This study was approved by the University of Auckland’s Animal Ethics Committee (AEC#1299 and AEC#1969) and experimental procedures were in adherence to the ARRIVE guidelines. Shank3^{ex13–16−/−} mice (B6.129-Shank3^tm2Gfng/J), created by deleting exons 13–16 in the Shank3 gene (Peca et al., 2011), were imported from Jackson Laboratories, Bar Harbor, ME, United States and housed in the Vernon Jansen Unit animal facility at the Faculty of Medical & Health Sciences, University of Auckland. Heterozygous x heterozygous breeding pairs were bred to generate wild type (WT), heterozygous and homozygous knockout (KO) offspring mice. WT and KO were genotyped by PCR as described previously [52]. After weaning (~20 days post-birth), wild-type Shank3B^+/+ (WT) mice and Shank3B^−/− knockout (KO) mice were randomly assigned to one of two dietary experimental groups: a control zinc diet [30 ppm; (Tallman and Taylor, 2003; Vyas and Montgomery, 2016; Fourie et al., 2018)] or a supplemented zinc diet (150 ppm), for 6 weeks (Research Diets, Brunswick, NJ, United States) (D19410B and D06041101, Supplementary Table 1). Diets were identical in composition except for the zinc level. Four experimental groups were examined: (1) WT30 (wildtype control mice fed 30 ppm zinc; n = 46); (2) WT150 (wildtype control mice fed 150 ppm zinc; n = 48); (3) KO30 (Shank3B^−/− mice fed 30 ppm zinc; n = 38); (4) KO150 (Shank3B^−/− mice fed 150 ppm zinc; n = 46). Knockout and wildtype mice of the same sex were housed together in individually ventilated cages (maximum of 6 mice/cage), with a 12:12 h light:dark cycle and food available ad libitum.

After 6 weeks on the control or supplemental zinc diet, faecal samples from 9-week-old WT and Shank3B^−/− KO mice were collected and immediately frozen at −20°C. Mice were euthanised by CO₂ and cervical dislocation, then gut tissue samples (ileum, caecum, colon) were collected into tubes containing RNAlater (Ambion, Inc.).

The Qiagen AllPrep DNA/RNA Mini Kit (Qiagen, United States) was used to extract DNA and RNA according to the manufacturer’s instructions, with minor alterations as follows. Firstly, bead beating was carried out using MP Biomedicals lysing matrix E bead tubes (containing 1.4 mm ceramic spheres, 0.1 mm silica spheres, and one 4 mm glass bead) in the Qiagen TissueLyser II at a frequency of 30 Hz for 2 min. Extracted RNA was treated with DNaseI (Life Technologies, Auckland, New Zealand) and checked for DNA contamination using long cycle PCR targeting the V3-V4 region of the 16S rRNA gene with the following conditions: 95°C for 3 min, 50 cycles consisting of 30 s at 95°C, 30 s at 55°C, and 30 s at 72°C, followed by 72°C for 5 min.

The hypervariable V3-V4 region of bacterial 16S rRNA genes was amplified using primers 341F and 785R (Klindworth et al., 2013) synthesised with an Illumina adaptor for multiplex sequencing (Supplementary Table 2). The fungal Internal Transcribed Spacer 2 (ITS2) genomic region was amplified using the ITS4 and ITS3mix1–5 primers with an Illumina adaptor (Supplementary Table 1) (Hoggard et al., 2018). Paired-end sequencing generating 2 × 300 bp reads was conducted by Auckland Genomics Ltd. (University of Auckland, New Zealand) on an Illumina MiSeq with V3 chemistry.

Extracted RNA of intestinal tissue from a subset of mice (4 treatment groups × 2 gut locations (ileum, colon) × 3 replicates = 24 samples) was sent to Otago Genomics (University of Otago, New Zealand) for ribosomal depletion (Ovation Complete Prokaryotic with mouse AnyDeplete module add-in, Tecan, United States), generation of one equimolar library pool, and paired-end sequencing generating 2 × 125 bp reads, using four lanes of two rapid flow cells on an Illumina HiSeq 2500 with V4 chemistry.

