Small-Molecule Inhibitor of Glycogen Synthase Kinase 3β 6-Bromoindirubin-3-oxime Inhibits Hematopoietic Regeneration in Stem Cell Recipient Mice

Abstract

Small-molecule inhibitors of glycogen synthase kinase 3β (GSK3β) have demonstrated strong anti-leukemia effects in preclinical studies. Here, we investigated the effect of GSK3β inhibitor 6-Bromoindirubin-3-oxime (BIO) previously shown to inhibit leukemia cell growth in vitro and of animal models on hematopoietic regeneration in recipients of stem cell transplant. BIO administered to immunocompromised mice transplanted with human hematopoietic stem cells inhibited human stem cell engraftment in the bone marrow (BM) and peripheral blood. BIO reduced CD34⁺ progenitor cells in the BM, and primitive lymphoid progenitors re-populated host thymus at later stages post-transplant. The development of all T-cell subsets in the thymus was suppressed in BIO-treated mice. Human cell engraftment was gradually restored after discontinuation of BIO treatment; however, T-cell depletion remained until the end of experiment, which correlated with the attenuated thymic function in the host. BIO delayed CD34⁺ cell expansion in stroma-supported or cytokine-only cultures. BIO treatment delayed progenitor cell divisions and induced apoptosis in cultures with sub-optimal cytokine support. In addition, BIO inhibited B- and T-cell development in co-cultures with MS5 and OP9-DL1 BM stroma cells, respectively. These data suggest that administration of GKS3β inhibitors may act to delay hematopoietic regeneration in patients who received stem cell transplant.

Introduction

Glycogen synthase kinase 3β (GSK3β) is a serine/threonine kinase that regulates a number of signaling pathways in the cell and has been shown to play a pivotal role in hematopoiesis [1 –3]. Inhibition of GSK3β in preclinical murine models of acute myeloid leukemia and acute lymphoblastic leukemia provided promising evidence of efficacy and identified GSK3β as a candidate cancer drug target for different types of leukemia [4 –8]. We have demonstrated that a novel potent inhibitor of GSK3β, 6-Bromoindirubin-3-oxime (BIO), suppressed cell growth in leukemia cell lines of diverse origin and, more importantly, primary leukemia samples [8]. Pretreatment with BIO reduced human leukemia stem cell engraftment, and in vivo administration of BIO delayed tumor formation in a mouse leukemia model [8]. Our data supported further evaluation of BIO as a promising novel agent for therapeutic intervention in leukemia.

Allogeneic stem cell transplantation (allo-SCT) is the only curative therapy for patients with high risk or recurrent leukemia. Leukemia relapse after allo-SCT, however, occurs in 30% of patients [9]. Small-molecule inhibitors of GSK3β such as BIO can be used as a second-line treatment for the allo-SCT-transplant patients resistant to conventional chemotherapy. In addition to its anti-leukemia effect, administration of BIO was shown to prevent acute graft versus host disease (aGVHD) in the xenograft mouse model [10]. Here, we have examined the effect of BIO on regeneration of human hematopoiesis in immunocompromised mice transplanted with human hematopoietic stem cells. We have previously reported that cord blood-derived CD34⁺ stem cells rapidly reconstitute human multilineage hematopoiesis in sub-lethally irradiated NOD-SCID IL2Rγ^−/− (NSG) mice [11,12]. Administration of BIO at early stages post-transplant inhibited human cell engraftment in the bone marrow (BM) and peripheral blood (PB). BIO treatment reduced the numbers of primitive CD34⁺ progenitor cells in the BM and the numbers of CD34⁺ progenitors re-populated in the host thymus at later stages post-transplant. T-cell development was severely suppressed in BIO-treated mice. Human cell counts in PB were gradually restored after discontinuation of BIO-treatment, while CD3⁺ T-cell counts remained low until the end of the experiment, which correlated with the reduced thymic function in the host mice. BIO given at later stages post-transplant still reduced human cell reconstitution in the BM but did not affect T-cell counts in PB. It is relevant, that at late stages post-transplant, T-cell balance in PB does not depend on thymopoiesis, but rather is associated with the homeostatic expansion of mature T cells [13]. BIO suppressed hematopoietic progenitor cell proliferation in stroma- and cytokine-only supported cultures, suggesting progenitor cell autonomous response to GSK3β inhibition independent of stroma. BIO suppressed B- and T-cell development in stem cells co-cultured with BM stroma MS5 and OP9-DL1 cells, respectively. Collectively, our results show that GSK3β inhibitors, before being implemented in clinics, should be carefully tested, particularly in conjunction with stem cell transplantation and adoptive T-cell immunotherapy.

Materials and Methods

Human cell processing and culture

Umbilical cord blood (UCB) for research was provided by Sydney Cord Blood Bank. PB for research was provided by Sydney Red Cross. All experiments were approved by the Sydney Children's Hospital Human Research Ethics Committee. Mononuclear cells (MNCs) were purified from UCB using Lymphoprep (Axis-Shield). CD34⁺ cells were isolated from UCB to greater than 95% purity by positive selection using a magnetic cell purification system (autoMACS; Miltenyi Biotec).

