In Vivo Tracking and 1H/19F Magnetic Resonance Imaging of Biodegradable Polyhydroxyalkanoate/Polycaprolactone Blend Scaffolds Seeded with Labeled Cardiac Stem Cells.

Medium-chain length polyhydroxyalkanoates (MCL-PHAs) have demonstrated exceptional properties for cardiac tissue engineering (CTE) applications. Despite prior work on MCL-PHA/polycaprolactone (PCL) blends, optimal scaffold production and use as an alternative delivery route for controlled release of seeded cardiac progenitor cells (CPCs) in CTE applications in vivo has been lacking. We present herein applicability of MCL-PHA/PCL (95/5 wt %) blends fabricated as thin films with an improved performance compared to the neat MCL-PHA. Polymer characterization confirmed the chemical structure and composition of the synthesized scaffolds, while thermal, wettability, and mechanical properties were also investigated and compared in neat and porous counterparts. In vitro cytocompatibility studies were performed using perfluorocrown-ether-nanoparticle-labeled murine CPCs and studied using confocal microscopy and 19F magnetic resonance spectroscopy and magnetic resonance imaging (MRI). Seeded scaffolds were implanted and studied in the postmortem murine heart in situ and in two additional C57BL/6 mice in vivo (using single-layered and double-layered scaffolds) and imaged immediately after and at 7 days postimplantation. Superior MCL-PHA/PCL scaffold performance has been demonstrated compared to MCL-PHA through experimental comparisons of (a) morphological data using scanning electron microscopy and (b) contact angle measurements attesting to improved CPC adhesion, (c) in vitro confocal microscopy showing increased SC proliferative capacity, and (d) mechanical testing that elicited good overall responses. In vitro MRI results justify the increased seeding density, increased in vitro MRI signal, and improved MRI visibility in vivo, in the double-layered compared to the single-layered scaffolds. Histological evaluations [bright-field, cytoplasmic (Atto647) and nuclear (4',6-diamidino-2-phenylindole) stains] performed in conjunction with confocal microscopy imaging attest to CPC binding within the scaffold, subsequent release and migration to the neighboring myocardium, and increased retention in the murine myocardium in the case of the double-layered scaffold. Thus, MCL-PHA/PCL blends possess tremendous potential for controlled delivery of CPCs and for maximizing possible regeneration in myocardial infarction.

Introduction

Stem cell (SC) delivery has been proposed and applied as a novel and promising therapeutic approach for cardiac diseases.[1−3] However, to-this-date, there is continued speculation over its efficacy, given the disparity of published preclinical and clinical results,[4,5] despite scientific evidence for the existence of paracrine effects associated with beneficial functional improvements in the infarcted myocardium.[6] Therefore, controlled SC administration and release still presents tremendous challenges toward a therapeutically successful cell engineering approach.

Polymer scaffolds have been introduced as a novel, biomimetic approach for administration and controlled release of viable SCs to the diseased myocardium,[7] ultimately aiming to replace scar tissue with engrafted healthy cells with pluripotent/multipotent capacity.

Historically, the earliest attempts for tissue regeneration date to the early 1970s with the use of grafts over the injured myocardium.[8] The first efforts to deliver cells within scaffolds were pioneered by Souren et al.,[9] while more recent attempts target the use of inducible pluripotent SCs.[10]

To this-date, collective efforts in cardiac tissue engineering (CTE) have employed natural and synthetic materials and have been tested with a multitude of SC types[8,11] in different hosts, including the mouse, rat, rabbit, pig, sheep, dog, and human.[8] The scaffold substrates have been synthesized using materials, such as collagen, fibrin, chitosan, alginate,[12] hyaluronic acid, gelatin, matrigel, decellularized extracellular matrix,[13,14] and hydrogels.[15,16]

In recent years, the scientific preference has shifted toward scaffolds synthesized using natural materials given their biodegradable and biocompatible properties, emulating closely the myocardial microenvironment. Specifically, medium-chain length polyhydroxyalkanoates (MCL-PHA) have generated great interest over recent years as functional materials in cardiac tissue engineering, with biodegradable, biocompatible, and synthetically tunable physical properties,[17] including flexibility, crystallinity, melting point, and glass transition temperatures.[18] Induced porosity and functionalization of these materials with growth factors and active peptide molecules (without and with electrospun blend fibers) has also led to the fabrication of cardiac patches with excellent morphological properties and upregulation of cues from the innate matrix structure, thereby promoting enhanced SC (endothelial, inducible pluripotent) adhesion and proliferation.[17,19] Furthermore, MCL-PHA patches have been shown to exhibit similar mechanical properties to myocardial muscle, with a flexibility and rigidity that can sustain the cyclic strain patterns developed during the contractile–relaxation phases of the cardiac cycle, over prolonged periods that span multiple weeks.[19] In addition, these polymers exhibit degradation by surface erosion, leading to the maintenance of the scaffold’s structure for a longer period of time, and hence provide better support functionality. Overall, MCL-PHAs are versatile, biocompatible, biodegradable, sustainable, display thermoplastic and elastomeric properties, and have predictable mechanical and physical characteristics.[20] They can be mass-produced using controlled fermentations,[21] and their biodegradation products are much less acidic (with hydroxyalkanoic acids as the main degradation byproducts) as compared to those of poly(l-lactic acid) (PLLA) and PLGA, thereby leading to less severe inflammatory responses. However, MCL-PHAs are associated with high production costs.

In comparison to MCL-PHAs, synthetic materials, such as PLLA,[22] are readily available and are associated with well-established processing conditions; however, PLLA films are much stiffer than MCL-PHAs and exhibit poorer degradation responses (bulk erosion),[19] ultimately causing plastic deformation and failure during long-term exposure to cyclic strain.[17] Additionally, PLLA degradation products are acidic in nature, which ultimately leads to inflammation and other undesirable reactions.[23]

Polycaprolactone (PCL)[24] is an elastomeric synthetic polymer that is adaptable, versatile, produced in large scale, is associated with precise synthetic control using easily accessible materials and prolonged degradation time,[25,26] has good miscibility with various polymers, and possesses good and predictable mechanical and physical characteristics, often used in load-bearing tissues to enhance stiffness,[25] although associated with very low glass transition and melting temperatures.[27]

Despite these advantages, PCL is a nonsustainable polymer, and its production methods include solvent- and catalyst-based synthesis. In addition, PCL does not exhibit structural variability and, consequently, the resulting range of material properties exhibited by MCL-PHAs.

