Self‐Assembling Peptide Hydrogels Support Stromal Vascular Fraction Viability to Promote In Vivo Nerve Regeneration
Liam A. McMorrow, Steffan Llewellyn, Jared McSweeney, Alessandro Faroni, Aline F. Miller, Alberto Saiani, Adam J. Reid
TL;DR
Injectable peptide gels help deliver cells that improve nerve repair in rats, matching the effectiveness of traditional grafts.
Contribution
A self-assembling peptide hydrogel is shown to enhance stromal vascular fraction cell viability and promote nerve regeneration comparable to autografts.
Findings
SVF delivered in optimized SAPH improved motor and sensory recovery in rat nerve defects.
SAPH-supported SVF outperformed collagen controls and SAPH without SVF in nerve regeneration.
qPCR suggests SAPH-delivered SVF has increased longevity compared to collagen-delivered SVF.
Abstract
Autograft, the gold standard in nerve reconstruction, outperforms the alternatives of acellular allograft/nerve conduits likely due to the transplanted Schwann cell population within. The stromal vascular fraction (SVF) is a heterogenous cell isolate, that may improve nerve regeneration outcomes when transplanted at the site of a nerve defect. Self‐Assembling Peptide hydrogels (SAPH) are synthetic materials derived from short chains of biological amino acids. They are safe, injectable hydrogels whose charge and mechanical properties are easily tuned. Our hypothesis is that SVF, when transplanted within an SAPH, tailored toward nerve regeneration and SVF viability, will improve outcomes of nerve regeneration. In vitro modelling is used to select SAPH that support the viable 3D culture of SVF and outgrowth from neuronal explants. In vivo experimentation with a 10 mm rat sciatic nerve…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7| Group | Axon Density (n/mm, Mean ± SEM) | Nerve fiber Diameter (µm, Mean ± SEM) |
Axon Diameter (µm, Mean ± SEM) | Myelin Thickness(µm, Mean ± SEM) |
|---|---|---|---|---|
| Contralateral | 16680 ± 978 | 6.052 ± 0.191 | 3.694 ± 0.111 | 1.179 ± 0.052 |
| Autograft | 21885 ± 5076 | 3.281 ± 0.125 | 1.802 ± 0.087 | 0.738 ± 0.059 |
| Collagen | 17487 ± 1960 | 3.070 ± 0.120 | 1.615 ± 0.139 | 0.705 ± 0.028 |
| α2%–25% | 19828 ± 1983 | 3.173 ± 0.147 | 1.709 ± 0.097 | 0.732 ± 0.040 |
| α2%–25% + SVF | 18882 ± 2343 | 4.027 ± 0.327 | 2.143 ± 0.160 | 0.943 ± 0.092 |
| Pairwise Comparison | Axon Density | Nerve fiber Diameter | Axon Diameter | Myelin Thickness |
|---|---|---|---|---|
| Contralateral vs. Autograft | 0.6812 |
|
|
|
| Contralateral vs. Collagen | 0.9996 |
|
|
|
| Contralateral vs. α2%–25% | 0.9286 |
|
|
|
| Contralateral vs. α2%–25% + SVF | 0.9798 |
|
| 0.0634 |
| Autograft vs. Collagen | 0.8031 | 0.9494 | 0.8264 | 0.9959 |
| Autograft vs. α2%–25% | 0.9849 | 0.9969 | 0.9878 | >0.9999 |
| Autograft vs. α2%–25% with SVF | 0.9414 | 0.0909 | 0.3188 | 0.1262 |
| Collagen vs. α2%–25% | 0.9758 | 0.9952 | 0.9781 | 0.9978 |
| Collagen vs. α2%–25% with SVF | 0.9966 |
|
| 0.0594 |
| α2%–25% vs. α2%–25% with SVF | 0.9993 |
| 0.1353 | 0.1116 |
- —UK Regenerative Medicine Platform10.13039/501100019326
- —Medical Research Council10.13039/501100000265
- —British Society for Surgery of the Hand10.13039/501100023294
- —British Association of Plastic, Reconstructive and Aesthetic Surgeons10.13039/100014919
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNerve injury and regeneration · Supramolecular Self-Assembly in Materials · Silk-based biomaterials and applications
Introduction
1
Peripheral nerve injury (PNI) is typically caused by trauma or surgery and has a reported incidence of 1.2 per 10,000 annually [1]. Division of a peripheral nerve results in loss of its associated function, potentially substantial morbidity and loss of work [2]. Following PNI, surgical coaptation of the divided nerve allows for some recovery, but this is not possible when a gap in nerve continuity exists. In such cases, the defect must be bridged using a biomedical device, such as a conduit or decellularized allograft, or an autologous nerve graft (ANG). Reconstruction with ANG is the gold standard and, in the setting of a large defect, is the only treatment option. ANG however is limited by a finite supply and donor site morbidity. The superiority of ANG over current conduits and decellularized cadaveric graft lies in the presence of a transplanted Schwann Cell (SC) population within ANG. This inbuilt cellular therapy is essential for supporting and guiding regenerating axons. Without cellular therapy, it is unlikely that another treatment will offer outcomes comparable to autograft.
Although tempting to improve bridging devices by seeding them with autologous SC, the requirements for harvesting (nerve biopsy) and prolonged culture make them impractical. A wide range of other cell types, particularly autologous adipose‐derived mesenchymal stromal cells (AD‐MSC), have been successfully trialed in experimental nerve reconstruction models. However, translating these findings to clinical trials has been challenging as the need for an in vitro processing and culture step adds regulatory complexity and uncertainty. The stromal vascular fraction (SVF), obtainable directly from lipoaspirate, is rich in regenerative progenitors and includes endothelial cells, endothelial progenitors, pericytes, mesenchymal stem cells, smooth muscle cells, fibroblasts, lymphocytes, and monocytes/macrophages [3]. Many of these groups are crucial in orchestrating Wallerian degeneration and subsequent nerve regeneration. SVF can be harvested and delivered in one operation, making it more directly translatable than cell therapies requiring lab processing. While SVF holds great promise for nerve injury, a suitable scaffold to support these cells in a nerve injury environment does not exist in currently approved devices. Only bovine collagen has been used in FDA‐approved conduits for nerve injury, and while collagen delivery of SVF has shown promise in some animal models [4, 5], collagen gels also suffer from drawbacks, such as inconsistent mechanical properties and batch‐to‐batch variation [6, 7].
Self‐assembling peptide hydrogels (SAPHs) are fully defined synthetic, i.e.: non‐animal derived, soft materials that exploit the fibrillar self‐assembly of short peptides and their subsequence entanglement and association into 3D percolated networks. They offer a compromise between the biocompatibility of animal and human derived materials and the consistency and predictability of synthetic materials. The mechanical properties of SAPH can be varied through the choice of peptide sequence but also more simply through dilution. The ability to control both cellular behavior and the fate of regenerative cell progenitors through mechanotransduction is well documented, and toward the purpose of neuronal and glial compatibility, there is evidence to support the use of hydrogels with storage moduli ranging from 1 to 10 kPa [8, 9, 10, 11].
In the early stages of nerve regeneration, migration of SC, fibroblasts, endothelial cells, and inflammatory cells leads to the creation of an extracellular matrix (ECM) bridge connecting the two ends [12]. This bridge is then reorganized and matured to accept the regenerating front. The role of the SAPH is not to replace this matrix but to boost regeneration through the delivery of cell therapy and to support its initial viability and proliferation. In addition to delivering SVF, the SAPH should also allow migration of all of these key cell types into the scaffold, before degrading at the rate in which it is expected to be replaced by this regenerating bridge/nerve.