BM transplantation model

Approval from the UNSW Animal Care and Ethics Committee was obtained for animal experiments. NOD-SCID IL2Rγ^−/− (NSG) and NOD/SCID mice were obtained from the Jackson Laboratory and bred in the animal facility at Children's Cancer Institute Australia for Medical Research. Mice at 6 to 8 weeks of age were irradiated with a sub-lethal dose of 2.5 Gy from a Cobalt-60 source for 16–24 h before being transplanted with UCB cells via intravenous injection. UCB CD34⁺ cells were inoculated at a cell dose of 1×10⁵ per mouse. Weekly bleeding was performed from the tail vein to monitor human cell engraftment in PB of transplanted mice. Human and murine cells were distinguished by the species-specific expression of the pan-hematopoietic marker CD45. The proportion of cells labeled with anti-human CD45 (hCD45) antibody was used to determine human cell engraftment. Antibodies specific to human CD33, CD19, CD34, and CD3 were used to examine multilineage reconstitution in PB, BM, and spleen at the endpoint of the experiment. PB, BM, and spleen were collected, and cell suspensions were subjected to red blood cell lysis (FACS lyse; BD). Human leukocyte populations were analyzed by direct immunofluorescence using PE, FITC, or APC-conjugated antibodies purchased from BD Pharmingen. Mouse IgG1 isotype antibodies were used as controls. 1×10⁵ events were acquired on an FACS Canto, and data were analyzed with FACSDiva software (Becton Dickinson). Viable cells were gated based on forward versus side scatter. BIO or DMSO were injected at either early or late stages post-transplant (see the text). BIO (Sigma) was dissolved in DMSO at a concentration of 10–25 mg/mL and administered to mice via intraperitoneal injection. BIO was administered at a dose indicated in the text at different stages post-transplant. Mice that received an equivalent amount of DMSO-only treatment were used as a control.

To analyze the effect of BIO on thymopoiesis, sub-lethally irradiated mice were transplanted with human stem cells (10⁵ CD34⁺ cells per mouse), and seven consecutive daily injections of BIO (10 mg/kg) or DMSO were administered beginning from week 5 post-transplant. Analysis of thymus was performed the next day after the last BIO injection was administered. Human CD45⁺, hCD34⁺, CD34⁺CD7⁺ T-cell precursors, double-positive CD4⁺CD8⁺, single-positive (SP) CD4⁺, and CD8⁺ and TCR⁺ T cells were measured in the thymocytes using flow cytometry.

To analyze the effect of BIO on proliferation and apoptosis of human hematopoietic stem/progenitor cells, 11 week-old female NOD/SCID mice were irradiated and transplanted with UCB CD34⁺ cells as described earlier. Seven consecutive daily injections of BIO (10 mg/kg) or DMSO were administered beginning from week 5 post-transplant. In vivo BrdU labeling of BM cells was done the next day after the last BIO injection by an intraperitoneal injection of BrdU (BD Pharmingen) at 1.5 mg/150 μL of PBS per mouse at 1 h before BM cells harvest. BM cells were then stained for BrdU, 7-AAD, and Annexin V-PE (BD Pharmingen). At least 30×10⁵ events were acquired on an FACS Canto, and data were analyzed with FACSDiva software (Becton Dickinson).

In vitro culture of CD34⁺ progenitor cells

CD34⁺ cells were seeded at a density of 1–2×10⁵cells/mL in Iscove's modified Dulbecco's medium (IMDM; Invitrogen) with 10% heat-inactivated fetal calf serum (FCS; Invitrogen), 100 U/mL penicillin/streptomycin, and a cytokine cocktail of Flt3 ligand (Flt-3L), stem cell factor (SCF), and thrombopoietin (TPO), all of which were at a concentration of 100 ng/mL (R&D Systems). The same cytokine combination was used in suspension cultures. CD34⁺ cells were exposed to 0.25, 0.5, and 1 μM BIO or a matched concentration of 1-methyl-6-bromoindirubin-3′oxime (MeBIO; Merck KGaA), or in some experiments a matched concentration of DMSO (see text for details). Placenta-derived mesenchymal stem cells (MSCs) were obtained and used as previously described [12]. To perform high-resolution division tracking, CD34⁺ cells were stained with carboxyfluorescein in diacetate, succinimidyl ester (CFSE; Invitrogen) and re-sorted as previously reported [12]. Flow cytometric analysis was performed on day 5 post-staining.

Colony-forming unit assay

UCB-derived CD34⁺ cells were plated in MethoCult GF H4434 [1% methylcellulose in IMDM, 30% fetal bovine serum (FBS), 1% bovine serum albumin (BSA), 10⁻⁴ M 2-mercaptoethanol, 2 mM glutamine, 50 ng/mL SCF, 10 ng/mL granulocyte-macrophage colony-stimulating factor, 10 ng/mL interleukin (IL)-3, supplemented with 3 U/mL of erythropoietin (Stem Cell Technologies)]. 2×10⁴ cells suspended in 1.0 mL of MethoCult were plated in a 35-mm dish in triplicate, and cultured for 11 days. Colony-forming units were scored on day 11.

In vitro analysis of B-cell development

B-cell development was assayed using the CD34⁺ cells (1,000 cells per well)/murine MS5 stromal cell (2×10³ per well) culture in 96-well plates with Dulbecco modified Eagle medium-10% FCS supplemented with human granulocyte-colony stimulating factor (10 ng/mL) and SCF (10 ng/mL) as previously described [14]. CD34⁺ and CD19⁺ cell subsets were analyzed once weekly within 4 weeks. Total nucleated cell numbers and the proportion of CD34⁺ progenitor cells and CD19⁺ B cells were measured using flow cytometry.

In vitro analysis of T-cell development

UCB- and mobilized PB-derived CD34⁺ cells were co-cultured with OP9 or OP9-DL1 cells for 2 weeks as previously reported [15]. Cell were growing in alpha-minimal essential medium (Invitrogen) supplemented with 20% FCS, 50 U/mL penicillin/streptomycin (Invitrogen). Flt-3L, 5 ng/mL and interleukin-7 (IL-7, 5 ng/mL; PeproTech) were added to the cultures first on day 0 and then every subsequent 3 to 4 days during medium changes. BIO or DMSO were added to the cultures at different time points as indicated in the figure legends. Duplicate cultures were performed for control and BIO-treated cells. Two experiments of each kind were performed using CD34⁺ cells isolated from different UCB donors. CD34⁺ progenitor cells, CD34⁺CD7⁺ precursor cells, and CD4⁺CD8⁺ double-positive (DP) cells were analyzed using flow cytometry at different time points.

Statistical analysis

Results are expressed as mean±SD. Differences between groups were examined for statistical significance using Student's t-test.

