Comparative Analysis of Gut Microbiota and Metabolome of Captive Male Malayan Pangolins with Normal and Abnormal Reproduction

TL;DR

This study found that gut bacteria and metabolites in male Malayan pangolins are linked to their reproductive success in captivity, offering insights for better conservation practices.

Contribution

The study identifies specific gut microbiota and metabolites associated with normal and abnormal reproduction in captive male Malayan pangolins.

Findings

Certain gut bacteria like Absiella and Butyribacter are more abundant in pangolins with normal reproduction.

Metabolites such as gamma-Aminobutyric acid are less present in pangolins with abnormal reproduction.

94 differentially expressed metabolites were identified, including those upregulated in abnormal reproduction.

Abstract

This study investigated the effects of gut microbiota and metabolites on the reproductive performance (normal and abnormal reproduction) of male pangolins under captive conditions. The results showed that certain key microbiota and metabolites do indeed have a significant impact on reproductive performance. These findings provide a reference for the captive breeding of pangolins and even other endangered wild animals and have important implications for their husbandry and management.

Ex-situ conservation and captive breeding are important measures for protecting endangered species. However, captive conditions inhibit reproduction in some wild animals, especially males. Under captive conditions, which differentially expressed microbiota and metabolites significantly influence or are key to reproductive performance? This study aimed to investigate the effects of differentially expressed microbiota and metabolites on reproductive performance, including male Malayan pangolins with normal reproductive behavior (NR, with natural mating behavior) and those with abnormal reproductive behavior (AR, without natural mating behavior). Five male Malayan pangolins with normal reproductive behavior and eight with abnormal reproductive behavior were divided into NR and AR groups. Fresh fecal samples were collected for metagenomic and metabolomic analysis. The results showed that Bacillota, Pseudomonadota, Bacteroidota, Actinomycetota, and Fusobacteriota were the dominant phyla. Escherichia, Lactobacillus, Lactococcus, Limosilactobacillus, and Ligilactobacillus were the dominant genera. The abundance of Absiella, Butyribacter, and Candidatus Scatovivens in the gut of the NR group was significantly higher than in the AR group (p < 0.05), while the abundance of Clostridium, Mycoplasmopsis, and Facklamia in the gut of the AR group was significantly higher than in the NR group. A total of 94 differentially expressed metabolites were identified, with Argininosuccinic acid and Cortol significantly upregulated in the AR group, and gamma-Aminobutyric acid and gamma-Glutamylglutamic acid significantly downregulated. These microbiota and metabolites can be screened as potential key biomarkers that might influence the reproductive performance of male pangolins. These findings provide important insights into the husbandry and management of captive pangolins and other endangered wild animals.

Figures

Full text

1. Introduction

The Malayan pangolin is mainly distributed in Southeast Asia [1,2]. Due to overhunting caused by the increasing demand for its meat as a high-end food and its scales used in traditional medicine, the number of wild pangolins has decreased significantly [3,4]. As the most traded mammal in Asia, it has become one of the most endangered species globally [5]. Ex situ conservation and captive breeding are important measures for the protection of endangered species [6,7]. However, captive pangolins face significant reproductive challenges, with most male Malayan pangolins showing no reproductive intention [8]. This has severely hampered the effectiveness and development of captive conservation projects.

Under captive conditions, pangolins cannot select as many food groups as they do in the wild, and they face chronic stress from captivity [9,10], which may lead to health problems such as gut microbiota dysbiosis and increased abundance of harmful bacteria [9]. We hypothesize that chronic stress in captivity and a mismatched diet are two important factors leading to abnormal reproduction in captive Malayan pangolins. Pangolins have a large home range in the wild [11], and the limited space in captive environments restricts their natural behaviors, triggering chronic stress [12,13]. This chronic stress disrupts the hypothalamic–pituitary–adrenal (HPA) axis balance, impairs immune function, and further perturbs the composition and diversity of gut microbiota [12,13]. Pangolins have a highly specialized diet, primarily feeding on ants and termites in the wild [14], but replicating this natural diet in captivity is extremely difficult. The substitute diets often lack essential nutrients or have an imbalanced nutritional structure, which alters gut microbiota homeostasis [5,13].

These two key factors (chronic stress and dietary mismatch) are intrinsically connected to gut microbiota, forming a cascade that impacts reproductive health. Gut microbiota, often regarded as the “second genome” of animals, is a complex bacterial community crucial for maintaining host homeostasis [9]. It regulates host metabolism, nutrient absorption, and immune response—all of which are closely associated with reproductive function [15]. In mammals, gut microbiota imbalance can impair the development and function of the male reproductive system by altering metabolic pathways or triggering inflammatory responses [15,16]. Previous studies have confirmed significant differences in gut microbiota composition between captive and wild Malayan pangolins [5,13], but the link between these microbiota changes and the high rate of reproductive abnormalities in captive individuals remains unclear, which is a key gap this study aims to address.

Metagenomics can analyze the composition and diversity of microbial communities, annotate the functional characteristics of genes, and study the genetics and function of microbial communities [17]. Metabolomics studies the types, quantities, and changes in endogenous metabolites caused by external stimuli, pathophysiological changes, and gene mutations [18]. Metabolomics identifies altered metabolites, studies the biological processes associated with these altered metabolites, and elucidates their underlying biological mechanisms [19]. To fill the existing research gap and clarify the impact of gut microbiota and metabolites on the reproductive health of male Malayan pangolins in captivity, this study investigated the gut microbiota and metabolite profiles of male individuals with normal and abnormal reproduction. We aimed to explore differences in gut microbiota composition and functional pathways, as well as associated metabolite changes, between the two groups. By identifying key microbial taxa, differential metabolites, and core functional pathways linked to reproduction, we sought to reveal the potential mechanism by which gut microbiota and metabolites regulate male pangolin reproductive health. This study will deepen our understanding of the relationship between microbiota and reproduction in pangolins and provide a scientific basis for optimizing captive management strategies (e.g., improving habitat design, optimizing dietary formulas) and enhancing the effectiveness of ex situ conservation for this endangered species.

2. Materials and Methods

2.1. Study Species, Area and Husbandry

Our study was carried out in the Guangxi Forestry Research Institute (22°33′ N, 108°13′ E), located in Nanning City, Guangxi Zhuang Autonomous Region, China. Since 2013, the institute has been engaged in raising trade-rescued pangolins. The diet consisted mainly of black ants (Polyrhachis vicina) and domestic silk moth (Bombyx mori); water was supplied ad libitum.

In this study, 13 captive male adult Malayan pangolins from the Guangxi Forestry Research Institute Base of Pangolin Breeding (Pangolin Base) were selected as research subjects. The animals are divided into two groups: NR (normal reproduction) group, 5 had normal reproductive behavior, AR (abnormal reproduction) group, another 8 male adult Malayan pangolins did not have normal reproductive behavior (Detailed individual information (Table S1)). Our previous research and experience indicate that Malayan pangolins are non-seasonal breeding animals [20]. Experiments have shown that when male and female pangolins are randomly and irregularly paired, the reproductively normal group exhibits reproductive behavior, mating with different female individuals. Meanwhile, the reproductively abnormal group does not exhibit reproductive behavior, regardless of the female individuals. The study protocol was approved by the Institutional Animal Care Committee of Guangxi Research Base of Pangolin Breeding (approved number: 2023003). The sampling period was from 23 to 29 September 2023.

Pangolins are nocturnal animals. They were kept in indoor cages, each consisting of three areas: activity area (120 cm × 80 cm × 50 cm), insulated wooden winter den (40 cm × 35 cm × 28 cm), and underground summer den (40 cm × 35 cm × 28 cm). Lay a disposable plastic film under the cage and collect feces every 2 h from 6:00 pm to 6:00 am the next day. The samples were stored at −80 °C until further metagenomic and metabolomics analysis (Brief photos of the captive environment (Figure S1).

2.2. DNA Extraction

Total DNA was extracted from feces using the Mag-Bind^®^ Soil DNA Kit (Omega Bio-tek, Norcross, GA, USA). The concentration and purity of extracted DNA were determined using a TBS-380 (Turner BioSystemsInc., Sunnyvale, CA, USA) and a NanoDrop2000 (Thermo Fisher Scientific, Waltham, MA, USA), respectively. DNA extract quality was checked on a 1% agarose gel. DNA samples were stored at −80 °C until subsequent library construction and sequencing were performed.

