Treatment of B-cell precursor acute lymphoblastic leukemia with the Galectin-1 inhibitor PTX008

All BP-ALL studies, except where noted, were conducted using a co-culture system in which the leukemia cells are grown with mitotically inactivated murine OP9 stromal cells in αMEM with 20% FBS, 1% L-glutamine and 1% penicillin/streptomycin (Life Technologies, Grand Island, NY). OP9 cells were mitotically inactivated by exposure to ≈ 8000 cGy γ irradiation. In other experiments (Figs. 1e, 5e and 6c, d) OP9 cells were mitotically inactivated by treatment with 10 μg/ml mitomycin C (Sigma cat M4287) for 3 h in complete medium. Selected lots of FBS that supported the co-culture included JR Scientific cat 43,640 lot K051–6; Life Technologies cat 16,140,071 lot 1,642,038 US Qual HI; and Atlanta cat S11550H lots D14047 and L15124. OP9 (CRL-2749) was obtained from the American Type Culture Collection. Patient-derived BP-ALLs include previously described [14‐17] TXL2 (diagnosis sample, Ph-positive, p210 BCR-ABL1 wild type), ICN06 (diagnosis; tel-AML1), US7 (no known karyotype abnormalities; bone marrow sample, adult, at diagnosis), LAX57 (diagnosis; 96% blasts; CD45 +/−(dim to negative), CD19 + CD10 + CD20+/− (dim to negative), CD22 + CD34 + SIg- with [t(1;9)(q44;p22)]), and LAX56 (relapse; 89% blasts; CD45+/−(dim to negative), CD19 + CD10 + CD20+/− (dim to negative), CD22 + CD34 + SIg-; with [t(Y;7)(p11.3;p13)]). LAX56 and LAX57 grew directly on OP9 stroma from Ficoll-purified bone marrow mononuclear cells. TXL2, ICN06 and US7 were passaged in NOD.Cg-Prkdcscid IL2rgtm1Wjl/SzJ (NSG) mice (Jackson Labs, Bar Harbor, ME), then grown on OP9 stroma. BM50 was a CD19-CD22+ relapse after bone marrow transplant and CART19 treatment. BM50 cells were passaged in NSG mice as PDX, then grown on OP9 stroma. Human specimen collection protocols were reviewed and approved by Children’s Hospital Los Angeles Institution Review Board (IRB) [Committee on Clinical Investigations] (CCI). Collections were in compliance with ethical practices and IRB approvals. Nilotinib (AMN107), obtained from Novartis (Basel, Switzerland), was dissolved in DMSO and stored at − 20 °C. A vincristine sulfate solution was obtained from Hospira Worldwide Inc. (Lake Forest, IL, USA). PTX008 (− 1 in Fig. 5e) was from Kevin Mayo. PTX008 (− 2 in Fig. 5e) was purchased from Axon Medchem LLC (Reston, VA 20191, cat # 2332) and was also used in Fig. 6c, d and in Additional file 1: Figure S2 panel c.

Gene expression analysis- Meta-analysis was performed for LGALS1 (Galectin-1) expression on a data set of RNAs from 270 pediatric bone marrow samples of ALL at diagnosis and 4 normal B-cell progenitor bone marrow samples selected for CD19 and CD10 on HG-U133A Affymetrix arrays [18]. Processed data in the series matrix files from GEO Datasets accession GSE28497 [18] represent values normalized by MAS5.0 with MFI values calculated to a median target intensity of 500. Txt file values for probe set ID 201105_at imported into Excel were manually extracted into Prism5.0. Coustan-Smith et al. reported (Table S2 in [18]) that 78.9% of the leukemia samples overexpressed LGALS1, defined as 25% or more of B-lineage ALLs compared to normal precursor B-cells.

