Lipid metabolism and microbial regulation analyses provide insights into the energy-saving strategies of hibernating snakes

Yuting Wei; Huirong Mao; Qiuhong Liu; Wenjie Fang; Tianxiang Zhang; Yongtao Xu; Weiwei Zhang; Biao Chen; Yunlin Zheng; Xiaolong Hu

doi:10.1038/s42003-025-07493-2

. 2025 Jan 12;8:45. doi: 10.1038/s42003-025-07493-2

Lipid metabolism and microbial regulation analyses provide insights into the energy-saving strategies of hibernating snakes

Yuting Wei ^1,^#, Huirong Mao ^1,^#, Qiuhong Liu ¹, Wenjie Fang ¹, Tianxiang Zhang ², Yongtao Xu ^3,⁴, Weiwei Zhang ^3,⁴, Biao Chen ¹, Yunlin Zheng ¹, Xiaolong Hu ^1,^3,^✉

PMCID: PMC11725596 PMID: 39800781

Abstract

Hibernation is a necessary means for animals to maintain survival while coping with low temperatures and food shortages. While most studies have largely focused on mammalian hibernation, its reptilian equivalent has been less studied. In order to provide insights into the energy metabolism and potential microbial regulatory mechanisms in hibernating snakes, the serum, liver, gut content samples were measured by multi-omic methods. Here we show the active snakes have more vigorous lipid metabolism, whereas snakes in hibernation groups have higher sphingolipid metabolism. Furthermore, the results indicate that the potential energy supply pathway was gluconeogenesis. Microbial analysis reveals that Proteobacteria and Firmicutes showed dynamic changes with the transformation among active, pre-hibernation and hibernation periods. The correlation analysis reveals the potential role of Romboutsia, Providencia and Vagococcus in regulating above metabolism by producing certain metabolites. The results advance the understanding of the complex energy-saving strategy in hibernating poikilotherms.

graphic file with name 42003_2025_7493_Figa_HTML.jpg

Subject terms: Animal physiology, Herpetology

Multi-omics approach provides insights into unique energy-saving strategies and the potential mechanisms of gut microbiota regulating sphingolipid and bile acid metabolism in the energy storage and supply in hibernating snakes.

Introduction

Hibernation is the biological process in which animals overwinter in a suspended state as an adaptation to seasonal energy shortages and temperature drops. However, Mammals and poikilotherms (like reptiles and amphibians) hibernate very differently: mammals lower their body temperature and metabolic rate but maintain some activity and respond to environmental cues; poikilotherms essentially match their body temperature to the environment and can be in a suspended state where they do not respond to ref. ¹. Many studies have explored the physiological regulatory mechanisms of mammalian hibernation, and found that available endogenous resources and the innate physiological mechanism determine its survival in hibernation², and the energy supply gradually changes from carbohydrate metabolism to high-energy lipid metabolism after the rapid depletion of glycogen^3,4. That is, lipid metabolism is crucial for mammals that alternately live in active and hibernation seasons⁵, and gut microbiota plays an important regulatory role⁶. Nevertheless, the reptile hibernation is largely understudied, and whether the mechanisms of hibernation are shared between mammals and reptiles needs validation. Though several studies^7,8 underscore the unique role of lipids in the energy supply of snakes during hibernation, the regulatory mechanism of lipid metabolism warrants further elucidation.

In recent years, accumulating studies have demonstrated the important role of the gut microbiota in regulating energy metabolism and proposed the microbial control strategy of hosts’ energy balance⁹, indicating that the gut microbiota is the key factor regulating fat metabolism in animals¹⁰. Further studies have shown that the gut microbiota regulates lipid metabolism by regulating short-chain fatty acid (SCFA) production, fatty acid oxidation¹¹, and bile acid (BA) function¹². Gut microbes utilize a wide range of food matrix and intestinal secretions to produce diverse metabolites involved in host energy metabolism, which directly or indirectly affect lipid metabolism^12,13. Therefore, a comprehensive analysis of the composition, function, and metabolites of the gut microbiota may more accurately reveal its role in the regulatory mechanism of lipid metabolism.

