- Email: [email protected]
Impact of lenalidomide on the functional properties of human mesenchymal stromal cells
Experimental Hematology 2012;40:867–876
Impact of lenalidomide on the functional properties of human mesenchymal stromal cells Manja Wobusa, Gwendolin Benatha, Ruben A. Ferrera, Rebekka Wehnerb, Marc Schmitzb,c, Lorenz C. Hofbauerd, Martina Raunerd, Gerhard Ehningera,c, Martin Bornh€ausera,c, and Uwe Platzbeckera a
Medical Clinic and Polyclinic I, University Hospital, Dresden, Germany; bInstitute of Immunology, Medical Faculty, Technical University, Dresden, Germany; cCenter for Regenerative Therapies Dresden, Dresden, Germany; dMedical Clinic and Polyclinic III, University Hospital, Dresden, Germany (Received 8 March 2012; revised 15 May 2012; accepted 11 June 2012)
Objective. Lenalidomide (LEN) has emerged as a promising therapeutic option for the management of various hematologic malignancies. Although its direct mechanisms of action on malignant cells have been studied intensively, its effects on the stromal compartment of bone marrow have not yet been analyzed systematically. Therefore, we investigated whether LEN alters the functional capacity of mesenchymal stromal cells (MSCs) as the main cellular component of the bone marrow microenvironment. In addition to their growth and differentiation characteristics, we focused on the ability of MSC to modulate T-cell function and support hematopoietic stem cells (HSCs). Materials and Methods. Bone marrow–derived MSCs were exposed to LEN (10 mM), and differences in proliferation, phenotype, inhibition of T-cell proliferation, and differentiation capacity were analyzed. A Boyden chamber assay was used to test the migratory potential of HSC toward the conditioned medium of LEN-treated or untreated MSCs, and the stromal cell–derived factor-1 (SDF-1) concentrations in these supernatants were determined by enzyme-linked immunosorbent assay. Results. Treatment of MSCs with LEN did not affect their growth rate, proliferation, osteogenic and adipogenic differentiation potential, or capacity to inhibit T-cell proliferation. However, LEN treatment increased the average of mean fluorescence intensity of CD29 and CD73 by 15 and 22%, respectively. Interestingly, LEN reduced SDF-1 by MSCs by 32% compared to that of control cells. As a functional consequence, the serum-free supernatant of LEN-treated MSCs had a significantly lower potential to induce the directed migration of CD34+ HSCs. Conclusion. LEN can modulate the expression of cell surface molecules and the chemokine secretion of MSCs in vitro. These effects might contribute to the clinical effects of the compound in vivo for patients with hematological malignancies. Ó 2012 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc.
Mesenchymal stromal cells (MSCs) are a major cellular component of the formation and function of the bone marrow (BM) microenvironment. MSCs are the progenitor cells of osteoblasts and represent the hematopoieticsupporting stroma components of the marrow [1]. In addition to their capacity to support hematopoiesis [2,3], MSCs display systemic immunoregulatory and immunosuppressive properties [4,5]. The main functions of MSCs are Offprint requests to: Manja Wobus, Ph.D., Medical Clinic and Polyclinic I, University Hospital Dresden, 01307 Dresden, Germany; E-mail: manja. [email protected]
mediated by chemokines and cytokines in an autocrine or paracrine manner. The immunomodulatory drug lenalidomide (LEN) has been used for the successful treatment of a variety of hematological malignancies. In fact, LEN has been demonstrated to reduce and eliminate the need for red blood cell transfusions in the majority of myelodysplastic syndrome (MDS) patients with del(5q) [6]. Furthermore, it performs several important biological actions, including inhibiting angiogenesis and suppressing the production of proinflammatory cytokines, such as tumor necrosis factor–a [7]. LEN has also been reported to abrogate the stromal cell support of myeloma cells by interfering with
0301-472X/$ - see front matter. Copyright Ó 2012 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.exphem.2012.06.004
868
M. Wobus et al./ Experimental Hematology 2012;40:867–876
cytokines, e.g., interleukin-6 and vascular endothelial growth factor, leading to the growth arrest of neoplastic cells [8]. By contrast, it can augment the proliferation of healthy hematopoietic progenitors [9]. Therefore, LEN is currently under investigation as a treatment for various hematological diseases, including those without direct marrow involvement. These approaches include maintenance strategies with long-term usage of the compound. To obtain detailed insights into the regulatory role of LEN in the stromal compartment, we investigated its effects on healthy MSCs. Our findings of increased expression of CD29 and an inhibited migration of hematopoietic stem cells (HSCs) in response to a LEN-induced alteration of stromal cell–derived factor-1 (SDF-1) secretion provide new insights into the mechanism of the effects of LEN on hematologic diseases.
Material and methods Cell culture MSCs were isolated from BM aspirates that were derived and cultured from healthy donors after receiving informed consent and approval from the local ethics committee as described previously [10]. All experiments were carried out with MSCs derived from passage 2 to 4. The age of the donors was 21 to 52 years (median 5 39 years). LEN (Celgene, Munich, Germany) was dissolved in 10% dimethyl sulfoxide, and the stock solution was stored at 20 C. After being thawed, the sample was protected from light and kept at 4 C for a maximum of 2 weeks. On the basis of multiple myeloma and del(5q) MDS data [11,12], a LEN concentration of 10 mM was chosen for the experiments after conducting a dose-response experiment using concentrations of 0.1, 1,0, 10, and 50 mM. Proliferation MSCs (5 103/well in a 96-well plate) were cultured in the presence or absence of LEN for 5 days, and an MTT assay was performed according to the protocol (Roche Diagnostics GmbH, Mannheim, Germany). The absorbance was measured at 570 and 650 nm. Medium was used as a negative control. Immunophenotypic characteristics of MSCs Flow cytometric analysis was performed with a FACSCalibur (BD Biosciences, Heidelberg, Germany) using BD CellQuest Pro and FlowJo software. At subconfluency, cells were incubated with or without LEN for 5 days and then characterized for the expression of positive (CD29-phycoerythrin [PE], CD73-PE, CD90–fluorescein isothiocyanate [FITC], CD105-FITC, and CD166-PE) and negative MSC markers (CD14-PE, CD34-PE, and CD45-FITC). IgG1-negative control-FITC (AbD Serotec, D€ usseldorf, Germany) and IgG1-PE (BD Biosciences) were used as isotype controls. Anti–CD90-FITC and anti–CD105FITC were bought from AbD Serotec. All other antibodies were purchased from BD Biosciences. Mixed lymphocyte reaction Blood samples were obtained from healthy donors after informed written consent. Peripheral blood mononuclear cells (PBMCs)
were prepared by Ficoll-Hypaque (Biochrom, Berlin, Germany) density centrifugation. PBMCs were pooled from five different donors, irradiated (30 Gy), and used as allogeneic stimulators for nonirradiated responder PBMCs. To analyze the effect of LEN on MSC-mediated inhibition of lymphocyte proliferation, stimulator PBMCs (1 105/well) and responder PBMCs (1 105/well) were cocultured in the presence or absence of thirdparty MSCs (5 103/well) and 10 mM LEN. Before culture, MSCs were harvested from culture flasks by trypsinization and subsequently irradiated (30 Gy). The cells were incubated in 96well plates using RPMI-1640 medium (Biochrom) supplemented with 2 mM L-glutamine, 10 mM sodium pyruvate, 1% nonessential amino acids, 100 mg/mL penicillin, 100 mg/mL streptomycin (all from Biochrom), and 10% fetal calf serum for 7 days. One microcurie of [3H]-thymidine (Hartmann Analytic, Braunschweig, Germany) was added to each well for the last 18 hours of culture. Cells were harvested and [3H]-thymidine incorporation was determined using a beta counter (Wallac, Freiburg, Germany). MSC differentiation assays MSCs in passage 2 were seeded in six-well plates (5 103 MSCs/ cm2), cultured to subconfluency for approximately 4 days in Dulbecco’s modified Eagle’s medium, and subjected to adipogenic (0.5 mM 1-methyl-3-butylisoxanthine, 1 mM dexamethasone, 100 mM indomethacin, and 10 mM insulin) or osteogenic (0.1 mM dexamethasone, 0.2 mM ascorbate-2-phosphate, 10 mM bglycerophosphate) differentiation in the presence or absence of LEN. The start of differentiation with the first appearance of lipid vacuoles was recorded, and adipogenesis was assessed by Oil Red O staining after 21 days. Differentiation was graded according to microscopic analysis as follows: 0 5 no differentiation; 1 5 25% of the culture well contained differentiated cells; 2 5 50% of the culture well contained differentiated cells; 3 5 75% of the culture well contained differentiated cells; 4 5 100% of the culture well contained differentiated cells. Moreover, cells were collected for messenger RNA isolation and quantitative real-time polymerase chain reaction was performed for the FABP4 gene (sense primer: 50 -tac tgg gcc agg aat ttg ac-30 ; reverse primer: 50 -tgg ttg att ttc cat ccc at-30 ) as described recently [13]. The first day on which calcium deposits were observed was documented as a sign of osteogenic differentiation. After 21 days, osteogenesis was determined by van Kossa staining, and the catalytic activity of alkaline phosphatase (ALP) was determined as described previously [14]. The medium was changed twice a week, and LEN was freshly added each time. All reagents for differentiation induction were purchased from Sigma-Aldrich (Hamburg, Germany). Chemokine measurements in cell culture supernatants Standardized 1-mL aliquots from serum-free culture supernatants were collected from untreated MSCs or MSCs treated with 10 mM LEN for 5 days. The protein levels of SDF-1 were quantified using an enzyme-linked immunosorbent assay (R&D Systems GmbH, Wiesbaden, Germany) according to the manufacturer’s protocol. HSC transwell migration assay Purification of CD34þ HSCs and Transwell migration toward the conditioned medium of either untreated or LEN-treated MSCs were performed as described previously [13]. Briefly, 2 105
M. Wobus et al./ Experimental Hematology 2012;40:867–876
869
HSCs were added to the upper chamber in 0.1 mL medium and allowed to migrate for 4 hours at 37 C toward 600 mL conditioned medium through a 5-mm polycarbonate membrane. Afterward, cells in the lower chamber were counted. The CXCR4 antagonist AMD3100 (Sigma) was used at a concentration of 10 mM. Statistical analysis Data were analyzed using the GraphPad Prism statistical PC program (GraphPad, San Diego, CA, USA). Results from at least three independent experiments are presented as the mean 6 standard error of the mean (presented as error bars). Statistical comparisons were performed using either a paired Wilcoxon test or one-way analysis of variance with Bonferroni’s post-test.
