,,]}1. Introduction

Sickle cell disease (SCD) is a deadly hereditary blood disorder, characterized by sickle-shaped red blood cells, anemia, and multiple organ failure.19,45,60,105 The disease affects a significant proportion (approximately 1 in 500 individuals) of the African American population. Hispanic and some other populations are also affected. Sickle cell disease manifests as a chronic inflammatory state9 with recurrent episodes of excruciating acute pain and chronic ongoing pain, negatively affecting the quality of life and contributing to early mortality in these patients.24,25,55,110 Pain is a lifelong companion of people living with SCD.88 The complex nature of SCD pain has been explored in several studies50,52,58 with considerable progress in understanding the acute pain episodes experienced by patients with SCD96,120; however, marked deficiencies in the knowledge regarding chronic pain mechanisms and the transition from acute to chronic pain in SCD have been realized and contribute to inadequate management of chronic pain in SCD.8,39,76

Chronic inflammation and oxidative stress are central to SCD pathology.81 Studies on chronic inflammatory conditions have found gut microbiota as the crucial regulatory feature in disease pathogenesis.40,42,116,127 The gut–brain axis, a well-established bidirectional network of signaling pathways, undergoes dysregulation resulting in altered blood–brain barrier permeability and neuroinflammation as a consequence of gut dysbiosis.94 The altered microbiota composition has also been implicated in changing the expression of innate pathways such as toll-like receptors and genes underlying nociceptive and pain responses, including the endocannabinoid, opiate, nerve growth factor, and vanilloid receptor pathways.1–3,83,91,93,113

In SCD, the presence of intestinal injury, increased intestinal permeability, and a major dysbiosis in the gut microbiota have been reported.16,37 This is particularly relevant for understanding the pathogenesis of acute pain episodes, wherein gut microbiota–mediated neutrophil activation has been shown to play a crucial role.130 The increased intestinal permeability allows for more translocation of the gut bacteria into the circulation, thus activating an inflammatory cascade.12 Although some investigations into the role of gut microbiota in SCD have been pursued, the impact of gut microbiota dysbiosis on pain is unknown.35,38,107 Investigating the mechanisms through which the gut microbiome–host interactions influence the pain phenotype and underlying pathways would provide a deeper understanding on the role of gut microbiota in SCD pain.

In this study, we aimed to investigate whether gut microbiota–host interactions influence the pain phenotype in sickle cell disease. We used the humanized Townes (TOW) sickle cell transgenic mice123 for 16S rRNA gene amplicon sequencing to provide insights into the compositional differences in the gut microbial communities of TOW mice and the littermate wild-type nonsickle control mice. In addition, we performed antibiotic-mediated gut microbiota depletion and fecal microbiota transplantation (FMT) to reshape the gut microbiota of these mice to examine the effects of microbiota-mediated changes on the sensory and affective pain and comorbidity anxiety-like behaviors in TOW mice and the latter's influence on the wild-type mice by FMT.

2. Methods 2.1. Materials

Ampicillin (BP176025) was procured from Fisher BioReagents (Waltham, MA). Vancomycin, USP Grade (V-200-1) was from Gold Biotechnology (Olivette, MO). Neomycin (N5285) and L-cysteine (168149) were obtained from Sigma (St. Louis, MO). Gabapentin (G117250) was purchased from Toronto Research Chemicals (Toronto, ON).

