Skip to main content

Columbia's Allikmets on Developing a Suite of APEX Retinal Disease Chips


Rando Allikmets
Director, Laboratory of Molecular Genetics
Columbia University

Name: Rando Allikmets

Position: Director, Laboratory of Molecular Genetics, Columbia University

Background: 2004 - present, Acquavella associate professor of ophthalmic science and director, Laboratory of Molecular Genetics, Columbia University; 2003 - present, research director, Department of Ophthalmology, Columbia University; 2003-2004, Acquavella assistant professor of ophthalmic science, Columbia University; 1996-1999, scientist, Science Applications International Corporation, Frederick, Md.; 1992-1996, visiting fellow, Human Genetics Section, Laboratory of Genomic Diversity, National Cancer Institute, Frederick, Md.

Education: 1988 — PhD, molecular biology, Institute of Bioorganic Chemistry, Moscow; 1983 — biochemistry and virology, Moscow State University; 1982 — BA, molecular biology and virology, Moscow State University.

Rando Allikmets is a molecular biologist who has worked in labs from Moscow to Maryland. For the past two years he has served as director of the Laboratory of Molecular Genetics as well as an Acquavella associate professor of ophthalmic science at Columbia University.
While Allikmets’ function at Columbia is heavily administrative, he has also been using APEX (array primer extension), a genotyping technology commercialized by Tartu, Estonia-based Asper Biotech, to develop a series of prognostic and diagnostic microarrays for various retinal diseases, including Stargardt’s disease, Leber congenital amaurosis, and retinitis pigmentosa.
BioArray News recently spoke with Allikmets to learn more about his efforts to develop new chips for retinal diseases and the application of APEX in diagnostics.
How did you find the shift from doing science to also being an administrator?
I don’t think it’s a big change. If you are a PI and you run your own lab you are an administrator in the first place anyway. And especially with the situation with tough funding — the NIH budget has been stagnant or even reduced recently — so the federal funding opportunities have dropped dramatically in the past few years. It’s a tough life for any scientist to start with. When you have to run an institute and be responsible for a large number of people, it definitely adds a lot of work to what you normally do. But if you are able to guide them and to get them to be successful it is pretty rewarding at the same time.
So the workload did increase, but at the same time I am content.
What kind of array resources do you have at Columbia?
At Columbia I think our resources are pretty good. I have not interacted lately with our core facilities but as far as a whole, I think it’s a pretty comprehensive collection. Mainly, it’s the Affymetrix arrays and that includes all necessary machinery. If you want to run expression arrays or arrays to your liking, you can get all the assistance you need.
I have not used it however, because I have my own small array facility in my laboratory. My specific interests are not expression arrays but in genotyping arrays. We are mainly looking into SNP typing and techniques that allow you to quickly and efficiently screen patients with various eye diseases.
You use array primer extension. Why have you decided to use that method instead of some other genotyping methods?
I have been using that technology for about five years now. The reason I picked it at that time was that I was looking for a technology that is affordable and realistic to apply in an average laboratory. I was looking for something I could use in my lab or use in conjunction with other technology immediately and apply that to my everyday needs.
When looking around I was scanning a lot of technologies. When picking APEX there were two reasons. One was that I liked the basic setup [because it] was a technique you could use in an average laboratory that was flexible and affordable.
The other thing was that a small company, Asper Biotech, had been established in Estonia that had pioneered the application of APEX and, being an Estonian, I always like to give something back to the country. I thought it would be doubly beneficial. I thought that I will get something, and the Estonians will get something. I thought it would be a good idea to try it out at first.
I like this technology especially in terms of diagnostics because you have to have very good certainty of calling a mutation if you want to report it back to a patient. I did not expect it to be like it is now, where we have five or six arrays that I have designed, planned, and implemented. And they have been used for diagnostic screening in thousands of patients with eye diseases.
Which chips are available?
They are pretty much all for retinal diseases. We started with a chip for the ABCR gene — now called ABCA4 — which encodes an ATP-binding cassette transporter involved in the visual cycle of the eye that transports vitamin A. So it’s a very important function.
