Kenneth Williams, PhD

A major focus of Dr. Williams is overseeing and continuously improving the Yale/NIDA Neuroproteomics Center that brings exceptionally strong Yale programs in proteomics and signal transduction in the brain together with neuroscientists from nine other institutions across the U.S. to identify adaptive changes in protein signaling that occur in response to substances of abuse. Twenty-three faculty with established records of highly innovative research into the molecular actions of psychoactive addictive drugs, as well as of other basic aspects of neurobiology, are working together in a unique synergy with the Keck Foundation Biotechnology Laboratory to support the Yale/NIDA Neuroproteomics Center. The main goal of the Center, whose theme is “Proteomics of Altered Signaling in Addiction”, is to use cutting edge proteomic technologies to analyze neuronal signal transduction mechanisms and the adaptive changes in these processes that occur in response to drugs of abuse. With Co-Directors Drs. Angus Nairn (Psychiatry) and Kenneth Williams (Mol. Biophys. & Biochem.) in the Administration Core, the Center includes Discovery Proteomics (DPC) and Targeted Proteomics (TPC) Cores. Biophysical technologies from the DPC extend protein profiling analyses into the functional domain while lipid analyses from the DPC positively leverage proteome level analyses to provide an increasingly biological systems level approach. A Bioinformatics and Biostatistics Core, which includes high performance computing and the Yale Protein Expression Database, provides essential support that positively leverages the value of each of the proteomic technology cores. A Pilot Research Project Core is a cornerstone in the Center’s efforts to encourage strong mentoring relationships that help attract and train future outstanding scientists. Behavioral adaptations that accompany drug addiction are believed to result from both short and long-term adaptive changes in brain reward centers. Thus, exposure to drugs of abuse regulates intracellular signaling processes that alter gene expression, protein translation, and protein post-translational modifications. Repeated exposure to drugs of abuse leads to stable alterations in these signaling systems that are critical for the changes in brain chemistry and structure of the addicted brain. The Center’s research goals include analysis of the actions of cannabis, cocaine, nicotine, and opioids on these intracellular signaling pathways in brain reward areas and development of methods that enable proteomic analysis of the single types of neurons that define the circuits that underlie the actions and addictive properties of drugs of abuse. Targeted and data-independent mass spectrometry analyses of signaling proteins implicated in the actions of drugs of abuse are being used to analyze the impact of substance abuse on the neuroproteome with motif-based, “Top-Down” MS/MS, and other approaches being used to study protein post-translational modifications. A major initiative led by the Bioinformatics and Biostatistics Core is to develop novel methods for deep integration of genomic, transcriptomic, and proteomic data with brain region and cell type-specificity.

A second area of interest for Dr. Williams is identifying the early protein biomarkers of Delayed Graft Function (DGF) following kidney transplant that are needed to improve the treatment of patients at highest risk of DGF. Kidney function during the first week following renal transplant varies tremendously, with some recipients experiencing immediate graft function (IGF, characterized by a rapid fall in serum creatinine), while others exhibit DGF and require at least one treatment of dialysis post-transplant. While DGF occurs infrequently in living donor kidney transplants, its incidence in deceased donor transplants is 20 to 33%. Recent strategies for increasing the recipient (e.g., elderly) and donor pools have also increased the risk of sub-optimal allograft function. Hence, both “extended-criteria donor” (ECD) and “donation after circulatory determination of death” (DCD) kidneys are associated with higher rates of DGF as compared with standard-criteria kidneys. The short term negative impact of DGF, which is caused primarily by ischemia-reperfusion injury (IRI) during allograft procurement and transplantation, includes increased lengths of stay and hospital costs primarily because of the need for dialysis. Over the longer term, DGF is associated with a >40% increased risk of graft loss. Current approaches for diagnosing DGF or SGF often include need for dialysis, changes in serum creatinine, and urine output. However, all three approaches are retrospective and can be confounded by residual native kidney function. As with other forms of acute kidney injury (AKI) caused by IRI, the delay in diagnosis necessitated by these retrospective approaches greatly impedes efforts to prevent or treat renal injury. Such a delay is particularly pernicious in the setting of transplant as the most common immunosuppression regimens utilize nephrotoxic calcineurin inhibitors. Rapidly distinguishing DGF from IGF post-operatively could allow early tailoring of immunosuppressants, both agents and doses, to renal function. Current research centers on the use of the Targeted Urine Proteome Assay (TUPA) that was described by Cantley et al [(2016) Proteomics Clinical Applications 10, 58-74 (PMCID: PMC5003777)] to identify protein biomarkers of delayed recovery from kidney transplant. Potential biomarkers were identified by using the TUPA Multiple Reaction Monitoring (MRM) assay to interrogate the relative DGF/IGF levels of expression of 167 proteins in urine taken 12-18 hours after kidney implantation from 21 DGF, 15 SGF (slow graft function), and 16 IGF patients. An iterative Random Forest analysis approach evaluated the relative importance of each biomarker, which was then used to identify an optimum biomarker panel that provided the maximum sensitivity and specificity with the least number of biomarkers. Four proteins (C4b-binding protein alpha, guanylin, immunoglobulin superfamily member 8, and serum amyloid P-component) were identified that together distinguished DGF with a sensitivity of 82.6%, specificity of 77.4% and AUC of 0.891. This panel represents an important step towards identifying DGF at an early stage so that more effective treatments can be developed to improve long term graft outcomes. Future studies will be directed at validating these results in an independent patient cohort and at further improving this panel.