A widely investigated example of self‐assembling peptide (SAP) is RADA‐16 (R: arginine, A: alanine, D: aspartic acid), which has shown promise in supporting neuronal growth in 2D but is not ideal as a 3D cell scaffold as cells must be exposed to a peptide stock solution with a pH of 2 – 2.5 before gelation is triggered at pH 7 [13]. Over the last few years, Saiani et al. [14]. have developed a family of SAPHs based on phenylalanine (F), lysine (K), and glutamic acid (E) as constituent amino acids that are formulated at pH 5 to 7 eliminating the need to expose cells to low pH solutions upon encapsulation making them more suitable for 3D cell culture.
Experimentation with neuronal cell culture has demonstrated that positively charged surfaces and hydrogels improve neuronal and Schwann cell (SC) attachment [15, 16, 17, 18, 19, 20]. Faroni et al. [21]. demonstrated that this was also the case for SAPHs, showing that the injectable (Figure S1) positively charged (FEFK)_2_K SAPH (trade name: PeptiGel Alpha 2, Cell Guidance Systems, Cambridge) was particularly suitable for the 3D culture of neuronal and glial cells. In addition, Morris et al. also showed that this particular peptide and its corresponding SAPH are also non‐immunogenic and can be degraded, metabolized and eliminated by the body [22]. Full material characterization of Alpha 2 including rheology, TEM and circular dichroism has also been published in the literature [23, 24].
The overarching hypothesis of this paper is that an appropriately optimized combination of SVF and SAPH, when transplanted into a nerve defect using a nerve conduit will potentiate nerve regeneration and consequently, functional recovery. Working with the same positively charged SAPH Alpha 2 as Faroni et al. [21] we first explored the optimal mechanical properties for an SAPH/SVF combination that supports cellular delivery, viability, and proliferation, and neuronal cell migration into the defect.
We then examined the effectiveness of this combination in vivo using a rat sciatic nerve defect model. We repaired the defect using a microgrooved PCL/PLA conduit developed in our group that has already undergone first in man preliminary trials, Polynerve (available from Polynerve Ltd.) [25]. We injected the optimized SVF/SAPH combination within the conduit before it was sutured in place with microsurgical techniques. Using short‐term assessments of axon and myelin morphometrics, and long‐term functional measures (the primary outcome of maximum isometric tetanic force, then muscle weights, and fine touch and temperature sensory testing), we demonstrate that SAPH‐SVF composite hydrogels enable equivalent nerve regeneration to ANG and significantly superior regeneration to acellular (without SVF) collagen and SAPH controls. Furthermore, using Y chromosome PCR, we demonstrated incorporation of transplanted cells into the regenerated nerve when delivered in SAPH.
Results
2
Softer SAPH Support Dorsal Root Ganglia Outgrowth and SVF Viability and Proliferation in 3D Culture
2.1
Alpha 2 positively charged SAPH precursor solution can be obtained commercially as a standard stock solution. On gelation through addition of cell culture media it produces a SAPH with shear modulus, Gʹ, of ∼ 10 kPa. Dilution prior to gelation of the precursor solution with HPLC water allows us to modulate the shear storage modulus of the SAPH. Alongside Alpha2 (α2), two‐ (α2%–50%), and four‐ (α2%–25%) fold diluted precursor solutions were produced. In Figure 1A the Gʹ of the 3 SAPH measured after 24 h incubation in cell culture media are given (Mean ± SEM: α2 = 11.0 ± 0.7 kPa, α2%–50% = 3.5 ± 0.8 kPa, α2%–25% = 1.2 ± 0.4 kPa,). These results confirm that the produced SAPH Gʹ fall within the targeted range of 1–10 kPa.
DRG Explant with SVF response to FEK‐based SAP hydrogels. (A) Storage modulus of SAP conditioned with media for 24 h(B) Brightfield Images of DRG encapsulated in SAP at (from left to right) day 1, 7 and 14. Far right column is confocal microscopy at day 14, scale bar is 500 µm (C) Mean cellular outgrowth distance from encapsulated DRG at 14 days (D) Viability of SVF in SAP hydrogels from fluorometric quantification of calcein. (E) Proliferation of SVF in SAP hydrogels from fluorometric quantification of dsDNA.(F) Migration of SVF from SAP hydrogels. (G) SVF cell count from gels and supernatant over 21 days culture from fluorometric quantification of dsDNA. Data are mean ± SEM, n = 3 gels per condition. Statistical significance designated as * = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001.
To screen these gels for their capacity to support nerve regeneration, adult rat dorsal root ganglion explants (n = 3) were encapsulated in the SAPH for 14 days. For nerve regeneration to occur across a defect in vivo, a mixed cellular population, inclusive of SC, fibroblasts and axons must migrate into the nerve defect from the proximal point of injury. Quantification of cellular migration, in vitro, from encapsulated Dorsal Root Ganglia (DRG) therefore provides an assay of the capacity for SAP to support nerve regeneration. Unfortunately, due to the charged structure of the investigated gels, limited antibody penetration does not allow standard immunofluorescent labelling of the different cell types e.g. axons/SC however smaller molecules such a phalloidin and DAPI did allow observation of overall cellular outgrowth. Mean cellular outgrowth from DRG explants from weekly brightfield and endpoint confocal (phalloidin) micrographs are shown in Figure 1B with values ± SEM for α2%–25% = 428 ± 55 µm, α2%–50% = 236 ± 85 µm and α2 = 8 ± 3 µm over 14 days (Figure 1C). Explant outgrowth was significantly greater in α2%–25% than α2 (p = 0.003). Complete pairwise comparisons shown in Table S1.
We assessed for differences in the viability and proliferation of SVF, from rat adipose tissue, encapsulated prior to gelation, within the three SAPH. Viability of cells were quantified through a calcein assay (Figure 1D) with α2%–25% supporting a significant increase in the number of viable cells by day 14 increasing from 10300 ± 1300 to 28900 ± 4700 (n, mean ± SEM). A drop in the number of viable cells between week 2 and 3 in α2%–25% coincided with gel degradation and cell migration out of the gel which we quantified though dsDNA in the supernatant (Figure 1F). A viable cell population was present in α2%–50% and α2 over the 3‐week period but we did not see any significant proliferation (full data and p‐value pairwise comparisons in Tables S2 and S3 respectively). Total cell proliferation of encapsulated SVF over 3 weeks (Figure 1G) was quantified by measuring dsDNA both in the hydrogel (Figure 1E) and in the supernatant (Figure 1F). Quantification of dsDNA in the supernatant captures cells that exit the gel either due to migration or gel degradation. All experimentation was conducted with low‐adherence well plates to prevent 2D plastic adherent cell proliferation. During the 3‐week culture period, macroscopic degradation of both α2%–25% and to a lesser extent α2%–50% was observed. This coincides with large and significant increases in quantity of dsDNA in the supernatant (Figure 1F) in both α2%–25% and to a lesser extent α2%–50% at this timepoint. We could not observe any visible macroscopic degradation for α2 across the three weeks. There was proliferation of encapsulated SVF in α2%–25% over the 3 weeks (including cells migrating from the gel). Total mean ± SEM % increases in cell number as measured through dsDNA were 233 ± 26 %, 69 ± 18 %, and 25 ± 16 % for α2%–25%, α2%–50%, and α2 respectively. This data combines cell counts from within the gels (Figure 1E), and in the supernatant (Figure 1F). Full data and p‐values from pairwise comparisons are shown in Tables S4–S7. Consequently, we deliberated that α2%–25% was optimal to take forward to in vivo investigation.