2.2. Animals

Animal care and experimentation were performed in accordance with the International Association for the Study of Pain (IASP) recommendations and the NIH Guide for the Care and Use of Laboratory Animals and after approval from the Institutional Animal Care and Use Committee at the University of Illinois. Adult age-matched and sex-matched TOW sickle cell transgenic mice and littermate wild-type control mice were used in the study unless mentioned otherwise. The TOW mice were generated by targeted knock-in and replacement of mouse α and β-globin genes with the human α- and human Aγ and βS (sickle) globin genes, resulting in extensive RBC sickling, severe anemia, and multiple organ damage in the homozygous sickle (TOW) mice, similar to the severe sickle cell anemia in humans.47,109 The model has been used in several studies to explore various pain pathways to understand the pain in sickle cell disease.53,61,95 The breeders for TOW mice were purchased from the Jackson Laboratory, and heterozygous mice were bred within our facility to obtain progeny that were homozygous for the human bS alleles (hbS/hbS) and control mice that were homozygous for the human bA allele (hbA/hbA).53 The mice were 3 to 5 months of age for all experiments and were housed in a 12-hour light–dark cycle with ad libitum access to food and water. The mice were divided into groups randomly, and the experimenter was blinded to the genotype and treatment conditions, which were coded for all evoked and nonevoked pain and anxiety behavior experiments. The codes were revealed to the experimenter during data analysis.

2.3. Fecal collection and 16S rRNA gene sequencing analysis

Fresh fecal pellets were collected from genotype-housed mice in sterile tubes, immediately frozen and stored at −80 °C until the DNA extraction. QIAamp Power Fecal Pro DNA kit (Qiagen, Hilden, Germany) was used to extract DNA from each sample. The 16S rRNA sequencing was performed at the High-Throughput Genome Analysis Core of Argonne National Laboratory using Illumina Miseq platform. Paired-end (2 × 150 bp) approach was used for sequencing the V4 region of the 16S rRNA. The raw sequencing reads were analyzed using the Quantitative Insights Into Microbial Ecology 2 (QIIME 2), an open-source microbiome bioinformatics analysis package.13 In brief, the raw reads were denoised using DADA2 for each sample amplicon sequence variants (ASVs). High-quality reads were selected for bioinformatic analysis. Microbial diversity was estimated by evaluating alpha diversity (Shannon, Observed features, and Faith PD indexes) and beta diversity (Bray–Curtis, Jaccard, and UniFrac distance metrics) and visualized by principal coordinate analysis using the specific tools implemented in the QIIME 2 pipeline and R studio (R 4.2.0- ALDEx2,41 qiime2R,11 tidyverse,122 ggpubr,59 ggfortify,106 and ggsignif4 packages). Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2),34 a tool developed to predict the functional potential of a bacterial community based on marker gene sequencing profiles, was used by using the QIIME2 plugin q2-picrust2.34 The ggpicrust2-package126 was used to perform ALDEx2 to identify significantly enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolic pathways and for visualization.

2.4. Antibiotic treatment

The TOW mice were randomly divided into control and treatment groups where mice in the control group received drinking water without antibiotics and the treatment group received drinking water containing ampicillin (1 g/L), neomycin (1 g/L), and vancomycin (0.5 g/L) for 6 weeks.107,130 Drinking water with or without the antibiotics was changed every 3 to 4 days, and the mice were tested for evoked hypersensitivities (von Frey test, Hargreaves test, and cold plantar assay) and nonevoked pain (conditioned placed preference).

2.5. Fecal material transplantation paradigm

The mice were divided into 2 broader groups, donor group and the recipient group. The donor group consisted of both TOW and wild-type littermate control mice. These mice were housed genotype-wise and were used for fecal material collection. Wild-type mice were the recipient group for the first study, and TOW mice were the recipient group for the second study. Study 1 comprised wild-type mice that were divided into 2 groups, control and treatment, wherein the treatment group received FMT from TOW mice (FMTSCD→WT) and the control group received FMT from healthy wild-type mice (FMTWT→WT). For study 2, TOW mice were divided into control and treatment groups with the treatment group receiving FMT from healthy wild-type donors (FMTWT→SCD) and the control group receiving FMT from TOW mice (FMTSCD→SCD).