Results LEN has no influence on MSC growth and proliferation In the initial experiments, we investigated whether LEN influences the growth characteristics of normal MSCs. The treated cells displayed the same fibroblastoid growth pattern as the control cultures and reached confluency at the same time point (not shown). Proliferation was measured using MTT assay after 5 days of culture in 96well plates. No significant differences were detected between LEN-treated and control MSC, as shown in Figure 1. Expression of CD29 and CD73 can be increased by LEN MSCs were cultured in the presence or absence of LEN for 5 days, and the expression of cell surface molecules was examined by flow cytometry. The expression of hematopoietic markers, such as CD14, CD34, and CD45, was !5% of control levels in all cases (not shown). No differences were detected between treated and untreated cells. The cultures were composed of a homogenous cell population with similar forward/sideward scatter characteristics, and the expression of typical MSC surface antigens, e.g., CD29, CD73, CD90, CD105, and CD166, was observed in 95% to 100% of the cells. The mean fluorescence intensities (MFIs) were different for the molecules, as demonstrated in Figure 2A, where a representative measurement for each antigen is shown. Interestingly, for CD29 and CD73, mean MFIs were significantly increased by LEN treatment (CD29: 453.4 vs 528.4; **p ! 0.005; CD73: 227.2 vs 291.2; *p ! 0.05; Fig. 2B, Table 1). LEN-treated MSCs inhibit the proliferation of lymphocytes In further experiments, we explored the effect of LEN on MSC-mediated lymphocyte proliferation. Therefore, we incubated MSCs with allogenic stimulator and responder PBMCs in the presence of LEN. As shown in Figure 3, LEN did not modulate the ability of MSCs to impair lymphocyte proliferation. Furthermore, LEN did not
Figure 1. The effect of LEN on MSC proliferation. The proliferative capacity of MSCs was determined by MTT assay after 5-day cultivation in the presence or absence of LEN. As shown for three different MSC preparations, no significant differences could be detected after LEN treatment.
modify the expansion of lymphocytes in the absence of MSC (data not shown). Adipogenic and osteogenic differentiation of MSCs are affected by LEN To examine the influence of LEN on the functional characteristics of MSCs, we induced adipogenic and osteogenic differentiation for 21 days. Adipogenic differentiation was first detectable between days 7 and 20 in untreated MSCs. In 2 of 10 cases, differentiation began earlier in adipogenic cultures treated with LEN. Differentiation as assessed by the differentiation grade after Oil Red O staining exhibited no differences between LEN-treated and untreated cultures in 6 of 10 cases. However, in four cases, we observed a trend for a higher differentiation state in the LEN-treated cells (Fig. 4A, C). This observation was confirmed by quantifying FABP4 messenger RNA levels after adipogenic differentiation. Using real-time polymerase chain reaction, significantly higher FABP4 expression after LEN incubation was detected in 2 cases (i.e., 5 and 6), and slightly increased expression was detected in 2 other samples (i.e., 1 and 3). However, in the other samples, no difference or decreased FABP4 expression was observed among the LEN-treated MSCs (Fig. 4D). Osteogenic differentiation started between days 4 and 10 in 6 of 10 cases irrespective of LEN treatment. In 4 cases, no obvious osteogenic differentiation was detected by van Kossa staining (Fig. 5A–C). The staining of the calcium deposits tended to be more intensive in four samples of LEN-treated MSCs (Fig. 5A, C). Furthermore, osteogenic differentiation was quantified by assessing ALP catalytic activity. ALP catalytic activity was highly variable among the MSC samples. As expected, all differentiated MSCs had higher ALP levels than the undifferentiated cells, independent of LEN treatment. In some cases, the increase in ALP levels after osteogenic differentiation was higher in LEN-treated samples (Fig. 5D, E).
870
M. Wobus et al./ Experimental Hematology 2012;40:867–876
Figure 2. Phenotypic characterization of MSCs after LEN treatment by flow cytometry. (A) Flow cytometric analysis of MSCs from one donor with monoclonal antibodies against CD29-PE, CD73-PE, CD90-FITC, CD105-FITC, CD166-PE, and isotype controls (solid black line) after 5 days of LEN treatment (dotted black line) in comparison to untreated control (Co) cultures (solid gray line). (B) Mean 6 standard error of mean of the MFIs of CD29 and CD73 for all MSC cultures. LEN treatment significantly increased the expression levels for both molecules (**p ! 0.005; *p ! 0.05).
LEN decreases SDF-1 secretion by MSCs, thereby inhibiting HSC migration Chemokines and cytokines are important intermediates in the inflammatory response, and as regulatory proteins, they participate in the recruitment and mobilization of HSCs to and from the BM microenvironment. One of the most important molecules is SDF-1, which modulates the migration and mobilization of HSCs by mediating chemo-
taxis through interactions with its receptor CXCR4 [15]. Therefore, we investigated whether SDF-1 secretion by MSCs is regulated by LEN. After 5 days of incubation with or without LEN, SDF-1 protein levels were significantly reduced in the supernatants of treated MSCs (**p ! 0.01, ***p ! 0.001, Fig. 6A). As SDF-1 secretion is inhibited by LEN, we tested whether LEN treatment of MSCs also affects the
M. Wobus et al./ Experimental Hematology 2012;40:867–876
871
Table 1. Depiction of all MFI for CD29 and CD73 expression as analyzed by flow cytometry in three independent experiments and their mean of 10 MSC samples with and without LEN treatment
CD29 MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC CD73 MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC
1, Co 1, LEN 2, Co 2, LEN 3, Co 3, LEN 4, Co 4, LEN 5, Co 5, LEN 6, Co 6, LEN 7, Co 7, LEN 8, Co 8, LEN 9, Co 9, LEN 10, Co 10, LEN 1, Co 1, LEN 2, Co 2, LEN 3, Co 3, LEN 4, Co 4, LEN 5, Co 5, LEN 6, Co 6, LEN 7, Co 7, LEN 8, Co 8, LEN 9, Co 9, LEN 10, Co 10, LEN
MFI
MFI
MFI
Mean
427.4 561.8 395.8 488.0 420.5 449.8 340.7 384.3 379.8 480.0 449.2 451.7 244.5 540.3 283.4 622.9 529.3 566.1 574.2 674.3 MFI 109.9 112.7 130.6 121.8 185.5 228.4 90.1 193.9 179.2 182.3 127.8 133.5 197.7 185.7 309.4 351.0 230.8 359.3 305.1 371.6
473.1 572.1 446.5 415.5 460.7 480.1 409.9 453.2 336.3 407.5 513.0 591.8 486.4 598.9 933.0 635.5 568.4 620.8 413.4 543.9 MFI 148.6 224.0 418.4 402.1 198.5 301.0 111.4 123.2 210.9 307.5 258.6 351.0 340.8 455.0 418.5 457.0 321.1 319.1 214.0 467.7
444.8 543.3 372.0 533.2 485.6 484.7 321.4 399.6 448.4 520.5 452.8 638.4 473.8 497.9 589.7 675.1 431.0 521.2 499.6 502.0 MFI 185.9 260.1 113.9 255.6 194.3 264.5 104.0 195.4 299.4 236.9 282.7 262.7 281.4 438.9 373.8 419.3 197.5 390.8 294.2 362.1
448.4 559.1 404.8 478.9 455.6 471.5 357.3 412.4 388.2 469.3 471.7 560.6 401.6 545.7 602.0 644.5 509.6 569.4 495.7 573.4 mean 148.1 198.9 221.0 259.8 192.8 264.6 101.8 170.8 229.8 242.2 223.0 249.1 273.3 359.9 367.2 409.1 249.8 356.4 271.1 400.5
Co 5 control.
migration of HSCs toward conditioned medium by using a Transwell migration assay. Interestingly, freshly isolated CD34þ cells migrated more slowly when induced by the conditioned medium from LEN-treated MSCs than when induced by the conditioned medium from untreated cells (Fig. 6B), suggesting that LEN modulates the homing capacity of CD34þ cells. The dependence of migration capacity on the SDF-1 concentration was confirmed by blocking the SDF-1/CXCR4 interaction using AMD3100. When added to the conditioned medium of untreated MSCs, AMD3100 significantly reduced the number of migrated HSCs, whereas no difference was observed with
Figure 3. Impact of LEN on MSC-mediated inhibition of lymphocyte proliferation. Responder PBMCs were stimulated by a pool of allogenic PBMCs and cocultured in the presence or absence of MSCs and LEN. After 7 days, lymphocyte proliferation was determined by [3H]-thymidine incorporation. The results of four different MSC preparations are presented. Values represent the mean 6 standard error of triplicate samples.