The third area of interest for Dr. Williams is identifying serum protein biomarkers for ovarian cancer. With an incidence of 12.1 and death rate of 7.7 per 100000, ovarian cancer is the deadliest gynecological cancer and the fourth most frequent cause of cancer death in women. Ovarian cancer has been termed the “silent killer” because of the lack of early warning symptoms. Although ~90% of patients have symptoms (e.g. frequent urination, pelvic pain, fatigue, abdominal distension) before diagnosis; the symptoms usually are too vague to prompt a visit to a physician or are easily confused with other illnesses. Hence, ~70% of women diagnosed with this cancer have advanced disease, where the 5-year survival rates are <30%. In contrast, for the ~15% of patients who are diagnosed early when the cancer is confined to the primary site (i.e., Stage 1), the 5-year survival rate is >90%. The >3-fold increase in survival rates for patients with localized disease and the >14,000 deaths annually in the U.S. from ovarian cancer provide compelling justification for supporting the research needed to identify improved biomarkers for early stage detection. CA-125 and imaging are the most common approaches for ovarian cancer screening. However, these approaches, either alone or in combination, are not useful for routine screening due to their low specificity and/or sensitivity. For example, serum CA-125 has a sensitivity and specificity of only 69% and 84% respectively for detecting ovarian cancer. Due to the low prevalence of ovarian cancer, a useful screening strategy must have a sensitivity >80% for early-stage disease and specificity >99.6%. Our review of 36 published serum/plasma biomarker panels for ovarian cancer identified 11 panels that each used from 1-6 biomarkers to achieve >90% sensitivity and specificity [Rauniyar et al (2017) Biomarkers Insight 12, 1-12 (PMCID: PMC5462478)]. Since most of these panels share few, if any biomarkers in common, we reason that inclusion of as many of the biomarkers in these, and other previously reported panels, in a single biomarker panel would leverage >40 years of research by providing an opportunity to more rigorously compare the relative efficacies of each of these biomarkers that are detectable by mass spectrometry in the non-fractionated serum that we believe is the best biological sample for these studies and for then choosing the best biomarker panel with the highest possible sensitivity and specificity. In addition to screening, there is also a critical need for improved biomarkers for diagnosis of ovarian cancer. It has been estimated that 5-10% of women in the U.S. will undergo surgery for a suspected ovarian neoplasm during their lifetime and that 13-21% of these patients have ovarian cancer. Since most adnexal masses are benign, it is important to identify preoperatively those patients who are at high risk of ovarian cancer and who will benefit from referral to a gynecologic oncologist to ensure the best possible care. Although ovarian cancer patients operated on by gynecologic oncologists have a 6- to 9-month median survival benefit, only about one third of women with ovarian cancer are referred to a gynecologic oncologist for primary surgery. To meet the need for improved diagnosis of high risk ovarian tumors, the FDA has approved three multivariate index assays. However, even the most recently approved assay, Overa, has a specificity of only 69%. To identify biomarkers that will allow earlier screening and improved diagnosis, a Data Independent Acquisition (DIA) and Parallel Reaction Monitoring (PRM) mass spectrometry workflow was implemented to determine differentially regulated proteins in ovarian cancer versus control sera and to validate these and other literature biomarkers. DIA identified Apolipoprotein A-IV, which had an ovarian cancer/control fold change of 0.52, as the most significantly differentially regulated protein (Rauniyar et al, 2017). PRM analyses of 10 biomarkers with the Targeted Ovarian Cancer Proteome Assay (TOCPA) and Random Forest (RF) analyses validated these results and showed that C-reactive protein, transferrin, and transthyretin are the next best biomarkers. Based on TOCPA analyses, ApoA-IV has a larger fold-change than determined by immunological assays and it is a more reliable biomarker than ApoA-I, which is in the Overa test for detecting ovarian cancer in pelvic masses. All samples were classified correctly using a breakpoint at ~54.4% of the mean level of ApoA-IV in the controls. This research suggests a way to improve the Overa test and it provides a PRM platform and RF approach together with four promising biomarkers to speed the development of a clinical test for diagnosing ovarian cancer.