Composite Polynerve + α2‐25% Fabrication
2.2
For composite conduit fabrication and seeding, SVF is resuspended in α2%–25% prior to gelation with a positive displacement pipette and injected into the conduit. Fluorescent imaging through the conduit wall for SVF labelled for DNA and actin demonstrated an even distribution of encapsulated SVF. It also showed adherence of cells onto the microgrooved conduit wall (Figure 2A).
Addition of α2%–25% to create a composite conduit (A) Representative fluorescent microscopy of composite conduit seeded with SVF. (B) Apparatus for 3‐point compression test. (C) Force measured to achieve maximum displacement was normalized to a Polynerve hollow conduit. Data is Mean ± SEM, n = 3, 7 cycles (D) Measured stiffness on 3‐point compression test. Data is Mean ± SEM, n = 3, 7 cycles. Statistical significance if present designated as * = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001.
Polynerve's material properties, developed by this research group, have already undergone extensive preclinical and clinical testing [24]. We assessed for differences in the mechanical properties of Polynerve and Polynerve with α2%–25% in a three‐point compression test (Figure 2B) to model the expected deforming forces in vivo (particularly if placed over a joint). The force measured to achieve maximum displacement was normalized to the hollow Polynerve conduit. No difference was observed (Figure 2C) between the hollow and the filled tubes (hollow conduit: 1.00 ± 0.23, filled conduit: 1.10 ± 0.54, p = ns). Similarly, no effect of filling the conduit with α2%–25% was observed with regards to the stiffness of the tube (Figure 2 D—hollow conduit: 4.59 ± 0.92 × 10^−4^ N m^−1^, filled conduit: 4.40 ± 1.04 × 10^−4^ N m^−1^, p = ns).
Nerve Regeneration In Vivo is Significantly Improved by SAPH Delivered SVF at 6 Weeks
2.3
In this experiment, we investigated whether the inclusion of SAPH ± SVF within a nerve conduit, during the repair of a 10 mm rat sciatic nerve defect, would improve nerve regeneration at 6 weeks. We assessed this by using TEM to count axons and measure morphometrics in the distal stump, 14 mm from the proximal injury (Figure 3). An early timepoint of 6 weeks is important because in this animal model, regeneration and recovery of function is good and the rate of recovery is important. Whilst 6 weeks is too early to assess functional outcomes, as target organs will not yet have reinnervated, early signs of myelination and axon counts provide essential data on the early stages of regeneration. Within this experimentation, we have used composite conduits containing bovine collagen as a control as all current FDA approved filled conduits utilize collagen.
TEM assessment of in vivo nerve regeneration at 6 weeks (A) Representative TEM micrographs from the distal stump of rat sciatic nerves reconstructed with experimental and control conduits at 6 weeks (n = 3, Scale bar 10 µm, yellow arrows; unmyelinated axon. blue arrows; myelinated axon). (B) Myelinated axon counts in the distal stump, 14 mm from proximal injury at 6 weeks (n = 3, Mean ± SEM) (C) Nerve fiber diameter distribution in the distal stump, 14 mm from proximal injury at 6 weeks (n = 3, annotated with median and interquartile range) (D) Myelin thickness distribution in the distal stump, 14 mm from proximal injury at 6 weeks (n = 3, annotated with median and interquartile range). Statistical significance designated as * = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001. Where no group comparison is demonstrated, p values for statistical significance = > 0.05.
Animals (Lewis rats, 180–200 g, n = 3) tolerated the procedure well with no surgical site complications, no autotomy or device extrusion. During the 6‐week experiment, animals gained weight as expected, given feeding ad libitum. Representative TEM micrographs (Figure 3A) show advanced nerve regeneration with multiple myelinated fibers in groups utilizing SVF (both collagen and α2%–25%) compared to filled conduits without SVF.
Mean myelinated axon counts (Figure 3B) (n/mm expressed as mean ± SEM) in the distal stump were 356 ± 356, 153 ± 153, 1551 ± 340, 10054 ± 1856 for groups; collagen, α2%–25%, collagen with SVF and α2%–25% with SVF, respectively. The α2%–25% with SVF group axon count was significantly greater than all other groups (p = <0.0001, 0.0003, <0.0001 when compared to collagen, collagen with SVF and α2%–25%, respectively). Axon morphometrics were not calculable for groups without SVF due to the sparsity of myelinated axons.
Mean nerve fiber diameter (Figure 3C) was 2.58 ± 0.1 and 2.43 ± 0.1 µm for collagen with SVF and α2%–25% with SVF (± SEM, p = ns) respectively and mean myelin thickness (Figure 3D) was 0.40 ± 0.03 µm and 0.42 ± 0.01 µm for collagen with SVF and α2%–25% with SVF respectively (± SEM, p = ns). This is evidence that SVF was required to regenerate axons into the distal stump at 6 weeks with collagen or α2%–25% hydrogels. The significantly increased number of myelinated axons at 6 weeks in rats reconstructed with α2%–25% with SVF is important because this demonstrated earlier advanced nerve regeneration compared to controls with or without SVF and we would expect that this will translate to earlier recovery of function.
Motor Recovery is Significantly Improved by SVF at 12 Weeks
2.4
In this experiment, we set out to investigate whether the inclusion of α2%–25% with and without SVF, within a nerve conduit (Polynerve) when used for the repair of a 10 mm rat sciatic nerve defect would improve motor recovery, when compared to the clinical gold standard autograft, and the closest industry equivalent device – a collagen‐filled conduit. Animals (Lewis rats, 180–200 g, n = 6) tolerated the procedure well with no surgical site complications, no autotomy or device extrusion. During the 6‐week experiment, animals gained weight as expected given feeding ad libitum. Maximum isometric tetanic force (MITF) expressed as a % of the contralateral limb (mean ± SEM, n = 6) values were recorded as 57 ± 11%, 39 ± 9%, 22 ± 5%, and 83 ± 10% for autograft, collagen, α2%–25% and α2%–25% with SVF groups respectively (Figure 4A). MITF was significantly greater in defects reconstructed with α2%–25% with SVF than α2%–25% without SVF (p = 0.006) and collagen (p = 0.014). These findings were corroborated by combined gastrocnemius and soleus wet muscle weights (WMW) at experimental endpoint (Figure 4B—D). WMW, expressed as a % of the contralateral limb (mean ± SEM, n = 6) were 48 ± 5%, 43 ± 5%, 27 ± 4%, 63 ± 7% for autograft, collagen, α2%–25% and α2%–25% with SVF groups respectively. Combined gastrocnemius and soleus WMW were significantly greater in defects reconstructed with α2%–25% with SVF than without SVF (0.013). All p‐value data for pairwise comparisons in both MITF and WMW are available in Table S8.
α2%–25% with SVF improves long‐term motor recovery (A) Composite gastrocnemius and soleus maximum isometric tetanic force (mean ± SEM, n = 6) as a percentage of the uninjured limb 12 weeks post sciatic nerve reconstruction (B) Wet muscle mass (gastrocnemius and soleus combined, (mean ± SEM, n = 6) as a percentage of the uninjured limb. (C) Representative example image of mass differences in comparison to uninjured control between α2%–25% with SVF and (D) α2%–25% groups. Statistical significance designated as * = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001s. Where no group comparison is demonstrated, p values for statistical significance = > 0.05.
Sensory Recovery Occurs Earlier with the SVF Group
2.5
In this experiment, we set out to investigate whether the inclusion of α2%–25%, with and without SVF, within a nerve conduit (Polynerve) when used for the repair of a 10 mm rat sciatic nerve defect would improve sensory recovery over 12 weeks, when compared autograft and collagen controls. Animals (Lewis rats, 180–200 g, n = 6) tolerated the procedure well with no surgical site complications, no autotomy or device extrusion.
