Hematopoiesis is supported in adults by a stromal microenvironment consisting of mesenchymal stem cells (MSCs) and their descendants that include fibroblast Colony-Forming Unit (CFU-F) progenitors of intermediate maturity and specialized differentiated cells. MSCs are stem cells with the capacity to differentiate into all elements of the stromal microenvironment [1]. More mature descendants of MSCs such as CFU-F are able to maintain hematopoiesis by differentiating into osteogenic and adipogenic lineages [2,3]. The stromal microenvironment has higher radio- and chemoresistance than hematopoietic cells because of its low self-renewal frequency [4].

The stromal microenvironment is often damaged in patients with hematological diseases [5-8]. The following changes in the stromal microenvironment of acute leukemia patients were previously described: disturbances of signaling pathways [9], genetic abnormalities [10-12] and functional changes [13-15]. Alterations in the stromal microenvironment were also observed in chronic leukemia patients [16,17], myelodysplastic syndrome patients [18,8,12,19], and multiple myeloma patients [7]. However, other investigators have not found pathological changes in the stromal microenvironment of acute and chronic leukemia patients [20,21]. The treatments for patients with hematological malignancies include high doses of cytotoxic drugs and allogeneic hematopoietic stem cell transplantation (allo-HSCT). Both chemotherapy and pretransplant conditioning affect stromal progenitor cells [22-25]. A damaged stromal microenvironment may impair hematopoiesis in patients after allo-HSCT [26-28].

Acute lymphoblastic leukemia (ALL) is the most common neoplastic disease in children [29] and is well characterized in this age group. Stromal precursor cells of ALL patients have also been studied and their functional changes were previously identified [14,15]. However, the state of the stromal microenvironment in adult ALL patients has not been studied in detail.

This study examines the elements of the MSCs compartment including the multipotent mesenchymal stromal cells (MMSCs) [30] and CFU-F in patients with ALL before and after allo- HSCT.

ALL patients and healthy donors

This study included 15 ALL patients (Table 1) with the following immunophenotypes: Pro-B (1 patient), Pre-B (4 patients), B common (5 patients), Mature B (2 patients), T (1 patient), and Pro -T (2 patients). The age of the patients ranged from 18 to 39 years (median age 28.6 years). Bone marrow (BM) was obtained during routine diagnostic aspiration after receiving the patients’ informed consent. The routine diagnostic aspiration was performed prior to conditioning and at +30, 60, 90, 120, 180 and 365 days after allo- HSCT.

BM samples were analyzed from 64 healthy donors and included 34 men and 30 women aged from 18 to 59 years (median-34 years). All BM samples were obtained during exfusion for allo-HSCT in the FGBU Hematological Scientific Center after receiving the patients’ signed informed consent. These samples were used as controls. Characteristics of ALL patients are shown in (Table 1).

Table 1: Characteristics of acute lymphoblastic leukemia (ALL) patients

MMSCs cultivation

MMSCs were derived from 5-10 ml of the BM cells. To obtain mononuclear cells the BM was mixed with an equal volume of alpha- ??? (ICN, Costa Mesa, USA) media containing 0.2% methylcellulose (1500 cP, Sigma-Aldrich, St. Louis, USA). After 40 min, the erythrocytes and granulocytes had precipitated and the mononuclear cells remained in suspension. The suspended (upper) fraction was aspirated and centrifuged for 10 minutes at 450 × g.

The cells from the sediment were resuspended in a standard cultivation medium composed of alfa-MEM supplemented with 10% fetal calf serum (Hyclone, Waltham, USA), 2 m? L-glutamine (ICN), 100 U/ml penicillin (Ferein, Moscow, Russia) and 50 mg/ml streptomycin (Ferein). The cells were cultured at 3 × 106 cells per T25 cm2 culture flask (Corning-Costar, Tewksbury, USA). When a confluent monolayer of cells had formed (passage 0), the cells were washed with 0.02% EDTA (ICN) in a physiologic solution (Sigma-Aldrich) and then trypsinized (ICN) with 0.25% trypsin (ICN). The cells were seeded at 4 ? 103 cells per cm2 in culture flasks. The cultures were maintained at 37°C in 5% CO2. MMSCs were cultured until the second passage.

