The intermittent fasting-dependent gut microbial metabolite indole-3 propionate promotes nerve regeneration and recovery after injury


 The regenerative potential of mammalian peripheral nervous system (PNS) neurons after injury is critically limited by their slow axonal regenerative rate1. Since a delayed target re-innervation leads to irreversible loss of function of target organs2, accelerated axonal regeneration is required to enhance functional outcomes following injury. Regenerative ability is influenced by both injury-dependent and injury-independent mechanisms3. Among the latter, environmental factors such as exercise and environmental enrichment have been shown to affect signalling pathways that promote axonal regeneration4. Several of these pathways, including modifications in gene transcription and protein synthesis, mitochondrial metabolism and release of neurotrophins, can be activated by intermittent fasting (IF)5,6. IF has in turn been shown to increase synaptic plasticity7,8 and neurogenesis9, partially sharing molecular mechanisms with axonal regeneration. However, whether IF influences the axonal regenerative ability remains to be investigated. Here we show that IF promotes axonal regeneration after sciatic nerve crush in the mouse via an unexpected mechanism that relies upon the gram + gut microbiome and an increase of the gut bacteria-derived metabolite indole-3-propionic acid (IPA) in the serum. IPA production by Clostridium sporogenes is required for efficient axonal regeneration, and delivery of IPA after sciatic injury significantly enhances axonal regeneration, accelerating recovery of sensory function. Mechanistically, RNA sequencing analysis from sciatic dorsal root ganglia suggested a role for neutrophil chemotaxis in the IPA-dependent regenerative phenotype that was confirmed by the inhibition of neutrophil chemotaxis. Our results demonstrate for the first time the ability of a microbiome derived metabolite, such as IPA, in facilitating regeneration and functional recovery of sensory axons via an immune-mediated mechanism. Since IPA is a naturally occurring metabolite with a very favourable toxicity profile, it represents a realistic translational possibility for human axonal injuries.


Abstract
The regenerative potential of mammalian peripheral nervous system (PNS) neurons after injury is critically limited by their slow axonal regenerative rate 1 . Since a delayed target re-innervation leads to irreversible loss of function of target organs 2 , accelerated axonal regeneration is required to enhance functional outcomes following injury. Regenerative ability is in uenced by both injury-dependent and injury-independent mechanisms 3 . Among the latter, environmental factors such as exercise and environmental enrichment have been shown to affect signalling pathways that promote axonal regeneration 4 . Several of these pathways, including modi cations in gene transcription and protein synthesis, mitochondrial metabolism and release of neurotrophins, can be activated by intermittent fasting (IF) 5,6 . IF has in turn been shown to increase synaptic plasticity 7,8 and neurogenesis 9 , partially sharing molecular mechanisms with axonal regeneration. However, whether IF in uences the axonal regenerative ability remains to be investigated. Here we show that IF promotes axonal regeneration after sciatic nerve crush in the mouse via an unexpected mechanism that relies upon the gram + gut microbiome and an increase of the gut bacteria-derived metabolite indole-3-propionic acid (IPA) in the serum. IPA production by Clostridium sporogenes is required for e cient axonal regeneration, and delivery of IPA after sciatic injury signi cantly enhances axonal regeneration, accelerating recovery of sensory function. Mechanistically, RNA sequencing analysis from sciatic dorsal root ganglia suggested a role for neutrophil chemotaxis in the IPA-dependent regenerative phenotype that was con rmed by the inhibition of neutrophil chemotaxis. Our results demonstrate for the rst time the ability of a microbiome derived metabolite, such as IPA, in facilitating regeneration and functional recovery of sensory axons via an immune-mediated mechanism. Since IPA is a naturally occurring metabolite with a very favourable toxicity pro le, it represents a realistic translational possibility for human axonal injuries.

