Huda Y. Zoghbi, MD
Ralph D. Feigin Professor, Baylor College of Medicine
Investigator, Howard Hughes Medical Institute
Director, Jan and Dan Duncan Neurological Research Institute at Texas Children’s Hospital
(November 1, 2016)
Rett Syndrome: From the Clinic to Genomes, Epigenomes, and Neural Circuits
Rett syndrome is a developmental disorder that becomes apparent in girls, early in life. Like autism, children with Rett syndrome appear to develop normally only to regress as they get older. Language and social interaction skills are lost first, followed by movement and other neurological problems. Dr. Zoghbi discussed her work focused on the genetics behind the syndrome. An early discovery was the identification of the MECP2 gene mutation responsible for Rett syndrome (as well as many other neuropsychiatric disorders). Dr. Zoghbi has developed a mouse model of Rett syndrome, which has allowed her to study the brain circuits and signaling chemicals (neurotransmitters) affected by the mutation. This model has helped her to pinpoint areas of possible treatment. Deep brain stimulation (inducing activity in a brain area) of the hippocampus improves learning and memory in Rett mice. Treatments to reduce levels of the MECP2 protein, which are very high in Rett syndrome, reduced many symptoms in mice. These exciting advances may offer future areas for treatment in children who develop this syndrome.
When I was a resident in pediatric neurology at Texas Children’s Hospital in Houston, I met a patient who changed the course of my life. Her name was Ashley, and she had been perfectly healthy until she was about two years old. Then she lost interest in socializing, stopped greeting her father when he came home from work, and seemed to forget the words she had just been happily babbling only weeks before. Instead, she just sat and wrung her hands incessantly. She had a hard time walking, alternated between holding her breath and hyperventilating, and developed features of autism. Her symptoms and disease course matched those in a paper describing a new condition called Rett Syndrome, and it was clear to me that I was seeing this syndrome in real life. A week later, I saw another girl who was diagnosed with cerebral palsy, and I became seriously interested in research, because it was just heartbreaking not to be able to offer the families any sort of treatment or hope.
I soon found a number of patients with Rett Syndrome, and two things convinced me that it had to be genetic in origin, despite the fact that the cases I was seeing were all sporadic. First, the disease appeared to affect only females, which suggested the gene could be on the X chromosome. Second, the timeline of progression — loss of language and social interaction first, then development of motor problems, then epilepsy and autonomic dysfunction — was remarkably consistent, despite the variability in the severity of symptoms. But some clinicians doubted that Rett Syndrome was actually a distinct clinical entity at all. Many scientists considered it a fool’s errand to pursue the genetic basis of a (mostly) sporadic disorder, so the early years were pretty discouraging. When I located a couple of families with more than one affected individual, this helped us narrow our search to a specific region on the X chromosome. Finally, 16 years after seeing my first Rett patient, my lab identified the responsible gene as methyl cytosine-binding protein 2 (MECP2). This gene was discovered in the early ’90s by Adrian Bird, who found that it is involved in control of gene expression. Our discovery opened up the field of studying epigenetics in neurodevelopment.
Identifying the gene enabled us to make mouse models of different MECP2 mutations and helped us really begin to understand Rett syndrome and related diseases. The variety of phenotypes that can be caused by mutations in MECP2 is very broad, ranging from neonatal encephalopathy and death within one year in male infants, to classic Rett syndrome, to simple autism or very mild learning disability, if the female child has a very mild mutation that preserves most of the protein’s function or very favorable X inactivation (i.e., the chromosome with the mutated MECP2 gene is preferentially silenced, so that it is mostly the “good” MeCP2 protein that gets expressed). Some children even develop an early-onset schizophrenia or bipolar disorder. To see what happens when the brain gets too much MeCP2, we created mice that overexpress the protein at twice or three times its normal levels. These mice developed a distinct neurological phenotype that enabled us to go back to the clinic to look for children who had the same symptoms. It turns out that chromosomal duplications of the region of the chromosome containing MECP2 also cause a neurodevelopmental disability, a syndrome of cognitive, affective, and motor impairments in boys. We now know that MeCP2 Duplication Syndrome is one of the most common X-linked intellectual disability syndromes in male children.
