Like an oasis in the desert, the splendor of precision medicine seems perpetually on the horizon. Even as technological advances bring genome sequencing into routine clinical use, it’s becoming clearer than ever that genomics reveals only part of the clinical picture.
“Genes are static,” said Vinit Mahajan, MD, PhD, an associate professor of ophthalmology at Stanford University. “They don’t tell us when the disease is active, when it’s going to start, when it’s going to stop.” For real-time analysis of a patient’s disease state, he said, “You need to look at proteins.”
Proteomics is a lot harder than genomics, because of the complex chemistry, but it can be richly informative. A new generation of multiplexed antibody arrays is helping cut through some of that complexity. Quantitative, customizable multiplex ELISA arrays allow researchers to cast a net into a pool of potential biomarkers. That net can be small and focused or large and diverse, for a variety of research applications.
For Mahajan, a vitreoretinal surgeon, protein testing reveals much that’s not apparent from visual imaging. For example, similar eye symptoms arise from four very different conditions: metastatic lymphoma, or infections with bacteria, fungus, or virus. Each of those requires a different treatment plan. “Despite our amazing clinical imaging, those four diseases have overlapping phenotypes,” he said. Time spent testing for each possibility is wasted treatment time. A protein assay could cut down diagnosis time and speed up treatment. “What we’ve found is that even though, clinically, they look similar, the fluid in the eye has very obvious proteomic signatures for each of those conditions,” said Mahajan.
Inflammatory eye disease presents a similar diagnostic challenge, because the symptoms are nonspecific. Broad spectrum anti-inflammatory drugs carry nasty side effects, so it’s better to pick a targeted drug. But often, the process comes down to trial and error.
Antibody-based assays are needed to identify important proteins. Multiplexing enables a screening approach rather than forcing the researcher to hypothesize about what is the important protein. Multiplex immunoassays help eliminate that guesswork by testing for dozens or hundreds of proteins simultaneously.
Recently, Mahajan treated a patient with poor vision due to inflammation of unknown cause. Using a RayBiotech antibody array, his team tested for 200 different cytokines in the vitreous fluid. By comparing the profiles of 15 patients with similar inflammation and 5 healthy controls, they identified several distinct patterns associated with different sources of inflammation. This particular patient’s cytokine profile, they found, indicated autoimmune retinopathy.
By modifying the treatment accordingly, they restored the patient’s vision. Proteomic profiling allowed the doctors to arrive at a diagnosis much faster than conventional methods. “By measuring the proteins in their eye fluid, we can pick the right drug,” Mahajan said, “and not put patients through some long, uncertain, complicated process.”
Urine Luck: Testing for hidden kidney disease
Kidney disease can be even harder to detect and diagnose than eye disease. Most people don’t spend a lot of time thinking about how their kidneys feel. And because we have two kidneys, if one slows down, the other still handles the job pretty well. Kidney damage, then, can build up from a variety of causes, and the patient won’t know until the organs begin to fail.
“If 5-10% of your kidney is injured by a given drug, you would not even see a difference,” said Chirag Parikh, MBBS, PhD, the director of the Division of Nephrology at Johns Hopkins University School of Medicine. Dr Parikh is studying acute interstitial nephritis (AIN), a kidney problem often caused by allergic reaction to medication. Currently, diagnosis requires a kidney biopsy, an invasive procedure that can lead to complications. Without decisive symptoms, clinician might be reluctant to order such a test, and any related condition can go undetected.
“We want to find this because it’s treatable,” Dr Parikh said. “You can stop the drug; you can give steroids. And if you don’t treat it, you are left with irreversible kidney damage.”
To make detection easier, Dr Parikh and his team went looking for biomarkers associated with the disease. Using a 10-cytokine array, they tested both plasma and urine; they also developed their own urine assay to detect 2 additional cytokines. In the plasma, they found no biomarkers that reliably indicated the disease, but 2 cytokines associated with AIN turned up in patients’ urine: TNF-α and IL-9.
