Elsawa, Sherine F.

Abstract Aberrant activation of GLI transcription factors has been implicated in the pathogenesis of different tumor types including pancreatic ductal adenocarcinoma. However, the mechanistic link with established drivers of this disease remains in part elusive. In this study, using a new genetically engineered mouse model overexpressing constitutively active mouse form of GLI2 and a combination of genome-wide assays, we provide evidence of a novel mechanism underlying the interplay between KRAS, a major driver of pancreatic ductal adenocarcinoma development, and GLI2 to control oncogenic gene expression. These mice, also expressing KrasG12D, show significantly reduced median survival rate and accelerated tumorigenesis compared with the KrasG12D only expressing mice. Analysis of the mechanism using RNA sequencing demonstrate higher levels of GLI2 targets, particularly tumor growth–promoting genes, including Ccnd1, N-Myc, and Bcl2, in KrasG12D mutant cells. Furthermore, chromatin immunoprecipitation sequencing studies showed that in these cells KrasG12D increases the levels of trimethylation of lysine 4 of the histone 3 (H3K4me3) at the promoter of GLI2 targets without affecting significantly the levels of other major active chromatin marks. Importantly, Gli2 knockdown reduces H3K4me3 enrichment and gene expression induced by mutant Kras. In summary, we demonstrate that Gli2 plays a significant role in pancreatic carcinogenesis by acting as a downstream effector of KrasG12D to control gene expression.

The Hedgehog-regulated transcription factors GLI1 and GLI2 play overlapping roles in development and disease; however, the mechanisms underlying their interplay remain elusive. We report for the first time that GLI1 and GLI2 physically and functionally interact in cancer cells. GLI1 and GLI2 were shown to co-immunoprecipitate in PANC1 pancreatic cancer cells and RMS13 rhabdomyosarcoma cells. Mapping analysis demonstrated that the zinc finger domains of both proteins are required for their heteromerization. RNAi knockdown of either GLI1 or GLI2 inhibited expression of many well-characterized GLI target genes (BCL2, MYCN, PTCH2, IL7 and CCND1) in PANC1 cells, whereas PTCH1 expression was only inhibited by GLI1 depletion. qPCR screening of a large set of putative canonical and non-canonical Hedgehog/GLI targets identified further genes (e.g. E2F1, BMP1, CDK2) strongly down-regulated by GLI1 and/or GLI2 depletion in PANC1 cells, and demonstrated that ANO1, AQP1 and SOCS1 are up-regulated by knockdown of either GLI1 or GLI2. Chromatin immunoprecipitation showed that GLI1 and GLI2 occupied the same regions at the BCL2, MYCN and CCND1 promoters. Furthermore, depletion of GLI1 inhibited GLI2 occupancy at these promoters, suggesting that GLI1/GLI2 interaction is required for the recruitment of GLI2 to these sites. Together, these findings indicate that GLI1 and GLI2 co-ordinately regulate the transcription of some genes, and provide mechanistic insight into the roles of GLI proteins in carcinogenesis.

