Gut instinct: Sex differences in the gut microbiome are associated with changes in adolescent nociception following maternal separation in rats

Adolescent chronic pain is a growing public health epidemic. Our understanding of its etiology is limited; however, several factors can increase susceptibility, often developing in response to an acute pain trigger such as a surgical procedure or mild traumatic brain injury (mTBI), or an adverse childhood experience (ACE). Additionally, the prevalence and manifestation of chronic pain is sexually dimorphic, with double the rates in females than males. Despite this, the majority of pre‐clinical pain research focuses on males, leaving a gap in mechanistic understanding for females. Given that emerging evidence has linked the gut microbiome and the brain–gut–immune axis to various pain disorders, we aimed to investigate sex‐dependent changes in taxonomic and functional gut microbiome features following an ACE and acute injury as chronic pain triggers. Male and female Sprague Dawley rat pups were randomly assigned to either a maternal separation (MS) or no stress paradigm, then further into a sham, mTBI, or surgery condition. Chronically, the von Frey test was used to measure mechanical nociception, and fecal samples were collected for 16S rRNA sequencing. Animals in the surgery group had an increase in pain sensitivity when compared to mTBI and sham groups, and this was complemented by changes to the gut microbiome. In addition, significant sex differences were identified in gut microbiome composition, which were exacerbated in response to MS. Overall, we provide preliminary evidence for sex differences and ACE‐induced changes in bacterial composition that, when combined, may be contributing to heterogeneity in pain outcomes.


INTRODUCTION
Adolescent chronic pain is a growing public health epidemic (King et al., 2011).Current understanding of the etiology of chronic pain is poor, although it is more prevalent in females and often manifests from an acutely painful trigger.For example, up to 54% of youth who undergo a surgical procedure develop chronic pain (Landman et al., 2011).Chronic pain is also one of the most common and debilitating symptoms following adolescent mild traumatic brain injury (mTBI) (Defrin, 2014;Kwan et al., 2018;Uomoto et al., 1993).Given that children and adolescents have the highest rates of mTBI (Yeates, 2010), this is alarming.The source of mTBI pain is generally unknown and may occur as a direct result of the injury or be a secondary consequence of the trauma (Defrin, 2014).Irrespective of the acute trigger (surgical or mTBI), we do not understand why these injuries result in persistent pain for some individuals, but not others.Differences in individual susceptibility to chronic pain may begin early in life.Adverse childhood experiences (ACEs) are one of the most serious threats to lifelong health and are linked to mortality, as well as a staggering number of mental and physical health problems (Anda et al., 2006;Dube et al., 2003;Felitti et al., 1998).ACEs are often categorized as emotional or physical abuse or neglect and household dysfunction (Felitti et al., 1998).Emerging research suggests that certain ACEs, such as neglect, may be more predictive of specific health outcomes, including chronic pain and PTSD symptomology (Beveridge et al., 2018).ACEs may prime the brain to be hyper-responsive to later insults, thereby modifying neural circuitry responsible for nociception.Maternal separation (MS) is a rodent model of early life neglect that reduces maternal care (Muhammad et al., 2012) and is known to modify neuroplasticity and increase pain sensitivity and hyperalgesia in adolescence and adulthood (Amini-Khoei et al., 2017;Mizoguchi et al., 2019;Salberg et al., 2021;Salberg et al., 2023;Weaver et al., 2007), with effects being more prominent in females (Burke et al., 2017;Salberg et al., 2020;Salberg et al., 2023).Although there is an established link between early adversity and chronic pain, we do not understand the systemic mechanisms linking ACEs to chronic pain in adolescence.
Recent literature, from both animal and clinical studies, has described the gut-brain-immune axis, a complex communication system between the gut microbiota, brain, and immune system, that plays an integral role in many neuropathological and homeostatic processes (Cryan et al., 2012(Cryan et al., , 2019;;Sgro et al., 2022).Importantly, the specific composition and functionality of the gut microbiome and its microbial metabolites have been linked to optimal brain development in children (Christian et al., 2015), psychological conditions such as anxiety in animals and humans (Foster et al., 2013), and even the perception of pain in young adults (Shiro et al., 2021).Given that the gut microbiome matures in a sex-dependent manner in both rodent and human populations (Jasarevic et al., 2016;Shin et al., 2019) and evidence links it to other pain disorders, such as migraine (Sgro et al., in press) and fibromyalgia (Minerbi et al., 2019), some of the heterogeneity in adolescent chronic pain development, particularly the increased prevalence in females, may manifest in response to microbial modifications within the gut.
Therefore, this study aimed to examine the taxonomic and functional gut microbiome features implicated in the chronification of pain, with a specific emphasis on how these are modified by sex and early life adversity.We hypothesized that sex differences in the specific microbial features are associated with chronic pain phenotypes and that these gut microbiome changes would be exacerbated in animals exposed to MS and differ between the mTBI and surgery groups.