Main
Injuries to the peripheral nervous system (PNS) have a high prevalence with millions of affected individuals worldwide and are typically followed by long-lasting neurological disability without effective treatments beyond surgical reconstruction, which is only effective in a small percentage of cases [10][11][12] .
These injuries frequently result in partial or total loss of sensory, motor and autonomic function due to ine cient and slow axonal regeneration 13,14 . The study of cellular and molecular responses to injury 15,16 , of neurodevelopmental pathways 17,18 , of regenerative organisms 18,19 , as well as molecular screening approaches 20 have contributed to the identi cation of a number of mechanisms that in uence regenerative ability. However, our knowledge of the cellular and molecular mechanisms underpinning axonal regeneration and functional recovery remains incomplete, undermining the development of effective treatments for clinical injuries.
Recently, accumulating evidence implicates healthy lifestyle choices such as exercise and environmental enrichment in priming and enhancing the regenerative potential of sensory neurons 4 . Similarly, dietary regimens including intermittent fasting (IF) have been shown to promote metabolic and signalling pathways with the potential to favour wound repair and recovery in several disease states, including following nervous system injury [21][22][23][24] . Therefore, we investigated whether IF would prime sensory neurons for enhanced axonal regeneration and recovery in a model of sciatic nerve injury and whether it would enable the identi cation of speci c molecular mechanisms with translational potential for nerve repair.
Intermittent fasting promotes axonal regeneration following sciatic nerve crush via the gram + gut microbiome Axonal regeneration was initially assessed 24 hours after a sciatic nerve crush (SNC) following 10 or 30 days of IF versus an ad libitum (AL) diet. Both 10 and 30 days IF regimens signi cantly enhanced regeneration past the crush site to a similar extent (Extended data Fig. 1a-c), therefore all subsequent experiments were performed following 10 days of IF. Importantly, IF also promoted axonal regeneration when the recovery time was extended to 72 hours after SNC (Fig. 1a-c). In con rmation of these ndings, IF was followed by an increase in neurite outgrowth in cultured dorsal root ganglia (DRG) neurons (Extended data Fig. 1d-f). Importantly, IF did not alter critical modi ers of axonal regeneration such as Schwann cell and macrophage recruitment 25,26 to the nerve crush site (Extended data Fig. 2a- Since IF modi es cell metabolism 32,33 it was hypothesised that changes in metabolism may underpin the IF-dependent regenerative ability. To this end, semi-targeted gas chromatography coupled to mass spectrometry (GC-MS) of extracted serum following IF vs AL was performed, enabling identi cation of 79 metabolites, 14 of which were differentially enriched upon IF ( Fig. 1d-f, Extended data Fig. 3). These metabolites could be clustered into 2 groups: microbiome derived and host metabolites (Fig. 1e).
Interestingly, the 4 most enriched metabolites upon IF were the microbiome derived 3-indolelactic acid 2, 2,3-butanediol 2, xylose 2 and indole-propionic acid (IPA) (Fig. 1f). To assess whether the IF-dependent microbiome promotes sciatic nerve regeneration, microbiota collected from IF fed animals were transplanted via faecal transplantation (FT) into AL mice: indeed, FT from IF-mice led to a signi cant increase in regenerating bres past the crush site in AL mice after SNC (Extended data Fig. 4). Since indole metabolites are primarily synthesized from tryptophan by the gram-positive bacteria Bi dobacterium (indole-3-lactic acid) 34 , Lactobacillus and Clostridium sporogenes (indole-propionic acid) 35 , the requirement of gram + bacteria for IF-dependent axonal regeneration was investigated.
Vancomycin treatment, which depletes gram + bacteria, abolished IF-dependent axonal regeneration (Fig. 2a-c), indicating that gram + bacteria are required for IF-dependent regeneration.
To identify metabolites underpinning the microbiome-dependent regenerative phenotype, serum samples were pro led by GC-MS following IF vs AL with and without vancomycin ( Fig. 2d-k, Extended data Fig. 5). Remarkably, IPA was the only metabolite signi cantly affected by vancomycin treatment with the most signi cant interaction p-value, suggesting that its amount critically depended on IF and the presence of the gram + gut microbiome (Fig. 2e, Table 1). Indeed, following IF, serum IPA was found among the highest increased metabolites, which was completely depleted by vancomycin (Fig. 2f), as con rmed by targeted Liquid Chromatography-tandem mass spectrometry (LC-MS/MS) (Fig. 2g, h). As expected, vancomycin signi cantly reduced bacterial diversity as shown by 16S rDNA sequencing (Extended data Fig. 6a, b, Supp File 1 and 2) and decreased the number of Bacteroidetes and Firmicutes, which were increased after IF (Extended data Fig. 6c). Importantly, the vast majority of Firmicutes belongs to the order of Clostridiales (Extended data Fig. 6d), and the top 5 most increased bacteria following IF are also Clostridiales (Extended data. Figure 6e), con rming Vancomycin treatment impacts IF-dependent pathways.
Together, these data suggest that gram + bacteria producing IPA are responsible for IF dependent axonal regeneration.
