food preference...... 151
4.4.1.2 Behavioural analysis...... 151
4.4.2 Experiment 2: Immediate Post-acquisition EB Administration ...... 154
4.4.3 Experiment 3: Delayed Post-acquisition EB Administration...... 156
4.4.4 EB Effects on Food Intake ...... 156
4.5 Discussion ...... 158
4.5.1 Conclusions ...... 162
4.6 Appendix 3: Effects of EB on Social Interactions in Experiment 1...... 164
vii Chapter 5: The involvement of estrogen receptors alpha and beta in the memory for a socially acquired food preference in female mice ...... 170
5.1 Abstract...... 171
5.2 Introduction...... 173
5.3 Methods...... 176
5.3.1 Subjects ...... 176
5.3.2 Surgery ...... 176
5.3.4 Apparatus...... 177
5.3.5 Drugs...... 178
5.3.6 General Procedure...... 179
5.3.6.1 Phase 1 (DEM Feeding and Social Interaction)...... 179
5.3.6.2 Phase 2 (Delay) ...... 179
5.3.6.3 Phase 3 (Choice Test)...... 182
5.3.7.1 Pre-acquisition DPN (Experiment 1) ...... 182
5.3.7.3 Immediate Post-acquisition Administration of PPT and DPN (Experiment 2a and 2b)...... 183
5.3.7.4 Immediate Post-acquisition Individual Learning Control Study (Experiment 2c) ...... 183
5.3.7.5 24h Delayed Post-acquisition Administration of PPT and DPN (Experiment 3a and 3b)...... 184
5.3.8 Statistical Analyses ...... 184
5.4 Results...... 187
5.4.1 Experiment 1: Pre-acquisition DPN...... 187
5.4.1.1 Food preferences...... 187
5.4.1.2 Behavioural analysis...... 188
5.4.3 Experiment 2b: Immediate Post-acquisition Administration of DPN ...... 191
5.4.4 Experiment 2c: Immediate Post-acquisition Control Study...... 194
viii 5.4.5 Experiment 3a: 24h Delayed Post-acquisition Administration of PPT...... 194
5.4.6 Experiment 3b: 24h Delayed Post-acquisition Administration of DPN ...... 194
5.4.7 Effects on Food Intake ...... 196
5.5 Discussion ...... 199
5.5.1 Effects of PPT on Social Learning ...... 199
5.5.2 Effects of DPN on Social Learning ...... 202
5.5.2.1 Pre-acquisition ERβ Agonists...... 202
5.5.2.2 Post-acquisition DPN Effects ...... 203
5.5.3 Overall Conclusions...... 206
Chapter 6: Effects of chronic estrogen and estrogen receptor agonists on the behaviour of ovariectomized female mice in the social transmission of food preferences paradigm...... 208
6.1 Abstract...... 209
6.2 Introduction...... 210
6.3 Methods...... 213
6.3.1 Subjects ...... 213
6.3.2 Surgery ...... 215
6.3.3 Drugs...... 215
6.3.3 Diets ...... 216
6.3.4 Apparatus...... 217
6.3.5 Procedure...... 217
6.3.6 Histomorphometric Analysis...... 218
6.3.8 Statistical Analyses ...... 221
6.3.8.1 Cinnamon Preference ...... 221
6.3.8.2 Behaviour analysis ...... 222
6.3.8.3 Histomorphometric Analysis ...... 222
6.4 Results...... 223 ix 6.4.1 Treatment Effects on Food Preference...... 223
6.4.1.1 Chronic EB ...... 223
6.4.1.3 Chronic PPT...... 223
6.4.1.3 Chronic DPN...... 225
6.4.2.1 Chronic EB ...... 228
6.4.2.2 Chronic PPT...... 230
6.4.2 Chronic DPN ...... 232
6.4.3 Feeding Effects...... 234
6.4.4 Vaginal Cytology...... 234
6.4.5 Chronic PPT and DPN Effects on Uterine Histomorphology...... 235
6.5 Discussion ...... 235
6.5.1 Overall Conclusions...... 239
Part III:
Chapter 7: Oxytocin, vasopressin and estrogen receptor gene expression in relation to social recognition in female mice ...... 240
7.1 Abstract...... 241
7.2 Introduction...... 242
7.3 Materials and Methods...... 245
7.3.1 Subjects ...... 245
7.3.2 Surgery ...... 246
7.3.3 Apparatus...... 248
7.3.4 Behavioural Testing Procedure...... 249
7.3.5 Quantitative Real Time Reverse Transcription Polymerase Chain Reaction...... 250
7.3.6 Statistical Analysis...... 253
7.3.6.1 Behavioural analysis...... 254
x 7.3.6.2 Molecular analysis...... 255
7.3.6.3 Behaviour/molecular correlations...... 255
7.4 Results...... 255
7.4.1 Behavioural Analysis ...... 255
7.4.2. Molecular Analysis ...... 258
7.4.3 Behaviour/Molecular Correlations ...... 258
7.4.3.1. Lateral amygdala...... 258
7.4.3.2. Medial preoptic area...... 258
7.4.3.3. Dorsolateral Septum...... 263
7.5 Discussion ...... 263
7.5.1 Lateral Amygdala ...... 263
7.5.2 Medial Preoptic Area ...... 265
7.5.3 Dorsolateral Septum...... 267
7.5.4 Overall Conclusions...... 268
Chapter 8: General Discussion and Conclusions...... 269
8.1 Summary of Social Interaction and Aggression...... 270
8.2 Summary of Social Learning of Food Preferences...... 275
8.3 Summary of Social Recognition...... 278
8.4 Estrogens and Social Cognition ...... 279
8.5 Significance and Applications...... 281
Reference List...... 285
xi LIST OF ABBREVIATIONS
5HT: serotonin
αERKO: mice in which the gene for ERα is inactivated (ERα knockout mice)
αβERKO: mice in which the genes for ERα and ERβ are inactivated (ERαβ knockout mice)
βERKO: mice in which the gene for ERβ is inactivated (ERβ knockout mice)
AMPA: alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid
ASD: autism spectrum disorders
AVP: vasopressin
AVPR1a: vasopressin receptor 1a
AVPR1b: vasopressin receptor 1b
BNST: bed nucleus of the stria terminalis
CDP: chlordiazepoxide
CS2: carbon disulphide
DA: dopamine
DATKO: mice in which the gene for DA transporter is inactivated (DA transporter knockout mice)
DEM: demonstrator mouse in the STFP paradigm
DHT: dihydrotestosterone
DNA: deoxyribonucleic acid
DPN: ERβ agonist 2,3-bis(4-hydroxyphenyl)propionitrile
ER: estrogen receptor
ERα: estrogen receptor alpha
ERβ: estrogen receptor beta gonadex: gonadectomized
GPER: G-protein-coupled estrogen receptor, aka GPR30, mER
GPR30: G-protein-coupled receptor 30, aka GPER xii HR: high recognizers; mice that had a preference for the novel mouse of >85% in the social discrimination test
HRT: hormone replacement therapy icv: intracerebroventricular infusion
KO: knockout
LR: low recognizers; mice that had no preference for the novel mouse in the social discrimination test (chance performance)
MPOA: medial preoptic area
NMDA: N-methyl-D-aspartate
OBS: observer mouse in the STFP paradigm obx: olfactory bulbectomized
OT: oxytocin
OTR: oxytocin receptor ovx: ovariectomized
PPT: ERα agonist 1,3,5-tris(4-hydroxyphenyl)-4-propyl-1H-pyrazole
PVN: paraventricular nucleus of the hypothalamus
STFP: social transmission of food preferences
T: testosterone
VMH: ventromedial hypothalamus
WAY-200070: ERβ agonist 7-Bromo-2-(4-hydroxyphenyl)-1,3-benzoxazol-5-ol
WHI: Women's Health Initiative
WT: wild type
xiii LIST OF TABLES
Chapter 1
Table 1: Effects of estrogen receptor manipulations on social recognition...... 32
Chapter 2
Table 2: Distribution of treatment groups...... 46
Table 3: Descriptions of scored behaviours...... 50
Table 4: Descriptions of grouped and composite behaviours...... 51
Table 5: Overview of PPT effects...... 55
Table 6 (Appendix 1): Differences between gonadally intact male and female uninjected mice...... 71
Table 7 (Appendix 1): Differences between gonadally intact and castrated male uninjected mice...... 73
Table 8 (Appendix 1): Differences between gonadally intact and ovariectomized female uninjected mice...... 75
Table 9 (Appendix 1): Effects of PPT in gonadally intact males ...... 76
Table 10 (Appendix 1): Effects of PPT in gonadally intact females...... 77
Table 11 (Appendix 1): Effects of the estrus cycle in gonadally intact female mice...... 79
Table 12 (Appendix 1): Effects of PPT in castrated males...... 83
Table 13 (Appendix 1): Effects of PPT in ovariectomized females ...... 85
Chapter 3
Table 14: Distribution of treatment groups ...... 96
Table 15: Description of scored behaviours ...... 100
Table 16: Descriptions of grouped and composite behaviours...... 101
Table 17 (Appendix 2): Effects of WAY-200070 in gonadally intact males ...... 121
Table 18 (Appendix 2): Effects of WAY-200070 in gonadally intact females ...... 123
Table 19 (Appendix 2): Effects of WAY-200070 in castrated males ...... 126
xiv Table 20 (Appendix 2): Effects of WAY-200070 in ovariectomized females ...... 130
Table 21 (Appendix 2): Differences between gonadally intact male and female uninjected mice...... 131
Table 22 (Appendix 2): Differences between gonadally intact and castrated male uninjected mice...... 133
Table 23 (Appendix 2): Differences between gonadally intact and ovariectomized uninjected mice...... 135
Chapter 4
Table 24: Number of observer mice in each treatment group...... 146
Table 25: Behaviours collected during the social interaction...... 168
Chapter 5
Table 26: Distribution of treatment groups ...... 180
Table 27: Description of selected performed and calculated behaviours in DPN pre- acquisition experiment ...... 186
Chapter 6
Table 28: Distribution of treatment groups ...... 214
Table 29: Descriptions of behaviours during the social interaction...... 219
Table 30: Effects of chronic PPT and DPN on uterine histomorphology...... 236
Chapter 7
Table 31: Quantity of total RNA used in reverse transcription reactions by brain region...... 247
Table 32: Descriptions of behaviours collected during behavioural testing...... 251
Table 33: Differences in cycle thresholds between high recognition and low recognition mice by brain region...... 258
Table 34: Effects of manipulations on aggression...... 272
Table 35: Effects of estrogens and estrogen receptor specific agonists on social transmission of food preferences...... 276
xv LIST OF FIGURES
Chapter 1
Figure 1: Steroidogenesis of estradiol from cholesterol ...... 8
Chapter 2
Figure 2: Sex differences and gonadectomy effects on agonistic behaviours...... 57
Figure 3: PPT effects on gonadally intact and castrated males ...... 59
Figure 4: PPT effects on gonadally intact and ovariectomized (ovx) females...... 61
Chapter 3
Figure 5: WAY-200070 effects on gonadally intact males...... 105
Figure 6: Effects of WAY-200070 on gonadally intact females ...... 107
Figure 7: WAY-200070 effects on gonadectomized male and female mice...... 109
Figure 8: Sex differences and gonadectomy effects on agonistic behaviour ...... 111
Chapter 4
Figure 9: Timeline of pre-acquisition, immediate post-acquisition and 24h delayed post- acquisition paradigms...... 143
Figure 10: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the pre-acquisition EB experiment...... 152
Figure 11: Effects of acute pre-acquisition EB on behaviour during the social interaction ...... 153
Figure 12: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a coca or cinnamon diet in the immediate post-acquisition EB experiment...... 155
Figure 13: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon diet in the 24h post-acquisition EB experiment ...... 157
Figure 14 (Appendix 3): Total food intake (g) by 2h intervals over the 8h choice test...... 167
Chapter 5
Figure 15: Timeline of pre-, immediate post- and 24h delayed post-acquisition paradigms.....181
Figure 16: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon diet in the pre-acquisition DPN experiment...... 189
xvi Figure 17: Effects of DPN on behaviours in the pre-acquisition DPN experiment...... 190
Figure 18: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the immediate post-acquisition PPT experiment ...... 192
Figure 19: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the immediate post-acquisition DPN experiment ...... 193
Figure 20: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the 24h delayed post-acquisition PPT experiment ...... 195
Figure 21: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the 24h delayed post-acquisition DPN experiment ...... 197
Figure 22: Total food intake (g) by 2h intervals over the 8h choice test ...... 198
Chapter 6
Figure 23: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the chronic EB experiment ...... 224
Figure 24: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the chronic PPT experiment...... 226
Figure 25: Cinnamon preference by 2h intervals in observers whose demonstrators had been fed either a cocoa or cinnamon flavoured diet in the chronic DPN experiment...... 227
Chapter 7
Figure 26: Behavioural differences between high and low recognizing mice ...... 257
Figure 27: Levels of mRNA for genes in high and low recognizing mice, relative to the average amount of gene product in HR animals (set at 1) ...... 260
Figure 28: Correlations between mRNA and behaviour in the dorsolateral septum and lateral amygdala ...... 261
Figure 29: Correlations between mRNA and behaviour in the medial preoptic area...... 262
xvii
CHAPTER 1: GENERAL INTRODUCTION
1 This thesis focuses on social cognition and its modulation by hormones. Specifically, I investigated how the gonadal steroids estrogens, the nonapeptides oxytocin (OT) and vasopressin (AVP), acting via their receptors, affected three aspects of social cognition: aggression and agonistic behaviour (as tested with the resident-intruder test [Ch 2-3]); social learning (as tested with the social transmission of food preferences (STFP) paradigm [Ch 4-
6]); and social recognition (as tested with the social discrimination task [Ch 7]). Numerous social behaviours have been previously shown to be influenced by estrogens, OT and/or AVP
(see Choleris et al., 2009; Choleris et al., 2008a for recent reviews), but a number of aspects of social cognition and social behaviours, and how they are affected by these hormones, required further elucidation.
1.1. Social Cognition
Mammals range in sociality from solitary animals, which are intolerant of conspecifics except where needed for reproduction (i.e., mating, and offspring care in females), to highly social animals that share territories and/or home ranges and live in groups (Blumstein et al.,
2010; Ebensperger, 2001); this variation in sociality may relate to different circulating levels of steroids, including gonadal hormones (e.g., testosterone, “T”) and/or glucocorticoids, variation in neurotransmitter and hormone receptor distribution, expression and density (e.g., oxytocin, “OT”), and/or steroid-binding globulin proteins that may facilitate or impede steroids reaching target tissues (reviewed in Blumstein et al., 2010). While they range along the spectrum from social to solitary (Ebensperger, 2001), rodents, including mice and rats, generally fall into the former group (Latham & Mason, 2004), although there is variability in sociality between individuals as well as between species (Young, 2009).
2 There are both costs and benefits associated with living in social groups (see
Ebensperger, 2001 for a review). Some potential costs are increased competition for resources, including food, territories, paternal care, and breeding rights or mates (Alexander,
1974), as well as higher stress levels associated with determining and maintaining a dominance hierarchy in some species (Creel, 2001). Other costs can include increased visibility resulting in increased predation (Alexander, 1974), and an increased rate of parasitic infections due to close contact with potentially infected conspecifics (Kavaliers et al., 2004). However, in those species that do live socially, the benefits of group life likely outweigh these and other costs (Alexander, 1974; Blumstein et al., 2010; Ebensperger, 2001).
Benefits of group living can include better defence of resources, including territory and food, as well as improved defence against predators (Ebensperger, 2001). Predator risks can also be reduced by increased vigilance (because there are more eyes watching for them), as well as an increased number of targets and a potentially safer location at the centre of the group (Alexander, 1974; Ebensperger, 2001). Similarly, whether because of more successful hunting with cooperative group techniques (e.g., lions, wolves), or an increased number of group members seeking out scattered large food caches (e.g., bees), groups may find it easier to successfully acquire food (Alexander, 1974; Rieucau & Giraldeau, 2011). Alternatively, group living may be necessary because critical resources, such as safe sleeping sites or appropriate breeding locations, may be rare and/or concentrated in certain areas (Alexander,
1974), or because the energy costs of solitary living are too high (e.g., digging burrows, thermoregulation; Ebensperger, 2001).
Another significant advantage of social living is that one may learn from conspecifics.
3 When individually learned information is considered unreliable or out-of-date, or when it is too risky or costly to obtain (e.g., sampling potentially toxic new foods or choosing a mate of unknown quality), often animals will behave according to socially acquired knowledge, if it is available (reviewed in Rieucau & Giraldeau, 2011). Costs of individual learning can include increased exposure and vulnerability to predation and/or loss of energy while foraging or evaluating breeding or sleeping sites (Rieucau & Giraldeau, 2011), as well as increased risk of infection and likelihood of toxic substance ingestion (Choleris et al., 2009).
Group members can socially acquire a wide range of information, from nesting sites and migratory routes, to mate choices and tool use, to feeding sites, behavioural strategies and food preferences (Galef & Laland, 2005; Russon, 1997).
In addition to attending to and/or learning from the social environment (social learning), there are other social cognitive abilities that are necessary for or involved in group living.
One of the most crucial of these, especially when dominance hierarchies must be established and determine the nature of intragroup interactions, is the recognition and/or identification of specific others, including their status and emotional state. This so-called social recognition can range from true individual recognition to simple species recognition, and is critical for the development of social bonds, like those between mated pairs or of a mother to her offspring. Additionally, social recognition can facilitate the avoidance of group members that are infected or carry parasites (Choleris et al., 2004), thus reducing the risk of contagion associated with group living (reviewed in Kavaliers et al., 2004). Furthermore, social recognition of varying degrees is essential for determining if other individuals are related and/or familiar, potential mates (i.e., adults of the opposite sex), and/or sexually receptive
(e.g., females in behavioural estrus). It can also allow one to distinguish between the animal
4 from which one can take resources (lower hierarchical status) and the animal to which one should relinquish resources (higher status), those that are healthy and those that are not, and mates or group members and strangers or intruders (reviewed in Choleris et al., 2009).
Agonistic behaviour includes multiple aspects of the aggressive encounter, including the preliminary behaviour (e.g., tail rattle), chasing, the attacks themselves, defensive behaviour, and other behaviours (e.g., aggressive grooming, pinning) that are involved in the establishment of a dominance hierarchy (Grant & Mackintosh, 1963; Scott & Fredericson,
1951). In animals that share territory, the nature of agonistic interactions with conspecifics is often governed by social recognition. For example, social or familiarity recognition is used to determine the appropriate response to a conspecific, with tolerance of familiar animals, but aggression towards and exclusion of unfamiliar animals (Miczek et al., 2001; Miczek et al.,
2007). Although aggression is typically high when a new social group is initially formed, mice, especially females, will usually settle into a hierarchy within a few days (Creel, 2001;
Uhrich, 1938). As dominance is established, often through repeated agonistic encounters between the same individuals, attacks decline (Scott & Fredericson, 1951), and the high risks of injury and energy costs associated with fighting are avoided (Enquist & Leimar, 1990).
Once established, maintenance of these dominance hierarchies tends to consist of threats and aggression in high ranking animals, and submissive, defensive and evasive behaviours by lower-ranked individuals (Miczek et al., 2007), and both dominant and subordinate mice may be less stressed and anxious in stable than unstable hierarchies, especially if the stable groups consist of siblings (Bartolomucci et al., 2001; Bartolomucci et al., 2002; Vekovishcheva et al., 2000). However, both males and females in stable hierarchies will attack strangers introduced into the group cage (Uhrich, 1938).
5 Social cognition, whether social recognition, social learning, communication, or choosing appropriate social interactions, is important for successful social living. Elucidation of the underlying neurobiological and genetic mechanisms by which it is mediated is of particular importance given the predominance of social living in our own species (Morgan et al., 2011).
While some research on these mechanisms has been done, there is still a great need for further investigations. Sex differences have been observed in a number of social cognitive tasks, and sex hormones are also known to affect social cognition (e.g., Choleris & Kavaliers,
1999; Fugger et al., 1998; Sánchez-Andrade & Kendrick, 2011; Temple et al., 2001; see
Choleris et al., 2009; Choleris et al., 2008a; Choleris et al., 2011c; Gabor et al., 2011 for recent reviews). Further, many of the neurotransmitters and neurochemical systems that mediate social cognition and behaviour can be affected by estrogens, whether through increases or decreases in transcription or by other means, including epigenetic mechanisms
(e.g., histone modifications; Choleris et al., 2009; Choleris et al., 2011c; Gabor et al., 2011;
Jacome et al., 2010b; Johnson et al., 2010; Morissette et al., 2008). These neurotransmitters include serotonin (5HT; de Almeida et al., 2001; de Almeida & Miczek, 2002; De Boer et al.,
1999; De Boer & Koolhaas, 2005; Fish et al., 1999; Holmes et al., 2002a; Kantak et al.,
1980; Kim et al., 2005; Lasley & Thurmond, 1985; Miczek et al., 1998; Olivier et al., 1990;
Saudou et al., 1994), dopamine (DA; Choleris et al., 2011a; de Almeida et al., 2005; Miczek et al., 2002; Rodriguiz et al., 2004), OT and AVP (reviewed in Choleris et al., 2009; Dore et al., 2011; Gabor et al., 2011). This suggests that estrogens may modulate or control social cognition and social behaviour in general.
6 1.2. Estrogens
Gonadal hormones, including estrogens, are known to be involved in social behaviour and social cognition. Aside from their well-characterized role in reproductive behaviour, estrogens have been implicated in aggression, social recognition, and social learning, as well as other, non-social, cognitive tasks.
1.2.1. Estrogens: What They Are
Estrogens are a class of steroid hormones, of which 17β-estradiol is the most common.
Along with progesterone, the other main female gonadal steroid, estradiol is produced primarily in the ovaries. Like all steroid hormones, estrogens have the characteristic three six-carbon ring plus one conjugated five-carbon ring chemical structure (Nelson, 2005).
Physiological estrogens are derived from cholesterol, which is metabolized to pregnenolone via the cytochrome P450 side-chain cleavage enzyme (cytochrome P450scc; Steimer, 1993).
Pregnenolone is then converted to 4-androstenedione, by way of either progesterone or 17α- hydroxypregnenolone (see Figure 1, modified from Steimer, 1993). Synthesis of 17β- estradiol from 4-androstenedione can occur either through the aromatization of 4- androstenedione to estrone, which is reversibly converted to estradiol by 17β-hydroxysteroid dehydrogenase, or by aromatization (by cytochrome P450 19, also known as aromatase) of T to estradiol following the reversible conversion of 4-androstenedione to T by 17β- hydroxysteroid dehydrogenase (see Figure 1; Rodgers, 1990; Steimer, 1993).
Estrogens were initially thought to bind to a single estrogen receptor (ER), but the discovery of a second ER in the mid-1990s (Kuiper et al., 1996) opened the door to the possibility of further receptors. Estrogens are now known to bind to two classic intranuclear
7
Figure 1. Steroidogenesis of estradiol from cholesterol, modified from Steimer (1993). 8 receptors, alpha (ERα) and beta (ERβ), and activation of these receptors can have genomic effects taking place in the hours-to-days timeframe. Estrogens can also act through membrane-bonded versions of these receptors, as well as membrane-specific receptors like the recently described G-protein-coupled estrogen receptor (GPER, formerly known as
GPR30 or mER; Filardo & Prossnitz, 2009; Prossnitz & Barton, 2009). These actions often take place rapidly (e.g., seconds to hours), and can affect both dendritic spines and learning and memory (Phan et al., 2011).
Steroid hormones typically exert their genomic effects by binding to cytoplasmic or nuclear receptors, which dimerize, and the resulting complex then acts as a transcription factor by binding to hormone response elements in gene promoters and then activating or suppressing gene transcription (London & Clayton, 2010; Nelson, 2005). ERα and ERβ can form both homo- and heterodimers when they are co-expressed, and the impact of each receptor may be partly mediated by differences in the properties of the homo- and heterodimers formed (Hall & McDonnell, 1999). ERα is typically a stronger transcriptional activator than ERβ (Cowley et al., 1997), especially when saturating or subsaturating levels of estradiol are present (Hall & McDonnell, 1999), and may be activated more efficiently by
17β-estradiol, as ERβ requires 1.5 times as much ligand for equivalent activation (Hall &
McDonnell, 1999). However, ERβ can suppress the potency and/or efficiency of ERα's response to estrogens in the presence of physiological levels of estrogens, such that coexpression of both receptors leads to intermediate levels of transcription (Hall &
McDonnell, 1999). ERβ also decreases the sensitivity of ERα cells to estradiol, and heterodimers have been found to act in opposition to ERα homodimers (Hall & McDonnell,
1999). There is also evidence from the KO literature that suggests that the two receptors may
9 regulate transcription of each other reciprocally in vivo (e.g., Couse et al., 1997; Nomura et al., 2003; Temple et al., 2001), as, for example, ERβ can act as a transdominant repressor of
ERα transcription (Hall & McDonnell, 1999).
Nuclear steroid hormone receptors, like ERα and ERβ, typically have three major regions, which include the steroid hormone binding C-terminal domain, the central domain that binds to deoxyribonucleic acid (DNA), and the N-terminal domain, which affects transcriptional activation by interacting with other DNA-binding proteins (Nelson, 2005).
Alpha and beta ERs are encoded by similar but different genes, and share approximately 95% of their DNA-binding domain and approximately 55% of their C-terminal ligand binding domain (Kuiper et al., 1996). More specifically, only two amino acids differ in the ligand- binding cavities on the two ERs, although the remainder of the binding domains only share
56% of their sequence (Choleris et al., 2008a).
The two receptors overlap in some brain regions in mice and rats (e.g., bed nucleus of the stria terminalis “BNST”, medial amygdala), but show largely distinct brain distributions; for example, the ventromedial hypothalamus (VMH) and arcuate nucleus contain predominantly
ERα, and ERβ is more prevalent in the suprachiasmatic and supraoptic nuclei, the paraventricular nucleus of the hypothalamus (PVN), the hippocampus and the cortex (Handa et al., 2011; Merchenthaler et al., 2004b; Mitra et al., 2003b). They are also expressed differently during development (Shugrue et al., 1997), and it has been suggested that ERα is responsible for the masculinization, and ERβ for the defeminization, of males (Bodo &
Rissman, 2006). The roles of the two receptors are varied, sometimes working in the same direction (e.g., social recognition) and sometimes having opposing effects (e.g., aggression)
10 on physiology and behaviour (reviewed in Choleris et al., 2008a; Gustafsson, 2006a).
1.2.2. Estrogens: Effects on Non-social Learning
In addition to affecting reproductive behaviours and physiology, estrogens have been the subject of numerous investigations of learning and memory. These have led to largely inconsistent results, ranging from enhancements to no effects to impairments of performance on cognitive tasks (reviewed in Choleris et al., 2008a; Daniel, 2006). For example, estradiol administration to ovariectomized (ovx) mice and/or rats has been shown to impair (Chesler &
Juraska, 2000; Daniel & Lee, 2004; Frick et al., 2004; Frye, 1995; Fugger et al., 1998;
Gresack et al., 2007), improve (Bowman et al., 2002; Daniel et al., 2006a; Daniel et al.,
1997; Heikkinen et al., 2004; Heikkinen et al., 2002; Luine et al., 1998; Markham et al.,
2002; Rhodes & Frye, 2006; Sandstrom & Williams, 2004) or have no effect on (Luine et al.,
1998; Luine & Rodriguez, 1994) performance on the Morris water maze and radial arm maze spatial learning tasks. A number of possible explanations have been suggested for these diverse effects, including the use of different strategies in different learning tasks (e.g., Korol et al., 2004; Korol, 2004), different treatment administration paradigms (e.g., chronic versus acute; Luine et al., 1998), different species, the involvement of different learning systems
(e.g., Davis et al., 2005), opposing effects of different receptors (e.g., Clipperton et al., 2008;
Rissman et al., 2002a), and genomic vs non-genomic actions (Phan et al., 2011).
The involvement of the two receptors in cognitive tasks has been investigated using both mice in which the gene for ERα or ERβ was inactivated or “knocked out” (KO; αERKO and
βERKO mice, respectively), and compounds which specifically target one receptor over the other (i.e., ER specific agonists). Findings that estradiol administration to βERKO but not
11 αERKO mice impaired Morris water maze performance (Fugger et al., 1998; Rissman et al.,
2002a), coupled with observed improvements of performance on this task by administration of both estrogens and ERβ selective agonists, but not ERα agonists (Rhodes & Frye, 2006), suggests that estradiol's improving effects may be largely through actions at ERβ, with its impairing effects being more controlled by ERα (Rissman et al., 2002a).
1.3. Social Interaction and Aggression
1.3.1. Social Interaction and Aggression: Introduction
As mice are often housed together in group cages in the laboratory, and will live socially in the wild (Latham & Mason, 2004), they are a good model species for the study of the neurobiology of affiliative and agonistic behaviour and of social interactions in general, although strain differences in agonistic behaviour must be considered. Aggression is one of the most commonly studied types of non-reproductive social interaction. Excessive aggression can have significant negative consequences in our own species, and knowledge of its underlying neurobiology could prove very beneficial. Laboratory mice and rodents have been used to study aggression since the early 1920s (e.g., Craig, 1921; Howard, 1920), and these studies have typically focused on measures of overt aggressive behaviour, such as the latency to or frequency of attack, which are aimed at excluding others from one's territory
(e.g., Beeman, 1947; Ginsburg & Allee, 1942; Nomura et al., 2002; Ogawa et al., 1998b).
However, some researchers have also looked at other, more subtle agonistic behaviours aimed at establishing a dominance hierarchy amongst group members in a shared territory
(e.g., Alleva, 1993; Branchi et al., 2006; Clipperton et al., 2008; Clipperton-Allen et al.,
2010; Clipperton-Allen et al., 2011a; Grant & Mackintosh, 1963; Miczek et al., 2001; Mos &
Olivier, 1989; Olivier et al., 1989; Pietropaolo et al., 2004).
12 The type and severity of these aggressive and agonistic interactions can be affected by a number of factors. Older and/or larger mice tend to be dominant (Uhrich, 1938), and, in what is known as the prior residency effect (Archer, 1988), an animal is more likely to win agonistic encounters and/or become dominant if tested in its home cage against an intruder
(Uhrich, 1938). Mice will also respond to the behaviour of their opponents (Ginsburg &
Allee, 1942). The likelihood and intensity of aggression can also be increased by isolation, pain, or housing with the opposite sex (Albert et al., 1989; Ginsburg & Allee, 1942; Scott &
Fredericson, 1951). Additionally, more aggressive males may be more anxious than less aggressive males (Ferrari et al., 1998; Guillot & Chapouthier, 1996), and benzodiazepines can reduce both anxiety and aggression in these animals (Parmigiani et al., 1999).
One of the most common tasks used to assess aggression is the resident-intruder paradigm, in which an “intruder” animal is introduced into the home cage of a “resident” conspecific and their interactions are observed and scored. The intruders are frequently group-housed, gonadectomized (gonadex) and/or olfactory-bulbectomized (obx) conspecifics, which show little or no aggression (Beeman, 1947; Ginsburg & Allee, 1942;
Ogawa et al., 1998b). This makes for more consistent intruders and thus better assessment of the aggression level of the experimental animals, which are typically gonadally intact and singly-housed.
The majority of aggression research has used male subjects (Parmigiani et al., 1998), as it has long been claimed that females are not agonistic and do not fight (e.g., Ginsburg & Allee,
1942; Scott & Fredericson, 1951). An exception to this is so-called maternal or postpartum aggression, which occurs when females are protecting their offspring (e.g., Palanza et al.,
13 1995), and is qualitatively different from the territorial aggression typically studied in males, being less ritualized and more violent (Parmigiani et al., 1998).
However, females do also show agonistic behaviour even when not defending their litters; in one study, 50% of female CD1 mice that were primed by exposure to male odours attacked female intruders (Parmigiani et al., 1999), and administration of T to females also stimulates aggression (e.g., Albert et al., 1990; Ogawa et al., 2004; Pibiri et al., 2006). These results notwithstanding, females of most laboratory rodent species do, in general, fail to attack and demonstrate the overt aggression seen in males and in wild-caught female mice, likely because female laboratory mice appear to have had most of their overt intrasexual aggression bred out to facilitate breeding (Parmigiani et al., 1999). Instead of attacks, they frequently perform other agonistic behaviours aimed at establishing a dominance hierarchy instead, such as following or chasing, pinning down, or aggressively grooming the intruder
(Grant & Mackintosh, 1963; Palanza et al., 2005). Thus, as has long been recommended
(e.g., Scott & Fredericson, 1951), the study of agonistic behaviour in female laboratory animals requires the use of more sensitive and comprehensive ethological analyses which include both overt and more subtle forms of agonistic behaviour and the contexts in which they occur.
Several neurochemical systems have been identified as important for aggression, including 5HT, DA, and AVP. Aggression is reduced by treatment with 5HT receptor agonists, and reversed by antagonist treatments (De Boer et al., 1999; Fish et al., 1999;
Miczek et al., 1998), and manipulations of the tryptophan (5HT precursor) content of animals' diets found an inverse relationship, such that increased tryptophan, and thus more
14 5HT, resulted in decreased aggression, and vice versa (Kantak et al., 1980; Lasley &
Thurmond, 1985). Additionally, 5HT receptor 1B KO mice exhibit increased aggression specifically, without sedative side effects (de Almeida et al., 2001; de Almeida & Miczek,
2002; De Boer & Koolhaas, 2005; Fish et al., 1999; Olivier et al., 1990; Saudou et al., 1994), as do serotonin transporter KO mice (Holmes et al., 2002a; Kim et al., 2005).
Serotonergic projections to the ventral tegmental area and substantia nigra can increase
(through serotonin 1B/1D receptors) or decrease (through serotonin 2-type receptors) DA activity (Cobb & Abercrombie, 2003; Ugedo et al., 1989). DA can also mediate aggression, and in fact may be necessary for aggressive behaviour (Miczek et al., 2002). Aggression is decreased by both D1 and D2 DA receptor antagonists (de Almeida et al., 2005), and DA transporter KO mice are more aggressive than WT, despite reduced D1 and D2 DA receptor expression in some brain regions (Rodriguiz et al., 2004).
Two other peptides that modulate aggression are OT and AVP. OT and AVP are nonapeptides that differ in only two amino acids (CITE). Both are synthesized primarily in the magnocellular supraoptic nucleus and both magnocellular and parvocellular neurons in the PVN, and are released to act as neurohormones through projections to the posterior pituitary (Choleris et al., 2009). AVP is also produced by parvocellular neurons in the
BNST, medial amygdala, and suprachiasmatic hypothalamic nucleus (Frank & Landgraf,
2008). Vasopressinergic neurons from the BNST and medial amygdala project to the hippocampus, lateral septum, olfactory bulb and other areas, and modulate social recognition and social memory (Kogan et al., 2000; Pierman et al., 2008; Rood & De Vries, 2011).
OT decreases aggression (Ferris, 2005; McCarthy, 1990), and KO of OT or OTR, or OT 15 antagonism, increase aggression in both sexes (Giovenardi et al., 1998; Lubin et al., 2003;
Nishimori et al., 2008; Ragnauth et al., 2005; Winslow et al., 2000). Contrarily, AVP increases aggression in males (Ferris & Delville, 1994). Further, inactivation or antagonism of either the AVPR1a or AVPR1b receptors decreases aggression in males, and AVPR1b receptors in females (Blanchard et al., 2005; Caldwell & Albers, 2004; Ferris et al., 2006;
Ferris et al., 1992; Ferris et al., 1997; Ferris & Potegal, 1988; Wersinger et al., 2003;
Wersinger et al., 2002; Wersinger et al., 2004; Wersinger et al., 2007a; Wersinger et al.,
2007b).
These systems may interact in aggression-relevant brain regions, such as the anterior hypothalamus (5HT and AVP) or nucleus accumbens (5HT and DA; reviewed in Miczek et al., 2007). Other brain regions have also been implicated in aggression, including the lateral septum, BNST, VMH and medial preoptic area (MPOA; Cologer-Clifford et al., 1997;
Spiteri et al., 2010; Trainor et al., 2006), several of which are sexually dimorphic (e.g., medial amygdala, BNST, MPOA, VMH; Ahmed et al., 2008; Bonthuis et al., 2010) and express hormone receptors. As mentioned above, many of these neurochemical systems are modulated by estrogens, which appear to play an important role in the mediation of social behaviour.
1.3.2. Social Interaction and Aggression: Effects of Estrogens
Gonadal hormones are known to affect aggression and social interactions in mice.
Removing the endogenous source of T by castrating males both eliminates aggression and increases grooming or sniffing the opponent, and this can be reversed by T administration
(e.g., Beeman, 1947; Brain & Bowden, 1979; Luttge, 1972; Simon & Whalen, 1986; Tollman
16 & King, 1956); the highly consistent effects of castration even led to the suggestion that T was responsible for aggression (Beeman, 1947; Scott & Fredericson, 1951). T's main androgenic metabolite, dihydrotestosterone (DHT), can also restore aggression in castrated males (Brain & Bowden, 1979; Brain & Poole, 1976; Luttge, 1972; Luttge & Hall, 1973;
Simon & Whalen, 1986).
However, estrogens, the prototypical female hormones, have also been found to affect agonistic behaviour. Thus, T effects on aggression may be mediated through its estrogenic, as well as its androgenic, metabolites (Bowden & Brain, 1978; Brain & Bowden, 1979; Brain
& Poole, 1976; Edwards & Burge, 1971; Luttge, 1972; Luttge & Hall, 1973; Simon &
Whalen, 1986). Like DHT, estrogens restore aggression in castrated males (Simon &
Whalen, 1986), and anti-estrogen treatment of T-replaced castrates reduces aggression (Clark
& Nowell, 1979; Luttge, 1979). Further, mice lacking aromatase, the enzyme that converts androgens to estrogens, show reduced male aggression, an effect that can be reversed with chronic estrogen treatment beginning shortly after birth (Toda et al., 2001). While it must be noted that the effects of T and estrogens on aggression and other behaviours may vary depending on the strain used (e.g., Parmigiani et al., 1999; Simon & Whalen, 1986), the results described above strongly suggest a role for estrogens in aggression and agonistic behaviour.
The involvement of estrogens and the two intranuclear ERs in the mediation of aggression have been studied using ER-specific KO mice. In male mice, αERKOs show decreased aggression, but no less social interaction, than their wild type (WT) counterparts
(Ogawa et al., 1998b). This deficit was not rescued by castration and exogenous T
17 administration in αERKO mice (Ogawa et al., 1998b), suggesting that ERα may be necessary for the development of normal social interactions. Given that these mice have higher endogenous levels of T than WT mice (Imwalle et al., 2002), this suggests that T's effects on aggression in males may be mediated by its estrogenic metabolites acting on ERα (Ogawa et al., 1998b; Scordalakes & Rissman, 2004). This hypothesis is further supported by the finding that levels of aggression in males with androgen receptor mutations did not differ from WT, and were greater than αERKO males, when all mice were castrated and administered estradiol (Scordalakes & Rissman, 2004). Interestingly, this reduction in aggression appears to be specific to obx male intruders, as the opposite is found if the intruder is a sexually receptive female: αERKO males will spend more time attacking and less time investigating these receptive females than WT males do (Scordalakes & Rissman,
2003). Thus, ERα may also mediate male social preference or sex recognition, such that its inactivation leads to indiscriminant aggression towards intruders, instead of the typical male- directed aggression (Scordalakes & Rissman, 2003). This would be consistent with the inability of αERKO mice to recognize familiar conspecifics (Choleris et al., 2003; Imwalle et al., 2002; Sánchez-Andrade & Kendrick, 2011; discussed in section 1.4.2 below); if αERKO males are unable to recognize the sex of the intruder, then they would attack males and receptive females equally, viewing them only as unfamiliar animals and not potential mates.
This would result in increased aggression towards the females, and less intense aggression towards the males, compared to that performed by WT males.
Levels of aggression have also been assessed in βERKO males, which, unlike male
αERKO mice, have normal levels of gonadal steroids (Couse & Korach, 1999). When sexually experienced, male βERKO mice show increased attacks and aggression during their
18 first exposure to an obx intruder, although this enhanced aggression disappears with repeated testing and/or with increased age, as pubertal and 3 month old mice showed more aggression than WT, but 6 month old mice did not (Nomura et al., 2002; Ogawa et al., 1999; Ogawa et al., 2000). Interestingly, when exposed to a stud male, βERKO male mice exhibit increased feminine behaviour relative to WT, suggesting that males may be defeminized through ERβ
(Bodo & Rissman, 2006).
As in the αERKO males, mice lacking both ERs (αβERKOs) showed reduced aggression and a near-total absence of attacks (Ogawa et al., 2000). This, taken together with the results of the αERKO and βERKO males discussed above, suggests that ERα may be necessary for the expression of overt attacks, but that agonistic behaviour may be modulated by ERβ in male mice (Ogawa et al., 2000).
There has been considerably less investigation of female aggression, particularly of the territorial kind tested in males. Female αERKO mice do demonstrate increased aggression towards same-sex intruders (Ogawa et al., 1998a); these results are consistent with their impairments in social recognition (Choleris et al., 2003; Imwalle et al., 2002), as decreased social recognition may relate to increased aggression in females (Ogawa et al., 2004).
Although, to the best of my knowledge, female βERKOs have not been tested in the resident- intruder test, they appear to have lower levels of maternal and T-mediated aggression
(Ogawa et al., 2004). However, it must be noted that these types of aggression are quite distinct from the territorial aggression tested in males (Al-Maliki et al., 1980; Svare &
Gandelman, 1973), and may not reflect the same motivational states or be controlled by the same systems as territorial aggression (Al-Maliki et al., 1980; Parmigiani et al., 1998; Svare
19 & Gandelman, 1973). Hence there is a need to use more in-depth ethological analyses.
Taken together, the research using ER-specific KO mice suggest that the two ERs have opposite effects in the two sexes. In males, ERα may increase aggression, while ERβ might decrease it, whereas in females, ERα appears to decrease aggression, and it is possible that
ERβ increases agonistic behaviour. This is consistent with previous findings that the regulation of aggression by hormones can be sex-dependent (Parmigiani et al., 1998).
Clearly, the use of these genetically modified mice has provided insight into the mediation of aggression by the two ERs. However, the use of global KO mice has a number of problems, including that it is difficult to dissect developmental from activational effects or to rule out the involvement of other flanking genes or compensatory mechanisms
(Waddington et al., 2005). Thus, investigations of the effects of pharmacological treatments targeting each receptor may shed further light on the relationship between estrogens, their receptors, and agonistic behaviour. Additionally, the use of an ethological analysis of behaviour in the resident-intruder test will allow for the sexes to be compared under identical testing conditions.
1.4. Social Learning of Food Preferences
1.4.1. Social Learning of Food Preferences: Introduction
One type of social learning that can be assessed in the laboratory is the acquisition of knowledge about novel foods from conspecifics. Rat pups have been found to bypass the potential risks of choosing their first solid meal by following adults and “copying” their food choices for their first solid meal, thus transitioning easily between a monophagus milk diet and an appropriate, polyphagus and varied adult diet by learning from adults (Galef &
20 Laland, 2005). The higher risk associated with individual learning may favour social learning, although the latter is less advantageous when the environment is less stable
(reviewed in Galef & Laland, 2005). Rodents often exhibit caution or neophobia when encountering novel potential foods, and must expend energy and risk toxin exposure by sampling only small amounts of a novel food and waiting to incorporate the food into their repertoire until they have received gastric feedback that it is safe (Kronenberger & Médioni,
1985). Thus, to use socially acquired information instead of individually sampling to expand one's food repertoire is particularly advantageous. This type of social learning is assessed in the laboratory using the STFP paradigm (Galef & Wigmore, 1983). In this paradigm, an observer animal interacts with a demonstrator animal that has recently consumed a distinctively flavoured food that is novel to the observer. When given a subsequent choice between two novel flavours, the observer will typically prefer the food that its demonstrator had eaten (e.g., Galef & Wigmore, 1983). The STFP is a robust and ecologically relevant paradigm, and is an exclusively social type of learning (Alvarez et al., 2001; Choleris et al.,
2011a; Galef et al., 1984; Galef et al., 1988; Galef, 1996; Valsecchi & Galef, 1989).
Additionally, it is mediated by oronasal investigation and stronger than individual learning of a taste aversion (discussed below; Galef, 1986; Galef et al., 1988), and as only a single exposure to the demonstrator is required to produce the preference, the STFP does not require the extensive training common to many other learning paradigms (Alvarez et al.,
2001).
Numerous rodent species have been shown to socially acquire a food preference, including Belding ground squirrels (Sermophilus beldingii; Peacock & Jenkins, 1988), spiny mice (Acomys cahirinus; McFadyen-Ketchum & Porter, 1989) and Mongolian gerbils
21 (Meriones unguiculatus; Valsecchi et al., 1996), although the majority of research has been done with rats (Rattus norvegicus; e.g., Galef & Wigmore, 1983) or house mice (Mus musculus; e.g., Valsecchi & Galef, 1989). Galef began characterizing this paradigm in 1983 and has continued to investigate the phenomenon, finding that the critical association formed during the social interaction is between the food odour and carbon disulphide (CS2), a semiochemical produced by digestion (Bean et al., 1988; Galef et al., 1988). Several controls have demonstrated the necessity of this association, showing that observer animals will not develop a preference from interactions with the food odour in isolation or the food odour in the presence of a conspecific (Choleris et al., 2011a), with the flavoured-food-dusted hind quarters of anaesthetized conspecific (Galef et al., 1988), or with food-powdered cotton batton surrogates (Valsecchi & Galef, 1989). The social nature is further supported by findings that in some cases, the nature of the demonstrator can affect the socially acquired food preference. Mongolian gerbils will only learn from a familiar or related demonstrator, not an unrelated stranger, unless treated with an anxiolytic (Choleris et al., 1998b; Valsecchi et al., 1996), and there is some evidence that mice will learn better from a dominant than a subordinate demonstrator (Clipperton et al., 2008; Clipperton-Allen et al., 2011b; Kavaliers et al., 2005a). Additionally, adult rats and mice tend to be better demonstrators than juveniles (Choleris et al., 1997; Galef et al., 2001; Galef & Clark, 1971; Gerrish & Alberts,
1995).
Interestingly, mice and rats will prefer a food smelled on the head of an anaesthetized
(but still breathing) or ill conspecific (Galef et al., 1988; Valsecchi & Galef, 1989), or a food to which CS2 has been added (Bean et al., 1988; Galef et al., 1988). Thus, this learning does not distinguish between demonstrators of different healthiness, but rather follows a general
22 rule of “prefer foods consumed by living conspecifics.” Furthermore, animals have been shown to use socially acquired information when it is contradicted by individually learned knowledge: even when the demonstrated food was followed by lithium chloride-induced illness (Galef et al., 1988) or had been previously associated with illness through conditioned taste avoidance (Galef, 1986), the observers still preferred it. The similar results found with both rats and mice suggests that common mechanisms could be at work, especially in light of other similarities between the species, including their dietary generality, and tendency to live in close proximity to humans (Latham & Mason, 2004; Valsecchi and Galef, 1989). Food neophobia has become particularly important, and social learning particularly advantageous, as the habitats of these species increasingly overlap with humans, which thus increases their likelihood of poisoning by rodenticides (Latham & Mason, 2004; Valsecchi & Galef, 1989).
While it has been the subject of less investigation than non-social learning (Choleris &
Kavaliers, 1999), several brain regions and neurochemical systems have been implicated in the STFP. These include the hippocampus and associated regions, including the entorhinal, perirhinal and postrhinal cortices (Alvarez et al., 2001; Alvarez et al., 2002; Bunsey &
Eichenbaum, 1995; Clark et al., 2002; Countryman et al., 2005b; Countryman et al., 2005a;
Ross & Eichenbaum, 2006; Winocur et al., 2001), the basolateral amygdala (Wang et al.,
2007), the frontal, piriform and orbitofrontal cortices (Ross et al., 2005; Ross & Eichenbaum,
2006), and the basal forebrain (Berger-Sweeney et al., 2000; Boix-Trelis et al., 2006; Boix-
Trelis et al., 2007; Ricceri et al., 2004; Vale-Martínez et al., 2002). In the latter region, the cholinergic system appears to be particularly important for the STFP, as components of the cholinergic basal forebrain mediate the acquisition, consolidation and/or retrieval of the demonstrated food preference (Berger-Sweeney et al., 2000; Boix-Trelis et al., 2007; Ricceri
23 et al., 2004; Ross et al., 2005; Vale-Martínez et al., 2002).
As with aggression, the DA and AVP systems may also play a role in the STFP, as do the
OT system and the cyclic AMP response element binding protein (CREB; Countryman et al.,
2005b; Countryman & Gold, 2007; Kogan et al., 1996). CREBKO mice are impaired in the
STFP when they are tested after a delay (Kogan et al., 1996), and DA transporter KO
(DATKO) mice show a preference for the novel food not consumed by their demonstrator
(Rodriguiz et al., 2004), and administration of a D1 but not a D2 DA receptor antagonist blocks the STFP (Choleris et al., 2011a). Additionally, both OT and AVP improve STFP performance as long as the task is of sufficient difficulty; if the STFP is easily performed by controls, then AVP appears to impair it (Bunsey & Strupp, 1990; Cushing & Wynne-
Edwards, 2006; Popik & van Ree, 1991; Strupp et al., 1990).
1.4.2. Social Learning: Effects of Estrogens
While natural variations in estrogens levels have been shown to affect social memory for a food preference, there has been little research on estrogens and their effect on this biologically significant phenomenon (Choleris et al., 2011a; Sánchez-Andrade et al., 2005).
Postpartum females showed a stronger preference for the demonstrated food than nulliparous, virgin females when tested 8 days post-interaction (Fleming et al., 1994), suggesting that parturitional hormones, including estrogens, may be involved (Choleris et al., 2011c).
Further, our laboratory has shown that mice in proestrous or diestrous phases of the estrous cycle, which involve higher estrogen levels (Walmer et al., 1992), maintain a demonstrated food preference 1.5-2.5 times longer than lower-estrogen estrous or ovariectomized (ovx) females when no delay is present between social interaction and test (Choleris et al., 2011a).
24 Similarly, when tested 24h following interaction with a demonstrator, only observer mice in the high-estrogen (Walmer et al., 1992) proestrous phase of the cycle showed a significant preference for the demonstrated food (Sánchez-Andrade et al., 2005). This led us to investigate the role of estrogens and their receptors in this paradigm, and we have previously shown that administration of an ERα agonist, PPT, blocks the STFP, while the observer maintains a preference for the demonstrated food longer following treatment with ERβ agonist WAY-200070 (Clipperton et al., 2008). This suggests that female gonadal hormones in general, and estrogens particularly, may mediate the STFP as they do other social behaviours or forms of learning.
1.5. Social Recognition
1.5.1. Social Recognition: Introduction
While the above tasks deal with aspects of social cognition and social life, social recognition is the basis for social living and all social behaviours (Colgan, 1983). In order to interact appropriately and form hierarchies and other social relationships, one must be able to recognize and distinguish between conspecifics, and adjust one’s behaviour based on one’s previous experience with that individual (Choleris et al., 2009; Choleris et al., 2004; Pierman et al., 2008).
Social recognition can be classified into different levels, including reproductive state, health, emotional, social hierarchical status, genetic relatedness, and familiarity, and has been observed in many rodent species, including gerbils, hamsters, rats and mice (reviewed in
Choleris et al., 2009). Some researchers have attempted to look at individual recognition specifically by using stimulus animals with which the experimental animals have had some
25 previous experience (for example, the experimental animal has either defeated or been defeated by the stimulus animal; e.g., Lai & Johnston, 2002). Thus, different behaviour by the experimental animal is re-exposed to a conspecific that previously defeated it than when re-exposed to a conspecific that it defeated implies individual recognition is implied, as this suggests that it remembers the specific animal and their previous encounters. However, the majority of laboratory research has focused on familiarity recognition (Choleris et al., 2009).
There are three main ways in which social (familiarity) recognition is assessed in the laboratory, all of which take advantage of the tendency of rodents to preferentially investigate a novel conspecific over a familiar one. The first of these, which is more commonly used, is the Habituation/Dishabituation paradigm (e.g., Choleris et al., 2003). In this task, the experimental animal is repeatedly exposed to a conspecific; as this stimulus animal becomes more familiar, the total amount of time the experimental animal spends investigating it decreases (Habituation). Following the final exposure of the familiar animal, the experimental animal is presented with a novel conspecific, which restores investigation to baseline or near-baseline levels (Dishabituation). This second step is particularly important to ascertain whether the experimental animal is habituating to the testing experience per se, or to the stimulus animal specifically.
A more sensitive test of social recognition is the social discrimination paradigm (e.g.,
Choleris et al., 2006; Engelmann et al., 1998; Phan et al., 2011). In this task, the animal is repeatedly presented with two conspecifics, which become familiar and to which it habituates. In the final trial, one of the familiar animals is replaced with a novel conspecific, and the investigation of the familiar and unfamiliar animals is compared. If the test animal
26 preferentially investigates the unfamiliar stimulus, then this is taken as social recognition.
One of the advantages of this task is that it bears greater similarity to other, non-social paradigms (i.e., the object recognition and object placement tasks), thus providing the option for subsequent testing to determine if effects are specifically social, or on more general recognition (Choleris et al., 2009).
Thirdly, a combination of the two tests, the habituation/discrimination paradigm, has also been used to simultaneously assess both short- and long-term social recognition memory. In this version, the experimental animal is repeatedly exposed to a stimulus animal (habituation) and then receives a single exposure to a novel stimulus conspecific (dishabituation), thus testing short-term memory. Twenty-four hours later, the test animal is given a binary choice test (social discrimination) between a familiar and a novel conspecific to assess long-term memory (Sánchez-Andrade et al., 2005; Sánchez-Andrade & Kendrick, 2011; Sánchez-
Andrade & Kendrick, 2009).
A number of brain regions and systems have been implicated in social recognition. In one pathway, social memory begins with the detection of semiochemicals and pheromones with receptors in the vomeronasal organ by the accessory olfactory bulb (Insel & Fernald,
2004). The bed nucleus of the accessory olfactory tract, BNST, and vomeronasal amygdala all receive mitral cell projections from the accessory olfactory bulb (Insel & Fernald, 2004).
The main olfactory system also detects social odours; the main olfactory bulb receives input from the main olfactory epithelium, and then projects to a number of brain regions, including the piriform olfactory cortex, entorhinal cortex, and anterior olfactory nucleus (Sánchez-
Andrade & Kendrick, 2009). Olfactory inputs from both the main and accessory systems
27 converge at the medial amygdala (Beauchamp & Yamazaki, 2003; Dulac & Torello, 2003;
Johnston & Peng, 2008; Kang et al., 2009); the amygdala then projects to the MPOA, VMH, and the remainder of the neuroendocrine hypothalamus (Insel & Fernald, 2004). Other brain regions that are important for social recognition and social memory include the lateral septum and hippocampus (Engelmann et al., 1996; Kogan et al., 2000; van Wimersma
Greidanus, 1982; van Wimersma Greidanus et al., 1983); additionally, the olfactory bulb and piriform cortex are required for recall and/or expression of the recognition memory when observers were tested 8h after the acquisition phase (Sánchez-Andrade et al., 2005).
The formation of social memories can be blocked by inhibiting the cholinergic system or the N-methyl-D-aspartate (NMDA) and alpha-amino-3-hydroxy-5-methyl-4-isoxazole- propionic acid (AMPA) glutamate receptors (Sánchez-Andrade et al., 2005). Additionally, two of the main systems involved in rodent social recognition are the OT and AVP systems.
OT has similar effects on social recognition in both sexes in rats and mice (reviewed in
Choleris et al., 2009). Physiological doses of OT and OT agonists improved social recognition when administered systemically or infused intracerebroventricularly (i.c.v.) or to the lateral septum, ventral hippocampus and MPOA (Popik et al., 1992; Popik & van Ree,
1998; Popik & van Ree, 1991; van Wimersma Greidanus & Maigret, 1996). Consistently,
OT antagonists reversed the OT effects when co-administered with OT or OT agonists, and impaired social recognition when administered alone (Benelli et al., 1995; Engelmann et al.,
1998). Further, OTKO and OT receptor KO (OTRKO) mice, both global and conditional forebrain OTRKO, show deficits in social recognition specifically, as their behaviour is otherwise normal (Choleris et al., 2003; Choleris et al., 2006; Ferguson et al., 2001; Ferguson
28 et al., 2000; Lee et al., 2008; Takayanagi et al., 2005).
Additionally, in the case of both OTKO and conditional OTRKO mice, with sufficient repeated exposures some degree of social recognition can become apparent (Choleris et al.,
2003; Dugatkin & Godin, 1992; Kavaliers et al., 2001; Lee et al., 2008). The more limited impairment of the conditional forebrain OTRKO mice, which show normal levels of OTR in the medial amygdala, olfactory bulbs and nucleus, and neocortex (Lee et al., 2008) suggests that social recognition may also be mediated by non-forebrain limbic structures like the medial amygdala; OT administration to this region restored social recognition in global
OTKO mice, and administration of an OT antagonist or OTR antisense DNA to the medial amygdala impaired social recognition in WT mice (Choleris et al., 2008a; Ferguson et al.,
2000). It has been proposed that the medial amygdala is actually both necessary and sufficient for OT mediation of familiarity recognition, as neither OT infusion into the olfactory bulbs of OTKO nor OT antagonist administration to the olfactory bulbs of rats affected this behaviour (Dluzen et al., 2000; Ferguson et al., 2000).
AVP also improved social recognition, both systemically and when administered i.c.v., to the lateral septum, or to the olfactory bulbs (Bluthé & Dantzer, 1992; Dantzer et al., 1987;
Dluzen et al., 1998b; Dluzen et al., 1998a; Engelmann et al., 1996; Engelmann & Landgraf,
1994; McCarthy et al., 1996). Additionally, AVP antiserum or antagonist treatment impairs social recognition when infused into the hippocampus, lateral septum, i.c.v. or injected peripherally, and elimination of AVP cells in the olfactory bulb also impairs social recognition (Appenrodt et al., 2002; Dantzer et al., 1987; Dantzer et al., 1988; Everts &
Koolhaas, 1999; Tobin et al., 2010; van Wimersma Greidanus & Maigret, 1996). Thus, in
29 rats AVP appears to have different effects on the sexes: peripheral AVP treatment improves social recognition in both sexes, but AVP antagonist treatment only impairs males and does not affect females (Bluthé et al., 1990; Bluthé & Dantzer, 1990; Engelmann et al., 1998), suggesting that while AVP may facilitate social recognition in females, it is not necessary for females like it is in males (Choleris et al., 2009).
AVP binds primarily to two receptors in the brain, AVPR1a and AVPR1b, and appears to mediate social recognition more through its AVPR1a than AVPR1b receptor. AVPR1a overexpression in the lateral septum improved social recognition (Landgraf et al., 2003), and inactivation of this receptor by antisense DNA, pharmacological antagonism, or genetic manipulation (AVPR1aKO) impairs or blocks social recognition (Appenrodt et al., 2002;
Bielsky et al., 2004; Dantzer et al., 1988; Everts & Koolhaas, 1999; Landgraf et al., 1995;
Landgraf et al., 2003; Wersinger et al., 2004). Re-expression of the AVPR1a gene in the lateral septum rescues social recognition of AVPR1aKO males (Landgraf et al., 2003).
These impairments appear to be fairly specific to social behaviour, as AVPR1aKO male mice are not impaired in olfaction, sensorimotor gating, locomotion or spatial or olfactory learning, and show decreased anxiety (Bielsky et al., 2004). Thus, it appears that AVPR1a in the lateral septum is necessary and sufficient for the mediation of social recognition by AVP.
AVPR1bKO males also have social recognition deficits, although less severe than those of AVPR1aKOs, as they will habituate and dishabituate given enough habituation sessions and a sufficiently short delay between habituation and test (Bielsky et al., 2004; Wersinger et al., 2002; Wersinger et al., 2004). Additionally, unlike AVPR1aKOs, the AVPR1bKO mice show impairments in social behaviour including aggression, social motivation and social
30 recognition, but show normal spatial learning, predation and olfaction (Caldwell et al., 2008;
Wersinger et al., 2002; Wersinger et al., 2004). This suggests that AVPR1b may modulate social behaviour in general, as opposed to specifically affecting social recognition.
1.5.2. Social Recognition: Effects of Estrogens
Estrogens have also been found to mediate social recognition, and appear to affect acquisition, as well as short- and long-term social memory (Sánchez-Andrade & Kendrick,
2009). Female mice trained in proestrus show improvements in both short- and long-term social recognition memory when tested in the habituation/discrimination paradigm (Sánchez-
Andrade et al., 2005; Sánchez-Andrade & Kendrick, 2011), and ovx mice can have their deficits in social recognition rescued by estrogens or estrogens plus progesterone treatment
(Spiteri & Ågmo, 2009; Tang et al., 2005).
Research with KO mice suggests that both ERs modulate social recognition, although
ERα appears to be more important (see Table 1). Female αERKO mice show no evidence of short- or long-term social recognition, as they do not appear to habituate, dishabituate or discriminate between familiar and unfamiliar stimulus animals, regardless of the paradigm used (Choleris et al., 2003; Choleris et al., 2004; Choleris et al., 2006; Sánchez-Andrade et al., 2005; Sánchez-Andrade & Kendrick, 2011; Sánchez-Andrade & Kendrick, 2009). Male
αERKO mice are also impaired in long-term social discrimination, but may be able to habituate and dishabituate; when exposed to conscious stimuli in the habituation/ dishabituation paradigm, they showed no evidence of social recognition, but did habituate and dishabituate to an unconscious stimulus animal in the habituation/discrimination paradigm (Choleris et al., 2004; Imwalle et al., 2002; Sánchez-Andrade & Kendrick, 2011;
31 Table 1. Effects of estrogen receptor manipulations on social recognition. ERα ERβ Test Males Females Males Females Gene Inactivation Habituation/Dishabituation Nob,1 No2 No2 (STa) Social Discrimination Partial No3 (ST) Impairmentc,3 Habituation/Discrimination Yesd,4 No4 Yes4 Yes4 (Habituation, ST) Habituation/Discrimination Proestrus No4 No4 Yes4 (Social Discrimination, LTe) Onlyf,4 ER-selective Agonist Social Discrimination Improvedg,5 Impairedh,5 Rapid Effects (<40 min) Social Discrimination No Effect6 Improved6 Improved6 Improved6 LT Effects (48-72h) a “ST” indicates short-term. b “No” indicates that the experimental animals did not display evidence of social recognition. c “Partial Impairment” indicates that there was a partial impairment in social recognition. d “Yes” indicates that the experimental animals showed evidence of social recognition. e “LT” indicates long-term. fIn this experiment, only mice trained in the proestrous phase showed the long-term social recognition memory, regardless of genotype. g “Improved” indicates that the treatment improved social recognition learning. h “Impaired” indicates that the treatment impaired social recognition learning. 1Imwalle et al., 2002. 2Choleris et al., 2003. 3Choleris et al., 2006. 4Sánchez-Andrade & Kendrick, 2011. 5Phan et al., 2011. 6Cragg et al., 2011.
32 Sánchez-Andrade & Kendrick, 2009).
Like the αERKO males, βERKO female mice fail to show social recognition in the habituation/dishabituation paradigm (Choleris et al., 2003); however, βERKO mice of both sexes both habituate and dishabituate when tested in the habituation/discrimination paradigm
(Sánchez-Andrade & Kendrick, 2011; Sánchez-Andrade & Kendrick, 2009). In the latter paradigm, male βERKOs also show long-term social recognition in the social discrimination component of the task (Sánchez-Andrade & Kendrick, 2011; Sánchez-Andrade & Kendrick,
2009). Female βERKOs are partially impaired in the social discrimination paradigm
(Choleris et al., 2006), and only βERKO or WT mice showed long-term social recognition
(Sánchez-Andrade & Kendrick, 2011; Sánchez-Andrade & Kendrick, 2009). These results suggest that ERα is critical for social recognition in females, while ERβ may play a modulatory role depending on the duration of the memory, difficulty of the task or other experimental factors. In males, however, the relationship is less clear, but both ERs may play a modulatory role in social recognition.
Involvement of specific estrogen receptors in social recognition has also been investigated by pharmacological means. Knockdown of ERα by short hairpin RNA injection into the medial amygdala, but not the VMH, blocked social recognition (Spiteri et al., 2010)
Acute ERα agonist 1,3,5-tris(4-hydroxyphenyl)-4-propyl-1H-pyrazole (PPT) improved social discrimination both rapidly (<40 min; Phan et al., 2011) and when administered in low doses
48h before testing (Cragg et al., 2011). Acute treatment with ERβ agonist 2,3-bis(4- hydroxyphenyl)propionitrile (DPN) may have rapidly impaired social discrimination (Phan et al., 2011), but performance was improved by administration of ERβ agonist
33 7-Bromo-2-(4-hydroxyphenyl)-1,3-benzoxazol-5-ol (WAY-200070) 72h prior to testing
(Cragg et al., 2011). Taken together, these results suggest that ERα is likely necessary for social recognition, while ERβ may play a more modulatory role.
OT is involved in social recognition and expression of both the hormone and its receptor can be altered by estrogens; thus, it has been suggested that the estrogen receptors affect social recognition by modulating the OT system (Choleris et al., 2003). Specifically, a so- called “four-gene micronet” underlying social recognition has been proposed, with estrogens controlling the expression of OTR in the medial amygdala through ERα, and regulating the production of OT in the PVN through ERβ (Choleris et al., 2003).
1.6. Aims
The overall aim of this thesis is to further elucidate the involvement of estrogens and their classic intranuclear receptors ERα and ERβ, as well as associated systems, in social cognition and social behaviours. The three types of social cognition outlined above were each explored in separate sets of experiments.
Part I: Involvement of the estrogen receptors in social interactions and aggression was investigated using pharmacological techniques. Specifically, the ERα specific agonist PPT
(Chapter 2) and the ERβ specific agonist WAY-200070 (Chapter 3) were administered to gonadally intact and gonadex outbred CD1 mice of both sexes, and their effects on the interaction of these treated resident mice with same-sex, gonadex, unfamiliar conspecific intruders were assessed using an ethological analysis in order to investigate effects on both overt aggression (i.e., attacks) and the more subtle, dominance-related agonistic behaviours
(e.g., pushing down the intruder), as well as on affiliative social behaviours, non-social 34 behaviours and broader, more general behavioural measures (e.g., total activity).
Part II: Estrogens' effects on social learning, which involves social interaction and is one of the main advantages of social living, has received less investigation than aggression.
Thus, in addition to a series of experiments in which we treated ovx female observers with the ERα specific agonist PPT and the ERβ specific agonist DPN (Chapters 5-6), we also tested the effects of the general estrogen estradiol benzoate (Chapters 4, 6). Studies in which the drug was administered prior to the social interaction, as in the aggression studies above, also included an ethological analysis of the interaction between the observer animal and her familiar demonstrator cage-mate. Thus, we were able to investigate the effects of estrogens and their receptors on the social interaction of familiar ovx female CD1 mice as well as unfamiliar animals and link behavioural features of the social interaction to social learning
(e.g., more submissive OBS learn better than less submissive OBS).
We had previously found effects of estrogen receptor agonists given pre-acquisition (i.e.,
48h (PPT) and 72h (WAY-200070) prior to the social interaction and choice test) on the social transmission of food preferences (Clipperton et al., 2008). Thus, we wanted to elucidate which of the phases of memory, acquisition, consolidation or retrieval, were being affected by the compounds. Drugs were therefore administered immediately or 24h after the social interaction took place, to assess the involvement of estrogens, ERα and ERβ in consolidation, retrieval and/or expression but not acquisition, or to affect only retrieval and/or expression, respectively (Chapters 4-5).
Part III: Perhaps the best characterized of the three types of social cognition studied here, particularly in terms of the involvement of estrogens, is social recognition. Both ERs and 35 their potential downstream effector systems, OT and AVP, have been identified as likely mediators. However, it is unclear how or if behaviours associated with social recognition relate to variations in the expression of these hormones and/or receptors in female outbred mice. To explore this, we assessed the behaviour and expression of receptors for the female gonadal hormones (i.e., ERα, ERβ, progesterone receptor), as well as OT, AVP and their receptors, in female mice that naturally performed very well (high recognizers, HR; >85% preference for the novel conspecific) or at chance levels (low recognizers, LR; 40-60% preference for the novel conspecific) on the social discrimination task (Chapter 7). Gene expression was analyzed in brain regions known to modulate social behaviours, and the expression of the genes in the different brain areas was compared between HR and LR mice.
Between-group differences in behaviour were also analyzed, and behaviours were also correlated with gene expression. The results of this work support the involvement of ERs,
OT and AVP in natural variations of social recognition. Our data reveal additional brain regions and targets whose specific involvement in social recognition could be further investigated, for example, in conditional KO studies or with intracranial pharmacological treatments.
Previous research has suggested that oxytocin, vasopressin and their modulators, estrogens, are involved in many types of social behaviour. This thesis increases our knowledge of the effects estrogens and their downstream effectors on social cognition, and contributes to the overall goal of gaining a comprehensive understanding of the neuroendocrinological factors underlying social cognition and social behaviour.
36
CHAPTER 2:
EFFECTS OF AN ESTROGEN RECEPTOR AGONIST ON AGONISTIC BEHAVIOUR
IN INTACT AND GONADECTOMIZED MALE AND FEMALE MICE
This manuscript is published in: Clipperton-Allen, A.E., Almey, A., Allen, C.P., Melichercik,
A., and Choleris, E. (2011). Effects of an estrogen receptor agonist on agonistic
behaviour in intact and gonadectomised male and female mice.
Psychoneuroendocrinology 36: 981-995. 37 2.1 Abstract
Gonadal hormones mediate both affiliative and agonistic social interactions. Research in estrogen receptor alpha (ERα) or beta (ERβ) knockout (KO) mice suggests that ERα increases and ERβ decreases male aggression, while the opposite is found for female αERKO and βERKO mice. Using a detailed behavioural analysis of the resident-intruder test, we have shown that the ERβ selective agonist WAY-200070 increased agonistic behaviours, such as aggressive grooming and pushing down a gonadectomized (gonadex) intruder, in gonadally intact but not gonadex male and female resident mice, while leaving attacks unaffected. The role of acute activation of ERα in agonistic behaviour in adult non-KO CD1 mice is presently unknown. The current study assesses the effects of the ERα selective agonist 1,3,5-tris(4- hydroxyphenyl)-4- propyl-1H-pyrazole (PPT) on the social and agonistic responses of gonadally intact and gonadex male and female CD1 mice to a gonadex, same-sex intruder.