2.3. Metagenomic Sequencing and Data Analysis

DNA samples were fragmented by Covaris ME220 (Covaris, Woburn, MA, USA), library was constructed using NEXTFLEX^®^ Rapid DNA-Seq (Bioo Scientific, Austin, TX, USA), bridge PCR was amplified, sequencing was performed using Illumina NovaSeq (Illumina, San Diego, CA, USA), and the raw data were preprocessed using Fastp software (version 0.23.0) [21]. Reads were aligned to host DNA sequences using BWA software [22] (version 0.7.9a), and contaminating reads with high alignment similarity were removed. Sequence assembly was performed using MEGAHIT (version 1.1.2) [23] to obtain contigs longer than 500 bp, which were then screened for subsequent analysis. Open reading frames (ORFs) were predicted for spliced contigs using Prodigal software (v1.0) [24], and the sequences were translated into amino acid sequences. The gene sequences predicted from all samples were clustered using CD-HIT [25] (version 4.6.1) (parameters: 90% identity, 90% coverage). The longest gene in each group was taken as the representative sequence to construct a non-redundant gene set.

The amino acid sequences of the non-redundant gene set were compared with the NR, KEGG, and eggNOG databases using Diamond [26] (version 0.8.35) (BLASTP alignment parameter setting expected value e-value is 1 × 10^−5^). (1) Species annotations were obtained and species abundance was calculated; (2) KEGG functions were obtained and the abundance of KO, Pathway, EC, and Module functional categories was calculated; (3) COG functions were obtained and COG abundance was calculated. The corresponding tool for the CAZy database, hmmscan (), was used to compare the amino acid sequences of the non-redundant gene set with the CAZy database (alignment parameter setting expected e-value is 1 × 10^−5^) to obtain carbohydrate-active enzyme annotation information and calculate carbohydrate-active enzyme abundance. The Wilcoxon rank-sum test was used to test the differences in functional abundance data between groups, and the p value was calculated. A p < 0.05 was considered statistically significant, and a p < 0.01 was considered highly statistically significant.

2.4. Metabolomics Analysis

Accurately weigh 50 mg of fecal sample into a 2 mL centrifuge tube, add 600 µL MeOH (Containing 2-Amino-3-(2-chloro-phenyl)-propionic acid (4 ppm), vortex for 30 s. Add steel balls, placed in a tissue grinder for 120 s at 50 Hz. Room temperature ultrasound for 10 min. Centrifuge for 10 min at 12,000 rpm and 4 °C, filter the supernatant by 0.22 μm membrane and transfer into the detection bottle for LC-MS detection. The LC analysis was performed on a Vanquish UHPLC System (Thermo Fisher Scientific, Waltham, MA, USA). Chromatography was carried out with an ACQUITY UPLC^®^ HSS T3 (2.1 × 100 mm, 1.8 µm) (Waters, Milford, MA, USA). The column maintained at 40 °C. The flow rate and injection volume were set at 0.3 mL/min and 2 μL, respectively. For LC-ESI (+)-MS analysis, the mobile phases consisted of (B2) 0.1% formic acid in acetonitrile (v/v) and (A2) 0.1% formic acid in water (v/v). Separation was conducted under the following gradient: 01 min, 8% B2; 18 min, 8%98% B2; 810 min, 98% B2; 1010.1 min, 98%8% B2; 10.112 min, 8% B2. For LC-ESI (-)-MS analysis, the analytes was carried out with (B3) acetonitrile and (A3) ammonium formate (5 mM). Separation was conducted under the following gradient: 01 min, 8% B3; 18 min, 8%98% B3; 810 min, 98% B3; 1010.1 min, 98%8% B3; 10.112 min, 8% B3 [27].