Abstract MicroRNAs (miRNAs) are small noncoding RNAs that are approximately 20–22 nucleotides with critical functions in cell growth, survival, and differentiation. These conserved sequences can regulate expression of multiple genes and are often tissue specific and dysregulated in malignancies. Thus, miRNA profiling has been used to create signatures for many solid tumors. These profiles have been used to classify tumors and to help predict survival and outcome. In the present study, we utilized the DiscovArray miRNA profiling service (Asuragen Services, Austin, TX) which utilizes a custom-manufactured Affymetrix GeneChip® from Ambion that covers miRNAs derived from the Sanger miRBase (http://microrna.sanger.ac.uk/sequences/index.shtml) and over 11,000 predicted miRNAs derived from published reports. The signal processing implemented was a multi-step process involving probe-specific signal detection calls, background estimation and correction, constant variance stabilization and global normalization. For each probe, an estimated background value was subtracted derived from the median signal of a set of G-C-matched anti-genomic controls. Arrays within a specific experiment were normalized together according to variance stabilization method. Detection calls were based on a Wilcoxon rank-sum test of the miRNA probe signal compared to the distribution of signals from GC-content matched anti-genomic probes. For statistical hypothesis testing, a two-sample t-test, with assumption of equal variance, was applied. One-way ANOVA was used for multiple group comparison. Probes were considered to be differentially expressed based on two criteria: a p-value of < 0.001 and glog2 difference > 1. miRNA expression was analyzed in all malignant B lineage cells (CD19+ CD138+) (n=8), malignant B cells alone (CD19+) (n=6) and plasma cells alone (CD138+) (n=3) from Waldenström macroglobulinemia (WM) patients. The expression was compared to malignant CD19+ B cells from chronic lymphocytic leukemia (CLL) patients (n=5), malignant plasma cells (CD138+) from multiple myeloma (MM) patients (n=5) and to B lineage cells (CD19+ CD138+) (n=4), CD19+ B-lymphocytes (n=3) and CD138+ plasma cells (n=6) from healthy donors. Data analysis based on a total of approximately 11,000 miRNAs analyzed shows that CD19+ CD138+ cells (double sorting) from WM patients did not cluster as a unique group. Some samples had a pattern similar to CLL, some similar to MM and others similar to CD19+ CD138+ cells from healthy controls. This lack of clear signature was observed by others in gene expression profiling and CGH arrays. We therefore hypothesized this lack of clustering was due to the lymphoplasmacytic nature of WM cells and therefore we analyzed B cells (CD19+) and plasma cells (CD138+) separately. miRNA expression in B cells (CD19+) identifies a signature in normal B cells that is absent in both WM B cells (CD19+) and CLL cells. There is also a set of miRNAs that are absent in normal B cells that are expressed in WM B cells and CLL. In addition, WM B cells had a unique miRNA signature that is unique compared to CLL and normal B cells. An additional set of miRNAs were expressed and clustered only in CLL patients. Similar to B cells, plasma cell (CD138+) analysis in WM, MM and healthy donors shows a clustering pattern that identifies normal plasma cells from MM plasma cells. WM plasma cells had a miRNA signature that is unique only to WM patients, however, a subset of miRNAs shared an expression pattern with MM plasma cells. While miRNAs can target multiple genes, some of the genes that are targets of the miRNAs identified in this analysis include XBP-1, Blimp-1, IRF-4, Bcl-6 and TACI. These target genes are known to be important in B cell and plasma cell development. In summary, we have analyzed miRNA expression in malignant B cells (CD19+) and malignant plasma cells (CD138+) from WM patients and compared their expression pattern to their normal counterpart as well as malignant counterpart in CLL B cells and MM plasma cells. Our analysis shows that WM B cells have a miRNA signature unique to WM only and one that is shared by CLL cells. Similarly, WM plasma cells have a unique miRNA signature but also has some miRNAs that are shared by malignant plasma cells in MM. These miRNAs target genes involved in B cell differentiation. Analysis of the functional roles of these miRNAs will and their regulation will further our understanding of the regulation of B cells development in normal and malignant conditions.