Real-time RT/PCR- For RNA analysis, cells were cultured in complete medium for 24 h or plated on irradiated OP9 cells. They were then stimulated with recombinant GST or GST-Gal3 purified with a step to remove endotoxin as described, [5] for 24 h and harvested. RNA was isolated using Trizol. cDNA was generated with random primers and an ABi High Capacity cDNA reverse transcriptase kit according to the supplier’s instructions. Primers used for reverse transcription/real time PCR of human LGALS1 were: forward 5-CTC TCG GGT GGA GTC TTC TG-3 and reverse: 5-GAA GGC ACT CTC CAG GTT TG-3. Samples were run in triplicate and data were analyzed using comparative Ct, with all samples being normalized to non-treated US7 cells.

Western blotting- Cells were lysed by a 20-min incubation in RIPA buffer (50 mM Tris-HCL, pH 8.0, 150 mM NaCl, 1% Triton X-100, 0.5% deoxycholate, 0.1% SDS) containing 2 μg/mL aprotinin, 10 μg/mL leupeptin, 1 μg/mL pepstatin A, 1 mM PMSF, 10 mM sodium fluoride, 1 mM sodium orthovanadate. Membranes were reacted with antibodies against Galectin-1 (cat 101,566, Genetex, Irvine, CA), pSrc (cat 2101S Cell Signaling Technology [CST] (Boston, MA), p44/42 Erk (cat 4377S CST), Src (cat 2109 CST), Erk (sc-94, Santa Cruz Biotechnology, Dallas, TX), Galectin-3 (cat 125,402, Biolegend, San Diego, CA). Gapdh (cat 627,408, Genetex, Irvine, CA) was used as a loading control.

To determine the levels of Galectin-1 on the surface of ALL cells, all ALL cells in the plate (except where indicated), were collected from OP9 cell co-cultures. ALL cells were washed 1× with FACS buffer (PBS, 2% BSA, 0.1% sodium azide) prior to blocking with human FcR blocking reagent (Miltenyi Biotech, San Diego, CA) according the manufacturer’s instructions. Cells (1 × 10⁶) were then incubated with 2.5 μg Galectin-1 antibody (Fig. 1: cat AF1152 R&D Systems, Minneapolis, MN; Fig. 3: cat 101566 Genetex, Irvine, CA) conjugated to CF647 (Mix-n-stain, Sigma-Aldrich, St. Louis, MO), for 45 min at 4 °C. Control cells were stained with rabbit APC-conjugated IgG. Cells were washed 2× with FACS buffer prior to analysis on a BD Accuri C6 cytometer (BD Biosciences, San Jose, CA). PE-conjugated Galectin-3 antibodies were from Biolegend (cat 126706).

DTSSP (3,3′-dithiobis[sulfosuccinimidylpropionate]; Thermo Scientific, Waltham, MA) was dissolved in PBS. LAX56 cells (6 × 10⁶) were resuspended in PBS and incubated with 2 mM DTSSP for 30 min at room temperature (RT). The reaction was terminated by addition of 20 mM Tris, pH 7.5 and incubation for 15 min at RT. For Galectin displacement assays, GST-Galetin-3 obtained from [19] was generated as previously described. A plasmid encoding bacterially-expressed His-tagged Lgals-1-pMCSG7 (clone HSCD00343182) was obtained from DNASU (Arizona State University). Galectin-1 protein was purified using standard procedures. US7 and LAX56 cells in co-culture with OP9 stromal cells were treated for one hour with DMSO or 10 μM PTX008, followed by further exposure to medium alone or to 50 mM lactose, 20 μM recombinant human Galectin-1 or 20 μM recombinant human Galectin-3 for 24 h. Suspension cells were collected and analyzed by FACS for cell surface Galectin-1 or Galectin-3.