The raw mass spectrometry files were converted to mzXML file format using the MSConvert tool in the Proteowizard software package (v3.0.8789) [28]. Peak detection, peak filtering, and peak alignment were performed using R XCMS software (v3.12.0) package [29]. To discover biomarkers, the relative standard deviation (RSD) of potential characteristic peaks in QC samples, i.e., the coefficient of variation, should not exceed 30%, and characteristic peaks that do not meet the requirements should be deleted. HMDB [30], massbank [31], LipidMaps [32], mzcloud [33], KEGG [34] databases were used for substance identification. With the parameters set to ppm < 30 ppm, qualitative results for metabolites were obtained. The specific principle involves determining the molecular weight of the metabolite based on the mass-to-charge ratio (m/z) of the parent ion in the primary mass spectrum. Molecular formula prediction is then performed using mass deviation (ppm) and adduct ion information, followed by matching against a database. Simultaneously, in the quantitative list, the detected metabolites’ secondary mass spectra are matched against the fragment ion information of each metabolite in the database, enabling secondary identification of the metabolites. FDR (False Discovery Rate) is the value obtained after correcting for false positives, calculated using the BH (Benjamini and Hochberg) method. A higher value indicates a higher probability of false positives. The analytical methods used included principal component analysis (PCA) and partial least squares discriminant analysis (OPLSDA). The overfitting of the model was evaluated by calculating the R^2^ and Q^2^ values of the corresponding PLSDA model (Q^2^ < 0, R^2^Y > 0.4). The variable weights (VIP, projected variable importance) obtained from the OPLS-DA model were used as auxiliary screening criteria. VIP > 1 and p < 0.05 were used as screening criteria for differential metabolites.

2.5. Data Analysis

The Wilcoxon rank-sum test was used with IBM SPSS Statistics 20.0 software to assess the differences in microorganisms and metabolites between the two groups. * indicates a significant difference (p < 0.05), and ** indicates an extremely significant difference (p < 0.01). We used Pearson correlation analysis to analyze the correlation between differentially expressed microorganisms and differentially expressed metabolites, with R > 0.8 and p < 0.05 as the screening criteria.

3. Results

In this study, metagenomics generated 649,570,474 clean reads and 2,090,981 contigs. At the Kingdom level, 98.735% were bacteria (Figure 1A,B). There were no significant differences in the Shannon and Simpson alpha diversity indices between AR and NR (p > 0.05) (Figure 1C,D). At the phylum level, the annotated genes mainly belonged to Bacillota (formerly known as Firmicutes) and Pseudomonadota (formerly known as Proteobacteria), followed by Actinomycetota, Bacteroidota, and Fusobacteriota (Figure 2A). At the genus level, they were mainly distributed in Escherichia, Lactobacillus, Lactococcus, Limosilactobacillus, and Ligilactobacillus (Figure 2B). At the species level, they are mainly distributed in Escherichia coli, Ligilactobacillus salivarius, Limosilactobacillus fermentum (Figure 2C).

Gene annotation results from databases show that in the KEGG level 1 database, the number of genes annotated as metabolism is the largest (Figure 3A), while in the CAZy database, the number of non-redundant genes annotated as glycoside hydrolases (GH) is the largest, and the number of genes annotated as Cellulosome Modules (CM) is the smallest (Figure 3B). In contrast, in the COG database, non-redundant genes are annotated in almost all functional categories (Figure 3C).

Wilcoxon further analyzed the differences in the microbial communities between the two groups. At the genus level, differential abundance analysis revealed that the abundance of Clostridium, Absiella, Butyribacter, and Candidatus Scatovivens in the gut of the NR group was significantly higher than in the AR group (p < 0.05), while the abundance of Mycoplasmopsis, and Facklamia in the gut of the AR group was significantly higher than in the NR group (p < 0.05) (Figure 4A). At the species level, differential abundance analysis revealed that the abundance of Clostridium butyricum, Clostridium saccharobutylicum, and Clostridium cochlearium in the gut of the NR group was significantly higher than in the AR group (p < 0.05) (Figure 4B, Table S2).

