Irwin Gelman, PhD

Gelman Lab and Members

The role of SSeCKS/AKAP12 in metastasis suppression.
Data from our lab indicate that the SSeCKS/AKAP12 gene product functions as a metastasis suppressor, likely by inhibiting tumor-induced expression of pro-angiogenic factors such as VEGF and HIF-1a (work by a previous postdoc, Bing Su, Ph.D.). We and other have amassed data showing that SSeCKS expression is typically lost in metastatic prostate cancers, in about a third of the cases due to gene deletion, and in two-thirds, due to epigenetic silencing (associated with hypermethylation of promoter CpG islands). In human prostate cancer cases with metastasis, we published that SSeCKS loss correlates with earlier onset of metastatic disease and poorer long-term survival. We also showed that SSeCKS re-expression in prostate cancer cells suppresses neovascularization at metastatic sites although it has little effect on growth of tumors at the primary site. The avascular micrometastases induced by SSeCKS-re-expressing tumor cells exhibit downregulated VEGF expression, and indeed, the forced re-expression of VEGF in these cells could rescue full macrometastasis-forming activity.
We also noted that higher grade human prostate tumors (Gleason sum >7) exhibited not only downregulated SSeCKS expression in prostate epithelial cells but also in adjoining stroma defining the so-called tumor microenvironment (TME). To address whether SSeCKS controls metastatic progression through a role in TME cells, we produced AKAP12-null mice and then tested these for increased metastatic potential when challenged with tumor xenografts or carcinogens. Our data indicate that carcinogens such as DMBA and TPA not only induce much greater skin cancer (squamous cell carcinomas) in AKAP12-null compared to WT hosts, but they also induce increased progression to spontaneous metastasis formation, namely spindle cell carcinomas. AKAP12-null hosts exhibit increased levels of spontaneous peritoneal and lung metastases arising from orthotopic, subcutaneous B16F10 melanomas (work by a previous postdoc, Shin Akakura, Ph.D.) or from i.v. tumor injections.
We also noted that higher grade human prostate tumors (Gleason sum >7) exhibited not only downregulated SSeCKS expression in prostate epithelial cells but also in adjoining stroma defining the so-called tumor microenvironment (TME). To address whether SSeCKS controls metastatic progression through a role in TME cells, we produced AKAP12-null mice and then tested these for increased metastatic potential when challenged with tumor xenografts or carcinogens. Our data indicate that carcinogens such as DMBA and TPA not only induce much greater skin cancer (squamous cell carcinomas) in AKAP12-null compared to WT hosts, but they also induce increased progression to spontaneous metastasis formation, namely spindle cell carcinomas. AKAP12-null hosts exhibit increased levels of spontaneous peritoneal and lung metastases arising from orthotopic, subcutaneous B16F10 melanomas (work by previous postdocs, Shin Akakura, Ph.D. and Masashi Muramatsu, Ph.D.) or from i.v. tumor injections. We published that the increased spontaneous B16F10 metastasis from orthotopic sites (skin) to the peritoneal membrane in AKAP12-null hosts is caused by the increased secretion of CXCL9 and CXCL10 by peritoneal mesenchymal cells, whereas the increased spontaneous metastasis to the lungs in AKAP12-null mice results from the increased adhesion of tumor cells to endothelial E-selectin.

Identification and characterization of novel metastasis suppressors in breast cancer.
We are using a so-called 3rd generation genomic shRNA and CRISPRi libraries (cloned into high-tier lentivirus vectors) to identify genes that i) suppress oncogenic invasiveness in poorly metastatic breast (T47D) or prostate cancer (LNCaP) cells, or that identify drug susceptibility in SYT-SSX fusion oncogene-driven synovial cell sarcoma. For the first project (a collaboration between postdoc Bing Su, PhD and grad student, Henry Withers, infected cell clones were isolated based on their increased invasiveness through Matrigel and increased metastasis from orthotopic sites (mammary glands or prostates) in nude mice. One of the metastasis suppressor genes we identified was FOXO4, which unlike its FOXO1 and FOXO3 family members, suppressed metastatic invasiveness in in vitro and in vivo models. Henry continued these screens after receiving his PhD to identify PDGFRA as a targetable driver for survival and growth of synovial sarcoma tumor cells.