If you have mutations in that gene you get a disease called Stargardt disease, or other related phenotypes, such as cone-rod dystrophy and retinitis pigmentosa. Variants of this gene are also associated with age-related macular degeneration. So mutations in one gene cause a variety of different phenotypes in the eye. Some of them are pretty devastating.
The gene is large, comprised of 50 exons, so screening it in a substantial number of patients by other means than arrays is time-consuming and expensive; that is almost unrealistic in an average academic setting. So we picked this gene as a model to start our array technology.
We started with 50 mutations for that gene and now we have about 500 on the array. So we’ve updated that chip about twice a year. It definitely is the best method to screen patients even now, five years later.
That was the first model and since it was successful we turned our sights to other eye diseases where the situations were the opposite. You would have one phenotype, for example retinitis pigmentosa (a disease affecting about 1 in 3,000 people), or Leber congenital amaurosis (a disease where infants are often born blind) caused by many genes.
So now we have expanded the chip technology to these diseases; for example, we have designed chips for autosomal recessive and autosomal dominant retinitis pigmentosa. Each of these arrays contains all known mutations from about 15 to 20 genes, resulting in hundreds of mutations. So these chips contain from 400 to 800 variants.
Except for direct sequencing, which is extremely time-consuming and expensive if performed on 20 genes, these arrays are the only tools right now available for screening patients with these diseases. In addition to Stargardt, Leber congenital amaurosis, and retinitis pigmentosa, there are chips now available for Usher syndrome, where in addition to blindness you also have hearing loss, and for Bardet Biedl syndrome.
So there are chips available to screen most of the retinal diseases that we know of.
Who is using them?
They are used by many different laboratories around the world. The laboratories that are investigating eye diseases use them for basic science purposes. They screen entire patient collections and then correlate the genetic findings with clinical findings. They do what we call genotype-phenotype studies for prognostic evaluation of patients.
For example, if you have a patient who has pigmentosa and you find a specific mutation on a specific gene, you can give some idea to the patient how the disease will progress and what they should expect in a certain timeframe.
The same studies are also geared for clinical trials. There is a gene therapy approach being developed where they add the normal functioning gene to affected patients, so they can actually restore the visual function in those people. We don’t know the results on people yet, but studies on the dog model have been very successful.
And, finally, many ophthalmologists use the arrays for diagnostic screening purposes.
Has the technology changed much since you and Asper started providing the chips?
Not really. Usually you can’t change a technology; you have to replace it with a new one. And this technology is, in essence, the old-fashioned standard dideoxy sequencing reaction. Of course the handling of the material and the chip-making has changed, allowing for a faster turnaround of a sample, et cetera. But in principle it’s the same technology.
What does Asper give you once you become a customer?
Asper mainly does the screening themselves. You send them DNA and you get back a report with all identified variants from the sample. For a regular customer it’s pretty much a screening service.
You mentioned that former Secretary of State Henry Kissinger is on the advisory board of your institute. Has he taken an interest in array technology?
He’s on the advisory board of the Eye Institute at Columbia University, which consists of about 20 people. The head of the advisory board is Lou Gerstner, who is the former CEO of IBM. 
We usually have two meetings a year. I presented the array technology at one of those meetings where both were present. I started with a joke that explaining the chip-making technology to Mr. Gerstner is a humbling experience.  I think all the people including Kissinger took interest in our work in this field. I do recall that we had a lively discussion after my presentation.
File Attachments
The Scan

Response Too Slow, Cautious

A new report criticizes the global response to the threat of the COVID-19 pandemic, Nature News reports.

Pushed a Bit Later

Novavax has pushed back its timeline for filing for authorization for its SARS-CoV-2 vaccine, according to Bloomberg.

AMA Announces Anti-Racism Effort

The Associated Press reports that the American Medical Association has released a plan to address systemic racism in healthcare.

Nucleic Acids Research Papers on miRMaster 2.0, MutationTaster2021, LipidSuite

In Nucleic Acids Research this week: tool to examine small non-coding RNAs, approach to predict ramifications of DNA variants, and more.