PPT had few effects in gonadally intact mice, but seemed to increase sex-typical aggression
(i.e., attacks in males, other dominance-related behaviours in females) in gonadex mice. In untreated mice, we confirmed our previous findings that gonadally intact males attacked the intruder more than females, but females spent more time engaged in agonistic behaviour than males. As in our previous results, we observed that gonadex mice generally show behaviour patterns more like those of the gonadally intact opposite sex, while leaving overall levels of agonistic behaviour unaffected. Taken together, our current and previous results show that exogenous activation of ERα had no effect in gonadally intact mice, but increased sex-typical agonistic behaviour in gonadex mice, while ERβ had no effect in gonadex mice, but increased non-attack agonistic behaviour in gonadally intact animals. This suggests that, as in social recognition, ERα may be necessary for the activation of agonistic responses, while
38 ERβ may play a modulatory role.
39 2.2 Introduction
Most laboratory research on agonistic behaviour has focused on the behaviour of males, typically within the resident-intruder paradigm in rodents (e.g.,(Brain et al., 1981; Simon &
Whalen, 1986). In this test, an animal (the intruder) is placed in the home cage of a conspecific resident; often the resident is isolated prior to the test. The frequency of or latency to attack, and/or other measures of overt aggression or fighting, are recorded (e.g.,
Morè, 2008), sometimes in addition to other measures which include both the aggression and the response to it (i.e., “agonistic” behaviour; Alleva, 1993; Branchi et al., 2006; Olivier et al., 1991). As male mice often dominate a shared territory or establish and maintain an exclusive territory in the wild, this test measures natural behaviours in laboratory mice
(Latham & Mason, 2004). Male mice of inbred (e.g., C57BL/6J, C-3H agoutis, BALB/C albinos) and outbred (e.g., Swiss CD1, Swiss-Webster, CF-1, Tuck TO) strains, and wild- caught or wild-descended mice, will consistently attack male intruders, especially if these residents have been isolated or housed with females (e.g., Brain & Bowden, 1979), with the highest levels of aggression exhibited by wild-caught, wild-descended and outbred Swiss
CD1 mice (Parmigiani et al., 1999).
Female aggressive or agonistic behaviour has been studied less frequently, and predominantly in reproduction-related conditions, including following housing with a male conspecific (e.g., Haney et al., 1989; Haney & Miczek, 1989). Protecting their litter
(“maternal” or “postpartum” aggression; e.g., Haney et al., 1989; Haney & Miczek, 1989) has been shown to be qualitatively different from the intrasexual territorial aggression studied in males, which is less violent and more ritualized than maternal aggression (Al-Maliki et al.,
1980).
40 In the few studies on non-reproduction-related female aggression, virgin or non- reproductively- active outbred albino females have been observed to fight during group formation, following isolation, or with lactating females (Morè, 2008). However, it is rare for virgin females of some inbred strains (e.g., C57BL/6J, 129) to attack same-sex intruders
(Ogawa et al., 1998a). This result, when taken with the observation that wild-caught or wild- descended females will exhibit higher levels of intrasexual aggression than both inbred and outbred laboratory females, suggests that interfemale aggression may have been selectively bred out of domesticated mice (Palanza et al., 2005). However, these domesticated females may be using other agonistic behaviours that are often not assessed in tests of aggression, such as chasing, pinning, or aggressively grooming the intruder (Alleva, 1993; Clipperton et al., 2008; Clipperton-Allen et al., 2010; Cologer-Clifford et al., 1997; Grant & Mackintosh,
1963). These behaviours are aimed at establishing dominance over the intruder, rather than violently excluding them from the territory. A comprehensive or “ethological” analysis, based on the full behavioural repertoire of the mouse (the ethogram; Grant & Mackintosh,
1963) and thus including more than just attack measures, is therefore required for the study of female agonistic interactions (Branchi et al., 2006; Miczek et al., 2001; Mos & Olivier,
1989; Olivier et al., 1989; Pietropaolo et al., 2004).
A role for the sex hormones in aggression has been demonstrated. Castration has long been known to radically reduce intermale aggression in inbred, outbred and wild-caught mice, and the administration of testosterone (T), the main male gonadal hormone, will restore aggression in these castrated males (e.g., Beeman, 1947; Brain & Bowden, 1979).
Restoration of aggression can be obtained either by T directly, or through one of its main metabolites, the estrogen estradiol or the androgen dihydrotestosterone (DHT). Intermale
41 aggression can be restored to near pre-castration levels in some strains by DHT (e.g., Swiss-
Webster, CD1, Tuck TO; Brain & Bowden, 1979; Brain & Poole, 1976; Luttge, 1972; Luttge
& Hall, 1973; Simon & Whalen, 1986), and in others by estradiol (CF-1, CFW, Tuck TO,
Swiss-Webster; Bowden & Brain, 1978; Brain & Bowden, 1979; Brain & Poole, 1976;
Edwards & Burge, 1971; Simon & Whalen, 1986). Furthermore, male mice who cannot convert T to estradiol due to an inactivation of the gene for the conversion enzyme aromatase show an estradiol-reversible reduction in aggression (Toda et al., 2001). Both androgens and estrogens thus appear to be involved in the mediation of aggressive behaviour in male mice.
Estrogens bind to two well-characterized intranuclear receptors, alpha (ERα) and beta
(ERβ). Results from studies with mice in which the gene for one of the two receptors has been inactivated (so-called “knock out,” KO, mice) suggest that the two ERs are differently involved in the mediation of aggression and agonistic behaviour (Choleris et al., 2008a;
Ogawa et al., 1998b; Ogawa et al., 1998a; Ogawa et al., 1999). When confronted with an olfactory bulbectomized (OBX) or intact male intruder, male αERKO mice show reduced aggressive responses, and do not show T-enhanced aggression (Ogawa et al., 1998b).
Contrarily, βERKO males have normal estrogen-induced aggression, and show increased aggression towards OBX male intruders in puberty and early adulthood (Nomura et al.,
2002). Thus, in male mice, activation of ERβ appears to decrease aggression in an experience and age dependent manner, while ERα seems to increase aggression.
In one of the rare studies on the neurobiology of female aggression, increased aggression towards ovariectomized (ovx), hormone-primed same-sex intruders, but not OBX males, was observed in virgin female αERKO mice, and this was reduced by estrogen replacement
42 following ovx (Ogawa et al., 1998a). Unlike αERKO females, gonadally intact female
βERKO mice show reduced aggression following T administration, and no aggression towards ovx female intruders (Ogawa et al., 2004; Ogawa et al., 1999). Thus it appears that opposite effects are found in males and females following chronic KO of the genes for ERα and ERβ, suggesting that estrogens may differentially mediate intrasexual aggression in the two sexes. This could be partially due to sex differences in the regulation of the ERs in different brain regions (Arteaga-Silva et al., 2007), or, since these studies used KO mice, possibly because of contributions of developmental vs activation effects, flanking genes and/or compensatory mechanisms (Choleris et al., 2008a; Waddington et al., 2005).
We have recently investigated the effects of acute pre-test administration of the ERβ agonist WAY-200070 on agonistic and social behaviour in gonadally intact and gonadectomized (gonadex) male and female CD1 mice (Clipperton-Allen et al., 2010). The responses of treated resident mice to gonadex, same-sex intruders were collected in the resident-intruder test with a comprehensive behavioural analysis to assess the effects of
WAY-200070 on different types of agonistic behaviour (Clipperton-Allen et al., 2010). Our results suggest that acute ERβ agonist administration to gonadally intact CD1 mice of both sexes can mediate dominance-related agonistic behaviours, while leaving overt attacks largely unaffected (Clipperton-Allen et al., 2010).
The effects of acute activation of ERα by PPT on agonistic interactions in mice are currently unknown. In the present study, we investigate these effects in normally developed, adult, gonadally intact and gonadex male and female CD1 mice. As we had previously found that males and females spend the same amount of time performing agonistic behaviours, but
43 that females use non-attack behaviours to attempt to establish dominance over a same-sex conspecific (Clipperton-Allen et al., 2010), we also compared gonadally intact animals of the two sexes, and same-sex gonadally intact and gonadex mice, in untreated controls to confirm these findings. As in Clipperton-Allen et al. (2010 [Ch 3]), we used a resident-intruder test, in which we included a comprehensive behavioural analysis in addition to standard measures of aggression.
2.3 Methods
2.3.1 Animal Housing
Adult CD1 mice (Mus musculus) were obtained from Charles River, Q.C., Canada. Mice used as residents were individually housed in clear polyethylene cages (26 x 16 x 12 cm) for two weeks in order to develop a home cage territory. Mice used as intruders were housed in same-sex group cages of four to six mice (46 x 26 x 21 cm). All cages were provided with tap water and food (Teklad Global 14% Protein Rodent Maintenance Diet, Harlan Teklad,
Madison, WI) ad libitum, environmental enrichment (plastic containers and paper nesting material) and corn cob bedding. The mice were held under a 12:12h reversed light/dark cycle
(lights on at 2000h) at a temperature of 21±1°C in the Central Animal Facility at the
University of Guelph.
2.3.2 Experimental Subjects
Subjects were 143 female and 141 male experimentally naïve CD1 mice. The supplier ovx 88 of the female mice and castrated 89 of the male mice, both at 2-3 months of age, 1-2 days prior to shipping to the University of Guelph, and at least 2 weeks before the commencement of the experiment. Vaginal cytology confirmed that the ovx mice were not
44 cycling. Thirty-two ovx and 32 castrated males were randomly selected to be intruders. To avoid the potential confounding effects of T producing and eliciting attacks in males (Luttge,
1972), and the variability of aggression produced by the estrous cycle (Hyde & Sawyer,
1977), group-housed, same-sex, gonadex intruders were used in order to have more
“standardized” intruders with uniformly low levels of circulating sex hormones.
Additionally, as little aggression is typically initiated by gonadex mice (e.g., Beeman, 1947;
Gray et al., 1978), the use of these mice as intruders allowed us to determine the level of agonistic behaviour initiated by the resident mouse, while avoiding the need to interrupt testing due to escalations in aggression. This enabled us to perform a full behaviour analysis throughout the duration of the test in all sexes and gonadal conditions. Intruders were replaced when they reached four months of age, following no more than four re-uses and with a minimum interval of 3 days between tests.
The remaining 111 female and 109 male mice were used as residents. Each resident experimental mouse was assigned to an independent treatment group, as outlined in Table 2, and no resident was reused. Intact female residents were at mixed stages of the estrous cycle
(see Table 2). This research was conducted in accordance with the regulations of the
Canadian Council on Animal Care and was approved by the Institutional Animal Care and
Use Committee of the University of Guelph.
2.3.3 Apparatus
Social interactions between residents and intruders took place in the resident’s home cage during the dark phase of the light/dark cycle. The interactions were videotaped from above through clear Plexiglas lids, using an 8 mm Sony Handycam (in Nightshot), for subsequent
45 Table 2. Distribution of treatment groups. Numbers indicate mice per group. Males Females Gonadally Gonadex Gonadex Treatment Intact (Castrated) Gonadally Intact (Ovx) n = 11 Uninjected n = 10 n = 12 n = 13 (3 P2, 6 E3, 2 D4) Sesame Oil n = 12 n = 12 n = 12 n = 11 Vehicle (3 P, 6 E, 2 D) 0.025 mg/kg n = 12 n = 11 n = 12 n = 12 PPT (3 P, 5 E, 3 D) 0.05 mg/kg n = 11 n = 11 n = 12 n = 13 PPT (2 P, 7 E, 1 D) n = 12 0.1 mg/kg PPT n = 12 n = 12 n = 12 (0 P, 9 E, 3 D) 1Ovx = ovariectomized. 2P = proestrus phase. 3E = estrus phase. 4D = diestrus phase.
46 behavioural analysis.
2.3.4 Drug
We used the selective estrogen receptor α agonist 1,3,5-tris(4-hydroxyphenyl)-4-propyl-
1H-pyrazole (PPT), which is 410 times more selective for ERα than ERβ in mice (Gray et al.,
1978), and was suspended in sesame oil.
For each sex and gonadal condition (gonadally intact males, gonadally intact females, castrated males, and ovx females), five independent groups of resident mice were formed: a sesame oil (vehicle) control (OIL; gonadally intact males n = 12, gonadally intact females n
= 12, castrated males n = 12, ovx females n = 11), three drug-treated groups, each receiving one of three doses of PPT: 0.025 mg/kg (LOW; gonadally intact males n = 11, gonadally intact females n = 12, castrated males n = 12, ovx females n = 12), 0.05 mg/kg (MID; gonadally intact males n = 11, gonadally intact females n = 11, castrated males n = 12, ovx females n = 13), or 0.1 mg/kg (HIGH; gonadally intact males n = 12, gonadally intact females n = 12, castrated males n = 12, ovx females n = 12), and an uninjected control group
(gonadally intact males n = 10, gonadally intact females n = 11, castrated males n = 12, ovx females n = 13). This last group was included in order to assess potential effects of sesame oil on behaviour (Anton et al., 1974; Clipperton-Allen et al., 2010; Colafranceschi et al.,
2007; Curtis & Wang, 2005), and to allow comparisons between the sexes and between gonadally intact and gonadex mice in untreated animals. The distribution of doses in each sex and gonadal condition, and the estrous cycle phase of the gonadally intact females, are outlined in Table 2. Injections were administered subcutaneously (s.c.) at 10 ml/kg 48h prior to the experiment to allow the drug to fully exert its genomic and other effects. Each
47 treatment group consisted of a unique set of resident mice, and each resident mouse was tested only once. This pattern of administration parallels the administration procedure that has shown effects of this drug on social and non-social learning tasks in rats and mice, and on antianxiety behaviour in rats (e.g., Clipperton et al., 2008; Walf et al., 2004). The specific doses were chosen on the basis of previously effective doses in rats (10 µg of PPT, or approximately 0.03 mg/kg PPT; e.g., Walf et al., 2004) and mice (Clipperton et al., 2008).
2.3.5 Procedure
All resident mice were individually housed for 7 days prior to testing, with no cage changes taking place in the last 3 or more days, to allow the establishment of a territory by each resident. Two days before testing, the resident mice of each of the four conditions
(gonadally intact males, gonadally intact females, castrated males, and ovx females) were randomly assigned to one of the five groups (uninjected control, sesame oil vehicle, 0.025 mg/kg PPT, 0.05 mg/kg PPT, or 0.1 mg/kg PPT) and treated. At least 12h before testing, all residents and intruders were moved into the testing room and left undisturbed. At this time, to assist in identification, the intruder mice were coloured with black magic marker.
During their active phase, each resident mouse had a same-sex, gonadex, age-matched intruder placed into his or her home cage. The mice were left undisturbed to freely interact for 15 min while being videotaped.
The videotaped social interactions were scored for 21 behaviours based on the ethogram by Grant and Mackintosh (1963; see Table 3 for behaviour descriptions, modified from
Clipperton et al., 2008; Clipperton-Allen et al., 2010) by two trained observers, who were unaware of the animals’ treatment. The Observer Video Analysis software (Noldus 48 Information Technology, Wageningen, the Netherlands) was used to score the videotapes, and the behavioural analysis focused on the treated mouse, the resident. The behaviour of the intruder was collected only in relation to the behaviour of the resident (i.e., Attacks Received,
Reciprocal Attacks, Aggressive Postures, Social Inactivity), and in the reciprocal pairs of behaviours (Dominant/Submissive Behaviour, Chasing/Avoidance of the Intruder).
Dominant behaviour on the part of one mouse was typically met by submissive behaviour on the part of the other.
In order to assess potential non-specific effects of the drug on the total activity and total social and non-social explorations of the mice, as well as on the overall levels of agonistic behaviour displayed during the interaction, 10 categories of individual behaviours were formed (see Table 4, modified from Clipperton et al., 2008 and Clipperton-Allen et al.,
2010). These categories were created to provide an overall view of the animals’ behaviour, and also allow for comparison to non-ethological types of analysis (e.g., Cushing et al.,
2008). However, with this paper we wish to emphasize the importance of looking at the component behaviours within these categories, as this will determine whether the effects observed are due to specific individual behaviours or a general effect across a category. This is particularly true in the case of agonistic behaviours, as an increase in total levels of agonistic behaviour may be due to effects on types of aggression-related behaviour that have potentially different functional implications, such as overt attacks for the establishment of territories, and/or behaviours aimed at establishing a dominance
49 Table 3. Descriptions of scored behaviours. Behaviour Description Social Behaviours Physical attacks which include box/wrestle, offensive and defensive Aggressive Postures postures, lateral sideways threats and tail rattle. Physical attacks with a locked fight including tumbling, kick-away Reciprocal Attacks and counterattack where the attacker cannot be identified. The resident mouse actively follows, or pursues and chases the Chasing the Intruder intruder; reciprocal to avoid. The resident mouse is in control; includes pinning of the intruder, Dominant Behaviour aggressive grooming, crawling over/on top, and mounting attempt. Physical attacks, including dorsal/ventral bites. Only the frequency Attacks Delivered of attacks was measured. Avoidance of the The resident withdraws and runs away from the intruder while the Intruder intruder is chasing. The intruder mouse is in control; includes crawl under, supine Submissive posture (ventral side exposed), prolonged crouch, and any other Behaviour behaviour in which the intruder is dominant (e.g., the intruder pins, aggressively grooms, etc., the resident). Physical attacks including bites to dorsal/ventral regions. Only the Attacks Received frequency of attacks was measured. Defensive Upright Species-typical defensive behaviour; upright with the head tucked Posturing and the arms ready to push away. Social Inactivity Includes sit/lie/sleep together. Oronasal Active sniffing of the intruder’s oronasal area. Investigation Body Investigation Active sniffing of the intruder’s body. Anogenital Active sniffing of the intruder’s anogenital region. Investigation Risk assessment behaviour; back feet do not move and front feet Stretched Approaches approach the intruder. Only the frequency of stretched approaches was measured. Often from across the cage; the resident’s attention is focused on the Approaching and/or intruder, head tilted toward the intruder and movements toward the Attending to the intruder; this becomes “Chasing the Intruder” once along the tail or Intruder sniff if within 1.5 cm of the intruder. Non-social Behaviours Horizontal Movement around the cage; includes active sniffing of air and Exploration ground. Movement to investigate upwards, both front feet off the ground; Vertical Exploration includes sniffing, wall leans and lid chews (less than 3). Digging Rapid stereotypical movement of forepaws in the bedding. Abnormal “Strange” behaviours, including spinturns, repeated jumps/lid Stereotypies chews/head shakes (more than 3). Solitary Inactivity No movement; includes sit, lie down and sleep. Self-Grooming Rapid movement of forepaws over facial area and along body.
50 Table 4. Descriptions of grouped and composite behaviours. Composite Behaviour Behaviours Included All behaviours involving activity, both social and non-social. Total Activity Excluded from this group are Inactive Alone, Inactive Together, and Self-Groom. Follow Intruder, Dominant Behaviour, Attack Delivered, Aggressive Postures, Reciprocal Attacks, Avoid Intruder, Total Social Submissive Behaviour, Attack Received, Defensive Upright Behaviours Posturing, Inactive Together, Oronasal Investigation, Body Investigation, Anogenital Investigation, Stretched Approaches, and Attend To/Approach Intruder. Follow Intruder, Dominant Behaviour, Attack Delivered, Avoid Intruder, Submissive Behaviour, Attack Received, Defensive Upright Posturing, Aggressive Postures, and Reciprocal Attacks. Total Agonistic This composite behaviour represents the overall levels of agonism Behaviours present in the resident-intruder interactions and does not indicate the direction of the agonistic behaviour (i.e., whether agonistic behaviour is directed toward the resident or toward the intruder). Agonistic Behaviour Follow Intruder, Dominant Behaviour, and Attack Delivered. Delivered Agonistic Behaviour Avoid Intruder, Submissive Behaviour, Attack Received, and Received Defensive Upright Posturing. Total agonistic behaviour delivered minus total agonistic behaviour received. A negative score indicates that the resident was the Dominance Score submissive animal in the pair, while a positive score signifies that the resident was the dominant animal. Oronasal Investigation, Body Investigation, Anogenital Social Investigation Investigation, Stretched Approaches, and Attend To/Approach Intruder. Horizontal Exploration, Vertical Exploration, Dig, Stereotypies, Non-social Behaviour Inactive Alone, and Self-Groom. Non-social Locomotor Horizontal Exploration, Vertical Exploration, and Dig. Behaviour Non-social Non- Inactive and Self-Groom. locomotor Behaviour
51 hierarchy within a shared territory. The direction of the agonistic behaviour was defined whenever possible, and a “Dominance Score” was also calculated for each pair of mice by subtracting the amount of Agonistic Behaviour Received from the amount of Agonistic
Behaviour Delivered (as described in Table 4 and modified from Clipperton et al., 2008;
Clipperton-Allen et al., 2010). The Dominance Score thus took into account the reciprocity of “dominant” and submissive behaviour involved in agonistic interactions, and a high positive Dominance Score means the resident delivered the majority of the agonistic behaviour, while receiving very little, and thus is obviously dominant over the other.
To determine the phase of the estrous cycle in intact females, and to confirm that the ovx mice were not cycling, vaginal smears were taken from all females immediately following testing. These slides were stained with Giemsa stain (Sigma-Aldrich, Oakville, ON) and examined under a microscope using 100x magnification. Proestrus was defined as consisting of predominantly nucleated cells, estrus of primarily non-nucleated, cornified cells, and diestrus of chiefly leukocytes.
2.3.6 Statistical Analysis
In order to assess the time course of the behaviours, the 15 min social interaction was divided into three 5 min intervals. The frequency (FREQ), duration (DUR), and latency
(LAT) of each behaviour were measured. If a behaviour was not performed, a latency of 900s was assigned. Levene’s test for equality of variance revealed that there were significant differences between group variances for many behaviours (Levene statistics between 13.10-
2.56, p values from less than 0.001 to 0.0496), therefore violating the assumption of equality of variance necessary for parametric statistics. Results were thus analyzed using
52 non-parametric statistics, which do not require assumptions to be made about the underlying population. Specifically, the Kruskal-Wallis test (the non-parametric equivalent of analysis of variance) was used, followed by the non-parametric Mann Whitney U test for binary comparisons. Additionally, as the only castrated males to perform Attacks Delivered were those treated with PPT, the data from PPT-treated mice were pooled and compared to zero
(the number of Attacks Delivered by the vehicle-treated mice) using the Wilcoxon Signed
Ranks test (a non-parametric test that can be used like a one-sample t-test).
Two sets of analyses were performed to assess 1) the effects of drug treatment on each mouse condition separately, and 2) differences between sexes and between intact and gonadex same-sex mice in untreated animals. First, for each condition (gonadex or intact male or female mice), the effects of the drug are reported in comparison to the sesame oil control group. Effects of sesame oil per se are reported in comparison to the uninjected group. The variable of interest in this analysis was the treatment (untreated controls, sesame oil, 0.025 mg/kg PPT, 0.05 mg/kg PPT, 0.1 mg/kg PPT). A second variable was included for the gonadally intact females, the phase of the estrous cycle (proestrus, estrus, diestrus).
The second set of analyses was performed to assess sex differences and to compare gonadex and gonadally intact mice in the uninjected control groups. Variables were sex
(gonadally intact male, gonadally intact female) and gonadal condition (gonadally intact, gonadex; analyzed separately for each sex). In the gonadally intact females, effects of the estrous cycle (proestrus, estrus, diestrus) were also analyzed.
In all of the analyses in which it was assessed, the phase of the estrous cycle did not have significant main effects. Additionally, the number of mice in each treatment group at each 53 phase of the cycle was very low. Therefore, any reported drug effects are collapsed across estrous cycle phases for the gonadally intact females. When the results of the analyses on
DUR and FREQ were in agreement, only those of DUR are reported; thus, if not specified, results are DUR. If the time course data were reflected in the total data, they are not reported.
All analyses were performed using SPSS 13.0 for Windows (SPSS Inc., Chicago, IL).
2.4 Results
For each experimental group (uninjected animals, drug treated gonadally intact or gonadex males or females), we report the key findings below. Additional statistically significant results are reported in detail in Appendix 1. We found several differences between sexes and between intact and gonadex mice. As expected, gonadally intact males attacked the intruder more than females, but females actually spent more time engaged in agonistic behaviour than males. Males also switched between behaviours more frequently
(“behavioural shifting”), and spent more time performing non-social behaviours (any behaviour in which there is no interaction with the intruder, e.g., horizontal exploration, self- grooming; see Table 4). As we have found previously (Clipperton-Allen et al., 2010), gonadectomy appears to make mice behave more like gonadally intact animals of the opposite sex: castrated males showed no attacks, less behavioural shifting and less non-social behaviour than males, and ovx females showed more behavioural shifting and non-social behaviour than females, while leaving agonistic behaviour unaffected. We also found that, overall, PPT had few effects in gonadally intact mice, but appears to increase sex-typical agonistic behaviour (i.e., attacks in males and other dominance behaviours in females) in gonadex mice. Please see Table 5 for a summary of the main results of PPT treatment.
54 Table 5. Overview of PPT effects. Behaviour Malesa Femalesb Castratedc Ovxd LAT Total Agonistic Behaviours — — Agonistic Behaviour Delivered — — — — Attacks Delivered — — — Dominant Behaviour — — — Agonistic Behaviour Received — — — Avoidance of the Intruder — — Social Investigation — — — Anogenital Investigation — — — Non-social Non-locomotor Behaviours — — a “Males” indicates gonadally intact males. b “Females” indicates gonadally intact females. c “Castrated” indicates castrated males. d “Ovx” indicates ovariectomized females. indicates a decrease in the behaviour. indicates an increase in the behaviour.
55 2.4.1 Sex Differences in Gonadally Intact Mice
As has been well documented (e.g., Scott & Fredericson, 1951), males performed more frequent Attacks Delivered (U = 30.00, z = -2.21, p = 0.027; Figure 2b) and more Aggressive
Postures (U = 26.00, z = -2.34, p = 0.019) than females. However, females actually spent more time performing Agonistic Behaviour Delivered than males (U = 25.00, z = -2.11, p =
0.035; Figure 2d), although this measure does not include attacks, which were scored without durations; when attacks were included (in the FREQ measures), there were no sex differences in Agonistic Behaviour Delivered (Figure 2a). While both males and females had a positive
Dominance Score, indicating that they delivered more agonistic behaviour than they received, the females had a higher Dominance Score than the males (U = 25.00, z = -2.11, p
= 0.035; Figure 2e).
Males performed more FREQ Approaching and/or Attending to the Intruder (U = 18.00, z
= - 2.61, p = 0.009; Figure 2c), more behavioural shifting (U = 15.00, z = -2.82, p = 0.005;
Figure 2f), and more Non-social Behaviours (U = 9.00, z = -3.24, p = 0.035; Figure 2g) than females. Please see Appendix 1, Table 6 for additional results.
2.4.2 Differences Between Gonadally Intact and Castrated Males
In the uninjected animals, castrated males showed more “female-like” behaviour than gonadally intact males, i.e., their pattern of behaviour was more similar to that of gonadally intact females. Like the intact females, castrated males showed less behavioural shifting (U =
15.50, z = -2.94, p = 0.003; Figure 2f) and Approaching and/or Attending to the Intruder
(FREQ: U = 19.00, z = -2.71, p = 0.007; Figure 2c). Castrated males also performed fewer
Total Non-social Behaviours, in which they were not in contact with the intruder (FREQ: U =
56 80 a Agonistic Behaviour Delivered d Agonistic Behaviour Delivered 300 * 60
200 40 Frequency Duration 100 20
0 (s) 0 20 b Attacks Delivered e Dominance Score 300 * 15
200 10 Frequency Duration 100 5
* * (s) 0 0 500 c Approaching and/or Attending f Behavioural Shifting 80 to the Intruder 400 + 60 * * * * 300
40 200 Frequency Frequency
20 100
0 0 Male Castrated Female Ovx g Total Non-social Behaviours 200
Gonadally Intact Males 150 + Castrated Males * Gonadally Intact Females 100 *
Ovariectomized Females Frequency 50
0 Male Castrated Female Ovx Figure 2. Sex differences and gonadectomy effects on agonistic behaviours. a) Frequency of agonistic behaviour delivered. b) Frequency of attacks delivered. c) Frequency of approaching and/or attending to the intruder. d) Duration of agonistic behaviour delivered. e) Dominance score (duration). f) Behavioural shifting. g) Frequency of total non-social behaviours. * significant difference vs gonadally intact males, p < 0.05. + significant difference vs gonadally intact females, p < 0.05.
57 20.00, z = - 2.64, p = 0.008; Figure 2g), and, as expected (e.g., Beeman, 1947), none of the
“male-typical” Attacks Delivered (U = 32.50, z = -2.31, p = 0.021; Figure 2b), thus again paralleling the observed sex differences in intact animals. Please see Appendix 1, Table 7 for additional results.
2.4.3 Differences Between Gonadally Intact and Ovariectomized Females
Similar to the differences between castrated and gonadally intact males, ovx females showed more “male-like” patterns of behaviour than gonadally intact females, with their behavioural results more closely resembling those of gonadally intact males. Ovx females showed more behavioural shifting (U = 35.00, z = -2.12, p = 0.034; Figure 2f), and performed more Total Non-social Behaviours (FREQ: U = 33.00, z = -2.23, p = 0.026; Figure
2g) than gonadally intact females. However, unlike in the comparison between gonadally intact males and females, no differences between gonadally intact and ovx females were observed on any specific agonistic behaviours. For additional results, please see Appendix 1,
Table 8.
2.4.4 Effects of PPT in Gonadally Intact Males
LOW increased Agonistic Behaviour Received (FREQ: U = 26.00, z = -2.49, p = 0.013; see Figure 3c), likely due to increased Avoidance of the Intruder (U = 28.00, z = -2.48, p =
0.013; see Figure 3d); the LAT to these behaviours was also reduced (Agonistic Behaviour
Received: U = 24.00, z = -2.62, p = 0.009; Avoidance of the Intruder: U = 32.00, z = -2.22, p
= 0.027). Neither Attacks Delivered (see Figure 3b) nor overall levels of Agonistic
Behaviour Delivered were affected by PPT. Few other behaviours were affected. Please see
Appendix 1, Table 9 for additional results.
58 Figure 3. PPT effects on gonadally intact and castrated males. a) Latency to total agonistic behaviour in gonadally intact males. b) Frequency of attacks delivered in gonadally intact males. c) Duration of agonistic behaviour received in gonadally intact males. d) Duration of avoidance of the intruder in gonadally intact males. e) Frequency of non-social non- locomotor behaviour in gonadally intact males. f) Latency to total agonistic behaviour in castrated males. g) Frequency of attacks delivered in castrated males. h) Duration of agonistic behaviour received in castrated males. i) Duration of avoidance of the intruder in castrated males. j) Frequency of non-social non-locomotor behaviour in castrated males. *significant difference vs vehicle, p < 0.05. **significant difference vs vehicle, p < 0.01. †significant difference vs zero, p < 0.05.
59 2.4.5 Effects of PPT in Gonadally Intact Females
There were no major effects of PPT on behaviours in gonadally intact females. Minor results are listed in Appendix 1, Table 10.
2.4.6 Estrous Cycle Effects
Due to the extremely low numbers of mice in each treatment group and phase of the cycle
(see Table 2), the minor effects of the estrous cycle on each treatment group are likely not meaningful, and thus are only reported in Appendix 1, Table 11.
2.4.7 Effects of PPT in Castrated Males
PPT decreased the LAT to Total Agonistic Behaviours in castrated mice (LOW: U =
30.00, z = -2.43, p = 0.015; MID: U = 14.00, z = -3.35, p = 0.001; see Figure 3f), and MID increased the FREQ of Non-social Non-locomotor Behaviours (U = 35.50, z = -2.11, p =
0.035; see Figure 3j), which include Solitary Inactivity and Self-Grooming. Additionally, as the only castrated males showing Attacks Delivered were PPT-treated, data were collapsed across the three doses of PPT and the Wilcoxon Signed Ranks test showed that PPT-treated mice performed more FREQ Attacks Delivered (Z = -2.37, p = 0.018), and had a shorter LAT to perform them (Z = -5.55, p < 0.001), than the OIL mice, who performed no Attacks
Delivered (see Figure 3g). PPT thus slightly increased agonistic behaviour in castrated males.
Please see Appendix 1, Table 12 for additional results.
2.4.8 Effects of PPT in Ovariectomized Females
In ovx females, HIGH decreased Avoidance of the Intruder (U = 36.00, z = -2.04, p =
0.041; see Figure 4f). LOW increased Dominant Behaviour at 0-5 min (U = 32.00, z = -2.09,
60 Figure 4. PPT effects on gonadally intact and ovariectomized (ovx) females. a) Duration of dominant behaviour at 0-5 min in gonadally intact females. b) Duration of avoidance of the intruder in gonadally intact females. c) Duration of social investigation in gonadally intact females. d) Duration of Anogenital investigation in gonadally intact females. e) Duration of dominant behaviour at 0-5 min in ovx females. f) Duration of avoidance of the intruder in ovx females. g) Duration of social investigation in ovx females. h) Duration of anogenital investigation in ovx females. * significant difference vs vehicle, p < 0.05.
61 p = 0.036; Figure 4e), the time at which most dominance hierarchies are beginning to be established, indicating that these residents may have been more motivated to assert their dominance, and that PPT increased agonistic behaviour. LOW also increased Social
Investigation (U = 25.00, z = - 2.52, p = 0.012; Figure 4g), likely due to increased Anogenital
Investigation (U = 34.00, z = -1.97, p = 0.049; Figure 4h) in LOW-treated mice. Additional minor results are listed in Appendix 1, Table 13.
2.5 Discussion
Our comprehensive behavioural assessment of the resident-intruder test allowed us to investigate the effects of the ERα agonist PPT on agonistic, non-agonistic, social and non- social mouse behaviour in response to a gonadex, same-sex intruder, and to show that acute activation of ERα increases sex-typical agonistic behaviour in gonadex but not gonadally intact male and female mice (see Table 5). In addition, the use of an uninjected control group enabled us to compare the sexes and gonadal conditions, and to confirm our previous findings (Clipperton-Allen et al., 2010) that CD1 females respond to their opponents using agonistic behaviours other than attacks, but spend as much, if not more, time in agonistic interactions with same-sex, gonadex intruders as males.
2.5.1 Sex and Gonadectomy Differences
Among the numerous sex differences observed, the gonadally intact male CD1 mice attacked and made more aggressive postures towards the same-sex gonadex intruder than did the gonadally intact females, as has often been seen in the literature (e.g., Anton et al., 1974;
Clipperton-Allen et al., 2010; Edwards, 1970). As these are the typical measures of agonistic behaviour studied in the mouse resident-intruder test, and are performed almost exclusively
62 by males, it is not surprising that females are thus often considered to be unaggressive.
However, when other types of non-attack agonistic behaviour were included in the analysis, such as pushing down the intruder or aggressive allogrooming (Clipperton et al., 2008;
Clipperton-Allen et al., 2010; Grant & Mackintosh, 1963), then females were actually more agonistic than males, and also had a higher dominance score, i.e., a more pronounced difference between delivered and received agonistic behaviour (although in both sexes the resident delivered more and received less agonistic behaviour than the intruder). While these and other results (e.g., Clipperton-Allen et al., 2010) indicate that the overall levels of intrasexual agonistic behaviour in male and female CD1 mice are similar, the function of these interactions is likely different. Females may engage in extended sessions of dominant agonistic behaviour in order to establish a dominance hierarchy, thus allowing them to share territory with other females, while males may fight to regulate space use and establish exclusive territories (Scott & Fredericson, 1951). This is consistent with the tendency of males to attack in short bursts and then retreat to the opposite sides of the cage (Ginsburg &
Allee, 1942; Miczek et al., 2001), and with the present finding that the males showed higher level of non-social behaviour than females.
As we have previously observed (Clipperton-Allen et al., 2010), we found that male mice shifted between behaviours and approached the intruder more often than females did. This suggests that the presence of a same-sex intruder may elicit a greater general arousal in the males due to the social challenge the intruders represent to these territorial animals (Branchi et al., 2006; Clipperton-Allen et al., 2010; Pfaff et al., 2008). Consistently, in our previous results (Clipperton-Allen et al., 2010 [Ch 3]), the less territorial castrated males approached the intruder less than gonadally intact males. Hence, it is possible that the differences in the
63 type of agonistic behaviours displayed by the sexes may be related to different motivational responses to unknown intruders.
Several differences were also noted between gonadally intact and gonadex uninjected
CD1 mice. In agreement with the literature on several strains of mice, castrated males in the present study showed the typical absence of attacks (e.g., Beeman, 1947; Clipperton-Allen et al., 2010; Luttge, 1972), but ovx females did not differ from intact females on agonistic behaviours. Behavioural shifting and non-social behaviour were lower in castrated males than gonadally intact males, but higher in ovx than gonadally intact females. Additionally, like gonadally intact females and in our previous results (Clipperton-Allen et al., 2010), castrated males show reduced approaching of the intruder. Both sexes thus appear to show patterns of behaviour more similar to those of gonadally intact animals of the opposite sex
(i.e., castrated males like gonadally intact females, ovx females like gonadally intact males) following the removal of gonadal hormones, as we have found previously (Clipperton-Allen et al., 2010).
2.5.2 PPT Effects in Gonadectomized Mice
In castrated males, PPT induced agonistic behaviour, as the only castrated males that delivered attacks were those treated with PPT, and PPT also reduced the latency to both agonistic behaviour and delivered attacks in these mice. Consistently, research with αERKO and βERKO male mice also suggests that ERα increases and ERβ decreases agonistic behaviour (Nomura et al., 2006; Ogawa et al., 1998b; Ogawa et al., 1998a). Furthermore,
ERα also increased another behaviour performed more by gonadally intact than castrated males, as PPT-treated castrated males in this study spend more time in solitary, non-
64 locomotor behaviours (e.g., self-grooming, solitary inactivity); males will frequently retreat to opposite sides of the cage following an agonistic interaction with a castrated intruder
(Ginsburg & Allee, 1942; Miczek et al., 2001). However, while ERα and ERβ agonists both increased solitary or non-social behaviour, the ERα agonist increased those non-social behaviours that were stationary, like self-grooming or solitary inactivity, whereas the ERβ agonist instead increased non-social behaviours that involved locomotion, such as horizontal or vertical exploration (Clipperton-Allen et al., 2010).
PPT also increased agonistic behaviour in ovx females, who performed slightly more behaviours aimed at establishing a dominance hierarchy, a type of agonistic behaviour that is more typical of females (Clipperton-Allen et al., 2010), and this was paralleled by decreased avoidance of the intruder. This differs from the results of female KO mice, which suggested that ERα reduces and ERβ increases aggression and the KO of ERβ reduced it (Ogawa et al.,
1998a; Ogawa et al., 2004; Ogawa et al., 1999), and could imply different developmental and activational effects of these compounds. In our previous work, acute activation of ERβ increased non-attack agonistic behaviour, which suggests that regulation of overt attacks and non-overt agonistic behaviours aimed at establishing a dominance hierarchy may not completely overlap (Clipperton-Allen et al., 2010), and that ERα may act differently on behaviour aimed at establishing a dominance hierarchy than on the more typically studied overt attacks.
2.5.3 PPT Effects in Gonadally Intact Mice
In gonadally intact mice, PPT did not modulate delivery of agonistic behaviour or attacks, unlike an ERβ agonist, which increased agonistic behaviour (Clipperton-Allen et al.,
65 2010 [Ch 3]). The lack of pro-aggression effects of PPT in gonadally intact animals may be due to concurrent activation of ERβ by endogenous hormones and/or the naturally circulating estrogens may be competing for binding sites on the receptors. As is the case for other physiological and behavioural processes (Gustafsson, 2006a), ERβ may be acting on agonistic behaviour in opposition to ERα (as reviewed in Choleris et al., 2008a).
Consistently, pharmacological activation of ERα leads to enhanced agonistic behaviour in gonadex males (in which ERβ is not endogenously activated), but an ERβ agonist does not increase agonistic responses in castrated males, which have no ERα activation (Clipperton-
Allen et al., 2010) .
The opposing effects of ERα and ERβ may also explain the U- or inverted-U shaped dose-response curve observed in several behaviours affected by PPT in all sexes and gonadal conditions, as it is possible that this is due to non-selective binding by the high dose, thus counteracting the effect of the targeted receptor. These curves are not uncommon with estrogens and ER-specific agonists, as we and others have previously found similar patterns of effects with estradiol on wheel-running activity (unpublished data), and the ERβ agonists
WAY-200070 and DPN (Clipperton et al., 2008; Neese et al., 2010).
2.5.4 ERα and ERβ in Mouse Agonistic Behaviour
Taken together, our previous findings and those of the current study show that in gonadex animals, PPT increases sex-typical agonistic behaviours, attacking for castrated males and dominance behaviours for ovx females, while the ERβ agonist WAY-200070 did not affect any agonistic behaviours in gonadex mice (Clipperton-Allen et al., 2010). In gonadally intact animals the pattern is reversed, as the ERβ agonist WAY-200070 increased non-attack
66 agonistic behaviour in both sexes (Clipperton-Allen et al., 2010), while PPT had no effects on agonistic behaviour in gonadally intact male and female mice. These effects of selective
ERα and ERβ activation suggest that while ERα seems to be necessary for the activation of agonistic responses, ERβ may act to fine-tune such responses (Ogawa, 2004). This interpretation would be consistent with similar roles of the ERs in other social behaviours
(Choleris et al., 2006; Choleris et al., 2009; Choleris et al., 2008a).
One possible explanation for our finding that ERα affected gonadex mice and ERβ affected gonadally intact animals is that the activation of both ERs by endogenous hormones may make the gonadally intact mice less responsive to PPT stimulation of ERα because of
ERβ-mediated reductions in ERα activity. While ERα may normally have higher transcriptional activity than ERβ (Cowley et al., 1997), at physiological levels of estrogens
(as in gonadally intact mice), ERβ suppresses the efficiency and potency of ERα’s response to estrogens (Hall & McDonnell, 1999), such that coexpression of both ERs results in an intermediate level of transcription (Cowley et al., 1997). This may be due to reciprocal transcriptional regulation of the two receptors (e.g., Nomura et al., 2003), or to the way the two receptors function after binding with estradiol, as hetero- or homodimers (Hall &
McDonnell, 1999). ERβ show low levels of homodimerization (Cowley et al., 1997) and may need to form heterodimers to function properly (Couse et al., 1997). ERα homodimers and
ERα/ERβ heterodimers show higher affinities for DNA than the ERβ homodimer (Cowley et al., 1997). These data could explain why, in the present study, selective pharmacological activation of ERα is effective in gonadex animals, with sub-physiological levels of estrogens and little or no ERβ activation, while ERβ activation has more effects in animals with natural levels of circulating gonadal hormones (Clipperton-Allen et al., 2010). 67 2.5.5 Vehicle Effects
As in our previous study (Clipperton-Allen et al., 2010), mice that received the common gonadal hormone vehicle sesame oil showed some behavioural differences when compared to the untreated mice. In males, both sesame oil-treated gonadally intact and gonadex males showed a decrease in approaching the intruder. Gonadally intact males treated with sesame oil also spent less time in horizontal exploration, while castrated males received more agonistic behaviour and spent less time in social investigation.
We found that females showed more vehicle effects than males, with both gonadex and gonadally intact females showing reduced dominance scores (i.e., less of a difference between the amount of agonistic behaviour delivered and received by the resident) and more non-social non-locomotor behaviours when treated with sesame oil. In some behaviours, including total activity, sesame oil acted in opposite directions depending on the gonadal condition of the female mice. A number of agonistic behaviours were reduced in gonadally intact females, while social investigative behaviours were either increased or decreased.
Sesame oil-treated ovx females received more agonistic behaviour relative to untreated females. As there are few studies that include an uninjected control group, many vehicle effects may be overlooked, particularly if a vehicle other than physiological saline is employed. The few studies to use an uninjected control group or to directly examine the effects of sesame oil have found effects with different modes of administration (oral, intraperitoneal, s.c.; Anton et al., 1974; Clipperton-Allen et al., 2010; Colafranceschi et al.,
2007; Curtis & Wang, 2005) and in various species. Sesame oil is known to have metabolic effects, although in the case of the current study, it seems unlikely that these effects would persist 48h after administration. Similarly, this time frame makes it unlikely that these effects
68 are the result of injection stress.
2.5.6 Conclusions
The use of an ethological analysis allowed for a comprehensive assessment of ERα’s role in agonistic interactions with same-sex conspecifics in CD1 mice. Our current and previous results suggest that activation of ERα will increase sex-typical agonistic behaviour, i.e., attacks in males and other dominance behaviours in females, but only in gonadex mice, while non-attack agonistic behaviour is increased by ERβ activation in gonadally intact mice only
(Clipperton-Allen et al., 2010). This may be due to different homo- and heterodimer efficiency and/or binding affinity, or to regulation of ERα by ERβ. When the present results and those of αERKO and βERKO research are taken together, they suggest that in females,
ERα may have different activational effects than developmental ones, or the receptor may act differently on overt attacks than on other dominance behaviours, possibly by modulating the different underlying motivations (Choleris et al., 2009; Choleris et al., 2008a). Clearly, more investigations on female agonistic behaviour are needed before the involvement of ERα and
ERβ can be fully understood, potentially by employing different selective agonists and antagonists as they are developed by the pharmaceutical industry.
Our observations that CD1 female mice are as agonistic as males, but use behaviours likely aimed at establishing dominance rather than overt attacks is in agreement with our previous findings (Clipperton-Allen et al., 2010). However, additional research is needed to determine the specific developmental and activational effects of ERα and ERβ in female intrasexual aggression and agonistic behaviours, especially in light of the attack- or maternal aggression-focused approaches typically used in previous research with female αERKOs and
69 βERKOs (Ogawa et al., 2004; Ogawa et al., 1999). Taken together with previous results on agonistic behaviour, social recognition and social learning (Choleris et al., 2006; Choleris et al., 2008a; Choleris et al., 2009; Clipperton et al., 2008; Clipperton-Allen et al., 2010), our findings suggest that the function of ERα may be to activate or inactivate social behaviour, while ERβ may play a more modulatory role.
70 2.6 Appendix 1: Supplementary Material
Table 6. Differences between gonadally intact male and female uninjected mice. Behaviour Effect FREQ1: Malesa > Femalesb (U = 15.00, z = -2.82, p = 0.005) and all Total Activity timepoints Total Social DUR2: Females > Males (U = 9.00, z = -3.24, p = 0.001) and all timepoints Behaviours Approaching FREQ: Males > Females (U = 18.00, z = -2.61, p = 0.009) and/or FREQ: Males > Females at 0-5 min (U = 17.00, z = -2.68, p = 0.007) Attending to FREQ: Males > Females at 5-10 min (U = 24.50, z = -2.15, p = 0.031) the Intruder DUR: Males > Females (U = 26.00, z = -2.34, p = 0.019) Aggressive DUR: Males > Females at 5-10 min (U = 27.50, z = -2.59, p = 0.010) Postures FREQ: Males > Females (U = 26.50, z = -2.30, p = 0.022) FREQ: Males > Females at 5-10 min (U = 27.50, z = -2.59, p = 0.010) Dominance DUR: Females > Males (U = 25.00, z = -2.11, p = 0.035) Score Agonistic Behaviour DUR: Females > Males (U = 25.00, z = -2.11, p = 0.035) Delivered DUR: Females > Males (U = 22.00, z = -2.33, p = 0.020) Chasing the DUR: Females > Males at 0-5 min (U = 19.00, z = -2.53, p = 0.011) Intruder FREQ: Females > Males (U = 24.50, z = -2.15, p = 0.031) FREQ: Females > Males at 0-5 min (U = 24.00, z = -2.19, p = 0.028) Attacks FREQ: Males > Females (U = 30.00, z = -2.21, p = 0.027) Delivered Social DUR: Females > Males (U = 26.00, z = -2.04, p = 0.041) Investigation DUR: Females > Males at 0-5 min (U = 24.00, z = -2.18, p = 0.029) Oronasal FREQ: Males > Females (U = 21.00, z = -2.40, p = 0.017) Investigation DUR: Females > Males (U = 18.00, z = -2.61, p = 0.009) Anogenital DUR: Females > Males at 0-5 min (U = 11.00, z = -3.10, p = 0.002) Investigation FREQ: Females > Males (U = 21.50, z = -2.36, p = 0.018) FREQ: Females > Males at 0-5 min (U = 8.50, z = -3.28, p = 0.001) Total Non- DUR: Males > Females (U = 9.00, z = -3.24, p = 0.001) and all timepoints social FREQ: Males > Females (U = 5.00, z = -3.52, p < 0.001) and all timepoints Behaviours Non-social DUR: Males > Females (U = 13.00, z = -2.96, p = 0.003) and all timepoints Locomotor FREQ: Males > Females (U = 10.00, z = -3.17, p = 0.002) and all Behaviours timepoints DUR: Males > Females (U = 14.00, z = -2.89, p = 0.004) Horizontal DUR: Males > Females at 0-5 min (U = 8.00, z = -3.31, p = 0.001) Exploration DUR: Males > Females at 5-10 min (U = 15.00, z = -2.82, p = 0.005)
FREQ: Males > Females (U = 7.50, z = -3.35, p = 0.001) and all timepoints
71 DUR: Males > Females (U = 21.00, z = -2.39, p = 0.017) DUR: Males > Females at 0-5 min (U = 14.50, z = -2.87, p = 0.004) Vertical DUR: Males > Females at 5-10 min (U = 27.00, z = -1.97, p = 0.049) Exploration FREQ: Males > Females (U = 11.00, z = -3.10, p = 0.002) and all timepoints LAT3: Females > Males (U = 24.00, z = -2.18, p = 0.029) DUR: Males > Females at 5-10 min (U = 29.00, z = -2.10, p = 0.036) Digging FREQ: Males > Females at 5-10 min (U = 29.50, z = -2.06, p = 0.039) LAT: Females > Males (U = 27.00, z = -2.01, p = 0.045) DUR: Males > Females at 0-5 min (U = 27.00, z = -2.09, p = 0.037) Non-social FREQ: Males > Females (U = 25.00, z = -2.12, p = 0.034) Non-locomotor FREQ: Males > Females at 0-5 min (U = 25.00, z = -2.24, p = 0.025) Behaviours FREQ: Males > Females at 10-15 min (U = 25.00, z = -2.13, p = 0.034) LAT: Females > Males (U = 9.00, z = -3.24, p = 0.001) FREQ: Males > Females (U = 18.50, z = -2.59, p = 0.010) Self-Grooming FREQ: Males > Females at 5-10 min (U = 25.00, z = -2.14, p = 0.033) LAT: Females > Males (U = 15.00, z = -2.82, p = 0.005) a “Males” indicates gonadally intact males. b “Females” indicates gonadally intact females. 1FREQ represents frequency. 2DUR represents duration. 3LAT represents latency.
72 Table 7. Differences between gonadally intact and castrated male uninjected mice. Behaviour Effect FREQ1: Malesa > Castratedb (U = 15.50, z = -2.94, p = 0.003) Total Activity FREQ: Males > Castrated at 5-10 min (U = 17.50, z = -2.80, p = 0.005) FREQ: Males > Castrated at 10-15 min (U = 14.00, z = -3.04, p = 0.002) DUR2: Castrated > Males (U = 27.50, z = -2.20, p = 0.028) Social Inactivity DUR: Castrated > Males at 10-15 min (U = 24.50, z = -2.46, p = 0.014) Approaching FREQ: Males > Castrated (U = 19.00, z = -2.71, p = 0.007) and/or Attending FREQ: Males > Castrated at 5-10 min (U = 16.50, z = -2.88, p = 0.004) to the Intruder Total Agonistic FREQ: Males > Castrated at 5-10 min (U = 28.00, z = -2.12, p = 0.034) Behaviour DUR: Males > Castrated (U = 26.00, z = -2.71, p = 0.007) DUR: Males > Castrated at 0-5 min (U = 30.00, z = -2.69, p = 0.007) DUR: Males > Castrated at 5-10 min (U = 33.50, z = -2.23, p = 0.026) Aggressive FREQ: Males > Castrated (U = 26.00, z = -2.71, p = 0.007) Postures FREQ: Males > Castrated at 0-5 min (U = 30.00, z = -2.69, p = 0.007) FREQ: Males > Castrated at 5-10 min (U = 33.00, z = -2.27, p = 0.023) LAT3: Castrated > Males (U = 27.00, z = -2.63, p = 0.009) DUR: Castrated > Males (U = 29.00, z = -2.05, p = 0.041) Chasing the DUR: Castrated > Males at 0-5 min (U = 21.00, z = -2.57, p = 0.010) Intruder FREQ: Castrated > Males (U = 24.00, z = -2.38, p = 0.017) FREQ: Castrated > Males at 0-5 min (U = 16.00, z = -2.92, p = 0.004) Dominant FREQ: Males > Castrated at 0-5 min (U = 29.00, z = -2.05, p = 0.040) Behaviour FREQ: Males > Castrated (U = 32.50, z = -2.31, p = 0.021) Attacks FREQ: Males > Castrated at 5-10 min (U = 42.00, z = -1.99, p = 0.047) Delivered LAT: Castrated > Males (U = 36.50, z = -1.97, p = 0.048) DUR: Castrated > Males (U = 23.00, z = -2.44, p = 0.015) Social DUR: Castrated > Males at 0-5 min (U = 19.00, z = -2.70, p = 0.007) Investigation DUR: Castrated > Males at 5-10 min (U = 22.50, z = -2.47, p = 0.013) DUR: Males > Castrated at 0-5 min (U = 29.00, z = -2.04, p = 0.041) Oronasal FREQ: Males > Castrated (U = 21.00, z = -2.57, p = 0.010) Investigation FREQ: Males > Castrated at 0-5 min (U = 22.00, z = -2.51, p = 0.012) FREQ: Males > Castrated at 10-15 min (U = 29.50, z = -2.02, p = 0.043) Body DUR: Castrated > Males (U = 29.00, z = -2.04, p = 0.041) Investigation DUR: Castrated > Males at 5-10 min (U = 24.00, z = -2.38, p = 0.018) DUR: Castrated > Males (U = 4.00, z = -3.69, p < 0.001) and all timepoints Anogenital FREQ: Castrated > Males (U = 5.50, z = -3.60, p < 0.001) Investigation FREQ: Castrated > Males at 0-5 min (U = 3.00, z = -3.76, p < 0.001) FREQ: Castrated > Males at 5-10 min (U = 21.50, z = -2.55, p = 0.011) FREQ: Males > Castrated (U = 20.00, z = -2.64, p = 0.008) Total Non-social FREQ: Males > Castrated at 5-10 min (U = 27.50, z = -2.14, p = 0.032) Behaviours FREQ: Males > Castrated at 10-15 min (U = 12.50, z = -3.14, p = 0.002)
73 Non-social FREQ: Males > Castrated (U = 27.00, z = -2.18, p = 0.030) Locomotor FREQ: Males > Castrated at 5-10 min (U = 23.50, z = -2.41, p = 0.016) Behaviours FREQ: Males > Castrated at 10-15 min (U = 21.50, z = -2.54, p = 0.011) DUR: Males > Castrated (U = 19.00, z = -2.70, p = 0.007) DUR: Males > Castrated at 5-10 min (U = 14.00, z = -3.03, p = 0.002) Horizontal DUR: Males > Castrated at 10-15 min (U = 21.00, z = -2.57, p = 0.010) Exploration FREQ: Males > Castrated (U = 16.50, z = -2.87, p = 0.004) FREQ: Males > Castrated at 5-10 min (U = 15.00, z = -2.97, p = 0.003) FREQ: Males > Castrated at 10-15 min (U = 16.00, z = -2.91, p = 0.004) Vertical FREQ: Males > Castrated at 10-15 min (U = 23.50, z = -2.41, p = 0.016) Exploration FREQ: Males > Castrated (U = 28.5, z = -2.10, p = 0.036) Self-Grooming FREQ: Males > Castrated at 5-10 min (U = 27.50, z = -2.17, p = 0.030) a “Males” indicates gonadally intact males. b “Castrated” indicates castrated males. 1FREQ represents frequency. 2DUR represents duration. 3LAT represents latency.
74 Table 8. Differences between gonadally intact and ovariectomized female uninjected mice. Behaviour Effect FREQ1: Ovxa > Femalesb (U = 35.00, z = -2.12, p = 0.034) Total Activity FREQ: Ovx > Females at 0-5 min (U = 27.00, z = -2.58, p = 0.010) Total Social DUR2: Females > Ovx at 0-5 min (U = 35.00, z = -2.12, p = 0.034) Behaviours Total Agonistic FREQ: Ovx > Females at 0-5 min (U = 36.00, z = -2.06, p = 0.039) Behaviours Total Non- DUR: Ovx > Females at 0-5 min (U = 35.00, z = -2.12, p = 0.034) Social FREQ: Ovx > Females (U = 33.00, z = -2.23, p = 0.026) Behaviours FREQ: Ovx > Females at 0-5 min (U = 26.00, z = -2.64, p = 0.008) Non-social FREQ: Ovx > Females (U = 37.00, z = -2.00, p = 0.046) Locomotor FREQ: Ovx > Females at 0-5 min (U = 28.50, z = -2.50, p = 0.013) Behaviours Horizontal FREQ: Ovx > Females at 0-5 min (U = 34.00, z = -2.18, p = 0.030) Exploration DUR: Ovx > Females at 0-5 min (U = 35.50, z = -2.10, p = 0.036) Vertical FREQ: Ovx > Females (U = 30.50, z = -2.38, p = 0.018) Exploration FREQ: Ovx > Females at 0-5 min (U = 28.50, z = -2.52, p = 0.012) DUR: Ovx > Females at 5-10 min (U = 40.00, z = -2.10, p = 0.036) Digging FREQ: Ovx > Females at 5-10 min (U = 41.50, z = -2.00, p = 0.045) a “Ovx” indicates ovariectomized females. b “Females” indicates gonadally intact females. 1FREQ represents frequency. 2DUR represents duration.
75 Table 9. Effects of PPT in Gonadally Intact Males Behaviour Effect Total Agonistic LAT1: Decreased by LOWa (U = 18.00, z = -2.96, p = 0.003) Behaviours Aggressive FREQ2: Increased by PPTb at 10-15 min (U = 105.00, z = -2.11, p = Postures 0.035) DUR3: Effect of treatment (χ2(4) = 10.21, p = 0.037) DUR: Effect of treatment at 5-10 min (χ2(4) = 12.08, p = 0.017) DUR: Increased by LOW at 0-5 min (U = 28.50, z = -2.36, p = 0.018) DUR: Increased by LOW at 5-10 min (U = 36.00, z = -2.56, p = 0.011) DUR: Increased by HIGHc at 5-10 min (U = 48.00, z = -2.13, p = 0.033) Agonistic FREQ: Effect of treatment (χ2(4) = 12.62, p = 0.013) Behaviour FREQ: Effect of treatment at 0-5 min (χ2(4) = 10.59, p = 0.032) Received FREQ: Effect of treatment at 5-10 min (χ2(4) = 11.92, p = 0.018) FREQ: Increased by LOW (U = 26.00, z = -2.49, p = 0.013) FREQ: Increased by LOW at 0-5 min (U = 20.00, z = -2.91, p = 0.004) FREQ: Increased by LOW at 5-10 min (U = 36.00, z = -2.56, p = 0.011) FREQ: Increased by HIGH at 5-10 min (U = 48.00, z = -2.14, p = 0.033) LAT: Decreased by LOW (U = 24.00, z = -2.62, p = 0.009) DUR: Effect of treatment (χ2(4) = 10.59, p = 0.032) DUR: Increased by LOW (U = 28.00, z = -2.48, p = 0.013) DUR: Increased by LOW at 0-5 min (U = 29.00, z = -2.41, p = 0.016) DUR: Increased by LOW at 5-10 min (U = 42.00, z = -2.24, p = 0.025) Avoidance of the FREQ: Effect of treatment (χ2(4) = 10.17, p = 0.038) Intruder FREQ: Increased by LOW (U = 28.00, z = -2.48, p = 0.013) FREQ: Increased by LOW at 0-5 min (U = 31.00, z = -2.31, p = 0.021) FREQ: Increased by LOW at 5-10 min (U = 42.00, z = -2.24, p = 0.025) LAT: Decreased by LOW (U = 32.00, z = -2.22, p = 0.027) LAT: Decreased by HIGH (U = 39.00, z = -2.00, p = 0.045) Non-social Non- locomotor DUR: Increased by LOW (U = 31.00, z = -2.16, p = 0.031) Behaviours Horizontal Exploration DUR: Decreased by OILd (U = 30.00, z = -1.98, p = 0.048) DUR: Decreased by OIL at 5-10 min (U = 25.50, z = -2.27, p = 0.023) FREQ: Effect of treatment at 5-10 min (χ2(4) = 9.91, p = 0.042) Horizontal FREQ: Decreased by OIL at 5-10 min (U = 25.00, z = -2.31, p = 0.021) Exploration a “LOW” indicates 0.025 mg/kg PPT. b “PPT” indicates all drug doses collapsed together. c “HIGH” indicates 0.1 mg/kg PPT. d “OIL” indicates sesame oil vehicle. 1LAT represents latency. 2FREQ represents frequency. 3DUR represents duration.
76 Table 10. Effects of PPT in Gonadally Intact Females Behaviour Effect Total Activity DUR1: Increased by OILa at 5-10 min (U = 30.00, z = -2.22, p = 0.027) Total Social DUR: Increased by LOWb at 10-15 min (U = 38.00, z = -1.96, p = 0.0496) Behaviour DUR: Increased by MIDc at 5-10 min (U = 35.00, z = -2.17, p = 0.030) DUR: Increased by HIGHd at 5-10 min (U = 34.00, z = -2.45, p = 0.014) Social Inactivity FREQ2: Increased by MID at 5-10 min (U = 36.00, z = -2.10, p = 0.036) FREQ: Increased by HIGH at 5-10 min (U = 37.50, z = -2.23, p = 0.026) Approaching FREQ: Increased by OIL at 5-10 min (U = 34.00, z = -1.98, p = 0.048) and/or Attending FREQ: Increased by MIDc at 5-10 min (U = 27.00, z = -1.98, p = 0.048) to the Intruder Dominance DUR: Decreased by OIL at 0-5 min (U = 31.00, z = -2.15, p = 0.031) Score Agonistic Behaviour DUR: Decreased by OIL at 0-5 min (U = 31.00, z = -2.15, p = 0.031) Delivered Chasing the FREQ: Decreased by HIGH at 10-15 min (U = 32.50, z = -2.35, p = 0.019) Intruder Dominant DUR: Decreased by OIL at 0-5 min (U = 30.00, z = -2.22, p = 0.027) Behaviour Agonistic FREQ: Effect of treatment (χ2(4) = 13.49, p = 0.009) Behaviour FREQ: Effect of treatment at 10-15 min (χ2(4) = 11.77, p = 0.019) Received Avoidance of DUR: Effect of treatment (χ2(4) = 10.01, p = 0.040) the Intruder FREQ: Effect of treatment (χ2(4) = 9.75, p = 0.045) Submissive FREQ: Effect of treatment (χ2(4) = 9.75, p = 0.045) Behaviour FREQ: Effect of treatment at 5-10 min (χ2(4) = 9.67, p = 0.046) Oronasal FREQ: Decreased by OIL (U = 30.00, z = -2.01, p = 0.045) Investigation FREQ: Increased by HIGH at 5-10 min (U = 33.00, z = -2.26, p = 0.024) Body LAT3: Increased by OIL (U = 30.00, z = -2.22, p = 0.027) Investigation Total Non-social DUR: Decreased by LOW at 10-15 min (U = 38.00, z = -1.96, p = 0.0496) Behaviour Vertical LAT: Decreased by OIL (U = 31.00, z = -2.15, p = 0.031) Exploration DUR: Effect of treatment at 5-10 min (χ2(4) = 11.11, p = 0.025) DUR: Increased by OIL (U = 24.00, z = -2.59, p = 0.010) DUR: Increased by OIL at 0-5 min (U = 33.00, z = -2.15, p = 0.031) Non-social DUR: Increased by OIL at 5-10 min (U = 16.50, z = -3.05, p = 0.002) Non-locomotor FREQ: Effect of treatment (χ2(4) = 11.97, p = 0.018) and all timepoints Behaviours FREQ: Increased by OIL (U = 19.00, z = -2.73, p = 0.006) and all
timepoints LAT: Effect of treatment (χ2(4) = 15.57, p = 0.004) LAT: Decreased by OIL (U = 26.00, z = -2.46, p = 0.014)
77 DUR: Increased by OIL at 0-5 min (U = 36.00, z = -2.27, p = 0.023) Solitary FREQ: Increased by OIL at 0-5 min (U = 37.00, z = -2.19, p = 0.028) Inactivity FREQ: Increased by MID at 10-15 min (U = 34.50, z = -1.99, p = 0.047) LAT: Effect of treatment (χ2(4) = 10.02, p = 0.040) DUR: Increased by OIL (U = 23.00, z = -2.65, p = 0.008) DUR: Increased by OIL at 5-10 min (U = 22.50, z = -2.68, p = 0.007) DUR: Increased by HIGH at 5-10 min (U = 36.50, z = -2.06, p = 0.040) Self-Grooming FREQ: Increased by OIL (U = 20.00, z = -2.67, p = 0.008) FREQ: Increased by OIL at 5-10 min (U = 24.50, z = -2.57, p = 0.010) -LAT: Decreased by OIL (U = 29.00, z = -2.28, p = 0.023) a “OIL” indicates sesame oil vehicle. b “LOW” indicates 0.025 mg/kg PPT. c “MID” indicates 0.05 mg/kg PPT. d “HIGH” indicates 0.1 mg/kg PPT. 1DUR represents duration. 2FREQ represents frequency. 3LAT represents latency.