Results

In vivo administration of BIO delayed human cell engraftment in the BM and PB and inhibited T-cell development in mice transplanted with human hematopoietic stem cells

Sub-lethally irradiated NSG mice transplanted with UCB-derived CD34⁺ progenitor cells efficiently reconstituted human hematopoiesis. Human CD45⁺ (hCD45⁺) cells were detectable in PB at week 6 poststem cell infusion, and the proportion of hCD45⁺ cells progressively increased, reaching 45% at week 21 post-transplant (Fig. 1A, B). The majority of hCD45⁺ expressed CD19 with the fraction of cells expressing immature B-cell marker CD10 (Fig. 1C). In experiment 1, BIO, 30 mg/kg, was administered twice a week beginning from day 1, and treatments were discontinued at 3.5 weeks post-transplant (seven injections in total were given to each mouse). DMSO was administered as a vehicle control. It is relevant that BIO administered at a similar schedule exhibited strong anti-leukemia effects and prevented aGVHD in xenograft mouse models [8,10]. Mice treated with BIO exhibited significantly reduced hCD45⁺ cell counts in PB from week 4 to week 12 post-transplant (Fig. 1A). By week 17, 13 weeks after discontinuation of BIO treatment, the proportion of hCD45⁺ cells in PB of BIO-treated mice reached the same level as in control mice (Fig. 1A). Human CD3⁺ T cells were readily detectable in PB of mice infused with human stem cells at week 8 post-transplant, and their numbers progressively increased in control mice, reaching more than 20% of hCD45⁺ cells at week 24 (Fig. 1B, D). CD3⁺ cell counts were significantly lower in BIO-treated mice, and T-cell counts remained low after discontinuation of BIO treatment (Fig. 1D).

FIG. 1.

In vivo administration of BIO delays human cell engraftment in mouse BM transplantation model. Sub-lethally irradiated mice were transplanted with human UCB CD34⁺ progenitor cells. Seven injections of BIO, 30 mg/kg, or DMSO were administered to mice (n=5 per group). BIO was administered twice weekly within first 3.5 weeks post-transplant. (A) Mice treated with BIO exhibited reduced proportion of hCD45⁺ cells in PB at week 4, 6, and 12 (*p<0.05). The percentage of hCD45⁺ cells did not significantly differ at week 17 and 21 post-transplant. (B) Human CD45⁺ and CD3⁺ cell analysis in PB of the mice transplanted with human stem cells. One representative flow cytometry analysis is shown. (C) The majority of hCD45⁺ cells in PB co-expressed CD19. The significant fraction of CD19⁺ B cells co-expressed immature B cell marker CD10. (D) Delayed hCD3⁺T-cell engraftment in mice administered with BIO. (E) Postmortem analysis of the BM was conducted at week 23 post-transplant. The proportion and absolute numbers of hCD45⁺ cells in the PB, spleen (SPL), and BM is shown in the top left and right panels. The proportion of CD34⁺ cells in the BM (left bottom panel) and hCD45⁺, CD19⁺, CD33⁺, and CD34⁺ cell numbers in the BM (left and middle bottom panels) are shown. One representative flow cytometry analysis of CD34⁺ cells in the BM is shown in the right bottom panel. (F) Mice were injected with BrdU to identify cycling cells. BrdU was added when seven daily injections of BIO or DMSO, 10 mg/kg, were completed. Three mice were injected with DMSO, and four mice were injected with BIO. The proportion of FITC⁺/BrdU⁺ cells in hCD45⁺ cells was significantly reduced in mice treated with BIO (top panel). BIO-treated mice exhibited a higher percentage of Annexin V⁺ apoptotic cells (bottom panel). BIO, 6-Bromoindirubin-3-oxime; BM, bone marrow; PB, peripheral blood; UCB, umbilical cord blood.

BM analysis performed at 5 weeks after discontinuation of BIO treatment revealed no differences in hCD45⁺, CD19⁺, and CD33⁺ cell numbers between control and BIO mice and did not affect total cellularity of the BM, suggesting that the host hematopoiesis was not affected by BIO (data not shown).

The suppressive effect of BIO given at early stages poststem cell infusion could be due to inhibited stem cell homing to the BM of the graft recipient mice. To exclude the effect of BIO on stem cell homing to the BM, mice were injected with BIO (10 mg/kg), at 19 weeks post-transplant. The reduced proportion and absolute numbers of hCD45⁺ cells were seen in PB, BM, and the spleen at week 23 post-transplant, although the difference in percentage and numbers between control and BIO-treated mice was significant only in the BM and spleen, respectively (Fig. 1E, top left and middle panels). Given that the suppressive effect of BIO did not depend on the timing of BIO administration relative to stem cell infusion, we concluded that the suppressive effect of BIO is unlikely to be mediated through the reduced BM homing. Significantly reduced percentage and numbers of primitive hCD34⁺ progenitor cells were identified in the BM of BIO mice, suggesting that GSK3β inhibition primarily targets primitive progenitor cells in the BM (Fig. 1E, bottom panels).

The reduced proliferation and apoptosis appear to account for the suppressive effect of BIO on human stem cell engraftment in the BM. This was shown in the experiment where mice were injected with BrdU to identify the dividing cells. BrdU was added when seven daily injections of BIO, 10 mg/kg, were completed. The proportion of BrdU⁺ cells in hCD45⁺ cells was significantly reduced in mice treated with BIO (Fig. 1F, top panel). In addition, BIO-treated mice exhibited a higher percentage of Annexin V⁺ apoptotic cells (Fig. 1F, bottom panel).