Non-targeted metabolic analysis was performed on fecal samples from different groups of male pangolins. Orthogonal partial least squares discriminant analysis (OPLS-DA) showed significant differences between NR and AR. PLSDA analysis showed good separation of fecal samples from the two groups. The R^2^Y (R^2^Y > 0.65) and Q^2^ values of the PLS-DA model were both satisfactory, indicating that the model has high explanatory and predictive power (Figure 5B,D). To more accurately compare the differential metabolites of the two groups of pangolins (Figure 5A,C), we used PLS-DA score plots to show the classification effect of the two groups; both groups of samples showed a bilevel distribution, indicating that the classification effect was significant and the sample clustering was obvious, suggesting that the metabolites of the two groups of pangolins were significantly differentiated. VIP > 1 and p < 0.05 were used as criteria for screening differentially expressed metabolites. A total of 94 differentially expressed metabolites were identified in the AR vs. NR groups (Figure 6A). Metabolic pathway analysis (Figure 6B) revealed the top 5 significantly enriched pathways (p < 0.05), which were associated with Alanine, aspartate, and glutamate metabolism; isoquinoline alkaloid biosynthesis; cell cycle-yeast; mineral absorption; taste transduction; and ABC transporters. Compared with the NR group, 25 metabolites were significantly upregulated in the pangolin gut of the AR group, including Argininosuccinic acid, Glutaric acid, Harman, Cortol, 9alpha-Hydroxyandrosta-1,4-diene-3,17-dione, and Phosphate (Table 1); while 69 metabolites were significantly downregulated, including gamma-Aminobutyric acid, gamma-Glutamylglutamic acid, (+)-Limonene, 6-Phosphonoglucono-D-lactone, Indoxyl sulfate, 6-Hydroxymelatonin, Soyasaponin Ba, Kaempferide, Salidroside, Apigenin 7,4′-dimethyl ether, Zerumbone, cis-Ocimene, Nevirapine, and Pindolol (Table 1).

The metagenomic-metabolomics association analysis results are shown in Figure 7. Dihydro-O-methylsterigmatocystin was significantly positively correlated with Mycoplasmopsis; 7alpha-Hydroxycholesterol was significantly positively correlated with Absiella, Hominicoprocola, Alicyclobacillus, and Oceanispirochaeta; Indoleacetaldehyde was significantly positively correlated with Candidatus_Scubalenecus; Tyramine was significantly positively correlated with Qiania and Candidatus_Faecimonas; Dhurrin was significantly positively correlated with Absiella, Pseudodesulfovibrio, Candidatus_Scatovivens, Alkalicoccobacillus, Candidatus_Scubalenecus, and Candidatus_Faecimonas; Benzyl isothiocyanate was significantly positively correlated with Absiella, Pectinatus, and Alkalicoccobacillus.

4. Discussion

4.1. The Dominant Microbiota in the Two Groups

This study explored the impact of gut microbiota on reproductive abnormalities in male Malayan pangolins. Results showed that besides Bacillota (Firmicutes), Pseudomonadota (Proteobacteria) was the most abundant phylum. Typically, mammals are dominated by Bacillota and Bacteroidota (Bacteroidetes) (for carbohydrate metabolism and SCFA production), with low Pseudomonadota abundance [35]. Myrmecophagous mammals have higher Pseudomonadota to adapt to protein/fat metabolism from ant diets [36], but its excessive abundance in captive pangolins is problematic. Captive individuals had significantly higher Pseudomonadota than wild ones [36], likely due to captive stress and insufficient dietary fiber [9], and its overgrowth indicates gut dysbiosis [37]. Escherichia was the dominant genus, consistent with previous studies including wild pangolins [5,9,10,13,38]. Escherichia coli, the most dominant species here [39], participates in nitrogen cycling and amino acid metabolism [9], and its dominance may reflect weakened immunity from long-term captivity [9]. E. coli produces cytotoxins and LPS, inducing inflammation to damage testicular tissue and sperm function [40,41]. During dysbiosis, LPS activates immune signaling [16,40] and enters the bloodstream to trigger pro-inflammatory cytokines in testes, causing seminiferous tubule damage, testicular necrosis, and reduced sperm production [40,42].

Shigella and Salmonella may result from long-term artificial feeding and chronic stress [9]. Under captivity, diet, environment, and stress make them contributors to pangolin gut dysbiosis [9,13], leading to leaky gut. This allows LPS and metabolites to enter the bloodstream, exacerbating systemic inflammation—linked to erectile dysfunction-related vascular complications—by increasing oxidative stress and reducing NO bioavailability, impairing erectile function [43,44,45]. Streptococcus is more abundant in myrmecophagous than non-myrmecophagous species [46], enriched in erectile dysfunction patients and associated with inflammation. It boosts pro-inflammatory cytokines, damaging NO signaling and vascular function [43,47,48].

Lactobacillus has anti-inflammatory effects [49], and pangolin gut Lactobacillus and Bacteroides are linked to digestion and health [9]. Wild pangolins have more stable gut microbiota with higher functional genera (e.g., Bacteroides, Lactobacillus) adapting to natural dietary carbohydrates and fiber [9,19], and Lactobacillus aids chitin degradation [13,50]. Lactobacillus reuteri maintains testicular volume and serum testosterone; L casei and L. coagulans alleviate CCl_4_-induced testicular toxicity by enhancing antioxidant activity, reducing apoptosis, and improving sperm parameters [50,51]. Lactobacillus-containing probiotics promote testicular interstitial cell expansion and testosterone synthesis [52].