78 Table 11. Effects of the Estrus Cycle in Gonadally Intact Female Mice. Behaviour Effect Uninjected Female Mice DUR1: Proestrusa > Estrusb (U = 3.00, z = -2.12, p = 0.034) DUR: Proestrus > Estrus at 0-5 min (U = 3.00, z = -2.12, p = 0.034) Aggressive FREQ2: Proestrus > Estrus (U = 3.00, z = -2.12, p = 0.034) Postures FREQ: Proestrus > Estrus at 0-5 min (U = 3.00, z = -2.12, p = 0.034) LAT3: Estrus > Proestrus (U = 3.00, z = -2.12, p = 0.034) Social DUR: Estrus > Proestrus (U = 1.00, z = -2.07, p = 0.039) Investigation Anogenital DUR: Estrus > Proestrus (U = 1.00, z = -2.07, p = 0.039) Investigation DUR: Estrus > Proestrus at 10-15 min (U = 1.00, z = -2.07, p = 0.039) DUR: Proestrus > Estrus at 0-5 min (U = 1.00, z = -2.16, p = 0.031) FREQ: Effect of estrus cycle at 0-5 min (χ2(2) = 6.35, p = 0.042) Vertical FREQ: Proestrus > Estrus at 0-5 min (U = 1.00, z = -2.20, p = 0.028) Exploration LAT3: Effect of estrus cycle (χ2(2) = 6.23, p = 0.044) LAT: Estrus > Diestrusc (U = 0.00, z = -2.00, p = 0.046) OIL Female Miced DUR: Estrus > Proestrus at 5-10 min (U = 1.00, z = -2.07, p = 0.039) FREQ: Proestrus < Diestrus (U = 0.00, z = -1.96, p = 0.0495) Total Activity FREQ: Proestrus < Diestrus at 10-15 min (U = 0.00, z = -1.96, p = 0.0495) FREQ: Estrus < Diestrus at 10-15 min (U = 1.00, z = -2.07, p = 0.039) DUR: Estrus > Proestrus at 0-5 min (U = 1.00, z = -2.07, p = 0.039) Total Social FREQ: Proestrus < Diestrus (U = 0.00, z = -1.96, p = 0.0495) Behaviours FREQ: Proestrus < Diestrus at 0-5 min (U = 0.00, z = -1.96, p = 0.0495) DUR: Effect of estrus cycle (χ2(2) = 6.08, p = 0.048) DUR: Effect of estrus cycle at 5-10 min (χ2(2) = 6.55, p = 0.038) DUR: Effect of estrus cycle at 10-15 min (χ2(2) = 6.08, p = 0.048) DUR: Estrus > Proestrus at 5-10 min (U = 3.00, z = -2.12, p = 0.034) DUR: Proestrus > Diestrus (U = 0.00, z = -2.09, p = 0.037) Social Inactivity DUR: Proestrus > Diestrus at 10-15 min (U = 0.00, z = -2.09, p = 0.037) FREQ: Effect of estrus cycle at 5-10 min (χ2(2) = 6.55, p = 0.038) FREQ: Proestrus > Estrus at 5-10 min (U = 3.00, z = -2.12, p = 0.034) FREQ: Proestrus > Diestrus at 10-15 min (U = 0.00, z = -2.12, p = 0.034) FREQ: Proestrus > Diestrus (U = 0.00, z = -2.09, p = 0.037) LAT: Proestrus < Diestrus (U = 0.00, z = -2.09, p = 0.037) Approaching FREQ: Effect of estrus cycle at 5-10 min (χ2(2) = 6.45, p = 0.040) and/or Attending FREQ: Proestrus > Estrus at 5-10 min (U = 0.00, z = -2.37, p = 0.018) to the Intruder Dominance FREQ: Proestrus < Diestrus at 0-5 min (U = 0.00, z = -1.99, p = 0.046) Score Chasing the DUR: Proestrus < Diestrus at 0-5 min (U = 0.00, z = -1.96, p = 0.0495) Intruder Agonistic Behaviour LAT: Proestrus > Diestrus (U = 0.00, z = -1.96, p = 0.0495) Received
79 DUR: Proestrus < Diestrus (U = 0.00, z = -1.99, p = 0.046) Avoidance of DUR: Proestrus < Diestrus at 0-5 min (U = 0.00, z = -1.99, p = 0.046) the Intruder LAT: Proestrus > Diestrus (U = 0.00, z = -1.99, p = 0.046) DUR: Proestrus > Estrus at 0-5 min (U = 1.00, z = -2.10, p = 0.036) DUR: Proestrus > Estrus at 5-10 min (U = 1.00, z = -2.26, p = 0.024) Submissive FREQ: Proestrus > Estrus (U = 1.00, z = -2.26, p = 0.024) Behaviour FREQ: Proestrus > Estrus at 0-5 min (U = 0.00, z = -2.41, p = 0.016) FREQ: Proestrus > Estrus at 5-10 min (U = 1.50, z = -2.13, p = 0.033) DUR: Estrus > Proestrus at 0-5 min (U = 0.00, z = -2.32, p= 0.020) Social FREQ: Proestrus < Diestrus (U = 0.00, z = -1.96, p = 0.0495) Investigation FREQ: Proestrus < Diestrus at 0-5 min (U = 0.00, z = -1.96, p = 0.0495) FREQ: Proestrus < Diestrus at 10-15 min (U = 0.00, z = -1.96, p = 0.0495) FREQ: Proestrus < Diestrus (U = 0.00, z = -1.96, p = 0.0495) Body FREQ: Proestrus < Diestrus at 0-5 min (U = 0.00, z = -1.96, p = 0.0495) Investigation FREQ: Proestrus < Diestrus at 10-15 min (U = 0.00, z = -1.99, p = 0.046) Anogenital FREQ: Diestrus > Estrus at 10-15 min (U = 0.50, z = -2.22, p = 0.026) Investigation FREQ: Proestrus < Diestrus at 10-15 min (U = 0.00, z = -1.96, p = 0.0495) Total Non- DUR: Proestrus > Estrus at 0-5 min (U = 1.00, z = -2.07, p = 0.039) social FREQ: Proestrus > Estrus at 0-5 min (U = 0.00, z = -2.34, p = 0.019) Behaviours FREQ: Proestrus > Estrus at 5-10 min (U = 1.00, z = -2.07, p = 0.038) Non-social DUR: Proestrus > Estrus at 0-5 min (U = 1.00, z = -2.07, p = 0.039) Locomotor FREQ: Proestrus > Estrus at 0-5 min (U = 0.00, z = -2.36, p = 0.018) Behaviours FREQ: Proestrus > Estrus at 5-10 min (U = 1.00, z = -2.09, p = 0.036) Horizontal DUR: Proestrus > Estrus at 0-5 min (U = 0.00, z = -2.32, p = 0.020) Exploration FREQ: Proestrus > Estrus at 0-5 min (U = 0.00, z = -2.36, p = 0.018) Vertical DUR: Proestrus > Estrus at 0-5 min (U = 1.00, z = -2.07, p = 0.038) Exploration FREQ: Proestrus > Estrus at 0-5 min (U = 1.00, z = -2.16, p = 0.031) DUR: Proestrus > Estrus at 5-10 min (U = 1.00, z = -2.07, p = 0.039) Non-social FREQ: Effect of estrus cycle at 5-10 min (χ2(2) = 6.81, p = 0.033) Non-locomotor FREQ: Proestrus > Estrus at 5-10 min (U = 1.00, z = -2.07, p = 0.038) Behaviours FREQ: Diestrus > Estrus at 5-10 min (U = 1.00, z = -2.07, p = 0.038) FREQ: Proestrus < Diestrus (U = 0.00, z = -1.96, p = 0.0495) Self-Grooming FREQ: Proestrus < Diestrus at 5-10 min (U = 0.00, z = -2.02, p = 0.043) FREQ: Proestrus < Diestrus at 10-15 min (U = 0.00, z = -1.96, p = 0.0495) FREQ: Effect of estrus cycle at 5-10 min (χ2(2) = 6.75, p = 0.034) Solitary FREQ: Proestrus > Estrus (U = 1.50, z = -1.96, p = 0.0499) Inactivity FREQ: Proestrus > Estrus at 5-10 min (U = 0.00, z = -2.44, p = 0.015) LOW Female Micee Chasing the DUR: Proestrus > Diestrus at 0-5 min (U = 0.00, z = -1.12, p = 0.034) Intruder FREQ: Proestrus > Diestrus at 0-5 min (U = 0.00, z = -2.12, p = 0.034) DUR: Proestrus > Diestrus (U = 0.00, z = -2.12, p = 0.034) DUR: Proestrus > Diestrus at 10-15 min (U = 0.00, z = -2.12, p = 0.034) Oronasal FREQ: Effect of estrus cycle at 10-15 min (χ2(2) = 6.59, p = 0.037) Investigation FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.26, p = 0.024) FREQ: Proestrus > Diestrus at 10-15 min (U = 0.00, z = -2.16, p = 0.031)
80 DUR: Effect of estrus cycle (χ2(2) = 6.37, p = 0.041) DUR: Effect of estrus cycle at 5-10 min (χ2(2) = 6.27, p = 0.044) DUR: Proestrus > Estrus (U = 0.00, z = -2.24, p= 0.025) Body DUR: Proestrus > Diestrus (U = 0.00, z = -2.12, p = 0.034) Investigation DUR: Proestrus > Diestrus at 0-5 min (U = 0.00, z = -2.12, p = 0.034) DUR: Proestrus > Diestrus at 5-10 min (U = 0.00, z = -2.12, p = 0.034) DUR: Proestrus > Estrus at 5-10 min (U = 0.00, z = -2.24, p = 0.025) FREQ: Proestrus > Estrus (U = 0.00, z = -2.24, p = 0.025) MID Female Micef Total Activity FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.05, p = 0.040) Total Social FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.05, p = 0.040) Behaviours DUR: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.07, p = 0.039) Social Inactivity FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.07, p = 0.039) Stretched FREQ: Effect of estrus cycle (χ2(2) = 6.66, p = 0.036) Approaches FREQ: Diestrus > Estrus (U = 0.50, z = -2.28, p = 0.023) DUR: Effect of estrus cycle (χ2(2) = 6.60, p = 0.037) Approaching DUR: Effect of estrus cycle at 5-10 min (χ2(2) = 7.36, p = 0.025) and/or DUR: Estrus > Proestrus at 5-10 min (U = 0.00, z = -2.05, p = 0.040) Attending to the DUR: Estrus < Diestrus (U = 0.00, z = -2.05, p = 0.040) Intruder DUR: Estrus < Diestrus at 5-10 min (U = 0.00, z = -2.05, p = 0.040) FREQ: Estrus > Proestrus at 5-10 min (U = 0.00, z = -2.08, p = 0.037) Agonistic DUR: Estrus > Proestrus (U = 0.00, z = -2.05, p = 0.040) Behaviour DUR: Estrus > Proestrus at 10-15 min (U = 0.00, z = -2.06, p = 0.040) Received FREQ: Estrus > Proestrus (U = 0.00, z = -2.08, p = 0.038) DUR: Proestrus > Estrus (U = 0.00, z = -2.25, p = 0.025) Avoidance of DUR: Proestrus > Estrus at 0-5 min (U = 0.00, z = -2.25, p = 0.025) the Intruder FREQ: Proestrus > Estrus (U = 1.00, z = -1.96, p = 0.0495) FREQ: Proestrus > Estrus at 0-5 min (U = 0.00, z = -2.27, p = 0.023) DUR: Effect of estrus cycle (χ2(2) = 6.60, p = 0.037) Submissive DUR: Estrus > Proestrus (U = 0.00, z = -2.05, p = 0.040) Behaviour DUR: Estrus > Proestrus at 10-15 min (U = 0.00, z = -2.06, p = 0.040) FREQ: Estrus > Proestrus (U = 0.00, z = -2.07, p = 0.039) Social FREQ: Proestrus > Estrus (U = 0.00, z = -2.06, p = 0.040) Investigation FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.06, p = 0.040) Oronasal FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.06, p = 0.040) Investigation Body FREQ: Proestrus > Estrus (U = 0.00, z = -2.05, p = 0.040) Investigation FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.05, p = 0.040) FREQ: Proestrus > Estrus at 10-15 min (U = 0.00, z = -2.08, p = 0.037) Anogenital LAT: Effect of estrus cycle (χ2(2) = 6.24, p = 0.044) Investigation LAT: Estrus < Diestrus (U = 0.00, z = -2.05, p = 0.040) Non-social Locomotor DUR: Estrus > Proestrus at 5-10 min (U = 0.00, z = -2.05, p = 0.040) Behaviour
81 Vertical DUR: Estrus > Proestrus at 5-10 min (U = 0.00, z = -2.05, p = 0.040) Exploration DUR: Effect of estrus cycle at 10-15 min (χ2(2) = 6.66, p = 0.036) Solitary DUR: Estrus > Proestrus at 10-15 min (U = 0.00, z = -2.07, p = 0.039) Inactivity FREQ: Effect of estrus cycle at 10-15 min (χ2(2) = 6.10, p = 0.047) FREQ: Estrus > Proestrus at 10-15 min (U = 0.00, z = -2.09, p = 0.036) HIGH Female Miceg DUR: Effect of estrus cycle (χ2(1) = 5.34, p = 0.021) DUR: Effect of estrus cycle at 5-10 min (χ2(1) = 5.34, p = 0.021) DUR: Effect of estrus cycle at 10-15 min (χ2(1) = 4.52, p = 0.033) Total Activity DUR: Estrus > Diestrus at 5-10 min (U = 1.00, z = -2.31, p = 0.021) DUR: Estrus > Diestrus at 10-15 min (U = 2.00, z = -2.13, p = 0.033) DUR: Estrus > Diestrus (U = 1.00, z = -2.31, p = 0.021) Agonistic FREQ: Effect of estrus cycle at 10-15 min (χ2(1) = 4.02, p = 0.045) Behaviour FREQ: Diestrus > Estrus at 10-15 min (U = 3.00, z = -2.01, p = 0.045) Received Oronasal LAT: Effect of estrus cycle (χ2(1) = 4.17, p = 0.041) Investigation LAT: Estrus > Diestrus (U = 2.50, z = -2.04, p = 0.041) Non-social FREQ: Effect of estrus cycle 5-10 min (χ2(1) = 4.17, p = 0.041) Non-locomotor FREQ: Diestrus > Estrus at 5-10 min (U = 2.50, z = -2.04, p = 0.041) Behaviours DUR: Effect of estrus cycle (χ2(1) = 4.52, p = 0.033) DUR: Effect of estrus cycle at 5-10 min (χ2(1) = 5.42, p = 0.020) DUR: Diestrus > Estrus (U = 2.00, z = -2.13, p = 0.033) DUR: Diestrus > Estrus at 5-10 min (U = 1.00, z = -2.33, p = 0.020) Self-Grooming FREQ: Effect of estrus cycle (χ2(1) = 4.54, p = 0.033) FREQ: Effect of estrus cycle at 5-10 min (χ2(1) = 5.92, p = 0.015) FREQ: Diestrus > Estrus (U = 2.00, z = -2.13, p = 0.033) FREQ: Diestrus > Estrus at 5-10 min (U = 0.50, z = -2.43, p = 0.015) a “Proestrus” indicates mice in the proestrus phase of the estrus cycle. b “Estrus” indicates mice in the estrus phase of the cycle. c “Diestrus” indicates mice in the diestrus phase of the estrus cycle. d “OIL” indicates sesame oil vehicle. e “LOW” indicates mice treated with 0.025 mg/kg PPT. f “MID” indicates 0.05 mg/kg PPT. g “HIGH” indicates mice treated with 0.1 mg/kg PPT. 1FREQ represents frequency. 2DUR represents duration. 3LAT represents latency.
82 Table 12. Effects of PPT in Castrated Males Behaviour Effect Approaching and/or DUR: Decreased by OIL at 10-15 min (U = 33.50, z = -2.22, p = 0.026) Attending to the Intruder LAT1: Effect of treatment (χ2(4) = 12.45, p = 0.014) Total Agonistic LAT: Decreased by LOWa (U = 30.00, z = -2.43, p = 0.015) Behaviours LAT: Decreased by MIDb (U = 14.00, z = -3.35, p = 0.001) Attacks FREQ2: Increased by PPT (Z = -2.37, p = 0.018; Wilcoxon Signed Ranks) Delivered LAT: Decreased by PPT (Z = -5.55, p < 0.001; Wilcoxon Signed Ranks) DUR3: Effect of treatment (χ2(4) = 13.22, p = 0.010) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.55, p = 0.032) DUR: Effect of treatment at 10-15 min (χ2(4) = 11.93, p = 0.018) DUR: Increased by OILd (U = 31.00, z = -2.38, p = 0.017) Agonistic DUR: Increased by OIL at 10-15 min (U = 33.00, z = -2.59, p = 0.010) Behaviour FREQ3: Effect of treatment (χ2(4) = 10.53, p = 0.032) Received FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.04, p = 0.040) FREQ: Effect of treatment at 10-15 min (χ2(4) = 12.81, p = 0.012) FREQ: Increased by OIL (U = 37.00, z = -2.04, p = 0.041) FREQ: Increased by OIL at 10-15 min (U = 33.00, z = -2.60, p = 0.009) FREQ: Increased by LOW at 0-5 min (U = 39.50, z = -2.07, p = 0.039) Avoidance of LAT: Increased by OIL (U = 39.00, z = -2.00, p = 0.045) the Intruder LAT: Increased by LOW (U = 39.00, z = -2.00, p = 0.045) DUR: Effect of treatment (χ2(4) = 9.47, p = 0.050) DUR: Increased by OIL (U = 27.50, z = -2.62, p = 0.009) Submissive DUR: Increased by OIL at 0-5 min (U = 34.00, z = -2.28, p = 0.023) Behaviour DUR: Increased by OIL at 10-15 min (U = 33.00, z = -2.59, p = 0.010) FREQ: Increased by OIL (U = 29.50, z = -2.51, p = 0.012) FREQ: Increased by OIL at 10-15 min (U = 33.50, z = -2.57, p = 0.010) Social DUR: Decreased by OIL at 5-10 min (U = 34.00, z = -2.19, p = 0.028) Investigation Body DUR: Increased by OIL at 10-15 min (U = 34.50, z = -2.17, p = 0.030) Investigation DUR: Decreased by LOW at 0-5 min (U = 37.00, z = -2.02, p = 0.043) DUR: Effect of treatment at 0-5 min (χ2(4) = 9.61, p = 0.048) Digging DUR: Decreased by OIL at 0-5 min (U = 48.00, z = -2.13, p = 0.033) FREQ: Decreased by OIL at 0-5 min (U = 48.00, z = -2.14, p = 0.033) DUR: Increased by MID at 0-5 min (U = 27.50, z = -2.57, p = 0.010) DUR: Increased by MID at 5-10 min (U = 35.00, z = -2.14, p = 0.033) DUR: Increased by HIGHe at 5-10 min (U = 35.00, z = -2.14, p = 0.033) Non-social FREQ: Effect of treatment (χ2(4) = 10.44, p = 0.034) Non-locomotor FREQ: Effect of treatment at 5-10 min (χ2(4) = 9.70, p = 0.046) Behaviours FREQ: Effect of treatment at 10-15 min (χ2(4) = 10.68, p = 0.030) FREQ: Increased by MID (U = 35.50, z = -2.11, p = 0.035) FREQ: Increased by MID at 0-5 min (U = 30.50, z = -2.43, p = 0.015) FREQ: Increased by MID at 5-10 min (U = 33.50, z = -2.24, p = 0.025)
83 DUR: Effect of treatment at 5-10 min (χ2(4) = 11.18, p = 0.025) DUR: Increased by LOW at 5-10 min (U = 32.00, z = -2.37, p = 0.018) DUR: Increased by MID (U = 31.00, z = -2.37, p = 0.018) DUR: Increased by MID at 5-10 min (U = 26.50, z = -2.70, p = 0.007) DUR: Increased by MID at 10-15 min (U = 35.50, z = -2.15, p = 0.032) DUR: Increased by HIGH (U = 34.50, z = -2.17, p = 0.030) DUR: Increased by HIGH at 5-10 min (U = 33.50, z = -2.31, p = 0.021) Solitary DUR: Increased by HIGH at 10-15 min (U = 38.00, z = -2.02, p = 0.044) Inactivity FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.44, p = 0.034) FREQ: Increased by LOW at 5-10 min (U = 33.50, z = -2.32, p = 0.020) FREQ: Increased by MID (U = 35.00, z = -2.15, p = 0.031) FREQ: Increased by MID at 5-10 min (U = 31.00, z = -2.45, p = 0.014) FREQ: Increased by HIGH (U = 32.00, z = -2.32, p = 0.020) FREQ: Increased by HIGH at 5-10 min (U = 36.00, z = -2.17, p = 0.030) FREQ: Increased by HIGH at 10-15 min (U = 36.50, z = -2.11, p = 0.035) DUR: Increased by MID at 0-5 min (U = 32.00, z = -2.33, p = 0.020) FREQ: Effect of treatment (χ2(4) = 12.34, p = 0.015) FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.00, p = 0.040) Self-Grooming FREQ: Effect of treatment at 10-15 min (χ2(4) = 12.01, p = 0.017) FREQ: Increased by MID at 0-5 min (U = 35.00, z = -2.20, p = 0.028) LAT: Decreased by MID (U = 38.00, z = -1.96, p = 0.0496) a “LOW” indicates 0.025 mg/kg PPT. b “MID” indicates 0.05 mg/kg PPT. c “PPT” indicates all drug doses collapsed together. d “OIL” indicates sesame oil vehicle. e “HIGH” indicates 0.1 mg/kg PPT. 1LAT represents latency. 2FREQ represents frequency. 3DUR represents duration.
84 Table 13. Effects of PPT in Ovariectomized Females Behaviour Effect DUR1: Decreased by OILa (U = 29.00, z = -2.46, p = 0.014) Total Activity DUR: Decreased by OIL at 5-10 min (U = 35.00, z = -2.12, p = 0.034) FREQ: Increased by OIL (U = 52.00, z = -1.97, p = 0.049) Stretched FREQ: Increased by OIL at 5-10 min (U = 52.00, z = -1.97, p = 0.049) Approaches LAT3: Decreased by OIL (U = 52.00, z = -1.97, p = 0.049) Dominance DUR: Decreased by OIL at 5-10 min (U = 34.00, z = -2.17, p = 0.030) Score Agonistic Behaviour DUR: Decreased by OIL at 5-10 min (U = 37.00, z = -2.00, p = 0.046 Delivered DUR: Increased by OIL (U = 28.00, = -2.52, p = 0.012) Chasing the DUR: Increased by OIL at 0-5 min (U = 26.50, z = -2.61, p = 0.009) Intruder FREQ2: Increased by OIL at 0-5 min (U = 30.00, z = -2.42, p = 0.016) DUR: Decreased by OIL (U = 35.00, z = -2.12, p = 0.034) DUR: Decreased by OIL at 0-5 min (U = 33.00, z = -2.23, p = 0.026) Dominant DUR: Decreased by OIL at 5-10 min (U = 37.00, z = -2.00, p = 0.046) Behaviour DUR: Increased by LOWb at 0-5 min (U = 32.00, z = -2.09, p = 0.036) FREQ: Effect of treatment at 0-5 min (χ2(4) = 9.59, p = 0.048) DUR: Effect of treatment at 0-5 min (χ2(4) = 9.81, p = 0.044) DUR: Increased by OIL at 5-10 min (U = 37.00, z = -2.08, p = 0.038) DUR: Increased by OIL at 10-15 min (U = 37.00, z = -2.18, p = 0.029) Agonistic DUR: Decreased by LOW at 5-10 min (U = 31.00, z = -2.20, p = 0.028) Behaviour FREQ: Increased by OIL (U = 36.00, z = -2.09, p = 0.036) Received FREQ: Increased by OIL at 5-10 min (U = 29.50, z = -2.54, p = 0.011) FREQ: Increased by OIL at 10-15 min (U = 39.00, z = -2.06, p = 0.039) FREQ: Decreased by LOW at 5-10 min (U = 31.00, z = -2.22, p = 0.026) DUR: Effect of treatment at 5-10 min (χ2(4) = 13.44, p = 0.009) DUR: Increased by OIL (U = 40.00, z = -1.99, p = 0.047) DUR: Increased by OIL at 0-5 min (U = 41.00, z = -1.97, p = 0.048) DUR: Increased by OIL at 5-10 min (U = 39.00, z = -2.65, p = 0.008) DUR: Decreased by LOW at 5-10 min (U = 40.50, z = -2.03, p = 0.042) DUR: Decreased by MIDc at 5-10 min (U = 44.00, z = -2.09, p = 0.036) Avoidance of the DUR: Decreased by HIGHd (U = 36.00, z = -2.04, p = 0.041) Intruder DUR: Decreased by HIGH at 5-10 min (U = 39.00, z = -2.15, p = 0.031) FREQ: Effect of treatment at 5-10 min (χ2(4) = 12.51, p = 0.014) FREQ: Increased by OIL (U = 40.00, z = -2.00, p = 0.045) FREQ: Increased by OIL at 5-10 min (U = 39.00, z = -2.66, p = 0.008) FREQ: Decreased by LOW at 5-10 min (U = 41.00, z = -2.01, p = 0.045) FREQ: Decreased by HIGH at 5-10 min (U = 41.00, z = -2.01, p = 0.045) Submissive DUR: Increased by OIL at 10-15 min (U = 37.00, z = -2.18, p = 0.029) Behaviour FREQ: Increased by OIL at 10-15 min (U = 36.50, z = -2.27, p = 0.023)
85 DUR: Increased by LOW (U = 25.00, z = -2.52, p = 0.012) DUR: Increased by LOW at 5-10 min (U = 26.00, z = -2.46, p = 0.014) DUR: Increased by LOW at 10-15 min (U = 32.00, z = -2.09, p = 0.036) Social DUR: Increased by HIGH at 5-10 min (U = 29.00, z = -2.28, p = 0.023) Investigation DUR: Increased by HIGH at 10-15 min (U = 33.00, z = -2.03, p = 0.042) FREQ: Decreased by OIL at 10-15 min (U = 34.50, z = -2.15, p = 0.032) FREQ: Increased by LOW at 5-10 min (U = 33.50, z = -2.00, p = 0.045) FREQ: Increased by LOW at 10-15 min (U = 28.50, z = -2.31, p = 0.021) DUR: Effect of treatment (χ2(4) = 13.20, p = 0.010) DUR: Effect of treatment at 0-5 min (χ2(4) = 9.60, p = 0.048) DUR: Effect of treatment at 5-10 min (χ2(4) = 11.08, p = 0.026) Body DUR: Increased by HIGH at 5-10 min (U = 29.50, z = -2.25, p = 0.025) Investigation DUR: Increased by HIGH at 10-15 min (U = 28.50, z = -2.31, p = 0.021) FREQ: Increased by LOW at 5-10 min (U = 33.50, z = -2.01, p = 0.045) FREQ: Increased by HIGH at 10-15 min (U = 32.50, z = -2.07, p = 0.039) DUR: Effect of treatment at 5-10 min (χ2(4) = 12.71, p = 0.013) DUR: Decreased by OIL at 5-10 min (U = 24.00, z = -2.75, p = 0.006) DUR: Increased by LOW (U = 34.00, z = -1.97, p = 0.049) Anogenital DUR: Increased by LOW at 5-10 min (U = 25.00, z = -2.52, p = 0.012) Investigation DUR: Increased by HIGH (U = 28.00, z = -2.34, p = 0.019) DUR: Increased by HIGH at 5-10 min (U = 13.00, z = -3.26, p = 0.001) FREQ: Increased by LOW at 10-15 min (U = 30.00, z = -2.22, p = 0.026) FREQ: Increased by HIGH at 5-10 min (U = 30.00, z = -2.22, p = 0.026) Horizontal DUR: Decreased by LOW at 0-5 min (U = 33.00, z = -2.03, p = 0.042) Exploration DUR: Effect of treatment (χ2(4) = 11.55, p = 0.021) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.89, p = 0.028) Non-social Non- DUR: Effect of treatment at 10-15 min (χ2(4) = 11.36, p = 0.023) locomotor DUR: Increased by OIL at 5-10 min (U = 32.00, z = -2.29, p = 0.022) Behaviours FREQ: Effect of treatment (χ2(4) = 11.45, p = 0.022) FREQ: Effect of treatment at 10-15 min (χ2(4) = 19.45, p = 0.001) FREQ: Increased by OIL at 10-15 min (U = 32.00, z = -2.30, p = 0.022) Solitary DUR: Decreased by LOW at 0-5 min (U = 33.50, z = -2.16, p = 0.031) Inactivity FREQ: Increased by OIL (U = 37.00, z = -2.02, p = 0.044) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.32, p = 0.035) DUR: Increased by OIL at 5-10 min (U = 28.00, z = -2.53, p = 0.011) FREQ: Effect of treatment (χ2(4) = 13.98, p = 0.007) Self-Grooming FREQ: Effect of treatment at 5-10 min (χ2(4) = 9.63, p = 0.047) FREQ: Effect of treatment at 10-15 min (χ2(4) = 16.13, p = 0.003) FREQ: Increased by OIL (U = 37.50, z = -1.98, p = 0.047) FREQ: Increased by OIL at 5-10 min (U = 37.00, z = -2.03, p = 0.043) a “OIL” indicates sesame oil vehicle. b “LOW” indicates 0.025 mg/kg PPT. c “MID” indicates 0.05 mg/kg PPT. d “HIGH” indicates 0.1 mg/kg PPT. 1DUR represents duration. 2FREQ represents frequency. 3LAT represents latency.
86
CHAPTER 3:
Agonistic behaviour in males and females:
Effects of an estrogen receptor beta agonist in gonadectomized and gonadally intact mice
This manuscript is published in: Clipperton-Allen, A.E., Cragg, C.L., Wood, A.J., Pfaff,
D.W., and Choleris, E. (2010). Agonistic behaviour in males and females: Effects of an
estrogen receptor beta agonist in gonadectomised and gonadally intact mice.
Psychoneuroendocrinology 35: 1008-1022.
87 3.1 Abstract
Affiliative and agonistic social interactions are mediated by gonadal hormones. Research with estrogen receptor alpha (ERα) or beta (ERβ) knockout (KO) mice show that long-term inactivation of ERα decreases, while inactivation of ERβ increases, male aggression.
Opposite effects were found in female αERKO and βERKO mice. The role of acute activation of ERα or ERβ in the agonistic responses of adult non-KO mice is unknown. We report here the effects of the ERβ selective agonist WAY-200070 on agonistic and social behaviour in gonadally intact and gonadectomized (gonadex) male and female CD1 mice towards a gonadex, same-sex intruder. All 15 min resident-intruder tests were videotaped for comprehensive behavioural analysis. Separate analyses assessed: 1) effects of WAY-200070 on each sex and gonadal condition; 2) differences between sexes, and between gonadally intact and gonadex mice, in untreated animals. Results show that in gonadally intact male and female mice WAY-200070 increased agonistic behaviours such as pushing down and aggressive grooming, while leaving attacks unaffected. In untreated mice, males attacked more than females, and gonadex animals showed less agonistic behaviour than same-sex, gonadally intact mice. Overall, our detailed behavioural analysis suggested that in gonadally intact male and female mice, ERβ mediates patterns of agonistic behaviour that are not directly involved in attacks. This suggests that specific aspects of aggressive behaviour are acutely mediated by ERβ in adult mice. Our results also showed that, in resident-intruder tests, female mice spend as much time in intrasexual agonistic interactions as males, but use agonistic behaviours that involve extremely low levels of direct attacks. This non-attack aggression in females is increased by acute activation of ERβ. Thus, acute activation of ERβ similarly mediates agonistic behaviour in adult male and female CD1 mice.
88 3.2 Introduction
Aggression, and its underlying neurobiological mechanisms, has long been a topic of interest and a target of research (e.g., Craig, 1921; Howard, 1920). Aggression research with laboratory rodents has typically focused on males’ behaviour, usually employing the resident-intruder test (e.g., Brain et al., 1981; Simon & Whalen, 1986; Thor & Flannelly,
1976), where an intruder is introduced into the home cage of another, usually isolated, animal
(the resident). Measures of fighting or overt aggression such as latency to, or frequency of, attack, are taken (e.g., Morè, 2008; Parmigiani et al., 1999; Simon & Whalen, 1986; Thor &
Flannelly, 1976), sometimes together with other measures of “agonistic” behaviour, which include both the aggression and the reaction to it (Alleva, 1993; Branchi et al., 2006; Olivier et al., 1989; Olivier et al., 1991; Pietropaolo et al., 2004). In mice, this paradigm takes advantage of the fact that males will often establish and maintain an exclusive territory or dominate a shared territory under natural conditions (Miczek et al., 2001; Miczek et al.,
2007; see Latham & Mason, 2004 for a review). Male laboratory mice of various inbred (e.g.,
C57BL/6J, C-3H agoutis, BALB/C albinos) and outbred (e.g., Swiss CD1, Swiss-Webster,
CF-1, Tuck TO, unspecified albino) strains, as well as wild-descended mice, consistently attack male intruders, especially following isolation or housing with a female, although wild- descended and outbred Swiss CD1 males show greater aggression than some inbred strains
(e.g., Brain & Bowden, 1979; Edwards, 1969; Ginsburg & Allee, 1942; Parmigiani et al.,
1999; Peters et al., 1972; Scott & Fredericson, 1951; Uhrich, 1938).
There has been much less research conducted on aggression or agonistic behaviour in female than male mice. Most of this research has focused on circumstances related to reproduction, such as following housing with a male or when protecting their litter, in
89 so-called maternal or postpartum aggression (e.g., Haney et al., 1989; Haney & Miczek,
1989; Palanza et al., 1995; Palanza et al., 2005). Studies with outbred Swiss albino mice show that this type of female aggression is qualitatively different from the typical intrasexual aggression studied in males, as it is less ritualized and more violent (Al-Maliki et al., 1980;
Svare & Gandelman, 1973). Non-reproduction-related female aggression, while less frequently studied than males, also exists. Virgin or non-reproductively-active outbred albino females will fight during group formation, following isolation, or if the intruder is a lactating female (Brain & Haug, 1992; Morè, 2008; Uhrich, 1938), although virgin females of some inbred strains (e.g., C57BL/6J, 129) only rarely attack same-sex intruders (Ogawa et al.,
1998a) (Ogawa et al., 2004).
Higher levels of intrasexual aggression are observed in wild or wild-descended females than in both inbred and outbred laboratory mice, suggesting that interfemale aggression has been largely eliminated from these domesticated mice (Palanza et al., 2005; Parmigiani et al.,
1999; Scott, 1966). Alternatively, these domesticated females may use other agonistic behaviours than attacks, such as chasing, pinning, or aggressively grooming the intruder
(Alleva, 1993; Clipperton et al., 2008; Grant & Mackintosh, 1963) that are often not assessed in tests for aggression. These behaviours would not lead to the violent exclusion of the intruder from the territory, but rather would be aimed at establishing dominance over the intruder. The study of female agonistic interactions in tests of aggression therefore requires the use of a comprehensive analysis, often referred to as “ethological” analysis, based on the mouse’s full behavioural repertoire (the ethogram; Grant & Mackintosh, 1963) and that involves more than just measures of attacks (Branchi et al., 2006; Miczek et al., 2001; Mos &
Olivier, 1989; Olivier et al., 1989; Pietropaolo et al., 2004).
90 Steroid hormones are known to mediate social and aggressive interactions in mice. It has long been known that castration eliminates intermale aggression in wild as well as inbred and outbred laboratory mice (e.g., Beeman, 1947; Brain & Bowden, 1979; Luttge, 1972; Simon
& Whalen, 1986; Uhrich, 1938). The administration of prototypical male gonadal hormone, testosterone (T), will restore aggression in these castrated males (Beeman, 1947; Brain &
Bowden, 1979; Luttge, 1972; Simon & Whalen, 1986; Tollman & King, 1956).
Testosterone can affect behaviour either directly, or indirectly through one of its two main metabolites: the androgen dihydrotestosterone (DHT) and the estrogen estradiol. In castrated males of some strains (e.g., Swiss-Webster, CD1, Tuck TO; Brain & Bowden,
1979; Brain & Poole, 1976; Luttge, 1972; Luttge & Hall, 1973; Simon & Whalen, 1986),
DHT restored intermale aggression near pre-castration levels, while in other strains these effects of DHT were not found (e.g., CFW, CF-1; Cologer-Clifford et al., 1999; Simon &
Whalen, 1986).
Estrogens were also shown to mediate aggression. Estradiol administration reinstated aggression in castrated male mice of a number of strains (CF-1, CFW, Tuck TO, Swiss-
Webster; Bowden & Brain, 1978; Brain & Bowden, 1979; Brain & Poole, 1976; Edwards &
Burge, 1971; Simon & Whalen, 1986). This effect is weaker in castrated male CD1 mice
(Finney & Erpino, 1976; Simon & Whalen, 1986). Similarly, anti-estrogen treatment reduces aggression in castrated CD1 or Tuck TO males given T replacement (Clark & Nowell, 1979;
Luttge, 1979). Furthermore, male mice in which the gene for aromatase has been inactivated, who therefore cannot convert T to estradiol, show an estradiol-reversible reduction in
91 aggression (Toda et al., 2001). Thus both androgens and estrogens have been shown to be involved in the mediation of aggressive behaviour in a number of strains of mice.
Work with C57BL/6J and 129 inbred mice and Swiss black mice in which either of the two intranuclear estrogen receptors (ER), ERα or ERβ, had been made non-functional (so- called “knockout,” KO, mice) suggested that the two ERs are differently involved in the mediation of aggression (Ogawa et al., 1998b; Ogawa et al., 1998a) (Choleris et al., 2008a;
Clipperton et al., 2008; Ogawa et al., 1999). Male αERKO mice showed reduced aggression towards an olfactory bulbectomized (OBX) or intact male intruder. In addition, αERKO males did not show T-enhanced aggressive responses (Ogawa et al., 1998b) and displayed female-directed aggression (Scordalakes & Rissman, 2003). Unlike αERKO males, male
βERKO mice showed increased aggression in puberty and early adulthood towards OBX male intruders, and had normal estrogen-induced aggression (Nomura et al., 2002; Nomura et al., 2006). In male mice, therefore, activation of ERα appears to increase aggression, while
ERβ seems to decrease it in a manner that is dependent upon the experience and age of the animal. Both effects on aggression may be related to the impairment of the αERKO and
βERKO mice in social recognition (Choleris et al., 2003; Choleris et al., 2006; Choleris et al.,
2009).
There have been few studies on female aggression using ER-specific KO mice. In one study, virgin αERKO females showed increased aggression towards ovariectomized (ovx), hormone-primed same-sex intruders, but not toward OBX males, and this aggression was reduced by estrogen replacement following ovariectomy (Ogawa et al., 1998a). Gonadally intact female βERKO mice showed no aggression towards ovx female intruders, and
92 exhibited reduced aggression following T administration (Ogawa et al., 2004; Ogawa et al.,
1999). Chronic KO of the genes for ERα and ERβ thus have the opposite effects in males and females, suggesting that intrasexual aggression may be differentially mediated by estrogens in the two sexes. However, these studies with KO mice do not allow dissection of developmental from activational effects, nor do they allow the contributions of flanking genes or compensatory mechanisms to be ruled out (Choleris et al., 2008a; Waddington et al.,
2005).
Single acute administration of the ERβ agonist 7-Bromo-2-(4-hydroxyphenyl)-1,3- benzoxazol-5-ol (WAY-200070; Malamas et al., 2004) to adult ovx female CD1 mice reduced Dominance Scores and increased submissive behaviour during their interactions with a familiar, ovx same-sex cagemate (Clipperton et al., 2008). This suggests that the acute activation of ERβ in adult mice affects aggressive behaviour. In the present study, we used the resident-intruder test to investigate the effects of acute administration of WAY-200070 on the responses of gonadally intact and gonadectomized (gonadex) male and female mice to unfamiliar, gonadex same-sex intruders. We combined the use of a standardized test for aggression with a comprehensive behavioural analysis, to assess the effects of WAY-200070 on different types of agonistic behaviour.
3.3 Method
3.3.1 Animal Housing
Adult CD1 mice (Mus musculus) were obtained from Charles River, QC, Canada. Mice used as intruders were housed in same-sex group cages of four to six (46 x 26 x 21 cm). Mice used as residents were individually housed for two weeks in order to develop a home cage
93 territory in clear polyethylene cages (26 x 16 x 12 cm). All cages were provided with corn cob bedding, environmental enrichment (plastic containers and paper nesting material), and food (Teklad Global 14% Protein Rodent Maintenance Diet, Harlan Teklad, Madison, WI) and tap water ad libitum. The mice were held in the Central Animal Facility at the University of Guelph, under a 12:12h reversed light/dark cycle (lights on at 2000h) at a temperature of
21±1°C.
3.3.2 Experimental Subjects
Subjects were 132 female and 152 male CD1 mice. Seventy-seven of the female mice were ovariectomized (ovx), and 95 of the male mice were castrated, both at two to three months of age by the supplier, one or two days before being shipped to the University of
Guelph, and at least two weeks before the experiment. The vaginal cytology of the ovx mice confirmed that they were not cycling.
Since both T and estrogens have been implicated in aggression, 28 ovx and 40 castrated males were randomly selected to be intruders. The use of group-housed, same-sex, gonadex mice results in more “standardized” intruders with uniformly low levels of circulating sex hormones. In males, this avoids possible confounding effects due to the fact that T both produces and elicits attacks (Luttge, 1972; Mugford, 1974; Mugford & Nowell, 1970; Simon
& Whalen, 1986). The use of ovx female intruders eliminates the estrous cycle dependent variability that has been shown to affect aggression in outbred, wild-descended mice (Hyde
& Sawyer, 1977). Furthermore, as both outbred and inbred gonadex mice typically initiate little aggression themselves (e.g., Beeman, 1947; Gray et al., 1978), using them as intruders enabled us to assess the level of aggression initiated by the resident mouse, while preventing
94 the encounter from escalating to the point at which testing would have to be interrupted. This allowed us to perform a full behavioural analysis over the entire duration of the test in all sexes and gonadal conditions. Intruders were reused no more than four times, at a minimum interval of three days between tests, and were replaced when they reached four months of age.
The remaining 104 female and 112 male mice were used as residents. No cage changes took place for at least three days prior to testing to allow the residents to establish a territory in their home cages. No resident experimental mouse was reused, and each resident was assigned to an independent treatment group, as outlined in Table 14. Intact female residents were at mixed stages of the estrous cycle. This research was conducted in accordance with the regulations of the Canadian Council on Animal Care and was approved by the
Institutional Animal Care and Use Committee of the University of Guelph.
3.3.3 Apparatus
Social interactions between residents and intruders took place in the resident’s home cage during the dark phase of the light/dark cycle. Clear Plexiglas lids were used to allow videotaping from above with an 8 mm Sony Handycam (in Nightshot) for subsequent behavioural analysis of the interactions.
3.3.4 Drug
The selective estrogen receptor β agonist 7-Bromo-2-(4-hydroxyphenyl)-1,3-benzoxazol-
5-ol (WAY-200070), which is 78 times more selective for ERβ than ERα in mice (compound
92 in Malamas et al., 2004; Harris, 2007), was obtained courtesy of Wyeth Laboratories, Inc.,
Madison, NJ, and was suspended in sesame oil. 95 Table 14. Distribution of treatment groups. Numbers indicate mice remaining in each group after the removal of outliers (number in parentheses) Males Females Gonadally Gonadex Gonadally Gonadex Treatment Intact (Castrated) Intact (Ovx) Uninjected n = 9 (0) n = 12 (0) n = 10 (3) n = 10 (1) Sesame Oil Vehicle n = 13 (0) n = 10 (0) n = 11 (1) n = 8 (2) 30 mg/kg WAY-200070 n = 11 (1) n = 13 (0) n = 12 (0) n = 10 (0) 90 mg/kg WAY-200070 n = 10 (1) n = 12 (0) n = 10 (0) n = 10 (1) 180 mg/kg WAY-200070 n = 12 (0) n = 8 (2) n = 12 (0) n = 11 (1)
96 For each of the four conditions (gonadally intact males, gonadally intact females, castrated males, and ovx females), five independent groups of resident mice were formed: a sesame oil vehicle control (OIL) and three drug-treated groups, each receiving one of three doses of WAY-200070, 30 mg/kg (LOW), 90 mg/kg (MID), or 180 mg/kg (HIGH). In addition, in order to assess potential effects of sesame oil on behaviour (Anton et al., 1974;
Colafranceschi et al., 2007; Curtis & Wang, 2005), an uninjected group was included for each gonadal condition. This also allowed us to compare gonadex and gonadally intact mice in untreated animals. The distribution of doses in each sex and gonadal condition is outlined in Table 14. Treatment was administered via a single intraperitoneal (i.p.) injection, at a volume of 10 ml/kg, approximately 72h prior to the experiment.
Each treatment group consisted of a unique set of resident mice, and each resident mouse was tested only once. This pattern of administration of the drug allows for the assessment of the acute, rather than chronic, behavioural effects of steroid hormones (Choleris et al., 2009;
Jasnow et al., 2007; Mickley & Dluzen, 2004; Priest et al., 1997; Walf & Frye, 2005) and parallels the administration procedure that showed effects of this drug on social learning and agonistic behaviours between familiar ovx females (Clipperton et al., 2008). Intraperitoneal administration of compounds dissolved in sesame oil has been used in a number of other studies (e.g., Chu et al., 2006; Clipperton et al., 2008; Cragg et al., 2011; Curtis & Wang,
2005; Eroschenko et al., 1997; Hoyk et al., 2006; Löfgren et al., 2008; Mahieu et al., 2004;
Rankin et al., 1997; Valentovic et al., 1994) and was chosen to ensure the mice were receiving the proper dosage, since in preliminary investigations we observed excessive leakage of the oily substance following subcutaneous injections.
97 The specific doses of WAY-200070 and the time of administration were chosen on the basis of information obtained directly from the provider. Lower doses of WAY-200070 do not cause nuclear translocation of ERβ receptors and do not seem to be biologically active
(Hughes et al., 2008). Consistently, in previous research both single and repeated administration of lower doses had no effects on either social (Choleris et al., 2009; Cragg et al., 2011) or non-social behaviour (Hughes et al., 2008). The acute i.p. administration of the doses used in this study, however, did affect social learning, social interactions (Choleris et al., 2009; Clipperton et al., 2008) and social recognition (Choleris et al., 2009; Cragg et al.,
2011). These effects were different from those of a highly specific ERα agonist (Choleris et al., 2009; Clipperton et al., 2008; Cragg et al., 2011). These behavioural results, together with the finding that high doses of WAY-200070 do not affect ERα-rich tissue (Malamas et al.,
2004; Mewshaw et al., 2005), indicate ERβ-specific effects of WAY-200070 at these doses.
3.3.5 Procedure
Three days before testing, the resident mice of each of the four conditions (gonadally intact males, gonadally intact females, castrated males, and ovx females) were randomly assigned to one of the five groups and administered treatment (uninjected control, sesame oil vehicle, 30 mg/kg WAY-200070, 90 mg/kg WAY-200070, or 180 mg/kg WAY-200070). To assist in identification, the intruder mice were coloured with black magic marker at least 12h prior to testing. At this time, all residents and intruders were moved into the testing room and left undisturbed for at least 12h. During their active phase, each resident mouse had a same- sex, gonadex intruder placed into his or her home cage. The mice were left undisturbed to freely interact for 15 min while being videotaped.
98 The videotaped social interactions were scored by two trained observers, who each watched half of each treatment group, were unaware of the animals’ treatment, and had an interobserver reliability correlation of 0.92. The interactions were analyzed for 21 behaviours based on the ethogram by Grant and Mackintosh (1963; see Table 15, modified from
Clipperton et al., 2008, for behaviour descriptions) using The Observer Video Analysis software (Noldus Information Technology, Wageningen, the Netherlands). The behavioural analysis focused on the treated mouse, the resident. The behaviour of the intruder was collected only in relation to the behaviour of the resident (i.e., Attacks Received, Reciprocal
Attacks, Aggressive Postures, Social Inactivity), and in the reciprocal pairs of behaviours
(Dominant/Submissive Behaviour, Chasing/Avoidance of the Intruder). Dominant behaviour on the part of one mouse was typically met by submissive behaviour on the part of the other.
Ten categories of individual behaviours were formed (see Table 16, modified from
Clipperton et al., 2008) in order to assess potential non-specific effects of the drug on the total activity and total social and non-social explorations of the mice, as well as on the overall levels of agonistic behaviour displayed during the interaction. These categories are meant to provide an overall view of the animals’ behaviour, and thus allow comparison to non-ethological types of analysis (e.g., Cushing et al., 2008; Sanchis-Segura et al., 2000;
Walf et al., 2008b). However, with this paper we wish to emphasize the importance of looking at the component behaviours of these categories, as this will determine whether the effects observed are due to specific individual behaviours or a general effect across a category. This is particularly true in the case of aggression, as an increase in total levels of agonistic behaviour may be due to effects on types of aggression-related behaviour that have potentially different functional implications; such as overt attacks for the establishment of
99 Table 15. Description of scored behaviours Behaviour Description Social Behaviours The resident mouse actively follows, or pursues and chases the Chasing the Intruder intruder; reciprocal to avoid. The resident mouse is in control; includes pinning of the intruder, Dominant Behaviour aggressive grooming, crawling over or on top, and mounting attempt. Physical attacks, including dorsal/ventral bites. Only the frequency Attacks Delivered of attacks was measured. Physical attacks which include box/wrestle, offensive and defensive Aggressive Postures postures, lateral sideways threats and tail rattle. Physical attacks with a locked fight including tumbling, kick-away Reciprocal Attacks and counterattack where the attacker cannot be identified. Avoidance of the The resident withdraws and runs away from the intruder while the Intruder intruder is chasing. The intruder mouse is in control; includes crawl under, supine Submissive posture (ventral side exposed), prolonged crouch, and any other Behaviour behaviour in which the intruder is dominant (e.g., the intruder pins, aggressively grooms, etc., the resident). Physical attacks including bites to dorsal/ventral regions. Only the Attacks Received frequency of attacks was measured. Defensive Upright Species-typical defensive behaviour; upright with the head tucked Posturing and the arms ready to push away. Social Inactivity Includes sit/lie/sleep together. Oronasal Active sniffing of the intruder’s oronasal area. Investigation Body Investigation Active sniffing of the intruder’s body. Anogenital Active sniffing of the intruder’s anogenital region. Investigation Risk assessment behaviour; back feet do not move and front feet Stretched approach the intruder. Only the frequency of stretched approaches Approaches was measured. Often from across the cage; the resident’s attention is focused on the Approaching and/or intruder, head tilted toward the intruder and movements toward the Attending to the intruder; this becomes “Chasing the Intruder” once along the tail or Intruder sniff if within 1.5 cm of the intruder. Non-social Behaviours Horizontal Movement around the cage; includes active sniffing of air and Exploration ground. Movement to investigate upwards, both front feet off the ground; Vertical Exploration includes sniffing, wall leans and lid chews (less than 3). Digging Rapid stereotypical movement of forepaws in the bedding. Abnormal “Strange” behaviours, including spinturns, repeated jumps/lid Stereotypies chews/head shakes (more than 3). Solitary Inactivity No movement; includes sit, lie down and sleep. Self-Grooming Rapid movement of forepaws over facial area and along body.
100 Table 16. Descriptions of grouped and composite behaviours Behaviour Descriptions All behaviours involving activity, both social and non-social. Total Activity Excluded from this group are Inactive Alone, Inactive Together, and Self-Groom. Follow Intruder, Dominant Behaviour, Attack Delivered, Aggressive Postures, Reciprocal Attacks, Avoid Intruder, Submissive Behaviour, Attack Received, Defensive Upright Posturing, Inactive Total Social Behaviour Together, Oronasal Investigation, Body Investigation, Anogenital Investigation, Stretched Approaches, and Attend To/Approach Intruder. Follow Intruder, Dominant Behaviour, Attack Delivered, Avoid Intruder, Submissive Behaviour, Attack Received, Defensive Upright Posturing, Aggressive Postures, and Reciprocal Attacks. Total Agonistic This composite behaviour represents the overall levels of agonism Behaviours present in the resident-intruder interactions and does not indicate the direction of the agonistic behaviour (i.e., whether agonistic behaviour is directed toward the resident or toward the intruder). Agonistic Behaviour Follow Intruder, Dominant Behaviour, and Attack Delivered. Delivered Agonistic Behaviour Avoid Intruder, Submissive Behaviour, Attack Received, and Received Defensive Upright Posturing. Total agonistic behaviour delivered minus total agonistic behaviour received. A negative score indicates that the resident was the Dominance Score submissive animal in the pair, while a positive score signifies that the resident was the dominant animal. Oronasal Investigation, Body Investigation, Anogenital Social Investigation Investigation, Stretched Approaches, and Attend To/Approach Intruder. Horizontal Exploration, Vertical Exploration, Dig, Stereotypies, Non-social Behaviours Inactive Alone, and Self-Groom. Non-social Locomotor Horizontal Exploration, Vertical Exploration, and Dig. Behaviours Non-social Non- Inactive and Self-Groom. locomotor Behaviour
101 territories and/or behaviours aimed at establishing a dominance hierarchy within a shared territory. The direction of the agonistic behaviour was defined whenever possible, and a
“Dominance Score” was also calculated for each pair of mice by subtracting the amount of agonistic behaviour received from the amount of agonistic behaviour delivered (as described in Table 16 and modified from Clipperton et al., 2008). The Dominance Score thus took into account the reciprocity of “dominant” and submissive behaviour involved in agonistic interactions.
To determine the phase of the estrous cycle in intact females, and to confirm that the ovx mice were not cycling, vaginal smears were taken from all females immediately following testing. These slides were stained with Giemsa stain (Sigma-Aldrich, Oakville, ON) and examined under a microscope using 100x magnification. Proestrus was defined as consisting of predominantly nucleated cells, estrus of primarily non-nucleated, cornified cells, and diestrus of chiefly leukocytes.
3.3.6 Statistical Analysis
The 15 min social interaction period was divided into three 5 min intervals in order to assess the time course of the behaviours. The frequency (FREQ), duration (DUR), and latency (LAT) of each behaviour were measured. If a behaviour was not performed, a latency of 900 s was assigned. For numerous behaviours, Levene’s test for equality of variance revealed that there were significant differences between group variances (Levene statistics between 10.39-2.63, p values from less than 0.001 to 0.045), therefore violating the assumption of equality of variance necessary for parametric statistics. Results were thus analyzed using non-parametric statistics, which do not require assumptions to be made about
102 the underlying population. Specifically, we used the non-parametric equivalent of analysis of variance, the Kruskal-Wallis test, followed by the non-parametric Mann Whitney U test for binary comparisons. When the results of the analyses on DUR and FREQ were in agreement, only those of DUR are reported; thus, if not specified, results are DUR. Only the main results are reported in the text. All other results, including the time course data, are in Appendix 2.
The data from 14 mice that were two standard deviations from the mean on four or more behaviours were removed; these mice were distributed across conditions and treatment groups (see Table 14).
Two sets of analyses were performed to assess 1) the effects of drug treatment on each mouse condition separately, and 2) differences between sexes and between intact and gonadex same-sex mice in untreated animals. First, for each condition (gonadex or intact male or female mice), the effects of the drug are reported in comparison to the sesame oil control group. Effects of sesame oil per se are reported in comparison to the uninjected group. The variable of interest in this analysis was the treatment (sesame oil, 30 mg/kg
WAY-200070, 90 mg/kg WAY-200070, 180 mg/kg WAY-200070). For the gonadally intact females, a second variable, the phase of the estrous cycle (proestrus, estrus, diestrus), was included.
The second set of analyses was performed to assess sex differences and to compare gonadex and gonadally intact mice in the uninjected control groups. Variables were sex
(gonadally intact male, gonadally intact female) and gonadal condition (gonadally intact, gonadex; analyzed separately for each sex). In the gonadally intact females, effects of the estrous cycle (proestrus, estrus, diestrus) were also analyzed.
103 In all of the analyses in which it was assessed, the phase of the estrous cycle did not have significant effects, neither as a main factor nor in interaction with drug treatments, probably because of the low number of mice in each group. Consequently, the phase of the estrous cycle was removed from the models, and results of the analyses of the data obtained from gonadally intact females without this variable are reported below. All analyses were performed using SPSS 13.0 for Windows (SPSS Inc., Chicago, IL).
3.4 Results
For each experimental group (gonadally intact or gonadex males and females) we report the main results observed below. Additional statistically significant minor results are reported in detail in Appendix 2.
3.4.1 Effects of WAY-200070 in Gonadally Intact Males
Overall, in gonadally intact males WAY-200070 specifically increased agonistic behaviour other than overt attacks and reduced anxiety-like behaviour without affecting non- agonistic social behaviour or non-social activity. Dominant Behaviour was increased by
HIGH (U = 42.00 z = -1.96, p = 0.050; see Figure 5a). Stretched Approaches were decreased by HIGH (U = 39.00, z = -2.15, p = 0.039; see Figure 5b). OIL also had some effects, e.g., increasing Aggressive Postures (U = 25.00, z = -2.27, p = 0.023) and decreasing the FREQ of
Received Agonistic Behaviour (U = 28.00, z = -2.04, p = 0.041). Please see Appendix 2,
Table 17 for additional results.
3.4.2 Effects of WAY-200070 in Gonadally Intact Females
Overall, in gonadally intact females, WAY-200070 increased both agonistic and non- agonistic social behaviours as well as non-social activity. 104 Figure 5. WAY-200070 effects on gonadally intact males. a) Duration of dominant behaviour. b) Frequency of stretched approaches. * significant difference vs vehicle, p < 0.05.
105 2 There was an effect of treatment on Total Activity (χ (4) = 11.72, p = 0.020), both the
DUR (U = 32.00, z = -2.09, p = 0.036; see Figure 6c) and FREQ (U = 10.00, z = -3.45, p =
0.001) of which were increased by HIGH. This was predominantly due to effects of treatment
2 on the frequencies of Total Social Behaviour (χ (4) = 24.72, p < 0.001; U = 8.00, z = -3.57, p
2 < 0.001; see Figure 6d) and Social Investigation (χ (4) = 20.54, p < 0.001; U = 6.00, z = -
3.69, p < 0.001; see Figure 6e), which were increased by HIGH. There was a treatment effect
2 on the FREQ of Agonistic Behaviour Delivered (χ (4) = 11.24, p = 0.024) and the frequencies of both this category (U = 28.50, z = -2.31, p = 0.021; see Figure 6a) and the
Dominance Score were increased by HIGH (U = 33.50, z = -2.00, p = 0.045; see Figure 6b).
2 There was a treatment effect on the FREQ of Non-social Non-locomotor Behaviours (χ (4) =
15.88, p = 0.003), which were decreased by HIGH (U = 30.00, z = -2.22, p = 0.026; see
Figure 6f), and this dose also increased the FREQ of the non-social locomotor behaviour
Digging (U = 34.00, z = -2.08, p = 0.038). OIL also affected a number of behaviours, including increasing Total Social Behaviours (U = 16.00, z = -2.75, p = 0.006) and decreasing Digging (U = 23.00, z = -2.35, p = 0.019). Please see Appendix 2, Table 18 for additional results.
3.4.3 Effects of WAY-200070 in Castrated Males
Overall, WAY-200070 effects on castrated males are similar to those in females and show that non-social as well as social non-agonistic behaviours were increased by treatment, although unlike gonadally intact females, there were few effects on agonistic social behaviours.
106 Figure 6. Effects of WAY-200070 on gonadally intact females. a) Frequency of agonistic behaviour delivered. b) Frequency of dominance score. c) Frequency of total social behaviour. d) Frequency of social investigation. e) Duration of total activity. f) Frequency of non-social non-locomotor behaviour. * significant difference vs vehicle, p < 0.05. ** significant difference vs vehicle, p < 0.01.
107 2 The FREQ of Total Activity (χ (4) = 14.80, p = 0.005; U = 8.00, z = -2.84, p = 0.004; see
Figure 7a) was affected by treatment and increased by HIGH, as were the FREQ of Non-
2 social Behaviour (χ (4) = 12.14, p = 0.016; U = 13.00, z = -2.40, p = 0.016; see Figure 7c) and several other elements of Total Activity. The FREQ of Social Investigation was affected
2 by treatment (χ (4) = 19.37, p = 0.001; see Figure 7b), decreased by OIL (U = 9.00, z = -3.36, p = 0.001) and increased by LOW (U = 33.00, z = -1.99, p = 0.047) and HIGH (U = 33.00, z
2 = -1.99, p = 0.047). The DUR of Social Investigation was also affected by treatment (χ (4) =
12.80, p = 0.012), but was decreased by HIGH (U = 16.00, z = -2.13, p = 0.033). OIL had a number of effects, including decreasing Total Activity (U = 21.50, z = -2.54, p = 0.011) and the FREQ of Social Investigation (U = 9.00, z = -3.36, p = 0.001), and increasing Total Non-
Social Behaviours (U = 29.00, z = -2.04, p = 0.041). Please see Appendix 2, Table 19 for additional results.
3.4.4 Effects of WAY-200070 in Ovariectomized Females
Overall, WAY-200070 had similar effects on ovx females to those observed in castrated males, with both non-social and social non-agonistic behaviours being increased.
2 The FREQ of Total Activity was affected by treatment (χ (4) = 11.14, p = 0.025; see
Figure 7d), with OIL decreasing (U = 16.00, z = -2.13, p = 0.033), and the LOW (U = 13.50, z = -2.36, p = 0.018) and HIGH (U = 12.00, z = -2.64, p = 0.008) doses of the drug increasing this behaviour, likely due to increases in social and non-social behaviours. The FREQ of
2 Total Social Behaviours was affected by treatment (χ (4) = 11.24, p = 0.024) and increased by HIGH (U = 10.00, z = -2.81, p = 0.005), in part because of similar results in the FREQ of
108 Figure 7. WAY-200070 effects on gonadectomized male and female mice. a) Frequency of total activity in castrated males. b) Frequency of social investigation in castrated males. c) Frequency of non-social behaviour in castrated males. d) Frequency of total activity in ovx females. e) Frequency of social investigation in ovx females. f) Frequency of non-social behaviour in ovx females. * significant difference vs vehicle, p < 0.05. ** significant difference vs vehicle, p < 0.01.
109 2 Social Investigation, which was also affected by treatment (χ (4) = 13.06, p = 0.011; see
Figure 7e) and increased by LOW (U = 17.00, z = -2.05, p = 0.041) and HIGH (U = 11.50, z
= -2.69, p = 0.007). The FREQ of Total Non-social Behaviours was increased by LOW (U =
16.50, z = -2.09, p = 0.037; see Figure 7f). Additionally, Total Activity was decreased by
OIL (U = 16.00, z = -2.13, p = 0.033). Please see Appendix 2, Table 20 for additional results.
3.4.5 Sex Differences in Uninjected, Gonadally Intact Mice
Overall, while males and females showed qualitatively different types of agonistic behaviour, the two sexes devoted a similar amount of time to agonistic responses to the intruder. In uninjected, gonadally intact CD1 mice, males showed more Reciprocal Attacks
(U = 10.00, z = -3.30, p = 0.001), and a shorter LAT to Reciprocal Attacks (U = 12.00, z = -
3.56, p < 0.001) and to Aggressive Postures (U = 32.50, z = -2.05, p = 0.041) than females.
Males also performed Attacks Delivered, which females did not (U = 10.00, z = -2.91, p =
0.004; see Figure 8b), while females showed more Submissive Behaviour (U = 21.00, z = -
1.96, p = 0.0499; see Figure 8c). However, despite these differences in the components of the categories, gonadally intact male and female CD1 mice did not significantly differ in the
DUR of Total Agonistic Behaviours (see Figure 8a) or Agonistic Behaviour Delivered.
Additionally, males switched between behaviours more frequently (U = 16.00, z = -2.37, p =
0.018), and females showed a longer LAT to Social Investigation (U = 13.50, z = -2.89, p =
0.004; see Figure 8f). Please see Appendix 2, Table 21 for additional results.
3.4.6 Differences Between Gonadally Intact and Castrated Uninjected Males
Castrated males were more like gonadally intact females than gonadally intact males in some respects. Castrated males performed fewer Attacks Delivered (U = 14.00, z = -2.89, p =
110
Figure 8. Sex differences and gonadectomy effects on agonistic behaviour. a) Duration of total agonistic behaviour. b) Frequency of attacks delivered. c) Duration of submissive behaviour. d) Frequency of approaching/attending to the intruder. e) Frequency of stretched approaches. f) Latency to social investigation. * significant difference vs gonadally intact males, p < 0.05. ** significant difference vs gonadally intact males, p < 0.01. + significant difference vs gonadally intact females, p < 0.05. 111 0.004; see Figure 8b), less Approaching and/or Attending to the Intruder (U = 9.00, z = -3.20, p = 0.001; see Figure 8d), and more Submissive Behaviours (U = 21.00, z = -2.35, p = 0.019; see Figure 8c) than the gonadally intact males. Castrated males also showed a longer LAT to
Social Investigation (U = 17.00, z = -2.63, p = 0.008; see Figure 8f) than the gonadally intact males. Please see Appendix 2, Table 22 for additional results.
3.4.7 Differences Between Gonadally Intact and Ovariectomized Uninjected Females
Ovx females appeared to be less anxious than gonadally intact females, showing fewer
Stretched Approaches (U = 21.50, z = -2.19, p = 0.029; see Figure 8e) and more frequent
Oronasal Investigation (U = 21.00, z = -2.20, p = 0.028) than the gonadally intact females.
Please see Appendix 2, Table 23 for additional results.
3.5 Discussion
By performing an ethological analysis where we looked at a range of behaviours, we were able to obtain a comprehensive assessment of the effects of the ERβ agonist WAY-
200070 on agonistic and non-agonistic social and non-social behaviour in mouse responses to a gonadectomized (gonadex), same-sex intruder, and the specific behaviours that make up these categories (e.g., the specific type of agonistic behaviour performed). It also allowed us to compare the responses of male and female mice in the same testing conditions and to determine that, over a 15 min test, CD1 females do not spend less time in agonistic interactions with a gonadex, same-sex intruder than males, but rather respond to an opponent using agonistic behaviours other than attacks.
112 3.5.1 WAY-200070 Effects
The ERβ agonist WAY-200070 affected behaviour in the sexes and gonadal conditions differently. WAY-200070 increased intrasexual agonistic behaviour in both gonadally intact
CD1 males and females while having little effect on agonistic behaviour in gonadex CD1 mice. In both gonadally intact and ovx CD1 females, WAY-200070 increased overall activity, but this effect was on both some social and some non-social behaviours in gonadally intact females, and specific to non-agonistic behaviours in ovariectomized (ovx) females.
Similarly, WAY-200070 increased several social behaviours in castrated males while specifically increasing agonistic behaviour, and no other social behaviours, in gonadally intact males.
Our results showing that acute activation of ERβ increased non-attack agonistic behaviour appear to be in contrast to the increased aggression seen in βERKO males of the same age (but a different strain) as the CD1 mice in the current study (Nomura et al., 2002).
This difference suggests that the effects seen in the βERKO males are at least in part developmental, not activational, or that the timing of the ERβ manipulation may affect its mediation of aggression. Alternatively, ERβ may be differently involved in different types of agonistic behaviours, as suggested by the present finding that WAY-200070 had clear enhancing effects on non-attack agonist behaviours while having no effects on direct attacks in gonadally intact CD1 mice of both sexes. A similar pattern of effects was found in ethopharmacological studies of maternal aggression in rats, in which diazepam affected agonistic behaviours but not biting attacks (Mos & Olivier, 1989), suggesting that attacks and non-attack agonistic behaviours may be differently regulated. It is possible that similar effects on non-attack agonistic behaviours occurred with the βERKO mice, but were simply
113 not recorded. Clearly, more studies conducting comprehensive behavioural analyses are needed before a full understanding of the involvement of ERβ in different aspects of agonistic behaviours can be achieved.
In gonadally intact males, WAY-200070 reduced the number of stretched approaches performed towards the intruder, suggesting an anxiolytic effect (Blanchard et al., 1993;
Choleris et al., 2001). This is in agreement with findings that activation of ERβ reduced anxiety-like behaviours in mice on numerous different tasks (see Choleris et al., 2008a for a review). Thus, our WAY-200070-treated gonadally intact males may have been more likely to perform agonistic behaviours because they were less afraid of the intruder. As WAY-
200070 did not affect the amount of non-social active behaviours or non-agonistic social behaviours in gonadally intact males, neither the increase in agonistic behaviour nor the decrease in anxiety-like behaviour can be ascribed to generalized effects on overall activity.
Like the gonadally intact males, intact females treated with WAY-200070 showed increased agonistic behaviour and had a higher Dominance Score, indicating that they delivered more aggression than they received. However, this was accompanied by an increase in overall activity. Although this non-specific effect could account for the increase in agonistic behaviour, the fact that there were no increases in received agonistic behaviours and such active behaviours as avoidance suggests that ERβ may, in fact, increase agonistic behaviours specifically in gonadally intact females. This would be in agreement with previous studies, which found a decrease in interfemale aggression following testosterone administration in gonadally intact female βERKO mice, and increased aggression toward ovx same-sex intruders in gonadally intact αERKO female mice (Ogawa et al., 1998a) (Ogawa et
114 al., 2004). Other ethopharmacological studies on aggression in rats, involving serotonin receptor agonists, also found analogous effects on activity and agonistic behaviour, while leaving other behaviours unaffected (e.g., Olivier & Mos, 1992). These results further highlight the advantages of ethological assessments of animal behaviours, as they can help determining the underlying effects on specific behaviours that may drive observed increases or decreases in generalized activity.
The specificity of WAY-200070 effects on non-attack agonistic behaviours suggest that
ERβ may be involved in the mediation of aspects of social agonistic interactions that are aimed at asserting dominance (Alleva, 1993; Branchi et al., 2006; Miczek et al., 2001;
Miczek et al., 2007; Pietropaolo et al., 2004). In our study, there was typically a clear predominance of agonistic behaviour delivered, rather than received, by the resident, which suggests that the resident would have established dominance over the same-sex intruder had our study involved a long-term pairing and observation of the mice. Dominance hierarchies, in which dominant females gain preferential access to space and resources, and possibly even exclusive breeding rights, have been observed in wild-descended female mice (Hurst, 1987).
More long-term studies with comprehensive behavioural analyses in males as well as females are needed for an understanding of ERβ mediation of different facets of agonistic behaviours.
The overall increase in the duration of all active behaviours by WAY-200070 observed in the gonadally intact CD1 females was also observed in the frequency measures, reflecting the fact that these mice switched between behaviours more frequently (“behavioural shifting”).
This is in agreement with the known facilitatory effects of estrogens on running wheel activity in several strains of female mice (e.g., Garey et al., 2001; Morgan & Pfaff, 2002).
115 Conversely, gonadally intact αERKO and βERKO females displayed increased activity in response to social stimuli (Choleris et al., 2006), suggesting that, as for male aggression, female activity in a social context may be differentially affected by the timing of the manipulation of ERβ. Alternatively, ERβ involvement in activity may differ in social and non social contexts.
In both male and female gonadex CD1 mice, WAY-200070 increased the amount of behavioural shifting, social investigation, and individual behaviours, although it did not affect agonistic behaviour. These results suggest that activation of ERβ in adult CD1 mice with extremely low levels of endogenous sex hormones enhances their overall activity and arousal (Pfaff et al., 2008) without affecting agonistic behaviours.
3.5.2 Sex and Gonadectomy Differences
In addition to the WAY-200070 effects, we observed a number of sex differences. In agreement with the literature on aggression in various strains of mice, in the present study the gonadally intact male CD1 mice attacked the same-sex, gonadex intruder more and made more reciprocal attacks than the gonadally intact females (e.g., Anton, 1969; Edwards, 1970).
This supports the notion that the aggression (i.e., attacks) commonly studied in the mouse resident-intruder test is typical only for males; females instead tend to use aggressive postures and other non-attack agonistic behaviours, such as aggressive allogrooming or pushing down the intruder (Clipperton et al., 2008; Grant & Mackintosh, 1963). When looking at both types of aggression, we found that in both sexes, the resident mouse delivered more and received less agonistic behaviour than the same-sex intruder. This occurred in males through attacks and in females through other agonistic behaviours. When looking at
116 the total amount of time allocated to agonistic behaviours, female CD1 mice were not different from males in their response to a gonadex, same-sex intruder. Hence, the overall levels of intrasexual agonistic behaviour are similar in male and female CD1 mice. However, the functional outcome of these interactions is likely to be different. In males, fighting may allow for the establishment of territories and regulation of space use (Scott & Fredericson,
1951), as male attacks tend to be in short bursts followed by periods when they retreat to the opposite sides of the cage (Ginsburg & Allee, 1942; Miczek et al., 2001). In females, the extended sessions of dominant agonistic behaviour performed by the resident could lead to the establishment of a dominance hierarchy which, in turn, would allow them to share the same territory with other females.
In gonadally intact CD1 mice, we found males showed more behavioural shifting and a shorter latency to social investigation than females. This may reflect greater general arousal induced by the social challenge represented by the presence of a same-sex intruder in gonadally intact, territorial males (Branchi et al., 2006; Pfaff et al., 2008), as castrated males, which are less territorial, also display a longer latency to social investigation and approach the intruder less than gonadally intact males. Thus, the observed sex differences in the type of agonistic behaviours displayed may be due to different underlying motivational responses to an unknown intruder in males and females.
In agreement with previous findings that castration affects a number of behaviours in many strains of mice (e.g., Beeman, 1947; de Catanzaro, 1987; Luttge, 1972; Uhrich, 1938), in the uninjected CD1 mice in the present study, some differences were noted between the gonadally intact and gonadex males. Castrated CD1 males exhibited reduced agonistic
117 behaviour when compared to intact males, not only by showing the well-established absence of attacks (e.g., Beeman, 1947; Luttge, 1972; Uhrich, 1938), but also by displaying an increase in submissive behaviour, similar to that seen in the gonadally intact CD1 females.
Ovariectomized CD1 females, overall, appeared to be less afraid of the same-sex intruder than gonadally intact females, as indicated by fewer stretched approaches (Blanchard et al.,
1993; Choleris et al., 2001) and more investigations directed towards the facial area. This is consistent with the well documented involvement of estrogens in the mediation of anxiety and anxiety-like behaviours in rodents and other species (reviewed in Choleris et al., 2008a).
3.5.3 Note on Vehicle Effects
When compared to the uninjected mice used in this study, the mice that had received sesame oil (a common vehicle for gonadal hormones) showed some behavioural differences.
In intact CD1 males, the sesame oil vehicle increased Aggressive Postures and decreased the amount of Received Agonistic Behaviour, while in intact CD1 females, oil increased social behaviour, and decreased some non-social active behaviours. Gonadex CD1 mice showed a number of vehicle effects as well, with sesame oil decreasing activity. Sesame oil also decreased social investigation in castrated males, while increasing non-social behaviours in general. This is in general agreement with studies showing that i.p. injection of sesame oil has been shown to produce partner preferences in prairie voles (Curtis & Wang, 2005) and oral ingestion of sesame oil by CD1 mice reduced norepinephrine levels (Anton et al., 1974).
Thus, sesame oil may affect social behaviour by increasing overall arousal (Pfaff et al.,
2008). Alternatively, it is possible that the effects found in the present study are due to metabolic effects of the sesame oil, which have been observed with s.c. injections in rats
118 (Colafranceschi et al., 2007), although why they would be seen 72h after administration is unclear. For the same reason, it is unlikely that these effects are due to the stress of the injection. An intriguing speculative hypothesis is that sesame oil may have as yet undescribed light hormonal effects.
Most research on the effects of gonadal hormones does not include an uninjected control group, and thus there may be vehicle effects that are overlooked, especially when a vehicle other than physiological saline is used. In the few studies in which an uninjected control group was used, or the effects of sesame oil were directly examined, vehicle or sesame oil effects were found in various species and with different modes of administration (oral, i.p., s.c.; Anton et al., 1974; Colafranceschi et al., 2007; Curtis & Wang, 2005). Given the implications this may have for endocrinological studies, further investigations into these effects are warranted.
