Steven D. Rosen, PhD
Professor Emeritus, Department of Anatomy, UCSF
I have a longstanding interest in biological sulfation and its roles in immunology and cancer. The origin of this interest began 40 years ago with our investigation of molecular mechanisms involved in lymphocyte homing to lymph nodes. Our early experiments established that lymphocyte attachment to high endothelial venules (HEVs) in lymph nodes involves a calcium-dependent lectin-like receptor on lymphocytes and carbohydrate-based ligands displayed on HEVs. With Genentech, we cloned the receptor, now known as L-selectin and found a C-type lectin domain, thus confirming its lectin nature. This work inaugurated the elucidation of the selectin family, which is comprised of E- and P-selectin, as well as L-selectin. We went on to define the HEV-ligands for L-selectin and showed that they are mucin-like glycoproteins that express O-glycans with 6-sulfo sialyl Lewis x as the key recognition determinant. We also identified the critical HEV-associated sulfotransferases that elaborate the essential sulfation modifications of the ligands. HEVs with the same biochemical phenotype as in lymph nodes are also found in tertiary lymphoid organs (TLOs) that are associated with tumors. The presence of HEVs in tumor-associated TLOs is correlated with good outcomes for patients with breast cancer and melanoma, probably reflecting immune-protection by the TLOs that are in proximity to the tumors and highlighting the importance of our elucidation of fundamental mechanisms of lymphocyte-HEV interactions.
Our study of the sulfatases known as the SULFs (SULF1 and SULF2) represent an extension of my interest in the regulation of sulfation at the cell surface. We originally cloned the human and mouse versions of these enzymes. We demonstrated that these enzymes are secreted and act as neutral pH glucosamine 6-O-endosulfatases for the GAG chains of heparan sulfate proteoglycans (HSPGs). By removing 6OS groups from glucosamines in GAG chains, the SULF are able to mobilize growth factors/chemokines/morphogens from HS sequestration and modulate multiple signaling pathways in the cells, including canonical Wnt, TGF-beta, and PDGF signaling. The two SULFs are over-expressed in multiple cancers at the mRNA and protein levels. For several of these cancers, SULF2 levels are associated with poor clinical outcomes. Our studies were the first to implicate SULF2 as an oncogenic driver in pancreatic cancer, NSCLC and glioma (with Joanna Phillips and Zena Werb). Subsequent studies have shown pro-oncogenic roles for SULF2 in hepatocellular carcinoma, neuroblastoma, breast cancer, cervical cancer, and colorectal cancer. We have developed function-blocking mAbs against human/mouse SULF2, which will allow further in vitro and in vivo investigation of the oncogenic activities of this enzyme.
Since SULF2 is secreted and is overexpressed in multiple cancers, this protein may have utility as a diagnostic or prognostic biomarker. We have developed a sensitive capture ELISA for SULF2 and have detected the enzyme in human blood and body fluids. Preliminary evidence obtained with the ELISA indicates overexpression of SULF2 in mouth-wash samples of patients with head & neck squamous cell carcinoma. Ongoing collaborations are aimed at determining the biomarker utility of SULF2 in several cancers.
University of California, Berkeley, CA, B.A., 1966, Physics
Cornell University, Ithaca, NY, Ph.D., 1972, Neurobiology
University of California, San Diego, CA, Postdoc, 1972-76, Cell Biology