LEN-treated MSCs in which SDF-1 secretion was inhibited (Fig. 6C).
Discussion The therapeutic benefit of LEN, a derivative of thalidomide, in various hematopoietic diseases indicates that the drug may affect multiple pathways that are both disease- and patient-specific [16]. However, the effects of LEN on nonhematopoietic BM progenitor cells, such as MSCs, have not been systematically investigated. In this study, we searched for a direct effect of LEN on the functional characteristics of BM-derived MSCs from healthy donors. For the first time, we provide evidence for a LEN-induced modulation of integrin b1 (CD29) expression and secretion of SDF-1, suggesting a thus far undescribed influence of LEN on signaling pathways in normal MSCs. First, we did not detect a significant effect of LEN on the growth pattern and their proliferation capacity of normal MSCs. These observations support the hypothesis that LEN, which selectively inhibits the growth of clonal MDS del(5q) HSC progenitors [12], does not change the hematopoietic stem cell environment in a quantitative manner. MSCs and LEN display an immunomodulatory capacity [17], and recent studies revealed that MSCs affect several properties of T cells. In this context, it has been demonstrated that MSCs efficiently suppress the proliferation of CD4 and CD8 T cells [5]. The reduction of T-cell proliferation is dependent on cell cycle arrest in G0/G1 phase [18]. In our setting, MSCs retained the ability to suppress T-cell proliferation in a mixed lymphocyte reaction after LEN treatment, thus excluding an MSC-mediated immunomodulatory effect of LEN.
872
M. Wobus et al./ Experimental Hematology 2012;40:867–876
Figure 4. Adipogenic differentiation of MSCs under the influence of LEN. (A–C) After 21 days of differentiation, the cultures were stained with Oil Red O, and differentiation was graded as described in the Methods. In four cases of LEN-treated cultures, a tendency for higher differentiation was observed. (D) The FABP4 messenger RNA expression levels in the differentiated MSCs were quantified by real-time PCR. Each bar demonstrates the fold expression in LENtreated cells compared to the expression in untreated controls, which was set as 1. The LEN-treated differentiated MSC samples 1, 3, 5, and 6 displayed higher FABP4 expression levels, with significant differences in 5 and 6 (*p ! 0.05).
M. Wobus et al./ Experimental Hematology 2012;40:867–876
873
Figure 5. Osteogenic differentiation of MSCs under the influence of LEN. (A–C) After 21 days of differentiation, the cultures were stained by the van Kossa method, and the differentiation was graded as described in the Methods. In four cases, forced differentiation was observed in the LEN-treated cultures. (D, E) ALP activity in the lysates of 2 exemplary shown MSC samples after 21 days of differentiation with 4-nitrophenylphosphate as a substrate. Co basic 5 undifferentiated MSCs without LEN; Co osteo 5 differentiated MSCs without LEN; LEN basic 5 undifferentiated MSCs with LEN; LEN osteo 5 differentiated MSCs with LEN. ALP release was determined for all samples photometrically at 405 nm and reported according to the protein concentration of the lysates. The results are presented as the mean 6 standard error of mean (n 5 3). Significant differences are indicated by **p ! 0.01 and ***p ! 0.001.
874
M. Wobus et al./ Experimental Hematology 2012;40:867–876
Figure 6. (Continued). (C) The SDF-1 dependence of HSC migration was demonstrated by the addition of the CXCR4 antagonist AMD3100 to the MSC-conditioned medium in the lower chamber of the Transwell system. The number of migrated cells was significantly inhibited by the addition of AMD3100 to the medium of untreated MSCs. By contrast, AMD3100 had no influence on HSC migration toward the conditioned medium of LENtreated MSCs in which SDF-1 levels were already reduced significantly (*p ! 0.05; **p ! 0.01; ***p ! 0.001).
Figure 6. LEN affects SDF-1 secretion by MSCs and influences HSC migration. (A) SDF-1 protein levels in the MSC supernatants were determined by enzyme-linked immunosorbent assay, which revealed a significant reduction after LEN treatment in three samples. Results are shown for five different samples and expressed as the mean 6 standard error of mean (SEM) of three independent experiments performed in triplicate. Significant differences are indicated by **p ! 0.01 and ***p ! 0.001. (B) Fresh CD34þ HSC migration induced by the conditioned medium from LEN-treated or untreated MSCs was observed using the in vitro Transwell migration assay. Results are shown for five different samples and expressed as the mean 6 SEM of two independent experiments performed in duplicate. Significant differences are indicated by **p ! 0.01 and ***p ! 0.001.(C) The SDF-1 dependence of HSC migration was demonstrated by the addition of the CXCR4 antagonist AMD3100 to the MSC-conditioned medium in the lower chamber of the Transwell system. The number of migrated cells was significantly inhibited by the addition of AMD3100 to the medium of untreated MSCs. By contrast, AMD3100 had no influence on HSC migration toward the conditioned medium of LENtreated MSCs in which SDF-1 levels were already reduced significantly (*p ! 0.05; **p ! 0.01; ***p ! 0.001).
Our observation that LEN might enhance adipogenic differentiation could be one positive side effect of LEN therapy, as it was demonstrated that the ability of MSCs to undergo adipogenic differentiation is significantly lower in elderly MDS patients than in normal MSCs [19]. A correlation between increased adipogenic differentiation
after LEN treatment and the age of the donors could not be detected in our study. Different concentrations of immunomodulatory compounds, e.g., CC-4047 and CC-6032, and LEN were studied for their effect on the osteoblastic differentiation of MSCs and the in vitro generation of osteoclasts from BM mononuclear cells [20]. At a concentration of 10 mM, the compounds inhibited osteoclastogenesis without suppressing osteogenic differentiation. After 42 days of differentiation, the treated osteoprogenitors exhibited similar nodule formation and ALP activity as the controls [20]. Although we differentiated the MSCs for only 21 days, we observed a trend for higher ALP activity after LEN treatment in some cases. Potential reasons for the varying results could be the generally high heterogeneity of primary MSCs, as well as differences in the MSC isolation and cultivation methods. The expression of a panel of surface molecules and the absence of hematopoietic markers such as CD34 and CD45 are, by definition, required to characterize MSCs [21]. As expected, CD34 and CD45 remained absent after LEN treatment. Although no differences in the MFIs of CD90, CD105, and CD166 were detected, the mean MFIs of CD29 and CD73 were increased by LEN. The GPI-anchored protein CD73 (ecto-50 -nucleotidase) is a signal transduction molecule that functions in the human immune system and is a mediator of cell–cell and cell–matrix interactions. In MSCs, it was demonstrated that CD73 is associated with their migratory behavior in response to mechanical loading [22]. CD73 is negatively regulated by tumor necrosis factor–a [22], and the repression of tumor necrosis factor–a is a crucial component of
M. Wobus et al./ Experimental Hematology 2012;40:867–876
many of the anti-inflammatory properties of LEN [23,24]. Interestingly, an in vitro study revealed that CD73 promoted osteoblastic differentiation, consistent with earlier reports indicating that the activation of adenosine receptors regulated the proliferation and differentiation of osteoblasts [25]. Our observation of a possible enhancement of the osteogenic differentiation of MSCs in response to LEN might be mediated by CD73. Whether these cellular effects translate to improved skeletal health remains to be determined. CD29 (integrin b1) is another interesting target of LEN, as it was demonstrated that integrin signaling is one of the most significantly deregulated pathways in del(5q) MDS HSCs [26]. Important cell–cell interactions and adhesion to the extracellular matrix and soluble ligands are regulated by integrins. Integrin b1 can associate with all other b-subunits and is therefore a central integrin within this class of receptors. The LEN-mediated enhancement of CD29 expression might result in the retention of HSCs, which will improve the maintenance of their residing niche. Recently, we demonstrated that integrins are critical for HSC adhesion to the MSC surface and further migration beneath the MSC layer [3]. Moreover, lymphocyte function–associated antigen-1 expression was significantly upregulated in HSCs on the MSC surface under low oxygen conditions [13]. Cell motility inside the BM and migration to or from this microenvironment are controlled by a close interaction between selective chemokines and adhesion molecules. Changes in the expression of these molecules could potentially affect the migration pattern of cells trafficking in the BM. The chemokine SDF-1 is expressed by MSCs, and it plays key roles in the migration of mature and immature hematopoietic precursor cells and leukemic cells that express the SDF-1 receptor CXCR4 [27]. Moreover, SDF-1 enhances the survival/antiapoptosis of hematopoietic progenitor cells in vitro as well as in vivo, suggesting that the SDF-1/CXCR4 axis plays an important role in hematopoietic cell regulation [28]. We observed a significant decrease in the concentration of SDF-1 in response to LEN, which has not been described previously. Interestingly, BM plasma of MDS patients contains higher SDF-1 levels than that from healthy volunteers [28]. Although the responsiveness of MDS progenitors to SDF-1 is highly controversial in the literature, it might be speculated that the reduction of SDF-1 levels by LEN could be one further mechanism to detach HSCs from their niche and thus reduce their survival advantage compared to healthy progenitors. The decreased migration of HSCs toward the conditioned medium of LEN-treated MSCs containing a lower concentration of SDF-1 suggests that the migratory activity of HSCs can be modulated by LEN administration. Changes in chemotactic gradients might control CD34þ cell localization and retention in the BM [13,27]. This might in part explain the side effects of the drug, such as cytopenia, in
875
several hematological malignancies. Additionally, there are some concerns about the potential alteration of stem cell mobilization in patients with myeloma after longterm LEN treatment [29,30]. Conclusions LEN can modulate some important characteristics of normal MSCs in vitro. The effect on SDF-1 secretion is functionally coupled to the stromal support toward HSCs, which might be considered an important and novel effect of LEN on the bone microenvironment.