Previous studies on the gut microbiota of snakes focused on the comparison between wild and captive populations¹⁴, the influence of food¹⁵, the description of the structure and function^16,17, the comparison of different parts of the digestive tract^18,19 and the comparison among species²⁰. Changes in the composition and function of gut microbiota could be of great biological significance during hibernation when snakes undergo a long-term fasting process and face the survival challenge of low temperature and hypoxia. At present, little is known about the changes in the gut microbiota in snakes between active and hibernation stages and the regulatory mechanism of the gut microbiota in lipid metabolism.

The Oriental rat snake (Ptyas mucosus) is a type of non-venomous snake that has a broad distribution in the southern regions of the Yangtze River within China and South and Southeast Asia. The rat snakes begin preparing for hibernation around November each year, fully enter hibernation from December to February, and emerge from hibernation in March. The 23–30 °C is the most active temperature range for rat snakes, and some snakes start to refuse food below 22 °C, and snakes stop all activity below 12 °C. The present study used the multi-omics approach combined microbial 16S rRNA gene sequencing, targeted metabolome, and transcriptome to profile serum lipid metabolism indices, liver fat metabolism genes expression, gut microbiota, and metabolites of oriental rat snake, to address the following questions: (1) are there significant differences in the snakes’ lipid metabolism during active and hibernating periods? (2) Are these microbial differences closely related to lipid metabolism? (3) Which microbes play a regulatory role? (4) How do these microbes play the regulatory role?

Results

Serum lipidome

The overall experimental design is shown in Fig. 1A. The serum lipidome was conducted to profile the lipid metabolism among active (Active), pre-hibernation (Pre-Hib), and hibernation (Hib) snakes. Partial least squares discriminant analysis (PLS‐DA) was performed on total lipids of serum and revealed significant differences in lipid profiles among active, pre-hibernation and hibernation groups (p < 0.01; Fig. 1B). The quantitative comparison of conventional serum lipid indicators was performed, including free fatty acids, diacylglycerol (DAG), triacylglycerol (TAG), acylcarnitine (ACar) and BA, which differed significantly with higher levels of these indices observed in active snakes and lower concentrations in hibernation snakes (Kruskal–Wallis, p < 0.05, Fig. 1C). Further, the stress indicator corticosterone showed the highest levels in hibernation snakes (Kruskal–Wallis, p < 0.05, Fig. 1C).

Analysis of individual lipid species demonstrated that 433 species significantly differed between the Active and Hib groups, and 9 of the top 10 differential species were significantly upregulated, with most of these species belonging to subclasses phosphatidylethanolamine (PE) and BA (Fig. 1D). Similarly, 379 species significantly differed between the Active and Pre-Hib groups and 9 of the top 10 differential species were significantly upregulated, with most of these species belong to subclasses PE and BA (Fig. 1E). However, only 66 species were significantly differed between the Pre-Hib and Hib groups (Supplementary Fig. 1A), and 4 of 5 downregulated species belong to phosphatidylcholine (PC) and phytosphingosine, which are both the intermediate metabolites of sphingolipid (SP).

Further analysis of lipid subclasses and categories highlighted above found higher BA metabolism in active snakes (Fig. 2A, B), and higher SP metabolism in hibernation snakes (Fig. 2A and Supplementary Fig. 1B). Furthermore, the analysis of KEGG functions also confirmed the higher BA metabolism in active snakes (Fig. 2C, D). Though the SP metabolism showed no significant difference, it was also at the forefront of hibernating snake metabolic indicators (Supplementary Fig. 1C). Thus, the serum lipidome results could conclude that active snakes showed higher lipid metabolism mediated by BA secretion, and hibernating snakes showed higher SP metabolism.