In view of these facts, polymeric MCL-PHA/PCL blends were used in this work to overcome the limitations of each of these polymer families, allowing the alteration of the mechanical properties of PCL and thereby enhancing the processability of the combined polymer product. Despite the existence of prior published work on MCL-PHA/PCL blends,[25,26,28−32] the reported compositions, aging, testing, and targeted applications have been distinctly different from this work, and they have not been previously used for CTE applications in either unseeded or seeded forms.

Correspondingly, in this effort, we aimed to synthesize MCL-PHA/PCL blend scaffold compositions seeded for the first time with cardiac progenitor cells (CPCs) enhancing the elicited benefits by capitalizing on the advantages of each material class.

Critical to the successful administration and controlled release of SCs by the scaffold is our ability to monitor them noninvasively, particularly in a temporal manner.[33] Even though bioluminescence imaging using reporter genes (such as luciferase) has allowed real-time, in vivo monitoring of viable SCs implanted on scaffolds,[34] the spatial localization of the scaffold and its degradability pattern are not easily discernible.

In this work, we propose the use of novel, functional biodegradable, biocompatible natural/synthetic polymer blend scaffolds (composed of MCL-PHA and PCL) to (a) benefit from the material properties of natural and synthetic polymers, (b) achieve controlled delivery of homologous CPCs—previously shown to elicit beneficial functional effects (in the chronic phase) following acute, reperfused myocardial infarction (MI), in preclinical[6] and clinical studies[4]—thereby aiming to increase retention of delivered cells to the murine myocardium, (c) extend the temporal window over which the release of labeled CPCs is achieved compared to traditional direct injection techniques,[35] and (d) use 19F magnetic resonance imaging (MRI)/magnetic resonance spectroscopy (MRS) to noninvasively detect and monitor the cells temporally.

The choice of the blend material underlies one of the important novelties of this work, given the increased ability of structural control in terms of its elastomeric nature and mechanical properties, the increased glass transition temperatures compared to neat PCL, and its potential to integrate with the myocardial network and be conjugated with bioactive molecules, such as vascular endothelial growth factor (VEGF) and Arg-Gly-Asp (RGD)/Tyr-Lle-Gly-Ser-Arg (YIGSR) peptides to further increase cellular attachment, viability, and proliferation.[17,19] Another novelty relates to the visualization and monitoring of the repair process using 19F MRI. The latter aim is pursued in a physiological model, despite the challenge of the low seeding density, as this is governed by the small size of the murine heart, stringent anatomical limitations regarding the scaffold’s placement, the endogenous hypoxic conditions, and the temporally dependent migration/dispersion of cells within the murine epicardium/mesocardium.

Experimental Procedures

Production of Poly-3-hydroxyalkanoates (PHAs)

Production of MCL-PHAs

MCL-PHAs were produced by Pseudomonas mendocina CH50 using glucose as the sole carbon source under nitrogen limiting conditions.[36] Batch fermentation was carried out in a 15 L bioreactor with an operating fermenter volume of 10 L. MCL-PHA production was completed in three stages:

Preparation of Inoculum

A single colony of P. mendocina CH50 was used to inoculate sterile nutrient broth. The broth was incubated at 30 °C, at 200 rpm for 16 h.

Preparation of Second Stage Seed Culture

The inoculum prepared at the first stage was used to inoculate a second stage mineral salt medium (MSM) with glucose as the carbon source. It was incubated at 30 °C, at 200 rpm for 24 h.

Preparation and Inoculation of Production Stage Media

Prior to the inoculation of the production media, the fermenter was sterilized at 121 °C for 30 min. The sterile mixture of MSM, glucose, magnesium sulfate, and trace element solution was aseptically added to the fermenter. The second stage seed culture was used to inoculate the production media. The culture was grown for a period of 48 h at 30 °C and at 200 rpm.

Extraction of MCL-PHA

The biomass was recovered by centrifuging the fermented cultures at a speed of 4600 rpm for approximately 45 min. The obtained biomass was lyophilized prior to extraction. MCL-PHA was extracted using the two-stage soxhlet extraction method. The powdered biomass was refluxed in methanol for 24 h to remove the impurities. The polymer was then extracted in chloroform for another 24 h. The solution was concentrated using a rotary vacuum evaporator and the polymer was precipitated using ice-cold methanol (1:10 polymer solution to ice-cold methanol).

Polymer Characterization

Chemical Characterization

Fourier Transform Infrared Spectroscopy (FTIR)

FTIR analysis was also conducted using a Perkin-Elmer Spectrum Two spectrometer, as described earlier.[17]

Nuclear Magnetic Resonance (NMR)

NMR was conducted on a 16.4 T (700 MHz), high-field spectrometer (Bruker, Billerica, MA, USA) at University College, London. Samples were prepared in accordance to standard methodological procedures, as reported earlier.[36,37]

Thermal Characterization

Thermal properties of the polymer, such as the melting temperature (Tm) and the glass transition temperature (Tg), were determined using a differential scanning calorimetry (DSC) system (model DSC 214, Polyma, Netzsch, Germany), equipped with an intracooler IC70 cooling system. A polymer sample (5 mg) was heated from −70 to 170 °C at a heating rate of 10 °C per minute over two successive heating–cooling cycles. The analysis was completed at the end of two heating cycles. DSC thermograms were analyzed using the Proteus 7.0 software.

Molecular Weight Analysis

The number average molecular weight, (Mn), and the weight average molecular weight, (Mw), of the polymer were determined using gel permeation chromatography (GPC, model 1260 Infinity GPC, Agilent Technologies). The polymer solution (2 mg/mL) was introduced into the GPC system at a flow rate of 1 mL/min. The system was equipped with a 5 μm PLgel MIXED-C (300 × 7.5 mm) column calibrated using narrow molecular weight polystyrene standards from 162 Da to 15 kDa. The eluted polymer was detected with a refractive index detector. The data were analyzed using the Agilent GPC/SEC software.