Human bone marrow and PB were collected in EDTA tubes and the cellular component removed by centrifugation. Plasma samples were diluted 10× and values calculated based on a standard curve as described below. We tested the same samples independently using a commercial human Galectin-1 ELISA (used in Additional file 1: Figure S1) and an ELISA made as follows (used in Fig. 1f). As described above, human recombinant His-tagged and TEV protease-site containing Galectin-1 protein was produced in E-coli Rosetta from pMCSG-7-Gal1 plasmid (DNASU clone HsCD00343182). Recombinant protein was isolated on a Ni-NTA agarose (Thermo-Fisher) column and treated overnight with TEV protease (5 u/ml). TEV protease, undigested His-Galectin-1 protein and His tag were removed by collecting the flow-through of a second Ni-NTA column. Samples were run on SDS-PAA gels to check the concentration and purity of the protein. Galectin-1 was serially diluted from 100 ng/ml to 0.03 ng/ml to generate a standard curve. A 96-well plate was coated with 10 μg/ml capture antibody (GeneTex, GTX101566) overnight at 4 °C. After blocking with 5% BSA in PBST (containing 0.05% Tween-20) at RT for 1 h, plasma samples were diluted in 10 times with PBS and applied into each well in 100 μl for 2 h incubation at RT. After washing, 2.5 μg/ml detection antibody (Abcam, ab58085) was added and the plate was incubated at RT for 2 h. HRP-conjugated goat anti-mouse IgG (H + L) (Invitrogen, 62–6520) as secondary antibody was diluted 1:400, and added to the plate for 1 h RT incubation. To measure absorbance, 100 μl ABTS Peroxidase Substrate (KPL, 50–66-00) was added into each well followed by 30 min incubation. The result was detected by absorbance at 405 nm and sample Galectin-1 levels determined using the standard curve. Human plasma samples from peripheral blood and bone marrow were as described previously [4]. For the results shown in Additional file 1, an ELISA was purchased from R&D Systems.

LAX56 cells (2 × 10⁴) were seeded into wells of a 24-well plate and incubated with DMSO or 10 μM PTX008 in the absence of OP9 cells. Cells were then treated with 20 μM Galectin-1 or GST-Galectin-3 or GST recombinant proteins for four hours. Cells were imaged using an Olympus 1X71 inverted microscope with digital camera. Agglutinated cell clusters defined as cell aggregates of 200 μm or larger were counted from three different images of three wells.

A 96-well plate was coated overnight at 4 °C with 5 μg/mL fibronectin in PBS. Negative control wells were coated with 0.01% poly-L-lysine (Sigma-Aldrich, St. Louis, MO), for 20 min at RT. All wells were washed with 0.1% BSA in PBS prior to blocking with 2% BSA in PBS for 1 h at RT. After blocking, wells were washed twice with 0.1% BSA in PBS. LAX56 cells were labeled with 5 μM Calcein AM (ThermoFisher Scientific, Waltham, MA) for 30 min at 37 °C. Cells were washed 2× with PBS prior to seeding at 5 × 10⁴ cells/well in αMEM base media +/− PTX008 or DMSO for 30 min. As a positive control for Galectin-1 inhibition, cells were incubated with various concentrations of lactose (Sigma-Aldrich, St. Louis, MO). After 30 min, the plate was read at 485 nM on a Synergy HTX Multi-Mode Reader (BioTek, Winooski, VT). The plate was washed two more times with 0.1% BSA in PBS, prior to reading again at 485 nM. The percent adherent cells was calculated using the following formula: adhesion = [(MFI _{post wash}-MFI _background)/ (MFI _{pre wash}-MFI _background)] × 100.

Migration of US7 and LAX56 cells was tested using a Transwell assay (5 μm pore size). US7 and LAX56 cells were washed 1× with base αMEM prior to resuspension in αMEM + 1% FBS and either DMSO vehicle or 10 μM PTX008. 100 μL of media containing 5 × 10⁵ cells was placed in the upper chamber and 600 μL of media with or without 200 ng/mL SDF-1α (PeproTech, Rocky Hill, NJ) in the lower chamber. For migration of leukemia cells towards OP9 cells, a confluent layer of irradiated OP9 cells was plated in the well 24 h prior to performing the Transwell assay. After overnight incubation, upper chambers were removed and cells that had migrated to the lower chamber were counted using Trypan blue.

