On the Pulse of Angelman Syndrome Gene Therapy Development Research:
BACKGROUND
Disease Profile & Epidemiology
Originally documented in 1965 by Dutch physician Dr. Harry Angelman, the namesake neurodevelopmental disorder is characterized by severe mental disability, seizures, absent speech, ataxia, and paroxysmal laughter (Kishino, 1997). Perhaps Angelman Syndrome’s most definitive clinical presentation is the characteristic constantly smiling, easily excitable demeanor. Facial expression is uncontrollable for Angelman patients. Detectable delays emerge by 12 months: beyond visible microcephaly, Angelman patients suffer neuromotor deficits, jerky movements, and hand-flapping, outstretched arms (Bernard, 2008). Cognitive impairment in learning, sensitivity to social cues, expression, and communication forms the basis for behavioral abnormality. Notably, Angelman Syndrome is not associated with developmental regression (affected individuals do not experience loss of previously attained milestones); developmental capacity rarely surpasses that of stage 12 months, even as Angelman individuals grow into adulthood (Clayton-Smith, 2003).
Angelman Syndrome does not have a significant impact on mortality, and as long as individuals medically manage their seizures, life expectancy is largely unaffected. The genetic disorder has an incidence of 1 in 15,000 births (Bernard, 2008). Prevalent misdiagnosis—in 49% of cases, as reported by the 2014 Angelman Syndrome Foundation survey—remains a challenge due to phenotypic overlap with Rett Syndrome, cerebral palsy, and autism (pathophysiological mechanisms this paper will explore), and general lack of awareness.
Molecular Genetics & Genomic Imprinting
A single-gene disorder, Angelman Syndrome is caused by functional loss of the ubiquitin ligase E3A (UBE3A) gene, located in the q11-q13 region of chromosome 15, and primarily arising de-novo via non-Mendelian inheritance mechanisms due to sporadic error during meiosis and chromosome segregation. Not to say the disease is un-inheritable, the data only suggests a 7% familial incidence—notably skipping generations along the maternal lineage, reflecting the mechanism of genomic imprinting (Bernard, 2008). Dubbed the Prader-Willi/Angelman Critical Region (PWCR) after the pair of sister disorders, this chromosomal region is subject to genomic imprinting—epigenetic selective expression and silencing based on parental origin. Wild-type chromosome 15 expression inactivates the paternally inherited UBE3A allele while the maternally inherited copy retains expression; the opposite imprinting pattern occurs in the SNRPN gene responsible for Prader-Willi Syndrome (Kishino, 1997).
A deletion in the UBE3A-containing region on maternal chromosome 15 results in total loss-of-function, as the remaining paternal allele is silent. Responsible for 70% of AS incidence, regional deletion presents with the most severe phenotype and is characteristic of classical AS. UBE3A gene mutations in the maternal allele account for 11% of Angelman cases. Roughly 7% of AS incidence results from paternal uniparental disomy, a phenomenon in which afflicted individuals inherit two inactive paternal copies of UBE3A due to missegregation. Imprinting defects, perhaps a mutation in the maternal imprinting center, account for another 3% of cases. Chromosomal rearrangement such as translocation and inversion have a rare incidence in Angelman Syndrome, about 2% (Bernard, 2008). Interestingly, a maternal duplication, resulting in UBE3A overexpression, is consistent with autism and certain effects opposite of AS, such as high dendritic spine density (Samaco et al., 2005).
Encoding the widespread, extensive protein ubiquitin ligase E3A or E6-AP, UBE3A experiences global biallelic expression; genomic imprinting of UBE3A is tissue-specific to the central and peripheral nervous systems. Furthermore, the paternal allele inactivation mechanism is exclusive to neurons, as expression is maintained in glia (Ehlers, 2009). As it turns out, recent research in the UBE3A maternal knockout mouse model suggests incomplete paternal allele inactivation in neurons during initial neurodevelopment. Imprinting is activated by the transcription of antisense RNA UBE3A-ATS later than previously estimated, as the recent determination of the spatiotemporal intricacies behind UBE3A expression discerned distinct developmental windows (Silvia-Santos et al., 2015). Findings in the maternal knockout mouse model are consistent with the hypothesis that, in AS-afflicted individuals, just enough paternal UBE3A is initially expressed to maintain functions of neurogenesis, until antisense transcription kicks in for paternal inactivation. Blocking UBE3A-ATS transcript entirely is the key target for therapeutic rescue of the silenced paternal allele in hopes of UBE3A functional restoration in Angelman Syndrome.
UBE3A Protein Functional Overview
During neurodevelopment, ubiquitin ligase E3A, also known as E6-AP, tags substrates for degradation by proteasomes, or ubiquitination. Disruption of the crucial function leads to the accumulation of aggregates, interfering with neurogenesis, membrane receptor turnover, and signal transduction. (Bingol and Sheng, 2011). As E6-AP is a cornerstone to the 26S proteolytic pathway, the only active proteolysis during neurodevelopment, failure to degrade effectors leads to postsynaptic kinase pathway dysfunction in the transgenic mouse model (Ehlers, 2009). Experience-dependent maturation and neural plasticity depends on ubiquitination; its action is required for synaptic maintenance, long-term potentiation, and plasticity.