Liver transcriptome

Secondly, in order to reveal the regulatory pathways of BA secretion and SP metabolism, we conducted a liver metabolome to investigate the key regulatory genes. The principal component analysis (PCA) plots display the overall gene expression of each sample among the three groups. It was found that the overlap of Active vs Hib, and Active vs Pre-Hib was smaller than that of Hib vs Pre-Hib (Supplementary Fig. 2). Then, the volcano plots exhibited that prosaposin (PSAP) gene, which was the key gene of SP degradation was upregulated in hibernation snakes (Fig. 3A and Supplementary Fig. 3A). Interestingly, the volcano plots identified a key gene (fructose-1,6-bisphosphatase, FBP) related to gluconeogenesis was upregulated in pre-hibernating (Fig. 3B) and hibernating snakes (Fig. 3A). However, these volcano plots failed to identify genes related to BA metabolism in active snakes, thus multiple-group volcano analyses were performed. The results identified some genes related to BA metabolism in active snakes, including cytochrome P450 2A (CYP2A), CYP3A5, CYP7B, CYP17A, solute carrier family 27 A6 (SLC27A6), and SLC4A5 (Fig. 3C, D). Further, the multiple-group volcano plots also verified the upregulated PSAP in hibernating snakes (Fig. 3C, D, Supplementary Fig. 3B), and FBP in pre-hibernating (Fig. 3D) and hibernating snakes (Fig. 3C). The differences in relative expression of CYP7B, PSAP and FBP were verified by qRT-PCR (Supplementary Fig. 4). Then, the correlation analysis combined serum lipidome with liver transcriptome highlighted the BA metabolism (Fig. 3E) and SP metabolism (Fig. 3F) when the snakes transformed from the active season to the hibernating season, which was in line with the dot plots of gene function (Supplementary Fig. 5).

Fig. 3 — The volcano plots identify the significantly up/down regulated genes in liver of paired comparisons of hibernation vs active snakes (Hib vs Active, A), and pre-hibernation vs active snakes (Pre-Hib vs Active, B). The upregulated genes closely related to energy metabolism are marked with red names, and the downregulated genes are marked with green names. The multiple-groups volcano plots identify the significantly up/down regulated genes in liver of paired comparisons of hibernation vs active snakes (Hib vs Active, C), and pre-hibernation vs active snakes (Pre-Hib vs Active, D) after the genes are pooled into fatty acids metabolism, bile acids metabolism and carbohydrates metabolism. The correlation analysis of functions of serum lipidome (Met, cycle) and liver transcriptome (Tran, triangle) revealed the highly enriched bile acid metabolism function in active snakes (E), and highly enriched sphingolipid metabolism in hibernating snakes (F). Redder colors indicate more significant correlations, while larger triangles or circles represent more genes or metabolites annotated to that function.

Gut microbiome

Thirdly, the microbiomics was performed to explore the potential role of gut microbiota in regulating the above BA metabolism and SP metabolism. The results of principal coordinate analysis (PCoA, based on Bray–Curtis dissimilarities) showed a different cluster among active, pre-hibernation, and hibernation snakes, and Analysis of similarities (ANOSIM) confirmed these significant differences from the statistical perspective (Supplementary Fig. 6). The α-diversity of the luminal microbiome was significantly higher in hibernating snakes than in active snakes, including the Shannon index and Simpson index (one-way analysis of variance, ANOVA, p < 0.05, Fig. 4A).

Taxonomic analysis at the phylum level revealed that the Proteobacteria (average relative abundance: 60.77%), Firmicutes (average relative abundance: 33.46%), and Bacteroidota (average relative abundance: 5.03%) were the dominant microbial phyla in all the 30 snakes, and the relative abundance of Proteobacteria was significantly lower in active snakes than that in pre-hibernation snakes (Kruskal–Wallis, p = 0.012), whereas the Firmicutes (Kruskal–Wallis, p = 0.023) showed significantly higher abundance in active snakes than in pre-hibernation (Kruskal–Wallis, p = 0.023) and hibernating (Kruskal–Wallis, p = 0.044) snakes, and Bacteroidota (Kruskal–Wallis, p = 0.013) showed significantly higher abundance in hibernating snakes than in pre-hibernation snakes (Fig. 4B and Supplementary Table 1). The linear discriminant analysis effect size (LEfSe) analysis confirmed the above differences in dominant phyla among the three groups (Supplementary Fig. 7).