2.5.1. Donor mice

Fresh fecal pellets were collected daily from the donor mice and pooled, followed by dilution with chilled phosphate-buffered saline (PBS) solution containing 0.05% cysteine (100 mg feces/mL buffer).63,80,90 The samples were homogenized, vortexed for 1 minute, and then centrifuged at 800g for 2 minutes,30,67 and the supernatant was collected. This supernatant was given to the recipient mice (200 μL/mouse) by oral gavage every day for 14 days with 1 day rest in between the weeks (6 days on, 1 day off, 7 days on).14

2.5.2. Recipient mice

The mice in the recipient group were treated with the antibiotic cocktail (0.5 g/L of vancomycin, 1 g/L of neomycin 1 g/L, and 1 g/L of ampicillin) in drinking water for 3 days to decrease the bacterial load.14 To validate the effect of antibiotic treatment, the decrease in the bacterial load was quantified using qPCR.57 After the antibiotic treatment, the mice received 200 μL of donor gut microbiota supernatant by gavage repeated daily for 14 days with 1 day rest in between the weeks (6 days on, 1 day off, 7 days on). During the treatments, body weight and water intake were monitored daily. No adverse effect was observed during the entire experimental period.

2.6. Behavior tests 2.6.1. Mechanical allodynia: von Frey test

The mice were placed in individual plexiglass containers on a wire mesh platform and acclimatized before beginning the experiment.22 The midplantar region of the hind paw was stimulated by a series of calibrated von Frey monofilaments (Stoelting, Wood Dale, IL), by pressing upward to the midplantar region for 5 seconds or until a withdrawal response happened. These filaments apply a fixed force onto the left hind paw, and a clear paw withdrawal is considered as a positive response. The 50% paw withdrawal threshold was calculated using the up-down algorithm.33

2.6.2. Heat hyperalgesia: Hargreaves test

The mice were placed in 10 × 10 × 10 cm plexiglass enclosures with matte light grey walls, front windows, perforated lids on a clear glass surface that was maintained at 30 °C. The mice were allowed to acclimate before the experiment. A radiant heat source (UGO BASILE Model 7372, Stoelting)22,48 was applied to the plantar surface of the left hind paw through the glass, and the latency to the paw withdrawal was recorded. A cutoff time of 20 seconds was imposed to prevent tissue damage.

2.6.3. Cold allodynia: cold plantar assay

The mice were placed in 10 × 10 × 10 cm plexiglass enclosures on a clear glass surface similar to the apparatus used for the Hargreaves test and were allowed to acclimate before the experiment. The dry ice probe was prepared as previously mentioned.15 In brief, the dry ice powder was converted into a pellet using a 3-mL BD syringe and applied to the mouse hind paw by extending the tip of the dry ice pellet past the end of the syringe and pressing into the glass. Light but consistent pressure is applied to the syringe plunger. The withdrawal latency was measured with a stopwatch, and a clear paw withdrawal was considered as a positive response. An interval of 15 minutes was allowed between trials to allow adequate time to return to a resting state after stimulation. Each mouse was measured 3 times. A cutoff time of 20 seconds was imposed to avoid potential tissue damage.

2.6.4. Assessment of ongoing spontaneous pain

The conditioned place preference (CPP) method was used to evaluate the ongoing spontaneous pain with some modifications.51 In brief, the mice were placed in the middle of the CPP apparatus (San Diego Instruments, Place preference systems) and allowed to freely explore the entire chamber for 30 minutes on day 1. A preconditioning bias test was performed by recording the mice movement and duration spent in each chamber in the first 15 minutes. Animals that spent more than 80% or less than 20% of the total time in a particular chamber were removed from the experiment due to preexisting bias. On the conditioning day, the mice received saline (i.p.) and were paired with a randomly chosen chamber for 30 minutes in the morning session. Four hours later, the mice received gabapentin (150 mg/kg, i.p.) and were paired with the opposite chamber for 30 minutes. On the testing day, 20 hours after the afternoon pairing, the mice were placed in the middle chamber of the CPP apparatus with access to all the chambers. The movement and duration of time spent by the mice in each chamber was recorded for 15 minutes for the analysis of chamber preference. The difference scores were calculated as test time – preconditioning time spent in the drug-paired chamber.

