Phenotypic, Morphological and Adhesive Differences of Human Hematopoietic Progenitor Cells Cultured on Murine versus Human Mesenchymal Stromal Cells

Abstract

Xenogenic transplantation models have been developed to study human hematopoiesis in immunocompromised murine recipients. They still have limitations and therefore it is important to delineate all players within the bone marrow that could account for species-specific differences. Here, we evaluated the proliferative capacity, morphological and physical characteristics of human CD34(+) hematopoietic stem and progenitor cells (HSPCs) after co-culture on murine or human bone marrow-derived mesenchymal stromal cells (MSCs). After seven days, human CD34(+)CD133(-) HSPCs expanded to similar extents on both feeder layers while cellular subsets comprising primitive CD34(+)CD133(+) and CD133(+)CD34(-) phenotypes are reduced fivefold on murine MSCs. The number of migrating HSPCs was also reduced on murine cells suggesting that MSC adhesion influences cellular polarization of HSPC. We used atomic force microscopy-based single-cell force spectroscopy to quantify their adhesive interactions. We found threefold higher detachment forces of human HSPCs from murine MSCs compared to human ones. This difference is related to the N-cadherin expression level on murine MSCs since its knockdown abolished their differential adhesion properties with human HSPCs. Our observations highlight phenotypic, morphological and adhesive differences of human HSPCs when cultured on murine or human MSCs, which raise some caution in data interpretation when xenogenic transplantation models are used.

Figure 1. Expansion of human HSPCs cultured on murine or human MSCs and release of CD133⁺ membrane vesicles.

(a) CD34⁺ HSPCs (6 × 10⁴ cells) isolated from mobilized peripheral blood (starting material) were cultured for one week either on murine (m) or human (h) MSCs or fibronectin prior to their analysis by flow cytometry for CD34 and CD38 or CD133. Total cell count, number of cells harboring a given phenotype (as indicated) and their expansion are shown. Fold change between cultivation with m or hMSCs or without versus starting material is indicated. Ratio of HSPC expansion with and without feeder cell layer is shown. Mean and standard deviation from four independent experiments using four distinct donors of HSPCs and two m or hMSCs preparations are shown. Note the abscissa is at a logarithmic scale (top panel). Representative dot plots are displayed (bottom panels). In the starting material, most of hHSPCs are CD34⁺CD133⁺ while CD133⁺CD34^– ones are very rare consistent with CD34 selection. The colored quadrants correspond to the cell phenotypes indicated in the upper panel. (b) After 7 days of HSPC–MSC co-culture, conditioned media were subjected to differential centrifugation for 5 min at 300 × g, 20 min at 1,200 × g, 30 min at 10,000 × g and 60 min at 200,000 × g. Adherent hematopoietic cells (HSPCs) were harvested, and centrifuged 5 min at 300 × g. All pellets were analyzed by immunoblotting for CD133 (arrow) and actin. Note that only half of the material was loaded for the HSPC fraction, and 1,200 and 10,000 × g fractions are not shown since no CD133 was detected therein. A representative experiment is displayed. The amount of CD133-immunoreactive material in each fraction was quantified, and ratios of vesicles (200,000 × g)/cells (HSPCs + 300 × g) are presented (mean ± standard deviation). Four independent experiments using distinct donors of HSPCs and MSC preparations were performed. Asterisks indicate a significant difference as calculated by two-tailed paired student’s t-test (*p < 0.05; **p < 0.01). N.s., not significant (p ≥ 0.05).

Figure 2. Polarization and migration of human HSPCs growing on murine or human MSCs.

(a) Human (h) HSPCs growing on murine (m) MSCs for 7 days were analyzed by SEM. Six cellular shapes are depicted (top panel, models I-VI) with the corresponding photographs (middle and bottom panels). Adherent hematopoietic cells were either round with microvilli on their surface (I) or displayed large plasma membrane protrusions such as filopodium (II, arrowhead), magnupodium (III-IV, star) and lamellipodium (V, solid line). They could also develop a uropod at the rear pole (VI, asterisk). Models I-II show non-migrating cells (stop) whereas models III-VI exemplify migrating ones. Green arrows represent the direction of migration. (b) Individual frames taken from a time-lapse video show a hHSPC (green) moving on the surface of h or mMSCs after 24 hours of co-culture. Asterisk and white line indicate the rear and front pole of a migrating hHSPC, respectively (first frame). White spot marks its starting point (last frame). The elapsed time is shown in the upper-left corner. (c) Tracking diagrams depict the movement of 10 cells for 30 min. They were tracked after 24 hours and 7 days of co-culture on MSCs. (d) Box-and-whisker plots show the migration distance within 30 min of cells cultured on MSCs for 7 days. The boxes represent data from 25th–75th percentiles and 100% within the whiskers. Horizontal lines within the box represent median values. 100 cells were counted per condition (n = 3; CD34⁺ HSPCs were isolated from distinct donors and cultured on two different MSC preparations). No significant (n.s.) difference was observed as calculated by two-tailed paired student’s t-test (p ≥ 0.05). Scale bars: 5 μm (a), 20 μm (b,c).

Figure 3. The relative amount of human HSPCs exhibiting a migrating morphology is reduced on murine MSCs in comparison to human ones.

(a) Phase contrast pictures of human (h) HSPCs growing on either h (top panels) or murine (m; bottom panels) MSCs for 7 days. The absolute amounts of migrating and non-migrating hematopoietic cells, which have an elongated (green) or spherical (red) morphology, respectively, observed in each frames are indicated. Left panels show higher magnification views of areas indicated in the corresponding images on the right. (b,c) Numbers of hematopoietic cells exhibiting morphologies characteristic of non-migrating (NM, red) and migrating (M, green) cells, respectively (see Fig. 2), were counted after 24 hours and 7 days of co-culture on MSCs (b). Numbers CD133⁺ hHSPCs with a given morphology after 7 days are indicated (c). Three to four experiments were performed using CD34⁺ hHSPCs isolated from different donors, and at least 400 (b) or 100–200 (c) cells were analyzed per condition. Three asterisks indicate a significant difference as calculated by two-tailed paired student’s t-test (p < 0.001). N.s., not significant (p ≥ 0.05). Scale bars: 10 μm.

Figure 4. The adhesion forces of human HSPCs on murine or human MSCs are measured by AFM.

(a) Phase contrast image of a human (h) HSPC attached to a tipless AFM cantilever. MSCs growing on fibronectin-coated coverslips are visible (see Supplementary Fig. S1). The inset illustrates a side-view of the contact between a single hHSPC and a MSC. (b) Representative retract F-D curves of an adhesion measurement between hHSPC and h (top panel) or murine (m, bottom panel) MSC are depicted. The contact time was 60 sec. The minimum of the curves (with respect to the zero force level) is the maximum detachment force (F_D) needed to separate both cells. (c) Box-whisker plots show F_D of freshly isolated CD34⁺ hHSPC measured from h (purple bars) or m (blue bars) MSC after a contact time of 2, 10 or 60 sec in absence or presence of 5 mM EGTA (EGTA, 60 sec contact time). (d) Box-whisker plots show F_D of CD34⁺ HSPCs pre-cultured for seven days either on h (left panel) or m (right panel) MSCs prior to force measurements on h (purple bars) or m (blue bars) MSC with a contact time of 60 sec. Box-whisker plots represent half of the data points within the box and 80% within the whiskers. Horizontal lines within the box represent median values. Numbers within bars (<n>) show the total number of analyzed F_D curves. Four experiments were performed for each condition using CD34⁺ hHSPCs isolated from different donors. MSCs were derived from one donor or one murine cell preparation, but additional data with other donors/cell preparations are presented in Supplementary Fig. S6. The pairwise comparison of data was performed using Mann-Whitney-U-test (**p < 0.01; ***p < 0.001; n.s., not significant, ≥0.05).

Figure 5. Mesenchymal cell-associated N-cadherin is the major player in the differential adhesion of human HSPCs.

(a) Human (h) or murine (m) MSCs were incubated without (–) or with (+) sulfo-NHS-LC-biotin prior to solubilization. Biotinylated proteins were detected using HRP-conjugated streptavidin. The purple and blue asterisks indicate the major cell surface proteins in h and mMSCs, respectively. The black ones indicate the potential N-cadherin signal. (b) Expression level of N-cadherin in h versus mMSCs. N-cadherin was detected by immunoblotting using 32/N-cadherin antibody and its expression was normalized to α-tubulin. Values are relative to that of hMSC (n = 5, MSCs were derived from three distinct donors or murine preparations). (c) Flow cytometry analysis of PFA-fixed, saponin-permeabilized cells labeled with two distinct N-cadherin antibodies (clones 32/N-cadherin and GC-4) followed by an anti-mouse DyLight^®649-conjugated secondary antibody. The antigen expression (cyan) and the appropriate isotype control (black) are shown. The mean fluorescence intensity appears in the right-top corner. (d,e) Human (d) or m (e) MSCs were either untransfected or transfected without (MOCK) or with negative control or N-cadherin siRNAs. N-cadherin knockdown was confirmed 20 hours after transfection by immunoblotting, and its expression normalized to α-tubulin (n = 3). (f) Box-whisker plots show F_D of freshly isolated CD34⁺ HSPC from h (purple) or m (blue) MSC measured 16–20 hours after transfection with a contact time of 60 sec. For comparison, the data obtained with untransfected MSCs and EGTA were presented (Fig. 4). For each transfection experiments (n = 3), CD34⁺ HSPCs were derived from distinct donors. Box-whisker plots represent half of the data points within the box and 80% within the whiskers. Horizontal lines within the box represent median values. Numbers within bars (<n>) show the total number of analyzed F_D curves. The pairwise comparison of data was performed using Mann-Whitney-U-test (***p < 0.001; n.s., not significant, p ≥ 0.05).