Abstract 2690

Diffuse large B cell lymphoma (DLBCL) is the most common type of adult non-Hodgkin lymphoma. Aberrant gene expression and mutations are involved in the pathogenesis of DLBCL and may have prognostic and therapeutic relevance. We have shown that STAT3 (by immunohistochemistry) is constitutively activated in approximately 50% of new untreated DLBCL cases. To investigate whether genetic mutations play a role in the activation of JAK2/STAT3 signaling pathway in DLBCL, we sequenced the coding regions of the JAK2 gene and found no novel missense mutations.

Since we did not find any missense mutations in the JAK2 gene, we proceeded to bi-directionally sequence all the 24 exons of the STAT3 gene in the same DLBCL patient tumors (n=40). We discovered mutations in 2 patients, one in exon 7 (T>A) and the other at the stop codon in exon 24 (G>A). While the G to A mutation in exon 24 was a silent mutation, the T to A mutation in exon 7 led to methionine substitution by lysine at codon 206 (M206K). Codon 206 is located in the coil-coiled domain of STAT3, which is a conserved amino acid in STAT3 from many different species. In order to delineate the functional significance of this STAT3 M206K mutation we generated STAT3 expression plasmids with the lentiviral expression vector pLEX-MCS. STAT3 constructs carrying the M206K mutation were generated by site directed mutagenesis of wild type STAT3. Since there are two STAT3 isoforms in humans, full-length STAT3α and the splicing variant STAT3β (which lacks the C-terminal transactivation domain), we constructed stable cell lines containing mutant and wild type gene of both STAT3α and STAT3β in HEK-293T cells. The mutation was found to increase the phosphorylation of Tyr705 in STAT3α mutant cell lines with no effect on STAT3β mutant cell line. Moreover the mutant STAT3α cell line was more sensitive to IL-10 in the activation of Tyr705 compared to its wild type while STAT3β mutant cell line was insensitive to IL-10. Both the STAT3 mutants (STAT3α and STAT3β) were insensitive to IL-6, which suggests that the mutant can sense the signal specifically from IL-10. Furthermore JAK2 inhibition through the pharmacological inhibitor TG 101348 dephosphorylated STAT3 in WT cells; however, had minimal effect in both the STAT3 mutants. Immunofluorescence staining showed that the M206K mutation did not affect the nuclear translocation of both STAT3α and STAT3β, which was confirmed by immuno-blotting on the cytosol and nucleus of the cells. Interestingly, the mutation activated the transcriptional activation activity of both STAT3α and STAT3β by luciferase reporter assay.

Overall these results suggest that the M206K STAT3 gene mutation is rare (2.5%) in DLBCL tumors, leads to STAT3 activation, and enhances STAT3 transcriptional activation activity. Thus, STAT3 mutations in addition to cytokines can produce STAT3 pathway activation in DLBCL. Our data suggest that clinical trials of novel STAT3 inhibitors need not be restricted to tumors with STAT3 mutations but rather should focus on tumors with demonstrated aberrant STAT3 activation.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.