Lactococcus aids chitin digestion [36,39], breaks down lactose to produce lactic acid [9], and acts as an immunoprotective probiotic that improves gut microbiota and inhibits harmful microbes/inflammation [9,53,54,55,56]. Similar to Lactobacillus, it may reduce LPS-induced reproductive inflammation, with metabolites potentially affecting hormone regulation or cellular energy metabolism.

For the genus Limosilactobacillus, Limosilactobacillus fermentum, like Lacticaseibacillus paracasei and Lacticaseibacillus rhamnosus, enhances sperm motility via mitochondria-related signaling [57]. In healthy male dogs, oral probiotics increased Ligilactobacillus salivarius abundance and improved sperm parameters [51]. In infertile men, Ligilactobacillus salivarius PS11610 altered reproductive tract microbiota and raised pregnancy rates [58].

Bifidobacterium (such as Bifidobacterium longum) has immune functions [10,59].

Its supplementation improved testicular morphology, sperm count/motility, antioxidant capacity, and reduced apoptosis in diabetic models [59], while increasing serum testosterone, LH, and FSH [59]. In idiopathic male infertility trials, Bifidobacterium-containing probiotics improved sperm parameters, reduced ROS and DNA damage [60], possibly by regulating blood–testis barrier permeability and FSH levels [61].

4.2. Differential Microbiota Between the Two Groups

The results of this study indicate that the abundance of Mycoplasmopsis, and Facklamia in the AR group was significantly higher than in the NR group, while Clostridium, Absiella, Butyribacter, and Candidatus Scatovivens in the NR group was significantly higher than in the AR group.

Clostridium has complex roles in pangolins: contributing to cellulose metabolism [9,12], acting as chitin-degrading symbionts [62], metabolizing carbohydrates and amino acids [12], synthesizing short-chain fatty acids [46], though some species are pathogenic [5]. For reproduction, Clostridium butyricum restores blood–testis barrier permeability via cell adhesion proteins, regulates this barrier and potential testicular endocrine function [63], and produces butyrate [64]. Additionally, Clostridium coli combined with Lactobacillus acidophilus improved metabolism and gut microbiota in high-fat diet-fed male mice [65].

Absiella is key for alleviating flavonoid-mediated oxidative stress and inflammation [66]. Mycoplasmopsis (formerly Mycoplasma spp. [67]) is associated with male infertility [68]. Facklamia causes respiratory infections in pets [69], and Facklamia hominis links to male reproductive infections like balanitis [70]. Butyrobacter produces butyric acid, which enhances gut development, barrier integrity, and reduces intestinal inflammation [71,72]. Feeding yeast cultures to calves increased Candidatus Scatovivens faecipullorum abundance, improving immunity, reducing inflammation, and enhancing antioxidant status [73].

Differences in metabolites between the two groups Argininosuccinic acid is a direct precursor of arginine. When there is no function of Argininosuccinate lyase (ASL) (or ASL is defective), Argininosuccinic acid accumulates (unable to cleave to generate arginine), resulting in reduced arginine production [74]. We hypothesize that the AR group had significantly higher levels of arginine, which may indicate a problem with the arginine synthesis pathway in the AR group, leading to a large accumulation of arginine. Arginine supplementation improves sperm quality and serum testosterone levels [75,76,77], and generates nitric oxide (NO) and polyamines [74]. NO enhances testicular microcirculation and regulates sexual behavior [78,79], while polyamines affect spermatogenesis, sperm motility, and semen antioxidant status [80,81,82]. Glutaric acid induces oxidative stress by inhibiting antioxidant activity [83], and Harman regulates intestinal inflammation and gut health [84]. Elevated Cortol (a metabolite of cortisol) in the AR group may suppress the hypothalamic–pituitary–gonadal (HPG) axis, impair testicular function, and induce oxidative stress and apoptosis [85,86,87,88]. The increased 9α-hydroxyandrost-1,4-diene-3,17-dione suggests enhanced testosterone degradation [89], and high phosphorus impairs testicular function and sperm quality [90], implying the need to reduce phosphorus in male pangolin feed.