A subset of faecal and ileum samples was selected for shotgun DNA metagenomic sequencing, matching the samples selected for RNA sequencing but with additional ileum samples due to low DNA concentration (3 faecal samples per experimental group; 6 ileum samples for WT150 & KO 30; 7 ileum samples for WT30 & KO150). Rubicon Thruplex DNA nano libraries were prepared to create 550 bp fragments for 2 × 125 bp paired-end sequencing using the Illumina HiSeq 2500 platform with V4 chemistry at Otago Genomics.

Amplicon sequencing of 16S rRNA genes and ITS2 genomic regions was carried out to describe the bacterial and fungal communities, respectively. We observed that there was considerable overlap in bacterial (Figure 1) and fungal (Supplementary Figure S1) community compositions among the four experimental groups. The dominant bacterial phyla across the entire dataset were Verrucomicrobia (35.3 mean ± 7.8% S.D.), Bacteroidetes (31.9 ± 6.2%) and Firmicutes (22.2 ± 8.4%) (Figure 1A). The genus Akkermansia dominated across the entire dataset regardless of genotype or dietary zinc level (35.0 mean ± 8.1% S.D.), with Faecalibaculum (6.8 ± 6.5%), Muribaculum (4.4 ± 3.6%), Escherichia/Shigella (4.3 ± 2.6%) and Dubosiella (4.1 ± 3.8%) also abundant (Figure 1B).

Although the majority of detected fungal ZOTUs (67.3 mean ± 39.9% S.D.) could not be classified beyond kingdom level, the phylum Ascomycota was dominant among those taxa that could be assigned (20.1 ± 31.9%), with other ZOTUs classified to the phyla Basidiomycota (9.6 ± 25.7%) and Zygomycota (3.0 ± 8.1%) (Supplementary Figure S1). At genus level, the fungi Rhodotorula (6.4 ± 22.4%) and Cyberlindnera (6.2 ± 20.0%) were most prevalent among those identified, with nine other genera ranging from 0.01–2.70% relative abundance across the data set (Supplementary Figure S1B).

The lack of distinct microbial (bacterial and fungal) signatures among the four experimental groups, as shown by the taxonomic bar charts, was also evident in the non-metric multidimensional scaling plots for both bacteria and fungi (Figures 2A,B). Nevertheless, there were marked differences among communities from different gut sections, with the fungal communities notably differing along the gastrointestinal tract and even more so between gut sections and faecal samples (Figure 2B). Additionally, bacterial communities associated with ileum samples were overall less similar compared to those from the caecum, colon and faecal samples within an individual mouse, indicating a longitudinal shift in community compositions (Figure 2A).

Bacterial community richness and diversity varied among samples, with 74 to 148 distinct bacterial ASVs observed per sample (Figure 3A). When aggregating samples from all gut locations for a given experimental group, there was significantly higher (p < 0.05) bacterial diversity in the WT compared to the Shank3B^−/− KO mice (Figure 3A), irrespective of alpha-diversity metric and zinc diet. For the fungal ITS2 data, significantly lower diversity—supported by all four alpha-diversity metrics—was observed in the Shank3B^−/− KO mice fed 30 ppm zinc compared to the wild-type mice fed the same amount of zinc (WT30) (Figure 3B), showing that gut fungal diversity was significantly decreased in Shank3B^−/− KO mice. However, this difference in fungal diversity was no longer observed when Shank3B^−/− KO mice were fed the supplemented zinc diet (KO150) and there was no significant difference in alpha diversity compared to WT30 control mice. This suggests that dietary zinc may aid in the increase of fungal diversity to levels observed in WT mice.

In this study, mice of different genotypes were cage housed together and fed one of the treatment diets (control or supplemented zinc), with each cage comprising members of one sex. Based on bacterial 16S rRNA gene amplicon data, individual cages contributed the greatest amount (12.8%) of microbiota compositional variation (p = 0.001; Table 1). Gut sample location, mouse genotype, dietary zinc level and the interaction between genotype and zinc diet were also significant (p ≤ 0.008), explaining 11.4, 3.9, 2.3 and 1.1% of microbiota variation, respectively. A broadly similar trend was observed for the fungal data, whereby dietary zinc levels explained 3.7% of observed variation overall (p = 0.007), while gut sample location accounted for almost one-quarter (23.4%, p = 0.001) of variation (Table 2; Figure 2B). These results reflect substantially greater discrimination of fungal communities based on gut location than was observed for bacteria.

