NV 0605.indd

596 VOLUME 11 | NUMBER 6 | JUNE 2005 NATURE MEDICINE persistently activated in a human disease called anaplastic large cell lymphoma10. Chromosomal translocations join the nucleoplasmin gene to the gene that encodes ALK kinase, producing the persistently active NMP-ALK oncogene. To investigate the role of STAT3 in malignant transformation by NMP-ALK and in cancer cell survival, Chiarle et al.5 for the first time have studied bone marrow–derived cancer cells in which the STAT3 gene has been deleted in bone-marrow derived cancer cells. They first showed that fibroblasts containing one functional STAT3 allele could be transformed by the NMP-ALK oncogene, as measured by growth in soft agar. In contrast, STAT3-deficient fibroblasts could not be transformed. NMP-ALK activity can be expressed in lymphocytes both in cultured cells and in the whole animal. The authors studied animals bearing NPM-ALK and molecularly marked STAT3 genes capable of being deleted specifically in lymphocytes (a Cre recombinase driven by the T cell–specific CD4 promoter and homozygously loxP-marked STAT3). T lymphocytes retaining a single STAT3 allele and T lymphocytes lacking both STAT3 alleles formed NMP-ALK–induced lymphomas that killed the animals in about 40 weeks. In culture, however, the lymphomas without STAT3 grew less well than those still retaining one STAT3 allele. Most important, in the NMPALK–bearing animals with a single STAT3 allele, every one of the induced lymphomas had persistently active, tyrosine phosphorylated STAT3—and when the lymphoma cells in these animals were engineered to lose the STAT3 gene, the cells promptly underwent apoptosis. The authors went on to show that NMPALK can induce myeloma, a B cell–specific tumor. In mice that had STAT3 ablated in about 80% of B cells, all NMP-ALK–induced myeloma cells still had STAT3—and STAT3 was persistently activated. These genetic experiments indicate clearly that the presence of persistently active STAT3 favors the development of Tand B-cell tumors and is required for their maintenance. In addition, Chairle et al. injected antisense oligonucleotides to STAT3 at a site distant from transplanted lymphoma or myeloma tissue. This treatment successfully restricted growth of either T-cell lymphomas or B-cell myelomas. It is noteworthy that the antisense oligonucleotides against STAT3 killed tumor cells but had little effect on the animal. These experiments with hematologic tumors build on other studies suggesting that anti-STAT3 therapy would limit tumor growth. For instance, tumors in mice resulting from transplantation of human head and neck squamous carcinoma cells shrink when injected directly with an antisense STAT3 synthetic DNA derivative11. All of these results—together with the requirement of STAT3 for survival in many tumor cells in culture—promise that whereas all tumors may not have persistently active STAT3, the very large number that do may well respond to anti-STAT3 therapy. To this end, efforts are underway to discover small molecules capable of directly inhibiting STAT3. Early success has been reported with platinum compounds12 that block the DNA binding of phosphorylated STAT3 and with natural products that block STAT3 activation13 or possibly even disrupt STAT3 phosphodimers14. Given the preliminary success in developing small molecules—plus the evidence from several labs, including the current report of effective antisense treatment in whole animals—it is not at all unreasonable to hope that effective anticancer therapy can come form inhibiting persistently active STAT3.