Global Reactivation and Targeted Preservation of Mecp2 Expression in a Mouse Model of Rett Syndrome

Global Reactivation and Targeted Preservation of MeCP2 Expression in a Mouse Model of Rett Syndrome

Min Lang

A thesis submitted in conformity with the requirements for the degree of Master of Science

Graduate Department of Physiology University of Toronto

Ubiquitous Reactivation and Targeted Preservation of MeCP2 Expression in a Mouse Model of Rett Syndrome

Min Lang Master of Science, 2012 Department of Physiology University of Toronto

Abstract

Rett syndrome is a neurodevelopmental disorder that is predominately caused by mutations of the MECP2 gene. As neuronal apoptosis is not observed in RTT patients and MeCP2-deficient mice, the neurological deficits may be reversible. To address this, we reactivated MeCP2 expression ubiquitously in MeCP2-deficient mice after symptom onset. Our results showed that life span, behavioural performances, EEG activity, thermoregulation, and daily rhythmic activity were significantly improved after MeCP2 reactivation. Furthermore, the extent of improvement was dependent upon the efficiency of MeCP2 reactivation. To assess the role of the catecholaminergic system in Rett syndrome pathophysiology, we selectively preserved MeCP2 function within tyrosine hydroxylase expressing cells. We observed a significant improvement in the life span of male rescue mice and reduced sudden unexplained death rates in female rescue mice. Behavioural performances and EEG patterns were also significantly improved.

Acknowledgements

I would like to thank Dr. Eubanks for giving me the opportunity to pursue research in his

lab and taking the time to teach me and guide me throughout my Master’s project. Dr. Zhang for

teaching me about the field of EEG and guiding me through my projects. Chiping Wu for

implanting all of the mice that was used for the studies. Richard Logan and Guanming Zhang for teaching me laboratory techniques that was used for my research. Dr. Hampson and Dr. Mount for their guidance and advice throughout my project. Elena Sidorova, Natalya Shulyakova, and

Robert Wither for their help, advice, and support during my Master’s.

III

Table of Contents

Title Page ...... I

Abstract ...... II

Acknowledgements ...... III

Table of Contents ...... IV

List of Figures ...... IX

1. Introduction ...... 1

1.1. General Overview (RTT) ...... 1

1.2. Clinical features of Rett syndrome ...... 2

1.3. Pathophysiology of Rett syndrome ...... 6

1.3.1. Neuropathology...... 6

1.3.2. Neurochemistry ...... 7

1.4. Electroencephalogram abnormalities and epileptic seizures ...... 8

1.5. Autonomic deficits ...... 9

1.6. Rett syndrome and MECP2 ...... 10

1.7. MeCP2 function ...... 11

1.8. Mouse models of Rett syndrome ...... 15

1.9. Gross phenotypes of Rett syndrome mouse models ...... 21

1.10. Targeted deletion of MeCP2 expression ...... 22

1.11. Reversibility of deficits in mouse models of Rett syndrome ...... 26

1.12. Gene therapy ...... 28

1.13. Pharmacological treatments ...... 28

1.14. Rationale and hypothesis ...... 31

1.14.1. Project1: Delayed ubiquitous reactivation of MeCP2 ...... 31

1.14.2. Project 2: Preservation of MeCP2 function in catecholaminergic cells ...... 33

1.15. Project aims ...... 35

2. Materials and Methods ...... 38

2.1. Mice ...... 38

2.2. Western blotting ...... 39

2.3. Immunohistochemistry ...... 40

2.4. Tamoxifen treatment ...... 40

2.5. Electrophysiology data collection ...... 41

2.6. Behavioural assessments ...... 41

2.7. Phenotypic severity scoring ...... 42

2.8. Cell counting ...... 43

2.9. Telemetry probe implantation protocol ...... 43

2.10. Tethered electrode implantation ...... 44

2.11. Electroencephalographic recordings and analysis ...... 44

2.12. Statistics ...... 45

3. Results ...... 46

3.1.1 Rescue of MeCP2 expression in Stop/y,cre mice ...... 46

3.1.2. Restoration of MeCP2 rescues life span and gross phenotypic severity ...... 49

3.1.3. Extent of behavioral rescue is dependent upon MeCP2 reactivation percentage ....52

3.1.4. Epileptiform discharges are significantly attenuated after MeCP2 reactivation .....57

3.1.5. Reactivation of MeCP2 in female MeCP2-deificient mice rescues behavioral performances ...... 60

3.1.6. MeCP2 reactivation improves daily rhythmic activity and thermoregulation in adult female MeCP2-deficient mice ...... 63

3.2.1. MeCP2 is selectively preserved in tyrosine hydroxylase-expressing neurons in the “Rescue” mouse brain ...... 71

3.2.2. Preservation of MeCP2 in catecholaminergic cells extends the lifespan of male MeCP2-deficient mice ...... 77

3.2.3. Preservation of MeCP2 in catecholaminergic cells decreases the rate of sudden unexpected death in female MeCP2-deficient mice ...... 80

3.2.4. Catecholaminergic preservation of MeCP2 improves deficits in ambulatory rate, motor coordination, and anxiety-like behavior in male MeCP2-deficient mice ...... 83

3.2.5. Catecholaminergic preservation of MeCP2 improves the ambulatory and anxiety-like behavioral deficits of adult female MeCP2-deficient mice ...... 86

3.2.6. Preservation of MeCP2 in catecholaminergic cells improves cortical EEG abnormalities in male, but not female, MeCP2-deficient mice ...... 90

3.2.7. Preservation of MeCP2 in catecholaminergic cells improves peak hippocampal theta frequency in male, but not female, MeCP2-deficient mice ...... 96

3.2.8. Preservation of MeCP2 in catecholaminergic cells rescues deficits in hippocampal gamma band oscillatory activity in male, but not female, MeCP2-deficient mice ...... 100

4. Discussion...... 105

4.1. Part 1: Delayed global reactivation of MeCP2 expression ...... 105

4.2. Part 2: Selective preservation of MeCP2 functions in catecholaminergic cells ...... 109

4.3. Future Directions ...... 114

References ...... 118

List of Figures

Introduction

Figure 1. Clinical progression of Rett syndrome ...... 4

Figure 2. MeCP2 function...... 13

Figure 3. Different mouse models of Rett syndrome ...... 19

Figure 4. Different transgenes expressed in different mouse lines ...... 36

Results

Figure 5. MeCP2 is reactivated in Stop,cre mice following tamoxifen treatment ...... 47

Figure 6. Survival and gross phenotypic severity is significantly improved in Stop/y,cre mice after tamoxifen treatment ...... 50

Figure 7. Behavioural performances are improved in Stop/y,cre mice following tamoxifen treatment ...... 54

Figure 8. Epileptiform-like discharge incidence rate is significantly improved in Stop/y,cre mice after tamoxifen treatment ...... 58

Figure 9. Behavioural performances are improved in Stop/+,cre mice following MeCP2 reactivation...... 61

Figure 10. Daily activity is significantly improved in Stop/+,cre mice after MeCP2 reactivation ...... 65

Figure 11. Core body temperature is improved in Stop/+,cre mice after MeCP2 reactivation ...... 67

Figure 12. Temperature and mobility correlation is significantly improved in MeCP2 reactivated female mice ...... 69

Figure 13. MeCP2 is selectively preserved in catecholaminergic neurons of "Rescue" mice ...... 73

Figure 14. MeCP2 expression is not preserved in non-catecholaminergic neurons ...... 75

VII

Figure 15. Survival and gross phenotypic behavior are improved in male and female "Rescue" mice ...... 78

Figure 16. The phenotypic severity score of female MeCP2+/- mice does not correlate with the time of their sudden and unexpected death ...... 81

Figure 17. Behavioral performances are improved in male "Rescue" mice ...... 84

Figure 18. Behavioral performances are improved in female "Rescue" mice ...... 87

Figure 19. The incidence rate of cortical epileptiform discharge activity is reduced in male "Rescue" mice ...... 92

Figure 20. The incidence rate of epileptiform discharge activity is not improved in female "Rescue" mice ...... 94

Figure 21. The peak hippocampal theta frequency and the total hippocampal gamma activity power are significantly improved in male "Rescue" mice ...... 97

Figure 22. The peak hippocampal theta frequency and the total power of hippocampal gamma activity are not improved in female "Rescue" mice ...... 102

VIII

Chapter 1

Introduction

1.1. General overview

“Ich bin Rett, ich bin Rett, ich will jetzt mit Ihnen sprechen” – (English – “It’s Rett, It’s

Rett, I want to talk to you”), shouted Andreas Rett as he rushed towards Bengt Hagberg at the

International Association Scientific Study Mental Deficiency conference in Toronto. Andreas

Rett was the pediatric neurologist who first described the syndrome in 1966. It went relatively

unnoticed, however, until Bengt Hagberg shared his clinical observation in 1983. Rett syndrome

(RTT) is now recognized as a leading cause of mental retardation in females with a prevalence rate of 1 in 10,000 live female births (Hagberg et al., 1985; Matijevic et al., 2009; Rett, 1966).

RTT is characterized by apparently normal development up to 6-18 months of age,

followed by a period of regression and loss of previously acquired skills. The patients lose

purposeful hand skills and develop stereotypical hand movements. Additional symptoms include

absence of speech, autistic-like features, impaired sleep patterns, cold feet and/or hands,

respiratory dysfunction, bruxism, back deformities, motor impairments, and seizures (Smeets et al., 2012; Weaving et al., 2005; Williamson and Christodoulou, 2006). After the stage of regression, which varies from weeks to years, the conditions stabilize and the patients usually survive into adulthood (Matijevic et al., 2009). The rate of sudden unexplained death is significantly greater in RTT patients than controls of the same age. (Matijevic et al., 2009). The reasons of these deaths may be associated with sudden respiratory failure, abnormal cardiac arrhythmia, or seizures (Asthana et al., 1990; Hagberg and Witt-Engerstrom, 1986; Matijevic et al., 2009).

Mutations in the MECP2 gene cause > 90% of typical RTT cases (Amir et al., 1999; Neul

et al., 2008). The protein product, MeCP2 (methyl-CpG binding protein 2), binds to methylated

dinucleotides and act as a transcription repressor (Amir et al., 1999; Chahrour and Zoghbi, 2007).

However, recent evidence has suggested that MeCP2 may also activate the transcription of several genes (Chahrour et al., 2008). In addition, mutations involving the FOXG1 and CDKL5

gene make up less than 10% of RTT cases. The congenital variant of RTT is related to mutations

of the FOXG1 gene and the infantile seizure variant is related to mutations of the CDKL5 gene

(Samaco and Neul, 2011; Smeets et al., 2012). It is still unclear whether the gene products of

FOXG1 and CDKL5 share a convergent pathway as MeCP2 or whether it induces RTT

phenotypes in a different pathway altogether.

The lack of observable signs of atrophy, degeneration, or demyelination differentiates

RTT from neurodegenerative diseases (Armstrong et al., 1999). The absence of

neurodegeneration and the lack of progressive brain weight loss support the hypothesis of RTT is

associated with neurodevelopmental arrest. Mouse models of RTT have been generated to

explore avenues of treatment and to investigate the reversibility of neurological symptoms.

1.2 Clinical features of Rett syndrome

RTT is characterized by normal development for the first 6-18 months of age. The

affected individuals meet developmental milestones with no obvious signs of developmental

disturbances (Smeets et al., 2003). However, retro-analysis of home videos reveals hypotonia

and inadequate hand-eye coordination patterns in some RTT girls before 6 months of age

(Smeets et al., 2012). RTT is divided into four developmental stages. During the first

preregression stage, subtle signs and symptoms begin to manifest, such as reduced eye contact, less demand for attention, and delays in crawling and standing. Overall, gross development appears to be normal.

The following regression stage usually occurs between 1 and 4 years of age (Glaze, 2005).

The rapid loss of acquired skills occurs acutely and the period of regression can last for days or

months. Affected individuals show an obvious decline in communicative and motor skills.

Cognitive deficiency also becomes apparent. Exploratory behaviour is lost and sleep pattern

becomes disturbed. Panting, hyperventilation, and other respiratory abnormalities also begin to

manifest. Seizures may become present during this period.

Conditions plateau and stabilize during the third post-regression stage of RTT. RTT girls

begin to display a loss of purposeful hand movement and stereotypic hand wringing becomes

prominent. Eye contact returns and irritability is slightly improved. Seizures (epilepsy) become a

common feature that requires medical treatment. Girls with milder phenotypes may have

preserved purposeful hand use and speech. Many affected individuals remain in this stage for the

rest of their lives.

The final stage of RTT is marked by motor deterioration. RTT patients develop motor

weakness, rigidity, and scoliosis. Parkinson-like features such as ataxia, dystonia, and

hypomimia becomes prominent. Many RTT patients survive into their 40s and 50s through

constant care and assistance, however, sudden unexpected death is often reported (Smeets et al.,

2012).

Figure 1

Figure 1. Clinical progression of Rett syndrome.

RTT girls usually develop normally up to 6-18 months of age. Development then begins to stagnate and learning is delayed. This is followed by the second regression stage, in which previous acquired skills are lost, cognitive deficits become obvious, motor impairments become apparent, and stereotypic hand movements begin to manifest. Symptoms stabilize during clinical stage III and seizures onset is common during this period. The last clinical stage is characterized by motor deterioration. Patients develop Parkinson-like motor impairments, lose the ability to walk, and often become wheel chair bound.

1.3 Pathophysiology of Rett syndrome

1.3.1 Neuropathology

The brain anatomy of RTT girls is grossly normal, however, the brain size and weight are

well below that of age-matched controls (Armstrong, 2005). At birth, though, the head

circumference of RTT girls is relatively normal but decelerated head growth begins at 2-3

months of age (Armstrong, 2005). Atrophy and progressive decline of brain weight is not

observed, which is consistent with impaired brain growth (Armstrong et al., 1999). The decrease

in the rate of brain growth, however, is not uniform as the cerebral hemispheres are affected

more than cerebellar regions. Specifically, brain volume does not increase at normal rates in the

prefrontal, anterior, and the posterior temporal regions, while posterior temporal and occipital

regions remain relatively preserved (Reiss et al., 1993; Subramaniam et al., 1997). Magnetic

imaging studies have also shown a significant decrease of grey matter in RTT brains (Reiss et al.,

1993; Subramaniam et al., 1997). Though there are numerous alterations within RTT brains, the

effects are generally subtle with no overall decrease in the number of neurons. Neuronal packing

density is increased whereas synaptic density, dendritic complexity, and neuronal size are

significantly diminished in the RTT brain (Armstrong, 1997; Bauman et al., 1995; Jellinger et al.,

1988; Naidu, 1997). Cortical hyperexcitability as well as giant amplitude somatosensory and

visual evoked potentials are reported in RTT patients (Glaze, 2005; Guerrini et al., 1998;

Yoshikawa et al., 1991). These alterations likely contribute to the EEG abnormalities and clinical

seizure onset in RTT (Cooper et al., 1998; Glaze, 2005; Niedermeyer et al., 1997).

1.3.2 Neurochemistry

During neural development, neurotransmitter/neuromodulators are critical for guiding

neuronal migration, formation of synapses, and formation of neural networks (Berger-Sweeney et al., 1998; Misgeld et al., 2002). Examinations of the cerebrospinal fluid and autopsy of RTT patients revealed a multitude of neurotransmitter abnormalities, including dopamine, serotonin, noradrenaline, glutamate, GABA, substance P, and acetylcholine (Armstrong, 2005; Gadalla et al., 2011). Cerebrospinal fluid collected from RTT girls shows an increase in glutamate levels and an increase in glutamate receptor density in the cerebral cortex has been reported (Armstrong,

2005). GABAergic dysregulation, including reduced GABA content as well as reduced glutamic acid decarboxylase 1 and 2, is evident in brains from young female RTT patients (Johnston et al.,

2005; Chao et al., 2010). Disruption of glutamate and GABAergic systems may underlie the

imbalance between excitatory and inhibitory transmission in RTT and contribute to seizure

genesis (Gatto and Broadie, 2010).

Bioamine deficits underlying RTT-like phenotypes have been suggested since 1985 as many symptoms overlap with disorders involving dopamine or noradrenergic deficits (Nomura et al., 1985). Currently, there are contradicting findings as Zoghbi et al. reported a significant reduction in homo-vanillic acid (catecholamine metabolite) and 3-methoxy-4-hydroxy- phenylethylene glycol (noradrenaline metabolite) in the CSF of RTT girls, whereas Perry et al. reported no significant reduction of bioamine levels in RTT girls (Perry et al., 1988; Zoghbi et al.,

1989). More recent findings showed a significant reduction of bioamine levels in the substantia nigra of RTT patients and reduced dopamine and its metabolites in the cortex and basal ganglia

(Wenk, 1996). Dopamine receptors, D2 receptors in particular, have also been reported to be

increased during early years (4-15 years of age) and decreased during adulthood (15-39 years of

age) in RTT patients (Chiron et al., 1993; Naidu et al., 2001). The alteration in receptor levels is

consistent with the developmental stages of RTT. It is important to note the differential age and

phenotypic severity of the patients in the bioamine studies, which may partially explain the conflicting findings (Lekman et al., 1990; Percy, 1992). Further investigation is required to establish the relationship between age, phenotypic severity, and neurochemical changes in RTT

(Armstrong, 2005).

1.4. Electroencephalogram abnormalities and epileptic seizures

Epilepsy is a common co-morbidity observed amongst RTT patients and has a prevalence rate of 60-90% (Glaze et al., 2010). Many reported seizure occurrences, however, are non-

epileptic in origin as parents and caregiver mistake breath holding, hyperventilation, blank stares

or vacant episodes, and motor abnormalities for seizures (Garofalo et al., 1988; Niedermeyer and

Naidu, 1998; Witt Engerstrom, 1992). Seizures negatively affect the patient’s quality of life and bring difficulty for caregivers (Bahi-Buisson et al., 2008). The onset of seizures is usually between 2 to 3 years of age and during clinical Stages II and III (Witt Engerstrom, 1992). The severity and onset of seizures appears to diminish after puberty (Glaze et al., 2010). Seizures that develop before 1 year of age tends to be more severe and typically results in intractable epilepsy

(Steffenburg et al., 2001). Furthermore, early seizure onset occurs less frequently within RTT patients affected by MECP2 mutations, and is more frequent within atypical RTT cases (Glaze et al., 2010). The infantile seizure variant of RTT is characterized by early seizure onset prior to the development of clinical RTT features (Aicardi, 1997). RTT associated seizures are often difficult

to treat. Different types of seizures may manifest, including complex partial, tonic-clonic, tonic,

and myoclonic (Glaze, 2005; Steffenburg et al., 2001). Commonly used antiepileptic drugs

include valproate, lamotrigine, and carbamazepine. However, RTT-associated seizures remain poorly controlled (Glaze, 2005).

EEG abnormalities are reported in most RTT cases, including individuals without a history of seizures (Glaze et al., 1998). Rhythmic spike wave discharge events associated with absence episodes are characteristics of generalized electrographic seizures in RTT patients

(Glaze, 2005). It has been suggested that these seizures originate from limbic structures (Boison,

2012). Hyperexcitability is observed within the hippocampus and the cortex of RTT girls.

Alternate hypotheses implicate an imbalance between excitability and inhibitory tone in the RTT brain as the source of seizure genesis (Dani et al., 2005). Further investigation is required to deduce the origin of these EEG abnormalities and epileptic seizure events in order to develop more effective methods of treatment.

1.5. Autonomic deficits

Cold extremities, anxiety-like behaviour, respiratory dysfunction, and cardiac abnormalities are prevalent features of RTT (Hagberg et al., 2002). These autonomic dysfunctions may contribute to sudden unexplained deaths in RTT patients (Glaze, 2005).

Electrocardiogram studies on RTT girls have shown that the Q-T intervals are prolonged compared to age-matched controls (Ellaway et al., 1999; Sekul et al., 1994). The prevalence of the Q-T interval alteration increases during progressive stages of RTT: 36% of patients in Stage

II, 38% of patients in Stage III, and 50% of those in Stage IV (Sekul et al., 1994). Heart rate variability and high frequency power in ECG recordings were also found to be diminished in a

study of 25 RTT patients (Johnsrude et al., 1995). The autonomic function, including cardiac

function and respiration, is modulated by the parasympathetic and sympathetic nervous system

through acetylcholine and norepinephrine release, respectively. The observed autonomic

impairments in RTT may be caused by an imbalance between parasympathetic and sympathetic

input (Glaze, 2005).

Hyperventilation, apneas, and breath holding are commonly observed amongst RTT girls

(Glaze, 2005). These symptoms are more pronounced during wakefulness and become less

severe during sleep (Glaze, 2005). Julu et al. found a decrease in cardiac vagal tone and sensitivity to baroreflex stemming which may contribute to the irregular breathing patterns (Julu

et al., 2001). Uncoupling of heart rate and breathing patterns are also observed during day time and night time (Weese-Mayer et al., 2006). The imbalance between sympathetic and parasympathetic tone, as well as compromised autonomic reflexes, are likely the underlying cause of these symptoms and sudden deaths in RTT patients (Glaze, 2005).

1.6. Rett syndrome and MECP2

MECP2 is critical for normal neurodevelopment and disruption of its expression causes a

wide array of neurological deficits. In 1999, Amir et al identified mutations of the gene, MECP2,

as the predominate cause of typical RTT (>90%) (Amir et al., 1999). Altered MeCP2 expression

is also observed in other neurodevelopmental disorders, including X-linked mental retardation, encephalopathy, Angelman’s syndrome, and autism (Gonzales and LaSalle, 2010). MECP2 is located on the X-chromosome, Xq28 region (Guy et al., 2011), and thus exhibits an X-linked inheritance pattern. Most MECP2 mutations, however, transmit paternally through de novo

mutations in male germ cells (Girard et al., 2001). Although more rare, MEPC2 mutations can be

transmitted maternally through mildly affected maternal carriers (Gonzales and LaSalle, 2010).