Scaffold Synthesis and Characterization

MCL-PHA films were prepared using the solvent casting method.[38] Nonporous MCL-PHA films were prepared by dissolving the polymer (0.5 g) in chloroform (10 mL). The MCL-PHA solution was poured into a glass Petri dish and was allowed to dry in a closed chamber. Polymer blends were prepared by dissolving 0.5 g of polymer with 0.26 g of PCL (Sigma-Aldrich, UK).

To prepare porous films, 1.7 g of sodium chloride with particular sizes (75, 100 μm) was used as the porogen. These were added into the polymer solution, mixed, and then poured into glass Petri dishes. Upon drying, these films were immersed in water to allow sodium chloride to leach out of the films.[39,40] These films were dried in a closed chamber.

The ultimate choice of the 95/5 wt % MCL-PHA/PCL blend composition was attributed to the outcomes elicited from empirical optimization tests (at different blend compositions and porosities) and was based on a previous study on PHA-based blends in relation to nerve tissue engineering using 95/5 and 75/25 wt % MCL-PHA/PCL blends.[41]

Mechanical Characterization

Mechanical properties of the porous and nonporous films were tested using an Instron tensile testing system (Instron, model 5942 Testing Systems, Buckinghamshire, UK). This analysis was carried out on solvent cast film strips of specified widths and lengths (n = 3, >23 mm in length and 5 mm in width). Tensile strength, elongation at break (%), and Young’s modulus values were determined from the stress–strain curves using the Instron’s analysis package (BlueHill 3) or via offline analysis.

Morphological Properties

Surface properties of the porous and the nonporous films were studied using a FEI XL30 FEG scanning electron microscope (FEI, Eindhoven, Netherlands). MCL-PHA film samples were mounted on conducting aluminum stubs and were coated with gold–platinum using an Polaron E5000 Sputter Coater (Quorum Technologies Ltd, Newhaven, East Sussex, UK) for approximately 2 min before they were imaged using the SEM. The images were acquired using an acceleration voltage of 10 kV at a 10 cm working distance at the Eastman Dental Institute, University College, London.

Hydrophobicity-Contact Angle Measurements

Contact angle (θ) measurements were performed using a KSV Cam 200 optical contact angle measurement system (KSV Instruments Ltd.) on both porous as well as nonporous MCL-PHA films to determine their wettability. Distilled water and cell media (200 μL) were placed on the surface of the film sample using a gas tight syringe. Ten images of the water/media droplets dispersing on the surface of the film sample were captured within a frame interval of 1 s. The analyses of the images were performed using the KSV Cam software. All work was completed at the Eastman Dental Institute, University College, London.

Cardiac Progenitor SCs

Isolation of CPCs, Labeling, and Scaffold Seeding

Cell Isolation

CPCs were isolated from adult, C57BL/6, mouse atria. Specifically, after hearts were excised, they were washed and digested with 0.05% trypsin–EDTA, and the tissue explants were plated on fibronectin-coated Petri dishes. They were expanded in culture as collagenase and trypsin digestion cells (CT) in accordance to standard methods described previously.[42]

Labeling

Cells were then plated in Iscove’s modified Dulbecco’s medium (IMDM, Thermo Fisher Scientific, UK) and incubated in culture with perfluoro-crown-ether (PFCE)-containing fluorescent nanoparticles (NPs) (containing Atto647) (10 mg/mL in 1 million cells)[43] and FuGENEHD (Promega, Madison, WI, USA) for approximately 24 h before trypsinization, isolation, and pelleting. Final cell pellet suspensions containing approximately 1 million cells each were maintained in cell media solutions and transferred to Eppendorf volumes containing 1.7 mL IMDM. Labeled cells were used to seed the scaffolds for SEM, MRI/MRS, and confocal microscopy experiments.

Cell Seeding

Cells (unlabeled or labeled) were seeded on scaffolds overnight after isolation and pelleting. The seeding density was 20k/scaffold for confocal-epifluorescence imaging and 300–500k/scaffold for in vitro MRI studies, with scaffolds cut at sizes spanning 2–8 × 2–8 mm2. Scaffolds were subsequently washed with PBS, and fixed cells were seeded [in 2% paraformaldehyde (PFA)/PBS solution (1:7 v/v)], while live cells were seeded in IMDM media. Scaffolds were subsequently prepared for high-content, confocal imaging, SEM or for MRS.

In vivo scaffolds were cut into a trapezoidal shape (the smaller side was implanted toward the apex) with a size of 2 × 5 mm2, with a height of 4 mm. The optimal seeding density was found to be 300–350k cells in 100 μm porous scaffolds (the ratio of the final number of cells to the number of the originally seeded cells was ∼0.6). This estimate was based on in vitro experiments where Trypan blue cell counts were conducted upon initial seeding/incubation and on corresponding counts of the freely floating cells in the IMDM media after the incubation period followed by the transfer of the scaffold in a new Eppendorf tube with fresh media. To maximize the cell density, a double-layered scaffold was implanted in a second mouse and studied in vivo, as reported below. The double-layered scaffold was composed of two single layer scaffolds that were glued at their four corners using surgical glue (Histoacryl, Braun Surgical S.A., Spain).

In Vitro Cell Adhesion and Proliferation Studies of Seeded Scaffolds

High-Content Microscopy-Epifluorescence Imaging

Epifluorescence Imaging

Live cells were stained with Calcein (CellTrace Calcein Red-Orange, ThermoFisher Scientific, UK) for high-content imaging and plated in 96-well plates. Cells (n = 3, ∼20–50k cells/well) were maintained in culture up to 7 days (D), and a time course study (D1–D7) of live cells was conducted to assess cell survival (calcein) using a high-content imaging system (Operetta, Perkin-Elmer, UK) (results not shown).

Label Detection—Confocal Microscopy

Fluorescent NPs were imaged [excitation wavelengths: λgreen = 488 nm, λred = 633 nm, emission ranges: 500–550 nm (green) and 650–700 nm (red)] using phase contrast and red/green excitations in control cell samples and in samples with and without FuGENE, using a Leica TCS SP8 confocal microscope (Leica-Microsystems, Mainhem, UK) with HyD detectors and an objective with numerical aperture = 1.4, 63×.

In Vitro, Postmortem, and in Vivo MRI/MRS

Animal Ethics

All experimental procedures involving animals were approved by the Home Office (UK) and were in accordance to the guidelines under The Animals (Scientific Procedures) Act, 1986, the European Animal Research Directive 2010/63/EU, and with local institutional guidelines.