Recovery of fine touch sensation (as measured with a Dynamic Plantar Aesthesiometer, Ugo Basile, Italy) is charted in Figure 5A and occurs earlier in rats reconstructed with either autograft or α2%–25% with SVF than collagen and α2%–25% groups. This difference in recovery is most notable at 10 weeks, where, as a % of the contralateral limb, mean ± SEM values were 77 ± 16 (autograft), 39 ± 13 (collagen), 19 ± 12 (α2%–25%), 67 ± 20 (α2%–25% with SVF). At this timepoint, sensory recovery in α2%–25% with SVF was significantly greater than without SVF (p = 0.0003). Recovery in rats reconstructed with autograft was significantly greater than those with collagen (p = 0.0101) or α2%–25% (p = 0.0001) but not α2%–25% with SVF. Full normalized data for fine touch recovery is shown in Table S9 with p‐values for multiple comparisons in Table S10.
SVF increases the rate of sensory recovery. (A) Fine touch recovery, n = 6, Left graph; a line graph demonstrating mean ± SEM as a percentage of baseline readings. Right graph; line graph demonstrating mean % of baseline additionally normalized to the mean of all groups which is a straight line at y = 0. This graph, shown without error bars, allows clearer visualization of the differences in the rate of sensory recovery between the two groups. (B) Demonstrates the same information as (A) but for pain/temperature recovery. The presence of statistically significant differences at each timepoint between each group is not presented above due to the volume of comparisons but is instead present in the results text and supplementary data.
Similarly, pain/temperature sensation (as measured with Hargreaves apparatus, Ugo Basile, Italy) also recovered earlier in autograft and α2%–25% with SVF groups than in collagen and α2%–25% groups. This is visualized in (Figure 5B). At week 9, as a % of baseline, mean ± SEM values were 68 ± 8 (autograft), 46 ± 13 (collagen), 45 ±12 (α2%–25%), 69 ± 14 (α2%–25% with SVF). At this timepoint, sensory recovery in α2%–25% with SVF was significantly greater than without SVF (p = 0.0251) and collagen (p = 0.0263) with recovery in rats reconstructed with autograft being significantly greater than those with collagen (p = 0.0474) or α2%–25% (p = 0.0456) but not α2%–25% with SVF. Full normalized data for fine touch recovery is shown in Table S11 with p‐values for multiple comparisons in Table S12.
SVF Significantly Increases Nerve Fiber Diameter at 12 weeks
2.6
Twelve weeks after reconstruction with autograft or conduits filled with either collagen or α2%–25% with and without SVF, myelinated axon morphometrics in the distal stump, 14 mm away from the proximal injury were calculated from TEM images for each group (n = 6) and are shown in Figure 6 with data described and p‐values from multiple comparisons in Tables 1 and 2 respectively.
Myelinated Axon Count and Axon/Myelin Morphometric data from 12‐week study. Violin plots (A–C) are marked with median and interquartile ranges (n = 6). (D) Column chart demonstrating mean ± SEM axon counts/mm (n = 6). Where shown = p ≤ 0.05; ** = p ≤ 0.01; *** = p ≤ 0.001). Where no group comparison bracket is demonstrated, p values for statistical significance = > 0.05.*
Mean nerve fiber diameter, myelin thickness and axon diameter were greater in the α2%–25% with SVF group than all other groups including autograft however this finding was only statistically significant in nerve fiber diameter versus collagen (p = 0.02) and α2%–25% without SVF (p = 0.04) and in axon diameter versus the collagen group (p = 0.04).
Delivery of SVF in SAP Extends Longevity of SVF in Nerve Conduits In Vivo
2.7
In this experimentation, the in vivo fate of male SVF, when transplanted within a Polynerve conduit, in either α2%–25% or collagen, in a female rat sciatic nerve defect (n = 3) was investigated with Y chromosome PCR. By week 3 all conduits contained a fully regenerated nerve, without any evidence of residual hydrogel. Primer specificity was confirmed by examining day 0 qPCR reaction products with agarose gel electrophoresis (Figure 7A). Male gDNA was detected within collagen filled conduits till week 1 only. Male gDNA was present within α2%–25% conduits for the full 6 weeks (Figure 7B). B2M is not a housekeeping gene but acts as a positive control at day 0 only, as it will be present within host cells infiltrating the conduit. There was an initial large increase in B2M at week one, likely representing the inflammatory stage of nerve regeneration with this falling by week 3, then increasing toward week 6 (Figure 7C). SRY and B2M gDNA concentration as a % of day 0 values are described in Tables S13 and S14, respectively.
Detection of Male SVF fate with qPCR for SRY study. (A) Representative agarose gel after electrophoresis of day 0 qPCR products. (B) Line graph of SRY gDNA concentration as a percentage of day 0 with individual biological replicates staggered and line representative of mean. (C) Line graph of B2M gDNA concentration as a % of Day 0 with individual biological replicates staggered and line representative of mean.
Discussion
3
We demonstrate that a positively charged self‐assembling peptide hydrogel (SAPH) to deliver stromal vascular fraction (SVF) in a composite hydrogel conduit promotes peripheral nerve regeneration in a rat in vivo nerve injury model.
Through a four‐fold dilution, the shear modulus of the positively charged (FEFK)_2_K SAPH (trade name: PeptiGel Alpha 2, Cell Guidance Systems, Cambridge) was reduced from 10 to 1 kPa. The stepwise decrease in stiffness, from α2 to α2–25 % significantly increased viability and proliferation of the encapsulated SVF, increased the rate of SAPH degradation and consequent SVF migration from the gel. The 3‐week period, before which, in vitro and in vivo, α2%–25% appeared to degrade, matched the 3‐week time frame by which a 10 mm rat sciatic nerve defect will normally regenerate its own primitive ECM bridge [26]. While previous work by Morris et al. [22], demonstrated in vivo the subcutaneous biodegradability of FEK‐based SAP hydrogels, the primary objective of the present study was to evaluate their efficacy in supporting nerve regeneration rather than to quantify degradation kinetics within the nerve environment. For larger nerve gaps the SAPH properties will likely have to be tuned further to ensure the hydrogel degradation matches the regenerative timeline. We will plan to next study these hydrogels in a large animal model to assess efficacy across larger defects and will investigate the in vivo degradation kinetics specific to the nerve defect at this timepoint. The ease and predictability with which SAPH properties can be fine‐tuned make them an ideal scaffold for cellular therapy, where a one‐size‐fits‐all approach is unlikely to be successful.
The α2 SAPH fiber network typically form pores in the 15–30 nm scale [27] considerably smaller than the cells themselves. Clearly the encapsulated cells can remodel and ‘digest’ this scaffold, as shown by Faroni et al. [21, 28], to allow migration. Decreasing the concentration of the peptide in the SAPH therefore, most likely increases the ease in which cells can migrate through the gel. Corroborating this, we observed that stepwise dilution of α2, increased migration of cells from encapsulated neuronal tissue explants into the SAPH.
Protocols to optimize antibody staining within the SAPH failed to improve penetration, likely related to the gel's charge; therefore, we were unable to differentiate between cell types and instead relied upon actin‐staining phalloidin to visualise cell migration. We proceeded to select α2%–25% for further in vivo study both due to superior SVF compatibility and our theory that the rate of hydrogel degradation and therefore the delivery of transplanted cells, should be commensurate with the time for the regenerating nerve to cross the defect.
In the in vivo study, at 12 weeks, the use of α2%–25% with SVF, within a nerve conduit, led to a significant increase in motor recovery when compared to collagen and SAPH without SVF, with mean MITF reaching 83 % of the uninjured contralateral limb in the α2%–25% with SVF group. This value was greater than the autograft group (mean MITF 57 %) although p = 0.20. These findings were corroborated with similar changes in wet muscle weights and nerve fiber diameters in α2%–25% with SVF leading to significant improvements in both. Sensory recovery, both in fine touch and pain/temperature occurred faster in autograft and α2%–25% with SVF than collagen and α2%–25% alone groups.