IPA promotes axonal regeneration after sciatic nerve crush via neutrophil chemotaxis and IFNγ Next, it was investigated whether IPA synthesis by the gut microbiome is required for axonal regeneration after SNC. Following vancomycin-mediated depletion of gram + bacteria in mice, the gut was recolonised with either a mutant Clostridium sporogenes dC (C.s. dC) bacterial strain that cannot produce IPA 36 or with a C.s. WT strain and axonal regeneration was measured 72 hours after a SNC (Fig. 3a-d). Depletion of IPA from the serum of mice recolonised with the C.s. dC (Fig. 3b) resulted in signi cantly reduced axonal regeneration compared with the C.s. WT (Fig. 3c, d). Conversely, IPA gavage signi cantly and robustly increased axonal regeneration of the sciatic nerve ( Fig. 3e-h).
Together these studies demonstrate that bacterial production of IPA in the gut critically affects axonal regeneration. Since changes in IPA production in the gut are re ected by variations in the serum, IPA was administered by intraperitoneal (i.p.) injection and axonal regeneration was assessed 72 hours after SNC.
Remarkably, systemic IPA delivery mirrored the increase in axonal regeneration elicited by IPA gavage (Extended data Fig. 7a-c), indicating that serum IPA might affect DRG neuron growth either directly or indirectly via neuronal intrinsic or extrinsic mechanisms respectively.
When IPA was administered directly into the media of cultured DRG neurons no changes in neurite outgrowth were observed (Extended data Fig. 7d-f), suggesting that IPA affects growth via an extrinsic regenerative mechanism to the DRG neurons.
To investigate IPA-dependent molecular signatures potentially affecting the regenerative ability of DRG neurons, RNA sequencing was performed from DRG preceding (naïve) or 72 hours after SNC following IPA or PBS oral gavage. IPA treatment affected the DRG gene expression programme as shown by differential gene expression analysis (Extended data Fig. 8a-c, Supp File 3 and 4), increased the expression of the gene encoding for the IPA nuclear receptor Pregnane X Receptor (Nr1i2), supporting activation of IPA signalling, and it affected mainly genes related to immune-regulatory GO categories, with the most signi cant enrichment for "neutrophil chemotaxis" at 72 hours following SNC (Fig. 4a, b, Supp File 4 and 5). Notably, the neutrophil ligand Cd177 and endothelial neutrophil chemoattractant chemokine (C-X-C motif) ligand 1 (Cxcl1) were selectively upregulated by IPA after SNC (Extended data Fig. 8c), suggesting that IPA might promote neutrophil chemotaxis towards the DRG via a CXCR2 mediated mechanism.
In support of this model, we found that IPA treatment increased the number of neutrophils within the DRG tissue, but not at the nerve crush site (Extended data Fig. 9a-h). The number of other immune cells including CD4 and CD8 T-cells, macrophages, NK and B-cells, did not differ between IPA and vehicle (Extended data Fig. 10). DRG immunolabelling showed that the majority of neutrophils were positive for CXCR2 (Extended data Fig. 9i, j), a marker of neutrophil activation and chemotaxis, suggesting that they might be activated and attracted by a CXCR2 mediated mechanism. To test this hypothesis, neutrophil chemotaxis was inhibited at the time of IPA treatment by using an anti-CXCR2 neutralising monoclonal antibody that signi cantly impaired IPA-dependent regeneration after SNC ( Fig. 4c-g, Extended data Fig. 11j-k). Accordingly, neutrophil depletion with anti-Ly6G monoclonal antibody signi cantly decreased axonal regeneration mediated by IPA compared with IgG control antibody (Extended Data Fig. 11a-i). Together these data indicate that IPA-dependent neutrophil chemotaxis is required for axonal regeneration after SNC.
RNA sequencing also identi ed an IPA-dependent increase in IFNγ signalling in the DRG, indicating a potential signalling mechanism downstream of neutrophil recruitment (Extended data Fig. 12a). Indeed, monoclonal antibody mediated neutralisation of IFNγ signi cantly reduced IPA-dependent sciatic nerve regeneration (Extended data Fig. 12b-c). In line with these data, IFNγ treatment in cultured DRG neurons and in vivo IFNγ delivery followed by ex vivo analysis of DRG neuronal outgrowth in culture revealed a signi cant increase in regenerative growth (Extended data Fig. 12d-g).
IPA delivery after sciatic nerve crush injury improves sensory neurological recovery and epidermal innervation Next, whether IPA administration at a clinically suitable time 24 hours after injury was able to accelerate the rate of sensory recovery following sciatic nerve injury was investigated. Both recovery of physiological sensory function by assessing responses to thermal stimulation and the presence of mechanical allodynia, which can be an undesired consequence of maladaptive epidermal re-innervation, were evaluated ( Fig. 4h-k). IPA treated mice showed a faster speed of recovery of thermal nociception as shown by the Hargreaves test, displaying a reduced withdrawal response time compared to PBS treated mice between day 7 and day 19 (Fig. 4j), however, no changes in mechanical allodynia, measured with Von Frey laments, between IPA-and PBS treated groups were observed (Fig. 4k). These data show that IPA leads to an accelerated recovery in thermal nociception while it does not induce mechanical allodynia following sciatic nerve injury. Importantly, anti-PGP9.5 immunostaining of the hind paw interdigital skin at 16 days post-SNC revealed a signi cantly increased epidermal innervation following IPA vs PBS, providing an anatomical substrate for functional recovery (Fig. 4l-n). Together, these experiments indicate that IPA induces sensory recovery including by improving epidermal innervation without causing neuropathic pain in a sciatic crush model of peripheral nerve injury.