3.5.4 Conclusions
The ethological analysis employed allowed a complete assessment of the role of ERβ in agonistic interactions in male and female CD1 mice and allowed us to establish not only generalized effects of the ERβ agonist WAY2000-70 on social and non-social behaviours, but also and more importantly, drug effects on specific aspects of the mice’s social responses in the intruder test. Our results suggest that in both sexes, acute stimulation of ERβ 72 h before testing can mediate dominance-related agonistic behaviours, while leaving overt attacks largely unaffected. These results of acute ERβ agonist treatment in normally developed adult mice suggest that at least part of the effects of gene knockout of ERβ on aggression (Nomura et al., 2002; Ogawa et al., 1999) may be due to an activational, rather
119 than a developmental, role of ERβ and further emphasize the need for the ethological assessments of behaviour in order to achieve a full understanding of the results of neurobiological studies.
Our ethological assessment of female interactions showed that CD1 females spend as much time performing agonistic behaviours as males do, but use behaviours other than attacks in what is likely their attempt to establish dominance over a same-sex conspecific.
We also showed that, similar to males, these agonistic behaviours are affected by acute activation of ERβ. However, more studies with females are needed in order to elucidate the respective roles of ERα and ERβ in intrasexual aggression and agonistic behaviours, as previous research with αERKO and βERKO females tended to focus on attacks and/or maternal aggression, and potential effects of the genetic manipulation on other agonistic behaviours would not have been observed (Ogawa et al., 1999; Ogawa et al., 2004).
Additionally, further research with different strains and with intruders of the opposite sex are needed to determine if these results generalize to all strains of mice and intruders. By gaining a more complete picture of the behaviour of male and female mice in the resident-intruder paradigm, we would be better able to understand this important aspect of behaviour, as well as the roles played by ERα and ERβ in functionally different types of aggression.
120 3.6. Appendix 2: Differences in behaviour between sexes and gonadal conditions, and effects of WAY-200070 treatment on behaviour.
Table 17. Effects of WAY-200070 in Gonadally Intact Males Behaviour Effect DUR1: Increased by HIGHa at 10-15 min (U = 32.50, z = -2.48, p = 0.013) Total Activity DUR: Decreased by MIDb at 5-10 min (U = 26.00, z = -2.42, p = 0.016) Total Social DUR: Effect of treatment at 5-10 min (χ2(4) = 11.02, p = 0.026) Behaviours DUR: Decreased by MID at 5-10 min (U = 32.00, z = -2.05, p = 0.041) Aggressive DUR: Increased by OILc (U = 25.00, z = -2.27, p = 0.023) Postures FREQ: Increased by OIL (U = 25.50, z = -2.26, p = 0.024) Chasing the FREQ: Decreased by MID at 10-15 min (U = 29.00, z = -2.38, p = 0.017) Intruder Dominant DUR: Increased by HIGH (U = 42.00 z = -1.96, p = 0.050) Behaviour LAT3: Decreased by HIGH (U = 17.00, z = -3.32, p = 0.001) DUR: Effect of treatment at 0-5 min (χ2(4) = 10.85, p = 0.028) Agonistic DUR: Increased by MID at 0-5 min (U = 29.00, z = -2.24, p = 0.025) Behaviour DUR: Decreased by OIL at 0-5 min (U = 26.00, z = -2.17, p = 0.030) Received FREQ: Decreased by OIL (U = 28.00, z = -2.04, p = 0.041) FREQ: Decreased by OIL at 0-5 min (U = 24.50, z = -2.27, p = 0.023) DUR: Increased by MID at 0-5 min (U = 23.50, z = -2.61, p = 0.009) Submissive DUR: Decreased by LOWd at 10-15 min (U = 45.00, z = -2.25, p = 0.024) Behaviour FREQ: Increased by MID at 0-5 min (U = 17.50, z = -3.02, p = 0.002) LAT: Decreased by MIDc (U = 37.00, z = -2.00, p = 0.045) Attacks FREQ: Decreased by LOW at 10-15 min (U = 39.00, z = -3.01, p = 0.003) Received LAT: Increased by HIGH (U = 35.00, z = -2.36, p = 0.018) Oronasal DUR: Decreased by HIGH (U = 37.00, z = -2.23, p = 0.026) Investigation DUR: Decreased by HIGH at 5-10 min (U = 41.00, z = -2.01, p = 0.044) Body FREQ2: Effect of treatment at 5-10 min (χ2(4) = 10.54, p = 0.032) Investigation FREQ: Increased by HIGH at 10-15 min (U = 35.50, z = -2.32, p = 0.021) FREQ: Effect of treatment at 10-15 min (χ2(4) = 10.07, p = 0.039) Stretched FREQ: Decreased by HIGH (U = 39.00, z = -2.15, p = 0.032) Approaches FREQ: Decreased by HIGH at 10-15 min (U = 27.50, z = -2.97, p = 0.003) DUR: Increased by HIGH at 10-15 min (U = 39.50, z = -2.10, p = 0.036) Digging FREQ: Increased by HIGH (U = 36.50, z = -2.27, p = 0.024) FREQ: Increased by HIGH at 10-15 min (U = 36.50, z = -2.27, p = 0.023) DUR: Effect of treatment (χ2(4) = 11.02, p = 0.026) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.72, p = 0.030) DUR: Effect of treatment at 10-15 min (χ2(4) = 10.42, p = 0.034) Non-social DUR: Increased by MID at 5-10 min (U = 29.00, z = -2.23, p = 0.026) Non-locomotor DUR: Decreased by HIGH (U = 36.00, z = -2.29, p = 0.022) Behaviours DUR: Decreased by HIGH at 10-15 min (U = 27.00, z = -2.78, p = 0.006) FREQ: Increased by MID at 5-10 min (U = 21.50, z = -2.71, p = 0.007) FREQ: Increased by HIGH at 10-15 min (U = 35.50, z = -2.32, p = 0.020) 121 DUR: Effect of treatment (χ2(4) = 11.83, p = 0.019) DUR: Effect of treatment at 10-15 min (χ2(4) = 14.57, p = 0.006) DUR: Decreased by HIGH (U = 29.00, z = -2.67, p = 0.008) Self-Grooming DUR: Decreased by LOW at 10-15 min (U = 37.00, z = -2.44, p = 0.015) DUR: Decreased by HIGH at 10-15 min (U = 17.00, z = -3.32, p = 0.001) FREQ: Decreased by HIGH at 10-15 min (U = 33.50, z = -2.43, p = 0.015) a “HIGH” indicates 180 mg/kg WAY-200070. b “MID” indicates 90 mg/kg WAY-200070. c “OIL” indicates sesame oil vehicle. d “LOW” indicates 30 mg/kg WAY-200070. 1DUR represents duration. 2FREQ represents frequency. 3LAT represents latency.
122 Table 18. Effects of WAY-200070 in Gonadally Intact Females Behaviour Effect DUR1: Effect of treatment (χ2(4) = 11.72, p = 0.020) DUR: Effect of treatment at 10-15 min (χ2(4) = 13.83, p = 0.008) DUR: Increased by HIGHa (U = 32.00, z = -2.09, p = 0.036) DUR: Increased by HIGH at 10-15 min (U = 34.00, z = -1.97, p = 0.049) Total Activity FREQ2: Effect of treatment (χ2(4) = 23.48, p < 0.001) and all timepoints FREQ: Decreased by OILb (U = 26.00, z = -2.04, p = 0.041) FREQ: Increased by HIGH (U = 10.00, z = -3.45, p = 0.001) and all timepoints DUR: Effect of treatment (χ2(4) = 13.95, p = 0.007) DUR: Effect of treatment at 5-10 min (χ2(4) = 11.97, p = 0.018) DUR: Effect of treatment at 10-15 min (χ2(4) = 13.12, p = 0.011) Total Social DUR: Increased by OIL (U = 16.00, z = -2.75, p = 0.006) Behaviours DUR: Increased by OIL at 10-15 min (U = 13.00, z = -2.96, p = 0.003) FREQ: Effect of treatment (χ2(4) = 24.72, p < 0.001) and all timepoints FREQ: Increased by HIGH (U = 8.00, z = -3.57, p < 0.001) and all timepoints DUR: Effect of treatment at 5-10 min (χ2(4) = 11.98, p = 0.018) Social DUR: Increased by HIGH at 5-10 min (U = 37.00, z = -2.19, p = 0.029) Inactivity FREQ: Effect of treatment at 5-10 min (χ2(4) = 11.64, p = 0.020) FREQ: Increased by HIGH at 5-10 min (U = 38.00, z = -2.12, p = 0.034) DUR: Increased by OIL at 10-15 min (U = 25.00, z = -2.11, p = 0.035) DUR: Decreased by HIGH at 0-5 min (U = 32.00, z = -2.05, p = 0.041) Total FREQ: Effect of treatment (χ2(4) = 11.31, p = 0.023) Agonistic FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.47, p= 0.033) Behaviours FREQ: Effect of treatment at 10-15 min (χ2(4) = 9.80, p = 0.044) LAT: Increased by LOW (U = 28.00, z = -2.34, p = 0.019) FREQ: Effect of treatment (χ2(4) = 11.24, p = 0.024) Agonistic FREQ: Increased by HIGH (U = 28.50, z = -2.31, p = 0.021) Behaviour FREQ: Increased by HIGH at 0-5 min (U = 28.50, z = -2.31, p = 0.021) Delivered LAT: Increased by MIDd (U = 25.00, z = -2.33, p = 0.020) DUR: Effect of treatment (χ2(4) = 11.10, p = 0.025) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.86, p = 0.028) DUR: Effect of treatment at 10-15 min (χ2(4) = 11.00, p = 0.027) DUR: Increased by HIGH (U = 33.00, z = -2.03, p = 0.042) DUR: Increased by HIGH at 5-10 min (U = 32.50, z = -2.06, p = 0.039) Chasing the FREQ: Effect of treatment at 0-5 min (χ2(4) = 9.92, p = 0.042) Intruder FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.81, p = 0.029) FREQ: Effect of treatment at 10-15 min (χ2(4) = 13.89, p = 0.008) FREQ: Increased by HIGH at 0-5 min (U = 33.50, z = -2.00, p = 0.045) FREQ: Increased by HIGH at 5-10 min (U = 33.00, z = -2.04, p = 0.041) FREQ: Increased by HIGH at 10-15 min (U = 22.50, z = -2.70, p = 0.007) Attacks LAT: Decreased by MID (U = 26.50, z = -2.44, p = 0.015) Delivered
123 DUR: Effect of treatment (χ2(4) = 11.08, p = 0.026) DUR: Effect of treatment at 0-5 min (χ2(4) = 11.38, p = 0.023) Agonistic DUR: Increased by OIL (U = 22.00, z = -2.32, p = 0.020) Behaviour DUR: Increased by OIL at 0-5 min (U = 27.00, z = -1.97, p = 0.049) Received FREQ: Effect of treatment (χ2(4) = 9.66, p = 0.047) FREQ: Increased by OIL (U = 21.50, z = -2.36, p = 0.018) FREQ: Increased by OIL at 0-5 min (U = 26.00, z = -2.05, p = 0.041) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.87, p = 0.028) DUR: Increased by OIL (U = 27.50, z = -1.97, p = 0.048) Avoidance of DUR: Increased by OIL at 0-5 min (U = 28.00, z = -1.98, p = 0.048) the Intruder DUR: Increased by OIL at 5-10 min (U = 30.00, z = -2.36, p = 0.018) FREQ: Increased by OIL at 5-10 min (U = 30.00, z = -2.36, p = 0.018) DUR: Increased by OIL (U = 25.00, z = -2.11, p = 0.035) DUR: Decreased by LOW (U = 29.50, z = -2.25, p = 0.025) Submissive DUR: Decreased by LOW at 5-10 min (U = 32.00, z = -2.10, p = 0.036) Behaviour FREQ: Increased by OIL at 10-15 min (U = 25.50, z = -2.12, p = 0.034) FREQ: Decreased by LOW at 5-10 min (U = 33.00, z = -2.04, p = 0.041) Defensive DUR: Effect of treatment at 5-10 min (χ2(4) = 10.06, p = 0.039) Upright DUR: Increased by OIL at 0-5 min (35.00 -2.05 0.040 Posturing FREQ: Increased by OIL at 0-5 min (U = 35.00, z = -2.05, p = 0.040) FREQ: Effect of treatment at 0-5 min (χ2(4) = 10.09, p = 0.039) Dominance FREQ: Increased by HIGH (U = 33.50, z = -2.00, p = 0.045) Score FREQ: Increased by HIGH at 0-5 min (U = 30.00, z = -2.22, p = 0.027) Social FREQ: Effect of treatment (χ2(4) = 20.54, p < 0.001) and all timepoints Investigation FREQ: Increased by HIGH (U = 6.00, z = -3.69, p < 0.001) and all timepoints DUR: Increased by HIGH at 0-5 min (U = 29.00, z = -2.23, p = 0.026) Oronasal FREQ: Increased by LOWc at 5-10 min (U = 31.00, z = -2.17, p = 0.030) Investigation LAT3: Decreased by OIL (U = 33.00, z = -2.03, p = 0.042) FREQ: Effect of treatment (χ2(4) = 13.22, p = 0.010) Body FREQ: Effect of treatment at 0-5 min (χ2(4) = 10.87, p = 0.028) Investigation FREQ: Effect of treatment at 10-15 min (χ2(4) = 12.61, p = 0.013) FREQ: Increased by HIGH (U = 17.00, z = -3.02, p = 0.003) DUR: Increased by HIGH at 0-5 min (U = 34.00, z = -1.97, p = 0.049) FREQ: Effect of treatment (χ2(4) = 20.43, p < 0.001) FREQ: Effect of treatment at 0-5 min (χ2(4) = 14.69, p = 0.005) Anogenital FREQ: Effect of treatment at 5-10 min (χ2(4) = 16.40, p = 0.003) Investigation FREQ: Increased by HIGH (U = 19.50, z = -2.86, p = 0.004) FREQ: Increased by HIGH at 0-5 min (U = 20.00, z = -2.84, p = 0.005) FREQ: Increased by HIGH at 5-10 min (U = 23.00, z = -2.65, p = 0.008) Stretched FREQ: Decreased by HIGH at 5-10 min (U = 35.00, z = -2.18, p = 0.030) Approaches DUR: Effect of treatment (χ2(4) = 12.57, p = 0.014) Total Non- DUR: Effect of treatment at 5-10 min (χ2(4) = 11.92, p = 0.018) social DUR: Decreased by OIL (U = 25.00, z = -2.11, p = 0.035) Behaviours DUR: Decreased by OIL at 10-15 min (U = 21.00, z = -2.39, p = 0.017) FREQ: Decreased by OIL (U = 27.00, z = -1.97, p = 0.049) 124 DUR: Effect of treatment at 10-15 min (χ2(4) = 10.23, p = 0.037) DUR: Decreased by OIL at 10-15 min (U = 19.50, z = -2.50, p = 0.012) Non-social FREQ: Effect of treatment at 10-15 min (χ2(4) = 11.56, p = 0.021) Locomotor FREQ: Decreased by OIL (U = 25.50, z = -2.08, p = 0.038) Behaviours FREQ: Decreased by OIL at 10-15 min (U = 24.00, z = -2.18, p = 0.029) FREQ: Increased by HIGH at 10-15 min (U = 31.50, z = -2.13, p = 0.033) DUR: Decreased by OIL at 10-15 min (U = 27.00, z = -1.97, p = 0.049) FREQ: Effect of treatment at 10-15 min (χ2(4) = 13.53, p = 0.009) Horizontal FREQ: Decreased by OIL (U = 24.00, z = -2.18, p = 0.029) Exploration FREQ: Decreased by OIL at 10-15 min (U = 21.50, z = -2.37, p = 0.018) FREQ: Increased by HIGH at 10-15 min (U = 28.50, z = -2.31, p = 0.021) Vertical DUR: Decreased by OIL at 10-15 min (U = 27.00, z = -1.97, p = 0.049) Exploration DUR: Decreased by LOW at 0-5 min (U = 30.50, z = -2.15, p = 0.032) DUR: Effect of treatment at 10-15 min (χ2(4) = 10.29, p = 0.036) DUR: Decreased by OIL (U = 23.00, z = -2.35, p = 0.019) DUR: Decreased by OIL at 10-15 min (U = 21.00, z = -2.59, p = 0.010) DUR: Increased by HIGH at 10-15 min (U = 30.50, z = -2.36, p = 0.018) FREQ: Effect of treatment at 10-15 min (χ2(4) = 10.59, p= 0.032) FREQ: Decreased by OIL (U = 24.50, z = -2.25, p = 0.024) Digging FREQ: Decreased by OIL at 10-15 min (U = 21.00, z = -2.59, p = 0.010) FREQ: Increased by HIGH (U = 34.00, z = -2.08, p = 0.038) FREQ: Increased by HIGH at 10-15 min (U = 27.00, z = -2.60, p = 0.009) LAT: Increased by OIL (U = 24.00, p = -2.70, p = 0.007) LAT: Decreased by LOW (U = 34.00, z = -2.13, p = 0.034) LAT: Decreased by HIGH (U = 28.00, z = -2.48, p = 0.013) DUR: Effect of treatment (χ2(4) = 20.67, p < 0.001) DUR: Decreased by HIGH (U = 18.00, z = -2.96, p = 0.003) DUR: Decreased by HIGH at 5-10 min (U = 26.00, z = -2.46, p = 0.014) Non-social DUR: Decreased by HIGH at 10-15 min (U = 19.00, z = -2.89, p = 0.004) Non-locomotor FREQ: Effect of treatment (χ2(4) = 15.88, p = 0.003) Behaviours FREQ: Effect of treatment at 10-15 min (χ2(4) = 19.36, p = 0.001) FREQ: Increased by LOW at 10-15 min (U = 32.50, z = -2.07, p = 0.038) FREQ: Decreased by HIGH (U = 30.00, z = -2.22, p = 0.026) FREQ: Decreased by HIGH at 10-15 min (U = 24.50, z = -2.57, p = 0.010) DUR: Effect of treatment (χ2(4) = 19.57, p = 0.001) DUR: Decreased by HIGH (U = 15.00, z = -3.14, p = 0.002) DUR: Decreased by HIGH at 5-10 min (U = 23.00, z = -2.65, p = 0.008) DUR: Decreased by HIGH at 10-15 min (U = 17.00, z = -3.02, p = 0.003) Self-Grooming FREQ: Effect of treatment (χ2(4) = 14.38, p = 0.006) FREQ: Effect of treatment at 0-5 min (χ2(4) = 9.58, p = 0.048) FREQ: Effect of treatment at 10-15 min (χ2(4) = 16.70, p = 0.002) FREQ: Decreased by HIGH (U = 32.50, z = -2.07, p = 0.039) FREQ: Decreased by HIGH at 10-15 min (U = 27.50, z = -2.38, p = 0.017) a ““HIGH” indicates 180 mg/kg WAY-200070. b “OIL” indicates sesame oil vehicle. c “LOW” indicates 30 mg/kg WAY-200070. d “MID” indicates 90 mg/kg WAY-200070. 1DUR represents duration. 2FREQ represents frequency. 3LAT represents latency.
125 Table 19. Effects of WAY-200070 in Castrated Males Behaviour Effect DUR1: Effect of treatment (χ2(4) = 12.95, p = 0.012) DUR: Effect of treatment at 5-10 min (χ2(4) = 13.30, p = 0.010) DUR: Effect of treatment at 10-15 min (χ2(4) = 10.58, p = 0.032) DUR: Decreased by OILa (U = 21.50, z = -2.54, p = 0.011) Total Activity DUR: Decreased by OIL at 5-10 min (U = 10.00, z = -3.30, p = 0.001) DUR: Decreased by OIL at 10-15 min (U = 24.00, z = -2.37, p = 0.018) FREQ: Effect of treatment (χ2(4) = 14.80, p = 0.005) FREQ: Effect of treatment at 10-15 min (χ2(4) = 20.45, p < 0.001) FREQ2: Increased by HIGHb (U = 8.00, z = -2.84, p = 0.004) DUR: Effect of treatment (χ2(4) = 10.94, p = 0.027) DUR: Effect of treatment at 5-10 min (χ2(4) = 12.51, p = 0.014) DUR: Decreased by OIL (U = 24.00, z = -2.37, p = 0.018) DUR: Decreased by OIL at 5-10 min (U = 21.50, z = -2.54, p = 0.011) Total Social DUR: Decreased by OIL at 10-15 min (U = 30.00, z = -1.98, p = 0.048) Behaviours FREQ: Effect of treatment (χ2(4) = 24.72, p < 0.001) FREQ: Effect of treatment at 5-10 min (χ2(4) = 20.18, p > 0.001) FREQ: Decreased by OIL (U = 26.00, z = -2.24, p = 0.025) FREQ: Decreased by OIL at 5-10 min (U = 17.00, z = -2.84, p = 0.005) LAT3: Decreased by HIGH (U = 20.00, z = -2.27, p = 0.023) DUR: Effect of treatment (χ2(4) = 14.94, p = 0.005) DUR: Effect of treatment at 5-10 min (χ2(4) = 19.18, p = 0.001) DUR: Increased by HIGH (U = 13.00, z = -2.73, p = 0.006) DUR: Increased by HIGH at 5-10 min (U = 18.50, z = -2.28, p = 0.023) DUR: Increased by HIGH at 10-15 min (U = 25.00, z = -2.05, p = 0.040) Social FREQ: Effect of treatment (χ2(4) = 16.18, p = 0.003) Inactivity FREQ: Effect of treatment at 5-10 min (χ2(4) = 20.07, p > 0.001) FREQ: Increased by HIGH (U = 12.00, z = -2.84, p = 0.005) FREQ: Increased by HIGH at 5-10 min (U = 17.50, z = -2.39, p = 0.017) FREQ: Increased by HIGH at 10-15 min (U = 25.00, z = -2.05, p = 0.040) LAT: Effect of treatment (χ2(4) = 10.07, p = 0.039) LAT: Decreased by HIGH (U = 24.00, z = -2.31, p = 0.021) DUR: Effect of treatment at 10-15 min (χ2(4) = 12.16, p = 0.016) DUR: Decreased by OIL at 10-15 min (U = 29.00, z = -2.04, p = 0.041) Total Agonistic DUR: Decreased by LOW at 0-5 min (U = 32.00, z = -2.05, p = 0.041) Behaviours DUR: Increased by MID at 10-15 min (U = 28.00, z = -2.11, p = 0.035) FREQ: Effect of treatment at 5-10 min (χ2(4) = 11.34, p = 0.023) Aggressive DUR: Effect of treatment at 5-10 min (χ2(4) = 11.97, p = 0.018) Postures FREQ: Effect of treatment at 5-10 min (χ2(4) = 11.97, p = 0.018) Agonistic Behaviour FREQ: Effect of treatment at 5-10 min (χ2(4) = 11.07, p = 0.026) Delivered
126 DUR: Effect of treatment (χ2(4) = 17.96, p = 0.001) DUR: Effect of treatment at 0-5 min (χ2(4) = 15.77, p = 0.003) DUR: Effect of treatment at 5-10 min (χ2(4) = 17.98, p = 0.001) DUR: Decreased by OIL (U = 26.00, z = -2.24, p = 0.025) DUR: Decreased by OIL at 0-5 min (U = 30.00, z = -1.98, p = 0.048) Chasing the DUR: Increased by HIGH (U = 14.00, z = -2.31, p = 0.021) Intruder DUR: Increased by HIGH at 0-5 min (U = 13.00, z = -2.40, p = 0.016) DUR: Increased by HIGH at 5-10 min (U = 17.50, z = -2.00, p = 0.045) FREQ: Effect of treatment at 0-5 min (χ2(4) = 9.90, p = 0.042) FREQ: Effect of treatment at 5-10 min (χ2(4) = 11.14, p = 0.025) FREQ: Increased by HIGH at 0-5 min (U = 17.00, z = -2.05, p = 0.040) Dominant DUR: Increased by OIL at 0-5 min (U = 28.00, z = -2.11, p = 0.035) Behaviour DUR: Effect of treatment (χ2(4) = 12.80, p = 0.012) DUR: Effect of treatment at 10-15 min (χ2(4) = 13.09, p = 0.011) DUR: Decreased by HIGH (U = 16.00, z = -2.13, p = 0.033) DUR: Decreaesd by HIGH at 10-15 min (U = 11.00, z = -2.58, p = 0.010) FREQ: Effect of treatment (χ2(4) = 19.37, p = 0.001) FREQ: Effect of treatment at 0-5 min (χ2(4) = 14.02, p = 0.007) Social FREQ: Effect of treatment at 5-10 min (χ2(4) = 20.20, p < 0.001) Investigation FREQ: Decreased by OIL (U = 9.00, z = -3.36, p = 0.001) FREQ: Increased by LOWc (U = 33.00, z = -1.99, p = 0.047) FREQ: Increased by LOW at 0-5 min (U = 33.00, z = -1.99, p = 0.047) FREQ: Increased by HIGH (U = 6.00, z = -3.02, p = 0.003) FREQ: Increased by HIGH at 0-5 min (U = 9.50, z = -2.71, p = 0.007) FREQ: Increased by HIGH at 5-10 min (U = 8.00, z = -2.85, p = 0.004) LAT: Decreased by HIGH (U = 19.00, z = -2.35, p = 0.019) DUR: Effect of treatment (χ2(4) = 17.43, p = 0.002) and all timepoints DUR: Decreased by LOW at 0-5 min (U = 29.00, z = -2.23, p = 0.026) DUR: Decreased by MIDd at 0-5 min (U = 23.00, z = -2.44, p = 0.015) Oronasal DUR: Decreased by HIGH (U = 1.00, z = -3.47, p = 0.001) and all Investigation timepoints FREQ: Decreased by HIGH at 0-5 min (U = 11.50, z = -2.55, p = 0.014) FREQ: Decreased by HIGH at 10-15 min (U = 17.50, z = -2.01, p = 0.045) FREQ: Effect of treatment (χ2(4) = 15.21, p = 0.004) FREQ: Effect of treatment at 5-10 min (χ2(4) = 14.08, p = 0.007) FREQ: Decreased by OIL (U = 14.50, z = -3.00, p = 0.003) and all timepoints Body FREQ: Increased by LOW (U = 25.50, z = -2.45, p = 0.14) Investigation FREQ: Increased by LOW at 0-5 min (U = 29.00, z = -2.23, p = 0.025) FREQ: Increased by LOW at 5-10 min (U = 30.50, z = -2.14, p = 0.032) FREQ: Increased by HIGH (U = 14.50, z = -2.70, p = 0.023) FREQ: Increased by HIGH at 0-5 min (U = 14.00, z = -2.32, p = 0.021) FREQ: Increased by HIGH at 5-10 min (U = 15.50, z = -2.18, p = 0.029)
127 DUR: Effect of treatment at 5-10 min (χ2(4) = 12.05, p = 0.017) DUR: Decreased by OIL (U = 21.00, z = -2.57, p = 0.010) DUR: Decreased by OIL at 0-5 min (U = 22.50, z = -2.47, p = 0.013) DUR: Decreased by OIL at 5-10 min (U = 10.00, z = -3.30, p = 0.001) DUR: Decreased by MID at 10-15 min (U = 28.50, z = -2.08, p = 0.038) Anogenital FREQ: Effect of treatment (χ2(4) = 17.45, p = 0.002) and all timepoints Investigation FREQ: Decreased by OIL (U = 12.50, z = -3.13, p = 0.002) and all timepoints FREQ: Increased by HIGH (U = 14.50, z = -2.27, p = 0.023) FREQ: Increased by HIGH at 0-5 min (U = 17.00, z = -2.05, p = 0.041) FREQ: Increased by HIGH at 5-10 min (U = 12.50, z = -2.45, p = 0.014) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.11, p = 0.039) FREQ: Effect of treatment (χ2(4) = 12.21, p = 0.016) FREQ: Effect of treatment at 0-5 min (χ2(4) = 9.52, p = 0.049) Approaching FREQ: Effect of treatment at 5-10 min (χ2(4) = 16.35, p = 0.003) and/or FREQ: Increased by HIGH (U = 15.00, z = -2.23, p = 0.026) Attending to FREQ: Increased by HIGH at 0-5 min (U = 15.00, z = -2.23, p = 0.026) the Intruder FREQ: Increased by HIGH at 5-10 min (U = 16.50, z = -2.10, p = 0.036) LAT: Effect of treatment (χ2(4) = 13.16, p = 0.011) LAT: Decreased by HIGH (U = 19.00, z = -2.35, p = 0.019) DUR: Effect of treatment (χ2(4) = 14.80, p = 0.005) DUR: Effect of treatment at 5-10 min (χ2(4) = 17.50, p = 0.002) DUR: Effect of treatment at 10-15 min (χ2(4) = 13.70, p = 0.008) DUR: Increased by OIL (U = 29.00, z = -2.04, p = 0.041) DUR: Increased by OIL at 5-10 min (U = 24.00, z = -2.37, p = 0.018) Total Non- DUR: Increased by OIL at 10-15 min (U = 25.00, z = -2.31, p = 0.021) Social DUR: Increased by HIGH at 5-10 min (U = 17.00, z = -2.04, p = 0.041) Behaviours FREQ: Effect of treatment (χ2(4) = 12.14, p = 0.016) FREQ: Effect of treatment at 0-5 min (χ2(4) = 11.70, p = 0.020) FREQ: Effect of treatment at 5-10 min (χ2(4) = 13.15, p = 0.011) FREQ: Increased by LOW at 0-5 min (U = 32.00, z = -2.05, p = 0.040) FREQ: Increased by HIGH (U = 13.00, z = -2.40, p = 0.016) and all timepoints DUR: Effect of treatment (χ2(4) = 12.88, p = 0.012) DUR: Effect of treatment at 5-10 min (χ2(4) = 14.36, p = 0.006) DUR: Effect of treatment at 10-15 min (χ2(4) = 11.74, p = 0.019) DUR: Increased by HIGH (U = 12.00, z = -2.49, p = 0.013) DUR: Increased by HIGH at 0-5 min (U = 15.00, z = -2.22, p = 0.026) DUR: Increased by HIGH at 5-10 min (U = 15.00, z = -2.22, p = 0.026) Non-social FREQ: Effect of treatment (χ2(4) = 12.42, p = 0.014) Locomotor FREQ: Effect of treatment at 0-5 min (χ2(4) = 12.78, p = 0.012) Behaviours FREQ: Effect of treatment at 5-10 min (χ2(4) = 13.00, p = 0.011) FREQ: Increased by LOW at 0-5 min (U = 30.00, z = -2.17, p = 0.030) FREQ: Increased by HIGH (U = 13.50, z = -2.36, p = 0.018) and all timepoints FREQ: Increased by HIGH at 0-5 min (U = 12.50, z = -2.45, p = 0.014) FREQ: Increased by HIGH at 5-10 min (U = 12.50, z = -2.45, p = 0.014)
128 DUR: Effect of treatment (χ2(4) = 11.21, p = 0.024) DUR: Effect of treatment at 5-10 min (χ2(4) = 15.39, p = 0.004) DUR: Increased by OIL at 5-10 min (U = 27.00, z = -2.18, p = 0.030) DUR: Increased by LOW at 0-5 min (U = 30.50, z = -2.15, p = 0.032) DUR: Increased by HIGH (U = 15.00, z = -2.22, p = 0.026) DUR: Increased by HIGH at 0-5 min (U = 18.00, z = -1.98, p = 0.048) Vertical DUR: Increased by HIGH at 5-10 min (U = 15.00, z = -2.22, p = 0.026) Exploration FREQ: Effect of treatment (χ2(4) = 11.63, p = 0.020) FREQ: Effect of treatment at 0-5 min (χ2(4) = 13.05, p = 0.011) FREQ: Effect of treatment at 5-10 min (χ2(4) = 12.41, p = 0.015) FREQ: Increased by LOW at 0-5 min (U = 26.50, z = -2.41, p = 0.016) FREQ: Increased by HIGH (U = 17.50, z = -2.00, p = 0.045) and all timepoints DUR: Effect of treatment (χ2(4) = 11.36, p = 0.023) DUR: Effect of treatment at 5-10 min (χ2(4) = 12.27, p = 0.015) DUR: Increased by HIGH at 5-10 min (U = 17.00, z = -2.04, p = 0.041) Horizontal FREQ: Effect of treatment (χ2(4) = 13.30, p = 0.010) Exploration FREQ: Effect of treatment at 0-5 min (χ2(4) = 11.47, p = 0.022) FREQ: Effect of treatment at 5-10 min (χ2(4) = 13.09, p = 0.011) FREQ: Increased by LOW at 0-5 min (U = 30.00, z = -2.18, p = 0.029) FREQ: Increased by HIGH (U = 9.00, z = -2.76, p = 0.006) DUR: Effect of treatment (χ2(4) = 9.67, p = 0.046) DUR: Increased by OIL at 10-15 min (U = 31.00, z = -2.09, p = 0.037) Digging FREQ: Effect of treatment at 10-15 min (χ2(4) = 9.58, p = 0.048) FREQ: Increased by OIL at 10-15 min (U = 32.00, z = -2.02, p = 0.043) DUR: Effect of treatment at 5-10 min (χ2(4) = 11.49, p = 0.022) DUR: Increased by OIL (U = 20.50, z = -2.61, p = 0.009) DUR: Increased by OIL at 5-10 min (U = 10.00, z = -3.30, p = 0.001) Non-social DUR: Increased by OIL at 10-15 min (U = 24.00, z = -2.37, p = 0.018) Non-locomotor FREQ: Effect of treatment (χ2(4) = 9.89, p = 0.042) Behaviours FREQ: Effect of treatment at 5-10 min (χ2(4) = 12.66, p = 0.013) FREQ: Increased by OIL at 5-10 min (U = 20.50 z = -2.64, p = 0.008) LAT: Decreased by OIL (U = 30.00, z = -1.98, p = 0.048) DUR: Effect of treatment at 5-10 min (χ2(4) = 10.50, p = 0.033) Solitary DUR: Increased by OIL at 5-10 min (U = 20.50, z = -2.67, p = 0.008) Inactivity FREQ: Increased by OIL (U = 28.00, z = -2.12, p = 0.034) FREQ: Increased by OIL at 5-10 min (U = 22.50, z = -2.57, p = 0.010) DUR: Increased by OIL (U = 23.00, z = -2.44, p = 0.015) Self-Grooming DUR: Increased by OIL at 10-15 min (U = 28.00, z = -2.11, p = 0.035) FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.42, p = 0.034) a “OIL” indicates sesame oil vehicle. b “HIGH” indicates 180 mg/kg WAY-200070. c “LOW” indicates 30 mg/kg WAY-200070. d “MID” indicates 90 mg/kg WAY-200070. 1DUR represents duration. 2FREQ represents frequency. 3LAT represents latency.
129 Table 20. Effects of WAY-200070 in Ovariectomized Females Behaviour Effect FREQ1: Effect of treatment (χ2(4) = 11.14, p = 0.025) FREQ: Effect of treatment at 0-5 min (χ2(4) = 14.22, p = 0.007) FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.09, p = 0.039) Total FREQ: Decreased by OILa (U = 16.00, z = -2.13, p = 0.033) Activity FREQ: Decreased by OIL at 5-10 min (U = 14.50, z = -2.27, p = 0.023) FREQ: Increased by LOWb at 0-5 min (U = 13.50, z = -2.36, p = 0.018) FREQ: Increased by HIGHc (U = 12.00, z = -2.64, p = 0.008) and all timepoints FREQ: Effect of treatment (χ2(4) = 11.24, p = 0.024) Total Social FREQ: Effect of treatment at 0-5 min (χ2(4) = 9.61, p = 0.048) Behaviours FREQ: Increased by HIGH (U = 10.00, z = -2.81, p = 0.005) and all timepoints LAT2: Decreased by HIGH (U = 21.50, z = -2.05, p = 0.041) DUR3: Increased by HIGH at 10-15 min (U = 15.00, z = -2.40, p = 0.016) Chasing the FREQ: Effect of treatment at 10-15 min (χ2(4) = 10.55, p = 0.032) Intruder FREQ: Increased by HIGH at 10-15 min (U = 17.50, z = -2.22, p = 0.027) FREQ: Effect of treatment (χ2(4) = 13.06, p = 0.011) FREQ: Effect of treatment at 0-5 min (χ2(4) = 12.89, p = 0.012) Social FREQ: Effect of treatment at 5-10 min (χ2(4) = 10.76, p = 0.029) Investigation FREQ: Decreased by OIL at 5-10 min (U = 15.00, z = -2.22, p = 0.026) FREQ: Increased by LOW (U = 17.00, z = -2.05, p = 0.041) FREQ: Increased by HIGH (U = 11.50, z = -2.69, p = 0.007) and all timepoints Body FREQ: Effect of treatment (χ2(4) = 11.01, p = 0.026) Investigation FREQ: Increased by HIGH (U = 11.50, z = -2.69, p = 0.007) and all timepoints FREQ: Effect of treatment (χ2(4) = 13.85, p = 0.008) Anogenital FREQ: Effect of treatment at 0-5 min (χ2(4) = 10.73, p = 0.030) Investigation FREQ: Increased by HIGH (U = 8.00, z = -2.98, p = 0.003) and all timepoints Total Non- social FREQ: Increased by LOW (U = 16.50, z = -2.09, p = 0.037) Behaviours DUR: Decreased by OIL at 5-10 min (U = 17.00, z = -2.05, p = 0.041) DUR: Increased by LOW at 0-5 min (U = 20.00, z = -2.25, p = 0.025) FREQ: Effect of treatment at 10-15 min (χ2(4) = 10.55, p = 0.032) Digging FREQ: Decreased by OIL (U = 12.50, z = -2.45, p = 0.014) FREQ: Decreased by OIL at 5-10 min (U = 15.00, z = -2.26, p = 0.024) FREQ: Decreased by OIL at 10-15 min (U = 16.50, z = -2.10, p = 0.036) FREQ: Increased by LOW at 0-5 min (U = 20.00, z = -2.26, p = 0.024) DUR: Effect of treatment at 10-15 min (χ2(4) = 9.95, p = 0.041) Solitary FREQ: Effect of treatment (χ2(4) = 10.25, p = 0.036) Inactivity FREQ: Effect of treatment at 10-15 min (χ2(4) = 11.60, p = 0.021) LAT: decreased by HIGH (U = 21.00, z = -2.09, p = 0.037) a “OIL” indicates sesame oil vehicle. b “LOW” indicates 30 mg/kg WAY-200070. c “HIGH” indicates 180 mg/kg WAY-200070. 1FREQ represents frequency. 2LAT represents latency. 3DUR represents duration.
130 Table 21. Differences between gonadally intact male and female uninjected mice. Behaviour Effect FREQ1: Malesa > Femalesb (U = 16.00, z = -2.37, p = 0.018) Total Activity FREQ: Males > Females at 0-5 min (U = 16.00, z = -2.37, p = 0.018) Total Social FREQ: Males > Females at 0-5 min (U = 12.50, z = -2.66, p = 0.008) Behaviours LAT2: Females > Males (U = 12.50, z = -2.96, p = 0.003) DUR3: No difference between Males and Females Total Agonistic FREQ: Males > Females (U = 11.50, z = -2.74, p = 0.006) Behaviours FREQ: Males > Females at 0-5 min (U = 7.50, z = -3.07, p = 0.002) FREQ: Males > Females at 5-10 min (U = 19.50, z = -2.09, p = 0.037) Aggressive LAT: Females > Males (U = 32.50, z = -2.05, p = 0.041) Postures DUR: Males > Females (U = 10.00, z = -3.30, p = 0.001) and all timepoints Reciprocal FREQ: Males > Females (U = 10.00, z = -3.30, p = 0.001) and all Attacks timepoints LAT: Females > Males (U = 12.00, z = -3.56, p < 0.001) Chasing the FREQ: Males > Females (U = 10.00, z = -2.91, p = 0.004) Intruder Attacks FREQ: Males > Females (U = 10.00, z = -2.91, p = 0.004) and all Delivered timepoints Agonistic FREQ: Males > Females (U = 9.50, z = -2.91, p = 0.004) Behaviour FREQ: Males > Females at 0-5 min (U = 7.50, z = -3.07, p = 0.002) Received FREQ: Males > Females at 10-15 min (U = 20.00, z = -2.08, p = 0.037) DUR: Males > Females (U = 16.50, z = -2.36, p = 0.018) Avoidance of the FREQ: Males > Females (U = 17.00, z = -2.33, p = 0.020) Intruder FREQ: Males > Females at 5-10 min (U = 20.00, z = -2.64, p = 0.008) FREQ: Males > Females at 10-15 min (U = 22.50, z = -2.01, p = 0.045) Submissive DUR: Females > Males (U = 21.00, z = -1.96, p = 0.0499) Behaviour DUR: Females > Males at 5-10 min (U = 19.50, z = -2.10, p = 0.036) DUR: Males > Females (U = 22.00, z = -2.28, p = 0.023) DUR: Males > Females at 0-5 min (U = 20.00, z = -2.63, p = 0.008) Defensive DUR: Males > Females at 5-10 min (U = 25.00, z = -2.29, p = 0.022) Upright FREQ: Males > Females (U = 22.00, z = -2.28, p = 0.023) Posturing FREQ: Males > Females at 0-5 min (U = 20.00, z = -2.64, p = 0.008) FREQ: Males > Females at 5-10 min (U = 25.00, z = -2.29, p = 0.022) LAT: Females > Males U = 30.00, z = -2.03, p = 0.042 Social LAT: Females > Males (U = 13.50, z = -2.89, p = 0.004) Investigation Body LAT: Females > Males (U = 25.50, z = -2.03, p = 0.043) Investigation Anogenital DUR3: Females > Males (U = 13.00, z = -2.61, p = 0.009) Investigation DUR: Females > Males at 5-10 min (U = 19.00, z = -2.12, p = 0.034) Approaching and/or Attending FREQ: Males > Females at 0-5 min (U = 17.00, z = -2.29, p = 0.022) to the Intruder
131 Non-social Locomotor FREQ: Males > Females at 0-5 min (U = 19.50, z = -2.09, p = 0.037) Behaviours Horizontal FREQ: Males > Females at 0-5 min (U = 20.50, z = -2.01, p = 0.045) Exploration DUR: Males > Females at 0-5 min (U = 25.00, z = -1.98, p = 0.048) Solitary FREQ: Males > Females at 0-5 min (U = 24.00, z = -2.11, p = 0.035) Inactivity LAT: Females > Males (U = 22.00, z = -2.41, p = 0.016) a “Males” indicates gonadally intact males. b “Females” indicates gonadally intact females. 1FREQ represents frequency. 2LAT represents latency. 3DUR represents duration.
132 Table 22. Differences between gonadally intact and castrated male uninjected mice. Behaviour Effect DUR1: Castrateda > Malesb (U = 21.00, z = -2.35, p = 0.019) Total Activity DUR: Castrated > Males at 0-5 min (U = 22.00, z = -2.29, p = 0.022) DUR: Castrated > Males at 5-10 min (U = 17.00, z = -2.63, p = 0.009) DUR: Castrated > Males (U = 23.00, z = -2.20, p = 0.028) Total Social DUR: Castrated > Males at 10-15 min (U = 20.00, z = -2.42, p = 0.016) Behaviours LAT2: Castrated > Males (U = 18.00, z = -2.56, p = 0.010) Social DUR: Males > Castrated (U = 32.50, z = -2.05, p = 0.041) Inactivity Total Agonistic FREQ: Males > Castrated at 0-5 min (U = 23.50, z = -2.17, p = 0.030) Behaviours DUR: Males > Castrated at 0-5 min (U = 36.00, z = -2.10, p = 0.036) Aggressive FREQ: Males > Castrated at 0-5 min (U = 36.00, z = -2.10, p = 0.035) Postures LAT: Castrated > Males (U = 32.50, z = -2.05, p = 0.041) DUR: Males > Castrated (U = 14.00, z = -3.25, p = 0.001) and all timepoints Reciprocal FREQ: Males > Castrated (U = 16.00, z = -3.09, p = 0.002) and all Attacks timepoints LAT: Castrated > Males (U = 32.50, z = -2.05, p = 0.041) Chasing the FREQ2: Castrated > Males (U = 14.00, z = -2.89, p = 0.004) Intruder Dominant LAT: Males > Castrated (U = 15.00, z = -2.77, p = 0.006) Behaviour FREQ: Males > Castrated (U = 14.00, z = -2.89, p = 0.004) Attacks FREQ: Males > Castrated at 0-5 min (U = 21.00, z = -2.69, p = 0.007) Delivered FREQ: Males > Castrated at 10-15 min (U = 12.50, z = -3.13, p = 0.002) LAT: Castrated > Males (U = 20.00, z = -2.44, p = 0.015) Avoidance of DUR: Males > Castrated at 10-15 min (U = 25.00, z = -2.29, p = 0.022) the Intruder FREQ: Males > Castrated at 10-15 min (U = 27.00, z = -2.14, p = 0.033) DUR: Castrated > Males (U = 21.00, z = -2.35, p = 0.019) Submissive DUR: Castrated > Males at 10-15 min (U = 16.00, z = -2.75, p = 0.006) Behaviour FREQ: Castrated > Males at 10-15 min (U = 20.00, z = -2.47, p = 0.013) FREQ: Males > Castrated (U = 23.50, z = -2.49, p = 0.013) Attacks FREQ: Males > Castrated at 0-5 min (U = 30.00, z = -2.49, p = 0.013) Received LAT: Castrated > Males (U = 22.00, z = -2.60, p = 0.009) Defensive DUR: Males > Castrated at 0-5 min (U = 26.00, z = -2.49, p = 0.013) Upright FREQ: Males > Castrated at 0-5 min (U = 26.00, z = -2.50, p = 0.013) Posturing DUR: Castrated > Males (U = 5.00, z = -3.48, p < 0.001) and all timepoints Social FREQ3: Castrated > Males at 5-10 min (U = 25.00, z = -2.06, p = 0.039) Investigation LAT: Castrated > Males (U = 17.00, z = -2.63, p = 0.008) DUR: Castrated > Males (U = 10.00, z = -3.13, p = 0.002) and all Body timepoints Investigation FREQ: Castrated > Males (U = 23.00, z = -2.20, p = 0.028) FREQ: Castrated > Males at 5-10 min (U = 21.00, z = -2.35, p = 0.019)
133 Anogenital DUR: Castrated > Males (U = 2.00, z = -3.70, p < 0.001) and all timepoints Investigation FREQ: Castrated > Males (U = 7.50, z = -3.31, p = 0.001) and all timepoints Approaching DUR: Males > Castrated (U = 9.00, z = -3.20, p = 0.001) and all timepoints and/or FREQ: Males > Castrated (U = 22.00, z = -2.28, p = 0.023) Attending to FREQ: Males > Castrated at 0-5 min (U = 13.50, z = -2.89, p = 0.004) the Intruder FREQ: Males > Castrated at 10-15 min (U = 24.00, z = -2.14, p = 0.032) Total Non- DUR: Males > Castrated (U = 17.00, z = -2.63, p = 0.009) and all timepoints social FREQ: Males > Castrated (U = 8.00, z = -3.27, p = 0.001) and all timepoints Behaviours DUR: Males > Castrated (U = 17.00, z = -2.63, p = 0.009) Non-social DUR: Males > Castrated at 5-10 min (U = 23.00, z = -2.20, p = 0.028) Locomotor DUR: Males > Castrated at 10-15 min (U = 14.00, z = -2.84, p = 0.004) Behaviours FREQ: Males > Castrated (U = 10.00, z = -3.13, p = 0.002) and all timepoints DUR: Males > Castrated (U = 12.00, z = -2.99, p = 0.003) and all timepoints Horizontal FREQ: Males > Castrated (U = 11.00, z = -3.06, p = 0.002) and all Exploration timepoints DUR: Males > Castrated (U = 22.00, z = -2.27, p = 0.023) DUR: Males > Castrated at 0-5 min (U = 24.00, z = -2.15, p = 0.032) DUR: Males > Castrated at 10-15 min (U = 18.00, z = -2.56, p = 0.010) Vertical FREQ: Males > Castrated (U = 17.00, z = -2.63, p = 0.008) Exploration FREQ: Males > Castrated at 0-5 min (U = 17.50, z = -2.62, p = 0.009) FREQ: Males > Castrated at 10-15 min (U = 19.50, z = -2.46, p = 0.014) LAT: Castrated > Males (U = 23.00, z = -2.20, p = 0.028) DUR: Males > Castrated (U = 16.50, z = -2.78, p = 0.005) DUR: Males > Castrated at 5-10 min (U = 26.00, z = -2.15, p = 0.032) DUR: Males > Castrated at 10-15 min (U = 12.50, z = -3.12, p = 0.002) Digging FREQ: Males > Castrated (U = 20.00, z = -2.52, p = 0.012) FREQ: Males > Castrated at 5-10 min (U = 28.00, z = -2.01, p = 0.045) FREQ: Males > Castrated at 10-15 min (U = 15.50, z = -2.90, p = 0.004) LAT: Castrated > Males (U = 26.00, z = -2.07, p = 0.038) DUR: Males > Castrated (U = 26.00, z = -1.99, p = 0.047) Non-social DUR: Males > Castrated at 0-5 min (U = 24.00, z = -2.15, p = 0.032) Non- DUR: Males > Castrated at 10-15 min (U = 20.50, z = -2.38, p = 0.017) locomotor FREQ: Males > Castrated (U = 17.00, z = -2.64, p = 0.008) Behaviours FREQ: Males > Castrated at 10-15 min (U = 22.00, z = -2.29, p = 0.022) DUR: Males > Castrated (U = 20.00, z = -2.42, p = 0.016) DUR: Males > Castrated at 0-5 min (U = 24.00, z = -2.22, p = 0.026) Self- DUR: Males > Castrated at 10-15 min (U = 16.50, z = -2.67, p = 0.008) Grooming FREQ: Males > Castrated (U = 7.50, z = -3.32, p = 0.001) FREQ: Males > Castrated at 0-5 min (U = 22.50, z = -2.34, p = 0.019) FREQ: Males > Castrated at 10-15 min (U = 13.00, z = -2.96, p = 0.003) a “Castrated” indicates castrated males. b “Males” indicates gonadally intact males. 1DUR represents duration. 2LAT represents latency. 3FREQ represents frequency.
134 Table 23. Differences between gonadally intact and ovariectomized uninjected mice. Behaviour Effect Total Social LAT1: Ovxa > Femalesb (U = 18.00, z = -2.77, p = 0.006) Behaviours Total Agonistic DUR: Females > Ovx at 5-10 min (U = 22.00, z = -2.12, p = 0.034) Behaviour Avoidance of DUR: Ovx > Females at 5-10 min (U = 30.00, z = -2.16, p = 0.031) the Intruder FREQ: Ovx > Females at 5-10 min (U = 30.00, z = -2.16, p = 0.030) DUR: Females > Ovx at 0-5 min (U = 21.50, z = -2.16, p = 0.031) Submissive DUR: Females > Ovx 5-10 min (U = 20.00, z = -2.29, p = 0.022) Behaviour FREQ: Females > Ovx at 0-5 min (U = 18.50, z = -2.43, p = 0.015) Oronasal FREQ2: Ovx > Females (U = 21.00, z = -2.20, p = 0.028) Investigation FREQ: Ovx > Females at 5-10 min (U = 22.50, z = -2.08, p = 0.037) DUR3: Females > Ovx (U = 20.00, z = -2.27, p = 0.023) Anogenital DUR: Females > Ovx at 5-10 min (U = 19.00, z = -2.34, p = 0.019) Investigation FREQ: Females > Ovx (U = 24.00, z = -1.98, p = 0.048) FREQ: Females > Ovx at 5-10 min (U = 13.50, z = -2.79, p = 0.005) FREQ: Females > Ovx (U = 21.50, z = -2.19, p = 0.029) Stretched FREQ: Females > Ovx at 5-10 min (U = 17.00, z = -2.82, p = 0.005) Approaches FREQ: Females > Ovx at 10-15 min (U = 23.00, z = -2.40, p = 0.016) Non-social Locomotor DUR: Ovx > Females at 5-10 min (U = 17.00, z = -2.50, p = 0.013) Behaviours Horizontal DUR: Females > Ovx at 10-15 min (U = 24.00, z = -1.97, p = 0.049) Exploration DUR: Ovx > Females (U = 20.00, z = -2.27, p = 0.023) DUR: Ovx > Females at 5-10 min (U = 12.00, z = -2.91, p = 0.004) Digging FREQ: Ovx > Females (U = 21.00, z = -2.20, p = 0.028) FREQ: Ovx > Females at 5-10 min (U = 13.00, z = -2.85, p = 0.004) LAT: Females > Ovx (U = 29.00, z = -2.05, p = 0.041) a “Ovx” indicates ovariectomized females. b “Females” indicates gonadally intact females. 1LAT represents latency. 2FREQ represents frequency. 3DUR represents duration.
135
CHAPTER 4:
Acute effects of estradiol benzoate on the social transmission of food preferences
in female mice
This manuscript is submitted to the International Journal of Neuropsychopharmacology as
Clipperton-Allen, A.E., Flaxey, I., Webster, H.K., Nediu-Mihalache, C., and Choleris, E.
Acute effects of estradiol benzoate on the social transmission of food preferences in
female mice.
136 4.1 Abstract
Learning from conspecifics is one advantage of social living, and comprises most of human learning. However, the neurobiological bases of social learning are not well understood, and may differ from those of the more commonly studied individual learning.
The social transmission of food preferences (STFP) paradigm is a social learning test in which an observer develops a preference for a novel food through social interaction with a demonstrator animal that previously consumed that food. There is evidence that estrogens may be involved, as estrogen receptor alpha (ERα) activation blocks STFP whether administered before, immediately following or 24h after the social interaction phase, while
ERβ activation prolongs demonstrated food preferences when administered pre-acquisition
(i.e., pre-social interaction), blocks them when administered immediately post-acquisition, and blocks or prolongs them when administered 24h post-acquisition. Effects of simultaneous ERα/ERβ activation by estradiol on STFP in ovariectomized females are unknown. Forty-eight hours before the choice test, ovariectomized observer mice received acute subcutaneous injections of sesame oil vehicle or 0.04mg/kg, 0.2mg/kg, or 0.4mg/kg estradiol benzoate (EB) 48h before choice test. Injections occurred 48h before (pre- acquisition), immediately following (immediate post-acquisition), or 24h following social interaction (24h delayed post-acquisition). Mice receiving EB pre- or immediately post- acquisition preferred the demonstrated food for twice as long as controls, while STFP was unaffected by 24h delayed post-acquisition EB. Estrogen that binds to both receptors thus may affect memory consolidation, but did not affect retrieval or expression of this preference specifically. Similarities between pre-acquisition results of EB and ERβ agonist treatment suggest that EB may be acting on STFP predominantly through ERβ, possibly by affecting
137 dopamine D1-type receptors, which have been implicated in the STFP.
138 4.2 Introduction
Learning from conspecifics is one of the advantages of social living, and has been observed in many species (see Galef & Laland, 2005 for a review). Indeed, in humans social learning is very prominent, and impairment of this type of learning has severe consequences
(e.g., in individuals suffering from autism spectrum disorders; Senju et al., 2007; Shane &
Albert, 2008). However, despite its importance, the neurobiological bases of social learning are not well understood, and may differ substantially from those of individual learning, which is more often studied in the laboratory.
One task that requires social learning and is amenable to detailed laboratory investigations is the social transmission of food preferences (STFP) paradigm, which has demonstrated the ability of numerous rodent species (e.g., house mice, rats; Galef et al.,
1988; Galef & Wigmore, 1983; Valsecchi & Galef, 1989) to increase their food repertoire by learning from conspecifics, thereby avoiding the potential costs of trial-and-error individual learning. In this task, an “observer” (OBS) animal interacts with a “demonstrator” (DEM) that has recently eaten a flavoured food. The OBS is then given a choice between two flavoured foods, both of which are novel to the OBS, and one of which was on the breath of its DEM (Galef & Wigmore, 1983). The OBS will typically prefer the DEM food, as long as it has previously encountered its odour on the breath of a live conspecific (Galef et al., 1988;
Valsecchi & Galef, 1989); the food preference will not develop as a consequence of interaction with a food powdered surrogate (Galef et al., 1988; Valsecchi & Galef, 1989), the food odour alone, or exposure to the odour during social interaction with a conspecific that did not consume the flavour (Choleris et al., 2011a). Thus, the STFP taps into mechanisms of learning that specifically depend upon the social context, and allows for the assessment of
139 the neurobiological underpinnings of this distinctively social learning. Additionally, we have previously found that a delay between the social interaction and the choice test does not affect the STFP, with OBS showing equal socially learned preferences with a post-learning delay of up to 30 days (Ervin et al., 2011).
While estrogens are known to be involved in social behaviour, learning/memory and cognition, the role of these hormones in social learning has only recently come under investigation. We have previously shown that this biologically relevant and specifically social paradigm is mediated by the estrogen receptors alpha (ERα) and beta (ERβ; Clipperton et al., 2008; Clipperton-Allen et al., 2011b), which are expressed differently in development
(Kudwa et al., 2006), differently distributed in the brain (Merchenthaler et al., 2004b; Mitra et al., 2003b), and encoded by different genes (Kuiper et al., 1996). These receptors also have different involvement in social learning and memory; the ERα agonist propyl pyrazole triol (PPT) blocked social learning, whether administered before, immediately after, or 24h after the social interaction during which the food preference is acquired. While all phases of memory could be impaired independently, these results, together with PPT’s long-lasting effects (Jacome et al., 2010b; Rhodes & Frye, 2006; Santollo et al., 2007; Walf & Frye,
2005), suggest that activation of ERα may predominantly interfere with expression or retrieval of the socially acquired food preferences (Clipperton-Allen et al., 2011b).
The involvement of ERβ is more complex: mice treated with the ERβ agonists 7-Bromo-
2-(4-hydroxyphenyl)-1,3-benzoxazol-5-ol (WAY-200070) and diarylpropionitrile (DPN) at testing preferred the DEM food for twice as long as controls when treated prior to the social interaction, potentially promoting acquisition by affecting the nature of the social interaction
140 (Clipperton et al., 2008; Clipperton-Allen et al., 2011b). However, when administered immediately following the social interaction, DPN blocked or increased the duration of the preference for the DEM food expressed at testing, depending on the dose administered, suggesting that activation of ERβ may interfere with consolidation or retrieval of the memory
(Clipperton-Allen et al., 2011b).
The effects of PPT and DPN in the STFP clearly point at an important role for estrogens in social learning. However, to the best of our knowledge, the involvement of estrogens in the STFP and other social learning paradigms has never been directly investigated. Since selective activation of ERα and ERβ has different effects on this paradigm, the effects of estrogens, which bind equally to the two receptors, cannot be inferred from the results of studies with PPT and DPN, especially as estrogens could also be acting through other receptors (e.g., the G-protein coupled estrogen receptor; Thomas et al., 2010). An involvement of estrogens in social learning would be consistent with its role in normal social recognition, which is necessary for complex social interactions like the formation of dominance hierarchies (Choleris et al., 2009; Pierman et al., 2008), and social hierarchies do affect social learning (Clipperton et al., 2008; Kavaliers et al., 2005a). Further, action of estrogens or selective agonists on ERs modulate the nature of social interactions, both affiliative and agonistic (e.g., Clipperton-Allen et al., 2011a; Clipperton-Allen et al., 2010).
All of these can affect how animals learn from others (Clipperton et al., 2008; Valsecchi et al., 1996). The often conflicting data on non-social learning suggest a complex relationship or interactions between estrogens, the various phases of memory processing (acquisition, consolidation, retrieval), and motivational factors involved in learning, including those related to the social context in which social learning occurs (see Choleris et al., 2008a;
141 Daniel, 2006; Frick, 2009 for recent reviews).
In the current experiments, we investigated the involvement of estrogens in the three phases of processing of a memory that is acquired within a social context. To this end, EB was administered prior to the social interaction (during which the food preference is acquired) to affect all three phases of memory (Experiment 1), immediately following the social interaction, to affect consolidation and retrieval (Experiment 2), or 24h following the social interaction, to affect only memory retrieval (Experiment 3). In all cases, ovariectomized (ovx) female CD1 mice were treated with EB 48h prior to the choice test (see
Figure 9 for an experimental timeline). Additionally, when administered before the acquisition phase (Experiment 1), EB effects on the mice’s behaviour during the social interactions were assessed with an ethological analysis, as we have previously seen relationships between aspects of the social interaction and the preference for the DEM food
(Clipperton et al., 2008; Clipperton-Allen et al., 2011b).
4.3 Method
4.3.1 Subjects
Experimentally naïve female mice (Mus musculus) of the CD1 strain were obtained from
Charles River, QC, Canada; 20 mice were randomly assigned the role of DEM, and 298 mice were OBS. Mice were only used as OBS once, but vehicle-treated OBS were subsequently used as DEMs. All animals were ovx under isofluorane anaesthesia (as described in
Clipperton-Allen et al., 2011d; Clipperton-Allen et al., 2011b), then were individually housed for at least 7 days before being paired into OBS-DEM dyads three days before the beginning of the experiment. Thus, the OBS were tested 10-15 days post-ovx. Housing consisted of
142 Figure 9. Timeline of a) pre-acquisition, b) immediate post-acquisition and c) 24h delayed post-acquisition paradigms. In all cases, 48h elapsed between observers receiving their injection and their choice test.
143 clear polyethylene cages (26 x 16 x 12 cm) equipped with corn cob bedding, environmental enrichment and free access to tap water and food (Teklad Global 14% Protein Rodent
Maintenance Diet, Harlan Teklad, Madison, WI, USA).
The colony room was kept on a 12:12h reversed light/dark cycle (lights on at 2000h) at
21±1°C. All procedures were in accordance with the Canadian Council on Animal Care’s regulations, and approved by the University of Guelph Institutional Animal Care and Use
Committee.
4.3.2 Diets
The diets used consisted of rodent chow, finely ground and mixed with either 1% ground cinnamon (CIN; McCormick Ground Cinnamon, McCormick Canada, London, Canada) or
2% powdered cocoa (COC; Fry’s Premium Cocoa, Cadbury Ltd., Mississauga, Canada) by weight. The two foods have equivalent metabolic and physical features, and were equipalatable for other female CD1 mice from Charles River (Choleris et al., 2011a;
Clipperton-Allen et al., 2011b).
4.3.3 Apparatus
DEM feeding was carried out in polyethylene cages (26 x 16 x 12 cm), which contained 5 cm high cylindrical jars, with a 7.5 cm diameter (Dyets Inc, Bethlehem, PA, USA), which were equipped with a stainless steel ring (hole 2.5 cm in diameter) and a sleeve collar to prevent spillage. Clear Plexiglas lids with air holes were used during OBS-DEM social interactions, which were videotaped with an 8 mm Sony Handycam in Nightshot. For the
OBS choice tests, each mouse was placed in a polyethylene cage (42.5 x 26.5 x 18.5 cm), with a stainless steel wire lid and ad libitum water. The two flavoured diets were contained
144 in two plastic feeders (Tecniplast, Varese, Italy), which were weighed every 2h on a balance accurate to 0.01 g (Sartorius Analytical Balance, Sartorius Inc., United Kingdom). These feeders, described previously (Valsecchi & Galef, 1989), allow for very precise measurement of food consumption.
4.3.4 Hormone Treatments
OBS were treated with a single 10 ml/kg subcutaneous injection of estradiol benzoate
(EB) or sesame oil vehicle 48h before the choice test to allow the long-term effects to take place. Four independent groups were formed for each administration time, consisting of a sesame oil vehicle control, 0.04 mg/kg EB, 0.2 mg/kg EB, and 0.4 mg/kg EB (see Table 24 for treatment distributions); these doses correspond to physiological levels (e.g., Akinci &
Johnston, 1997; Jansson et al., 1990). To prevent leakage of the oil-based solutions, injection sites were sealed with superglue (Instant Krazy Glue, Elmer’s Products Inc., Columbus, OH,
USA).
4.3.5 Procedure
4.3.5.1 Phase 1 (DEM Feeding and Social Interaction)
The DEM-OBS dyads were moved into the experimental room, the DEM fur was coloured with black magic marker for identification, and all mice were food deprived and left undisturbed for at least 12h. Early in the active phase, DEM mice were moved for 1h into an empty cage containing either COC- or CIN-flavoured food, which they were allowed to consume; following this, DEM were returned to their home cages for 30 min of free social interaction with OBS. The data of any OBS that interacted with DEM that consumed less than 0.1 g of the flavoured food would have been removed from the experiment, although
145 Table 24. Number of observer (OBS) mice in each treatment group. Treatment Number of OBS Experiment 1 Pre-acquisition Administration Vehicle COC1 DEM2 = 12, CIN3 DEM = 12 0.04 mg/kg EB COC DEM = 12, CIN DEM = 14 0.2 mg/kg EB COC DEM = 11, CIN DEM = 12 0.4 mg/kg EB COC DEM = 11, CIN DEM = 12 Experiment 2 Immediate Post-acquisition Administration Vehicle COC DEM = 13, CIN DEM = 13 0.04 mg/kg EB COC DEM = 12, CIN DEM = 11 0.2 mg/kg EB COC DEM = 13, CIN DEM = 13 0.4 mg/kg EB COC DEM = 13, CIN DEM = 13 Experiment 3 Delayed Post-acquisition Administration Vehicle COC DEM = 12, CIN DEM = 12 0.04 mg/kg EB COC DEM = 12, CIN DEM = 15 0.2 mg/kg EB COC DEM = 12, CIN DEM = 11 0.4 mg/kg EB COC DEM = 13, CIN DEM = 10 1COC indicates that the observer interacted with a cocoa-fed demonstrator. 2DEM represents demonstrator. 3CIN indicates that the observer interacted with a cinnamon-fed demonstrator.
146 this was not necessary in these experiments. Additionally, at least 1h prior to the social interaction, vaginal swabs were taken and later examined under a microscope at 100x magnification following Giemsa staining (Sigma-Aldrich, Oakville, ON) to determine if any mice were cycling or if there were effects of treatment on vaginal cytology.
4.3.5.2 Phase 2 (Delay)
Following the social interaction in Experiment 1, the OBS, which had been treated with
EB or vehicle 48h prior to the test day, were immediately given a choice between the COC and CIN diets (Phase 3). In Experiment 2, OBS were injected immediately post-interaction, while in Experiment 3, OBS received their treatment 24h after the social interaction (see
Figure 9 for a timeline). In all cases, treatment took place 48h before the choice test, and the duration of the delay between the social interaction and the choice test was 0h (Experiment
1), 48h (Experiment 2) or 72h (Experiment 3). When the delay was 48h or 72h, OBS mice were singly housed following the social interaction, and received ad libitum unflavoured food access until approximately 12h prior to the test, when they were moved into the testing room and food deprived. We have previously shown that delays of this duration do not affect the STFP, with OBS showing equal socially learned preferences with a post-learning delay of up to one week (Clipperton-Allen et al., 2011b), and the results of our control groups in the present study confirm this.
4.3.5.3 Phase 3 (Choice Test)
OBS were placed into the testing cages, in which both flavoured diets were available, and underwent an 8h continuous choice test. Every 2h, feeders were manually weighed, and the consumption of each diet was calculated. This allowed us to assess the duration of the
147 preference for the DEM food.
4.3.6 Behavioural Analysis
The videotaped social interactions in Experiment 1 were scored by two trained, blind-to- treatment observers (interrater reliability correlation of 0.97) using a 21-behaviour ethological analysis based on Grant and Mackintosh’s (Grant & Mackintosh, 1963) ethogram
(see Table 25, based on Clipperton et al., 2008) with The Observer Video Analysis software
(Noldus Information Technology, Wageningen, The Netherlands). All behaviours were collected based on the OBS, and behaviour of the DEM was collected only in relation to the
OBS behaviour. In addition to the individual behaviours, composite behavioural categories were also calculated to assess more general treatment effects (see Table 25, modified from
Clipperton et al., 2008).
4.3.7 Statistical Analyses
Only results that are significant at an α level of 0.05 are reported. All analyses were performed with PASW 18.0 for Windows (SPSS Inc., an IBM Company, Chicago, IL, USA).
4.3.7.1 Food preferences and food consumption
For each 2h interval, the quantity of each diet consumed was determined, and both a CIN preference ratio [CIN diet consumed/(COC + CIN diets consumed)] and a DEM food preference ratio [DEM's diet consumed/(total COC + CIN diets consumed)] were calculated.
The analysis of the preference for the DEM food allows us to compare the strength of the socially learned food preference in the different treatment groups, while the analysis on the
CIN preference ratio allows us to assess whether or not the effects of social learning are present in each group at each 2h interval. In addition, the total intake per 2h interval was 148 calculated (COC + CIN diets) to assess if overall food consumption was affected by treatment.
Three-way mixed design analyses of variance (ANOVAs) to ascertain the effects of time
(2h, 4h, 6h, 8h), DEM food (COC, CIN), and EB treatment (0.04 mg/kg, 0.2 mg/kg, 0.4 mg/kg EB, vehicle) on the OBS CIN preference were not performed as approximately 15% of animals (distributed across treatment groups) did not eat during at least one of the 2h intervals (most often during the last of these intervals, which was shortly before the light cycle change), greatly reducing the sample size when repeated measures tests were used
(Clipperton et al., 2008; Clipperton-Allen et al., 2011b). Thus, for each experiment, each 2h interval was analyzed using two-way ANOVAs to determine effects of DEM food and EB treatment on CIN preference, with Bonferroni post hoc tests as necessary (Clipperton et al.,
2008). Additionally, planned comparisons of the effect of DEM food on CIN preference was assessed separately for each treatment group at each 2h interval using independent sample t- tests (Clipperton et al., 2008; Clipperton-Allen et al., 2011b), as this allows us to assess whether or not the effect of social learning is observable in each group at each time interval.
Two-way mixed-model ANOVAs were also used to analyze the effects of treatment and time on the DEM food preference ratio. Food intake results were analyzed using two-way mixed design ANOVAs (treatment, time), and Bonferroni and paired-sample t-tests were used as post hocs as appropriate.
4.3.7.2 Behavioural analysis
In Experiment 1, EB effects on the social interaction were also analyzed, as the OBS had been treated prior to this phase. Effects of EB are reported in comparison to the vehicle
149 controls. The duration, frequency and latency of each behaviour were assessed, both in six 5- min intervals and in the entire 30 min social interaction. As the majority of behavioural elements were not normally distributed (Clipperton et al., 2008; Clipperton-Allen et al.,
2010; Clipperton-Allen et al., 2011a), behavioural data were analyzed using the nonparametric Kruskal-Wallis and Mann-Whitney U tests. We have previously found a relationship between OBS submissiveness and the degree or duration of DEM food preference in the group showing a longer preference for the DEM food (Clipperton-Allen et al., 2011b), so Pearson product-moment correlations were calculated between the degree and duration of the DEM food preference and submissive behaviour, agonistic behaviour received, and dominance score, an assessment of whether the OBS (positive score) or DEM
(negative score) is dominant in the pair (behaviours described in Clipperton et al., 2008) for each treatment group separately.