Gene expression analysis performed on BM samples using real-time (PCR) revealed the reduced expression of cyclin D1 in BIO-treated mice (Fig. 2A). Expression of the cyclin-dependent kinase inhibitor (cdki) p21 was not modulated by BIO (Fig. 2A). In addition, we have previously shown that growth-suppressive effect of BIO on CD34⁺ cell cultures correlates with upregulation of cdki p57 [12]. P57 expression was not found to be modulated in BIO mice when analyzed postmortem at 5 days after completion of the treatment (data not shown). The role of p57 in mediating growth-suppressing effects in vivo, however, cannot be completely excluded. We speculate that p57 expression can be modulated at earlier time points after administration of BIO. Bcl2 protein expression previously shown to be downregulated by BIO in leukemia cells [8] was reduced, and the expression of pro-apoptotic protein bax expression was not modulated by BIO (Fig. 2B). Thus, downregulation of cyclin D1 expression may account for anti-mitogenic effect of BIO in BM immature progenitor cells.

FIG. 2.

Administration of BIO modulates the levels of cyclin D1mRNA and bcl2 but not bax protein in the BM of stem cell-reconstituted mice (the same experiment as it is shown in Fig. 1E). (A) Real-time polymerase chain reaction analysis was performed using total RNA isolated from the BM cells. Human-specific primers were used to analyze the expression of cyclin D1 (Hu-cyclinD1), cyclin-dependent kinase inhibitor p21 (Hu-p21), and β-actin (Hu-actin). Cord blood-derived CD34⁺ cells previously shown to express high levels of p21 and detectable levels of cyclin D1were used as a positive control; mRNA-free samples were used as negative controls. Hu-cyclin D1 and Hu-p21 expression was normalized to β-actin mRNA expression. (B) Western blot analysis of bcl2 and bax expression in human cells repopulated mouse BM. Analysis was performed using the lysates from the BM derived from mice from the experiment described in Fig. 1D. Cells from five mice per condition were mixed before the lysis. Human bcl2 and bax-specific antibodies were used to determine protein expression.

BIO suppressed thymopoiesis in mice transplanted with human CD34⁺ progenitor cells

The proportion of primitive CD34⁺ progenitor cells homed to the BM re-populates host thymus and gives rise to human T cells [13]. Human CD34⁺ progenitor cells acquire CD7 expression before losing CD34 expression and developing double-positive CD4⁺CD8⁺ and SP T-cell subsets [13]. The peak of thymopoiesis was registered at week 5–12 post-transplant [13]. Here, we examine the effect of BIO treatment on thymopoiesis in mice transplanted with CD34⁺ cells (10⁵ per mouse). Mice received seven consecutive daily injections of BIO (10 mg/kg), beginning from week 5, and were analyzed at week 7 post-transplant. BIO treatment reduced the numbers of hCD45⁺, hCD34⁺, and CD34⁺CD7⁺ T-cell precursors as well as SP CD4⁺ cells recovered from the thymus compared with controls (Fig. 3A–D). Human CD3⁺, TCR⁺, double-positive CD4⁺CD8⁺, and SP CD4⁺ and CD8⁺ cells were readily detectable in control but undetectable in BIO-treated mice (data not shown). BM analysis revealed a significant reduction in the proportion and absolute numbers of primitive hCD34⁺ progenitors compared with controls (Fig. 3E). Thus, the reduced numbers of T-cell subsets in the thymus correlate with the reduced numbers of CD34⁺ progenitors in the BM, suggesting that reduced repopulation of the thymus by BM CD34⁺ progenitor cells accounts for the suppression of thymopoiesis in BIO-treated mice.

FIG. 3.

Administration of BIO inhibits thymopoiesis. Sub-lethally irradiated mice were transplanted with human UCB-derived CD34⁺ stem cells (10⁵ CD34⁺ cells per mouse), and were given seven consecutive daily IP injections of BIO (10 mg/kg) or DMSO beginning from week 5 post-transplant. Analysis of thymus was performed the next day after the last BIO injection was administered. Human T-cell subsets were analyzed using flow cytometry. Mice treated with BIO exhibited a dramatic reduction in the numbers of hCD45⁺ (A), CD34⁺ (B), CD7⁺CD34⁺ (C), and SP CD4⁺ thymocytes (D) recovered from thymus, as well as the reduction of hCD34⁺ cells (×10⁶) in BM (E). (F) BIO given at late-stage post-transplant (19 weeks) did not modulate CD3⁺T-cell numbers in PB of the graft recipient mice. SP, single positive.

Interestingly, BIO treatment at late stages post-transplant (19 weeks post-transplant) did not affect T-cell numbers in PB of stem cell-reconstituted mice (Fig. 3F). At this stage of hematopoietic regeneration, homeostatic T-cell proliferation rather than thymopoiesis determines T-cell balance in PB, as the thymic function is reduced at late stages post-transplant due to thymic involution in aging mice [13]. Thus, BIO given at late stages post-transplant does not attenuate T-cell regeneration. In fact, we have recently shown that BIO treatment acts to preserve naive T-cell phenotype in BM reconstituted mice [16]. Thymus-independent T-cell expansion is associated with defective generation of naive T cells and memory T-cell skewing, resulting in decreased diversity in the T-cell repertoire. Thus, late application of the inhibitor may be beneficial for the patient who received stem cell transplant, as it may delay late memory T-cell skewing.

BIO suppressed CD34⁺ progenitor cell proliferation

Post-transplant hematopoiesis depends on regenerative capacity of donor progenitor cells and the supporting capacity of the BM stroma cells that is strongly impaired by preconditioning. We hypothesized that BIO may either inhibit the expansion of donor progenitor cells or/and modulate hematopoiesis supporting capacity of BM stroma cells. Therefore, we have examined the effect of BIO on human CD34⁺ progenitor cells co-cultured with MSCs, and compared its effect on CD34⁺ cells cultured in cytokine-only supported culture. CD34⁺ cells were stained with CFSE to monitor cell divisions as previously described [12]; BIO (0.5 μM) or DMSO were added on day 0 to the cultures, and cell division tracking was performed at day 5. Continuous treatment with BIO inhibited total, CD34⁺, and more primitive CD34⁺CD38⁻ HPC cell expansion (Fig. 4A). BIO but not its inactive derivative MeBIO inhibited CD34⁺ and CD34⁺CD38⁻ progenitor cell expansion in a dose-dependent manner (Fig. 4B).