Our finding that SVF potentiates nerve regeneration and consequent functional recovery adds to the findings of Shimizu et al. [5] who demonstrated improved axon diameters and CMAPs when using SVF in a collagen gel to reconstruct a 7 mm rat facial nerve defect (critical nerve defect in a rat facial nerve is lower than the sciatic nerve). In our study, at the earlier timepoint of 6 weeks, although limited by a smaller sample size for TEM, myelinated axon counts in the distal stump were significantly greater following reconstruction with α2%–25% with SVF, than collagen with SVF (and collagen or α2%–25% without SVF) suggestive that the medium by which SVF was transplanted could have impact on its efficacy. Using male SVF in female rats, we demonstrated that SVF was detectable for at least 6 weeks when delivered in α2%–25% as opposed to just one week when delivered in collagen. Although simple detection of male gDNA does not allow us to comment on SVF viability, given that PCR extraction, at the 3‐ and 6‐week timepoints, involved tissue solely from regenerated nerve tissue, this would imply that SVF had become incorporated into the regenerated nerve. While we speculate that improved engraftment may be secondary to improved viability and proliferation in the SAPH of choice, we have not explored the mechanism behind this, which will be a focus for our further research.
Conclusion
4
In summary, the data presented in this study, and the primary outcome measure; motor recovery as measured through MITF, supports the hypothesis that SVF, when transplanted into a nerve defect, via the injectable, positively charged, Alpha 2 PeptiGel diluted SAPH, potentiates nerve regeneration through a conduit and this SVF is subsequently incorporated into the regenerated nerve and significantly improves functional recovery.
Numerous experimental studies have justified the need for cellular therapy in nerve regeneration, yet no human study has occurred. It is well established that SC and other cell types create, remodel, and control the ECM bridge in nerve regeneration, yet new devices still focus on replacing this bridge with an acellular scaffold. The superiority of ANG, over all other devices, is the transplanted cell population within. Consequently, to target outcomes similar to ANG, we must shift our focus toward devices that can also provide cell therapy. The ideal scaffold for cellular therapy should be biocompatible, biodegradable, non‐toxic, carry a 0 % risk of disease transmission, and demonstrate batch‐to‐batch consistency. SAPH are well suited for this role. Given the potential to harvest, and redeliver SVF in a single operative setting, the use of SVF, delivered in SAPH, is a realistic, achievable therapeutic goal, which we have shown can improve outcomes in nerve reconstruction. Further study towards human trials, initially in a large animal model, is certainly warranted.
Experimental Section
5
Self Assembling Peptide Hydrogel Preparation
5.1
PeptiGel Alpha 2 was obtained from Manchester BIOGEL and are now available from Cell Guidance Systems, Cambridge, UK). Two‐ and four‐ fold dilutions of the stock solution were performed using HPLC water to produce the SAPHs with lower shear moduli referred to as α2%–50%, α2%–25% respectively.
Bovine Collagen Hydrogel Preparation
5.2
Collagen I, Bovine (Gibco, ThermoFisher Scientific, MA, USA) gels at a concentration of 3 mg/mL at pH 7 were prepared according to manufacturer's instructions in a Class 2 biosafety cabinet then used immediately.
SAPH Rheology
5.3
Rheological tests were performed on a Discovery Hybrid 2 (DHR‐2) rheometer (TA Instruments, USA) using a 20 mm parallel plate geometry with a gap size of 500 µm. Samples were prepared by pipetting 200 µL of hydrogel into ThinCert well inserts (1 µm pore size Greiner Bio‐One Ltd, Gloucestershire, UK). The inserts were then placed into 24‐well plates and incubated at 37°C overnight in 1 mL cell culture media. Following media exposure, samples were removed from the inserts by peeling‐off the bottom membrane of the insert and transferred onto the rheometer plate. Rheometer head was then lowered until the desired gap size was reached and samples left to equilibrate for 3 min at 37°C before performing the measurements. The hydrogel shear storage moduli were measure at 1 Hz and 0.1% strain. For all experiments, a solvent trap was used to minimize sample evaporation. Measurements were repeated at least 3 times to ensure reproducibility.
Ethics and Animal Welfare
5.4
Adult Lewis rats, 180–200 g (ENVIGO) or Adult Sprague Dawley rats (200–250 g), dependent on experiment were maintained in accordance with the Animal (Scientific Procedures) Act (1986) in a licensed facility with 24/7 access to animal technicians and veterinary care. All experimentation was carried out in accordance with a Home Office approved project license (PP96445155) and by personal license holders.
Animals were housed in groups, in a calm environment with 12‐h light/dark cycles, provision of enrichment activities, toys, nesting, tubes in housing and additional twice weekly playpen time with further enrichment provision. All animals were inspected for signs of poor health daily and weighed weekly.
Dorsal Root Ganglia Harvest
5.5
Adult Sprague‐Dawley rats were terminated in concordance with schedule 1; appropriate methods of humane killing as defined in the Animal (Scientific Procedures) Act (1986) with concussion of the brain by striking the cranium followed by cervical dislocation. DRG were extracted as described in the protocol by de Luca et al. [29], The spine was extracted, and all spinal musculature dissected free. Dissection was conducted in a laminar flow cabinet with the aid of a low‐powered operating microscope (Leica, Germany). The spine was cut in half with the transection passing though the spinous processes dorsally and the vertical bodies ventrally. Spinal cord tissue was removed exposing the DRG. DRG were removed with microforceps and transferred to a 35 mm petri‐dish (Corning, NY, USA) containing warm Ham's F12 (ThermoFisher Scientific, MA, USA) media. DRG were separated from their roots and encapsulated in 70 µL SAPH in a 35 mm glass‐bottomed petri‐dish containing 3‐well silicon inserts (Culture‐Insert 3 Well in µ‐Dish 35 mm, Ibidi, Gräfelfing, Germany). Following encapsulation, gels were coated in Ham's F12 (ThermoFisher Scientific, MA, USA) media supplemented with 10% Foetal Bovine Serum (ThermoFisher Scientific, MA, USA), 0.01% 100× N2 supplement (Gibco, ThermoFisher Scientific, MA, USA) and NGF 50 ng/mL (ThermoFisher Scientific, MA, USA). Explants were kept in culture conditions (37°C 5% C0_2_) and media changed hourly three times then every 72 h till experimental endpoint.
Stromal Vascular Fraction Preparation
5.6
Adult Lewis Rats were sacrificed by stunning (cranial concussion) then cervical dislocation in accordance with Schedule 1 of the Animals (Scientific Procedures) act 1986. In a laminar flow cabinet, following sacrifice, the inguinal subcutaneous fat pad was dissected and harvested. In a class 2 biosafety cabinet, adipose tissue was mechanically fragmented using sterile scissors and then digested for 90 min at 37°C in HBSS with 2 mg/mL collagenase type‐I (ThermoFisher Scientific, MA, USA) and mechanical agitation. Following enzyme neutralization with serum (Foetal Bovine Serum, ThermoFisher Scientific, MA, USA), digested tissue is passed through a 100 µm filter (Steriflip, Merk Millipore/ and centrifuged at 400 g for 10 min. After aspiration of the supernatant, the cell pellet containing SVF was resuspended in α‐MEM to form 10% of the volume of gel that it was seeded to. Resuspended SVF is gently suspended in the respective SAP gel triturating with a positive displacement tip, to form a homogenous mixture. Seeded SAP gels were dispensed into the relevant tissue culture vessel (experiment dependent) with media changes hourly three times then every 72 h.