2.6.5. Anxiety-like behaviors

The influence of gut microbiota on the pain comorbidity anxiety-like behaviors was evaluated by using the Elevated Plus Maze and the Open Field Test.

2.6.5.1. Elevated Plus Maze

The mice behavior was tested as previously described by Walf and Frye.115 The mice were placed in the center of an elevated plus maze (Stoelting) consisting of 2 open and 2 closed arms opposite each other in a plus-shaped formation. The mice were allowed to freely explore the maze for a period of 5 minutes. The time spent in each arm and the number of arm entries were recorded using ANY-maze (Stoelting).

2.6.5.2. Open field test

The mice were placed in a square box with dimensions 50 × 50 × 50 cm, with the top of the box uncovered. The square box was divided into 16 square fields, and the 4 middle square fields were denoted as the central area, while the rest was considered peripheral area. The time spent in each chamber and the total distance travelled were noted using ANY-maze tracking software. The mice were allowed to move freely, and the movements were recorded over the course of 15-minute sessions.

2.6.6. Statistics

All data are presented as mean ± SEM. For evoked pain behavior data, differences between groups were analyzed using the Student t test (2 groups) or multiple unpaired t test, followed by the Bonferroni–Dunn correction for multiple comparisons. To analyze the CPP data, multiple unpaired t test was used followed by the Bonferroni–Dunn post hoc test. Difference scores were analyzed using paired t test by computing the differences between test time and preconditioning time for each mouse. Data processing and statistical analysis of the raw reads from 16S sequencing was performed using QIIME2 followed by differential abundance analysis using ALDEx2 R package.41 The alpha diversity between the groups was analyzed using the nonparametric Mann–Whitney test. The beta diversity analysis was performed using permutational ANOVA, ie, PERMANOVA, to test the significance between sample clusters observed after the principal coordinate analysis (PCoA); the significance was determined through 999 permutations. For differential abundance analysis and q2-picrust2 metabolic pathway identification, ALDEx2 was used, which uses the Monte Carlo Dirichlet sampling approach, followed by nonparametric t test, returning P values and Benjamini–Hochberg–adjusted P values (q < 0.05). Statistical analyses were performed using GraphPad Prism version 9.4.1 (San Diego, CA) and R (version 4.2.0), with P < 0.05 considered significant.

3. Results 3.1. Difference in the gut microbiota composition of mice with sickle cell disease

The diversity and complexity of the gut microbiota are recognized as essential components of normal physiology, with important roles in homeostasis.114 Dysbiosis in the gut microbiota composition have been shown to influence the pain behavioral response77,108 either through interactions with the immune system or through direct neuronal activation by the bacteria.23

In this study, we first aimed to determine the gut microbiota composition of mice with and without sickle cell disease. Fecal samples were collected from genotype-housed mice and then sequenced to characterize the gut microbial communities present. The alpha diversity was significantly different between the 2 cohorts (Fig. 1A), as quantified by the Shannon diversity index, which was decreased in the mice with sickle cell disease (P = 0.021, Fig. 1A), as well as the decreased observed features (P = 0.01, Fig. 1A) and lower Faith PD index (Fig. 1A) for the TOW mice when compared with the wild-type nonsickle mice.

Figure 1.:

Analysis of alpha diversity, beta diversity, and taxonomic profiles of gut microbiota from the sickle cell mice and littermate wild-type control mice. (A) Alpha diversity as depicted by (i) Shannon entropy, (ii) Observed features, and (iii) Faith PD for WT mice compared with TOW mice with SCD. (B) Principal coordinate analysis plots for beta diversity based on Bray–Curtis, Jaccard, Weighted Unifrac, and Unweighted Unifrac distance matrices for WT group vs SCD group. *P < 0.05, **P < 0.01, ***P < 0.001, n = 15 for WT group, n = 16 SCD group.

