Abstract 1362

Epigenetic or transcriptional silencing of important tumor suppressors has been described to contribute to cell survival and tumorigenesis in chronic lymphocytic leukemia (CLL). We investigated gene silencing in CLL using gene expression microarray analysis, and found that over 2000 genes are repressed more than 2-fold in CLL B cells compared to normal B cells, including genes involved in cell signaling and survival (Jun/Fos signaling, ATF family, cyclin dependent kinases and suppressors of cytokine signaling). In particular, the suppressor of cytokine signaling protein 3 (SOCS3) was decreased 60-fold in CLL B cells compared to peripheral blood B cells from normal donors. Despite this profound effect, few therapeutic approaches have focused on reversing this gene silencing in CLL. NF-κB has been shown to transcriptionally silence gene expression in several types of cancer, and our group has previously identified a similar role for this transcription factor in CLL. In addition, we have previously shown that the Hsp90 inhibitor 17-DMAG targets NF-κB signaling to induce apoptosis in CLL through transcriptional down-regulation of oncogenes such as MCL1 and BCL2. However, our microarray analysis revealed that treatment with 17-DMAG also leads to the re-expression of nearly 200 genes silenced in CLL compared to normal B cells. One of the genes significantly silenced in CLL and re-expressed by 17-DMAG is SOCS3. This increase in SOCS3 was evident as early as 8 hours following treatment with 17-DMAG, and peaking at 16–24 hours just prior to 17-DMAG induced cell death (up-regulated 5.6 fold at 8 hours, 59.8 fold at 16 hours, and 25.7 fold at 24 hours). The magnitude of induction in CLL cells was significantly greater than SOCS3 induction in peripheral blood B cells from normal donors, which correlates with a decreased apoptotic response of normal B cells to 17-DMAG (25.7 fold up-regulated in CLL versus 3.5 fold in normal B cells). While little is known about its regulation or functional impact in CLL, SOCS3 has been shown to be methylated in solid tumors as well as myeloid leukemia, leading to aberrant cytokine production and cell survival. While it is known that promoter hypermethylation and subsequent gene silencing contributes to CLL disease progression, we found that there was no significant methylation of the SOCS3 promoter in CLL compared to normal B lymphocytes, indicating an alternative mechanism of SOCS3 silencing in CLL. In order to further characterize the downstream effect of SOCS3 regulation, we investigated the pathways known to be regulated by this protein, specifically IL-6 and CXCR4 signaling. We found that 17-DMAG prevents phosphorylation of STAT3 induced by IL-6 stimulation, which leads to decreased production of pro-survival cytokines including negative feedback by decreasing IL-6 itself. While STAT3 is a known Hsp90 client protein, the effect on phosphorylation of STAT3 was evident before a decrease in the total protein was observed, indicating a distinct effect on the signaling pathway independent of Hsp90's role as a protein chaperone. SOCS3 has also been shown to prevent phosphorylation of focal adhesion kinase (FAK) and therefore block both integrin and CXCR4 signaling pathways. We found that 17-DMAG prevents constitutive phosphorylation of FAK in primary CLL cells, and subsequently reduces AKT phosphorylation following recombinant SDF-1 stimulation. In order to determine if 17-DMAG inhibits migration of CLL cells towards recombinant SDF-1 as well as the direct signaling through the CXCR4 receptor, we performed transwell migration assays and found that 17-DMAG significantly inhibits migration towards both recombinant SDF-1 and CXCL13 (migration towards CXCL12, 12.4% with Vehicle vs. 8.9% with 17-DMAG, p=0.0061, towards CXCL13, 12.4% with Vehicle vs. 6.1% with 17-DMAG, p<0.0001). Similar results were obtained by over-expression of SOCS3 in a CLL B cell line, suggesting that 17-DMAG inhibits migration through an increase in SOCS3. Based on these results, we suggest that 17-DMAG reverses gene silencing in CLL, and through re-expression SOCS3 inhibits the migration and signaling associated with SDF-1/CXCR4, an important factor in the tumor microenvironment that contributes to CLL cell survival. Therefore Hsp90 inhibitors represent a novel approach to target transcriptional silencing in CLL and other B cell lymphoproliferative disorders.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.