Heterozygous mutations of MECP2 in females leads to RTT, whereas MECP2 mutations

in males lead to infantile encephalopathy and early death (Weng et al., 2011). Most RTT females

show a mosaic expression of MeCP2 throughout the body due to random X-chromosome

inactivation, which contributes to the variability of phenotypic severity in RTT patients

(Hoffbuhr et al., 2002). In murine models of RTT, the mutant allele seems to be favourably inactivated (Braunschweig et al., 2004). Recent evidence, however, have shown that X-

chromosome inactivation is insufficient to explain the variability in phenotypic severity and

therefore, other genetic modifiers must be at play (Takahashi et al., 2008). More than 300

mutations of the MECP2 gene have been related to RTT, and certain types of mutation seem to

correlate with the level of phenotype manifestation. Nonsense mutations, which ablate normal

MeCP2 function, are found in severe cases of RTT, and truncated mutations, which may

preserve partial MeCP2 function, are found in milder manifestations (Bebbington et al., 2010;

Huppke et al., 2000; Smeets et al., 2005).

1.7. MeCP2 function

MECP2 gene encodes for MeCP2, a 53 kDa nuclear protein belonging to the methyl-CpG

binding domain family of proteins (D'Esposito et al., 1996; Quaderi et al., 1994). MeCP2 exists

as two isoforms, e1 and e2, with different N-termini (Gadalla et al., 2011). Both isoforms,

however, contains a nuclear localization signal, CpG binding domain, and a transcriptional

repression domain that is important for formation of the repressor complex. MeCP2 has been

shown to interact with corepressors such as Sin3a, cSki, NcoR, and CoREST (Guy et al., 2011).

Histone deacetylases, HDAC1 and HDAC2, can be recruited to the repressor complex which results in condensation of chromatin structure and hindering of gene transcription (Samaco and

Neul, 2011). Though MeCP2 is generally considered to be a global repressor, genetic studies have shown only subtle changes in gene transcription profiles. This suggests that MeCP2 may function more than a proximal repressor of methylated genes. Alternate evidence has shown

MeCP2 to interact with CREB, a transcription activator (Chahrour et al., 2008). Further, MeCP2 have also been shown to interact with RNA splice site regulators such as RNA-binding protein Y box-binding protein 1 (Young et al., 2005). Though MeCP2 is characterized by a methyl-CpG binding domain, ChIP assays have shown MeCP2 binding to non-methylated DNA sequences, although it remains unclear whether this is affected by other intermediary proteins (Guy et al.,

2011).

MeCP2 is expressed throughout the body but it is most abundantly found in post-mitotic neurons (Amir et al., 1999). Glia cells of the neural system also express MeCP2, albeit at significantly lower levels (Ballas et al., 2009). MeCP2 expression varies during different stages of development, with low expression levels during embryogenesis and increasing expression levels during neuronal development and synaptogenesis (Guy et al., 2011). Further, the absence of MeCP2 does not seem to affect neuronal precursor formation, migration, or development

(Kishi and Macklis, 2004). These findings suggest that MeCP2’s primary function is maintaining neuronal maturity rather than neuronal development (Guy et al., 2011). The observation that

RTT patients show normal development up to 6-18 months of age, followed by a period of regression, is consistent with the proposed function and expression profile of MeCP2.

Figure 2

Figure 2. MeCP2 function

MeCP2 functioning as a transcription repressor. MeCP2 binds to methylated CpG islands and recruits co-repressor complexes, including Sin3a and histone deacetylases. Chromatin condenses and transcription of genes is inhibited.

1.8. Mouse models of Rett syndrome

Our understanding of RTT remained limited due to the variability of phenotypic severity,

different forms of MECP2 mutations, different variants of RTT, age-dependent factors, and environmental effects. These factors complicate findings and at times produce conflicting results.

Several animal models of RTT have been produced that allow investigators to properly control conditions to reach more definite conclusions (Chen et al., 2001; Guy et al., 2007; Guy et al.,

2001; Zoghbi, 2005). The utilization of animal models has provided more insight into the mechanisms underlying the development, pathophysiology, and alterations of RTT. However, it is important to remember there are fundamental difference between the anatomy, biochemistry, and morphology in animal models and human patients (Peters et al., 2007). These differences occur at the molecular level to phenotypic behaviours. Although animal models can increase our understanding of different disorders, there are also limitations to the information that we can extract.

After uncovering that MECP2 mutations are the underlying cause of most RTT cases, several mouse models with MECP2 mutations are generated to recapitulate the human condition

(Chen et al., 2001; Guy et al., 2007; Guy et al., 2001; Shahbazian et al., 2002). MeCP2-null male mice were generated from both Dr. Jaenisch and Dr. Bird’s group (Chen et al., 2001; Guy et al.,

2001). The Jaenisch mouse line, MeCP2Jae was produced by deleting exon 3 of the MECP2 gene, whereas both exon 3 and 4 are both deleted in the Bird’s mouse line, MeCP2tm1Bird. Both mouse models recapitulate many features of RTT. These mice develop motor impairments, tremors, respiratory dysfunction, and stereotypic limb movements. Similar to the human condition, these mice seems to develop normally up to 6 weeks of age, after which observable symptoms quickly develop resulting in early lethality (~10-20 weeks of age). Chen et al. also generated a brain-

specific MeCP2 knockout mouse line through the use of a Nestin-cre transgene (Chen et al.,

2001). Mice lacking MeCP2 only in the brain are phenotypically similar to MeCP2-null mice with identical symptom development, suggesting that RTT is mainly a neurological disorder and the function of MeCP2 is primarily within the brain.

Mouse models mimicking the different human MECP2 mutations in humans have been generated. Dr. Zoghbi’s group produced a transgenic mouse model of RTT (MeCP2308), which

expresses a truncated mutation, MeCP2-308, that is commonly found in RTT patients

(Shahbazian et al., 2002). In this mouse model, a premature stop codon was inserted after codon

308 of the MECP2 gene. These mice develop milder symptoms than MeCP2-null mice but show

RTT-like phenotypes, including stereotypic limb movements, and impairments in social and

spatial memory (Moretti et al., 2006; Moretti and Zoghbi, 2006). A mouse model expressing

another truncated mutation of MECP2, MeCP2-168, was also produced (Lawson-Yuen et al.,

2007). MeCP2168 male mice develop milder symptoms than MeCP2-null mice but typical RTT-

like phenotypes are still present. MeCP2168 female mice develop normally up to 6 months of age

and survive past 1 year of age. Another mouse model is the MeCP2A140V, which possess a

point mutation that causes the 140th alanine to be replaced by valine (Jentarra et al., 2010). Only

0.6% of RTT patients exhibit this form of mutation and show a distinct variant phenotype of

RTT that resembles X-linked mental retardation. A more common form of MECP2 mutation,

T158M, encompasses 10% of all RTT cases (Goffin et al., 2012). A mouse line is generated with

the same mutation, which substitutes a threonine with an alanine at the C-terminus of the MBD,

and clinical RTT manifestations are observed in this transgenic mouse model (Goffin et al.,

2012). Lastly, a complete MeCP2-null mouse line, MeCP2Tam is produced by disrupting the

MBD of MECP2, which essentially ablates the function of the MeCP2 protein (Pelka et al.,

2006). RTT-like learning and cognitive deficits are features of this mouse model.

In 2007, Guy et al. engineered a new mouse model of RTT, MeCP2tm2Bird, by inserting a

floxed-stop cassette between exon 2 and 3 of the MECP2 gene (Guy et al., 2007). These mice are

phenotypically similar to MeCP2tm1Bird mice. The floxed-stop cassette enables conditional

reactivation of the MECP2 gene through Cre recombinase activity. By crossing MeCP2tm2Bird

with mouse lines expressing different transgenes of Cre recombinase driven by various

promoters, MECP2 reactivation can be controlled temporally and spatially. This has led to

several reactivation and rescue studies that have provided further insight on the pathophysiology

and reversibility of RTT.

Female counterparts of the different mouse models are also generated. MECP2 expression is disrupted in each model as described previously, however, these female mice still possess a functional MECP2 allele (Guy et al., 2001). X-chromosome inactivation occurs in every cell. Thus, depending upon the inactivation profile, the phenotypic severity can be significantly skewed. Female mouse models of RTT develop, on average, milder symptoms than male counterparts and are phenotypically normal up to 6 months of age (Chen et al., 2001; Guy et al., 2001). Tremors, breathing abnormalities, motor impairments, anxiety-like behaviour, and cognitive deficits develop progressively throughout adulthood. Unlike the human condition, female MeCP2-deficient mice are more prone to becoming obese (Guy et al., 2001).

Female mouse models are more clinically relevant, as RTT is a disorder that almost exclusively affects female humans. Most animal work, however, has been conducted on hemizygous male mice with a disrupted MECP2 allele (Ricceri et al., 2008). Male models of

RTT eliminate the confounding effects of phenotypic variability due to X-chromosome inactivation, and represent the more severe cases of RTT. Furthermore, the more aggressive symptom progression in male mice allows for quicker and more efficient studies, as female mice do not develop overt symptoms until several months of age (Chen et al., 2001; Guy et al., 2001).

The use of female mouse models, however, should be incorporated as there are fundamental differences between severe MeCP2 deficiency in male mice and partial MeCP2 deficiency in both female mouse models and RTT patients.

Figure 3

Figure 3. Different mouse models of Rett syndrome.

MeCP2 expression is knocked out of using different genetic methods. Early lethality is a feature of the male mutant mice. Female mice of the models display normal lifespan and survive well into adulthood.

1.9. Gross phenotypes of Rett syndrome mouse models

The cardinal feature of RTT is the normal development during early life followed by periods of regression and stabilization. Mouse models of RTT recapitulate this feature. Male mouse models, however, develop symptoms significantly earlier than female mice and are more severely (Chen et al., 2001; Guy et al., 2001).

Male and female mice of the MeCP2tm1Bird develop RTT-like features at different ages with different rates of symptom progression (Guy et al., 2001). Male MeCP2tm1Bird mice develop symptoms between 4th and 7th week of age. The symptoms progress rapidly along with a decline in body weight, which leads to early death at approximately 8 weeks of age. Gait impairment, hind limb clasping, and misalignment of the jaw are commonly observed. In contrast, female

MeCP2tm1Bird mice develop symptoms much later, at approximately 12th weeks of age, and usually survive well into adulthood. Female mice develop milder symptoms, including deficits in inertia and hindlimb clasping. Other features such as respiratory abnormalities and tremor are only present in a portion of the heterozygous females by 9 months of age.

The MeCP2Jae mouse model is very similar to the MeCP2tm1Bird mice (Chen et al., 2001).

In male MeCP2Jae mice, symptoms such as gait impairment can be detected as early as 4 weeks of age. Breathing irregularities, body tremors, and shaking paws are subsequently observed. The body weight of the male mice is decreased compared to wild-type littermates. The heterozygous female mutant mice develop normally up to 4 months of age. Reduced activity, hindlimb clasping, and respiratory deficits are observed after 6 months of age. Female mutant mice also display elevated weight gain.

In the MeCP2Tam mouse line, mutant male mice do not survive past 20 weeks of age

(Pelka et al., 2006). Gait impairment, dishevelled fur, breathing difficulties, and gait deficits are

reported during development. Seizures have also been reported in this model. Female mutant

mice show a delayed onset of these symptoms at around 13 weeks of age.

Mice expressing the MeCP-308 mutation also recapitulate many RTT-like phenotypes

(Shahbazian et al., 2002). Male MeCP308 mice develop normally until 6 weeks of age. Tremors soon develop and become visibly apparent. Abnormal posture and dishevelling of fur becomes noticeable between 5 and 8 months of age. Myoclonic jerks and seizure events are observed in some mice. Handwringing movement, similar to the stereotypic hand movements in RTT patients, motor deficits, and respiratory dysfunction are reported in mutant MeCP2308 male mice.

Female MeCP2308 mice develop symptoms at approximately 6 months of age and exhibit milder phenotypes than the mutant male mice.

1.10. Targeted deletion of MeCP2 expression

MECP2 is expressed in all tissues but most abundantly in the brain. The brain can be divided into different sectors, each with distinct as well as overlapping functions. Through targeted deletion studies, we gained valuable knowledge as to which brain regions are critical for the pathogenesis in RTT and how different neuronal groups may contribute to different phenotypic impairments.

A brain specific deletion of MeCP2 was first produced by Chen et al. This study is consistent with notion that RTT is primarily a disorder of neurodevelopment and MeCP2 function is most critical within the central nervous system. In 2006, Gemelli et al. generated a

forebrain MeCP2 knockout mouse model through the use of a CamK-Cre93 transgene (Gemelli

et al., 2006). These mice developed many RTT-like phenotypes with however, normal life span.

Though gross phenotypes were relatively normal in the forebrain knockout (KO) mice, impairments were reported in the rotarod test, anxiety-like behaviour tests (elevated plus-maze and open field test), cue-dependent fear conditioning, and social interaction tests. Taken together, the forebrain likely contributes to a wide spectrum of RTT related symptoms (Gemelli et al.,

2006).

The basolateral amygdala is known to be central for the regulation of emotion and RTT patients are often reported to exhibit mood swings and anxiety-like behaviour (Sansom et al.,

1993). It is hypothesized that MeCP2 mutations disrupt basolateral amygdala function which contributes to the reported symptoms. Adachi et al. selectively ablated MeCP2 expression in the

BLA by targeted injection of an adeno-associated virus that express Cre recombinase (Adachi et al., 2009). Consistent with the hypothesis, the targeted MeCP2 KO mice exhibited heightened anxiety-like behaviour and impaired cue-dependent fear learning. Motor and social interaction impairments were not observed. This finding links basolateral amygdala deficits to anxiety and fear learning impairments in RTT, but at the same time, it demonstrates its lack of effect on motor performance and social interaction.

Other neuronal populations that may be involved in the pathophysiology of RTT include the catecholaminergic and serotoninergic system (Nomura et al., 1985; Wenk and Hauss-

Wegrzyniak, 1999). Clinical features of RTT, such as ataxia, dystonia, and respiratory irregularities, overlaps with disorders involving dopaminergic or noradrenergic deficits (Fahn,

2008; Segawa, 2001). Heightened anxiety, mood swings, and aggression observed in RTT may be attributed to serotonergic abnormalities (Gordon and Hen, 2004; Lucki, 1998; Popova, 2008).

Samaco et al. selectively ablated MeCP2 function in catecholaminergic and serotonergic cells by

crossing transgenic mice that express loxp flanked MECP2 allele with TH-cre and PC12 ets

factor 1 (PET1) expressing mouse lines, respectively (Samaco et al., 2008). Tyrosine hydroxylase specific MeCP2 KO mice displayed reduced levels of dopamine and norepinephrine relative to wild-type controls. Reduced locomotive activity was also reported. PET1-specific

MeCP2 KO mice displayed reduced levels of serotonin in addition to altered pattern of social interaction and increased aggression. Both KO mouse models did not show early lethality, suggesting that MeCP2 deficiency within these systems is not necessary for normal life span.

The findings from this study demonstrate the importance of MeCP2 in the serotonergic and catecholaminergic systems. More generally, the lack of MeCP2 disrupts neuronal systems and causes subsequent alterations in the phenotypes and behaviour which the system normally regulates.

Recently it has been shown that MeCP2 is required within the HoxB1-domain as the lack of MeCP2 within the region caused respiratory dysfunction, autonomic irregularities, motor impairment, and premature death (Ward et al., 2011). The HoxB1-specific MeCP2 KO mice displayed impairments in motor coordination tasks, decreased heart rate, and increased respiration rate during hypoxia challenge. The specificity of the deficits confers with the normal function of the brain regions targeted for MeCP2 ablation. Thus the deficits of the HoxB1- domain likely contribute to the motor and autonomic impairments associated with RTT.

Excitatory and inhibitory neurotransmission balance is altered in RTT patients

(Monteggia and Kavalali, 2009). The GABAergic system is the main source of inhibition in the

brain and whether it is compromised by a loss of MeCP2 or if it had any behavioural

consequences remained poorly investigated. A mouse line with MeCP2 specifically knocked out

in GABAergic neurons, viaat-MeCP2-/y, was generated (Chao et al., 2010). These mice develop repetitive behaviours, motor impairment, compulsive grooming, learning deficits, abnormal EEG hyperexcitability, respiratory dysfunction, and premature death. Additionally, mice with MeCP2 deficiency only in forebrain GABAergic neurons develop a subset of the deficits, including altered sensorimotor gating and arousal, impaired motor function, increased sociability, and repetitive behaviours. Quantal size release of GABA is decreased and glutamate decarboxylase 1 and glutamate decarboxylase 2 levels are reduced in MeCP2-deficient GABAergic neurons

(Chao et al., 2010). The GABAergic system regulates synaptic signaling throughout the entire brain, thus it is not surprising that MeCP2 deficiency in such a system produces a large array of

RTT-like phenotypes. These findings highlight the importance of MeCP2 in GABAergic neurons and that disturbances of this system are central to the pathogenesis of RTT.

Though studies of RTT have focused on neuronal alterations, MeCP2 is also expressed, albeit at very low levels, in glia cells (Ballas et al., 2009). MeCP2-deficient glia exerts negative non-cell autonomous effects on neuronal properties, including dendritic morphology.

Specifically, astrocytes lacking MeCP2 expression display abnormalities in BDNF and neuronal dendritic regulation (Maezawa et al., 2009). By crossing mice expressing a floxed MECP2 allele with mice expressing hGFAPcreT2 transgene, Lioy et al. (2011) generated mice that have

MeCP2 selectively ablated from astrocytes. These mutant mice develop many RTT-like phenotypes such as breathing irregularities, hindlimb clasping, and decreased body weight.

However, lifespan, locomotive activity, and anxiety-like behaviour are not affected. The findings of this study revealed a non-neuronal effect in the pathogenesis of RTT. It remains to be investigated whether MeCP2 is required for proper function in other cell types and whether these deficits contribute to the pathophysiology of RTT.

All of these targeted knockout studies have demonstrated the cell and tissue-autonomous function for MeCP2 within different brain regions. Local MeCP2 deficiency only produced deficits that are normally related to the affected region. These deletion studies are complemented with reactivation/rescue studies that further describe how different neural systems contribute to the various symptoms of RTT.

1.11. Reversibility of deficits in mouse models of Rett syndrome

An important finding in RTT patients and mouse models of RTT is the lack of neuronal death and apoptosis (Johnston et al., 1995). Rather the neurons appear to be preserved in an immature state. These reports lend to the hypothesis that the phenotypes of RTT may be reversible (Guy et al., 2007). As RTT is mainly caused by mutations of the MECP2 gene, it may be possible to achieve phenotypic rescue by re-introduction of the MeCP2 function. In 2007,

Adrian Bird’s group generated a global inducible rescue model of RTT (Guy et al., 2007). This rescue mouse model was generated by crossing the MeCP2TM2bird mouse line with a mouse line

that expressed the ESRcre construct, which encodes Cre recombinase enzyme fused with a

mutated ligand binding domain of the estrogen receptor). Tamoxifen (an estrogen receptor

antagonist) treatment allowed the translocation of Cre recombinase into the nucleus, which in

turn reactivated MeCP2 expression through excision of the floxed-stop cassette. Guy et al. (2007) reactivated MeCP2 expression in the double transgenic rescue mice after the development of symptoms. MeCP2 reactivation significantly improved life span and gross phenotypic severity in the male rescue mice and reversed the gross phenotypes of the female rescue mice. In addition, long-term potentiation was also significantly improved in the female rescue mice. This study was

26 the first to demonstrate the reversibility of RTT and that the alterations of the neural networks are not irremediable (Guy et al., 2007). Rather, the systems may be preserved in an immature state that still consisted of the necessary machinery for further development. Robinson et al.

(2012) further extended upon this study by demonstrating that specific sensorimotor performances and neuronal morphology are also improved in the global MeCP2 rescue. However, these studies focused on male MeCP2-deficient mice and the extent of phenotypic rescue in the female mutant mice remains unclear.

Following the global rescue studies, several groups have generated targeted rescue models of RTT. Jugloff et al. (2008) reintroduced MeCP2 into the forebrain of MeCP2 deficient female mice. Exploratory and rearing activity was significantly improved. Further, reactivation of MeCP2 expression in glia cells partially rescued behaviour deficits, breathing abnormalities, and neural anatomical abnormalities (Lioy et al., 2011). This was an important finding as it demonstrates that MeCP2 is critical for non-neuronal cells and reactivation of MeCP2 expression within astrocytes can benefit the neuronal network and improve several phenotypes of RTT.

Lastly, reactivation of MeCP2 in the HoxB1 domain led to significant autonomic improvements, and most importantly, rescued early lethality (Ward et al., 2011). These genetic rescue mouse models of RTT complement the knockout studies and further demonstrate the role of each brain region in the pathophysiology of RTT. The symptoms of RTT are various and are often affected differentially. Understanding the neural origin of the specific symptoms will allow for more selective treatment strategies tailored to individual patients.

1.12. Gene therapy

Genetic rescue studies in mouse models of RTT suggest gene therapy as a viable option

(Cobb et al., 2010). The neuronal network seems to remain intact even in the absence of MeCP2.

The methylated DNA targets of MeCP2 are suggested to be preserved, therefore as newly synthesized MeCP2 become present, gene transcription should return to normal (Lewis et al.,

1992; Nan et al., 1997). Further, studies have shown that overexpression of MeCP2 leads to the development of motor impairments and tremors, whereas modest overexpression of MeCP2 levels enhances motor performance and reduces anxiety-like behaviour (Collins et al., 2004).