All MMSCs were immunophenotyped as described [31] with the following markers: CD105, CD73, CD45, CD34, CD14, and HLADR using standard protocols. The antibodies were purchased at BD Pharmingen, Franklin Lakes, USA (CD105, CD59, CD73, CD90, CD31, CD34 and CD14), Sigma, St. Louis, USA (CD45, FSP) and DAKO, Glostrup, Denmark (HLA-DR). The level of nonspecific antibody binding was analyzed using a mouse immunoglobulin IgG1 isotype control (BD Pharmingen, Franklin Lakes, USA). To assess viability the cells were stained with 7 -aminoactinomycin D (7-AAD), (Sigma). The analysis was performed on a FACSCanto II (BD Biosciences, Franklin Lakes, USA) and the results were analyzed with BD FACSDiva 6.1.3 software (Franklin Lakes, USA).

Analysis of CFU-F

Mononuclear BM cells were seeded at 10⁶ and 5 × 10⁵ per T25 flask in alpha -MEM with 20% fetal calf serum (Hyclone), 2 mM L- glutamine (ICN), 100 U/ml penicillin (Ferein), 50 mg/ml streptomycin (Ferein) and analyzed after 14 days. The colony count was performed using a microscope (Opton, Oberkochen , Germany) after staining the cells with 1% crystal violet in 20% methanol

RNA isolation and quantitative reverse transcriptionpolymerase chain reaction

Total RNA was extracted from the MMSCs at passage 1 using the standard guanidine isothiocyanate method [32]. The cDNA was synthesized using a mixture of random hexamers and oligo(dT) primers. Gene expression levels were quantified by real-time quantitative PCR (qRT-PCR) using hydrolysis probes (Taqman) on a Rotor-Gene 6000 (Corbett Research, Concorde, USA). The gene-specific primers were designed by the authors and synthesized by Syntol R&D (Moscow, Russia). The primers and probes are provided in (Table 2). The relative gene expression levels were determined by normalizing the expression of each target gene to the levels of BACT (beta- actin) and GAPDH (glyceraldehyde -3- phosphate dehydrogenase) and calculated using the ΔΔCt method [33] for each MMSC sample. The reaction was conducted using the following PCR protocol: initial denature at 95 oC for 10 min, cyclic denaturation at 95 oC for 20 sec, hybridization with primers and template extension at 60 °C for 60 sec.

Table 2: Primers and probes.

Statistical analysis

All values are expressed as the means ± SEM. The data were analyzed using independent Student’s t-tests in Microsoft Excel.

Cell production in MMSC cultures from ALL patients

The total cell production of MMSCs derived from ALL BM prior to allo-HSCT conditioning was not significantly different than BM from healthy donors (Figure 1). At the moment of allo-HSCT, 14 of 15 studied patients were in remission after treatment with the standard protocol ALL-2009 and 1 patient was in relapse. There were 2 patients with Ph+ ALL that received additional tyrosine kinase inhibitor therapy. Thus, the treatment protocol used did not inhibit MMSC proliferation. Notably the total MMSCs cell production from the patient in relapse was substantially reduced. These results are consistent with the data on cell production in MMSC cultures from newly diagnosed ALL patients [34]. However, studies of ALL in children at the time of diagnosis [35] and after chemotherapy [14,15] demonstrated that MMSCs were strongly suppressed. This result might occur because in children the disease developed simultaneously with the formation and growth of the hematopoietic microenvironment. In adults with acute myeloid leukemia (AML) [36], chronic myeloid leukemia (CML) [20] and B-cell chronic leukemia (B-CLL) [13] growth characteristics of MMSCs were not changed.

Figure 1: Alterations in cumulative cell production in MMSCs cultures from ALL patients before and after allo-HSCT. Cumulative MMSC production after 3 passages is presented as the mean ± SEM. The data summarize the results of MMSCs production from 64 BM donors and 15 patients with ALL before and after allo-HSCT. *indicates a significant difference (p < 0.05) between MMSC production after allo-HSCT and MMSC production both from donors and patients before allo-HSCT. Time point “0” indicates MMSC samples obtained before the start of pretransplant conditioning, and time points “30-365” indicate samples of MMSCs derived from BM of ALL patients obtained on corresponding days after allo-HSCT.

The pretransplant conditioning significantly reduced the proliferative potential of MMSCs (Figure 1). There was no recovery of MMSCs proliferation observed during the year after allo- HSCT. There was a significant increase in the time needed to reach a confluent monolayer of MMSCs from ALL patients compared with cultures of control MMSCs. However, in the subsequent passages there were no significant differences in the time to confluence (Table 3).