Radiofrequency Coils

For MRI studies, a 4 × 4 cm2 single-turn, transmit/receive butterfly coil (implemented on a 28 mm diameter plastic former) [in vitro/postmortem/in vivo studies] and a 5 (diameter) × 8 (length) mm2 solenoid coil [in vitro studies] were fabricated in-house using flexible copper laminate sheaths, tuned, and matched to the 19F resonant frequency at 375.8 MHz. The broad frequency response of the coil allowed intermittent imaging on the 1H and 19F nuclei.

MRI/MRS of Nonporous and Porous Scaffolds

In Vitro Studies

Unseeded and seeded scaffolds were maintained in IMDM media and placed in 0.2–0.7 mL Eppendorf tubes. 1H and 19F MRI/MRS were then conducted. For postmortem studies, 1H/19F MRI measurements were performed on hearts with control (unseeded) and labeled (seeded) scaffolds positioned using fibrin glue (Baxter, UK) on the anterior epicardial surfaces.

In Vivo Studies

Healthy mice were anesthetized and maintained using 1.5% isoflurane. They were then intubated and underwent a lateral thoracotomy. Scaffolds were positioned on the anterior myocardium using histoacryl surgical glue (B. Braun Surgical S.A., Spain). Mice were recovered, monitored for adequate postsurgical recovery, and transferred to MRI for imaging. They were then re-anesthetized in accordance to standard imaging protocols. Imaging parameters for all MRI/MRS acquisitions are listed below.

(a) In vitro studies 1H MRI (unseeded/seeded scaffolds): 1H MRI was completed with two-dimensional (2D) segmented k-space, double-gated spoiled gradient echo (SPGR), and three-dimensional (3D) ungated sequences.

(b) 19F-MRI/MRS: Work was performed on a 9.4 T Varian scanner. 19F spectra were acquired using nonlocalized acquisitions (ungated and gated for in vivo scans) with the following parameters: repetition time (TR) = 800–1000 ms, number of excitations (NEX) = 64 or 256, 512 points, bandwidth (BW) = 20 kHz, and receiver gain (RG) = 30.

(c) In vitro studies 19F MRI (unseeded/seeded scaffolds): Correspondingly, the 19F MRI acquisitions [SPGR, steady state free precession (SSFP)] were TR = 8.3 ms, TE = 4.17 ms, flip angle = 50°, NEX = 1024, matrix = 32 × 32, ST = 10 and 40 mm, BW = 4 kHz, RG = 30, and total acquisition time = 4.3 min.

(d) Postmortem studies (unseeded/seeded and labeled scaffolds): The 2D 1H MRI acquisition parameters were TR = 2.73 or 3.13 ms, TE = 1.58 ms, flip angle = 50°, NEX = 32, matrix = 128 × 128, FOV = 40 × 40 mm2, ST = 1 mm, BW = 100 kHz, and pulse width (pw) = 1500 μs (total acquisition time = 12.8 s). The 3D 1H MRI acquisition parameters were TR = 2.63 ms or 2.73 ms, TE = 1.33 or 1.38 ms, flip angle = 20°, NEX = 4, matrix = 128 × 128 × 128, FOV = 40 × 40 × 40 mm3, BW = 100 kHz, and total acquisition time = 2.5 min.

The corresponding 19F acquisition parameters were TR = 8.31 ms, TE = 4.17 ms, flip angle = 50°, NEX = 796, matrix = 32 × 32, ST = 5 mm, BW = 4 kHz, and total acquisition time = 3.31 min. 19F MRS was acquired with nonselective excitation using TR = 800 ms, NEX = 256, 512 points, BW = 20 kHz, and RG = 30.

(e) In vivo murine studies: 3D ungated scans were acquired using the following imaging parameters: TR = 3 ms, TE = 1.68 ms, flip angle = 30°, NEX = 4, matrix = 128 × 128 × 128, FOV = 40 × 40 × 40 mm2, BW = 100 kHz, and total acquisition time = 5 min.

Histology

Cellular Retention

Postmortem histological evaluation was performed at D1 and D7 postscaffold implantation to assess CPC retention. Mice were euthanized by cervical dislocation under general anesthesia and the hearts excised. The hearts were then dehydrated and fixed (either in a 15% sucrose, 0.4% PFA solution, or in a 4% PFA solution) after which they were embedded in paraffin and stored (at −80 °C or room temperature). Serial transverse paraffin sections of 10–17 μm were cut, from base to apex for histological staining using a nuclear stain 4′,6-diamidino-2-phenylindole (DAPI). Imaging and analyses were performed on a bright-field optical and on a confocal microscope [nuclear (DAPI), label (Atto647)].

Image Processing

Image and Spectral Analyses

Low-resolution 19F MR images were imported and interpolated in ImageJ (NIH, Bethesda, USA) using bicubic splines to match the 1H matrix size. Thoracic muscle 1H and 19F MRI were overlaid in ImageJ (opacity = 40–70%). In vitro and in vivo spectra were read and processed in CSX (P. Barker-Kennedy Krieger Institute, Johns Hopkins USA) and using the interactive data language software (IDL, Harris Geospatial, USA). Signal and signal-to-noise (SNR) ratio values were estimated using standard methodologies.[44]

High-field polymer spectral processing was conducted using the Mnova software package (v12, Mestrelab Research, S.L., A Coruna, Spain). The chemical shifts were referenced against the residual solvent signals at 7.26 and 77.0 ppm for the 1H and 13C spectra, respectively.

Statistical Analyses

All results are reported as mean ± standard deviation (SD). Paired and unpaired, two-tailed Student’s t-tests, were also used (XLSTAT, Addinsoft, New York) for mean comparisons (α = 5%).

Results

Production of PHAs and Physical Characterization

The polymer was produced by the fermentation of P. mendocina CH50, purified, and structurally characterized, as previously described[36] (Figure ). The concentration of the obtained biomass at the end of fermentation was 1.5 g/L. The final PHA concentration was 0.52 g/L.

