Acute leukemias with Mixed Lineage Leukemia gene (MLL, also called KMT2A) translocations are associated with poor outcomes. These leukemias are most frequently encountered in infants and as secondary malignancies, and can be either lymphoid or myeloid. In spite of aggressive treatments, including bone marrow transplantation, infants with MLL-leukemias face a grim prognosis, with predicted survival of only 20%. Novel, effective therapies are thus urgently needed. In search for molecular targets, we observed that MLL-leukemias uniquely display over-expression of Lysosome-associated Membrane Protein 5 (LAMP5, also known as C20orf103). This observation was consistent across several gene-expression-profiling studies and occurred in both ALL and AML. Moreover, data from the TCGA study on AML showed that patients with LAMP5 expression suffered worse prognosis compared to those lacking LAMP5. We first confirmed LAMP5 expression in human leukemia cell lines by immunoblot and RT-qPCR assays. We readily detected LAMP5 mRNA and protein in MLL-fusion leukemia cell lines (MV4;11, MOLM13, THP1, RS4;11, KOPN8), while no expression of LAMP5 was found in the non-MLL-cell lines (Kasumi, K562, REH). Published ChIP-seq studies on leukemia cell lines show the MLL-fusion protein directly bound at the promotor region of LAMP5. To validate that the MLL-fusion protein activates the expression of LAMP5, we transformed human CD34+ cord blood cells with an inducible (Tet-off) retrovirus carrying the MLL-AF9 fusion cDNA. In this system, addition of Doxycycline represses expression of the MLL-AF9 oncogene. We found that LAMP5 expression directly correlated with MLL-AF9 levels, with levels of both decreasing upon addition of Doxycycline. To investigate the role of LAMP5 in MLL-fusion leukemia, we studied the effect of shRNA-mediated knockdown. By screening several hairpin sequences, we identified one construct that efficiently inhibited LAMP5 expression in MLL-fusion leukemia cells but had no effect on LAMP5-negative cells, implying specificity. All MLL-fusion leukemia cell lines tested showed growth inhibition with LAMP5 knockdown. Specifically, growth of MV4;11, THP1 and MOLM13 cells was decreased by 69%, 73% and 80% respectively compared to controls (non-targeting shRNA). When cultured in semi-solid methylcellulose media for 10 days, LAMP-5-depleted MV4;11 cells formed significantly fewer colonies than control cells (64.3 ± 25.98 and 245.3 ± 27.42 colonies per 1000 cells respectively). To investigate the role of LAMP5 in leukemia-propagation in vivo, we transplanted control and LAMP5-depleted MV4;11 cells into Busulfan-conditioned immune-deficient (NRGS) mice (2x105cells/mouse). Preliminary results from bone marrow aspirates of transplanted mice at weeks 4 post-transplant showed abundant human leukemia cells in mice receiving control cells while the mice receiving LAMP5-knockdown cells showed near-absence of human leukemia cells. Collectively, these results demonstrate that LAMP5, a novel target of MLL-fusion proteins is required for the propagation of leukemia. In normal hematopoiesis, LAMP5 expression is restricted to non-activated plasmacytoid dendritic cells (pDC), where it localizes to the ER-Golgi intermediate compartment (ERGIC). In leukemia cells, using immunofluorescent confocal microscopy we detected LAMP5 in the perinuclear zones where it co-localized with ERGIC-53, a marker of the ERGIC compartment. While little is known about the functions of LAMP5 in normal pDC, studies suggest that it functions as a co-chaperone with UNC93B1, a known Toll-like Receptor (TLR) chaperone, to shuttle the TLRs to their respective locations in the plasma membrane or endosomes. In ongoing experiments, we are determining the functions of LAMP5 in leukemia, including its association with UNC93B1 and with the TLR-NFKB signaling pathway. Overall, based on our results and the limited expression in normal hematopoiesis, we postulate that LAMP5 could potentially serve as a therapeutic target with a wide therapeutic-window to treat MLL-leukemias.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution