Clonal analysis of multipotent stromal cells derived from CD271+ bone marrow mononuclear cells: functional heterogeneity and different mechanisms of allosuppression

Abstract

Previous reports demonstrated a relationship between proliferation potential and trilineage differentiation in mesenchymal stromal cell-derived clones generated using plastic adherence (PA-MSCs). However, there are no reports presenting a clonal analysis of the proliferative potential, differentiation potential and allosuppressive effects of human mesenchymal stromal cell subsets. In this study, we performed a clonal analysis of mesenchymal stromal cells generated from human CD271(+) bone marrow mononuclear cells (CD271-MSCs). After transfection with the gene encoding green fluorescent protein, the cells were single-cell sorted and cultured for 2-4 weeks. A population doubling analysis demonstrated that 25% of CD271-MSC clones are fast-proliferating clones compared to only 10% of PA-MSC clones. Evaluation of the allosuppressive potential demonstrated that 81.8% of CD271-MSC clones were highly allosuppressive compared to only 58% of PA-MSC clones. However, no consistent correlation was observed between allosuppression and proliferative potential. Prostaglandin E2 levels were positively correlated with the allosuppressive activity of individual clones, suggesting that this molecule may be a useful predictive biomarker for the allosuppressive potential of mesenchymal stromal cells. In contrast, inhibitory studies of indoleamine 2,3 dioxygenase indicated that none of the clones used this enzyme to mediate their allosuppressive effect. Differentiation studies revealed the presence of tripotent, bipotent and unipotent CD271-MSC and PA-MSC clones which suppressed the allogeneic reaction to differing extents in vitro. In conclusion, our findings demonstrate differences between CD271-MSCs and PA-MSCs and indicate that neither proliferation potential nor differentiation potential represents a consistent predictive parameter for the immunomodulatory effects of either type of mesenchymal stromal cells.

Figure 1.

The experimental protocol used in this study. Passage 1 MSCs of both types were transfected with GFP and then single-cell sorted by FACSAria cell sorter. After 16–35 days in culture single-cell derived MSC-clones were assessed for their proliferative potential (PD: population doublings; DT: doubling time), immunosuppressive potential in mixed lymphocyte culture (MLR) and their differentiation potential along different lineages

Figure 2.

Clonogenic potential of single sorted MSCs and their phenotype. (A) On Day 1 post-sorting each well was monitored for GFP-positive cells under a fluorescence microscope, and wells containing more than one cell were excluded from the study (magnification: 100x). (B) Developing colony-forming unit-fibroblast (CFU-F) on Day 7 (magnification: 40x). (C) Confluent CFU-F colony at Day 21. These CFU-Fs were detached, expanded and then used to assess their proliferative, differentiation and allosuppressive potential (magnification 20x). (D) Clonogenic potential of single-sorted CD271-MSCs and PA-MSCs. Data are presented as a mean±SEM of single-sorted CD271-MSCs or PA-MSCs that were able to give rise to CFU-Fs. (E) Clonally-derived CD271-MSCs were negative for CD45 and CD34, while expressed typical MSC antigens (colored histograms represent isotype control, whereas red lines represent percentage of MSCs expressing the antigens).

Figure 3.

Composition of single derived colonies from CD271-MSCs. (A) Most of the clones showed a typical spindle-shaped MSC-morphology at week 3 in culture. The majority of the cells expressed high levels of CD73 (magnification: 40x). (B) A percentage of these clones revealed an endothelial-like morphology being positive for von Willebrand factor (magnification: 100x). (C) The other clones demonstrated an osteoblast-like morphology and expressed CD9 antigen (magnification: 100x).

Figure 4.

Cellularity and proliferation kinetics of single-cell derived MSC-clones. Clonally-derived MSCs of both types were harvested at different time-points and analyzed for their cellularity (A), proliferation rate during the culture (B), number of population doublings achieved within their expansion phase (C) and accordingly the time needed to duplicate themselves-doubling time (D). Circles present the CD271-MSC- derived clones, whereas triangles present the PA-MSC-derived clones.

Figure 5.

Allosuppressive potential of single-MSC derived clones and non-cloned MSCs. (A). The allosuppressive effect of clones derived from CD271-MSC and PA-MSCs which have been generated from 2 bone marrow donors. Triangles represent the MSC-clones of the first donor, whereas circles represent MSC-clones from the second donor. (B) CD271-MSC derived clones, which whose allosuppressive effects (red bars) are completely mediated by PGE2, because treatment with indomethacin abrogated completely this effect (green bars). (C) CD271-MSC derived clones which use partially PGE2 as a mediator of allosuppressive activity (red bars), while treatment with indomethacin can only partially reverse the inhibitory effect on proliferation of MNCs (green bars). (D) CD271-MSC derived clones, which are PGE2-independent, because the treatment with indomethacin as an inhibitor of PGE2 synthesis (green bars), was not able to abrogate the inhibitory potential of clones on proliferation of the mononuclear cells in the presence of MSC-clones (red bars).

Figure 6.

Allosuppressive and proliferation potential of the individual CD271-MSC derived clones and non-cloned CD271-MSCs. (A) In this figure is shown the relationship between the allosuppressive effect of single clones in MLR and PGE2 levels as a mediator of this effect in the appropriate supernatants (n=6 clones). (B) Relationship between the allosuppressive and proliferation potential of the individual CD271-MSC derived clones (n=20). (C) Inhibition of indoleamine 2, 3 dioxygenase (IDO) in a PGE2-dependent clone and a PGE2-independent clone (D). (E) Allosuppressive potential of non-cloned CD271-MSCs from 2 bone marrow donors, whose MSCs were used as a source for generation of the clones after single-cell sorting. Approximately 60% of this effect was mediated by PGE2, as the treatment with indomethacin reversed partially this effect. (F) PGE2 levels increased significantly (P<0.0006) when allogeneic MNCs were co-incubated with third-party CD271-MSCs, and decreased significantly after addition of indomethacin as an inhibitor of PGE2-synthesis (P<0.0007).

Figure 7.

Trilineage differentiation of MSC-derived clones and their allosuppressive potential. (A) Clones generated from CD271-MSC (n=15) and PA-MSCs (n=10) were cultured in induction media for adipocytic, osteoblastic and chondrocytic differentiation. Green bars represent the clones derived from CD271-MSCs, red bars represent the clones derived from PA-MSCs. A: adipocytic; O: osteoblastic; C: chondrocytic; ND: non-differentiated. In addition to trilineage differentiation, six CD271-MSC derived clones (B) and six PA-MSC derived clones (C) were assessed for their allosuppressive potential in mixed lymphocyte reaction (MLR). (D) Representative examples for trilineage differentiation of CD271-MSCs and PA-MSC derived clones (E) into adipocytes, osteoblasts and chondrocytes. Magnification used for microscopic examination of adipocytes was 20x, for osteoblasts 10× and for chondrocytes 40x.