In a landmark study for the field of personalized medicine, using “Personal Omics Profiling” a Stanford geneticist and his colleagues, analyzed his genome to predict a genetic disposition to type 2 diabetes, tracked at the molecular level how it developed in his body, and then went away again after dietary and lifestyle changes.
Dr Michale Snyder, Professor and Chair of Genetics at Stanford University in California, and also Director of Stanford’s Center for Genomics and Personalized Medicine, and colleagues published the results of their two-year long study of the most intimate secrets of Snyder’s DNA, RNA, bloodborne proteins, metabolites and signaling molecules, in the 14 March online issue of Cell.
They observed, as he succumbed to two viral infections, and how his immune system responded.
And much to Snyder’s shock, they discovered he had a genetic predisposition to type 2 diabetes. This prompted tests of blood sugar, which shot up as he developed the condition during the course of the study, and then came down again as his timely changes in lifestyle and diet took effect before any lasting tissue damage.
The account gives us a glimpse into a possible future of medicine. Snyder told the press:
“I was not aware of any type 2 diabetes in my family and had no significant risk factors.”
But, because of the detailed genomic sequencing, he learned he had a genetic predisposition to the condition:
“Therefore, we measured my blood glucose levels and were able to watch them shoot up after a nasty viral infection during the course of the study,” he explained.
So, after getting a confirmation diagnosis from a doctor, he changed his diet and exercise regimen and gradually the levels of glucose returned to normal. Had the disease gone undiagnosed, even for a couple of years, the tissue damage that occurs would already have started.
“This is the first time that anyone has used such detailed information to proactively manage their own health,” said Snyder.
“It’s a level of understanding of health at the molecular level that has never before been achieved,” he added.
The study describes the mind-boggling number of data points that would have to be collected in a personalized medicine approach to health compared to the cursory workup most of us receive when we have our regular physical checkup.
“Currently, we routinely measure fewer than 20 variables in a standard laboratory blood test,” said Snyder.
“We could, and should, be measuring many, many thousands,” he added.
It appears, that to find what makes an individual unique, as a far as trying to predict and analyze health and disease, you need more than DNA and genes: you have to investigate the surrounding sequences as well, say the researchers.
They call the detailed analysis they carried out on Snyder an “integrative Personal Omics Profile”, or “iPOP”. The word “omics” indicates that a large body of information, such as that acquired from a genome (a map of all the DNA in a cell) or the proteome (all the proteins).
An iPOP relies on collecting and analyzing billions of individual bits of data.
Snyder’s iPOP also included his metabolome (a map of all his metabolites), his transcriptome (RNA transcripts), auto-antibody profiles, and other things.
The researchers say that the diabetes predisposition they discovered in Snyder’s case is but one of many, many problems that an iPOP can uncover. They believe such “dynamic monitoring” of individuals will soon be quite common.
Over the course of the study, Snyder gave about 20 blood samples (about once every two months, and more often when he was ill).
Each sample underwent a variety of tests (assays) for tens of thousands of biological variables.
But it was what they found on day 301 that Snyder says was particularly informative. This was about 12 days after a viral infection: his glucose regulation seemed to have gone awry. Not long after this, his blood glucose shot up. So he went to his own doctor, which is how on day 369, he was diagnosed with type 2 diabetes.
Snyder said normally he would go for a checkup with his doctor every two or three years:
“So, under normal circumstances, my diabetes wouldn’t have been diagnosed for one or two years.”
“But with this real-time information, I was able to make diet and exercise changes that brought my blood sugar down and allowed me to avoid diabetes medication,” he explained.
Snyder started his study a few months after arriving at Stanford in 2009, when whole-genome sequencing of individuals was just beginning to be looked at seriously as a medical option.
The iPOP approach takes the genome-sequencing a step further. A person’s genome is a DNA blueprint that changes little over time. It has some predictive power to hint at future events, such as the risk of developing certain diseases.
But organisms are dynamic: they change over time. The DNA blueprint has to be interpreted by RNA, and then expressed in proteins, which have to send signals and drive chemistry to sustain life and interact with the environment.