Among the dominant genera (Fig. 4B), the relative abundance of Providencia was significantly lower in active snakes than in pre-hibernation (Kruskal–Wallis, p = 0.011) and hibernating snakes (Kruskal–Wallis, p = 0.004), and Romboutsia showed a significantly higher abundance in active snakes than in pre-hibernation (Kruskal–Wallis, p = 0.010) and hibernation snakes (Kruskal–Wallis, p = 0.021), and Vagococcus (Kruskal–Wallis, p < 0.001) was more enriched in pre-hibernation snakes (Fig. 4C and Supplementary Table 1). In addition, the random forest was further applied to classify groups and selected the 10 most important bacterial genera ranked by mean decrease accuracy of classification of bacterial communities (Fig. 4D), and confirmed Providencia and Vagococcus were the two most representative genera for the pre-hibernation and hibernating snakes, whilst Romboutsia was the most representative genus for active snakes. Then, correlation analysis revealed that Romboutsia was positively correlated with BA level; Providencia was positively related with carnitine and Gangliosides (GM), while Vagococcus was negatively correlated with Ceramide (Cer) and phytosphingosine (Fig. 4E and Supplementary Fig. 8). In addition, Romboutsia was positively correlated with genes related with BA biosynthesis, and Providencia was positively related with key gene of gluconeogenesis (FBP) and SP degradation (PSAP); and Vagococcus was negatively correlated with most BA genes and PSAP, while positively correlated with FBP (Fig. 4F and Supplementary Fig. 9).

Gut trimethylamine N-oxide (TMAO) determination and N300 metabolomes

Finally, targeted metabolome was used to investigate how Providencia, Vagococcus, and Romboutsia get involved in SP degradation and BA biosynthesis. The concentrations of total BAs, dominant primary BAs (including cholic acid CA and chenodeoxycholic acid CDCA) and numerous secondary BAs (including deoxycholic acid DCA, lithochalic acid LCA, taurodeoxycholic acid TDCA, ursocholic acid UCA, and ursodeoxycholic acid UDCA) were significantly higher in active snakes than in pre-hibernation and hibernation snakes (Figs. 5A and S10). Furthermore, the concentrations of propionic acid, butyric acid, isovaleric acid, ethylmethylacetic acid, and valeric acid were dramatically higher in active snakes (Fig. 5B and Supplementary Table 2). The levels of phenylpropanoic acids (Supplementary Fig. 11A), organic acids (Supplementary Fig. 11B), and fatty acids (Supplementary Fig. 11C) were higher in active snakes. The levels of carnitine and acetylcarnitine, which are the predominant carnitine metabolites, were significantly higher in hibernation snakes than in active and pre-hibernation snakes (Supplementary Fig. 11D). The levels of glucose 6-phosphate and fructose 6-phosphate—the dominant carbohydrate metabolites—were highest in hibernation snakes and lowest in active snakes (Supplementary Fig. 11E). TMAO and creatinine levels were substantially higher in pre-hibernation snakes, whereas betaine and l-carnitine levels were the highest in hibernation snakes (Supplementary Fig. 11F).

Fig. 5 — The comparisons of dominant bile acids (A) and all kinds of SCFAs (B) in the gut of snakes during active (Active), pre-hibernation (Pre-Hib), and hibernation (Hib) period, n = 10 biologically independent samples. *Indicate p < 0.05, **p < 0.01 and more * means more significant differences, and ns means no significant differences. The error bars represent the standard deviation of each group of data.

