Harrison Harkins was born with a never-before-diagnosed gene mutation that left him with a host of abnormalities that took his life when he was 9 months old. However, because of cutting edge technology and collaboration that spanned national borders and oceans, researchers that included a team from Baylor College of Medicine identified a new (de novo) mutation in a previously unknown gene and determined that it was the cause of Harrison’s problems. Although the infant died soon after the gene was found, Tim Harkins was able to reassure his son.
“I could honestly look him in the eye and say, ‘You have a better chance of being hit by lightning. This was a random act of nature,'” said Tim Harkins.
A report that appears online in BioMed Central’s open access journal Genome Medicine outlines the process by which this infant and three other children from around the world were identified as having a mutation in this gene.
“Using exome sequencing (finding the genetic sequence of the protein-coding portion of gene) we identified a mutation in the geneASXL3 that results in a novel syndrome,” said Dr. Matthew Bainbridge, a postdoctoral associate in the BCM Human Genome Sequencing Center and first author of the report. “This the first time we have done that outside of a purely research setting. The BCM Whole Genome Laboratory has shown how the mysteries of a novel syndrome can be solved in this way in a diagnostic laboratory.”
The syndrome is unnamed but has symptoms similar to two very rare diseases, Cornelia-de-Lange and Bohring-Opitz syndromes. These include distinctive facial features and posture, small size at birth and subsequent failure to thrive, problems feeding and severe intellectual disabilities.
A genetics expert who had previously seen Harrison Harkins had told the family that a diagnosis could take as long as three years and might never be possible. Tim Harkins, who works for Life Technologies, a firm that produces sequencing technology, sought the aid of Dr. Richard Gibbs, director of the BCM Human Genome Sequencing Center, and Bainbridge.
Looking at the infant’s physical symptoms (phenotype) did not provide an answer, and Bainbridge supervised the sequencing of the child’s exome (the protein-coding region of the genetic blue print) and genome as well as those of both his parents.
“We found a de novo mutation in a gene called ASXL3,” said Bainbridge. This new mutation, not found in either parent, caused the genetic code to be “cut off” or truncated, altering the “recipe” from which the cell could make a protein. No diseases were known to be associated with that gene. Mutations in a similar gene called ASXL1 cause a disorder called Bohring-Opitz syndrome, which had some of the symptoms seen in this child.
However, without other people with the same mutation, it was difficult to know if this gene mutation was the source of the infant’s problems. To solve this problem, Bainbridge and Gibbs did a global search that would connect them to doctors in Europe. They first contacted Dr. Alexander Hoischen of Radboud University Medical Centre in Nijmegen, The Netherlands, as he had reported on a similar gene, ASXL1, to see if he had ever seen a similar case. He responded that a group in Germany had asked him the same question about ASXL3. Dr. H. Hilger Ropers of the Max Planck Institute for Molecular Genetics in Berlin was himself on a quest to find an answer to a family whose child was extremely ill. He did both whole exome and whole genome sequencing in the family and had found a very similar mutation in the same gene.
“Two pieces of evidence pointed to this mutation being the right one,” said Ropers. “One was some resemblance I found between our patient and the patients with mutations in ASXL1. That gene was found by my former colleagues in Holland, and so it was natural to contact them and discuss whether they saw a clinical similarity here. The other link was that we were dealing with twin genes in a way as they were from the same gene family.”
No one knows what function the genes have in the cell, but his Dutch colleagues agreed that he might have a point. Then they called to tell him that a group in the United States had found a similar case. “I figured it might be Jim Lupski (another BCM author on the report) and Gibbs,” he said. Lupski is vice chair of molecular and human genetics at BCM.
The parents of the child he had seen wanted to know if they could have other children with the same condition. After further tests, Ropers was able to tell them that the risk of having another child with the problem was nearly zero.
A further search of the BCM Whole Genome Laboratory database found another patient with a mutation in the same gene, although the mutation was in a different area of the gene. The child has some of the symptoms of the other two but they are not as severe. As Bainbridge was putting his data together and submitting his paper for publication, yet another case turned up whose symptoms were very similar to those of the first two children, and highlights the value research can play in clinical diagnostics at an institution such as BCM.
“This shows that there are many individuals with a condition or abnormality caused by a gene whose function is unknown,” said Dr. V. Reid Sutton, associate professor of molecular and human genetics at BCM, who saw the first infant clinically and followed the sequencing process.”“It takes these sorts of collaborative efforts where we identify patients from around the world and communicate with many international experts to understand what the results mean.”
“Once we could talk to Dr. Ropers and Dr. (Karen) Gripp (of A.I. duPont Hospital for Children in Wilmington, Delaware), who were evaluating children with changes in the same gene and could compare the patients to one another, it solidified that this gene does seem to be the basis of the problems that these children were having,” said Sutton. “The bottom line is that there are still many genes for which we don’t know the function and the number of patients affected is small. It sometimes takes an international collaboration to understand what the results mean.”
Data sharing important
He and Bainbridge both know that there is concern about privacy in establishing genetic databases, but they point out that such information and data sharing are an important part of providing answers to patients and parents.
“In the broader context, my win will be helping other patients,” said Harkins. “It’s the unknown that hurts. If I just help one family by telling our story, that will be important. It would be a professional win if we could encourage the hospital where Harrison was born to embrace exome sequencing to eliminate the guessing for genetic diseases in newborn infants.”
“This is a fabulous diagnostic tool,” said Ropers. “In Germany, people are reluctant to adopt novel technologies.” He hopes that by showing that the technology can provide answers, he will overcome that reticence. Having parents who are willing to share their stories is important.
Bainbridge points out that sequencing was the only way to find an answer in this case. The phenotype or symptoms were very non-specific and were found in other disorders.
“The genotype was very specific. The value of the phenotype to diagnosis is questionable when the genotype is so specific,” he said. Because the “language” of genetics is so specific, it is easy to share information, he said.
Quality of life
As long as researchers and patients or their parents are willing to share information of this kind, it will become easier and more common to answer these questions that so often plague geneticists and their patients.
Having an answer enables patients to balance treatment and quality of life, said Tim Harkins. “If you know what is afflicting your child, you can make the right decision for quality of life.”
Others who took part in this research include Donna M. Muzny, Brett H. Graham and Yaping Yang, all of BCM; Hao Hu, Luciana Musante, Wei Chen and Thomas F. Wienker, all of Max-Planck Institute for Molecular Genetics in Berlin and Karen W. Gripp and Kim Jenny of A.I. DuPont Hospital for Children in Wilmington, Delaware.
Funding for this work came from the Baylor-Hopkins Center for Mendelian Genomics (NIH/NHGRI 1U54HG006542-02) and the Baylor Human Genome Sequencing Center (NIH/NHGRI U54HG0033273), Max Planck Society and from EU-FP7 project GENCODYS, grant no. 241995.