The impact of oxygen exposure on clinical biomarkers - an underrecognized source of pre-analytic variability

When preclinical research fails to replicate human biology, scientific progress stalls, clinical trials falter, and patients continue to suffer. While many factors contribute to these failures, lack of attention to pre-analytic variability is a seminal issue. We recently reported that even short-term exposure to ambient air is sufficient to

trigger signaling changes in tumor and non-malignant biospecimens. Those changes in turn alter their biology and responsiveness to targeted therapies. Oxygen (O2) tension in various organs ranges from 3-9%, whereas O2 in ambient air is 21%. Using transgenic models of breast cancer and ascitic fluid from patients with ovarian

cancer, we demonstrated differential expression of various signaling molecules including YAP1, NRF2/KEAP1 and WNT/b-CATENIN when biospecimens are collected and processed under physioxia (3-5% O2) compared to the same biospecimen collected and processed under ambient air. We also observed that basal and drug-

induced signaling networks that determine cellular response to targeted therapies are impacted by exposure to ambient air. Key signaling molecules affected include pEGFR(Y1068), pPDGFRb(Y751), pAKT(S473), pERK(T202/Y204), DNMT3A, TET2, and BRD4. Based on these results, we hypothesize that exposure of

biospecimens to ambient air during collection and processing incorrectly estimates the levels of many signaling molecules that are used as biomarkers to define tumor characteristics and to determine the potential benefit of targeted therapies. The effects of ambient air exposure during collection and processing extends to

biospecimens from normal tissues, impacting characterization of the cell-of-origin of their associated cancers. Experiments designed in three aims will test these hypotheses. In the first aim, we will collect breast tumor biopsies, malignant ascites, and pleural effusions under 1, 3 and 5% O2 reflecting variable O2 tension in tumors,

then perform comparative analysis with and without subsequent exposure to ambient air for signaling molecules listed above. Since our preliminary studies have shown the effects of ambient air exposure on the levels of DNMT3A and TET2, two major determinants of DNA methylation, aim 2 will investigate the effects of ambient

air exposure on the DNA methylome and transcriptome. In the third aim, using the unique institutional resource of Komen Tissue Bank, the world’s only repository of normal breast tissue donated by healthy women to support research, we will characterize the effects of ambient air exposure on the levels of biomarkers from Aim 1 and

transcriptome. Successful completion of these studies will change how samples are collected and processed, ensuring that O2 exposure mimics the O2 concentration of the organ to limit the introduction of ambient air O2- induced changes. In addition, this work will force the development of new collection devices that limit the

exposure of biospecimens to ambient air. With respect to this specific FOA, this proposal addresses three major topics: improving clinical biomarker assays for treatment decisions, addressing limited translation of pre-clinical findings into clinic, and evidence-based standardization of biospecimen handling procedures.