Harvard-Partners Center for Genetics and Genomics
Harvard Medical School
Name: Raju Kucherlapati
Position: Scientific director, Harvard-Partners Center for Genetics and Genomics, since 2001; Professor of genetics and medicine, Harvard Medical School, since 2001
Experience and Education:
— Professor and chairman, department of molecular genetics, Albert Einstein College of Medicine, 1989-2001
— Professor, department of genetics, University of Illinois College of Medicine, 1982-1989
— Assistant professor, department of biochemical sciences, Princeton University, 1975-1982
— Postdoctoral fellow in genetics, Yale University, 1972-1975
— PhD in genetics, University of Illinois, 1972
— Undergraduate degree in biology, P.R. College, Kakinada, India, 1960
As scientific director of the Harvard-Partners Center for Genetics and Genomics, Raju Kucherlapati helps to translate new research findings in genetics to the clinic.
The center also evaluates new technology platforms for cheaper and faster genetic testing. Recently, HPCGG’s Laboratory for Molecular Medicine, a CLIA-certified clinical diagnostic laboratory that offers a variety of genetic tests, switched two multi-gene tests from Sanger capillary electrophoresis sequencing to an Affymetrix resequencing chip platform, followed by confirmatory Sanger sequencing.
In Sequence spoke with Kucherlapati last week about the new tests, and the potential of new sequencing technologies in genetic testing.
Tell me briefly about the Harvard Partners Center for Genetics and Genomics, and the role of the Laboratory for Molecular Medicine within it.
One of the major goals of the center is to bring genetics and genomics to clinical medicine. We felt that it is very important for us to do that because we recognize that genetics is now well established to play a very important role in virtually all human diseases. Genetic information can help in assessing an individual’s risk for particular types of diseases, genetics will help us with diagnosis for diseases, and genetics can help determining the appropriate drugs for patients, and making decisions about the right doses of the drugs. The center wants to be able to find methods and strategies where this type of information can indeed be brought to the clinic.
We felt that diagnostics would be an important component of bringing genetics to the clinic. To facilitate that, we established this Laboratory for Molecular Medicine, which is a CLIA-approved diagnostics laboratory. That facility provides a variety of genetic and genomic tests that is in support of this notion of how to bring genetics to the bedside.
What kinds of testing does the LMM provide, and how many of them involve DNA sequencing?
We do a variety of different types of tests. Some of them involve genotyping, others involve detection of deletions, others involve sequencing. The genotyping tests could involve a few genotypes, or very large numbers of genotypes. We have a variety of different types of platforms that we use. Certainly, a large number of tests involve DNA sequencing.
What platforms do you use right now at the HPCGG?
We have many different platforms. For genotyping, we use our own TaqMan systems, or we use Sequenom, and we use Illumina, and Affymetrix.
And then we have also some specific systems for specific types of test. We provide testing for Warfarin dosing, and we use a machine that is produced by a company named AutoGenomics to do that. And then we have evaluated other companies that make other types of systems, like Osmetech. We are completely technology-agnostic.
For sequencing, we use capillary electrophoresis, and currently, we use Illumina’s Solexa machines. But we are completely agnostic, and we might consider using ABI’s SOLiD systems, or 454’s system, or Helicos’ system. We are [also] evaluating Polonators that are developed here [at Harvard by George Church’s group].
Tell me how some sequence-based genetic tests have recently evolved at the LMM.
When we were established, all of the sequencing tests were performed on the basis of capillary electrophoresis. One of the things that we wanted to be able to tackle [was] complex problems in DNA and sequence-based diagnostics. One of the decisions that we made at the time [was] how to deal with types of genetic disorders in which there is a lot of heterogeneity.
The first example that we chose was hypertrophic cardiomyopathy. We chose [HCM] for two reasons: One is, much of the discovery of the genes that are known to be involved in [HCM] have been made by our colleagues Christine Seidman and Jon Seidman. Both of them are in the genetics department that I am a member of, and both of them are also members of the Harvard Partners Center for Genetics and Genomics. At that time, this testing was not offered at any palace in the world. We understood the importance of it, and the complexity of the testing, so we decided that we wanted to begin to offer the test.
The testing is complicated because this particular disorder is an autosomal dominant disorder, and as a result, there are lots of new mutations that occur. So it was necessary to sequence all of the genes in their entirety to be able to assess whether there are mutations in any one of those genes. We developed a test to sequence 11 genes, but the issue that we dealt with is the cost of providing such tests. We are always mindful of the cost of genetic testing.
So we decided to offer the testing, initially, in a tiered fashion: Take those genes which account for the majority of the known mutations as tier 1, and then, if a patient’s DNA is found to be negative for all of the first tier of genes, then we go to the second tier. And if they are negative in that place, then go to the third tier. The minimum for each sample was to go to tier 1, and that cost $3,500. And if they had to go through all of the genes, then it cost as much as $6,000 to do the test. So we felt that it is important for us to think of ways in which that can be significantly reduced. [Also], the time that it takes to complete all of these tests is very significant, [weeks or months], because you do it in a serial fashion. So we wanted to explore reliable alternative technologies that would enable use to examine all of these different genes to reduce the cost and to reduce the turnaround time.