FIG. 4.

BIO suppressed in vitro proliferation of CD34⁺ HPCs. (A) Addition of BIO (0.5 μM) delayed total, CD34⁺, and CD34⁺CD38⁻ cell expansion. Growth-suppressive effects of BIO were observed in cytokine-only suspension culture and in co-culture with MSC. Two experiments were performed using different cord blood donors. One representative experiment is shown. Each culture condition was done in duplicate. Mean value of two measurements is shown. (B) BIO but not MeBIO inhibited CD34⁺ and CD3⁺CD38⁻ progenitor cell expansion in a dose-dependent manner. (C) BIO inhibited CD34⁺CD38⁻ cell division by reducing the proportion of rapidly dividing cells and reciprocally increasing the proportion of slowly dividing cells (right panel). CFSE analysis was performed in CD34⁺CD38⁻ gate (left panel). (D) BIO inhibits CD133⁺ (P2 gate) and CD133⁻ (P3 gate) cell division, reducing the proportion of rapidly dividing cells and increasing the proportion of slowly dividing cells (P5 and P6 gates, middle and right panels). (E) CD34⁺ cells were expanded in cytokine only-suspension culture for 5 days, re-sorted, and treated with BIO, 0.5 μM, overnight. Cells then were placed in fresh medium and examined on day 11 for apoptosis and cell cycling. BIO treatment did not modulate apoptosis (cells with sub-genomic DNA shown in P4 gate) and cell cycling. (F) UCB CD34⁺ progenitor cells were cultured in the presence of 0.5 μM BIO or DMSO, 5 days. 2×10⁴ cells that include both CD34⁺ and differentiated CD34⁻ cells were plated in 1 mL of MethoCult GF H4434 (1% methylcellulose in IMDM, 30% FBS, 1% BSA, 10⁻⁴ M 2-mercaptoethanol, 2 mM glutamine, 50 ng/mL SCF, 10 ng/mL granulocyte-macrophage colony-stimulating factor, 10 ng/mL IL-3, supplemented with 3 U/mL of erythropoietin) (Stem Cell Technologies). Colonies were scored at day 11. Total number of CFU per each MethoCult culture is presented. Colonies were pooled after scoring, cells were re-suspended in PBS, and the number of cells per colony was determined by dividing the cell numbers per number of colonies. Clonogenic activity of CD34⁺ cells grown in methylcellulose cultures was not affected by the treatment with BIO. (G) BIO induced apoptosis in CD34⁺ cells growing in sub-optimal concentrations of cytokines (20 ng/mL instead of 100 ng/mL)—increased proportion of cells with sub-genomic DNA content (shown in P5) was seen at 48 h after treatment. BSA, bovine serum albumin; CFSE, carboxyfluorescein in diacetate, succinimidyl ester; CFU, colony-forming unit; FBS, fetal bovine serum; IL, interleukin; IMDM, Iscove's modified Dulbecco's medium; MeBIO, 1-methyl-6-bromoindirubin-3′oxime; MSC, mesenchymal stem cell; SCF, stem cell factor.

Cell division tracking demonstrated that BIO delayed CD34⁺CD38⁻ cell division; BIO reduced the proportion of rapidly dividing cells (five and more divisions) and increased the proportion of slowly dividing cells (one to four divisions) (Fig. 4C, right panel). BIO delayed cell division in primitive CD34⁺CD133⁺ and differentiated CD34⁺CD133⁻ cells; BIO reduced the proportion of rapidly dividing cells and increased the proportion of slowly dividing cells in both subsets (shown in P5 and P6 gates for CD34⁺CD133⁺ and CD34⁺CD133⁻ cells, respectively) (Fig. 4D). Importantly, growth-suppressive effects of BIO were observed both in co-cultures with MSCs and in suspension culture without stroma support, suggesting that BIO is likely to directly affect HPCs rather than modulate BM stroma.

When CD34⁺ cells were prestimulated for 5 days with 100 ng/mL cytokines and then treated with BIO (0.5 μM), O/N, no difference in cell cycle profile was observed, and clonogenic activity of plated progenitor cells was not affected by the treatment (Fig. 4E, F). However, when BIO was administered to CD34⁺ cells growing in sub-optimal concentrations of cytokines (20 ng/mL instead of 100 ng/mL), an increased proportion of cells with sub-genomic DNA content was seen at 48 h after treatment (Fig. 4G). Collectively, our results demonstrate that BIO treatment can induce both cytostatic and cytotoxic effects on human primitive hematopoietic progenitor cells.

BIO delays B-cell development in CD34⁺ progenitor cells co-cultured with MS5 stroma cells

The effect of BIO on B-cell development was examined in CD34⁺ cells co-cultured with murine BM stroma MS5 cells [14]. CD34⁺ cells were treated with BIO (1 μM, 5 days), then plated on a confluent MS5 feeder layer, and analyzed at week 1–4 postplating. BIO-treated cells exhibited a significant delay in total cell expansion determined in a nonadherent fraction of co-cultured cells, at week 1 and 2 after plating; at later stages, there were no differences in total cell numbers in control and BIO-treated cell cultures (Fig. 5A). The absolute numbers of CD34⁺ and CD19⁺ cells were reduced in BIO-treated cells, at the first 2 weeks of culture; after that, CD34⁺ cells were hardly detectable in both groups (Fig. 5B, C). The proportion of CD19⁺ cells adherent to MS5 cells fraction was also lower in BIO-treated cells (Fig. 5D). Thus, BIO delays CD34⁺ cell expansion and inhibits B-cell cell development from human CD34⁺ HPCs.

FIG. 5.

