Bone marrow-derived mesenchymal stromal cells differ in their attachment to fibronectin-derived peptides from term placenta-derived mesenchymal stromal cells

Abstract

Introduction: Human mesenchymal stromal cells (MSCs) can be isolated from different sources including bone marrow and term placenta. These two populations display distinct patterns of proliferation and differentiation in vitro. Since proliferation and differentiation of cells are modulated by cell-matrix interactions, we investigated the attachment of MSCs to a set of peptide-coated surfaces and explored their interactions with peptides in suspension.

Methods: Human MSCs were isolated from bone marrow and term placenta and expanded. Binding of MSCs to peptides was investigated by a cell-attachment spot assay, by blocking experiments and flow cytometry. The integrin expression pattern was explored by a transcript array and corroborated by quantitative reverse transcription polymerase chain reaction and flow cytometry.

Results: Expanded placenta-derived MSCs (pMSCs) attached well to surfaces coated with fibronectin-derived peptides P7, P15, and P17, whereas bone marrow-derived MSCs (bmMSCs) attached to P7, but barely to P15 and P17. The binding of bmMSCs and pMSCs to the peptides was mediated by β1 integrins. In suspension, expanded bmMSCs barely bind to P7, P13, P15, and less to P14 and P17. Ex vivo, bmMSCs failed to bind P7, but displayed a weak interaction with P13, P14, and P15. In suspension, expanded pMSCs displayed binding to many peptides, including P4, P7, P13, P14, P15, and P17. The differences observed in binding of bmMSCs and pMSCs to the peptides were associated with significant differences in expression of integrin α2-, α4-, and α6-chains.

Conclusions: Human bmMSCs and pMSCs show distinct patterns of attachment to defined peptides and maintain differences in expression of integrins in vitro. Interactions of ex vivo bmMSCs with a given peptide yield different staining patterns compared to expanded bmMSCs in suspension. Attachment of expanded MSCs to peptides on surfaces is different from interactions of expanded MSCs with peptides in suspension. Studies designed to investigate the interactions of human MSCs with peptide-augmented scaffolds or peptides in suspension must therefore regard these differences in cell-peptide interactions.

Fig. 1

Attachment assay with human MSCs. Human MSCs from bone marrow (a) and term placenta (b) were incubated on spots coated with peptides P2–P17, laminin-111 (LM), or fibronectin (FN) as indicated. After 15 min of incubation non-adherent cells were washed away and MSCs attached are visualized by microphotography. Spots coated with activated bovine serum albumin (aBSA) served as controls

Fig. 2

Overview on attachment patterns of human mesenchymal stromal cells (MSCs). a Human bmMSCs (n = 14 donors, black bars) or pMSCs (n = 10 donors, grey bars) were expanded to investigate their attachment to peptides P2–P17 as indicated. Attachment of MSCs to laminin-111 (LM) and fibronectin (FN) served as positive control (100 %), attachment to activated bovine serum albumin (aBSA) served as negative control (0 %). Attachment of human DF to the same substratum was tested in comparison to the MSCs (n = 5 donors, hatched bars). b Attachment of primary culture human MSCs to peptides P7, P14, P15, and P17 was tested prior to deep freezing (naïve MSCs) and compared to the same cells after cryopreservation and revitalization (cryo MSCs). Overall, naïve MSCs attached to the peptides at 43 ± 18 % mean efficacy compared to FN (=100 %). Cryopreservation caused a reduced overall attachment to these peptides of all MSCs investigated (19.2 ± 14.2 %). Thus, cryopreservation yielded a significant drop of attachment of human MSCs (n = 10, mean reduction 0.42 ± 0.20, p ≤ 0.0048)

Fig. 3

Contribution of integrins to protein- and peptide- dependent cell attachment. Human bmMSCs (a) or pMSCs (b) were pre-incubated with a function-blocking antibody to CD29 (integrin β1 chain) or remained untreated as indicated (ø). The cells were then incubated on peptide- or fibronectin (FN)-coated spots. Pre-incubation of bmMSCs (a) or pMSCs (b) with anti-CD29 mAb completely blocked the attachment of the cells, whereas untreated MSCs attached well to peptides or proteins. Pre-incubation of bmMSCs with anti-CD90 mAb failed to block attachment of the cells confirming the specificity of the blocking reaction (a, upper right). Incubation of pMSCs on activated bovine serum albumin (aBSA) did not cause unspecific binding of cells to this reagent (b). Human bmMSCs were labeled with PKH26 and attachment of PKH26-loaded cells to FN was confirmed (c, left). Human fibroblasts were labeled with PKH67, mixed 1:1 with PKH26-labeled MSCs and incubated on fibronectin (c, right). DF competed for binding sites and displaced the MSCs (c, right). Human bmMSCs were loaded with Calcein-AM and EthD-1 to discriminate viable cells (green cytoplasm, v) from dead cells (red nuclei, d) and added to LM-111-coated spots (d). The MSCs attached presented a bright green fluorescence indicating a high viability of the population studied (d, v). Only a few dead cells were observed (d, d)

Fig. 4

Interaction of peptides with human MSCs in suspension. Human bmMSCs (a) and pMSCs (b) were expanded to the second passage of in vitro culture, harvested by mild proteolysis, and stained with titrated amounts of FITC-labeled peptides. In addition, bmMSCs were counterstained with PE-labeled anti-CD90 (a), and pMSCs with PE-labeled anti-CD73 (b) to confirm their mesenchymal origin. Moderate binding of peptides P7, P13, and P15 was observed on bmMSCs (a). In contrast, pMSCs displayed a bright staining with all peptides investigated (b). FITC Fluorescein isothiocyanate, FSC Forward scatter, PE Phycoerythrin, SSC Side scatter

Fig. 5

Interaction of peptides with human bmMSCs ex vivo. Mononuclear cells were purified from human bone marrow by Ficoll® gradient centrifugation, washed and incubated with peptides as indicated and counterstained with AF-647-labeled mAb to CD271, a marker for bmMSCs. Peptide P13 stained a small subset of CD271^pos bone marrow cells ex vivo, whereas peptides P4, P14, and P15 stained only very few CD271^pos cells. Peptides P7 failed to bind to CD271^pos cells, but all peptides interacted with CD271^neg bone marrow cells. FITC Fluorescein isothiocyanate, FSC Forward scatter, SSC Side scatter

Fig. 6

Detection of integrins on bmMSCs and pMSCs. Human bmMSCs (a) or pMSCs (b) were expanded to the second passage to investigate the expression of integrin α2, α4, α6, α7, and β1 by flow cytometry. Expression of integrin α2, α4, α6, and β1 was lower on bmMSCs compared to pMSCs (solid histograms). MSCs incubated with PE-labeled anti-Ig antibody served as control (dotted histograms). For integrin α7, no difference in expression levels was observed by flow cytometry. PE Phycoerythrin