STAT3 Activates Anti-Apoptotic Genes But When In Access Induces Apoptosis Of CLL Cells

In chronic lymphocytic leukemia (CLL) cells, signal transducer and activator of transcription 3 (STAT3) is constitutively phosphorylated on serine 727 residues, and phosphoserine STAT3 induces the transcription and proliferation of anti-apoptosis genes including c-Myc, BCL2, Bcl-XL, and Mcl-1 (Hazan-Halevy, 2010). Therefore, we hypothesized that the CLL cells of patients with high peripheral blood absolute lymphocyte counts (ALCs) have a high level of STAT3 protein expression and a low rate of spontaneous apoptosis. Using Western blotting and flow cytometry, we quantitated STAT3 levels and apoptosis rates in CLL cells from 64 patients with high (N=32) and low (N=32) ALCs. As expected, the levels of STAT3 expression were significantly higher in cells from patients with high ALCs (145,000 ± 49,738/µL) than in cells from patients with low ALCs (12,800 ± 4,654/µL). However, contrary to our hypothesis, Annexin V/PI staining revealed that the levels of procaspase-3, its activated (cleaved) form caspase-3, and cleaved PARP as well as the apoptosis rates of cells from patients with high ALCs were significantly higher than those in cells from patients with low ALCs. These findings suggest that cells from patients with high ALCs are prone to spontaneous apoptosis. An RNA microarray analysis revealed that the levels of apoptotic pathway genes were upregulated in cells from patients with high ALC (P = 0.002), and Reverse-transcriptaction PCR of Caspase-3, Calpain 9, MAPK 8, KRAS, PLCγ-2, and PKC validated the array data. Intrigued by these findings, we sought to determine whether high levels of intracellular STAT3 induce apoptosis. Indeed, overexpression of STAT3 in interleukin-6–stimulated MM1 cells upregulated caspase-3 and caspase-3 protein levels and induced apoptosis. Similarly, in CLL cells levels of caspase3 and procaspase3 remained stable across a wide range of STAT3 levels, but when STAT3 reached a threshold, level of caspase3 and procaspase3 markedly increased.

Because sequence analysis revealed that the caspase-3 promoter harbors γ-interferon activation sequence-like elements, we cloned the caspase-3 promoter in MM1 cells and, using a luciferase assay, identified regions with putative STAT3 binding sites. Chromatin immunoprecipitation (ChIP) and an electrophoretic mobility shift essay (EMSA) confirmed that STAT3 binds to the promoter of caspase-3. However, caspase-3 was activated only in cells with high STAT3 expression levels, suggesting that STAT3 binds to caspase-3 with low affinity. To assess STAT3’s binding affinity to the caspase-3 promoter, we prepared serial dilutions of CLL cell DNA and, using ChIP and EMSA, found that STAT3’s binding affinities to p21 and c-Myc were 8 and 4 times as high, respectively, as its binding affinity to caspase-3, suggesting that high levels of STAT3 protein are required to activate caspase-3. Taken together, these findings suggest that activated STAT3 has a previously unknown pro-apoptotic function. At high intracellular levels, rather than providing CLL cells with survival advantage, STAT3 induces apoptosis by activating caspase-3. Thus, because CLL cell proliferation is coupled with spontaneous CLL cell apoptosis, the number of circulating CLL cells based on the cells’ proliferation rate is often lower than expected.