Table 3: Time needed to achieve confluence of the layer of MMSCs

These data suggested that in the BM of ALL patients the number of MMSCs had been reduced after allo-HSCT. It was previously shown that the total cell production of MMSCs was also reduced in patients with various hematological malignancies after allo-HSCT [31]. Therefore, the observed decline in proliferative potential of MMSCs was not associated with the disease itself and was caused by the pretransplant conditioning. The standard conditioning protocol included busulfan and cyclophosphamide. Studies using an animal model demonstrated that busulfan impaired stromal progenitor cells (CFU-F) and their function was not subsequently restored. However, the mesenchymal stem cells were insensitive to cytostatic treatments [23]. The data indicating that human MMSCs were sensitive to high doses of cytotoxic drugs used in transplantation once again confirmed that the population of MMSCs was not the population of true stem cells. Thus, MMSCs consist of more mature stromal precursor cells [37]. The growth characteristics of MMSCs from ALL patients in remission before allo-HSCT were slightly changed. So one can conclude that the pretransplant conditioning led to prolonged damage to MMSCs proliferation and reduced MMSCs number in the BM of patients.

Gene Expression

We next examined the expression levels of several genes regulating stromal cell division in the MMSCs of ALL patients because of the observed changes in their ability to proliferate. Prior to allo-HSCT, we observed reduced expression levels of FGFR1, FGFR2, PDGFRA and PDGFRB genes (Figure 2 A,B,C,D). However, in some cases the reduction was not significant. After allo-HSCT the expression level of these genes remained significantly reduced. These genes encode receptors for growth factors regulating the proliferation of MMSCs. It has been shown that the inhibition of at least one of these signaling pathways leads to MMSC growth reduction [38]. The reduction of gene expression in MMSCs of ALL patients indicates a possible dysregulation of signal transduction in the FGF2 and PDGF pathways. This result may explain the observed decline in total cellular production of MMSCs. The expression level of FGF2 did not vary significantly between the donors and ALL patients before and after allo-HSCT (Table 4). It is known that IL-1 beta is involved in the growth regulation of stromal progenitor cells [39-41]. Moreover, IL-1 beta is expressed by fibroblasts [42]. The expression level of the IL-1 beta receptor in MMSCs from ALL patients was reduced during the entire observation period. However, the changes were significant only at 60 and 180 days after allo- HSCT (Figure 2E). The expression of IL -1 beta itself was not altered significantly (Table 4).

Figure 2: a,b,c,d,e,f: Relative expression level of genes in MMSCs from donors and ALL patients before and after allo-HSCT. Gene expression was analyzed by qRT-PCR with TaqMan probes. The relative expression level was calculatedusing the ΔΔCt method. The results were normalized according to the expression of BACT and GAPDH. Genes analyzed are shown as follows: A-FGFR1, B-FGFR2, C- PDGFRA, D-PDGFRB, E-IL1R1, F-SDF1. * indicates a significant difference (p < 0.05) between MMSC production before and after allo-HSCT and MMSC production from donors. Time point “0” indicates MMSC samples obtained before the start of pretransplant conditioning, and time points “30-365” indicate samples of MMSCs derived from BM of ALL patients obtained on corresponding days after allo-HSCT.

Table 4: Alterations of relative expression level of gene expression in MMSCs of ALL patients compared to donor’s MMSCs.

A reduction in gene expression of receptors for growth factors, but not growth factors themselves, was identified. This result suggested that autocrine secretion of these growth factors did not determine the proliferative ability of MMSCs from ALL patients. These pathways could be important for the proliferation of MMSCs due to the coincident reduction in receptor expression and cells proliferation. Conversely, the expression of all these genes tended to decrease in MMSCs before allo- HSCT, while total cell production in these cultures was not reduced. Probably the genes we examined were not vital for the proliferation of MMSCs, and there still were some other regulatory mechanisms of their division that had not been included in this study.

The expression level of the chemokine SDF1 was significantly reduced in MMSCs of ALL patients before and after allo-HSCT (Figure 2F). A similar effect was also described in patients with AML [43]. This chemokineplays a key role in the homing of hematopoietic cells to the BM niche [44]. Following autologous and allogeneic HSCT both hematopoietic dysfunction and long-term cytopenia are often observed in patients [45–47]. We predict that these disorders are associated with the changes in the interaction between hematopoietic stem cells and the elements of the stromal microenvironment. These changes may be specifically related to the decrease in SDF1 expression.

Characteristics of CFU-F in ALL patients

The concentration of CFU-F in the BM of ALL patients before allo-HSCT slightly but insignificantly increased (p=0.15) compared to donors (Figure 3). This result might be caused by the mild differences in the mean age as ALL patients which ranged from 18 to 39 years (median - 28.6 years), whereas the donors ranged from 13 to 59 years (median - 34 years). It is known that the CFU-F concentration decreases significantly with the age of the donor [48,49]. Several authors have analyzed BM before chemotherapy and did not find differences in CFU-F concentration between patients and healthy donors [50,51]. Other studies have revealed a decline or complete exhaustion in CFU-F concentration in the BM of patients with acute leukemia [52-54,24]. The same result was observed in mice injected with acute leukemia cells [55,56]. However, it is unclear whether chemotherapy treatment occurred in the human studies and the protocol followed was not described. This information is crucial when analyzing the changes in the stromal microenvironment [25,57,58].