Conclusions
The present study identi es a novel dietary-dependent regenerative mechanism that relies on the gut gram + Clostridiales bacterial production of the metabolite IPA. IPA accelerates axonal regeneration, epidermal innervation and sensory neurological recovery after SNC via the recruitment of neutrophils to the DRG. The gut microbial metabolite IPA, a product of bacterial metabolism of tryptophan in the lower gut, and its xenobiotic nuclear receptor PXR have been implicated in a variety of immune responses in several tissues and cell types [37][38][39][40][41] , including in the gut microenvironment, where an immune suppressive and anti-in ammatory function for IPA and PXR signalling have been described [37][38][39] . Conversely, PXR activation in vascular endothelial cells has been shown to initiate an innate in ammatory response 41 , which results in neutrophil recruitment 42 , supporting our model that suggests an IPA-dependent increase of trans-endothelial neutrophil chemotaxis.
While neutrophils are required for axonal regeneration in the PNS 43 and CNS [44][45][46] via changes in their recruitment at the lesion site, our study supports a regenerative role for neutrophils by communicating regenerative signals such as IFNγ at the level of the DRG where the neuronal cell bodies reside.
IFNγ-dependent axonal regeneration is in line with increased in ammatory signalling by neutrophils, however, a role for additional cytokines supporting IPA-dependent axonal regeneration cannot be ruled out.
Since the gut microbiome undergoes profound changes in human subjects upon nervous system injury 47 , identifying gut microbiome derived regenerative small molecules for oral or intravenous administration, could provide promising new therapeutic possibilities to heal nerve injury, promote axonal regeneration and enhance neurological recovery.

