Improved Left Ventricular Aneurysm Repair with Cell- and Cytokine-Seeded Collagen Patches

Abstract

Background: Engineered heart tissues (EHTs) present a promising alternative to current materials for surgical ventricular restoration (SVR); however, the clinical application remains limited by inadequate vascularization postimplantation. Moreover, a suitable and economic animal model for primary screening is another important issue.

Methods: Recently, we used 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride chemistry (EDC) to initiate a strengthened, cytokine-conjugated collagenous platform with a controlled degradation speed. In vitro, the biomaterial exhibited an enhanced mechanical strength maintaining a porous ultrastructure, and the constant release of cytokines promoted the proliferation of seeded human mesenchymal stem cells (hMSCs). In vivo, with the hMSC-seeded, cytokine-immobilized patch (MSCs + GF patch), we performed modified SVR for rats with left ventricular aneurysm postmyocardial infarction (MI). Overall, the rats that underwent modified SVR lost less blood and had lower mortality. After 4 weeks, the rats repaired with this cell-seeded, cytokine-immobilized patch presented preserved cardiac function, beneficial morphology, enhanced cell infiltration, and functional vessel formation compared with the cytokine-free (MSC patch), cell-free (GF patch), or blank controls (EDC patch). Furthermore, the degradable period of the collagen patch in vivo extended up to 3 months after EDC treatment.

Conclusions: EDC may substantially modify collagen scaffold and provide a promising and practical biomaterial for SVR.

Figure 1

Characterization of the scaffolds. (a) Scan electron microscopy (SEM 250x) showed that the collagen sponge maintained a similar porosity and porous size after EDC chemistry and cytokine immobilization. (b) Enzyme-linked immunosorbent assay (ELISA) demonstrated that both vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF) immobilized in scaffolds were released in a constant and controlled modes.

Figure 2

Proliferation of hMSCs in scaffolds in vitro. (a) Representative micrographs (100x) of HE staining for human mesenchymal stem cells (hMSCs) in PBS-treated scaffold (PBS), crosslinker (EDC/Sulfo-NHS)-treated scaffold, and scaffold immobilized with VEGF + PDGF (GF) after 2-day and 4-day cultivation. (b) The bar graph shows that cells increased in all scaffolds over time and grew most quickly in cytokine-immobilized scaffold (GF) (4D: ^∗ p < 0.05 GF versus PBS, EDC). No difference was identified between EDC and PBS scaffolds. (c) The CCK₈ assay performed on day 3 and day 7 after cell seeding illustrated similar results quantitatively (7 D, ^∗ p < 0.05 GF versus PBS, EDC).

Figure 3

Experimental design. Schematic representation of patch preparation, MI, SVR, and patch implantation, with experimental test groups shown.

Figure 4

Modification of animal model and patch degradation in vivo. (a) and (b) Surgical mortality and blood loss significantly decreased by changing the approach of SVR from the midsternum (control group) to the 4th left anterolateral intercostal space (experimental group). (mortality: ^∗ p < 0.05 Exp versus Con; blood loss: ^∗∗ p < 0.01 Exp versus Con). (c)–(e) 3 days after patch implantation, one rat in each group was sacrificed. Interventricular image identified the full-thickness replacement of the left ventricle by the patch. Four weeks after implantation, the patch had been endothelialized and became a part of the left ventricle. The dotted line (red) indicates the approximate edge of neoendocardium and collagen patch. (f) and (g) Representative micrographs of Masson Trichrome's staining (100x) of crosslinked patch. At 4 weeks after implantation, the collagenous networks in the core of the patch could be clearly identified; however, most of them degraded, and limited residual debris remained after 3 months (white arrow indicated).

Figure 5

Morphologic and recellularized analysis of patch 28 days after implantation in vivo. (a) Representative images of heart slices stained with Masson's Trichrome to indicate the patch thickness and patch recellularization (100x). Thicker neoendocardium and denser cellular distribution in scaffold were identified in the GF patch, MSC patch, and MSC + GF patch compared with the EDC control (dotted yellow line separated collagen patch from newly formed endocardium and epicardium). (b) Representative images of rat hearts showed the outer border of the patches (yellow broken line indicated). (c) Compared with the EDC patch, cytokine conjugation or hMSC seeding induced an insignificant increase in thickness. The combination of cytokines and cells resulted in the thickest patch (^∗ p < 0.05 MSC + GF versus GF; ^∗∗ p < 0.01 MSC + GF versus EDC); however, there was no significant difference among the EDC, GF, and MSC patches or between the MSC and MSC + GF patches. (d) Comparatively, hMSC seeding resisted patch expansion more effectively than cytokine immobilization because the patch area in both cell seeding groups was significantly smaller than its cell-free control (^∗ p < 0.05 MSC versus EDC; ^∗ p < 0.05 MSC + GF versus GF). Eventually, the combination of cells and cytokines led to the smallest patch (^∗∗ p < 0.01 MSC + GF versus EDC). Black broken line indicates patch area at the time of implantation.

Figure 6

Cardiac function 28 days after patch implantation in vivo. (a) Representative M mode echocardiographic images before myocardial infarction, before patch implantation (D₀), 7 days (D₇) and 28 days (D₂₈) after patch implantation. (b) Fractional shortening (FS%) and (c) fractional area change (FAC%) substantially decreased in all groups at a similar speed to a comparable level before patch implantation (Day₀). However, after patch implantation, the cardiac function depressed relatively slower in both cell seeding groups, which presented as the more properly preserved FS% and FAC% at 28 days after patch implantation (FS%: ^∗∗ p < 0.01 MSC + GF versus GF, EDC; ^∗∗ p < 0.01 MSC versus GF, EDC. FCA%: ^∗∗ p < 0.01 MSC + GF versus EDC; ^∗ p < 0.05 MSC versus EDC). (d) and (e) Representative pressure-volume loop in a steady state or in response to vena cava occlusion. The end-systolic elastance in the MSC + GF group (blue line) was compared with that in the other three groups (red lines).

Figure 7

Analysis of blood perfusion in tissues peri-patch 28 days after implantation in vivo. (a) Representative gross images indicated the vasculogenesis around the patch, and representative micrographs (200x) of immunostaining for α-SMA (red) and DAPI (blue) identified arterioles and nuclei, respectively. (b) Density of mature arterioles (defined as vessels with a cross-sectional area greater than 100 μm²) was significantly higher in the GF and MSC patches (^∗ p < 0.05 GF versus EDC; ^∗ p < 0.05 MSC versus EDC) and highest in the MSC + GF group (^∗∗ p < 0.01 MSC + GF versus EDC); however, there was no significant difference among the GF, MSC, and MSC + GF groups. (c) The average total cross-section area of mature arterioles and their corresponding area fraction illustrated the individual or synergetic effects of hMSCs or cytokines on patch arteriogenesis (^∗∗ p < 0.01 GF, MSC versus EDC; ^∗ p < 0.05 MSC + GF versus GF, MSC; ^∗∗∗ p < 0.001 MSC + GF versus EDC).