The dosage at which MeCP2 is reintroduced becomes critically important. Distinct brain regions have been shown to contribute to different symptoms of RTT, and selective rescue of specific neuronal systems can improve the deficits associated with the network. Genetically targeting specific brain regions is a potential avenue for clinical investigation.

1.13. Pharmacological treatments

As current gene therapy methods cannot provide the level of control and specificity required for MECP2 reintroduction, many are seeking alternative pharmacological methods of treatment. Several studies have identified molecules and systems that are affected downstream of

MeCP2 function. However, improving downstream pathways is unlikely to completely compensate for the absence of MeCP2. Several candidates for therapeutic treatments are suggested.

Brain derived neurotrophic factor (BDNF) is essential for neural development and its levels are significantly decreased in RTT brains (Acheson et al., 1995; Chen et al., 2003;

Martinowich et al., 2003). Mouse studies support pharmacological methods of enhancing BDNF

activity in RTT as increased BDNF level positively affects MeCP2-null mice by delaying onset

of symptoms, improving life span, and rescuing electrophysiological abnormalities (Chang et al.,

2006). Ogier et al. (2007) tested ampakine CX546, an enhancer of glutamergic AMPA receptor which in turn increases BDNF levels, in MeCP2-null mice. Respiratory function was restored in the mouse, however, other deficits remains unexplored.

Another candidate for therapeutic treatment of RTT is the insulin-like growth factor 1.

IGF1 is known to be involved in neuronal maturation and synaptic development (Gadalla et al.,

2011). Itoh et al. (2007) found elevated levels of IGF3 in RTT patients and MeCP2-null mice. It has been hypothesized that the increase is a consequence of IGF1 reduction. Tropea et al. (2009)

investigated whether systemic administration of IGF1 can improve RTT-associated deficits.

Indeed, cortical spine density and excitatory current amplitudes are significantly improved in

MeCP2-null mice after IGF1 treatment. Life span, locomotion activity, and autonomic functions

are improved as well.

The severe and potentially fatal respiratory abnormalities of RTT are hypothesized to be

associated with noradrenaline deficits (Brucke et al., 1987; Lekman et al., 1989; Nomura et al.,

1985). Reduced bioamine levels are reported in both RTT brains and MeCP2-null mice (Samaco

et al., 2008). Further, immunohistochemistry studies have shown a significant reduction in the

number of tyrosine hydroxylase expressing neurons within the medulla, where respiratory

centers are located (Viemari et al., 2005). Following this logic, desipramine, a noradrenaline

reuptake inhibitor, was tested in MeCP2-null mice to test whether respiratory abnormalities can be improved (Roux et al., 2007). The treated mice showed a delayed onset of breathing

abnormalities and life span was improved as well. The rescue was not complete, however, as only a small system was targeted in a wide array of impaired neural networks.

Lastly, the cholinergic system is also altered in RTT brains (Gadalla et al., 2011). Studies have shown a reduction of choline acetyltransferase and vesicular transporter binding (Wenk and

Mobley, 1996). To test the beneficial effects of improving the cholinergic system in RTT,

MeCP2-null mice were provided with choline enriched diet during early development (Nag and

Berger-Sweeney, 2007). The effects were subtle as there was only a modest improvement in

locomotion activity and motor function.

These various pharmacological interventions showed the beneficial effects of targeting different systems in RTT. However, most of the improvements were modest and only a subset of

RTT symptoms is improved. As mentioned previously, the most effective treatment would require direct improvement of MeCP2 function. However, treatment methods are limited by our incomplete understanding of MeCP2 function in RTT.

1.14. Rationale and hypothesis

1.14.1 Project 1: Delayed ubiquitous reactivation of MeCP2

Many clinical signs of RTT are recapitulated in MeCP2 deficient mouse models (Guy et al., 2001). MeCP2-deficient male mice develop severe symptoms and have significantly shortened lifespan (Guy et al. 2007). In contrast, MeCP2 deficient female mice resemble clinical

RTT progression as they show delayed onset of symptoms at 4-12 months of age (Robinson et al

2012). Phenotypically, male MeCP2 deficient mice show severe behaviour impairments in ambulation, motor coordination, anxiety, and social interaction in addition severe respiratory abnormalities (Samaco et al 2009; Ward et al 2011; Wither et al 2012). At a more cellular and anatomical level, neuron electrical properties, dendritic branching, and synaptic transmission are also impaired in MeCP2 deficient male mice. Milder phenotypes are observed in female MeCP2 deficient mice due to random X-chromosome inactivation profiles. Hemizygous male mouse models of RTT do not exhibit the phenotypic variability as female models and thus, are an efficient model for RTT investigations. However, heterozygous female mutants are more clinically relevant and should be incorporated into investigations to demonstrate the clinically applicability of different treatment methods.

Deficits caused by a lack of MeCP2 are not irreversible as reactivation of MeCP2 expression in 3-4 week old male MeCP2-deficient mice rescued early lethality, gross phenotypic severity, behavioural, and neural morphological deficits (Guy et al, 2007; Robinson et al 2012).

These previous studies reactivated MeCP2 to approximately 70% of wild-type levels. As work from our lab have recently shown a correlation between MeCP2 levels and behavioural performances in female MeCP2-deficient mice, the extent of improvements in MeCP2-deficient

male mice may correlate with MeCP2 reactivation efficiency. In addition, little is known about

the potential remedial effects of restoring MeCP2 to female RTT mouse models. Only

improvements of gross phenotypic severity and LTP have been reported in female mice following MeCP2 reactivation (Guy et al., 2007). The full extent of rescue achieved by MeCP2

reactivation remains unclear. Epileptiform discharge events, impaired daily rhythmic activity,

and reduced core body temperature are reported in MeCP2-deficient mice and recapitulate the clinical conditions (D’Cruz et al., 2010; Wither et al., 2012). Given that seizures, abnormal biological rhythms, and poor thermoregulation are severe hindrance to RTT girls in everyday life, it is important to determine the reversibility of these symptoms. Here, we tested whether MeCP2 reactivation in the same rescue mouse model described in Guy et al. (2007) can reverse these deficits. Furthermore, we investigated whether MeCP2 reactivation efficiencies correlate with the extent of improvements in MeCP2-deficient male mice. We found that reactivation of

MeCP2 post symptom onset leads to significant improvements in sensorimotor tasks, electroencephalography abnormalities, core body temperature, and daily rhythmic activity. In addition, we observed that greater MeCP2 reactivation levels resulted in more pronounced improvements.

1.14.2. Project 2: Preservation of MeCP2 function in catecholaminergic cells

Mutations within the X-linked gene encoding methyl-CpG-binding protein 2 (MeCP2),

have been identified as the predominate causes of RTT (Amir et al., 1999). Cardinal RTT

phenotypes are recapitulated in both male and female MeCP2 deficient mice (Chen et al., 2001;

Guy et al., 2001). By utilizing these models, studies have shown that that global reintroduction

of MeCP2 (Guy et al., 2007, Robinson et al., 2012) and targeted reintroduction of MeCP2 into specific populations of neurons or glia (Alvereez-Savaadra et al., 2007; Jugloff et al., 2008;

Giacommetti et al., 2005, Ward et al., 2011; Lioy et al., 2011) can improve the behavioural

deficits of MeCP2-deficient mice. Collectively, these studies show that the RTT-like phenotype

of MeCP2-deficient mice is not irreversible, and raises the possibility that gene reintroduction strategies may have clinical potential. While encouraging, repopulating large regions of the brain with MeCP2 remains a challenging prospect clinically. Thus the next logical step would be to determine whether preserving MeCP2 function within small populations of defined neurons

would be sufficient to improve the deficits associated with RTT.

The catecholaminergic system has been strongly implicated in the pathophysiology of

RTT (Nomura et al., 1985; Zoghbi et al., 1989; Viemari et al., 2005). The majority of

norepinephrine projections within the brain originate from neurons residing within the locus

ceruleus and lateral tegmental area (Hokfelt et al., 1984), while the majority of dopamine

projections arise from neurons within the ventral tegmental area, arcuate nucleus, or substantia

nigra (Bjorklund et al., 2007). These regions are well defined anatomically, and while comprised

of relatively small numbers of neurons, their functions influence the activity of numerous cell

types and neural systems throughout the brain. Many RTT deficits and symptoms are consistent

with phenotypes caused by abnormalities in catecholaminergic system. Motor deficits and

rigidity could be associated with dopaminergic dysfunction and autonomic abnormalities can be

attributed to adrenergic impairments (Fahns et al 2008; Tanaka et al 2000). Consistent with this

idea, decreased levels of dopamine and norepinephrine have been previous reported in RTT

patients as well as MeCP2-null animals (Samaco et al, 2009). The importance of MeCP2 function in catecholaminergic cells has been recently demonstrated, as the selective ablation of

MeCP2 from the catecholaminergic system induced RTT-like phenotypic impairments in mice

(Samaco et al., 2009). Importantly, no apoptotic neurons are observed within catecholaminergic systems (Roux et al., 2010). Since the targeted deletion of MeCP2 function in catecholaminergic cells can produce RTT-like deficits, we hypothesized that selective preservation of MeCP2

function in catecholaminergic cells of MeCP2 deficient mice will lead to retention of normal

phenotypic behaviours.

To test this, we selectively preserved MeCP2 function in TH-positive cells of a murine

RTT mouse model. Our results show that the preservation of MeCP2 function in this population

of cells is sufficient to improve phenotypic deficits in both male and female MeCP2-deficient

mice.

1.15. Project Aims

Project 1: to determine whether the delayed ubiquitous reactivation of MeCP2 can improve

- behaviour deficits

- epileptiform-like discharge events

- circadian activity

- core body temperature

Project 2: to determine whether selective preservation of MeCP2 expression in catecholaminergic cells can improve

- gross phenotypic severity

- behavioural performances

- electroencephalography abnormalities

Figure 4

Figure 4. Different transgenes expressed in different mouse lines.

Panel A: MeCP2flox-stop transgene is inserted in the Stop/y or Non-rescue mouse line. In the presence of cre recombinase, the floxed-stop cassette is excised and MECP2 transcription is re- established. Panel B: THcre transgene is expressed in the TH-cre mouse line. Cre recombinase is only expressed in tyrosine hydroxylase positive cells as it is driven by a TH-specific promoter. Panel C: Rosa26-ESRcre transgene expressed in the Rosa-cre mouse line. Cre recombinase, in this case, is normally sequestered in the cytoplasm. When tamoxifen (an estrogen antagonist) binds to the estrogren receptor, cre recombinase is allowed to enter the nucleus and target loxp flanked sites.

Chapter 2

Materials and Methods

2.1 Mice. All animal procedures and protocols were approved by the Canadian Council on

Animal Care and Toronto Western Research Institute’s Animal Facility. Mice were housed in a

controlled facility on a 12 hour light/dark cycle and were provided with food and water ad

libitum. CreESR and MeCP2stop/+ mice were obtained from The Jackson Laboratory and

maintained on a C57Bl/6 background. MeCP2stop/+ female mice, which is heterozygous for

MeCP2 allele silenced by a neo-stop cassette, is crossed with male CreESR to obtain male

MeCP2stop/yESRcre and female MeCP2stop/+ESRcre mice.

TH-cre mice were obtained as a gift from Dr. Joseph Savitt (Johns Hopkins University) and

maintained on a pure C57Bl/6 background. Mecp2tm2Bird mice (Guy et al., 2007) were obtained

from The Jackson Laboratory (Bar Harbor, Maine) and also maintained on a pure C57Bl/6 background. TH-Cre male mice were crossed with Mecptm2Bird female mice to generate

experimental subjects of both genders. Polymerase chain reaction (PCR) was used to identify the

genotype of these mice. DNA samples were prepared using the HotSHOT genomic DNA

preparation method (Truett et al., 2000) on tissues collected from ear punches. The floxed-stop

sequence is identified using the primer set: “5’-CTTCAGTGACAACGTCGAGC” and “5’-

CATTCTGCACGCTTCAAAAG”. The presence of cre recombinase sequence was identified

using the primer set: ”5’-AAATGTTGCTGCTGGATAGTTTTTACTGC” and ”5’-

GGAAGGTGTCCAATTTACTGACCGTA”.

2.2 Western blotting. Animals were anesthetized under isoflurane and sacrificed for tissue collection. Brain samples were collected on ice and snap frozen in dry ice. Quarter brains were homogenized in approximately 300µl of RIPA buffer (50 mM Tris-HCL[pH 7.5], 150 mM NaCl,

1% NP40, 2 mM EDTA, 0.5% sodium deoxycholate, and 0.1% SDS) supplemented with a mixture of protease inhibitors (PMSF 40 ng/ml, Antipain 2 ng/ml, PepstatinA 2 ng/ml, Leupeptin

20 ng/ml, Aprotinin 20 ng/ml, and MDL28170 20 ng/ml). Samples were then spun down at

12,000g for 5 minutes and the supernatant was collected. Samples were stored at -80°C until used. Protein concentrations of samples were determined using the Folin method (Bio-Rad,

Hercules, CA, Cat # 500-0116).

Protein samples were resolved by sodium dodecyl sulphate polyacrylamide gel electrophoresis

and transferred to a polyvinylidene fluoride membrane in standard transfer buffer (25 mM TRIS,

192 mM glycine, 20% methanol) overnight at 4°C and pre-hybridized for 2 hours at room temperature in blocking solution (TRIS-buffered saline containing 0.05% Tween-20 (v/v) (TBST) and 5% (w/v) non-fat dry milk). Membranes were hybridized overnight at 4°C with primary antibody (1/1000; Cell Signaling Technology, Danvers, MA, Cat # 3456S) in blocking solution.

After washes in TBST, membranes were hybridized with HRP-linked secondary antibodies

(1/5000; GE Healthcare, Buckinghamshire, England, Cat # NA934 [anti-rabbit], Cat # NA931

[anti-mouse]) for 2 hours at room temperature. After extensive washing in TBST, immunoreactivity was visualized by enhanced chemiluminescence (GE Healthcare, Cat #

RPN2106).

2.3 Immunohistochemistry. Animals were anesthetized through inhalation of 2% isoflurane, and

transcardially perfused with 0.9% NaCl saline solution followed by perfusion with an ice-cold 2%

paraformaldehyde-PBS solution. Intact brains were dissected from the skull and equilibrated

overnight in a 30% sucrose/PBS solution at 4oC. The brain was then stored at -80oC until further

assessment. For sectioning, the brain was cut at the midline, and one hemisphere was embedded

in O.C.T. compound (Sakura, Torrance, California) and coronal sections (15 µm) were collected

with a Leica cryostat (model Jung CM 3000, Wetzlar, Germany) at -24oC. Sections were blocked

with 10% NGS + 2% BSA in 0.1% PBS-T for 1 hour, and then incubated with rabbit anti-

MeCP2 (1:500 Cell Signaling, #3456S) antibodies and/or mouse anti-tyrosine hydroxylase

(1:1000, Millipore, #NG1752067) in 0.1% PBS-T supplemented with 2% normal goat serum

overnight at 4oC. The sections were then washed using 0.1% PBS-T at room temperature 3 times,

and incubated with secondary antibodies conjugated to either DyLight 488 (Invitrogen, goat anti- mouse, #A11001) and/or DyLight 568 (Invitrogen, goat anti-rabbit, #A11011) for 1 hour at room temperature. Following incubation, the sections were washed with PBS, and then incubated

briefly with DAPI (5 µg/ml, Roche Diagnostics, Indianapolis, Indiana #10236276001) for 3 min.

Sections were then rinsed with PBS, and mounted atop slides with Dako Fluorescent Mounting

Media (Burlington, Ontario, Canada, #S302380). Imaging was done using a Zeiss Axioplan 2

deconvolution microscope (Carl Zeiss, Göttingen, Germany).

2.4 Tamoxifen treatment. Tamoxifen (Sigma) was dissolved in corn oil (6mg/ml) and stored at

4°C until use. Tamoxifen was administered to mice through peritoneal injection for five

consecutive days at 100mg/kg or once per week for three weeks at 33mg/kg followed by two

booster injections at 67mg/kg. Tamoxifen injections in male Stop/ESRcre mice were delayed

until phenotypic symptoms are fully developed (~50-60 days of age with a phenotypic score of at least 5). Phenotypic severities of female Stop/ESRcre mice were scored weekly and tamoxifen injections commenced when the mice reached a score of 5 and are at least 250 days of age. Wild- type, MeCP2stop/y, and MeCP2-/+ controls were treated with tamoxifen in the same manner. Age

matched Rosa male and female mice were also injected with tamoxifen to ensure no confounding

effects from the treatment.

2.5 Electrophysiology data collection. 24 hour activity, body temperature, and EEG were

collected from telemetrically implant female mice as described previously in Wither et al. (2012).

Briefly, the TA11ETA-F10 telemetry probe and a wireless receiver (RPC-1, DSI) were used to

collect waveform data. Locomotion activity was collected by calculating the standard deviation

of the transmitter signal strength relative to two perpendicularly arranged antennae in the RPC-1

wireless receiver. Body temperature was collected from an internal thermosensor in the

TA11ETA-F10 transmitter. Body temperature and locomotion activity data were transmitted at a

rate of 50 Hz, using a sampling frequency of 250 Hz. EEG waveform data was transmitted at 200

Hz and sampled at 1 kHz. (Implantations were done by Chiping Wu)

2.6 Behavioural Assessments. Animals were assessed in the open field and accelerating rotarod

test as described previously (Jugloff et al., 2008). For the open field ambulation test, subjects

were placed in a Plexiglass cage (20 x 30 cm2) and an automated movement detection system

(AM1053 activity monitors; Linton Instruments, United Kingdom) was used to record the motor

activities of the animals for a one-hour period. For the accelerating rotarod test, subjects were

placed on a rotating rod (MED Associates Inc., #ENV-575M, St. Albans, Vermont) that

accelerates linearly from 3.5 rpm to 35 rpm over a 5-minute period. The time at which the subject fell from the rotating rod was recorded via a laser beam sensor. Each subject was assessed on the accelerating rotarod three times a day for four consecutive days. Consecutive trials were separated by at least one hour to allow the animals to recover from physical fatigue.

Animals that circumnavigated the rod for three consecutive times were scored as having fallen off the apparatus upon the third rotation. For the light-dark placement preference test, mice were

placed into a box consisting of a dark compartment (20 x 14 cm) and a light compartment (20 x

28 cm) connected through a single small opening (4 cm2). The amount of time the animal spent

in the dark and light compartments, as well as the number of risk assessment behaviours (head pokes out of the dark compartment) each subject took while in the dark compartment, were recorded from videotaped 10 minute sessions. For the nest-building behaviour test, subjects

were placed into a new cage containing a single piece of nestlet, and the volume of each

assembled nest was calculated the following day. All behaviour tests were conducted between

9:00am and 13:00pm to minimize circadian effects. Female subjects were assayed after 280

days of age, and male animal subjects were assessed between 50 and 70 days of age.

2.7 Phenotypic severity scoring. Animals were scored using the deficit scoring system described previously (Cheval et al., 2012; Guy et al., 2007; Robinson et al., 2012). In short, mice were

scored from 0-2 according to the following scheme: Mobility score: 0 =same as wild-type; 1 =

slower movements than wild-type with intermittent freezing periods; 2 = severely reduce or no

movement at all. Gait score: 0 = same as wild-type; 1 = hind-limbs are spread wider than wild-

types and slips or double tapping of the same feet is observed; 2 = dragging of hind-limbs or

constant slips. Hind-limb clasp score: 0 = hind-limbs are spread out when lifted by the tail, same

as wild-type; 1 = one hind-limb is pulled towards the body and clasped or both hind-limbs are

mildly pulled inwards towards the body; 2 = both hind-limbs are clasped and pulled tightly into

the body. Breathing score: 0 = no noticeable respiratory abnormalities, same as wild-type; 1 =

occasional respiratory jerks or some irregularity in breathing patterns; 2 = severe respiratory

difficulties involving strong body jerks. Tremor score: 0 = no tremor; 1 = intermittent mild

tremors; 2 = severe and constant tremors. General condition score: 0 = well-groomed and shiny

fur, opened eyes, normal body posture; 1 = slightly dishevelled fur, squinty eyes, and slightly

hunched posture; 2 = extremely dishevelled fur, eyes closed, and severely hunched posture.

2.8 Cell Counting. MeCP2 and tyrosine hydroxylase expression was determined through

random sampling of every 5th section in brain slices containing substantia nigra or locus ceruleus.

Substantia nigra was identified from Bregma -2.48mm to -3.88 mm, and locus ceruleus was identified from Bregma -5.38 mm to -5.85 mm. MeCP2 expression was assessed only in clearly identified nuclei that displayed DAPI staining. MeCP2 antibody specificity was confirmed by comparing the staining patterns of wild-type and MeCP2-/y animals. Counts were made for cells

displaying either tyrosine hydroxylase or MeCP2 immunoreactivity alone, and for cells

expressing both tyrosine hydroxylase and MeCP2. Cell counts were conducted by two

independent examiners blinded to condition, and their individual counts averaged for analysis.

2.9 Telemetry Probe Implantation Protocol. Female mice were implanted with a wireless

telemetry probe TA11ETA-F10; Data Sciences International (DSI), St. Paul, MN) for long

duration electroencephalogram (EEG) and activity recordings as described in (Wither et al.,

2012). Animals were allowed to recover for at least 3 weeks prior to any data collection.

2.10 Tethered Electrode Implantation. Animals were implanted with electrode cap assemblies as described previously (Wu et al., 2008). Briefly, animals were anesthetized under 2-4% isoflurane through inhalation. Electrodes made from polyimide-insulated stainless steel were implanted in the hippocampal CA1 (Bregma, -2.3 mm; lateral, 1.7 mm; depth, 2.0 mm) and contralateral somatosensory cortex (Bregma, -0.8 mm; lateral, 1.8 mm; depth, 1.5 mm). A reference electrode was implanted in the frontal cortex (Bregma, -3.8 mm; lateral 1.8 mm; depth,

1.5 mm). Male mice were implanted between ages 40 and 60 days of age, during which the symptoms begin to. Female mice were implanted post 250 days of age, and after symptom onset.