Animal Husbandry
All animal experiments were carried out according to UK Home O ce regulations in line with the Animals (Scienti c Procedures) Act of 1986 under personal and project licenses registered with the UK Home O ce. Mice were maintained in-house under standard housing conditions (12-h light/dark cycles), 24 hr access to water and standard chow diet. All mice were of a C57BL6 background (male, 20-30 g, 6-8 weeks of age). Experimenters were blinded to experimental groups during scoring and quanti cations.
Sample sizes were chosen in accordance with similar previously published experiments.

Intermittent Fasting
Body weight-matched 6-8 week old male C57BL/6 mice were randomly assigned to intermittent fasting (IF), or ad libitum (control) treatment groups. The IF group did not have access to food (fasting) during the rst 24 h and then every second day after that (e.g. fasting during 0-24 h, 48-72 h, 96-120 h) with ad libitum access to food on the alternating days (24-48 h, 72-96 h, 120-144 h). Pre-weighed food was provided in the food hopper of their home cage at 9:30 am (unless a fasting day for the IF groups), and leftover food was weighed 24 h later.

Sciatic nerve crush (SNC) surgery
Mice were anesthetized with iso urane (5% induction, 2% maintenance), shaved on the hind limbs and lower back, sterilised with iodine, and an ophthalmic solution was applied to the eyes to prevent drying. An incision was made on the skin and the biceps femoris and the gluteus super cialis were opened by blunt dissection and the sciatic nerve was exposed using a surgical hook. The sciatic nerve crush was performed orthogonally for 20 seconds (45 seconds for reinnervation experiment) using a 5 mm surgery forceps (91150-20 Inox-Electronic). The crush was performed at approximately 20 mm distally from the sciatic DRG.

Faecal transplantation
Fresh faeces were collected every day from the cage of 5 healthy C57BL/6 mice, approximately 200 mg/ mouse. A fresh cage was used each day.. 1g of faeces was then homogenized in 10 ml of PBS for 30 seconds at 3000 rpm at 4°C. Thereafter, the supernatant was transferred to a new tube and centrifuged for 5 min at 12000 rpm at 4°C. The pellet was resuspended in 2.5 ml PBS and 500 µl gavaged to each mouse.

Antibiotic administration
Vancomycin (Sigma, V2002) was administered through the drinking water (50mg/kg/day; 214.27 mg/l), which was replenished every second day.

Clostridium sporogenes recolonization
WT and dC (mutant for (R)-phenyllactyl-CoA dehydratase beta subunit of the phenyllactate dehydratase complex [ dC]) Clostridium sporogenes were cultured overnight in trypticase yeast extract medium anaerobically. Bacterial cultures were mixed with glycerol to reach the nal concentrations of: WT 6.01E+06 CFU/ml and C.s. dC mutant 5.46E+06 CFU/ml in hermetically sealed glass vials. Mice were pre-treated with vancomycin in drinking water for 3 days. C.s. WT or C.s. were transplanted to the cecum via oral gavage for 10 consecutive days at 1E+06 CFU/day.

Serum and cecum preparation
Blood was extracted via heart punctuation and collected into a covered test tube. The blood was allowed to clot for a minimum of 30 min at room temperature and centrifuged at 2000 g for 10 min at 4 °C. The supernatant was then collected into a fresh tube and stored at -20 °C until use.
Cecum content was removed from the cecum using sterile forceps into a tube and snap-frozen in liquid nitrogen immediately.

Gas chromatography -Mass Spectrometry (GC-MS) untargeted metabolomics
Serum samples (100 μl) were prepared as follows: I) samples were spiked with 10 μL internal standard solution (myristic acid-d27, 750mg/ml), II) 850 μL of ice cold methanol were added, followed by centrifugation for 20 min (4 o C, 16000 g), III) 750 μL of supernatants were transferred to silanized dark 2mL autosampler vials and were evaporated to dryness in a rotational vacuum concentrator (45 o C, 20 mbar, 2 h), IV) 50 μL of methoxyamine solution (2% in pyridine) were added and the samples were incubated overnight at room temperature and nally, V) 100 μL of N-methyl-trimethylsilyltri uoroacetamide (MSTFA) +1% trimethylchlorosilane (TMCS) solution were added, the samples were incubated at 60 ºC for 1 h and were transferred to dark autosampler vials with 250 μL silanized inserts. The samples were analysed in an Agilent 7890B-5977B Inert Plus GC-MS system. 2 μL of each sample were injected in a split inlet (1:10 ratio). The chromatographic column was an Agilent ZORBAX DB5-MS (30 m X 250 µm X 0.25 µm + 10m Duraguard). The temperature gradient was 37.5 min long and the mass analyser was operated in full scan mode between 50-600 m/z. The detailed instrumental conditions are described elsewhere (Agilent G1676AA Fiehn GC/MS Metabolomics RTL Library, User Guide, Agilent Technologies, https://www.agilent.com/cs/library/usermanuals/Public/G1676-90001_Fiehn.pdf). Quality Control (QC) samples were created by pooling equal amounts of every serum sample of the study and were analysed interspaced in the analytical run. Study samples were randomized before sample preparation. Peak deconvolution, alignment and annotation and were performed with the use of the Fiehn library via the software packages AMDIS (NIST), Mass Pro ler Pro and Unknowns (Agilent technologies) in the pooled QC samples. Peak picking was performed with the GAVIN package (A software complement to AMDIS for processing GC-MS metabolomics data. doi: 10.1016/j.ab.2011.04.009) 48 . Non-reproducible (CV>30% in QC samples) and contaminated (blank > 20% of the mean QC levels) metabolic features were removed from the dataset. This data set was then used to build orthogonal partial least-squares-discriminant analysis (OPLS-DA) (IF vs AL) or partial least-squares-discriminant analysis (PLS-DA) (IF, IF-V, AL, AL-V), focusing on the differences among the experimental groups. The OPLS algorithm derives from the partial least-squares (PLS) regression method 49 . The method explains the maximum separation between class samples Y (n dummy variables for n classes) by using the GC-MS data X. Here the ropls R package (http://bioconductor.org/packages/release/bioc/html/ropls.html) was used, which implements the PCA, PLS-DA and OPLS-DA approaches 50 . It includes R 2 and Q 2 goodness-of-t and goodness-of-prediction statistics, permutation tests, as well as scores, loadings, predictions, diagnostics, outliers graphics 50 . R2X describes the percentage of predictive and orthogonal variation in X that is explained by the full model. Isotopes) as internal standard (IS) were used. Acetonitrile, methanol and formic acid were of ULC-MS grade was supplied from Bio-Lab. Water with resistivity 18.2 MΩ was obtained using Direct 3-Q UV system (Millipore). Standard curve was built using concentration range of 3-indole propionic acid 0.01-12 µg/mL, with nal concentration of IS 100 ng/mL.