Figure 1

Synthetic and chemical characteristics of polymer: (A) general chemical structure of polyhydroxyalkanoates (x = 1, 2, 3; n = 100–30 000; R1, R2 = alkyl groups; C1–C13 units). (B–E) NMR spectra of the MCL-PHA and MCL-PHA/PCL blend. (F,G) Corresponding FTIR spectra for MCL-PHA and MCL-PHA/PCL blend depicting the ester carbonyl bond and C–O stretching peaks. Only the characteristic peaks for PHAs and PCL are annotated in FTIR spectra.[45,46]

FTIR and NMR

The polymer was identified to be MCL-PHA using FTIR. The two characteristic peaks of MCL-PHAs (1726.2 cm–1, indicative of the ester carbonyl bond, and 1160.0 cm–1, indicative of C–O stretching) were present in the elicited FTIR spectrum. Final confirmation of the polymeric structure was carried out using 13C and 1H NMR spectroscopy (Figure and Table ). The observed 1H NMR peak area ratios for the MCL-PHA were 2:1:2:8:3 (a/b/c/d, d*/e, e*), which exactly corresponded to the expected ratios from the structure, as shown in Figure B. The elicited ratio value of 8 obtained for the 1Hs (annotated as d, d*) is the average of 6 and 10, that is, the number of protons in each monomeric unit type. In the case of the MCL-PHA/PCL blend, the polymeric peak ratios matched those of the pure MCL-PHA polymer. Hence, the MCL-PHA related peak area ratios were 2:1:2:8:3 (a/b/c/d, d*/e, e*), while those for the PCL were 1:2:1:1 (^a/^b/^c/^d), as shown in Figure C.

Table 1

Chemical Shifts (δ) for (A) 1H NMR Peaks and Corresponding Chemical Shifts for (B) 13C Peaks for MCL-PHA and Caprolactone[37,47]

(A)
proton atoms	3-hydroxyoctanoate (δ, ppm)	caprolactone (δ, ppm)
CH	5.20 (b, multiplet)	4.1 (^a, triplet adjacent to carbonyl group)
CH2	2.50 (a, eightfold peak)	2.30 (^d, triplet)
CH2	1.58 (c, multiplet)	1.65 (^b, eightfold peak)
other CH2	1.25 (d, d*, multiplet)	1.40 (^c, multiplet)
CH3	0.88 (e, e*, triplet)

Thermal Properties

DSC was used to determine the thermal properties of the synthesized materials (Table ). All thermograms showed the presence of two or more peaks corresponding to the melting and glass transition temperatures. Tg values for the MCL-PHA scaffolds (nonporous [n = 4] and porous (75 [n = 1]/100 μm [n = 4]) at 100%) were −44.5/–44.0/–44.5 °C, while the corresponding values for the MCL-PHA/PCL blends (with PCL at 95/5%) were −48.2 (n = 4)/–44.2 (n = 1)/–46.5 °C (n = 3). The corresponding Tm values were 52.1 (n = 4)/53.2 (n = 1)/53.6 °C (n = 4) (100% composition) and 50.4 and 58.3/52.8 and 56.0/51.9 and 59.4 °C for the PCL blends. Two values are reported for the blends corresponding to the two distinct peaks elicited in the PCL blend thermograms.

Table 2

Summary of Thermal Characterization Results of Synthesized Scaffolds

sample	glass transition temperature (°C)	melting temperature (°C)
MCL-PHA (100%) nonporous [n = 4]	–44.5 ± 1.6	52.1 ± 0.4
MCL-PHA/PCL (95/5%) nonporous [n = 4]	–48.2 ± 0.5	50.4 ± 0.1/58.3 ± 0.1a
MCL-PHA (100%) porous (100 μm) [n = 4]	–44.5 ± 1.9	53.6 ± 0.4
MCL-PHA/PCL (95/5%) porous (100 μm) [n = 3]	–46.5 ± 1.1	51.9 ± 0.1/59.4 ± 0.3a

Reported values correspond to the separate peaks of the MCL-PHA/PCL thermograms.

Mechanical Properties

2D patches [nonporous and porous (75 and 100 μm porosity)] were fabricated using the solvent casting technique. Controlled porosity is a critical morphological characteristic for biocompatible scaffolds for cell adhesion and growth. Therefore, porous patches were fabricated. The porosity was optimized based on the mechanical properties of the synthesized scaffolds (75, 100 μm), using tensile testing (Figure , Table ).

Figure 2

Mechanical characteristics of synthesized scaffolds: histogram plots of (A) ultimate tensile strength (MPa), (B) elongation at break (%), and (C) Young’s modulus (MPa) for the nonporous (n = 4), porous (75 μm [n = 3], and 100 μm [n = 7]) scaffolds. Results represent mean ± SD values over 3–7 independent tests (Table ).

Table 3

Summary of Mechanical Properties of the Synthesized Scaffolds

sample	tensile strength (MPa)	elongation at break (%)	Young’s modulus (MPa)
MCL-PHA (100%) nonporous	7.83	507.25	4.63
MCL-PHA/PCL (95/5%) nonporous	7.40	526.50	4.65
MCL-PHA (100%) porous (75 μm)	2.38	528.00	0.68
MCL-PHA (100%) porous (100 μm)	1.31	212.50	1.92
MCL-PHA/PCL (95/5%) porous (75 μm)	2.34	604.33	0.72
MCL-PHA/PCL (95/5%) porous (100 μm)	0.91	206.00	1.46

Surface Characterization

The surface characteristics of the neat and porous scaffolds were quantified in terms of wettability and imaged using SEM. Surface morphology and microstructural features of the optimized porous scaffolds were visualized in cases of seeded scaffolds with CPCs using SEM, as shown in Figure .

Figure 3

Surface morphology of porous scaffolds: morphological characterization of the synthesized (A,D) porous (MCL-PHA, MCL-PHA/PCL) scaffolds [(A,C) nonseeded, and (B,D) seeded with cardiac progenitor CT green fluorescent protein negative cells] using SEM. Indicative results are shown from the 75 μm porous scaffolds.

Surface wettability was quantified only for the neat scaffolds using water and IMDM media with a drop shape analyzer. The mean water contact angles for the nonporous scaffolds (MCL-PHA vs MCL-PHA/PCL) were 90.0 ± 11.6°/80.7 ± 4.2° (water) and 88.8 ± 6.7°/72.0 ± 2.5° (IMDM), as shown in Figure . The contact angles were significantly higher in the H2O versus the IMDM cases (MCL-PHA: p < 0.013, PCL: p < 0.015). Significantly decreased mean contact angles were also observed in the MCL-PHA/PCL vs MCL-PHA samples in the case of IMDM tests (p < 0.004), indicative of increased hydrophilicity.