The implanted mice were allowed to recover for at 7 days before any further experiments. Baytril antibiotics (Bayer Healthcare, Toronto, Ontario) were added to the water supply 2 days before surgery and 7 days after surgery to minimize infections. (All implantations were done by Chiping

Wu)

2.11 Electroencephalographic recordings and analysis. EEG recordings were collected as described previously (D’Cruz et al., 2010). Briefly, the implanted electrodes were connected to two independent head stages (Model-300, AM Systems Inc., Carlsborg, Washington). EEG signals were amplified 1000x, bandpass filtered (0.01 – 1000 Hz), and digitized (Digidata 1300,

Axon Instruments, Weatherford, Texas). EEG data were collected at 60 kHz and analyzed using

Clampfit software (Axon Instruments). Recording sessions were at least 2 hours in duration, and each subject was recorded for a minimum of two sessions on different days. The EEG recordings were decimated 10x via Clampfit 10.2 software before analysis. Epileptiform discharge-like events were counted manually using the following criteria: frequency between 6-12Hz, minimum duration of 0.5 seconds, and at least 1.5x the baseline amplitude, and high rhythmicity.

Hippocampal theta epochs during exploratory behaviours were bandpass filtered (0.5-200 Hz), and spectral plots (50% window overlap and frequency resolution of 0.25 Hz) were generated. A minimum of 10 epochs from at least two recording sessions were averaged to obtain peak theta frequency, total theta power, and total gamma power for each animal. The frequency between 6-

12 Hz with the greatest power was taken as the peak theta. Total theta wave power was calculated by taking the area underneath the spectral plot between 6-12 Hz. Similarly, total gamma wave activity is taken as the area underneath the spectral plot between 35 and 60 Hz. All

EEG data were calculated and analyzed using Clampfit 10.2 software.

2.12 Statistics. All statistical analysis was performed using PRISM or Microsoft Excel software on a PC. Paired student’s t-test was used to compare before and after Stop/+,cre and Stop/y,cre mice. Two-way ANOVA with Bonferroni post hoc correction was used to for multiple group comparison with that had time as a factor. Pearson’s product moment correlation was used to compare correlative strength between two groups. All behavioural data were analyzed using one- way ANOVA with Tukey’s post hoc comparison, or two-way ANOVA with Bonferroni’s post- hoc test. Cell counts as well as MeCP2 and tyrosine hydroxylase expression levels were analysed using unpaired student’s t-test. Kaplan-Meier survival plot was analyzed using

Wilcoxon rank sum tests. Spontaneous death rate was compared using population Chi-squared

tests with one degree of freedom. For all cases, the threshold for statistical significance was set at

p<0.05.

Chapter 3

Results

3.1.1. Rescue of MeCP2 expression in Stop/y,cre mice.

As described in Guy et al. (2007), the MeCP2Tm2bird mice have disrupted MeCP2 expression as a

“neo-stop” cassette is inserted between exon 2 and exon 3. Consistent with previous reports, we

confirmed that the MeCP2lox-stop allele behaves similar to a null mutation. Stop/y mice expressed

MeCP2 at ~1.5-3% of wild-type levels (Figure 5A). By crossing MeCP2+/- mice with mice that

expressed ESRcre, we generated rescue mice that allow MeCP2 expression to be reinstated

through tamoxifen injections. To determine whether the level of MeCP2 reactivation correlates

to the extent of rescue, we employed two tamoxifen injection paradigms. The low reactivation cohort of Stop/y,cre mice received tamoxifen injections once a week at a dosage of 50mg/kg for three weeks, followed by two consecutive booster injections of 100mg/kg (low TMX). The high

reactivation cohort of Stop/y,cre mice received five consecutive daily injections tamoxifen at the

dosage of 100mg/kg (high TMX). The low reactivation cohort (Stop/y,cre + low TMX) express

MeCP2 at 21 +/- 6% of wild-type levels, whereas the high reactivation cohort express MeCP2 at

65 +/- 18% of wild-type levels after tamoxifen injections (Figure 5C).

Figure 5

Figure 5. MeCP2 is reactivated in Stop,cre mice following tamoxifen treatment. Panel A:

Western blot of whole brain samples from a wild-type and Stop/y mice. Panel B: Western blot of whole brain samples from wild-type (n=3), Stop/y,cre + low TMX (n=3), and Stop/y,cre + high

TMX (n=3). Panel C: Level of MeCP2 expression in Stop/y,cre + low TMX (n=3) and Stop/y,cre

+ high TMX (n=3) mice as a percentage of male wild-type mice (n=3).

3.1.2. Restoration of MeCP2 rescues life span and gross phenotypic severity.

As MeCP2 is essentially silenced in Stop/y mice, early lethality is observed (Figure 6A).

The median survival age of Stop/y mice is 77 days. In contrast, survival is significantly improved in Stop/y,cre + low TMX mice compared to Stop/y mice (median survival age of 255 days, p<0.01, Figure 6A). Stop/y,cre + high TMX mice displayed a median survival age of 320 days, which is significantly improved relative to both Stop/y and Stop/y,cre + low TMX mice (p<0.01 and p<0.05 respectively, Figure 6A). Consistent with the improved life span, phenotypic severity scores of both the high and low TMX treated Stop/y,cre mice were significantly lower than

Stop/y mice, which showed continuous progression of severity until early death (Figure 6B). In addition, Stop/y,cre + high TMX mice scored on average significantly lower than Stop/y,cre + low TMX mice between 4-23 weeks of age.

Figure 6

Figure 6. Survival and gross phenotypic severity is significantly improved in Stop/y,cre mice after tamoxifen treatment. Panel A: Kaplan-Meier survival plot of male Stop/y mice (n=29),

Stop/y,cre + low TMX (n=10), and Stop/y,cre + high TMX (n=9). The life span of the Stop/y,cre

mice after tamoxifen treatment is significantly longer than Stop/y mice (p<0.01, Wilcoxon rank-

sum test). The life span of Stop/y,cre + low TMX and Stop/y,cre + high TMX is not significantly

different. Panel B: the gross phenotypic severity score of Stop/y mice (n=13), Stop/y,cre + low

TMX (n=6), Stop/y,cre + high TMX (n=5). Stop/y,cre + low TMX scored significantly lower

than Stop/y mice starting at and after 4 weeks of age (one-way ANOVA with Tukey’s post-hoc

correction). Stop/y,cre + low TMX scored significantly lower than Stop/y at and after 5 weeks of

age, and significantly higher than Stop/y,cre + high TMX between 5 and 20 weeks of age (one-

way ANOVA with Tukey’s post-hoc correction).

3.1.3. Extent of behavioural rescue is dependent upon the degree of MeCP2 reactivation.

We further examined the reversibility of behavioural impairments by assessing the mice

in specific sensory-motor tests – open field ambulation test, accelerating rotarod test, light/dark

preference test, and nest building test (Figure 7). Stop/y mice were assayed between 50-70 days

of age and the rescue mice are assayed post 100 days of age and at least two months after

tamoxifen treatment to avoid confounding effects. As expected, the behavioural performances of

Stop/y mice are significantly impaired compared to wild-types in all tests. We further tested low

TMX treated and high TMX treated Stop/y,cre mice in the same behavioural tests and found that

high TMX treated Stop/y,cre mice displayed greater levels of behaviour rescue than low TMX treated Stop/y,cre mice. In the open field test, activity and rearing counts were significantly improved in Stop/y,cre + low TMX compared to Stop/y (one way ANOVA, p<0.05), whereas ambulation mobility was not improved (one way ANOVA, p>0.05 for each) (Figure 7A). High

TMX treated Stop/y,cre mice’s performances were significantly improved in every parameter compared to Stop/y (one way ANOVA, p<0.01 for each). Additionally, high TMX treated

Stop/y,cre mice displayed significantly more activity and rear counts than low TMX-treated

Stop/y,cre mice (one way ANOVA, p<0.01). Motor deficits are cardinal features of RTT and we assayed whether MeCP2 reactivation rescues these impairments by testing the rescue mice on the accelerating rotarod test (Figure 7B). Stop/y mice showed severe motor impairment as their initial latency to fall and their final day latency to fall was 101 +/- 9 seconds and 158 +/- 11 seconds, compared to wild-types’ 196 +/- 7 and 295 +/- 4 seconds of initial and final latency to fall (two-way ANOVA, p<0.001). Although low TMX treated Stop/y,cre mice remained significantly below wild-type performance levels, they were able to last significantly longer on the rotarod than Stop/y mice on trial days 1,2, and 4 (p<0.05, two-way ANOVA). The

52 performances of high TMX treated Stop/y,cre were significantly better than Stop/y mice on all days of trial and not significantly different than wild-type levels (p<0.001, p>0.05, respectively; two-way ANOVA). In the light and dark place preference test, the number of head pokes performed by the mouse while in the dark compartment was taken as an indicator of risk assessment/anxiety-like behaviour. Only high TMX treated Stop/y,cre mice, not low TMX treated, performed significantly more risk assessments per time spent in the dark compartment than Stop/y mice (one way ANVOA, p<0.05) (Figure 7C). Sociability was tested using the nest building behaviour test by measuring the volume of the nest size 24 hours after introducing the mouse to the new cage. Similar to the light/dark preference test, only high TMX treated,

Stop/y,cre mice showed a significant improvement in the average nest volume built (high TMX treated versus low TMX treated Stop/y,cre mice versus Stop/y: 46.6 +/- 13.2 cm3 versus 9.4 +/-

2.5 cm3 versus 3.5 +/- 0.3 cm3; one way ANOVA, p<0.05 for high TMX treated Stop/y,cre mice versus Stop/y).

Figure 7

Figure 7. Behavioural performances were improved in Stop/y,cre mice following tamoxifen

treatment. Panel A: Histogram showing the mean and SEM of male Stop/y mice (n=17),

Stop/y,cre + low TMX (n=9), and Stop/y,cre + high TMX (n=7) in the open field test. The behavioural performances of the mice are normalized to mean wild-type levels. Stop/y,cre + high

TMX mice performed significantly better than Stop/y and Stop/y,cre + low TMX in all parameters (p<0.01, two-way ANOVA with Bonferroni’s post-hoc correction). Stop/y,cre + low

TMX displayed great total activity and rearing counts than Stop/y mice (p<0.05, two-way

ANOVA with Bonferroni’s post-hoc correction). Panel B: Motor performance of Stop/y mice

(n=27), Stop/y,cre + low TMX (n=11), Stop/y,cre + high TMX (n=9) on the accelerating rotarod.

Stop/y,cre + high TMX mice displayed a longer latency to fall than Stop/y mice on all trial days

(p<0.01, two-way ANOVA with Bonferroni’s post-hoc correction). Stop/y,cre + low TMX

displayed a longer latency to fall than Stop/y mice on trial days 1, 2, and 4. No significant

difference was detected between Stop/y,cre mice with low TMX and high TMX. Panel C:

Anxiety-like behaviour was assessed using the light-dark place preference test. The histogram

shows the mean and SEM of the number of risk assessments (head pokes) performed by the

different cohort of mice per minute of time spent in the dark chamber. Stop/y,cre + high TMX

(n=6) performed significantly more head pokes than Stop/y mice (n=15). Panel D: Social

behaviour was assessed using the nest building test. The histogram shows the mean and SEM of

the nest volume built by the different cohort of mice 24 hours being introduced into the test cage.

Stop/y,cre + high TMX (n=6) built larger nests than Stop/y mice (n=13) (p<0.05, one-way

ANOVA with Tukey’s post-hoc correction), but still smaller than male wild-type mice (n=14)

(p<0.05, one-way ANOVA with Tukey’s post-hoc correction). Stop/y,cre + low TMX mice (n=6)

did not assemble nests that were significantly larger than Stop/y mice (p>0.05, one-way ANOVA

55 with Tukey’s post-hoc correction). For panel A, C, and D, * indicates p<0.05. For panel B, * indicates p<0.05 compared between Stop/y,cre + high TMX and Stop/y. # indicate p<0.05 compared between Stop/y,cre + low TMX and Stop/y.

3.1.4. Epileptiform discharges are significantly attenuated after MeCP2 reactivation

Epileptiform-like discharge events were reported in MeCP2 deficient mice (D’Cruz et al.,

2010, Wither et al., 2012). Here, we observed these events in all of the Stop/y mice after 50 days

of age. On average, Stop/y mice exhibited 42 +/- 7.8 discharge events per hour with an average

duration of 2.8 +/- 0.71 seconds per event (Figure 8A,D,E). Similar epileptiform-like discharge

events were observed in tamoxifen-treated Stop/y,cre mice. The average duration of individual

discharge events, however, is significantly reduced in both low TMX and high TMX treated

Stop/y,cre mice (0.96 +/- 0.1 seconds and 0.99 +/- 0.14 seconds, respectively; p<0.05, one-way

ANOVA) (Figure 8B,C,D). Low TMX treated Stop/y,cre mice displayed 28.7 +/- 4.2 discharges per hour and high TMX treated Stop/y,cre mice displayed an average of 5.4 +/- 1.4 discharges per hour. The incident rate was only significantly reduced in high TMX treated Stop/y,cre mice

compared to Stop/y (p<0.05, one-way ANOVA). There was a strong trend towards significance

in low TMX treated Stop/y,cre mice compared to Stop/y (p=0.056, one-way ANOVA). We

observed severe long discharge events, which were characterized by at least 5 seconds in

duration, in Stop/y mice. These severe long discharge events coincide with movement arrest and

blank gaze. 88.9% (8 out of 9) of our Stop/y cohort displayed these severe long discharge events,

which was significantly higher than the 11.1% (1 out of 9) observed in low TMX treated

Stop/y,cre mice. However, no long discharge events were observed in any high TMX treated

mice (0 out 6) (Figure 8F).

Figure 8

Figure 8. Epileptiform-like discharge incidence rate is significantly improved in Stop/y,cre mice after tamoxifen treatment. Panel A, B, and C: Representative EEG recording traces (5 seconds)

showing a single discharge event. Panel A trace is collected from a Stop/y mice, panel B trace is

collected from a Stop/y,cre + low TMX mice, and panel C trace is collected from a Stop/y,cre +

high TMX mice. Panel D: Histogram showing the mean and SEM of the average discharge

duration of individual epileptiform event in Stop/y mice (n=9), Stop/y,cre + low TMX mice

(n=6), and Stop/y,cre + high TMX mice (n=5). The average epileptiform duration is significantly

reduced in Stop/y,cre mice with low and high TMX compared to Stop/y mice (p<0.05, one-way

ANOVA with Tukey’s post-hoc correction). Panel E: Histogram showing the mean and SEM for

the number of epileptiform events per hour in Stop/y mice (n=9), Stop/y,cre + low TMX mice

(n=8), and Stop/y,cre + high TMX mice (n=5). The number of discharge events per hour is

significantly reduced in Stop/y,cre + high TMX mice compared to Stop/y mice (p<0.05, one-way

ANOVA with Tukey’s post-hoc correction). There is a strong trend in the reduction of

epileptiform discharges per hour in Stop/y,cre + low TMX mice compared to Stop/y mice

(p=0.056, one-way ANOVA with Tukey’s post-hoc correction). Panel F: Histogram showing the

percentage of population of Stop/y mice (8 out of 9), Stop/y,cre + low TMX (1 out of 9), and

Stop/y,cre mice + high TMX (0 out of 9) that exhibit severe long duration discharge events. For

panel E and D, * indicates p<0.05, for panel F, it indicates p<0.05 compared between Stop/y

versus Stop/y,cre + low TMX and Stop/y,cre + high TMX.

3.1.5. Reactivation of MeCP2 in female MeCP2-deificient mice rescues behavioural

performances

Female Stop/+ and Stop/+,cre display neurological symptoms that recapitulate clinical

features of RTT syndrome (Guy et al., 2007). Consistent with symptoms such as impaired inertia,

gait, and anxiety behaviour, our Stop/+ female mice showed significant impairment in the open field, rotarod, light/dark placement preference, and nest building behavioural test, compared to wild-type littermates (p<0.01 for all behavioural test) (Figure 9). Stop/+,cre females were similarly impaired before tamoxifen treatments (<270 days of age) and showed no significant differences from Stop/+ females in all the behavioural tests (p>0.05 for all tests). Stop/+,cre females were treated with tamoxifen post 270 days of age and assayed for the same behavioural test 2 months after treatments. Activity, rearing, and mobile counts in the open field test were all significantly improved following MeCP2 reactivation (p<0.05 for all parameters, paired t-test)

(Figure 9A). However, no improvements were seen in the light and dark place preference test after treating Stop/+,cre mice with tamoxifen (Figure 9B). Motor coordination was impaired in

Stop/,cre mice, as the mice showed a shorter latency to fall on all days of trial compared to wild- type controls (p<0.01, two-way ANOVA). Mild improvements were seen in motor performance as tamoxifen-treated Stop/+,cre mice showed a longer latency to fall on day one of the trials

(p<0.05, two-way ANOVA) (Figure 9C). Stop/+,cre females showed a significant impairment in nest building ability compared to wild-type controls (52 +/- 6cm3 versus 118 +/- 7.1cm3 in nest

volume, respectively; p<0.01 student’s t-test) (Figure 9D). Following tamoxifen treatment, nest building behaviour of Stop/+,cre females was significantly improved as the average nest volume was 73.7 +/- 7.9cm3 (p<0.05, paired t-test) (Figure 9D).

Figure 9

Figure 9. Behavioural performances are improved in Stop/+,cre mice following MeCP2

reactivation. Panel A: Histogram showing the mean and SEM of female Stop/+,cre mice before

and after tamoxifen treatment in the open field test. The behavioural performances of the mice

are normalized to average wild-type levels. Stop/+,cre mice performed significantly better in all

parameters after tamoxifen treatment (p<0.05, paired t-test). Panel B: Anxiety-like behaviour

was assessed using the light-dark place preference test. The histogram shows the mean and SEM

of the number of risk assessments (head pokes) performed per minute of time spent in the dark

chamber. Stop/+,cre mice did not show an improvement in risk assessment behaviour after

MeCP2 reactivation. Panel C: Motor performance was assayed using the accelerating rotarod.

Latency to fall was significantly improved on day one of the rotarod trial in Stop/+ mice after

tamoxifen treatment (p<0.05). Panel D: Social behaviour was assessed using the nest building

test. The histogram shows the mean and SEM of the nest volumes built by the mice 24 hours

being introduced into the test cage. Stop/+,cre mice built significantly larger nest post MeCP2

reactivation. For all panels, n = 7 for Stop/+,cre mice. Paired student’s t-test was used for all statistical comparisons between pre and post-tamoxifen treated Stop/+,cre mice. * indicates p<0.05 compared between pre and post tamoxifen treated Stop/+,cre mice.

3.1.6. MeCP2 reactivation improves daily rhythmic activity and thermoregulation in adult female MeCP2-deficient mice.

Impaired movement and activity are cardinal features of RTT syndrome. These impairments were recapitulated in Stop/+ female mice (Wither et al., 2012) and our cohort of

Stop/+,cre mice before tamoxifen treatment as total daily (24 hours) activity count is significantly lower in Stop/+,cre female mice than wild-type littermates (139 +/- 8.7 versus

204.38 +/- 20.2, p<0.05, one way ANOVA) (Figure 10C). Activity counts during the light phase

(6:00am – 6:00pm) and the dark phase (6:00pm-6:00am) were significantly lower in Stop/+,cre than in wild-type female mice (42.8 +/- 3.7 arbitrary units versus 74.8 +/- 6.6 arbitrary units during light phase, and 96.2 +/- 5.8 arbitrary units versus 139.6 +/- 13.9 arbitrary units during dark phase, Stop/+,cre versus wild-type, p<0.05 Student’s t-test) (Figure 10A,B). After treating

Stop/+,cre female mice with tamoxifen, activity during the light and dark phase as well as total daily activity counts were significantly improved and restored to wild-type levels (207.2 +/- 21.9 arbitrary units and 79.5 +/- 10.6 arbitrary units; p>0.05, compared against wild-type, one way

ANOVA). The amount of time spent being active was significantly lower in Stop/+,cre mice compared to wild-type controls (23.25 +/- 0.74% versus 32.32 +/- 3.29%, p<0.01, Student’s t- test). After treating Stop/+,cre with tamoxifen, the percent of day spent being active was significant increased to 30.88 +/- 1.96% (p<0.05, paired t-test) and was not significantly different from wild-type levels (p>0.05, student’s t-test) (Figure 10D).

Thermoregulation and core body temperature are abnormal in Stop/+,cre female mice

(Figure 11C,D). The daily temperature maximum, minimum, and average of Stop/+,cre female mice (37.98 +/- 0.03°C, 34.49 +/- 0.13°C, and 36.11 +/- 0.08°C, respectively) were all significantly lower than wild-type levels (38.54 +/- 0.07°C, 35.56 +/- 0.2, 36.97 +/- 0.09°C;

p<0.01 for each, compared between Stop/+,cre and wild-type, Student’s t-test) (Figure

11A,B,C,D). Additionally, core body temperature variation is significantly smaller in wild-type than Stop/+,cre mice (2.98 +/- 0.19°C versus 3.49 +/- 0.13°C, respectively; p<0.05, Student’s t- test). After treating Stop/+,cre with tamoxifen, the maximum, minimum, and average daily temperatures were significantly improved to 38.31 +/- 0.06°C, 34.91 +/- 0.18°C, and 36.39 +/-

0.09°C, respectively (p<0.01, p<0.05, and p<0.01, respectively, compared between pre and post tamoxifen treatment in Stop/+,cre mice, paired t-test) (Figure 11A,B,C,D). However, the daily temperature range was not significantly improved (p>0.05, paired t-test) (Figure 11D).