To facilitate that, we entered into a relationship with Affymetrix and developed a resequencing chip that would contain all of the 11 genes on a single chip. That took a significant amount of effort. However, in November of 2007, we validated the chip, and we [have been using] those chips to resequence all of the genes. [We first run the] Affymetrix chip-based testing. If we find a mutation, that is confirmed by capillary sequencing. The accuracy of the testing is better than 99.9 percent. The [total] cost is $3,000, and the turnaround time is about five weeks. So we were able to significantly reduce the cost of providing this [test]. And because it’s a chip, [it provides] highly reproducible results.
Are you planning to use the Affy platform for other tests as well?
[The CardioChip] is the first test that we have launched on the Affy [platform]. The second one is another disorder called Noonan syndrome. I was involved in the discovery of the first Noonan syndrome gene in 2001. However, it turns out that the mutations in that gene account only for about 50 percent of Noonan syndrome patients. At that time, we did not know what the other genes are that cause Noonan syndrome. Noonan is also an autosomal dominant syndrome.
Since 2001, we and other people have discovered mutations in a number of other genes that cause Noonan syndrome. During that period of time, we also discovered that a couple of other disorders — one is called Cardio-Facio-Cutaneous, or CFC, syndrome —are very closely related to Noonan, and the genes in which mutations can be found all belong to one class of genes in the so-called MAP kinase pathway.
So suddenly, there was a situation where we were able to [analyze] only one gene, and that was able to detect only 50 percent [of patients], and now have about seven genes that need to be tested for these patients. [The test] for one gene, at the time it was offered, [cost] about $1,000. And again, if we offer testing for each of those genes, as we have been doing, the time and the cost of providing the testing are just prohibitively expensive.
So we have continued to work with Affymetrix and generated a resequencing chip called the Noonan Spectrum Chip that simultaneously sequences all of these genes in one shot. And again, anything that is found in those is confirmed by capillary electrophoresis [sequencing]. We just launched that [test] about a month ago, and the cost of that is $1,500. Again, I think that we are able to offer a test that is much more comprehensive than what anybody else is offering, and at a price that we feel is much more affordable for a lot of people.
What is the potential of next-generation sequencing platforms in genetic testing?
First of all, we recognize, as we have recognized very early on, that testing can become more complex. An ultimate possibility is that instead of doing sequencing for a few genes, maybe in the not-too-distant-future, we would sequence the entire genome, and all of the information pertaining to that individual can be obtained at one time.
If that is the case, then clearly, the capillary electrophoresis systems are not going to be able to do that and we need the next-generation sequencing technologies. So we are very interested in these technologies. Obviously, the cost of doing whole-genome sequencing with these next-generation sequencing technologies is still pretty high if you deal with it on a patient basis. But we are exploring these technologies, and we work with George Church, [another] colleague here, and Jon Seidman. We are all interested in applying these technologies for these types of purposes, and we are continuously evaluating new technologies to see which of them would be feasible.
It may go stepwise: One step, maybe to sequence 10, 15, 20, or 30 genes, initially; then it is possible that we will sequence all of the exons in a genome; and then, finally, move towards sequencing the entire human genome.
The issues about doing all of these different things is, we are doing some of them on a research basis, but offering them as a clinical tests is completely different, because in a CLIA environment and a CLIA laboratory, you have to validate all the tests that you do, and write appropriate operating procedures for them, and be able to have a really reproducible way to collect and analyze information and interpret the information to provide a comprehensive understandable report to the ordering physician.
We are evaluating [next-gen sequencing] for a variety of different types of approaches, small numbers of genes as well as very large numbers of genes. If we find that those methods are very robust, and that we would be able to provide our results for the sets of genes that we are interested in at a lower cost with equally high levels of accuracy, and with a significantly short turnaround time, we will offer such testing.
What are the pros and cons of using a resequencing chip, compared to a second-generation sequencing technology, for genetic tests?
The resequencing chips are less flexible but they are highly portable. Because [of] the manufacturing process, they are highly controlled, so once you make the masks, they can be produced as many times as you want, and then the results are going to be absolutely consistent from then on.
Also, it is conceivable that some of these types of tests may go to FDA for approval. And if that is the case, then having not a process, but actually a physical element such as a chip, it might be easier. Whereas with the sequencing [platforms], it is not a physical product, it is just a process by which you do that, like doing capillary sequencing. It’s more based upon the individual expertise of the laboratory rather than a product that you can use. That’s an important difference.
It’s like getting software for your computer. If Microsoft makes it, it’s reliable, and everybody can use it. Similarly, in this particular case, if it is a chip like the one that Affymetrix makes, then we know what the parameters are, and everybody has experience with them, and once we have developed this chip, it can, at least theoretically, be used by anybody in the world. So the dissemination of that is easier, whereas with the next-gen sequencing, each of the laboratories has to be able to develop the expertise to do that. Of course, people can develop the appropriate expertise, and it would have [greater] flexibility. But you know, one of the things about diagnostics, it’s not that you can add just a gene at any time that you want to, because we need to develop the processes for them and document them very carefully and validate all of the results very carefully, and each of them takes a significant amount of time.
What platform will be mostly used for genetic testing in the future?
We don’t believe that any one platform is going to be the predominant platform. There are going to be situations when you want to look at one nucleotide, or one gene, or five genes, or 50 genes. Next-gen sequencing is not going to be good if you want to be able to look at a single nucleotide polymorphism in one gene, whereas if you want to look at 20 genes, it might be very good.
It may turn out that a different platform is most appropriate for a particular application. So at the center, we continuously evaluate them, and then the appropriate types of technologies can be easily translated to the Laboratory for Molecular Medicine.