DRG Explant Outgrowth
5.7
DRG explants (n = 3) were cultured in α2, α2%–50% and α2%–25% for 14 days as described in Section 5.5. After 14 days, media was aspirated, and gels with explants were fixed in 4% paraformaldehyde for 90 min at room temperature with 3x PBS 1‐hour interval washes to remove PFA. Gels were permeabilized with 0.2% Triton X‐100 in PBS (Sigma‐Aldrich, Merck, Germany) for 30 min at 37°C. Gels were washed three times with PBS for 15 min before incubation with conjugated Alexa Fluor 488 Phalloidin (ThermoFisher Scientific, MA, USA) for 3 h in the dark at 37°C. After staining, samples were washed 3 times with PBS, with 1‐h incubation periods between then imaged with a Leica inverted SP8 confocal microscope (Danaher, Washington, D.C. USA). For data analysis, z stacked images were produced in ImageJ (NIH). Mean cellular outgrowth was calculated first by hand segmenting images into DRG and cellular outgrowth areas, then converting these (using the morphometric feature of AxonDeepSeg [30]) into equivalent circle diameters for both the DRG and the cellular outgrowth area. The mean cellular outgrowth distance was then defined as half the difference between equivalent circle diameters for the DRG, and DRG + cellular outgrowth area. Mean ± SEM data were analyzed in GraphPad Prism V10 with statistical testing performed with Multiplicity adjusted p values for Tukey's multiple comparisons tests following one‐way ANOVA generated.
Calcein Assay for SVF Viability in SAP Hydrogels
5.8
Cells (1 × 10^4^) (n = 3, biological and technical triplicates) from freshly harvested SVF was suspended in 3D in 50 uL of α2, α2%–50% and α2%–25% and transferred to sterile, black‐walled 96 well plates (ThermoFisher Scientific, MA, USA). For each SAP gel, a standard curve of cells in gel with 0, 5, 10, 20, 40 k cells were also established as above. Gels were covered with 100 µL αMEM (Gibco, ThermoFisher Scientific, MA, USA) which were changed hourly 3× then every 72 h. Plates were kept in culture conditions (37°C 5% C0_2_) until experimental endpoint. At days 1, 7, 14, and 21 media were aspirated and replaced with 100 µL 1x PBS containing 0.05 µL calein AM and incubated for 30 min in culture conditions (37°C 5% C0_2_). Calcein fluorescence were measured with a fluorescent plate reader with ex/em 485/530 (FLUOstar Omega, BMG LabTech, Aylesbury, UK). Linear extrapolation of standard curves was calculated for each gel utilizing GraphPad Prism (San Diego, CA, USA) to allow determination of cell numbers. Statistical presentation of Mean ± SEM and a 2‐way ANOVA with Dunnett's correction for multiple comparisons was also performed with GraphPad Prism (San Diego, CA, USA).
Quantification of dsDNA in/on SAP Hydrogels (Quantification of Cells in Gels and Having Migrated into Media)
5.9
This methodology, in addition to the PicoGreen (ThermoFisher Scientific, MA, USA) manufacturer's instructions, adapts a protocol developed by Burgess [31]. 5 × 10^4^ cells (n = 3, biological and technical triplicates) of freshly harvested SVF were suspended in 3D in 50 ul of α2, α2%–50%, and α2%–25% were transferred to sterile 96 well plates (ThermoFisher Scientific, MA, USA). For each SAP gel, a standard curve of cells in gel with 0, 12.5, 25, 50, 100, 150 k cells was also established as above. Gels were covered with 100 µL phenol‐free αMEM (Gibco, ThermoFisher Scientific, MA, USA) which was initially changed hourly 3 × and kept in culture conditions (37°C 5% C0_2_). At days 1, 7, 14, and 21, media were aspirated from gels and 100 uL of 10 mg/mL pronase (Merck, NJ, USA) in distilled water was added to the gels and triturated with a pipette till a homogenous mixture was formed. Homogenized gels were incubated in culture conditions (37°C 5% C0_2_) for 5 min. Hundred microliters 2× TE buffer were added to gels and 100 uL of this transferred to a black 96 well plate and 100 uL of 1 in 100 PicoGreen reagent in 1X TE buffer (Quant‐iT PicoGreen dsDNA Reagent, Thermo Fisher Scientific, MA, USA) was added. Media from each change (twice weekly and at end point) was mixed with 100 µL 2x TE buffer of which 100 µL was transferred to a separate black walled 96 well plate and 100 uL of 1 in 100 PicoGreen reagent in 1X TE buffer (Quant‐iT PicoGreen dsDNA Reagent, Thermo Fisher Scientific, MA, USA) was added
After 5 min of adding the PicoGreen reagent, measurements were taken with a fluorescent plate reader excitation/emission 480/520 nm. Linear extrapolation of standard curves were calculated for each gel utilizing GraphPad Prism (San Diego, CA, USA) to allow determination of cell numbers. Statistical presentation of Mean±SEM and a 2‐way ANOVA with Dunnett's correction for multiple comparisons was also performed with GraphPad Prism (San Diego, CA, USA).
Fabrication of Microgrooved PLA/PCL Conduits (Polynerve)
5.10
Production of the Polynerve conduit followed pre‐existing protocols as described for the production of “SL” grooved PCL/PLA conduits previously [26]. Films were rolled around a 1.5 mm steel rod and sealed also as described previously [26]. Conduits were either used as empty conduits or combined with SAPH as described below.
Addition of an Intraluminal Filler to (Polynerve)
5.11
All filled conduits were prepared 24 h before planned surgical use.
In a class‐2 biosafety cabinet (Envair, UK), A 33 mm petri‐dish was prepared with two‐sided adhesive tape placed (3M, MN, USA) on the tissue culture surface and exposed to UV sterilization from the culture hood for 30 min. Bovine‐collagen hydrogels were prepared immediately before use as described in Section 5.2. For an acellular, 1.5 mm diameter, composite conduit, 40 µl of α2%–25% or bovine collagen was transferred into the conduit with a positive displacement pipette. The filled conduit was transferred to the prepared 33 mm petri‐dish and secured to the two‐sided tape. Media (αMEM) was added to the dish to cover the conduit and the conduit was stored at (37°C, 5% C0_2_) till surgical implantation the next day. For conduits containing SVF, the gel of choice, either α2%–25% or bovine collagen (dependent on experiment) was allowed to equivalate to room temperature for 30 min. As per acellular conduits above, in a class‐2 biosafety cabinet, A 33 mm petri‐dish was prepared with two sided adhesive tape placed (3M, MN, USA) on the tissue culture surface and exposed to UV sterilization from the culture hood (Envair, UK) for 30 min. SVF is harvested, extracted and seeded to the gel of choice immediately prior to conduit preparation at a concentration of 1 × 10^4^ per µl as described in Section 5.6. The total volume of the SVF in media prior to seeding to the gel was 10% of the planned gel volume. For example, to produce 1 mL of gel with 1 × 10^4^ per µl SVF concentration, 1 × 10^7^ cells of SVF were suspended in 100 µl of αMEM and suspended in 900 µl gel of interest. Forty microliters of α2%–25% or bovine collagen, dependent on experiment, was then transferred into the conduit with a positive displacement pipette. The filled conduit was transferred to the prepared 33 mm petri‐dish and secured to the two‐sided tape. Media (αMEM) was added to the dish to cover the conduit and the conduit was stored at (37°C, 5% C0_2_) overnight till surgical implantation the next day.
