Comparative Neuromorphology and Electrophysiology of Purkinje Cells in the Leopard
Gecko, Mouse, and Chicken
By
Yi Fan Liu
A Thesis
presented to
The University of Guelph
In partial fulfilment of requirements
for the degree of
Master of Science
in
Biomedical Sciences
Guelph, Ontario, Canada
©Yi Fan Liu, June 2021
ABSTRACT
COMPARATIVE NEUROMORPHOLOGY AND ELECTROPHYSIOLOGY OF
PURKINJE CELLS IN THE LEOPARD GECKO, MOUSE, AND CHICKEN
Yi Fan Liu Advisor: Dr. M. K. Vickaryous
University of Guelph, 2021 Co-Advisor: Dr. C. D. C. Bailey
Purkinje cells (PCs) are one of the largest neurons in mammals, and play a key role in coordinating motor control, sensory perception, and various higher cognitive functions. While PCs are widely recognized as having elaborate dendrite branching patterns, for most species, details of their neuromorphology and function are rarely quantified. Here, I investigate PC dendrite neuromorphology in three distantly-related species and the electrophysiological properties of PCs from the leopard gecko. Using Golgi-Cox staining, I quantified neuromorphology in three species: the leopard gecko, laboratory mouse, and domestic chicken. I determined that there are species- specific differences in both PC size and dendrite complexity. Next, I investigated the electrophysiological properties of PCs in the leopard gecko. Based on dendrite neuromorphology and electrophysiological properties, I found that geckos have at least two PC subtypes. Taken together, my findings reveal that both within and between species, PCs are heterogeneous in form and function.
iii
ACKNOWLEDGEMENTS
To Matt: I never would have imagined that joining a research lab would feel so much like gaining a new family, yet the Eublephosphere welcomed me with open arms. Thank you for taking me on as both an undergrad and a graduate student, sometimes I’m still not quite sure what you saw in me, but I can now say that the skills and confidence I’ve gained since joining the lab speaks volumes about the kind of advisor you are. You never failed to support me even when I doubted myself, and I cannot stress enough how much I appreciate your patience and insight for both research and life advice. I believe we can all agree that the humour and enthusiasm you bring into the lab everyday has fostered a positive, relaxed atmosphere, which is largely the reason we feel comfortable enough to make niche memes and jokes around you (that you are so kind enough to tolerate). Thank you for being such a humble advisor and one of the kindest people I know, YOU are AWESOME for being able to juggle a million things at once and yet still find the time to be there for us all. Right on! *finger guns*
To Jalene, Isla, Piper, Miss Marbles, et al.: Thank you for all the kindness you have shown me, and all the support you give to the lab. I don’t think I will ever stop talking about Jae’s mulled wine that she made my first ever Festivus (that I showed up embarrassingly early for), because it really is that good. Isla and Piper, your drawings are amazing and I am incredibly grateful for all of your fan art (or should I say Yifan-art...haha), don’t forget me when you’re both famous!
To Craig: First of all, I want to let you know that you have an incredibly good poker face, and I’d like to say I’ve improved at knowing when you’re joking but the fact is I still only have a 50% success rate of guessing. Thank you for accepting me into your lab and letting me use your iv facilities even though I am technically part of The Gecko Lab, your kindness and humour really made me feel welcome even when I was a petrified young grad student. I will always remember our hallway chit chats and how you like to drink from the (kind of gross) water fountain with your beaker mug, and how alarmingly quiet you walk in our squeaky hallway. Thank you for taking us to extracurricular activities like bowling and treetop trekking, two things I learned that
I am quite terrible at but fully enjoyed with the lab nonetheless. You are an amazing co-advisor and I am forever grateful for your support with my project, keep being Craig-cellent!
To Kathy: Hello Buckwheat! To say that you’ve been a great help to my project is an understatement- I think all of us in the lab would be lost without your insight and expertise.
Every time any one of us comes to you with a problem, you either have the solution, or are willing to help us find it. I also appreciate your ability to have a funny story on hand every time we see you, because you just bring that much joy and hilarity into the lab. Kathy, you are an amazing mentor and an even better friend, and one of the most dedicated scientists I know (who really knows how to work hard AND play hard). I am so honoured to have worked alongside you for all these years, and I hope you know that I will be thinking of you every time I see a groundhog or drink rosé wine.
To Ashutosh: I would like to formally apologize for the number of times I have asked you to repeat concepts of ephys to me, but I would also like to thank you for your infinite patience and wisdom. I used to be intimidated by every person I met who was really smart and good at absolutely everything, but after meeting you, I am intimidated by you and only you now. You have been such a huge help to both parts of my project, for which I will forever be grateful for.
Your love for science and your affinity for ephys is really unparalleled, and I know that you will v go on to accomplish amazing things in the future. Thank you for all that you do for the lab, and for taking the time to help me with my project!
To Laura: Pepe Pepe Pepe...where do I even begin. I think the fact that we are pretty much the same person says enough about how much you mean to me, but I’ll keep this short and sweet: you made the bad days better, and the good days amazing. I cannot believe the amount of work we were able to get done when left alone, but we really somehow managed to juggle doing actual science with water gloves and Naruto runs. Thank you for being there for me through the mental breakdowns and the caffeine-fueled shenanigans, I wish you as much murder podcasts and hummus wraps as your heart desires for the rest of your life.
To Chloe: Never have I met someone who radiates so much Horse Girl and Instagram Influencer
Energy, but somehow I can’t imagine you any other way. I have to admit you really are good at having a witty comeback for every situation, and are able to bring endlessly positive energy and kind-heartedness with you everywhere you go (I see you cringing right now, I’m sorry). Thank you for helping me try new things and getting me out of my comfort zone, these past few years were never boring with you around. I want you to know that you are an excellent scientist, and that I will forever be thankful for and appreciate your vine references and niche memes.
To Elizabeth: Thank you for being my adventure buddy, and fellow Timmies enthusiast. You understood exactly what I was going through with my project, and I genuinely don’t think I could’ve finished this degree without you there. Never have I thought that belting out medieval covers of pop songs at 2am would be so therapeutic, but you really opened my eyes to many new things. I am so grateful for the support system that we’ve developed, you are an amazingly talented researcher and a great friend and I cannot wait to see you in the tunnels again someday. vi
To MinhHanh: You started out as an undergraduate summer student, and yet you managed to
EMBED yourself into the HEARTS of everyone in the lab (sorry, I had to). Knowing you now, I wish we started talking to each other earlier because you are extremely witty and personable without even trying. I am so thankful for all of the support and consultation you have given me, you may be younger than me but you have such an insightful perspective on things that I often fail to see. For being one of the smartest and most hardworking people I know, I trust that everything you do will pay off someday, and that you will accomplish anything you reach for.
To Sarah: My anatomy queen, thank you for all the support and laughs you’ve given me. You never fail to go above and beyond to help any member of our lab, even when you are busy with your job downstairs. I appreciate your impeccable taste in music as well as YouTubers, and I cannot wait to discuss new episodes with you as they come out. You are as talented as you are funny, and I thank you for always matching the energy I give you without fail every time.
To all the past and present members of the Eublephosphere: To Gabby, Keeley, Catherine,
Jordyn, Karemna, Rebecko, and Stef: thank you for being some of the smartest, funniest, most friendly, and supportive group of scientists I have worked with. You all have truly made my graduate experience an unforgettable one, and I will be forever grateful for your friendship and support.
To the Bailey Lab: Electrophysiology and neuroscience are not easy subjects to study, but you guys make it seem like a piece of cake. To Cassandra, Chloe, Anthony, Anna, Aariez, Ashley, and Ryan: thank you for accepting me into your lab, you are all undoubtedly some of the most talented researchers I have had the pleasure of working with, and I am thankful to have been friends with you all. vii
To my advisory and examination committees: Dr. Descalzi, Dr. MacLusky, and Dr. Saleh, thank you for taking the time to provide valuable input and insightful comments on my thesis, and for asking all the right questions to challenge my knowledge.
To my family and friends: I would like to thank my parents for their constant support these past couple of years. While I know it hadn’t been easy for you, I appreciate you both taking the time to drive two hours to bring me homemade meals and groceries because you “didn’t want me to starve”. Although we weren’t able to see each other for these past few months/years, I would like to extend a huge thank you to my friends for constantly checking in to ask me how I’m doing, and for enthusiastically congratulating me on all my achievements.
To all my friends in the Biomedical Science department, as well as anyone else I might have missed: thank you for making this project possible.
viii
TABLE OF CONTENTS
ABSTRACT……………………………………………………………………………………...ii
ACKNOWLEDGEMENTS…………………………………………………………………….iii
TABLE OF CONTENTS……………………………………………………………………...viii
LIST OF FIGURES……………………………………………………………………….……xii
LIST OF TABLES……………………………………………………………………….…….xiv
LIST OF APPENDICES…………………………………………………………………….…xv
LIST OF ABBREVIATIONS…………………………………………………………………xvi
DECLARATION OF WORK PERFORMED………………………………………………xvii
CHAPTER 1: LITERATURE REVIEW………………………………………………………1
1.1 The Cerebellum………………………………………………………………………1
1.1.1 Development of the cerebellum…………………………………………...1
1.1.2 Morphological variation of the cerebellum………………………………..3
1.1.3 Cytoarchitecture of the cerebellum………………………………………..4
1.2 The Purkinje Cell…………………………………………………………………….5
1.3 Visualizing Neurons………………………………………………………………….8
1.4 Neuromorphology and Electrophysiology………………….…….………….……11
1.5 Rationale, Hypothesis, & Objectives………………………………………………16
CHAPTER 1 FIGURES………………………………………………………………………..18
CHAPTER 2: COMPARATIVE NEUROMORPHOLOGY OF PURKINJE CELLS IN THE
LEOPARD GECKO (EUBLEPHARIS MACULARIUS), LABORATORY MOUSE (MUS
MUSCULUS), AND DOMESTIC CHICKEN (GALLUS GALLUS DOMESTICUS)…………...20 ix
2.1 Introduction…………………………………………………………………………20
2.2 Materials and Methods……………………………………………………………..23
2.2.1 Animal Care……………….……………………………………………..23
2.2.2 Tissue Collection and Preparation ………………………………………24
2.2.3 Golgi-Cox Impregnation…………………………………………………25
2.2.4 Golgi-Cox Brain Tissue Slicing………………………………………….25
2.2.5 Golgi-Cox Brain Tissue Processing…………………………….…..……25
2.2.6 Golgi-Cox Brian Tissue Mounting and Dehydrating…………………….26
2.2.7 Microscopy and Imaging…………………………………….…………..26
2.2.8 Neuron Analysis and Statistics…………………………………………..27
2.3 Results……………………………………………………………………………….28
2.3.1 Cerebellar Structure and Anatomy……………………………………….28
2.3.2 Purkinje Cell Neuromorphology: Qualitative Observations……………..29
2.3.3 Purkinje Cell Neuromorphology: Quantitative Observations……………30
2.4 Discussion……………………………………….……….….………………………32
2.4.1 Intraspecific variation……………………………………………………33
2.4.2 Purkinje cells differ in both size and complexity across species………...34
2.4.3 Differences in Purkinje cell neuromorphology across species…………..35
CHAPTER 2 TABLES…………………………………………………………………………38
CHAPTER 2 FIGURES………………………………………………………………………..39
CHAPTER 3: ELECTROPHYSIOLOGICAL PROPERTIES AND
NEUROMORPHOLOGY OF PURKINJE CELLS IN THE LEOPARD GECKO
(EUBLEPHARIS MACULARIUS): EVIDENCE FOR MULTIPLE CELL TYPES……….48 x
3.1 Introduction…………………………………………………………………………48
3.2 Materials and Methods……………………………………………………………..50
3.2.1 Animal Care…………………………………………………….…….….50
3.2.2 Tissue Collection and Preparation……………………………………….51
3.2.3 Electrophysiology………………………………………………………..52
3.2.4 Neuron Morphology…………………….………………………………..53
3.2.5 Microscopy and Imaging………………………………………….……..54
3.2.6 Neuron and Statistical Analyses…………………………………………54
3.3 Results……………………………………………………………………………….55
3.3.1 Electrophysiological Properties………………………………………….55
3.3.2 Neuromorphology………………………………………………………..57
3.4 Discussion………………………………………………………………………...…58
3.4.1 Gecko Purkinje cells have distinct action potential firing patterns..…….60
3.4.2 Neuromorphological variation is present between subtypes…………….61
CHAPTER 3 TABLES…………………………………………………………………………62
CHAPTER 3 FIGURES………………………………………………………………………..63
CHAPTER 4: CONCLUDING STATEMENTS……………………………………………..70
4.1 Neuromorphological variation of Purkinje cells………………………………….70
4.2 Form and function of gecko Purkinje cells………………………………………..72
4.3 Differences between Golgi-Cox stained and Biocytin filled neurons……………73
REFERENCES………………………………………………………………………………….74
APPENDIX I: DETAILED HISTOCHEMICAL PROTOCOLS…………………………...92
APPENDIX II: TWO-DIMENSIONAL PURKINJE CELL TRACINGS……………….…96 xi
APPENDIX III: LOCATION OF LEOPARD GECKO PURKINJE CELLS OBTAINED
FROM WHOLE-CELL RECORDINGS……………………………………………………100
xii
LIST OF FIGURES
CHAPTER 1
Figure 1
The layers of the chicken cerebellum.
Figure 2
Tracings of Purkinje cells in different species.
CHAPTER 2
Figure 1
Golgi-Cox stained cerebellar slices from a gecko, mouse, and chicken.
Figure 2
Representative tracings of Purkinje cells from geckos, mice, and chickens.
Figure 3
Variation in Purkinje neuron size between geckos, mice, and chickens.
Figure 4
Outlier Purkinje cells within geckos, mice, and chickens.
Figure 5
Branched structure analyses of gecko, mouse, and chicken Purkinje cells.
Figure 6
Sholl analyses of gecko, mouse, and chicken Purkinje cells.
Figure 7
Differences in PCL to pial surface distance between geckos, mice, and chickens.
CHAPTER 3
Figure 1 xiii
Electrophysiological properties of type 1 and type 2 gecko Purkinje cells.
Figure 2
Action potential traces of type 1 and type 2 gecko Purkinje cells.
Figure 3
Representative neuron tracings of type 1 and type 2 gecko Purkinje cells.
Figure 4
Branched structure analyses of type 1 and type 2 gecko Purkinje cells.
Figure 5
Sholl analyses of type 1 and type 2 gecko Purkinje cells.
APPENDIX II
All two-dimensional gecko Purkinje cell tracings.
Figure 2
All two-dimensional mouse Purkinje cell tracings.
Figure 3
All two-dimensional chicken Purkinje cell tracings.
Figure 4
All two-dimensional gecko Purkinje cell tracings from whole-cell recordings.
APPENDIX III
Figure 1
Distribution of type 1 and type 2 gecko Purkinje cells by location in slice, and by animal.
xiv
LIST OF TABLES
Chapter 2
Table 1
Outlier Purkinje cells within geckos, mice, and chickens.
Table 2
Inter-species branched structure analyses of gecko, mouse, and chicken Purkinje cells.
Chapter 3
Table 1
Electrophysiological properties of type 1 and type 2 gecko Purkinje cells.
Table 2
Branched structure analyses of type 1 and type 2 gecko Purkinje cells.
xv
LIST OF APPENDICES
Appendix I
Detailed histochemical protocols and solutions.
Appendix II
Two-dimensional tracings of all leopard gecko, mouse, and chicken Purkinje cells.
Appendix III
Location of type 1 and type 2 leopard gecko cells.
xvi
LIST OF ABBREVIATIONS
°C degrees Celsius
ACSF artificial cerebrospinal fluid
AHP afterhyperpolarization diH2O deionized water
EtOH ethanol
GABA gamma-aminobutyric acid
GFP green fluorescent protein
GL granular layer
H&E hematoxylin and eosin
HRP horseradish peroxidase
ML molecular layer
PCL Purkinje cell layer
RMP resting membrane potential rpm revolutions per minute sACSF sucrose artificial cerebrospinal fluid
SEM standard error of the mean
SK small conductance calcium-activated potassium channels
TBS Tris-buffered saline
WM white matter
xvii
DECLARATION OF WORK PERFORMED
I declare that I performed all of the work in this thesis with the exception of the following: Stefanie Bradley assisted with the extraction of gecko brains, processing, imaging, and tracing of leopard gecko Purkinje cells used for the Golgi-Cox study. Craig Bailey assisted with the extraction of mouse brains. Matthew Vickaryous, Laura Austin, Sarah Donato, and
Craig Bailey assisted with the extraction of chicken brains. Laura Austin assisted with the extraction of gecko brains for whole-cell electrophysiology. Ashutosh Patel assisted with the extraction, slicing, and processing of gecko brains for electrophysiology, and performed all electrophysiological recordings and provided post-recording data for further analysis.
CHAPTER 1
LITERATURE REVIEW
1.1 The Cerebellum
The cerebellum is a prominent region of the vertebrate hindbrain classically defined by its role in providing sensory integration, motor coordination, and motor learning (Butts et al., 2011; Hibi et al., 2017). In addition to its well-known motor functions, emerging evidence has revealed that the cerebellum is also involved in higher cognitive functions as well, such as memory and speech
(Schmahmann, 1997; Diamond, 2000; Hayter et al., 2007; Baillieux et al., 2008; Cantalupo &
Hopkins, 2010; Buckner, 2013; Macrì et al., 2019). These diverse roles are enabled, in part, by an unusually large number of neurons. For example, in humans the cerebellum accounts for
~10% of brain mass but 80% of all brain neurons, the majority of which are one cell type – granule cells (=granule neurons) (Azevedo et al., 2009; Herculano-Houzel, 2009; Herculano-
Houzel, 2010). By comparison, the human cerebral cortex represents ~80% of the brain mass, but only 19% of brain neurons (~16 x 109; Azevedo et al., 2009). A similar pattern of distribution is also present in other mammals (and other species; Yopak et al., 2017), with the majority of neurons (up to 97.5% in elephants) residing within the cerebellum (Herculano-Houzel et al.,
2014; Herculano-Houzel et al., 2015).
1.1.1 Development of the cerebellum
The cerebellum develops from the dorsal region of rhombomere 1, the most cranial region of the embryonic hindbrain (Butts et al., 2011; Hashimoto & Hibi, 2012; Butts et al.,
2014). In vertebrates, including reptiles and mammals, the cerebellum is first identifiable as a
1
pair of thickenings, the cerebellar plates (=anlagen), that gradually join together to cover the fourth ventricle (Larsen, 1926; Ten Donkelaar, 1998; see also Schoenwolf et al, 2009). The boundaries of the developing cerebellum are largely defined by a gradient of expression from various homeobox genes (e.g., Otx2, Hox2a, Gbx2), as well as the diffusible morphogen Fgf8
(Hashimoto & Hibi, 2012; White & Sillitoe, 2013; Butts et al., 2014). With continued growth, the cerebellar plates fuse and begin to protrude into the fourth ventricle, before rapidly expanding outwards (Ten Donkelaar, 1998).
Cerebellar specific precursor cells are first identified along the dorsoventral axis of the neural tube (Hatten & Heintz, 1995; Butts et al., 2014). Migration and ultimate organization of these cell populations is regulated by the signaling molecules semaphorin and reelin (Hashimoto
& Hibi, 2012). Under the influence of these molecules, two main populations or germinal zones become recognized: inhibitory gamma-aminobutyric acid (GABA)-ergic cells (which includes
Purkinje cell precursors, as well as Golgi and Stellate cells) and excitatory glutamatergic cells
(granule cells) (Hibi & Shimizu, 2012; White & Sillitoe, 2013). Whereas the glutamatergic cells are produced in the rhombic lip, located between the roof plate and the fourth ventricle, the
GABAergic populations come from the ventricular zone (Hatten & Heintz, 1995; White &
Sillitoe, 2013; Butts et al., 2014). Purkinje cell precursors migrate from the ventricular zone by embryonic day 13 (E13) with the help of cerebellar radial glia cells, whose radial processes extend towards the cerebellar plate. These radial processes serve as scaffolds for Purkinje cell migration (White & Sillitoe, 2013; Rahimi-Balaei et al., 2018). As revealed by the study of rodent development, Purkinje cells first form clusters along the cerebellar plate (by E17.5 in mice), before reorganizing into a single monolayer (by P4-5). Concurrently, Purkinje cells begin growing their axons and dendrites (White & Sillitoe, 2013; Rahimi-Balaei et al., 2018). Granule
2
cell precursors migrate tangentially from the rhombic lip towards the external germinal layer
(adjacent to the pial surface) (Rahimi-Balaei et al., 2018). From E12.5 to E17, the granule cell precursors then undergo mass proliferation to differentiate into post-mitotic granule cells, which migrate radially along the processes of Bergmann glial cells inwards past the Purkinje cell layer to form the granule cell layer (Butts et al., 2014; Hatten & Heintz, 1995).