Correlation analysis revealed that Romboutsia showed a robust correlation with BAs, fatty acids, and partly SCFAs (Fig. 6A), in line with the higher concentrations of BAs, fatty acids, and SCFAs in active snakes. Providencia was closely related to metabolites with higher concentrations in hibernation snakes, including carnitines, carbohydrates, organic acids, and fatty acids, which should be the metabolites degraded from SPs (Fig. 6B). Vagococcus was more abundant in pre-hibernation snakes and were closely related to TMAO, which is an important substance that inhibits fat degradation, and this result might explain the negative relationship between Vagococcus and serum SP lipids (Fig. 4E) and liver genes (Fig. 4F and Fig. 6B). Finally, based on data of liver gene expressions, serum lipid levels and gut metabolite concentrations, three more detailed and focused pathway maps of primary BA biosynthesis (Fig. 7A), SP degradation (Fig. 7B) and glycolysis/gluconeogenesis (Fig. 7C) were constructed.

Fig. 7 — The hand-painted heatmap pathway maps of bile acid biosynthesis (A), sphingolipid degradation (B), gluconeogenesis (C), and their potential roles in the energy metabolism of snakes’ annual life cycle based on the identified genes, serum lipids, and gut metabolites from liver transcriptome, serum lipidome, and gut metabolome. Words in boxes are the names of genes, and the words in ellipses are the names of metabolites, and the heatmaps near them show the differences in the gene expression, or metabolites levels/concentrations among active (AL, AM, AF), pre-hibernation (PL, PM, PF), and hibernation (HL, HM, HF) snakes.

Discussion

Higher BA secretion in active snakes promotes lipid metabolism: The higher levels of serum lipids in active snakes found in the present study were consistent with a previous study²¹, in which active Naja haje haje exhibited higher concentrations of serum total lipids, free fatty acids, and triglycerides. The present study also found higher levels of gut fatty acids, phenylpropanoic acids, and organic acids in active snakes, which are direct or indirect metabolites produced from the host’s digestion or microbes’ fermentation and can affect lipid metabolism²². The highly robust lipid metabolism suggested the digestion, absorption, and accumulation of more fat in the active phase, and this physiological phenomenon has been reported in poikilotherms²³ and endotherms^3,24. Nevertheless, a previous study found that hibernating mammals always showed high lipid degradation to supply energy during hibernation and storage lipids in active season⁶, whereas the present study found lower lipid levels in snakes during pre-hibernation and hibernation. Although no other study has explored serum and gut lipid concentrations in reptiles, a transcriptomic study on alligators found that the fat catabolism pathway was highly suppressed during hibernation²³. This unique energy strategy in snakes may result from the distinct hibernation pattern: Large hibernators, such as bears, can only lower their body temperature by a few degrees during hibernation. In other hibernating mammals, although their body temperatures will approach the environmental temperature, the ability to regulate body temperature is partially retained. Therefore, their energy consumption may be higher than that of hibernating reptiles^24,25. Passive hibernators mainly include some poikilotherms, such as snakes, turtles, crocodiles, and frogs, whose body temperature passively changes with the ambient temperature, and this very low body temperature greatly prevents vast energy consumption^1,26. Another possible reason is that snakes only have white adipose tissue without brown adipose tissue, and only the latter can respond to cold by activating thermogenesis and burning itself to produce heat, initiating the catabolism of white adipose tissue^27,28.