We identified specific microbial taxa that were differentially abundant among the four experimental groups, using Kruskal-Wallis rank-sum and Dunn tests with adjustments for multiple pairwise comparisons. A total of 74 bacterial ASVs differed significantly in abundance between the WT30 controls and the KO30 groups (Supplementary Table 2), with the majority classified as members of the phyla Bacteroidetes and Firmicutes, and the families Muribaculaceae and Lachnospiraceae. Individual ASVs with the lowest p-values, indicative of between-treatment differences (p < 3.78 × 10⁻⁹) were ASVs 182/155/135 (all Akkermansia), ASV 90 (Lachnoclostridium) and ASV 69 (Muribaculaceae). Sixty-five ASVs differed significantly (p < 0.05) in relative abundance between the KO30 and KO150 groups, and 65 ASVs also differed between the WT30 and KO150 groups, including the aforementioned bacterial taxa (Supplementary Table 2). Across all experimental groups, 24 fungal ZOTUs were identified as being differentially abundant (Supplementary Table 3). Only five of these ZOTUs could be classified below kingdom level: Rhizopus ZOTUs 5, 7 and 502, and Aspergillus ZOTUs 11 and 19 (Supplementary Table 3).

Differences in host gene expression were observed between gut sections and treatment groups. Host gene expression patterns correlated strongly with gut section (colon versus ileum). Compared to the colon, ileal samples exhibited lower expression of functional gene groups related to the digestive system, cell motility, and metabolism of amino acids, while genes relating to energy metabolism were typically more highly expressed in the colon (Figure 4). KEGG functional groups of genes related to signal transduction, endocrine and metabolic diseases, and neurodegenerative diseases were also highly expressed in all colon and ileum samples, encompassing ~ 86–89% of all mouse-associated transcripts (KEGG Level 2, Figure 4).

There were differences in KEGG functional groups associated with host gene expression between the WT and Shank3B^−/− KO genotypes as well as between the zinc diets within a genotype, with differences most notable among colon samples. Genes involved in energy metabolism were enriched in the colon of Shank3B^−/− KO mice (on both zinc diets) compared to wild-type mice on the control zinc diet. Additionally, there was decreased expression of immune disease genes on the supplemented zinc diet (150 ppm) in both KO and WT mice compared to the control zinc diet (30 ppm) in the colons of mice with these genotypes (Figure 4).

Across the mouse host-associated data set, a total of 1,387 genes were significantly differentially expressed. Consistent with the clustering shown in Figure 4, ileum and colon samples were clearly separated in two-dimensional space (Figure 5A), with greater variability in the expression of these genes in the colon. There were 1,271 genes among the ileum samples and 741 genes among the colon samples which were significantly differentially expressed. Many of these were related to antimicrobial interactions (alpha-defensins, mucin 2, carbonic anhydrase 1, intelectin) and host metabolism genes that may be regulated by the gut microbiota (e.g., NADH dehydrogenases). By contrast, distribution of samples in the ordination did not correlate with experimental group, which encompasses zinc diet and mouse genotype (Figure 5B).

Therefore, to further investigate differences between experimental groups, linear discriminant analysis of effect size (LEfSe) was used with host and microbiome gene transcript data. This identified 54 genes (biomarkers) that were significantly differentially expressed between the KO30 and KO150 groups (Figure 6A). NADH dehydrogenases, DNA/RNA binding proteins, zinc finger and transmembrane proteins were among the most differentially expressed gene categories. We focused our subsequent analysis on two main categories: alpha defensins, due to their importance in the control of bacteria at the mucosal barrier, and genes related to tight junction proteins, which play an integral role in mediating gastrointestinal permeability (Chelakkot et al., 2018). Overall, alpha defensin genes were more highly expressed in the ilea of Shank3B^−/− (KO) mice compared with the ilea of wild-type mice, with slightly higher expression in the supplemented zinc diet KO mice than without supplementation (Figure 6B). Expression of tight junction genes was significantly and markedly lower in the colon samples from Shank3B^−/− KO mice receiving zinc supplementation compared to their wild-type counterparts on the same diet (p = 0.0139), while the ileum samples showed the opposite trend, and these ilea genes were overall more highly expressed than those in the colon (Figure 6C).