1.1.2 Morphological variation of the cerebellum
The cerebellum demonstrates considerable variation in absolute size and gross morphology across different taxa. In jawless cyclostomes, the cerebellum is either microscopic (lamprey) or possibly absent (hagfish) (reviewed by Sugahara et al., 2017; Voogd et al., 2017). In many teleost fish and chondrichthyans (sharks and rays), the cerebellum resembles a simple dome, while in amphibians and reptiles it is a smooth-surfaced, leaf-shaped outgrowth (Ten Donkelaar,
1998; Hashimoto & Hibi, 2012; Yopak et al., 2016; Hibi et al., 2017; Macrì et al., 2019; Macri and Di-Poi, 2020). Among birds, mammals, mormyrid fish, and as well as some chondrichthyans, the cerebellum is characterized by the presence of transverse gyri/sulci (folds) or folia (Hashimoto & Hibi, 2012; Yopak et al., 2016; Hibi et al., 2017). As for gyri of the cerebral cortex, folia serve to increase the surface area of the cerebellar cortex, without a commensurate expansion in volume brain (Sereno et al., 2020). Curiously, both birds and mammals have ten folia (numbered, from rostral to caudal, I to X; Iwanuik et al., 2005;
Hashimoto & Hibi, 2012). In humans, this pattern of foliation is so tightly folded that only 10% of the cerebellum is externally visible (the rest is wedged into the sulci; Sultan & Glickstein,
2007). Indeed, a recent computer-generated model of a female human cerebellum estimate that the flattened and unfolded surface area of the cerebellar cortex to be only 10 cm wide but almost
3
100 cm long (Sereno et al., 2020). Among the most celebrated examples of cerebellar diversity is seen in mormyrid teleosts, a group of weakly-electric freshwater fish. Mormyrids (along with some other ray-finned fish) have specialized rostral region of the cerebellum known as the valvula (Nieuwenhuys and Nicholson,1967; Nieuwenhuys and Nicholson,1969; Meek and
Nieuwenhuys, 1998). In mormyrids, the valvula is hypertrophied and has expanded from its typical position (within the ventricle of the midbrain) to cover over the dorsal surface of the brain. Further, the outer surface of the valvula is heavily folded (Nieuwenhuys and
Nicholson,1967; Nieuwenhuys and Nicholson,1969; Meek and Nieuwenhuys, 1998). As currently understood, this massive expansion of the cerebellum (sometimes referred to as a
‘gigantocerebellum’; Nieuwenhuys and Nicholson,1969) is likely related to electroreception.
Among amniotes, the cerebellum is dominated by the corpus cerebelli (the main body of the cerebellum) and a small pair of laterally located flocculi (=auricles, flocculonodular lobes). In mammals and birds, the corpus cerebelli is further divided into a central vermis and a pair of bilateral hemispheres (Voogd & Glickstein, 1998; Sultan & Glickstein, 2007). In reptiles, the corpus cerebelli is organized into a central region, the pars interposita, flanked by laterally positioned pars lateralis (Larsell, 1926). Whether the vermis and pars interposita (and hemispheres and pars lateralis) are homologous structures (with comparable functions) remains unclear (although see Goodman, 1964).
1.1.3 Cytoarchitecture of the cerebellum
With the exception of cyclostomes, the histological organization of the cerebellum consists of a relatively thin (and sometimes foliated) cerebellar cortex surrounding a central core of white matter. In turn, the cerebellar cortex is organized into three layers: a cell-dense granular layer, a
4
cell-sparse molecular layer adjacent to the pial surface, and nested between, the Purkinje cell layer (Figure 1.1). This trilaminar organization is so widely conserved that it is described as a
“biological law” (Ramón y Cajal, 1909). These three layers are populated by a small number of neuron types, including Golgi cells, granule cells, molecular layer inhibitory interneurons, and
Purkinje cells (Hashimoto & Hibi, 2012). Golgi cells are large inhibitory GABA-ergic interneurons found in the granular layer. Golgi cells provide inhibitory input to granule cells and receive excitatory input from both granule cells and mossy fibers (D’Angelo et al., 2013).
Granule cells are small glutamatergic neurons with round soma found within the granular layer, and are the most abundant neuron in the cerebellar cortex (and the entire brain; Voogd &
Glickstein, 1998; Azevedo et al., 2009; Herculano-Houzel, 2009; Herculano-Houzel, 2010).
Granule cell dendrites receive afferent inputs from mossy fibres, while granule cell axons pass into the molecular layer and then bifurcate, giving rise to T-shape parallel fibers. Parallel fibres then make synaptic connections with dendrites of Purkinje cells and other interneurons (Voogd
& Glickstein, 1998; Sultan & Glickstein, 2007). Molecular layer inhibitory interneurons, including both basket and stellate cells, are small inhibitory GABAergic interneurons that act on
Purkinje cells (Voogd & Glickstein, 1998). The last neuronal type, Purkinje cells, are large, inhibitory GABAergic cells located in the Purkinje cell layer. These cells are the focus of the current investigation.
1.2 The Purkinje Cell
Purkinje cells are one of the largest and most distinct cell types in the vertebrate brain.
Purkinje cells have a relatively large cell body (soma; 25-40 microns in diameter in rodents), one or more primary dendrites, and a complex arborization of dendrites (Herndon, 1963). While the
5
cell body resides within the Purkinje cell layer, Purkinje cells dendrites branch into the molecular layer towards the pial surface (Sultan & Glickstein, 2007; Fujishima et al., 2018). Purkinje cell dendrites are essentially two-dimensional, and while they ramify in a space-filling manner transversely, they are non-overlapping in the parasagittal plane (Ramón y Cajal, 1911). As a result of this self-avoidant pattern, Purkinje cell dendrites are optimized for non-redundant synaptic connections with parallel fibers (Fujishima et al., 2018). In birds, mammals and some lizards (but not snakes or crocodylians), Purkinje cells demonstrate a discontinuous pattern of expression of the glycolytic isozyme Zebrin II (=aldolase C) across the cerebellum (Aspden et al., 2015; Gutierrez-Ibáñez et al., 2020). Consequently, immunostaining with Zebrin II generates alternating longitudinal stripes or zones reflecting expressing and non-expressing populations
(Voogd & Glickstein, 1998; White & Sillitoe, 2013).
Purkinje cells function as the sole output of the cerebellar cortex that operates through the release of inhibitory neurotransmitter GABA. While some Purkinje cells project their axons directly to the vestibular nuclei, most pass into the deep cerebellar nuclei within the white matter of the cerebellum. Purkinje cells provide inhibitory modulation to counteract excitatory input from mossy and climbing fibers to the deep cerebellar nuclei (Ito, 2006; Hibi et al., 2017). Each
Purkinje cell receives upwards of 100,000 excitatory synapses from parallel fibers, which occur at membranous protrusions known as spines (Ramón y Cajal, 1899; Benavides-Piccione et al.,
2002; Nishiyama & Linden, 2004; O’Brien & Unwin, 2005). Spines are unevenly distributed along the length of the dendrite, with regions close to the soma being essentially smooth (i.e., spineless) (Herndon, 1963; O’Brien & Unwin, 2005; Fujishima et al., 2018). Inhibitory inputs
(from basket and stellate cells) synapse on both the Purkinje cell soma and dendrites. The extent
6
to which information processing occurs within the cerebellum is directly linked to the number of parallel fiber synapses made with dendrites (Huang et al., 2014).
Despite their stereotypical neuromorphology (Smith et al., 1993), there are obvious qualitative species-specific differences in Purkinje cell structure (Figure 1.2). For example, the dendrite arbor of mammalian Purkinje cells is commonly reported to be more complex than those of fish, birds, or reptiles (Takeda et al., 1992; Kidd, 2017). Among mammals, Purkinje cells of the humpback whale appear to be unique in that they lack the ramifying network of other species. Instead, they have a simpler palisade pattern of arborization, with vertically oriented tertiary branches that closely resembles the neuromorphology of Purkinje cells in mormyrid fish
(Jacobs et al., 2014; Kidd, 2017). Among reptiles, less is known. To date, qualitative descriptions of Purkinje cells have only been reported for the American alligator (Alligator missisppiensis) and two species of turtle (Pseudemys scripta and Chrysemys picta) (Llinas & Hillman, 1969;
Chan & Nicholson, 1986). While the neuromorphology of alligator Purkinje cells was likened to those of birds, the dendritic arbor of turtles was reported to more closely resembled that of sharks and rays (Chan & Nicholson, 1986).
In addition to variation between species, the size and shape of the dendritic arbor can also vary within a species. For example, Purkinje cells in mice, bullfrogs, and humans may have one, two or even four primary dendrites (Uray & Gona, 1978; Kato et al., 1985; Nedelescu &
Abdelhack, 2013; Jacobs et al., 2014; Nedelescu et al., 2018). Research from mice suggests that this variation may be associated with different regions of the cerebellum. For example, in mice, the neuromorphology of Purkinje cells is reported to vary between different lobules of the vermis
(Nedelescu et al., 2018) and between cerebellar sulcus and apex (Nedelescu & Abdelhack,
2013).
7
One of the best-understood sources of neuromorphological variation in Purkinje cells is pathogenesis and injury. For example, spinocerebellar ataxia alters the dendrite arbor of Purkinje cells in both mice (Trzesniewski et al., 2018) and humans (Yang et al., 2000). Similarly, the dendritic arbors of Purkinje cells in white leghorn chickens (Gallus gallus domesticus) with congenital cerebellar hypoplasia are less organized when compared to controls (Nakamura et al.,
2014). In humans, a reduction or partial loss of dendrite arborization and dendrite spines, along with the presence of heterotopic Purkinje cells in the molecular layer, are characteristic of essential tremor (Louis et al., 2014; Louis et al., 2018). These pathogenic conditions underscore the importance of Purkinje cell neuromorphology in cerebellar function.
1.3 Visualizing Neurons
At the level of histology, the ability to visualize individual neurons can be accomplished using a variety of different techniques. One of the most common histochemical approaches for staining neural tissue is the Nissl method (Shimono & Tsuji, 1987; Kádár et al., 2009).
Developed by the histopathologist Franz Nissl (1860-1919) in 1894, the Nissl method uses a basic dye (such as cresyl violet or methylene blue) to stain the rough endoplasmic reticulum (so- called Nissl bodies), and nucleic acids (and therefore DNA and RNA) present in neurons (Kádár et al., 2009; Koyama, 2013). Hematoxylin and eosin (H&E) staining is another common staining procedure, which stains nucleic acids blue (hematoxylin) and proteins in surrounding cytoplasm and extracellular matrices various shades of pink (eosin) (Day, 2014). The Nissl method and
H&E stains reveal features such as the shape of the nucleus and soma, which can then be used to screen for pathological features (De Biase & Paciello, 2015). However, both methods fail to
8
capture the complexity and diversity of neurons, especially with respect to features such as dendrite arborization and axonal processes.
Another strategy for visualizing neurons involves labeling with a fluorescent marker
(such as green fluorescent protein, GFP) via a transferred reporter gene or viral transduction
(Czechowska et al., 2019; Farhoodi et al., 2019), or immunostaining for a neuron-specific antigen. Fluorescent labeling strategies can be used to target individual or small populations of neurons in vivo or ex vivo. Depending on the approach, such strategies may require the use of transgenic animal models, or species for which genetic markers are readily available
(Czechowska et al., 2019; Vints et al., 2019; Farhoodi et al., 2019). While immunostaining allows for a broad range potential antigen targets (De Biase & Paciello, 2015; Farhoodi et al.,
2019), results often come with a loss of cellular resolution and reveal highly convoluted neuronal networks in which it can be difficult to outline individual neuron morphology (De Biase &
Paciello, 2015; Czechowska et al., 2019; Vints et al., 2019; Farhoodi et al., 2019). This is likely due to the inability of the antibodies to recognize the protein of interest, as well as proper selection of secondary antibodies that result in the least amount of background staining, that result in poorer image quality and reduced ease of quantification (Glynn & McAllister, 2006).
Without question, one of the most important staining protocols for the study of neurons is the Golgi staining method. First developed by Camillo Golgi in 1873, this method involved impregnating neurons with heavy metals (Shimono & Tsuji, 1987; DeFelipe, 2015). The resulting “black reaction” permits neurons to be visualized using light microscopy. Of particular importance (and for reasons that remain unclear), Golgi’s method only stains a small percentage
(0.73-3%) of neurons (Pasternak & Woolsey, 1975; Zaquot & Kaindl, 2016). The ability to visually isolate single neurons marked one of the most important advancements in neuroscience,
9
and enabled (for the first time) detailed studies of neuronal circuitry, as well as the identification of dendrite spines (Ramón y Cajal, 1888; Vints et al., 2019). Further, use of the Golgi method provided Cajal with irrefutable evidence in support of the neuron doctrine – that the nervous system is composed of discrete cellular components (neurons) and was not a single reticulated network (Koyama, 2013; DeFelipe, 2015). Subsequent refinements to the original Golgi method have resulted in a number of related protocols, including the Rapid Golgi technique, developed by Ramón y Cajal, as well as the Golgi-Cox method, developed by Cox in 1891 (Koyama, 2013;
Czechowska et al., 2019). Of these, the Golgi-Cox method remains one of the most commonly employed methods for visualizing neurons.
The Golgi-Cox method differs from the original Golgi method in two main ways. First, while the Golgi method was performed on whole brains or large pieces of neural tissue, the
Golgi-Cox method was used to stain free-floating tissue sections (Vints et al., 2019). Second, the original Golgi method stained using silver nitrate, while the Golgi-Cox method uses a solution of potassium chromate, potassium dichromate, and mercuric chloride (Koyama, 2013; Vints et al.,
2019). As a result, the Golgi-Cox method demonstrates improved resolution of various neuromorphological details, including dendrite spines (García-López et al., 2007; Koyama, 2013;
Vints et al., 2019).
Another strategy for visualizing individual neurons involves use of an intracellular injection. This strategy builds off, and can be used in combination with, whole-cell recording of neurons. During electrophysiological recordings, various types of labeling compounds can be diffused into the neuron, thus tying together the neuromorphology and electrophysiological properties of the cell under investigation (Farhoodi et al., 2019). Commonly used labeling compounds include horseradish peroxidase (HRP), Lucifer yellow, and biocytin. HRP generates
10
a dark, Golgi-like reaction post-processing, but has the disadvantage that the molecule is too large to pass through gap junctions, and therefore it blocks high-resistance intracellular electrodes (Horikawa & Armstrong, 1988; Thomson & Armstrong, 2011). Lucifer yellow is strongly fluorescent and readily diffuses, but is prone to rapid fading and may alter the electrophysiological properties of recorded cells due to its solubility as a lithium salt (Horikawa
& Armstrong, 1988; Tasker et al., 1991; Thomson & Armstrong, 2011). Biocytin is a complex formed with biotin and lysine and was first introduced as an intracellular marker in 1988
(Horikawa & Armstrong, 1988). As an intracellular label, biocytin has two main advantages.
First, it is a small-sized and highly water-soluble molecule that is easily injected through pipettes
(Horikawa & Armstrong, 1988). Second, biocytin may serve as a high contrast marker, as it may be used with chromogenic precipitates or fluorescent labels to clearly distinguish the injected neuron from the surrounding tissue (Marx et al., 2012).
1.4 Neuromorphology and Electrophysiology
Neuronal electrophysiology is the study of the physiological function and electrical properties of individual neuronal cells. The basis of cellular communication resides in transmission of electric impulses through both interactions between synaptic (i.e. transmission of neurotransmitters) and intrinsic (i.e. function of sodium, calcium, and potassium ion channels) cellular excitability, spread by propagation of action potentials to deliver electrical signals
(Llinás, 1988; Kole & Stuart, 2012; Segev et al., 2016). In a series of studies published by
Hodgkin, Huxley, and Katz, the basis of action potential generation was described in the giant axon of the squid, where they demonstrated that membrane excitability is determined by the passive flow of ions across biological membranes according to their chemical gradients
11
(Hodgkin & Katz, 1949; Hodgkin & Huxley, 1952; Hodgkin & Huxley, 1952). Action potentials were found to be mediated by the conductance of sodium and potassium ions as a function of time (such that the action potential rises when the membrane is depolarized, and falls when repolarized) (Hodgkin & Huxley, 1952). The rapid activation of voltage-gated sodium channels at the threshold of the action potential outpacing the delayed activation of voltage-gated potassium channels gives rise to an all-or-none event that, with minor variations, is roughly similar amongst most species (Hodgkin & Huxley, 1952; Peterson, 2017). These early studies were conducted using the voltage clamp technique, where microelectrodes were inserted into the axon to measure currents through the cell membrane while holding the voltage constant
(Jackson, 1997; Sontheimer & Olsen, 2007). Modern electrophysiological studies often use the whole-cell recording technique, which involves rupturing the cell membrane with a single electrode to permit direct access to the neuronal cytoplasm (Hamill et al., 1981; Jackson, 1997;
Segev et al., 2016; Peterson, 2017). The whole cell voltage-clamp technique thus allowed for real-time visualization of the direct effects of cellular manipulations, such as drug applications, on cell functions and ion channels (Sontheimer & Olsen, 2007; Segev et al., 2016). A related technique, the current-clamp mode, allows for measurement of intrinsic excitability due to variations in membrane excitability induced by ion currents by injecting known current and measuring changes in voltage (Louth et al., 2016; Segev et al., 2016). As a result, whole-cell recording is widely accepted as a fundamental method for analyzing the electrophysiological properties and responses to different manipulations of neuronal cells.
Although the relationship between neuron morphology and function has long been recognized (beginning with the work of Ramón y Cajal, 1899), early electrophysiological studies were unclear as to the role of the dendrite arbor. Whereas some concluded that dendrites were
12
essentially passive, and played little role in electrical signalling within neurons (Llinas, 1988), this view is now obsolete. It is now well-understood that neuronal function does relate to neuron morphology, and the ability of these cells to integrate and relay signals (Levick, 1975; Miller &
Jacobs, 1984; Vetter et al., 2000; Krichmar et al., 2002). Among neuron types, the overall branching pattern (arborization), spatial distribution, length of dendrites, as well as the number and distribution of dendrite spines varies greatly. As such, neurons are commonly identified by a combination of their overall morphology, as well as their intrinsic electrophysiological properties, including their ion channels and firing activity (Llinas, 1988; Connors & Regehr,
1996).
The relationship between neuromorphology and function has frequently been modeled in silico based on pyramidal cells of the mammalian hippocampus. While not unique to the hippocampus, pyramidal cells are multipolar cells which function in the hippocampus to integrate special, contextual, and emotional information, and to transmit hippocampal output to other brain regions (Graves et al., 2012; Benavides-Piccione et al., 2020). Computer simulations of rat hippocampal CA3 pyramidal cells have demonstrated that the size of the dendritic arbor has a profound effect on the electrophysiological response of cells (Krichmar et al., 2002). More specifically, pyramidal cells with larger dendritic arbors were found to have a higher spiking frequency of action potentials, and tend to display plateau bursting patterns as opposed to regular bursting of cells with smaller arbors (Krichmar et al., 2002). Further, burst duration was influenced by the overall distribution of dendrites, as well as the number of dendritic terminals
(Krichmar et al., 2002). Similar results have also been reported for simulations of layer 5 pyramidal cells of the cat visual cortex, with a reduction in the size of the dendrite arbor resulting in a shift from burst firing to tonic firing (van Elburg & van Ooyen, 2010). Computational
13
modeling of hippocampal CA1 pyramidal cells in rats found that increasing the size of the soma and the length of the axon and dendrites resulted in a decrease of the firing threshold of the cell
(Yi et al., 2017).
Pyramidal cells also demonstrate regional differences in neuromorphology and electrophysiological properties. For example, in the tree shrew (Tupaia belangeri) compared to pyramidal cells of the primary visual cortex, pyramidal cells of the prefrontal cortex have more regular firing patterns, and larger dendrite trees with larger dendritic spines (Parra et al., 2019).
Similarly, in humans, dendrite tree size affected cell physiology such that larger dendrite trees resulted in fast action potential kinetics that enabled cells to track synaptic input activity with higher temporal precision than cells with smaller dendrite trees (Goriounova et al., 2018).