Figure 4

Contact angle measurements in the synthesized scaffolds: mean contact angles (mean ± SD) over 6–11 independent measurements for three different samples of each of the synthesized scaffolds (neat MCL-PHA and MCL-PHA/PCL blend at a composition of 95/5 wt %) using water and IMDM media.

Cardiomyocyte Cytocompatibility Studies

CPC Cell Density

The biocompatibility, cell surface adherence characteristics, and cell seeding density of the optimal porous scaffolds (with a porosity of 100 μm) were assessed using high-content epifluorescent imaging in vitro, using unlabeled and 19F FuGENE-labeled cells (50–70k). The first set of experimental tests assessed the differences of cell adherences on the two types of the porous scaffolds having an optimal porosity of 100 μm. Elicited results are summarized in Figure A,B. Figure clearly shows the increased cell adherence (and correspondingly increased viable cell density) in the MCL-PHA/PCL (Figure B) compared to the MCL-PHA neat scaffold (Figure B). Quantification of the total number of viable cells was challenging, given the 3D porous structure of the scaffold, and its optical scattering characteristics.

Figure 5

Comparison of in vitro CPC attachment on the two types of synthesized scaffolds: high-content (epifluorescent) imaging of porous MCL-PHA scaffolds (with a 100 μm porosity) seeded with (A) unlabeled (∼50–70k cells) and CT cells [cytoplasmic Calcein stain (orange), nuclear Hoechst stain (blue)]. Results indicate the lower affinity of porous MCL-PHA scaffolds for unlabeled cardiac progenitor SCs. (B) Increased cell density is observed for the porous MCL-PHA/PCL scaffold.

Noninvasive, Temporal Monitoring of Scaffolds Using 19F MRS/MRI

In Vitro, Postmortem, and in Vivo Scaffold Characterization

Of increased interest are the MRS/MRI results from in vitro tests of the scaffold with optimal porosity (100 μm), summarized in Figures and 7, over a temporal period of 9 days (D1–D9) following seeding with FUGENE-labeled cells. Constancy of cell retention is justified in Figure by the 25% integrated MRS area difference between D1 and D3. The cell density decreased to 44% at D6 and to 30% at D9. MRI also achieved visualization and clear delineation of the scaffolds using both 1H and 19F MRI, as indicated in Figure . Indicative is the decreased signal (and SNR) responses at D6 compared to earlier days (Figure G–I).

Figure 6

Temporal, in vitro MRS characterization of persistence of CPC attachment on the optimal type of porous scaffolds: magnetic resonance 19F spectra (MRS) of MCL-PHA/PCL (100 μm, 95/5%) scaffolds that were initially seeded with using 300k FuGENE-labeled CT cells each at (A–D) days 1 (D1), D3, D6, and D9.

Figure 7

In vitro MR visualization and temporal monitoring of CPC seeded porous scaffolds: (A–C) indicative axial (A) and coronal/sagittal (B,C) axis 1H MRI views of a 75 μm porous MCL-PHA/PCL 95/5 wt % blend scaffold (arrows) showing in vitro detectability (samples immersed in IMDM media in 0.7 mL Eppendorf tubes). (D–F) Corresponding axis and coronal/sagittal views of a 100 μm porous scaffold in vitro. (G,I) 19F MRI of seeded scaffold (MCL-PHA/PCL 95/5 wt % blend) with labeled cells. Axial views have a slice thickness of (G–I) 10 mm at D1, D3, and D6.

The seeded scaffolds were successfully tested in the postmortem mouse using both 19F MRS/MRI, as shown in Figure A–D. Results indicate responses at the first day following scaffold implantation. More interesting are the elicited results from the in vivo tests conducted in two C57BL/6 mice, indicating the ability of 19F MRS to detect and track the labeled, seeded cells, within a week following initial implantation (Figure E–H). The two peaks on the left-most part of the PFCE-NPs represent the accumulated isoflurane peaks, as reported earlier.[48] The double-layered scaffold was also clearly and distinctly identified in 1H MRI at D1 and D7 (the detected 19F area doubled compared to the single-layered scaffold), as shown in Figure G,H. The single-layered scaffold was not visible using 19F MRI as a result of the low seeding density (<300k at D1) [ultimately dependent on the scaffold’s size in the case of the mouse] and the heterogeneous distribution of seeded cells within the scaffold, an effect that worsens owing to the rhythmic and cyclical contraction–relaxation pattern of the mouse heart at rates exceeding 350 beats per minute.

Figure 8

In vivo MR visualization and monitoring of signal responses from CPC seeded scaffolds: in vivo 1H MRI of seeded scaffolds (with labeled CT cells) on the antero-lateral murine myocardium of normal C57BL/6 mice at day 1 (D1) from (A) single-layered and (B) double-layered porous MCL-PHA/PCL 95.5 wt % (100 μm) scaffolds seeded with FuGENE-labeled CT cells. (C) In vitro axial 19F MRI of the single-layered scaffold. (D) In vitro comparison of single- and double-layered scaffolds using 19F MRS, and (E–H) corresponding ungated 19F MRS from the in vivo mouse (E,F) (single-layered scaffold), and (G,H) double-layered scaffold at D1 and D7. (I,J) Two separate axial views of the fused 1H–19F MRI of the double-layered scaffold at D1 and D7.

Histological Characterization of Implanted Scaffolds

Figure shows a histological short-axis, mid-apical view of the postmortem heart indicating the sites of cellular migration and retention of CPCs within the myocardium and the scaffold. Also shown are the existing cellular entities at the interface of the scaffold and myocardium, as identified using bright field and confocal imaging (Atto647, DAPI).