DNA shotgun sequencing data were reconstructed into 124 metagenome-assembled genomes (MAGs). Seven were 100% complete (with 0–0.94% contamination), 104 were 90–99.78% complete (0–9.68% contamination), and 13 were 76.62–89% complete (0–14.75% contamination). Of these, 124 MAGs with ≤5% contamination and at least 70% estimated completeness were analysed further (Figure 7). Overall the majority of the MAGs were assigned to the families Muribaculaceae, Lachnospiraceae and Erysipelatoclostridiaceae (including Faecalibaculum and Clostridium genera), which was largely consistent with the faecal 16S rRNA gene amplicon data. As expected, general positive gut health indicator taxa Akkermansia and Bifidobacteriaceae were also present in the MAGs (Figure 7). MAGs were also obtained for members of the Atopobiaceae, Eggerthellaceae, Borkfalkiaceae, Oscillospiraceae and Anaerotignaceae, despite their not previously being detected in Shank3^−/− mice bacterial amplicon data (Figure 1). The most highly abundant MAGs belonged to collections of Akkermansia muciniphila followed by Paramuribaculum (including P. intestinale) in faecal samples. In the ileum samples, MAG 22 (Muricomes sp900604355) was highly represented in WT30, WT150 and KO150 groups compared to the KO30 group. This same trend was also observed for MAGs representing Akkermansia, Porphyromonas bacterium UBA7173 and Faecalibaculum. Overall, we observed increased abundances of Akkermansia, Porphyromonas and Faecalibaculum in bacterial communities from KO mice on the supplemented zinc diet.

Differences in bacterial community function were determined using KEGG database annotations of MAGs. In terms of functional capacity, the KO150 group on the supplemented zinc diet displayed a greater relative abundance of genes for metabolism, environmental information processing, and cellular processes than other groups. For example, cellular process gene relative abundances differed significantly between KO150 and the groups KO30 (p = 0.037) and WT30 (p = 0.00018), and were highest in the KO150 group. Significant genotype differences were observed relating to two of these categories, environmental processing (p = 0.015) and cellular processes (p = 0.039). Between the dietary zinc treatments, cellular processes (p = 0.00084) and environmental processing (p = 0.0000056) differed most significantly. Cellular process genes differed significantly between KO150 and the groups KO30 (p = 0.037) and WT30 (p = 0.00018), with the same trend for environmental processing genes (KO30 p = 0.000091; WT30 p = 0.000021)—i.e. relative abundances were again highest in the KO150 group. By contrast, the abundance of mucin genes was greater in WT mice compared to KO mice, with increased presence in faecal samples compared to ileum samples within all experimental groups.

While sample location did not significantly affect the distribution of genes involved in carbohydrate metabolism, genotype (p = 0.0046), zinc diet (p = 0.011) and the interaction between genotype and zinc diet (p = 0.021) all significantly affected the number of protein-coding sequences annotated with carbohydrate metabolism in the microbial community. Zinc-related genes, including zinc transporters, metallopeptidases, and zinc-binding cytidine monophosphate, were more abundant in the KO150 group compared to the KO30 group, especially in the ileum. Saturated and unsaturated fatty acid synthesis genes were found in 93 bacterial taxa, with Akkermansia muciniphila, Muribaculum intestinale, and Alistipes finegoldii representing the top three taxa across the entire dataset with these genes. In the ileum samples, normalised count data revealed a lower abundance of fatty acid-related genes in KO30 (total number of genes = 1,346) compared to WT30 (total = 2,371), but an increase in KO150 mice (total = 2,026) towards WT30 levels. This trend was also seen in the faecal samples, WT30 (total = 2,843); KO30 (total = 2,166); KO150 mice (total = 2,368). Interestingly, in the KO30 faecal samples and KO150 ileum samples, A. finegoldii was not among the top three taxa with respect to fatty acid synthesis genes, instead being replaced by Bacteroides salanitronis (identified from KO30 faeces) and Lachnoclostridium sp. YL32 (KO150 ileum sample). Genes related to synthesis of propionate were significantly different between zinc diets (p = 0.029) and the interaction between genotype and zinc diet (p = 0.033), but not between the genotypes; these genes were associated with Escherichia coli K-12 MG1655, Desulfitobacterium dehalogenans and Flavonifractor plautii. In KO30 mice, the main propionate-associated organism was F. plautii, while Escherichia coli K-12 MG1655 was the top bacterium identified that is associated with propionate production and transport in both WT30 and KO150 mice.