4.4 Results
Mice receiving the high dose of EB at testing showed a preference for their DEM’s food for twice as long as control mice when the drug was administered either before or immediately after the acquisition (Phase 1), but it did not affect the socially acquired preference when OBS were treated 24h post-acquisition. Effects of EB on behaviour in the social interaction did not correlate with the duration of DEM food preference. Food intake decreased across the test, as we have repeatedly noted (e.g., Choleris et al., 2011a; Clipperton et al., 2008; Clipperton-Allen et al., 2011b), and some minor effects of EB on intake were also observed.
150 4.4.1 Experiment 1: Pre-acquisition EB Administration
4.4.1.1 OBS food preference
Two-way ANOVAs found significant effects of DEM food on the CIN preference of
OBS at 2h (F(1,66) = 28.07, p < 0.001) and 4h (F(1,86) = 9.78, p = 0.002), but no effects of
DEM food or drug at 6h or 8h. Planned comparisons of the CIN preference in OBS that interacted with CIN-fed DEM to that of OBS with COC-fed DEM for each treatment group found that all groups preferred the food their DEM ate during the first 2h interval (vehicle: t(22) = 4.70, p < 0.001, 0.04 mg/kg EB: t(24) = 2.61, p = 0.015, 0.02 mg/kg EB: t(21) = 3.26, p = 0.005, 0.4 mg/kg EB: t(21) = 3.44, p = 0.002; see Figure 10), while the 0.4 mg/kg EB group also showed a preference at 4h (t(21) = 2.37, p = 0.027). No effect of DEM food on
CIN preference was found in any other time period for any other group. Two-way ANOVAs also found a significant effect of time on the DEM food preference ratio (F(3,255) = 9.99, p
< 0.001). There were neither a significant effect of treatment nor a treatment x time interaction on the DEM food preference ratio, likely due to the fact that only one group
(0.4mg/kg EB) still showed a socially acquired food preference at 4h and to a ceiling effect, since all groups, including controls, have a high level of performance on this task in the first
2h interval.
4.4.1.2 Behavioural analysis
In general, EB decreased social behaviour, social investigation and oronasal investigation, and increased non-social behaviour, both locomotor and non-locomotor.
Details of the effects of EB on behaviour during the social interaction are in Appendix 3 and
Figure 11.
151 Figure 10. Cinnamon preference by 2h intervals of observers in Experiment 1 (pre- acquisition EB). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers were treated with a) Sesame oil vehicle. b) 0.04 mg/kg EB. c) 0.2 mg/kg EB. d) 0.4 mg/kg EB.
152 1.0 ac 26h Interval Cocoa Demonstrators Cinnamon Demonstrators 0.8 Cocoa Demonstrators * Cinnamon Demonstrators * * 0.6 *
0.4
0.2 Cinnamon Preference Ratio Cinnamon Preference Ratio 0.0 Oil 0.04 0.2 0.4 mg/kg EB
1.0 bd 48h Interval
Cocoa Demonstrators 0.8 Cioncnoaam Doenm Doenmsotrnastotrrastors Cinnamon Demonstrators 0.6 *
0.4
0.2 Figure 11. Effects of acute pre-acquisition EB on behaviour during the social interaction. Cinnamon Preference Ratio a) Duration of oronasal investigation by 5 min 0intervals..0 b) Duration of dominance behaviour. c) Duration of total non-social behaviour byO i5l min intervals.0.04 d) 0Frequency.2 0 .4of self- grooming. * indicates a significant difference from vehicle, p < 0.05. +m indicatesg/kg EB a significant difference between the vehicle control group and all drug-treated groups, p < 0.05.
153 Unlike our previous findings with ERβ agonists (Clipperton-Allen et al., 2011b), there were no significant correlations between the three measures of submissive behaviour and the duration or degree of preference for the DEM food in the 0.4 mg/kg EB group, which showed the prolonged DEM food preference. However, we did observe correlations between behaviour and the degree or duration of DEM food preference in other groups. In the vehicle control group, the Dominance Score correlated with the DEM food preference at 2h (r(19) =
0.499, p = 0.021) and 4h (r(19) = 0.548, p = 0.010), although as a group, the vehicle treated mice did not show a significant preference for the DEM food at 4h. Similarly, agonistic behaviour received correlated with the duration of preference (r(15) = 0.487, p = 0.047), and submissive behaviour (r(15) = 0.487, p = 0.020) and agonistic behaviour received (r(15) =
0.632, p = 0.007) correlated with DEM food preference at 4h in the 0.04 mg/kg EB treated mice, although this group also did not show a preference at this time.
4.4.2 Experiment 2: Immediate Post-acquisition EB Administration
Two-way ANOVAs found significant effects of DEM food on the CIN preference of
OBS at 2h (F(1,90) = 51.72, p < 0.001), and a significant DEM food x treatment interaction at 4h (F(3,92) = 3.14, p = 0.029), but no effects of DEM food or drug at 6h or 8h. Planned comparisons indicated that all groups showed a significant effect of DEM food during the first 2h interval (vehicle: t(24) = 2.17, p = 0.041; 0.04 mg/kg EB: t(20) = 5.39, p < 0.001; 0.2 mg/kg EB: t(23) = 2.34, p = 0.028; 0.4 mg/kg EB: t(23) = 4.85, p < 0.001; see Figure 12). At
4h, only the 0.4mg/kg EB group still showed a significant effect of DEM food (t(24) = 2.10, p = 0.046). Two-way ANOVAs also found a significant effect of time on the DEM food preference ratio (F(3,252) = 7.80, p < 0.001). Neither the main effect of treatment nor the time x treatment interactions were significant for the DEM food preference ratio data, likely 154 Figure 12. Cinnamon preference by 2h intervals of observers in Experiment 2 (immediate post-acquisition EB). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers were treated with a) Sesame oil vehicle. b) 0.04 mg/kg EB. c) 0.2 mg/kg EB. d) 0.4 mg/kg EB.
155 because only the 0.4mg/kg EB group still showed a socially acquired food preference at 4h and because of the high level of performance in all OBS, including the vehicle-treated mice in the first interval.
4.4.3 Experiment 3: Delayed Post-acquisition EB Administration
Two-way ANOVAs found significant effects of DEM food on the CIN preference of
OBS at 2h (F(1,82) = 59.66, p < 0.001) and 4h (F(1,88) = 5.12, p = 0.026), but no effects of
DEM food or drug at 6h or 8h. Planned comparisons indicated that all groups showed a significant effect of DEM food on the OBS CIN preference during the first 2h (vehicle: t(20)
= 4.61, p < 0.001; 0.04 mg/kg EB: t(24) = 6.71, p < 0.001; 0.2 mg/kg EB: t(21) = 3.37, p =
0.003; 0.4 mg/kg EB: t(17) = 2.32, p = 0.033; see Figure 13), but at no other time points, indicating that delayed post-acquisition EB treatment did not affect the STFP. Two-way
ANOVAs also found a significant effect of time on the DEM food preference ratio (F(3,219)
= 6.69, p < 0.001).
4.4.4 EB Effects on Food Intake
In all three experiments, food intake decreased over time (all Fs > 88.00, all ps < 0.001; see Appendix 3, Figure 14), which Bonferroni post hoc tests indicated was due to significant differences between all 2h intervals (all ps < 0.006). Additionally, there were significant treatment x time interactions in all three studies (all Fs > 2.00, all ps < 0.035), and a significant main effect of treatment in Experiment 2 (F(3,98) = 2.93, p = 0.037). Bonferroni post hoc tests showed that in Experiment 2, OBS receiving 0.04 mg/kg EB (p = 0.002) and
0.4 mg/kg EB (p = 0.015) consumed significantly less food than the vehicle controls during the first 2h interval, and the 0.04 mg/kg EB group also ate significantly less overall during the choice test than the vehicle controls (p = 0.031). In Experiment 3, Bonferroni post hoc 156 Figure 13. Cinnamon preference by 2h intervals of observers in Experiment 3 (24h post- acquisition EB). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers were treated with a) Sesame oil vehicle. b) 0.04 mg/kg EB. c) 0.2 mg/kg EB. d) 0.4 mg/kg EB.
157 tests indicated that mice receiving 0.04 mg/kg EB ingested significantly more food than the vehicle controls during the last 2h interval (p = 0.002).
4.5 Discussion
Like in our previous results with an ERβ agonist (Clipperton et al., 2008; Clipperton-
Allen et al., 2011b), we did not find a direct effect of treatment on the strength of the socially acquired food preference. We found that mice treated with the estrogen EB preferred the
DEM food for twice as long as controls during the choice test when treated before (Figure
10) or immediately after the social interaction (Figure 12), but did not differ from controls when EB was administered 24h post-interaction (Figure 13). While EB affected a number of social behaviours, these effects did not appear to relate directly to the duration of DEM food preference.
The observed results with pre-acquisition EB are similar to those observed with pre- acquisition ERβ agonists (Clipperton et al., 2008; Clipperton-Allen et al., 2011b) and in high estrogen phases of the estrous cycle (Choleris et al., 2011a). The occurrence of similar results with systemic EB and ERβ agonists when administered pre-acquisition is consistent with some of the literature on non-social learning paradigms. Both EB and ERβ agonists
DPN and/or WAY-200070, but not ERα agonist PPT, improved performance on the radial arm, Y and Morris water mazes, as well as on object recognition and object placement tasks
(Jacome et al., 2010b; Liu et al., 2008; Rhodes & Frye, 2006; Walf et al., 2008a).
Additionally, neither EB nor DPN administration to ERβ knockout (βERKO) mice enhanced object recognition, object placement or Morris water maze tasks, while both treatments improved performance on these paradigms in wild-type (WT) littermates (Rissman et al.,
158 2002a; Walf et al., 2008a), which suggests that estrogens’ improving effects on learning and memory paradigms may be through actions on ERβ, not ERα. Additionally, while ERα is typically more transcriptionally active than ERβ (Cowley et al., 1997), physiological levels of EB, as in our study, can suppress ERα’s estrogen response (Hall & McDonnell, 1999), and heterodimers of the two ERs may also regulate protein levels of the ERs themselves (Nomura et al., 2003). This may explain why we see similar results with EB, which acts on both receptors, as with an ERβ selective agonist alone.
Our previous finding that both ERα and ERβ agonists block the STFP when administered immediately following the social interaction suggests that both receptors may need to be activated for the preference of the OBS for the DEM’s food to be showed for a longer time during testing or that other ERs, such as the recently described GPER (formerly GPR30;
Filardo & Prossnitz, 2009; Revankar et al., 2005), are also involved in the STFP. This remains to be investigated. The results of both immediate and 24h delayed post-acquisition administrations are consistent with some previous results indicating that when EB is administered immediately after training, it improves learning and memory, but that EB has no effects when animals are treated more than 2h after training (reviewed in Packard, 1998).
A single injection immediately post-training, as in Experiment 2, can improve subsequent performance on both working and reference memory aspects of spatial and non-spatial tasks, but not if administered 2-3h after training (reviewed in Frick, 2009).
Observer mice treated with the same dose (0.4mg/kg) of EB showed a socially acquired food preference for a longer time during the choice test whether EB was administered pre- or immediately post-acquisition. Interestingly, this dose (0.4 mg/kg) produces proestrous levels
159 of estradiol (Akinci & Johnston, 1997; Jansson et al., 1990). Consistently, proestrous mice, which have elevated estrogen and progesterone levels (Walmer et al., 1992) during testing express the socially acquired food preference for at least twice as long as diestrous and estrous mice (Choleris et al., 2011a), and when tested 24h after the social interaction, only proestrous mice showed a preference for their DEM food (Sánchez-Andrade et al., 2005).
However, in the present study mice receiving the proestrus-like dose of EB did not show a preference for the DEM food for as long as the gonadally intact female mice in proestrus
(Choleris et al., 2011a), suggesting that progesterone or its neuroactive metabolites may also play an enhancing role in the STFP paradigm.
That OBS mice treated with the same dose of EB showed a DEM food preference during testing for longer than controls when EB was administered either before or immediately after the social acquisition of the food preference also suggests that EB may be acting through the same or similar mechanisms in early phases of memory encoding, possibly by rapidly improving memory consolidation. Consistently, our laboratory has recently shown that estradiol can rapidly (within 40 min) improve performance on social recognition, object recognition, and object placement tasks, as well as increasing dendritic spines in the CA1 region of the hippocampus in ovx female CD1 mice (Phan et al., 2011). This suggests that estrogens may be modulating these tasks by affecting the hippocampus, a region known to be involved in the STFP task (e.g., Alvarez et al., 2001; Alvarez et al., 2002; Bunsey &
Eichenbaum, 1995). This is also consistent with the presence of both ERs in the hippocampus, albeit at low levels (Merchenthaler et al., 2004b; Mitra et al., 2003b).
Estrogens can increase baseline synaptic excitability and LTP, inhibit γ-aminobutyric acid
(GABA) synthesis and decrease LTD (reviewed in Frick, 2009), actions on synaptic protein
160 activation and LTP enhancement which appear to depend on ERβ (Liu et al., 2008), which is also consistent with our previous results with ERβ agonists (Clipperton et al., 2008;
Clipperton-Allen et al., 2011b).
Unlike our previous results with ERβ agonists (Clipperton et al., 2008; Clipperton-Allen et al., 2011b), OBS mice that received pre-acquisition EB and expressed a longer preference for the DEM food during testing did not show a significant relationship between submission to the DEM and duration or degree of socially acquired food preference. It is possible that when ERβ acts in the absence of ERα activation, it affects the acquisition through modulating the social interaction between the OBS and the DEM, but estrogens have different effects via their activation of both receptors, possibly due to an interaction of the effects on social interactions mediated by the different ERs or to estrogens’ activation of other receptors (e.g.,
GPER).
There are several possible mechanisms by which estrogens and ERβ could be mediating learning and memory. Hippocampal and/or neocortical plasticity may be increased by estrogens’ increase of cholinergic input from cholinergic basal forebrain (CBF) neurons that project to these regions, as post-ovx administration of estradiol increases levels of mRNA for choline acetyltransferase (ChAT; enzyme that synthesizes acetylcholine) in the basal forebrain, as well as increasing ChAT activity, acetylcholine (ACh) release, and high affinity choline uptake in the CBF, neocortex and hippocampus (reviewed in Frick, 2009). Estrogens can also modulate dopamine (DA), norepinephrine (NE) and serotonin (5HT) systems
(Jacome et al., 2010b). Both ERβ agonists and estrogens increase 5HT and DA activity in several brain regions (Jacome et al., 2010b; Johnson et al., 2010; Morissette et al., 2008), and
161 also increase NE activity in both the amygdala and the hippocampus (reviewed in Frick,
2009). Our previous finding that systemic administration of DA D1-type (but not D2-type) receptor antagonist specifically impaired the STFP without affecting feeding (Choleris et al.,
2011a) implicates the DA system in the STFP. DA receptors and transporters have also been shown to be modulated by estradiol and DPN, but not PPT (Jacome et al., 2010b; Morissette et al., 2008). Taken together, this suggests that EB may affect the STFP by modulating the effect of DA, possibly through ERβ, and may be acting specifically on the D1-like receptors.
4.5.1 Conclusions
Mice receiving treatment with estrogens at testing prefer the DEM food for twice as long as controls, and this appears to be independent of the nature of the social interaction, possibly by affecting the consolidation of the memory. Estrogens do not appear to affect the retrieval or expression of the socially acquired food preference, as EB administration 24h after the acquisition phase (i.e., the social interaction) had no effect on the STFP paradigm. We have previously found that activation of ERα blocks the STFP regardless of administration time, while at testing ERβ treated mice showed a preference for their DEM food for longer than control mice when administered pre-acquisition and possibly 24h post-acquisition, but blocks this preference when administered immediately and possibly 24h post-acquisition
(Clipperton et al., 2008; Clipperton-Allen et al., 2011b). The similarity between the pre- acquisition results of EB and ERβ agonist treatment suggest that EB may be acting on the
STFP predominantly through ERβ. The beneficial effects of estrogens, likely through ERβ, on social learning and social cognition (reviewed in Choleris et al., 2009; Choleris et al.,
2008a), suggest that targeting this receptor specifically could retain the beneficial cognitive and emotional effects of hormone replacement therapy, while avoiding some of the negative
162 side effects associated with ERα (e.g., breast cancer; Koehler et al., 2005).
163 4.6 Appendix 3: Effects of EB on Social Interactions in Experiment 1
The duration of Total Social Behaviour was decreased by all three doses of EB in the first
5 min (0.04 mg/kg EB: U = 120.00, z = -2.15, p = 0.031; 0.2 mg/kg EB: U = 138.50, z = -
2.06, p = 0.039; 0.4 mg/kg EB: U = 116.00, z = -2.26, p = 0.024), and was increased by the
0.4 mg/kg dose at 5-10 min (U = 120.00, z = -2.15, p = 0.031) and 25-30 min (U = 124.50, z
= -2.04, p = 0.042). The duration of social inactivity at 5-10 min was affected by the treatment (χ2(3) = 10.25, p = 0.020), due to the 0.4 mg/kg dose increasing this behaviour at that time (U = 107.00, z = -2.51, p = 0.12). There was an effect of treatment on the frequency of total social investigation overall (χ2(3) = 8.56, p = 0.036) and at 10-15 min
(χ2(3) = 9.27, p= 0.026). The 0.04 mg/kg EB group showed less frequent investigation overall (U = 119.50, z = -2.17, p = 0.030) and in the middle third of the interaction (10-15 min: U = 120.50, z = -2.15, p = 0.031; 15-20 min: U = 124.00, z = -2.10, p = 0.036), and also performed less frequent oronasal investigation in the last 5 min of the trial (U = 131.00, z = -
2.01, p = 0.044). The duration of both total social investigation (χ2(3) = 9.48, p = 0.024) and oronasal investigation (χ2(3) = 9.48, p = 0.024) at 0-5 min were also affected by treatment, due to all three doses of EB decreasing these behaviours at this time (social investigation:
0.04 mg/kg EB: U = 107.00, z = -2.51, p = 0.012; 0.2 mg/kg EB: U = 129.00, z = -2.30, p =
2.30, p = 0.021; 0.4 mg/kg EB: U = 105.00, z = -2.56, p = 0.010; oronasal investigation: 0.04 mg/kg EB: U = 127.00, z = -1.96, p = 0.050; 0.2 mg/kg EB: U = 133.00, z = -2.20, p = 0.028;
0.4 mg/kg EB: U = 95.50, z = -2.82, p = 0.005). In addition, a treatment effect on the duration of approaching and/or attending to the intruder was found for the 5-10 min interval
(χ2(3) = 9.46, p = 0.024), largely due the 0.2 mg/kg group spending more time performing this behaviour at this time (U = 100.50, z = -3.02, p = 0.003) and overall (U = 141.00, z =
164 -2.00, p = 0.046).
The frequency of total aggression was affected by treatment in the first 10 min of the interaction (0-5 min: χ2(3) = 9.06, p = 0.029; 5-10 min χ2(3) = 9.16, p = 0.029), with all three doses increasing total aggression in the first 5 min (0.04 mg/kg EB: U = 96.50, z = -2.79, p =
0.005; 0.2 mg/kg EB: U = 137.0, z = -2.1, p = 0.036; 0.4 mg/kg EB: U = 118.00, z = -2.21, p
= 0.027), and 0.2 mg/kg EB increasing the frequency of this behaviour at 5-10 min (U =
134.00, z = -2.18, p = 0.029). The 0.2 mg/kg EB dose decreased agonistic behaviour delivered at 10-15 min (U = 156.50, z = -2.08, p = 0.037), and follow at 15-20 min (U =
168.00, z = -2.35, p = 0.019). Dominance behaviour was affected by treatment (overall: χ2(3)
= 8.01, p = 0.046; 5-10 min: χ2(3) = 7.81, p = 0.050) and was decreased by the 0.04 mg/kg
EB dose, both at 5-10 min (U = 152.00, z = -2.24, p = 0.025) and overall (U = 131.00, z = -
2.23, p = 0.026), and the latency to dominance behaviour was increased (U = 136.00, z = -
2.07, p = 0.039). Both 0.04 mg/kg EB (U = 168.00, z = -1.96, p = 0.050) and 0.2 mg/kg EB
(U = 145.50, z = -2.20, p = 0.028) decreased the dominance score at 10-15 min.
Total non-social behaviours were affected by treatment in the first 5 min (χ2(3) = 11.40, p
= 0.010), with all three doses increasing the amount of time spent in non-social behaviour at this time (0.04 mg/kg EB: U = 101.00, z = -2.67, p = 0.008; 0.2 mg/kg EB: U = 117.50, z = -
2.59, p = 0.010; 0.4 mg/kg EB: U = 97.00, z = -2.78, p = 0.006). The duration of non-social locomotor behaviours were affected by treatment during the first 10 min (0-5 min: χ2(3) =
14.29, p = 0.003; 5-10 min: χ2(3) = 9.54, p = 0.023), increased by all three doses at 0-5 min
(0.04 mg/kg EB: U = 76.00, z = -3.35, p = 0.001; 0.2 mg/kg EB: U = 112.00, z = -2.73, p =
0.006; 0.4 mg/kg EB: U = 103.00, z = -2.51, p = 0.009) and by the 0.2 mg/kg dose at 5-10
165 min (U = 125.50, z = -2.39, p = 0.017). Both horizontal and vertical exploration also showed effects of treatment in the first 10 min (horizontal: 0-5 min χ2(3) = 7.83, p = 0.050; 5-10 min
χ2(3) = 9.27, p = 0.026; vertical: 0-5 min χ2(3) = 13.35, p = 0.004; 5-10 min χ2(3) = 8.32, p =
0.040), with exploration increased by 0.04 mg/kg EB in the first 5 min (horizontal: U =
103.00, z = -2.61, p = 0.009; vertical: U = 81.00, z = -3.21, p = 0.001) and by 0.2 mg/kg EB at 5-10 min (horizontal: U = 118.00, z = -2.58, p = 0.010; vertical: U = 140.00, z = -2.03, p =
0.043). The 0.2 mg/kg and 0.4 mg/kg EB doses also increased vertical exploration at 0-5 min
(0.2 mg/kg EB: U = 101.00, z = -3.01, p = 0.003; 0.4 mg/kg EB: U = 118.00, z = -2.21, p =
0.027). Latency to vertical exploration was also affected by treatment (χ2(3) = 12.34, p =
0.006) due to being decreased by all three doses (0.04 mg/kg EB: U = 85.00, z = -3.10, p =
0.002; 0.2 mg/kg EB: U = 116.00, z = -2.63, p = 0.009; 0.4 mg/kg EB: U = 102.50, z = -2.63, p = 0.009).
The frequency of non-social non-locomotor behaviour was affected by treatment at 5-10 min (χ2(3) = 10.13, p = 0.017), and increased by 0.04 and 0.2 mg/kg EB at 0-5 min (0.04 mg/kg EB: U = 119.50, z = -2.17, p = 0.030; 0.2 mg/kg EB: U = 131.00, z = -2.25, p =
0.024), although the 0.04 mg/kg dose of EB decreased the overall duration of non-social non- locomotor behaviour (U = 126.00, z = -1.99, p = 0.047). Self-grooming showed a treatment effect in the first 5 min (χ2(3) = 8.18, p = 0.042), with an effect on its frequency overall (χ2(3)
= 8.88, p = 0.031). Self-grooming was decreased by 0.04 mg/kg EB overall (U = 127.00, z =
-1.96, p = 0.050) and at 0-5 min (U = 107.50, z = -2.50, p = 0.012). The 0.4 mg/kg EB dose also decreased self-grooming (U = 105.00, z = -2.59, p = 0.010) and increased the frequency of solitary inactivity (U = 121.00, z = -2.32, p = 0.021) in the last 5 min.
166 Figure 14. Total food intake (g) by 2h intervals over the 8h choice test. a) Experiment 1 (pre- acquisition EB). OBS received sesame oil vehicle (black square), 0.04 mg/kg EB (white circle), 0.2 mg/kg EB (white triangle), or 0.4 mg/kg EB (white square). b) Experiment 2 (immediate post-acquisition EB). OBS received sesame oil vehicle (black square), 0.04 mg/kg EB (white circle), 0.2 mg/kg EB (white triangle), or 0.4 mg/kg EB (white square). c) Experiment 3 (24h delayed post-acquisition EB). OBS received sesame oil vehicle (black square), 0.04 mg/kg EB (white circle), 0.2 mg/kg EB (white triangle), or 0.4 mg/kg EB (white square). Means and standard errors are shown. * significant difference between OBS that received sesame oil vehicle and those that received 0.04 mg/kg EB, p < 0.05. † significant difference between OBS that received sesame oil vehicle and those that received 0.4 mg/kg EB, p < 0.05. ** significant difference between OBS that received sesame oil vehicle and those that received 0.04 mg/kg EB over the entire choice test, p < 0.05. 167 Table 25. Behaviours collected during the social interaction. Scored Description Composite/Calculated Behaviours Behaviour OBS mouse is in control; includes Agonistic Behaviour Delivered, Dominance pinning of DEM, aggressive Total Agonistic Behaviour, Total Behaviours grooming, crawling over or on top, Social Behaviour, Total Activity and mounting attempt. OBS mouse actively follows, or Agonistic Behaviour Delivered, Chasing pursues and chases DEM; reciprocal Total Agonistic Behaviour, Total DEM1 to Avoiding DEM. Social Behaviour, Total Activity Physical attacks, including Agonistic Behaviour Delivered, Attack dorsal/ventral bites. Only the Total Agonistic Behaviour, Total Delivered frequency of attacks was measured. Social Behaviour, Total Activity DEM mouse is in control; includes crawl under, supine posture (ventral Agonistic Behaviour Received, Submissive side exposed), prolonged crouch, and Total Agonistic Behaviour, Total Behaviours any other behaviour in which DEM is Social Behaviour, Total Activity dominant (e.g., DEM pins, aggressively grooms, etc., OBS). Agonistic Behaviour Received, Avoiding OBS withdraws and runs away from Total Agonistic Behaviour, Total DEM DEM while DEM is chasing. Social Behaviour, Total Activity Physical attacks including bites to Agonistic Behaviour Received, Attack dorsal/ventral regions. Only the Total Agonistic Behaviour, Total Received frequency of attacks was measured. Social Behaviour, Total Activity Defensive Species-typical defensive behaviour; Agonistic Behaviour Received, Upright upright with the head tucked and the Total Agonistic Behaviour, Total Posture arms ready to push away. Social Behaviour, Total Activity Physical attacks with a locked fight Reciprocal including tumbling, kick-away and Total Agonistic Behaviour, Total Attacks counterattack where the attacker Social Behaviour, Total Activity cannot be identified. Physical attacks which include Aggressive box/wrestle, offensive and defensive Total Agonistic Behaviour, Total Postures postures, lateral sideways threats and Social Behaviour, Total Activity tail rattle. Agonistic Behaviour Delivered – N/A Dominance Score Agonistic Behaviour Received Oronasal Active sniffing of DEM’s oronasal Social Investigation, Total Social Investigation area. Behaviour, Total Activity Body Social Investigation, Total Social Active sniffing of DEM’s body. Investigation Behaviour, Total Activity Anogenital Active sniffing of DEM’s anogenital Social Investigation, Total Social Investigation region. Behaviour, Total Activity
168 Often from across the cage; OBS’s attention is focused on DEM, head Approach tilted toward DEM and movements Social Investigation, Total Social and/or Attend toward DEM; this becomes “Chasing Behaviour, Total Activity to DEM DEM” once along the tail or sniff if within 1.5 cm of DEM. Risk assessment behaviour; back feet Stretch do not move and front feet approach Social Investigation, Total Social Approach DEM. Only the frequency of Behaviour, Total Activity stretched approaches was measured. Social Includes sit/lie/sleep together. Total Social Behaviour Inactivity Movement to investigate upwards, Total Non-social Locomotor Vertical both front feet off the ground; Behaviour, Total Non-social Exploration includes sniffing, wall leans and lid Behaviour, Total Activity chews (less than 3). Total Non-social Locomotor Horizontal Movement around the cage; includes Behaviour, Total Non-social Exploration active sniffing of air and ground. Behaviour, Total Activity Total Non-social Locomotor Rapid stereotypical movement of Digging Behaviour, Total Non-social forepaws in the bedding. Behaviour, Total Activity Total Non-social Non-locomotor Self- Rapid movement of forepaws over Behaviour, Total Non-social Grooming facial area and along body. Behaviour Total Non-social Non-locomotor Solitary No movement; includes sit, lie down Behaviour, Total Non-social Inactivity and sleep. Behaviour “Strange” behaviours, including “Abnormal” spinturns, repeated jumps/lid N/A Stereotypies chews/head shakes (more than 3).
169
CHAPTER 5:
The involvement of estrogen receptors alpha and beta in the memory
for a socially acquired food preference in female mice
This manuscript is submitted to the Journal of Neuroendocrinology as Clipperton-Allen,
A.E., Flaxey, I., Foucault, J.N., Hussey, B., Rush, S.T., Tam, C., Brown, A., and
Choleris, E. The involvement of oestrogen receptors alpha and beta in the memory for a
socially acquired food preference in female mice.
170 5.1 Abstract
Social learning is the predominant form of learning for humans, and is pervasive among mammalian species. Access to conspecifics from which to learn, thus providing an alternative to risky trial-and-error learning, is a major advantage of group living. Despite its clear importance, the neurobiology of this cognitive ability is largely unknown. The social transmission of food preferences paradigm (STFP) assesses the ability of “observer” animals to expand their food repertoire by preferring novel foods smelled on the breath of a “demonstrator” animal. We have previously shown that this social learning is modulated by activation of the estrogen receptors (ERs), such that an ERα agonist blocked the STFP, while an ERβ agonist induced a prolonged preference for the demonstrated food. However, it is unclear what aspect(s) of learning (acquisition, consolidation, retrieval) were being affected.
We thus introduced a delay between the social interaction (acquisition) phase and the food preference test, and administered either the ERα agonist propyl pyrazole triol
(PPT) or the ERβ agonist diarylpropionitrile (DPN) to observers immediately, or 24 h after, the social interaction. Observers were given an 8 h continuous food choice test
48 h after treatment. Results suggested that PPT blocks the STFP regardless of administration time. DPN may interfere with consolidation, as it blocked learning when administered immediately post-acquisition, but when administered 24 h later, it appears to have either blocked or prolonged the expression of the acquired preference, depending on dose. ERβ thus may promote acquisition by affecting social behaviour, but may impair consolidation, and its effects on retrieval appear to vary, possibly due to variations in anxiolytic actions. Possible mechanisms of action for
171 the two ERs are discussed.
172 5.2 Introduction
Social learning is the predominant form of learning for humans, and impairments in learning of this type lead to considerable deficits in living a functional life (e.g., autism spectrum disorders; Senju et al., 2007; Shane & Albert, 2008). Social learning is also pervasive among mammalian species (reviewed in Galef & Laland, 2005), and one major advantage of social living. Although social learning is essential for normal life in modern society, neurobiological studies of learning and behaviour have focused more often on individual learning, which may use different mechanisms and likely involve different brain areas; the neurobiology of this important cognitive ability is largely unknown.
The social transmission of food preference (STFP) paradigm assesses one example of this under-studied type of learning, in which an animal can avoid the risks involved in trial-and- error learning by choosing foods that a conspecific has already consumed (Galef, 1996). In this paradigm, which has been utilized in several rodent species (e.g., rats, house mice, gerbils; Choleris et al., 1998b; Galef, 1996), a “demonstrator” (DEM) is fed a flavoured food and then interacts briefly with an “observer” (OBS), who will typically then prefer the food consumed by the DEM (Galef, 1996). For both mouse and rat OBS to learn the preference, it is necessary for the food odour to be carried in the mouth of a live DEM from which the semiochemical product of digestion, carbon disulphide, is emitted (Galef, 1996; Valsecchi &
Galef, 1989). Exposure to the food odour alone, to a conspecific in the vicinity of the food
(Choleris et al., 2011a), or to a food powdered gauze and cotton batten surrogate DEM
(Galef, 1996; Valsecchi & Galef, 1989) is insufficient for the development of the food preference. Thus, the STFP is both a specifically social type of learning and a biologically relevant task.
173 This task has been shown to be mediated by estrogens, which bind to two mostly intracellular receptors, alpha (ERα) and beta (ERβ), which are encoded by different genes, expressed differently in development, and which have largely but not completely overlapping brain distributions (Choleris et al., 2008a; Mitra et al., 2003a). Like in other aspects of physiology and behaviour (e.g., Choleris et al., 2008b; Gustafsson, 2006b; Rissman et al.,
2002b), the two ERs differentially affect the STFP (Clipperton et al., 2008). When administered prior to the STFP paradigm, ERα agonist propyl pyrazole triol (PPT) blocked social learning. ERβ agonist WAY-200070, instead, induced a prolonged preference for the
DEM food in a manner, and these effects may be due to changes in aspects of the social interaction (i.e., increased submissive behaviour) that are related to prolonged expression of the DEM food preference (Clipperton et al., 2008). These results are similar to those observed for non-social learning paradigms (Rissman et al., 2002b). However, given that these agonists are still present in plasma 3-4 h after administration (Harris et al., 2002a;
Patisaul et al., 2009) and have much further extended genomic effects, it is not clear which of the three stages of memory (acquisition, consolidation, and retrieval; Abel & Lattal, 2001a) were affected by these pre-acquisition treatments (Clipperton et al., 2008). Elucidation of the phase(s) affected by estrogens could have implications for treatment of peri- and post- menopausal memory impairments; for example, there is some evidence of impaired acquisition of new memories during the peri-menopausal period (Greendale et al., 2009).
Different molecular mechanisms have been implicated in these stages of memory in a number of learning and memory tasks; for example, acquisition may involve protein kinases
A (PKA) and C (PKC) and the n-methyl-d-aspartate receptor (NMDAR; Abel & Lattal,
2001b). Consolidation may also be mediated by PKA and NMDAR, as well as the
174 α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR), cyclic adenosine monophosphate response element binding protein (CREB), dopamine and/or extracellular signal-related kinase (ERK)/mitogen-activated protein kinase (MAPK) signalling cascades (Abel & Lattal, 2001b; Guzowski & McGaugh, 1997), while retrieval appears to involve norepinephrine, hippocampal ERK, acetylcholine (ACh), and dopamine
(Huang et al., 2010; Li et al., 2010; Mahmoodi et al., 2010; Murchison et al., 2004).
In order to assess the effects of ERα and ERβ agonists on different aspects of the STFP, we needed to isolate each phase of the STFP paradigm. This can be achieved by introducing a delay between the acquisition (during the social interaction) and the retrieval (choice test).
To ensure that the long-term genomic effects had taken place, drugs required administration
48 h prior to the experiment. To accommodate this, and to be able to test both pre- and post- consolidation effects of all drugs, we needed a delay of at least 72 h. While it has been previously shown that OBS rats will demonstrate a preference for their DEM food when tested up to 30 days after a single 30 min social interaction (Galef & Whiskin, 2003), the maximum duration of delay between acquisition and choice test that still allows mice to show a preference had not been determined. Pilot data indicated that this delay could be increased to at least 30, but less than 40, days without affecting the OBS preference for the DEM food
(Ervin et al., 2011). Additionally, because of the prolonged duration of the drugs’ effects
(e.g., Jacome et al., 2010b; Rhodes & Frye, 2006), we could not administer treatment for the consolidation or acquisition phases in isolation. Thus, we administered the agonists either immediately post-acquisition to attempt to affect the consolidation and retrieval (but not the acquisition), or 24 h post-acquisition to attempt to affect only the retrieval (Guzowski &
McGaugh, 1997).
175 In the current experiments, we therefore investigated the involvement of ERα and ERβ in the acquisition, consolidation and retrieval of a socially learned food preference by administering the ERα agonist PPT or the ERβ agonist diarylpropionitrile (DPN) to ovariectomized (ovx) CD-1 female mice before, immediately after, or 24 h after the social interaction. In all cases, social learning was tested 48 h post-injection.
5.3 Methods
5.3.1 Subjects
Experimentally naïve female CD-1 mice (Mus musculus) were obtained from Charles
River, QC, Canada at 2-3 months of age. DEM were randomly chosen naïve animals or OBS who had previously received vehicle treatments, while mice were only used as OBS once.
Mice were housed individually following ovx surgery (described below) until three days prior to testing, when they were paired into OBS-DEM dyads. They were provided with corn cob bedding and environmental enrichment in clear polyethylene cages (26 x 16 x 12 cm) and given access to tap water and food (Teklad Global 14% Protein Rodent Maintenance
Diet, Harlan Teklad, Madison, WI) ad libitum. Mice were housed at 21±1°C under a 12:12 h reversed light/dark cycle (lights off at 0800 h). This research, and all procedures, were approved by the University of Guelph Institutional Animal Care and Use Committee and conducted in accordance with the regulations of the Canadian Council on Animal Care.
5.3.2 Surgery
The mice underwent bilateral ovx as described in Clipperton-Allen et al. (2011d). Briefly, under isoflurane gas anaesthesia (Benson Medical Industries, Markham, ON) and with
50mg/kg carprofen injected subcutaneously (s.c.) at 10ml/kg (Rimadyl, Pfizer Canada Inc,
176 Kirkland, QC) as analgesic and anti-inflammatory, ovaries were removed through a single dorsal incision in the skin and two small incisions in the muscles over the ovaries. A few drops of local anaesthetic marcaine (25% bupivacaine hydrochloride in saline, diluted to
12.5%, Hospira, Inc, Lake Forest, IL) were dripped onto the skin incision, which was then stapled with 1-2 MikRon Autoclip 9mm wound clips (MikRon Precision Inc, Gardena, CA), and followed by a rehydrating intraperitoneal injection of 0.5ml saline. Additional analgesic treatment was provided when necessary. Following at least 7 days of single-housed recovery, OBS were followed by at least 3 days of paired with DEM for at least 3 days housing before being tested (approximately 10 to 15 days following ovx).
5.3.3 Diets
The diets used for both the OBS choice tests and the DEM feeding were composed of ground standard rodent diet, mixed with either 2% powdered cocoa (COC; Fry’s Premium
Cocoa, Cadbury Ltd., Mississauga, Canada) or 1% ground cinnamon (CIN; McCormick
Ground Cinnamon, McCormick Canada, London, Canada) by weight. Other female CD-1 mice obtained from Charles River, QC found these flavoured diets equipalatable (e.g.,
Choleris et al., 2011a).
5.3.4 Apparatus
Cylindrical jars, 5 cm high and 7.5 cm in diameter (Dyets Inc, Bethlehem, PA), fitted with a stainless steel ring (hole 2.5 cm in diameter) and a sleeve collar to prevent spillage, were used to feed one of the flavoured diets to the DEM, which occurred in clean polyethylene cages (26 x 16 x 12 cm). During the social interactions between OBS and
DEM in their home cages, clear Plexiglas lids with air holes were used. OBS choice tests
177 were run in large polyethylene cages (42.5 x 26.5 x 18.5 cm) with stainless steel wire lids.
Water was available ad libitum, and two plastic metabolic feeders (described in Valsecchi &
Galef, 1989; Tecniplast, Varese, Italy) held the two flavoured diets. These feeders allow for very precise measurement of the mice’s food consumption (Valsecchi et al., 1989), and were weighed on a balance accurate to 0.01 g (Sartorius Analytical Balance, Sartorius Inc., United
Kingdom) to measure the consumption of each diet.
5.3.5 Drugs
Two ER specific agonists were used in Experiments 2 and 3: the ERα agonist 1,3,5- tris(4-hydroxyphenyl)-4-propyl-1H-pyrazole (PPT), which is 410 times more selective for
ERα than ERβ in mice (Stauffer et al., 2000), and the ERβ agonist 2,3-bis(4-hydroxyphenyl)- propionitrile (DPN), which is 78 times more selective for ERβ than ERα (Meyers et al.,
2001). For each drug, a vehicle (sesame oil) control group and three drug-dose groups (0.025 mg/kg, 0.05 mg/kg, 0.1 mg/kg) were formed. These treatments were administered to independent groups of OBS by s.c. injection at 10 ml/kg 48 h prior to the choice test in order for the genomic effects of the drugs to be exerted (see Table 26 for treatment distributions).
The timing and doses of PPT were the same as those that have been previously shown to affect the STFP (Clipperton et al., 2008). We chose the DPN doses to be the same as PPT, based on the literature (e.g., Walf et al., 2008a), and to facilitate DPN’s dissolution, the sesame oil for this drug contained a minute amount of ethanol (less than 0.05%) for both drug and vehicle treated groups in the DPN experiments (Experiments 2b, 3b). Experiment 1
(below) further confirmed that this timing and these doses of DPN have comparable effects on the STFP to those of WAY-200070 (Clipperton et al., 2008). The injection sites were sealed with superglue (Instant Krazy Glue, Elmer’s Products Inc., Columbus, OH, United
178 States of America) to prevent leakage of the oil-based solutions.
5.3.6 General Procedure
5.3.6.1 Phase 1 (DEM Feeding and Social Interaction)
In Phase 1, all mice were moved into the experimental room and food deprived, the DEM fur was coloured with black magic marker, and the mice were left undisturbed for at least 12 h. DEM mice were removed from the home cage and placed into an empty cage with a jar of either COC- or CIN-flavoured food, and allowed to eat the food for 1 h. If any DEM ate less than 0.1 g, they and their OBS data were removed from the experiment (n = 3 across all studies). After feeding, the DEM were returned to the cages with the OBS and allowed to freely interact for 30 min.
Additionally, early in the active phase, vaginal swabs were taken from all mice; these were later stained with Giemsa stain (Sigma-Aldrich, Oakville, ON) and examined under a microscope at 100x magnification, thus using vaginal cytology to confirm that the ovx mice showed no estrous cycle; the data from 9 OBS across all six studies (less than 1.5%) were removed due to OBS cycling. It is unlikely that these effects are due to PPT or DPN treatment, as 4/9 were in vehicle groups, and the remaining five were distributed across drug treatments (see Table 26).
5.3.6.2 Phase 2 (Delay)
Following the social interaction, the OBS were housed individually for the appropriate period of time (see details of various experiments below and Figure 15) before beginning the choice test.
179 Table 26. Distribution of treatment groups. Numbers indicate the remaining number of mice in each group after the removal of OBS that were cycling (numbers in parentheses). Experiment 1a: Post-interaction Delay and DEM Familiarity Effects 0h Delay DEM1 Familiar COC3 DEM: n = 10 (1), CIN4 DEM: n = 11 0h Delay DEM Unfamiliar COC DEM: n = 11, CIN DEM: n = 10 (1) 90 min Delay DEM Familiar COC DEM: n = 12, CIN DEM: n = 12 90 min Delay DEM Unfamiliar COC DEM: n = 11, CIN DEM: n = 11 48h Delay DEM Familiar COC DEM: n = 14, CIN DEM: n = 12 (1) 48h Delay DEM Unfamiliar COC DEM: n = 11, CIN DEM: n = 8 (3) 7 days Delay DEM Familiar COC DEM: n = 12, CIN DEM: n = 12 7 days Delay DEM Unfamiliar COC DEM: n = 11 (1), CIN DEM: n = 12 30 days Delay DEM Familiar COC DEM: n = 13, CIN DEM: n = 13 Experiment 1b: DPN at the Chosen Doses Has the Same Effect as WAY-200070 Vehicle (sesame oil) COC DEM: n = 6, CIN DEM: n = 7 DPN 0.025 mg/kg COC DEM: n = 7, CIN DEM: n = 7 (ERβ 0.05 mg/kg COC DEM: n = 7, CIN DEM: n = 6 agonist) 0.1 mg/kg COC DEM: n = 7, CIN DEM: n = 6 Experiment 2a: Immediate Post-acquisition Administration of PPT Vehicle (sesame oil) COC DEM: n = 11 (1), CIN DEM: n = 10 (2) PPT 0.025 mg/kg COC DEM: n = 13, CIN DEM: n = 14 (ERα 0.05 mg/kg COC DEM: n = 14, CIN DEM: n = 14 (2) agonist) 0.1 mg/kg COC DEM: n = 12, CIN DEM: n = 12 Experiment 2b: Immediate Post-acquisition Administration of DPN Vehicle (sesame oil) COC DEM: n = 15, CIN DEM: n = 14 DPN 0.025 mg/kg COC DEM: n = 13, CIN DEM: n = 12 (ERβ 0.05 mg/kg COC DEM: n = 14, CIN DEM: n = 15 agonist) 0.1 mg/kg COC DEM: n = 15, CIN DEM: n = 12 (1) Experiment 2c: Immediate Post-acquisition Control Study PPT/DPN Vehicle (sesame oil) COC DEM: n = 12, CIN DEM: n = 13 PPT 0.05 mg/kg COC DEM: n = 12, CIN DEM: n = 12 PPT 0.1 mg/kg COC DEM: n = 12, CIN DEM: n = 11 DPN 0.025 mg/kg COC DEM: n = 11, CIN DEM: n = 12 DPN 0.05 mg/kg COC DEM: n = 10, CIN DEM: n = 15 DPN 0.1 mg/kg COC DEM: n = 11, CIN DEM: n = 12 Experiment 3a: 24h Delayed Post-acquisition Administration of PPT Vehicle (sesame oil) COC DEM: n = 14 (1), CIN DEM: n = 13 PPT 0.025 mg/kg COC DEM: n = 12, CIN DEM: n = 11 (ERα 0.05 mg/kg COC DEM: n = 14, CIN DEM: n = 15 agonist) 0.1 mg/kg COC DEM: n = 12 (1), CIN DEM: n = 13 Experiment 3b: 24h Delayed Post-acquisition Administration of DPN Vehicle (sesame oil) COC DEM: n = 14, CIN DEM: n = 14 DPN 0.025 mg/kg COC DEM: n = 12, CIN DEM: n = 11 (1) (ERβ 0.05 mg/kg COC DEM: n = 12, CIN DEM: n = 11 agonist) 0.1 mg/kg COC DEM: n = 14, CIN DEM: n = 13 1DEM indicates demonstrator. 2OBS indicates observer. 3COC indicates that the OBS’s DEM ate the cocoa diet. 4CIN indicates that the OBS’s DEM ate the cinnamon diet.
180 Injection 47.5h 30 min 8h A Pre-acquisition Post-Treatment Social Choice Test Interval Interaction (acquisition)
Injection 30 min 48h 8h B Immediate Post-acquisition Social Post-Treatment Interval Choice Test Interaction (acquisition)
Injection C 24h Delayed 30 min 24h 48h 8h Post-acquisition Social Post-Interaction Post-Treatment Interval Choice Test Interaction Delay (acquisition)
Figure 15. Timeline of pre-, immediate post- and 24h delayed post-acquisition paradigms (Experiments 1b-3b and (Clipperton et al., 2008). In all paradigms, 48h elapsed between injection and choice test.
181 5.3.6.3 Phase 3 (Choice Test)
In experiments with a delay between the social interaction and the test, OBS received rodent chow and tap water ad libitum and were kept in the colony room until being returned to the experimental room and food deprived at least 12 h before the choice test. At least 30 min after undergoing a vaginal swab to assess possible effects of drug on cytology, the OBS were placed into the testing cages and given an 8 h choice test in which both flavoured diets were continuously available. The OBS food preferences were assessed by manually weighing the feeders every 2 h and the consumption of each flavour was recorded in order to gain time course information (Clipperton et al., 2008).
5.3.7 Study-specific Procedures
A timeline of the pre-acquisition, immediate post-acquisition, and 24 h delayed post- acquisition procedures is shown in Figure 15. In all cases, the choice test took place 48 h after drug treatment.
5.3.7.1 Pre-acquisition DPN (Experiment 1)
Because we were using a different ERβ agonist than in our previous study, this experiment was run to verify whether the effects of the high, middle and low doses of DPN would correspond to those previously seen with high, middle and low doses of WAY-200070
(Clipperton et al., 2008). We tested OBS injected with sesame oil (vehicle) control, 0.025 mg/kg DPN, 0.05 mg/kg DPN, and 0.1 mg/kg DPN. In order to replicate the previous study, there was no delay between the social interaction and the choice test. Additionally, the social interactions, which were videotaped in Nightshot using a Sony 8 mm Handycam, were analyzed by a blind, trained observer to assess effects of drug treatments on social and
182 non-social behaviour using The Observer Video Analysis software (Noldus Information
Technology, Wageningen, the Netherlands), as in Clipperton et al. (2008; see Table 27 for selected behaviour descriptions, and Tables 1 and 2 in Clipperton et al., 2008 for descriptions of all behaviours).
5.3.7.3 Immediate Post-acquisition Administration of PPT and DPN (Experiment 2a
and 2b)
In Experiment 2, OBS were injected immediately following the social interaction. Mice in Experiment 2a were treated with sesame oil vehicle, 0.025 mg/kg PPT, 0.05 mg/kg PPT, or 0.1 mg/kg PPT. In Experiment 2b, OBS received sesame oil vehicle, 0.025 mg/kg DPN,
0.05 mg/kg DPN, or 0.1 mg/kg DPN. All OBS began their choice tests 48 h after their social interaction and injection.
5.3.7.4 Immediate Post-acquisition Individual Learning Control Study (Experiment 2c)
Because both PPT and DPN blocked social learning when administered immediately following the acquisition phase, an individual learning control study was performed in order to determine if the drugs were producing an aversion to or avoidance of the odour of the
DEM food that was independent of the social interaction. Thus, instead of interacting with a
DEM, the experimental mice were exposed for 30 min to jars of one of the flavoured foods, which were covered with wire mesh that allowed the mice to smell the odour but prevented them from consuming the diet (as in Choleris et al., 2011a). Immediately following this exposure period, mice were treated with sesame oil vehicle (n = 25), or one of the doses of
PPT and DPN that had blocked the STFP: 0.05 mg/kg PPT (n = 24), 0.1 mg/kg PPT (n = 23),
0.025 mg/kg DPN (n = 23), 0.05 mg/kg DPN (n = 25), or 0.1 mg/kg DPN (n = 23). As in
183 Experiments 2a and 2b, all mice were administered a choice test 48 h after the injection.
5.3.7.5 24h Delayed Post-acquisition Administration of PPT and DPN (Experiment 3a
and 3b)
In Experiment 3, OBS were treated approximately 24 h following the social interaction, and started their choice tests 48 h after drug administration (72 h after their social interaction). Experiment 3a OBS received sesame oil vehicle, 0.025 mg/kg PPT, 0.05 mg/kg
PPT, or 0.1 mg/kg PPT, and those in Experiment 3b were injected with sesame oil vehicle,
0.025 mg/kg DPN, 0.05 mg/kg DPN, or 0.1 mg/kg DPN.
5.3.8 Statistical Analyses
The amount of the COC and CIN diets ingested in each 2 h interval was calculated, and then a CIN preference ratio [CIN diet consumed/(COC + CIN diets consumed)], arcsine transformed (as in Clipperton et al., 2008), was used to express OBS choice data. Total intake (COC + CIN diets consumed) was calculated for each 2 h interval to determine if the drugs affected overall food consumption.
A three-way mixed design analysis of variance (ANOVA) was performed to assess the effects of treatment (drug), demonstrated food, and time on the CIN preference of the OBS.
One- or two-way ANOVAs, paired sample t-tests and Bonferroni corrected multiple comparisons were used as post-hoc tests where appropriate. However, a substantial number of animals, distributed across treatment groups, did not consume any food during at least one of the four 2 h intervals; the most common time interval during which mice did not eat was the 8 h time interval, which occurred shortly before the beginning of the light or inactive phase; this is not surprising, as mice consume the majority of their food in the beginning of 184 the nocturnal active phase of the light/dark cycle (Clipperton et al., 2008; Latham & Mason,
2004). This greatly decreased the sample size available for repeated measures testing.
Because of this, planned comparisons of the effect of demonstrated food on the CIN preference in each group separately were performed using one-way ANOVAs for each 2 h interval (Clipperton et al., 2008).
To determine if there were effects of treatment or time on the total food intake, two- or three-way mixed design ANOVAs were performed, with post-hoc Bonferroni and paired sample t-tests as needed to determine the sources of significance. Mice that did not eat during a 2 h block had an intake of zero.
The behavioural data from the social interaction in Experiment 1 were analyzed as described in Clipperton et al. (2008). Briefly, the frequency, duration and latency of 31 observed or calculated behaviours (see Table 27 for selected behaviour descriptions, and
Tables 1 and 2 in Clipperton et al., 2008 for descriptions of all behaviours) were assessed across the entire 30 min interaction (Clipperton et al., 2008). Most behavioural elements were not normally distributed, so data were analyzed using the Kruskal-Wallis and Mann-
Whitney U nonparametric tests (equivalent to one-way ANOVA and independent sample t- tests, respectively). Drug effects are in comparison to the corresponding vehicle control group unless otherwise stated.
Because our previous pre-acquisition ERβ agonist results suggested that the submissiveness of the OBS was related to the duration of preference, we performed
Pearson’s product-moment correlations between the degree and duration of the preference for
185 the DEM food and submissive behaviour, agonistic behaviour received, and dominance score Table 27. Description of selected performed and calculated behaviours in Experiment 1b, modified from Clipperton et al. (2008). Behaviour Description All behaviours involving activity, both social and non-social. Total Activity Excluded from this group are Inactive Alone, Inactive Together, and (calculated) Self-Groom. Total Social All behaviours in which social interaction takes place (i.e., in which Behaviour two mice are involved), whether agonistic, investigatory or (calculated) affiliative. Agonistic Behaviour Avoidance of the DEM1, Submissive Behaviour, Attack Received, Received (calculated) and Defensive Upright Posture. Avoidance of the OBS2 withdraws and runs away from DEM while the DEM is DEM chasing or following. DEM mouse is in control; includes crawl under, supine posture (ventral side exposed), prolonged crouch, and any other behaviour Submissive Behaviour in which the DEM is dominant (e.g., DEM pins, aggressively grooms, etc, OBS). Defensive Upright Species-typical defensive behaviour; upright with the head tucked Posturing and the arms ready to push away. Total agonistic behaviour delivered (consists of Chasing the DEM, Dominant Behaviour i.e., where DEM is submissive, and Attack Dominance Score Delivered) minus total agonistic behaviour received. A negative (calculated) score indicates that the observer was the submissive animal in the pair, while a positive score signifies that the observer was the dominant animal. Oronasal Investigation Active sniffing of DEM oronasal (face/mouth) area. Risk assessment behaviour; back feet do not move and front feet Stretched Approaches approach demonstrator. Only frequency of stretched approaches was measured. Total Non-social All solitary behaviours (i.e., those in which only the OBS is Behaviour involved), including both active and inactive elements. (calculated) Non-social Locomotor Horizontal Exploration, Vertical Exploration, and Dig. Behaviour (calculated) Horizontal Movement around the cage; includes active sniffing of air and Exploration ground. Movement to investigate upwards, both front feet off the ground; Vertical Exploration includes sniffing, wall leans and lid chews (< 3). 1DEM represents demonstrator. 2OBS represents observer.
186 (as in Clipperton et al., 2008; see Table 27). In addition to correlations pooling all mice, correlations were also performed for each treatment group separately, and our previous data
(Clipperton et al., 2008) was also re-analyzed in this manner.
Only significant results are reported, and all analyses were performed with an α level of
0.05 using PASW 17.0 for Windows (SPSS Inc., an IBM Company, Chicago, IL).
5.4 Results
5.4.1 Experiment 1: Pre-acquisition DPN
5.4.1.1 Food preferences
Results of this experiment showed that all OBS mice preferred their DEM food for the first 2 h interval, but only the middle dose of DPN (0.05 mg/kg) produced a significant DEM food preference at 4 h. This replicates our results with WAY-200070, in which the middle dose (90 mg/kg) prolonged the expression of the preference to 4 h, while all other mice showed only a 2 h preference for the DEM food (Clipperton et al., 2008). In both cases, a relationship between the degree and/or duration of DEM food preference and submissive behaviour and/or agonistic behaviour received was found.
A three-way mixed design ANOVA performed on the CIN preference found a time x
DEM food interaction (F(3,96) = 4.70, p = 0.004) and a main effect of DEM food (F(1,32) =
9.52, p = 0.004). Post-hoc independent samples t-tests showed that the DEM food had a significant effect on the CIN preference at 2 h (t(51) = -5.67, p < 0.001) and 4 h (t(51) = -
2.79, p = 0.007), but not at any other time interval. Planned comparisons of the CIN preference in OBS with CIN-fed DEM and the CIN preference in OBS with COC-fed DEM in each treatment group found a significant effect of DEM food in all groups during the first 187 2 h (vehicle: F(1,11) = 9.57, p = 0.010; 0.025 mg/kg DPN: F(1,12) = 4.81, p = 0.049; 0.05 mg/kg DPN: F(1,11) = 15.09, p = 0.002; 0.1 mg/kg DPN: F(1,11) = 10.00, p = 0.009), but only in the middle dose of DPN did this extend to the 4 h interval (F(1,11) = 6.61, p = 0.026; see Figure 16).
5.4.1.2 Behavioural analysis
There were few drug effects on behaviour of the OBS during the social interaction. The latency to perform Stretched Approaches was increased by 0.05 mg/kg DPN (U = 40.00, z = -
2.28, p = 0.022; see Figure 17a), but no other behaviours were significantly affected by DPN treatment. Additionally, no OBS group showed significant dominance of or submission to the DEM (i.e., no OBS group had a Dominance Score significantly different from zero).
Correlations between the duration and degree of OBS preference for the DEM food and
Submissive Behaviour, Agonistic Behaviour Received and Dominance Score over the 30 min social interaction were not significant when all animals were included. However, when each treatment group was analyzed separately, significant positive correlations between the duration of preference and both Agonistic Behaviour Received (r(11) = 0.570, p = 0.042; see
Figure 17b) and Submissive Behaviour (r(11) = 0.604, p = 0.029; see Figure 17c) were found in the 0.05 mg/kg DPN group, but no significant correlations were observed in any other group.
188 Figure 16. Cinnamon preference by 2h intervals of observers in the pre-acquisition DPN study (Experiment 1b). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers were treated with a) sesame oil vehicle, b) 0.025 mg/kg DPN, c) 0.05 mg/kg DPN, or d) 0.1 mg/kg DPN, 48h prior to the interaction with their demonstrator.
189 Figure 17. Effects of DPN on behaviours in Experiment 1. a) Duration of oronasal investigation. b) Dominance score (duration). @ significant difference between 0.025 mg/kg DPN and vehicle groups, p < 0.05. # significant difference between 0.05 mg/kg DPN and vehicle groups, p < 0.05. + significant difference between 0.1 mg/kg DPN and vehicle groups, p < 0.05.
190 5.4.2 Experiment 2a: Immediate Post-acquisition Administration of PPT
In this experiment, OBS mice treated with vehicle or the low dose of PPT (0.025 mg/kg) showed a significant preference for the DEM food at 2 h, but no preference was observed for mice treated with 0.05 or 0.1 mg/kg PPT.
A three-way mixed ANOVA revealed a significant main effect of DEM food on the CIN preference of OBS mice (F1,47) = 5.558, p = 0.023). In the vehicle and 0.025 mg/kg PPT groups, planned comparisons of the preference for CIN in OBS mice that interacted with
CIN-fed DEM to that of OBS that interacted with COC-fed DEM showed a significant difference in CIN preference in the first 2 h (vehicle: (F(1,18) = 8.24, p = 0.010; 0.025 mg/kg
PPT: F(1,23) = 6.41, p = 0.019), but no other differences were found (i.e., no effect of DEM food was observed at any time for the 0.05 mg/kg PPT and 0.1 mg/kg PPT groups; see Figure
18).
5.4.3 Experiment 2b: Immediate Post-acquisition Administration of DPN
In this experiment, only mice treated with vehicle showed a significant preference for the
DEM food at 2 h. All DPN-treated mice showed no DEM food effects.
A three-way mixed ANOVA found no significant interactions or main effects. Planned comparisons of the CIN preference in OBS with COC-fed DEM to the CIN preference in
OBS with CIN-fed DEM showed a significant difference in the vehicle group during the first
2 h (F(1,25) = 5.99, p = 0.022), but no other differences were observed (i.e., OBS receiving any of the three doses of DPN did not show a difference in CIN preference because of different DEM foods; see Figure 19).
191
Figure 18. Cinnamon preference by 2h intervals of observers in the immediate post- acquisition PPT administration study (Experiment 2a). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa- flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon- flavoured diet (white triangle), p < 0.05. Observers were treated with a) sesame oil vehicle, b) 0.025 mg/kg PPT, c) 0.05 mg/kg PPT, or d) 0.1 mg/kg PPT, immediately after interacting with their demonstrator.
192 Figure 19. Cinnamon preference by 2h intervals of observers in the immediate post- acquisition DPN administration study (Experiment 2b). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers were treated with a) sesame oil vehicle, b) 0.025 mg/kg DPN, c) 0.05 mg/kg DPN, or d) 0.1 mg/kg DPN, immediately after interacting with their demonstrator.
193 5.4.4 Experiment 2c: Immediate Post-acquisition Control Study
Because post-acquisition administration of both PPT and DPN blocked the STFP, this experiment was run to determine if receiving an injection of either drug immediately following exposure to the odour of a food would induce an aversion to that food. Neither a three-way mixed ANOVA nor planned comparisons of CIN preference in mice that investigated COC- or CIN-containing jars found significant effects (data not shown), indicating that treatment did not affect food preference or avoidance when mice investigated the odour of the food in the absence of the DEM.
5.4.5 Experiment 3a: 24h Delayed Post-acquisition Administration of PPT
Results of this experiment showed that OBS mice treated with vehicle, 0.05 or 0.1 mg/kg
PPT only showed a significant preference for the DEM food at 2 h, while the 0.025 mg/kg
PPT group showed no food preference at any time.
A three-way mixed ANOVA found a significant time x DEM food interaction (F(3,93) =
3.96, p = 0.010). Planned comparisons of the CIN preference in OBS with COC-fed DEM to the CIN preference in OBS with CIN-fed DEM showed differences in CIN preference in the vehicle group (F(1,19) = 5.10, p = 0.036), the 0.05 mg/kg PPT group (F(1,25) = 5.11, p =
0.033), and the 0.1 mg/kg PPT group (F(1,19) = 22.48, p < 0.001) at 2 h, but no other differences (see Figure 20).
5.4.6 Experiment 3b: 24h Delayed Post-acquisition Administration of DPN
In this experiment, OBS mice treated with vehicle and 0.025mg/kg DPN preferred the
DEM food for 2 h, and mice administered 0.1 mg/kg DPN showed a prolonged preference which was expressed during the 2 h and 4 h intervals. However, the 0.05 mg/kg dose of 194 Figure 20. Cinnamon preference by 2h intervals of observers in the 24h delayed post- acquisition PPT administration study (Experiment 3a). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa- flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon- flavoured diet (white triangle), p < 0.05. Observers were treated with a) sesame oil vehicle, b) 0.025 mg/kg PPT, c) 0.05 mg/kg PPT, or d) 0.1 mg/kg PPT, 24h after interacting with their demonstrator.
195 DPN blocked the socially acquired food preference.
A three-way mixed ANOVA found a significant time x DEM food interaction (F(3,222)
= 3.62, p = 0.014), and a significant main effect of DEM food (F(1,74) = 4.14, p = 0.045) on the OBS preference for CIN. Independent samples t-tests showed that the DEM food affected the preference for CIN at 2 h (t(93) = -3.66, p < 0.001) and 4 h (t(94) = -2.29, p =
0.024), but at no other time-point. Planned comparisons of the CIN preference in the OBS of
COC- and CIN-fed DEM showed that the vehicle (F(1,23) = 5.15, p = 0.033), 0.025 mg/kg
DPN (F(1,19) = 7.26, p = 0.014) and 0.1 mg/kg DPN (F(1,25) = 4.99, p = 0.035) groups showed a difference in the first 2 h, with the 0.1 mg/kg DPN group showing an extended preference that lasted for 6 h (4 h: F(1,25) = 4.53, p = 0.043; 6 h: F(1,25) = 5.51, p = 0.027), while the 0.05 mg/kg DPN dose showed no evidence of social learning at any time point (see
Figure 21).
5.4.7 Effects on Food Intake
In all experiments, food intake was affected by time (all F > 56.00, all p < 0.001) and post-hoc tests revealed this was due to decreased intake between all time intervals, i.e., 2 h-4 h, 4 h-6 h, 6 h-8 h, and 2 h-8 h (all p < .005; see Figure 22). There were no treatment effects on food intake in any of the PPT or DPN social learning experiments. In the individual learning control study (Experiment 2c) we found minor drug effects on food intake, described below.
5.4.7.1 Immediate Post-acquisition Individual Learning Control Study
We found that the 0.025 and 0.1 mg/kg doses of DPN increased total consumption during the 4 h interval, but no effects of PPT on consumption at any time.
196 Figure 21. Cinnamon preference by 2h intervals of observers in the 24h delayed post- acquisition DPN administration study (Experiment 3b). Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers were treated with a) sesame oil vehicle, b) 0.025 mg/kg DPN, c) 0.05 mg/kg DPN, or d) 0.1 mg/kg DPN, 24h after interacting with their demonstrator.
197 Figure 22. Total food intake (g) by 2h intervals over the 8h choice test. Means and standard errors are shown. a) Experiment 1a: Delays between social interaction and test were 0h (familiar demonstrator (DEM): black square, unfamiliar DEM: white square), 90 min (familiar DEM: black triangle, unfamiliar DEM: white triangle), 48h (familiar DEM: black circle, unfamiliar DEM: white circle), 7 days (familiar DEM: black diamond, unfamiliar DEM: white diamond), or 30 days (familiar DEM: black hexagon). b) Experiment 1b: OBS received sesame oil vehicle (black square), 0.025 mg/kg DPN (white circle), 0.05 mg/kg DPN (white triangle), or 0.1 mg/kg DPN (white square). c) Experiment 2a: Observers (OBS) received sesame oil (black square), 0.025 mg/kg PPT (white diamond), 0.05 mg/kg PPT (white triangle), or 0.1 mg/kg PPT (white hexagon). d) Experiment 2b: OBS received sesame oil vehicle (black square), 0.025 mg/kg DPN (white circle), 0.05 mg/kg DPN (white triangle), or 0.1 mg/kg DPN (white square). e) Experiment 3a: OBS received sesame oil (black square), 0.025 mg/kg PPT (white diamond), 0.05 mg/kg PPT (white triangle), or 0.1 mg/kg PPT (white hexagon). f) Experiment 3b: OBS received sesame oil vehicle (black square), 0.025 mg/kg DPN (white circle), 0.05 mg/kg DPN (white triangle), or 0.1 mg/kg DPN (white square). g) Experiment 2c: Experimental mice received sesame oil vehicle (black square), 0.025 mg/kg DPN (white circle), 0.05 mg/kg DPN (white 'down' triangle), 0.1 mg/kg DPN (white square), 0.05 mg/kg PPT (white 'up' triangle), or 0.1 mg/kg PPT (white hexagon). *significant difference between sesame oil vehicle and 0.025 mg/kg DPN groups. †significant difference between sesame oil vehicle and 0.1 mg/kg DPN groups. 198 A two-way mixed ANOVA performed on the total intake found a significant time x drug interaction (F(15,405) = 3.88, p < 0.001), a significant main effect of time (F(3,405) =
112.25, p < 0.001), and a significant main effect of drug (F(5,135) = 4.29, p = 0.001; see
Figure 22b). Post-hoc one-way ANOVAs found a significant effect of drug at 2 h (F(5,135)
= 3.48, p = 0.005), 4 h (F(5,135) = 7.86, p < 0.001), 6 h (F(5,135) = 3.03, p = 0.013) and 8 h
(F(5,135) = 2.82, p = 0.019). Multiple comparison post-hoc tests found no differences in intake between vehicle and any treatment group at 2 h, 6 h, or 8 h, but significant increases in intake were found at 4 h by 0.025 mg/kg DPN (p < 0.001) and 0.1 mg/kg DPN (p = 0.001).
5.5 Discussion
We found that OBS treated with the ERα agonist PPT did not show a socially acquired food preference, regardless of the timing of drug administration. OBS that were given the
ERβ agonist DPN immediately following the social interaction also failed to show a STFP, as did OBS receiving the middle dose of DPN 24 h after the interaction. However, mice administered the high dose of DPN 24 h after the interaction expressed a preference for the
DEM food for three times as long as controls Control studies demonstrated that the immediate post-acquisition blockage of the STFP was not due to either PPT or DPN inducing a conditioned avoidance of the DEM food.
5.5.1 Effects of PPT on Social Learning
Mice treated with PPT were impaired in the STFP regardless of the time at which the drug was administered (Figures 18, 20). While our previous finding that pre-acquisition PPT impaired learning (Clipperton et al., 2008) was consistent with the non-social learning literature, the results of post-acquisition PPT treatments have been much less consistent,
199 resulting in improved (e.g., Walf et al., 2006), unaffected (e.g., Rhodes & Frye, 2006), or impaired (e.g., Hebert & Dash, 2004; Igaz et al., 2006) learning and memory. However, the effects of estrogen agonist/antagonist tamoxifen, which may have partial agonist effects at
ERα, and may be a pure ERβ antagonist (Barkhem et al., 1998), are consistent with our findings. When tamoxifen was administered either immediately or 24 h post-acquisition, it impaired passive avoidance learning (Chen et al., 2002), and both immediate and delayed post-acquisition administration also impaired morphine-induced conditioned place preference, which is, like the STFP, an appetitively motivated type of associative learning
(Esmaeili et al., 2009).
There are a number of possible explanations for the observed inhibition of the STFP by
PPT. ERα may interfere with consolidation by acting in the hippocampus, an area required for STFP learning (e.g., Alvarez et al., 2002). ERα is expressed in γ-aminobutyric acid
(GABA)-ergic inhibitory interneurons in the hippocampus ( McEwen et al., 2001; Spencer et al., 2008), thus modulating inhibitory tone in the hippocampus (Donner & Handa, 2009b;
McEwen et al., 2001) and possibly impairing memory processing. ERα activation inhibits
ERK phosphorylation (Singh et al., 2000), which has been linked to long- and short-term memory formation (Hebert & Dash, 2004; Igaz et al., 2006). PPT thus may block the STFP, as well as other types of learning (e.g., fear-mediated, spatial, food-reward, etc; Hebert &
Dash, 2004; Igaz et al., 2006) by interfering with memory formation in the hippocampus.
Alternatively, while all phases of memory could be affected by the treatment independently,
PPT’s long-lasting effects (e.g., Jacome et al., 2010b), taken together with the results of our current and previous studies demonstrating that PPT-treated mice showed no social learning whether the drug was administered pre-acquisition (Clipperton et al., 2008), immediately
200 post-acquisition (Figure 18), or 24 h post-acquisition (Figure 20), suggests that ERα could be impairing this paradigm through action on the retrieval or expression of the memory.