Three‐Point Compression Test
5.12
Conduits, with/without α2%–25% were tested by an Inston 3344 to measure flexural rigidity and bending strength. Speed was set at 5 mm/min and displacement of 2 mm with stiffness and force required to obtain maximum displacement measured in triplicate and averaged over 7 cycles. Mean ± SEM data were analyzed in GraphPad Prism V10 with statistical testing performed with Mann Whitney test for significance.
Intra‐Conduit SVF Distribution Imaging
5.13
Conduits containing α2%–25% and SVF (composite conduits) (n = 3) were prepared as described in Section 5.10. At 7 days, media was aspirated, and composite conduits were fixed in 4% paraformaldehyde for 90 min at room temperature with 3x PBS 1‐h interval washes to remove PFA. Composite conduits were permeabilized with 0.2% Triton X‐100 in PBS (Sigma‐Aldrich, Merck, Germany) for 30 min at 37°C. Composite conduits were washed three times with PBS for 15 min before incubation with conjugated Alexa Fluor 488 Phalloidin (Thermo Fisher Scientific, MA, USA) for 3 h in the dark at 37°C. After staining, samples were washed 3 times with PBS, with 1‐h incubations periods between then imaged with an Olympus IX‐51 fluorescent microscope (Olympus, UK).
Surgical Model
5.14
Lewis rats (180–200 g) were anaesthetized by inhalation of isoflourane (Abbott Laboratories, IL, USA) and received buprenorphine 0.05 mg/kg subcutaneously. The lateral thigh was shaved, and, under aseptic technique, the left sciatic nerve was exposed and a 10 mm section removed. In autograft groups, this segment was reversed and, with the aid of an operative microscope (Carl Zeiss, Germany), secured with epineural 9.0 nylon (Ethilon, Johnson and Johnson, NJ, USA). In conduit groups, distal and proximal stumps were advanced 2 mm into the 14 mm conduit to maintain a 10 mm defect between the two within the conduit and secured with a single 9.0 nylon (Ethilon, Johnson and Johnson, NJ, USA) horizontal mattress suture at each end under an operative microscope (Carl Zeiss, Germany). A 4.0 polygalactin 910 (Vicryl Johnson and Johnson, NJ, USA) suture was used to close the muscle and skin in layers. Animals were housed post‐operatively as described in Section 5.4.
Transmission Electron Microscopy
5.15
Adult Lewis Rats, 6 weeks (mid‐term experiment, n = 3, groups; Collagen, α2%–25%, Collagen & SVF, α2%–25% & SVF) or 12 weeks (long term experiment, n = 6, groups; Collagen, α2%–25%, α2%–25% & SVF, Autograft) after conduit or autograft repair of a 10 mm sciatic nerve defect, were sacrificed by stunning (cranial concussion) then cervical dislocation in accordance with Schedule 1 of the Animals (Scientific Procedures) act 1986. In a laminar flow cabinet, following sacrifice, the reconstructed left sciatic nerve was exposed, explanted, and a 1 mm portion of nerve, 4 mm into the distal nerve stump, fixed in 2.5% glutaraldehyde and 4% formaldehyde in 0.1 m HEPES buffer for 2 h. Samples are washed in dH_2_0 3 × 5 min then post‐fixed in 1% osmium tetroxide and 1.5% potassium ferrocyanide in 0.1 m cacodylate buffer for 2 h. Samples are washed in dH_2_0 3 × 5 min then incubated in 1% uranyl acetate overnight at 4°C. Samples are washed in dH_2_0 3 × 5 min then undergo dehydration in a graded series of 50%, 70%, 90%, 100%, 100% ethanol each for 20 min at room temperature. Samples were then washed in acetone twice for 20 min. Samples then underwent successive infiltration at room temperature with TAAB LV 25% in acetone for 2 h, TAAB LV 75% in acetone overnight, TAAB LV 100% for 6 h. Samples were then embedded in 100% TAAB LV labelled molds and cured at 60°C for 24 h. Ultrathin sectioning was performed and samples were imaged with TEM (Talos L120C). For determination of myelinated axon counts and measurements of axon morphology, five random images at 1250X magnification were used per sample (for 6‐week study, n = 3; for 12 week study n = 6). Higher magnifications were captured and used to confirm the identity of ultrastructural components where appropriate. All images were segmented by hand with Adobe Photoshop (Adobe, CA, USA) into axon and myelin components and then morphometrics calculated using AxonDeepSeg [30]. Although Axondeepseg can perform segmentation through use of an AI model, the software was only used for morphometric calculation to ensure accuracy.
Measurement of Maximum Tetanic Force
5.16
Adult Lewis Rats, 12 weeks after conduit or autograft repair of a 10 mm sciatic nerve defect (long term experiment, n = 6, groups; Collagen, α2%–25%, α2%–25% & SVF, Autograft), underwent terminal anesthesia and measurement of composite gastrocnemius, soleus, plantaris tetanic force as delivered through the Achilles tendon. An incision from hip to ankle was used to expose the reconstructed sciatic nerve and Achilles tendon. The Achilles tendon was transected at its distal insertion. A bipolar nerve stimulator was used to apply the stimulus just proximal to the reconstruction, The knee joint was immobilized to the platform with a pin across the joint. The platform was kept warm at 37°C and a 20 mL syringe filled with warm aMEM was loaded in a syringe driver to provide a slow drip over the exposed tissues to prevent desiccation. The distal Achilles tendon was sutured to the force transducer arm and the ideal resting length and voltage determined and set. The resting length was defined as the length at which maximum contractile force was obtained, and this was readjusted per technical reading group to produce the maximal twitch force. Prior to formal measurement, force frequency relationship was assessed across a range of frequencies 80 – 150 hz to obtain the maximum tetanic force. For each technical reading group, the measurement was then repeated 3 times with 10‐min rest intervals between measurements. Measurement of maximum tetanic force as above was repeated on the contralateral limb as a control. Maximum tetanic force was then assessed as a percentage of the contralateral control Maximum tetanic force was validated for use in the rat lower limb without evidence for interference from right/left dominance [32, 33]. Results were analyzed with GraphPad Prism V10 (GraphPad Software, USA) with data expressed as a % of the contralateral limb. A one‐way ANOVA followed by Tukey's post hoc analysis provided p‐values of the respective pairwise group comparisons.
Measurement of Muscle Weights
5.17
After measurement of maximum tetanic force (long term experiment, n = 6, groups; collagen, α2%–25%, α2%–25% & SVF, autograft) from soleus, gastrocnemius and plantaris muscles, these muscles, conjoined by their common insertion into the Achilles tendon, were detached at their origin and wet muscle mass readings (in grams) obtained with a digital weighing scale (OHAUS, China). Results were analyzed with GraphPad Prism V10 (GraphPad Software, USA) with data expressed as a % of the contralateral limb. A one‐way ANOVA followed by Tukey's post hoc analysis provided p‐values of the respective pairwise group comparisons.
Fine Touch Sensory Recovery
5.18
To quantify the recovery of fine touch sensation post sciatic nerve reconstruction, measurement of plantar response to an escalating force equivalent to 0 – 100 g through a fine filament and over 0 – 50 s with detection of paw withdrawal was performed preoperatively and weekly for 12 weeks post 10 mm sciatic nerve defect conduit or autograft repair (long term experiment, n = 6, groups; collagen, α2%–25%, α2%–25% & SVF, autograft) utilizing a dynamic plantar aesthesiometer (Ugo Basille, Italy). Lewis rats were placed in glass cages with a mesh floor for the filament to test through. All rats were placed in the holding cages 15 min before experimentation and then measurements were taken in technical triplicate of both left and right hind paws with at least 30 min between each technical repeat. Results were expressed in seconds and normalized as a percentage of the baseline mean for each rat such that 50 s is given a value of 0% and the preoperative mean 100%. A two‐way ANOVA was performed with multiplicity‐adjusted p values calculated following Tukey's multiple comparison test. All data analysis and statistical testing were performed in Prism v10 (GraphPad, MA, USA).