With alpha diversity depicting within-group variation, we also looked at beta diversity indices that mainly quantify the between-group similarities or dissimilarities. Analysis of the beta diversity revealed that TOW mice harbor an intestinal microbial composition distinct from the wild-type mice as visualized by the principal coordinate analysis (PcoA), depicting separation between the 2 groups (Fig. 1B). We used several measures to analyze beta diversity including the Bray–Curtis index for compositional data, Jaccard for unique feature identification regardless of abundance, and the UniFrac distance matrices,72 which also take into account the phylogenetic tree information (Fig. 1B). The rarefied ASV tables were used to investigate the differences in the communities between the 2 groups.

The Bray–Curtis matrix suggested that abundant species differ significantly (PERMANOVA P = 0.001, Fig. 1B) among the 2 groups being observed. However, we did not observe a significant difference when the same was tested using weighted UniFrac (PERMANOVA P = 0.07, Fig. 1B), which is a phylogenetically informed metric. We also analyzed the beta diversity using Jaccard and Unweighted UniFrac. Significant differences were observed between TOW and littermate wild-type mice for both Jaccard (PERMANOVA, P = 0.002) and unweighted UniFrac (P = 0.003 Fig. 1B). Using the SILVA-based 16S classifier, the ASVs were mapped to taxonomy where no significant differences at phylum or class level between TOW and the wild-type mice were observed. The relative abundance at family level displayed observable differences in abundance of specific families (Fig. 2A). Comparison of relative abundances of different taxa at genus level was completed using ALDEx2 (Table 1) that showed significantly lower ASVs associated with Turicibacter sp. and significantly higher ASVs associated with Lactobacillus, Odoribacter, and Monoglobus genera in mice with SCD relative to the wild-type mice (Fig. 2B). In addition, we were able to appreciate a visible trend in the abundance of other genera between TOW and wild-type mice (Fig. 2B), such as lower relative abundances of Acetatifactor sp. and Rhodospirillales sp. and higher relative abundances of Lachnoclostridium sp., Muribaculaceae, Clostridia_vadinBB60_group, Oscillibacter sp., and Muribaculum sp. in TOW mice (P < 0.05); however, due to a stringent false discovery rate (FDR), these genera were not statistically significant. The FDR method typically consists of adjusting the individual P values and calls taxa differentially abundant if their adjusted P values fall below a certain threshold,49 which in this case was <0.05.

Figure 2.:

Differential abundance of microbial taxa and inferred metagenome content for WT vs mice with SCD. (A) Relative taxonomic abundance at family level for WT group vs SCD group. (B) Heatmap showing the differential abundance of microbial taxa at the genus level between WT mice and mice with SCD, where rows represent the genera ordered by hierarchical clustering and columns (samples) represent individual mouse sample. The genera that were significantly different, FDR < 0.05 as analyzed with ALDEx2, are labeled in red, and those suggestively different (P < 0.05 but FDR > 0.05) are labeled in blue. (C) Predicted functional differences between WT and SCD microbiota. Of the 78 pathways identified as significantly different, the most relevant pathways are depicted here, with vertical axis of the bar plot representing the KEGG pathway class and names, while the horizontal axis represents the relative abundance of the predicted pathways (center) and log2 fold change values (right). P value (adjusted) represents the P value after the FDR control (ie, Q value < 0.05). FDR, false discovery rate; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Table 1 - Differentially abundant genera evaluated by ALDEx2. Genus Diff. btw Diff.win Effect Overlap we.ep wi.ep we.eBH wi.eBH Odoribacter* −1.128 0.989 −0.9881 0.1716 0.0013 0.0016 0.0299 0.036 Turicibacter* 8.627 6.164 1.1743 0.1227 0.0003 0.0004 0.0104 0.016 Lactobacillus* −2.059 2.445 −0.8795 0.1520 0.0004 0.0007 0.0170 0.026 Monoglobus* −1.755 1.770 −0.9397 0.1394 0.0004 0.0008 0.0152 0.024 Lachnoclostridium −0.953 1.331 −0.6935 0.2195 0.0059 0.0079 0.0832 0.098 Acetatifactor 2.266 4.012 0.5443 0.2393 0.0079 0.0163 0.1014 0.159 Muribaculum −0.947 1.216 −0.7124 0.2114 0.0066 0.0054 0.0878 0.077 Muribaculaceae −0.792 1.312 −0.5352 0.2528 0.0154 0.0251 0.1479 0.189 Clostridia_vadinBB60_group −0.860 1.359 −0.5479 0.2424 0.0412 0.0179 0.2549 0.152 Oscillibacter −0.612 1.145 −0.4865 0.2968 0.0458 0.0744 0.2623 0.304 Rhodospirillales 2.317 3.852 0.5030 0.2601 0.0490 0.0260 0.2466 0.184