Among Purkinje cells, there are two main types of action potentials: simple spikes, and complex spikes (de Ruiter et al., 2006; Arancillo et al., 2015; Otis, 2016). Simple spikes are generated both spontaneously within Purkinje cells and by parallel fiber input, and are characterized by a relatively regular pattern and high firing frequencies (~40-100 spikes/second in resting animals; de Ruiter et al., 2006; Otis, 2016). Even in the absence of parallel fiber input,
Purkinje cells were recorded to fire at a regular rate, indicating an endogenous pace-making ability (Gruol & Franklin, 1987; Otis, 2016). For example, weaver (wv) mutant mice are characterized with an agranular cerebellum (the granule cells fail to migrate) and yet their
Purkinje cells closely resembled those of non-mutant mice, both neuromorphologically and functionally (Siggins et al., 1976). In contrast to simple spikes, complex spikes are generated in response to a large excitatory input from climbing fibers originating from the inferior olivary nucleus, and consist of a large initial single spike followed by several small spikelets (de Ruiter et al., 2006; Arancillo et al., 2015). Although most research related to the action potentials of
14
Purkinje cells has focused on lab rodents, it is worth noting that turtles (e.g., Pseudemys scripta and Chrysemys picta) and the spectacled caiman (Caiman sclerops) also demonstrated similar basic firing properties (Llinas & Nicholson, 1971; Hounsgaard & Midtgaard, 1988).
Interestingly, there is increasing evidence of region-specific variability in Purkinje cell firing patterns (Kim et al., 2012). For example, in mice Purkinje cells from lobules III-V have two distinct types of firing patterns (tonic firing, complex bursting), whereas those from lobule X have four (tonic firing, complex bursting, initial bursting, gap firing; Kim et al., 2012). These differences point towards lobule-specific functional differences contributing to signal processing.
Whether these functional differences in mice are correlated with quantitative differences in neuromorphology remains unclear.
Regional differences in Purkinje cell function and neuromorphology have been well described for mormyrid teleosts. In the central lobes and the valvula of the mormyrid cerebellum,
Purkinje cells demonstrate a palisade pattern of dendrite branching (the dendrites are long and oriented parallel) and have three distinct types of spikes: small, narrow sodium spikes; large, broad sodium spikes; and large, broad calcium spikes (Han & Bell, 2003; Zhang & Han, 2007).
However, in the posterior caudal lobe of the cerebellum there are three subtypes of Purkinje cells, distinguished on the basis of neuromorphology and function: multipolar, fan-shaped, and small Purkinje cells (Zhang & Han, 2007). Multipolar Purkinje cells have three or more dendrites originating from the soma in random directions, with each dendrite branching at irregular intervals, and cells only fire large, narrow spikes. Fan-shaped Purkinje cells also fire large, narrow spikes but have a distinct neuromorphology: their tertiary dendrites are short and are oriented parallel to each other (Zhang & Han, 2007). The third subtype are the small Purkinje cells, that have a small soma, one primary dendrite and exhibited both small, narrow spikes, and
15
large, broad spikes (Zhang & Han, 2007). Multipolar and small (but not fan-shaped) Purkinje cells are also found in the anterior caudal lobe of the cerebellum (Zhang et al., 2008; Zhang et al., 2010). The identification of regionally distinct Purkinje cells subtypes may be indicative of functional differences within different areas of the mormyrid cerebellum.
1.5 Rationale, Hypothesis, & Objectives
The neuromorphology and electrophysiological properties of neurons varies both within and between vertebrate species. Purkinje cells of the cerebellum participate in regulating fine- scale motor coordination as well as various cognitive functions. Although Purkinje cells are typically recognized as having a large soma and a ramifying pattern of dendrite branching, outside a few select taxa details of their neuromorphology and function remain poorly understood. The aim of my study is to characterize Purkinje cells across a number of vertebrate species. To quantify neuromorphological variation of Purkinje cells within and between species,
I conducted a neuromorphological investigation, focusing on three distantly-related species: the leopard gecko (Eublepharis macularius), laboratory mouse (Mus musculus), and domestic chicken (Gallus gallus domesticus). The gecko is an emerging squamate reptile model most notably studied for their tissue regeneration capabilities, of which limited neuromorphology information is available. In contrast, the mouse is the standard mammalian model used extensively across biomedical studies. The chicken is a routinely available avian that was historically significant in neuromorphology studies conducted by Ramón y Cajal. To investigate the relationship between Purkinje cell structure and function, I focused on the leopard gecko, as the neurobiology of reptiles is an emerging field of which electrophysiological investigations are rare.
16
I hypothesize that there will be significant quantitative differences in the neuromorphology and electrophysiological properties of Purkinje cells both within and between species. Given the role of the cerebellum in motor and cognitive processing, I predict that mouse
Purkinje cells will have a greater degree of complexity than those of geckos and chickens. As multiple subtypes of Purkinje cells have been previously found in multiple species, I predict that
I will observe more than one type of Purkinje cell in the leopard gecko that will differ in both electrophysiology and neuromorphology.
The objectives of my study are:
1) To characterize and quantify Purkinje cell neuromorphology in geckos, mice, and
chickens using Golgi-Cox staining.
2) To investigate Purkinje cell function and neuromorphology in geckos using whole-
cell electrophysiology.
17
Chapter 1 Figures
ML
PCL
GL
WM
Figure 1. Layers of The Cerebellum.
Chicken cerebellum stained using a modified Golgi-Cox method. The molecular layer (ML), granular layer (GL) and white matter (WM) can be identified, with various interneurons dispersed throughout and two Purkinje cells (left and right) with their somas located in the Purkinje cell layer
(PCL).
18
Figure 2. Variation in Purkinje neuron morphology across different taxa.
Tracings of Purkinje cells from different taxa, set to scale. Across verterbrates Purkinje cells vary considerably in both overall size and shape. a) weakly electric fish (Teleostii, Mormyridae), b) pond turtle (Testudines, Emydidae), c) albino guinea pig (Mammalia, Caviidae), d) Sprague-
Dawley rat (Mammalia, Muridae), e) finch (Aves, Fringillidae), f) alligator (Reptilia,
Alligatoridae), g) bat (Mammalia, Chiroptera). Adapted from Kidd (2017). Scale bar = 50µm.
19
CHAPTER 2
COMPARATIVE NEUROMORPHOLOGY OF PURKINJE CELLS IN THE LEOPARD
GECKO (EUBLEPHARIS MACULARIUS), LABORATORY MOUSE (MUS
MUSCULUS), AND DOMESTIC CHICKEN (GALLUS GALLUS DOMESTICUS)
2.1 Introduction
The cerebellum is a distinctive region of the hindbrain located dorsal to the fourth ventricle. Classically recognized as essential for coordinating motor control, the cerebellum is also involved in sensory perception, as well as higher emotional and cognitive functions (Smith et al., 1993; Hibi et al., 2017; Macrì et al., 2019; see also Yopak et al., 2017). At the level of gross morphology, the cerebellum demonstrates considerable diversity ranging from essentially absent as a discrete organ in hagfish and lamprey, to large and foliated in mammals and birds
(Voogd & Glickstein, 1998; Iwaniuk et al., 2006; Sugahara et al., 2016; Yopak et al., 2017). In reptiles and amphibians, the cerebellum is typically described as a simple leaf-like or domed structure (Voogd & Glickstein, 1998; Macrì and Di-Poi, 2020). Despite this external variation, the cellular organization of the cerebellum is highly conserved (Ramón y Cajal, 1909; Smith et al. 1993; Voogd & Glickstein, 1998). Invariably, the cerebellum has a trilaminar cytoarchitecture consisting of a molecular layer, a Purkinje cell layer, and a granular cell layer (Sultan &
Glickstein, 2007; Hibi et al., 2017). Each layer is defined in part by a unique assemblage of neurons. The molecular layer is relatively cell-sparse, but does have populations of inhibitory interneurons, including stellate cells (superficially) and basket cells (in deeper positions) (Voogd
& Glickstein, 1998; Jacobs et al., 2014). The granule cell layer is the most neuron-rich region of the brain, and is densely packed with excitatory granule cells. Nested between the two is the
20
Purkinje cell layer, dominated by the soma of Purkinje cells. Within the Purkinje cell layer,
Purkinje cell soma are often reported as being organized into either an orderly monolayer or having a scattered arrangement (Larsell, 1926; Macrì and Di-Poi, 2020; see also Yopak et al.,
2017).
Purkinje cells are one of the largest and most distinctive neuron types, with a large teardrop-shaped soma and a highly ramified dendrite arbour (Smith et al., 1993; Voogd &
Glickstein, 1998; Simons & Pellionisz, 2006; Sultan & Glickstein, 2007; Hibi et al., 2017;
Nedelescu et al., 2017; Yopak et al., 2017). Functionally, Purkinje cells are gamma-aminobutyric acid (GABA)-ergic inhibitory cells that serve as the sole motor output of the cerebellum. The inhibitory activity of Purkinje cells is modulated by excitatory inputs from granule cell parallel fibers, as well as a climbing fiber that projects from the inferior olivary nucleus (Voogd &
Glickstein, 1998; Nishiyama & Linden, 2004). Whereas parallel fibres make upwards of 100,000 synaptic connections with the mid- to distal regions of the dendrite arbour of each Purkinje cell, the climbing fibre (one per Purkinje cell) makes synaptic contact on the cell body and proximal portion of the dendrite tree (Voogd & Glickstein, 1998; Nishiyama & Linden, 2004). The axons of Purkinje cells pass directly into the deep cerebellar nuclei within the white matter of the cerebellum, and axons originating from the flocculonodular lobe and vermis are able to bypass to the vestibular nuclei (De Zeeuw & Berrebi, 1995; Voogd & Glickstein, 1998; Hibi et al., 2017).
Purkinje cell axons then synapse onto both excitatory and inhibitory neurons in the cerebellar and vestibular nuclei to innervate areas including the thalamus, spinal cord, and oculomotor complex (De Zeeuw & Berrebi, 1995; Teune et al., 1998).
Qualitative descriptions reveal that Purkinje cells are neuromorphologically heterogenous, both between and even within species (Kato et al., 1985; Nedelescu et al., 2017;
21
Kidd, 2017). Key examples of variation include the absolute size of the cell, the number of primary dendrites, and the overall geometric shape of the dendrite arbor (Uray & Gona, 1978;
Kato et al., 1985; Meek & Nieuwenhuys, 1991; O’Brien & Unwin, 2005; Nedelescu et al., 2017).
For example, in mice each Purkinje cell may have one or two primary dendrites, while in humans each neuron may have as many as four (Uray & Gona, 1978; Kato et al., 1985; Nedelescu &
Abdelhack, 2013; Jacobs et al., 2014; Nedelescu et al., 2018). Evidence from mice suggests that
Purkinje cells may also demonstrate regional variation, differing in neuromorphology between lobules of the vermis (Nedelescu et al., 2018) and between cerebellar sulcus and apex (Nedelescu
& Abdelhack, 2013). Further, in some species of mormyrid teleosts (e.g., Gnathonemus petersii and Petrocephalus bovei) the stereotypical branching dendrite arbor is replaced by a palisade pattern: long, unbranched tertiary dendrites in a parallel arrangement (Nieuwenhuys &
Nicholson, 1967; Meek & Nieuwenhuys, 1991; O’Brien & Unwin, 2005). A similar palisade pattern of Purkinje cell dendrites is also reported for humpback whales (Megaptera novaeangliae; Jacob et al., 2014). However, owing to the challenges of visualizing and tracing the (oftentimes complex) dendrite arbors (Jacobs et al., 2014) the neuromorphology of Purkinje cells in most taxa however remains unknown.
Here, we use a modified Golgi-Cox method to visualize Purkinje cells from three distantly related vertebrate species: leopard geckos, laboratory mice, and domestic chickens. To document neuromorphological variation both within and between species we performed Sholl and branched structure analyses from digital tracings. Within species, the neuromorphology of individual Purkinje cells was found to be a significant source of variation. We also found significant quantitative differences between species. While gecko and mouse Purkinje cells are similar in size, those of mice have more complex dendrite arbours. Purkinje cells in chickens are
22
larger in absolute size and demonstrate a distinctive palisade pattern of dendrites, similar to that of mormyrid fish. Altogether, our findings reveal that Purkinje cells are a heterogenous neuronal population with quantitatively and qualitatively distinct neuromorphologies, even within species.
2.2 Materials and Methods
2.2.1 Animal Care
Captive bred leopard geckos (Eublepharis macularius; hereafter ‘geckos’) were purchased from Global Exotic Pets (Kitchener, Ontario). All geckos (n=5) were subadults
(female; ~200 days old) at the time of data collection. Geckos were individually housed according to the protocols of McLean and Vickaryous (2011). Briefly, geckos were housed in a secure, temperature-controlled environmental chamber at the Hagen Aqualab, University of
Guelph (12:12h light cycle, with an average ambient temperature of 27.5°C). Each gecko was individually housed in an 18.9 L polycarbonate tank, with heating cables (Hagen Inc., Baie d’Urfe, Quebec, Canada) set to 32°C located underneath one side of the tank to establish a temperature gradient. Geckos were fed three larval mealworms (Tenebrio spp.) dusted with powdered calcium and vitamin D3 (cholecalciferol; Zoo Med Laboratories Inc., San Luis
Obispo, California, USA) daily, with access to fresh drinking water at all times. All geckos were cared for in accordance with Animal Utilization Protocol 3947, and the procedures outlined by the Canadian Council on Animal Care.
CD1 strain mice (Mus musculus) were purchased from Charles River Canada (Saint-
Constant, Quebec) and maintained as a local breeding colony. All mice in this study (n=5) were males collected at postnatal day 22 (with the day of birth represented as day 0). Mice were housed in groups of up to 4 per cage following the protocols outlined by Louth et al. (2016).
23
Mice were raised in in a secure vivarium at the Animal Care Facility, University of Guelph
(12:12h reverse light cycle, with an ambient temperature ranging from 21-24°C). Mice had ad libitum access to fresh drinking water and food. All mice were cared for in accordance with
Animal Utilization Protocol 3622, and the procedures outlined by the Canadian Council on
Animal Care.
White leghorn layer chickens (Gallus gallus domesticus; hereafter ‘chickens’) were provided by the Arkell Poultry Research Facility (Guelph, Ontario). Chickens were mature 81- week-old layer females (hens) at the time of collection, and were collected post-euthanasia as part of a scheduled population culling.
2.2.2 Tissue Collection and Preparation
Geckos were euthanized with a 150µL injection of Alfaxalone (85mg/kg) into the epaxial muscles of the neck, followed by decapitation, and immediate removal of the whole brain. The brain was then immersed into Golgi-Cox solution. Mice were anesthetized with 5% isoflurane, followed by decapitation and immediate removal of the whole brain. The brain was then immersed into Golgi-Cox solution. Chickens were euthanized using CO2 gas with a flow rate of
10-30% chamber volume per minute (vol/min) until the animal is insensible, upon which the flow rate is increased until death. Chicken brains were dissected out ~60 minutes following euthanasia and immersed into Golgi-Cox solution. None of the brains collected demonstrated any obvious signs of gross neuroanatomical abnormalities.
24
2.2.3 Golgi-Cox Impregnation
Staining followed a modified Golgi-Cox method, as outlined in Louth et al. (2017).
Whole brains were incubated in a scintillation vial (geckos, mice) or a glass vial (chickens) containing Golgi-Cox solution (1% (w/v) potassium dichromate, 0.8% (w/v) potassium chromate, and 1% (w/v) mercuric chloride) for 26 days in the dark. Following the incubation period, brains were transferred to scintillation vials containing sucrose cryoprotectant (30% (w/v) sucrose) for 72hrs at 4°C in the dark.
2.2.4 Golgi-Cox Brain Tissue Slicing
Following staining, brains were removed from the sucrose cryoprotectant and trimmed.
Brains were oriented longitudinally, with the lateral surface facing down, and molten agar (3%
(w/v) heated in diH2O for 60 seconds) was added to embed the brain. Agar blocks were then trimmed, leaving a ~0.5cm edge surrounding each side of the brain, and mounted on the vibratome stage (Leica VT1000s) with ethyl 2-cyanoacrylate glue. The vibratome stage area was filled with sucrose cryoprotectant to cover the agar block, and the brain/agar block was sliced into 400µm thick slices. Slices were transferred with a paintbrush into a well of a 6-well plate containing mesh inserts (3 slices per well), which were pre-filled with 6% sucrose solution (6%
(w/v) sucrose in 0.1M phosphate buffer). Brain slices were then incubated overnight at 4°C in the dark.
2.2.5 Golgi-Cox Brain Tissue Processing
Brain slices were transferred from the sucrose solution into a new 6-well plate filled with
2% paraformaldehyde solution (2% (w/v) paraformaldehyde in 0.1M phosphate buffer), and
25
incubated on a rocker (slow speed, 15 revolutions per minute, rpm) for 15 minutes. Slices were then washed twice in new wells filled with diH2O for 5 minutes per wash on the rocker
(moderate speed, 20 rpm). Slices were transferred to new wells filled with 2.7% ammonium hydroxide and left on the rocker (slow speed, 15 rpm) for 15 minutes, and then washed twice in new wells filled with diH2O for 5 minutes each on the rocker (moderate speed, 20 rpm). Next, slices were transferred into new wells pre-filled with Kodak Fixative A (diluted 10x from stock concentration) and incubated on the rocker (slow speed, 15 rpm) for 25 minutes, and then washed twice in new wells filled with diH2O for 5 minutes each on the rocker (moderate speed,
20 rpm).
2.2.6 Golgi-Cox Brain Tissue Mounting and Dehydrating
Following tissue processing, slices were transferred with a paintbrush onto Superfrost
Plus microscope slides. Excess agar was removed with forceps, and excess water dabbed off with
Kimwipes. Slices were left to air-dry on slides for either 45 minutes (geckos) or 60 minutes
(mice, chickens). Next, slices were dehydrated using increasing concentrations of ethanol
(EtOH) solutions in Coplin staining jars: 2 minutes each in 50%, 75%, and 95% EtOH, twice in
100% EtOH for 5 minutes each, followed by two rounds of clearing in Citrisolv solution for 5 minutes each. Slides were then coverslipped using Permount mounting solution, and air-dried horizontally in the dark for 5 days.
2.2.7 Microscopy and Imaging
The use of 400µm slices allowed whole neurons to be fully contained within the slices, while still being small enough to image clearly at different depths. Neurons were visualized with
26
brightfield microscopy, using a BX53 Olympus microscope (Olympus, Richmond Hill, Ontario,
Canada). Using the 30x silicone oil objective lens, high resolution image stacks of neurons taken
1 µm apart in the Z axis were then stitched together in 3-dimensions using Neurolucida software
(MBF Biosciences, Williston, Vermont, USA). Once stitched, select Purkinje cells were traced using Neurolucida software. The selection criteria included: (1) the soma was fully located within the Purkinje cell layer; (2) there was minimal overlap between adjacent neurons; (3) each neuron exhibited dendrite arborization characteristic of Purkinje cells; and (4) each neuron was fully contained within the slice (i.e. dendrites were not cut off from visual field). The goal was to trace five neurons from each of five brains, for a total of 25 neurons, for each species. However, among the geckos sample only eight (8) Purkinje cells total (from only three brains) met all the selection criteria.
2.2.8 Neuron Analysis and Statistics
To quantify neuromorphological complexity, Neurolucida Explorer software (version 10;
MBF Biosciences, Williston, Vermont, USA) was used to analyze neurons. Each individual neuron of each species was treated as an independent variable (gecko, n=8; mouse, n=25; chicken, n=25). Both Sholl and branched structure analyses were employed to quantify key aspects of neuron morphology. Sholl analyses were generated using concentric spheres originating from the soma distanced 15µm apart, and measured dendrite intersections, average dendrite diameter, and dendrite volume. Branched structure analyses were used to measure total neuron dendrite length, number of dendrite terminals, total dendrite volume, distance of the furthest dendrite terminal, average terminal distance, and the convex hull volume and surface area.
27
GraphPad Prism 8 software (GraphPad Software, San Diego, CA) was used to quantify and graphically represent variation within and between the species. Within species, branched structure analyses were generated using column statistics, and Sholl analyses were generated using regular two-way ANOVA to assess main effects between individual neurons and the distance from soma. Between species, branched structure statistical analyses were generated using one-way ANOVA, and Sholl analyses were generated using regular two-way ANOVA to assess main effects between species and the distance from soma. Groups were considered to be statistically different where p<0.05. All post-hoc multiple comparisons were analysed using the
Bonferroni method.