Acknowledgments The study was supported by the collaborative research grants SFB 655, From Cells to Tissues (M.B.) and the DFG Research Group-1586 SKELMET (L.C.H., M.B., and M.W.). In part, this work was also supported by the BMBF-funded consortium Therapeutic Potential of Mesenchymal Stem Cells (M.S. and M.B.). The authors would like to thank Dr. Ute Hempel, Institute of Physiological Chemistry, Medical Faculty Dresden, for her help with ALP measurements and Katrin M€ uller and Kristin Heidel, University Hospital of Dresden, for excellent technical support.
Conflict of interest disclosure No financial interest/relationships with financial interest relating to the topic of this article have been declared.
References 1. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potential of adult human mesenchymal stem cells. Science. 1999;284:143–147. 2. Alakel N, Jing D, Muller K, et al. Direct contact with mesenchymal stromal cells affects migratory behavior and gene expression profile of CD133þ hematopoietic stem cells during ex vivo expansion. Exp Hematol. 2009;37:504–513. 3. Jing D, Fonseca AV, Alakel N, et al. Hematopoietic stem cells in coculture with mesenchymal stromal cellsdmodeling the niche compartments in vitro. Haematologica. 2010;95:542–550. 4. Uccelli A, Pistoia V, Moretta L. Mesenchymal stem cells: a new strategy for immunosuppression? Trends Immunol. 2007;28:219–226. 5. Zhao S, Wehner R, Bornhauser M, et al. Immunomodulatory properties of mesenchymal stromal cells and their therapeutic consequences for immune-mediated disorders. Stem Cells Dev. 2010;19:607–614. 6. List A, Dewald G, Bennett J, et al. Lenalidomide in the myelodysplastic syndrome with chromosome 5q deletion. N Engl J Med. 2006;355: 1456–1465. 7. Matsuoka A, Tochigi A, Kishimoto M, et al. Lenalidomide induces cell death in an MDS-derived cell line with deletion of chromosome 5q by inhibition of cytokinesis. Leukemia. 2010;24:748–755. 8. Davies F, Baz R. Lenalidomide mode of action: linking bench and clinical findings. Blood Rev. 2010;24(suppl 1):S13–S19. 9. Verhelle D, Corral LG, Wong K, et al. Lenalidomide and CC-4047 inhibit the proliferation of malignant B cells while expanding normal CD34þ progenitor cells. Cancer Res. 2007;67:746–755. 10. Oswald J, Boxberger S, Jorgensen B, et al. Mesenchymal stem cells can be differentiated into endothelial cells in vitro. Stem Cells. 2004;22:377–384.
876
M. Wobus et al./ Experimental Hematology 2012;40:867–876
11. Hideshima T, Chauhan D, Shima Y, et al. Thalidomide and its analogs overcome drug resistance of human multiple myeloma cells to conventional therapy. Blood. 2000;96:2943–2950. 12. Pellagatti A, Jadersten M, Forsblom AM, et al. Lenalidomide inhibits the malignant clone and up-regulates the SPARC gene mapping to the commonly deleted region in 5q- syndrome patients. Proc Natl Acad Sci U S A. 2007;104:11406–11411. 13. Jing D, Wobus M, Poitz DM, et al. Oxygen tension plays a critical role in the hematopoietic microenvironment in vitro. Haematologica. 2012; 97:331–339. 14. Hempel U, Hefti T, Kalbacova M, et al. Response of osteoblast-like SAOS-2 cells to zirconia ceramics with different surface topographies. Clin Oral Implants Res. 2010;21:174–181. 15. Lapidot T, Dar A, Kollet O. How do stem cells find their way home? Blood. 2005;106:1901–1910. 16. Schecter J, Galili N, Raza A. MDS: Refining existing therapy through improved biologic insights. Blood Rev. 2012;26:73–80. 17. Trento C, Dazzi F. Mesenchymal stem cells and innate tolerance: biology and clinical applications. Swiss Med Wkly. 2010;140:w13121. 18. Glennie S, Soeiro I, Dyson PJ, Lam EW, Dazzi F. Bone marrow mesenchymal stem cells induce division arrest anergy of activated T cells. Blood. 2005;105:2821–2827. 19. Varga G, Kiss J, Varkonyi J, et al. Inappropriate Notch activity and limited mesenchymal stem cell plasticity in the bone marrow of patients with myelodysplastic syndromes. Pathol Oncol Res. 2007; 13:311–319. 20. Munemasa S, Sakai A, Kuroda Y, et al. Osteoprogenitor differentiation is not affected by immunomodulatory thalidomide analogs but is promoted by low bortezomib concentration, while both agents suppress osteoclast differentiation. Int J Oncol. 2008;33:129–136. 21. Pontikoglou C, Deschaseaux F, Sensebe L, Papadaki HA. Bone marrow mesenchymal stem cells: biological properties and their role
22.
23.
24. 25.
26.
27.
28.
29.
30.
in hematopoiesis and hematopoietic stem cell transplantation. Stem Cell Rev. 2011;7:569–589. Ode A, Kopf J, Kurtz A, et al. CD73 and CD29 concurrently mediate the mechanically induced decrease of migratory capacity of mesenchymal stromal cells. Eur Cell Mater. 2011;22:26–42. Bodera P, Stankiewicz W. Immunomodulatory properties of thalidomide analogs: pomalidomide and lenalidomide, experimental and therapeutic applications. Recent Pat Endocr Metab Immune Drug Discov. 2011;5:192–196. Chanan-Khan AA, Cheson BD. Lenalidomide for the treatment of B-cell malignancies. J Clin Oncol. 2008;20:26. 1544–1552. Takedachi M, Oohara H, Smith BJ, et al. CD73-generated adenosine promotes osteoblast differentiation. J Cell Physiol. 2012;227:2622– 2631. Pellagatti A, Cazzola M, Giagounidis A, et al. Deregulated gene expression pathways in myelodysplastic syndrome hematopoietic stem cells. Leukemia. 2010;24:756–764. Wright N, de Lera TL, Garcia-Moruja C, et al. Transforming growth factor-beta1 down-regulates expression of chemokine stromal cellderived factor-1: functional consequences in cell migration and adhesion. Blood. 2003;102:1978–1984. Matsuda M, Morita Y, Hanamoto H, et al. CD34þ progenitors from MDS patients are unresponsive to SDF-1, despite high levels of SDF-1 in bone marrow plasma. Leukemia. 2004;18:1038–1040. Kumar S, Dispenzieri A, Lacy MQ, et al. Impact of lenalidomide therapy on stem cell mobilization and engraftment post-peripheral blood stem cell transplantation in patients with newly diagnosed myeloma. Leukemia. 2007;21:2035–2042. Popat U, Saliba R, Thandi R, et al. Impairment of filgrastiminduced stem cell mobilization after prior lenalidomide in patients with multiple myeloma. Biol Blood Marrow Transplant. 2009;15: 718–723.