Mounting evidence has demonstrated that BAs play a critical role in regulating energy absorption, lipid metabolism, and cholesterol homeostasis^29,30. CA and CDCA are the main downstream products of cholesterol catabolism in the liver and are conjugated with glycine and taurine to produce GCA, TCA, GCDCA, and TCDCA. These six BAs are classified as primary BAs³¹. Primary BAs are secreted into the intestine through the enterohepatic circulation and transformed into secondary BAs mediated by intestinal bacteria¹². The present study found that active snakes had the highest levels of both serum and gut BAs, which was consistent with the highly robust lipid metabolism; however, cholesterol showed no significant difference among the three groups. BAs not only aid in the digestion and absorption of fat and fat-soluble vitamins as a fat emulsifier^32,33 but also play a key role as signaling molecules in modulating lipid and glucose metabolism³⁴. Multiple genes regulate BA synthesis in the liver, including Farnesoid X Receptor (FXR), CYP, and SLC gene families^35,36. We discovered various upregulated CYP and SLC genes in active snakes. Generally, BAs are produced by the oxidation of cholesterol in liver parenchymal cells mediated by CYP genes, and the synthetic pathways include the neutral pathway mediated by CYP7A1 and the alternative pathway mediated by CYP27A1, as well as several other CYP genes. Thus, CYP genes are the rate-limiting factor of BA synthesis³⁷. SLC27 genes are well known to encode fatty acid transport proteins and can transport BAs³⁸. The multi-level results based on metabolite levels and gene expression fully confirmed the vigorous fat digestion, absorption, and lipid metabolism mediated by BAs in active snakes.

Higher SP degradation in hibernation snakes provides new insight into the energy-saving strategy: Given that lipid metabolites and BAs are present in very low levels, and in the absence of carbohydrates, it is therefore essential to explore the alternative energy source for hibernating snakes to maintain their basic life. Similar to mammalian hibernators, some hibernating ectotherms, such as common lizards (Lacerta vivipara) and geckos (Phyllodactylus marmoratus), rely on fat stores during hibernation^39,40, while some use glycogen, such as the side-spotted lizard (Uta stansburiana)⁴¹. Whether the glycogen can be used as an energy supplier during lengthy hibernation is the question, as the liver and muscle glycogen will be used for several hours to several days after the exhaustion of carbohydrates. The present study did not support the utilization of glycogen in hibernation, because of the limited change in lactic and pyruvate acids, which are the key intermediate of glycogen degradation. A recent study on alligators found that in addition to muscle glycogen, β-oxidation of uncertain lipids may also be a potential source of energy supply during hibernation²³. Another study reported higher levels of phospholipid in the serum of hibernating than active snakes²¹. These clues point to membrane lipids, mainly including phospholipids (including GPs and SM) and SP.

The present study found higher levels of phosphocholine (PC), phytosphingolipids, and several SPs (Cer-NS, Cer-NP, and GM3) in hibernation snakes. Further, the highly upregulated gene PSAP in hibernation snakes confirmed the changes in the above serum indexes from different levels. PSAP encodes a highly conserved lysosomal protein prosaposin, which plays a pivotal role in SP metabolism⁴². SPs are the core components of membrane lipids, including ceramides, SM, and other various forms of glycosphingolipids. SP metabolites, especially ceramides, can enter gluconeogenesis via the tricarboxylic acid (TCA) cycle and act as second messengers to regulate multiple cellular processes involved in inflammation, cell cycle, and metabolism⁴³. In poikilotherms, changes in membrane lipids composition in response to cold and starvation are associated with an increase in unsaturated fatty acid levels⁴⁴, which can also partly be reflected by serum-free fatty acid levels. The degradation of SPs found in the current study provides novel insight into the energy strategy of hibernation snakes.

Higher gluconeogenesis in hibernation snakes helping energy supply: Although we detected the degradation of SP during hibernation, the downstream pathway that drives the conversion of non-sugar substrates into directly usable glucose was not clarified. Gluconeogenesis is the most important process through which to convert non-sugar substances into sugars. The sugars then enter the glycolysis process leading to the production of energy for the body²⁶. In this study, we found two types of evidence indicating the enrichment of gluconeogenesis in hibernating snakes: firstly, the genes involved in gluconeogenesis were upregulated in hibernating snakes, especially FBP. FBP encodes the rate-limiting enzyme of gluconeogenesis^45,46. Secondly, the key substrate (acetate and lactate) and intermediates (fructose 6-phosphate, glucose 6-phosphate) of gluconeogenesis^47,48 were increased in hibernating snakes. Although these metabolites were identified in the intestine, they can be absorbed by intestinal epithelial cells into the bloodstream, with elevated concentrations indicating a dynamic state of gluconeogenesis at the metabolic level⁴⁹. Although the prevailing literature suggests that reptiles can significantly decrease their metabolism during extended fasting^23,26, our study uncovered a potential energy supply pathway during snake hibernation.