Human ALL cells cultured in a 24-well plate were provided with stromal OP9 support unless otherwise indicated. For proliferation and viability analysis, treatment was with PTX008 or DMSO for 3 days. Additional cultures were treated with 5 nM vincristine or in combination with PTX008. Cell number and viability were assessed using manual counting of Trypan blue-excluding cells on day 3. For cell cycle analysis with PTX008, vincristine, or a combination of PTX008 with vincristine, 1 × 10⁶ leukemia cells were plated with complete α-MEM media on a confluent layer of mitotically inactivated OP9 cells, or kept in complete medium. At 72 h, cells were incubated with 10 μM BrdU for 4 h. All cells were removed via trypsinization and stained following the manufacturer’s directions (BD FITC BrdU Flow kit, cat #559619, BD Pharmingen, San Jose, CA). Briefly, cells were fixed with BD Cytofix/Cytoperm for 30 min at RT. Cells were washed with Perm/Wash Buffer prior to incubation with Cytoperm Permeabilization Buffer Plus for 10 min on ice. After fixation and permeabilization, cells were treated with 30 μg RNase for 1 h at 37 °C. Cells were then incubated with 1:50 FITC anti-BrdU for 20 min at room temperature. Cells were washed with Perm/Wash buffer and then resuspended with 20 μL 7-AAD prior to addition of FACS buffer. For apoptosis analysis, LAX56 cells grown on mitomycin-C treated OP9 stromal cells or kept in complete medium were treated for 72 h with drugs, then stained for Annexin-V and 7-AAD according to the manufacturer’s instructions (APC Annexin V Apoptosis Detection kit, cat#640930, Biolegend, San Diego, CA). FACS analysis was used to identify early and late apoptotic cells (Annexin-V+, 7-AAD+). Gating in FSC/SSC was on all single cells. Data in Fig. 6a, b were acquired on a BD Accuri C6 cytometer (BD Biosciences, San Jose, CA); those in Fig. 6c and d were acquired on a BD LSR Fortessa X-20 Cell analyzer.

All animal experiments were carried out in concordance with Institutional IACUC and NIH guidelines. 2 × 10⁶ LAX57 ALL cells were injected i.v. into female NSG mice. On day 7 after transplant, mice were treated intraperitoneally 5× per week with PBS or 5 mg/kg PTX008, or 0.5 mg/kg vincristine once per week, or a combination therapy of daily 5 mg/kg PTX008 and once per week 0.5 mg/kg vincristine. Mice (n = 6 mice per group) were treated up to day 35 post-transplant and weighed daily during treatment. When controls lost 10% weight compared to the previous day, all mice were sacrificed. White blood cells from peripheral blood, spleen and bone marrow were analyzed by flow cytometry after red blood cell lysis with BD Pharm Lyse (BD Pharmingen, San Jose, CA). 1 × 10⁶ cells were stained with 1.25 μg CD19 FITC (clone H1B19, Biolegend, San Diego, CA), 0.5 μg LyG FITC (clone 1A8, BD Pharmingen, San Jose, CA), 0.5 μg CD11c PE (clone N418, eBioscience, San Diego, CA), 0.2 μg CD11b (clone M1/70, Biolegend, San Diego, CA), 0.9 μg Galectin-1 (cat 101,566, Genetex, Irvine, CA) conjugated to CF647, or the appropriate isotype controls.

Based on four ALL cell lines and patient samples, Juszczynski et al. reported that MLL-rearranged ALLs selectively express Galectin-1 [7]. Gene expression profiling by Coustan-Smith et al. [18] confirmed significantly elevated Galectin-1 mRNA in all 23 ALL samples with MLL rearrangement compared to controls. Meta-analysis of that data set however (Fig. 1a) shows that individual samples within other subtypes of ALL also contain high Galectin-1 mRNA levels including non-MLL rearranged samples, compared to normal bone marrow CD19⁺CD10⁺ cells from 4 healthy donors. In agreement with this, Western blotting (Fig. 1b) showed that Galectin-1 protein was present in different non-MLL rearranged BP-ALLs including TXL2, a Ph-positive ALL; diagnosis sample LAX56, relapse sample LAX57, and diagnosis sample US7.