Normally, core body temperature increases during activity and decreases during inactive periods. Thus a strong correlation between core body temperature and activity is usually observed. The Pearson’s correlation coefficient between activity and body temperature during 24 hour cycle, light phase, and dark phase were all significantly decreased in Stop/+,cre mice before tamoxifen treatment when compared to wild-type controls (0.56 +/- 0.05 versus 0.76 +/- 0.03, p<0.01, Student’s t-test) (Figure 12F). After treating Stop/+,cre mice with tamoxifen, the correlation between core body temperature and activity during light phase, dark phase, and 24 hour period were all significantly improved (p<0.01 for each, paired t-test) and were not significantly different than wild-type controls (p>0.05, Student’s t-test).

Figure 10

Figure 10. Daily activity is significantly improved in Stop/+,cre mice after MeCP2

reactivation. Panel A: Histograms showing the mean and SEM of activity counts during the dark

phase (6:00pm – 6:00am) of wild-type mice, Stop/+,cre mice, and Stop/+,cre mice after

tamoxifen treatment. Panel B: Histograms showing the mean and SEM of activity counts during

the light phase (6:00am – 6:00pm) of wild-type mice, Stop/+,cre mice, and Stop/+,cre mice after

tamoxifen treatment. Panel C: Histograms showing the mean and SEM of total activity counts

during a 24 hours cycle of wild-type mice, Stop/+,cre mice, and Stop/+,cre mice after tamoxifen

treatment. Panel D: Histograms showing the percent of the 24 hour duration the wild-type mice,

Stop/+,cre mice, and Stop/+,cre mice after tamoxifen treatment spent being active. For all panels,

n=8 for wild-types, and n=6 for Stop/+,cre mice (same mice for before and after tamoxifen

treatment groups). Paired Student’s t-test was used for all comparisons between Stop/+,cre mice before and after tamoxifen treatment. Student’s t-test was used for comparisons between wild- type mice and Stop/+,cre mice without tamoxifen. * indicates p<0.05.

Figure 11

Figure 11. Core body temperature is improved in Stop/+,cre mice after MeCP2

reactivation. Panel A: Core temperature maximum during 24 hour period of wild-type mice and

Stop/+,cre mice before and after tamoxifen treatment. Panel B: Core temperature minimum

during a 24 hour period of wild-type mice and Stop/+,cre mice before and after tamoxifen

treatment. Panel C: Average core body temperature during a 24 hour period of wild-type mice

and Stop/+,cre mice before and after tamoxifen treatment. Panel D: Scatter plot showing the

range of core body temperature in in wild-type mice and Stop/+,cre mice before and after

tamoxifen treatment. Each point represents the absolute temperature range of an individual

mouse. In all panels, n=8 for wild-type, and n=7 for Stop/+,cre mice. Student’s paired t-test was used to all statistical comparisons between Stop/+,cre mice before and after tamoxifen. Student’s t-test was used for comparisons between Stop/+,cre before tamoxifen with wild-type mice. * indicates p<0.05.

Figure 12

Figure 12. Temperature and mobility correlation is significantly improved in MeCP2 reactivated female mice. Panel A and B: Representative 24 hour activity cycle profile of a Stop/+,cre mice before (A) and after (B) tamoxifen treatment. Each spike indicates an instance of activity. Panel

C and D: Representative 24 hour temperature cycle of a Stop/+,cre mice before (C) and after (D) tamoxifen treatment. The cycle is plotted from 6:00am to 6:00am of the next day. Panel E:

Histograms showing the mean and SEM of Pearson’s correlation coefficient (r) for activity and core body temperature in Stop/+,cre mice before and after tamoxifen. Panel F: Scatter plot showing the range of Pearson’s correlation coefficient (r) for activity and core body temperature in wild-type mice (n=8), Stop/+,cre before (n=7) and after (n=7) tamoxifen treatment. Each point indicate the daily correlative strength between activity and body temperature of an individual mice. Paired Student’s t-test was used for all comparisons between Stop/+,cre mice before and after tamoxifen treatment. Student’s t-test was used for comparisons between wild-type mice and

Stop/+,cre mice without tamoxifen. * indicates p<0.05.

Project 2

3.2.1. MeCP2 is selectively preserved in tyrosine hydroxylase-expressing neurons in the

“Rescue” mouse brain.

To allow the selective preservation of functional MeCP2 expression within catecholaminergic cells, we crossed female MeCP2 deficient (MeCP2+/-) mice containing a

“stop-flox” MeCP2 allele (Guy et al., 2007) with transgenic mice expressing cre recombinase from an exogenous rat tyrosine hydroxylase (TH) promoter (TH-cre) (Savitt et al., 2005). For simplicity, we will refer to MeCP2-deficient mice as “Non-Rescue” mice, and MeCP2-deficient mice expressing cre recombinase in TH-positive cells as “Rescue” mice. To confirm the

reactivation efficiency of MeCP2 expression in “Rescue” mice, we employed dual-label

immunohistochemistry and quantified MeCP2 expression within the catecholaminergic regions

of the adult male “Rescue” mouse brain (Figure 13A-H). These results revealed that within the

substantia nigra, 87.5 ± 5.0% of the MeCP2-positive neurons stained positively for TH, and

conversely, 85.4 ± 3.4% of TH-positive neurons expressed MeCP2 (Figure 13C, D). In the locus ceruleus, 81.3 ± 6.2% of MeCP2-positive neurons expressed TH, and 81.9 ± 5.7% of TH-

positive neurons expressed MeCP2 (Figure 13G, H). These co-expression percentages were

comparable to that of adult wild-type littermate mice, where within the substantia nigra, 94.1 ±

3.1% of the MeCP2 positive neurons co-expressed TH, and 96.9 ± 0.9% of TH-positive neurons co-expressed MeCP2, and 92 ± 7.1% of MeCP2-positive neurons in the locus ceruleus co- expressed TH, and 95 ± 4.5% of TH-positive neurons co-expressed MeCP2 (Figure 13C, D, G,

H). Few, if any, MeCP2-positive cells were observed within the cortex, hippocampus, or

cerebellum of “Rescue” mice (Supplemental Figure 13A-C). However, although the large majority of MeCP2 expression was restricted to TH-positive cells throughout the “Rescue” mouse brain, reactivated MeCP2 expression was seen infrequently in some TH-negative cells within the periventricular and paraventricular nuclei of the hypothalamus, and in some neurons in the midbrain and brainstem of adult male “Rescue” mice (Figure 14 D-E). The ectopic expression of the rescue MeCP2 protein in a small cohort of TH-negative cells is not unexpected, however, as the TH-Cre transgenic mouse we employed has been shown previously to activate a stop-flox reporter gene in scattered cells within these same areas (Savitt et al., 2005).

Figure 13

Figure 13. MeCP2 is selectively preserved in catecholaminergic neurons of "Rescue"

mice. Panels A and B: Dual-label fluorescence micrographs at different magnifications from the

substantia nigra region showing immunoreactivity for MeCP2 (red channel), tyrosine

hydroxylase (green channel) and the nuclear stain DAPI (blue channel) in wild-type, "Non-

Rescue" and "Rescue" male mice. The scale bar for Panel A is 200 microns, and for Panel B is

50 microns. Panel C: The percentage of MeCP2-positive neurons that co-express TH in the

substantia nigra of male wild-type (n=3) and male "Rescue" (n=3) mice. Panel D: The

percentage of TH-positive neurons co-expressing MeCP2 in the substantia nigra of male wild- type (n=3) and male "Rescue" (n=3) mice. Panels E and F: Dual-label fluorescence micrographs

as above showing MeCP2 and tyrosine hydroxylase immunoreactivity in the locus ceruleus of

male wild-type, "Non-Rescue", and "Rescue" mice. Scale bar in Panel E is 200 microns, and is

50 microns in Panel F. Panel G: The percentage of MeCP2-positive neurons co-expressing TH

in the locus ceruleus of male wild-type (n=3) and male "Rescue" (n=3) mice. Panel H: The

percentage of TH-positive neurons co-expressing MeCP2 in the locus ceruleus of male wild-type

(n=3) and male "Rescue" (n=3) mice.

Figure 14

Figure 14. MeCP2 expression is not preserved in non-catecholaminergic neurons. Panel A:

Ectopic MeCP2 immunoreactivity in the cortex regions of a “Rescue” mouse at different magnifications. Panel B: MeCP2 immunoreactivity is not preserved within the hippocampus region of the “Rescue” mice. Panel C: Ectopic MeCP2 immunoreactivity is not observed within the cerebellum regions of the “Rescue” mice brain. Scale bars in all panels indicate 200 µm.

3.2.2. Preservation of MeCP2 in catecholaminergic cells extends the lifespan of male

MeCP2-deficient mice.

Analysis of Kaplan-Meyer survival plots revealed life expectancy to be significantly

longer in male “Rescue” mice than male “Non-Rescue” mice. In contrast to MeCP2-null “Non-

rescue” mice, which displayed a median survival age of 77 days, male “Rescue” mice displayed

a median survival age of 180 days, and more than 40% of the “Rescue” cohort survived longer

than 200 days (Figure 15A). Only one of the male “Non-Rescue” mice in our cohort lived longer than 100 days (Figure 15A). In addition to increased longevity, male “Rescue” mice displayed an overall improvement in general phenotypic severity compared to “Non-Rescue” mice. Using a phenotypic severity scale previously employed for MeCP2-deficient mice (Guy et al., 2007;

Robinson et al., 2012; Cheval et al., 2012) we found that male “Rescue” mice consistently displayed lower severity scores than “Non-Rescue” mice at and after 5 weeks of age (Figure 15B,

two-way ANOVA p<0.05). The increase in lifespan, and attenuation of phenotype severity,

occurred in the absence of body mass normalization, however, as male “Rescue” mice remained

significantly underweighted compared to age-matched wild-type mice, and not significantly

different from “Non-Rescue” mice (Figure 15C, two-way ANOVA p<0.001).

Figure 15

Figure 15. Survival and gross phenotypic behaviour are improved in male and female TH

"Rescue" mice. Panel A: Kaplan-Meier survival plot of male "Non-Rescue" (light grey line,

n=29) and male "Rescue" (dark grey line, n=17) mice. The life span of the "Rescue" mice is

significantly longer than "Non-Rescue" mice (p<0.01, Wilcoxon rank-sum test). Panel B: The

gross phenotypic severity score of male "Rescue" mice (n=10, dark grey line) is significantly

lower than male "Non-Rescue mice" (n=13, light grey line) at and after 35 days of age (p<0.05,

one-way ANOVA with Tukey's post-hoc test). The severity scores for male wild-type mice were between 0-1 over these ages (shown in closed circles). Panel C: The average body mass of male

"Rescue" mice (dark grey line, n=19-8 at different ages) does not significantly differ from male

"Non-Rescue" mice (light grey line, n=23-7 at different ages), and both groups remain

significantly underweighted compared to male wild-type mice (black line, n=23) (p<0.01, one- way ANOVA with Tukey's post-hoc test). Panel D: The spontaneous death rate in female

"Rescue" mice (3 of 33) is significantly lower than in female "Non-Rescue" mice (10 of 29;

p<0.05, Chi-Square test with one degree of freedom). Panel E: The gross phenotypic severity

score of female "Non-Rescue" mice (light grey line, n=8) and female "Rescue" mice (dark grey

line, n=10) does not significantly differ at any time between 30-50 weeks of age (one-way

ANOVA). Wild-type female mice severity scores were between 0-1 over these ages (shown in

solid circles). Panel F: The average body mass of female "Rescue" mice (dark grey line, n=12)

does not significantly differ from female "Non-Rescue" mice (light grey line, n=13), or female

wild-type mice (black line, n=17, one-way ANOVA with Tukey's post-hoc test).

3.2.3. Preservation of MeCP2 in catecholaminergic cells decreases the rate of sudden

unexpected death in female MeCP2-deficient mice

Although early lethality is not a typical phenotype of female MeCP2+/- mice (Chen et al.,

2001; Guy et al., 2001), female MeCP2-deficient mice are prone to sudden and unexpected death.

In our cohort, 34.5% (10 of 29) of the female MeCP2-deficient mice died suddenly and without

indication before reaching one-year of age, which is significantly higher than the rate of 4.8% (2

of 42) observed in female wild-type mice. In contrast, the spontaneous unexplained death rate in

female “Rescue” mice was significantly lower at 9.1% (3 of 33) (Figure 15D, p<0.05), and not

significantly different from the rate seen in female wild-type mice. This effect did not correlate

with significant improvements in the gross phenotypic severity of the cohort of “Rescue” mice,

however, as the general phenotypic severity score of the female “Rescue” mice did not

significantly differ from the female “Non-Rescue” mice (Figure 15E). Further, there was no correlation between the phenotypic severity score of a given mouse with its unexpected death; mice displaying low phenotypic severity scores were equally likely to die spontaneously as mice displaying higher cumulative severity scores (Figure 16). Analysis of growth rate in female

“Rescue” mice also failed to reveal any differences from female “Non-Rescue” mice. It should be noted, though, that on the pure C57Bl/6 background employed for this study, neither the

“Non-Rescue” or “Rescue” mice displayed significant body mass differences from female wild- type mice (Figure 15F).

Figure 16

Figure 16. The phenotypic severity score of female MeCP2+/- mice does not correlate with the time of their sudden and unexpected death. On the plot, the x-axis denotes the age of spontaneous death, while the y-axis indicates the severity score of the individual mouse one week before its death. Each point on the plot denotes an individual female MeCP2+/- mouse.

Linear regression analysis of these data revealed a Pearson R value of 0.133, indicating a lack of

correlation between severity score and time of spontaneous death (p=0.714).

3.2.4. Catecholaminergic preservation of MeCP2 improves deficits in ambulatory rate, motor coordination, and anxiety-like behaviour in male MeCP2-deficient mice

Consistent with previous reports (Samaco et al., 2009; Ward et al., 2011), 50-70 day old male “Non-Rescue” mice displayed clear impairments relative to wild-type and TH-cre mice in general activity, balance and coordinated movement, ambulatory rate, risk-assessment behaviour, and in nest-building performance. In the open field test, male “Non-Rescue” mice displayed on average a 72.3 ± 2% decrease in total activity counts, an 83.2 ± 3% decrease in rearing counts, and a 38.5 ± 3% decrease in ambulatory rate compared to age-matched male wild-type mice. In contrast, although remaining below wild-type levels, the general activity, total rearing behaviour, and ambulatory rate, of male “Rescue” mice were significantly improved from “Non-Rescue” mice (Figure 3A; p<0.05 for each behaviour). On the accelerating rotarod, male “Non-Rescue” mice displayed a shorter latency to fall time compared to wild-type or TH-cre mice on each of the four consecutive trial days. The latency to fall time for male “Rescue” mice was significantly longer than “Non-Rescue” mice on days 2-4 of the trial paradigm (Figure 17B, two- way ANOVA p<0.05). In the light/dark place preference test, male “Non-Rescue” mice conducted fewer risk-assessments while in the dark compartment compared to either wild-type or

TH-cre mice (Figure 17C). The risk assessment behaviour of male “Rescue” mice was significantly improved from “Non-Rescue” mice (2.34 ± 0.27 verses 1.2 ± 0.25 risk assessments/minute, respectively; p<0.05 one-way ANOVA; Figure 17B), but still below the performance of wild-type mice. Finally, in the nest-building test, male “Rescue” mice assembled nests with significantly larger volume than “Non-Rescue” mice (25.5 ± 7.3 cm3 versus 3.5 ± 0.3 cm3, respectively), (p<0.05 one-way ANOVA; Figure 17D).

Figure 17

Figure 17. Behavioural performances are improved in male "Rescue" mice. Panel A: Histogram

showing the mean and SEM of male "Rescue" mice (n=16, black) relative to male "Non-Rescue"

mice (n=17, dark grey) and male TH-cre control mice (n=24, light grey) in the open field test.

On the histogram, the average performance of male wild-type mice is denoted as 100% (dotted line). The general activity, rearing, and ambulatory rate of male "Rescue" mice was significantly

improved from male "Non-rescue" mice. Panel B: Motor coordination was assessed using the

accelerating rotarod test. Though remaining below the values of male wild-type (n=26) or TH- cre (n=21) control mice, male "Rescue" mice (n=20) displayed a significantly longer latency to fall than male "Non-Rescue mice" (n=17) on trial days 2, 3, and 4. Panel C: Anxiety-like

behaviour was assessed using the light-dark place preference test. The histogram shows the

mean and SEM of the number of risk assessment time (head poke) taken by the different cohorts

of male mice per minute of time spent in the dark chamber of the apparatus. The risk-assessment behaviour of male "Rescue" mice (n=12) was significantly above that of male "Non-Rescue" mice (n=12), but remained below the behaviour of male wild-type (n=20) or TH-cre controls

(n=7). Panel D: The nest-building test was used as an index of social behaviour. The histogram

shows the mean and SEM of the nest volume built by the different cohorts of mice 24 hours after

being placed in the test cage with a neslet. Male "Rescue" mice (n=13) assembled significantly

larger nests than male "Non-Rescue" mice (n=13), but smaller than male wild-type (n=14) or

TH-cre control mice (n=13). One-way ANOVA with Tukey's post-hoc test for multiple

comparisons was used for the open field, light-dark place preference, and nesting behaviour tests, and a two-way ANOVA (genotype verses trial) with Bonferroni's post-hoc test was used for the accelerating rotarod test. For each panel, * indicates p<0.05 compared between "Rescue" and

"Non-Rescue" mice.

3.2.5. Catecholaminergic preservation of MeCP2 improves the ambulatory and anxiety-like

behavioural deficits of adult female MeCP2-deficient mice

Although less well characterized to date, female MeCP2+/- mice also display significant

impairments in overall activity, ambulation rates, motor coordination, and anxiety-like behaviour after 8-12 months of age (Jugloff et al., 2008; Wither et al., 2012; Stearns et al., 2007).

Consistent with the results obtained in male “Rescue” mice, these behavioural deficits were also

largely improved in female “Rescue” mice. In the open field test, the general activity counts,

rearing behaviour, and average ambulation rate were each significantly improved compared to

female “Non-Rescue” mice (Figure 18A, one-way ANOVA p<0.01 for each). In the accelerating rotarod, female “Rescue” mice showed a partial rescue; their performance was significantly improved from female “Non-Rescue” mice on the 4th day of the trial paradigm (Figure 18B, two-way ANOVA p<0.05). In the light/dark place preference test, female “Rescue” mice displayed a significant improvement in risk-assessment behaviour compared to age-matched female “Non-Rescue” mice (Figure 18C, one-way ANOVA p<0.05). In fact, the risk assessment behaviour of female “Rescue” mice did not significantly differ from that of female wild-type

mice. In the nest-building test, female “Rescue” mice assembled nests that were significantly

larger in total volume than those of female “Non-Rescue” mice (Figure 18D, one-way ANOVA

p<0.05), and equivalent in volume to those assembled by female wild-type or TH-cre mice

(Figure 18D).

Figure 18

Figure 18. Behavioural performances are improved in female "Rescue" mice. Panel A:

Histogram showing the normalized mean and SEM of the performance of female wild-type

(n=19, denoted by dotted line at 100%), female TH-cre control (n=7, light grey), female "Non-

Rescue" (n=15, dark grey), and female "Rescue" (n=14, black) in the open field test. The total

activity levels, rearing, and ambulatory rate were each significantly improved in the "Rescue"

mice compared to "Non-Rescue" mice, and the rearing and ambulatory rate of the "Rescue" mice

were restored to wild-type control levels. Panel B: Graph showing the mean and SEM for the

latency to fall for the different cohorts of mice on the accelerating rotarod for each day of the

successive trial paradigm. On each trial day, female "Non-Rescue" mice (n=20) displayed significantly shorter latency to fall than either female wild-type (n=24) or female TH-cre (n=12) control mice. While the average performance of the Female "Rescue" mice (n=15) was above

"Non-Rescue" mice on each trial day, their performance reached statistical significance only on trial day 4 (p<0.05, two-way ANOVA (genotype verses trial) with Bonferroni's post-hoc correction). Panel C: Histogram showing the mean and SEM of the risk-assessment behaviour

(headpokes per minute in dark) of the female cohorts of mice. The anxiety-like behaviour of female "Rescue" mice (n=14) is significantly improved compared to female "Non-Rescue" mice

(n=19), and is not significantly different from female wild-type (n= 15) or female TH-cre control mice (n=7). Panel D: Histogram showing the mean and SEM of the nest volume built by the

different cohorts of female mice 24 hours after being placed in the test cage with a neslet. The

nest volume assembled by female "Non-Rescue" mice (n=9) was significantly less than either

female wild-type (n=9) or female TH-cre control mice (n=6). Female "Rescue" mice (n=14)

assembled nests with volumes significantly larger than "Non-Rescue" mice, that were not

significantly different from wild-type or TH-cre controls. For each of the behaviours assessed,

88 the performances of female wild-type and female TH-cre mice were not significantly different.

Statistical analysis for the open field, light-dark place preference, and nest building tests was done using one-way ANOVA with Tukey’s post hoc correction. On each panel, * indicates p<0.05 significance between female "Rescue" and female "Non-Rescue" mice.