Animals
Sprague Dawley rats were acquired from the Monash Animal Research Platform, with all experiments being approved by the Alfred Medical Research and Education Precinct Animal Ethics Committee and carried out in accordance with the Precinct Animal Centre (E/1928/2019/M).This study is reported, and all methods were performed, in accordance with the relevant guidelines and regulations and also as per ARRIVE guidelines (Percie du Sert et al., 2020).Animals were maintained on a 12:12 h light:dark cycle in a temperature-controlled room with ad libitum access to standard rat chow and water.Sires and dams were pair mated in-house, and their offspring were used for this study.Pups were utilized as part of a more comprehensive study, with additional groups and data published elsewhere (Salberg et al., 2023).All animal characteristics and behavior data shown include the numbers from the larger study, while a subset was used for all microbiome analysis.Male and female pups were randomly assigned to either an MS or no stress paradigm, then further into to a sham, mTBI, or surgery condition (n = 7-10/sex/group).Appropriate group sizes were determined from power calculations with a power of 0.80, medium effect size of 0.5, and  of .05,which gave a sample size of n = 6.5/sex/group.Given that animals were mated in-house and litters were not culled down, any additional animals were randomly allocated across groups.At post-natal day (p) 22, between the stress and injury procedures, pups were weaned into cages of three to four with cage-mates from the same experimental condition.To control for potential litter effects, a maximum of two pups per sex per litter were weaned into each experimental group.Thirty days later, von Frey testing was performed for mechanical pain sensitivity.Once completed, animals were humanely euthanized (40 days after surgery) via isoflurane anesthesia until they were unresponsive followed by rapid decapitation, at which point fecal and small intestine samples were collected for downstream processing.Given that a sample size of n = 6.5/sex/group was deemed sufficient from power calculations, a subset (n = 6-7/sex/group) of samples was randomly chosen for downstream processing.
See Figure 1 for experimental timeline.

Maternal separation
The MS procedure, as described elsewhere (Muhammad et al., 2012;Salberg et al., 2020), began on p2 and continued for 12 continuous days, ceasing on p13.Briefly, this involved removing the entire litter of pups from their mother for 4 h/day (0900-1300), and placing them together in a separate, half-heated cage.At the end of the 4 h, the pups were returned to their home-cage with their mother.The pups allocated to the no stress paradigm were left undisturbed with their mothers for the duration of the procedure.All pups were then weaned at p22.