2.3 Results
2.3.1 Cerebellar Structure and Anatomy
The gross neuroanatomical appearance of the gecko cerebellum is strikingly different from that of the mouse and chicken. As for other lizards, the gecko cerebellum has a smooth (i.e., unfoliated) surface, and resembles a simple sheet or leaf-like feature tilted rostrally over the optic tectum (Figure 1a). In contrast, the cerebellum of each of the mouse and chicken is proportionally a larger component of the brain, roughly spherical in overall shape with a prominent series of transverse fissures, or folia (Figures 1b,c). However, despite these outward differences all three species shared the same trilaminar organization, with a distinctive Purkinje cell layer nested between the granular layer and molecular layer.
28
2.3.2 Purkinje Cell Neuromorphology: Qualitative Observations
As expected, Purkinje cells were easily identifiable in all species, characterised by complex two-dimensional dendrite arbors and large soma closely packed within the Purkinje cell layer. Soma were arranged as either a monolayer (mice, chickens) or 1-3 cells deep (geckos).
The dendrite arbors were composed of a stereotypical ramified arrangement of primary, secondary, and tertiary dendrites, along with many fine, spiny branchlets, projecting into the molecular layer (Figures 2, 3). Unlike the dendrites, Purkinje cell axons were not always visualized.
Although all Golgi-Cox stained Purkinje cells shared a similar neuromorphology, especially within species, no two neurons were identical (Figure 2). Within species, individual
Purkinje cells differed in terms of the arrangement of dendrites and the resulting geometry of the entire arbor. Between species, the main qualitative differences included variation in the absolute size of each neuron, the number of primary dendrites and the dendrite geometry. For example, while gecko and mouse Purkinje cells were comparable in size, those of chickens were significantly larger in absolute terms (Figure 3). In addition, while Purkinje cells of both geckos and chickens have a single primary dendrite that passes into the molecular layer, those of mice had one or two primary dendrites (see Figure 2, and Appendix II, Figure 2). Compared to mice, the dendrite arbor of gecko Purkinje cells had fewer secondary and tertiary dendrites, giving rise to a less ramified appearance. In chickens, the dendrite arbor is characterized by a distinctly parallel and comparatively straight palisade arrangement of terminal dendrites. Further, in chickens, but not geckos or mice, the majority of dendrites appeared to terminate within the same horizontal plane of the molecular layer.
29
2.3.3 Purkinje Cell Neuromorphology: Quantitative Observations
Intraspecific Variability: Sholl Analyses
Sholl analyses revealed that one or more of the traced Purkinje cells was significantly different in dendrite measurements within each species (Figure 4; Table 1). Among geckos, one of the eight traced Purkinje cells was significantly different in the number of intersections (two- way ANOVA, F7,119=2.91, p=0.0076; Bonferroni’s multiple comparisons test) and the total dendrite volume (two-way ANOVA, F7,126=6.47, p<0.0001, Bonferroni’s multiple comparisons test), while two were significantly different in the average dendrite diameter (two-way ANOVA,
F7,126=25.45, p<0.0001, Bonferroni’s multiple comparisons test) (Figures 4a-c).
In mice, two of the 25 Purkinje cells traced were significantly different in the number of intersections (two-way ANOVA, F24,408=3.65, p<0.0001, Bonferroni’s multiple comparisons test), one was significantly different in dendrite diameter (two-way ANOVA, F24,432=2.37, p=0.0003, Bonferroni’s multiple comparisons test), and three were significantly different in dendrite volume (two-way ANOVA, F24,432=5.05, p<0.0001, Bonferroni’s multiple comparisons test) (Figures 4d-f).
In chickens, nine of the 25 Purkinje cells traced were significantly different in the number of intersections (two-way ANOVA, F24,744=12.95, p<0.0001, Bonferroni’s multiple comparisons test), four were significantly different in dendrite diameter (two-way ANOVA, F24,768=4.71, p<0.0001, Bonferroni’s multiple comparisons test), and six were significantly different in dendrite volume (two-way ANOVA, F24,768=23.02, p<0.0001, Bonferroni’s multiple comparisons test) (Figures 4g-i).
30
Interspecific Variability: Branched Structure and Sholl Analyses
Overall, the largest Purkinje cells traced were those of chickens (Figure 3). Compared to both geckos and mice, chicken Purkinje cells had significantly greater values for almost every branched structure and convex hull analysis, including average terminal distance, furthest terminal distance, total dendrite length, total dendrite volume, convex hull surface area, and convex hull volume (Figure 5, see Table 2; p<0.0001 for all, one-way ANOVA). In contrast,
Purkinje cells in geckos and mice were closely comparable with almost no significant differences in any of the branched structure and convex hull analyses. One notable exception was the total number of dendrite terminals, which was significantly different between all three species (one- way ANOVA, p<0.0001). Geckos had significantly fewer terminals compared to chickens
(p=0.0036, Bonferroni’s multiple comparisons test), while mice had significantly more dendrite terminals than either geckos (p<0.0001; Bonferroni’s multiple comparisons test) or chickens (see
Figure 5; p=0.0220; Bonferroni’s multiple comparisons test).
Across the three species, there were also significant differences in Sholl analyses (Figure
6; two-way ANOVA, p<0.0001). More specifically, we found that chickens had significantly more intersections than geckos and mice farther from the soma (105µm-375µm), while mice had significantly more intersections than geckos and chickens closer to the soma (45µm-90µm). We found a similar trend in dendrite volume, with significantly differences between the three species
(two-way ANOVA, p<0.0001). The dendrite volume of chicken Purkinje cells was significantly greater than those of geckos and mice farther from the soma (105µm -295µm), while the dendrite volume of mouse Purkinje cells was significantly greater than those of geckos and chickens closer to the soma (60µm -75µm). Finally, there were also significant differences in dendrite diameter between all three species (two-way ANOVA, p<0.0001). In this case, the dendrite
31
diameter of chicken Purkinje cells was greater than that of both geckos and mice at most distances from the soma (15µm-420µm). Mice and geckos have comparable dendrite diameters, and all three species demonstrated a trend of decreasing diameters with increasing distance from the soma.
Purkinje cell dendrites project into the molecular layer. To determine if the thickness of the molecular layer correlated with the size differences observed between the different species, we measured the distance from the Purkinje cell layer (PCL) to the pial surface of the cerebellum
(Figure 7). Cerebella slices (n=7 for geckos, n=15 for both mice and chickens) were randomly selected and the distance from the PCL to the pial surface was measured using Neurolucida. We determined that the distance from the PCL to the pial surface was significantly greater in chickens than either geckos and mice (one-way ANOVA, p<0.0001 for both, means=383.4µm
[chickens], 271.3µm [geckos], and 162.6µm [mice]), and significantly greater in geckos than mice (one-way ANOVA, p<0.0001, Bonferroni’s multiple comparisons test).
2.4 Discussion
Purkinje cells are the sole output of the cerebellum, and play a key role in integrating thousands of neural inputs. While it has long been recognized that the neuromorphology of these neurons varies between species (e.g., Ramón y Cajal, 1899; Jacobs et al., 2014), with few exceptions (viz. laboratory rodents) detailed quantitative investigations are lacking. Here, I conducted a comparative investigation of Purkinje cell neuromorphology focusing on the leopard gecko, laboratory mouse, and domestic chicken. My findings demonstrate that there are both intraspecific and interspecific differences in the dendrite arbor. Whereas all Purkinje cells share a similar basic neuromorphology, with a relatively large tear-shaped soma and ramifying network
32
of dendrite branching, there are measurable and significant differences in branching pattern and absolute neuron size. Taken together, my findings reveal an unexpected level of neuromorphological heterogeneity for a single neuronal cell type both across and within species.
2.4.1 Intraspecific variation
Although the majority of neurons sampled within each species shared a similar neuromorphology, Purkinje cells are not a homogenous population. As revealed by Sholl analyses, I identified one (geckos) or more (mice and chickens) Purkinje cells that have a significantly different Sholl profile, including the number of Sholl dendrite intersections, volume, and diameter. Although intraspecific variation is not entirely unexpected, previous research has primarily documented differences between distinct regions of the cerebellum. For example, Purkinje cells from the anterior lobule of the mouse cerebellum often have a single primary dendrite, whereas those of the posterior lobule frequently have two (Nedelescu et al.,
2018). Similarly, in both mouse and humans, Purkinje cells with single primary dendrite are commonly found in the (convex) cerebellar apex, while Purkinje cells with two primary dendrites are characteristic of the (concave) cerebellar sulci (Kato et al., 1985; Nedelescu &
Abdelhack, 2013). Among the Purkinje cells I traced, only those of mice varied in the number of primary dendrites present, having either one or two. In geckos and chickens, all the traced
Purkinje cells had only a single primary dendrite. Despite this, my Sholl analyses revealed that
Purkinje cells in mice were more homogenous (with only 4-12% of Purkinje cells showing significant differences) than those of geckos (12.5-25%) and chickens (16-36%) (Table 1).
Although the functional implications of this variation remain poorly understood, it may relate to differences in excitatory input, from both parallel fibers and climbing fibers (Nishiyama &
33
Linden, 2004; Kidd, 2017; Zhang et al., 2018). Further, differences in dendrite diameter and length are known to impact neurophysiological functions, including signal propagation, action potential amplitude, and axial resistance (Vetter et al., 2000; Branco & Hausser, 2010; Forrest et al., 2012).
2.4.2 Purkinje cells differ in both size and complexity across species
While all Purkinje cells share a similar basic neuromorphology (large soma, complex dendrite arbor), there are obvious species-specific differences. In particular, I found that the dendrite arbors of chickens are almost twice as long as those of geckos and mice (both in terms of average and furthest terminal distances). In contrast, Purkinje cells of geckos and mice are comparable in size. Previous research has reported that the scaling relationship between Purkinje cell size and overall body size is logarithmic but not isometric (i.e., neuron size does not increase proportionally with body size; Friede, 1963). For example, Purkinje cells in domestic cattle are only twice as large as those of mice (Friede, 1963). Among galliform birds (including various relatives of domestic fowl such as quail, turkey, and pheasant), the number of Purkinje cells, along with the size of their soma, increases with cerebellar volume (Cunha et al., 2020). Whether the same holds true for the dendrite arbor remains unclear. As Purkinje cell dendrites project from the Purkinje cell layer into the molecular layer, the thickness of the latter is predicted to be a determinant of furthest Purkinje cell dendrite length (Friede, 1963). Among the species under consideration, the Purkinje cell layer-pial surface distance was longest in chickens, and shortest in mice. It is worth noting however, that gecko Purkinje cell dendrites tend not to contact the pial surface, suggesting that the Purkinje cell layer-pial surface distance does not have an effect on determining their overall cell size.
34
Sholl analyses found that neuromorphological complexity of Purkinje cells varied at different distances from the soma. The peak of Sholl measurements indicate where the most complex branching patterns lie (Nedelescu et al., 2017). Mice and geckos showed similar trends of Sholl intersections and volume, with both peaking around 90µm-105µm, while chicken
Purkinje cell intersections and volume peaked at ~180µm and ~135µm, respectively. This is consistent with previous documentations of Sholl intersections of various cerebellar neurons across different species, which, on average, peaked at around 100µm from the soma (Jacobs et al., 2014). Sholl diameter of geckos and mice are similar, starting at a maximum diameter of
~3.2µm on average from the soma and steadily decreasing at farther distances from the soma to
~0.24µm on average. Chickens, on the other hand, exhibited a greater starting average diameter of ~5.0µm from the soma and thinning out to an average of ~0.10µm. This regional difference in neuron complexity may also result in different Purkinje cell innervation and function, where climbing fibers and parallel fibers have greater chances of making synaptic connections at regions that are dendrite-rich (Vetter et al., 2000; Kidd, 2017).
2.4.3 Differences in Purkinje cell neuromorphology across species
Comparative studies are the benchmark for exploring organismal diversity. My investigation includes representative members of the three largest clades of amniote vertebrates: squamate reptiles (10,851 species; Uetz et al., 2020), mammals (6,410 species; Burgin et al.,
2019), and birds (18,043 species; Barrowclough et al., 2016). These groups diverged from one- another early during amniote evolution, with the ancestor of modern mammals splitting from the ancestor of modern reptiles (including birds) ~311.9 million years ago, while the lineages leading to modern geckos and chickens diverged ~279.7 million years ago (Kumar et al., 2017).
35
However, despite these lengthy divergence times, it is clear that the neuromorphological diversity of the cerebellum in general, and Purkinje cells in particular, does not strictly correlate with phylogeny (Sultan & Glickstein, 2007; Jacobs et al., 2014; Hibi et al., 2017; Macrì et al.,
2019). For example, foliation of the cerebellum has evolved independently in mammals, birds, and some chondrichthyan and teleost fish species (Yopak et al., 2017). Similarly, studies of numerous vertebrate groups have shown that an increase in the size of the cerebellum (often as a result of foliation) is associated with increasing complexity of locomotion, habitat use, and prey capture, even among closely related species (see Sultan & Glickstein, 2007; Yopak et al., 2017).
By extension, I predict that variation in the neuromorphology of Purkinje cells (as the sole motor output of the cerebellum) may also correlate with differences in behaviour and the functional demands of each species (Eccles, 1969; Bell, 2002). For example, while both geckos and mice are terrestrial quadrupeds of comparable body size, mouse Purkinje cells are quantitatively more complex in terms of the number of dendrite terminals. It is tempting to speculate that this difference relates to species-level variation in fine motor control, tactile sensory systems (mice have whiskers [vibrissae], geckos have scale-based somatosensory organs called sensilla), or other forms of cerebellar-mediated behaviour. Future investigations targeting this form-function relationship are necessary in order to explore this hypothesis.
One of my most notable findings was the large and distinctive neuromorphology of the chicken Purkinje cell. The chicken Purkinje cell was the subject of Cajal’s first Golgi- impregnated drawings (see Fig. 1.1), but the unusual palisade arrangement of terminal dendrites appears to have been overlooked. A similar neuromorphology has also been reported for Purkinje cells in some mormyrid fish (Kidd, 2017) and humpback whales (Jacob et al., 2014). Previous authors had suggested that this dendrite arrangement may relate to the use of acoustic and
36
electrosensory systems in an aquatic environment (Jacobs et al., 2014). My findings call this prediction into question, and underscore the need for a more exhaustive comparative sampling of vertebrate species.
37
Chapter 2 Tables
Table 1. Outlier Purkinje cells for intraspecific Sholl intersections, diameter, and volume analyses. Numbers were arbitrarily assigned to cells within each species. Each neuron differed from others within each species at all distance from the soma. Refer to Appendix II,
Supplementary Figures 1-3 for the complete collection of Purkinje cell tracings.
Table 2. Inter-species branched structure analyses of geckos (n=8), mice (n=25), and chickens
(n=25). Values are reported as mean±SEM (minimum value-maximum value).
38
Chapter 2 Figures
a)
b)
c)
Figure 1. Variation in gross morphology of Golgi-Cox stained cerebellum slices.
Longitudinal images of Golgi-Cox stained cerebellum slices of a gecko (a), mouse (b), and chicken (c). Whereas the gecko cerebellum has a smooth surface, that of the mouse and chicken is heavily foliated. Scale bar = 0.1mm.
39
Gecko Purkinje Cells
Mouse Purkinje cells
Chicken Purkinje Cells
Figure 2. Tracings of representative Purkinje cells of the leopard gecko, mouse, and chicken.
While all species shared similar morphological features of Purkinje cells, such as a large soma and distinct arborized branching of dendrites, mouse Purkinje cells were found to have either one or two primary dendrites, while those of geckos and chickens had only one. Smaller secondary and tertiary branches can be seen branching off the primary dendrites. Gecko Purkinje cells have terminal endings that are more asymmetrically distributed around the primary dendrites compared to the more uniform branching patterns of the mouse, and chicken Purkinje cells retain a somewhat palisade pattern of branching reminiscent to that of mormyrid teleosts.
40
Figure 3. Differences in neuron size present between species.
Tracings of chicken (left), gecko (top right), and mouse (bottom right) Purkinje cells, set to size, to illustrate the size differences between the three species. The Purkinje cells of the chicken were almost twice as large as that of the mouse and gecko, whereas the latter two are more comparable in overall size. Scale bar = 20µm.
41 a) b) c)
Gecko Sholl Intersections Gecko Sholl Diameter Gecko Sholl Volume 20 5 Neuron 2 1500 Neuron 3 Neuron 3 Neuron 3 4
15 ) 3 m)
µ 1000 m
3 µ 10 2 500 Volume ( Volume 5 Diameter ( 1 Number of Intersections 0 0 0 30 60 90 120 150 180 210 240 270 30 60 90 120 150 180 210 240 270 30 60 90 120 150 180 210 240 270 Distance from Soma (µm) Distance from Soma (µm) Distance from Soma (µm) d) e) f)
Mouse Sholl intersections Mouse Sholl Diameter Mouse Sholl Volume 2500 40 8 Neuron 7 Neuron 3 Neuron 3 2000 Neuron 24 Neuron 24 ) 30 6 3 m)
m Neuron 25 µ
µ 1500 20 4 1000 Volume ( Volume Diameter ( 2 10 500 Number of Intersections 0 0 0 30 60 90 120 150 180 210 240 270 30 60 90 120 150 180 210 240 270 30 60 90 120 150 180 210 240 270 Distance from Soma (µm) Distance from Soma (µm) Distance from Soma (µm) g) h) Chicken Sholl Diameter i) Chicken Sholl Intersections Chicken Sholl Volume Neuron 2 15 80 Neuron 4 Neuron 14 4000 Neuron 2 Neuron 6 Neuron 15 Neuron 3
60 m) 10 Neuron 8 ) 3000 µ Neuron 19 3 Neuron 4 m
Neuron 20 µ Neuron 20 Neuron 5 40 2000 Neuron 21 5 Neuron 14
20 Neuron 22 Diameter ( Volume ( Volume 1000 Neuron 15 Neuron 23 Number of Intersections 0 Neuron 25 0 60 120 180 240 300 360 420 480 60 120 180 240 300 360 420 480 0 60 120 180 240 300 360 420 480 Distance from Soma (µm) Distance from Soma (µm) Distance from Soma (µm)
Figure 4. Outlier Purkinje cells in each species.
Sholl graphs of the number of intersections, diameter, and volume of outlier Purkinje cells within
all cells sampled in geckos (a-c), mice (d-f), and chickens (g-i). For each analysis, the outlier
cells with each species are denoted by coloured lines on the graph, and all other cells are in grey.
Outlier cells are significantly different from others at all distances from the cell body.
42
Average Terminal Distance Furthest Terminal Distance
✱✱✱ ✱✱✱ ✱✱✱ 400 600 ✱✱✱
300 m) m) 400 µ µ
200
200 Distance ( Distance ( 100
0 0 Geckos Mice Chickens Geckos Mice Chickens Total Dendrite Length Total Dendrite Volume ✱✱✱ 20000 ✱✱✱ 60000 ✱✱✱ ✱✱✱ m) µ
15000 ) 3 40000 m µ 10000
20000
5000 ( Volume Length of Dendrite ( 0 0 Geckos Mice Chickens Geckos Mice Chickens
Convex Hull Surface Area Convex Hull Volume ✱✱✱ ✱✱✱ 5000000 300000 ✱✱✱ ✱✱✱ )
2 4000000 ) m 3 µ
200000 m
µ 3000000
2000000 100000 Volume ( Volume 1000000 Surface Area (
0 0 Geckos Mice Chickens Geckos Mice Chickens
Number of Dendrite Terminals ✱✱ 250 ✱ ✱✱✱ 200
150 100
50 Number of Terminals 0 Geckos Mice Chickens
43
Figure 5. Inter-species branched structure analyses of Purkinje neurons between the gecko, mouse, and chicken.