Abstract Waldenström macroglobulinemia (WM) is a B-cell malignancy characterized by the aberrant production of a monoclonal IgM protein that may lead to hyperviscosity. Although this is a major factor causing significant morbidity in patients, little is known about the mechanisms that regulate monoclonal protein synthesis. Cytokines are protein mediators that are known to be involved in many biological processes, and can profoundly affect tumor cells and the tumor microenvironment. Many cytokines have been shown to have potent therapeutic efficacy in preclinical cancer models; however, the role of cytokine networks in WM is not fully understood and few studies have described the precise functional roles of cytokines in WM. To address this issue we performed a multiplex ELISA analysis to test cytokine levels in sera from patients with WM. We found that Rantes/CCL5 is significantly elevated in WM patients and correlates with disease activity. Elevated expression of RANTES in the serum was confirmed by ELISA and was also detected in the bone marrow of WM patients by ELISA and immunohistochemistry. RANTES expression serves as a marker for recruitment of immune cells and is associated with a wide range of immune-mediated diseases. However, the impact of RANTES in WM is not known. We analyzed the expression of receptors for Rantes by flow cytometry and found that WM B cells and stromal cells express CCR1 and CCR3, but not CCR5. Using a standard chemotaxis assay, we determined that Rantes had no effect on B cell or stromal cell recruitment. Rantes also had no effect on B cell or stromal cell survival, however it did promote stromal cell proliferation (p<0.04). The interaction between Rantes and IL-6 has been described in an autoimmune disease model. Since stromal cells secrete significantly more IL-6 than WM B cells, we treated stromal cells with Rantes for 24 hr and found that Rantes increases IL-6 secretion (p<0.03). To characterize the mechanism of Rantes-mediated IL-6 secretion, we transfected stromal cells with an IL-6 promoter construct and treated with Rantes for 12 hr and found a significant increase in IL-6 promoter activity (p<0.0162) indicating Rantes can regulate IL-6 transcription. Bioinformatics analysis of the IL-6 promoter indicates the presence of multiple candidate binding sites for transcription factors that have been previously shown to play a role in the biology of B cells, including NFkB, AP1, and GLI. Co-transfection of stromal cells with an IL-6 reporter construct and a plasmid expressing GLI1, GLI2 or GLI3 demonstrates that GLI2 and GLI3 proteins can regulate the IL-6 promoter. We then transfected stromal cells with a reporter construct containing 8X-GLI binding sites and demonstrate that Rantes can regulate GLI transcription further supporting a role for the interaction between Rantes and IL-6 through GLI transcription factors. IL-6 rich tumor microenvironment supports malignant cells. Elevated IL-6 levels have no effect on survival of BCWM.1 cells or CD19+ 138+ cells from WM patients, but leads to upregulation of IgM secretion by BCWM.1 cells and CD19+ CD138+ cells from WM patients, and increases their proliferation (p<0.0039). IL-6 activates Erk1/2 and Jak/ Stat in WM and stimulation of the IgM secreting cells BCWM.1 with IL-6 in the presence of PD98059 MAPK inhibitor had no effect on IgM secretion. However, stimulation of BCWM.1 cells with IL-6 in the presence of JakI inhibitor abolished the IL-6 mediated IgM secretion suggesting IL-6 mediated increase in IgM secretion occurs through Jak/ Stat signaling pathway. Analysis of Rantes levels in other B cell malignancies including follicular lymphoma, chronic lymphocytic leukemia, monoclonal gammopathy of undetermined significance and multiple myeloma indicates that Rantes is elevated in other B cell lymphoproliferative disorders and suggest Rantes may play a similar role in other malignancies. In summary, our data identifies a novel Rantes-GLI-IL-6 interplay in the stromal microenvironment that promotes IgM production by malignant B cells. This therefore provides multiple new potential therapeutic avenues, targeting both malignant cells and the microenvironment to control malignant cell growth, and immunoglobulin secretion in WM and Ig-mediated diseases.