Injury
Injury (sham/mTBI/surgery) was induced on p35.The sham procedure involved anesthetizing the animals with 5% isoflurane at 1 L/min O 2 , then allowing them to recover in a clean cage before being placed back into their home-cage.The mTBI was induced using the lateral impact model, as previously described (Mychasiuk et al., 2016).Briefly, animals were anesthetized in an induction chamber with 5% isoflurane at 1 L/min O 2 and then placed in a prone position with the left temporal lobe positioned to receive the impact.A small weight was then propelled toward the rats' head using pneumatic pressure, first impacting a small aluminum plate acting as a helmet.The impact caused the rat to be rotated 180˚, mimicking the acceleration and deceleration forces often seen in sports concussions.The animal was then placed in a clean cage and allowed to recover before being placed back in its home-cage.The surgery was induced using the Brennan model of plantar incision (Brennan et al., 1996).Animals were anesthetized in an induction chamber with 5% isoflurane at 1 L/min O 2 and then transferred to a nosecone at 2%-3% isoflurane at 1 L/min O 2 for maintenance throughout the surgery.Aseptic conditions were followed for the duration of the surgery, whereby the left hind paw was disinfected with ethanol and chlorohexidine and then an 11-blade scalpel was used to make a 1 cm incision longitudinally in the skin.The underlying plantaris muscle was uncovered with forceps and incised longitudinally 3× before the wound was closed with two simple interrupted 6-0 prolene sutures.The wound was then re-sterilized, and the animal was placed in a clean cage to recover before being returned to its home-cage.Time to right (loss of consciousness) was measured following each procedure as time required for the animal to flip from a supine to prone position.Given that the injury procedures were all minor, and designed to produce mild acute pain, no analgesics were provided for any of the injury conditions.

von Frey
The von Frey task was performed in adolescence (p65-67), 30 days post-injury, as a measure of mechanical nociceptive sensitivity (Salberg et al., 2020(Salberg et al., , 2021)).The test was run over 3 days, with the first two being habituation days whereby the animals were placed in the testing apparatus and allowed to explore, undisturbed for 20 min before being returned to their home-cage.On the testing day, filaments of increasing size were applied to the right hind paw (uninjured for the surgical group), with number of reactions (hind paw withdrawal) out of five being recorded by a researcher blinded to experimental conditions.The uninjured paw was used to allow for nociceptive comparisons between mTBI and surgery animals.Once a 5/5 reaction was observed, the test was stopped, and this filament size was used for analysis.A larger filament size was indicative of decreased sensitivity.