Memory retrieval, and/or other aspect(s) of the STFP, may be impaired due to PPT increasing long term depression in the hippocampus (Choleris et al., 2008a). The STFP in
PPT-treated mice thus may be blocked either by acting to impair retrieval of the memory, or by affecting consolidation and/or all phases of memory, possibly through different mechanisms.
Another mechanism by which PPT could be affecting the STFP is through modulation of the ACh system. Estrogens have been shown to affect ACh and its synthesizing enzyme, choline acetyltransferase, in mice, and it has been hypothesised that ERα may regulate ACh release and reuptake (Spencer et al., 2008). Basal forebrain regions, including the medial septum (MS) and the diagonal band of Broca (DB; specifically the vertical limb), project to the hippocampus, and cholinergic lesions of these areas impaired retrieval of the STFP (Vale-
Martínez et al., 2002). ACh also appears to mediate the STFP in other brain regions, as cholinergic lesions of the orbitofrontal cortex (Ross et al., 2005), or to the ventral tegmental area, the prelimbic cortex, and the ventral hippocampus immediately after the social interaction (Boix-Trelis et al., 2007; Carballo-Márquez et al., 2009a) impaired or blocked performance on the STFP. Further, when scopolamine was infused the prelimbic cortex or the ventral hippocampus immediately after the social interaction (Carballo-Márquez et al.,
2009b), the STFP was also blocked or impaired. Similarly, in an inhibitory avoidance task, administration of scopolamine to the ventral tegmental area impaired both consolidation, when administered immediately following acquisition, and retrieval, when administered pre- test (Mahmoodi et al., 2010).
201 It is interesting to note that different doses of PPT were effective in the current studies.
One possible explanation is that more ERα activation may be required to affect GABA and
ERK phosphorylation than the ACh system. This is consistent with the brain distribution of
ERα in the mouse, as ERα may be more highly expressed in the cholinergic basal forebrain than in the hippocampus (Merchenthaler et al., 2004a; Mitra et al., 2003a), which suggests that lower doses of PPT might have a greater effect in the former region than the latter..
5.5.2 Effects of DPN on Social Learning
5.5.2.1 Pre-acquisition ERβ Agonists
Our previously study with WAY-200070 (Clipperton et al., 2008), as well as the results of Experiment 1 with DPN (Figure 17), found that mice receiving pre-acquisition treatment with either ERβ agonist showed a prolonged preference for the DEM food, and that both ERβ agonists affected social behaviour, suggesting that ERβ may modulate the acquisition of this memory by altering the nature of the social interaction during which it occurs. Further analysis revealed that for both ERβ agonists, those groups that showed prolonged preferences for the DEM food, and only those groups, showed relationships between submissive and/or received agonistic behaviour and the duration or strength of the STFP. Reanalysis of the
WAY-200070 data, correlating behaviour with social learning of food preference measures revealed that in mice receiving the effective dose of WAY-200070, 90 mg/kg (Clipperton et al., 2008), the Dominance Score (which was significantly below zero, indicating that the
OBS were submissive to the DEM) was inversely related to the degree of preference at 4 h, while Agonistic Behaviour Received was positively associated with the degree of preference at both 2 h and 4 h. In other words, as the Dominance Score decreased (i.e., the OBS became more submissive to the DEM) and the Agonistic Behaviour Received by the OBS increased,
202 the degree of socially acquired food preference increased. This is consistent with other studies showing modulatory effects of social interactions on the STFP (e.g., Choleris et al.,
1998a) and other types of social learning, where subordinate OBS learned better than dominant OBS (Kavaliers et al., 2005b), discussed in Clipperton et al. (2008), as well as with other social paradigms (e.g., social recognition) that have found a modulatory role for ERβ
(reviewed in Choleris et al., 2009).
5.5.2.2 Post-acquisition DPN Effects
As with PPT, DPN-treated mice showed impaired social learning when injected immediately following the social interaction, but despite some evidence of estrogenic induction of taste avoidance (e.g., Miele et al., 1988), neither this nor PPT’s effects were due to an avoidance of the odour of the food used in our studies (Experiment 2c).
In our current results, DPN appeared to interfere with memory consolidation when administered immediately following the social interaction and acquisition. Memories are vulnerable to both interference and enhancing effects at this time (Abel & Lattal, 2001b), and several intracellular systems known to be mediated by estrogens are involved in consolidation processes. For example, CREB inactivation has been shown to impair memories if it occurs shortly after the acquisition phase (Guzowski & McGaugh, 1997).
Both PPT and DPN have been shown to inhibit some types of CREB phosphorylation
(Boulware et al., 2005), and CREB KO mice may be impaired in the STFP when tested 24 h after the social interaction (Kogan et al., 1996; but see Gass et al., 1998). Estrogens also modulate PKA activity and affect MAPK activation (Boulware et al., 2005; Spencer et al.,
2008), which are necessary for consolidation (Boulware et al., 2005).
203 Another possible mechanism by which DPN might be impairing the STFP is through increasing levels of the rate limiting enzyme for 5-HT production, tryptophan hydroxylase 2
(Donner & Handa, 2009b). Long term potentiation (LTP) is blocked by 5-HT, and hippocampal activity is influenced by 5-HT in the MS and DB (Leranth et al., 1999), areas known to be involved in the STFP (Vale-Martínez et al., 2002). This suggests that ERβ- mediated increases in 5-HT activity may be related to the STFP impairment observed in mice receiving DPN immediately post-acquisition.
When the highest dose of DPN was administered 24 h after the acquisition, the OBS preference for the demonstrated food lasted for 3 times as long as that of the vehicle group.
These results are similar to those observed in mice in the high estrogen, low progesterone proestrous and diestrous phases of the estrous cycle (Choleris et al., 2011a), and with pre- acquisition ERβ agonist administration (Clipperton et al., 2008). This supports the interpretation that the prolonged preference for the DEM food shown by proestrous and diestrous females may be due to acting primarily through ERβ in the regulation of this task
(Choleris et al., 2011a; Clipperton et al., 2008).
ERβ activation may have induced prolonged DEM food preferences directly by enhancing LTP in the hippocampus (Milner et al., 2010), and/or by acting on the monoamine neurotransmitters dopamine and norepinephrine, which affect LTP (reviewed in Kirkwood,
2007). Both norepinephrine and dopamine are involved in memory processes and social behaviour, including the STFP (Choleris et al., 2011a; Marino et al., 2005; Ross et al., 2005;
Sánchez-Andrade et al., 2005; Shea et al., 2008; Vale-Martínez et al., 2002). Norepinephrine is necessary for retrieval of a memory (Murchison et al., 2004), and DPN increases norepinephrine activity in a number of brain regions, including the STFP-relevant
204 hippocampus and prefrontal cortex (Alvarez et al., 2002; Jacome et al., 2010b). Dopamine has also been shown to mediate the STFP via the D1-like receptors (Choleris et al., 2011a), and this could be modulated by ERβ increasing dopamine release from the ventral tegmental area (Liu & Xie, 2004; McDermott et al., 1994), which expresses ERβ but lacks ERα (Mitra et al., 2003b), or by decreasing dopamine reuptake from the extracellular space by inhibiting the dopamine transporter (reviewed in Karakaya et al., 2007). This involvement of dopamine suggests that the socially learned food preference, or the interaction during which it is acquired, may be rewarding.
It is unclear why the middle and high doses of DPN would have opposing effects when administered 24 h after acquisition, although dopamine transporter knock-out mice, which have chronically elevated dopamine levels, do show a reversed preference in the STFP
(Rodriguiz et al., 2004). One possibility is that that the different dose effects may be related to distribution of the ERβ receptors in the brain, such that areas that contain more receptors may be affected at different doses than those containing fewer receptors. It is worth noting that the biphasic pattern has also been observed with hormones and social behaviours.
Medium and low doses of estradiol improved non-social learning in female mice lacking
ERα, but high estradiol doses impaired it, possibly through actions on ERβ (Rissman et al.,
2002a), and oxytocin, whose production is regulated by estrogens via ERβ, shows the same pattern in its effects on social recognition, impairing at high doses and improving at low ones
(reviewed in Choleris et al., 2009).
Both oxytocin and ERβ have anxiolytic effects, and this may relate to their biphasic effects (Choleris et al., 2008a; Donner & Handa, 2009a; Jacome et al., 2010a). The
205 benzodiazepine anxiolytic chlordiazepoxide (CDP) showed a similar biphasic pattern of effects on the STFP in gerbils. When administered 2.5 mg/kg CDP 30 min prior to the social interaction, male (but not female) gerbils showed a preference for the food not consumed by their DEM, but when treated with 5 mg/kg CDP, both males and females showed an increased preference for the DEM food (Choleris et al., 1998). It is possible that biphasic
DPN effects in mice would be similar to the effects of benzodiazepine/GABAA activation in gerbils (Choleris et al., 1998a).
5.5.3 Overall Conclusions
Activation of ERα appears to always block social learning, as OBS treated with PPT had impaired social acquisition of food preferences regardless of the time of administration.
These impairments may be through ERα effects on learning processes in the hippocampus or other areas. ERβ’s involvement in social learning is more complex, as OBS receiving ERβ agonists the social interaction show a prolonged DEM food preference, possibly due to ERβ promoting acquisition through its effects on social behaviour and/or the dopamine system
(Choleris et al., 2011b; Clipperton et al., 2008). When ERβ agonist DPN was administered immediately following the social interaction, it may have interfered with consolidation or short-term memory, while when administered 24 h later, treated OBS show either blocked or prolonged expression of the socially acquired food preference, depending on the dose they received. It is currently unclear why different doses have opposing effects on the paradigm, but it may be related to DPN’s anxiolytic effects. Taken together, these findings suggest that
ERα activation interferes with social learning, while ERβ may play a more modulatory role.
Further, these results suggest that while ERβ treatment may help with the formation of new memories, it is less likely to prove beneficial in treating deficits in previously-established
206 memories.
207 .
CHAPTER 6:
Effects of chronic estrogen and estrogen receptor agonists on the behaviour of
ovariectomized female mice in the social transmission of food preferences paradigm
This manuscript will be submitted as: Clipperton-Allen, A.E., Mikloska, K.V., Roussel, V.R.,
Power, K.A., and Choleris, E. Effects of chronic estrogen and estrogen receptor agonists
on the behaviour of ovariectomized female mice in the social transmission of food
preferences paradigm.
208 6.1 Abstract
The ability to learn from conspecifics, thus reducing the reliance on risky trial-and-error learning, is one of the advantages of social life. The social transmission of food preferences
(STFP) paradigm assesses a specifically social type of learning, in which “observer” mice interact with “demonstrators” that have eaten a novel food, and then prefer that food when given a subsequent choice test. Previous research indicates that acute pre-acquisition administration of estradiol benzoate (EB) or estrogen receptor beta (ERβ) agonist prolonged the preference for the demonstrated food, while this preference was blocked by acute administration of ERα agonist. Effects of estrogenic compounds are often assessed using chronic rather than acute treatments, so in the current study, ovariectomized (ovx) observer mice received Silastic capsules containing EB, ERα agonist PPT, ERβ agonist DPN, sesame oil vehicle, or no implant 9-17 days prior to testing. Following a 30 min interaction with a demonstrator that had just consumed either cocoa- or cinnamon-flavoured rodent chow, observers were given an 8h choice test in which the two flavours (both novel to the observer) were continuously available. Results indicate that chronic ERα and ERβ agonists, as well as chronic EB, prolong the preference for the demonstrator’s food, suggesting that the long-term exposure to these compounds may result in a loss of specificity. Additionally, our current and previous results suggest that ERβ effects may prevail over those of ERα in the STFP.
209 6.2 Introduction
Numerous animals, including humans (Morgan et al., 2011), possess the ability to avoid trial-and-error learning, and any associated potential costs, by acquiring information from conspecifics. Social learning is one of the advantages of social living, and has been shown to be increasingly relied upon when humans are uncertain or when incorrect choices carry large costs (Morgan et al., 2011). Additionally, knowledge acquired from conspecifics often has stronger effects than individual learning (e.g., Galef, 1986; Galef et al., 1997; Morgan et al.,
2011), but the neurobiological mechanisms underlying social learning have been less well characterized.
One way to investigate these mechanisms in animal models is by using the social transmission of food preferences (STFP) paradigm, which has been employed to assess social learning in a number of species, including mice (Mus musculus; Valsecchi & Galef, 1989) and rats (Rattus norvegicus; Galef et al., 1988; Galef & Wigmore, 1983). This paradigm involves the interaction of an “observer” (OBS) animal with a “demonstrator” (DEM) that ate a flavoured food immediately prior to the interaction. Following this social interaction,
OBS is given a choice between two novel flavours, one of which had been consumed by its
DEM; typically, OBS will prefer the food consumed by DEM (Galef & Wigmore, 1983).
Research by Galef and others (e.g., Choleris et al., 2011a; Galef et al., 1988; Galef &
Wigmore, 1983) has shown that interacting with the food odour alone or in the presence of an unflavoured-chow-fed conspecific, or with a food powdered surrogate, are insufficient to elicit this preference, indicating that this is a specifically social learning.
Work from our laboratory has previously shown that estrogens and their intranuclear
210 receptors alpha (ERα) and beta (ERβ) mediate the STFP (Clipperton et al., 2008; Clipperton-
Allen et al., 2011c; Clipperton-Allen et al., 2011b). ERα and ERβ are encoded by different genes (Kuiper et al., 1996), distributed differently in the brain (Merchenthaler et al., 2004b;
Mitra et al., 2003b), expressed differently in development (Kudwa et al., 2006), and affect some learning and memory paradigms differently (e.g., Clipperton et al., 2008; see Choleris et al., 2008a for a review). In the STFP, acute treatment with ERα agonist propyl pyrazole triol (PPT) before, immediately after or 24h after the acquisition (social interaction) phase of the task blocked social learning, while treatment with the estrogen estradiol benzoate (EB) or
ERβ agonists induced a prolonged preference for DEM food when administered prior to the acquisition phase (Clipperton et al., 2008; Clipperton-Allen et al., 2011c; Clipperton-Allen et al., 2011b). EB also prolonged the preference for DEM food when administered immediately after the interaction, but had no effect when administration was delayed for 24h, while ERβ agonist diarylpropionitrile (DPN) blocked social learning when administered immediately post-interaction, but blocked or prolonged the socially acquired food preference when administered 24h later, depending on the dose (Clipperton-Allen et al., 2011c; Clipperton-
Allen et al., 2011b). Taken together, these results suggest that EB may act primarily through
ERβ and may affect the consolidation and possibly the acquisition phases of the STFP memory, ERβ agonists may differently affect the paradigm depending on the stage of learning and memory at which they are present, and ERα activation in the absence of ERβ function may impair the expression or retrieval of the STFP memory, possibly in addition to blocking acquisition and/or consolidation (Clipperton et al., 2008; Clipperton-Allen et al.,
2011c; Clipperton-Allen et al., 2011b).
Varying effects of estrogens have also been shown in other, non-social, learning tasks.
211 Depending on the nature of the task, the administration regimen, and numerous other experimental and procedural factors, estrogens may impair, improve or have no effect on learning and memory (see Choleris et al., 2008a; Daniel, 2006 for recent reviews). Some of these differences in results may be due to different effects of chronic and acute treatments.
For example, acute estrogens treatment impaired (Galea et al., 2001; Holmes et al., 2002b) or did not affect (Holmes et al., 2002b; Luine et al., 1998) performance on the radial arm maze, while chronic estrogens treatment improved it (Bowman et al., 2002; Daniel et al., 2006b;
Daniel et al., 1997; Davis et al., 2005; Heikkinen et al., 2004; Luine et al., 1998) or had no effect (Daniel et al., 2006a; Daniel et al., 2006b; Luine & Rodriguez, 1994).
Whether chronic estrogens and ER-specific agonists affect social learning, and the effects of chronic ER-specific agonists on social behaviour in general, is currently unknown. This is of particular importance given that estrogen-based hormone replacement therapy (HRT) is administered chronically to women, and because one of the reasons that women seek HRT is to treat the cognitive, emotional and social symptoms of menopause (Kumari et al., 2005).
Because of the negative side effects, like increased risks of uterine and breast cancer
(Koehler et al., 2005), that accompany treatment with HRT using estrogens, there is an increasing interest in the use of ER-selective agonists to treat cognitive and social impairment and other symptoms of menopause, while avoiding the negative consequences of these treatments.
However, little is known of the effects of chronic selective activation of ERα and ERβ.
To the best of our knowledge, only one study has investigated their effects on behaviour.
This study found that estradiol, PPT and DPN all improved rats’ performance on a delayed
212 matching-to-position T-maze task (Hammond et al., 2009). Thus, more study of the effects of treatments on behaviour is needed, especially in light of findings that chronic and acute estrogens differently affect some tasks (e.g., radial arm maze; Bowman et al., 2002; Daniel et al., 2006b; Daniel et al., 1997; Davis et al., 2005; Galea et al., 2001; Heikkinen et al., 2004;
Holmes et al., 2002b; Luine et al., 1998; Luine & Rodriguez, 1994).
The current study, therefore, examined the effects of chronic treatment with the estrogen
EB, ERα agonist PPT, or ERβ agonist DPN on the STFP. In addition to those on food preferences, we also assessed the effects of EB, PPT and DPN on the social interactions between OBS and DEM, which were analyzed using a comprehensive ethological analysis, as we have previously shown that the nature of the interaction can be related to the subsequent socially acquired food preference (Clipperton et al., 2008; Clipperton-Allen et al.,
2011b). Previous studies have found that PPT can dose-dependently increase uterus weight in prepubertal and OVX rodents, while DPN elicits minimal effects on uterine growth
(although these studies used higher doses than the current study, and these drugs were administered acutely instead of chronically; Frasor et al., 2003; Harris et al., 2002b).
Therefore, uterus growth was assessed by histomorphometric analysis.
6.3 Methods
6.3.1 Subjects
Subjects were obtained from Charles River, QC, Canada and consisted of 442 female
CD1 mice (Mus musculus), experimentally naïve and 2.5-4 months old. Fifty mice were randomly chosen to be DEM, while the remainder were used as OBS. Those OBS receiving no treatment were also re-used as DEM (see Table 28 for experimental treatment conditions).
213 Table 28. Distribution of treatment groups. Numbers indicate the number of observer Drug Dose n of OBS1 Sham COC DEM2: n = 11, CIN DEM3: n = 12 Sesame Oil Vehicle COC DEM: n = 16, CIN DEM: n = 16 Chronic Estradiol 1.25 µg COC DEM: n = 12, CIN DEM: n = 12 Benzoate 12.5 µg COC DEM: n = 14, CIN DEM: n = 13 25 µg COC DEM: n = 12, CIN DEM: n = 11 50 µg COC DEM: n = 12, CIN DEM: n = 13 Chronic ER Agonist Sham COC DEM: n = 11, CIN DEM: n = 11 Control Groups DPN Vehicle COC DEM: n = 9, CIN DEM: n = 11 3 µg COC DEM: n = 11, CIN DEM: n = 12 Chronic ERα Agonist 6 µg COC DEM: n = 15, CIN DEM: n = 15 PPT 12 µg COC DEM: n = 11, CIN DEM: n = 11 24 µg COC DEM: n = 12, CIN DEM: n = 12 3 µg COC DEM: n = 11, CIN DEM: n = 11 Chronic ERβ Agonist 6 µg COC DEM: n = 12, CIN DEM: n = 12 DPN 12 µg COC DEM: n = 12, CIN DEM: n = 12 24 µg COC DEM: n = 13, CIN DEM: n = 14 1OBS represents observer. 2COC DEM indicates that OBS interacted with a demonstrator (DEM) that consumed the cocoa diet. 3CIN DEM indicates that OBS interacted with a DEM that consumed the cinnamon diet.
214 Following surgery, mice were singly housed until at least 3 days before testing, at which time each OBS was paired with a DEM. In all cases, mice were housed in clear polyethylene cages (26 x 16 x 12 cm), provided with corn cob bedding, environmental enrichment, and ad libitum access to food (Teklad Global 14% Protein Rodent Maintenance Diet, Harlan Teklad,
Madison, WI) and tap water, and maintained at 21±1°C under a 12:12h reversed light/dark cycle (lights on 8:00 pm). This research was approved by the University of Guelph
Institutional Animal Care and Use Committee and conducted in accordance with the regulations of the Canadian Council on Animal Care.
6.3.2 Surgery
All mice were bilaterally ovariectomized, as described in Clipperton-Allen et al. (2011b).
Briefly, while under isoflurane gas anaesthesia (Benson Medical Industries, Markham, ON), a single dorsal incision in the skin was made. Ovaries were removed through two small incisions over the dorsal muscles, and OBS mice were implanted with a subcutaneous
Silastic capsule along the left dorsolateral side. Mice in the sham group were ovariectomized, but received no capsule. The skin incision was stapled with 1-2 MikRon
Autoclip 9 mm wound clips (MikRon Precision Inc, Gardena, CA). Mice were allowed at least 7 days of recovery in single housing before pairing into OBS-DEM dyads.
6.3.3 Drugs
Treatments were administered via Silastic capsules, made of 3 cm long pieces of Silastic laboratory tubing (1.98 mm inner diameter, 3.18 mm outer diameter; Dow Corning
Corporation, Midland, MI, USA) with 0.5 cm of Silastic medical adhesive (silicone type A,
Dow Corning Corporation, Midland, MI, USA) at each end; the total drug reservoir was thus
215 2 cm long. Capsules were filled with 0.7 ml of sesame oil vehicle (estradiol benzoate [EB],
PPT) or sesame oil + approximately 0.04% EtOH vehicle (DPN) or EB, ERα agonist PPT or
ERβ agonist DPN. Mice were implanted at the time of ovariectomy surgery, 9-17 days before testing. Treatment distributions are outlined in Table 28.
The estrogen EB was chronically administered at one of four doses (1.25 µg, 12.5 µg, 25
µg, 50 µg) that correspond to levels of estradiol ranging from low to above-proestrus, and have been shown to increase running wheel activity, induce sexual receptivity, and improve spatial learning (Hammond et al., 2009; Morgan et al., 2000; Ribeiro et al., 2009).
The ERα agonist 1,3,5-tris(4-hydroxyphenyl)-4-propyl-1H-pyrazole (PPT) is 410 times more selective for ERα than ERβ in mice (Stauffer et al., 2000), while ERβ agonist 2,3-bis(4-
Hydroxyphenyl)-propionitrile (DPN) is 78 times more selective for ERβ than ERα (Meyers et al., 2001). The dissolution of DPN was facilitated by the addition of a small amount of ethanol (less than 0.05%), and the same amount of ethanol was added to the appropriate vehicle control. Both drugs were administered in a dose range (3 µg, 6 µg, 12 µg, 24 µg; approximately 2.9, 5.7, 11.4 and 22.9ug/kg/day in mice) that extended above and below the dose used in the behavioural study of Hammond, et al. (2009); approximately 17ug/kg/day in rats), and also had the same relationship to acute PPT and DPN that chronic EB did to acute
EB (Clipperton-Allen et al., 2011c; Clipperton-Allen et al., 2011b).
6.3.3 Diets
Regular ground rodent chow, mixed with either 1% ground cinnamon (CIN; McCormick
Ground Cinnamon, McCormick Canada, London, Canada) or 2% powdered cocoa (COC;
Fry’s Premium Cocoa, Cadbury Ltd., Mississauga, Canada) by weight, were used for both 216 DEM feeding and OBS choice tests. We have previously found these flavoured diets to be equipalatable to other female CD1 mice from Charles River, QC (Choleris et al., 2011a;
Clipperton-Allen et al., 2011b).
6.3.4 Apparatus
DEM feeding took place in clean, empty polyethylene cages (26 x 16 x 12 cm), into which were placed 5 cm high, 7.5 cm diameter cylindrical jars (Dyets Inc, Bethlehem, PA), fitted with a sleeve collar and stainless steel ring (hole 2.5 cm in diameter) to prevent spillage, and containing one of the two flavoured diets. Clear Plexiglas lids with air holes were used during the videotaped social interactions between OBS and DEM, which took place in their home cages. Polyethylene cages (42.5 x 26.5 x 18.5 cm) with stainless steel wire lids and ad libitum water access were used for OBS choice tests. These cages were equipped with two plastic feeders (Tecniplast, Varese, Italy) that each held one of the flavoured diets (as described in Valsecchi & Galef, 1989) and allow for very precise food consumption measurements; using a balance accurate to 0.01 g (Sartorius Analytical
Balance, Sartorius Inc., United Kingdom), these feeders were weighed every 2h during the choice test.
6.3.5 Procedure
Mice were moved into the experimental room, food deprived, the fur of DEM was coloured with black magic marker at least 12h before testing, and then left undisturbed.
Early in the active phase, DEM were placed into a cage containing a jar of either COC- or
CIN-flavoured food for 1h. Data from OBS whose DEM ate less than 0.1 g (n = 2) were removed from the experiment. Immediately following the consumption of one of the
217 flavoured foods, DEM were returned to their home cage to interact for 30 min with their
OBS. These interactions were videotaped for subsequent behavioural analysis, and were immediately followed by placement of OBS into the testing cages for an 8h continuous choice test between the two diets. Consumption of each diet was assessed every 2h.
Videotaped interactions were scored by trained observers (average inter-observer reliability correlations = 0.98) as described previously (Clipperton et al., 2008; Clipperton-
Allen et al., 2011c; Clipperton-Allen et al., 2011b). Briefly, an ethological analysis based on the ethogram of Grant and Mackintosh (Grant & Mackintosh, 1963) and using The Observer
Video Analysis software (Noldus Information Technology, Wageningen, The Netherlands) collected 21 behaviours focused on OBS’s behaviour (see Table 29, modified from
Clipperton et al., 2008); DEM’s behaviour was only collected relative to that of OBS.
Composite behavioural categories were also calculated and provided information about more general drug effects (see Table 29, modified from Clipperton et al., 2008).
Early in the active (dark) phase, approximately 1h prior to the experiment, vaginal swabs were taken from all mice and subsequently Giemsa stained (Sigma-Aldrich, Oakville, ON) and examined under a microscope at 100x magnification to assess effects of treatment on vaginal cytology. Diestrus was defined as consisting primarily of leucocytes, proestrus was classified as predominantly nucleated cells, and estrus consisted of primarily non-nucleated, cornified cells (as in Clipperton et al., 2008).
6.3.6 Histomorphometric Analysis
Uteri were collected after CO2 asphyxiation, cleaned of soft tissue, formalin fixed, paraffin embedded, and cross sectioned (5µm) using a Thermo Scientific Finesse 325 Manual 218 Table 29. Behaviours during the social interaction. Scored Behaviour Description Composite(s) Behaviours OBS mouse is in control; includes pinning of DEM, aggressive Dominance Behaviours grooming, crawling over or on top, and mounting attempt. Agonistic Behaviour OBS mouse actively follows, or Delivered, Total Agonistic Chasing DEM1 pursues and chases DEM; Behaviour, Total Social reciprocal to Avoiding DEM. Behaviour, Total Activity Physical attacks, including dorsal/ventral bites. Only the Attack Delivered frequency of attacks was measured. DEM mouse is in control; includes crawl under, supine posture (ventral side exposed), prolonged Submissive Behaviours crouch, and any other behaviour in which DEM is dominant (e.g., DEM pins, aggressively grooms, etc., OBS). Agonistic Behaviour OBS withdraws and runs away Avoiding DEM Received, Total Agonistic from DEM while DEM is chasing. Behaviour, Total Social Physical attacks including bites to Behaviour, Total Activity dorsal/ventral regions. Only the Attack Received frequency of attacks was measured. Species-typical defensive Defensive Upright behaviour; upright with the head Posture tucked and the arms ready to push away. Physical attacks with a locked fight including tumbling, kick- Reciprocal Attacks away and counterattack where the Total Agonistic Behaviour, attacker cannot be identified. Total Social Behaviour, Physical attacks which include Total Activity box/wrestle, offensive and Aggressive Postures defensive postures, lateral sideways threats and tail rattle. Agonistic Behaviour Delivered – N/A Dominance Score Agonistic Behaviour Received Active sniffing of DEM’s oronasal Oronasal Investigation area. Social Investigation, Total Body Investigation Active sniffing of DEM’s body. Social Behaviour, Total Active sniffing of DEM’s Activity Anogenital Investigation anogenital region.
219 Often from across the cage; OBS’s attention is focused on DEM, head tilted toward DEM and Approach and/or Attend movements toward DEM; this to DEM becomes “Chasing DEM” once Social Investigation, Total along the tail or sniff if within 1.5 Social Behaviour, Total cm of DEM. Activity Risk assessment behaviour; back feet do not move and front feet Stretch Approach approach DEM. Only the frequency of stretched approaches was measured. Social Inactivity Includes sit/lie/sleep together. Total Social Behaviour Movement to investigate upwards, both front feet off the ground; Vertical Exploration includes sniffing, wall leans and lid chews (less than 3). Total Non-social Locomotor Movement around the cage; Behaviour, Total Non-social Horizontal Exploration includes active sniffing of air and Behaviour, Total Activity ground. Rapid stereotypical movement of Digging forepaws in the bedding. Rapid movement of forepaws over Self-Grooming Total Non-social Non- facial area and along body. locomotor Behaviour, Total No movement; includes sit, lie Solitary Inactivity Non-social Behaviour down and sleep. “Strange” behaviours, including “Abnormal” spinturns, repeated jumps/lid N/A Stereotypies chews/head shakes (more than 3).
220 Microtome. Sections were cleared in xylene, rehydrated with graded alcohol solutions, and subject to routine hematoxylin and eosin (H&E) staining. At least six different uterine sections per group were analyzed for various histomorphological parameters, including uterine area (perimeter of whole uterine section at 100x) and luminal epithelial height
(average of at least 10 different measurements per section collected at mid-uterine horn at
400x magnification). The uterine sections were observed under light microscope (Olympus
BH51) equipped with a Digital Camera System (Olympus DP72), and the measurements were performed with Image J calibrated against scale bars in the images (Rasband, 2011).
6.3.8 Statistical Analyses
6.3.8.1 Cinnamon Preference
Consumption of each flavour during each 2h interval was measured, and OBS choice data was expressed as a CIN preference ratio [CIN diet consumption/(COC + CIN diet consumption)]. If an OBS’s total intake for a 2h interval was less than 0.1 g, no preference ratio was calculated. To determine if treatment affected overall food consumption, the total intake (COC + CIN) in each 2h interval was also calculated.
Two-way mixed-model analyses of variance (ANOVAs) were used to assess the effect of treatment and demonstrated food on OBS CIN preference for each time interval (Clipperton et al., 2008; Clipperton-Allen et al., 2011c; Clipperton-Allen et al., 2011b), and planned comparisons were carried out to assess the effect of DEM food in each group and each interval using paired sample t-tests. The effects of treatment and time on overall food consumption were assessed with two-way ANOVAs, with post-hoc paired sample t-tests and
Bonferroni post-hoc tests used to determine the sources of significance for main effects and
221 interactions.
6.3.8.2 Behaviour analysis
Behaviours were predominantly not normally distributed (Levene’s statistics from 2.30-
13.72, ps from <0.001-0.049), and so the duration, frequency and latency of each behaviour was analyzed using the non-parametric Kruskal-Wallis and Mann-Whitney U tests
(Clipperton et al., 2008; Clipperton-Allen et al., 2010; Clipperton-Allen et al., 2011a).
Additionally, as we have previously seen relationships between Submissive Behaviour,
Agonistic Behaviour Received, and/or Dominance Score (see Table 29) and the degree or duration of preference for DEM food, we examined the relationship between these factors using Pearson’s product-moment correlations, both across all mice and for each treatment group separately (Clipperton-Allen et al., 2011c; Clipperton-Allen et al., 2011b).
6.3.8.3 Histomorphometric Analysis
Differences between PPT or DPN treatment groups and OVX controls were assessed by one-way ANOVAs followed by Bonferroni post hoc tests. The results are presented as the mean ± the standard error of the mean (SEM).
Only significant results are reported, and drug effects are in comparison to the corresponding vehicle control group, unless otherwise stated. All analyses were performed with an α level of 0.05 using PASW 18.0 for Windows (SPSS Inc., an IBM Company,
Chicago, IL).
222 6.4 Results
6.4.1 Treatment Effects on Food Preference
Chronic treatment with EB, PPT and DPN all resulted in preferences for the DEM food that lasted for 4h or 6h (2 or 3 times the duration of the preference shown by control groups).
6.4.1.1 Chronic EB
Two-way mixed design ANOVAs revealed significant main effects of DEM food at 2h
(F(1,139) = 105.68, p < 0.001), 4h (F(1,138) = 34.46, p < 0.001), and 6h (F(1,140) = 4.567, p = 0.034). Main effects of treatment were significant at 4h (F(5,138) = 2.43, p = 0.038) and showed a trend at 6h (F(5,140) = 2.20, p = 0.057). No significant drug or DEM food effects or interactions were found at 8h. Planned comparisons of the CIN preference in OBS that interacted with COC- and CIN-fed DEM revealed a significant effect of DEM food in all groups during the first 2h (Sham: t(21) = 3.68, p = 0.001; Vehicle: t(29) = 6.22, p < 0.001;
1.25 µg: t(22) = 3.04, p = 0.006; 12.5 µg: t(24) = 4.34, p < 0.001; 25 µg: t(21) = 3.99, p =
0.001; 50 µg: t(22) = 4.52, p < 0.001; see Figure 23). All OBS receiving EB treatment also preferred their DEM food at 4h (1.25 µg: t(22) = 2.81, p = 0.010, 12.5 µg: t(24) = 3.79, p =
0.001; 25 µg: t(20) = 2.51, p = 0.021; 50 µg: t(22) = 2.90, p = 0.008), and this preference was maintained for 6h in the 12.5 µg group (t(25) = 2.28, p = 0.032).
6.4.1.3 Chronic PPT
Two-way ANOVAs found significant main effects of DEM food on the CIN preference at 2h (F(1,128) = 105.44, p < 0.001) and 4h (F(1,128) = 24.28, p < 0.001), with no effects of drug or DEM food at 6h and 8h, and no drug effects at any 2h interval. Planned comparisons showed significant differences in the preference for CIN between OBS that interacted with
223 Figure 23. Cinnamon preference by 2h intervals of observers that received chronic estradiol benzoate (EB) treatment. Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers received no treatment (a) or a Silastic capsule containing sesame oil vehicle (b), 1.25 µg EB (c), 12.5 µg EB (d), 25 µg EB (e) or 50 µg EB (f).
224 COC-fed DEM and OBS that interacted with CIN-fed DEM in all groups for the first 2h
(Sham: t(20) = 2.78, p = 0.011; Vehicle: t(17) = 4.07, p = 0.001; 3 µg PPT: t(21) = 3.77, p =
0.001; 6 µg PPT: t(28) = 5.31, p < 0.001; 12 µg PPT: t(20) = 3.54, p = 0.002; 24 µg PPT: t(22) = 7.23, p < 0.001; see Figure 24). The groups receiving the higher three doses of PPT also showed a DEM food preference at 4h (6 µg PPT: t(27) = 2.07, p = 0.049; 12 µg PPT: t(20) = 3.42, p = 0.003; 24 µg PPT: t(22) = 2.74, p = 0.012).
6.4.1.3 Chronic DPN
Significant main effects of DEM food on the CIN preference were found at 2h (F(1,124)
= 68.70, p < 0.001) and 4h (F(1,125) = 27.86, p < 0.001), with a trend to a main effect of treatment at 4h (F(5,125) = 2.21, p = 0.057). No effects of drug were found at any time, and
DEM food effects were not significant for the 6h and 8h intervals. Planned comparisons showed significant differences in the preference for CIN between OBS that interacted with
COC-fed DEM and OBS that interacted with CIN-fed DEM in all groups for the first 2h
(Sham: t(20) = 2.78, p = 0.011; Vehicle: t(17) = 4.07, p = 0.001; 3 µg DPN: t(19) = 2.55, p =
0.020; 6 µg DPN: t(22) = 4.20, p < 0.001; 12 µg DPN: t(17.5) = 3.25, p = 0.004; 24 µg DPN: t(24) = 4.32, p < 0.001; see Figure 25). All mice administered DPN also showed a DEM food preference at 4h (3 µg DPN: t(19) = 3.01, p = 0.007; 6 µg DPN: t(22) = 2.27, p = 0.033;
12 µg DPN: t(19.8) = 2.89, p = 0.009; 24 µg DPN: t(24) = 2.17, p = 0.041), and the 12 µg
DPN group maintained this preference for the 6h interval as well (t(22) = 2.62, p = 0.016).
225 Figure 24. Cinnamon preference by 2h intervals of observers receiving chronic ERα agonist PPT treatment. Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers received no treatment (a) or a Silastic capsule containing sesame oil vehicle (b), 3 µg PPT (c), 6 µg PPT (d), 12 µg PPT (e) or 24 µg PPT (f).
226 Figure 25. Cinnamon preference by 2h intervals of observers receiving ERβ agonist DPN treatment. Means and standard errors are shown. * significant difference between observers whose demonstrators had been fed the cocoa-flavoured diet (black circle) and those whose demonstrators had been fed the cinnamon-flavoured diet (white triangle), p < 0.05. Observers received no treatment (a) or a Silastic capsule containing sesame oil vehicle (b), 3 µg DPN (c), 6 µg DPN (d), 12 µg DPN (e) or 24 µg DPN (f).
227 6.4.2 Treatment Effects on Behaviour
6.4.2.1 Chronic EB
The highest dose of EB increased the amount of shifting between behaviours
(“behavioural shifting”; Clipperton-Allen et al., 2010), as well as increasing social inactivity, the dominance score, and social investigation, especially of the oronasal area. The 50 µg dose of EB also increased non-social locomotor behaviours, including horizontal and vertical exploration and digging, as well as non-social non-locomotor behaviours, specifically solitary inactivity. Fewer behavioural effects were seen with the lower doses; 25 µg EB increased submissive behaviour, reduced the latency to dominance behaviours, and decreased self-grooming, and the two lower EB doses increased anogenital investigation and solitary inactivity.
Frequency of Total Activity was increased by the 50 µg dose of EB (U = 180.00, z = -
2.22, p = 0.027). Social Inactivity frequency was increased by 50 µg (U = 160.50, z = -2.63, p = 0.009). The latency to Dominant Behaviour was reduced by 25 µg EB (U = 161.00, z =
-2.09, p = 0.037), and the frequency of Submissive Behaviour was increased by 25 µg EB (U
= 151.00, z = -2.32, p = 0.020). The Dominance Score (frequency) was increased by 50 µg
EB (U = 184.00, z = -2.14, p = 0.011). Social Investigation (frequency) was increased by 50
µg EB (U = 182.50, z = -2.17, p = 0.030) and the latency to Social Investigation was reduced by 12.5 µg EB (U = 146.50, z = -2.76, p = 0.006). Oronasal Investigation was increased by
50 µg EB (U = 178.50, z = -2.25, p = 0.024). There was a treatment effect on Anogenital
Investigation (χ2(5) = 12.22, p = 0.032), and this behaviour was increased by 1.25 µg (U =
131.00, z = -2.41, p = 0.016) and 12.5 µg EB (U = 153.50, z = -2.61, p = 0.009).
Approaching and/or Attending to DEM was increased in frequency by 50 µg EB (U =
228 191.00, z = -1.99, p = 0.046).
There was a treatment effect on the latency to Non-social Behaviour (χ2(5) = 11.91, p =
0.036), due to a decreased latency in the 12.5 µg EB group (U = 174.50, z = -2.16, p =
0.031). Non-social Locomotor Behaviour was increased by 50 µg EB (U = 181.00, z = -
2.20, p = 0.028). Horizontal (U = 192.50, z = -1.96, p = 0.050) and Vertical Exploration were both increased by 50 µg EB (U = 191.00, z = -1.99, p = 0.046), as was Digging (U =
195.50, z = -2.00, p = 0.045), and the latency to dig was decreased by both the 1.25 µg (U =
142.00, z = -2.28, p = 0.023) and 50 µg EB doses (U = 181.50, z = -2.30, p = 0.021). Non- social Non-locomotor Behaviour was increased by 50 µg EB (U = 190.50, z = -2.00, p =
0.045). There was a treatment effect on Self-Grooming (χ2(5) = 11.63, p = 0.040), which was increased by Vehicle (U = 155.00, z = -2.03, p = 0.042) and decreased by 12.5 µg EB
(U = 156.50, z = -2.54, p = 0.011) and 25 µg EB (U = 150.50, z = -2.33, p = 0.020). Solitary
Inactivity was increased by 1.25 µg (U = 149.00, z = -1.97, p = 0.049), 12.5 µg (U = 157.00, z = -2.53, p = 0.011), and 50 µg (U = 179.00, z = -2.24, p = 0.025), and its latency was decreased by 12.5 µg EB (U = 152.50, z = -2.63, p = 0.009).
When all mice were included, the Dominance Score correlated weakly but positively with the duration of preference for DEM food (r(132) = 0.174, p = 0.044), and the degree of
DEM food preference at 2h (r(132) = 0.254, p = 0.003) and 8h (r(122) = 0.211, p = 0.019).
Agonistic Behaviour Received also had weak positive correlations with the duration (r(134)
= 0.175, p = 0.044) and 8h degree (r(124) = 0.217, p- 0.016) of DEM food preference.
When each treatment group was analyzed separately, DEM food preference at 4h was negatively correlated with the Dominance Score in the Vehicle group (r(21) = -0.484, p = 229 0.019), which did not express the preference at 4h; we found the opposite, a positive correlation between Dominance Score and the 4h DEM food preference, when mice received acute vehicle treatment (Clipperton-Allen et al., 2011c). In the 1.25 µg (r(20) = 0.441, p =
0.040) and 12.5 µg (r(22) = -0.443, p = 0.030) EB groups, the socially acquired food preference at 6h correlated with the frequency of Agonistic Behaviour Received, and in the
1.25 µg EB group, the preference for DEM food at 8h also correlated with the Dominance
Score (r(19) = 0.484, p = 0.026). It should be noted that there was no socially acquired food preference expressed at 6h or 8h by the 1.25 µg EB group. In the 25 µg EB group, the
Dominance Score and Agonistic Behaviour Received correlated positively with the duration
(Dominance Score: r(20) = 0.477, p = 0.025; Agonistic Behaviour Received: r(20) = 0.429, p
= 0.046) and 2h degree of DEM food preference (Dominance Score: r(20) = 0.708, p <
0.001; Agonistic Behaviour Received: r(20) = 0.434, p = 0.044). The positive correlations with Agonistic Behaviour Received is consistent with acute EB effects (Clipperton-Allen et al., 2011c).
6.4.2.2 Chronic PPT
Vehicle decreased, and the three higher doses of PPT increased Stretched Approaches, with vehicle also increasing and these PPT doses decreasing the latency to Stretch
Approaches. Additionally, Total Agonistic Behaviour was also decreased by the 12 µg PPT dose, the 24 µg PPT group were slower to investigate the oronasal region, and both 6 µg and 24 µg PPT reduced less solitary inactivity. Correlations between behaviour in the interaction and the duration and/or degree of socially acquired food preference were also significant in some groups. In the vehicle group, the Dominance Score correlated positively with the duration of DEM food preference, and negative correlations between Submissive
230 Behaviour and Agonistic Behaviour Received and the degree of preference at 4h were found in the 3 µg PPT group, despite the lack of preference expressed at this interval.
Additionally, a negative correlation was found between the Dominance Score and the 2h
DEM food preference, as well as between Submissive Behaviour and Agonistic Behaviour
Received and the 6h degree of DEM food preference, although this group also did not show a preference for DEM food at this time.
Total Agonistic Behaviour was decreased by 12 µg PPT (U = 284.00, z = -2.07, p =
0.039). The latency to Oronasal Investigation was increased by 24 µg PPT (U = 278.00, z =
-2.16, p = 0.031). Stretch Approaches were decreased by Vehicle (U = 160.00, z = -2.08, p =
0.038) and increased by 6 µg (U = 480.00, z = -2.45, p = 0.014), 12 µg (U = 360.00, z = -
2.43, p = 0.015) and 24 µg (U = 360.00, z = -2.43, p = 0.015). The latency to Stretch
Approaches was increased by Vehicle (U = 160.00, z = -2.08, p = 0.038) and decreased by 6
µg (U = 500.00, z = -2.40, p = 0.016), 12 µg (U = 360.00, z = -2.43, p = 0.015) and 24 µg
PPT (U = 360.00, z = -2.43, p = 0.015). Vertical Exploration was increased by 6 µg PPT (U
= 410.00, z = -2.07, p = 0.039). There was a treatment effect on Solitary Inactivity (χ2(5) =
11.59, p = 0.041), which was increased by 6 µg (U = 369.00, z = -2.57, p = 0.010), 12 µg (U
= 226.00, z = -2.95, p = 0.003) and 24 µg PPT (U = 274.00, z = -2.22, p = 0.027).
When all mice were included, the duration of DEM preference correlated significantly but weakly with the Dominance Score (r(126) = 0.198, p = 0.025). When analyzed separately by treatment, the Dominance Score correlated positively with DEM preference duration in the Vehicle group (r(17) = 0.512, p = 0.025) and negatively with the preference for DEM food at 2h in the 24 µg PPT group (r(19) = -0.496, p = 0.022). Additionally, DEM
231 food preference at 6h correlated negatively with Submissive Behaviour (r(19) = -0.433, p =
0.050) and Agonistic Behaviour Received (r(19) = -0.480, p = 0.028) in the 24 µg PPT group, despite failing to show a DEM food preference at this time. In the 3 µg PPT group,
DEM food preference at 4h correlated negatively with the frequencies of Submissive
Behaviour (r(18) = -0.524, p = 0.018) and Agonistic Behaviour Received (r(18) = -0.517, p =
0.020), despite this group not showing a significant preference for DEM food at this time.
6.4.2 Chronic DPN
The majority of DPN’s effects on behaviour during the social interaction were by the highest dose (24 µg), although it was the 12 µg DPN group that showed the longest preference for DEM food. Social Behaviour, including Social inactivity, was increased by 24
µg DPN, as were Non-social Behaviours, specifically Non-social Locomotor Behaviours, including horizontal and vertical exploration. Solitary Inactivity was decreased by 24 µg
DPN. The 12 µg DPN dose increased Non-social Non-locomotor behaviours, including solitary inactivity and self-grooming. Finally, as with chronic PPT, Stretched Approaches were decreased by vehicle and increased by the three lower doses of DPN, with the reverse effects on latency.
Social Behaviour was increased by 24 µg DPN (U = 322.00, z = -2.40, p = 0.016).
Social Inactivity was decreased by 24 µg DPN (U = 348.00, z = -2.05, p = 0.040), and both this (U = 314.00, z = -2.51, p = 0.012) and the 12 µg dose of DPN (U = 260.00, z = -2.65, p
= 0.008) increased the latency to Social Inactivity. The latency to Chasing DEM was decreased by 24 µg DPN (U = 354.00, z = -1.98, p = 0.047). Stretch Approaches were decreased by Vehicle (U = 160.00, z = -2.08, p = 0.038), and increased by 3 µg (U = 300.00,
232 z = -3.27, p = 0.001), 6 µg (U = 360.00, z = -2.77, p = 0.006), and 12 µg DPN (U = 360.00, z = -2.77, p = 0.006). The latency to Stretch Approaches was increased by Vehicle (U =
160.00, z = -2.08, p = 0.038) and decreased by 3 µg (U = 300.00, z = -3.27, p = 0.001), 6 µg
(U = 360.00, z = -2.76, p = 0.006) and 12 µg DPN (U = 360.00, z = -2.76, p = 0.006).
Non-social Behaviour (U = 322.00, z = -2.40, p = 0.016) and Non-social Locomotor
Behaviour (U = 328.00, z = -2.32, p = 0.020) were increased by 24 µg DPN. Horizontal (U
= 316.00, z = -2.48, p = 0.013) and Vertical Exploration (U = 326.00, z = -2.35, p = 0.019) were increased by 24 µg DPN. Non-social Non-locomotor Behaviour (U = 260.00, z = -
2.65, p = 0.008), and both Solitary Inactivity (U = 243.00, z = -2.90, p = 0.004) and Self-
Grooming (U = 283.00, z = -2.31, p = 0.021), were increased by 12 µg DPN. Solitary
Inactivity was decreased by 24 µg DPN (U = 305.00, z = -2.63, p = 0.009). The latency to
Non-social Non-locomotor Behaviour was increased by 24 µg DPN (U = 329.00, z = -2.31, p = 0.021).
When all mice were included, the preference for DEM food at 4h correlated positively with Submissive Behaviour (r(124) = 0.182, p = 0.042) and Agonistic Behaviour Received
(r(124) = 0.184, p = 0.039), while the 6h DEM food preference correlated negatively with
Submissive Behaviour (r(124) = -0.255, p = 0.004) and Agonistic Behaviour Received
(r(124) = -0.249, p = 0.005), and positively with Dominance Score (r(124) = 0.191, p =
0.032), although none of these correlations were very strong.
When the treatment groups were analyzed separately, the 6h DEM food preference correlated negatively with both Submissive Behaviour (r(17) = -0.466, p = 0.045) and
Agonistic Behaviour Received (r(17) = -0.456, p = 0.050) in the 3 µg DPN group, which did 233 not show a DEM food preference at this time, and in the 12 µg DPN group (Submissive
Behaviour: r(20) = -0.465, p = 0.029; Agonistic Behaviour Received: r(20) = -0.472, p =
0.027), which did significantly prefer DEM food at this time. These findings are unlike our results with acute ERβ agonists, which indicated that increased submissive behaviours and categories were positively related to DEM food preference degree and duration.
Additionally, in the 24 µg DPN group the preference for DEM food at 8h correlated positively with Submissive Behaviour (r(19) = 0.452, p = 0.040), although this group did not show a significant DEM food preference at this time.
6.4.3 Feeding Effects
In all analyses of effects of EB, PPT and DPN on feeding, two-way ANOVAs found a significant main effect of time (all Fs > 124.4, all ps < 0.001) with significant differences in intake between all intervals (all ps < 0.001). Additionally, there was a significant main effect of treatment in the EB study (F(5,147) = 3.09, p = 0.011), due to significantly higher food intake in the vehicle controls compared to the sham group (p = 0.011), and a time x treatment interaction (F(15,441) = 2.01, p = 0.014), which was due to effects of treatment at
4h (F5,147) = 3.32, p = 0.007) and 8h (F(5,147) = 3.71, p = 0.003). However, post-hoc tests showed that this treatment effect was not due to significant differences between the vehicle controls and any of the hormone-treated groups or the sham controls.
6.4.4 Vaginal Cytology
Control mice which vaginal cytology showed to be cycling were removed from the data pool (n = 6 across all studies). Mice receiving chronic EB treatment showed estrus-like vaginal cytology, except for one mouse in the 50 µg EB group, which was removed. Mice
234 receiving chronic PPT or DPN treatment were inconsistent in their vaginal cytology and all were included.
6.4.5 Chronic PPT and DPN Effects on Uterine Histomorphology
Chronic EB of similar duration and at the doses used in this study have previously been shown to increase uterine wet weight and increase the thickness of the endometrial lining
(Morgan et al., 2000). However, there were no significant differences between ovx vehicle control and PPT- or DPN-treated mice observed in any of the uterus histomorphological endpoints collected (see Table 30).
6.5 Discussion
Chronic administration of EB, PPT and DPN prolonged the preference for DEM food beyond the 2h long preference shown by the sham and vehicle control groups. In the case of the 12.5 µg EB and the 12 µg DPN groups, the socially acquired food preference lasted for three times as long (see Figure 23d and Figure 25e), while all other EB-, PPT- and DPN- treated groups, with the exception of 3 µg PPT, expressed this preference for twice as long as control mice (see Figures 23-25). Correlations between submissive behaviour, agonistic behaviour received and/or dominance score and the duration and/or degree of DEM food preference were found in some treatment groups, supporting the notion that the nature of the social interaction may relate to the expression of socially acquired food preferences, and that drug effects on social learning may in part be due to modifications of social behaviour, as more subordinate observers may learn better from dominant demonstrators (Clipperton et al.,
2008; Kavaliers et al., 2005; Choleris et al., 1998)
Several studies suggest that estrogens’ positive effects on learning and memory may be 235 mediated by ERβ. Estrogens and ERβ agonists have largely similar effects, whether they are
Table 30. Effects of chronic PPT and DPN on uterine histomorphology. Dose Uterine Area (mm2) Luminal Epithelial Height (µm) Sham 1.55 23.31 Ovx Vehicle 0.47±0.03 14.47±0.81 3 µg PPT 0.50±0.03 16.07±0.83 6 µg PPT 0.55±0.03 17.63±0.75 12 µg PPT 0.59±0.04 15.72±0.49 24 µg PPT 0.50±0.03 15.87±0.54 3 µg DPN 0.54±0.05 14.54±0.80 6 µg DPN 0.47±0.03 15.24±0.45 12 µg DPN 0.48±0.03 15.43±0.74 24 µg DPN 0.53±0.04 16.17±0.52
236 administered acutely or chronically. Acute administration has been found to improve learning and memory, especially when tested on hippocampus-mediated spatial tasks like the radial arm maze (Holmes et al., 2002b) and Morris water maze (Markham et al., 2002;
Sandstrom & Williams, 2004). Similarly, in the STFP, we have found that OBS treated with chronic (see Figures 23 and 25) and acute pre-acquisition EB, DPN and ERβ agonist WAY-
200070 show longer preferences for the DEM (Clipperton et al., 2008; Clipperton-Allen et al., 2011c; Clipperton-Allen et al., 2011b). In the present study, chronic EB and DPN treatments had stronger effects on the STFP than the corresponding acute treatments, as all doses prolonged the socially acquired food preference (2-3 times controls), while in the acute studies the DEM food preference was twice as long as that of controls, and only in one dose group (Clipperton-Allen et al., 2011c; Clipperton-Allen et al., 2011b).
Unlike estrogens and ERβ specific agonist treatments, chronic ERα activation appears to produce different effects than those seen with acute treatments. We have previously found that PPT will block the STFP whether administered before, immediately after or 24h after social interaction (Clipperton et al., 2008). However, similar to the other chronic PPT study
(Hammond et al., 2009), chronic doses of PPT induced a longer expression of DEM food preference. In non-social paradigms, PPT often has no effects on performance of hippocampal and spatial learning tasks (Jacome et al., 2010b; Rhodes & Frye, 2006).
However, administration of estrogens to mice in which the gene for ERβ has been inactivated
(knocked out, “KO”; βERKO mice) or wild type mice, but not to αERKO mice, impaired performance on the Morris water maze, suggesting that ERα may have detrimental effects on learning, especially when activated in the absence of ERβ (Fugger et al., 1998; Rissman et al., 2002a). Further, high doses of PPT reduced performance during non-cued alternation
237 (Neese et al., 2010). Additionally, high doses of PPT have been found to impair social recognition (Cragg et al., 2011).
Similar effects of EB and the ERβ agonist DPN may be related to ERβ’s role as a transdominant repressor of ERα transcriptional activity, that can also decrease the sensitivity of ERα cells to estradiol (Hall & McDonnell, 1999). Additionally, ERα and ERβ form heterodimers when both are activated (Cowley et al., 1997; Hall & McDonnell, 1999), and there is some evidence that these heterodimers may act in opposition to ERα homodimers
(Hall & McDonnell, 1999; Lindberg et al., 2002).
ERβ appears to be more responsive to estrogens than ERα in the hippocampus (Bora et al., 2005), suggesting that actions of estrogens on ERβ in the hippocampus may be mediating these effects. Further, the observed similarities in results between hippocampus-dependent spatial tasks and the STFP are not entirely surprising; although the STFP is an exclusively social task, requiring associations between food odours and breathing DEM to be learned
(Galef et al., 1988), it is mediated by the hippocampus (Alvarez et al., 2001; Alvarez et al.,
2002; Bunsey & Eichenbaum, 1995). Additionally, the hippocampus receives inputs from the cholinergic basal forebrain (CBF), an area that has also been implicated in the STFP
(Berger-Sweeney et al., 2000; Ricceri et al., 2004; Vale-Martínez et al., 2002). This suggests a possible mechanism by which estrogens may be acting to affect these learning and memory tasks, as they have been found to increase CBF acetylcholine release, as well as increasing activity of choline acetyltransferase, the enzyme that synthesizes acetylcholine (e.g., Gibbs et al., 1994; Luine, 1985; Frick et al., 2002a; Gibbs, 2000a; Gibbs et al 1997; reviewed in Frick,
2009).
238 We have previously found relationships between the submissiveness of OBS and the degree and/or duration of DEM food preference following acute pre-acquisition administration of DPN in the doses that prolonged expression of the socially acquired food preference (Clipperton et al., 2008; Clipperton-Allen et al., 2011b). However, such interactions were not found with acute EB (Clipperton-Allen et al., 2011c), or in the current study with chronic EB, PPT or DPN treatments. This, together with the different effects of chronic and acute PPT, further supports the notion that the longer exposure to all three compounds may lead to a loss of receptor selectivity (as discussed in Jacome et al., 2010b), thus inducing similar effects on the STFP that are unrelated to the nature of the social interaction.
6.5.1 Overall Conclusions
Our results suggest that chronic targeting of specific estrogen receptors using the currently available agonists in order to receive ERβ’s cognitive benefits may not avoid negative side effects like ERα-associated breast cancer, as there may be a loss of specificity when these receptors are subjected to constant sustained agonism. Further research is clearly required to determine if cyclic treatment, more similar to the natural cycle of women, could prove more beneficial than artificially high constant hormone levels.
239
CHAPTER 7:
Oxytocin, vasopressin and estrogen receptor gene expression in relation to
social recognition in female mice
This manuscript is published in: Clipperton-Allen, A.E., Lee, A.W., Reyes, A., Devidze, N.,
Phan, A., Pfaff, D.W., and Choleris, E. Oxytocin, vasopressin and estrogen receptor gene
expression in relation to social recognition in female mice. Physiology & Behavior 105:
915-924.
240 7.1 Abstract
Inter- and intra-species differences in social behaviour and recognition-related hormones and receptors suggest that different distribution and/or expression patterns may relate to social recognition. We used qRT-PCR to investigate naturally occurring differences in expression of estrogen receptor-alpha (ERα), ER-beta (ERβ), progesterone receptor (PR), oxytocin (OT) and receptor, and vasopressin (AVP) and receptors in proestrous female mice.
Following four 5 min exposures to the same two conspecifics, one of which was replaced with a novel mouse in the final trial (T5). Gene expression was examined in mice showing high (85-100%) and low (40-60%) social recognition scores (i.e., preferential novel mouse investigation in T5) in eight socially-relevant brain regions. Results supported OT and AVP involvement in social recognition, and suggest that in the medial preoptic area, increased OT and AVP mRNA, together with ERα and ERβ gene activation, relate to improved social recognition. Initial social investigation correlated with ERs, PR and OTR in the dorsolateral septum, suggesting that these receptors may modulate social interest without affecting social recognition. Finally, increased lateral amygdala gene activation in the LR mice may be associated with general learning impairments, while decreased lateral amygdala activity may indicate more efficient cognitive mechanisms in the HR mice.
241 7.2 Introduction
Social recognition, the ability to identify or recognize conspecifics, is the basis for complex social relationships (Choleris et al., 2009; Pierman et al., 2008). In order to facilitate dominance hierarchy and social bond development, one needs to modulate one's behaviour based on previous experience with an individual (Choleris et al., 2006; Choleris et al., 2009; Choleris et al., 2007; Choleris et al., 2004).
Sex differences in sociality in our own and other species, as well as the dramatic sexual dimorphism in social disorders (e.g., 90% of ASD individuals are male; Gillberg, 2002), suggest that gonadal hormones may be involved. The animal literature has supported this, and implicated female gonadal hormones in social recognition and other types of social cognition like social learning (see Choleris et al., 2008a for a recent review). Typically, rodent social recognition is assessed in the laboratory by measuring habituation (reduction in investigation) to familiar conspecifics and dishabituation (increase in investigation) to novel animals, and/or by examining whether novel conspecifics are preferentially investigated over familiar animals in a choice test (social discrimination; Choleris et al., 2003; Choleris et al.,
2006; Choleris et al., 2009).
Administration of estrogens to ovx mice increases long-term social recognition memory
(Tang et al., 2005), and both main intranuclear estrogen receptors, alpha (ERα) and beta
(ERβ), appear to mediate social recognition, ERα more than ERβ (Choleris et al., 2003;
Choleris et al., 2004; Choleris et al., 2006; Imwalle et al., 2002; Sánchez-Andrade &
Kendrick, 2011; Spiteri & Ågmo, 2009). Consistently, systemic treatment with ERα or ERβ agonists improved social recognition in mice when administered 48-72h prior to testing
242 (Choleris et al., 2008a; Clipperton et al., 2006; Cragg et al., 2011), but only the ERα agonist improved performance rapidly (within 40 min; Choleris et al., 2008a; Phan et al., 2011).
Although it has been the subject of little to no research, progesterone may also modulate social recognition. Treatment with estradiol benzoate and progesterone increased habituation rate in ovariectomized (ovx) rats (Spiteri & Ågmo, 2009), and progesterone receptor (PR) is expressed in social recognition-related brain areas (e.g., medial amygdala; Spiteri & Ågmo,
2009).
Downstream effectors of gonadal hormone action on social recognition may include nonapeptides oxytocin (OT), vasopressin (AVP), and their receptors, which have been shown to mediate social recognition in rodents and are both under the control of sex hormones (see
Choleris et al., 2009 for a recent review). Administration of antagonists or inactivation of the
OT gene (“knock out”, KO), its receptor (OTR), or either AVP receptor (AVPR1a more than
AVPR1b) resulted in specific social recognition deficits (Bielsky et al., 2004; Bielsky et al.,
2005; Choleris et al., 2003; Choleris et al., 2007; Choleris et al., 2006; Dantzer et al., 1987;
Ferguson et al., 2001; Lee et al., 2008; Roper et al., 2011; Samuelsen & Meredith, 2011;
Takayanagi et al., 2005; Wersinger et al., 2002). Exogenous OT rescued social recognition in OTKO mice and improved it in male (but not female) wild type (WT) mice (Ferguson et al., 2002) and rats (Engelmann et al., 1998), while AVP improved social recognition in both sexes (Bluthé et al., 1990; Bluthé & Dantzer, 1990; Engelmann et al., 1998). However, AVP antiserum or antagonist administration blocked male but not female social recognition
(reviewed in Choleris et al., 2009).
While the involvement of these hormones and receptors in social recognition has largely
243 been established, whether their functions and/or distributions are linked to behaviour in social situations has received less attention. Species differences exist in both hormones and their receptors in relation to social behaviour. Male vole and hamster species that are monogamous and/or paternal express more ERα than polygynous species in socially-related brain regions like the bed nucleus of the stria terminalis (BNST) and medial amygdala
(Cushing & Wynne-Edwards, 2006; Insel et al., 1991). It is also possible to induce pair bonding in both male mice and promiscuous male voles by introducing the monogamous prairie vole AVPR1a gene into the ventral forebrain or central nervous system (Cushing &
Wynne-Edwards, 2006; Insel et al., 1991; Lim et al., 2004; Young et al., 1999), although variations in the AVPR1a gene may not be sufficient or necessary for this type of mating behaviour in voles and deer mice (Turner et al., 2010). Within-species variations in hormones and their receptors are also linked to social behaviours; voles that spent more time investigating a novel female, and did not habituate, had more AVPR1a and less OTR in the septum (Ophir et al., 2009), and more social males express less ERα in the BNST and medial amygdala (Cushing et al., 2004).
To the best of our knowledge, whether social-recognition-associated behaviours are linked to within-species variation in ERs and PR, or putative downstream effector OT and
AVP systems, in female house mice (Mus musculus) is unknown. Thus, we tested outbred female CD1 mice in the proestrous phase of the estrous cycle on the social discrimination paradigm. This phase was chosen partly because it involves high estrogen levels (Walmer et al., 1992) and leads to improved social recognition learning (Sánchez-Andrade et al., 2005;
Sánchez-Andrade & Kendrick, 2011). Additionally, social approach and social choice are particularly important at this reproductively active phase, as remembering conspecific and
244 environmental interactions facilitates assessment of the social and physical environment’s suitability for reproduction (Sánchez-Andrade et al., 2005; Sánchez-Andrade & Kendrick,
2009).
Brain tissue obtained from mice that performed very well (high recognizers, HR) and from those that performed close to chance (low recognizers, LR) was analyzed for levels of
ERα, ERβ, PR, OT, OTR, AVP, AVPR1a, and AVPR1b mRNA expression. The following brain regions were selected for this analysis, taking into account their role in social behaviour and social recognition: BNST, dorsolateral septum (DS), frontal cortex, lateral amygdala
(LA), medial amygdala, medial preoptic area (MPOA), paraventricular nucleus (PVN) of the hypothalamus, ventromedial hypothalamus (VMH) and hippocampus (Choleris et al., 2003;
Choleris et al., 2004; Choleris et al., 2006; Choleris et al., 2007; Ferguson et al., 2002;
Hammock & Young, 2007; Kogan et al., 2000; Pierman et al., 2008; Popik & van Ree, 1991;
Ross & Young, 2009; Spiteri et al., 2010). Gene expression was assessed via quantitative real-time reverse-transcription polymerase chain reaction (qRT-PCR). Behaviours in the social discrimination test and gene expression levels were compared between HR and LR mice, and social recognition and social interest were correlated with gene expression.
7.3 Materials and Methods
7.3.1 Subjects
Fifty-eight adult female mice of the outbred CD1 strain were obtained from Charles
River, QC, Canada, and held in the Central Animal Facility at University of Guelph, under a
12:12h reversed light/dark cycle (lights off at 0800h) at 21±1°C. Mice were kept in clear polyethylene cages (26 x 16 x 12 cm), and provided with corn cob bedding, environmental
245 enrichment (plastic containers, paper nesting material), and food (Teklad Global 14% Protein
Rodent Maintenance Diet, Harlan Teklad, Madison, WI) and tap water ad libitum.
The 49 experimental mice were individually housed for at least 7 days to develop a home cage territory, with no cage changes taking place for at least 3 days prior to testing, and no experimental mouse was used more than once. Nine ovx stimulus mice were housed in same-sex groups of three, and mice were reused with at least one day between exposures.
Ovx stimulus mice, which we have previously shown will be highly investigated by female mice (Choleris et al., 2003; Choleris et al., 2006), were chosen in order to avoid estrous cycle-induced variability and to have more “standardized” stimuli with more consistent levels of circulating gonadal hormones (Clipperton-Allen et al., 2010). Brains were collected from experimental mice with an estimated preference for the novel mouse of 85% or higher in the HR group or between 40-60% in LR group, which consisted of approximately 16%
(n=8) and 20% (n=10) of the population, respectively (see Table 31).
7.3.2 Surgery
Stimulus mice underwent bilateral ovx under isoflurane gas anaesthesia (Benson Medical
Industries, Markham, ON), with subcutaneous (s.c.) carprofen (50 mg/kg; Rimadyl, Pfizer
Canada Inc, Kirkland, QC) in saline at 10 ml/kg as an analgesic and anti-inflammatory. The lower back skin was shaved and cleaned, and a single 1.5-2 cm incision was made in the skin to expose the back muscles. A small 0.5-1 cm incision was made in the muscles overlying the ovaries on each side, and then the ovary and the tip of the uterus were carefully drawn out of the hole. The ovary was removed by cutting just above a clamp on the uterus, and then the uterus was replaced in the abdominal cavity. A few drops of 12.5% marcaine (bupivacaine
246 Table 31. Quantity of total RNA used in reverse transcription reactions by brain region. Total RNA/20 µl Brain Region High Recognition Low Recognition reaction mixture (µg) BNSTa 1.5 n=7 n=7 Dorsolateral 1.0 n=4 n=5 Septum Hippocampus 1.5 n=6 n=9 Lateral Amygdala 1.5 n=4 n=9 Medial Amygdala 1.5 n=6 n=9 MPOAb 1.0 n=5 n=8 PVNc 1.0 n=6 n=9 VMHd 1.5 n=7 n=10 Frontal cortex 1.0 n=6 n=10 aBed Nucleus of the Stria Terminalis. bMedial Preoptic Area. cParaventricular Nucleus of the Hypothalamus. dVentromedial Hypothalamus.
247 hydrochloride; Hospira, Inc, Lake Forest, IL) were dripped onto the back incision as a local anaesthetic, and the incision was stapled with 1-2 MikRon Autoclip 9 mm wound clips
(MikRon Precision Inc, Gardena, CA). Mice received a rehydrating 0.5 ml intraperitoneal saline injection, and were allowed 7 days of single-housed recovery, ovx stimulus mice were group-housed at least 7 days before the experiment.
7.3.3 Apparatus
Trials were run in home cages of experimental mice, and videotaped from above with an
8 mm Sony Handycam (in Nightshot) through clear Plexiglas cage lids. Stimulus mice were placed into clear Plexiglas cylinders (12 cm high, 7 cm in diameter) that had 36 holes (4 mm diameter) drilled into their bottom third, thus allowing for passage of olfactory cues while preventing direct interactions between stimulus and experimental mice (Choleris et al.,
2006). This ensured that experimental mice did not deposit their own scent onto stimulus mice (Insel & Fernald, 2004), and also reduced stimulus-induced variability in behaviour, while still eliciting high social interest and investigation from experimental mice (Choleris et al., 2003; Choleris et al., 2006; Choleris et al., 2007; Kudryavtseva et al., 2002). Stimulus mice were repeatedly habituated to cylinders prior to testing until they spontaneously entered them with no observable distress (Choleris et al., 2006).
Clean cylinders were used for all trials, while the same empty cylinders were used in inter-trial intervals. Cylinders were washed thoroughly with unscented soap (Alconox, VWR
International, West Chester, PA) and baking soda (Arm & Hammer, Church & Dwight
Canada Corp., Mississauga, ON), then paper towel- and air-dried.
248 7.3.4 Behavioural Testing Procedure
Gonadally intact female experimental mice were tested during the proestrous phase of the estrous cycle. The vaginal cytology of the experimental mice was analyzed as described in
Clipperton et al. (2008), and they were retested until they entered this phase of the estrous cycle, and vaginal cytology was also used to confirm that ovx mice were not cycling, determined as described in Clipperton et al. (2008). All procedures followed the regulations of the Canadian Council on Animal Care and were approved by the Institutional Animal Care and Use Committee of the University of Guelph.
We followed the procedure described in Choleris et al. (2006), with all trials taking place between the second and sixth hour of the dark (active) phase of the light/dark cycle (i.e., between 0900 and 1400). Briefly, both experimental and stimulus mice were moved into an antechamber of the testing room and left undisturbed for at least 90 min. For testing, all mice were moved into the testing room, enrichment was removed from the experimental mouse’s cage, and two clean empty cylinders were placed into two corners of her cage for 10 min.