We have also learned that a few thousand genes are affected by changes in MeCP2 function. Many of the ~2500 genes that we’ve identified are also associated with autism and other neuropsychiatric disorders. With so many genes affected simultaneously, any gene-based treatment, or efforts to target one specific pathway pharmacologically would be very unlikely to succeed. So we decided to study the neurotransmitters and brain circuits that are most affected in MeCP2 disorders. We discovered that partial reductions of GABAergic or glutamatergic signaling reproduces much of the Rett syndrome phenotype, which led us to propose that many genes involved in GABAergic or glutamatergic signaling (including synthesizing enzymes or transporters) are candidate genes for neuropsychiatric disorders. In studying the electrophysiological behavior of these neurons, which are inhibitory and excitatory, respectively, we also realized that modulating neural circuits in Rett syndrome with deep-brain stimulation (DBS) might be able to provide an avenue for treatment. To test this hypothesis, we focused on learning and memory, which are functions of a healthy hippocampus. We took female Rett syndrome mice and, in collaboration with Dr. Jianrong Tang, performed DBS in the hippocampus for one hour a day over a period of two weeks. Remarkably, after this treatment the Rett mice performed as well as wild-type mice in specific tests of learning and memory, such as the Morris water maze. The treated mice also showed improvements in long-term potentiation (LTP), a sign of enhanced plasticity at the physiological level, and also showed a restoration of hippocampal neurogenesis to the level observed in wild-type animals. DBS is an invasive but well-established procedure that has the potential to improve life for Rett children, but much more testing needs to be done.
Our most exciting prospect for treatment is in the context of MeCP2 duplication syndrome. Here the problem is that there is twice as much MeCP2 protein in the brain as there should be, and we need to reduce the levels back to normal. We know that if you genetically delete one of the two MeCP2-coding alleles in the duplication mice after they are symptomatic, all the features of the disease disappear — the mice become normal. We have therefore collaborated with Ionis Pharmaceuticals to develop antisense oligonucleotides to suppress production of MeCP2 protein. Antisense oligonucleotides, or ASOs, are small, modified DNA molecules that bind the mRNA. When this happens, an enzyme called RNAseH will detect the RNA-DNA pairing and will degrade the RNA so that no protein will be made from this allele. Because children with MeCP2 Duplication Syndrome already have had symptoms for years by the time of diagnosis, we allowed the duplication mice to mature into adults and develop all the features of the syndrome, before administering ASOs that target the extra copy of MeCP2. We infused the ASO directly into the ventricles of the brain over a period of four weeks and succeeded in normalizing the levels of MeCP2. The ASO treatment also eradicated the symptoms of MeCP2 duplication mice, from their extreme anxiety and avoidance of social interaction, to their seizures. Even older adult mice of six or seven months of age that have had the disease features for many months can be restored to health with the ASO treatment.
In conclusion, we’ve learned that the brain is highly sensitive to the levels of MeCP2 protein. I call it the “Goldilocks” principle. You have to have it just right, and I am sure there are some patients that might have 60 percent, or 40 percent of the protein, and they might present with partial symptoms. I hope that one day we will be able to give back to the patients and their families by developing better therapeutics. With that, I would like to thank those who have contributed to the work, particularly Ruthi Amir, who stuck with me and kept sequencing genes until she found the right one; and all the students, fellows and collaborators who have contributed to various aspects of our Rett studies over the years.
I am very grateful to the Rett syndrome and MECP2 disorders families, and to all the funding agencies (particularly to the Howard Hughes Medical Institute whose support really enabled me to continue to study Rett), as well as the NIH, the Rett syndrome Research Trust, International Rett Syndrome foundation, the Keck Foundation, and the Simons Foundation.