When considering the gut-associated microbiota, differences among individual mice were substantial. Even within a given dietary zinc level or genotype group, phylum-and genus-level bacterial community compositions varied greatly, albeit with overall dominance by certain taxa, most notably the genus Akkermansia (phylum Verrucomicrobia). Despite such inter-individual variation, which is consistent with previous autism gut microbiota studies (Bundgaard-Nielsen et al., 2020), our data reveal significant effects of host genotype, dietary zinc level, and their interaction. Specifically, for both bacterial and fungal data, across all alpha-diversity indices WT mice displayed significantly increased diversity compared with Shank3B^−/− KO mice. This trend of decreased fungal and bacterial alpha-diversity in Shank3B^−/− KO mice regardless of dietary zinc treatment suggests that host-associated gut microbes may be highly regulated by ASD-linked genetic variations occurring in key synaptic proteins in the central nervous system which can ultimately affect immunological interplay between host and microbe.

Akkermansia muciniphila is a mucin-degrading gut bacterium that improves gut barrier function by stimulating host mucous production and turnover, ultimately reducing GI tract inflammation via regulation of antimicrobial peptides in the gut stimulating host Toll-like receptors that are integral to host immunity (Garcia-Vello et al., 2024). In our study, Akkermansia was one of the most dominant bacterial taxa for all experimental groups and was also widely represented in the shotgun metagenomic data. Decreases in A. muciniphila have been observed in ASD, irritable bowel syndrome and Crohn’s disease, leading to the development of probiotic treatments with this microbe for these gut disorders (Gu et al., 2021).

The clearest differences related to dietary zinc levels and/or genotype were evident not at the level of entire bacterial communities, but rather at the level of specific ASVs, consistent with other gut microbiome studies (de Theije et al., 2014; Chen et al., 2017). For example, 74 ASVs, mostly from the families Muribaculaceae and Lachnospiraceae, were significantly differentially represented between the WT and KO groups that were both fed 30 ppm zinc. When comparing the two genotypes on the control zinc diet, Muribaculaceae, Blautia and Erysipelotrichaceae were more abundant in WT30 mice than KO30 mice. Blautia produces butyrate, a short-chain fatty acid (SCFA), and is considered beneficial for host health. Butyrate has positive anti-inflammatory effects at the mucosa and fortifies the epithelial barrier, with implications for the gut-microbiota-brain axis as it is able to cross the blood–brain barrier and produce neuromodulatory effects (Canani et al., 2011; Oleskin and Shenderov, 2016; Silva et al., 2020). Other SCFAs such as acetate, propionate and succinate can also be produced by members of the family Muribaculaceae. This family represents one of the largest proportions of gut bacteria found in mice, and has been suggested to play a similar role to the genus Bacteroides in humans, occupying similar niches and metabolising the same dietary source—complex carbohydrates (Ormerod et al., 2016; Smith et al., 2019). It is worth considering the extent to which the gut microbiota may be directly utilising zinc from the diet, potentially in competition with the host animal. Smith et al. (1972) estimated that GI bacteria utilise 20% of the dietary zinc consumed in a germ-free/pathogen-free rat study (Smith et al., 1972). Zinc supplementation of the host diet (150 ppm) potentially results in increased zinc availability along the GI tract for the microbial community, affecting its overall community structure and function. Microbial high-affinity membrane transport systems and zinc-binding proteins for competitive uptake of zinc between gut microbes are crucial for their persistence and growth along the GI tract. Balancing host and gut microbiome requirements for zinc is therefore essential for survival of both symbiotic components.

Among the Shank3B^−/− KO mice, the bacterial families Enterobacteriaceae and Coriobacteriaceae were the top two significant biomarkers that were more abundant in the supplemented zinc diet (KO150) mice compared with the control zinc diet (KO30) mice. KO30 mice, instead, had increased abundances of Muribaculaceae and Ileibacterium. Diet changes can alter host production of bile acids, which can shift the gut community in favour of organisms that can metabolise bile acids such as members of the Coriobacteriaceae (Stenman et al., 2013). This, in turn, can affect gut barrier permeability by interaction of bile acids with the epithelial cells (Raimondi et al., 2008). Similarly, members of Muribaculaceae, Blautia and Faecalibaculum were more abundant in WT30 than KO150 mice, which again had higher abundances of Coriobacteriaceae and Escherichia/Shigella.

