Background: Mesenchymal stem cells (MSC) protect acute myeloid leukemia (AML) cells from chemotherapy. Galectin 3 (LGALS3) is a beta-galactoside binding protein that regulates cell adhesion, apoptosis, cell cycle, and mRNA processing. When secreted, LGALS3 can modulate receptor kinase activity, influence cell adhesion, and can suppress immune cells. MSC are a rich source of LGALS3. In the present study we address the role of MSC derived LGALS3 in the leukemia microenvironment.

Methods: The expression of LGALS3 and 150 other proteins was examined by reverse phase protein array analysis (RPPA) in AML (N = 106) and healthy donor (N = 71) MSC. To examine the putative role of MSC-derived proteins in therapy resistance, RPPA was used to determine levels of the 151 proteins in MSC obtained at diagnosis and from relapsed/refractory patients relapse. In vitro studies were performed to examine the effects of hypoxia or co-culture with AML cell lines on LGALS3 expression by immunofluorescence (IF) microscopy and immunoblot analysis. MSC containing lentiviral control shRNA or LGALS3 shRNA were used to assess effects of LGALS3 on cell viability, apoptosis, and cell adhesion by flow cytometry assays and effects on survival signaling by immunoblot analysis.

Results: LGALS3 levels were found to be higher in AML MSC compared to normal MSC (p = 0.0001). The increase in protein expression was not correlated with gene expression as mRNA levels were similar in AML and normal MSC, suggesting a post-translational mechanism. To identify MSC proteins associated with relapse or refractory status, RPPA was utilized to compare protein expression in MSC from newly diagnosed with those from relapsed and refractory (salvage) patients. RPPA identified LGALS3 as one of only three proteins increased in MSC from relapse/refractory patients. Pearson Correlation of LGALS3 with the other 150 proteins in AML MSC revealed that LGALS3 was negatively correlated with expression of integrin beta 3 and LYN. These 2 proteins were also among the 6 proteins found to have lower expression in salvage as compared to MSC from newly diagnosed patients. LGALS3 positively correlated with 13 proteins including phosphorylated beta-catenin which was also higher in salvage as compared to newly diagnosed AML MSC. . The LGALS3 promoter contains HIF1 alpha response elements. MSC from healthy donors grown in 1% oxygen displayed > 4 fold increase in LGALS3 protein after 24 hours. Hypoxic MSC exhibited a re-localization of LGALS3 from the nucleus to the cytoplasm. LGALS3 activates RAS signaling so hypoxia induced expression and nuclear export of the galectin would be expected to activate diverse survival signaling pathways and would also allow the molecule to be secreted where it could then suppress immune cells. Using in vitro co-culture experiments of MSC with human OCI-AML3 cells to mimic the leukemia microenvironment, LGALS3 levels were induced in MSC suggesting that cross talk from leukemia cells could contribute to the increased LGALS3 levels in MSC cells. The consequences of suppressing LGALS3 in normal MSC were tested using lentivral shRNA. Suppression of LGALS3 resulted in a 2-fold reduction of MYC and S473 phosphorylated AKT, and a 2-fold increase in PPP2R2A (an AKT phosphatase) compared to control shRNA. LGALS3 suppression rendered MSC less adherent to OCI-AML3 cells (~ 50% reduction in MSC with LGALS3 shRNA compared to control shRNA).

Conclusions: The data presented here demonstrate that LGALS3 is elevated in AML MSC especially in relapse/refractory samples. Possible mechanisms of up-regulation of LGALS3 in AML MSC may be due to the hypoxic nature of the leukemia microenvironment and/or contact with AML cell. Hypoxia induced nuclear export of the galectin to the cytoplasm could contribute to MSC survival and anti-tumor functions (e.g. suppression of immune cells by secreted galectin). Knock down of LGALS3 in MSC suppresses AKT activation and MYC expression suggesting it has a key role in MSC cell survival. Experiments are underway to determine if MSC derived LGALS3 might impact AML cell homing (as a regulator of cell adhesion) and/or immune surveillance (as a suppressor of T and NK cells). The identification of LGALS3 as an important component of AML MSC suggests that this molecule could be a target in a tumor microenvironment based therapy concept of AML.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.