The table summarizes the outcomes from ALDEx2 depicting the differentially abundant genera in TOW sickle cell mice in comparison with the littermate wild-type mice (*q value <0.05, significant genera).

diff.btw (median difference in clr values between SCD and WT groups), diff.win (median of the largest difference in clr values within SCD and WT), effect (median effect size: diff.btw/max (diff.win) for all instances), overlap (proportion of effect size that overlaps 0 (ie, no effect), we.ep (expected P value of Welch t test), we.eBH (expected Benjamini–Hochberg–corrected P value of Welch t test), wi.ep (expected P value of Wilcoxon rank test), wi.eBH (expected Benjamini–Hochberg–corrected P value of Wilcoxon test).

To further predict the functional potential of the bacterial communities based on marker gene sequencing profiles, we used q2-picrust2 analysis to explore the microbiome function. Of the 207 KEGG pathways identified, 78 pathways were found to be significantly different between TOW and wild-type mice (q < 0.05). These pathways mainly included pathways related to carbohydrate, lipid, and amino acid metabolism, xenobiotics biodegradation, bile acid synthesis, as well as genetic information processing. The most relevant pathways associated with sickle cell disease and pain are displayed in Figure 2C. We found that TOW mice showed higher abundance of arachidonic acid metabolism, propanoate metabolism, butanoate metabolism, and linoleic acid metabolism pathways, while pathways related to bile acid biosynthesis, amino acid metabolism, and cofactor and vitamin metabolism such as retinol metabolism pathway were more enriched in wild-type mice. Notably, biofilm formation pathway and retinoic acid–inducible gene I (RIG-I)–receptor signaling pathways were more enriched in TOW mice.

3.2. Antibiotic-mediated depletion of gut microbiota attenuates ongoing spontaneous pain, without affecting evoked pain, in sickle cell mice

The TOW mice were treated with a cocktail of broad-spectrum antibiotics for 6 weeks, resulting in efficient depletion and alteration of the gut microbiota (Fig. 3A). No significant changes in the mechanical allodynia (0.05 ± 0.01 g vs 0.07 ± 0.02 g at 6 weeks, Fig. 3B-i), heat hyperalgesia (7.11 ± 0.90 seconds vs 6.83 ± 0.77 seconds at 6 weeks, Fig. 3B-ii), or the cold allodynia (1.55 ± 0.1 seconds vs 1.51 ± 0.063 seconds at 6 weeks, Fig. 3B-iii) were observed between the controls and the mice receiving the antibiotic treatment.

Figure 3.:

Effect of antibiotic-mediated gut microbiota depletion on SCD pain. (A) Schematic depicting the experimental procedures, where mice in Abx-treated group received broad-spectrum antibiotics in drinking water for 6 weeks, while the control group was given drinking water without antibiotics. (B) Time courses for evaluating changes to (i) mechanical allodynia, (ii) heat hyperalgesia, and (iii) cold allodynia. No significant changes were observed after antibiotics for 6 weeks. (C) Gabapentin (150 mg/kg, i.p.) generated CPP in the control group but not in the Abx-treated group. The mice were evaluated for CPP at the end of the 6-week treatment. (D) Difference score analysis (test time - preconditioning time spent in the drug-paired chamber) confirmed the presence of significant difference core in the mice with SCD that received saline, but not in those that received Abx, indicating that antibiotic treatment attenuated ongoing pain in mice with SCD. *P < 0.05, **P < 0.01, ***P < 0.001, n = 8. Abx, antibiotic; CPP, conditioned place preference; SCD, sickle cell disease.