Two cylinders, each containing a stimulus mouse from a different cage, replaced the empty cylinders for five 5 min trials, with the empty cylinders returned during the 15 min inter-trial intervals. For the first four trials (T1-T4), the same stimulus mice were used in the same locations, while in the final trial (T5), one of the stimulus mice was replaced with a novel mouse from a different cage than the other two stimuli. The location of the replaced mouse was counterbalanced across subjects. The number and location of stimulus mice was maintained for consistent social stimulation. During all trials, mice were left undisturbed in the room and their behaviour was videotaped for subsequent behavioural analysis. During test (T5), the experimenter also sat quietly in an antechamber and manually timed the
249 duration of investigation of the novel and familiar stimulus mice with two Jumbo Digital
Traceable Stopwatches (Fisher Scientific, Ottawa, Ontario) to gain a quick preference estimate.
Immediately following T5, any mice with an estimated preference for the novel mouse of approximately 85% or above (HR mice) or 40-60% (LR mice) were sacrificed and brains were extracted as described previously (Phan et al., 2011). Briefly, as per institutional
Animal Care Committee guidelines, animals were rapidly asphyxiated with CO2 (typically less than 1 min in duration, thus limiting possible CO2 effects on brain tissues), cervically dislocated and decapitated. Brains were carefully removed, flash frozen on dry ice and stored at -80°C.
Videotaped trials were subsequently scored by a trained observer unaware of the initial rough estimate of preference, using The Observer Video Analysis software (Noldus
Information Technology, Wageningen, the Netherlands), for eight behaviours based on the mouse ethogram (Grant & Mackintosh, 1963; see Table 32, modified from Choleris et al.,
2006; Phan et al., 2011). Social investigation and cylinder investigation were calculated as ratios (investigation of novel mouse or cylinder/investigation of both mice or cylinders) and totals. Habituation and dishabituation measures of total social behaviour were also calculated (see Table 32).
7.3.5 Quantitative Real Time Reverse Transcription Polymerase Chain Reaction
Nine brain regions (BNST, DS, hippocampus, LA, medial amygdala, MPOA, PVN,
VMH, and a section of frontal cortex as a control region) were isolated as described in
250 Table 32. Descriptions of behaviours collected during behavioural testing. Behaviour Description Scored Behaviours Movement to investigate upwards, lifting both front feet off the Vertical Exploration ground; includes sniffing, wall leans and lid chews. Horizontal Exploration Locomotor activity; includes active sniffing of air and ground. Rapid stereotypical movement of forepaws in the bedding; this Digging behaviour was not highly performed. Self-Grooming Rapid movement of forepaws over facial area and along body. Inactivity No movement; includes sit, lie down and sleep. Risk assessment behaviour; back feet do not move and front feet approach the stimulus mouse. Only the frequency of stretched Stretched Approaches approaches was measured, and this behaviour was not highly performed. The experimental mouse actively sniffs the holes at the bottom part Social Investigation of the cylinder containing the stimulus mouse. Investigation of each stimulus mouse was scored separately for all trials. The experimental mouse sniffs the part of cylinder without holes. Cylinder Investigation Investigations of the two cylinders were scored separately in all trials. Calculated Behaviours This ratio was created by calculating the ratio of Total Social Investigation that was focused on the “novel” stimulus, i.e., the cylinder that would contain the novel stimulus mouse in the test trial: (Investigation of “Novel” stimulus/Total social Social Investigation Investigation). In the training trials, both stimulus mice should be Ratio equally familiar, and thus the ratio is expected to be around 0.5 (random chance). In the test trial, however, if the mouse can discriminate between the novel and familiar stimulus mice, it is expected that the mouse would explore the novel mouse significantly more than chance. Total Social The Total Social Investigation of the two stimulus mice; this was Investigation also evaluated as an overall measure of ‘social interest’. The preference for the cylinder that at the test trial will contain the Cylinder Investigation novel mouse was calculated as described for the Social Ratio Investigation Ratio above. Total Cylinder The total investigation of both cylinders. Investigation The Total Social Investigation at T1 – the Total Social Habituation (Social) Investigation at T4. The Total Social Investigation at T5 – the Total Social Dishabituation (Social) Investigation at T4. The Total Cylinder Investigation at T1 – the Total Cylinder Habituation (Cylinder) Investigation at T4. Dishabituation The Total Cylinder Investigation at T5 – the Total Cylinder (Cylinder) Investigation at T4.
251 Ribeiro et al. (2009). Briefly, frozen brains were sliced coronally at 1 mm intervals, and areas of interest were dissected out using a scalpel or razor blade and fine forceps.
Ribonucleic acid (RNA) from each region was isolated using Trizol reagent according to the manufacturer’s instructions (Invitrogen, San Diego, CA). The amount and quality of total
RNA were assayed using an Agilent Bioanalyzer 2100 (Agilent Technologies, Palo Alto,
CA, USA) and a Nanodrop UV spectrophotometer. Total RNA from each sample was reverse transcribed using a High Capacity cDNA Reverse Transcription Kit (Applied
Biosystems, Foster City, CA) according to the manufacturer’s specifications. Briefly, 20 µl of reaction mixture contained 200 or 300 ng of total RNA (see Table 31), 2 µl 10X Reverse
Transcription buffer, 0.8 µl 25X dNTP mix (100 mM), 2 µl 10X Reverse Transcription random primers, 1 µl of MultiScribe Reverse Transcriptase, and 4.2 µl nuclease-free water.
The cycling conditions for reverse transcription reaction were: 10 min at 25°C, 120 min at
37°C, 5s at 85°C, and then maintenance at 4°C until removed from the PTC-100
Programmable Thermal Controller (MJ Research Inc., Waltham, MA). For each gene, qRT-
PCR was carried out in all brain regions, using Applied Biosystems’ gene expression assays
(Foster City, CA): ERα (Assay ID Mm00433149_m1), ERβ (Assay ID Mm00599819_m1),
PR (Assay ID Mm00435628_m1), OT (Assay ID Mm01329577_g1) , OTR (Assay ID
Mm01182684_m1), AVP (Assay ID Mm00437761_g1), AVPR1a (Assay ID
Mm00444092_m1), and AVPR1b (Assay ID Mm01700416_m1). Eukaryotic 18S was used as an endogenous control (Assay ID Mm03928990_g1). Q-PCR reactions were carried out using the Taqman detection system (ABI Prism® 7000 Sequence Detection System, Applied
Biosystems, Foster City, CA) and the following conditions: 2 min at 50°C, 10 min at 95°C, followed by 40 cycles of 15s at 95°C and 1 min at 60°C. This resulted in each cycle
252 producing a 1.87x increase in quantity of cDNA. In some brain regions, the samples for specific animals were of insufficient quality or quantity for qRT-PCR, resulting in smaller sample sizes for these regions.
7.3.6 Statistical Analysis
For all statistical analyses, only significant results are reported, and SigmaStat 3.5 (Systat
Software, Inc., Chicago, IL), SPSS 13.0 (SPSS Inc., Chicago, IL), and G*Power 3 (Faul et al., 2009; Faul et al., 2007) were used. For the comparisons of the HR and LR groups on behavioural or molecular measures, the alpha level was set at p < 0.05. To reduce the number of comparisons, and thus the possibility of type I error, selected mean comparisons were planned a priori. No multiple comparison correction was performed, as Bonferroni and similar corrections are too conservative (i.e., increase the probability of a type II error above acceptable limits). This decision follows recent concerns about the widespread use of such corrections and the consequent increase in type II errors (e.g., Moran, 2003; Nakagawa,
2004; Rothman, 1990), and is in line with the statistical analyses in the literature (e.g., Bales et al., 2007; Cushing et al., 2004; Cushing & Wynne-Edwards, 2006; Litvin et al., 2011;
Murakami et al., 2011; Ribeiro et al., 2009). We report effect sizes (Cohen's d for t-tests, partial η2 for analyses of variance, ANOVAs) and power for each significant comparison and allow our readers to interpret the data accordingly.
Due to the large number of behavioural/molecular correlations, a more stringent criterion of p < 0.01 was applied to correct for multiple correlations. We believe that this reduces the likelihood of type I error, while maintaining an acceptable power level (greater than 0.80;
Talamini et al., 1999).
253 7.3.6.1 Behavioural analysis
Because ratios violate the homogeneity of variance assumption of parametric statistics, the duration and frequency of the Social Investigation and Cylinder Investigation Ratios were arcsine-transformed prior to analysis [arcsine (√ratio)]. Digging and Stretched
Approaches were only occasionally observed, so they were not analyzed. For all other behaviours, durations and frequencies were analyzed, and as their results were generally consistent, only the duration results will be reported unless differences were found. Mixed- model ANOVAs assessed the effect of trial number (T1, T2, T3, T4, T5) and group (HR, LR) on the Social Investigation Ratio, the Total Social Investigation and the Total Cylinder
Investigation,. The Greenhouse-Geisser correction for repeated measures was applied, and
Sidak post-hoc tests were performed to assess sources of significance when main effects and/or interactions were found. For each behaviour performed in the test trial (T5), group mean comparisons were planned a priori and conducted using independent samples t-tests.
Comparisons between the HR and LR mice were also planned a priori for habituation
(Total Social Investigation at T1 - Total Social Investigation at T4) and dishabituation (Total
Social Investigation at T5 - Total Social Investigation at T4), and one-sample t-tests determined if there was a significant preference for the novel mouse or its cylinder during T5
(comparison with chance, 0.50), and if habituation occurred (comparison with no change,
0.00). Assessing effects on habituation/dishabituation was important, as some studies use this as the measure of social recognition (e.g., Choleris et al., 2003), and impaired social discrimination may be related to an impairment in habituation and/or dishabituation (e.g.,
Choleris et al., 2006).
254 7.3.6.2 Molecular analysis
As all q-PCR reactions were run in triplicate, any cycle threshold (Ct) values for a particular gene in a particular brain region that was more than 0.5 from both other replicates was removed, as per the manufacturer’s instructions. Ct values (2 or 3) were averaged for each mouse, and normalized by dividing each subject’s average Ct value for each gene product in each brain region by the average Ct for 18S messenger (mRNA) for that subject and region. For each gene product in each brain region, the average amount of gene product in HR animals was set at 1, and the fold difference from average HR was calculated for each mouse using the formula: F(x) = 1.87-(X-Y), where X is the mouse’s normalized Ct (nCt), Y is the average nCt for the HR group, and 1.87 is the increase in cDNA quantity per cycle with our equipment (modified from Murakami et al., 2011). Higher Cts mean that there was less mRNA in the sample, since more cycles (approximately doubling the amount of cDNA with each cycle) were required to reach threshold. The nCt values for each gene in each brain region were compared between HR and LR groups using independent sample t-tests.
7.3.6.3 Behaviour/molecular correlations
Pearson product-moment correlations were calculated between nCts of each gene product within each brain region and the frequency and duration of the habituation, dishabituation, and the Social Investigation Ratio and Total Social Investigation at T1 and T5.
7.4 Results
7.4.1 Behavioural Analysis
Mice in HR group had significantly higher Social Investigation (t(10.09) = 9.95, p <
0.001, Cohen's d = 4.86, power = 1.00) and Cylinder Investigation Ratios at T5 (t(16) = 3.20,
255 p = 0.006, Cohen's d = 1.53, power = 0.86) than LR mice (see Figure 26a). In addition, HR mice preferentially investigated the perforated parts of the two cylinders containing stimulus mice (m(across trials)=109s, m(T5)=101s) over the unperforated parts (m(across trials)=28s, m(T5)=24s), indicating that social investigation and recognition are driven predominantly by olfactory cues passing through the perforations. Planned comparisons indicated that HR and
LR groups differed significantly in social habituation (t(16) = 2.22, p = 0.041, Cohen's d =
1.05, power = 0.54; see Figure 26b), and one-sample t-tests indicated that HR, but not LR, mice habituated (t(7) = 3.47, p = 0.010, Cohen's d = 1.23, power = 0.84) and had a significant preference for the novel mouse (t(7) = 25.88, p < 0.001, Cohen's d = 5.91, power = 1.00) and its cylinder (t(7) = 5.39, p = 0.001, Cohen's d = 2.71, power = 1.00) during T5. This indicated that only HR mice displayed social recognition.
A mixed-model ANOVA showed a significant effect of trial (F(2.5,39.9) = 11.874, p <
0.001, partial η2 = 0.43, power = 1.00), group (F(1,16) = 6.352, p = 0.023, partial η2 = 0.28, power = 0.66), and a significant trial x group interaction (F(2.5,39.9) = 10.753, p < 0.001, partial η2 = 0.40, power = 0.99) for the Social Investigation Ratio (see Figure 26c). Post hoc tests showed that there were significant differences between T5 and all other trials (all p <
0.002), primarily due to effects of trial in HR group (F(1.9,13.2) = 14.056, p = 0.001, partial
η2 = 0.67, power = 0.99), which also showed a significant increase in preference for the novel mouse in T5 compared with all other trials (all p < 0.001). LR mice did not show any significant changes in Social Investigation Ratio across trials. A mixed-model ANOVA showed a significant effect of trial (F(2.8,44.2) = 5.769, p = 0.003, partial η2 = 0.27, power =
0.92) on Total Social Investigation (see Figure 26d). Post hoc tests determined that there were significant differences between T1 and T4 and between T1 and T5 (all p < 0.050),
256 Figure 26. Behavioural differences between high (HR; filled bars/circles) and low (LR; empty bars/triangles) recognizing mice. a) Ratios of investigating the novel mouse (Social Investigation Ratio) or cylinder containing the novel mouse (novel/(novel + familiar) in the test trial (T5). b) Habituation (Total Social Investigation of both stimuli at T4 – Total Social Investigation of both stimuli at T1). c) Social Investigation Ratio across trials. d) Total Social Investigation of both novel and familiar stimulus mice across trials. * indicates a significant difference between HR and LR mice, p < 0.05. † indicates a significant habituation effect (significant difference from 0), p < 0.05.
257 indicating that Total Social Investigation decreased with repeated trials (see Figure 26d).
There were no significant trial, group or interaction effects on Total Cylinder Investigation.
Other behaviours did not differ between HR and LR mice. Importantly, Total Social
Investigation was the same for both groups at T1 (t(16) = 0.986, p = 0.339, Cohen's d = 0.46, power = 0.15), showing equivalent initial social motivation.
7.4.2. Molecular Analysis
In the LA, HR mice had significantly higher Cts than LR mice (indicating that they expressed less mRNA) for all gene products except PR and OT (see Table 33 for exact statistics; see Figure 27a). In the MPOA, HR mice expressed more mRNA for OT and AVP than LR mice, as indicated by HR group’s significantly lower Cts (see Table 33 and Figure
27b). Only significant results are reported.
7.4.3 Behaviour/Molecular Correlations
7.4.3.1. Lateral amygdala
Social Investigation Ratio frequency during T5 correlated positively with AVP nCt (r(13)
= 0.729, p = 0.005, power = 0.86; see Figure 28e), indicating that higher expression of AVP mRNA in the LA is associated with lower Social Investigation Ratio frequency. No significant relationships with the duration measures were found.
7.4.3.2. Medial preoptic area
Total Social Investigation frequency during T5 correlated positively with ERα nCt (r(13)
= 0.758, p = 0.003, power = 0.91; see Figure 29a), indicating that higher expression of ERα mRNA in the MPOA is associated with lower Total Social Investigation frequency during
T5. Habituation frequency correlated negatively with ERα nCt (r(13) = -0.769, p = 0.002, 258 Table 33. Differences in cycle thresholds (Cts) between High Recognition (HR) and Low Recognition (LR) mice by brain region. Gene Product T-Test Effect Size Power Lateral Amygdala ERαa HR > LR (t(11) = 2.26, p = 0.045) 1.307 0.510 ERβb HR > LR (t(10) = 2.28, p = 0.046) 1.308 0.488 OTRc HR > LR (t(11) = 2.66, p = 0.022) 1.301 0.506 AVPd HR > LR (t(11) = 2.72, p = 0.020) 1.593 0.676 AVPR1ae HR > LR (t(11) = 2.41, p = 0.035) 1.348 0.534 AVPR1bf HR > LR (t(11) = 2.96, p = 0.013) 1.701 0.732 Medial Preoptic Area OTg LR > HR (t(10) = -2.49, p = 0.032) 1.391 0.426 AVP LR > HR (t(11) = -2.49, p = 0.030) 1.325 0.437 aEstrogen receptor alpha. bEstrogen receptor beta. cOxytocin receptor. dVasopressin. eVasopressin receptor 1a. fVasopressin receptor 1b. gOxytocin.
259 Figure 27. Levels of mRNA for genes in HR (filled bars) and LR (empty bars) mice, relative to the average amount of gene product in HR animals (set at 1). a) Gene expression in the lateral amygdala. b) Gene expression in the medial preoptic area. * indicates a significant difference between HR and LR mice, p < 0.05.
260
Figure 28. Correlations between social behaviour and mRNA expression in the dorsolateral septum (DS) and lateral amygdala (LA). a) Total Social Investigation in the first trial correlated negatively with DS ERα mRNA, (r(9) = 0.860, p = 0.003). b) Total Social Investigation in the first trial correlated negatively with DS ERβ mRNA, (r(9) = 0.815, p = 0.007). c) Total Social Investigation in the first trial correlated negatively with DS PR mRNA, (r(9) = 0.818, p = 0.007). d) Total Social Investigation in the first trial correlated negatively with DS OTR mRNA, (r(9) = 0.796, p = 0.010). e) Social Investigation Ratio at test (T5) correlated negatively with LA AVP mRNA, (r(13) = 0.729, p = 0.005).
261 Figure 29. Correlations between social behaviour and mRNA expression in the medial preoptic area (MPOA). a) Total Social Investigation in the test trial (T5) correlated negatively with MPOA ERα mRNA, (r(13) = 0.758, p = 0.003). b) Habituation (Total Social Investigation in T4 – Total Social Investigation in T1) correlated positively with MPOA ERα mRNA, (r(13) = -0.769, p = 0.002). c) Habituation (Total Social Investigation in T4 – Total Social Investigation in T1) correlated positively with MPOA ERβ mRNA, (r(13) = -0.747, p = 0.003). d) Habituation (Total Social Investigation in T4 – Total Social Investigation in T1) correlated positively with MPOA OT mRNA, (r(12) = -0.832, p = 0.001). e) Habituation (Total Social Investigation in T4 – Total Social Investigation in T1) correlated positively with MPOA AVP mRNA, (r(13) = -0.761, p = 0.003). 262 power = 0.92; see Figure 29b), ERβ nCt (r(13) = -0.747, p = 0.003, power = 0.89; see Figure
29c), OT nCt (r(12) = -0.832, p = 0.001, power = 0.96; see Figure 29d), and AVP nCt (r(13)
= -0.761, p = 0.003, power = 0.91; see Figure 29e) indicating that mRNA expression and
Habituation vary together. No significant relationships were found with the duration measures.
7.4.3.3. Dorsolateral Septum
Duration of Total Social Investigation during T1 correlated positively with ERα nCt (r(9)
= 0.860, p = 0.003, power = 0.92; see Figure 28a), ERβ nCt (r(9) = 0.815, p = 0.007, power =
0.84; see Figure 28b), PR nCt (r(9) = 0.818, p = 0.007, power = 0.85; see Figure 28c), and
OTR nCt (r(9) = 0.796, p = 0.010, power = 0.81; see Figure 28d), indicating that higher mRNA expression in the DS is associated with lower Total Social Investigation duration during T1.
7.5 Discussion
Overall, consistent with our selection of HR and LR groups, we found that HR mice scored high on all measures of social recognition (see Figure 26), while the behaviour of the
LR mice did not demonstrate any social recognition. We also found that social recognition and levels of social investigation (indicative of social interest) were closely linked to gene expression in the brain.
7.5.1 Lateral Amygdala
Lower expression of mRNA for all genes expressed, except for PR and OT, in the LA of
HR mice (see Figure 27a), taken together with the observed inverse relationship between
AVP and the social investigation ratio at test (see Figure 28e), supports the hypothesis that 263 estrogens may modulate social recognition by acting on the OT and/or AVP systems in this brain region.
The LA has been implicated in both aversive and appetitive associative learning (e.g.,
Cardinal et al., 2002; Gaskin & White, 2006), and it is possible that the LR mice may have a general learning impairment due to the increased activation of the majority of the genes we studied in this region. Several studies have found that improved learning is associated with decreased gene expression or brain region activation (e.g., Cacioppo & Decety, 2011; Mong et al., 2003a; Mong et al., 2003b; Ribeiro et al., 2009; Sun et al., 2005), including, for example, those associated with estrogens administration (Mong et al., 2003a; Ribeiro et al.,
2009). This suggests that the lower LA gene expression observed in HR mice may indicate increased efficiency of LA-mediated learning mechanisms that would reflect better performance in the social recognition learning paradigm.
Alternatively, as the LA has been repeatedly shown to mediate fear and anxiety (e.g.,
Markram et al., 2008; Misslin & Ropartz, 1981; Nachman & Ashe, 1974; Slotnick, 1973), the higher expression of genes in the LR group suggests these mice may experience greater anxiety than HR mice. When considered with previous results implying that anxiety may modulate social recognition, affiliation and/or investigation (e.g., Everts & Koolhaas, 1999;
Landgraf & Wigger, 2002; Ophir et al., 2009; Taylor et al., 2000), this suggests that the observed results could be partially explained by differences in anxiety. However, we observed no differences in anxiety-related behaviours, including stretched approaches and amount of investigation at T1, when both animals were novel. Thus, it is unlikely that anxiety was increased sufficiently to impair social recognition by itself.
264 Most research on social recognition in rodents has focused on the medial amygdala
(Choleris et al., 2007; Ferguson et al., 2001; Samuelsen & Meredith, 2011), and it is unclear why no differences were found in the current study. Differences in mRNA do not always translate to differences in proteins, and it is possible that had we measured levels of protein, we would have found more ERα and OTR in the medial amygdala of the high recognition mice as predicted by previous studies (e.g., Choleris et al., 2007; Spiteri et al., 2010). The present results suggest that optimal levels of activation of the genes for receptors for estrogens and their possible downstream targets OT and AVP in the LA may be required for social recognition; hence, the LA may be part of the brain circuit involved in the detection and recognition of social stimuli. Indeed, a general social behaviour circuit has been proposed that includes the MPOA, DS, and amygdala, and these structures appear to be involved in key social interactions and their modulation by sex-steroid hormones (Newman,
1999).
7.5.2 Medial Preoptic Area
In the MPOA, ER mRNA expression was related to habituation, as were levels of both
OT and AVP mRNA (see Figure 29b-e). Additionally, HR mice expressed more OT and
AVP than LR mice in this region (see Figure 27b), which is a source of both peptides (Bosch et al., 2010; Frank & Landgraf, 2008) and exhibits sexual dimorphism in estrogens, progesterone, OT and AVP (Bales et al., 2007; Dubois-Dauphin et al., 1996; Hill et al., 2008;
Rood et al., 2008; Sakuma, 2009; Yamamoto et al., 2006).
Increased OT in the MPOA may be related to social interest in HR mice. No Fos activation has been observed in the MPOA of recognition-impaired OTKO males, unlike WT
265 mice, following a brief social interaction, and OT administration to this region rescued social recognition in OTKO mice (Ferguson et al., 2002). In addition, OT administration to the
MPOA also prolonged (Popik & van Ree, 1991; Popik & van Ree, 1992) and improved
(Ferguson et al., 2002) social recognition in male rats and mice, while administration of an
OT antagonist to the MPOA blocked social recognition in WT mice (Ferguson et al., 2002).
AVP antagonist or antiserum treatment only impaired males (Bluthé et al., 1990; Bluthé
& Dantzer, 1990; Ferguson et al., 2002; Kogan et al., 2000), which suggests that female social recognition does not depend on AVP (Choleris et al., 2009). However, systemic AVP can improve social recognition in both sexes (Choleris et al., 2009; Ferguson et al., 2002;
Kogan et al., 2000; Le Moal et al., 1987), and although AVP in the MPOA, especially in females, has received little investigation, our results suggest that higher levels of AVP in the
MPOA may relate to better social recognition even in female mice. The correlations with
ERα and ERβ, as well as with OT and AVP mRNA, in our female mice suggest that, as has been proposed for other brain regions (e.g., Choleris et al., 2003; Choleris et al., 2006),
MPOA OT and AVP involvement in female social recognition may be mediated by ERs.
The involvement of MPOA OT and AVP in social recognition, as well as in numerous social behaviours, including sexual and parental behaviours (Bosch & Neumann, 2008;
Bosch & Neumann, 2010; Champagne et al., 2001; Champagne et al., 2003; Curtis et al.,
2007; de Bono, 2003; Kendrick et al., 1988; Lubin et al., 2003; Neumann, 2008; Nunes,
2007; Pedersen, 1997; Shahrokh et al., 2010) suggests a fundamental role of these hormones in this brain region in the evolution of sociality. Complex social behaviours that depend upon social recognition (e.g., dominance hierarchies) have likely evolved from basic forms
266 of social behaviour involved in mating and reproduction (Curtis et al., 2007), probably developing first through philopatry (the delay of dispersal; Solomon, 2003) and cooperation with kin and then expanding to more complex social behaviour and social systems (Nunes,
2007). Consistently, AVP and OT are highly involved in maternal behaviour: MPOA OTR and AVPR1a binding are higher in lactating than virgin dams, and MPOA release of AVP is increased in the presence of pups; similar results have been observed in the BNST (Bosch et al., 2010; Bosch & Neumann, 2008). Extension of MPOA social function from reproduction to social recognition may have been a first step in the evolution of sociality. Subsequent involvement of other so-called “social brain” regions would have led to the evolution of complex social behaviours and animal societies (Alexander, 1974; Amdam et al., 2006;
Dunbar & Schultz, 2007). The involvement of MPOA OT and AVP in social recognition suggests that LR mice might be generally impaired in social behaviour, including more evolutionarily primitive reproduction-related social behaviours from which other social behaviours may have developed (Curtis et al., 2007).
7.5.3 Dorsolateral Septum
While no group differences in behaviour or gene expression were observed in the DS, possibly due to small DS sample size (nHR = 4, nLR = 5), social investigation at T1 had an inverse relationship with mRNA for receptors of both ovarian hormones and for OTR in this region (see Figure 28a-d). These receptors are known to be related to social behaviours
(reviewed in Choleris et al., 2009), and our results are consistent with lower OTR levels observed in high investigating prairie voles (Ophir et al., 2009). Taken together, these results point to the involvement of septal female gonadal hormone receptors and OTR in social interest and approach.
267 7.5.4 Overall Conclusions
Our results support the idea that variation in ovarian hormone receptor expression could be related to individual differences in social recognition, at least partly through modulation of the OT and/or AVP systems, particularly in the DS, LA and MPOA. As mRNA levels do not always reflect differences in protein levels, it should be noted that additional studies are necessary to determine how the observed differences in gene expression relate to levels of receptors and/or hormones, and whether these occur in the same brain regions as the mRNA changes or in regions to which they project.
Receptors for estrogens, progesterone and OT may be important for the regulation of initial social interest in the DS, which could relate to increased attention to social stimuli.
Additionally, the LR's increased activation of the majority of gonadal and hypophyseal hormones tested in the LA may relate to general learning impairments, while lower mRNA levels in the HR mice could be indicative of increased efficiency of social recognition learning and/or memory mechanisms. Higher OT and AVP mRNA levels in the MPOA could facilitate social recognition by releasing more OT and AVP into the lateral septum and
LA, thus facilitating social recognition, as both the DS and the amygdala receive projections from hypothalamic OT- and AVP-expressing neurons (Gimpl & Fahrenholz, 2001; Rood &
De Vries, 2011).
268
CHAPTER 8:
General Discussion and Conclusions
269 The goal of this thesis was to explore the involvement of estrogens, their downstream effectors OT and AVP, and their receptors in social cognition and social behaviour. This was pursued using pharmacological techniques and analysis of gene expression, and investigated three types of social behaviour, social interactions (aggression), social learning, and social recognition.
8.1 Summary of Social Interaction and Aggression
Our use of an ethological analysis, which looked at more than just attacks as a measure of agonistic or aggressive behaviour, allowed us to show that while both sexes perform agonistic behaviour, its nature differs. This is likely due to different functions of such behaviour, as overt attacks are used by males that need to establish exclusive territories and regulate space use (Clipperton-Allen et al., 2010; Clipperton-Allen et al., 2011a; Ginsburg &
Allee, 1942; Miczek et al., 2001; Scott & Fredericson, 1951), while females tend to establish a dominance hierarchy that will allow them to share territory, and thus use behaviours aimed at allowing a dominance hierarchy to form (Alleva, 1993; Branchi et al., 2006; Clipperton et al., 2008; Clipperton-Allen et al., 2010; Clipperton-Allen et al., 2011a; Grant & Mackintosh,
1963; Miczek et al., 2001; Miczek et al., 2007; Pietropaolo et al., 2004). These findings highlight the need for more research on female intrasexual aggression using ethologically- based, sex-appropriate measures to allow for better comparison between sexes and understanding of agonistic behaviour, especially in light of recent increases in attention to and investigation of so-called “relational aggression” (i.e., bullying) and aggression in school-aged girls (Leckie, 1997; National Center for Mental Health Promotion and Youth
Violence Prevention, 2011; O'Neil, 2008; Young Women's Christian Association, 2002).
270 We thus investigated the effects of ER-specific agonists on aggression in both sexes using identical resident-intruder tests. Our results indicated that acute activation of ERα increased sex-typical agonistic behaviour in gonadex but not gonadally intact male and female mice (Clipperton-Allen et al., 2011a). These results would be consistent with ERα activation “turning on” agonistic behaviour in gonadex mice, which have little circulating endogenous estrogens and little or no ERβ activation (Clipperton-Allen et al., 2011a).
Activation of ERβ, on the other hand, may be more effective in animals with naturally circulating gonadal hormone levels (thus activating agonistic behaviour through ERα), allowing for additional ERβ activation to modulate these behaviours (Clipperton-Allen et al.,
2011a).
These results are consistent with the findings from αERKO and βERKO male mice
(Nomura et al., 2006; Ogawa et al., 1998b; Ogawa et al., 1999; Scordalakes & Rissman,
2003), and together suggest that ERα activation leads to increased aggression and agonistic behaviour (see Table 34). However, our results are inconsistent with those in the literature regarding αERKO and βERKO females (see Table 34); we found that PPT-treated females exhibited more dominance-related, sex-typical agonistic behaviour, while the KO literature suggests that ERα decreases female aggression (Ogawa et al., 1998a; Ogawa, 2004; Ogawa et al., 2004; Ogawa et al., 1999). There are several possible explanations for this, including
1) that the KO studies primarily analyzed overt attacks, and did not assess other dominance- related behaviours, 2) that they were based on different paradigms, as maternal and T- mediated agonistic behaviour can be distinctly different (Al-Maliki et al., 1980; Svare &
Gandelman, 1973), or 3) that developmental and activational effects of ERα may be different in females.
271 Table 34. Effects of manipulations of estrogen receptors: implied effects of estrogen receptors on aggression. ERα ERβ Manipulation Males Females Males Females KO ↑ ↓ ↓ ↑ ER-Selective Agonist ↑ ↑ ↑ ↑
272 Like ERα in females, ERβ appears to have different effects in males when acutely activated than when the gene for ERβ is KO. We found that acute ERβ activation by WAY-
200070 increased non-attack agonistic behaviour in gonadally intact mice (Clipperton-Allen et al., 2010), but male βERKO mice were more aggressive, and αERKOs less aggressive, than WT (Nomura et al., 2002; Nomura et al., 2006; Ogawa et al., 1998b; Ogawa et al., 1999;
Scordalakes & Rissman, 2003), suggesting that ERβ activation would reduce agonistic behaviour (see Table 34). Although this could suggest different outcomes from developmental and activational manipulations of ERβ, as with the ERα effects in females above, this discrepancy could also be due to different behavioural analyses, as the studies typically did not look at aggression other than attacks. Our observed increase in agonistic behaviour in WAY-200070 treated gonadally intact females (Clipperton-Allen et al., 2010) is more consistent with the KO literature, in which βERKO females showed less aggression in response to T, and αERKO females exhibited increased territorial aggression (Ogawa et al.,
1998a; Ogawa et al., 2004; Ogawa et al., 1999). This suggests that ERβ may modulate the less overt agonistic behaviours used to establish dominance.
Taken together, the above results suggest that ERs may regulate overt and non-overt aggression differently (e.g., Mos & Olivier, 1989), and that they may have different effects in the two sexes. Further research is required to clarify these differences, particularly in terms of studying agonistic territorial behaviour in similar situations in both males and females.
ERα may be required to 'activate' aggression, while ERβ may have more subtle modulatory effects (Ogawa, 2004). The two ERs also appear to have different efficacy in different gonadal conditions, with ERα activation being more effective in gonadex animals, and ERβ having more effects in gonadally intact mice. This could be due to reciprocal regulation of
273 transcription by the two ERs (e.g., Nomura et al., 2003) or modulation of each receptor's response to estrogens by the other receptor (e.g., ERβ decreases the response of ERα to estrogens; Cowley et al., 1997; Hall & McDonnell, 1999). Similarly, PPT may be more effective in gonadex animals, with little to no circulating estrogens, because ERα's transcriptional activity is not reduced by ERβ (Cowley et al., 1997; Hall & McDonnell,
1999). ERα is also unlikely to form heterodimers, which are as effective as ERα homodimers (Cowley et al., 1997) but may have opposite effects (Hall & McDonnell, 1999;
Lindberg et al., 2002), as the absence of circulating estrogens results in little or no activation of ERβ. WAY-200070, however, may be more effective in gonadally intact animals, which have physiological levels of estrogens, resulting in activation of both ERα and ERβ; this increases the likelihood of the formation of heterodimers, which are more effective and form more easily that ERβ homodimers (Couse et al., 1997; Cowley et al., 1997). Indeed, it has been suggested that ERβ may need to heterodimerize to function (Couse et al., 1997), although work with αERKO mice and ERβ agonists in ovariectomized animals do not support this.
One way ERs may affect aggression is through modulation of the OT and/or AVP systems, both of which have been implicated in agonistic behaviours. OTRKO (Nishimori et al., 2008), and possibly OTKO (Ragnauth et al., 2005; Winslow et al., 2000; but see DeVries et al., 1997; Nishimori et al., 2008), mice show increased aggression. Further, OT inhibits aggression (Ferris, 2005; McCarthy, 1990), although it increases maternal aggression (Ferris et al., 1992), and OT antagonism increases aggression (Giovenardi et al., 1998; Lubin et al.,
2003). Septal release of both OT and AVP have also been shown to affect aggression
(Engelmann et al., 2000). Further, septal AVP correlates negatively with intermale
274 aggression in mice (Compaan et al., 1993), and its effects on aggression are sexually dimorphic and under the control of gonadal steroids (DeVries et al., 1997). Consistently,
AVP administration increases aggression (Ferris & Delville, 1994), and KO or antagonism of
AVPR1a or AVPR1b blocked or decreased aggression (Blanchard et al., 2005; Caldwell &
Albers, 2004; Ferris et al., 2006; Ferris et al., 1992; Ferris et al., 1997; Ferris & Potegal,
1988; Wersinger et al., 2003; Wersinger et al., 2002; Wersinger et al., 2004; Wersinger et al.,
2007a; Wersinger et al., 2007b).
8.2 Summary of Social Learning of Food Preferences
Estrogens were clearly shown to affect social transmission of food preferences, with acute estrogen and ERβ agonists generally prolonging the preference for the demonstrated food, acute ERα agonist blocking it, and a prolonged preference shown by all mice receiving chronic treatment with any of the compounds (see Table 35).
Our results suggest that acute estrogens treatment 48h prior to testing may improve consolidation, as treatment before or immediately after acquisition prolonged the demonstrator food preference, but had no apparent effect on the retrieval or expression of this preference. This is consistent with non-social learning research, which has failed to find effects of estrogens when administered more than 2h post-training or post-acquisition
(reviewed in Frick, 2009; Packard, 1998), and with improvements on both social and non- social recognition tasks within 40 min of estrogens treatment (Phan et al., 2011).
Acute PPT treatment may block the STFP by impairing acquisition, consolidation, retrieval and/or expression, or some combination of the three (Clipperton et al., 2008;
Clipperton-Allen et al., 2011e), as it blocked the STFP regardless of its time of 275 Table 35. Effects of estrogens and estrogen receptor specific agonists on social transmission of food preferences. Treatment Estradiol ERα agonist ERβ agonists Administration Paradigm Benzoate PPT WAY-200070, DPN Prolonged Prolonged Prolonged (DPN) Chronic [4-6 h] [4 h] [4-6 h] Prolonged Prolonged Pre-acquisition Blocked [4 h] [4 h] Prolonged Immediate Post-acquisition Blocked Blocked (DPN) [4 h] 24h Delayed Post- Blocked (DPN) and No Effect Blocked acquisition Prolonged (DPN) [6 h]
276 administration. Further investigation, with treatments that last for a briefer time to allow for the system to return to baseline, are needed to tease this apart. While this is consistent with some non-social learning paradigms (e.g., passive avoidance, morphine-induced conditioned place preference; Chen et al., 2002; Esmaeili et al., 2009; Hebert & Dash, 2004; Igaz et al.,
2006), PPT also often has no effect on non-social learning tasks in rats (e.g., Morris water maze, radial arm maze; Jacome et al., 2010b; Rhodes & Frye, 2006).
While the involvement of estrogens and ERα appear clearer, ERβ's effects on the STFP are more complicated. When administered prior to the social interaction, ERβ may affect motivation or the salience of the social cue, as the duration of the preference for the demonstrator's food may be related to the nature of the social interaction, specifically the submissiveness of the observer (Clipperton et al., 2008; Clipperton-Allen et al., 2011b).
Consistently, learning from dominant conspecifics has been previously found to be better than learning from submissive demonstrators (Choleris et al., 1998b; Kavaliers et al., 2005a).
When administered at either post-acquisition time, however, ERβ agonist DPN may interfere with consolidation, as memories are vulnerable to disruption by drug treatment immediately following acquisition (Abel & Lattal, 2001b), or impair performance by affecting 5HT or MAPK activation.
Unlike the middle dose of DPN, the high DPN dose prolonged the preference to 3x that of controls when administered 24h after the interaction. It is unclear why these doses had opposite effects, but as the benzodiazepine chlordiazepoxide (CDP) had similar opposing dose effects on social learning in gerbils (Choleris et al., 1998b), it may relate to the anxiolytic effects of both ERβ and CDP (reviewed in Choleris et al., 1998b; Choleris et al., 277 2008a).
When activated chronically, ERα, ERβ or both prolonged the preference to 2-3x the duration of vehicle and uninjected controls (Clipperton-Allen et al., 2011e), effects that are consistent with improving effects of chronic EB, PPT and DPN on non-social learning tasks
(Bowman et al., 2002; Daniel et al., 1997; Daniel, 2006; Gibbs, 2000; Hammond et al., 2009;
Heikkinen et al., 2004; Heikkinen et al., 2002; Luine et al., 1998; Markham et al., 2002;
Sandstrom & Williams, 2004). The effects of chronic EB and DPN parallel the acute pre- acquisition treatments, although only acute ERβ agonist treatment led to a relationship between submissiveness of the observer during the social interaction and the duration of preference; chronic PPT, however, affects the STFP differently depending on the length of treatment. The similarity of treatment effects suggest that chronic treatment with PPT or
DPN may result in a loss of specificity.
As with social interaction and aggression above, and social recognition below, the ERs could be mediating the STFP through modulation of OT, AVP and their receptors. Both OT and AVP have been shown to improve performance on this task, although if the task is too easy, AVP may impair performance (Bunsey & Strupp, 1990; Cushing & Wynne-Edwards,
2006; Popik & van Ree, 1991; Strupp et al., 1990).
8.3 Summary of Social Recognition
As previous research has indicated that ERs, as well as their downstream targets OT,
AVP and their receptors mediate social recognition (reviewed in Choleris et al., 2009), the relationship between these proteins and natural variation in social recognition in an outbred mouse strain was investigated. Results indicated that improved social recognition may relate 278 to decreased expression of ERs, OTRs, AVPRs and especially AVP in the LA. In the sexually dimorphic MPOA, higher OT and AVP mRNA were present in the mice that showed recognition than those that appeared unable to recognize familiar conspecifics, and habituation, an indication of social recognition, was related to ERs as well as OT and AVP.
These data thus support the idea that ERs modulate social recognition through modulations of OT and AVP systems (e.g., Choleris et al., 2003). Expression of receptor mRNA for estrogens, progesterone and oxytocin in the DS related to initial social interest, which could indicate increased motivation for social contact or increased attention to social stimuli.
These results are consistent with a proposed social behaviour circuit in which the MPOA expresses more OT and AVP mRNA, and then releases increased amounts of these proteins into the DS and LA, both of which receive hypothalamic OT and AVP neuron projections
(Gimpl & Fahrenholz, 2001; Rood & De Vries, 2011), leading to social recognition facilitation.
8.4 Estrogens and Social Cognition
The results of our and others' research suggest that estrogens, OT and AVP are all involved in social behaviours and mediate social recognition, social learning, social interactions, and aggression. Both ERs act in the same direction, albeit to different degrees, on social recognition, and have similar effects on agonistic behaviour, but appear to act in opposition on social learning. Thus, ERs differently modulate the two types of social learning investigated here, as ERα is critical for social recognition, but impairs social learning, while ERβ is less important in social recognition, and prolongs the preference for the demonstrated food in the STFP. One possibility is that these differences relate to the key brain areas involved in the different types of social cognition. For example, the MPOA and
279 lateral septum appear more important for social recognition than the STFP, and both areas may express more ERα than ERβ (Merchenthaler et al., 2004b; Mitra et al., 2003b). For areas like the hippocampus and parahippocampal cortex, which appear more involved in the
STFP than social recognition, we see the opposite pattern of receptor expression, with higher
ERβ than ERα expression (Merchenthaler et al., 2004b; Mitra et al., 2003b).
The results of investigation of the two ERs suggest that they may play different roles in the behaviours they modulate, regardless of the direction of their effects. For example, ERα may be predominantly involved in promoting reproduction and sexual behaviour, possibly in part by improving recognition (to assist in potential mate identification), and inhibits any behaviours which might reduce the likelihood of successful reproduction (e.g., feeding; reviewed in Choleris et al., 2009; Choleris et al., 2008a). ERβ may also promote reproduction but less directly, inhibiting the effects of potentially interfering experiences or emotions (e.g., fear, anxiety, pain), and promoting behaviours that allow for good analysis of environmental conditions for reproduction, including, for example, the availability of food sources and the familiarity of nearby conspecifics (i.e., social and non-social learning, social recognition; Choleris et al., 2009; Choleris et al., 2008a). Consistently, females perform best on social recognition and social learning tasks in proestrus, when they are reproductively active, and both ERs may indirectly promote reproduction by increasing dominance-related behaviours, which may lead to a greater likelihood of mating (Creel, 2001; Miczek et al.,
2001). When taken together with the fact that ERs, OT and AVP are all involved in social, sexual and parental behaviour, this suggests that these hormones may have evolved from basic reproduction-related mechanisms to modulate more advanced social behaviours, possibly through a general social behaviour circuit that includes the MPOA, DS and
280 amygdala, brain regions that are involved in key social interactions and modulated by gonadal hormones (Newman, 1999).
The precise mechanisms by which ERs, OT and AVP modulate social behaviour and social learning are as yet unclear, although there are several excellent candidates. Estrogens have been found to modulate synapses through increases in dendritic spines as well as LTP,
LTD and GABA synthesis (reviewed in Frick, 2009; Phan et al., 2011), possibly through
ERβ (Liu et al., 2008). Estrogens also modulate PKA and MPAK activity (Boulware et al.,
2005), and CREB phosphorylation is inhibited by both ERα and ERβ selective agonists
(Boulware et al., 2005). Additionally, estrogens mediate a number of neurotransmitter systems in addition to OT and AVP, including the cholinergic, serotonin, dopamine and norepinephrine systems (Frick, 2009; Jacome et al., 2010b; Johnson et al., 2010; Morissette et al., 2008), again potentially through ERβ action (Frick, 2009; Jacome et al., 2010b;
Johnson et al., 2010; Morissette et al., 2008). ERα may modulate hippocampal inhibitory tone (Donner & Handa, 2009b; McEwen, 2001), increase LTD in the hippocampus (Mukai et al., 2007), regulate acetycholine release and reuptake (Spencer et al., 2008), and inhibit ERK phosphorylation (Singh et al., 2000).
8.5 Significance and Applications
Many cognitive behaviours are related to, or dependent on, social interactions, and the majority of human learning is either social, or takes place in a social context (Morgan et al.,
2011). However, relatively little is known of the neurobiology underlying social behaviour, and more investigation is needed to elucidate the mechanisms by which it functions both normally and abnormally. The serious impact of impaired social functioning is clear when
281 one looks at individuals with autism spectrum disorders (ASD), who have deficits in social interactions, communication and social learning (Autism Society of Canada, 2010).
One of the most significant problems in this highly prevalent disorder, which may affect up to 1:110-200 individuals (CDC & Autism and Developmental Disabilities Monitoring
Network Surveillance, 2009; Fombonne, 2003a; Fombonne, 2003b), is impaired social recognition, as this is key to forming human relationships. Many ASD individuals appear not to notice or recognize people (Baird et al., 2003), and have deficits in social learning and memory, including imitation (Oberman & Ramachandran, 2007; Senju et al., 2007). ASD children pay more attention to and learn better from a computer than in a social context or from a teacher (Moore & Calvert, 2000; Rehfeldt et al., 2004; Shane & Albert, 2008;
Stephens & Ludy, 1975).
ASD are sexually dimorphic, with 80-90% males (Gillberg, 2002). This suggests gonadal hormone involvement, and is consistent with estrogens' mediation of OT, which has been strongly implicated, with AVP, in ASD. Lower plasma OT and alterations in the OT system have been observed in ASD children (reviewed in Choleris et al., 2004), and intravenous OT administration in human adults has been shown to reduce some ASD symptoms (Lim et al., 2005). Additionally, rodents with deficient OT and/or AVP, which are impaired in social recognition, have been proposed as models of ASD (Lim et al., 2005).
Social behaviour is particularly important for women, as social interactions are frequently used to cope with stress (Taylor et al., 2000). Peri- or post-menopausal disorders, which can include reduced sociability, cognitive decline, and mood disorders (Imamov et al., 2005;
Koehler et al., 2005), are an increasing problem as the “baby boom” generation ages. With 282 the increase in life expectancy, many women will spend more than 35% of their lives in a post-menopausal state (Morrison et al., 2006), and 75% of women will experience symptoms associated with this lack of gonadal hormones (Stovall & Pinkerton, 2008). While hormone replacement therapy (HRT) can counteract many of these symptoms (e.g., cognitive decline;
Resnick et al., 2009; Sherwin, 1988; Wharton et al., 2009), the publication of HRT investigations, like the Women's Health Initiative (WHI) and the Million Women Study, that emphasized the potential negative side effects (e.g., increased cancer risks; Beral & Million
Woman Study Collaborators, 2003; Koehler et al., 2005; Resnick et al., 2004; Shumaker et al., 2003; Writing Group for the WHI Investigators, 2002), despite methodological concerns with these studies (Asthana et al., 2009; Robb-Nicholson, 2009; Sherwin, 1988), resulted in an 18-37% decline in HRT use from 2000-2004 (Hersh et al 2004; Usher et al 2006; Faber et al 2006). However, some doctors, particularly those who emphasize quality over quantity of life, still recommend HRT as a treatment for menopausal symptoms (Sturmburg & Pond,
2009). There is a need for new treatments to avoid the negative side effects (Nath & Sitruk-
Ware, 2009), and the development of drugs that specifically target ERβ, which has positive cognitive effects (reviewed in Choleris et al., 2008a), but avoids, for example, the increased breast cancer risk, has been proposed (Hertrampf et al., 2008; Koehler et al., 2005).
However, the results of my investigations of estrogens and estrogen receptor specific agonists suggests that caution must be employed, as major differences in outcomes may be found between acute and chronic treatments, especially as long-term exposure to the currently available estrogen receptor specific agonists may lead to a loss of specificity and thus eliminate the advantage of selectively targeting one receptor.
While much is still undiscovered about social cognition and behaviour, and their
283 neurobiological underpinnings, the results of this thesis suggest that estrogens, through the various systems they modulate (e.g., OT, AVP, DA, 5HT), have a key role to play. Further investigations of how estrogens effect change in these systems at the molecular and cellular level, as well as the critical brain areas and downstream effectors involved in these complex behaviours, are needed, especially because of the critical importance of sociality, social behaviour and social cognition in our own lives.
284 Reference List
Abel T, Lattal KM (2001b). Molecular mechanisms of memory acquisition, consolidation and retrieval. Curr Opin Neurobiol, 11: 180-187.
Abel T, Lattal KM (2001a). Molecular mechanisms of memory acquisition, consolidation and retrieval. Current Opinion in Neurobiology, 11: 180-187.
Ahmed EI, Zehr JL, Schulz KM, Lorenz BH, DonCarlos LL, Sisk CL (2008). Pubertal hormones modulate the addition of new cells to sexually dimorphic brain regions. Nat Neurosci, 11: 995-997.
Akinci MK, Johnston GAR (1997). Sex differences in the effects of gonadectomy and acute swim stress on GABAA receptor binding in mouse forebrain membranes. Neurochem Int, 31: 1-10.
Al-Maliki S, Brain PF, Childs G, Benton D (1980). Factors influencing maternal attack on conspecific intruders by lactating female “TO” strain mice. Aggress Behav, 6: 103-117.
Albert DJ, Jonik RH, Walsh ML (1990). Aggression by ovariectomized female rats with testosterone implants: competitive experience activates aggression toward unfamiliar females. Physiol Behav, 47: 699-703.
Albert DJ, Petrovic DM, Walsh ML (1989). Ovariectomy attenuates aggression by female rats cohabiting with sexually active sterile males. Physiol Behav, 45: 225-228.
Alexander RD (1974). The evolution of social behavior. Annu Rev Ecol Evol Sys, 34: 517- 547.
Alleva E (1993). Assessment of aggressive behavior in rodents. In: Conn PM (ed.), Paradigms for the study of behavior. Methods in neurosciences, Vol. 14, pp. 111-137.
Alvarez P, Lipton PA, Melrose R, Eichenbaum H (2001). Differential effects of damage within the hippocampal region on memory for a natural, nonspatial odor-odor association. Learn Mem, 8: 79-86.
Alvarez P, Wendelken L, Eichenbaum H (2002). Hippocampal formation lesions impair performance in an odor-odor association task independently of spatial context. Neurobiol Learn Mem, 78: 470-476.
Amdam GV, Csondes A, Fondrk MK, Page REJ (2006). Complex social behaviour derived from maternal reproductive traits. Nature, 439: 76-78.
Anton AH (1969). Effect of group size, sex and time on organ weights, catecholamines and behavior in mice. Physiol Behav, 4: 483-487.
285 Anton AH, Serrano A, Beyer RD, Lavappa KS (1974). Tetrahydrocannabinol, sesame oil and biogenic amines: A preliminary report. Life Sci, 14: 1741-1746.
Appenrodt E, Juszczak M, Schwarzberg H (2002). Septal vasopressin induced preservation of social recognition in rats was abolished by pinealectomy. Behav Brain Res, 134: 67- 73.
Archer J (1988). The behavioural biology of aggression.
Arteaga-Silva M, Rodriguez-Dorantes M, Baig S, Morales-Montor J (2007). Effects of castration and hormone replacement on male sexual behavior and pattern of expression in the brain of sex-steroid receptors in BALB/c AnN mice. Comp Biochem Physiol A, 147: 607-615.
Asthana S, Brinton RD, Henderson VW, McEwen BS, Morrison JH, Schmidt PJ et al. (2009). Frontiers proposal. National Institute on Aging “bench to bedside: estrogen as a case study”. Age, 31: 199-210.
Autism Society of Canada (2010). Understanding ASDs. Autism Society of Canada,
Baird G, Cass H, Slonims V (2003). Diagnosis of autism. BMJ, 327: 488-493.
Bales KL, Plotsky PM, Young LJ, Lim MM, Grotte N, Ferrer E et al. (2007). Neonatal oxytocin manipulations have long-lasting, sexually dimorphic effects on vasopressin receptors. Neuroscience, 144: 38-45.
Barkhem T, Carlsson B, Nilsson Y, Enmark E, Gustafsson J-Å, Nilsson S (1998). Differential response of estrogen receptor α and estrogen receptor β to partial estrogen agonists/antagonists. Mol Pharmacol, 54: 106-112.
Bartolomucci A, Palanza P, Gaspani L, Limiroli E, Panerai AE, Ceresini G et al. (2001). Social status in mice: behavioral, endocrine and immune changes are context dependent. Physiol Behav, 73: 401-410.
Bartolomucci A, Palanza P, Parmigiani S (2002). Group housed mice: are they really stressed? Ethol Ecol Evol, 14: 341-350.
Bean NJ, Galef BGJ, Mason JR (1988). The effect of carbon-disulfide on food-consumption by house mice. J Wildl Manag, 52: 502-507.
Beauchamp JK, Yamazaki K (2003). Chemical signalling in mice. Biochem Soc Trans, 31: 147-151.
Beeman EA (1947). The effect of male hormone on aggressive behavior in mice. Physiol Zool, 20: 373-405.
286 Benelli A, Bertolini A, Poggioli R, Menozzi B, Basaglia R, Arletti R (1995). Polymodal dose-response curve for oxytocin in the social recognition test. Neuropeptides, 28: 251- 255.
Beral V, Million Woman Study Collaborators (2003). Breast cancer and hormone- replacement therapy in the Million Women Study. Lancet, 362: 419-427.
Berger-Sweeney J, Stearns NA, Frick KM, Baxter MG (2000). Cholinergic basal forebrain is critical for social transmission of food preferences. Hippocampus, 10: 729-738.
Bielsky IF, Hu S-B, Ren X, Terwilliger EF, Young LJ (2005). The V1a vasopressin receptor is necessary and sufficient for normal social recognition: a gene replacement study. Neuron, 47: 503-513.
Bielsky IF, Hu S-B, Szegda KL, Westphal H, Young LJ (2004). Profound impairment in social recognition and reduction in anxiety-like behaviour in vasopressin V1a receptor knockout mice. Neuropsychopharmacology, 29: 483-493.
Blanchard RJ, Griebel G, Farrokhi C, Markham C, Yang M, Blanchard DC (2005). AVP V1b selective antagonist SSR149415 blocks aggressive behaviors in hamsters. Pharmacol Biochem Behav, 80: 189-194.
Blanchard RJ, Yudko EB, Rodgers RJ, Blanchard DC (1993). Defense system psychopharmacology: an ethological approach to the pharmacology of fear and anxiety. Behav Brain Res, 58: 155-165.
Blumstein DT, Ebensperger LA, Hayes LD, Vásquez RA, Ahern TH, Burger JR et al. (2010). Toward an integrative understanding of social behavior: new models and new opportunities. Front Behav Neurosci, 4: 1-9.
Bluthé R-M, Dantzer R (1992). Chronic intracerebral infusions of vasopressin and vasopressin antagonist modulate social recognition in the rat. Brain Res, 572: 261-264.
Bluthé R-M, Dantzer R (1990). Social recognition does not involve vasopressinergic neurotransmission in female rats. Brain Res, 535: 301-304.
Bluthé R-M, Schoenen J, Dantzer R (1990). Androgen-dependent vasopressinergic neurons are involved in social recognition in rats. Brain Res, 519: 150-157.
Bodo C, Rissman EF (2006). New roles for estrogen receptor β in behavior and neuroendocrinology. Front Neuroendocrinol, 27: 217-232.
Boix-Trelis N, Vale-Martínez A, Guillazo-Blanch G, Costa-Miserachs D, Martí-Nicolovius M (2006). Effects of nucleus basalis magnocellularis stimulation on a socially transmitted food preference and c-Fos expression. Learn Mem, 13: 783-793.
287 Boix-Trelis N, Vale-Martínez A, Guillazo-Blanch G, Martí-Nicolovius M (2007). Muscarinic cholinergic receptor blockade in the rat prelimbic cortex impairs the social transmission of food preference. Neurobiol Learn Mem, 87: 659-668.
Bonthuis PJ, Cox KH, Searcy BT, Kumar P, Tobet S, Rissman EF (2010). Of mice and rats: key species variations in the sexual differentiation of brain and behavior. Front Neuroendocrinol, 31: 341-358.
Bora SH, Liu Z, Kecojevic A, Merchenthaler I, Koliatsos VE (2005). Direct, complex effects of estrogens on basal forebrain cholinergic neurons. Exp Neurol, 194: 506-522.
Bosch OJ, Neumann ID (2010). Vasopressin released within the central amygdala promotes maternal aggression. Eur J Neurosci, 31: 883-891.
Bosch OJ, Neumann ID (2008). Brain vasopressin is an important regulator of maternal behavior independent of dams' trait anxiety. Proc Natl Acad Sci U S A, 105: 17139- 17144.
Bosch OJ, Pförtsch J, Biederbeck DI, Landgraf R, Neumann ID (2010). Maternal behaviour is associated with vasopressin release in the medial preoptic area and bed nucleus of the stria terminalis in the rat. J Neuroendocrinol, 22: 420-429.
Boulware MI, Weick JP, Becklund BR, Kuo SP, Groth RD, Mermelstein PG (2005). Estradiol activates group I and II metabotropic glutamate receptor signaling, leading to opposing influences on cAMP Response Element-Binding Protein. J Neurosci, 25: 5066- 5078.
Bowden NJ, Brain PF (1978). Blockade of testosterone-maintained intermale fighting in albino laboratory mice by an aromatization inhibitor. Physiol Behav, 20: 543-546.
Bowman RE, Ferguson D, Luine VN (2002). Effects of chronic restraint stress and estradiol on open field activity, spatial memory, and monoaminergic neurotransmitters in ovariectomized rats. Neuroscience, 113: 401-410.
Brain PF, Benton D, Childs G, Parmigiani S (1981). The effect of the type of opponent in tests of murine aggression. Behav Process, 6: 319-327.
Brain PF, Bowden NJ (1979). Sex steroid control of intermale fighting in mice. Curr Devel Psychopharmacol, 5: 403-465.
Brain PF, Haug M (1992). Hormonal and neurochemical correlates of various forms of animal “aggression”. Psychoneuroendocrinology, 17: 537-551.
Brain PF, Poole AE (1976). The role of endocrines in isolation-induced intermale fighting in albino laboratory mice: II. Sex steroid influences in aggressive mice. Aggress Behav, 2: 55-76.
288 Branchi I, D'Andrea I, Fiore M, Di Fausto V, Aloe L, Alleva E (2006). Early Social Enrichment Shapes Social Behavior and Nerve Growth Factor and Brain-Derived Neurotrophic Factor Levels in the Adult Mouse Brain. Biol Psychiatry, 60: 690-696.
Bunsey M, Eichenbaum H (1995). Selective damage to the hippocampal region blocks long- term retention of a natural and nonspatial stimulus-stimulus association. Hippocampus, 5: 546-556.
Bunsey M, Strupp BJ (1990). A vasopressin metabolite produces qualitatively different effects on memory retrieval depending on the accessibility of the memory. Behav Neur Biol, 53: 346-355.
Cacioppo JT, Decety J (2011). Social neuroscience: challenges and opportunities in the study of complex behaviour. Ann N Y Acad Sci, 1224: 162-173.
Caldwell HK, Albers HE (2004). Effect of photoperiod on vasopressin-induced aggression in Syrian hamsters. Horm Behav, 46: 444-449.
Caldwell HK, Lee HJ, Macbeth AH, Young WSI (2008). Vasopressin: behavioral roles of an original neuropeptide. Prog Neurobiol, 84: 1-24.
Carballo-Márquez A, Vale-Martínez A, Guillazo-Blanch G, Martí-Nicolovius M (2009b). Muscarinic receptor blockade in ventral hippocampus and prelimbic cortex impairs memory for socially transmitted food preference. Hippocampus, 19: 446-455.
Carballo-Márquez A, Vale-Martínez A, Guillazo-Blanch G, Martí-Nicolovius M (2009a). Muscarinic transmission in the basolateral amygdala is necessary for the acquisition of socially transmitted food preferences in rats. Neurobiol Learn Mem, 91: 98-101.
Cardinal RN, Parkinson JA, Hall JM, Everitt BJ (2002). Emotion and motivation: the role of the amygdala, ventral striatum, and prefrontal cortex. Neurosci Biobehav Rev, 26: 321- 352.
CDC, Autism and Developmental Disabilities Monitoring Network Surveillance Y2PI (2009). Prevalence of autism spectrum disorders---Autism and Developmental Disabilities Monitoring Network, United States, 2006. MMWR 2009;58(No. SS-10). MMWR Surveill Summ, 58: 1-20.
Champagne FA, Diorio J, Sharma S, Meaney MJ (2001). Naturally occurring variations in maternal behavior in the rat are associated with differences in estrogen-inducible central oxytocin receptors. Proc Natl Acad Sci U S A, 98: 12736-12741.
Champagne FA, Francis DD, Mar A, Meaney MJ (2003). Variations in maternal care in the rat as a mediating influences for the effects of environment on development. Physiol Behav, 79: 359-371.
Chen D, Wu CF, Shi B, Xu YM (2002). Tamoxifen and toremifene cause impairment of learning and memory function in mice. Pharmacol Biochem Behav, 71: 269-276.
289 Chesler EJ, Juraska JM (2000). Acute administration of estrogen and progesterone impairs the acquisition of the spatial Morris water maze in ovariectomized rats. Horm Behav, 38: 234-242.
Choleris E, Clipperton AE, Phan A, Kavaliers M (2008b). Estrogen receptor β agonists in neurobehavioral investigations. Current Opinion in Investigational Drugs, 733-773.
Choleris E, Clipperton AE, Phan A, Kavaliers M (2008a). Estrogen receptor β agonists in neurobehavioral investigations. Curr Opin Investig Drugs, 9: 733-773.
Choleris E, Clipperton-Allen AE, Gray DG, Diaz-Gonzalez S, Welsman RG (2011b). Differential effects of dopamine receptor D1-type and D2-type antagonists and phase of the estrous cycle on social learning of food preferences, feeding and social interactions in mice. Neuropsychopharmacology,
Choleris E, Clipperton-Allen AE, Gray DG, Diaz-Gonzalez S, Welsman RG (2011a). Differential effects of dopamine receptor D1-type and D2-type antagonists and phase of the estrous cycle on social learning of food preferences, feeding and social interactions in mice. Neuropsychopharmacology, 36: 1689-1702.
Choleris E, Clipperton-Allen AE, Phan A, Kavaliers M (2009). Neuroendocrinology of social information processing in rats and mice. Front Neuroendocrinol, 30: 442-459.
Choleris E, Clipperton-Allen AE, Phan A, Kavaliers M (2011c). Estrogenic involvement in social learning, social recognition and pathogen avoidance. Front Neuroendocrinol,
Choleris E, Guo C, Liu H, Mainardi M, Valsecchi P (1997). The effect of demonstrator age and number on duration of socially-induced food preferences in house mouse (Mus domesticus). Behav Process, 41: 69-77.
Choleris E, Gustafsson J-Å, Korach KS, Muglia LJ, Pfaff DW, Ogawa S (2003). An estrogen-dependent four-gene micronet regulating social recognition: a study with oxytocin and estrogen receptor-alpha and -beta knockout mice. Proc Natl Acad Sci U S A, 100: 6192-6197.
Choleris E, Kavaliers M (1999). Social learning in animals: sex differences and neurobiological analysis. Pharmacol Biochem Behav, 64: 767-776.
Choleris E, Kavaliers M, Pfaff DW (2004). Functional genomics of social recognition. J Neuroendocrinol, 16: 383-389.
Choleris E, Little SR, Mong JA, Puram SV, Langer R, Pfaff DW (2007). Microparticle-based delivery of oxytocin receptor antisense DNA in the medial amygdala blocks social recognition in female mice. Proc Natl Acad Sci U S A, 104: 4670-4675.
Choleris E, Ogawa S, Kavaliers M, Gustafsson J-Å, Korach KS, Muglia LJ et al. (2006). Involvement of estrogen receptor α, β and oxytocin in social discrimination: A detailed behavioral analysis with knockout female mice. Genes Brain Behav, 5: 528-539. 290 Choleris E, Thomas AW, Kavaliers M, Prato FS (2001). A detailed ethological analysis of the mouse open field test: effects of diazepam, chlordiazepoxide and an extremely low frequency pulsed magnetic field. Neurosci Biobehav Rev, 25: 235-260.
Choleris E, Valsecchi P, Wang Y, Ferrari P, Kavaliers M, Mainardi M (1998a). Social learning of a food preference in male and female Mongolian gerbils is facilitated by the anxiolytic, chlordiazepoxide. Pharmacology, Biochemistry, and Behavior, 60: 575-584.
Choleris E, Valsecchi P, Wang Y, Ferrari PF, Kavaliers M, Mainardi M (1998b). Social learning of a food preference in male and female Mongolian gerbils is facilitated by the anxiolytic, chlordiazepoxide. Pharmacol Biochem Behav, 60: 575-584.
Chu SH, Goldspink P, Kowalski J, Beck J, Schwertz DW (2006). Effects of estrogen on calcium-handling proteins, beta-adrenergic receptors, and function in rat heart. Life Sci, 79: 1257-1267.
Clark CR, Nowell NW (1979). The effect of the antiestrogen CI-628 on androgen-induced aggressive behavior in castrated male mice. Horm Behav, 12: 205-210.
Clark RE, Broadbent NJ, Zola SM, Squire LR (2002). Anterograde amnesia and temporally graded retrograde amnesia for a nonspatial memory task after lesions of hippocampus and subiculum. J Neurosci, 22: 4663-4669.
Clipperton AE, Cragg CL, Wood AJ, Langmo A-M, Choleris E (2006). Influence of gonadal hormones on social behavior in male and female mice. Soc Behav Neuroendocrinol Abstr, 10: 22-
Clipperton AE, Spinato JM, Chernets C, Pfaff DW, Choleris E (2008). Differential effects of estrogen receptor alpha and beta specific agonists on social learning of food preferences in female mice. Neuropsychopharmacology, 33: 2362-2375.
Clipperton-Allen AE, Almey A, Melichercik A, Allen CP, Choleris E (2011a). Effects of an estrogen receptor alpha agonist on agonistic behaviour in intact and gonadectomized male and female mice. Psychoneuroendocrinology, 36: 981-995.
Clipperton-Allen AE, Cragg CL, Wood AJ, Pfaff DW, Choleris E (2010). Agonistic behavior in males and females: effects of an estrogen receptor beta agonist in gonadectomized and gonadally intact mice. Psychoneuroendocrinology, 35: 1008-1022.
Clipperton-Allen AE, Flaxey I, Foucault JN, Hussey B, Rush ST, Tam C et al. (2011b). The involvement of estrogen receptors alpha and beta in the memory for a socially acquired food preference in female mice. in preparation:
Clipperton-Allen AE, Flaxey I, Webster HK, Nediu-Mihalache C, Choleris E (2011c). Acute effects of estradiol benzoate on the social transmission of food preferences in female mice. J Neuroendocrinol, in preparation:
291 Clipperton-Allen AE, Lee AW, Reyes A, Devidze N, Phan A, Pfaff DW et al. (2011d). Oxytocin, vasopressin and estrogen gene expression in relation to social recognition in female mice. Physiol Behav, 105: 915-924.
Clipperton-Allen AE, Mikloska KV, Roussel VR, Ying HL, Choleris E (2011e). Effects of chronic estrogen and estrogen receptor agonists on the behaviour of ovariectomized female mice in the social transmission of food preferences paradigm. in preparation, in preparation:
Colafranceschi M, Capuani G, Miccheli A, Campo S, Valerio M, Tomassini A et al. (2007). Dissecting drug and vehicle metabolic effects in rats by a metabonomic approach. J Biochem Biophys Methods, 70: 355-361.
Colgan PW (1983). Comparative Social Recognition.
Cologer-Clifford A, Simon NG, Lu S-F, Smoluk SA (1997). Serotonin agonist-induced decreases in intermale aggression are dependent on brain region and receptor subtype. Pharmacol Biochem Behav, 58: 425-430.
Cologer-Clifford A, Simon NG, Richter ML, Smoluk SA, Lu S-F (1999). Androgens and estrogens modulate 5-HT1A and 5-HT1B agonist effects on aggression. Physiol Behav, 65: 823-828.
Compaan JC, Buijs RM, Pool CW, De Ruiter AJ, Koolhaas JM (1993). Differential lateral septal vasopressin innervation in aggressive and nonaggressive male mice. Brain Res Bull, 30: 1-6.
Countryman RA, Gold PW (2007). Rapid forgetting of social transmission of food preferences in aged rats: relationship to hippocampal CREB activation. Learn Mem, 14: 350-358.
Countryman RA, Kaban NL, Colombo PJ (2005a). Hippocampal c-fos is necessary for long- term memory of a socially transmitted food preference. Neurobiol Learn Mem, 84: 175- 183.
Countryman RA, Orlowski JD, Brightwell JJ, Oskowitz AZ, Colombo PJ (2005b). CREB phosphorylation and c-Fos expression in the hippocampus of rats during acquisition and recall of a socially transmitted food preference. Hippocampus, 15: 56-67.
Couse JF, Korach KS (1999). Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev, 20: 358-417.
Couse JF, Lindzey J, Grandien K, Gustafsson J-Å, Korach KS (1997). Tissue distribution and quantitative analysis of estrogen receptor-α (ERα) and estrogen receptor-β (ERβ) messenger ribonucleic acid in the wild-type and ERá-knockout mouse. Endocrinology, 138: 4613-4621.
292 Cowley SM, Hoare S, Mosselman S, Parker MG (1997). Estrogen receptors α and β form heterodimers on DNA. Journal of Biological Chemistry, 272: 19858-19862.
Cragg CL, Fissore E, Pfaff DW, Choleris E (2011). Effects of α and β estrogen receptor agonists on social recognition in intact and gonadectomized male and female mice.
Craig W (1921). Why do animals fight? Int J Ethics, 31: 264-278.
Creel S (2001). Social dominance and stress hormones. Trends Ecol Evol, 16: 491-497.
Curtis JT, Liu Y, Aragona BJ, Wang Z (2007). Neural regulation of social behavior in rodents. In: Wolff JO, Sherman PW (eds.), Rodent Societies: An Ecological & Evolutionary Perspective, pp. 185-194.
Curtis JT, Wang Z (2005). Glucocorticoid receptor involvement in pair bonding in female prairie voles: the effects of acute blockade and interactions with central dopamine reward systems. Neuroscience, 134: 369-376.
Cushing BS, Perry A, Musatov S, Ogawa S, Papademetriou E (2008). Estrogen receptors in the medial amygdala inhibit the expression of male prosocial behavior. J Neurosci, 28: 10399-10403.