Figure 3: Alterations in CFU-F concentration in BM of donors and ALL patients before and after allo-HSCT * indicates the significant difference (p < 0.05) between MMSCs production after allo-HSCT and MMSCs production both from donors and patients before allo- HSCT. Time point “0” indicates MMSCs samples obtained before the start of pretransplant conditioning and time points “30-365” indicate samples of MMSCs derived from BM of ALL patients obtained on corresponding days after allo-HSCT.

After allo-HSCT the concentration of CFU-F in the BM ALL patients was reduced by more than 10-fold and was not restored during the subsequent year (Figure 3). The results were consistent with the changes in the concentration of CFU-F caused by the transplantation protocol and are likely not reparable in patients more than 5 years old [48]. The effect could be associated with the significant decrease in the expression level of FGFR1 in CFU-F of ALL patients after allo- HSCT (Figure 4A). The autocrine expression of FGF2 was significantly upregulated before allo-HSCT, but it was not reduced significantly after transplantation (Figure 4B). We found a range of expression levels for this gene in the MMSCs from different patients. Due to the low total number of patients we cannot make a straightforward conclusion. The observed effect was consistent with the finding that FGF2 was described as the growth factor for stromal progenitor cells [59,60]. We found the simultaneous decrease in the expression of FGF2 and its receptors in CFU-F cells. This result suggests the reduced concentration of CFU-F in the BM after allo-HSCT was associated with impaired FGF2 signaling. The reduced number of stromal precursor cells could be caused by the changes in their proliferative potential. The decrease in proliferative capacity reflected the loss of undifferentiated status of precursor cells and was accompanied by increased expression of differentiationassociated genes. In this study, we analyzed 2 markers for osteogenic differentiation (BGLAP, SPP1) and one for each of adipogenic (PPARG) and chondrogenic (SOX9) lineages. The expression of SOX9 only was changed in CFU-F from BM of ALL patients before allo-HSCT. We found that SOX9 was increased, but the changes were not significant (Figure 4F). The expression level of this gene remained slightly and nonsignificantly increased up to 120 days after allo-HSCT. The expression then decreased to nearly zero before returning to baseline one year after allo-HSCT. The expression level of osteogenic and adipogenic markers of differentiation tended to increase with time after allo-HSCT (Figure 3 C, D, E). The non-significant changes were associated with a range of values in CFU-Fs from ALL patients. Nevertheless, we suppose a shift towards more differentiated cells had occurred in the population of CFU-F derived colonies after allo-HSCT. It was shown that the conditioning before allo-HSCT was accompanied by the accumulation of reactive oxygen species in the cells of the organism [61,62], which could cause premature “aging” of the cells and tissues [63]. The reduction of the MMSCs and CFU-F number in BM of ALL patients after allo-HSCT suggest the damage to those stromal precursor cells that were attributed to the top of the mesenchymal stem cells hierarchy. It is known that the HSCT patients suffer from disorders of the musculoskeletal system including bone loss and osteoporosis [45,46,64]. The results of this study revealed possible mechanisms of bone marrow stroma damage after allo-HSCT in ALL patients. The number of stromal precursors was reduced and their growth and regeneration were inhibited after allo- HSCT. These changes were accompanied by an increased number of more differentiated progenitors with decreased proliferative potential. The damage to stromal cells by different protocols for treatment of leukemia should be taken into consideration.

Figure 4: Relative expression level of genes in CFU-F derived colonies from donors and ALL patients before and after allo-HSCT. Gene expression was analyzed by qRT-PCR with TaqMan probes. The relative expression level was calculated using the ΔΔCt method. The results were normalized by the expression of BACT and GAPDH. Genes analyzed are shown as follows: A-FGFR1, B-FGF2, C- BGLAP, D-SPP1, E-PPARG, F-SOX9.* indicates a significant difference (p < 0.05) between MMSC production before and after allo-HSCT and MMSC production from donors. Time point “0” indicates MMSC samples obtained before the start of pretransplant conditioning, and time points “30-365” indicate samples of MMSCs derived from BM of ALL patients obtained on corresponding days after allo-HSCT.

The authors thank the staff of the Bone Marrow Transplantation Department. This study was supported by grant from the Russian Foundation for Basic Research (12-04-00457a).