BIO delays generation of B cells from CD34⁺ progenitor cells in co-cultures with MS5 stroma cells. CD34⁺ HPCs expanded in suspension culture±DMSO or BIO, both at 0.5 μM, 5 day expansion, were plated on confluent MS5 cells and analyzed once weekly within 4 week period. Duplicate cultures were performed for control and BIO-treated cells. Two experiments were performed using CD34⁺ cells isolated from different cord blood donors. The results from one representative experiment are shown. BIO treatment reduced the numbers of total nucleated (A), CD19⁺ B (B), and CD34⁺ cells (C) recovered from co-cultures with MS5 cells on week 1 and 2 from nonadherent fraction. Reduced proportion of CD19⁺ cells was also registered in adherent fraction of HPCs co-cultured with MS5 cells (D).

BIO delays T-cell development in CD34⁺ progenitor cells co-cultured with OP9-DL1 stroma cells

The effect of BIO on T-cell development was examined in CD34⁺ cells co-cultured with OP9-DL1 cells, supporting T-cell differentiation [15]. BIO was added to the cultures at day 0 at 1, 2, and 5 μM, and equivalent DMSO was used as a vehicle control. Duplicate cultures were performed for control and BIO-treated cells. Two experiments were performed using CD34⁺ cells isolated from different UCB donors. The results from one representative experiment are shown in Fig. 6. Significantly reduced numbers of total CD34⁺ and CD34⁺CD7⁺ precursor cells were recovered from co-cultures with OP9-DL1 cells, at 2 weeks postseeding (Fig. 6A). In addition, BIO treatment reduced the numbers of differentiating T cells when cells were co-cultured with OP9-DL1 stromal layer for 1 month, followed by co-culturing in the presence or absence of 0.1 μM BIO for 7 days and further maintained in co-culture with OP9-DL1 for an additional 70 days. Treatment with BIO reduced the proportion and absolute number of CD4⁺CD8⁺ DP cells (Fig. 6B). BIO treatment of mobilized PB-derived CD34⁺ cells also reduced progenitor cell numbers and delayed T-cell differentiation during co-culture with OP9-DL1 cells. PB-derived CD34⁺ cells (n=3) were exposed to 0.5 μM BIO for the first week (BIO+−) or 2 weeks (BIO++) of co-culture, and cultures were maintained for the next 6 weeks. BIO treatment reduced the number of cells recovered from OP9-DL1 co-culture, and significantly decreased generation of CD7⁺ T-cell precursors (Fig. 6C).

FIG. 6.

BIO delays T-cell development from CD34⁺ progenitor cells in co-cultures with OP9-DL1 cells. (A) UCB-derived CD34⁺ cells co-cultured with OP9 or OP9-DL1 cells for 2 weeks. BIO or DMSO was added to the cultures at day 1 at 1, 2, and 5 μM. Duplicate cultures were performed for control and BIO-treated cells. Two experiments were performed using CD34⁺ cells isolated from different cord blood donors. The results from one representative experiment are shown. Reduced numbers of total CD34⁺ (A) and differentiating CD34⁺CD7⁺ precursor cells were recovered from co-cultures with OP9-DL1 cells at 2 weeks postseeding. (B) BIO treatment of UCB-derived CD34⁺ cells resulted in reduced numbers of CD4⁺CD8⁺ double-positive (DP) cells (n=2) when cells were co-cultured with OP9-DL1 stromal layer for 1 month, followed by co-culturing in the presence or absence of 0.1 μM BIO for 7 days (bar). After exposure to BIO, cells were maintained in OP9-DL1 co-culture for a further 35 days. Treatment with BIO reduced the number of cells recovered and the proportion of DP cells generated. (C) BIO treatment of mobilized PB-derived CD34⁺ cells reduced progenitor cell numbers and delayed T-cell differentiation during co-culture with OP9-DL1 cells. PB-derived CD34⁺ cells (n=3) were exposed to 0.5 μM BIO for the first week (BIO+−) or 2 weeks (BIO++) of co-culture, and cultures were maintained for 6 weeks. BIO treatment reduced the number of cells recovered from OP9-DL1 co-culture (^#), and it significantly decreased the generation of CD7⁺ T-cell precursors (*).

Discussion

Small-molecule inhibitors of GSK3β have demonstrated remarkable anti-leukemia activity, and some of them are approaching clinical trials [3 –7]. Here, we have examined the effect of a small-molecule inhibitor of GSK3β, BIO, previously shown to produce strong anti-leukemia effects, both in vitro and in animal models [8], on hematopoietic regeneration in immunocompromised mice transplanted with human stem cells. We have earlier demonstrated that BIO efficiently inhibits GSK3β in human CD34⁺ progenitor cells both in vitro and in vivo [5,7,8,10,12, 16]. Here, we show that BIO reduced human cell engraftment in PB and BM. The discontinuation of treatment, however, resulted in gradual restoration of human cell counts. Importantly, BIO treatment reduced hCD34⁺ cell expansion in the BM and suppressed CD34⁺ progenitor cell expansion in co-cultures with MSC. Since the suppressive effect of BIO was also observed in cytokine-only cultures when stroma support was removed, we have suggested that BIO directly modulates CD34⁺ progenitor cells rather than BM stroma. Reduced proportion of BrdU⁺ and increased proportion of AnnexinV⁺ apoptotic cells were identified in the BM of mice treated with BIO, suggesting that inhibited proliferation of donor progenitor cells and apoptosis may account for the reduced human cell engraftment.

BIO reduced the numbers of total CD34⁺ and CD34⁺CD38⁻ progenitor cells and delayed cell division of primitive CD34⁺CD133⁺ and differentiating CD133⁻ cells. The effect of BIO was dose dependent and correlated with GSK3β inhibition, as kinase inactive derivative MeBIO did not suppress the expansion of hematopoietic progenitor cells. Mechanistically, BIO targets cyclin D1 in BM progenitor cells, and it downregulates cyclin D1 in cytokine-stimulated human CD34⁺ progenitor cells [12]. We have previously shown that the growth-suppressive effect of BIO in CD34⁺ cell cultures correlates with upregulation of cyclin-dependent kinase inhibitor p57 and downregulation of cyclin D1 [12]. Although p57 expression was not found to be modulated in BIO mice (data not shown) when analyzed postmortem 5 days after completion of the treatment, it cannot be excluded; we speculate that p57 expression can be modulated at earlier time points after administration of BIO.