Chicken Purkinje cells had significantly greater measurements than both the gecko and mouse for average terminal distance, furthest terminal distance, total dendrite length, total dendrite volume, convex hull surface area, and convex hull volume, whereas mice and geckos were closely comparable in these analyses with no significant differences. However, mice had the greatest number of dendrite terminals compared to both chickens and geckos, and geckos had fewer terminals compared to chickens. Significance is denoted with asterisks (***p<0.0001,
**p=0.0036, *p=0.0220; one-way ANOVA, Bonferroni’s multiple comparisons test).
44
Number of Intersections * 40 * Geckos Mice 30 Chickens
20
10
Number of Intersections 0 60 120 180 240 300 360 420 480
-10 Distance from Soma (µm)
Sholl Dendrite Volume
2000 * * Geckos Mice
) 1500 3 Chickens m µ 1000
500
Dendritic Volume ( 0
60 120 180 240 300 360 420 480 -500 Distance from Soma (µm)
Sholl Average Diameter 8 * Geckos 6 Mice Chickens
m) µ 4
2
Diameter ( 0 60 120 180 240 300 360 420 480
-2 Distance from Soma (µm)
45
Figure 6. Inter-species Sholl analyses of Purkinje neurons.
Sholl analysis of dendrite intersections significantly differed between species (p<0.0001, two- way ANOVA), where chickens had a greater number of intersections than both geckos and mice farther from the soma (105µm-375µm), whereas mice had greater number of intersections than geckos closer to the soma (45µm-90µm). Dendrite volume was significantly different between all three species (p<0.0001, two-way ANOVA), where chickens had dendrite volumes greater than that of geckos and mice farther from the soma (105µm-295µm), but mice had greater dendrite volume than both chickens and geckos closer to the soma (60µm-75µm). Dendrite diameter was also significantly different between species (p<0.0001, two-way ANOVA). While all species demonstrated a trend of decreasing dendrite diameter farther from the soma, chickens had greater dendrite diameter than both geckos and mice at most distances from the soma (15µm-420µm), whereas the dendrite diameters of mice and geckos were more comparable overall. Significance where chickens have greater Sholl measurements than both geckos and mice are denoted with a black bar and asterisk, and where mice have greater Sholl measurements than both chickens and geckos are denoted with a grey bar and asterisk.
46
✱✱✱ 600 ✱✱✱ ✱✱✱
m) 400 µ
200 Distance (
0 Geckos Mice Chickens
Figure 7. Differences in distance from Purkinje cell layer to the pial surface between species.
Significant differences were found between all three species, where chickens had greater distance than geckos and mice (p<0.0001 for both, one-way ANOVA, Bonferroni’s multiple comparisons test), and geckos had greater distance than mice (p<0.0001, one-way ANOVA,
Bonferroni’s multiple comparisons test). Means were compared between groups
(geckos=271.3µm, mice=162.6µm, chickens=271.3µm).
47
CHAPTER 3
ELECTROPHYSIOLOGICAL PROPERTIES AND NEUROMORPHOLOGY OF
PURKINJE CELLS IN THE LEOPARD GECKO (EUBLEPHARIS MACULARIUS):
EVIDENCE FOR MULTIPLE CELL TYPES
3.1 Introduction
Purkinje cells are large inhibitory gamma-aminobutyric acid (GABA)-ergic neurons that serve as the sole output of the cerebellar cortex. Located between the molecular and granular layers of the cerebellar cortex, Purkinje cells are characterized by a relatively large soma, and a ramifying branching pattern of dendrite arborization. The complexity of the dendrite arbor underscores the role of Purkinje cells as central integrators of the cerebellar circuit. In mammals, dendrite arbors of each Purkinje cell receives as many as 100,000 excitatory synapses from adjacent parallel fibers (Nishiyama & Linden, 2004). As evidenced by the study of Purkinje cell connectivity and pathophysiology, these neurons are key participants in the regulation of fine motor movements, as well as various higher order cognitive and emotional functions (Smith et al., 1993; Hibi et al., 2017; Hoxha et al., 2018; Macrì et al., 2019; Locke et al., 2020; see also
Yopak et al., 2017).
Although Purkinje cells and their participation in the cerebellar circuit are widely recognized as being highly conserved across vertebrates (e.g., Ramón y Cajal, 1909; Eccles,
1969; Voogd & Glickstein, 1998; Yopak et al., 2017), these neurons vary in both form and function even within a species. At the level of neuromorphology, Purkinje cells may demonstrate variability in the number of primary dendrites (Uray & Gona, 1978; Kato et al., 1985; Nedelescu
& Abdelhack, 2013; Jacobs et al., 2014; Nedelescu et al., 2018) and in quantitative
48
measurements of the dendrite arbor (Nedelescu & Abdelhack, 2013; Nedelescu et al., 2018; see
Chapter 2). Purkinje cells also vary at the level of electrophysiological properties (Kim et al.,
2012; Zhou et al., 2014: Chang et al., 2020).
In response to stimulation, Purkinje cells fire one of two different types of action potentials: simple or complex spikes (Eccles et al., 1966; de Ruiter et al., 2006; Arancillo et al.,
2015; Otis, 2016). Simple spikes are generated both endogenously and in response to parallel fiber input, and are regulated by sodium channels (de Ruiter et al., 2006; Otis, 2016). In contrast, complex spikes are produced in response to a large excitatory input from climbing fibers originating from the inferior olivary nucleus and are regulated by calcium channels (Llinás &
Sugimori, 1980; Hounsgaard & Midtgaard, 1988; Knöpfel et al., 1990; Masoli et al., 2015).
Across species, these two types of action potentials contribute to different variants of firing patterns. For example, Purkinje cells in anterior lobules of the mouse cerebellum have two distinctive types of firing patterns (tonic firing and complex bursting), while those in posterior lobules have four (tonic firing, complex bursting, initial bursting, gap firing; Kim et al., 2012).
Emerging evidence suggests that, at least in mice, differences in Purkinje cell firing patterns may correlate with zebrin II (glycolytic enzyme aldolase c) expression (Zhou et al., 2014). Purkinje cells with different firing patterns have also been reported for various teleost fish, including the mormyrid Gnathonemus petersii (Han & Bell, 2003; Zhang & Han, 2007; Zhang et al., 2008;
Zhang et al., 2010), and the cyprinid Danio rerio (zebrafish; Chang et al., 2020). Interestingly, in teleosts, Purkinje cells with different firing patterns also demonstrate distinctive neuromorphologies, with at least three subtypes in G. petersii and four in D. rerio (Han & Bell,
2003; Zhang & Han, 2007; Zhang et al., 2008; Zhang et al., 2010; Chang et al., 2020). For most species however, form-function relationship of Purkinje cells remains poorly understood.
49
Here, I use whole-cell recordings and intracellular biocytin labeling to characterize the electrophysiological properties and neuromorphology of Purkinje cells from a representative squamate reptile, the leopard gecko (Eublepharis macularius). Previous research on the gecko has focused on characterizing the neuromorphology of Purkinje cells following the use of Golgi-
Cox staining (Chapter 2; see also Bradley et al., submitted). These data reveal that, at the level of the dendrite arbor, gecko Purkinje cells are structurally diverse. To better understand this diversity, I focused on characterizing the form and function of these neurons, combining basic electrophysiological properties with quantitative neuromorphology. I determined that there are at least two distinct types of Purkinje cells in adult geckos. Type 1 Purkinje cells are less excitable than type 2 cells, with significant differences in sag ratio as well as input resistance. Overall, type 1 cells are also larger in size, with greater dendrite volumes, lengths, and convex hull measurements. Sholl analyses revealed that type 1 cells had peaks of dendrite intersections, volume, and length that are farther from the soma, indicating variation in neuromorphological complexity in different cell regions. Taken together, our data indicate that functionally and structurally distinct Purkinje cell populations are present in leopard geckos, hinting at differences in the cerebellar microcircuitry.
3.2 Materials and Methods
3.2.1 Animal Care
Captive bred leopard geckos (E. macularius; hereafter ‘geckos’) were purchased from
Global Exotic Pets (Kitchener, Ontario). All geckos (n=3) were adults (all females; >1-year-old) at the at the time of data collection. Geckos were individually housed according to the protocols of McLean and Vickaryous (2011). Briefly, geckos were housed in a secure, temperature-
50
controlled environmental chamber at the Hagen Aqualab, University of Guelph (12:12h light cycle, with an average ambient temperature of 27.5°C). Each gecko was individually housed in an 18.9 L polycarbonate tank, with heating cables (Hagen Inc., Baie d’Urfe, Quebec, Canada) set to 32°C located underneath one side of the tank to establish a temperature gradient. Geckos were fed three larval mealworms (Tenebrio spp.) dusted with powdered calcium and vitamin D3
(cholecalciferol; Zoo Med Laboratories Inc., San Luis Obispo, California, USA) daily, with access to fresh drinking water at all times. All geckos were cared for in accordance with Animal
Utilization Protocol 3947, and the procedures outlined by the Canadian Council on Animal Care.
3.2.2 Tissue Collection and Preparation
Geckos were euthanized with a 150µL injection of Alfaxalone (85mg/kg) into the epaxial muscles of the neck, followed by decapitation. Brains were then rapidly extracted in cold oxygenated sucrose artificial cerebrospinal fluid (sACSF; 254mM sucrose, 10mM D-glucose,
26mM NaHCO3, 2mM CaCl2, 2mM MgSO4, 3mM KCl, and 1.25mM NaH2PO4, pH 7.4).
Cerebella were then dissected out, mounted longitudinally on the vibratome stage (Leica
VT1200 vibrating microtome; Leica Microsystems, Richmond Hill, ON, Canada), and sliced into
400µm thick slices. Cerebellar slices were then transferred into a recovery chamber containing oxygenated ACSF maintained at 30°C (128mM NaCl, 10mM D-glucose, 26mM NaHCO3, 2mM
CaCl2, 2mM MgSO4, 3mM KCl, and 1.25mM NaH2PO4). Slices were allowed to recover for at least 2 hours before whole-cell electrophysiology. Both sACSF and ACSF were continually oxygenated with carbogen (95% O2, 5% CO2) throughout the experiment.
51
3.2.3 Electrophysiology
To perform whole-cell electrophysiology, brain slices were transferred into a recording chamber (Warner Instruments, Hamden, CT) mounted on a stage that was attached to an
Axioskop FS2 microscope (Carl Zeiss Canada, Toronto, Ontario). Experiments were performed at room temperature, and oxygenated ACSF was perfused over the slices at a rate of ~5mL/min.
Borosilicate pipette electrodes (outer diameter 1.5mm, 3-5 MW internal resistance; Sutter
Instrument Company, Novato, CA) containing 120 mM K-gluconate, 5mM KCl, 2mM MgCl2,
4mM K2-ATP, 400µM Na2-GTP, 10mM Na2-phosphocreatine, 10mM HEPES buffer (pH 7.3, adjusted with KOH), and 0.3% (w/v) biocytin (Tocris Bioscience) were used. Using infrared differential interference contrast microscopy, Purkinje neurons were located by finding the
Purkinje cell layer and identifying the large somas with a prominent primary dendrite. The following basic passive and active electrophysiological properties were recorded in current clamp mode in response to positive and negative current steps: resting membrane potential, input resistance, afterhyperpolarization, input/output, and rheobase. Membrane resistance was calculated by injecting a hyperpolarizing current and measuring the steady-state deflection in membrane potential. Sag ratio was calculated as the difference between the minimum value and the steady state decrease in voltage divided by the peak voltage deflection during hyperpolarizing current steps. All recordings were made using a Multiclamp 700B amplifier, signals were acquired at 20 kHz and low-pass filtered at 2 kHz using a Digidata 1440A data acquisition system (Molecular Devices, Sunnyvale, CA). All neurons were corrected for liquid junction potential prior to recording.
52
3.2.4 Neuron Morphology
Purkinje cells were able to be visualized using an adapted protocol by Chung et al.
(2016). Neurons were recorded with pipettes containing patch solution with 0.3% (w/v) biocytin, held for electrophysiological recording for at least 30 minutes to allow the biocytin to diffuse throughout the neuron, and the pipette withdrawn carefully to allow the cell membrane to re-seal.
Neurons were recorded at intervals throughout the length of the brain slice to reduce chances of overlap. Brain slices were then fixed in 4% (w/v) paraformaldehyde in 0.1M phosphate buffer solution for up to 3 weeks at 4°C. Slices were then washed 3 times for 10 minutes each on the rocker (fast speed, 25 rpm) using Tris-buffered saline (0.1M Tris, 0.15M NaCl, pH 7.5). Next, slices were incubated with 0.5% (v/v) H2O2 solution (0.5% (v/v) H2O2 and 0.25% (v/v) Triton X-
100 in TBS) on the rocker for 15 minutes (slow speed, 15rpm) to supress endogenous peroxidase activity. Slices were then washed 3 times for 10 minutes each on the rocker (fast speed, 25rpm) using TBS, and incubated with Vectastain Elite ABC solution (Vector Labs, Burlington, Ontario) containing 0.25% (v/v) Triton X-100, 0.25% (v/v) ABC Elite Reagent A, and 0.25% (v/v) ABC
Elite Reagent B in TBS at room temperature for 24 hours on the rocker (slow speed, 15rpm).
Slices were washed 3 times for 10 minutes each on the rocker (fast speed, 25rpm) using TBS, and then incubated in a solution containing 0.05% (w/v) 3,3’-diaminobenzidine tetrahydrochloride hydrate (DAB), 0.04% (w/v) nickel chloride hexahydrate, and 0.25% (v/v)
Triton X-100 in TBS for 5 minutes on the rocker (slow speed, 15rpm). The solution was then replaced with the same solution with the addition of 0.001% (v/v) H2O2 and incubated for 15 minutes on the rocker (slow speed, 15rpm). Slices were then washed 3 times for 10 minutes each in TBS on the rocker (fast speed, 25rpm). Slices were mounted onto Superfrost Plus microscope slides, and coverslipped using Dako mounting solution (DAKO, Glostrup, Denmark).
53
3.2.5 Microscopy and Imaging
Neurons were visualized using brightfield microscopy with a BX53 Olympus microscope
(Olympus, Richmond Hill, Ontario, Canada), using the 30X objective silicon oil immersion lens and Neurolucida software (version 10; MBF Biosciences, Williston, Vermont, USA). High resolution image stacks of neurons taken 1µm apart in the Z-axis were stitched together in 3- dimensions, and then traced using Neurolucida software. Neuron inclusion criteria were as follows: (1) the neuron was fully visible (i.e. Biocytin dye was properly diffused throughout the neuron); (2) each neuron exhibited dendrite arborization characteristic of Purkinje cells; (3) each neuron was fully contained within the slice (i.e. dendrites were not cut off from visual field); and
(4) there must be viable electrophysiological recordings for each neuron. A total of 8 neurons were recorded from each of three geckos, with a total neuron count of 24. However, in one gecko, only 3 neurons met all the selection criteria, resulting in a final traced neuron count of 19.
3.2.6 Neuron and Statistical Analyses
Neurolucida Explorer software (MBF Biosciences, Williston, Vermont, USA) was used to analyze neurons, with each individual neuron treated as an independent variable. Sholl and branched structure analyses were used to quantify key aspects of neuron morphology. Sholl analyses were generated using concentric spheres originating from the soma distanced 15µm apart, measuring dendrite intersections, volume, length, and diameter. Branched structure analyses were used to measure total dendrite length, number of dendrite terminals, total dendrite volume, distance of the furthest dendrite terminal, and the convex hull volume and surface area.
54
GraphPad Prism 8 software (GraphPad Software, San Diego, CA) was used to quantify and graphically represent variation between neurons. Electrophysiology data was analysed using unpaired t-tests, with the exception of afterhyperpolarization and input/output, which were analyzed using regular two-way ANOVA. Sholl analyses were generated using regular two-way
ANOVA to assess main effects between the two groups and distance from soma, and branched structure analyses were generated using unpaired t-tests. Groups were considered to be statistically different where p<0.05. All post-hoc multiple comparisons were analysed using the
Bonferroni method.
3.3 Results
3.3.1 Electrophysiological Properties
The whole-cell recording technique was used to determine the electrophysiological properties of gecko Purkinje cells. It was found that while some Purkinje cells exhibited spontaneous firing activity, others were less consistent. The resting membrane potential (RMP) was then recorded for all cells (Figure 1a, Table 1; all measurements reported as mean ± SEM).
The RMPs for all the Purkinje cells sampled ranged from -70.68mV to -37.14mV, with a mean of -57.55 ± 2.09mV. Among Purkinje cells that exhibited spontaneous firing activity (hereafter, type 1 cells), the RMP was consistently greater than -55mV (n=15; mean RMP -63.72 ±
1.18mV). In contrast, among Purkinje cells that exhibited less spontaneous firing activity
(hereafter, type 2 cells), the RMP was consistently less than -55mV (n=8; mean RMP -45.96 ±
2.23mV). Both types of neurons were evenly distributed throughout the Purkinje cell layer of all three gecko cerebella sampled, and neither type was specific to a particular location within the cerebellum (see Appendix III).
55
Next, the electrophysiological characterization of the cells was expanded to include their rheobase, input/output, afterhyperpolarization (AHP), input resistance, and sag ratio. The minimum amount of current needed to reach the depolarization threshold of the neuron from the
RMP, the rheobase, for all the sampled Purkinje cells ranged from 1pA to 400pA (mean=75.27 ±
23.85pA), and was not significantly different between type 1 cells (mean=93 ± 33.74pA) and type 2 cells (mean=37.29 ± 14.53pA) (Figure 1b). The input/output curve for both Purkinje cell types were generated by injecting cells with increasing current (50pA to 500pA), and quantifying the number of action potentials fired as a measure of excitability (Figure 1c). As for the rheobase, the input/output curves were not significantly different between type 1 and type 2 cells
(Figure 1c). Subsequently, current pulses were injected (in a doubling sequence starting from 1 to 32) to determine the AHP. AHP is a post-spike feedback mechanism that helps define the refractory period (Dumenieu et al., 2015), and did not differ significantly between the two
Purkinje cell types (Figure 1d). Next, the current flow and cell excitability was quantified as measured by input resistance. The input resistance for all the Purkinje cells sampled ranged from
71.61MW to 514.1MW (mean=279.6MW), with type 2 cells (mean=378.7 ± 52.83MW) having a significantly greater input resistance than type 1 cells (mean=226.8 ± 22.96MW; unpaired t-test, p=0.0057) (Figure 1e). The sag ratio was quantified as a measure of intrinsic excitability calculated by comparing the difference between the maximum decrease in voltage and the steady state voltage following a hyperpolarizing current step. The sag ratio for all Purkinje cells sampled ranged from -11.14mV to 1.24mV (mean=-3.15 ± 0.61mV), and was significantly different between the two cell types (type 1cells mean=-4.13 ± 0.79mV; type 2 cells mean=-1.35
± 0.55mV; unpaired t-test, p=0.0258) (Figure 1f).
56
In response to a series of depolarizing current injections, type 1 and type 2 Purkinje cells demonstrated different action potential firing patterns. Type 1 cells were found to have three distinct firing patterns, while type 2 cells typically exhibited a fourth firing pattern not observed in type 1 cells (Figure 2). At 200pA, type 1 cells fired both small and large narrow spikes, as well as small, broad spikes. The three firing patterns include: tonic firing, the continuous firing of large and/or smaller narrow spikes or small broad spikes (Figure 2a; n=5 cells); initial firing,
1-3 initial large narrow spikes followed by 0-5 small broad spikes near the end of current injection (Figure 2b;n=7 cells); and complex/phasic firing, large narrow spikes that decrease to small narrow spikes and transition back into large narrow spikes (Figure 2c; n=1 cell). Two cells with an RMP of greater than -55mV did not display any spikes with current injection. Compared to type 1 cells, the firing pattern of type 2 cells was more erratic and consisted primarily of small spikes, both narrow and wide (Figure 2d). Although not identifiable as a conventional firing pattern, it is worth noting that all but one (out of eight) type 2 cells had this same irregular firing of action potentials; the eighth cell with an RMP of less than -55mV did not display any spikes with current injection.
3.3.2 Neuromorphology
A neuromorphological investigation was conducted to explore the variation of Purkinje cell morphology, and to determine if the differences in function (i.e., electrophysiological properties) correlated with differences in form. To begin, biocytin was diffused into each
Purkinje cell following the electrophysiological recording. Each neuron was then traced (using
Neurolucida), and the resulting dendrite arbor quantified using branched structure and Sholl analyses (Figures 4 and 5; see also Table 2). All gecko Purkinje cells are readily identifiable
57
based on their large, oval soma, single primary dendrite, and ramifying dendrite arborizations extending into the molecular layer (Figure 3; see also Appendix II, Figure 4). With one exception
(convex hull volume), the two Purkinje cell types had comparable branched structure profiles; i.e., there were no significant differences between type 1 and type 2 Purkinje cells with respect to the total number of dendrite terminals, furthest terminal distance, total dendrite length, total dendrite volume, and convex hull surface area (Figure 4). The convex hull volume was significantly greater in type 1 cells compared to type 2 cells (unpaired t-test, p=0.0483).