Figure 9

Histological assessment of single- and double-layered porous scaffolds seeded with CPCs: (A) histological short-axis, mid-apical view of a fixed heart indicating the sites of scaffold (single layer) attachment, seeded scaffold cells, and cell retention of CPCs in the neighboring myocardium. (B–D) Arrows in the middle and right sides of the subfigures indicate the locations of the scaffold and the scaffold–myocardial interface. (E–G) Black arrows indicate the locations of detected cells (CPCs and red blood cells). (H–P) Fluorescent views [nuclear (DAPI), and label (Atto) presented in grayscale, and fused shown in color (Atto-red, DAPI-blue)], indicating CPC retention (H, K, N) within the scaffold and myocardium (white arrows). (Q–V) Corresponding histological and fluorescent views of the double-layered scaffold (shown in grayscale). Indicative is the increased number of cells within the myocardium compared to the single-layered case. Bright field images have been pseudocolored in different instances for better visualization.

Discussion

In this work, a novel, polymeric MCL-PHA/PCL blend material has been proposed for use as the substrate of a controlled delivery patch/scaffold for controlled SC release in cardiac tissue applications.[16,36] Both MCL-PHA and MCL-PHA/PCL polymer types have been studied previously and have shown exceptional attributes to adhesion and cell proliferation for SCs.[16,18,24] In particular, MCL-PHAs have exhibited increased elastomeric responses, low crystallinity, low tensile strength, low melting points, and high elongations at break and have been extensively studied in various applications to-date.[49−51] However, blend MCL-PHA/PCL compositions have not been explored previously for their suitability and responses in applications in vivo or for temporal imaging with CPCs. We show improved properties from synthesized scaffolds, capitalizing on the merits of each material class, as discussed below.

Chemical, Thermal, and Morphological Characteristics

The signaling interplay between the seeded CPCs on scaffolds and the local microenvironment has been shown to be deterministic for the cell fate, including migration, proliferation, differentiation, apoptosis, and/or homing/engraftment.[6] Correspondingly, the physical, morphological, and functional characteristics are deterministic for such responses.

Furthermore, the biocompatibility and degradability are essential to silence inflammatory/infectious responses, especially in cardiac diseases. The scaffold’s long-term stress/strain and thermoplastic responses are also critical in view of the cardiac cyclic activity.

DSC was used to assess the polymer’s thermal properties, yielding (as previously shown) two distinct peaks corresponding to the Tg and Tm values observed within the ranges of 44–47 and 51–56 °C, respectively. Low glass transition and melting temperatures have a direct elastomeric impact on the material’s response and are typical for MCL-PHA. Nevertheless, similar thermal properties were elicited for MCL-PHA and MCL-PHA/PCL blends that are most suitable for the proposed in vivo scaffold applications.

Porosity is also invariably linked with seeding density and the ability of cells to infiltrate/cross-link, migrate, and for cell ingrowth (through maximization of the surface area and active binding sites). Despite pore size disparity, all samples (neat and blends at 95/5% compositions) were similar in pore and cellular morphology, thus confirming batch reproducibility.

A pore size of 100 μm was chosen, which is larger than the sizes of cells contained in the heterogeneous mixture of CPCs [endothelial, fibroblasts, innate cardiac SCs, spanning a size range of approximately 10–40 μm (mean pore size (±SD) = 19.5 ± 8.8 μm)] in each spatial dimension. The porogen concentration used was optimal, in accordance to prior published work.[17] Smaller pore sizes are prohibitive for cellular seeding and permeation, whereas large pore sizes may prove detrimental to the scaffold’s cell retention capacity.

Mechanical Characteristics

The scaffold’s rigidity/flexibility is of upmost importance for eliciting proper functional responses (support for adhesion of seeded cells) and attaining a proper coupling constant[52] with the injured myocardium. Correspondingly, the ability to control, study, and quantify its mechanical characteristics is critical, and these attributes have been extensively investigated herein.

The constitutive law manifested in terms of the stress–strain response of the material is fully deterministic for its mechanical response under all operational conditions. To this effect, stress–strain was quantified for these membranes at room temperatures. No experimentation was conducted using membrane samples immersed in buffer solutions, although prior work has shown additional beneficial effects.[17] The decrease of the Young’s modulus values for the nonporous versus the porous counterparts was anticipated. Of importance is the noted decreasing trend for these values in the case of the blend scaffolds compared to the MCL-PHA counterparts as reported previously,[53] indicative of a decreased stiffness. The documented disparity in the trend of Young’s modulus values for scaffolds with targeted porosities of 75 and 100 μm is attributed to the actual pore size disparity (from targeted values) and increased pore size variability in these scaffolds. It is thus likely that the pore distributions in the two cases at the spatial scale of uniaxial tensile testing are similar.

The elongation at break and tensile strength were comparable for the neat and porous scaffolds, but significantly smaller in value compared to PCL constructs,[54] and significantly higher than the reported ranges for human myocardium (100–300% for the elongation at break, and 3–15 kPa for tensile strength).[55,56] The achieved Young’s modulus values of the optimized (unseeded) porous scaffolds are high compared to targeted values of murine and human myocardium in the range of 0.02–0.5 MPa.[17,52,54−56] The stiffness of seeded scaffolds is anticipated to be even higher, exacerbating the scaffold-myocardium stiffness difference. While such a property is fundamental in achieving an appropriate mechanical stiffness gradient between the patch and myocardium, the observed Young’s modulus mismatch is not envisaged to be deleterious for the proposed applications. We did not anticipate that an exact match is required for the scaffold to be effective. Instead, we consider the chemical gradient (owing to the concentration difference of free, mobile SCs in the scaffold and the endogenous pool of SCs in the innate myocardium) to be the most critical factor in facilitating an efficient diffusional-migratory SC process, especially in the case of the injured myocardium. Reinforcing these arguments is the fact that the stiffness of MI tissue (as previously characterized using atomic force microscopy[57]) is greater than that of healthy myocardium (i.e., the stiffness mismatch is smaller). Correspondingly, the chemical gradient (increased SC density in scaffold vs lack of cells, or minimal number of recruited SC cells from the endogenous niche pool in MI tissue) is thus expected to be the predominant driving force for cell migration.