Abstract The tumor microenvironment in B-cell non-Hodgkin’s lymphoma (NHL) plays an important role in supporting the growth and survival of malignant lymphocytes. In contrast, the extent to which this is true in T-cell NHL is less clear. While professional monocytederived antigen-presenting cells (MDC) are required for T-cell activation and survival, a growing body of evidence demonstrates that MDC may promote the survival of malignant T cells. Immunohistochemical staining of skin biopsies from cutaneous T-cell lymphoma (CTCL) patients (n=11) demonstrated a rich infiltrate of CD68+ macrophages and CD1a+ dendritic cells (DC). As these cells are derived from CD14+ monocytes, we examined the role of monocytes in promoting survival in malignant T cells obtained from CTCL patients (n=9). Initially, peripheral blood mononuclear cells (PBMC) were either depleted of CD14+ monocytes (>85% depleted) or mock-depleted prior to being cultured. Viability of malignant CD4+CD7− T cells was determined by Annexin V/7-AAD staining. While 90.9% (S.D. +/− 0.4%) of mock-depleted malignant T cells were viable, only 73.9% (S.D. +/− 1.3%) of monocyte-depleted cells remained viable, suggesting that monocytes may promote the survival of malignant T cells. In a similar fashion, HuT-78 cells (Sézary Syndrome cell line) were cocultured with CD14+ monocytes from normal donors. Compared with HuT-78 cells alone, the presence of CD14+ monocytes resulted in a 2.7 fold increase in thymidine incorporation (n=5; S.D. +/− 0.6). This is largely due to cytokine production, as the same experiment performed with a transwell insert resulted in a 2.1 fold increase in HuT-78 thymidine incorporation (n=5; S.D. +/− 0.4). Similar results were obtained using additional Sézary Syndrome (i.e. SeAx) and Mycosis Fungoides (i.e. MyLa) cell lines. Monocytes obtained from normal donors were subsequently cultured and cell-free conditioned media was obtained 48 hours later. Media supplemented with monocyte-conditioned media (monocyte-CM) prevented the induction of apoptosis in CTCL cell lines as measured by thymidine incorporation, trypan blue exclusion and Annexin V/propidium iodide staining. This affect was monocyte-specific, as neither T-cell nor B-cell conditioned media were observed to promote survival of CTCL cell lines. Furthermore, monocyte-CM had no affect on resting or anti-CD3 activated normal T cells. We next cultured malignant T cells obtained from PBMC (n=6) or skin biopsies (n=3) from CTCL patients in the presence or absence of monocyte-CM and assessed cell viability by Annexin V staining. 32.1% (S.D. +/− 14.7%) and 60.5% (S.D. +/− 21.5%) of malignant T cells isolated from PBMC remained viable in the absence or presence of monocyte-CM, respectively; 13.0% (S.D. +/− 10.8%) of malignant T cells isolated from skin biopsies remained viable, compared with 26.5% (S.D. +/− 8.2%) of cells grown in the presence of monocyte-CM (p=.03). Given the prevalence of MDC within CTCL-involved skin and their potential role in promoting the survival of malignant T cells, one may predict that malignant T cells may recruit monocytes in a chemokinedependent fashion. Therefore, cell-free conditioned media from CTCL cell lines (HuT- 78, SeAx, MyLa) was collected and used to promote monocyte chemotaxis in a standard chemotaxis assay. Only neutralization of CCL5 (RANTES) was capable of abrogating monocyte chemotaxis, demonstrating that the production of CCL5 by malignant T cells promotes monocyte migration. Therefore, we examined serum CCL5 levels by ELISA from both normal donors (n=24) and CTCL patients (n=23) and found an approximately 3-fold elevation in serum CCL5 in CTCL patients (normal donors: mean CCL5 10.0 ng/mL; 95% CI 7.1–13.0; CTCL patients: mean 29.2 ng/mL; 95% CI 20.6–37.8; p<0.0001). Serial serum samples were available for 6 patients, 4 of whom had progressive disease. Serum CCL5 increased from a mean of 17.9 ng/mL before progression (S.D. +/− 7.0) to 44.8 ng/mL (S.D. +/− 20.0) after progression (p=.04). In contrast, serum CCL5 decreased by at least 50% in both patients with responsive disease. Our findings demonstrate that monocytes, recruited by malignant T cells in a CCL5-dependent fashion, promote the survival of lymphoma cells in CTCL, thus raising the possibility that targeting tumor-associated monocytes and MDC may represent a novel therapeutic strategy in CTCL.