Next, Sholl analyses were used to investigate the branching pattern and geometry of the dendrite arbor. Both Purkinje cell types have distinct Sholl profiles, with significant differences in each of the: number of dendrite intersections (two-way ANOVA, p=0.0058); dendrite volume
(two-way ANOVA, p<0.0001); dendrite diameter (two-way ANOVA, p<0.0001); and dendrite length (two-way ANOVA, p=0.0335). While there were no significant differences between the two Purkinje cell types at specific distances away from the soma, type 1 cells typically have longer dendrites, with more intersections, and a greater dendrite volume at distances farther from the soma than type 2 cells due to their characteristically longer dendrite projections (Figure 5).
For example, dendrite length and the number of intersections peaked at ~260µm from the soma for type 1 cells, and ~150µm for type 2 cells. Similarly, the maximum Sholl dendrite volume was observed at ~120µm from the soma for type 1 cells, and ~105µm for type 2 cells. Type 1 cells also had a greater Sholl dendrite diameter beyond 330µm away from the soma.
3.4 Discussion
This study sought to quantitatively characterize electrophysiological and neuromorphological properties of Purkinje cells in a squamate reptile. Using whole-cell
58
recordings and biocytin labeling, I determined that, similar to mice, mormyrid teleosts, and zebrafish (Kim et al., 2012; Zhang & Han, 2007; Chang et al., 2020), Purkinje cells in leopard gecko are not a homogenous population. As Purkinje cells represent the sole output of the cerebellum, characterization of these cells in the leopard gecko provides important insight into the functional circuitry of reptilian cerebellum, and evolution of this brain region across vertebrates.
In leopard geckos, there are at least two distinct types of Purkinje cells that differ both in electrophysiological properties (including their ability to fire spontaneously, along with their action potential firing patterns, RMP, and input resistance) and neuromorphology (particularly their Sholl profiles). It is worth noting however, that all the values for the basic electrophysiological properties of both types of leopard gecko Purkinje cells were found to be within range of reported values for other species in literature (Llinás & Sugimori, 1980; Storm et al., 1998; Netzeband et al., 1999; McKay & Turner, 2005; Kim et al., 2012). Compared with type
2 Purkinje cells, type 1 cells were less excitable, with a significantly lower input resistance and significantly greater sag ratio. While not significant, type 1 cells also had a greater rheobase than type 2 cells, indicating that type 1 cells required a higher current to initiate an action potential than type 2 cells, and a larger AHP. AHP participates as a feedback mechanism that regulates neuronal firing, and is often activated by SK channels (Kaffashian et al., 2011; Ohtsuki &
Hansel, 2018; Titley et al., 2020). One possible explanation for the differences in AHP amplitude between the two Purkinje cell types is variation in the expression of SK channels.
While input-output curves between the cell types were also not significantly different, type 1 cells generated more action potentials than type 2 cells, and both cell types had decreased action potential frequency with increasing current injections. This is in contrast to mice and rats,
59
where input-output curves have a more linear relationship between action potential frequency and injected current (Khaliq & Raman, 2006; Walter & Khodakhah, 2006; Yang et al., 2017).
This decrease in frequency may indicate that gecko Purkinje cells are able to undergo depolarization block with greater injected current, and that this population of cells may be more intrinsically excitable compared to Purkinje cells of other species (Bianchi et al., 2012).
3.4.1 Gecko Purkinje cells types have distinct action potential firing patterns
Similar to other species, Purkinje cells in geckos have multiple distinct patterns of action potentials. In mice, these firing patterns are associated with different regions (lobules) of the cerebellum: tonic firing and complex bursting cells are found in anterior lobules, while tonic firing, complex bursting, initial bursting, and gap firing are found in the posterior lobule (Kim et al., 2012). In the mormyrid fish Gnathonemus petersii, different firing patterns are associated with different Purkinje cell types. Multipolar and fan shaped Purkinje cells exhibit tonic firing, and large, narrow spikes, while small Purkinje cells exhibit tonic firing, small narrow spikes, and large broad spikes (Han & Bell, 2003; Zhang & Han, 2007). As demonstrated by the use of calcium and sodium channel blockers (i.e. cadmium and tetrodotoxin, respectively), small narrow are sodium spikes, while large broad are calcium spikes (Han & Bell, 2003; Zhang &
Han, 2007; Zhang et al., 2010). Type 1 leopard gecko cells have small and large narrow spikes and small broad spikes during tonic firing, large narrow spikes and small broad spikes during initial firing cells, and large and small narrow spikes during complex/phasic firing. In contrast, the type 2 cells only exhibited small narrow and wide spikes with no discernable firing patterns.
These differences in spike shape and firing patterns may be due to differences in ion channel regulation (type 2 cells seem to exhibit more small sodium spikes compared to type 1 cells), and
60
future investigations using pharmacological manipulations (such as ion channel blockers) are needed to confirm these findings. However, unlike mice and mormyrid fish, there are no regional differences in the distribution of type 1 and 2 cells (both types appear to be randomly distributed throughout the cerebellum) and there is no evidence of regional-dependent electrophysiological compartmentalization.
3.4.2 Neuromorphological variation is present between subtypes
In addition to being less excitable than type 2 Purkinje cells, type 1 cells have larger and more complex dendrite arbors. Sholl analyses also reveal that within type 1 cells, the peak measurements for Sholl length, intersections, and volume are shifted further away from the soma compared to type 2 cells. In addition, as revealed by the tracings of reconstructed cells, type 1 cells are typically longer and narrower than type 2 cells. The type 1 neuromorphology most closely resembles that of alligators, with the dendrites passing well into the molecular layer
(Llinás, 1969). In contrast, the neuromorphology of some type 2 cell tracings appears to conform to the dendrite arrangement of Purkinje cells from mice, especially those of the sulcus and posterior lobules (see Nedelescu & Abdelhack, 2013; Nedelescu et al., 2018). At this time, I can only speculate about the functional importance of having multiple Purkinje cell types in geckos.
Because the dendrite arbors occupy different areas within the molecular layer, and that these morphologies are associated with different electrophysiological properties, it is tempting to conclude that localized cellular heterogeneity replaces the need for regional compartmentalization. Future investigations targeting the roles of each cell type during cerebellar-mediated activities (e.g., motor coordination), may help unravel this issue.
61
Chapter 3 Tables
Table 1. Electrophysiological properties of type 1 and type 2 leopard gecko Purkinje cells.
Values are reported as mean±SEM, with significantly different values between types reported as p-values as determined using unpaired T-tests.
Property Type 1 Type 2 Significance RMP (mV) -63.72±1.176 -45.96±2.228 p<0.0001
Input Resistance (MW) 226.8±22.96 378.7±52.83 p=0.0057 Rheobase (pA) 93±33.74 37.29±14.53 n.s. Sag Ratio (mV) -4.129±0.7883 -1.346±0.5495 p=0.0258
Table 2. Branched structure analyses of type 1 and type 2 leopard gecko Purkinje cells. Values are reported as mean±SEM, with significantly different values between types reported as p- values as determined using unpaired T-tests.
Analysis Type 1 Type 2 Significance Number of Terminals (µm) 141.3±19.01 144.1±18.94 n.s. Furthest Terminal Distance (µm) 506.4±24.97 432.4±27.47 n.s. Total Dendrite Length (µm) 8436±1074 7160±654.8 n.s. Total Dendrite Volume (µm³) 13830±3704 8142±947.5 n.s. Convex Hull Volume (µm³) 1498272±250259 844766±98266 p=0.0483 Convex Hull Surface Area (µm²) 136711±12890 118359±5402 n.s.
62
Chapter 3 Figures
a) b) Resting Membrane Potential Rheobase 0 500
400 -20 300 -40 200 RMP (mV)
-60 Rheobase (pA) 100
-80 **** 0 Type 1 Type 2 Type 1 Type 2
c) d) Input/Output Afterhyperpolarization 40 5 Type 1 Type 1 Type 2 0 30 1 2 4 8 16 32 Type 2 -5 20 -10 10
AP Frequency (Hz) -15 Number of Current Pulses 0 Afterhyperpolarization (mV) -20 50 100 150 200 250 300 350 400 450 500 Current Injected (pA) e) f) Input Resistance Sag Ratio 600 ✱✱ 5 ) Ω 0 400
-5
200
Sag Ratio (mV) -10 Input Resistance (M 0 -15 * Type 1 Type 2 Type 1 Type 2
63
Figure 1. Differences in electrophysiological properties of type 1 and type 2 leopard gecko
Purkinje cells.
Type 1 cells had significantly higher resting membrane potentials (a). While not significantly different, type 1 cells had greater rheobase (b) than type 2 cells, indicating that more current is needed to elicit an action potential in this subtype. Although input-output curves were also not significantly different (c), type 1 cells generated greater number of action potentials than type 2 cells, and both subtypes had decreased action potential frequency with greater currents injected.
Type 1 cells had larger afterhyperpolarization than type 2 cells with greater current pulses injected (d), but the difference was not found to be significantly different between groups. Type
1 cells had significantly smaller input resistance (e) and greater sag ratio (f), indicating that these cells may be more excitable compared to type 2 cells. Significance is denoted with asterisks
(****p<0.0001, **p=0.0057, *p=0.0258; unpaired T-test).
64
1 2 43 1 2 43 40
40
20 20 a) 0 b) 0
-20 -20 (mV) (mV) Vm_primar Vm_primar
-40 -40
-60 -60
-80 -80
-100 -100 200 1 2 43 200
0
(pA) 40 0 Im_sec (pA) Im_sec 1 2 43 -200 -200 0.2 c)0.4 0.6 0.8 1 40 1.2 1.4 1.6 0.2 d) 0.4 0.6 0.8 1 1.2 1.4 1.6 20 Time (s) Sw eep:1 Visible:1 of 4 Time (s) Sw eep:1 Visible:1 of 5
0 20
-20 0 (mV) Vm_primar
-40 -20 (mV) Vm_primar
-60
-40
-80
-60
-100 200 -80
0 (pA) Im_sec
-100
-200 Figure 2. Different action potential200 patterns of type 1 and type 2 leopard gecko Purkinje 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Time (s) Sw eep:1 Visible:1 of 6
(pA) 0 Im_sec
cells. -200 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Time (s) Sw eep:1 Visible:1 of 6 All action potentials here were recorded at 200pA of current injected. Type 1 cells exhibited 3
different firing patterns: tonic firing (a), initial firing (b), and complex/phasic firing (c). In type 2
cells, there is a distinct firing pattern unseen in type 1 cells that is more erratic and irregular
compared to those of type 1 cells (d). Type 1 cells fired both small and large narrow spikes,
whereas type 2 cells fired only small spikes, both broad and narrow.
65
Type 1
Type 2
Figure 3. Tracings of representative type 1 and type 2 leopard gecko Purkinje cells.
All cells exhibited classic Purkinje cell characteristics, such as a large, oval soma, a single primary dendrite, and complex networks of ramifying dendrite arborizations. While there are no distinct morphological differences between subtypes, type 1 cells were on average longer and had the majority of their dendrites clustered farther from the soma, as compared to type 2 cells which had smaller arbors closer to the soma (the differences of which can also be reflected in
Sholl analyses).
66
Number of Terminals Furthest Terminal Distance
400 800
300 600 m) µ
200 400
100 Distance ( 200 Number of Terminals 0 0 Type 1 Type 2 Type 1 Type 2
Total Dendrite Length Total Dendrite Volume
20000 60000 )
15000 3 40000 m m) µ µ 10000
20000 Length ( 5000 ( Volume
0 0 Type 1 Type 2 Type 1 Type 2
Convex Hull Volume Convex Hull Surface Area
4000000 * 250000 ) 2 200000 ) 3000000 m 3 µ m
µ 150000 2000000 100000
Volume ( Volume 1000000 50000 Surface Area (
0 0 Type 1 Type 2 Type 1 Type 2
67
Figure 4. Branched structure analyses of type 1 and type 2 leopard gecko Purkinje cells.
While comparable in number of dendrite terminals, furthest terminal distance, total dendrite length, and convex hull surface area, type 1 cells still had greater values for each analysis, indicating similar degrees of complexity in dendrite arborization, but larger cell size in type 1 cells. Type 1 cells also had larger total dendrite volume and significantly greater convex hull volume compared to type 2 cells, further supporting the larger cell size in type 1 cells.
Significance is denoted by asterisk (*p=0.0483; unpaired T-test).
68
Sholl Dendrite Length Number of Intersections 40 800 Type 1 Type 1 Type 2 30 600 Type 2 m)
µ 400 20
200 10 Length ( 0 0
45 Number of Intersections 45 105 165 225 285 345 405 465 525 105 165 225 285 345 405 465 525 -200 -10 Distance from Soma (µm) Distance from Soma (µm)
Sholl Dendrite Volume Sholl Dendrite Diameter 5 1500 Type 1 Type 1 4 Type 2 Type 2 m) 1000 µ 3 ) 3 m
µ 2 500 Diameter ( 1 Volume ( Volume 0 0
45 45 105 165 225 285 345 405 465 525 105 165 225 285 345 405 465 525 -500 Distance from Soma (µm) Distance from Soma (µm)
Figure 5. Sholl analyses of type 1 and type 2 leopard gecko Purkinje cells.
Type 1 and type 2 cells are significantly different in dendrite length (two-way ANOVA, p=0.0335), number of dendrite intersections (two-way ANOVA, p=0.0058), dendrite volume
(two-way ANOVA, p<0.0001), and dendrite diameter (two-way ANOVA, p<0.0001). While significance does not lie at specific distances from the soma, type 1 cells seem to have peaks of length, intersections, and volume shifted to the right, such that most of the dendrite arborization is clustered farther from the soma compared to type 2 cells. Sholl diameter also shows that type 1 cells have greater dendrite diameter farther from the soma than type 2 cells.
69
CHAPTER FOUR
CONCLUDING STATEMENTS
This study sought to characterize Purkinje cell heterogeneity both within and between species. Using Golgi-Cox staining, I determined that there were significant differences in dendrite neuromorphology both within and between three distantly-related species: the leopard gecko (Eublepharis macularius; a squamate reptile), laboratory mouse (Mus musculus: a rodent mammal), and domestic chicken (Gallus gallus domesticus; a galliform bird). While Purkinje cells in geckos and mice were comparable in terms of overall size, they differed in various measure of dendrite complexity. Purkinje cells in chickens were almost twice as large as those of mice and geckos, including a greater dendrite volume, diameter, and length. The structure- function relationship of leopard gecko Purkinje cells was further explored using whole-cell electrophysiology, and intracellular biocytin injection. Based on details of neuromorphology, firing patterns, and intrinsic cellular properties, I identified two distinct subtypes of gecko
Purkinje cells. Taken together, my findings reveal that Purkinje cells are a heterogeneous neuron type both between and within species.
4.1 Neuromorphological variation of Purkinje cells
The intraspecific differences seen in this study highlights the diversity of Purkinje cell neuromorphology seen within a single species. Most of the outlier Purkinje cells for each species stood out in more than one Sholl analysis, indicating neuromorphological variation that is present at different regions of the cell arborization as well. However, while the outlier cells provide regional differences in neuromorphology, overall cell size within species was quite similar to each other. This is in contrast to one of the most striking interspecific differences in Purkinje cell
70
size found here, which was that chicken Purkinje cells were almost twice as large as those of mice and geckos. It has been previously observed that while Purkinje cell size does increase with body size, the extent of increase is not directly proportional to body size (Friede, 1963). This can be seen when comparing the Purkinje cell sizes of chickens (brain mass=3g) in this study to those of humans (brain mass=1435g), which were approximately three times as large in dendrite arborization diameter, and to humpback whales (brain mass=3606g), which were approximately twice as large (Jacobs, 2014). In addition to the size of the neuron, I also determined there were distinct differences in the overall complexity and pattern of dendrite branching. For example, while both gecko and mouse Purkinje cells showed a similar pattern of radial branching, consisting of secondary and tertiary branches that radiate away from the soma. In contrast,
Purkinje cells in chickens have a palisade pattern of relatively straight and vertically oriented tertiary branches, similar to those reported for mormyrid fish and whale species (Jacob et al.,
2014; Kidd, 2017).
These findings provide interesting insights into the differences in Purkinje cell structure in the context of evolutionary trends and ecological behaviour, where geckos have the least complex Purkinje cells compared to mice and chickens. The role of the cerebellum has been implicated in avian species for functions related to flight, such that several regions of the cerebellum are activated during flight which coincide with areas connecting to optic flow pathways (Feenders et al., 2008). As chickens have limited flight capabilities, the neuromorphology of chicken Purkinje cells may be correlated with fine tuning motor capabilities including flight, which requires integration of multiple senses to allow movement in the air. It is also tempting to speculate that the increased Purkinje cell neuromorphology of mice may be related to higher degrees of forelimb dexterity. Interestingly, brain size was not found to be
71
correlated to total forelimb dexterity in mammalian species, and cerebellar size was not a good indicator of flight capabilities in avian species (Iwaniuk et al., 1999; Walsh et al., 2013). Thus, the possibility remains that differences on a smaller scale (i.e. Purkinje cell neuromorphological variation) may contribute towards differences in ecological behaviour and functional capabilities.
Future studies examining the relationship between flight capabilities of various species and
Purkinje cell neuromorphology, as well as comparative studies on branching shape and functional output may help reveal the contributions of neuromorphology towards varying degrees of function.
4.2 Form and function of gecko Purkinje cells
To investigate the form-function relationship of Purkinje cells in leopard gecko, I used whole-cell electrophysiology and biocytin filling. I found that there were at least two distinct populations of Purkinje cells. Type 1 cells are less excitable than type 2 cells, had larger and have more complex dendrite arbors that clustered farther from the soma than type 2 cells. The presence of multiple Purkinje cell types have been previously documented in mouse and mormyrid fish species, both of which were morphologically and functionally distinct (Han &
Bell, 2003; Zhang & Han, 2007; Kim et al., 2012). However, those studies also revealed regional differences in the location of Purkinje cells in the cerebellum that was not seen in the leopard gecko, as both type 1 and type 2 cells were distributed evenly throughout the cerebellar slices sampled. To investigate the purpose of this heterogeneous distribution of cell subtypes, further studies using axon tracing to find the postsynaptic targets of gecko Purkinje cells may provide insight on the differences in regional axonal projections between cell subtypes.
72
4.3 Differences between Golgi-Cox stained and Biocytin filled neurons
Although not the subject of my research, it is worth noting that there were differences in the neuromorphology of gecko Purkinje cells based on the method of visualization used (i.e.,
Golgi-Cox staining vs. biocytin labelling). For example, the dendrite arbours labeled with biocytin (Chapter 3) were larger and had three-times as many dendrite terminals when compared to those stained by Golgi-Cox (Chapter 2). The differences in the two staining methods have been previously quantified, where it was found that biocytin staining resulted in significantly larger dendrite arbors when directly compared to cells stained using the Golgi-Cox method
(Klenowski et al., 2017). Here, it is worth noting that the geckos used for each experiment were not age-matched: geckos in the Golgi-Cox study were subadults, ~200 days old, while geckos in the biocytin study were adults, >1yr of age. In addition, there are also important differences in the methods themselves. Golgi-Cox staining includes the use of tissue dehydration, leading to dendrite shrinkage (Grace & Llinás, 1985; Louth et al., 2017). At this time, it remains unclear whether differences in chronological age and dehydration account for the variation observed between the two visualization methods employed. Moving forward, future investigations trialling both techniques on brain tissue from same subject and/or members of the same size- matched/age-matched cohort are necessary.
73
REFERENCES
Arancillo, M., White, J. J., Lin, T., Stay, T. L., & Sillitoe, R. V. (2015). In vivo analysis of
Purkinje cell firing properties during postnatal mouse development. Journal of
Neurophysiology, 113(2), 578–591.
Aspden, J. W., Armstrong, C. L., Gutierrez-Ibanez, C. I., Hawkes, R., Iwaniuk, A. N., Kohl, T.,
Graham, D. J., & Wylie, D. R. (2015). Zebrin II / Aldolase C expression in the cerebellum
of the western diamondback rattlesnake (Crotalus atrox). PLOS One, 10(2).
Azevedo, F. A. C., Carvalho, L. R. B., Grinberg, L. T., Farfel, J. M., Ferretti, R. E. L., Leite, R.
E. P., Filho, W. J., Lent, R., & Herculano-Houzel, S. (2009). Equal numbers of neuronal
and nonneuronal cells make the human brain an isometrically scaled-up primate brain. The
Journal of Comparative Neurology, 513(5), 532–541.
