EU-AIM leader talks about autism, finding the elusive SHANK3 mutation and whether PMS cases are common

Dr. Thomas Bourgeron is one of the world’s foremost experts in the genetics of neuropsychiatric conditions, especially autism. He works at the Pasteur Institute in Paris and is a leader in EU-AIMS, a European autism research consortium in which SHANK3 – the cause of the majority of Phelan-McDermid Syndrome cases — is an area of focus. April is Autism Awareness Month and PMSF Research Support Committee Chair Geraldine Bliss talked with Dr. Bourgeron about PMS, SHANK3 and autism.

BLISS: You recently led a team of thirty-seven researchers from around the world in a project where you looked for deletions and mutations of the genes SHANK1, 2 and 3. You compiled data from 14 published datasets and found that deletions and truncating mutations of SHANK3 occurred in 0.69 percent of people with autism.  When you looked at individuals who have both autism and intellectual disability, the rate is even higher, 2.12 percent.  (A truncating mutation is genetic spelling error that prevents the proper expression of the normal protein product.)

How do those figures compare to previous estimates about the rate of SHANK3 deletions and mutations in people with autism and developmental disability?

BOURGERON: This collaborative study is the first meta-analysis of SHANK mutations in ASD (Leblondet al. PLOS genetics 2014). There was especially a need to gather results of sequencing studies. A second objective of our study was to provide a clinical characterization of patients carrying SHANK mutations.

BLISS: What is the reason for the increase in the estimates of SHANK3 deletions and mutations from past estimates?

BOURGERON: We can divide SHANK3 genetic alterations in three types: First, large deletions encompassing SHANK3 (also known as 22q13 or 22qter deletions) are screened by standard karyotypes (a method to observe the chromosomes of one individual). Second, small deletions of SHANK3 (also known as copy number variants (CNV)) are well screened using DNA arrays (a method to identify small loss or gain of DNA). Finally, point mutations affecting a single nucleotide (letter) of the SHANK3 gene are much less studied since you need to use either classical Sanger or next generation sequencing methods.

In our study, we have increased the number of patients and controls screened for SHANK mutations (especially, we screened a new cohort of patients for SHANK3 point mutations by Sanger sequencing). This is one of the two reasons why the prevalence of SHANK3 mutations has raised. The second reason is that we have extended the screening of patients with autism and intellectual disability. Given that mutations of SHANK3 are found more often in patients with intellectual disability, we could detect more mutations than reported in previous studies.

BLISS: How does that rate compare to other genetic causes of autism and developmental disability?

BOURGERON: The group of Evan Eichler has recently published a map of genes associated with developmental disorders (Coe et al. Nat Genet 2014). The map is based on the analysis of 29,085 patients with developmental disorders and 19,584 controls. In this map, SHANK3 is the third locus associated with developmental disorders with a prevalence of the deletion in 0.5 percent of the patients. SHANK3 is one of the main genes associated with developmental disorders.

BLISS: The PMS community includes many individuals with deletions of 22q13 that span SHANK3 and a growing number of individuals with mutations of SHANK3.  Mutations of SHANK3 continue to be difficult to identify.  What are the challenges in sequencing SHANK3?

BOURGERON: There are unfortunately many challenges in sequencing SHANK3. First, the gene contains more than 22 exons (different fragments of the gene that you have to study individually). Second, the gene is extremely rich in nucleotides G and C. This feature increases the difficulty to sequence the DNA (even the exome sequencing strategy was not able to sequence some exons). Third, there were several mistakes in the SHANK3 sequence of the referenced human genome.

The good news is that we have corrected the sequences (the corrections are listed in Leblond et al. PLOS genetics 2014) and the new whole genome sequence methodologies greatly improve the screening of SHANK3 exons.

BLISS: Recently I’ve been hearing about new, very large sequencing studies.  Do you think the rates of SHANK3 deletions and mutations are likely to increase in larger studies?

BOURGERON: Unfortunately, the whole exome sequencing (a technology that enable the sequence of the coding regions of the genome) has difficulties to identify SHANK3 mutations (the same is true for SHANK1, less for SHANK2). The whole genome sequencing technology is much better, but still expensive (>$1,000 per genome). We are however confident that the price will drop soon.

BLISS: When you totaled up the data from the fourteen published datasets, you found that the prevalence of SHANK3 deletions in people with autism was 0.18 percent and the prevalence of SHANK3 truncating mutations was 0.51 percent.  I was surprised to see that mutations of SHANK3 occurred more than twice as often as deletions of SHANK3.  Could that have been a result of other factors in the studies (such as how they developed heir cohorts, perhaps including patients with autism but excluding patients with ID)?

BOURGERON: The IQ of the patients can greatly bias the results. This is why we have stratified the population for presence (IQ70) of intellectual disability.

BLISS: I was also surprised to see that in a cohort of patients with autism that you screened, 8 of 429 (1.86 percent) had truncating mutations of SHANK3. This is more than three times the rate the meta-analysis would have predicted.  What explains the very high yield of SHANK3 mutations in your cohort?

BOURGERON: The mean IQ of our cohort is lower than other cohorts, such as the Simon Simplex Collection (SSC). This is the main reason why our prevalence of SHANK3 mutations is higher. When IQ is taken into account, there was no difference between cohorts.

BLISS: What do your new findings mean for clinicians when they see patients with autism or developmental disability?

BOURGERON: Given the relatively high prevalence of SHANK3 mutations, we have suggested that SHANK3 should be routinely tested for patients with ASD and intellectual disability. As explained above, the screen for SHANK3 mutation remains difficult to perform, but hopefully the screening will be easier in the very near future.

BLISS: In the United States, we don’t have any surveillance programs for PMS or other rare neurodevelopmental disorders.  Do you care to wager a guess about the actual number of people who have deletions and mutations of SHANK3?

BOURGERON: This is an important question and unfortunately I have no precise answer on the number of patients with SHANK mutation. Based the current estimates, I would say that the prevalence of SHANK3 mutations is 1/10,000-20,000 births. This is less than Fragile X (1/4,000 males), but similar to Rett syndrome (1/10,000 females).

BLISS: Thank you for talking to us about your work. We wish you continued success.

3 main objectives of Dr. Bourgeron’s group

(1) Dr. Anne Claude Tabet aims at identifying all the patients with deletions or duplications including SHANK3 in France. We have already identified 50 patients with 22q13 deletions. This work is done in collaboration with all the cytogeneticists from France and should be completed this year. We will also extend this study to other European countries as we are currently doing so with the EU-AIMS project.

(2) Dr. Roberto Toro aims to identify modifier genes that influence the severity of SHANK3 mutations. As you know, some patients have difficulty walking and some go to school and can talk. We suspect that the genetic background of each individual will greatly influence the severity of the clinical outcomes. Because the genetic background is made of a large number of variants (>1,000 or maybe > 10,000 variants), we need to have genetic data on a very large population of patients carrying SHANK3 mutations. The PMS association will be very helpful for this project.

(3) Dr. Richard Delorme aims to initiate new clinical trials in patients carrying SHANK3 mutations. To identify new pharmacological treatments, Dr. Isabelle Cloëz-Tayarani, in collaboration with Alexandra Benchoua at iStem and Tobias Boeckers at Ulm university, is studying neurons derived from induced pluripotent stem cells (iPSC) of patients carrying a SHANK3 mutation. Dr. Elodie Ey is studying the social communication of mice mutated for SHANK2 or SHANK3. Using these cellular and animal models, we are currently testing the effect