16S rRNA
Animals were weighed and then euthanized on p75, and fecal samples were collected from the cage, flash frozen, and stored at −80˚C until processing.Microbial DNA was extracted from samples using the QIAamp PowerFaecal Pro DNA Kit (Qiagen), in accordance with the manufacturer's instructions and in conjunction with the QIAcube (Qiagen).Samples were processed using a dual-PCR approach, according to protocol (15044223; 16S Metagenomic Sequencing Library Preparation).Extracted microbial DNA was amplified using gene-specific sequences targeting the 16S V3 and V4 region.Primer sequences were as follows: PCR reactions were 25 μL each, comprising 2.5 μL Microbial DNA (5 ng/μL), 5 μL Amplicon PCR Forward Primer 1 μM, 5 μL Amplicon PCR Reverse Primer 1 μM, 12.5 μL 2× KAPA HiFi HotStart Ready Mix.Cycling parameters were as follows: 95˚C for 3 min, followed by 25 cycles of 95˚C for 30 s, 55˚C for 30 s, and 72˚C for 30 s and then 72˚C for 5 min.The Invitrogen Qubit and dsDNA HS chemistry were used to assess library quantity, and the Agilent Fragment Analyzer 5200 with the HS NGS Fragment Kit was used to assess quality.An Illumina MiSeq using a MiSeq Reagent Kit V3 (600 cycles) was used according to the manufacturer's instructions in order to pool libraries in an equimolar ratio, and denature, dilute, and sequence them.
Raw sequences were processed using the microbiome-dada2 pipeline (see availability of data and material section) using the DADA2 (version 1.22.0)R package (Callahan, Mcmurdie, et al., 2016).Briefly, indexed fastq files were demultiplexed using the iu-demultiplex function (version 2.8) from illumina-utils (Eren et al., 2013), primers and adapters removed with cutadapt (version 2.10) (Martin, 2011), reads filtered and trimmed, sequencing error models generated, sequences dereplicated, amplicon sequence variants (ASVs) inferred, paired-ends merged, and chimeras removed.Bacterial 16S ASVs were assigned a taxonomy using the SILVA database train set (version 138) and the SILVA species assignment dataset (version 138) for exact sequence matching.A phylogenetic tree based on ASV sequences was built by performing a multiple-alignment using DECIPHER (version 2.22.0),followed by the construction of a neighbor-joining tree using phangorn (version 2.9.0) before fitting a GTR+G+I (generalized time-reversible with Gamma rate variation) maximum likelihood tree using the neighbor-joining tree as a starting point as previously described (Callahan, Sankaran, et al., 2016).Samples with less than 5000 ASVs were excluded from the dataset, and ASVs below 1% prevalence or unassigned at the Phylum level were filtered out.Shannon index was determined using the estimate_richness function of the phyloseq (version 1.38.0)R package (McMurdie et al., 2013).ASV counts were normalized by cumulative sum scaling (CSS) using the calcNormFactors function from MetagenomeSeq (version 1.36.0)(Paulson et al., 2013) followed by log transformation.
Statistical analyses were performed in R (version 4.1.0)(Team, 2021), plots were generated using ggplot2 (version 3.3.5)(Wickham, 2016), and heatmaps were constructed with ComplexHeatmap (version 2.8.0) (Gu et al., 2016).For plots, the box represents the interquartile range (IQR, i.e., 25th percentile to 75th percentile), and the tails show the minimum and maximum.Differences in bacterial alpha-diversity were assessed by non-parametric Wilcoxon Rank Sum testing.Non-parametric tests were used because the assumption of data normality was not upheld (significant variation from normality as assessed by Shapiro-Wilk normality tests).
Principal coordinates analysis was performed on weighted UniFrac distances based on the inferred de novo phylogenetic tree constructed in the DADA2 pipeline using the ordinate function of the phyloseq package.Differential abundance testing for bacterial data was performed via a custom wrapper (see availability of data and material section) around the R limma (version 3.48.3)function (Ritchie et al., 2015).A fixed random-number seed value of 2 was set to ensure maximum reproducibility of tools requiring random pseudo-numbers.The alpha level of significance of .05 was used for all statistical tests.Multiple testing correction was not performed when running limma linear regression analyses that aimed to identify differentially abundant (DA) genera in the stool samples.Limma linear regression was used as a feature selection step, following which post hoc Wilcoxon ranked sum tests were performed to determine significance.

Reverse transcription quantitative real-time polymerase chain reaction
A ∼1 cm sample of proximal small intestine was collected, flash frozen at the time of sacrifice, and stored at −80˚C until processing.RNA was extracted from samples according to the manufacturer's protocol of the RNeasy Mini Kit (Qiagen) in conjunction with the QIAcube (Qiagen).RNA was then reverse transcribed to cDNA using qScript™ XLT cDNA SuperMix (Quantabio) for use in reverse transcription quantitative real-time polymerase chain reaction (qRT-PCR).The 2 −ΔΔCt method was used, with the housekeeping genes Ywhaz and Cyca being used for normalization Bonefeld et al., 2008;Pfaffl, 2001.The genes of interest were Occludin (OCC) and Tight Junction Protein 1 (TJP1).See Table 1 for primer sequences obtained from IDT and cycling parameters.All samples were run in duplicate on the QuantStudio7 (Thermo Fisher Scientific), with each well of a 384-well plate containing 20 ng cDNA, 1 X SYBR Green FastMix ROX, and 0.5 μM of both forward and reverse primers.
Three-way aligned rank transformation analyses of variance (ANOVAs) with sex (male; female), stress (no stress; MS), and treatment (sham; mTBI; surgery) as factors were run for all animal characteristics, behavior, and tight junction analysis, followed by post hoc Wilcoxon rank sum testing.ARTool R package (version 0.11.1) was used, and statistical significance was considered at p < .05.All animals were included in analyses.