We additionally evaluated TXL2 and US7 for Galectin-1 mRNA levels using real-time PCR. The human BP-ALL cells used in our studies are dependent on stromal support and have a high percentage of dying cells in the absence of a feeder layer when early freeze-downs are used. We and others use the murine bone marrow stromal cell line OP9 to stimulate the growth and also to promote drug resistance development of such human BP-ALL cells (e.g., [15‐17, 20‐22]). We compared Galectin-1 mRNA in ALL cells kept in complete medium without OP9 cells to those cultured with stromal support, but expression was comparable under these conditions (Fig. 1c). Interestingly, stimulation of US7 BP-ALL cells by the addition of exogenous Galectin-3 induced an increase in Galectin-1 mRNA expression (Fig. 1c, compare US7 control GST to GST-Galectin-3), indicating that in some ALLs, its levels are likely to be regulated by external factors.

Flow cytometric analysis was used to confirm expression of Galectin-1 by various ALL subtypes. Because of a possible low-affinity association of Galectin-1 with the cell surface, we compared the signals of cell surface Galectin-1 with and without chemical cross-linking. As shown in Fig. 1d, DTSSP cross-linking had no measurable effect on Galectin-1 detection, indicating that Galectin-1 is bound with relatively high affinity to these cells. Further FACS analysis confirmed that ALLs other than the MLL subtype express Galectin-1. BM50, from a patient who had relapsed on anti-CD19 CAR-T cell therapy, contained a high percentage of cells with cell surface Galectin-1, whereas ICN06, an ETV6-RUNX positive ALL had the lowest percentage. All the cells in these ALL samples except ICN06 were positive for intracellular Galectin-1 (Fig. 1e).

As increased levels of Galectin-1 were reported in serum samples of lymphoma and osteosarcoma patients [23], we also compared control and ALL bone marrow and peripheral blood plasma samples for Galectin-1. No statistically significant differences were found between ALL and control samples (Fig. 1f, also see Additional file 1: Figure S1). Bone marrow plasma had high levels of Galectin-1 (average, 163.6 ng/ml) compared to blood plasma (average 79.7 ng/ml). We found much higher Galectin-1 levels in normal blood plasma compared to those reported by others in serum (up to 7 ng/ml in healthy people e.g. [24]). This could be caused by depletion of Galectin-1 in serum during coagulation, as it was reported to bind to Factor VIII [25].

To investigate the function of Galectin-1 in ALL cells we made use of a specific small molecule inhibitor of Galectin-1, PTX008 (also known as 0118, OTX008). PTX008 is a non-peptidic, calix (4)arene-based topomimetic of Anginex that acts as an allosteric inhibitor of Galectin-1 by binding to the lectin at a site opposite of the carbohydrate binding domain [12]. PTX008 has been in clinical trials where it was found to be well-tolerated [26] [27].

We used an agglutination assay to confirm the specificity of this compound in inhibiting binding of Galectin-1 to BP-ALL cells. As shown in Fig. 2a, LAX56 cells do not spontaneously aggregate in DMSO or in the presence of control GST but exogenously added Galectin-1 or GST-Galectin-3 promote agglutination. Pretreatment with PXT008 significantly inhibited the ability of Galectin-1 but not of Galectin-3 to agglutinate the cells (Fig. 2b). We also tested if PTX008 could inhibit Galectin-1 and Galectin-3 binding to the cell surface of US7 and LAX56 cells by flow cytometry. As shown in Additional file 1: Table S1, PTX008 inhibited recombinant Galectin-1 but not Galectin-3 binding to cell surface glycoconjugates present on US7 and LAX56 cells.

