Bone marrow-derived mesenchymal stem cells facilitate engineering of long-lasting functional vasculature

Abstract

Vascular tissue engineering requires a ready source of endothelial cells and perivascular cells. Here, we evaluated human bone marrow-derived mesenchymal stem cells (hMSCs) for use as vascular progenitor cells in tissue engineering and regenerative medicine. hMSCs expressed a panel of smooth muscle markers in vitro including the cardiac/smooth muscle-specific transcription coactivator, myocardin. Cell-cell contact between endothelial cells and hMSCs up-regulated the transcription of myocardin. hMSCs efficiently stabilized nascent blood vessels in vivo by functioning as perivascular precursor cells. The engineered blood vessels derived from human umbilical cord vein endothelial cells and hMSCs remained stable and functional for more than 130 days in vivo. On the other hand, we could not detect differentiation of hMSCs to endothelial cells in vitro, and hMSCs by themselves could not form conduit for blood flow in vivo. Similar to normal perivascular cells, hMSC-derived perivascular cells contracted in response to endothelin-1 in vivo. In conclusion, hMSCs are perivascular cell precursors and may serve as an attractive source of cells for use in vascular tissue engineering and for the study of perivascular cell differentiation.

Figure 1

Transcriptional profile of endothelial and smooth muscle markers in hMSCs. (A) RT-PCR analysis of endothelial markers (VE-cadherin and PECAM1) in hMSCs at baseline and in hMSCs cocultured for 3 days with MS1 mouse endothelial cell line. (B) RT-PCR analysis of smooth muscle markers' mRNA in human mesenchymal stem cells, human aortic vascular smooth muscle cells (CRL-1999), and human dermal fibroblasts (CRL-2575). Water was used as negative control. (C-F) Protein expression of smooth muscle markers in hMSCs was confirmed by immunohistochemistry (α-smooth muscle actin [C], SM22α [D], desmin [E], and NG2 [F], yellow; DAPI, blue). Scale bar represents 50 μm.

Figure 2

Induction of myocardin transcription and migration in hMSCs by endothelial cells. (A) Quantitative real-time PCR was performed to assess the induction of myocardin in hMSCs by various conditions after 2 days (hMSCs alone, hMSCs cultured with MS1 endothelial cells with contact, hMSCs cultured with MS1 without contact by transwell culture, and hMSCs stimulated with TGFβ1 [10 ng/mL]). Values expressed as fold increase above hMSC-alone levels and normalized by GAPDH. Representative data of 3 separate experiments. (B) Transwell migration assay was performed to assess endothelial cell–induced hMSC migration. Imatinib mesylate at 0.2 μM, 2 μM, and 10 μM was added to test for inhibition of hMSC migration. #P < .001 (medium vs HUVECs); *P < .001 (2 μM imatinib mesylate vs HUVECs); **P < .001 (10 μM imatinib mesylate vs HUVECs).

Figure 3

hMSCs stabilized engineered blood vessels in vivo. EGFP-labeled HUVECs were either implanted alone (A) or coimplanted with hMSCs (B) or 10T1/2 cells (C) in a collagen gel onto cranial windows in SCID mice. Images were taken at periodic time points with multiphoton laser scanning microscope to monitor the in vivo dynamics of vascularization by the implanted endothelial cells. Green indicates HUVECs expressing EGFP; red, functional blood vessels contrast-enhanced by rhodamine-dextran. Scale bar represents 50 μm.

Figure 4

Quantification of functional engineered blood vessel density. Quantification of functional vessel density in SCID mice implanted with HUVECs alone or HUVECs with hMSCs (A) and HUVECs with hMSCs or 10T1/2 (B) (results are mean ± SEM, n = 3 in panel A and n = 5 in panel B; different batches of HUVECs were used in panels A and B, resulting in variation in vessel density between the 2 experiments). (C) The hMSC-stabilized vascular network remained functional for more than 130 days in vivo. Scale bar represents 100 μm.

Figure 5

Intravital monitoring of EGFP-hMSCs in a tissue-engineered vessel model. Fibronectin-collagen constructs with EGFP-hMSCs alone (A) or EGFP-hMSCs with DsRed-HUVECs (B-G) were implanted into cranial windows of SCID mice. Images were taken at different time points with multiphoton laser scanning microscopy (MPLSM) (B, day 7; C,D, day 19; E,F, day 31; G, day 83). Green indicates human mesenchymal stem cells expressing EGFP (A-G); red, HUVECs expressing DsRed-express fluorescent protein (B-E,G); and red, functional blood vessels contrast-enhanced by rhodamine-dextran (D,F). Scale bars represent (A,B,G) 100 μm; (C-F) 50 μm.

Figure 6

Expression of smooth muscle cell markers in hMSC-derived cells incorporated into the tissue-engineered vessels. Whole-mount staining was performed for the extracted tissue-engineered vessel constructs. (A-H) Confocal microscopy images. (A) Blue indicates DAPI staining; (B) red, DsRed-HUVECs; (C,F,I) green, EGFP-hMSCs; (D) blue, α-smooth muscle actin staining; (G) blue, SM22α staining; and (J) blue, desmin staining. Region of colocalized staining of EGFP and smooth muscle markers is highlighted as white (E,H,K). Panels A-E, F-H, and I,J are the same location/construct. Scale bar represents 100 μm.

Figure 7

Endothelin-1 stimulated vasoconstriction of the engineered blood vessels. HUVEC/hMSC-derived engineered blood vessels were superfused with 100 nM endothelin-1 to stimulate vasoconstriction in vivo. (A) Representative multiphoton microscopy images during endothelin-1 superfusion. Green indicates EGFP-hMSCs; red, functional blood vessels contrast-enhanced by rhodamine-dextran. Scale bar represents 50 μm. (B) The blood vessel diameter was quantified over a 20-minute period (n = 4). The value was expressed as a fraction of the original diameter. Scale bar represents 50 μm.