Impact of lenalidomide on the functional properties of human mesenchymal stromal cells Manja Wobusa, Gwendolin Benatha, Ruben A. Ferrera, Rebekka Wehnerb, Marc Schmitzb,c, Lorenz C. Hofbauerd, Martina Raunerd, Gerhard Ehningera,c, Martin Bornh€ausera,c, and Uwe Platzbeckera a
Medical Clinic and Polyclinic I, University Hospital, Dresden, Germany; bInstitute of Immunology, Medical Faculty, Technical University, Dresden, Germany; cCenter for Regenerative Therapies Dresden, Dresden, Germany; dMedical Clinic and Polyclinic III, University Hospital, Dresden, Germany (Received 8 March 2012; revised 15 May 2012; accepted 11 June 2012)
Objective. Lenalidomide (LEN) has emerged as a promising therapeutic option for the management of various hematologic malignancies. Although its direct mechanisms of action on malignant cells have been studied intensively, its effects on the stromal compartment of bone marrow have not yet been analyzed systematically. Therefore, we investigated whether LEN alters the functional capacity of mesenchymal stromal cells (MSCs) as the main cellular component of the bone marrow microenvironment. In addition to their growth and differentiation characteristics, we focused on the ability of MSC to modulate T-cell function and support hematopoietic stem cells (HSCs). Materials and Methods. Bone marrow–derived MSCs were exposed to LEN (10 mM), and differences in proliferation, phenotype, inhibition of T-cell proliferation, and differentiation capacity were analyzed. A Boyden chamber assay was used to test the migratory potential of HSC toward the conditioned medium of LEN-treated or untreated MSCs, and the stromal cell–derived factor-1 (SDF-1) concentrations in these supernatants were determined by enzyme-linked immunosorbent assay. Results. Treatment of MSCs with LEN did not affect their growth rate, proliferation, osteogenic and adipogenic differentiation potential, or capacity to inhibit T-cell proliferation. However, LEN treatment increased the average of mean fluorescence intensity of CD29 and CD73 by 15 and 22%, respectively. Interestingly, LEN reduced SDF-1 by MSCs by 32% compared to that of control cells. As a functional consequence, the serum-free supernatant of LEN-treated MSCs had a significantly lower potential to induce the directed migration of CD34+ HSCs. Conclusion. LEN can modulate the expression of cell surface molecules and the chemokine secretion of MSCs in vitro. These effects might contribute to the clinical effects of the compound in vivo for patients with hematological malignancies. Ó 2012 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc.
Mesenchymal stromal cells (MSCs) are a major cellular component of the formation and function of the bone marrow (BM) microenvironment. MSCs are the progenitor cells of osteoblasts and represent the hematopoieticsupporting stroma components of the marrow [1]. In addition to their capacity to support hematopoiesis [2,3], MSCs display systemic immunoregulatory and immunosuppressive properties [4,5]. The main functions of MSCs are Offprint requests to: Manja Wobus, Ph.D., Medical Clinic and Polyclinic I, University Hospital Dresden, 01307 Dresden, Germany; E-mail: manja. [email protected]
mediated by chemokines and cytokines in an autocrine or paracrine manner. The immunomodulatory drug lenalidomide (LEN) has been used for the successful treatment of a variety of hematological malignancies. In fact, LEN has been demonstrated to reduce and eliminate the need for red blood cell transfusions in the majority of myelodysplastic syndrome (MDS) patients with del(5q) [6]. Furthermore, it performs several important biological actions, including inhibiting angiogenesis and suppressing the production of proinflammatory cytokines, such as tumor necrosis factor–a [7]. LEN has also been reported to abrogate the stromal cell support of myeloma cells by interfering with
0301-472X/$ - see front matter. Copyright Ó 2012 ISEH - Society for Hematology and Stem Cells. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.exphem.2012.06.004
868
M. Wobus et al./ Experimental Hematology 2012;40:867–876
cytokines, e.g., interleukin-6 and vascular endothelial growth factor, leading to the growth arrest of neoplastic cells [8]. By contrast, it can augment the proliferation of healthy hematopoietic progenitors [9]. Therefore, LEN is currently under investigation as a treatment for various hematological diseases, including those without direct marrow involvement. These approaches include maintenance strategies with long-term usage of the compound. To obtain detailed insights into the regulatory role of LEN in the stromal compartment, we investigated its effects on healthy MSCs. Our findings of increased expression of CD29 and an inhibited migration of hematopoietic stem cells (HSCs) in response to a LEN-induced alteration of stromal cell–derived factor-1 (SDF-1) secretion provide new insights into the mechanism of the effects of LEN on hematologic diseases.
Material and methods Cell culture MSCs were isolated from BM aspirates that were derived and cultured from healthy donors after receiving informed consent and approval from the local ethics committee as described previously [10]. All experiments were carried out with MSCs derived from passage 2 to 4. The age of the donors was 21 to 52 years (median 5 39 years). LEN (Celgene, Munich, Germany) was dissolved in 10% dimethyl sulfoxide, and the stock solution was stored at 20 C. After being thawed, the sample was protected from light and kept at 4 C for a maximum of 2 weeks. On the basis of multiple myeloma and del(5q) MDS data [11,12], a LEN concentration of 10 mM was chosen for the experiments after conducting a dose-response experiment using concentrations of 0.1, 1,0, 10, and 50 mM. Proliferation MSCs (5 103/well in a 96-well plate) were cultured in the presence or absence of LEN for 5 days, and an MTT assay was performed according to the protocol (Roche Diagnostics GmbH, Mannheim, Germany). The absorbance was measured at 570 and 650 nm. Medium was used as a negative control. Immunophenotypic characteristics of MSCs Flow cytometric analysis was performed with a FACSCalibur (BD Biosciences, Heidelberg, Germany) using BD CellQuest Pro and FlowJo software. At subconfluency, cells were incubated with or without LEN for 5 days and then characterized for the expression of positive (CD29-phycoerythrin [PE], CD73-PE, CD90–fluorescein isothiocyanate [FITC], CD105-FITC, and CD166-PE) and negative MSC markers (CD14-PE, CD34-PE, and CD45-FITC). IgG1-negative control-FITC (AbD Serotec, D€ usseldorf, Germany) and IgG1-PE (BD Biosciences) were used as isotype controls. Anti–CD90-FITC and anti–CD105FITC were bought from AbD Serotec. All other antibodies were purchased from BD Biosciences. Mixed lymphocyte reaction Blood samples were obtained from healthy donors after informed written consent. Peripheral blood mononuclear cells (PBMCs)
were prepared by Ficoll-Hypaque (Biochrom, Berlin, Germany) density centrifugation. PBMCs were pooled from five different donors, irradiated (30 Gy), and used as allogeneic stimulators for nonirradiated responder PBMCs. To analyze the effect of LEN on MSC-mediated inhibition of lymphocyte proliferation, stimulator PBMCs (1 105/well) and responder PBMCs (1 105/well) were cocultured in the presence or absence of thirdparty MSCs (5 103/well) and 10 mM LEN. Before culture, MSCs were harvested from culture flasks by trypsinization and subsequently irradiated (30 Gy). The cells were incubated in 96well plates using RPMI-1640 medium (Biochrom) supplemented with 2 mM L-glutamine, 10 mM sodium pyruvate, 1% nonessential amino acids, 100 mg/mL penicillin, 100 mg/mL streptomycin (all from Biochrom), and 10% fetal calf serum for 7 days. One microcurie of [3H]-thymidine (Hartmann Analytic, Braunschweig, Germany) was added to each well for the last 18 hours of culture. Cells were harvested and [3H]-thymidine incorporation was determined using a beta counter (Wallac, Freiburg, Germany). MSC differentiation assays MSCs in passage 2 were seeded in six-well plates (5 103 MSCs/ cm2), cultured to subconfluency for approximately 4 days in Dulbecco’s modified Eagle’s medium, and subjected to adipogenic (0.5 mM 1-methyl-3-butylisoxanthine, 1 mM dexamethasone, 100 mM indomethacin, and 10 mM insulin) or osteogenic (0.1 mM dexamethasone, 0.2 mM ascorbate-2-phosphate, 10 mM bglycerophosphate) differentiation in the presence or absence of LEN. The start of differentiation with the first appearance of lipid vacuoles was recorded, and adipogenesis was assessed by Oil Red O staining after 21 days. Differentiation was graded according to microscopic analysis as follows: 0 5 no differentiation; 1 5 25% of the culture well contained differentiated cells; 2 5 50% of the culture well contained differentiated cells; 3 5 75% of the culture well contained differentiated cells; 4 5 100% of the culture well contained differentiated cells. Moreover, cells were collected for messenger RNA isolation and quantitative real-time polymerase chain reaction was performed for the FABP4 gene (sense primer: 50 -tac tgg gcc agg aat ttg ac-30 ; reverse primer: 50 -tgg ttg att ttc cat ccc at-30 ) as described recently [13]. The first day on which calcium deposits were observed was documented as a sign of osteogenic differentiation. After 21 days, osteogenesis was determined by van Kossa staining, and the catalytic activity of alkaline phosphatase (ALP) was determined as described previously [14]. The medium was changed twice a week, and LEN was freshly added each time. All reagents for differentiation induction were purchased from Sigma-Aldrich (Hamburg, Germany). Chemokine measurements in cell culture supernatants Standardized 1-mL aliquots from serum-free culture supernatants were collected from untreated MSCs or MSCs treated with 10 mM LEN for 5 days. The protein levels of SDF-1 were quantified using an enzyme-linked immunosorbent assay (R&D Systems GmbH, Wiesbaden, Germany) according to the manufacturer’s protocol. HSC transwell migration assay Purification of CD34þ HSCs and Transwell migration toward the conditioned medium of either untreated or LEN-treated MSCs were performed as described previously [13]. Briefly, 2 105
M. Wobus et al./ Experimental Hematology 2012;40:867–876
869
HSCs were added to the upper chamber in 0.1 mL medium and allowed to migrate for 4 hours at 37 C toward 600 mL conditioned medium through a 5-mm polycarbonate membrane. Afterward, cells in the lower chamber were counted. The CXCR4 antagonist AMD3100 (Sigma) was used at a concentration of 10 mM. Statistical analysis Data were analyzed using the GraphPad Prism statistical PC program (GraphPad, San Diego, CA, USA). Results from at least three independent experiments are presented as the mean 6 standard error of the mean (presented as error bars). Statistical comparisons were performed using either a paired Wilcoxon test or one-way analysis of variance with Bonferroni’s post-test.