While our main focus was on the influence of dietary zinc on the gut microbiota of Shank3B^−/− KO mice, this study also provided an opportunity to revisit the association between host genotype and microbiota composition in this animal model. An earlier study by Tabouy et al. (2018) described reduced bacterial alpha-diversity in Shank3B^−/− KO compared with wild-type mice, as well as differences in the relative abundances of several bacterial genera including Lactobacillus and Prevotella. The authors interpreted these differences as reflective of a dysbiosis, or dysregulation, of the microbiome. We also observed differences of a comparable magnitude between the Shank3B^−/− KO and WT mice microbiotas (on the control 30 ppm zinc diet), specifically a decrease in bacterial diversity in KO mice, but contend that these do not necessarily constitute a dysbiosis. Indeed, there were marked similarities at both phylum and genus levels between the Shank3B^−/− KO and wild-type bacterial communities in our data (Figure 1), as well as in the aggregated genus-level data of Tabouy et al. (2018) (Supplementary Figure 2 in their paper). Moreover, while statistically significant, alpha-diversity in the female mice (which displayed the strongest effect) still only decreased by <10% in Tabouy et al. (2018) data set. Widespread use of the term “dysbiosis” to describe any shift in microbial communities has been criticised (Olesen and Alm, 2016) and we suggest that this term would more appropriately apply to a situation of extreme microbiota depletion, such as seen in Clostridioides difficile infection (Antharam et al., 2013), rather than the more subtle changes evident in both our data and those of Tabouy et al. (2018). Nonetheless, whether or not the microbiota differences observed here and by Tabouy et al. (2018) constitute a dysbiosis per se, the functional implications of reduced gut-associated microbiota diversity in Shank3B^−/− KO mice warrant investigation. It is important to note that genotype and dietary zinc factors explained ~7% of variation in the bacterial biota, compared with almost 13% attributable to individual cage differences. The explanatory power of dietary zinc and genotype was therefore lower, despite striking host behavioural differences associated with zinc supplementation (Fourie et al., 2018), which were not influenced by mouse cage assignment.

While zinc deficiency can be difficult to diagnose in the clinic, plasma serum levels can be influenced by multiple independent factors and use of hair/nail samples have not been conclusive in the literature; therefore proposals for potential microbial biomarkers for zinc deficiency have been suggested (Chen et al., 2021).

Our data also show significant effects of genotype and dietary zinc on fungal diversity. Twenty-four fungal ZOTUs differed significantly in relative abundance across the entire dataset. The five that could be assigned to below kingdom level represented the genera Rhizopus and Aspergillus. We observed that, on the control diet, KO mice displayed significantly lower fungal diversity compared with WT mice, and moreover that dietary zinc treatment in KO mice reverted fungal diversity towards WT levels. Zinc is key for fungi and has been demonstrated to enhance growth of Rhizopus with higher zinc levels (Wegener and Romano, 1963) and Aspergillus (Vicentefranqueira et al., 2019) by stimulating RNA synthesis and transcription. Autism-associated behaviours of a child with ASD were reversed following treatment with antifungals to target Rhizopus, though the efficacy of this treatment for a wider population is yet to be tested (Baker and Shaw, 2020). Mycotoxins and fungi have been postulated to be involved in the pathobiology of ASD due to their high prevalence in the urine and blood of children on the autism spectrum, but this is largely understudied (De Santis et al., 2019; Markova, 2019). Potential targets for antifungal treatment include zinc transporters (due to their ubiquity and importance in fungi) as well as probiotic treatments such as Lactobacillus aimed at adsorbing mycotoxins or altering gut microbial communities (Amich and Calera, 2014; Guo et al., 2014; Vicentefranqueira et al., 2015). Therefore, dietary zinc supplementation may aid in maintaining the balance between a healthy fungal community and its interactions with the Shank3B-deficient host which ultimately may affect their phenotype. Together with our data, these findings highlight fungal diversity and zinc as an important focus for future studies.

Consideration of host–microbe interactions demands attention to not only the microbiota, but also the host itself. We therefore utilised RNA-Seq to investigate patterns of host (mouse) gene expression in the ileum and colon of both WT and Shank3B^−/− KO mice fed either 30 ppm or 150 ppm dietary zinc. As overall gene expression correlated most strongly with gut region (ileum vs colon) rather than experimental group (dietary zinc or genotype) (Figure 5), we focused our attention on those genes that were significantly differentially expressed and/or directly relevant to integrity of the gut epithelial barrier.