3.2.6. Preservation of MeCP2 in catecholaminergic cells improves cortical EEG abnormalities in male, but not female, MeCP2-deficient mice

Previous studies have demonstrated the presence of spontaneous epileptiform discharge activity in the cortex of both male and female MeCP2-deficient mice (D’Cruz et al., 2010;

Wither et al., 2012). In our current cohort, 100% (9 of 9) of the male “Non-Rescue” mice examined displayed cortical epileptiform-like discharges (Figure 19A, B), while no discharge activity was observed in any of the age-matched male wild-type mice (n=5). In these male

“Non-Rescue” mice, the average number of discharge events was 42.1 ± 7.9 per hour, with each discharge having an average duration of 2.8 ± 0.7 seconds, and an average frequency of 6.2 ± 0.3

Hz (Figure 19E, F). Each of the male “Rescue” mice retained spontaneous cortical discharge activity (Figure 19C, D), and there was no significant change in their average duration or frequency (2.2 ± 0.6 seconds and 6.5 ± 0.36 Hz, respectively) as compared to male “Non-Rescue” mice (Figure 19E, F). However, the average number of discharge events in the “Rescue” mice was significantly decreased from that of male “Non-Rescue” mice to 20.5 ± 6.3 discharges per hour (Figure 19G, p<0.05, one-way ANOVA).

While a significant decrease in average cortical discharge activity was seen in male

“Rescue” mice, the same was not observed in female “Rescue” mice. Cortical discharge activity was observed in female “Non-Rescue” mice (Figure 20A, B), with an average incidence rate of

69.6 ± 15.5 events per hour (n=5). Cortical discharge activity was also observed in female

“Rescue” mice (Figure 20C, D), and there were no significant differences in its average duration or frequency compared to female “Non-Rescue” mice (Figure 20E, F). Unlike the decreased incidence rate observed in male “Rescue” mice, however, the incidence rate of cortical discharge

90 activity in female “Rescue” mice (n=6) was 67.1 ± 13.9 per hour, which did not significantly differ from the incidence rate seen in female “Non-Rescue” mice (Figure 20G).

Figure 19

Figure 19. The incidence rate of cortical epileptiform discharge activity is reduced in male

"Rescue" mice. Panels A and B: Representative EEG recording traces showing 10 minutes of

cortical activity (Panel A), or 10 seconds of activity surrounding a discharge event (Panel B),

from a male "Non-Rescue" mouse. The upper and lower traces in Panel A are contiguous in time, and individual discharge events are highlighted above the trace. Panels C and D: Representative

EEG traces from a male "Rescue" mouse. As above, Panel C shows 10 minutes of continuous

activity with dicharges highlighted, and Panel D shows a single discharge event. Cortical EEG

activity in wild-type mice (n=6) was also assessed, and no discharge activity was seen in any of

the individual mice (traces not shown). Panel E: Histogram showing the mean and SEM of the

discharge duration for cortical epileptiform events in male "Non-Rescue" (n=9) and male

"Rescue" (n=8) mice. As shown, no differences in the average discharge duration of the individual discharge events was observed. Panel F: Histogram showing the mean and SEM of

the frequency observed for individual discharges in male "Rescue" and "Non-Rescue" mice. No

differences in the average individual discharge frequency was observed between groups. Panel

G: Histogram showing the mean and SEM for the number of epileptiform discharge events per

hour in male "Rescue" mice (n=8) and male "Non-Rescue" mice (n=9). The number of discharge

events in the "Rescue" mice per hour was significantly decreased from those of male "Non-

Rescue" mice (p<0.05, Student's unpaired t-test).

Figure 20

Figure 20. The incidence rate of epileptiform discharge activity is not improved in female

"Rescue" mice. Panels A and B: Representative trace showing 10 minutes of continuous cortical

EEG activity with discharge events highlighted (Panel A), and a representative 10 second trace

showing the waveform of a typical discharge event (Panel B) from a female "Non-Rescue"

mouse. Panels C and D: Representative 10 minute continuous trace (Panel C) and single

discharge event (Panel D) from a female "Rescue" mouse. Female wild-type mice of 280-400

days of age (n=8) were also assessed, but no discharge activity was detected in any subject (not

shown). Panel E: Histogram showing the mean and SEM of the cortical discharge duration seen

in female "Non-Rescue" (n=7) and female "Rescue" (n=8) mice. No differences in average

discharge duration was observed. Panel F: Histogram showing the mean and SEM of the

discharge event frequency for female "Rescue" and female "Non-Rescue" mice. No differences

in discharge frequency were observed between groups. Panel G: Histogram showing the mean

and SEM for the epileptiform discharge event number per hour seen in female "Rescue" mice

(n=8) and female "Non-Rescue" mice (n=7). No significant differences in the number of discharge events per hour were seen between groups (p=0.87, Student's unpaired t-test).

3.2.7. Preservation of MeCP2 in catecholaminergic cells increases peak hippocampal theta

frequency in male, but not female, MeCP2-deficient mice

In addition to possessing epileptiform-like EEG discharge activity, we have shown

previously that neural network oscillatory activity is altered in male and female MeCP2-deficient mice (D’Cruz et al., 2010; Wither et al., 2012). In our current cohort of male “Non-Rescue” mice, the peak hippocampal theta frequency was found to be 7.2 ± 0.1 Hz, compared to 8.7 ± 0.2 Hz for wild-type male mice (p<0.05, one-way ANOVA) (Figure 21B, E). In male “Rescue” mice, this peak theta frequency shift was partially restored. Although still slower than wild-type, the peak theta frequency of male “Rescue” mice during exploratory behaviour was 7.8 ± 0.1 Hz

(p<0.05 relative to “Non-Rescue”, one-way ANOVA). In female “Non-Rescue” mice, the alteration in peak hippocampal theta frequency was less pronounced, but still significantly slower than that of female wild-type mice (Figure 22B, E). Peak hippocampal theta frequency in female “Non-Rescue” mice was 7.9 ± 0.3 Hz, as compared to 8.9 ± 0.2 Hz in female wild-type mice (p<0.05, one-way ANOVA). In contrast to the positive effect seen in males, however, the peak hippocampal theta frequency during exploratory behaviour in female “Rescue” mice was

8.0 ± 0.1 Hz, which did not significantly differ from female “Non-Rescue” mice (Figure 22E).

Figure 21

Figure 21. The peak hippocampal theta frequency and the total hippocampal gamma activity power are significantly improved in male "Rescue" mice. Panel A: Representative power spectrum of hippocampal EEG activity during exploration from male wild-type (black),

"Rescue" (dark grey), and "Non-Rescue" (light grey) mice. For this plot, the highest peak in the theta range of the raw data was set at the 100% level, and the remaining frequency powers were normalzed to this peak value. Panel B: Representative normalized hippocampal power spectra showing specifically the theta range from male wild-type (black), "Rescue" (dark grey) and

"Non-Rescue" (light grey) mice during periods of exploratory behaviour. Note the shift in peak theta power frequency seen between the different mice. Panel C: Representative spectral plots showing raw activity power (e.g., non-normalized) within the gamma frequency range (35-60 Hz) in male wild-type (black), "Rescue" (dark grey), and "Non-Rescue" (light grey) mice. Panel D:

Histogram showing the mean and SEM of the total hippocampal non-normalized spectral power of the theta range during exploratory behaviour of male wild-type (n=5), "Non-Rescue" (n=9), and "Rescue" (n=7) mice. No significant differences in total theta power was observed, indicating that the MeCP2-deficiency does not cause a ubiquitous decrease in hippocampal EEG spectral power. Panel E: Histogram showing the mean and SEM for the observed peak theta frequency during explortatory behaviour for wild-type, "Non-Rescue", and "Rescue" mice. The n numbers for each group are the same as above. The peak theta frequency is significantly lower in "Non-Rescue" mice than in wild-type mice, while the peak theta frequency of "Rescue" mice is significantly greater than "Non-Rescue" mice. While improved, the peak theta frequency in the "Rescue" mice remained significantly below the peak frequency of wild-types. Panel F:

Histogram showing the mean and SEM of the total hippocampal non-normalized spectral power in the gamma range during exploratory behaviour of male wild-type (n=5), "Non-Rescue" (n=5),

98 and "Rescue" (n=7) mice. Unlike the preservation of total theta range power, total gamma range power was significantly diminished in male "Non-Rescue" mice. In "Rescue" mice, total gamma power was significantly improved from "Non-Rescue" mice, and did not significantly differ from the levels of wild-type mice. For these panels, * indicates significance levels of p<0.05 between the indicated groups (one-way ANOVA with Tukey’s post hoc correction).

3.2.8. Preservation of MeCP2 in catecholaminergic cells rescues deficits in hippocampal gamma band oscillatory activity in male, but not female, MeCP2-deficient mice

Power spectrum analysis was then conducted to compare the composite power of

frequencies in the theta and gamma ranges between wild-type, “Non-Rescue”, and “Rescue” mice of both genders. Analysis of total raw theta activity power revealed no significant changes in total power of the 6-12 Hz theta frequency range between groups (Figure 21D, 22D). Male wild-type mice displayed total theta power of 385 +/- 56 µV2/Hz, which was not significantly

different from “Non-Rescue” male mice, whose total theta power was 295 +/- 9 µV2/Hz (p>0.2,

one-way ANOVA). The total theta power for male “Rescue” mice was 345 +/- 29 µV2/Hz,

which also did not differ significantly from either wild-type or “Non-Rescue” male mice (Figure

21D). In contrast to the preservation of total theta power, however, significant changes were

observed between groups for total gamma band power. Analysis of total raw EEG power from

35-60 Hz revealed a significant decrease in total gamma band power in male “Non-Rescue”

compared to wild-type mice (124 +/- 6 µV2/Hz verses 235 +/- 25, respectively; p<0.01 one-way

ANOVA). In male “Rescue” mice, however, total gamma power in this range was restored to

261 +/- 31 µV2/Hz, which was significantly improved from “Non-Rescue” mice, and not significantly different from wild-type mice (Figure 21F).

As with male mice, no significant differences in total hippocampal theta power were observed between any of the female mouse groups during exploratory behaviour (Figure 22D).

Total theta power in female wild-type mice was 396 +/- 44 µV2/Hz (n=4), was 424 +/- 101

µV2/Hz for female “Non-Rescue” mice (n=5), and was 415 +/- 74 µV2/Hz for female “Rescue”

mice (n=5). Consistent with what was observed in male mice, the total hippocampal gamma

band power was also significantly diminished in female “Non-Rescue” mice compared to female

100 wild-type mice (118 +/- 11 µV2/Hz verses 295 +/- 47 µV2/Hz, respectively, p<0.01 one-way

ANOVA). However, unlike the improvement in gamma power seen in male “Rescue” mice, no improvement in total gamma band power was observed. The total gamma power in female

“Rescue” mice during exploratory behaviour was found to be 164 +/- 16 µV2/Hz, which did not differ significantly from female “Non-Rescue” mice, and remained significantly below the value of female wild-type mice (Figure 22F).

101

Figure 22

102

Figure 22. The peak hippocampal theta frequency and the total power of hippocampal gamma

activity are not improved in female "Rescue" mice. Panel A: Representative normalized power

spectrum of hippocampal EEG activity during exploration from female wild-type (black),

"Rescue" (dark grey), and "Non-Rescue" (light grey) mice. Panel B: Representative normalized

hippocampal power spectra of the theta range from male wild-type (black), "Rescue" (dark grey) and "Non-Rescue" (light grey) mice during periods of exploratory behaviour. Panel C:

Representative spectral plots showing raw gamma frequency range power distribution in female

wild-type (black), "Rescue" (dark grey), and "Non-Rescue" (light grey) mice. Panel D:

Histogram showing the mean and SEM of the total non-normalized hippocampal theta spectral

power during exploratory behaviour of female wild-type (n=4), "Non-Rescue" (n=5), and

"Rescue" (n=7) mice. No significant differences in total theta power was observed between groups. Panel E: Histogram showing the mean and SEM for the peak theta frequency during

explortatory behaviour for wild-type, "Non-Rescue", and "Rescue" mice. Consistent with what

was seen in male "Non-Rescue" mice, the peak theta frequency is significantly lower in female

"Non-Rescue" mice than in female wild-type mice. However, the peak theta frequency of female

"Rescue" mice is not significantly altered, and remains significantly below the peak theta

frequency of female wild-type mice. Panel F: Histogram showing the mean and SEM of the total

hippocampal non-normalized gamma range spectral power during exploratory behaviour in

female wild-type, "Non-Rescue", and "Rescue" mice. Total gamma-band power is significantly

reduced from wild-type levels in "Non-Rescue" mice. In contrast to the effect seen in male

"Rescue" mice, though, total gamma power was not significantly improved in female "Rescue"

mice, and remained significantly below the levels of wild-type mice. For these panels, *

103

indicates significance levels of p<0.05 between the indicated groups (one-way ANOVA with

Tukey’s post hoc correction).

104

Chapter 4

Discussion

4.1 Part 1: Delayed global reactivation of MeCP2 expression

Mice models with disrupted MeCP2 expression recapitulate many cardinal features of

RTT, including motor impairments and anxiety issues (Chen et al., 2001, Guy et al., 2001, Guy et al., 2007; Jugloff et al., 2008; Samaco et al., 2009; Ward et al., 2011). Dendritic branching, neuronal packing density, neuronal size, and membrane properties are similarly altered in

MeCP2-deficient mice and RTT patients. Neurodegeneration, however, is not observed (Kishi and Macklis, 2004). Neurons appear to remain in an immature state, and may not be irreversibly compromised. Indeed, as MeCP2 is ubiquitously reactivated in MeCP2-deficient mice, life span, phenotypic severity, respiratory measures, and neuronal morphology are significantly improved

(Guy et al., 2007; Robinson et al., 2012). However, the ubiquitously rescued mice did not completely revert to a wild-type like profile. The incomplete rescue may be partially explained by the incomplete restoration of MeCP2 levels, as well as potentially irreversible neurological deficits caused by a lack of MeCP2 during early development. These previous studies reactivated

MeCP2 at ~70% efficiency and yielded significant improvements. It is still unknown if lower

increases in MeCP2 levels will yield similar benefits and whether there is a threshold for MeCP2

reactivation efficiency for significant rescue. In addition, although RTT can predominately affect

female patients, many of the rescue investigations have been focused on male MeCP2-deficient mouse models. Due to the X-linked nature of RTT, MeCP2 deficient male mice display a much more severe phenotypic profile and may potentially have a larger window of improvement compared to the mildly affect female MeCP2-deficient mice (Guy et al., 2007). Thus, treatments

105

that improve the deficits of male MeCP2-deficient mice may not have the same efficacy in female MeCP2-deficient mice. Reactivation of MeCP2 in female MeCP2-deificient mice

completely rescued the gross phenotypic severity and hippocampal long-term potentiation

impairments (Guy et al., 2007). However, other phenotypic aspects, including motor

coordination, anxiety, ambulation, circadian activity, and thermoregulation may remain impaired

in MeCP2 reactivated female mice. The purpose of the present study was to investigate whether

low MeCP2 reactivation efficiency, <40%, yields similar beneficial effects as reactivating

MeCP2 at ~60-70% efficiency and if reactivation of MeCP2 in female MeCP2-deficient mice

improves the behavioural outcomes, daily rhythmic activity, and thermoregulation.

MeCP2-deficient mice were generated by inserting a floxed-stop cassette into the

endogenous MECP2 gene (Guy et al., 2007). These Stop/y male mice express MeCP2 at

approximately 1-5% of wild-type level, though they are phenotypically identical to MeCP2-null

mice. Previous work from our lab showed a strong correlation between MeCP2 levels and

behavioural performances; mice with greater MeCP2 levels performed better in sensory-motor

tests (Wither et al., 2012). Consistent with these findings, in this study we showed that a modest

increase in MeCP2 expression (~20% of wild-type protein levels) can significantly improve life

span and mildly improve motor skills. Greater MeCP2 reactivation levels (~60% of wild-type

levels) produced more pronounced improvements in sensory-motor tasks. If MeCP2 reactivation

efficiency correlates to phenotypic reversal, then to achieve complete rescue, MeCP2

reactivation must be close to 100%. As MeCP2 has been shown to be an important epigenetic

factor during brain development, there may be deficits that are not completely reversible through

MeCP2 reactivation (Guy et al. 2011).

106

It is known that certain systems such as the forebrain, Hox1b domain, and GABAergic

neurons have more stringent MeCP2 requirements than other regions, as ablation of MeCP2 from

these systems leads to the development of RTT phenotypes (Chao et al., 2010; Jugloff et al.,

2008; Ward et al., 2011). Our previous work has also shown that MeCP2 levels in different brain

regions correlate with different behaviour (Wither et al., 2012). In this study, it is highly likely

that MeCP2 was not reactivated equally throughout the brain. Thus, the variation between behavioural performances of mice with similar level of global MeCP2 reactivation may be attributed to differential MeCP2 reactivation profiles across brain regions. Further investigation is required to further examine what type of MeCP2 reactivation profile is sufficient, and the most

efficient, to rescue neurological symptoms.

RTT is predominately caused by mutations of the MECP2 gene on one allele (Amir et al.,

1999). MeCP2-deficient female mice express MeCP2 in a mosaic profile and are therefore a

more appropriate model of the disorder (Chen et al., 2001; Guy et al., 2001). MeCP2-deficient

female mice express MeCP2 at ~50% of wild-type, and as a result are more mildly affected than male MeCP2-deficient mice. In addition, early lethality is not a common feature of female

MeCP2-deficient mice. Like the human condition, the MeCP2-deficient female mice show delayed development and altered phenotypes including respiratory abnormalities, motor impairments, anxiety behaviour, and tremors (Guy et al., 2007). Our recent work has shown that thermoregulation, core body temperature, and circadian activity are also altered in MeCP2- deficient female mice (Wither et al., 2012). All of these impairments are significantly improved following MeCP2 reactivation, suggesting that these deficits are not irreversible and the underlying mechanisms required for proper function remains intact. Importantly, rearing activity and circadian activity was restored to wild-type levels. However, anxiety-like behaviour and

107

exploratory activity were not rescued, although targeted reactivation of MeCP2 in specific

neuronal groups has been shown to correct these deficits (Jugloff et al., 2008). This further suggests that not all parts of the brain was reactivated equally in Stop/+,cre mice.

Gene therapy, although an attractive avenue, is currently unavailable for the treatment of

RTT. Restoration of MeCP2 significantly improves various phenotypes, however,

overexpression of MeCP2 leads to detrimental effects (Collins et al., 2004; Na et al., 2012). Our

study extends upon previous studies, and shows that the level of improvement and rescue in

MeCP2-deficient male mice directly correlates with the extent of MeCP2 reactivation. Even moderate increases of MeCP2 levels provide beneficial results. Methods that can increase

MeCP2 expression at low levels may improve RTT-like phenotypes without the risk of

overexpression. Further, our study not only showed that the symptoms of male MeCP2-deficient mice can be rescued, but that the more clinically relevant female MeCP2-deficient animals can also benefit significantly from MeCP2 reactivation. A caveat of this study is that MeCP2 is reactivated after symptoms develop. It remains to be investigated whether early reactivation of

MeCP2 will lead to greater improvements of neurological symptoms. Our findings are in agreement with the hypothesis that RTT-like phenotypes can be remediated and further highlight

the need to understand the functions of MeCP2 in different regions of the brain.

108

4.2 Part 2: Selective preservation of MeCP2 functions in catecholaminergic cells

The catecholaminergic system has been postulated to play a role in RTT pathophysiology

(Nomura et al., 1985; Segawa et al., 1997). This is a reasonable speculation as many impairments

of RTT overlap with disorders involving dopaminergic or noradrenergic disturbances (Samaco et

al., 2008). In recent years, noradrenergic and dopaminergic neurons lacking MeCP2 have been

shown to be impaired morphologically and functionally (Taneja et al., 2009; Gantz et al., 2011).

Importantly, these neurons do not undergo neurodegeneration but lose their mature TH

expressing phenotype, suggesting that the deficits may not be irreversible (Roux et al., 2009). In

fact, targeted restoration of MeCP2 to other brain regions, including the forebrain, HoxB1

domain, and glia cells, were able to improve RTT associated impairments (Alvarez-Saavedra et

al., 2007; Jugloff et al., 2008; Ward et al., 2011; Lioy et al., 2011). Here we extend upon this

study by illustrating that the selective preservation of MeCP2 in catecholaminergic cells of both

male and female MeCP2 deficient mice can rescue specific RTT-like behavioural and

neurophysiological deficits. Our results therefore not only add to the growing evidence that

MeCP2 dysfunction in catecholaminergic cells plays a significant role in RTT pathophysiology,

but also provide the proof-of-principle that the selective targeting of these cells is sufficient to

improve several of the cardinal RTT-like deficits seen in MeCP2 deficient mice.

The positive effect on lifespan and survival rates associated with preserving MeCP2 function in catecholaminergic cells is consistent with previous suggestions that deficits in catecholaminergic neuronal function play a key role in the early lethality and sudden unexpected death seen in both this mouse model and in clinical RTT syndrome (Johnston et al., 2001; Julu et al., 1997). For example, treating MeCP2-null mice with the norepinephrine uptake inhibitor, desipramine, significantly extends lifespan (Roux et al., 2007; Zanella et al., 2008), and as

109

discussed above, the preservation of MeCP2 function in HoxB1-expressing neurons of the brain

stem and spinal cord also significantly extends the lifespan of male MeCP2-deficient mice (Ward

et al., 2011). Although our data show clearly that preserving MeCP2 expression in

catecholaminergic cells significantly extend survival times in male MeCP2-deficient mice, it is worth noting that the lifespan extension seen by MeCP2 reactivation in HoxB1 neurons was more robust than what we observed in the current study; only about 1/3 of male HoxB1 “Rescue” mice died prior to 45 weeks of age (Ward et al., 2011), while about half of our male TH “Rescue” mice had succumbed by 30 weeks of age. Thus, while our results show preserving MeCP2 function in catecholaminergic neurons provides benefit, the benefit does not reach the same magnitude as what can be achieved by preservation of MeCP2 expression in both catecholaminergic and subsets of non-catecholaminergic neurons. This suggests that impairments of non-catecholaminergic neural circuits (at least some of which express HoxB1) also contribute significantly to the longevity of male MeCP2-deficient mice, and that full rescue of this phenotype will require less restrictive cell-type preservation of MeCP2 function.