Gamma-Aminobutyric acid (GABA) improves sperm parameters, regulates the HPG axis, enhances gut health and antioxidant capacity, and its high level in the NR group may benefit reproduction [91,92,93,94,95]. Gamma-Glutamylglutamic acid (γ-GluGlu) has multiple biological activities and is upregulated in the testes of treated mice [96,97]. D-limonene impairs sperm quality and testicular structure [98]. Cysteine-S-sulfate, L-cysteine alleviates cisplatin-induced testicular damage [99]. Soybean saponin Ba improves sperm quality in diabetic mice [100]; kaempferide reverses reproductive toxicity by regulating oxidative stress and apoptosis markers [101]; Salidroside improves sperm parameters and reduces testicular oxidative stress [102]; Apigenin 7,4′-dimethyl ether enhances post-thaw sperm vitality [103]. Zerumbone increases sperm count in obese rats [104], and cis-Ocimene has multiple bioactivities [105]. Glycerol 3-phosphate (G3P) accumulation causes male infertility [106], while excessive 6-phosphonoglucono-D-lactone (upstream of NADPH generation) is detrimental to reproduction [107].

Indoxyl sulfate inhibits Nrf2 expression and aggravates oxidative stress [108]. 6-Hydroxymelatonin elevation indicates inhibited melatonin synthesis, and melatonin improves sperm quality and protects testicular function via antioxidant and anti-apoptotic effects [109,110,111,112]. Nevirapine reduces testicular weight, sex hormone levels, and induces histological changes in rats [113], while dovalol decreases rat sexual behavior parameters [114].

Dihydro-O-methylsterigmatocystin is an intermediate in the biosynthesis of sterigmatocystin and aflatoxins [115,116]. Sterigmatocystin, a mycotoxin with toxicity similar to aflatoxin, has genotoxic, carcinogenic, hepatotoxic, nephrotoxic, and cytotoxic effects [117]. 7α-hydroxycholesterol can enhance the expression and secretion of the inflammatory chemomolecule CCL2 [118]. Indoleacetaldehyde (indole-3-acetaldehyde) is an agonist of the aryl hydrocarbon receptor (AhR) [119]; AhR signaling is key for immune regulation, inflammation, and epithelial barrier function, maintaining gut and systemic immune homeostasis [120]. Tyramine produced by gut microbes disrupts the intestinal barrier and increases permeability [121]. Dhurrin, a typical cyanogenic glucoside mainly in sorghum bicolor, is non-toxic in plants but enzymatically decomposes to release hydrocyanic acid with severe effects on mammals [122]. Benzyl isothiocyanate exerts antitumor, anti-inflammatory, and metabolic regulatory effects at appropriate doses, but poses genotoxicity/cellular damage risks at excessive doses or specific conditions; dosage and exposure form are critical to its effects [123].

5. Conclusions

This study found that Pseudomonadota (Proteobacteria), Escherichia, and Escherichia coli are the dominant microbial taxa in male Malayan pangolins. The microbiota may also perform certain biological functions, but they can also pose significant health risks, possibly due to chronic stress from captive environments and inappropriate diets. The abundance of Clostridium, Absiella, Butyribacter, and Candidatus Scatovivens was significantly higher in the NR group than in the AR group, while the abundance of Mycoplasmopsis, and Facklamia was significantly higher in the AR group. Arginine succinate and cortisol were significantly upregulated in the AR group, while γ-aminobutyric acid and γ-glutamylglutamate were significantly downregulated. These microbiota and metabolites may serve as potential key biomarkers significantly affecting the reproductive performance of male pangolins. Given the small sample size, and despite identifying metabolites at the MS level 2, we cannot completely rule out the possibility of metabolite isomers, and the lack of detailed consideration of other confounders (such as stress history, lack of pre-captive data) means these results must be interpreted with caution, and this is a pilot study. Based on the principles of non-invasiveness and minimal disturbance, we did not collect semen samples, nor did we perform combined analysis of microbiota, metabolites, and semen. Even some medications may exist due to environmental pollution or errors in database annotation. Despite these limitations, this study lays the foundation for future research and highlights the need for more in-depth validation studies to further explore the effects of captivity on pangolin reproduction.