UBE3A disruption not only directly affects target substrates, it also triggers overall misregulation. Studies in the Drosophila model with ortholog dube3a knockdown, mimicking AS proteolytic pathway dysfunction, has aided the discovery of diverse UBE3A substrates and investigation into UBE3A’s regulatory role. (Lee et al., 2014). Recently, 43 new targets were identified in Drosophila, in addition to key substrates: Arc synaptic protein (Ferdousy et al., 2011), plic-1 regulator of GABAA receptor surface stability (Luscher, 2001), and cyclin-dependent kinase inhibitor p27 in cell proliferation (Mishra et al., 2009). Current research investigates the diverse interactions and pathways in UBE3A’s function as a transcriptional co-activator.
NEUROPATHOLOGY
At first glance, AS neuroanatomy reveals no morphological abnormalities discernable by MRI scan or cortical section, beyond microcephaly. Employing the latest techniques in data analysis of digital tractography imaging (DTI) via groupwise, observer-independent tract-based spatial statistics (TBSS), Tiwari et al. visualized the global impairment of white matter integrity in neuronal components along all axis. The 2012 discoveries briefed below effectively identified aberrant brain connectivity with the clinical presentations of Angelman Syndrome.
Consistent abnormalities in the arcuate fasciculus, integral in development of language pathways, corresponded with absent speech and lacking linguistic capacity in Angelman patients. Additional findings included aberrant networks of thalamocortical neurons; evident of error in axon guidance due to UBE3A and (consequently) proteolytic loss-of-function, these cortical clusters exhibited abnormal thalamic radiation. Loss of the inhibitory mechanism of the intra-thalamic reticular nucleus accounted for the abnormally synchronized thalamocortical oscillations characteristic of AS—linked to the epilepsy, tremors, uncontrolled paroxysms of laughter—and detectable by EEG readings, a critical component for Angelman diagnosis. Loss of frontal lobe expressive motor control and inhibition corresponded with the fixed, smiling facial expression in afflicted patients. Clinical and behavioral characterizations of AS patients as “lacking empathy [on account of] difficulty understanding [their] social environment, interacting with peers, and uncontrolled emotion” could be attributed to detected abnormalities in the anterior cingulum, known as the empathy center for its regulatory role in cognitive and emotional processes (Smit, 2009; Tiwari et al., 2012).
Despite displaying dendritic branching consistent with that of wild-type hippocampal and cortical pyramidal neurons and cerebellar Purkinje cells, AS-afflicted dendrites exhibit a substantial reduction in spines (Dindot et al., 2008). The lack of dendritic spines and abnormal morphology in the absence of maternal UBE3A not only presents problems in synaptic strength and long-term potentiation (LTP), but points to the gene’s function regulating local synaptic plasticity (Dindot et al., 2008). Since the 2008 Dindot et al. findings, research investigation into UBE3A’s downstream substrates have identified synaptic protein Arc as a crucial target for synaptic maintenance. Arc “promotes the internalization of AMPA glutamate receptors,” and Ube3A loss-of-function in neurons allows Arc accumulation and increased expression, triggering a subsequent decrease in the number of AMPA receptors at excitatory synapses (Greer, 2010). As a consequence of UBE3A disruption, AMPA receptor dysregulation at the synapses may contribute to the cognitive dysfunction of Angelman Syndrome.
Overlap & Insight into UBE3A Interactions
The intersection of AS’s manifestations with those of other neurological disorders makes for difficult clinical diagnosis. From a molecular perspective, however, the mechanisms of phenotypic overlap offer revealing insight to reinforce research findings in UBE3A regulatory pathways as a transcriptional co-activator.
Examine, for example, the behavioral neurogenetics of adenylosuccinate lyase (ADSL) deficiency, a mutation-driven disorder of disrupted purine metabolism cycle necessary for de novo AMP synthesis. The level of neurotoxic impairment, due to the accumulation of SAICAr (phosphoribosyl succinyl-aminoimidazole carboxamide), fascinatingly presents with the behavioral profile of the Angelman phenotype, minus the characteristic EEG oscillations (Gitiaux et al., 2012). Global developmental delay, motor disability, seizures, severe speech deficit, as well as uncontrolled smiling and laughter coincided with a single mutation in the ADSL gene of chromosome 22, without a detectable defect in either UBE3A sequence nor methylation pattern. These cases of misleading behavioral phenotype documented in the European Journal of Human Genetics bring us closer to grasping the downstream neurological and regulatory consequences of disrupted degradation pathways.
Maintenance of GABAergic synapses involves UBE3A’s interaction with the neighboring GABRB3 gene of the q12 region on chromosome 15. Defects in the GABRB3 gene encoding the GABAA receptor beta subunit, have been attributed to autism, and the gene’s location explains the exacerbated severity of symptoms and epilepsy in AS cases suffering regional deletions. Notably, GABRB3-induced changes in dendritic spine dynamics and synaptic stability are UBE3A dosage dependent: UBE3A deficiency in AS has the opposite effect of autism’s increased spine density and short-range connectivity. Ubiquitination degradation of interacting proteins and substrates in GABAergic synapses is the conceivable mechanism for this regulation (Samaco et al., 2005).