Animal characteristics and behavior
There were significant sex effects observed for body weight, and treatment effects were observed in time to right and  2c).The three-way aligned rank transformation ANOVA for the von Frey task demonstrated a significant treatment effect, F(2, 88) = 31.75,p = 4.16e-11.The Wilcoxon rank sum post hoc testing revealed that the treatment effect was driven by the surgery group, whereby animals that underwent the surgery required lower filament weights to respond compared to sham and mTBI animals, indicating increased sensitivity.No other significant effects were observed, p's > .05(see Figure 2d).

Bacterial composition and diversity
Seventy-six (100%) rat fecal samples were retained following sequencing quality control measures.In total, 2544 unique bacterial ASVs were identified from seven phyla and 44 genera.The most prevalent genera irrespective of early life stress were Lactobacillus, Bacteroides, Prevotellaceae NK3B31 group, and Turicibacter, accounting for 61.2% of total reads (Figure 3a).Alpha-diversity, measured by Shannon index, was similar for both male and female animals, regardless of early life stress or subsequent treatment (Figures 3b,c).Further, ordination of weighted UniFrac distances showed considerable post-treatment fecal microbiome overlap between rats that experienced early life MS and controls.The permutational multivariate ANOVA identified a modest difference (Sum-Sqs = 0.139; R 2 = 0.071; p = .041)in beta-diversity as a result of early life stress (MS) for the mTBI group (Figure 3d).

Effect of MS on bacterial composition
Early life MS itself resulted in changes in fecal microbiome composition after controlling for both treatment group and sex.Levels of Akkermansia and Lachnospiraceae NK4A136 group were particularly low in fecal samples from rats that experienced MS (p = .0026and p = .003,respectively), with Bacteroides (p = .0037)and

Tight Junction Proteins
The three-way aligned rank transformation ANOVAs for TJP1 and OCC demonstrated no significant effects, with post hoc Wilcoxon rank sum test statistical values shown in Figure 7, p's > .05(Figure 7a,b).