Extract preparation
Plasma (30 µL) and IS (10uL, 1 ug/mL) was incubated 10 min, then 500 µL of methanol was added. The mixture was shaken at 10°C for 30 min (ThermoMixer C, Eppendorf), and centrifuged at 21,000 g for 10 min. Collected supernatant was evaporated in speedvac and then in lyophilizer. Before LC-MC analysis, the obtained residue was re-suspended in 100 µL of 20%-aq methanol, centrifuged twice at 21,000 g for 5 min to remove insoluble material. The soluble part was placed to insert of LC-MS vial. Some samples were diluted 1/20 with 20%-aq methanol.

LC-MS analysis
The LC-MS/MS instrument consisted of an Acquity I-class UPLC system (Waters) and Xevo TQ-S triple quadrupole mass spectrometer (Waters) equipped with an electrospray ion source and operated in positive ion mode. MassLynx and TargetLynx software (version 4.1, Waters) were applied for the acquisition and analysis of data. Chromatographic separation was done on a 100 mm × 2.

16S rRNA amplicon sequencing data analysis
Only single-end R1 sequences were analysed. Data processing was carried out in R statistical environment version 3.5.2 according to the DADA2 pipeline 51 as follows; primers were trimmed from demultiplexed sequences, amplicon sequence variants (ASVs) were generated using DADA2 package version 1.10.1, and taxonomic assignment was performed using the SILVA rRNA database version 132 [https://www.arb-silva.de/]. ASVs with no phylum assigned were excluded from analyses to yield a nal number of n = 906 annotated ASVs. Alpha and beta diversity analyses were carried out using phyloseq package version 1.26.1 52 . Alpha diversity (Shannon Index) was calculated using raw counts 53 . Groups were compared using Mann Whitney U statistical tests. For beta diversity analyses, data was normalised by log-transformation with a pseudo count of +1. Differential abundance of groups were compared in the DESeq2 package version 1.22.2 54 using the Wald test with Benjamini and Hochberg adjustments for multiple testing (p value set as 0.05). Alpha diversity is a measure of diversity or "microbial variation" within a (single) sample while beta diversity is a measure of diversity across (several) samples.
RNA sequencing DRG were collected from animals that had undergone 10 days of IPA or PBS treatment and/or followed by sciatic nerve crush. Sciatic DRG were extracted 72h after sciatic nerve crush or from naïve animals (surgeries were performed as described above) and collected into RNAlater. DRG were crushed with RNase free micro pestle and RNA was then immediately extracted using RNAeasy kit (Qiagen), according to manufacturer's guidelines. Residual DNA contamination was removed by treating the spin column with 40 units of RNase-free DNase I (Qiagen) for 15 min at 23 °C prior to RNA elution. RNA concentrations and purity were veri ed for each sample following elution with the Agilent 2100 Bioanalyzer (Agilent). RNA with RIN factors above 8.5 were used for library preparation. cDNA libraries for each sample were generated by the Imperial BRC Genomics Facility using the TruSeq Sample Preparation Kit A (Illumina, San Diego CA) and sequenced using Illumina HiSeq 4000 (PE 2x75 bp) sequencing. Gene ontology (GO) was performed on differentially expressed genes with DAVID 6.8 (Database for Annotation, Visualization, and Integrated Discovery (http://david.abcc.ncifcrf.gov/)). Differentially expressed genes were selected using a threshold of P<0.05 and |1.5|<FC (fold change) or no FC cut-off. To identify IPA-dependent speci c differentially expressed genes following SNC, we compared IPA-SNCvsPBS and PBS-SNCvsPBS groups and selected the uniquely up-or downregulated genes in the IPA-SNCvsPBS group.