The beneficial effect on longevity was not restricted to male mutants, as heterozygous female “Rescue” MeCP2-deficient mice, which are the gender-appropriate model for clinical

RTT, also displayed a significant diminution in spontaneous death rates. In fact, the sudden and unexpected death rates in our female “Rescue” cohort did not differ significantly from that of wild-type female mice up to one year of age. This outcome is not necessarily surprising given the mosaic nature of female MeCP2-deficient mice, where on average endogenous MeCP2 is expressed in half of the cells in the body – which by itself prevents the early lethality seen in male MeCP2-deficient mice. Based on the average reactivation efficiency for MeCP2 observed in TH-positive neurons of the substantia nigra and locus ceruleus in male subjects (85%), the

110 expression of MeCP2 in catecholaminergic cells of female “Rescue” mice would be expected to increase from roughly 50% on average to more than 90% in these target neuronal populations.

Thus, our data suggest that this increase in MeCP2, occurring on the already existing mosaic expression pattern of MeCP2 throughout the brain of female MeCP2-deficient mice, is able to completely abrogate the sudden and unexpected death seen in this gender-appropriate RTT mouse model.

In addition to improving longevity and diminishing spontaneous death rates, our data also show that the catecholaminergic specific preservation of MeCP2 partially rescues behavioural deficits in ambulation, anxiety-like behaviour, motor coordination, and nest-building in both male and female MeCP2-deficient mice. As pharmacologic and genetic manipulations have established a role for noradrenergic and dopaminergic systems in these behaviours (Szczypka et al., 2001; Thomas and Palmiter, 1997), the observed improvements in the respective “Rescue” mice are consistent with heightened overall catecholaminergic function in the “Rescue” mice.

Intriguingly, Samaco et al. (2009) demonstrated that the selective ablation of MeCP2 from only catecholaminergic cells using a similar TH-cre transgenic mouse was sufficient to cause hypoactivity, impaired motor performance, and elevated anxiety-like levels in mice. Our results complement this study, therefore, by showing the preservation of MeCP2 in only catecholaminergic cells is sufficient to improve the performance of both male and female

MeCP2-deficient mice in a host of behavioural tasks. Collectively these results highlights the importance of MeCP2 function in the catecholaminergic system, and show that retaining MeCP2 function in this relatively small number of neurons is able to partially overcome the impact of a global loss of MeCP2 function across non-catecholaminergic neural systems of mice.

111

In addition to assessing behavioural outcomes, a novel component of our study was the

examination of cortical and hippocampal EEG activity in the male and female “Rescue” mice.

Our previous work revealed the presence of EEG abnormalities in both male and female MeCP2-

deficient mice, which include the presence of spontaneous high amplitude epileptiform-like

discharges in somatosensory cortex (D'Cruz et al., 2010; Wither et al., 2012), and a shift in the

average peak frequency of hippocampal theta wave activity during exploratory behaviour

(D’Cruz et al., 2010). In this study, we also show that hippocampal gamma band power during

exploratory behaviour is significantly reduced in both male and female MeCP2-deficient mice;

an observation that comports with data from a recent report that also showed gamma-band

abnormalities in a related strain of MeCP2+/- mice (Liao et al., 2012). Consistent with male

MeCP2-deficient mice being more severely affected than female MeCP2-deficient mice, the average discharge duration, the peak theta frequency, and the magnitude of total gamma power attenuation were more pronounced in male MeCP2-null mice than heterozygous female

MeCP2+/- mice.

Intriguingly, though, significant improvements in EEG abnormalities were only evident

in male “Rescue” mice. This gender-specific outcome was unexpected, as we anticipated the less robust phenotypic deficits of female MeCP2-deficient mice would be more amenable to correction by selective catecholaminergic MeCP2 reactivation. One possible explanation for the selective improvement in male “Rescue” mice might be a “ceiling” response, where the restoration of MeCP2 in catecholaminergic neurons was sufficient to partially improve these

EEG phenotypes of male MeCP2-null mice, but not sufficient to completely restore the phenotypes to wild-type levels. Due to their mosaic nature, these EEG phenotypes of female

MeCP2+/- mice might already be at the ceiling that could be afforded by catecholaminergic-

112 specific MeCP2 preservation. Irrespective of mechanism, these results do raise an important point: they provide evidence that interventions that show positive effects in male MeCP2- deficient mice may not show the same beneficial effects in mosaic female MeCP2-/+ mice. Given that the vast majority of clinical RTT cases affect females, these results highlight the importance of including female MeCP2-deficient mice in future translational studies.

113

4.3 Future Directions

Our study demonstrated that the extent of rescue is dependent upon the level of MeCP2

reactivation efficiency. However, we were not able to achieve a full spectrum of MeCP2

reactivation efficiency. Future investigations should utilize more diverse tamoxifen injection

schemes to yield diverse levels of MeCP2 reactivation and examine the correlation between

MeCP2 reactivation levels and phenotypic reversal. Tamoxifen was systematically injected to

induce MeCP2 reactivation. It remains unknown, however, whether MeCP2 was reactivated

equally throughout the brain or if different brain structures exhibited differential levels of

MeCP2 reactivation. As studies have shown that different brain regions contribute to different

symptoms of RTT and certain brain regions such as the forebrain and HoxB1, are critical for

survival, it would be worthwhile to assess MeCP2 reactivation levels across brain regions

(Alvarez-Saavedra et al., 2007; Jugloff et al., 2008; Ward et al., 2011).

In our study, MeCP2 was systematically reactivated after the onset of symptoms.

Irreversible damage, however, may be caused by the lack of MeCP2 during early development.

A future direction for the study is to reactivate MeCP2 at different times of development, from

neonatal stages to pre-symptomatic stages. Critical periods during development that requires

MeCP2 function have been suggested by Cheval et al. (2012). Their group ablated MeCP2 function from mice at different stages of life and found two sensitive time periods, 11 and 39 weeks of age, in which the deletion of MeCP2 resulted in premature death.

In addition, reversibility of electroencephalography abnormalities, neurotransmission, synaptic plasticity, neurochemistry, and various other deficits of RTT have not been examined.

Thus far, reactivation studies have focused on the reversibility of gross phenotypes and

114

behavioural deficits associated with RTT (Guy et al., 2007; Robinson et al., 2012). It remains

unknown whether the restoration of MeCP2 completely rescues neuronal deficits such as altered

membrane capacitance, soma size, and dendritic complexity. More detailed investigations are

required to explore the full extent of rescue achieved by delayed MeCP2 reactivation.

RTT is classified as a neurological disorder (Hagberg et al., 1985; Hagberg and Witt-

Engerstrom, 1986). This is supported by the finding that the deletion of MeCP2 from only the

CNS is able to recapitulate all the phenotypes of RTT (Chen et al., 2001). Recent findings have also shown that MeCP2 is critically important in glia cells (Ballas et al., 2009; Boison, 2012).

MeCP2, however, is known to be expressed in different cell types and tissue. Reports have shown that MeCP2 is expressed in the lungs and spleen as well (Shahbazian et al., 2001).

Various symptoms of RTT involve abnormalities in the periphery, although, dysregulation from the CNS is thought to be the source of these abnormalities (Glaze, 2005; Jellinger et al., 1990).

MeCP2, however, is a globally expressed transcriptional regulator and as such, it should exert an effect in all cell types (Amir et al., 1999; Guy et al., 2011). Further research is required to characterize the role of MeCP2 in the periphery and determine whether specific symptoms of

RTT are caused by alterations of peripheral functions. Specifically, MeCP2 function in the autonomic system should be carefully examined as disturbances of this system likely contribute to early lethality, anxiety-like behaviour, respiratory dysfunction, and cardiac arrhythmia (Ward et al., 2011; Weese-Mayer et al., 2006).

In this study, we demonstrated that the selective preservation of MeCP2 in catecholaminergic cells can improve life span, behavioural performances, and electroencephalography abnormalities. While these results suggest that the catecholaminergic system is a potential target for therapeutic treatment, additional parameters remain

115

uninvestigated. Bioamine levels are intrinsically linked to catecholaminergic centers of the brain

as they are the source of these neurotransmitters (Roux and Villard, 2010). Though bioamine

alterations in RTT patients are still under debate, the consensus is that dopamine and

norepinephrine levels are reduced (Panayotis et al., 2011; Samaco et al., 2008; Zoghbi et al.,

1989). Theoretically, if MeCP2 is returned to the catecholaminergic neurons, the production of bioamines should be improved. This remains an area to be investigated. It is also plausible that the preservation of MeCP2 in these neurons preserved their morphology and neural transmitter function. Since the catecholaminergic systems innervate the majority of the brain, the beneficial effects observed in our study may be due to a non-cell autonomous effect. Thus, it is worthwhile in the future to examine whether the catecholaminergic neurons with MeCP2 function preserved is similar to wild-type catecholaminergic neurons.

Respiratory and cardiac deficits are detrimental to RTT patients (Acampa and Guideri,

2006; Rohdin et al., 2007). These impairments have been suggested to be the cause of sudden unexpected death (Ohya et al., 2005). These phenotypes have not been examined in our study. It is worthwhile to explore the extent of rescue in respiratory patterns and cardiac function in our

“Rescue” mouse line. Previous study has shown that the administration of desipramine, a norepinephrine reuptake inhibitor, can improve the breathing abnormalities of MeCP2-deficient mice (Roux et al., 2007). Thus, it is logical that our “Rescue” mice should share similar improvements. This, however, remains to be investigated in the future. Cardiac function should be tested in the future using electrocardiography recordings. As Q-T intervals are prolonged in

RTT patients (Ellaway et al., 1999), we can test whether this phenomenon is corrected in our

“Rescue” mice.

116

Our study has added further evidence towards the reversibility of RTT. In conjunction

with other rescue studies, behavioural deficits, autonomic deficits, electroencephalography

abnormalities, respiratory dysfunction, neuronal properties, and premature death can all be

rescued, at least partially, through reactivation of MeCP2 to the whole brain or selective regions

(Guy et al., 2007; Ward et al., 2011; Lioy et al., 2011; Jugloff et al., 2008). There are several

other brain regions that may be of particular interest such as the hippocampus, thalamus,

brainstem, respiratory regulator nuclei, and hypothalamus. The normal function of these regions

suggests that they are involved with cognitive deficits, seizure development, motor impairments,

respiratory dysfunction, and autonomic dysregulation observed in RTT (Boison, 2012; Fyffe et al., 2008; Gadalla et al., 2011; Julu et al., 1997). Targeted deletion or rescue studies should be attempted in these regions to explore how these different brain regions contribute to the different symptoms of RTT. Further, these targeted studies may also reveal a critical center that may produce the greatest level of rescue. If so, gene therapy and pharmaceutical approaches may be greatly improved.

117

References:

Acampa, M., and Guideri, F. (2006). Cardiac disease and Rett syndrome. Arch Dis Child 91, 440-443.

Acheson, A., Conover, J.C., Fandl, J.P., DeChiara, T.M., Russell, M., Thadani, A., Squinto, S.P., Yancopoulos, G.D., and Lindsay, R.M. (1995). A BDNF autocrine loop in adult sensory neurons prevents cell death. Nature 374, 450-453.

Adachi, M., Autry, A.E., Covington, H.E., 3rd, and Monteggia, L.M. (2009). MeCP2-mediated transcription repression in the basolateral amygdala may underlie heightened anxiety in a mouse model of Rett syndrome. J Neurosci 29, 4218-4227.

Aicardi, J. (1997). [Risk factors in recurrent seizures and in mental retardation]. Rev Neurol 25, 754-756.

Alvarez-Saavedra, M., Saez, M.A., Kang, D., Zoghbi, H.Y., and Young, J.I. (2007). Cell-specific expression of wild-type MeCP2 in mouse models of Rett syndrome yields insight about pathogenesis. Hum Mol Genet 16, 2315-2325.

Amir, R.E., Van den Veyver, I.B., Wan, M., Tran, C.Q., Francke, U., and Zoghbi, H.Y. (1999). Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet 23, 185-188.

Armstrong, D.D. (1997). Review of Rett syndrome. J Neuropathol Exp Neurol 56, 843-849.

Armstrong, D.D. (2005). Neuropathology of Rett syndrome. J Child Neurol 20, 747-753.

Armstrong, D.D., Dunn, J.K., Schultz, R.J., Herbert, D.A., Glaze, D.G., and Motil, K.J. (1999). Organ growth in Rett syndrome: a postmortem examination analysis. Pediatr Neurol 20, 125-129.

Asthana, J.C., Sinha, S., Haslam, J.S., and Kingston, H.M. (1990). Survey of adolescents with severe intellectual handicap. Arch Dis Child 65, 1133-1136.

Bahi-Buisson, N., Guellec, I., Nabbout, R., Guet, A., Nguyen, G., Dulac, O., and Chiron, C. (2008). Parental view of epilepsy in Rett Syndrome. Brain Dev 30, 126-130.

Ballas, N., Lioy, D.T., Grunseich, C., and Mandel, G. (2009). Non-cell autonomous influence of MeCP2-deficient glia on neuronal dendritic morphology. Nat Neurosci 12, 311-317.

Bauman, M.L., Kemper, T.L., and Arin, D.M. (1995). Microscopic observations of the brain in Rett syndrome. Neuropediatrics 26, 105-108.

Bebbington, A., Percy, A., Christodoulou, J., Ravine, D., Ho, G., Jacoby, P., Anderson, A., Pineda, M., Ben Zeev, B., Bahi-Buisson, N., et al. (2010). Updating the profile of C-terminal MECP2 deletions in Rett syndrome. J Med Genet 47, 242-248.

118

Berger-Sweeney, J., Libbey, M., Arters, J., Junagadhwalla, M., and Hohmann, C.F. (1998). Neonatal monoaminergic depletion in mice (Mus musculus) improves performance of a novel odor discrimination task. Behav Neurosci 112, 1318-1326.

Boison, D. (2012). Is Intrinsic Hyperexcitability in CA3 the Culprit for Seizures in Rett Syndrome? Epilepsy Curr 12, 13-14.

Braunschweig, D., Simcox, T., Samaco, R.C., and LaSalle, J.M. (2004). X-Chromosome inactivation ratios affect wild-type MeCP2 expression within mosaic Rett syndrome and Mecp2- /+ mouse brain. Hum Mol Genet 13, 1275-1286.

Brucke, T., Sofic, E., Killian, W., Rett, A., and Riederer, P. (1987). Reduced concentrations and increased metabolism of biogenic amines in a single case of Rett-syndrome: a postmortem brain study. J Neural Transm 68, 315-324.

Chahrour, M., Jung, S.Y., Shaw, C., Zhou, X., Wong, S.T., Qin, J., and Zoghbi, H.Y. (2008). MeCP2, a key contributor to neurological disease, activates and represses transcription. Science 320, 1224-1229.

Chahrour, M., and Zoghbi, H.Y. (2007). The story of Rett syndrome: from clinic to neurobiology. Neuron 56, 422-437.

Chang, Q., Khare, G., Dani, V., Nelson, S., and Jaenisch, R. (2006). The disease progression of Mecp2 mutant mice is affected by the level of BDNF expression. Neuron 49, 341-348.

Chao, H.T., Chen, H., Samaco, R.C., Xue, M., Chahrour, M., Yoo, J., Neul, J.L., Gong, S., Lu, H.C., Heintz, N., et al. (2010). Dysfunction in GABA signalling mediates autism-like stereotypies and Rett syndrome phenotypes. Nature 468, 263-269.

Chen, R.Z., Akbarian, S., Tudor, M., and Jaenisch, R. (2001). Deficiency of methyl-CpG binding protein-2 in CNS neurons results in a Rett-like phenotype in mice. Nat Genet 27, 327-331.

Chen, W.G., Chang, Q., Lin, Y., Meissner, A., West, A.E., Griffith, E.C., Jaenisch, R., and Greenberg, M.E. (2003). Derepression of BDNF transcription involves calcium-dependent phosphorylation of MeCP2. Science 302, 885-889.

Cheval, H., Guy, J., Merusi, C., De Sousa, D., Selfridge, J., and Bird, A. (2012). Postnatal inactivation reveals enhanced requirement for MeCP2 at distinct age windows. Hum Mol Genet 21, 3806-3814.

Chiron, C., Bulteau, C., Loc'h, C., Raynaud, C., Garreau, B., Syrota, A., and Maziere, B. (1993). Dopaminergic D2 receptor SPECT imaging in Rett syndrome: increase of specific binding in striatum. J Nucl Med 34, 1717-1721.

119

Cobb, S., Guy, J., and Bird, A. (2010). Reversibility of functional deficits in experimental models of Rett syndrome. Biochem Soc Trans 38, 498-506.

Collins, A.L., Levenson, J.M., Vilaythong, A.P., Richman, R., Armstrong, D.L., Noebels, J.L., David Sweatt, J., and Zoghbi, H.Y. (2004). Mild overexpression of MeCP2 causes a progressive neurological disorder in mice. Hum Mol Genet 13, 2679-2689.

Cooper, R.A., Kerr, A.M., and Amos, P.M. (1998). Rett syndrome: critical examination of clinical features, serial EEG and video-monitoring in understanding and management. Eur J Paediatr Neurol 2, 127-135.

D'Cruz, J.A., Wu, C., Zahid, T., El-Hayek, Y., Zhang, L., and Eubanks, J.H. (2010). Alterations of cortical and hippocampal EEG activity in MeCP2-deficient mice. Neurobiol Dis 38, 8-16.

D'Esposito, M., Quaderi, N.A., Ciccodicola, A., Bruni, P., Esposito, T., D'Urso, M., and Brown, S.D. (1996). Isolation, physical mapping, and northern analysis of the X-linked human gene encoding methyl CpG-binding protein, MECP2. Mamm Genome 7, 533-535.

Dani, V.S., Chang, Q., Maffei, A., Turrigiano, G.G., Jaenisch, R., and Nelson, S.B. (2005). Reduced cortical activity due to a shift in the balance between excitation and inhibition in a mouse model of Rett syndrome. Proc Natl Acad Sci U S A 102, 12560-12565.

Ellaway, C.J., Sholler, G., Leonard, H., and Christodoulou, J. (1999). Prolonged QT interval in Rett syndrome. Arch Dis Child 80, 470-472.

Fahn, S. (2008). The history of dopamine and levodopa in the treatment of Parkinson's disease. Mov Disord 23 Suppl 3, S497-508.

Fyffe, S.L., Neul, J.L., Samaco, R.C., Chao, H.T., Ben-Shachar, S., Moretti, P., McGill, B.E., Goulding, E.H., Sullivan, E., Tecott, L.H., et al. (2008). Deletion of Mecp2 in Sim1-expressing neurons reveals a critical role for MeCP2 in feeding behavior, aggression, and the response to stress. Neuron 59, 947-958.

Gadalla, K.K., Bailey, M.E., and Cobb, S.R. (2011). MeCP2 and Rett syndrome: reversibility and potential avenues for therapy. Biochem J 439, 1-14.

Gantz, S.C., Ford, C.P., Neve, K.A., and Williams, J.T. (2011). Loss of Mecp2 in substantia nigra dopamine neurons compromises the nigrostriatal pathway. J Neurosci 31, 12629-12637.

Garofalo, E.A., Drury, I., and Goldstein, G.W. (1988). EEG abnormalities aid diagnosis of Rett syndrome. Pediatr Neurol 4, 350-353.

Gatto, C.L., and Broadie, K. (2010). Genetic controls balancing excitatory and inhibitory synaptogenesis in neurodevelopmental disorder models. Front Synaptic Neurosci 2, 4.

120

Gemelli, T., Berton, O., Nelson, E.D., Perrotti, L.I., Jaenisch, R., and Monteggia, L.M. (2006). Postnatal loss of methyl-CpG binding protein 2 in the forebrain is sufficient to mediate behavioral aspects of Rett syndrome in mice. Biol Psychiatry 59, 468-476.

Girard, M., Couvert, P., Carrie, A., Tardieu, M., Chelly, J., Beldjord, C., and Bienvenu, T. (2001). Parental origin of de novo MECP2 mutations in Rett syndrome. Eur J Hum Genet 9, 231- 236.

Glaze, D.G. (2005). Neurophysiology of Rett syndrome. J Child Neurol 20, 740-746.

Glaze, D.G., Percy, A.K., Skinner, S., Motil, K.J., Neul, J.L., Barrish, J.O., Lane, J.B., Geerts, S.P., Annese, F., Graham, J., et al. (2010). Epilepsy and the natural history of Rett syndrome. Neurology 74, 909-912.

Glaze, D.G., Schultz, R.J., and Frost, J.D. (1998). Rett syndrome: characterization of seizures versus non-seizures. Electroencephalogr Clin Neurophysiol 106, 79-83.

Goffin, D., Allen, M., Zhang, L., Amorim, M., Wang, I.T., Reyes, A.R., Mercado-Berton, A., Ong, C., Cohen, S., Hu, L., et al. (2012). Rett syndrome mutation MeCP2 T158A disrupts DNA binding, protein stability and ERP responses. Nat Neurosci 15, 274-283.

Gonzales, M.L., and LaSalle, J.M. (2010). The role of MeCP2 in brain development and neurodevelopmental disorders. Curr Psychiatry Rep 12, 127-134.

Gordon, J.A., and Hen, R. (2004). The serotonergic system and anxiety. Neuromolecular Med 5, 27-40.

Guerrini, R., Bonanni, P., Parmeggiani, L., Santucci, M., Parmeggiani, A., and Sartucci, F. (1998). Cortical reflex myoclonus in Rett syndrome. Ann Neurol 43, 472-479.

Guy, J., Cheval, H., Selfridge, J., and Bird, A. (2011). The role of MeCP2 in the brain. Annu Rev Cell Dev Biol 27, 631-652.