We have earlier shown that BIO treatment upregulates β-catenin in hematopoietic progenitor cells both in vitro and in vivo in the spleen of mice transplanted with human MNCs [7,10]. The role of β-catenin/Wnt signaling in hematopoiesis has been quite controversial because of conflicting results in various gain- and loss-of-function studies [17]. Wnt is regulated in a dose-dependent manner at key checkpoints in various lineages of the hematopoietic system [18]. It was recently shown that the most primitive hematopoietic progenitor cells tolerate only relatively low levels of Wnt signaling [18]. In addition, it is consistent with the data showing that Wnt signaling induces stem cell quiescence in homeostatic hematopoiesis [19]. Thus, our data showing growth-suppressive effect of BIO is consistent with these data.

Our data, however, contradict the results from Trowbridge et al. showing that in vivo administration of another small-molecule inhibitor of GSK3β, CHIR-911 (Chiro), acts to promote human cell repopulation in NOD/SCID mice [20]. The authors admitted, however, that Chiro suppressed stem cell proliferation in their in vitro cultures similar to BIO. It is relevant that Chiro and BIO have similar IC50, and both were used at 30 mg/kg. It is unlikely, although quite possible, that different mouse strains (NOD/SCID in Bhatia's study and NSG in our study) may account for the differences in recorded results. The delivery of the inhibitors, however, was different—Chiro was formulated in captisol and injected intraperitoneally; therefore, it was likely to provide a gradual release of the compound [20]. BIO was administered by an intravenous injection of the whole dose at once. We suggest that a gradual release of Chiro in [20] may prevent cytostatic effects known to be characteristic for both inhibitors, in addition to GSK3 inhibition. It is relevant that BIO inhibits IC50 cdk activity at approximately a 60-fold higher dose [21].

We have previously shown that human CD34⁺ cells pretreated (o/n) with BIO ex vivo before transplantation did not produce a cytostatic effect in vitro on CD34⁺ progenitor cells and, in fact, promoted stem cell engraftment [7]. Moreover, we have recently shown that pretreatment with BIO acts to promote short-term engraftment in NSG mice [22]. Thus, our data correlate with the results from Trowbridge et al. [20]. The engraftment-promoting effect of BIO pretreatment, however, was abrogated after continuous 5-day treatment with BIO ex vivo before transplantation. In this case, BIO mediated progenitor cell quiescence [7,12]. Thus, it appears that giving BIO at different experimental settings produces different effects. Further analysis of the exact mechanism/s triggered by BIO that are responsible for its growth-suppressive effects is needed.

BIO treatment suppressed generation of immature B cells in the BM and PB and suppressed thymopoiesis, reducing the numbers of early lymphoid CD34⁺CD7⁺ progenitor cells as well as more mature human T-cell subsets, including double-positive CD4⁺CD8⁺ and SP CD4⁺, CD8⁺, and TCRαβ⁺CD3⁺ T cells. BIO suppressed B- and T-cell development in vitro when primitive CD34⁺ progenitor cells were co-cultured with BM stroma MS5 or OP9-DL1 cells, respectively. In both cultures, BIO-mediated suppression primarily targeted CD34⁺ progenitor cells. The reduced production of immature B and T cells is likely to result from the CD34⁺ progenitor cell suppression. Collectively, our in vivo and in vitro results suggest that the reduced CD34⁺ progenitor cell expansion in the BM of BIO mice restricts their colonization of the thymus, and thymic involution at late stages post-transplant in aging mice prevents T-cell regeneration in BIO-treated mice after discontinuation of the treatment.

Importantly, T-cell numbers in PB of mice that received BIO treatment at late stages post-transplant (19 weeks post-transplant) were not affected by the treatment. Homeostatic T-cell proliferation rather than thymopoiesis determines T-cell balance in PB at late stages post-transplant [13]. Thus, BIO given at late stages post-transplant does not attenuate T-cell regeneration, but, in fact, produces a beneficial effect on mature T cells preserving a naive T-cell phenotype [16]. Mature T-cell expansion in PB is associated with memory T-cell skewing, resulting in decreased diversity in the T-cell repertoire, thus attenuating the immune responses. Thus, administration of BIO as an anti-leukemia agent may be beneficial for the patient who received stem cell transplant, as it may delay late memory T-cell skewing and improve T-cell immune responses in addition to its direct anti-leukemia effect.

Footnotes

Acknowledgments

The authors would like to thank Children's Cancer Institute Australia for Medical Research for supporting their research. They are grateful to Sydney Cord Blood Bank for providing UCB samples for their research. In addition, they would like to acknowledge support from Financial Markets Foundation for Children, Inner Wheel Foundation, St Vincent's Clinic Foundation, John Kirkpatrick Research Fund, and Maple-Brown Family Charitable Foundation.

Author Disclosure Statement

The authors declare that there are no conflicts to disclose and no competing financial interests exist.

References

Rayasam

, Tulasi

, Sodhi

, Davis

and Ray

. (2009). Glycogen synthase kinase 3: more than a namesake. Br J Pharmacol, 56:885–898.

Frame

and Cohen

. (2001). GSK3 takes centre stage more than 20 years after its discovery. Biochem J, 359:1–16.

Huang

, Zhang

, Bersenev

, O'Brien

, Tong

, Emerson

and Klein

. (2009). Pivotal role for glycogen synthase kinase-3 in haematopoietic stem cell homeostasis in mice. J Clin Invest, 119:3519–3529.

Ougolkov

and Billadeau

. (2008). Inhibition of glycogen synthase kinase-3. Methods Mol Biol, 468:67–75.