Abstract Waldenström macroglobulinemia (WM) is a monoclonal B cell disorder characterized by a circulating monoclonal IgM protein that may lead to serum hyperviscosity in association with an infiltration of lymphoplasmacytic cells into the bone marrow. Although proinflammatory and chemotactic cytokines can profoundly affect tumor cells and the tumor microenvironment, and many cytokines have been shown to have potent therapeutic efficacy in preclinical cancer models, the role of cytokine networks in WM is not fully understood. In this study, we used a high-throughput xMAP multiplex immunobead assay technology (Luminex Corp., Austin, TX) to simultaneously test 30 cytokines, chemokines, angiogenic factors as well as growth factors and soluble receptors in the sera of WM patients and compared them with other B cell malignancies including IgM monoclonal gammopathy of undetermined significance (MGUS), follicular lymphoma, chronic lymphocytic leukemia (CLL) as well as healthy controls. Using a Mann-Whitney U test to analyze the differences between the groups, 15 of the 30 cytokines tested had significantly different levels in WM compared to healthy controls. Of those 15 cytokines, 11 were elevated in WM patients and 4 were decreased. Cytokines were grouped into 3 groups; those with < 2-fold difference, 2–8 fold difference and those having > 8-fold difference in their cytokine levels compared to healthy donors. There was a greater than 8-fold increase in the serum levels of Rantes, G-CSF and IL-2R (p<0.0001) in WM patients. Furthermore, 3 cytokines had between 2–8-fold increase in WM patients including IL-4 (p<0.0001), IL-6 (p<0.0019) and IP-10 (p<0.0006). Five cytokines had statistically elevated levels in WM patients compared to healthy controls, however the fold increase was < 2 including HGF (p<0.0185), IL-10 (p<0.0002), MIP-1α (P<0.0484), IL-2 (P<0.0130) and IL-12 (P<0.0155). Of the cytokines that had significantly lower levels in the sera of WM patients, IL-8 (p<0.0001) and EGF (p<0.0001) were > 8-fold decreased, MCP-1 (p<0.0001) was 2–8 fold lower and Eotaxin (p<0.0004) was < 2-fold lower in WM patients. All of the cytokines that had the greatest fold difference (> 8-fold) in WM patients compared to healthy donors also differed significantly from the MGUS patients. Rantes, G-CSF, IL-2R and EGF had significantly different levels compared to other B cell malignancies. We tested for a correlation between the cytokines that had > 2-fold difference between the WM group and control group with clinical features of the disease and found the cytokines IL-6 and IL-2R had a significant correlation with β2-microglobulin levels (p<0.01). We analyzed cytokine levels in the bone marrow plasma of the same patients and found that high levels of IL-2R in the bone marrow microenvironment significantly correlated with anemia and elevated serum β2-microglobulin (p<0.01). In conclusion, we have simultaneously analyzed sera from WM patients for 30 cytokines and found the most significantly elevated cytokines are Rantes, G-CSF and IL-2R and the most significantly downregulated cytokines are IL-8 and EGF. Furthermore, we found that elevated serum levels of IL-6 and IL-2R correlated with β2-microglobulin levels, a measure of disease activity. Further analysis of the biological role of these cytokines in WM may offer insight into disease pathogenesis and provide a basis for novel targeted therapies.

Abstract Waldenström macroglobulinemia (WM) is a B-cell malignancy that is characterized by the production of a monoclonal IgM protein, a lymphoplasmacytic infiltrate in the bone marrow, and associated symptoms including anemia, lymphadenopathy and hyperviscosity. The aberrant production of a monoclonal IgM in the serum is a major factor causing significant morbidity in patients with this disease, yet little is known about the mechanisms that regulate monoclonal protein synthesis. While recent gene array studies and serum analysis have shown that IL-6 is elevated in WM patients suggesting an important role for this cytokine in this disease, the precise role played by IL-6 in WM is unknown. Using a multiplex ELISA approach to screen sera from WM patients, we confirmed that IL-6 was significantly elevated (p<0.0019) in patients (n=20) compared to controls (n=20). Serum levels of IL-6 in WM patients correlated with elevated levels of β2-microglobulin (p<0.0019). Additionally, we also found that serum levels of CCL5 (Rantes) were