An additional advantage of the scaffold is that it can act as a passive restraint explant to provide mechanical support or, equivalently, act as a “passive-assist construct”, to counteract the induced stress from the myocardial pressure-overload during remodeling in reperfused MI. The beneficial characteristics of the scaffold as a possible passive-assist explant are anticipated to be highly effective in the case of MI, whereby the load restraint imposed by the attached patch is expected to ameliorate and offset the dilatatory and hypertrophic response of the injured myocardium, elicited in the acute (dilation), and chronic MI phases (connective tissue deposition and scar formation) that ultimately lead to heart failure. The elongation at break and ultimate tensile strength characteristics were also within expected operational limits for both healthy and diseased states.[17,19] Of increased importance are perhaps the surface characteristics of the scaffold, as described next.

The material’s roughness and hydrophobicity are critical markers relevant to cellular adhesion and proliferation.[16] The surface roughness of this MCL-PHA was reported previously based on light interferometric techniques and measurements[17,19] with a noted increase in the case of porous versus neat patches.

SEM and surface contact analyses revealed that both the neat and nonporous blend films had smooth and regular surfaces and that the specific cardiac myocytes adhered well on the outer surfaces of the nonporous scaffolds.

Given that the measured contact angles were 90° or less, both material types are considered as hydrophilic. As expected, there was an increased surface hydrophilicity with the use of IMDM, an effect that is more prominent in the MCL-PHA/PCL blends compared to the MCL-PHA scaffolds. This is expected as a result of the content of such media in electrolytes and other proteins that facilitate surface binding and adhesion. Surface contact angle results (IMDM vs H2O) have also been justified by SEM analyses, indicating good outer surface adherence of CPCs in both cases (neat and porous). More importantly, increased hydrophilicity was documented in the case of neat MCL-PHA/PCL scaffolds compared to their MCL-PHA counterparts, confirmed by the excellent CPC attachment and increased retention, as shown by confocal microscopy and 19F MRS findings. Further hydrophilicity increases are expected from the functionalization of these scaffolds with VEGF and RGD/YIGSR peptides.[17]

Novel Functional Response in Cardiac Tissue Engineering

Cell Retention, 19F Temporal Persistence, and Noninvasive Monitoring

The material has been synthesized to attain optimal characteristics (including pore size, actual physical dimensions, and biochemical composition) to tailor its physical, chemical, and thermal characteristics and to optimize the seeding density of CPCs with fluorinated NPs for visualization and temporal tracking using MRS over periods spanning >7–9 days.

The pore size of the scaffold has been chosen to be large enough to accommodate all different types of cells contained in the heterogeneous population of the seeded CPCs (including primarily innate myocardial, endothelial cells, and fibroblasts), yet being able to accommodate other types of cardiac SCs (e.g., mesenchymal, embryonic, inducible-pluripotent, and others).

The scaffold’s seeding density (limited capacity determined by pore size, thickness, and size of patch ultimately limited by the murine myocardium) is approximately 350k cells/scaffold, which is at the limit of the capacity of 19F MRI to detect and monitor labeled cells using fast imaging with the described imaging methodologies at 9.4 T.

Overall, SC retention has been proven, yet the cellular attachment can be further improved using VEGF/RGD/YIGSR. Nevertheless, the ability to maintain cells viable over multiple weeks is/will continue to be a challenge, as will be the determination on whether the increased 19F signal observed at later time periods using the double-layered scaffold is due to released NPs from lysed cells within the pouch region between the two layers.

Temporal-dependent 19F MRI decreases from seeded scaffolds, likely attributed to cell death, decreased cell adherence, and increased leaching rates of nonviable and viable cells, need to be investigated further. Conclusive confirmation is envisaged to be provided by in vitro dead/live assays and postmortem immunohistochemical assays following in vivo imaging, as part of future planned work.

The engraftment also needs to be assessed once the optimal scaffold conditions are attained over periods of 1–8 weeks to assess N-cadherin, connexin,[43] and spatial gap junction formation and localization within the intercalated disc.[58] Although the exact mechanisms through which scaffolds exert beneficial effects are still not well understood, it is likely that they promote neogenesis and angiogenesis in conjunction with an associated increased SC viability and proliferation, amplifying the beneficial paracrine effects.

Overall, the elicited results from prior and ongoing clinical trials and preclinical studies of CPCs and SCs have been promising. Nevertheless, despite the existence of vast prior work in this field, important questions still remain, including the optimal choice of cells, growth factors, the scaffold’s mechanical coupling and directionality of the explants with respect to the injured myocardium,[58] and the exact timing of its administration after MI to elicit maximum benefits.[59]

Conclusions

We have presented applicability of MCL-PHA/PCL porous blends fabricated as thin films with an improved performance compared to MCL-PHAs, capitalizing on the excellent polymeric properties of natural MCL-PHA polymers, and the simplified synthetic process relevant to PCLs. We have also demonstrated superior MCL-PHA/PCL scaffold performance compared to MCL-PHA scaffolds through experimental comparisons of (a) morphological data using SEM and (b) contact angle measurements indicative of increased hydrophilic responses of the blend scaffolds attesting to improved CPC adhesion, (c) in vitro confocal microscopy showing increased SC proliferative capacity, (d) mechanical testing that elicited good overall responses, (e) improved in vitro NMR retention of seeded SCs, and (f) and in vivo applicability and MRI visualization of labeled SCs over periods spanning 8 days. The scaffold’s structural and morphological characteristics are tunable and could allow maximization of the seeding density for easier detection and temporal follow-up using direct 19F MRI/MRS in vivo, anticipated to be beneficial to larger animals/humans. The proposed scaffold can be potentially modified synthetically to address the induced CPC hypoxic state postscaffold implantation through the controlled delivery of exogenously administered oxygen via the scaffold, particularly following MI, as a result of the loss of perfusion pathways and myocardial vascular obstruction effects. Additionally, conjugation/functionalization of the scaffold with angiogenic, vascular growth factors, and peptides, is possible and is expected to lead to even further increases in cellular attachment and proliferation, as evidenced previously in MCL-PHA seeded with other SC types.[17]

On the basis of this work, and the elicited preclinical outcomes in normal mice in vivo, MCL-PHA/PCL blends are expected to have tremendous potential as future materials for cardiac patch development. A dual approach combining direct injections and controlled delivery of CPCs using these scaffolds is expected to maximize potential benefits following infarction. For translational purposes, refinement of the synthetic method of patch production is also envisaged, whereby the automation of the processing of polymeric synthesis and its direct, in situ injection can be implemented in accordance to recently published studies.[60,61] Studies on their suitability and effects on MI are ongoing.