Galectin-1 Is Highly Expressed By Myeloma Cells and the Bone Marrow Microenvironment and Its Suppression Delineates a New Therapeutic in Vitro and in Vivo Strategy in Multiple Myeloma

Galectin-1 (Gal-1) is a lectin, involved in several processes related to cancer, including immunosuppression, angiogenesis, hypoxia, and metastases. However, the expression profiles of Gal-1 and its pathophysiological role in multiple myeloma (MM) cell growth, in the relationship between MM cells and the bone marrow (BM) microenvironment and in the MM-induced angiogenesis are unknown and were investigated in this study.

Firstly we evaluatedGal-1 expression by CD138⁺ cells of a dataset of 133 MM patients at diagnosis (GSE16122) and 23 human myeloma cell lines (HMCLs) (GSE6205) or on a proprietary? dataset of primary mesenchymal stromal cells (MSCs) and osteoblasts (OBs) of 16 MM and 4 MGUS. CD138⁺ cells and HMCLs were positive for LGALS1 with no statistically significant differences. LGALS1 mRNA expression was positively correlated with 154 genes and negatively with 109 genes including ERG1 and SPARC. MSCs cells showed a higher expression of LGALS1 compared to the OBs and MM-OBs showed a higher expression of LGALS1 mRNA than that obtained from healthy subjects. Gene expression profiling (GEP) data were then validated by Real-Time PCR and western blot in freshly purified primary CD138⁺ and BM MSCs samples as well as in 6 HMCLs and in both human MSC (HS-5 and hMSC-Tert) and osteoblastic cell lines (HOBIT and HOB-01). Moreover, immunohistochemistry analyses on bone biopsies obtained from 12 MM, 9 smoldering MM, 9 MGUS and 3 plasma cell leukemia samples revealed an high level of Gal-1 protein expression by MM cells, OBs and vessels in all the patients tested. Secondly, we evaluated whether Gal-1 expression was regulated by hypoxia and by Hypoxia Inducible Factor-1a (HIF-1a) checking the effect of hypoxic treatment (1% of O₂) and HIF-1α inhibition by shRNA lentivirus. We found that Gal-1 was upregulated in HMCLs upon hypoxic treatment and consistently the re-oxygenation process significantly restored the expression level of Gal-1. Interestingly the stable knock-down of HIF-1a significantly down-regulated Gal-1 expression in HMCLs both in normoxic and hypoxic conditions.

Thereafter, we explored the effect of persistent Gal-1 inhibition in MM cells and BM microenvironment cells on cell proliferation, survival and the transcriptional and pro-angiogenic profiles. An anti-Gal-1 Lentivirus shRNA was used for Gal-1 stable knock-down in HMCLs (JJN3-anti-Gal-1 and OPM-2-anti-Gal-1) and MSC cell lines (HS-5 and HMSC-Tert) and the Scramble lentiviral vector (JJN3-Scramble and OPM-2-Scramble) was used as the empty control vector. The stable inhibition of Gal-1 did not affect the proliferation rate and viability of both HMCLs and MSC cell lines. On the other hand Gal-1 inhibition by shRNA lentivirus significantly modified the transcriptional profiles of HMCLs and HS-5, evaluated by U133 Plus2.0 Arrays (Affymetrix®) either in normoxic or hypoxic or re-oxygenation conditions. Among the genes significantly modulated by Gal-1 inhibition in HMCLs, we found that pro-angiogenic (eg. CCL2, MMP9) and adhesion molecules (eg. MCAM and STEAP1) were down-regulated by Gal-1 suppression in both normoxic and hypoxic conditions as well as some putative anti-tumoral genes, including EGR1, SPARC and TGFBI, and anti-angiogenic ones, including SEMA3A, were up-regulated by Gal-1 inhibition. In line with these observations, we found that Gal-1 suppression by shRNA significantly decreased the pro-angiogenic proprieties of HMCLs by an in vitro angiogenesis assay.

Finally, we found that mice, injected subcutaneously with JJN3-anti-Gal-1 and OPM-2-anti-Gal-1, showed a reduction in the weight and volume of the tumor burden compared to mice inoculated with the JJN3-Scramble and OPM-2-Scramble. Moreover, a significant reduction in the number of CD34 positive vessels X field was observed. In an intratibial mouse model, JJN3-anti-Gal-1, JJN3-Scramble and JJN3 wild type were injected: in the anti-Gal-1 group tumors grew in reduced number and size compared to the Scramble group, moreover JJN3 anti-Gal-1 mice developed fewer and smaller lytic lesions on x-ray compared to the controls.

Overall our data indicate that Gal-1 is highly expressed by MM cells and those of the BM microenvironment and that its expression is regulated by hypoxia. Gal-1 shows a role in MM-induced angiogenesis and its inhibition in MM cells significantly reduced tumor growth in vivo, suggesting that Gal-1 is a potential new therapeutic target in MM.