significantly elevated in WM patients (p<0.0001). CCL5 has been shown to regulate IL-6 secretion, and we therefore wanted to determine if CCL5 influenced IL-6 expression in WM and what the subsequent consequence of IL-6 stimulation was on WM cells. To define the source of IL-6 in the tumor microenvironment, we used stromal cells from the bone marrow of healthy donors, malignant cells from patients with WM, and the BCWM.1 WM cell line, and tested their ability to secrete IL-6 by ELISA. All cell types secreted IL-6, with stromal cells secreting the most. We then tested the ability of CCL5 to induce IL-6 secretion by WM and stromal cells. CCL5 significantly increased IL-6 secretion by stromal cells (p<0.03) and also increased IL-6 secretion by fresh CD19+ CD138+ cells from WM patients (p<0.02). Using fresh patient WM cells and the BCWM.1 WM cell line as a model, we then determined the effect of IL-6 on growth of WM cells. We found that IL-6 had a modest effect (mean=20% increase, range=5–41%) on cell proliferation (p<0.0039) but had no effect on cell viability. In contrast, when we addressed the role of IL-6 on IgM secretion, we found that IL-6 increased IgM secretion by BCWM.1 cells in a dose-dependent manner. The IL-6 mediated increase in IgM secretion was abolished in the presence of neutralizing antibodies to IL-6. When we analyzed the downstream signaling events activated by IL-6 in WM cells we found that stimulation of BCWM.1 cells, which express the IL-6R, resulted in phosphorylation of Stat1, Stat3 and Erk1/2, but not Akt. Using a mitogen activated protein kinase (MAPK) inhibitor, we could inhibit the IL-6-mediated phosphorylation of Erk1/2. Similarly, using a JAK1 Inhibitor, we could inhibit IL-6 mediated signaling through Stat1 and Stat3. In summary, we have clearly shown that IL-6 significantly upregulates IgM secretion by WM cells and increases their proliferation. We have also demonstrated the ability of both the malignant cells and the stromal cells to secrete IL-6, and that this secretion is regulated in part by CCL5. We have found that WM cells express IL-6R, and that IL-6 induced signaling is through both the MAPK and Jak/Stat pathways. Therapies targeting IL-6 secretion or the IL-6 signaling pathways may therefore provide clinical benefit to patients with WM; not just by inhibiting the malignant cells but by down regulating the production of the monoclonal protein.

Waldenström macroglobulinemia (WM) is a low-grade B-cell lymphoma associated with the accumulation of lymphoplasmacytic cells in the bone marrow (BM) and a high level of monoclonal IgM protein in the serum. Genome-wide studies revealed genomic alternations in WM, with the most prevalent being a point mutation in the myeloid differentiation primary response gene 88 (MyD88) resulting in an amino acid substitution (L265P) that is present in over 90% of WM patients. MyD88 is an effector of Toll-like receptor (TLR) signaling pathway that mediates inflammatory cytokine expression and secretion upon activation. Here, we found that activation of the TLR-MyD88 pathway in WM cells carrying the MyD88 L265P somatic mutation increases the expression of IL-6 and CCL2, two cytokines known to participate in the pathobiology of several neoplasms including B cell malignancies. Furthermore, IL-6 has been shown to modulate IgM secretion and malignant B cell growth in WM. We demonstrate that an active ERK1/2 pathway, which is elevated in WM patient samples, is required to maintain IL-6 and CCL2 expression in BCWM.1 (p=0.0339) and RPCI-WM1 (p=0.0005) cells. Further characterization identified a change in the chromatin landscape in response to TLR-MyD88 activation in WM cells. We found increased levels of trimethylation of histone 3 lysine 4 (H3K4me3) at the promoters of the inflammatory cytokines IL-6 and CCL2 by chromatin immunoprecipitation (ChIP) assay followed by qPCR (ChIP-qPCR). The H3K4me3 modification is catalyzed by six related homologs of the yeast histone methyltransferase (HMT) family. Further analysis identified the Mixed-lineage leukemia 1 (MLL1) as the enzyme bound to these promoters in response to TLR-MyD88 stimulation of WM cells. Analysis of WM cell lines and primary WM patient cells showed that MLL1 and its binding partner, menin, are expressed at significantly higher levels in CD19+CD138+cells from WM patients compared to CD19+B cells from peripheral blood of healthy donors (p<0.001). We also found an increase in H3K4me3 deposition