Temperature/Pain Sensory Recovery
5.19
To quantify the recovery of pain/temperature sensation post sciatic nerve reconstruction, measurement of plantar responses to an escalating thermal stimulus, generated by a focal light source applied through a glass pane to a maximum of 25 s were performed preoperatively and weekly for 12 weeks post 10 mm sciatic nerve defect repair with either nerve conduit or autograft (long term experiment, n = 6, groups; collagen, α2%–25%, α2%–25% & SVF, autograft) utilizing a Hargreaves apparatus (Ugo Basille, Italy) [34]. Lewis rats were placed in glass cages with a mesh floor for the filament to test through. All rats were placed in the holding cages 15 min before experimentation and then measurements were taken in technical triplicate of both left and right hind paws with at least 30 min between each technical repeat. All rats were given a break to return to cages for food/water between the second and third measurements. Results were expressed in seconds and normalized as a percentage of the baseline mean for each rat such that 25 s is given a value of 0% and the preoperative mean 100%. A two‐way ANOVA was performed with multiplicity‐adjusted p values calculated following Tukey's multiple comparison test. All data analysis and statistical testing were performed in Prism v10 (GraphPad, MA, USA).
SVF Tracking
5.20
Surgical Model
5.20.1
Female and Male SVF was harvested as described in Section 5.6. Polynerve conduits containing Male or Female SVF in α2%–25% were produced as described in Sections 5.10 and 5.11.
The animal model described in Section 5.14 (with animal welfare information in Section 5.4) was established with n = 3, Lewis rats, with two groups, reconstruction with Polynerve & α2%–25% & Male SVF or reconstruction with Polynerve & collagen & Male SVF and three time points 7, 21, and 42 days.
A day 0 time point was also established whereby the conduit groups (Polynerve & α2%–25% & Male SVF, Polynerve & collagen & Male SVF, Polynerve & α2%–25% & Female SVF, Polynerve & collagen & Female SVF,) were produced as per above but not implanted into the rat. Subsequent end‐point processing is described in Section 5.20.2.
DNA Extraction
5.20.2
At 7, 21, and 42 days after establishment of the surgical model described in Section 5.20.1 adult Lewis Rats (n = 3, groups), were sacrificed by stunning (cranial concussion) then cervical dislocation in accordance with Schedule 1 of the Animals (Scientific Procedures) act 1986. In a laminar flow cabinet, following sacrifice, the reconstructed left sciatic nerve was exposed and explanted. Alongside Day 0 controls, conduit contents were emptied into a 1.5 mL microcentrifuge tube containing 180 µL buffer ATL and 20 µL Proteinase K (both components of DNeasy Blood & Tissue kit, QIAGEN) and stored overnight at 56°C. DNA extraction steps continued as per manufacturers provided protocol [35]. DNA concentration and purity (through 260 nm/280 nm ratio; all > 1.9) was assessed using the NanoDrop ND‐100 (Fisher Scientific, NH, USA).
Real Time Quantitative PCR (qPCR)
5.20.3
qPCR of B2M and SRY gDNA extracted as described in 5.20.2 was performed as per manufacturer's instructions within the Quantinova SYBR PCR kit [36]. Cycling parameters were as per manufacturers instruction. Primers were sourced from Life technology, SRY; Fwd 5’‐AAGCGCCCCATGAATGCAT‐3’, Rev 5’‐CGATGAGGCTGATATTTATA‐3’, B2M; Fwd 5’‐CCCACCCTCATGGCTACTTC‐3’ [37]. A standard curve for both genes was established for SRY and B2M from known DNA concentrations on each plate. Data, including melt analysis was collected as per MIQE guidelines [38]. PCR Cq values were interpolated to concentration values (ng/uL) from standard curves and normalized to Day measurements for each rat with 100% representing measured gDNA concentration at day 0. All data analysis was performed with Prism 10 (GraphPad Software, USA).
Electrophoresis
5.20.4
Two percent Agarose (Thermo Fisher Scientific, MA, USA) in 0.5X TBE SYBR Safe DNA gel (Thermo Fisher Scientific, MA, USA) gel was cast. DNA samples, including loading buffer (Bioline) were added to wells alongside a 25–500 bp ladder (HyperLadder 25 bp, Merdian Bioscience, OH, USA). Gels were covered with TBE running buffer (VWR, UK). Samples were run at 100 V till loading buffer had visibly progressed across 80% of the gel.
Statistical Analysis
5.20.5
All quantitative results were analyzed with GraphPad Prism V10 (GraphPad Software, USA). Details of analytical techniques and statistical testing methods were provided within each respective technique.
Conflicts of Interest
A.J.R. is a founding shareholder of polynerve ltd.
Supporting information
Supporting file: adhm70599‐sup‐0001‐SuppMat.docx
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1R. N. A. Murphy , C. de Schoulepnikoff , J. H. C. Chen , et al., “The Incidence and Management of Peripheral Nerve Injury in England (2005–2020),” Journal of Plastic, Reconstructive & Aesthetic Surgery 80 (2023): 75–85.10.1016/j.bjps.2023.02.01736996504 · doi ↗ · pubmed ↗
- 2H. E. Rosberg , K. S. Carlsson , R. I. Cederlund , E. Ramel , and L. B. Dahlin , “Costs and Outcome for Serious Hand and Arm Injuries during the First Year after Trauma—A Prospective Study,” BMC Public Health [Electronic Resource] 13 (2013): 501.23706070 10.1186/1471-2458-13-501PMC 3681586 · doi ↗ · pubmed ↗
- 3V. M. Ramakrishnan and N. L. Boyd , “The Adipose Stromal Vascular Fraction as a Complex Cellular Source for Tissue Engineering Applications,” Tissue Engineering Part B: Reviews 24, no. 4 (2018): 289–299.28316259 10.1089/ten.teb.2017.0061 PMC 6080106 · doi ↗ · pubmed ↗
- 4M. Shimizu and M. H. Abstract , “Abstract: Facial Nerve Regeneration with PGA‐Collagen Nerve Conduit in a Rat Model,” Plastic and Reconstructive Surgery—Global Open 5, (2017): 17–18.
- 5M. Shimizu , H. Matsumine , H. Osaki , et al., “Adipose‐Derived Stem Cells and the Stromal Vascular Fraction in Polyglycolic Acid‐Collagen Nerve Conduits Promote Rat Facial Nerve Regeneration,” Wound Repair and Regeneration 26, no. 6 (2018): 446–455.30118577 10.1111/wrr.12665 · doi ↗ · pubmed ↗
- 6Š. Rýglová , M. Braun , T. Suchý , M. Hříbal , M. Žaloudková , and L. Vištějnová , “The Investigation of Batch‐to‐Batch Variabilities in the Composition of Isolates from Fish and Mammalian Species Using Different Protocols,” Food Research International 169 (2023): 112798.37254382 10.1016/j.foodres.2023.112798 · doi ↗ · pubmed ↗
- 7S. T. Kreger , B. J. Bell , J. Bailey , et al., “Polymerization and Matrix Physical Properties as Important Design Considerations for Soluble Collagen Formulations,” Biopolymers 93, no. 8 (2010): 690–707.20235198 10.1002/bip.21431 PMC 2880202 · doi ↗ · pubmed ↗
- 8A. J. Engler , S. Sen , and H. L. Sweeney , “Discher DE. Matrix Elasticity Directs Stem Cell Lineage Specification,” Cell 126, no. 4 (2006): 677–689.16923388 10.1016/j.cell.2006.06.044 · doi ↗ · pubmed ↗