Results LEN has no influence on MSC growth and proliferation In the initial experiments, we investigated whether LEN influences the growth characteristics of normal MSCs. The treated cells displayed the same fibroblastoid growth pattern as the control cultures and reached confluency at the same time point (not shown). Proliferation was measured using MTT assay after 5 days of culture in 96well plates. No significant differences were detected between LEN-treated and control MSC, as shown in Figure 1. Expression of CD29 and CD73 can be increased by LEN MSCs were cultured in the presence or absence of LEN for 5 days, and the expression of cell surface molecules was examined by flow cytometry. The expression of hematopoietic markers, such as CD14, CD34, and CD45, was !5% of control levels in all cases (not shown). No differences were detected between treated and untreated cells. The cultures were composed of a homogenous cell population with similar forward/sideward scatter characteristics, and the expression of typical MSC surface antigens, e.g., CD29, CD73, CD90, CD105, and CD166, was observed in 95% to 100% of the cells. The mean fluorescence intensities (MFIs) were different for the molecules, as demonstrated in Figure 2A, where a representative measurement for each antigen is shown. Interestingly, for CD29 and CD73, mean MFIs were significantly increased by LEN treatment (CD29: 453.4 vs 528.4; **p ! 0.005; CD73: 227.2 vs 291.2; *p ! 0.05; Fig. 2B, Table 1). LEN-treated MSCs inhibit the proliferation of lymphocytes In further experiments, we explored the effect of LEN on MSC-mediated lymphocyte proliferation. Therefore, we incubated MSCs with allogenic stimulator and responder PBMCs in the presence of LEN. As shown in Figure 3, LEN did not modulate the ability of MSCs to impair lymphocyte proliferation. Furthermore, LEN did not
Figure 1. The effect of LEN on MSC proliferation. The proliferative capacity of MSCs was determined by MTT assay after 5-day cultivation in the presence or absence of LEN. As shown for three different MSC preparations, no significant differences could be detected after LEN treatment.
modify the expansion of lymphocytes in the absence of MSC (data not shown). Adipogenic and osteogenic differentiation of MSCs are affected by LEN To examine the influence of LEN on the functional characteristics of MSCs, we induced adipogenic and osteogenic differentiation for 21 days. Adipogenic differentiation was first detectable between days 7 and 20 in untreated MSCs. In 2 of 10 cases, differentiation began earlier in adipogenic cultures treated with LEN. Differentiation as assessed by the differentiation grade after Oil Red O staining exhibited no differences between LEN-treated and untreated cultures in 6 of 10 cases. However, in four cases, we observed a trend for a higher differentiation state in the LEN-treated cells (Fig. 4A, C). This observation was confirmed by quantifying FABP4 messenger RNA levels after adipogenic differentiation. Using real-time polymerase chain reaction, significantly higher FABP4 expression after LEN incubation was detected in 2 cases (i.e., 5 and 6), and slightly increased expression was detected in 2 other samples (i.e., 1 and 3). However, in the other samples, no difference or decreased FABP4 expression was observed among the LEN-treated MSCs (Fig. 4D). Osteogenic differentiation started between days 4 and 10 in 6 of 10 cases irrespective of LEN treatment. In 4 cases, no obvious osteogenic differentiation was detected by van Kossa staining (Fig. 5A–C). The staining of the calcium deposits tended to be more intensive in four samples of LEN-treated MSCs (Fig. 5A, C). Furthermore, osteogenic differentiation was quantified by assessing ALP catalytic activity. ALP catalytic activity was highly variable among the MSC samples. As expected, all differentiated MSCs had higher ALP levels than the undifferentiated cells, independent of LEN treatment. In some cases, the increase in ALP levels after osteogenic differentiation was higher in LEN-treated samples (Fig. 5D, E).
870
M. Wobus et al./ Experimental Hematology 2012;40:867–876
Figure 2. Phenotypic characterization of MSCs after LEN treatment by flow cytometry. (A) Flow cytometric analysis of MSCs from one donor with monoclonal antibodies against CD29-PE, CD73-PE, CD90-FITC, CD105-FITC, CD166-PE, and isotype controls (solid black line) after 5 days of LEN treatment (dotted black line) in comparison to untreated control (Co) cultures (solid gray line). (B) Mean 6 standard error of mean of the MFIs of CD29 and CD73 for all MSC cultures. LEN treatment significantly increased the expression levels for both molecules (**p ! 0.005; *p ! 0.05).
LEN decreases SDF-1 secretion by MSCs, thereby inhibiting HSC migration Chemokines and cytokines are important intermediates in the inflammatory response, and as regulatory proteins, they participate in the recruitment and mobilization of HSCs to and from the BM microenvironment. One of the most important molecules is SDF-1, which modulates the migration and mobilization of HSCs by mediating chemo-
taxis through interactions with its receptor CXCR4 [15]. Therefore, we investigated whether SDF-1 secretion by MSCs is regulated by LEN. After 5 days of incubation with or without LEN, SDF-1 protein levels were significantly reduced in the supernatants of treated MSCs (**p ! 0.01, ***p ! 0.001, Fig. 6A). As SDF-1 secretion is inhibited by LEN, we tested whether LEN treatment of MSCs also affects the
M. Wobus et al./ Experimental Hematology 2012;40:867–876
871
Table 1. Depiction of all MFI for CD29 and CD73 expression as analyzed by flow cytometry in three independent experiments and their mean of 10 MSC samples with and without LEN treatment
CD29 MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC CD73 MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC MSC
1, Co 1, LEN 2, Co 2, LEN 3, Co 3, LEN 4, Co 4, LEN 5, Co 5, LEN 6, Co 6, LEN 7, Co 7, LEN 8, Co 8, LEN 9, Co 9, LEN 10, Co 10, LEN 1, Co 1, LEN 2, Co 2, LEN 3, Co 3, LEN 4, Co 4, LEN 5, Co 5, LEN 6, Co 6, LEN 7, Co 7, LEN 8, Co 8, LEN 9, Co 9, LEN 10, Co 10, LEN
MFI
MFI
MFI
Mean
427.4 561.8 395.8 488.0 420.5 449.8 340.7 384.3 379.8 480.0 449.2 451.7 244.5 540.3 283.4 622.9 529.3 566.1 574.2 674.3 MFI 109.9 112.7 130.6 121.8 185.5 228.4 90.1 193.9 179.2 182.3 127.8 133.5 197.7 185.7 309.4 351.0 230.8 359.3 305.1 371.6
473.1 572.1 446.5 415.5 460.7 480.1 409.9 453.2 336.3 407.5 513.0 591.8 486.4 598.9 933.0 635.5 568.4 620.8 413.4 543.9 MFI 148.6 224.0 418.4 402.1 198.5 301.0 111.4 123.2 210.9 307.5 258.6 351.0 340.8 455.0 418.5 457.0 321.1 319.1 214.0 467.7
444.8 543.3 372.0 533.2 485.6 484.7 321.4 399.6 448.4 520.5 452.8 638.4 473.8 497.9 589.7 675.1 431.0 521.2 499.6 502.0 MFI 185.9 260.1 113.9 255.6 194.3 264.5 104.0 195.4 299.4 236.9 282.7 262.7 281.4 438.9 373.8 419.3 197.5 390.8 294.2 362.1
448.4 559.1 404.8 478.9 455.6 471.5 357.3 412.4 388.2 469.3 471.7 560.6 401.6 545.7 602.0 644.5 509.6 569.4 495.7 573.4 mean 148.1 198.9 221.0 259.8 192.8 264.6 101.8 170.8 229.8 242.2 223.0 249.1 273.3 359.9 367.2 409.1 249.8 356.4 271.1 400.5
Co 5 control.
migration of HSCs toward conditioned medium by using a Transwell migration assay. Interestingly, freshly isolated CD34þ cells migrated more slowly when induced by the conditioned medium from LEN-treated MSCs than when induced by the conditioned medium from untreated cells (Fig. 6B), suggesting that LEN modulates the homing capacity of CD34þ cells. The dependence of migration capacity on the SDF-1 concentration was confirmed by blocking the SDF-1/CXCR4 interaction using AMD3100. When added to the conditioned medium of untreated MSCs, AMD3100 significantly reduced the number of migrated HSCs, whereas no difference was observed with
Figure 3. Impact of LEN on MSC-mediated inhibition of lymphocyte proliferation. Responder PBMCs were stimulated by a pool of allogenic PBMCs and cocultured in the presence or absence of MSCs and LEN. After 7 days, lymphocyte proliferation was determined by [3H]-thymidine incorporation. The results of four different MSC preparations are presented. Values represent the mean 6 standard error of triplicate samples.