Extracellular Galectin-1 interacts with glycosylated β1 integrins on BP-ALL Nalm6 cells [28]. As β1 integrins on BP-ALL cells regulate adhesion to extracellular matrix proteins, we also tested if PTX008 would affect integrin-mediated adhesion of BP-ALL cells to fibronectin (FN), a specific integrin α4β1 ligand, or to control poly-L-lysine. We found that PTX008 monotreatment did not affect the viability of the cells under these conditions (also see below). As a positive control for Galectin-1 inhibition, ALL cells were incubated with lactose, a widely used inhibitor for Galectin-1-carbohydrate interactions. As shown in Fig. 3a-c (middle panels), PTX008 had no effect on adhesion of any of the ALLs tested to poly-L-lysine. Also, the drug did not significantly inhibit adhesion of US7 ALL cells (Fig. 3a, left panel), to FN. Treatment of LAX56 cells, which had relatively high cell surface Galectin-1, with PTX008 reduced FN-mediated LAX56 adhesion better than 50 mM lactose (Fig. 3b). Similarly, PTX008 treatment produced a dose-dependent inhibition of LAX57 binding to fibronectin (Fig. 3c). These combined results (Fig. 3d) suggest that PTX008 specifically inhibits the interaction of cell surface Galectin-1 on ALL cells with receptors for fibronectin.

We also investigated if Galectin-1 inhibition with PTX008 affects ALL cell migration. We tested chemotaxis to SDF1α (CXCL12), as BP-ALL cells exhibit high levels of CXCR4 expression and SDF1α, the CXCR4 ligand, is responsible for ALL cell homing to and engraftment in the bone marrow [29]. As shown in Fig. 4a and b, PTX008-treated US7 and LAX56 cells exhibit significantly reduced migration towards SDF1α, when compared to DMSO controls.

OP9 stromal cells strongly stimulate the migration of BP-ALL cells: when BP-ALL cells are added to a stromal OP9 layer, they migrate to and underneath the stromal layer within a time frame of hours. This is mediated, to a large extent, by SDF1α secreted by the stroma [30]. We therefore also tested the effect of PTX008 in interfering with migration of BP-ALL cells to OP9. Consistent with its inhibition of migration towards SDF1α, as shown in Fig. 4, PTX008 also inhibited migration to OP9 cells compared to vehicle control.

We next investigated the effect of Galectin-1 inhibition by PTX008 on ALL cell proliferation and viability. As shown for LAX56 in Fig. 5a-b (left panels) and for LAX57 in Fig. 5c-d (left panels), based on cell numbers, PTX008 was cytostatic for LAX56 (Fig. 5a, b) and for US7 (Additional file 1: Figure S2a). The presence of OP9 stromal cells, as used here, strongly protects BP-ALL cells against eradication when these are treated with chemotherapeutic drugs [15‐17]. In agreement with this, in the presence of irradiated OP9 stroma, mono-treatment with this drug at concentrations up to 10 μM had no, or only a small effect on cell viability (Fig. 5a-d).

We also tested a combination treatment with vincristine. LAX56 cells were relatively resistant to a 3-day treatment with 5 nM vincristine (DMSO samples) but a combination with 10 μM PTX008 significantly reduced viability and cell proliferation (Fig. 5a, b right panels). Similarly, the combination also affected LAX57 (Fig. 5c, d right panels) and US7 as well as TXL2 (Additional file 1: Figure S2a, b, bottom panels). Moreover, the inclusion of 10 μM PTX008 allowed a dose reduction of vincristine, with 2.5 nM vincristine combination, achieving a similar cytostatic effect (Fig. 5e, left panel) as monotreatment with 5 nM vincristine.

Nilotinib is a targeted tyrosine kinase inhibitor of BCR-ABL1 that is used to treat Ph-chromosome positive leukemias. We also treated TXL2 cells with the combination of PXT008 and nilotinib. This analysis showed that the cytostatic effect of 20 nM nilotinib was enhanced by the presence of PTX008 (Additional file 1: Figure S2 panel c).