Moreover, the methyl-CpG-binding protein gene MECP2, associated with Rett Syndrome, has been implicated in this regulatory dynamic between UBE3A and GABRB3. In cases of MECP2 epigenetic dysregulation, associated with abnormal X-inactivation, UBE3A deficiency is exacerbated by skewed expression—which also affects a significant reduction in GABRB3, despite its typically global bi-allelic expression. MECP2 deficiency has a significant impact on UBE3A and GABRB3 expressions in the model mouse brain, without apparent alterations in allele-specific expression (Samaco et al., 2005). Angelman Syndrome, autism, and Rett syndrome share overlapping phenotypes of developmental delay, language impairment, seizures, and stereotypic behaviors—unsurprisingly, given their epigenetic regulatory overlap.
Towards a Treatment for Angelman Syndrome
Existing treatment approaches for Angelman Syndrome are aimed at managing symptoms, between rehabilitation and pharmacological agents. Extensive physical therapy from early childhood is necessary to address motor deficits and delayed development of walking. Affected individuals suffer from poor muscle tone and continually struggle with a stiff, jerky gait. Additionally, Angelman patients can benefit from intensive behavioral and occupational therapy, as well as communication therapy involving picture signals and sign language. Pharmacological targets of GABRB3 receptor activity are an effective form of epilepsy management for AS patients. A combination of valproic acid, benzodiazepines, and ethosuximide anti-seizure medications are commonly prescribed (Bernard, 2008).
The direction of research for AS therapy is geared towards paternal UBE3A rescue, in the face of maternal allele dysfunction and specifically via reduction of negative regulator UBE3A-antisense RNA. Upregulation of typically-silenced paternal UBE3A presents a host of potentially widespread, off-target effects—given the lack of control and specificity in epigenetic mechanisms of altering DNA methylation, histone modification and chromatin reorganization, or targeting the imprinting center on the paternal copy. Furthermore, delivery challenges include the high-risk nature of blood-brain barrier penetration and restrictive permeability exclusive to small, lipid-soluble particles or mediated transport. On account of the complications of genomic imprinting, target specificity must exceed tissue-specific, to act in a neuron-specific manner. Even then, dosage determination remains a challenge, along with limitations for functional recovery past certain neurodevelopmental stages.
One well-tolerated potential treatment involves the topoisomerase inhibitor topotecan, a camptothecin derivative (Huang et al., 2012). Previously employed for cancer therapies, topotecan displayed long-lasting effects up to 12 weeks post-delivery via mini-osmotic pump in the AS mouse model. Paternal UBE3A was upregulated to up to 50% of wild-type levels, as topotecan functionally—and apparently, selectively—blocks chromatin decondensation, and thus transcription of UBE3A-ATS. Most unexpectedly, topotecan application produced no widespread unintended effects on other imprinted genes. The apparent mechanism of R-loop formation at the regional locus has a stabilizing effect, mediating topotecan for allele-specific chromatin decondensation (Powell, 2013). One benefit of topotecan treatment is its long-lasting potential, requiring fewer doses and reducing overall toxicity (Huang et al., 2012).
The recommendation of RNA-directed ATFs or artificial transcription factors, ideal therapy for a single-gene disease such as AS was first raised by Philpot et al. in 2011. As of January 2015, Meng et al. demonstrated sustained unsilencing of paternal allele via antisense oligonucleotides (ASOs), sequence-specific for the reduction of UBE3A ATS transcript. The procedure in the AS mouse model displayed well-tolerated intrathecal delivery, resulting in broad neural tissue distribution and long duration of action, resisting nuclease degradation. The most important advantages of ASOs are high-specificity and no impact on DNA methylation and the imprinting center (Meng et al., 2015).
Even with such innovative developments towards AS therapies, the progress of research nonetheless unearths new limitations. Silvia-Santos et al., published in May 2015 a study on the limits of therapeutic intervention in adult AS mice, Paternal UBE3A unsilencing proved insufficient for recovery of most functions, evidence of a window early in development of critical UBE3A activity neurogenesis and circuitry. With the assistance of ErbB inhibitors and ampakine cognitive enhancers, LTP and plasticity were the only restorable features in mature AS mice; as UBE3A expression returned, so did proper degradation of effectors. Motor deficits were rescued to some degree in juvenile mice, but the “autistic manifestations” of anxiety, stereotypical and repetitive behavior, and epilepsy arose too early in temporal neurodevelopment for recovery by UBE3A reinstatement. These distinct developmental windows for UBE3A rescue contrast with results in MECP2 methylation rescue for Rett Syndrome, remaining effective in the adult mice model. Further investigation into UBE3A targets and downstream signalling pathways seems a warranted alternative, given these temporal limits to epigenetic rescue. Accordingly, the epileptic symptoms of AS have been better managed targeting UBE3A downstream interactions with GABAA (Silvia-Santos et al., 2015).