Sciatic nerve regeneration 24 or 72 hours following the surgery, sciatic nerves were dissected and post-xed in 4% PFA, incubated at 4°C for 1h and transferred into 30 % sucrose for at least 3 days. Subsequently, the tissue was embedded and frozen in Tissue-Tek OCT and maintained at -80°C until cut into 11 µm sagittal sections. Tissue sections were immunostained for SCG10 (1:1000, rabbit, Novus) a marker for regenerating axons. The crush site was identi ed by deformation of the nerve and disruption of axons coinciding with highest SCG-10 intensity. The SCG-10 intensity was measured in 500 µm intervals along the length of the nerve distal to sciatic nerve crush site. The intensity was normalised to the SCG-10 intensity before the crush site and plotted as fold-change. 4-6 sections per animal were analysed and imaged with a HWF1 -Zeiss Axio Observer with a Hamamatsu Flash 4.0 fast camera using 10x magni cation.
Glass coverslips were coated with 0.1 mg/ml PDL, washed and coated with mouse Laminin 2ug/ml (Millipore) for 1-2 hours each previous to the start of the experiment. Sciatic DRG from adult animals were dissected and collected in Hanks balanced salt solution (HBSS) on ice. The DRG were transferred into a digest solution (5mg/ml Dispase II (Sigma), 2.5 mg/ml Collagenase Type II (Worthington) and incubated in a 37 °C water bath for 45 min, with occasional shaking for 30 seconds . Thereafter, the DRG were washed and manually dissociated with a 1ml pipette in media containing 10 % heat inactivated FBS (Invitrogen) and 1x B27 (Invitrogen) in F12:DMEM (Invitrogen). Pipetting was continued until DRG were fully dissociated and no clumps could be observed. Next, the cell suspension was spun down at 1000 rpm for 4 min and resuspended in culture media containing 1x B27 and Penicillin/Streptomycin in F12 : DMEM. 3500 cells were plated on each coverslip (laminin and PDL coated) and maintained in a humidi ed culture chamber with 5% CO2 at 37 °C, for 12 hours before xed with 4% PFA and immunostained.

Microscopy
Photomicrographs were taken with a Nikon Eclipse TE2000 microscope with an optiMOS scMOS camera using 10x or 20x resolution Zeiss Axio Observer with a Hamamatsu Flash 4.0 fast camera using 10x or 20x magni cation. Confocal images were taken with a Leica TCS SP8 II confocal microscope at 40X or 60X magni cation and processed with the LAS-AM Leica software (Leica).

Image Analysis for IHC and ICC
Image analysis was conducted using ImageJ (Fiji) software. All analysis was performed by the same experimenter who was blinded to the experimental groups. DRG images were taken using a Nikon Eclipse TE2000 microscope with an optiMOS scMOS camera at either 10x or 20x magni cation. Images were analysed by counting the number of cells per area, measuring the intensity against the background or calculating percentage of cells with positive staining.
For neurite length analysis between 15 and 20 images were taken per coverslip and analysed using NeuronJ plugin for Image J software (Image J). All analyses were performed in blind. Approximately 45-60 cells were analysed per animal and condition.

Statistical analysis
Results are graphed as mean ± SEM. Statistical analysis was carried out using GraphPad Prism 7.
Normally distributed data was evaluated using a two-tailed unpaired Student's t-test or a one-way ANOVA when experiments contained more than two groups. Dunnett multiple comparisons test or multiple comparison testing corrected by FDR with Benjamini and Hochberg were applied when appropriate. The two-way ANOVA, Tukey's or Sidak's test, was applied when two independent variables on one dependent variable were assessed. A threshold level of signi cance was set at P<0.05. Signi cance levels were de ned as follows: * P<0.05; ** P<0.01; *** P<0.001; **** p<0.0001. All data analysis was performed blind to the experimental group.

Competing interests
The authors declare no competing interests.
Author contributions