Role of SSeCKS/AKAP12 in prostate cancer progression and metastasis.
Previous data from our lab indicate that AKAP12-null mouse embryo fibroblasts (MEF) exhibit Rb-dependent premature cell senescence marked by polyploidy and multinucleation. Moreover, AKAP12-null prostates are hyperplastic and they express markers of increased senescence such as senescence-associated β-galactosidase, p16ink4a or γ-H2AX. To investigate whether the combined loss of AKAP12 and Rb is sufficient to induce prostate cancer progression, a previous graduate student, Hyun-Kyung Ko, Ph.D. (with the help of Jennifer Peresie and a previous postdoc, Shin Akakura, Ph.D.) crossed our AKAP12-null mice with Pb-Cre;Rbfl/fl mice (or to Pb-Cre; p53fl/fl mice, as a control), which produce prostate-specific (PE) Rb deletions. The Akap12-/-;RbPE-/- mice displayed high grade prostatic intraepithelial neoplasias (HG-PIN) by 7 months of age, but there was little progression to adenocarcinoma. In contrast, all the mice showed evidence of lymph node metastasis, consisting of well-differentiated lesions expressing markers of a transitional basal/luminal prostate cell phenotype (cytokeratin 14, 8, AR and p63). This suggests that rare cells in the HG-PIN lesions are marked by the combined loss of Rb and SSeCKS/Akap12 for metastatic progression. Currently, we are crossing these mice with lineage-tracing transgenic stocks so that we can determine if the metastatic cells arise from either basal (CK14+) or luminal (CK8+) cells in the prostate, and then characterize the epigenetic signature of early metastatic dissemination. We are also crossing this model into a prostate-specific conditional loss of the Smad4 suppressor gene, based on evidence that human prostate cancer metastases with decreased SSeCKS and Rb also suffer from Smad4 loss. Another previous grad student, Karina Miller, PhD, compared Akap12/Rb-null prostate tumors to those formed in transgenic mice lacking prostate expression of Pten and Rb. This is because although both transgenic mouse strains had activated AKT plus Rb loss, the Akap12/Rb-null mice formed high-grade PIN lesions and indolent metastatic disseminations to local lymph nodes, the Pten/Rb-null mice formed aggressive prostate adenocarcinomas and systemic metastases. She showed that whereas the Pten/Rb-null tumors had increased activation of Akt2 plus Smad4 loss, the Akap12/Rb-null tumors had relatively increased Akt1 levels plus no change in Smad4. Together with current grad student, Seamus Degan, M.S., they showed that PTEN-deficient prostate cancer cells are dependent on PI3K-p110β plus AKT2 for metastatic growth and motility parameters, whereas the re-expression of PTEN shifted the dependence to PI3K-p110α plus AKT1. Seamus has performed a phosphoproteomic mass spectrometry analysis to identify potential AKT2-specific substrates that might drive metastatic progression in prostate cancer cells.

Identification of genes enabling breast cancer dormancy in endosteal bone.
Breast cancers often metastasize to the bone, and after long periods of dormancy, can develop into osteolytic lesions. Using 3D growths of stromal and bone cells that recapitulate the endosteal niche, Julie showed that a usually aggressive breast cancer cell line, MDA-MB-231, fails to proliferate. Using MDA-MB-231 cells infected with a genomics shRNA lentivirus library (DECIPHER), two Master’s students, Julie McGrath (who has now earned her PhD from the Arizona State University) and Louis Panzica (who has earned his JD degree from SUNY-Buffalo), several gene knockdowns were identified that facilitate the proliferation of the breast cancer cells in the 3D endosteal niche. Julie and Louis, respectively, characterized how HBP1 and WNT3A suppress the reawakening of dormant breast cancer cells in 3D bone microenvironment cultures. The long-term goal is to identify drugs that sustain HBP1 or WNT3A expression, thereby preventing recurrent breast cancer metastases in the bone, and presumably, increasing patient survival.
Other ongoing projects
Genomic Progression Signatures Regulating Castration-Recurrent Prostate Cancer (CR-CaP) or Barrett’s Esophagus.
There is growing evidence that CR-CaP is driven by the activation of the androgen receptor (AR) in the absence of serum androgen levels. Our group and others have shown that AR can be activated by direct phosphorylation by Src family or Ack1 tyrosine kinases. The current project is to use androgen-dependent (AD)/CR-CaP cell line and tumor pairs to analyze the AR cistrome by AR-ChIP-seq and then to identify differentially-expressed AR-engaged genes by transcriptome-seq. Bioinformatic analyses of these data will allow us to identify a CR-CaP progression signature that might explain the resistance of CR-CaP to conventional treatments, and that might identify novel targetable pathways. In parallel, postdoc Indranil Chattopadyhy, Ph.D. and collaborators (Kandel, Campbell, Liu, Morrison, Buck) performed transcriptome-seq, miRNA-seq, microbiome-seq and other next-gen analyses to identify potential genomic signatures that explain the progression of Barrett’s esophagus metaplasias to dysplasias and eventually, esophageal adenocarcinomas.
The role of FAK/Pyk2 activated of NFκB signaling in tumor-associated macrophages in metastatic progression of breast cancer.
We have studied the role of FAK/Src signaling in the control of cytoskeletal architecture, adhesion and oncogenic motility. To this end, grad student, Sheila Fiegel, Ph.D. sought to identify novel FAK substrates and to elucidate the FAK phosphorylation motif. After produced baculovirus FAK, she performed in situ-FAK kinase assays using protein microarrays containing thousands of GST-protein fusions or tyrosine-containing peptides, and has identified many potentially new FAK-specific substrates. In parallel, she has analyzed how the Src/FAK-induced tyrosine phosphorylation of PRAK/MAPKAP-5 affects adhesion of cancer cells.

Other projects

We are collaborating with labs outside Roswell Park to investigate the role of SSeCKS/AKAP12 in:
1. suppression of aortic restenosis by retinoids
2. exercise-induced cardiac failure
3. learning and memory
4. B-cell development
5. blood brain barrier formation
6. brain injury healing
7. chronic obstructive pulmonary disease
8. aging
9. glomerulonephritis
We are also collaborating to identify genomic signatures of melanoma metastasis.

Lab Members:

Seamus Degan, M.S.- grad student
Subrahmanya Anirudh Kaligotla Venkata- Master's student