In head and neck SQ20B and A2780-1A9 ovarian cancer cell lines, PTX008 treatment resulted in a decreased level of Galectin-1 expression at 48 and 72 h and reduced phospho-Erk starting as soon as 2 h post-treatment [11]. Because PTX008 interferes with Galectin-1 dimerization [12], and Galectin-1 dimers promote H-Ras nanoclustering in epithelial cells [31‐33] we also analyzed Galectin-1, pSrc and pErk1/2 protein levels in lysates of BP-ALL cells treated with PTX008. However, we did not measure marked changes in pSrc or Galectin-1 levels, although pErk1/2 was clearly decreased after 24 h of drug exposure (Additional file 1: Figure S3).

We next tested if treatment with PTX008 affects the cell cycle of ALL cells when these were co-cultured with OP9 stromal cells. LAX56 and LAX57 were treated with 10 μM PTX008 for 72 h and then assessed for cell cycle using BrdU and 7-AAD. DMSO-treated control cells showed active DNA synthesis, with a relatively high percentage of cells in S phase. Interestingly, PTX008 treatment significantly reduced this (Fig. 6a, b).

We also compared the effect of PTX008 on BP-ALL cells with and without OP9 stroma. Figure 6c (compare left and right panels, control DMSO-treated samples,) illustrates the impact of the presence of the stromal cells: without OP9 stromal support, more cells with < 2 N DNA content (apoptotic/necrotic- subG0G1) were present. PTX008 treatment had a rather dramatic effect on increasing this population in the absence of OP9 cells but also had a marked effect even in the presence of their support (Fig. 6c, PTX 10 μM samples, compare left and right panels sub G0G1).

Vincristine is a vinca alkaloid that binds to microtubules and spindle proteins in S-phase and interferes with mitosis. As shown in Fig. 6c (vin samples in left and right panels) and as expected, vincristine treatment affected the proliferation of the ALL cells more strongly when no OP9 was present, reducing the percentage of cells in S phase. When we combined the PXT008 and vincristine treatment on cells without stromal support, no additional toxicity was observed (Fig. 6c, combination, left panel). However, because the overall number of healthy cells was higher in the OP9 co-cultures, increased sub G0G1 and decreased S phase cell numbers were measurable in the combination-treated ALL cultures (Fig. 6c, right panel, combination samples).

In addition, we measured apoptosis using FACS and staining for Annexin V and PI for the single and combination drug treatments. Fig. 6d shows that, compared to cells grown on OP9 stroma (right panel), as expected, the percentage cells that were apoptotic in the absence of stromal cells (left panel) was higher. Without stroma, single and combination treatment resulted in increased apoptosis, but the effect of combination treatment was most prominent when the BP-ALL cells were protected by the OP9 cells (Fig. 6d, right panel).

Since PTX008 inhibited the proliferation of ALLs, we next investigated if PTX008 treatment could increase the cytotoxic effect of vincristine in an in vivo NSG transplantation model. We used the diagnosis BP-ALL LAX57 based on its relatively high Galectin-1 expression and sensitivity to vincristine treatment in vitro. After transplant, BP-ALL cells were allowed to proliferate for a one-week period. We then started treatment with PTX008 alone, vincristine alone, or a combination of PTX008 and vincristine for a total of 28 days and monitored body weight of the animals over the treatment period (Additional file 1: Figure S4a and b). At the end of the treatment, mice were euthanized and bone marrow, blood and spleen analyzed for leukemia burden by FACS for human CD19. We also examined expression of CD11c and Ly-6C as markers for murine myeloid cells, and CD11b as leukocyte marker, to address the possibility that inhibition of endogenously produced Galectin-1 activity could affect murine hematopoiesis. As shown in Fig. 7a, c, there was no significant difference between any of the treatments with respect to myeloid markers, or for leukemia burden in blood or bone marrow. This indicates that vincristine monotreatment was not able to control proliferation of these BP-ALL cells in vivo. Combination-treated mice did have the lowest spleen weight, which corresponded with a significantly (p≤0.001) reduced percentage of CD19+ cells in the spleen compared to PTX008 only treated mice (Fig. 7b, d).