Baillieux, H., Smet, H. J. D., Paquier, P. F., De Deyn, P. P., & Mariën, P. (2008). Cerebellar
neurocognition: Insights into the bottom of the brain. Clinical Neurology and Neurosurgery,
110(8), 763–773.
Barrowclough, G. F., Cracraft, J., Klicka, J., & Zink, R. M. (2016). How Many Kinds of Birds
Are There and Why Does It Matter? PLOS One, 11(11).
Bell, C. C. (2002). Evolution of Cerebellum-Like Structures. Brain, Behavior and Evolution,
59(5–6), 312–326.
Benavides-Piccione, R., Ballesteros-Yáñez, I., DeFelipe, J., & Yuste, R. (2002). Cortical area
and species differences in dendritic spine morphology. Journal of Neurocytology, 31, 337–
346.
Benavides-Piccione, R., Regalado-Reyes, M., Fernaud-Espinosa, I., Kastanauskaite, A., Tapia-
González, S., León-Espinosa, G., Rojo, C., Insausti, R., Segev, I., & DeFelipe, J. (2020).
74
Differential Structure of Hippocampal CA1 Pyramidal Neurons in the Human and Mouse.
Cerebral Cortex, 30, 730–752.
Bianchi, D., Marasco, A., Limongiello, A., Marchetti, C., Marie, H., Tirozzi, B., & Migliore, M.
(2012). On the mechanisms underlying the depolarization block in the spiking dynamics of
CA1 pyramidal neurons. Journal of Computational Neuroscience, 33(2), 207–225.
Branco, T., & Häusser, M. (2010). The single dendritic branch as a fundamental functional unit
in the nervous system. Current Opinion in Neurobiology, 20(4), 494–502.
Buckner, R. L. (2013). The Cerebellum and Cognitive Function: 25 Years of Insight from
Anatomy and Neuroimaging. Neuron, 80(3), 807–815.
Burgin, C. J., Colella, J. P., Kahn, P. L., & Upham, N. S. (2018). How many species of mammals
are there? Journal of Mammalogy, 99(1), 1–14.
Butts, T., Chaplin, N., & Wingate, R. J. T. (2011). Can Clues from Evolution Unlock the
Molecular Development of the Cerebellum? Molecular Neurobiology, 43(1), 67–76.
Butts, T., Green, M. J., & Wingate, R. J. T. (2014). Development of the cerebellum: Simple steps
to make a “little brain.” Development, 141(21), 4031–4041.
Cantalupo, C., & Hopkins, W. (2010). The cerebellum and its contribution to complex tasks in
higher primates: A comparative perspective. Cortex, 46(7), 821–830.
Chan, C. Y., & Nicholson, C. (1986). Modulation by applied electric fields of Purkinje and
stellate cell activity in the isolated turtle cerebellum. The Journal of Physiology, 371(1), 89–
114.
Chang, W., Pedroni, A., Hohendorf, V., Giacomello, S., Hibi, M., Köster, R. W., & Ampatzis, K.
(2020). Functionally distinct Purkinje cell types show temporal precision in encoding
locomotion. Proceedings of the National Academy of Sciences, 117(29), 17330–17337.
75
Chung, B. Y. T., Bignell, W., Jacklin, D. L., Winters, B. D., & Bailey, C. D. C. (2016).
Postsynaptic nicotinic acetylcholine receptors facilitate excitation of developing CA1
pyramidal neurons. Journal of Neurophysiology, 116(5), 2043–2055.
Connors, B. W., & Regehr, W. G. (1996). Neuronal firing: Does function follow form? Current
Biology, 6(12), 1560–1562.
Cunha, F., Racicot, K., Nahirney, J., Heuston, C., Wylie, D. R., & Iwaniuk, A. N. (2020).
Allometric scaling rules of the cerebellum in galliform birds. Brain, Behavior, and
Evolution, 95, 78–92.
Czechowska, N., van Rienen, A., Lang, F., Eiberger, B., & Baader, S. L. (2019). An update on
the Golgi staining technique improving cerebellar cell type specificity. Histochemistry and
Cell Biology, 151(4), 327–341.
D’Angelo, E., Solinas, S., Mapelli, J., Gandolfi, D., Mapelli, L., & Prestori, F. (2013). The
cerebellar Golgi cell and spatiotemporal organization of granular layer activity. Frontiers in
Neural Circuits, 7.
Day, C. E. (Ed.). (2014). Histopathology: Methods and Protocols (Vol. 1180). Springer New
York.
De Biase, D., & Paciello, O. (2015). Essential and current methods for a practical approach to
comparative neuropathology. Folia Morphologica, 74(2), 137–149.de Ruiter, M. M., De
Zeeuw, C. I., & Hansel, C. (2006). Voltage-Gated Sodium Channels in Cerebellar Purkinje
Cells of Mormyrid Fish. Journal of Neurophysiology, 96(1), 378–390.
DeFelipe, Javier. (2015). The dendritic spine story: An intriguing process of discovery. Frontiers
in Neuroanatomy, 9.
76
De Zeeuw, C. I., & Berrebi, A. S. (1995). Postsynaptic Targets of Purkinje Cell Terminals in the
Cerebellar and Vestibular Nuclei of the Rat. European Journal of Neuroscience, 7(11),
2322–2333.
Diamond, A. (2000). Close Interrelation of Motor Development and Cognitive Development and
of the Cerebellum and Prefrontal Cortex. Child Development, 71(1), 44–56.
Eccles, J. C. (1969). The development of the cerebellum of vertebrates in relation to the control
of movement. Die Naturwissenschaften, 56(11), 525–534.
Farhoodi, R., Lansdell, B. J., & Kording, K. P. (2019). Quantifying How Staining Methods Bias
Measurements of Neuron Morphologies. Frontiers in Neuroinformatics, 13, 36.
Feenders, G., Liedvogel, M., Rivas, M., Zapka, M., Horita, H., Hara, E., Wada, K., Mouritsen,
H., & Jarvis, E. D. (2008). Molecular Mapping of Movement-Associated Areas in the Avian
Brain: A Motor Theory for Vocal Learning Origin. PLoS ONE, 3(3), e1768.
Friede, R. L. (1963). The relationship of body size, nerve cell size, axon length, and glial density
in the cerebellum. Proceedings of the National Academy of Sciences of the United States of
America, 49(2), 187-193.
Fujishima, Kazuto, Kawabata Galbraith, K., & Kengaku, M. (2018). Dendritic Self-Avoidance
and Morphological Development of Cerebellar Purkinje Cells. The Cerebellum, 17(6), 701–
708.
García-López, P., García-Marín, V., & Freire, M. (2007). The discovery of dendritic spines by
Cajal in 1888 and its relevance in the present neuroscience. Progress in Neurobiology,
83(2), 110–130.
Glynn, M. W., & McAllister, A. K. (2006). Immunocytochemistry and quantification of protein
colocalization in cultured neurons. Nature Protocols, 1(3), 1287–1297.
77
Goodman, D. C., Horel, J. A., & Freemon, F. R. (1964). Functional localization in the
cerebellum of the bird and its bearing on the evolution of cerebellar function. The Journal of
Comparative Neurology, 123(1), 45–53.
Goriounova, N. A., Heyer, D. B., Wilbers, R., Verhoog, M. B., Giugliano, M., Verbist, C.,
Obermayer, J., Kerkhofs, A., Smeding, H., Verberne, M., Idema, S., Baayen, J. C.,
Pieneman, A. W., de Kock, C. P., Klein, M., & Mansvelder, H. D. (2018). Large and fast
human pyramidal neurons associate with intelligence. ELife, 7, e41714.
Graves, A. R., Moore, S. J., Bloss, E. B., Mensh, B. D., Kath, W. L., & Spruston, N. (2012).
Hippocampal pyramidal neurons comprise two distinct cell types that are countermodulated
by metabotropic receptors. Neuron, 76(4), 776–789.
Gruol, D., & Franklin, C. (1987). Morphological and physiological differentiation of Purkinje
neurons in cultures of rat cerebellum. The Journal of Neuroscience, 7(5), 1271–1293.
Gutiérrez-Ibáñez, C., Dannish, M. R., Kohl, T., & Kettler, L. (2020). Zebrin expression in the
cerebellum of two crocodilian species. Brain Behaviour and Evolution, 95(1), 1-11.
Hamill, O. P., Marty, A., Neher, E., Sakmann, B., & Sigworth, F. J. (1981). Improved patch-
clamp techniques for high-resolution current recording from cells and cell-free membrane
patches. Pflügers Archiv - European Journal of Physiology, 391(2), 85–100.
Han, V. Z., & Bell, C. C. (2003). Physiology of Cells in the Central Lobes of the Mormyrid
Cerebellum. The Journal of Neuroscience, 23(35), 11147–11157.
Hashimoto, M., & Hibi, M. (2012). Development and evolution of cerebellar neural circuits.
Development, Growth & Differentiation, 54(3), 373–389.
Hatten, M. E., & Heintz, N. (1995). Mechanisms of Neural Patterning and Specification in the
Development Cerebellum. Annual Review of Neuroscience, 18(1), 385–408.
78
Hayter, A. L., Langdon, D. W., & Ramnani, N. (2007). Cerebellar contributions to working
memory. 12.
Herculano-Houzel, S. (2009). The Human Brain in Numbers: A Linearly Scaled-up Primate
Brain. Frontiers in Human Neuroscience, 3.
Herculano-Houzel, S. (2010). Coordinated scaling of cortical and cerebellar numbers of neurons.
Frontiers in Neuroanatomy, 8.
Herculano-Houzel, S., Avelino-de-Souza, K., Neves, K., PorfÃrio, J., Messeder, D., Mattos
FeijÃ3, L., Maldonado, J., & Manger, P. R. (2014). The elephant brain in numbers.
Frontiers in Neuroanatomy, 8.
Herculano-Houzel, S., Catania, K., Manger, P. R., & Kaas, J. H. (2015). Mammalian Brains Are
Made of These: A Dataset of the Numbers and Densities of Neuronal and Nonneuronal
Cells in the Brain of Glires, Primates, Scandentia, Eulipotyphlans, Afrotherians and
Artiodactyls, and Their Relationship with Body Mass. Brain, Behavior and Evolution,
86(3–4), 145–163.
Herndon, R. M. (1963). The fine structure of the Purkinje cell. Journal of Cell Biology, 18(1),
167–180.
Hibi, M., & Shimizu, T. (2012). Development of the cerebellum and cerebellar neural circuits.
Developmental Neurobiology, 72(3), 282–301.
Hibi, M., Matsuda, K., Takeuchi, M., Shimizu, T., & Murakami, Y. (2017). Evolutionary
mechanisms that generate morphology and neural-circuit diversity of the cerebellum.
Development, Growth & Differentiation, 59(4), 228–243.
Hodgkin, A. L., & Katz, B. (1949). The effect of sodium ions on the electrical activity of the
giant axon of the squid. The Journal of Physiology, 108(1), 37–77.
79
Hodgkin, A. L., & Huxley, A. F. (1952a). Currents carried by sodium and potassium ions
through the membrane of the giant axon of Loligo. The Journal of Physiology, 116(4), 449–
472.
Hodgkin, A. L., & Huxley, A. F. (1952b). The components of membrane conductance in the
giant axon of Loligo. The Journal of Physiology, 116(4), 473–496.
Hodgkin, A. L., & Huxley, A. F. (1952c). The dual effect of membrane potential on sodium
conductance in the giant axon of Loligo. The Journal of Physiology, 116(4), 497–506.
Hodgkin, A. L., & Huxley, A. F. (1952d). A quantitative description of membrane current and its
application to conduction and excitation in nerve. The Journal of Physiology, 117(4), 500–
544.
Horikawa, K., & Armstrong, W. E. (1988). A versatile means of intracellular labeling: Injection
of biocytin and its detection with avidin conjugates. Journal of Neuroscience Methods,
25(1), 1–11.
Hounsgaard, J., & Midtgaard, J. (1988). Intrinsic determinants of firing pattern in Purkinje cells
of the turtle cerebellum in vitro. The Journal of Physiology, 402(1), 731–749.
Hoxha, E., Balbo, I., Miniaci, M. C., & Tempia, F. (2018). Purkinje Cell Signaling Deficits in
Animal Models of Ataxia. Frontiers in Synaptic Neuroscience, 10, 6.
Huang, C., Gammon, S. J., Dieterle, M., Huang, R. H., Likins, L., & Ricklefs, R. E. (2014).
Dramatic increases in number of cerebellar granule-cell-Purkinje-cell synapses across
several mammals. Mammalian Biology, 79(3), 163–169.
Ito, M. (2006). Cerebellar circuitry as a neuronal machine. Progress in Neurobiology, 78(3–5),
272–303.
80
Iwaniuk, A. N., Pellis, S. M., & Whishaw, I. Q. (1999). Brain Size Is Not Correlated with
Forelimb Dexterity in Fissiped Carnivores (Carnivora): A Comparative Test of the Principle
of Proper Mass. Brain Behaviour and Evolution, 54, 167–180.
Iwaniuk, A. N., Hurd, P. L., & Wylie, D. R. W. (2006). Comparative Morphology of the Avian
Cerebellum: I. Degree of Foliation. Brain, Behavior and Evolution, 68(1), 45–62.
Jackson, M. B. (1997). Whole-Cell Voltage Clamp Recording. Current Protocols in
Neuroscience, 6.6.1-6.6.30.
Jacobs, Bob, Johnson, N. L., Wahl, D., Schall, M., Maseko, B. C., Lewandowski, A., Raghanti,
M. A., Wicinski, B., Butti, C., Hopkins, W. D., Bertelsen, M. F., Walsh, T., Roberts, J. R.,
Reep, R. L., Hof, P. R., Sherwood, C. C., & Manger, P. R. (2014). Comparative neuronal
morphology of the cerebellar cortex in afrotherians, carnivores, cetartiodactyls, and
primates. Frontiers in Neuroanatomy, 8.
Kádár, A., Wittmann, G., Liposits, Z., & Fekete, C. (2009). Improved method for combination of
immunocytochemistry and Nissl staining. Journal of Neuroscience Methods, 184(1), 115–
118.
Kaffashian, M., Shabani, M., Goudarzi, I., Behzadi, G., Zali, A., & Janahmadi, M. (2011).
Profound Alterations in the Intrinsic Excitability of Cerebellar Purkinje Neurons Following
Neurotoxin 3-Acetylpyridine (3-AP)-Induced Ataxia in Rat: New Insights Into the Role of
Small Conductance K+ Channels. Physiological Research, 355–365.
Kato, T., Hirano, A., & Llena, J. F. (1985). A Golgi study of the human Purkinje cell soma and
dendrites. Acta Neuropathologica, 68(2), 145–148.
81
Khaliq, Z. M., & Raman, I. M. (2006). Relative Contributions of Axonal and Somatic Na
Channels to Action Potential Initiation in Cerebellar Purkinje Neurons. The Journal of
Neuroscience, 26(7), 1935–1944.
Kidd, K. (2017). A comparative analysis of Purkinje cells across species combining modelling,
machine learning, and information theory [Doctoral dissertation, University of
Hertfordshire].
Kim, C.-H., Oh, S.-H., Lee, J. H., Chang, S. O., Kim, J., & Kim, S. J. (2012). Lobule-specific
membrane excitability of cerebellar Purkinje cells: Lobular difference of Purkinje cell
excitability. The Journal of Physiology, 590(2), 273–288.
Klenowski, P., Wright, S., Mu, E., Noakes, P., Lavidis, N., Bartlett, S., Bellingham, M., &
Fogarty, M. (2017). Investigating Methodological Differences in the Assessment of
Dendritic Morphology of Basolateral Amygdala Principal Neurons—A Comparison of
Golgi–Cox and Neurobiotin Electroporation Techniques. Brain Sciences, 7(12), 165.
Knopfel, T., Vranesic, I., Staub, C., & Gahwiler, B. H. (1991). Climbing Fibre Responses in
Olivo-cerebellar Slice Cultures. II. Dynamics of Cytosolic Calcium in Purkinje Cells.
European Journal of Neuroscience, 3(4), 343–348.
Kole, M. H., & Stuart, G. J. (2012). Signal Processing in the Axon Initial Segment. Neuron,
73(2), 235–247.
Koyama, Y. (2013). The unending fascination with the Golgi method. OA Anatomy, 1(3).
Krichmar, J. L., Nasuto, S. J., Scorcioni, R., Washington, S. D., & Ascoli, G. A. (2002). Effects
of dendritic morphology on CA3 pyramidal cell electrophysiology: A simulation study.
Brain Research, 941(1–2), 11–28.
82
Larsell, O. (1926). The cerebellum of reptiles: Lizards and snake. The Journal of Comparative
Neurology, 41(1), 59–94.
Levick, W. R. (1975). Form and function of cat retinal ganglion cells. Nature, 254(5502), 659–
662.
Llinás, R., & Hillman, D. E. (1969). Neuronal organization of the cerebellar cortex in amphibia
and reptilia. In R. Llinas (Ed.), Neurobiology of Cerebellar Evolution and Development (pp.
279-327). American Medical Association.
Llinás, R., & Nicholson, C. (1971). Electrophysiological properties of dendrites and somata in
alligator Purkinje cells. Journal of Neurophysiology, 34(4), 532–551.
Llinás, R., & Sugimori, M. (1980). Electrophysiological properties of in vitro Purkinje cell
dendrites in mammalian cerebellar slices. The Journal of Physiology, 305(1), 197–213.
Llinás, R. (1988). The intrinsic electrophysiological properties of mammalian neurons: Insights
into central nervous system function. Science, 242(4886), 1654–1664.
Locke, T. M., Fujita, H., Hunker, A., Johanson, S. S., Darvas, M., du Lac, S., Zweifel, L. S., &
Carlson, E. S. (2020). Purkinje Cell-Specific Knockout of Tyrosine Hydroxylase Impairs
Cognitive Behaviors. Frontiers in Cellular Neuroscience, 14, 228.
Louis, E. D., Lee, M., Babij, R., Ma, K., Cortés, E., Vonsattel, J.-P. G., & Faust, P. L. (2014).
Reduced Purkinje cell dendritic arborization and loss of dendritic spines in essential tremor.
Brain, 137(12), 3142–3148.
Louis, E. D., Kuo, S.-H., Tate, W. J., Kelly, G. C., Gutierrez, J., Cortes, E. P., Vonsattel, J.-P. G.,
& Faust, P. L. (2018). Heterotopic Purkinje Cells: A Comparative Postmortem Study of
Essential Tremor and Spinocerebellar Ataxias 1, 2, 3, and 6. Cerebellum, 17, 104-110.
83
Louth, E. L., Bignell, W., Taylor, C. L., & Bailey, C. D. C. (2016). Developmental Ethanol
Exposure Leads to Long-Term Deficits in Attention and Its Underlying Prefrontal Circuitry.
ENeuro, 3(5).
Louth, E. L., Sutton, C. D., Mendell, A. L., MacLusky, N. J., & Bailey, C. D. C. (2017). Imaging
Neurons within Thick Brain Sections Using the Golgi-Cox Method. Journal of Visualized
Experiments : JoVE, 122.
Macrì, S., Savriama, Y., Khan, I., & Di-Poï, N. (2019). Comparative analysis of squamate brains
unveils multi-level variation in cerebellar architecture associated with locomotor
specialization. Nature Communications, 10(1), 5560.
Macrì, S., & Di-Poï, N. (2020). Heterochronic Developmental Shifts Underlying Squamate
Cerebellar Diversity Unveil the Key Features of Amniote Cerebellogenesis. Frontiers in
Cell and Developmental Biology, 8, 593377.
Marx, M., Günter, R. H., Hucko, W., Radnikow, G., & Feldmeyer, D. (2012). Improved biocytin
labeling and neuronal 3D reconstruction. Nature Protocols, 7(2), 394–407.
Masoli, S., Solinas, S., & D’Angelo, E. (2015). Action potential processing in a detailed Purkinje
cell model reveals a critical role for axonal compartmentalization. Frontiers in Cellular
Neuroscience, 9.
McKay, B. E., & Turner, R. W. (2005). Physiological and morphological development of the rat
cerebellar Purkinje cell: Purkinje cell output parallels dendritic development. The Journal of
Physiology, 567(3), 829–850.
McLean, K. E., & Vickaryous, M. K. (2011). A novel amniote model of epimorphic
regeneration: The leopard gecko, Eublepharis macularius. BMC Developmental Biology,
11(1), 50.