Guy, J., Gan, J., Selfridge, J., Cobb, S., and Bird, A. (2007). Reversal of neurological defects in a mouse model of Rett syndrome. Science 315, 1143-1147.

Guy, J., Hendrich, B., Holmes, M., Martin, J.E., and Bird, A. (2001). A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome. Nat Genet 27, 322-326.

Hagberg, B., Goutieres, F., Hanefeld, F., Rett, A., and Wilson, J. (1985). Rett syndrome: criteria for inclusion and exclusion. Brain Dev 7, 372-373.

Hagberg, B., Hanefeld, F., Percy, A., and Skjeldal, O. (2002). An update on clinically applicable diagnostic criteria in Rett syndrome. Comments to Rett Syndrome Clinical Criteria Consensus Panel Satellite to European Paediatric Neurology Society Meeting, Baden Baden, Germany, 11 September 2001. Eur J Paediatr Neurol 6, 293-297.

121

Hagberg, B., and Witt-Engerstrom, I. (1986). Rett syndrome: a suggested staging system for describing impairment profile with increasing age towards adolescence. Am J Med Genet Suppl 1, 47-59.

Hoffbuhr, K.C., Moses, L.M., Jerdonek, M.A., Naidu, S., and Hoffman, E.P. (2002). Associations between MeCP2 mutations, X-chromosome inactivation, and phenotype. Ment Retard Dev Disabil Res Rev 8, 99-105.

Huppke, P., Laccone, F., Kramer, N., Engel, W., and Hanefeld, F. (2000). Rett syndrome: analysis of MECP2 and clinical characterization of 31 patients. Hum Mol Genet 9, 1369-1375.

Itoh, M., Ide, S., Takashima, S., Kudo, S., Nomura, Y., Segawa, M., Kubota, T., Mori, H., Tanaka, S., Horie, H., et al. (2007). Methyl CpG-binding protein 2 (a mutation of which causes Rett syndrome) directly regulates insulin-like growth factor binding protein 3 in mouse and human brains. J Neuropathol Exp Neurol 66, 117-123.

Jellinger, K., Armstrong, D., Zoghbi, H.Y., and Percy, A.K. (1988). Neuropathology of Rett syndrome. Acta Neuropathol 76, 142-158.

Jellinger, K., Grisold, W., Armstrong, D., and Rett, A. (1990). Peripheral nerve involvement in the Rett syndrome. Brain Dev 12, 109-114.

Jentarra, G.M., Olfers, S.L., Rice, S.G., Srivastava, N., Homanics, G.E., Blue, M., Naidu, S., and Narayanan, V. (2010). Abnormalities of cell packing density and dendritic complexity in the MeCP2 A140V mouse model of Rett syndrome/X-linked mental retardation. BMC Neurosci 11, 19.

Johnston, M.V., Blue, M.E., and Naidu, S. (2005). Rett syndrome and neuronal development. J Child Neurol 20, 759-763.

Johnston, M.V., Hohmann, C., and Blue, M.E. (1995). Neurobiology of Rett syndrome. Neuropediatrics 26, 119-122.

Johnston, M.V., Jeon, O.H., Pevsner, J., Blue, M.E., and Naidu, S. (2001). Neurobiology of Rett syndrome: a genetic disorder of synapse development. Brain Dev 23 Suppl 1, S206-213.

Jugloff, D.G., Vandamme, K., Logan, R., Visanji, N.P., Brotchie, J.M., and Eubanks, J.H. (2008). Targeted delivery of an Mecp2 transgene to forebrain neurons improves the behavior of female Mecp2-deficient mice. Hum Mol Genet 17, 1386-1396.

Julu, P.O., Kerr, A.M., Apartopoulos, F., Al-Rawas, S., Engerstrom, I.W., Engerstrom, L., Jamal, G.A., and Hansen, S. (2001). Characterisation of breathing and associated central autonomic dysfunction in the Rett disorder. Arch Dis Child 85, 29-37.

122

Julu, P.O., Kerr, A.M., Hansen, S., Apartopoulos, F., and Jamal, G.A. (1997). Functional evidence of brain stem immaturity in Rett syndrome. Eur Child Adolesc Psychiatry 6 Suppl 1, 47-54.

Kishi, N., and Macklis, J.D. (2004). MECP2 is progressively expressed in post-migratory neurons and is involved in neuronal maturation rather than cell fate decisions. Mol Cell Neurosci 27, 306-321.

Lawson-Yuen, A., Liu, D., Han, L., Jiang, Z.I., Tsai, G.E., Basu, A.C., Picker, J., Feng, J., and Coyle, J.T. (2007). Ube3a mRNA and protein expression are not decreased in Mecp2R168X mutant mice. Brain Res 1180, 1-6.

Lekman, A., Witt-Engerstrom, I., Gottfries, J., Hagberg, B.A., Percy, A.K., and Svennerholm, L. (1989). Rett syndrome: biogenic amines and metabolites in postmortem brain. Pediatr Neurol 5, 357-362.

Lekman, A., Witt-Engerstrom, I., Holmberg, B., Percy, A., Svennerholm, L., and Hagberg, B. (1990). CSF and urine biogenic amine metabolites in Rett syndrome. Clin Genet 37, 173-178.

Lewis, J.D., Meehan, R.R., Henzel, W.J., Maurer-Fogy, I., Jeppesen, P., Klein, F., and Bird, A. (1992). Purification, sequence, and cellular localization of a novel chromosomal protein that binds to methylated DNA. Cell 69, 905-914.

Liao, W., Gandal, M.J., Ehrlichman, R.S., Siegel, S.J., and Carlson, G.C. (2012). MeCP2+/- mouse model of RTT reproduces auditory phenotypes associated with Rett syndrome and replicate select EEG endophenotypes of autism spectrum disorder. Neurobiol Dis 46, 88-92.

Lioy, D.T., Garg, S.K., Monaghan, C.E., Raber, J., Foust, K.D., Kaspar, B.K., Hirrlinger, P.G., Kirchhoff, F., Bissonnette, J.M., Ballas, N., et al. (2011). A role for glia in the progression of Rett's syndrome. Nature 475, 497-500.

Lucki, I. (1998). The spectrum of behaviors influenced by serotonin. Biol Psychiatry 44, 151-162. Maezawa, I., Swanberg, S., Harvey, D., LaSalle, J.M., and Jin, L.W. (2009). Rett syndrome astrocytes are abnormal and spread MeCP2 deficiency through gap junctions. J Neurosci 29, 5051-5061.

Martinowich, K., Hattori, D., Wu, H., Fouse, S., He, F., Hu, Y., Fan, G., and Sun, Y.E. (2003). DNA methylation-related chromatin remodeling in activity-dependent BDNF gene regulation. Science 302, 890-893.

Matijevic, T., Knezevic, J., Slavica, M., and Pavelic, J. (2009). Rett syndrome: from the gene to the disease. Eur Neurol 61, 3-10.

Misgeld, T., Burgess, R.W., Lewis, R.M., Cunningham, J.M., Lichtman, J.W., and Sanes, J.R. (2002). Roles of neurotransmitter in synapse formation: development of neuromuscular junctions lacking choline acetyltransferase. Neuron 36, 635-648.

123

Monteggia, L.M., and Kavalali, E.T. (2009). Rett syndrome and the impact of MeCP2 associated transcriptional mechanisms on neurotransmission. Biol Psychiatry 65, 204-210.

Moretti, P., Levenson, J.M., Battaglia, F., Atkinson, R., Teague, R., Antalffy, B., Armstrong, D., Arancio, O., Sweatt, J.D., and Zoghbi, H.Y. (2006). Learning and memory and synaptic plasticity are impaired in a mouse model of Rett syndrome. J Neurosci 26, 319-327.

Moretti, P., and Zoghbi, H.Y. (2006). MeCP2 dysfunction in Rett syndrome and related disorders. Curr Opin Genet Dev 16, 276-281.

Na, E.S., Nelson, E.D., Adachi, M., Autry, A.E., Mahgoub, M.A., Kavalali, E.T., and Monteggia, L.M. (2012). A mouse model for MeCP2 duplication syndrome: MeCP2 overexpression impairs learning and memory and synaptic transmission. J Neurosci 32, 3109-3117.

Nag, N., and Berger-Sweeney, J.E. (2007). Postnatal dietary choline supplementation alters behavior in a mouse model of Rett syndrome. Neurobiol Dis 26, 473-480.

Naidu, S. (1997). Rett syndrome: a disorder affecting early brain growth. Ann Neurol 42, 3-10.

Naidu, S., Kaufmann, W.E., Abrams, M.T., Pearlson, G.D., Lanham, D.C., Fredericksen, K.A., Barker, P.B., Horska, A., Golay, X., Mori, S., et al. (2001). Neuroimaging studies in Rett syndrome. Brain Dev 23 Suppl 1, S62-71.

Nan, X., Campoy, F.J., and Bird, A. (1997). MeCP2 is a transcriptional repressor with abundant binding sites in genomic chromatin. Cell 88, 471-481.

Neul, J.L., Fang, P., Barrish, J., Lane, J., Caeg, E.B., Smith, E.O., Zoghbi, H., Percy, A., and Glaze, D.G. (2008). Specific mutations in methyl-CpG-binding protein 2 confer different severity in Rett syndrome. Neurology 70, 1313-1321.

Niedermeyer, E., and Naidu, S.B. (1998). Rett syndrome, EEG and the motor cortex as a model for better understanding of attention deficit hyperactivity disorder (ADHD). Eur Child Adolesc Psychiatry 7, 69-72.

Niedermeyer, E., Naidu, S.B., and Plate, C. (1997). Unusual EEG theta rhythms over central region in Rett syndrome: considerations of the underlying dysfunction. Clin Electroencephalogr 28, 36-43.

Nomura, Y., Segawa, M., and Higurashi, M. (1985). Rett syndrome--an early catecholamine and indolamine deficient disorder? Brain Dev 7, 334-341.

Ogier, M., Wang, H., Hong, E., Wang, Q., Greenberg, M.E., and Katz, D.M. (2007). Brain- derived neurotrophic factor expression and respiratory function improve after ampakine treatment in a mouse model of Rett syndrome. J Neurosci 27, 10912-10917.

124

Ohya, T., Yamashita, Y., and Matsuishi, T. (2005). [Sudden death in Rett syndrome]. Nihon Rinsho 63, 1178-1182.

Panayotis, N., Pratte, M., Borges-Correia, A., Ghata, A., Villard, L., and Roux, J.C. (2011). Morphological and functional alterations in the substantia nigra pars compacta of the Mecp2-null mouse. Neurobiol Dis 41, 385-397.

Pelka, G.J., Watson, C.M., Radziewic, T., Hayward, M., Lahooti, H., Christodoulou, J., and Tam, P.P. (2006). Mecp2 deficiency is associated with learning and cognitive deficits and altered gene activity in the hippocampal region of mice. Brain 129, 887-898.

Percy, A.K. (1992). Neurochemistry of the Rett syndrome. Brain Dev 14 Suppl, S57-62.

Perry, T.L., Dunn, H.G., Ho, H.H., and Crichton, J.U. (1988). Cerebrospinal fluid values for monoamine metabolites, gamma-aminobutyric acid, and other amino compounds in Rett syndrome. J Pediatr 112, 234-238.

Peters, L.L., Robledo, R.F., Bult, C.J., Churchill, G.A., Paigen, B.J., and Svenson, K.L. (2007). The mouse as a model for human biology: a resource guide for complex trait analysis. Nat Rev Genet 8, 58-69.

Popova, N.K. (2008). From gene to aggressive behavior: the role of brain serotonin. Neurosci Behav Physiol 38, 471-475.

Quaderi, N.A., Meehan, R.R., Tate, P.H., Cross, S.H., Bird, A.P., Chatterjee, A., Herman, G.E., and Brown, S.D. (1994). Genetic and physical mapping of a gene encoding a methyl CpG binding protein, Mecp2, to the mouse X chromosome. Genomics 22, 648-651.

Reiss, A.L., Faruque, F., Naidu, S., Abrams, M., Beaty, T., Bryan, R.N., and Moser, H. (1993). Neuroanatomy of Rett syndrome: a volumetric imaging study. Ann Neurol 34, 227-234.

Rett, A. (1966). [On a unusual brain atrophy syndrome in hyperammonemia in childhood]. Wien Med Wochenschr 116, 723-726.

Ricceri, L., De Filippis, B., and Laviola, G. (2008). Mouse models of Rett syndrome: from behavioural phenotyping to preclinical evaluation of new therapeutic approaches. Behav Pharmacol 19, 501-517.

Robinson, L., Guy, J., McKay, L., Brockett, E., Spike, R.C., Selfridge, J., De Sousa, D., Merusi, C., Riedel, G., Bird, A., et al. (2012). Morphological and functional reversal of phenotypes in a mouse model of Rett syndrome. Brain.

Rohdin, M., Fernell, E., Eriksson, M., Albage, M., Lagercrantz, H., and Katz-Salamon, M. (2007). Disturbances in cardiorespiratory function during day and night in Rett syndrome. Pediatr Neurol 37, 338-344.

125

Roux, J.C., Dura, E., Moncla, A., Mancini, J., and Villard, L. (2007). Treatment with desipramine improves breathing and survival in a mouse model for Rett syndrome. Eur J Neurosci 25, 1915-1922.

Roux, J.C., and Villard, L. (2010). Biogenic amines in Rett syndrome: the usual suspects. Behav Genet 40, 59-75.

Samaco, R.C., Fryer, J.D., Ren, J., Fyffe, S., Chao, H.T., Sun, Y., Greer, J.J., Zoghbi, H.Y., and Neul, J.L. (2008). A partial loss of function allele of methyl-CpG-binding protein 2 predicts a human neurodevelopmental syndrome. Hum Mol Genet 17, 1718-1727.

Samaco, R.C., Mandel-Brehm, C., Chao, H.T., Ward, C.S., Fyffe-Maricich, S.L., Ren, J., Hyland, K., Thaller, C., Maricich, S.M., Humphreys, P., et al. (2009). Loss of MeCP2 in aminergic neurons causes cell-autonomous defects in neurotransmitter synthesis and specific behavioral abnormalities. Proc Natl Acad Sci U S A 106, 21966-21971.

Samaco, R.C., and Neul, J.L. (2011). Complexities of Rett syndrome and MeCP2. J Neurosci 31, 7951-7959.

Sansom, D., Krishnan, V.H., Corbett, J., and Kerr, A. (1993). Emotional and behavioural aspects of Rett syndrome. Dev Med Child Neurol 35, 340-345.

Segawa, M. (1997). Pathophysiology of Rett syndrome from the standpoint of early catecholamine disturbance. Eur Child Adolesc Psychiatry 6 Suppl 1, 56-60.

Segawa, M. (2001). Pathophysiology of Rett syndrome from the stand point of clinical characteristics. Brain Dev 23 Suppl 1, S94-98.

Sekul, E.A., Moak, J.P., Schultz, R.J., Glaze, D.G., Dunn, J.K., and Percy, A.K. (1994). Electrocardiographic findings in Rett syndrome: an explanation for sudden death? J Pediatr 125, 80-82.

Shahbazian, M., Young, J., Yuva-Paylor, L., Spencer, C., Antalffy, B., Noebels, J., Armstrong, D., Paylor, R., and Zoghbi, H. (2002). Mice with truncated MeCP2 recapitulate many Rett syndrome features and display hyperacetylation of histone H3. Neuron 35, 243-254.

Smeets, E., Schollen, E., Moog, U., Matthijs, G., Herbergs, J., Smeets, H., Curfs, L., Schrander- Stumpel, C., and Fryns, J.P. (2003). Rett syndrome in adolescent and adult females: clinical and molecular genetic findings. Am J Med Genet A 122A, 227-233.

Smeets, E., Terhal, P., Casaer, P., Peters, A., Midro, A., Schollen, E., van Roozendaal, K., Moog, U., Matthijs, G., Herbergs, J., et al. (2005). Rett syndrome in females with CTS hot spot deletions: a disorder profile. Am J Med Genet A 132A, 117-120.

Smeets, E.E., Pelc, K., and Dan, B. (2012). Rett Syndrome. Mol Syndromol 2, 113-127.

126

Steffenburg, U., Hagberg, G., and Hagberg, B. (2001). Epilepsy in a representative series of Rett syndrome. Acta Paediatr 90, 34-39.

Subramaniam, B., Naidu, S., and Reiss, A.L. (1997). Neuroanatomy in Rett syndrome: cerebral cortex and posterior fossa. Neurology 48, 399-407.

Szczypka, M.S., Kwok, K., Brot, M.D., Marck, B.T., Matsumoto, A.M., Donahue, B.A., and Palmiter, R.D. (2001). Dopamine production in the caudate putamen restores feeding in dopamine-deficient mice. Neuron 30, 819-828.

Takahashi, S., Ohinata, J., Makita, Y., Suzuki, N., Araki, A., Sasaki, A., Murono, K., Tanaka, H., and Fujieda, K. (2008). Skewed X chromosome inactivation failed to explain the normal phenotype of a carrier female with MECP2 mutation resulting in Rett syndrome. Clin Genet 73, 257-261.

Taneja, P., Ogier, M., Brooks-Harris, G., Schmid, D.A., Katz, D.M., and Nelson, S.B. (2009). Pathophysiology of locus ceruleus neurons in a mouse model of Rett syndrome. J Neurosci 29, 12187-12195.

Thomas, S.A., and Palmiter, R.D. (1997). Impaired maternal behavior in mice lacking norepinephrine and epinephrine. Cell 91, 583-592.

Tropea, D., Giacometti, E., Wilson, N.R., Beard, C., McCurry, C., Fu, D.D., Flannery, R., Jaenisch, R., and Sur, M. (2009). Partial reversal of Rett Syndrome-like symptoms in MeCP2 mutant mice. Proc Natl Acad Sci U S A 106, 2029-2034.

Truett, G.E., Heeger, P., Mynatt, R.L., Truett, A.A., Walker, J.A., and Warman, M.L. (2000). Preparation of PCR-quality mouse genomic DNA with hot sodium hydroxide and tris (HotSHOT). Biotechniques 29, 52, 54.

Viemari, J.C., Roux, J.C., Tryba, A.K., Saywell, V., Burnet, H., Pena, F., Zanella, S., Bevengut, M., Barthelemy-Requin, M., Herzing, L.B., et al. (2005). Mecp2 deficiency disrupts norepinephrine and respiratory systems in mice. J Neurosci 25, 11521-11530.

Ward, C.S., Arvide, E.M., Huang, T.W., Yoo, J., Noebels, J.L., and Neul, J.L. (2011). MeCP2 is critical within HoxB1-derived tissues of mice for normal lifespan. J Neurosci 31, 10359-10370.

Weaving, L.S., Ellaway, C.J., Gecz, J., and Christodoulou, J. (2005). Rett syndrome: clinical review and genetic update. J Med Genet 42, 1-7. Weese-Mayer, D.E., Lieske, S.P., Boothby, C.M., Kenny, A.S., Bennett, H.L., Silvestri, J.M., and Ramirez, J.M. (2006). Autonomic nervous system dysregulation: breathing and heart rate perturbation during wakefulness in young girls with Rett syndrome. Pediatr Res 60, 443-449.

Weng, S.M., Bailey, M.E., and Cobb, S.R. (2011). Rett syndrome: from bed to bench. Pediatr Neonatol 52, 309-316.

127

Wenk, G.L. (1996). Rett syndrome: evidence for normal dopaminergic function. Neuropediatrics 27, 256-259.

Wenk, G.L., and Hauss-Wegrzyniak, B. (1999). Altered cholinergic function in the basal forebrain of girls with Rett syndrome. Neuropediatrics 30, 125-129.

Wenk, G.L., and Mobley, S.L. (1996). Choline acetyltransferase activity and vesamicol binding in Rett syndrome and in rats with nucleus basalis lesions. Neuroscience 73, 79-84.

Williamson, S.L., and Christodoulou, J. (2006). Rett syndrome: new clinical and molecular insights. Eur J Hum Genet 14, 896-903.

Wither, R.G., Colic, S., Wu, C., Bardakjian, B.L., Zhang, L., and Eubanks, J.H. (2012). Daily rhythmic behaviors and thermoregulatory patterns are disrupted in adult female MeCP2-deficient mice. PLoS One 7, e35396.

Witt Engerstrom, I. (1992). Age-related occurrence of signs and symptoms in the Rett syndrome. Brain Dev 14 Suppl, S11-20.

Wu, C., Wais, M., Sheppy, E., del Campo, M., and Zhang, L. (2008). A glue-based, screw-free method for implantation of intra-cranial electrodes in young mice. J Neurosci Methods 171, 126- 131.

Yoshikawa, H., Kaga, M., Suzuki, H., Sakuragawa, N., and Arima, M. (1991). Giant somatosensory evoked potentials in the Rett syndrome. Brain Dev 13, 36-39.

Young, J.I., Hong, E.P., Castle, J.C., Crespo-Barreto, J., Bowman, A.B., Rose, M.F., Kang, D., Richman, R., Johnson, J.M., Berget, S., et al. (2005). Regulation of RNA splicing by the methylation-dependent transcriptional repressor methyl-CpG binding protein 2. Proc Natl Acad Sci U S A 102, 17551-17558.

Zanella, S., Mebarek, S., Lajard, A.M., Picard, N., Dutschmann, M., and Hilaire, G. (2008). Oral treatment with desipramine improves breathing and life span in Rett syndrome mouse model. Respir Physiol Neurobiol 160, 116-121.

Zoghbi, H.Y. (2005). MeCP2 dysfunction in humans and mice. J Child Neurol 20, 736-740.

Zoghbi, H.Y., Milstien, S., Butler, I.J., Smith, E.O., Kaufman, S., Glaze, D.G., and Percy, A.K. (1989). Cerebrospinal fluid biogenic amines and biopterin in Rett syndrome. Ann Neurol 25, 56- 60.

128