LEN-treated MSCs in which SDF-1 secretion was inhibited (Fig. 6C).
Discussion The therapeutic benefit of LEN, a derivative of thalidomide, in various hematopoietic diseases indicates that the drug may affect multiple pathways that are both disease- and patient-specific [16]. However, the effects of LEN on nonhematopoietic BM progenitor cells, such as MSCs, have not been systematically investigated. In this study, we searched for a direct effect of LEN on the functional characteristics of BM-derived MSCs from healthy donors. For the first time, we provide evidence for a LEN-induced modulation of integrin b1 (CD29) expression and secretion of SDF-1, suggesting a thus far undescribed influence of LEN on signaling pathways in normal MSCs. First, we did not detect a significant effect of LEN on the growth pattern and their proliferation capacity of normal MSCs. These observations support the hypothesis that LEN, which selectively inhibits the growth of clonal MDS del(5q) HSC progenitors [12], does not change the hematopoietic stem cell environment in a quantitative manner. MSCs and LEN display an immunomodulatory capacity [17], and recent studies revealed that MSCs affect several properties of T cells. In this context, it has been demonstrated that MSCs efficiently suppress the proliferation of CD4 and CD8 T cells [5]. The reduction of T-cell proliferation is dependent on cell cycle arrest in G0/G1 phase [18]. In our setting, MSCs retained the ability to suppress T-cell proliferation in a mixed lymphocyte reaction after LEN treatment, thus excluding an MSC-mediated immunomodulatory effect of LEN.
872
M. Wobus et al./ Experimental Hematology 2012;40:867–876
Figure 4. Adipogenic differentiation of MSCs under the influence of LEN. (A–C) After 21 days of differentiation, the cultures were stained with Oil Red O, and differentiation was graded as described in the Methods. In four cases of LEN-treated cultures, a tendency for higher differentiation was observed. (D) The FABP4 messenger RNA expression levels in the differentiated MSCs were quantified by real-time PCR. Each bar demonstrates the fold expression in LENtreated cells compared to the expression in untreated controls, which was set as 1. The LEN-treated differentiated MSC samples 1, 3, 5, and 6 displayed higher FABP4 expression levels, with significant differences in 5 and 6 (*p ! 0.05).
M. Wobus et al./ Experimental Hematology 2012;40:867–876
873
Figure 5. Osteogenic differentiation of MSCs under the influence of LEN. (A–C) After 21 days of differentiation, the cultures were stained by the van Kossa method, and the differentiation was graded as described in the Methods. In four cases, forced differentiation was observed in the LEN-treated cultures. (D, E) ALP activity in the lysates of 2 exemplary shown MSC samples after 21 days of differentiation with 4-nitrophenylphosphate as a substrate. Co basic 5 undifferentiated MSCs without LEN; Co osteo 5 differentiated MSCs without LEN; LEN basic 5 undifferentiated MSCs with LEN; LEN osteo 5 differentiated MSCs with LEN. ALP release was determined for all samples photometrically at 405 nm and reported according to the protein concentration of the lysates. The results are presented as the mean 6 standard error of mean (n 5 3). Significant differences are indicated by **p ! 0.01 and ***p ! 0.001.
874
M. Wobus et al./ Experimental Hematology 2012;40:867–876
Figure 6. (Continued). (C) The SDF-1 dependence of HSC migration was demonstrated by the addition of the CXCR4 antagonist AMD3100 to the MSC-conditioned medium in the lower chamber of the Transwell system. The number of migrated cells was significantly inhibited by the addition of AMD3100 to the medium of untreated MSCs. By contrast, AMD3100 had no influence on HSC migration toward the conditioned medium of LENtreated MSCs in which SDF-1 levels were already reduced significantly (*p ! 0.05; **p ! 0.01; ***p ! 0.001).
Figure 6. LEN affects SDF-1 secretion by MSCs and influences HSC migration. (A) SDF-1 protein levels in the MSC supernatants were determined by enzyme-linked immunosorbent assay, which revealed a significant reduction after LEN treatment in three samples. Results are shown for five different samples and expressed as the mean 6 standard error of mean (SEM) of three independent experiments performed in triplicate. Significant differences are indicated by **p ! 0.01 and ***p ! 0.001. (B) Fresh CD34þ HSC migration induced by the conditioned medium from LEN-treated or untreated MSCs was observed using the in vitro Transwell migration assay. Results are shown for five different samples and expressed as the mean 6 SEM of two independent experiments performed in duplicate. Significant differences are indicated by **p ! 0.01 and ***p ! 0.001.(C) The SDF-1 dependence of HSC migration was demonstrated by the addition of the CXCR4 antagonist AMD3100 to the MSC-conditioned medium in the lower chamber of the Transwell system. The number of migrated cells was significantly inhibited by the addition of AMD3100 to the medium of untreated MSCs. By contrast, AMD3100 had no influence on HSC migration toward the conditioned medium of LENtreated MSCs in which SDF-1 levels were already reduced significantly (*p ! 0.05; **p ! 0.01; ***p ! 0.001).
Our observation that LEN might enhance adipogenic differentiation could be one positive side effect of LEN therapy, as it was demonstrated that the ability of MSCs to undergo adipogenic differentiation is significantly lower in elderly MDS patients than in normal MSCs [19]. A correlation between increased adipogenic differentiation
after LEN treatment and the age of the donors could not be detected in our study. Different concentrations of immunomodulatory compounds, e.g., CC-4047 and CC-6032, and LEN were studied for their effect on the osteoblastic differentiation of MSCs and the in vitro generation of osteoclasts from BM mononuclear cells [20]. At a concentration of 10 mM, the compounds inhibited osteoclastogenesis without suppressing osteogenic differentiation. After 42 days of differentiation, the treated osteoprogenitors exhibited similar nodule formation and ALP activity as the controls [20]. Although we differentiated the MSCs for only 21 days, we observed a trend for higher ALP activity after LEN treatment in some cases. Potential reasons for the varying results could be the generally high heterogeneity of primary MSCs, as well as differences in the MSC isolation and cultivation methods. The expression of a panel of surface molecules and the absence of hematopoietic markers such as CD34 and CD45 are, by definition, required to characterize MSCs [21]. As expected, CD34 and CD45 remained absent after LEN treatment. Although no differences in the MFIs of CD90, CD105, and CD166 were detected, the mean MFIs of CD29 and CD73 were increased by LEN. The GPI-anchored protein CD73 (ecto-50 -nucleotidase) is a signal transduction molecule that functions in the human immune system and is a mediator of cell–cell and cell–matrix interactions. In MSCs, it was demonstrated that CD73 is associated with their migratory behavior in response to mechanical loading [22]. CD73 is negatively regulated by tumor necrosis factor–a [22], and the repression of tumor necrosis factor–a is a crucial component of
M. Wobus et al./ Experimental Hematology 2012;40:867–876
many of the anti-inflammatory properties of LEN [23,24]. Interestingly, an in vitro study revealed that CD73 promoted osteoblastic differentiation, consistent with earlier reports indicating that the activation of adenosine receptors regulated the proliferation and differentiation of osteoblasts [25]. Our observation of a possible enhancement of the osteogenic differentiation of MSCs in response to LEN might be mediated by CD73. Whether these cellular effects translate to improved skeletal health remains to be determined. CD29 (integrin b1) is another interesting target of LEN, as it was demonstrated that integrin signaling is one of the most significantly deregulated pathways in del(5q) MDS HSCs [26]. Important cell–cell interactions and adhesion to the extracellular matrix and soluble ligands are regulated by integrins. Integrin b1 can associate with all other b-subunits and is therefore a central integrin within this class of receptors. The LEN-mediated enhancement of CD29 expression might result in the retention of HSCs, which will improve the maintenance of their residing niche. Recently, we demonstrated that integrins are critical for HSC adhesion to the MSC surface and further migration beneath the MSC layer [3]. Moreover, lymphocyte function–associated antigen-1 expression was significantly upregulated in HSCs on the MSC surface under low oxygen conditions [13]. Cell motility inside the BM and migration to or from this microenvironment are controlled by a close interaction between selective chemokines and adhesion molecules. Changes in the expression of these molecules could potentially affect the migration pattern of cells trafficking in the BM. The chemokine SDF-1 is expressed by MSCs, and it plays key roles in the migration of mature and immature hematopoietic precursor cells and leukemic cells that express the SDF-1 receptor CXCR4 [27]. Moreover, SDF-1 enhances the survival/antiapoptosis of hematopoietic progenitor cells in vitro as well as in vivo, suggesting that the SDF-1/CXCR4 axis plays an important role in hematopoietic cell regulation [28]. We observed a significant decrease in the concentration of SDF-1 in response to LEN, which has not been described previously. Interestingly, BM plasma of MDS patients contains higher SDF-1 levels than that from healthy volunteers [28]. Although the responsiveness of MDS progenitors to SDF-1 is highly controversial in the literature, it might be speculated that the reduction of SDF-1 levels by LEN could be one further mechanism to detach HSCs from their niche and thus reduce their survival advantage compared to healthy progenitors. The decreased migration of HSCs toward the conditioned medium of LEN-treated MSCs containing a lower concentration of SDF-1 suggests that the migratory activity of HSCs can be modulated by LEN administration. Changes in chemotactic gradients might control CD34þ cell localization and retention in the BM [13,27]. This might in part explain the side effects of the drug, such as cytopenia, in
875
several hematological malignancies. Additionally, there are some concerns about the potential alteration of stem cell mobilization in patients with myeloma after longterm LEN treatment [29,30]. Conclusions LEN can modulate some important characteristics of normal MSCs in vitro. The effect on SDF-1 secretion is functionally coupled to the stromal support toward HSCs, which might be considered an important and novel effect of LEN on the bone microenvironment.