84
Meek, J., & Nieuwenhuys, R. (1991). Palisade pattern of mormyrid Purkinje cells: A correlated
light and electron microscopic study. The Journal of Comparative Neurology, 306(1), 156–
192.
Meek, J., & Nieuwenhuys, R. (1998). Holosteans and Teleosts. In R. Nieuwenhuys, H. J. ten
Donkelaar, & C. Nicholson (Eds.), The Central Nervous System of Vertebrates: Volume 1 /
Volume 2 / Volume 3 (pp. 759–937). Springer.
Miller, J. P., & Jacobs, G. A. (1984). Relationships between neuronal structure and function.
Journal of Experimental Biology, 112, 129–145.
Nakamura, S., Ochiai, K., Yabushita, H., & Abe, A. (2014). Congenital cerebellar dysplasia in
White Leghorn chickens (Gallus gallus domesticus). Avian Pathology, 43(2), 135–138.
Nedelescu, H., & Abdelhack, M. (2013). Comparative Morphology of Dendritic Arbors in
Populations of Purkinje Cells in Mouse Sulcus and Apex. Neural Plasticity, 2013, 1–12.
Nedelescu, H., Abdelhack, M., & Pritchard, A. T. (2018). Regional differences in Purkinje cell
morphology in the cerebellar vermis of male mice. Journal of Neuroscience Research,
96(9), 1476–1489.
Netzeband, J. G., Trotter, C., Parsons, K. L., & Gruol, D. L. (1999). Chronic ethanol treatment
alters AMPA-induced calcium signals in developing Purkinje neurons. Brain Research,
826(2), 270–280.
Nieuwenhuys, R., & Nicholson, C. (1967). Cerebellum of Mormyrids. Nature, 215, 764-765.
Nishiyama, H., & Linden, D. J. (2004). Differential Maturation of Climbing Fiber Innervation in
Cerebellar Vermis. Journal of Neuroscience, 24(16), 3926–3932.
O’Brien, J., & Unwin, N. (2006). Organization of spines on the dendrites of Purkinje cells.
Proceedings of the National Academy of Sciences, 103(5), 1575–1580.
85
Ohtsuki, G., & Hansel, C. (2018). Synaptic Potential and Plasticity of an SK2 Channel Gate
Regulate Spike Burst Activity in Cerebellar Purkinje Cells. IScience, 1, 49–54.
Otis, T. S. (2016). Simple Spikes and Complex Spikes. In D. L. Gruol, N. Koibuchi, M. Manto,
M. Molinari, J. D. Schmahmann, & Y. Shen (Eds.), Essentials of Cerebellum and
Cerebellar Disorders (pp. 299–303). Springer International Publishing.
Parra, A., Baker, C. A., & Bolton, M. M. (2019). Regional Specialization of Pyramidal Neuron
Morphology and Physiology in the Tree Shrew Neocortex. Cerebral Cortex, 29(11), 4488–
4505.
Pasternak, J. F., & Woolsey, T. A. (1975). On the selectivity Of the Golgi-Cox method. The
Journal of Comparative Neurology, 160(3), 307–312.
Petersen, C. C. H. (2017). Whole-Cell Recording of Neuronal Membrane Potential during
Behavior. Neuron, 95(6), 1266–1281.
Rahimi-Balaei, M., Bergen, H., Kong, J., & Marzban, H. (2018). Neuronal Migration During
Development of the Cerebellum. Frontiers in Cellular Neuroscience, 12, 484.
Ramón y Cajal, S. (1888). Estructura de los centros nerviosos de las aves. Revista Trimestral de
Histología Normal y Patológica, 1, 1-10.
Ramón y Cajal, S. (1899). Comparative Study of the Sensory Areas of the Human Cortex. Clark
University.
Ramón y Cajal, S. (1909). Histologie du système nerveux de l’homme & des vertébrés. Paris: A.
Maloine.
Schmahmann, J. (1997). The Cerebellum and Cognition. Academic Press.
Schoenwolf, G., & Larsen, W. (2009). Larsen’s Human Embryology (4th ed., Thoroughly rev.
and updated.). Churchill Livingstone/Elsevier.
86
Segev, A., Garcia-Oscos, F., & Kourrich, S. (2016). Whole-cell Patch-clamp Recordings in
Brain Slices. Journal of Visualized Experiments, 112, 54024.
Sereno, M. I., Diedrichsen, J., Tachrount, M., Testa-Silva, G., d’Arceuil, H., & De Zeeuw, C.
(2020). The human cerebellum has almost 80% of the surface area of the neocortex.
Proceedings of the National Academy of Sciences, 117(32), 19538–19543.
Shimono, M., & Tsuji, N. (1987). Study of the selectivity of the impregnation of neurons by the
Golgi method. The Journal of Comparative Neurology, 259(1), 122–130.
Siggins, G. R., Henriksen, S. J., & Landis, S. C. (1976). Electrophysiology of purkinje neurons
in the weaver mouse: Iontophoresis of neurotransmitters and cyclic nucleotides, and
stimulation of the nucleus locus coeruleus. Brain Research, 114(1), 53–69.
Simons, M., & Pellionisz, A. (2006). Genomics, morphogenesis and biophysics: Triangulation of
Purkinje cell development. The Cerebellum, 5(1), 27–35.
Smith, T. G., Brauer, K., & Reichenbach, A. (1993). Quantitative Phylogenetic Constancy of
Cerebellar Purkinje Cell Morphological Complexity. The Journal of Comparative
Neurology, 331, 402-406.
Sontheimer, H., & Olsen, M. L. (2007). Whole-Cell Patch-Clamp Recordings. In W. Walz (Ed.),
Patch-Clamp Analysis: Advanced Techniques (pp. 35-68). Humana Press Inc.
Storm, D. R., Hansel, C., Hacker, B., Parent, A., & Linden, D. J. (1998). Impaired Cerebellar
Long-Term Potentiation in Type I Adenylyl Cyclase Mutant Mice. Neuron, 20(6), 1199–
1210.
Sultan, F., & Glickstein, M. (2007). The cerebellum: Comparative and animal studies. The
Cerebellum, 6(3), 168–176.
87
Takeda, T., Ishikawa, A., Ohtomo, K., Kobayashi, Y., & Matsuoka, T. (1992). Fractal dimension
of dendritic tree of cerebellar Purkinje cell during onto- and phylogenetic development.
Neuroscience Research, 13(1), 19–31.
Tasker, J. G., Hoffman, N. W., & Dudek, F. E. (1991). Comparison of three intracellular markers
for combined electrophysiological, morphological and immunohistochemical analyses.
Journal of Neuroscience Methods, 38(2–3), 129–143. ten Donkelaar, H. J. (1998). Reptiles. In R. Nieuwenhuys, H. J. ten Donkelaar, & C. Nicholson
(Eds.), The Central Nervous System of Vertebrates: Volume 1 / Volume 2 / Volume 3 (pp.
1315–1524). Springer.
Teune, T. M., Van Der Burg, J., De Zeeuw, C. I., Voogd, J., & Ruigrok, T. J. H. (1998). Single
Purkinje cell can innervate multiple classes of projection neurons in the cerebellar nuclei of
the rat: A light microscopic and ultrastructural triple-tracer study in the rat. The Journal of
Comparative Neurology, 392, 164-178.
Thomson, A. M., & Armstrong, W. E. (2011). Biocytin-labelling and its impact on late 20th
century studies of cortical circuitry. Brain Research Reviews, 66(1–2), 43–53.
Titley, H. K., Watkins, G. V., Lin, C., Weiss, C., McCarthy, M., Disterhoft, J. F., & Hansel, C.
(2020). Intrinsic Excitability Increase in Cerebellar Purkinje Cells after Delay Eye-Blink
Conditioning in Mice. Journal of Neuroscience, 40(10), 2038–2046.
Trzesniewski, J., Altmann, S., Jäger, L., & Kapfhammer, J. P. (2019). Reduced Purkinje cell size
is compatible with near normal morphology and function of the cerebellar cortex in a mouse
model of spinocerebellar ataxia. Experimental Neurology, 311, 205–212.
Uetz, P., Slavenko, A., Meiri, S., & Heinicke, M. (2020). Gecko diversity: a history of global
discovery. Israel Journal of Ecology and Evolution, 66(3-4), 117-125.
88
Uray, N. J., & Gona, A. G. (1978). Golgi studies on Purkinje cell development in the frog during
spontaneous metamorphosis I. General pattern of development. The Journal of Comparative
Neurology, 180(2), 265–276. van Elburg, R. A. J., & van Ooyen, A. (2010). Impact of Dendritic Size and Dendritic Topology
on Burst Firing in Pyramidal Cells. PLoS Computational Biology, 6(5), e1000781.
Vetter, P., Roth, A., & Häusser, M. (2001). Propagation of Action Potentials in Dendrites
Depends on Dendritic Morphology. Journal of Neurophysiology, 85(2), 926–937.
Vints, K., Vandael, D., Baatsen, P., Pavie, B., Vernaillen, F., Corthout, N., Rybakin, V., Munck,
S., & Gounko, N. V. (2019). Modernization of Golgi staining techniques for high-
resolution, 3-dimensional imaging of individual neurons. Scientific Reports, 9(1), 130.
Voogd, J., & Glickstein, M. (1998). The anatomy of the cerebellum. Trends in Neurosciences,
21, 370-375.
Walsh, S. A., Iwaniuk, A. N., Knoll, M. A., Bourdon, E., Barrett, P. M., Milner, A. C., Nudds, R.
L., Abel, R. L., & Sterpaio, P. D. (2013). Avian Cerebellar Floccular Fossa Size Is Not a
Proxy for Flying Ability in Birds. PLoS ONE, 8(6), e67176.
Walter, J. T., & Khodakhah, K. (2006). The Linear Computational Algorithm of Cerebellar
Purkinje Cells. Journal of Neuroscience, 26(50), 12861–12872.
White, J. J., & Sillitoe, R. V. (2013). Development of the cerebellum: From gene expression
patterns to circuit maps: Development of cerebellum. Wiley Interdisciplinary Reviews:
Developmental Biology, 2(1), 149–164.
Willett, J. A., Cao, J., Dorris, D. M., Johnson, A. G., Ginnari, L. A., & Meitzen, J. (2019).
Electrophysiological Properties of Medium Spiny Neuron Subtypes in the Caudate-Putamen
of Prepubertal Male and Female Drd1a-tdTomato Line 6 BAC Transgenic Mice. 14.
89
Yang, Q., Hashizume, Y., Yoshida, M., Wang, Y., Goto, Y., Mitsuma, N., Ishikawa, K., &
Mizusawa, H. (2000). Morphological Purkinje cell changes in spinocerebellar ataxia type 6.
Acta Neuropathologica, 100(4), 371–376.
Yang, Z., Chen, N., Ge, R., Qian, H., & Wang, J.-H. (2017). Functional compatibility between
Purkinje cell axon branches and their target neurons in the cerebellum. Oncotarget, 8(42),
72424–72437.
Yi, G.-S., Wang, J., Deng, B., & Wei, X.-L. (2017). Morphology controls how hippocampal CA1
pyramidal neuron responds to uniform electric fields: A biophysical modeling study.
Scientific Reports, 7(1), 3210.
Yopak, K. E., Galinsky, V. L., Berquist, R. M., & Frank, L. R. (2016). Quantitative
Classification of Cerebellar Foliation in Cartilaginous Fishes (Class: Chondrichthyes) Using
Three-Dimensional Shape Analysis and Its Implications for Evolutionary Biology. Brain,
Behavior and Evolution, 87(4), 252–264.
Yopak, K. E., Pakan, J. M. P., & Wylie, D. (2017). 1.20—The Cerebellum of Nonmammalian
Vertebrates. In J. H. Kaas (Ed.), Evolution of Nervous Systems (Second Edition) (pp. 373–
385). Academic Press.
Zaqout, S., & Kaindl, A. M. (2016). Golgi-Cox Staining Step by Step. Frontiers in
Neuroanatomy, 10.
Zhang, Yueping, & Han, V. Z. (2007). Physiology of Morphologically Identified Cells in the
Posterior Caudal Lobe of the Mormyrid Cerebellum. Journal of Neurophysiology, 98(3),
1297–1308.
90
Zhang, Yueping, Han, P. F., & Han, V. Z. (2008). Local circuitry in the anterior caudal lobe of
the mormyrid cerebellum: A study of intracellular recording and labeling. The Journal of
Comparative Neurology, 509(1), 1–22.
Zhou, H., Lin, Z., Voges, K., Ju, C., Gao, Z., Bosman, L. W., Ruigrok, T. J., Hoebeek, F. E., De
Zeeuw, C. I., & Schonewille, M. (2014). Cerebellar modules operate at different
frequencies. ELife, 3, e02536.
91
APPENDIX I: DETAILED HISTOCHEMICAL PROTOCOLS
Golgi-Cox Protocol (adapted from the Bailey Lab, University of Guelph)
1) Subject is euthanized and the brain is removed.
2) Whole brain is placed into a 20 mL scintillation vial containing 20 mL of Golgi-Cox Solution (i.e., topped-up with solution until the container is full). Cover the vial with aluminum foil and incubate for 25 days at room temperature in the dark.
3) Place brain into a 20 mL scintillation vial containing 20 mL of Sucrose Cryoprotectant. Cover the vial with aluminum foil and incubate for 48 hours at 4 °C in the dark.
4) Remove brain from sucrose and block it for slicing by making a small parasagittal cut on one side of the brain. Place the brain cut-side down in a small weigh boat/dish. Heat the agar prior to blocking the brain and when it has cooled to near-solid phase, pour it into the weigh boat to encase the brain.
5) Once the agar has solidified, trim excess agar and glue the brain to the stage of the vibratome. Slice in Sucrose Cryoprotectant at 400 µm thickness. Place slices into a well of a 6-well plate (35 mm diameter) filled with 10 mL of 6% Sucrose and incubate overnight at 4 °C in the dark.
6) Place slices into a well with 5 mL of 2% paraformaldehyde. Slow rocking speed (15rpm) for 15 min.
7) Wash slices twice in wells containing 5 mL of H2O. Moderate rocking speed (20rpm), 5 mins each.
8) Place slices into a well filled with 5 mL of 2.7% NH4OH for 15 min. Slow rocking speed (15rpm) for 15 min.
9) Wash slices twice in wells containing 5 mL of H2O. Moderate rocking speed (20rpm), 5 mins each.
10) Place slices into a well filled with 5 mL of Kodak Fixative A (diluted 10x from purchased concentration). Slow rocking speed (15rpm) for 25 min.
11) Wash slices twice in wells containing 5 mL of H2O. Moderate rocking speed (20rpm), 5 mins each.
12) Mount slices onto gelatinized slides, or Superfrost Plus slides. Remove excess agar and water with a Kimwipe.
92
13) Allow slides to air dry for 45 mins (400 µm) or 90 mins (500 µm). This timing is critical, as too short a time leads to slices falling off of the slides and too long leads to cracked slices. Should look shiny, and have a bit of a halo where water has dried.
14) Dehydrate slices: a. 50% EtOH 2 min b. 75% EtOH 2 min c. 95% EtOH 2 min d. 100% EtOH 5 min e. 100% EtOH 5 min f. Citrisolv for 5 min g. Citrisolv for 5 min.
15) Coverslip each section or slide using Permount and allow slides to dry horizontally for five days in the dark.
Golgi-Cox Solution 1% (w/v) potassium dichromate 0.8% (w/v) potassium chromate 1% (w/v) mercuric chloride
0.4 M Phosphate Buffer Stock 0.9% (w/v) sodium phosphate monobasic (anhydrous) 8.7% (w/v) sodium phosphate dibasic (heptahydrate)
Sucrose Cryoprotectant 30% (w/v) sucrose in 0.1 M phosphate buffer
6% Sucrose 6% (w/v) sucrose in 0.1 M phosphate buffer
2% Paraformaldehyde 2% (w/v) paraformaldehyde in 0.1 M phosphate buffer
3% Agar 3% (w/v) agar in water
2.7% NH4OH Diluted 10x in H2O from purchased strength of 27%
Kodak Fixative A Diluted 10x in H2O from purchased concentration
93
Biocytin Labeling Protocol (courtesy of the Bailey Lab, University of Guelph)
1) Slice brains and perform electrophysiology using biocytin patch solution. (0.3% (w/v) biocytin in regular patch solution)
2) Fix sections overnight (up to three weeks) in 2 mL of Paraformaldehyde solution at 4ºC. (4% PFA in 0.1 M phosphate buffer)
3) Wash sections 3 x 10 min in 2 mL of TBS. (0.1 M Tris base, 0.15 M NaCl, pH 7.5) Fast rocking speed (25rpm)
4) Incubate sections in 2 mL of Endogenous Peroxidase Blocking Buffer for 15 mins at room temperature. (0.5 % (v/v) H2O2 in TBS) Slow rocking speed (15rpm)
5) Wash sections 3 x 10 min in TBS. Fast rocking speed (25rpm)
6) Incubate sections in 1.5 mL of ABC Solution for overnight at room temperature. Slow rocking speed (15rpm)
7) Wash sections 3 x 10 min in TBS. Fast rocking speed (25rpm)
8) Incubate sections in 1.5 mL of DAB / Nickel Solution for 5 mins at room temperature Slow rocking speed (15rpm)
9) Start the reaction by adding 1.5 mL of DAB / Nickel / H2O2 Solution to each well Try 15 min reaction Slow rocking speed (15rpm)
10) Stop the reaction by washing 3 x 10 min in TBS Fast rocking speed (25rpm)
11) Mount sections onto slides.
12) Coverslip using Dako mounting solution (aqueous).
94
TBS (1L) 0.1 M TRIS = 12.114 g 0.15 M NaCl = 8.766 g pH 7.5
4% PFA in Phosphate Buffer 16% PFA Stock: 2.5 mL 0.4 M Phosphate buffer stock: 2.5 mL H2O: 5.0 mL 10 mL
Endogenous Peroxidase Blocking Buffer (Mix Triton X-100 first)
0.25% (v/v) Triton x-100: 25 µL 0.5% (v/v) H2O2: 167 µL of 30% stock TBS: Balance 10 mL
ABC Solution 0.25% (v/v) Triton x-100: 25 µL 0.25% (v/v) ABC Elite Reagent A 25 µL 0.25% (v/v) ABC Elite Reagent B 25 µL TBS Balance 10 mL -Prepare Vector Labs ABC Solution 30 min before using
DAB / Nickel Solution 0.25% (v/v) Triton x-100: 25 µL 0.05% (w/v) DAB (stock is 50 mg/mL) (1 tube) 100 µL 0.04% (w/v) Nickel (stock is 40 mg/mL) (1 tube) 100 µL TBS Balance 10 mL
DAB / Nickel / H2O2 Solution First dilute 10 µL of 30% (v/v) H2O2 with 2,990 µL of TBS (3 mL total) 300 x dilution = 0.1% (v/v) H2O2 in TBS Second dilute 100 µL of the 0.1% solution above into 9.900 mL of DAB / Nickel Solution 100 x dilution = 0.001% (v/v) H2O2 in DAB / Nickel Solution 10 mL
95
APPENDIX II: TWO-DIMENSIONAL PURKINJE CELL TRACINGS
Figure 1: Two-dimensional tracings of leopard gecko Purkinje cells obtained from Golgi-Cox staining. Scale bar=20µm.
96
Figure 2: Two-dimensional tracings of mouse Purkinje cells. Scale bar=20µm.
97
Figure 3: Two-dimensional tracings of chicken Purkinje cells. Scale bar=20µm.
98
Figure 4: Two-dimensional tracings of type 1 (row 1 and 2) and type 2 (row 3 and 4) leopard gecko Purkinje cells obtained from biocytin staining, post-whole cell recording. Scale bar=20µm.
99
APPENDIX III: LOCATION OF LEOPARD GECKO PURKINJE CELLS OBTAINED FROM WHOLE-CELL RECORDINGS
a) Slice Location of Cells
4 Oct 4 3 Nov 1 Mar 10
2
1 Number of Cells
0
Top Top Top MiddleBottom MiddleBottom MiddleBottom
Location in Slice
b) Type by Animal 8 Type 1
6 Type 2
4
2 Number of Cells
0 Oct 4 Nov 1 Mar 10 Animal
Figure 1: (a) Purkinje cell location along the longitudinal axis of the gecko cerebellum in each gecko collected. (b) The distribution of type 1 and type 2 cells in each gecko collected. See
Chapter 3 for details on cell types.
100