on IL-6 and CCL2 promoters during early (1-3 hr) and late (12-24 hr) kinetics following TLR-MyD88 stimulation. This also coincided with increased deposition of MLL1 on these cytokine promoters by ChIP-qPCR. Disruption of menin-MLL1 using the MI-2 or MI-503 menin-MLL1 inhibitors significantly reduced IL-6 and CCL2 expression in WM cell lines (BCWM.1 p=0.039; RPCI-WM1 p=0.0041). This also results in significantly reduced IgM expression (p<0.0001) and secretion at inhibitor dose (5 μM) that has no effect on cell proliferation. Finally, MLL1 knockdown using RNAi significantly reduces IgM expression in BCWM.1 (p<0.0001), MWCL (p=0.0016) and RPCI-WM1 (p<0.0001) and IgM secretion in BCWM.1 (p<0.0001) and MWCL-1 (p<0.0001). Taken together, these results identify a novel role for menin-MLL1 in regulating inflammatory cytokines and IgM expression and secretion in WM and provide the rationale for targeting these molecules in WM patients. Disclosures Ansell: Mayo Clinic Rochester: Employment; Mayo Clinic Rochester: Employment; Regeneron: Research Funding; Seattle Genetics: Research Funding; Trillium: Research Funding; Seattle Genetics: Research Funding; Regeneron: Research Funding; Trillium: Research Funding; Bristol-Myers Squibb: Research Funding; Mayo Clinic Rochester: Employment; Trillium: Research Funding; Affimed: Research Funding; Affimed: Research Funding; Regeneron: Research Funding; Trillium: Research Funding; Mayo Clinic Rochester: Employment; Bristol-Myers Squibb: Research Funding; Mayo Clinic Rochester: Employment; LAM Therapeutics: Research Funding; Bristol-Myers Squibb: Research Funding; Seattle Genetics: Research Funding; LAM Therapeutics: Research Funding; LAM Therapeutics: Research Funding.

Abstract Monocytes play crucial supporting roles in a vast variety of malignancies ranging from solid tumors to hematological neoplasms. Monocytes contain repertoires of immune sensors on their cell surface mediating the inflammatory response in the tumor microenvironment. Of particular interest are Toll-like receptors (TLR), which have well-characterized and documented roles, in not only the innate immune response, but also in carcinogenesis. Our goal was to expand on these findings and further the understanding of the mechanisms mediating inflammation in the tumor microenvironment, especially, identification and characterization of pathways regulating pro-inflammatory cytokine genes in monocytes. Thus, we began our studies by identifying genes differentially regulated in monocyte cell lines treated with lipopolysaccharide (LPS), a known activator of TLR4 and subsequently the inflammatory response. After 3 hours, we found an increase in the expression of GLI3, an effector of the Hedgehog (HH) signaling pathway. Interestingly, treatment of the monocyte cell lines MonoMac6 (MM6) and THP-1 with the HH signaling inhibitor cyclopamine (10 μM) followed by LPS stimulation resulted in up-regulation of GLI3, suggesting that the regulation of GLI3 by TLR4 occurs in a HH-independent manner. TLR4 mediates signaling through MyD88-dependent or TRIF-dependent mechanisms. This is characterized by recruitment of TIRAP or TRAM to the Toll/Interleukin-1 Receptor (TIR) domain of the cytoplasmic portion of TLR4 and subsequent signaling through MyD88 or TRIF, respectively. To further characterize the mechanism of TLR4-GLI3 regulation, we examined the role of TLR signaling adaptors (TIRAP, MyD88, TRAM) in this process. Dominant negative (dn) forms of MyD88, TIRAP and TRAM were transfected into MM6 and THP-1 cells followed by LPS treatment for 3 hours. In response to LPS stimulation, GLI3 was up-regulated in samples transfected with dnMyD88 and dnTIRAP, while GLI3 expression was near basal levels in samples transfected with dnTRAM, the adaptor component for TRIF-mediated TLR4 signaling. This suggests TLR4 mediated GLI3 expression occurs in a TRIF-dependent manner. It has been previously shown that TLR4 signaling from endosomes can cause late NF-κβ and MAPK signaling in addition to classical TRIF-mediated IRF signaling. To further define the mechanism of regulation of GLI3, we utilized pharmacological inhibitors to target IRF, NF-κβ and MAPK signaling components. We found that inhibition of IRF and NF-κβ signaling had no effect on GLI3 up-regulation by LPS. However, p38 inhibition resulted in abolished GLI3 induction in response to LPS stimulation suggesting the p38 mediates TLR4-GLI3 regulation. Inhibition of Jnk did not have