Acknowledgments The study was supported by the collaborative research grants SFB 655, From Cells to Tissues (M.B.) and the DFG Research Group-1586 SKELMET (L.C.H., M.B., and M.W.). In part, this work was also supported by the BMBF-funded consortium Therapeutic Potential of Mesenchymal Stem Cells (M.S. and M.B.). The authors would like to thank Dr. Ute Hempel, Institute of Physiological Chemistry, Medical Faculty Dresden, for her help with ALP measurements and Katrin M€ uller and Kristin Heidel, University Hospital of Dresden, for excellent technical support.
Conflict of interest disclosure No financial interest/relationships with financial interest relating to the topic of this article have been declared.
References 1. Pittenger MF, Mackay AM, Beck SC, et al. Multilineage potential of adult human mesenchymal stem cells. Science. 1999;284:143–147. 2. Alakel N, Jing D, Muller K, et al. Direct contact with mesenchymal stromal cells affects migratory behavior and gene expression profile of CD133þ hematopoietic stem cells during ex vivo expansion. Exp Hematol. 2009;37:504–513. 3. Jing D, Fonseca AV, Alakel N, et al. Hematopoietic stem cells in coculture with mesenchymal stromal cellsdmodeling the niche compartments in vitro. Haematologica. 2010;95:542–550. 4. Uccelli A, Pistoia V, Moretta L. Mesenchymal stem cells: a new strategy for immunosuppression? Trends Immunol. 2007;28:219–226. 5. Zhao S, Wehner R, Bornhauser M, et al. Immunomodulatory properties of mesenchymal stromal cells and their therapeutic consequences for immune-mediated disorders. Stem Cells Dev. 2010;19:607–614. 6. List A, Dewald G, Bennett J, et al. Lenalidomide in the myelodysplastic syndrome with chromosome 5q deletion. N Engl J Med. 2006;355: 1456–1465. 7. Matsuoka A, Tochigi A, Kishimoto M, et al. Lenalidomide induces cell death in an MDS-derived cell line with deletion of chromosome 5q by inhibition of cytokinesis. Leukemia. 2010;24:748–755. 8. Davies F, Baz R. Lenalidomide mode of action: linking bench and clinical findings. Blood Rev. 2010;24(suppl 1):S13–S19. 9. Verhelle D, Corral LG, Wong K, et al. Lenalidomide and CC-4047 inhibit the proliferation of malignant B cells while expanding normal CD34þ progenitor cells. Cancer Res. 2007;67:746–755. 10. Oswald J, Boxberger S, Jorgensen B, et al. Mesenchymal stem cells can be differentiated into endothelial cells in vitro. Stem Cells. 2004;22:377–384.
876
M. Wobus et al./ Experimental Hematology 2012;40:867–876
11. Hideshima T, Chauhan D, Shima Y, et al. Thalidomide and its analogs overcome drug resistance of human multiple myeloma cells to conventional therapy. Blood. 2000;96:2943–2950. 12. Pellagatti A, Jadersten M, Forsblom AM, et al. Lenalidomide inhibits the malignant clone and up-regulates the SPARC gene mapping to the commonly deleted region in 5q- syndrome patients. Proc Natl Acad Sci U S A. 2007;104:11406–11411. 13. Jing D, Wobus M, Poitz DM, et al. Oxygen tension plays a critical role in the hematopoietic microenvironment in vitro. Haematologica. 2012; 97:331–339. 14. Hempel U, Hefti T, Kalbacova M, et al. Response of osteoblast-like SAOS-2 cells to zirconia ceramics with different surface topographies. Clin Oral Implants Res. 2010;21:174–181. 15. Lapidot T, Dar A, Kollet O. How do stem cells find their way home? Blood. 2005;106:1901–1910. 16. Schecter J, Galili N, Raza A. MDS: Refining existing therapy through improved biologic insights. Blood Rev. 2012;26:73–80. 17. Trento C, Dazzi F. Mesenchymal stem cells and innate tolerance: biology and clinical applications. Swiss Med Wkly. 2010;140:w13121. 18. Glennie S, Soeiro I, Dyson PJ, Lam EW, Dazzi F. Bone marrow mesenchymal stem cells induce division arrest anergy of activated T cells. Blood. 2005;105:2821–2827. 19. Varga G, Kiss J, Varkonyi J, et al. Inappropriate Notch activity and limited mesenchymal stem cell plasticity in the bone marrow of patients with myelodysplastic syndromes. Pathol Oncol Res. 2007; 13:311–319. 20. Munemasa S, Sakai A, Kuroda Y, et al. Osteoprogenitor differentiation is not affected by immunomodulatory thalidomide analogs but is promoted by low bortezomib concentration, while both agents suppress osteoclast differentiation. Int J Oncol. 2008;33:129–136. 21. Pontikoglou C, Deschaseaux F, Sensebe L, Papadaki HA. Bone marrow mesenchymal stem cells: biological properties and their role
22.
23.
24. 25.
26.
27.
28.
29.
30.
in hematopoiesis and hematopoietic stem cell transplantation. Stem Cell Rev. 2011;7:569–589. Ode A, Kopf J, Kurtz A, et al. CD73 and CD29 concurrently mediate the mechanically induced decrease of migratory capacity of mesenchymal stromal cells. Eur Cell Mater. 2011;22:26–42. Bodera P, Stankiewicz W. Immunomodulatory properties of thalidomide analogs: pomalidomide and lenalidomide, experimental and therapeutic applications. Recent Pat Endocr Metab Immune Drug Discov. 2011;5:192–196. Chanan-Khan AA, Cheson BD. Lenalidomide for the treatment of B-cell malignancies. J Clin Oncol. 2008;20:26. 1544–1552. Takedachi M, Oohara H, Smith BJ, et al. CD73-generated adenosine promotes osteoblast differentiation. J Cell Physiol. 2012;227:2622– 2631. Pellagatti A, Cazzola M, Giagounidis A, et al. Deregulated gene expression pathways in myelodysplastic syndrome hematopoietic stem cells. Leukemia. 2010;24:756–764. Wright N, de Lera TL, Garcia-Moruja C, et al. Transforming growth factor-beta1 down-regulates expression of chemokine stromal cellderived factor-1: functional consequences in cell migration and adhesion. Blood. 2003;102:1978–1984. Matsuda M, Morita Y, Hanamoto H, et al. CD34þ progenitors from MDS patients are unresponsive to SDF-1, despite high levels of SDF-1 in bone marrow plasma. Leukemia. 2004;18:1038–1040. Kumar S, Dispenzieri A, Lacy MQ, et al. Impact of lenalidomide therapy on stem cell mobilization and engraftment post-peripheral blood stem cell transplantation in patients with newly diagnosed myeloma. Leukemia. 2007;21:2035–2042. Popat U, Saliba R, Thandi R, et al. Impairment of filgrastiminduced stem cell mobilization after prior lenalidomide in patients with multiple myeloma. Biol Blood Marrow Transplant. 2009;15: 718–723.