Abstract 3119Poster Board III-56The frequency of extramedullary infiltration (EMI) in acute myeloblastic leukemia (AML) is reported up to 40% and most prevalent in myelomonoblastic and monoblastic subtypes of AML (M4 and M5 according to FAB classification). The majority of patients with EMI suffered poor prognosis. To explore mechanism underlying EMI, we analyzed SHI-1 cells, a highly invasive human acute monocytic leukemia cell line, and other leukemia cells such as NB4, K562, U937 and THP-1 cells. qRT-PCR revealed that the transcription of TIMP-2, MMP-2 and MT1-MMP in SHI-1 cells were all higher than other leukemic cells(P<.05). Western blotting analysis showed that expression of these genes in SHI-1 cells were also significantly higher than that in other leukemic cells at the protein level (P<.01). By using cell migration assay to determine the invasive capability of leukemic cells, and BMSCs co-culture to mimic the microenvironment of bone marrow, we found that the invasion rates of NB4, K562, U937 and THP-1 cells were only 5% to 60% of SHI-1 cells(P<.05). To elucidate the individual contribution of TIMP-2, MMP-2 and MT1-MMP to the invasive capability of SHI-1 cells, RNA interference (RNAi) was applied to individually knock down these genes. As determined by qRT-PCR 24 hours after transfection, the transcription of theses genes was inhibited to 85% to 98%, respectively, when compared with control cells transfected with non–target-directed siRNA(P<.001). Immunoblot analysis on cell lysates revealed a minimal expression of these genes at protein level. Cells being treated with RNAi were applied to the transwell assay to detect their migratory potential. We found that the down-regulation of TIMP-2, MMP-2 and MT1-MMP significantly decreased the invasion rates of SHI-1 cells by 60%-70%, 50%-60% and 40%-50%, respectively (P<.05). These findings suggested that all these three genes enhance the process of SHI-1 cells invasion. To determine the precise role of TIMP-2 in this specific leukemic cell, SHI-1 cells were transfected with human TIMP-2 cDNA. Following G418 selection and limited dilution, three stable subclones (S1, S2 and S3) were isolated and expanded. qRT-PCR analysis showed that the mRNA level of TIMP-2 increased about 3 fold, 2 fold and 1.5 fold in S1, S2 and S3 cells, respectively(P<.05), whereas the MMP-2 and MT1-MMP almost unchanged. Western blotting further exhibited that protein levels of TIMP-2 in S1, S2 and S3 cells increased 2.6 fold, 1.5 fold and 1.3 fold, respectively(P<.01) when compared with the control SHI-1 cells, while no difference of MMP-2 and MT1-MMP protein level could be found between three subclone cells and SHI-1 cells. These subclone cells co-cultured with BMSC were then subjected to transwell assay to detect the influence of TIMP-2 up-regulation on cell invasion. We found that the invasion rates of these subclone cells elevated about 1.5 fold to 2.5 fold (P<.05). To explore the underlying mechanism of high expression of TIMP-2, MMP-2 and MT1-MMP promoting SHI-1 cells invasion, by using zymography to detect proMMP-2 secretion and MMP-2 activation, we found that released proMMP-2 (72KDa) as well as activated MMP-2 (62KDa) from SHI-1 cells were both higher than that in other leukemic cells (P<.05). Furthermore, three TIMP-2 transfected SHI-1 subclones demonstrated about 1.5 fold to 3.0 fold (P<.01) increase in activated MMP-2 while the proMMP-2 secretion remained unchanged. Down-regulating the expression of MMP-2, but not MT1-MMP and TIMP-2, led to a markedly decline of secreted proMMP-2. Activated MMP-2 had not been found in the supernatants any of these transfected cell culture. These results suggested that up-regulation of TIMP-2 as well as elevation of MMP-2 and MT1-MMP-2 expression will promote SHI-1 cell invasion, at least partially via the increase of MMP-2 activation. In conclusion, our study demonstrated that constitutively high expression of MMP-2, MT1-MMP and TIMP-2 in SHI-1 leukemic cells facilitated cell invasion by promoting MMP-2 activation. Up-regulation of TIMP-2 exhibited not a repressive but an active effect on SHI-1 leukemic cells invasion. An elevated expression of TIMP-2 in AML patients may have higher risk of EMI, particularly in patients in company with high levels of MMP-2 and MT1-MMP. We suggest that a careful analysis of TIMP/MMP ratio in vivo may be useful to monitor and predict extramedullary infiltration of leukemia cells in AML patients. DisclosuresNo relevant conflicts of interest to declare.