DISCUSSION
Studies from our laboratory (Salberg et al., 2023) and others (for review, see Bartley et al., 2013) have demonstrated substantial sex differences in a variety of pain outcomes.However, the mechanisms driving pain have been founded predominantly in male-centric research, leaving the factors responsible for pain in females virtually unstudied (Mogil, 2012(Mogil, , 2020)).Previously, our laboratory demonstrated that pro-inflammatory cytokines and markers of microglia were increased in males, whereas this effect was not observed in females despite similar behavioral pain phenotypes (Salberg  , 2023).Therefore, this study aimed to investigate a novel pathophysiological mechanism that could be contributing to pain in females.Given the strong bidirectional relationship between the gut microbiome, the immune system, and the central nervous system, termed the brain-gut-immune axis, we aimed to investigate the sex effects underlying this link.
Both pre-clinical and clinical literatures have suggested that the gut microbiota plays a role in various pain conditions, from visceral, neuropathic, and chronic pain (Luczynski et al., 2017;Pusceddu et al., 2018;Rea et al., 2017;Yang et al., 2019) to irritable bowel syndrome (Ghaisas et al., 2016;Menees et al., 2018) and fibromyalgia (Minerbi et al., 2019), as well as neurological disorders such as traumatic brain injury, Alzheimer's disease, and Parkinson's disease (Ghaisas et al., 2016;Sharon et al., 2016).Our results support this literature, with injured groups displaying greater pain sensitivity and altered gut microbiome composition compared to shams.Further, stress is one of the primary agents known to influence the brain-gut-immune axis across the lifespan (Cryan et al., 2019).Given that stress increases intestinal permeability, activates mucosal immune responses, and alters gut microbiome composition (Dinan et al., 2012), we also examined the role of MS on bacterial composition and pain outcomes in our preclinical cohort.At the behavioral level, the plantar incision surgery increased pain sensitivity.We previously hypothesized (Salberg et al., 2023) that the sterility of the procedure could influence pain outcomes, with the non-sterile, open wound of the surgery likely increasing inflammation and microglial activity, perpetuating nociceptive responses.Conversely, the mTBI was induced using a closed-head model, which would have minimized systemic inflammation.A similar effect was observed on bacterial composition, whereby the injury mechanism altered genera abundance, with surgery inducing more substantial changes than the mTBI.For instance, in the no stress animals, only Akkermansia was increased in the mTBI group, whereas Prevotellaceae NK3B31, Lactobacillus, Akkermansia, and Romboutsia were altered in the surgery animals.Our increases in Prevotellaceae NK3B31 are consistent with previous literature, showing an association between Prevotellaceae NK3B31 and the inflammatory response (Jia et al., 2022).Alterations in Romboutsia are also consistent with literature suggesting a correlation with proinflammatory cytokines; however, we found decreases in bacterial abundance, but increases were observed in ulcerative colitis models (Wang et al., 2022;Wu et al., 2021).Thus, bacterial abundance was differentially altered based on the model of injury induced.
Effects of early adversity were also evident in bacterial composition, with MS in itself modifying the gut microbiome after controlling for all other factors.Early life stress has previously been linked to long-term deficits, associated with increased inflammation and microglial activity as well as an altered stress response, defined by changes in hypothalamicpituitary-adrenal (HPA) axis activity and gene expression (i.e., glucocorticoid receptor) functioning (Aisa et al., 2007; Diz- Chaves et al., 2013;Kalinichev et al., 2002).Given these established links between the gut and the brain, and the role of the early environment on gut microbiome colonization (Cho et al., 2020;Diaz Heijtz et al., 2011), it is not surprising that we see MS-induced changes in the gut microbiome.This is consistent with other groups that have shown increased intestinal permeability, gut dysbiosis, and gut microbiota changes following MS (Cong et al., 2015;Cresci et al., 2015).The release of cortisol following stress has also been shown to influence gut microbiome composition (Brenner et al., 2021;Dinan et al., 2012).Regardless of injury mechanism or sex, the MS animals demonstrated decreases in Akkermansia, Lachnospiraceae NK4A136, Bacteroides, and Lactobacillus and increases in Bifidobacterium and Prevotellaceae UCG-001.Given that we expected MS to induce negative changes in the gut microbiome, increases in Prevotellaceae UCG-001, which generates beneficial shortchain fatty acids (SCFAs) and is associated with reduced inflammation (Zou et al., 2020), were unexpected.However, this may have be a compensatory mechanism to counteract increased systemic inflammation that is often associated with MS (Salberg et al., 2023).Our decreases, however, are consistent with the literature.Given that these bacteria are all beneficial and associated with health and intestinal barrier integrity, the loss of these species following early adversity is expected.Lachnospiraceae NK4A136 has beneficial and probiotic properties, with its ability to produce butyrate, which is critically involved in microglia maturation and function (Erny et al., 2015;Kohler et al., 2016).Dietary increases in this species have also been shown to improve gut barrier function in ageing rats (J.Li et al., 2019), and perturbation of the gut microbiota in obese mice that resulted in reduced circulating proinflammatory cytokines also increased SCFA-producing bacteria such as Lachnospiraceae NK4A136 and Lactobacillus (Hu et al., 2019).Similarly, Akkermansia has been consistently associated with health, contributing to improved gut barrier function and reduced high-fat dietinduced weight gain (Everard et al., 2013;Plovier et al., 2017), strengthening of the intestinal barrier (Reunanen et al., 2015), correction of metabolic disorders (Everard et al., 2013), and reduction in inflammatory markers (Schneeberger et al., 2015).
Lastly, we observed significant sex differences in bacterial composition, with increases in Romboutsia and Turicibacter in no stress males and increases in Alloprevotella and Prevotellaceae UCG-001 in no stress females.Interestingly, sex interacted with the stress paradigm to produce opposing results, with Romboutsia increasing in MS females and Prevotellaceae UCG-001 increasing in MS males.Lachnospiraceae NK4A136, Bacteroides, Clostridium sensu stricto 1, and Akkermansia also increased in MS females, while Lactobacillus increased in MS males.Of note, Lachnospiraceae NK4A136, Clostridium sensu stricto 1, and Akkermansia were virtually undetected in MS males.These findings in particular are critical to interpreting sex differences in pain susceptibility and may provide an important foundation for our understanding of the mechanisms contributing to pain outcomes in females.Specifically, the increases in Lachnospiraceae NK4A136, Clostridium sensu stricto 1, and Akkermansia in MS females, and complete lack of these bacteria in MS males, suggest a mechanism whereby these bacteria contribute to increased risk for pain.The lack of Akkermansia and Lachnospiraceae NK4A136 in MS males, which are health promoting and associated with reduced inflammation (Hu et al., 2019;Schneeberger et al., 2015), may provide insight into the systemic inflammation observed in these animals previously (Salberg et al., 2023).Additionally, Clostridium sensu stricto 1 increases with stress (Zaytsoff et al., 2020), visceral hypersensitivity in irritable bowel disease (Y.-J.Li et al., 2020), and has been correlated with intestinal inflammation (Zou et al., 2020); thus, it is likely this bacteria contributed to increased inflammation in the MS females.Given that the gut microbiome plays a critical role in the immune response, the dysregulation of these pathogenic bacteria could lead to an inappropriate immune response to stressors (Valdes et al., 2018).Further, the production of metabolites, neurotransmitters, and neuromodulators from the gut can interact with neuroimmune system and lead to sensitization and chronic pain (Brenner et al., 2021;Chen et al., 2020).