Holmes

, O'Brien

, Knight

, Lindeman

, Symonds

and Dolnikov

. (2008). The role of glycogen synthase kinase-3β in normal haematopoiesis, angiogenesis and leukaemia. Curr Med Chem, 15:1493–1499.

Wang

, Smith

, Murphy

, Piloto

, Somervaille

TCP

and Cleary

. (2008). Glycogen synthase kinase 3 in MLL leukaemia maintenance and targeted therapy. Nature, 455:1205–1209.

Holmes

, O'Brien

, Knight

, Lindeman

, Shen

, Song

, Symonds

and Dolnikov

. (2008). Glycogen synthase kinase-3β inhibition preserves hematopoietic stem cell activity and inhibits leukemic cell growth. Stem Cells, 26:1288–1297.

Song

, Palladinetti

, Klamer

, Ko

K-H

, Lindeman

, O'Brien

and Dolnikov

. (2010). Glycogen synthase kinase-3β inhibitors suppress leukaemia cell growth. Exp Haematol, 38:908–921.

Gooley

, et al. (2010). Reduced mortality after allogeneic hematopoietic-cell transplantation. N Engl J Med, 363:2091–2101.

10.

Klamer

, Shen

, Song

, Rice

, Knight

, Lindeman

, O'Brien

and Dolnikov

. (2013). GSK3 inhibition prevents lethal GVHD in mice. Exp Hematol, 41:39–55.

11.

Holmes

, Yan

, Ko

K-H

, Nordon

, Song

, O'Brien

and Dolnikov

. (2012). Ex vivo expansion of cord blood progenitors impairs their short-term and long-term repopulating activity associated with transcriptional dysregulation of signalling network. Cell Prolif, 45:266–278.

12.

K-H

, Holmes

, Palladinetti

, Song

, Nordon

, O'Brien

and Dolnikov

. (2011). GSK-3β inhibition promotes engraftment of ex vivo expanded hematopoietic stem cells and modulates gene expression. Stem Cells, 29:108–118.

13.

Parietti

, Nelson

, Telliam

, et al. (2012). Dynamics of human prothymocytes and xenogeneic thymopoiesis in hematopoietic stem cell-engrafted nonobese diabetic-SCID/IL-2rγnull mice. J Immunol, 189:1648–1660.

14.

Hirose

, Kiyoi

, Itoh

, Kato

, Saito

and Naoe

. (2001). B-cell precursors differentiated from cord blood CD34⁺ cells are more immature than those derived from granulocyte colony-stimulating factor-mobilized peripheral blood CD34⁺ cells. Immunology, 104:410–417.

15.

La Motte-Mohs

, Herer

and Zuniga-Pflucker

. (2005). Induction of T-cell development from human cord blood hematopoietic stem cells by Delta-like 1 in vitro. Blood, 105:1431–1439.

16.

Shen

, Klamer

, Xu

, O'Brien

and Dolnikov

. (2013). GSK-3β inhibition preserves naive T cell phenotype in bone marrow reconstituted mice. Exp Hematol, 41:1016–1027.

17.

Lento

, Congdon

, Voermans

, Kritzik

and Reya

. (2013). Wnt signaling in normal and malignant hematopoiesis. Cold Spring Harb Perspect Biol, 5:pii: a008011.

18.

Luis

, Ichii

, Brugman

, Kincade

and Staal

. (2012). Wnt signaling strength regulates normal hematopoiesis and its deregulation is involved in leukemia development. Leukemia, 26:414–421.

19.

Fleming

, Janzen

, Lo Celso

, Guo

, Leahy

, Kronenberg

and Scadden

. (2008). Wnt signaling in the niche enforces hematopoietic stem cell quiescence and is necessary to preserve self-renewal in vivo. Cell Stem Cell, 2:274–283.

20.

Trowbridge

, Xenocostas

, Moon

and Bhatia

. (2006). Glycogen synthase kinase-3 is an in vivo regulator of hematopoietic stem cell repopulation. Nat Med, 12:89–98.

21.

Meijer

, Skaltsounis

, Magiatis

, Polychronopoulos

, Knockaert

, Leost

, Ryan

, Vonica

, Brivanlou

, et al. (2003). Greengard PGSK-3-selective inhibitors derived from Tyrian purple indirubins. Chem Biol, 10:1255–1266.

22.

Dolnikov

, Xu

, Shen

, Song

, Holmes

, Klamer

and O'Brien

. (2014). GSK-3β inhibition promotes early engraftment of ex vivo-expanded haematopoietic stem cells. Cell Prolif, 47:113–123.

Small-Molecule Inhibitor of Glycogen Synthase Kinase 3β 6-Bromoindirubin-3-oxime Inhibits Hematopoietic Regeneration in Stem Cell Recipient Mice

Abstract

Introduction

Materials and Methods

Human cell processing and culture

BM transplantation model

In vitro culture of CD34+ progenitor cells

Colony-forming unit assay

In vitro analysis of B-cell development

In vitro analysis of T-cell development

Statistical analysis

Results

In vivo administration of BIO delayed human cell engraftment in the BM and PB and inhibited T-cell development in mice transplanted with human hematopoietic stem cells

BIO suppressed thymopoiesis in mice transplanted with human CD34+ progenitor cells

BIO suppressed CD34+ progenitor cell proliferation

BIO delays B-cell development in CD34+ progenitor cells co-cultured with MS5 stroma cells

BIO delays T-cell development in CD34+ progenitor cells co-cultured with OP9-DL1 stroma cells

Discussion

Footnotes

Acknowledgments

Author Disclosure Statement

References

In vitro culture of CD34⁺ progenitor cells

BIO suppressed thymopoiesis in mice transplanted with human CD34⁺ progenitor cells

BIO suppressed CD34⁺ progenitor cell proliferation

BIO delays B-cell development in CD34⁺ progenitor cells co-cultured with MS5 stroma cells

BIO delays T-cell development in CD34⁺ progenitor cells co-cultured with OP9-DL1 stroma cells