Cushing BS, Razzoli M, Murphy AZ, Epperson PM, Le W-W, Hoffman GE (2004). Intraspecific variation in estrogen receptor alpha and the expression of male sociosexual behavior in two populations of prairie voles. Brain Res, 1016: 247-254.
Cushing BS, Wynne-Edwards KE (2006). Estrogen receptor-α distribution in male rodents is associated with social organization. J Comp Neurol, 494: 595-605.
Daniel JM (2006). Effects of oestrogen on cognition: what have we learned from basic research? J Neuroendocrinol, 18: 787-795.
Daniel JM, Fader AJ, Spencer AL, Dohanich GP (1997). Estrogen enhances performance of female rats during acquisition of a radial arm maze. Horm Behav, 32: 217-225.
Daniel JM, Hulst JL, Berbling JL (2006a). Estradiol replacement enhances working memory in middle-aged rats when initiated immediately after ovariectomy but not after a long- term period of ovarian hormone deprivation. Endocrinology, 147: 607-614.
Daniel JM, Lee CD (2004). Estrogen replacement in ovariectomized rats affects strategy selection in the Morris water maze. Neurobiol Learn Mem, 82: 142-149.
Daniel JM, Sulzer JK, Hulst JL (2006b). Estrogen increases the sensitivity of ovariectomized rats to the disruptive effects produced by antagonism of D2 but not D1 dopamine receptors during performance of a response learning task. Horm Behav, 49: 38-44.
Dantzer R, Bluthé R-M, Koob GF, Le Moal M (1987). Modulation of social memory in male rats by neurohypophyseal peptides. Psychopharmacology, 91: 363-368.
293 Dantzer R, Koob GF, Bluthé R-M, Le Moal M (1988). Septal vasopressin modulates social memory in male rats. Brain Res, 457: 143-147.
Davis DM, Jacobson TK, Aliakbari S, Mizumori SJY (2005). Differential effects of estrogen on hippocampal- and striatal-dependent learning. Neurobiol Learn Mem, 84: 132-137. de Almeida RMM, Ferrari PF, Parmigiani S, Miczek KA (2005). Escalated aggressive behavior: dopamine, serotonin and GABA. Eur J Pharmacol, 526: 51-64. de Almeida RMM, Miczek KA (2002). Aggression calculated by social instigation or by discontinuation of reinforcement (“frustration”) in mice: inhibition by anpirtoline–a 5HT1B receptor agonist. Neuropsychopharmacology, 27: 171-181. de Almeida RMM, Nikulina EM, Faccidomo SP, Fish EW (2001). Zolmitriptan-a 5-HT1B/D agonist, alcohol, and aggression in mice. Psychopharmacology, 157: 131-141.
De Boer SF, Koolhaas JM (2005). 5-HT1A and 5-HT1B receptor agonists and aggression: a pharmacological challenge of the serotonin deficiency hypothesis. Eur J Pharmacol, 526: 125-139.
De Boer SF, Lesourd M, Mocaer E, Koolhaas JM (1999). Selective antiaggressive effects of alnespirone in resident-intruder test are mediated via (5-hydroxytryptamine) 1A receptors: a comparative pharmacological study with 8-hydroxy-2-dipropylaminotetralin, ipsapirone, buspirone, eltoprazine, and WAY-100635. J Pharmacol Exp Ther, 288: 1125- 1133. de Bono M (2003). Molecular approaches to aggregation behavior and social attachment. J Neurobiol, 54: 78-92. de Catanzaro D (1987). Differential sexual activity of isolated and grouped male mice despite testosterone administration. Behav Neur Biol, 48: 213-221.
DeVries AC, Young SW, Nelson RJ (1997). Reduced aggressive behavior in mice with targeted disruption of the oxytocin gene. J Neuroendocrinol, 9: 363-368.
Dluzen DE, Muraoka S, Engelmann M, Ebner K, Landgraf R (2000). Oxytocin induces preservation of social recognition in male rats by activating alpha-adrenoceptors of the olfactory bulb. Eur J Neurosci, 12: 760-766.
Dluzen DE, Muraoka S, Engelmann M, Landgraf R (1998a). The effects of infusion of arginine vasopressin, oxytocin, or their antagonists into the olfactory bulb upon social recognition responses in male rats. Peptides, 19: 999-1005.
Dluzen DE, Muraoka S, Landgraf R (1998b). Olfactory bulb norepinephrine depletion abolishes vasopressin and oxytocin preservation of social recognition responses in rats. Neurosci Lett, 254: 161-164.
294 Donner N, Handa RJ (2009a). Estrogen receptor beta regulates the expression of tryptophan- hydroxylase 2 mRNA within serotonergic neurons of the rat dorsal raphe nuclei. Neuroscience, 163: 705-718.
Donner N, Handa RJ (2009b). Estrogen receptor beta regulates the expression of tryptophan- hydroxylase 2 mRNA within serotonergic neurons of the rat dorsal raphe nuclei. Neuroscience, 163: 705-718.
Dore R, Phan A, Clipperton-Allen AE, Kavaliers M, Choleris E (2011). The involvement of oxytocin and vasopressin in social recognition and social learning: an interplay with sex hormones. In: Choleris E, Pfaff DW, Kavaliers M (eds.), Oxytocin, Vasopressin and Related Peptides in the Regulation of Behavior, pp.
Dubois-Dauphin M, Barberis C, de Bilbao F (1996). Vasopressin receptors in the mouse (Mus musculus) brain: sex-related expression in the medial preoptic area and hypothalamus. Brain Res, 743: 32-39.
Dugatkin LA, Godin J-G (1992). Reversal of female mate choice by copying in the guppy (Poecillia reticulata). Proc R Soc B, 249: 179-184.
Dulac C, Torello T (2003). Molecular detection of pheromone signals in mammals: from genes to behaviour. Nat Rev Neurosci, 4: 551-562.
Dunbar RIM, Schultz S (2007). Evolution in the social brain. Science, 317: 1344-1347.
Ebensperger LA (2001). A review of the evolutionary causes of rodent group-living. Acta Theriol, 46: 115-144.
Edwards DA (1970). Post-neonatal androgenization and adult aggressive behavior in female mice. Physiol Behav, 5: 465-467.
Edwards DA (1969). Early androgen stimulation and aggressive behavior in male and female mice. Physiol Behav, 4: 333-338.
Edwards DA, Burge KG (1971). Estrogenic arousal of aggressive behavior and masculine sexual behavior in male and female mice. Horm Behav, 2: 239-245.
Engelmann M, Ebner K, Wotjak CT, Landgraf R (1998). Endogenous oxytocin is involved in short-term olfactory memory in female rats. Behav Brain Res, 90: 89-94.
Engelmann M, Landgraf R (1994). Microdialysis administration of vasopressin into the septum improves social recognition in Brattleboro rats. Physiol Behav, 55: 145-149.
Engelmann M, Wotjak CT, Ebner K, Landgraf R (2000). Behavioural impact of intraseptally released vasopressin and oxytocin in rats. Exp Physiol, 85: 125S-130S.
295 Engelmann M, Wotjak CT, Neumann ID, Ludwig M, Landgraf R (1996). Behavioral consequences of intracerebral vasopressin and oxytocin: focus on learning and memory. Neurosci Biobehav Rev, 20: 341-358.
Enquist M, Leimar O (1990). The evolution of fatal fighting. Anim Behav, 39: 1-9.
Eroschenko VP, Swartz WJ, Ford LC (1997). Decreased superovulation in adult mice following neonatal exposures to technical methoxychlor. Reprod Toxicol, 11: 807-814.
Ervin K, Clipperton-Allen AE, Choleris E (2011). Something about delays and making harder tasks.
Esmaeili B, Basseda Z, Gholizadeh S, Javadi Paydar M, Dehpour AR (2009). Tamoxifen disrupts consolidation and retrieval of morphine-associated contextual memory in male mice: interaction with estradiol. Psychopharmacology, 204: 191-201.
Everts HGJ, Koolhaas JM (1999). Differential modulation of lateral septal vasopressin receptor blockade in spatial learning, social recognition, and anxiety-related behaviors in rats. Behav Brain Res, 99: 7-16.
Faul F, Erdfelder E, Buchner A, Lang A-G (2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behav Res Methods, 41: 1149-1160.
Faul F, Erdfelder E, Lang A-G, Buchner A (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods, 39: 175-191.
Ferguson JN, Aldag JM, Insel TR, Young LJ (2001). Oxytocin in the medial amygdala is essential for social recognition in the mouse. J Neurosci, 21: 8278-8285.
Ferguson JN, Young LJ, Hearn EF, Matzuk MM, Insel TR, Winslow JT (2000). Social amnesia in mice lacking the oxytocin gene. Nat Genet, 25: 284-288.
Ferguson JN, Young LJ, Insel TR (2002). The neuroendocrine basis of social recognition. Front Neuroendocrinol, 23: 200-224.
Ferrari PF, Palanza P, Parmigiani S, Rodgers RJ (1998). Interindividual variability in Swiss male mice: relationship between social factors, aggression and anxiety. Physiol Behav, 63: 821-827.
Ferris CF (2005). Vasopressin/oxytocin and aggression. Novartis Found Symp, 268: 190- 198.
Ferris CF, Delville Y (1994). Vasopressin and serotonin interactions inthe control of agonistic behavior. Psychoneuroendocrinology, 19: 593-601.
Ferris CF, Foote KB, Meltser HM, Plenby MG, Smith KL, Insel TR (1992). Oxytocin in the amygdala facilitates maternal aggression. Ann N Y Acad Sci, 652: 456-457.
296 Ferris CF, Lu S-F, Messenger T, Guillon CD, Heindel N, Miller MM et al. (2006). Orally active vasopressin V1a receptor antagonist, SRX251, selectively blocks aggressive behavior. Pharmacol Biochem Behav, 83: 169-174.
Ferris CF, Melloni RH, Koppel G, Perry KW, Fuller RW, Delville Y (1997). Vasopressin/serotonin interactions in the anterior hypothalamus control aggressive behavior in golden hamsters. J Neurosci, 17: 4331-4340.
Ferris CF, Potegal M (1988). Vasopressin receptor blockade in the anterior hypothalamus suppresses aggression in hamsters. Physiol Behav, 44: 235-239.
Filardo EJ, Prossnitz ER (2009). Estrogen (G protein coupled) receptor, introductory chapter. IUPHAR-DB,
Finney HC, Erpino MJ (1976). Synergistic effect of estradiol benzoate and dihydrotestosterone on aggression in mice. Horm Behav, 7: 391-400.
Fish EW, Faccidomo SP, Miczek KA (1999). Aggression heightened by alcohol or social instigation in mice: reduction by the 5-HT1B receptor agonist CP-94,253. Psychopharmacology, 146: 391-399.
Fleming AS, Kuchera C, Lee AW, Winocur G (1994). Olfactory-based social-learning varies as a function of parity in female rats. Psychobiology, 22: 37-43.
Fombonne E (2003a). Modern views of autism. Can J Psychiatry, 48: 503-505.
Fombonne E (2003b). Epidemiology of autism and other pervasive developmental disorders: an update. J Autism Dev Disord, 33: 365-381.
Frank E, Landgraf R (2008). The vasopressin system - from antidiuresis to psychopathology. Eur J Pharmacol, 583: 226-242.
Frasor J, Barnett DH, Danes JM, Hess R, Parlow AF, Katzenellenbogen BS (2003). Response-specific and ligand dose-dependent modulation of estrogen receptor (ER) α activity by ERβ in the uterus. Endocrinology, 144: 3159-3166.
Frick KM (2009). Estrogens and age-related memory decline in rodents: what have we learned and where do we go from here? Horm Behav, 55: 2-23.
Frick KM, Fernandez SM, Bennett JC, Prange-Kiel J, MacLusky NJ, Leranth C (2004). Behavioral training interferes with the ability of gonadal hormones to increase CA1 spine synapse density in ovariectomized female rats. Eur J Neurosci, 19: 3026-3032.
Frye CA (1995). Estrus-associated decrements in a water maze task are limited to acquisition. Physiol Behav, 57: 5-14.
Fugger HN, Cunningham SG, Rissman EF, Foster TC (1998). Sex differences in the activational effect of ERα on spatial learning. Horm Behav, 34: 163-170.
297 Gabor CS, Phan A, Clipperton-Allen AE, Kavaliers M, Choleris E (2011). Interplay of oxytocin, vasopressin and sex hormones in the regulation of social recognition. Behav Neurosci,
Galea LAM, Wide JK, Paine TA, Holmes MM, Ormerod BK, Floresco SB (2001). High levels of estradiol disrupt conditioned place preference learning, stimulus response learning and reference memory but have limited effects on working memory. Behav Brain Res, 126: 115-126.
Galef BGJ (1986). Social interaction modifies learned aversions, sodium appetite, and both palatability and handling-time induced dietary preferences in rats (Rattus norvegicus). J Comp Psychol, 100: 432-439.
Galef BGJ (1996). Social enhancement of food preferences in Norway rats: a brief review. In: Heyes CM, Galef BGJ (eds.), Social Learning in Animals: The Roots of Culture., pp. 49-64.
Galef BGJ, Clark MM (1971). Parent-offspring interactions determine time and place of first ingestion of solid food in wild rat pups. Psychon Sci, 25: 15-16.
Galef BGJ, Kennett DJ, Wigmore SW (1984). Transfer of information concerning distant foods in rats: a robust phenomenon. Anim Learn Behav, 12: 292-296.
Galef BGJ, Laland KN (2005). Social learning in animals: Empirical studies and theoretical models. Biosci, 55: 489-499.
Galef BGJ, Marczinski CA, Murray KA, Whiskin EE (2001). Studies of food stealing by young Norway rats. J Comp Psychol, 115: 16-21.
Galef BGJ, Mason JR, Preti G, Bean NJ (1988). Carbon disulfide: a semiochemical mediating socially-induced diet choice in rats. Physiol Behav, 42: 119-124.
Galef BGJ, Whiskin EE (2003). Socially transmitted food preferences can be used to study long-term memory in rats. Learn Behav, 31: 160-164.
Galef BGJ, Whiskin EE, Bielavska E (1997). Interaction with demonstrator rats changes observer rats' affective responses to flavors. J Comp Psychol, 111: 393-398.
Galef BGJ, Wigmore SW (1983). Transfer of information concerning distant foods: a laboratory investigation of the 'information-centre' hypothesis. Anim Behav, 31: 748-758.
Garey J, Morgan MA, Frohlich J, McEwen BS, Pfaff DW (2001). Effects of the phytoestrogen coumestrol on locomotor and fear-related behaviors in female mice. Horm Behav, 40: 65-76.
Gaskin S, White NM (2006). Cooperation and competition between the dorsal hippocampus and lateral amygdala in spatial discrimination learning. Hippocampus, 16: 577-585.
298 Gass P, Wolfer DP, Balschun D, Rudolph D, Frey U, Lipp H-P et al. (1998). Deficits in memory tasks of mice with CREB mutations depend on gene dosage. Learn Mem, 5: 274-288.
Gerrish CJ, Alberts JR (1995). Differential influence of adult and juvenile conspecifics on feeding in weanling rats (Rattus norvegicus): a size-related explanation. J Comp Psychol, 109: 61-67.
Gibbs RB (2000). Long-term treatment with estrogen and progesterone enhances acquisition of a spatial memory task by ovariectomized aged rats. Neurobiol Aging, 21: 107-116.
Gillberg C (2002). A guide to Asperger Syndrome.
Gimpl G, Fahrenholz F (2001). The oxytocin receptor system: structure, function and regulation. Physiol Rev, 81: 629-683.
Ginsburg B, Allee WC (1942). Some effects of conditioning on social dominance and subordination in inbred strains of mice. Physiol Zool, 15: 485-506.
Giovenardi M, Padoin MJ, Cadore LP, Lucion AB (1998). Hypothalamic paraventricular nucleus modulates maternal aggression in rats: effects of ibotenic acid lesion and oxytocin antisense. Physiol Behav, 63: 351-359.
Grant EC, Mackintosh JH (1963). A comparison of the social postures of some common laboratory rodents. Behaviour, 21: 246-259.
Gray LEJ, Whitsett JM, Ziensenis JS (1978). Hormonal regulation of aggression toward juveniles in female house mice. Horm Behav, 11: 310-322.
Greendale GA, Huang M-H, Wight RG, Seeman T, Luetters C, Avis NE et al. (2009). Effects of the menopause transition and hormone use on cognitive performance in midlife women 1685. Neurology, 72: 1850-1857.
Gresack JE, Kerr KM, Frick KM (2007). Life-long environmental enrichment differentially affects the mnemonic response to estrogen in young, middle-aged, and aged female mice. Neurobiol Learn Mem, 88: 393-408.
Guillot PV, Chapouthier G (1996). Intermale aggression and dark/light preference in ten inbred mouse strains. Behav Brain Sci, 77: 211-213.
Gustafsson J-Å (2006b). ERβ scientific visions translate to clinical uses. Climacteric, 9: 156- 160.
Gustafsson J-Å (2006a). ERβ scientific visions translate to clinical uses. Climacteric, 9: 156- 160.
299 Guzowski JF, McGaugh JL (1997). Antisense oligodeoxynucleotide-mediated disruption of hippocampal cAMP response element binding protein levels impairs consolidation of memory for water maze training. Proc Natl Acad Sci U S A, 94: 2693-2698.
Hall JM, McDonnell DP (1999). The estrogen receptor β-isoform (ERβ) of the human estrogen receptor modulates ERα transcriptional activity and is a key regulator of the cellular response to estrogens and antiestrogens. Endocrinology, 140: 5566-5578.
Hammock EAD, Young LJ (2007). Neuroendocrinology, neurochemistry, and molecular neurobiology of affiliative behavior. In: Lathja A, Blaustein JD (eds.), Handbook of Neurochemistry and Molecular Neurobiology, pp. 247-284.
Hammond R, Mauk R, Ninaci D, Nelson D, Gibbs RB (2009). Chronic treatment with estrogen receptor agonists restores acquisition of a spatial learning task in young ovariectomized rats. Horm Behav, 56: 309-314.
Handa RJ, Ogawa S, Wang JM, Herbison AE (2011). Roles for estrogen receptor beta in adult brain function. J Neuroendocrinol, Accepted Article:
Haney M, DeBold JF, Miczek KA (1989). Maternal aggression in mice and rats towards male and female conspecifics. Aggress Behav, 15: 443-453.
Haney M, Miczek KA (1989). Morphine effects on maternal aggression, pup care and analgesia in mice. Psychopharmacology, 98: 68-74.
Harris HA (2007). Estrogen receptor-β: recent lessons from in vivo studies. Mol Endocrinol, 21: 1-13.
Harris HA, Katzenellenbogen JA, Katzenellenbogen BS (2002a). Characterization of the biological roles of the estrogen receptors, ERα and ERβ, in estrogen target tissues in vivo through the use of an ERα-selective ligand. Endocrinology, 143: 4172-4177.
Harris HA, Katzenellenbogen JA, Katzenellenbogen BS (2002b). Characterization of the biological roles of the estrogen receptors, ERα and ERβ, in estrogen target tissues in vivo through the use of an ERα-selective ligand. Endocrinology, 143: 4172-4177.
Hebert AE, Dash PK (2004). Nonredundant roles for hippocampal and entorhinal cortical plasticity in spatial memory storage. Pharmacol Biochem Behav, 79: 143-153.
Heikkinen T, Puoliväli J, Liu L, Rissanen A, Tanila H (2002). Effects of ovariectomy and estrogen treatment on learning and hippocampal neurotransmitters in mice. Horm Behav, 41: 22-32.
Heikkinen T, Puoliväli J, Tanila H (2004). Effects of long-term ovariectomy and estrogen treatment on maze learning in aged mice. Exp Gerontol, 39: 1277-1283.
300 Hertrampf T, Seibel J, Laudenbach U, Fritzemeier KH, Diel P (2008). Analysis of the effects of oestrogen receptor alpha (ERalpha)- and ERbeta-selective ligands given in combination to ovariectomized rats. Br J Pharmacol, 153: 1432-1437.
Hill RA, Simpson ER, Boon WC (2008). Evidence for the existence of an estrogen- responsive sexually dimorphic group of cells in the medial preoptic area of the 129SvEv mouse strain. Int J Impot Res, 20: 315-323.
Holmes A, Murphy DL, Crawley JN (2002a). Reduced aggression in mice lacking the serotonin transporter. Psychopharmacology, 161: 160-167.
Holmes MM, Wide JK, Galea LAM (2002b). Low levels of estradiol facilitate, whereas high levels of estradiol impair, working memory performance on the radial arm maze. Behav Neurosci, 116: 928-934.
Howard HE (1920). Territory in Bird Life.
Hoyk Z, Varga C, Párducz A (2006). Estrogen-induced region specific decrease in the density of 5-bromo-2-deoxyuridine-labelled cells in the olfactory bulb of adult female rats. Neuroscience, 141: 1919-1924.
Huang C-H, Chiang Y-W, Liang K-C, Thompson RF, Liu IY (2010). Extra-cellular signal- regulated kinase 1/2 (ERK1/2) activated in the hippocampal CA1 neurons is critical for retrieval of auditory trace fear memory. Brain Res, 1326: 143-151.
Hughes ZA, Liu F, Platt BJ, Dwyer JM, Pulicicchio CM, Zhang G et al. (2008). WAY- 200070, a selective agonist of estrogen receptor beta as a potential novel anxiolytic/antidepressant agent. Neuropharmacology, 54: 1136-1142.
Hurst JL (1987). Behavioural variation in wild house mice Mus domesticus Rutty: a quantitative assessment of female social organization. Anim Behav, 35: 1846-1857.
Hyde JJ, Sawyer TF (1977). Estrous cycle fluctuations in aggressiveness of house mice. Horm Behav, 9: 290-295.
Igaz LM, Winograd M, Cammarota M, Izquierdo LA, Alonso M, Izquierdo I et al. (2006). Early activation of extracellular signal-regulated kinase signaling pathway in the hippocampus is required for short-term memory formation of a fear-motivated learning. Cell Molecul Neurobiol, 26: 989-1002.
Imamov O, Shim G-J, Warner M, Gustafsson J-Å (2005). Minireview: estrogen receptor beta in health and disease. Biol Reprod, 73: 866-871.
Imwalle DB, Scordalakes EM, Rissman EF (2002). Estrogen receptor α influences socially motivated behaviors. Horm Behav, 42: 484-491.
Insel TR, Fernald RD (2004). How the brain processes social information: Searching the social brain. Annu Rev Neurosci, 27: 697-722.
301 Insel TR, Gelhard R, Shapiro LE (1991). The comparative distribution of forebrain receptors for neurohypophyseal peptides in monogamous and polygamous mice. Neuroscience, 43: 623-630.
Jacome LF, Gautreaux C, Inagaki T, Mohan G, Alves SE, Lubbers L et al. (2010a). Estradiol and ERβ agonists enhance recognition memory, and DPN, an ERβ agonist, alters brain monoamines. Neurobiology of Learning and Memory, 94: 488-498.
Jacome LF, Gautreaux C, Inagaki T, Mohan G, Alves SE, Lubbers LS et al. (2010b). Estradiol and ERβ agonists enhance recognition memory, and DPN, an ERβ agonist, alters brain monoamines. Neurobiol Learn Mem, 94: 488-498.
Jansson L, Mattsson A, Mattsson R, Holmdahl R (1990). Estrogen induced suppression of collagen arthritis V: physiological level of estrogen in DBA/1 mice is therapeutic on established arthritis, suppresses anti-type II collagen T-cell dependent immunity and stimulates polyclonal B-cell activity. J Autoimmun, 3: 257-270.
Jasnow AM, Mong JA, Romeo RD, Pfaff DW (2007). Estrogenic regulation of gene and protein expression within the amygdala of female mice. Endocrine, 32: 271-279.
Johnson ML, Ho CC, Day AE, Walker QD, Francis R, Kuhn CM (2010). Oestrogen receptors enhance dopamine neurone survival in rat midbrain. J Neuroendocrinol, 22: 226-237.
Johnston RE, Peng A (2008). Memory for individuals: hamsters (Mesocricetus auratus) require contact to develop multicomponent representations (concepts) of others. J Comp Psychol, 122: 121-131.
Kang N, Baum MJ, Cherry JA (2009). A direct main olfactory projection to the 'vomeronasal' amygdala in female mice selectively responds to volatile pheromones from males. Eur J Neurosci, 29: 624-634.
Kantak KM, Hegstrand LR, Eichelman BS (1980). Dietary tryptophan modulation and aggressive behavior in mice. Pharmacol Biochem Behav, 12: 675-679.
Karakaya S, Kipp M, Beyer C (2007). Oestrogen regulates the expression and function of dopamine transporters in astrocytes of the nigrostriatal system. J Neuroendocrinol, 19: 682-690.
Kavaliers M, Choleris E, Ågmo A, Pfaff DW (2004). Olfactory-mediated parasite recognition and avoidance: linking genes to behavior. Horm Behav, 46: 272-283.
Kavaliers M, Colwell DD, Choleris E (2005a). Kinship, familiarity and social status modulate social learning about “micropredators” (biting flies) in deer mice. Behav Ecol Sociobiol, 58: 60-71.
302 Kavaliers M, Colwell DD, Choleris E (2005b). Kinship, familiarity and social status modulate social learning about “micropredators” (bbiting flies) in deer mice. Behav Ecol Sociobiol, 58: 60-71.
Kavaliers M, Colwell DD, Choleris E, Ågmo A, Muglia LJ, Ogawa S et al. (2001). Impaired discrimination of and aversion parasitized male odors by female oxytocin knockout mice. Genes Brain Behav, 2: 220-230.
Kendrick KM, Keverne EB, Chapman C, Baldwin BA (1988). Intracranial dialysis measurement of oxytocin, monoamine and uric acid release from the olfactory bulb and substantia nigra of sheep during parturition, suckling, separation from lambs and eating. Brain Res, 439: 1-10.
Kim DK, Tolliver TJ, Huang SJ, Martin BJ, Andrews AM, Wichems C et al. (2005). Altered serotonin synthesis, turnover and dynamic regulation in multiple brain regions of mice lacking the serotonin transporter. Neuropharmacology, 49: 798-810.
Kirkwood A (2007). Neuromodulation of cortical synaptic plasticity. In: Tseng K-Y, Atzori M (eds.), Monoaminergic Modulation of Cortical Excitability, pp. 209-216.
Koehler KF, Helguero LA, Haldosén L-A, Warner M, Gustafsson J-Å (2005). Reflections on the discovery and significance of estrogen receptor β. Endocr Rev, 26: 465-478.
Kogan JH, Frankland PW, Blendy JA, Coblentz J, Marowitz Z, Schütz G et al. (1996). Spaced training induces normal long-term memory in CREB mutant mice. Curr Biol, 7: 1-11.
Kogan JH, Frankland PW, Silva AJ (2000). Long-term memory underlying hippocampus- dependent social recognition in mice. Hippocampus, 10: 47-56.
Korol DL (2004). Role of estrogen in balancing contributions from multiple memory systems. Neurobiol Learn Mem, 82: 309-323.
Korol DL, Malin EL, Borden KA, Busby RA, Couper-Leo J (2004). Shifts in preferred learning strategy across the estrous cycle in female rats. Horm Behav, 45: 330-338.
Kronenberger JP, Médioni J (1985). Food neophobia in wild and laboratory mice (Mus musculus domesticus). Behav Process, 11: 53-59.
Kudryavtseva NN, Bondar NP, Avgustinovich DF (2002). Association between experience of aggression and anxiety in male mice. Behav Brain Res, 133: 83-93.
Kudwa AE, Michopoulos V, Gatewood JD, Rissman EF (2006). Roles of estrogen receptors α and β in differentiation of mouse sexual behavior. Neuroscience, 138: 921-928.
Kuiper GGJM, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson J-Å (1996). Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci U S A, 93: 5925- 5930.
303 Kumari M, Stafford M, Marmot M (2005). The menopausal transition was associated in a prospective study with decreased health functioning in women who report menopausal symptoms. J Clin Epidemiol, 58: 719-727.
Lai W-S, Johnston RE (2002). Individual recognition after fighting by golden hamsters: a new method. Physiol Behav, 76: 225-239.
Landgraf R, Frank E, Aldag JM, Neumann ID, Sharer CA, Ren X et al. (2003). Viral vector- mediated gene transfer of the vole V1a vasopressin receptor in the rat septum: improved social discrimination and active social behaviour. Eur J Neurosci, 18: 403-411.
Landgraf R, Gerstberger R, Montkowski A, Probst JC, Wotjak CT, Holsboer F et al. (1995). V1 vasopressin receptor antisense oligodeoxynucleotide into septum reduces vasopressin binding, social discrimination abilities, and anxiety-related behavior in rats. J Neurosci, 15: 4250-4258.
Landgraf R, Wigger A (2002). High vs low anxiety-related behavior rats: an animal model of extremes in trait anxiety. Behav Genet, 32: 301-314.
Lasley SM, Thurmond JB (1985). Interaction of dietary tryptophan and social isolation on territorial aggression, motor activity, and neurochemistry in mice. Psychopharmacology, 87: 313-321.
Latham N, Mason G (2004). From house mouse to mouse house: The behavioural biology of free-living Mus musculus and its implications in the laboratory. Appl Anim Behav Sci, 86: 261-289.
Le Moal M, Dantzer R, Michaud B, Koob GF (1987). Centrally injected arginine vasopressin (AVP) facilitates social memory in rats. Neurosci Lett, 77: 353-359.
Leckie B (1997). Girls, bullying behaviours and peer relationships: the double edged sword of exclusion and rejection.
Lee HJ, Caldwell HK, Macbeth AH, Tolu SG, Young WSI (2008). A conditional knockout mouse line of the oxytocin receptor. Endocrinology, 149: 3256-3263.
Leranth C, Shanabrough M, Horvath TL (1999). Estrogen receptor-α in the raphe serotonergic and supramammillary area calretinin-containing neurons of the female rat. Exp Brain Res, 128: 417-420.
Li F, Wang LP, Shen X, Tsien JZ (2010). Balanced dopamine is critical for pattern completion during associative memory recall. PLoS ONE, 5: e15401-
Lim MM, Bielsky IF, Young LJ (2005). Neuropeptides and the social brain: potential rodent models of autism. Int J Dev Neurosci, 23: 235-243.
304 Lim MM, Wang Z, Olazábal DE, Ren X, Terwilliger EF, Young LJ (2004). Enhanced partner preference in a promiscuous species by manipulating the expression of a single gene. Nature, 429: 754-757.
Lindberg MK, Weihua Z, Andersson N, Movérare S, Gao H, Vidal O et al. (2002). Estrogen receptor specificity for the effects of estrogen in ovariectomized mice. J Endocrinol, 174: 167-178.
Litvin Y, Murakami G, Pfaff DW (2011). Effects of chronic social defeat on behavioral and neural correlates of sociality: vasopressin, oxytocin and the vasopressinergic V1b receptor. Physiol Behav, 103: 393-403.
Liu B, Xie J (2004). Increased dopamine release in vivo by estradiol benzoate from the central amygdaloid nucleus of Parkinson's disease model rats. J Neurochem, 90: 654-658.
Liu F, Day M, Muniz LC, Bitran D, Arias R, Revilla-Sanchez R et al. (2008). Activation of estrogen receptor-beta regulates hippocampal synaptic plasticity and improves memory. Nat Neurosci, 11: 334-343.
Löfgren M, Johansson IM, Meyerson B, Turkmen S, Bäckström T (2008). Withdrawal effects from progesterone and estradiol relate to individual risk-taking and explorative behavior in female rats. Physiol Behav, 96: 91-97.
London SE, Clayton DF (2010). Genomic and neural analysis of the estradiol-synthetic pathway in the zebra finch. BMC Neurosci, 11: 46-
Lubin DA, Elliott JC, Black MC, Johns JM (2003). An oxytocin antagonist infused into the central nucleus of the amygdala increases maternal aggressive behavior. Behav Neurosci, 117: 195-201.
Luine VN, Richards ST, Wu VY, Beck KD (1998). Estradiol enhances learning and memory in a spatial memory task and effects levels of monoaminergic neurotransmitters. Horm Behav, 34: 149-162.
Luine VN, Rodriguez M (1994). Effects of estradiol on radial arm maze performance of young and aged rats. Behav Neur Biol, 62: 230-236.
Luttge WG (1972). Activation and inhibition of isolation induced inter-male fighting behavior in castrate male CD-1 mice treated with steroidal hormones. Horm Behav, 3: 71-81.
Luttge WG (1979). Anti-estrogen inhibition of testosterone-stimulated aggression in mice. Cellular and Molecular Life Sciences, 35: 273-274.
Luttge WG, Hall NR (1973). Androgen-induced agonistic behavior in castrate male Swiss- Webster mice: Comparison of four naturally occurring androgens. Behav Biol, 8: 725- 732.
305 Mahieu ST, Navoni J, Millen N, del Carmen Contini M, Gonzalez M, Elias MM (2004). Effects of aluminum on phosphate metabolism in rats: a possible interaction with vitamin D3 renal production. Arch Toxicol, 78: 609-616.
Mahmoodi G, Ahmadi S, Pourmotabbed A, Oryan S, Zarrindast MR (2010). Inhibitory avoidance memory deficit induced by scopolamine: interaction of cholinergic and glutamatergic systems in the ventral tegmental area. Neurobiol Learn Mem, 94: 83-90.
Malamas MS, Manas ES, McDevitt RE, Gunawan I, Xu ZB, Collini MD et al. (2004). Design and synthesis of aryl diphenolic azoles as potent and selective estrogen receptor- beta ligands. J Med Chem, 47: 5021-5040.
Marino MD, Bourdélat-Parks BN, Liles LC, Weinshenker D (2005). Genetic reduction of noradrenergic function alters social memory and reduces aggression in mice. Behav Brain Res, 161: 197-203.
Markham JA, Pych JC, Juraska JM (2002). Ovarian hormone replacement to aged ovariectomized female rats benefits acquisition of the Morris water maze. Horm Behav, 42: 284-293.
Markram K, Rinaldi T, La Mendola D, Sandi C, Markram H (2008). Abnormal fear conditioning and amygdala processing in an animal model of autism. Neuropsychopharmacology, 33: 901-912.
McCarthy MM (1990). Oxytocin inhibits infanticide in female house mice (Mus domesticus). Horm Behav, 24: 365-375.
McCarthy MM, McDonald CH, Brooks PJ, Goldman D (1996). Anxiolytic action of oxytocin is enhanced by estrogen in the mouse. Physiol Behav, 60: 1209-1215.
McDermott JL, Liu BJ, Dluzen DE (1994). Sex differences and effects of estrogen on dopamine and DOPAC release from the striatum of male and female CD-1 mice. Exp Neurol, 125: 306-311.
McEwen BS (2001). Estrogens effects on the brain: multiple sites and molecular mechanisms. J Appl Physiol, 91: 2785-2801.
McEwen BS, Akama K, Alves SE, Brake WG, Bulloch K, Lee S et al. (2001). Tracking the estrogen receptor in neurons: implications for estrogen-induced synapse formation. Proc Natl Acad Sci U S A, 98: 7093-7100.
McFadyen-Ketchum SA, Porter RH (1989). Transmission of food preferences in spiny mice (Acomys cahirinus) via nose-mouth interaction between mothers and weanlings. Behav Ecol Sociobiol, 59-62.
Merchenthaler I, Lane MV, Numan S, Dellovade TL (2004a). Distribution of estrogen receptor α and β in the mouse central nervous sytem: in vivo audioradiographic and immunocytochemical analyses. The Journal of Comparative Neurology, 473: 270-291.
306 Merchenthaler I, Lane MV, Numan S, Dellovade TL (2004b). Distribution of estrogen receptor α and β in the mouse central nervous sytem: in vivo audioradiographic and immunocytochemical analyses. J Comp Neurol, 473: 270-291.
Mewshaw RE, Edsall RJ, Yang CJ, Manas ES, Xu ZB, Henderson RA et al. (2005). ER β ligands. 2. Exploiting two binding orientations of the 2-phenylnaphthalene scaffold to achieve ERβ selectivity. J Med Chem, 48: 3953-3979.
Meyers MJ, Sun J, Carlson KE, Marriner GA, Katzenellenbogen BS, Katzenellenbogen JA (2001). Estrogen receptor-beta potency-selective ligands: structure-activity relationship studies of diarylpropionitriles and their acetylene and polar analogues. J Med Chem, 44: 4230-4251.
Mickley KR, Dluzen DE (2004). Dose-response effects of estrogen and tamoxifen upon methamphetamine-induced behavioral responses and neurotoxicity of the nigrostriatal dopaminergic system in female mice. Neuroendocrinology, 79: 305-316.
Miczek KA, Faccidomo SP, Fish EW, DeBold JF (2007). Neurochemistry and molecular neurobiology of aggressive behavior. In: Lathja A, Blaustein JD (eds.), Handbook of Neurochemistry and Molecular Neurobiology: Behavioral neurochemistry, neuroendocrinology and molecular neurobiology., pp. 285-336.
Miczek KA, Fish EW, DeBold JF, de Almeida RMM (2002). Social and neural determinants of aggressive behavior: pharmacotherapeutic targets at serotonin, dopamine and gamma- aminobutyric acid systems. Psychopharmacology, 163: 434-458.
Miczek KA, Hussain S, Faccidomo SP (1998). Alcohol-heightened aggression in mice: attenuation by 5-HT1A receptor agonists. Psychopharmacology, 139: 160-168.
Miczek KA, Maxson SC, Fish EW, Faccidomo SP (2001). Aggressive behavioral phenotypes in mice. Behav Brain Res, 125: 167-181.
Miele J, Rosellini RA, Svare B (1988). Estradiol benzoate can function as an unconditioned stimulus in a conditioned taste aversion paradigm. Horm Behav, 22: 116-130.
Milner TA, Thompson LI, Wang G, Kievits JA, Martin E, Zhou P et al. (2010). Distribution of estrogen receptor beta containing cells in the brains of bacterial artificial chromosome transgenic mice. Brain Res, 1351: 74-96.
Misslin R, Ropartz P (1981). Effects of lateral amygdala lesions on the responses to novelty in mice. Behav Process, 6: 329-336.
Mitra SW, Hoskin E, Yudkovitz J, Pear L, Wilkinson HA, Hayashi S et al. (2003a). Immunolocalization of estrogen receptor β in the mouse brain: comparison with estrogen receptor α. Endocrinology, 144: 2055-2067.
307 Mitra SW, Hoskin E, Yudkovitz J, Pear L, Wilkinson HA, Hayashi S et al. (2003b). Immunolocalization of estrogen receptor β in the mouse brain: comparison with estrogen receptor α. Endocrinology, 144: 2055-2067.
Mong JA, Devidze N, Frail DE, O'Connor LT, Samuel M, Choleris E et al. (2003a). Estradiol differentially regulates lipocalin-type prostaglandin D synthase transcript levels in the rodent brain: evidence from high-density oligonucleotide arrays and in situ hybridization. Proc Natl Acad Sci U S A, 100: 318-323.
Mong JA, Devidze N, Goodwillie A (2003b). Reduction of lipocalin-type prostaglandin D synthase in the preoptic area of female mice mimics estradiol effects on arousal and sex behavior. Proc Natl Acad Sci U S A, 100: 15206-15211.
Moore M, Calvert S (2000). Vocabulary acquisition for children with autism: teacher or computer instruction. J Autism Dev Disord, 30: 359-362.
Moran MD (2003). Arguments for rejecting the sequential Bonferroni in ecological studies. Oikos, 100: 403-405.
Morè L (2008). Intra-female aggression in the mouse (Mus musculus domesticus) is linked to the estrous cycle regularity but not to ovulation. Aggress Behav, 34: 46-50.
Morgan MA, Dellovade TL, Pfaff DW (2000). Effect of thyroid hormone administration on estrogen-induced sex behavior in female mice. Horm Behav, 37: 15-22.
Morgan MA, Pfaff DW (2002). Estrogen's effects on activity, anxiety, and fear in two mouse strains. Behav Brain Res, 132: 85-93.
Morgan TJH, Rendell LE, Ehn M, Hoppitt W, Laland KN (2011). The evolutionary basis of human social learning. Proc R Soc B,
Morissette M, Le Saux M, D'Astous M, Jourdain S, Al Sweidi S, Morin N et al. (2008). Contribution of estrogen receptors alpha and beta to the effects of estradiol in the brain. J Steroid Biochem Mol Biol, 108: 327-338.
Morrison JH, Brinton RD, Schmidt PJ, Gore AC (2006). Estrogen, menopause, and the aging brain: how basic neuroscience can inform hormone therapy in women. J Neurosci, 26: 10332-10348.
Mos J, Olivier B (1989). Quantitative and comparative analyses of pro-aggressive actions of benzodiazepines in maternal aggression of rats. Psychopharmacology, 97: 152-153.
Mugford RA (1974). Androgenic stimulation of aggression eliciting cues in adult opponent mice castrated at birth, weaning, or maturity. Horm Behav, 5: 93-102.
Mugford RA, Nowell NW (1970). Pheromones and their effect on aggression in mice. Nature, 226: 967-968.
308 Mukai H, Tsurugizawa T, Murakami G, Kominami S, Ishii H, Ogiue-Ikeda M et al. (2007). Rapid modulation of long-term depression and spinogenesis via synaptic estrogen receptors in hippocampal principle neurons. J Neurochem, 100: 950-967.
Murakami G, Richard H, Fontaine C, Ribeiro AC, Pfaff DW (2011). Relationships among estrogen receptor, oxytocin and vasopressin gene expression and social interaction in male mice. Eur J Neurosci, 34: 469-477.
Murchison CF, Zhang X-Y, Zhang W-P, Ouyang M, Lee AW, Thomas SA (2004). A distinct role for norepinephrine in memory retrieval. Cell, 117: 131-143.
Nachman M, Ashe JH (1974). Effects of basolateral amygdala lesions on neophobia, learned taste aversions, and sodium appetite in rats. J Comp Physiol Psychol, 87: 622-643.
Nakagawa S (2004). A farewell to Bonferroni: the problems of low statistical power and publication bias. Behav Ecol, 15: 1044-1045.
Nath A, Sitruk-Ware R (2009). Pharmacology and clinical applications of selective estrogen receptor modulators. Climacteric, 12: 188-205.
National Center for Mental Health Promotion and Youth Violence Prevention (2011). Girls Bullying and Violence. http://www promoteprevent org/publications/prevention- briefs/girls-bullying-and-violence,
Neese SL, Korol DL, Katzenellenbogen JA, Schantz SL (2010). Impact of estrogen receptor alpha and beta agonists on delayed alternation in middle-aged rats. Horm Behav, 58: 878- 890.
Nelson RJ (2005). An Introduction to Behavioral Endocrinology. 3rd ed.:
Neumann ID (2008). Brain oxytocin: a key regulator of emotional and social behaviours in both females and males. J Neuroendocrinol, 20: 858-865.
Newman SW (1999). The medial extended amygdala in male reproductive behavior. Ann N Y Acad Sci, 877: 242-257.
Nishimori K, Takayanagi Y, Yoshida M, Kasahara Y, Young LJ, Kawamata M (2008). New aspects of oxytocin receptor function revealed by knockout mice: sociosexual behaviour and control of energy balance. Prog Brain Res, 170: 79-90.
Nomura M, Andersson S, Korach KS, Gustafsson J-Å, Pfaff DW, Ogawa S (2006). Estrogen receptor-β gene disruption potentiates estrogen-inducible aggression but not sexual behaviour in male mice. Eur J Neurosci, 23: 1860-1868.
Nomura M, Durbak L, Chan J, Smithies O, Gustafsson J-Å, Korach KS et al. (2002). Genotype/age interactions on aggressive behavior in gonadally intact estrogen receptor β knockout (βERKO) male mice. Horm Behav, 41: 288-296.
309 Nomura M, Korach KS, Pfaff DW, Ogawa S (2003). Estrogen receptor β (ERβ) protein levels in neurons depend on estrogen receptor α (ERα) gene expression and on its ligand in a brain region-specific manner. Mol Brain Res, 110: 7-14.
Nunes S (2007). Dispersal and philopatry. In: Wolff JO, Sherman PW (eds.), Rodent Societies: An Ecological & Evolutionary Perspective, pp. 150-162.
O'Neil S (2008). Bullying by tween and teen girls: a literature, policy and resource review. The Society for Safe and Caring Schools & Communities,
Oberman LM, Ramachandran VS (2007). The simulating social mind: the role of the mirror neuron system and simulation in the social and communicative deficits of autism spectrum disorders. Psychol Bull, 133: 310-327.
Ogawa S (2004). Differential roles of two types of estrogen receptors in the regulation of neuroendocrine functions and socio-sexual behaviors. Jpn J Anim Psychol, 54: 49-58.
Ogawa S, Chan J, Chester AE, Gustafsson J-Å, Korach KS, Pfaff DW (1999). Survival of reproductive behaviors in estrogen receptor beta gene-deficient (beta ERKO) male and female mice. Proc Natl Acad Sci U S A, 96: 12887-12892.
Ogawa S, Chester AE, Hewitt SC, Walker VR, Gustafsson J-Å, Smithies O et al. (2000). Abolition of male sexual behaviors in mice lacking estrogen receptors alpha and beta (alpha beta ERKO). Proc Natl Acad Sci U S A, 97: 14737-14741.
Ogawa S, Choleris E, Pfaff DW (2004). Genetic Influences on Aggressive Behaviors and Arousability in Animals. Ann N Y Acad Sci, 1036: 257-266.
Ogawa S, Eng V, Taylor JA, Lubahn DB, Korach KS, Pfaff DW (1998a). Roles of estrogen receptor-alpha gene expression in reproduction-related behaviors in female mice. Endocrinology, 139: 5070-5081.
Ogawa S, Washburn TF, Taylor JA, Lubahn DB, Korach KS, Pfaff DW (1998b). Modifications of testosterone-dependent behaviors by estrogen receptor-alpha gene disruption in male mice. Endocrinology, 139: 5058-5069.
Olivier B, Mos J (1992). Rodent models of aggressive behavior and serotonergic drugs. Prog Neuropsychopharmacol Biol Psychiatry, 16: 847-870.
Olivier B, Mos J, Miczek KA (1991). Ethopharmacological studies of anxiolytics and aggression. Eur Neuropsychopharm, 1: 97-100.
Olivier B, Mos J, Rasmussen D (1990). Behavioural pharmacology of the serenic, eltoprazine. Rev Drug Metab Drug Interact, 8: 31-83.
Olivier B, Mos J, Van der Heyden J, Hartog J (1989). Serotonergic modulation of social interactions in isolated male mice. Psychopharmacology, 97: 154-156.
310 Ophir AG, Zheng D-J, Eans S, Phelps SM (2009). Social investigation in a memory task relates to natural variation in septal expression of oxytocin receptor and vasopressin receptor 1a in prairie voles (Microtus ochrogaster). Behav Neurosci, 123: 979-991.
Packard MG (1998). Posttraining estrogen and memory modulation. Horm Behav, 34: 129- 139.
Palanza P, Della Seta D, Ferrari PF, Parmigiani S (2005). Female competition in wild house mice depends upon timing of female/male settlement and kinship between females. Anim Behav, 69: 1259-1271.
Palanza P, Parmigiani S, Vom Saal FS (1995). Urine marking and maternal aggression of wild female mice in relation to anogenital distance at birth. Physiol Behav, 58: 827-835.
Parmigiani S, Ferrari PF, Palanza P (1998). An evolutionary approach to behavioral pharmacology: using drugs to understand proximate and ultimate mechanisms of different forms of aggression in mice. Neurosci Biobehav Rev, 23: 143-153.
Parmigiani S, Palanza P, Rodgers RJ, Ferrari PF (1999). Selection, evolution of behavior and animal models in behavioral neuroscience. Neurosci Biobehav Rev, 23: 957-970.
Patisaul HB, Burke KT, Hinkle RE, Adewale HL, Shea D (2009). Systemic administration of diarylpropionitrile (DPN) or phytoestrogens does not affect anxiety-related behaviors in gonadally intact male rats. Horm Behav, 55: 319-328.
Peacock MM, Jenkins SH (1988). Development of food preferences: social learning by Beldingi's ground squirrel (Spermophilus beldingi). Behav Ecol Sociobiol, 22: 393-399.
Pedersen CA (1997). Oxytocin control of maternal behavior Regulation by sex steroids and offspring stimuli. Ann N Y Acad Sci, 807: 126-145.
Peters PJ, Bronson FH, Whitsett JM (1972). Neonatal castration and intermale aggression in mice. Physiol Behav, 8: 265-268.
Pfaff DW, Ribeiro AC, Matthews J, Kow L-M (2008). Concepts and mechanisms of generalized central nervous system arousal. Ann N Y Acad Sci, 1129: 11-25.
Phan A, Lancaster KE, Armstrong JN, MacLusky NJ, Choleris E (2011). Rapid effects of estrogen receptor α and β selective agonists on learning and dendritic spines in female mice. Endocrinology, 152: 1492-1502.
Pibiri F, Nelson M, Carboni G, Pinna G (2006). Neurosteroids regulate mouse aggression induced by anabolic androgenic steroids. Neuroreport, 17: 1537-1541.
Pierman S, Douhard Q, Bakker J (2008). Evidence for a role of early oestrogens in the central processing of sexually relevant olfactory cues in female mice. Eur J Neurosci, 27: 423-431.
311 Pietropaolo S, Branchi I, Cirulli F, Chiarotti F, Aloe L, Alleva E (2004). Long-term effects of the periadolescent environment on exploratory activity and aggressive behaviour in mice: Social versus physical enrichment. Physiol Behav, 81: 443-453.
Popik P, van Ree JM (1992). Long-term facilitation of social recognition in rats by vasopressin related peptides: a structure-activity study. Life Sci, 50: 567-572.
Popik P, van Ree JM (1998). Neurohypophyseal peptides and social recognition in rats. Prog Brain Res, 119: 415-436.
Popik P, van Ree JM (1991). Oxytocin but not vasopressin facilitates social recognition following injection into the medial preoptic area of the rat brain. Eur Neuropsychopharm, 1: 555-560.
Popik P, Vos PE, van Ree JM (1992). Neurohypophyseal hormone receptors in the septum are implicated in social recognition in the rat. Behav Pharmacol, 3: 351-358.
Priest CA, Borsook D, Pfaff DW (1997). Estrogen and stress interact to regulate the hypothalamic expression of a human proenkephalin promoter-beta-galactosidase fusion gene in a site-specific and sex-specific manner. J Neuroendocrinol, 9: 317-326.
Prossnitz ER, Barton M (2009). Signaling, physiological functions and clinical relevance of the G protein-coupled estrogen receptor GPER. Prostaglandins Other Lipid Mediat, 89: 89-97.
Ragnauth AK, Devidze N, Moy V, Finley K, Goodwillie A, Kow L-M et al. (2005). Female oxytocin gene-knockout mice, in a seminatural environment, display exaggerated aggressive behavior. Genes Brain Behav, 4: 229-239.
Rankin GO, Hong SK, Valentovic MA, Beers KW, Anestis DK, Nicoll DW et al. (1997). Effects of sodium sulfate on acute N-(3,5-dichlorophenyl)succinimide (NDPS) nephrotoxicity in the Fischer 344 rat. Toxicology, 123: 1-13.
Rasband WS (2011). ImageJ. U S NIH, 1997-2011:
Rehfeldt RA, Kinney EM, Root S, Stromer R (2004). Creating activity schedules using Microsoft Powerpoint. J Appl Behav Anal, 37: 115-128.
Resnick SM, Espeland MA, An Y, Maki PM, Coker LH, Jackson RD et al. (2009). Effects of conjugated equine estrogens on cognition and affect in postmenopausal women with prior hysterectomy. J Clin Endocrinol Metab, 94: 4152-4161.
Resnick SM, Espeland MA, Yang A, Maki PM, Coker LH, Jackson RD et al. (2004). Effects of conjugated equine estrogen in postmenopausal women with hysterectomy: the Women's Health Initiative randomized controlled trial. JAMA, J Am Med Assoc, 291: 1701-1712.
312 Revankar CM, Cimino DF, Sklar LA, Arterburn JB, Prossnitz ER (2005). A transmembrane intracellular estrogen receptor mediates rapid cell signaling. Science, 307: 1625-1630.
Rhodes ME, Frye CA (2006). ERβ-selective SERMs produce mnemonic-enhancing effects in the inhibitory avoidance and water maze tasks. Neurobiol Learn Mem, 85: 183-191.
Ribeiro AC, Pfaff DW, Devidze N (2009). Estradiol modulates behavioral arousal and induces changes in gene expression profiles in brain regions involved in the control of vigilance. Eur J Neurosci, 29: 795-801.
Ricceri L, Minghetti L, Moles A, Popoli P, Confaloni A, De Simone R et al. (2004). Cognitive and neurological deficits induced by early and prolonged basal forebrain cholinergic hypofunction in rats. Exp Neurol, 189: 162-172.
Rieucau G, Giraldeau L-A (2011). Exploring the costs and benefits of social information use: an appraisal of current experimental evidence. Philosophical Transactions of the Royal Society Series B, Biological Sciences, 366: 949-957.
Rissman EF, Heck AL, Leonard JE, Shupnik MA, Gustafsson J-Å (2002a). Disruption of estrogen receptor beta gene impairs spatial learning in female mice. Proc Natl Acad Sci U S A, 99: 399-4001.
Rissman EF, Heck AL, Leonard JE, Shupnik MA, Gustafsson J-Å (2002b). Disruption of estrogen receptor beta gene impairs spatial learning in female mice. Proceedings of the National Academy of Sciences of the United States of America, 99: 399-4001.
Robb-Nicholson C (2009). By the way, doctor. The “Women's Health Initiative” found that hormone therapy wasn't helpful for avoiding dementia; there was some suggestion that it might even cause cognitive problems. Am I at risk for dementia by continuing hormone therapy? Harv Womens Health Watch, 17: 8-
Rodgers RJ (1990). Steroidogenic cytochrome P450 enzymes and ovarian steroidogenesis. Reprod Fertil Dev, 2: 153-163.
Rodriguiz RM, Chu R, Caron MG, Wetsel WC (2004). Aberrant responses in social interaction of dopamine transporter knockout mice. Behav Brain Res, 148: 185-198.
Rood BD, De Vries GJ (2011). Vasopressin innervation of the mouse (Mus musculus) brain and spinal cord. J Comp Neurol, 519: 2434-2474.
Rood BD, Murray EK, Laroche J, Yang MK, Blaustein JD, De Vries GJ (2008). Absence of progestin receptors alters distribution of vasopressin fibers but not sexual differentiation of vasopressin system in mice. Neuroscience, 154: 911-921.
Roper JA, O'Carroll A-M, Young WSI, Lolait SJ (2011). The vasopressin Avpr1b receptor: molecular and pharmacological studies. Stress, 14: 98-115.
313 Ross HE, Young LJ (2009). Oxytocin and the neural mechanisms regulating social cognition and affiliative behavior. Front Neuroendocrinol, 30: 534-547.
Ross RS, Eichenbaum H (2006). Dynamics of hippocampal and cortical activation during consolidation of a nonspatial memory. J Neurosci, 26: 4852-4859.
Ross RS, McGaughy J, Eichenbaum H (2005). Acetylcholine in the orbitofrontal cortex is necessary for the acquisition of a socially transmitted food preference. Learn Mem, 12: 302-306.
Rothman KJ (1990). No adjustments are needed for multiple comparisons. Epidemiology, 1: 43-46.
Russon AE (1997). Exploiting the expertise of others. In: Whiten A, Byrne RW (eds.), Machiavellian Intelligence II: Extension and Evaluations, pp. 174-231.
Sakuma Y (2009). Gonadal steroid action and brain sex differentiation in the rat. J Neuroendocrinol, 21: 410-414.
Samuelsen CL, Meredith M (2011). Oxytocin antagonist disrupts male mouse medial amygdala response to chemical-communication signals. Neuroscience, 180: 96-104.
Sánchez-Andrade G, James BM, Kendrick KM (2005). Neural encoding of olfactory recognition memory. J Reprod Dev, 51: 547-558.
Sánchez-Andrade G, Kendrick KM (2011). Roles of α- and β-estrogen receptors in mouse social recognition memory: effects of gender and the estrous cycle. Horm Behav, 59: 114-122.
Sánchez-Andrade G, Kendrick KM (2009). The main olfactory system and social learning in mammals. Behav Brain Res, 200: 323-335.
Sanchis-Segura C, Correa M, Aragon CMG (2000). Lession on the hypothalamic arcuate nucleus by estradiol valerate results in a blockade of ethanol-induced locomotion. Behav Brain Res, 114: 57-63.
Sandstrom NJ, Williams CL (2004). Spatial memory retention is enhanced by acute and continuous estradiol replacement. Horm Behav, 45: 128-135.
Santollo J, Wiley MD, Eckel LA (2007). Acute activation of ERα decreases food intake, meal size, and body weight in ovariectomized rats. Am J Physiol Regul Integr Comp Physiol, 293: 2194-2201.
Saudou F, Ait Amara D, Dierich A, LeMeur M, Ramboz S, Segu L et al. (1994). Enhanced aggressive behavior in mice lacking 5-HT1B receptor. Science, 265: 1875-1878.
Scordalakes EM, Rissman EF (2004). Aggression and arginine vasopressin immunoreactivity regulation by androgen receptor and estrogen receptor α. Genes Brain Behav, 3: 20-26.
314 Scordalakes EM, Rissman EF (2003). Aggression in male mice lacking functional estrogen receptor alpha. Behav Neurosci, 117: 38-45.
Scott JP (1966). Agonistic behavior of mice and rats: a review. Am Zool, 6: 683-701.
Scott JP, Fredericson E (1951). The causes of fighting in mice and rats. Physiol Zool, 24: 273-309.
Senju A, Maeda M, Kikuchi Y, Hasegawa T, Tojo Y, Osanai H (2007). Absence of contagious yawning in children with autism spectrum disorder. Biol Lett, 3: 706-708.
Shahrokh DK, Zhang TY, Diorio J, Gratton A, Meaney MJ (2010). Oxytocin-dopamine interactions mediate variations in maternal behavior in the rat. Endocrinology, 151: 2276- 2286.
Shane HC, Albert PD (2008). Electronic screen media for persons with autism spectrum disorders: results of a survey. J Autism Dev Disord, 38: 1499-1508.
Shea SD, Katz LC, Mooney R (2008). Noradrenergic induction of odor-specific neural habituation and olfactory memories. J Neurosci, 28: 10711-10719.
Sherwin BB (1988). Estrogen and/or androgen replacement therapy and cognitive functioning in surgically menopausal women. Psychoneuroendocrinology, 13: 345-357.
Shugrue PJ, Lane MV, Merchenthaler I (1997). Comparative distribution of estrogen receptor-α and -β mRNA in the rat central nervous system. J Comp Neurol, 388: 507- 525.
Shumaker SA, Legault C, Rapp SR, Thal L, Wallace RB, Ockene JK et al. (2003). Estrogen plus progestin and the incidence of dementia and mild cognitive impairment in postmenopausal women: the Women's Health Initiative Memory Study: a randomized controlled trial. JAMA, J Am Med Assoc, 289: 2651-2662.
Simon NG, Whalen RE (1986). Hormonal regulation of aggression: Evidence for a relationship among genotype, receptor binding, and behavioral sensitivity to androgen and estrogen. Aggress Behav, 12: 255-266.
Singh M, Sétáló GJ, Guan X, Frail DE, Toran-Allerand CD (2000). Estrogen-induced activation of the mitogen-activated protein kinase cascade in the cerebral cortex of estrogen receptor-α knock-out mice. J Neurosci, 20: 1695-1700.
Slotnick BM (1973). Fear behavior and passive avoidance deficits in mice with amygdala lesions. Physiol Behav, 11: 717-720.
Solomon NG (2003). A reexamination of factors influencing philopatry in rodents. J Mammal, 84: 1182-1197.
315 Spencer JL, Waters EM, Romeo RD, Wood GE, Milner TA, McEwen BS (2008). Uncovering the mechanisms of estrogen effects on hippocampal function. Front Neuroendocrinol, 29: 219-237.
Spiteri T, Ågmo A (2009). Ovarian hormones modulate social recognition in female rats. Physiol Behav, 98: 247-250.
Spiteri T, Musatov S, Ogawa S, Ribeiro AC, Pfaff DW, Ågmo A (2010). The role of the estrogen receptor α in the medial amygdala and ventromedial nucleus of the hypothalamus in social recognition, anxiety and aggression. Behav Brain Res, 210: 211- 220.
Stauffer SR, Coletta CJ, Tedesco R, Nishiguchi G, Carlson KE, Sun J et al. (2000). Pyrazole ligands: structure-affinity/activity relationships and estrogen receptor-alpha-selective agonists. J Med Chem, 43: 4934-4947.
Steimer T (1993). Steroid hormone metabolism. In: Campana A, Dreifuss JJ, Sizonenko P, Vassalli JD, Villar J (eds.), Reproductive health: Second postgraduate course for training in reproductive medicine and reproductive biology, pp.
Stephens WE, Ludy IE (1975). Action-concept learning in retarded children using photographic slides, motion pictures, sequences and live demonstrations. Am J Ment Defic, 80: 277-280.
Stovall DW, Pinkerton JV (2008). Estrogen agonists/antagonists in combination with estrogen for prevention and treatment of menopause-associated signs and symptoms. Womens Health, 4: 257-268.
Strupp BJ, Bunsey M, Bertsche B, Levitsky DA, Kesler M (1990). Enhancement and impairment of memory retrieval by a vasopressin metabolite: an interaction with the accessibility of the memory. Behav Neurosci, 104: 268-276.
Sturmburg JP, Pond DC (2009). Impacts on clinical decision making - changing hormone therapy management after the WHI. Aust Fam Phys, 38: 249-255.
Sun W, Mercado EI, Wang P, Shan X, Lee T-C, Salvi RJ (2005). Changes in NMDA receptor expression in auditory cortex after learning. Neurosci Lett, 374: 63-68.
Svare B, Gandelman R (1973). Postpartum aggression in mice: Experiential and environmental factors. Horm Behav, 4: 323-334.
Takayanagi Y, Masahide Y, Bielsky IF, Ross HE, Kawamata M, Onaka T et al. (2005). Pervasive social deficits, but normal parturition, in oxytocin receptor-deficient mice. Proc Natl Acad Sci U S A, 102: 16096-16101.
Talamini LM, Koch T, Luiten PGM, Koolhaas JM, Korf J (1999). Interruptions of early cortical development affect limbic association and social behaviour in rats; possible relevance for neurodevelopmental disorders. Brain Res, 847: 105-120.
316 Tang AC, Nakazawa M, Romeo RD, Reeb BC, Sisti H, McEwen BS (2005). Effects of long- term estrogen replacement on social investigation and social memory in ovariectomized C57BL/6 mice. Horm Behav, 47: 350-357.
Taylor SE, Klein LC, Lewis BP, Gruenewald TL, Gurung RAR, Updegraff JA (2000). Biobehavioral responses to stress in females: tend-and-befriend, not fight-or-flight. Psychol Rev, 107: 411-429.
Temple JL, Fugger HN, Li X, Shetty SJ, Gustafsson J-Å, Rissman EF (2001). Estrogen receptor β regulates sexually dimorphic neural responses to estradiol. Endocrinology, 142: 510-513.
Thomas P, Alyea RA, Pang Y, Peyton C, Dong J, Berg AH (2010). Conserved estrogen binding and signaling functions of the G protein-coupled estrogen receptor 1 (GPER) in mammals and fish. Steroids, 75: 595-602.
Thor DH, Flannelly KJ (1976). Age of intruder and territorial-elicited aggression in male Long-Evans rats. Behav Biol, 17: 237-241.
Tobin VA, Hashimoto H, Wacker DW, Takayanagi Y, Langnaese K, Caquineau C et al. (2010). An intrinsic vasopressin system in the olfactory bulb is involved in social recognition. Nature, 464: 413-417.
Toda K, Saibara T, Okada T, Onishi S, Shizuta Y (2001). A loss of aggressive behaviour and its reinstatement by oestrogen in mice lacking the aromatase gene (Cyp19). J Endocrinol, 168: 217-220.
Tollman J, King JA (1956). The effects of testosterone propionate on aggression in male and female C57B1/10 mice. Br J Anim Behav, 4: 147-149.
Trainor BC, Greiwe KM, Nelson RJ (2006). Individual differences in estrogen receptor α in select brain nuclei are associated with individual differences in aggression. Horm Behav, 50: 338-345.
Turner LM, Young AR, Römpler H, Schöneberg T, Phelps SM, Hoekstra HE (2010). Monogamy evolves through multiple mechanisms: evidence from V1aR in deer mice. Mol Biol Evol, 27: 1269-1278.
Uhrich J (1938). The social hierarchy in albino mice. J Comp Psychol, 25: 373-413.
Vale-Martínez A, Baxter MG, Eichenbaum H (2002). Selective lesions of basal forebrain cholinergic neurons produce anterograde and retrograde deficits in a social transmission of food preference task in rats. Eur J Neurosci, 16: 983-998.
Valentovic MA, Williams P, Carl JI, Rankin GO (1994). Urinary enzyme excretion as a parameter for detection of acute renal damage mediated by N-(3,5- dichlorophenyl)succinimide (NDPS) in Fischer 344 rats. J Appl Toxicol, 14: 281-285.
317 Valsecchi P, Choleris E, Moles A, Guo C, Mainardi M (1996). Kinship and familiarity as factors affecting social transfer of food preferences in adult Mongolian gerbils (Meriones unguiculatus). J Comp Psychol, 110: 243-251.
Valsecchi P, Galef BGJ (1989). Social influences on the food preferences of house mice (Mus musculus). Int J Comparat Psychol, 2: 245-256.
Valsecchi P, Mainardi M, Sgoifo A, Taticchi A (1989). Maternal influences on food preferences in weanling mice: Mus domesticus. Behav Process, 19: 155-166. van Wimersma Greidanus TB (1982). Disturbed behavior and memory of the Brattleboro rat. Ann N Y Acad Sci, 394: 655-662. van Wimersma Greidanus TB, Maigret C (1996). The role of limbic vasopressin and oxytocin in social recognition. Brain Res, 713: 153-159. van Wimersma Greidanus TB, van Ree JM, de Weid D (1983). Vasopressin and memory. Pharmacol Ther, 20: 437-458.
Vekovishcheva OY, Sukhotina IA, Zvartau EE (2000). Co-housing in a stable hierarchical group is not aversive for dominant and subordinate individuals. Neurosci Behav Physiol, 30: 1371-1324.
Waddington JL, O'Tuathaigh C, O'Sullivan G, Tomiyama K, Koshikawa N, Croke DT (2005). Phenotypic studies on dopamine receptor subtype and associated signal transduction mutants: insights and challenges from 10 years at the psychopharmacology- molecular biology interface. Psychopharmacology, 181: 611-638.
Walf AA, Frye CA (2005). ERβ-selective estrogen receptor modulators produce antianxiety behavior when administered systemically to ovariectomized rats. Neuropsychopharmacology, 30: 1598-1609.
Walf AA, Koonce CJ, Frye CA (2008a). Estradiol or diarylpropionitrile administration to wild type, but not estrogen receptor beta knockout, mice enhances performance in the object recognition and object placement tasks. Neurobiol Learn Mem, 89: 513-521.
Walf AA, Koonce CJ, Frye CA (2008b). Estradiol or diarylpropionitrile decrease anxiety-like behavior of wildtype, but not estrogen receptor beta knockout, mice. Behav Neurosci, 122: 974-981.
Walf AA, Rhodes ME, Frye CA (2006). Ovarian steroids enhance object recognition in naturally cycling and ovariectomized, hormone-primed rats. Neurobiol Learn Mem, 86: 35-46.
Walf AA, Rhodes ME, Frye CA (2004). Antidepressant effects of ERβ-selective estrogen receptor modulators in the forced swim test. Pharmacol Biochem Behav, 78: 523-529.
318 Walmer DK, Wrona MA, Hughes CL, Nelson KG (1992). Lactoferrin expression in the mouse reproductive tract during the natural estrous cycle: correlation with circulating estradiol and progesterone. Endocrinology, 131: 1458-1466.
Wang Y, Fontanini A, Katz DB (2007). Temporary basolateral amygdala lesions disrupt acquisition of socially transmitted food preferences in rats. Learn Mem, 13: 794-800.
Wersinger SR, Caldwell HK, Christiansen M, Young WSI (2007a). Disruption of the vasopressin 1b receptor gene impairs the attack component of aggressive behavior in mice. Genes Brain Behav, 6: 653-660.
Wersinger SR, Caldwell HK, Martinez L, Gold PW, Hu S-B, Young WSI (2007b). Vasopressin 1a receptor knockout mice have a subtle olfactory deficit but normal aggression. Genes Brain Behav, 6: 540-541.
Wersinger SR, Christiansen M, Hu S-B, Gold PW, Young WSI (2003). Aggression is reduced in vasopressin 1b receptor null mice but is elevated in vasopressin 1a receptor null mice. Soc Neurosci Abstr, 838: 14-
Wersinger SR, Ginns EI, O'Carroll A-M, Lolait SJ, Young WSI (2002). Vasopressin V1b receptor knockout reduces aggressive behaviour in male mice. Mol Psychiatry, 7: 975- 984.
Wersinger SR, Kelliher KR, Zufall F, Lolait SJ, O'Carroll A-M, Young WSI (2004). Social motivation is reduced in vasopressin 1b receptor null mice despite normal performance in an olfactory discrimination task. Horm Behav, 46: 638-645.
Wharton W, Dowling M, Khosropour CM, Carlsson C, Asthana S, Gleason CE (2009). Cognitive benefits of hormone therapy: cardiovascular factors and healthy-user bias. Maturitas, 64: 182-187.
Winocur G, McDonald RM, Moscovitch M (2001). Anterograde and retrograde amnesia in rats with large hippocampal lesions. Hippocampus, 11: 18-26.
Winslow JT, Hearn EF, Ferguson JN, Young LJ, Matzuk MM, Insel TR (2000). Infant vocalization, adult aggression, and fear behavior of an oxytocin null mutant mouse. Horm Behav, 37: 145-155.
Writing Group for the WHI Investigators (2002). Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA, J Am Med Assoc, 288: 321-333.
Yamamoto Y, Carter CS, Cushing BS (2006). Neonatal manipulation of oxytocin affects expression of estrogen receptor alpha. Neuroscience, 137: 157-164.
Young Women's Christian Association (2002). If looks could kill: young women and bullying. http://www platform51 org/downloads/resources/briefings/iflookscouldkill pdf,
319 Young LJ (2009). The neuroendocrinology of the social brain. Front Neuroendocrinol, 30: 425-428.
Young LJ, Nilsen R, Waymire KG, MacGregor GR, Insel TR (1999). Increased affiliative response to vasopressin in mice expressing the V1a receptor from a monogamous vole. Nature, 400: 766-768.
320