To study spontaneous ongoing pain, we determined gabapentin-induced CPP in these mice. There was no existing chamber bias during preconditioning trials. When subjected to conditioning with gabapentin (150 mg/kg, i.p.), TOW mice without antibiotic treatment spent significantly more time in the gabapentin-paired chamber (490.86 ± 36.00 seconds) than that in the saline-paired chamber (317.25 ± 39.04 seconds, P < 0.05, Fig. 3C), indicative of the presence of ongoing pain in TOW mice. On the contrary, antibiotic-treated TOW mice did not show preference between the gabapentin-paired chamber (393.70 ± 76.39 seconds, Fig. 3C) and saline-paired chamber (443.05 ± 79.28 seconds, P > 0.05, Fig. 3C). When we calculated and analyzed the difference scores, we found a significant difference score in TOW mice without antibiotic treatment (P < 0.05, Fig. 3D). Townes mice that were treated with antibiotics showed no significant difference score, confirming that gut microbiota depletion by antibiotics diminished ongoing spontaneous pain. The effect seems to be selective for affective pain, without affecting evoked pain hypersensitivity to mechanical, heat, and cold stimuli.

3.3. Fecal material transplantation from sickle cell disease donor mice induces pain

To establish the relationship between gut microbiota and pain in sickle cell disease, we used FMT and transplanted fecal microbiota from TOW mice to healthy wild-type mice. We hypothesized that reshaping the gut microbiome of mice using FMT would affect the pain phenotype. To prepare the recipient mice, pretreatment with a cocktail of antibiotics was used for 3 days to deplete the gut microbiota (Fig. 4A). The fecal samples from the mice were collected before and after the antibiotic treatment and the number of 16s DNA copies were quantified using qPCR to quantify the number of bacteria in the gut after treatment (Fig. 4B). To limit the effects of antibiotics on pain phenotype (Fig. 4C), we chose the minimum time and dose necessary to ensure sufficient microbiota depletion before proceeding with FMT.14 Starting day 4, FMT was performed using feces from the donor mice. We determined the nociceptive response of the mice towards evoked mechanical and thermal stimuli. The wild-type mice received FMT from TOW mice (FMTSCD→WT), and the control group received FMT from healthy wild-type donors (FMTWT→WT). FMTSCD→WT significantly lowered paw withdrawal threshold to probing by normally innocuous von Frey filaments after 14 days’ FMT (0.08 ± 0.02 seconds for FMTSCD→WT vs 0.95 ± 0.17 seconds for FMTWT→WT, P < 0.001, Fig. 4D), indicative of the development of mechanical allodynia in wild-type mice that are otherwise healthy except for receiving FMT from SCD donor mice. Similarly, the FMTSCD→WT mice also displayed decreased paw withdrawal latency to noxious heat stimuli applied to the left hind paw (Hargreaves) on the testing day 14 (8.68 ± 1.17 seconds for FMTSCD→WT vs 15.10 ± 0.79 seconds FMTWT→WT, P < 0.01, Fig. 4E). Moreover, FMTSCD→WT-induced cold allodynia after 14 days’ FMT treatment (1.68 ± 0.08 seconds for FMTSCD→WT vs 2.75 ± 0.26 seconds for FMTSCD→WT, P < 0.01, Fig. 4F). The FMTSCD→WT mice exhibited pain phenotypes similar to those of the donor TOW mice.