Finding IL-9 made biological sense, because IL-9 is associated with mast cells, the immune cells that induce allergic responses by releasing histamine. The presence of IL-9 in the urine successfully distinguished AIN from other kidney disease, such as damage from diabetes, which would require different treatment. If larger studies confirm the findings, it could lead to a clinical urine test for AIN.
Easing chemotherapy’s harmful side effects
“In my lab, we’re very interested in the process of stem cell regeneration, particularly in the blood system,” said John Chute, MD, the director of the Outpatient Stem Cell Transplantation Program at UCLA. “We’re very interested in understanding what proteins regulate that process, because it’s largely a black box.”
Cancer patients frequently find their bone marrow stem cells under assault, either as a side effect of chemotherapy or radiation, or in preparation for stem cell transplant. To find ways to speed recovery for these patients, Chute and his team investigated what proteins swing into action to replenish these cells. First, they genetically modified mice to be radiation resistant. These mice recovered their hematopoietic stem cells much faster than wild-type mice after total body irradiation. Using a RayBiotech mouse cytokine array, the researchers discovered an excess of a protein called Dkk1 in the bone marrow of those mice. Dkk1 blocks the WNT signaling pathway, which is involved in development and regeneration of tissues after injury.
“That cytokine result was pretty helpful,” Dr Chute said, because it led the team to hypothesize that Dkk1 enabled the mice to fend off radiation damage. From this finding, the researchers went on to test the protein’s role in stem cell regeneration.
First, they showed that giving Dkk1 to mice that had been lethally irradiated would save them, speeding the recovery of the blood cells and girding the mice against infections. Conversely, eliminating Dkk1 delayed recovery. Bolstered by this finding, the researchers delved into how the protein accomplished this on a molecular level.
“What was perhaps the most interesting thing to us was that we discovered mechanistically that it was doing this in two different ways,” Dr Chute said. Dkk1 acts directly on blood stem cells, they found, to suppress the cell death processes. It also acts indirectly, by prompting endothelial cells to produce epidermal growth factor (EGF), which supports stem cell regeneration. Blocking the activity of EGF, they found, undid the benefits of Dkk1. Finding that interaction was gratifying, Dr Chute said. “This concept of crosstalk amongst the orchestra of niche cells had been much postulated, but not demonstrated yet,” he said. “It’s not just one cell acting on a stem cell to support it.”
Treating chemotherapy patients wouldn’t be as simple as just giving them Dkk1 but understanding the complex interactions at play could help guide drug discovery efforts.
Connecting patients with the right therapies
Few metastatic melanoma patients survive to the 5-year mark, but advances in immunotherapy continue to help people beat the odds. Still, many patients don’t respond to the therapy. “It’s a bad thing, because you’ve lost time for those patients,” said Olivier Michielin, MD, PhD, associate professor of oncology at the University of Lausanne and Group Leader of the Molecular Modelling group at SIB Swiss Institute of Bioinformatics, Switzerland. “They could have maybe done better on another treatment.” Michielin and his colleagues are searching for protein biomarkers that will help match patients to appropriate, personalized treatments. Using an antibody array, the researchers tested the serum of patients with advanced metastatic melanoma for 440 different proteins. “We have a very high degree of confidence in the detection,” Michielin said. “Antibodies are the best tools to recognize proteins that are highly specific, and they also are very high affinity, so they are very sensitive.” In addition, being able to quantify the results adds another layer of information about conditions inside the cell. “Clearly, these arrays are extremely powerful.”
The patients were given immunotherapy, consisting of ipilimumab and nivolumab, and their progress was followed for 36 months. Analysis revealed 3 protein biomarkers that appeared to predict which patients would respond to the treatment. In 90% of the patients possessing the biomarkers, treatment halted the cancer for at least 3 years, compared with 25% of those lacking the markers.
“We are quite confident that we are seeing a real signal,” Michielin said, but cautioned that this result is a first step. Larger trials will show how well the set of markers predicts outcomes. Ultimately, he said, the goal of precision oncology is to group patients based on their biomarkers, and ensure they receive the most effective individualized therapy.
“You want to have ways to fine tune your approach,” he said. “You are working in subpopulations, but if those subpopulations are molecularly defined, you can have a very high chance of success.”