CONCLUSIONS
Overall, sex, stress, and injury all altered gut microbiome composition, although the effects were generally not cumulative.These results, however, should be interpreted with caution as there may be limited independence of samples for the microbiome analysis.Given that rats were group housed, there may be overlap of microbiome features within cages.This study is also limited by a lack of microbial metabolomics to compliment the 16S rRNA results; however, it does provide preliminary evidence for our hypothesis, with findings supporting sex differences in bacterial composition which may be contributing to pain outcomes.Interestingly, both males and females exhibit behavioral pain sensitivity following injury; however, the mechanisms driving this persistent pain vary between the sexes.Previously, males were shown to have a robust inflammatory response, which was not observed in females (Salberg et al., 2023).Here, we found increases in different bacterial species that may begin to elucidate con-tributing mechanisms for female pain.These novel avenues should be further explored with the potential for developing more targeted and effective treatment strategies.Future studies could implement probiotics or diet as possible therapeutic strategies for chronic pain to compensate for early life adversity.

C O N F L I C T O F I N T E R E S T S T A T E M E N T
The authors declare no conflict of interest.

D A T A AVA I L A B I L I T Y S T A T E M E N T
The data that support the findings of this study are openly available in NCBI BioProject at https://dataview.ncbi.nlm.nih.gov/object/PRJNA940177?reviewer=dp02o1n1b7g8953 mlkotem1k2v., reference number PRJNA940177.Processing pipeline for 16S rRNA data is available at https:// github.com/respiratory-immunology-lab/microbiome-dada2,and data analysis pipelines are available at https://github.com/mucosal-immunology-lab/microbiome-analysis.

F
Experimental timeline for study.