Alpha defensins are antimicrobial peptides active against a wide range of microorganisms (De Smet and Contreras, 2005). Consistent with their known prevalence in the small intestine (Takakuwa et al., 2019; Nakamura et al., 2020), we documented higher expression of alpha defensin genes in the ileum compared with colon samples. Expression of several alpha defensins was higher in the Shank3B^−/− KO mice compared to the wild-type mice, especially those on the supplemented zinc diet. Whilst speculative, this could indicate enhanced antimicrobial activity in Shank3B^−/− KO mice, and that this is further enhanced by high dietary zinc. This suggests that the gut microbiome may be more strictly controlled in Shank3B^−/− KO mice compared to wild-type mice by various host–microbe interactions.

Given the significant research focus on potential leaky gut in ASD, we also focused our analysis on the expression of tight junction proteins between WT and Shank3B^−/− KO mice fed control or supplemented zinc levels. Tight junction proteins are crucial for integrity of the gut epithelium, preventing potentially harmful microorganisms or their metabolites from leaking out of the gut and into the bloodstream. Zinc supplementation enhances and maintains the intestinal barrier by regulating expression of several tight junction proteins (Miyoshi et al., 2016). Tight junction expression in the ileum samples from Shank3B^−/− KO mice was higher than in the corresponding WT30 samples, which may suggest the presence of intestinal inflammation or compromised barrier integrity that can ultimately impact the gut-microbiota-brain axis (Berkes et al., 2003). Additionally, Shank3B^−/− KO mice can have abnormal GI morphology, with alterations of the gut microbial community and increased bacterial LPS levels in the liver (Sauer et al., 2019); this suggests the importance of the SHANK3 protein in the gut and the consequences of KO SHANK3 involvement in disruption of the mucosal barrier. Dietary zinc supplementation correlated with higher tight junction gene expression, regardless of host genotype, particularly in the ileum. This is intuitive given the greater absorbance of dietary zinc in the small intestine (Maret, 2000; Dufner-Beattie et al., 2003; Liuzzi et al., 2004). Overall, our observations suggest that zinc may be playing a role in reinforcement of the gut epithelial barrier that may help to tightly regulate the gut-microbiota-brain axis and perhaps ameliorate the resulting ASD behaviours.

We sampled the mouse gut at several regions along the GI tract, namely the ileum, caecum and colon, as well as faecal samples. For bacteria, there was no clear separation of microbiota composition related to gut region in the ordination plot (Figure 3A), although for many individual mice the ileum samples were somewhat distinct. There was much greater separation for fungi, with the fungal composition of faecal samples being markedly different from that of other sample types (Figure 2B) and greater variation explained in the PERMANOVA models compared to the bacterial data. The bacterial community was substantially more diverse than fungal communities across the dataset (Figure 3).

Fatty acids produced from microbial fermentation/metabolism, such as butyrate and propionate, modulate gut barrier function and provide an anti-inflammatory effect via inhibition of host immune responses, both of which have been linked to the gut-microbiome-brain axis. In particular, Roseburia and Faecalibaculum were present across samples and sample groups, and encoded genes for fatty acid biosynthesis, with the greatest relative abundance of Faecalibaculum in the KO150 group. However, the main fatty acid producers found in this dataset were Akkermansia muciniphila, Muribaculum intestinale and Alistipes finegoldii. Therefore, important fatty acid functional genes of the gut microbiome may be expressed by different members of the bacterial community. This suggests that bacterial function rather than identity could be a more important contributor to the reversal of ASD-related behaviours observed in this mouse model.

Overall, a potential reciprocal relationship, with A. muciniphila degrading mucin and supplying butyrate to the host, may contribute to the observed effects of supplementary dietary zinc, in concert with direct zinc interactions alongside SHANK3 proteins in brain synapses subsequently supporting the ASD behavioural reversal in Shank3B^−/mice. Understanding crosstalk between a host and its gut microbiota is challenging due to the multitude of communication mechanisms across the gut-microbiota-brain axis, however elucidating specific beneficial functions could aid with GI disorders commonly associated with ASD via probiotic supplementation (Zhang et al., 2022).