The iPOP takes into account this more dynamic store of information, so if you like, it offers not only the snapshot of the genome, but also how the genome is working over time, within the context of a person’s environment: how they metabolize food, flex their muscles, breathe, react to infection, and all the little nuances and adjustments that keep the parameters of biology within what we may define as “healthy”.
To generate Snyder’s iPOP, the researchers had to sequence his genome at a level of accuracy that had not been achieved before.
After that, they took dozens of molecular snapshots, using a plethora of various techniques, to generate thousands of variables and track their progress over time.
It was a cluster of molecular cues that led to the discovery of diabetes. From the genome sequencing, they could see that Snyder had an increased risk for: high cholesterol, coronary artery disease, basal cell carcinoma, and type 2 diabetes.
Snyder was surprised by the diabetes prediction: he was already aware of the heart disease risk.
But, compared to other men of his age (he was 54 at the start of the study), he had a lower risk for: high blood pressure, obesity, and prostate cancer.
At the start of the study, Snyder also had high levels of triglycerides in his blood: 321 mg/dL. The levels dropped to 81-116 mg/dL when he took simvastatin, the cholesterol buster.
Also, at the start of the study, Snyder’s blood samples showed he had normal levels of blood sugar, but, because the genome analysis showed a type 2 diabetes risk, the researchers decided to continue monitoring this.
They report that Snyder acquired two viral infections over the period of the study. He started with a viral infection (rhinovirus, on day 0), and during the first year (on day 298), they record he acquired respiratory syncytial virus.
They note that both times, his immune system responded to the virus attacks by increasing blood levels of pro-inflammatory cytokines. These are proteins that cells produce to communicate and coordinate their responses to things around them such as attacks.
Snyder’s samples after the viral infections also showed he had higher levels of auto-antibodies: these react with the body’s own proteins. This is usually a normal temporary reaction after illness, but in Snyder’s case, the researchers were curious that one auto-antibody in particular had targeted an insulin receptor-binding protein.
As if this level of detail was not enough, the researchers went one step further, they took a peep at what was happening inside Snyder’s cells. This is where the RNA transcript analysis comes in. This analyses the transcriptome, a snapshot of the ongoing dynamic interpretation of the fairly static genome into the set of instructions that drive cellular chemisty.
“We generated 2.67 billion individual reads of the transcriptome, which gave us a degree of analysis that has never been achieved before,” explained Snyder.
Overall, Snyder and colleagues tracked nearly 20,000 distinct transcripts coding for 12,000 genes and measured the relative levels of more than 6,000 proteins and 1,000 metabolites.
From such an incredible level of detail, they noticed some processing and editing of instructions that nobody had suspected. For instance, each individual carries two copies of a gene (one from each biological parent), and the researchers discovered, using Snyder’s transcriptome data, that these behave quite differently during infection.
They found about 2,000 genes that appear to be more highly expressed during infection. Some of these are involved with immune processes and the devouring of infected cells.
They also identified another 2,200 genes that were expressed at a lower level during infection, including some involved in insulin signalling and responding.
In Snyder’s case, by cross-referring results from different “omes”, they also found some unexpected pathways and links between viral infection and type 2 diabetes.
The study, using Snyder’s iPOP, is intended as a “proof of principle”, and the hope is that it will open the door to more streamlined and less complex approaches that can be used in the clinic.
We may not need 40,000 variables in the future, says Snyder. It’s possible we can do the job with just a subset that prove to be really useful for predicting future health.
In the meantime, we will need more studies like this to whittle down those 40,000 or so to a more manageable number.
“Right now, this type of analysis is very expensive. But we have to expect that, like whole-genome sequencing, it will get much cheaper. And we also have to consider the savings to society from preventing disease,” said Snyder.
Answering the question why he chose himself as the subject of the study, Snyder said in a statement reported by Science NOW, that the reasons were mainly practical. He wanted someone local who could come in to undergo the frequent blood tests, and he also wanted someone who would not turn on his team if some devastating information came up:
“I wasn’t going to sue myself,” said Snyder.
Funds from Stanford University, the National Institutes of Health, the Spanish Ministry of Science and Innovation, the European Union, the European Research Council, the Korber Foundation, the Fundación Marcelino Botín and Fundación Lilly and the Breetwor Family Foundation helped pay for the research.
Written by Catharine Paddock PhD