F
Animal characteristics and behavior.Initial three-way aligned rank transformation analyses of variance (ANOVAs) were performed to assess the overall effect, followed by post hoc Wilcoxon rank sum test.(a) Violin plots of average body weight at sacrifice between treatment groups.(b) violin plots of average body weight at sacrifice between sexes.(c) Violin plots of time to right following injury between treatment groups.(d) Violin plots of average filament required to react on the von Frey task between treatment groups.

F
Bacterial diversity and composition.Diversity differences were assessed through Wilcoxon rank sum tests.(a) Bar plots of genus-level composition, grouped by early life stress, injury, and sex.(b) Violin plots of bacterial alpha diversity (Shannon index) within treatment groups, comparing both sex-related and early life stress-related differences.(c) Corresponding plots assessing sex-related and treatment-related differences in alpha diversity within early life stress groups.(d) Principal coordinates analysis of weighted UniFrac distances of amplicon sequence variants (ASVs) grouped by treatment, with ellipses of the 95% confidence intervals for early life stress groups shown.Lactobacillus (p = .0072)genera also decreased.Conversely, Bifidobacterium (p = .02)and Prevotellaceae UCG-001 (p = .034)showed modest increases in abundance (Figure 6a,b).

F I G U R E 4
Effect of animal sex on fecal genera.Linear regression modeling with R package limma was used for initial feature selection, followed by secondary confirmation of significant differences using pairwise Wilcoxon rank sum testing.(a) Box plots of differentially abundant genera between the sexes without early life stress, with the treatment group indicated by dot color.(b) Corresponding heatmap with samples separated by sex.Corresponding treatments are indicated at the top.(c and d) Corresponding box plots and heatmap for differentially abundant genera in rats that experienced maternal separation.et al.

F
Effect of treatment on fecal genera.Linear regression modeling with R package limma was used for initial feature selection, followed by secondary confirmation of significant differences using pairwise Wilcoxon rank sum testing.(a) Box plots of differentially abundant genera between treatment groups without early life stress, with the sex indicated by dot color.(b) Corresponding heatmap with samples separated by treatment group.Animal sex is indicated at the top.(c and d) Corresponding box plots and heatmap for differentially abundant genera in rats that experienced maternal separation.F I G U R E 6 Effect of maternal separation on fecal genera.Linear regression modeling with R package limma was used for initial feature selection, followed by secondary confirmation of significant differences using pairwise Wilcoxon ranke sum testing.(a) Box plots of differentially abundant genera between maternal separation and control groups, with the treatment group indicated by dot color.(b) Corresponding heatmap with samples separated by early life stress group.Animal sex is indicated at the top.

F
Effect of maternal separation on tight junction proteins.Initial three-way aligned-rank transformation analyses of variance (ANOVAs) were performed to assess the overall effect, followed by post hoc Wilcoxon rank sum test.p-Values represent results of group-wise Wilcoxon Rank Sum testing.(a) Violin plots of tight junction protein 1 (TJP1) between maternal separation and control groups.(b) Violin plots of occludin (OCC) between maternal separation and control groups.
The authors would like to acknowledge the GIN Discovery Program, as well as the facilities, scientific, and technical assistance of Micromon Genomics at Monash University.This work was supported by grants from the National Health and Medical Research Council (NHMRC) of Australia to Richelle Mychasiuk (#1173565) and Benjamin Marsland (#1154344); and the Canadian Institute of Health Research (CIHR; RMPJT153051).Open access publishing facilitated by Monash University, as part of the Wiley -Monash University agreement via the Council of Australian University Librarians.
Primer information and cycling parameters for qRT-PCR gene expression analysis.Tm -melting temperature.
T A B L E 1 p = 1.73e-3; however, post hoc Wilcoxon rank sum testing revealed no significance, p > .05(seeFigure2a,b).Results demonstrated a significant treatment effect for time to right, whereby animals in the surgery and mTBI groups took longer to right themselves than sham animals, with surgery animals taking the most time, F(2, 92) = 57.43,p = 2e-16.No other significant effects were observed, p's > .05(see Figure