an effect on LPS-induced GLI3 expression, suggesting that TLR4-TRIF-p38 axis mediates increased GLI3 expression. Further analysis indicates that LPS stimulation increases the expression of pro-inflammatory chemokines CCL2 and CCL7. Interestingly, overexpression of GLI3 increases the expression of these chemokines suggesting GLI3 may be required for chemokine expression in response to LPS. Taken together, these results identify a novel interaction between TLR4 signaling and GLI3 transcription factor and show a novel signaling pathway that can modulate the expression of GLI proteins independent of HH signaling. Targeted inhibition of this signaling crosstalk may provide therapeutic efficacy for patients with inflammatory and/or HH-related malignant disorders. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 613 Uncontrolled immunoglobulin (Ig) secretion by malignant cells is a characteristic feature of lymphoplasmacytic malignancies and associates with significant morbidity. IL6 has been shown to promote Ig secretion by malignant B cells, however, the mechanisms mediating this cellular event remain elusive. Here, we identified a CCL5 induced pathway that upregulates IL6 secretion in B cell/plasma cell malignancies including Waldenström macroglobulinemia (WM). Characterization of this mechanism revealed that activation of the CCL5 pathway increases IL6 expression and promoter activity in stromal cells (HS-5, Saka and primary stromal cells from WM patients). Conversely, stable RNAi knockdown of the CCL5 receptor CCR3 in stromal cells led to significantly reduced IL6 promoter activity and expression. Further analysis using a combination of bioinformatics analysis, genetic analysis and chromatin immunoprecipitation assays (ChIP) identified GLI2 as the mediator of this CCL5-IL6 functional interaction. CCL5 induces GLI2 expression via activation of the PI3K/AKT pathway. Constitutively active PI3K and AKT mutants increase expression of GLI2 through the canonical NFkB pathway. ChIP assay showed p65 binds to the GLI2 promoter and increases GLI2 levels in stromal cells. Dominant negative molecules antagonizing these pathways effectively blocked CCR3 induction of GLI2 expression. Interestingly, this CCL5-induced pathway was independent of the Hedgehog signaling. Blockade of this cascade using pharmacological or genetic tools did not affect CCL5 activation of GLI2 in stromal cells. To validate the biological significance of this newly identified CCL5-GLI2-IL6 signaling axis in the regulation of IgM secretion, we used an in vitro coculture system and an in vivo mouse model. Coculture of primary WM stromal cells, HS-5 or Saka cells with primary WM malignant cells or BCWM.1, an IgM producing cell line, led to a significant increase in IL6 and IgM secretion. Similar results were obtained in vivo using athymic nu/nu mice subcutaneously injected with BCWM.1 and HS-5 cells. We then confirmed the requirement of GLI2 in vitro and in vivo. We infected HS-5 cells with a lentivirus containing shRNA targeting GLI2 or a scramble control and cocultured them with BCWM.1 cells. Knockdown of GLI2 decreased IL6 and IgM secretion in coculture. This was confirmed in vivo by infecting HS-5 cells or primary WMsc with lentivirus targeting GLI2 or scrambled shRNA and coinjecting with BCWM.1 cells into athymic nu/nu mice. Mice injected with cells lacking GLI2 had lower IgM levels in the serum and tumors compared with mice injected with HS-5 cells with an intact GLI2, further supporting the hypothesis that CCL5 requires an intact GLI2 to increase IL6 and consequently IgM. Finally, we confirmed the importance of this phenomenon in other B cell/plasma cell malignancies such as MGUS and MM. Primary stromal cultures from these neoplasms were infected with lentivirus targeting GLI2 or scrambled shRNA and cocultured with BCWM.1 cells. Similar to WM, there was a significant decrease in IL6 and IgM secretion in the absence of GLI2 suggesting the role of this molecule is conserved in other B cell/plasma cell malignancies. In summary, our data identifies a novel role for GLI2, an effector of the Hedgehog signaling pathway, in mediating the interaction between CCL5 and IL6 in the tumor microenvironment that modulates Ig secretion by malignant B cells. Therefore therapies targeting this signaling axis in the tumor microenvironment might provide efficacy in patients with Ig-mediated diseases. Disclosures: No relevant conflicts of interest to declare.