Figure 4.:

Effect of fecal microbiota transplant (FMT) from sickle cell mice to healthy wild-type mice on pain phenotype. (A) Schematic depicting the experimental procedures, where mice in both groups were treated with broad-spectrum antibiotics in drinking water for 3 days followed by FMT for 14 days with 1 day off in between. The mice in the treatment group received FMT from the mice with SCD (FMTSCD→WT), whereas the mice in the control group received FMT from wild-type mice (FMTWT→WT). (B) Quantification of bacterial load before and after antibiotic treatment (PreABX vs PostABX). (C) (i) Evaluation of mechanical allodynia, (ii) heat hyperalgesia, and (iii) cold allodynia before and after antibiotic treatment (PreABX vs PostABX). No significant change was observed after 3-day antibiotic treatment, n = 5 for each group. (D) Mechanical allodynia was determined on day 7 and day 14 in FMTSCD→WT and FMTWT→WT mice. Paw withdrawal threshold in FMTSCD→WT mice was reduced after 14 days' FMT treatment. (E) Heat hyperalgesia was determined at day 7 and day 14 in FMTSCD→WT and FMTWT→WT groups. Paw withdrawal latency in FMTSCD→WT mice was reduced after 14 days' FMT treatment. (F) Cold allodynia was determined on day 7 and day 14 in FMTSCD→WT and FMTWT→WT groups. Paw withdrawal latency in FMTSCD→WT mice was reduced after 14 days' FMT treatment. **P < 0.01, ***P < 0.001, n = 8 for each group. SCD, sickle cell disease.

3.4. Anxiety-like behaviors in mice receiving fecal material transplantation from sickle cell disease donor mice

To determine whether gut microbiota from TOW mice may be linked with other pain comorbidity behavioral phenotypes, we tested the mice for anxiety-like behaviors. In the EPM test, the time spent in the open arms of the apparatus was reduced in FMTSCD→WT mice (36.15 ± 7.34 vs 76.18 ± 13.18 FMTWT→WT, P < 0.05, Fig. 5A), a sign of anxiety evident in the representative mouse tracks (Fig. 5B). In addition to the EPM test, the Open Field Test (OFT) was also performed, and the FMTSCD→WT mice showed a decrease in the time spent in central compartment indicative of anxiety-like behavior (53.98 ± 19.33 seconds vs 126.4 ± 35.3 seconds FMTWT→WT). This change in behavior is also apparent from the individual mouse tracks (Fig. 5C) in the apparatus where FMTSCD→WT spent more time in the corners of the apparatus. Taken together, FMTSCD→WT mice showed anxiety-like behaviors.

Figure 5.:

Effect of FMT from sickle cell mice to healthy wild-type mice on anxiety-like behaviors. (A) The EPM was used to evaluate anxiety-like behavior in FMTSCD→WT mice and FMTWT→WT mice. The amount of time spent in the open arm of EPM was decreased in FMTSCD→WT mice. *P < 0.05, n = 6 for each group. (B) Representative movement tracks of FMTWT→SCD and FMTSCD→SCD mice in EPM. (C) The Open Field Test was used to evaluate anxiety-like behavior. Movement of each mouse in the activity apparatus was tracked using the ANY-maze software. Reduction in time spent in the center of the chamber is an indication for anxiety-like behavior. SCD, sickle cell disease.

3.5. Attenuation of pain in sickle cell mice after fecal material transplantation from healthy wild-type donor mice

To further ascertain the influence of gut microbiota on the pain phenotype in sickle cell disease, we aimed to discern the effect of changing the gut microbiota of TOW mice. We hypothesized that FMT from healthy wild-type mice to TOW mice (FMTWT→SCD) would alleviate the pain phenotype. Similar to the first experiment, the mice were treated with antibiotics for 3 days followed by FMT starting on day 4. The TOW mice received FMT from healthy wild-type mice (FMTWT→SCD), and the control group received FMT from another group of TOW mice (FMTSCD→SCD). The evoked pain response was