Hibernation-induced microbial response and potential role in regulation metabolism: Although the α-diversity exhibited the biggest gap between hibernation and pre-hibernation, analysis of PCoA and phylum revealed a close distance between these two periods and far distance between active with hibernation and pre-hibernation. The predominant phyla were Proteobacteria and Firmicutes, followed by Bacteroidota. The results were consistent with the precious findings on the same species¹⁹. Similarly, other studies on snakes have reported the dominant niche of these three phyla^50,51. Herein, we found higher Proteobacteria and Bacteroidota abundance in hibernating snakes, but the abundance of Firmicutes was decreased. Although Proteobacteria has been recognized as a hallmark feature of dysbiosis in gut microbiota because it contains many opportunistic pathogens, it is closely related to a high-fat diet or obesity^52,53, indicating it plays a role in fat metabolism. The Firmicutes have been extensively known to ferment dietary fiber and indigestible carbohydrates⁵⁴. In contrast, Bacteroidota favors more digestible carbohydrates⁵⁵. Dynamic changes in these three phyla during hibernation have been reported in other animals^56,57, implying the phyla have specific functions in nutrition, immune, or thermogenesis that help animals to thrive. The dynamic changes in microbial genera and species may explain the changes in the phylum. The correlation analysis revealed that Romboutsia was the main genus involved in the BA metabolism of active snakes, and Providencia was mainly involved in SP degradation and gluconeogenesis of hibernating snakes, whereas Vagococcus was significantly related to the inhibition of SP degradation in pre-hibernation snakes.

Further measurements of gut microbial metabolites showed that Romboutsia promoted the metabolism of lipids into long-chain fatty acids by producing secondary BAs, and then fermented to produce SCFAs to provide energy. Romboutsia is a new genus isolated from animal gut and is classified into family Peptostreptococcaceae, which has the capability to transform primary BAs into secondary BAs⁵⁸, and previous studies reported its positive correlation with BAs produced in animals^59,60. Providencia facilitated the degradation of SP through the production of carnitine and then produced fatty acids, organic acids and carbohydrates as host energy. Carnitine is the key medium for fatty acids to enter the mitochondria for oxidative decomposition, known as “fat carrier”. In the process of lipolysis, l-carnitine first promotes the hydrolysis of fatty acids, and then transports the decomposed fatty acids through the blood circulation system to the various cells of the body, especially muscle tissue, for further oxidative decomposition⁶¹. We also found a highly positive correlation in hibernating snakes between Providencia and creatinine, which is an index to indicates the level of muscle metabolism. Vagococcus inhibited SP degradation by producing TMAO. TMAO is produced in the liver from trimethylamine (TMA), which is synthesized from dietary or endogenetic carnitines, betaine, creatinine, and choline under the catalysis of trimethylaminelyase produced by gut microbes⁶². Vagococcus may accelerate the conversion from carnitines into TMAO, resulting in the reduction of carnitines, and then inhibited the SP metabolism⁶³. Briefly, Romboutsia was involved in fat metabolism in active snakes by producing BAs and SCFAs in active snakes, and Providencia promoted SP degradation by producing carnitines in hibernation snakes, whereas Vagococcus inhibited SP degradation by producing TMAO in pre-hibernation snakes.

In this study, we show that snakes showed distinct characteristics of lipid and energy metabolism coping with cold and fasting hibernation, founding the certain microbes and their potential mechanism of BA metabolism and SP degradation. Specially, this study revealed the unique role of sphingolipid metabolism in snake hibernation, rather than relying on fat degradation for energy in mammalian hibernation. These findings advance our understanding of the complex energy-saving strategy in hibernating poikilotherm.

It is important to acknowledge a limitation that exists within this research. The investigation into the role of snake gut microbiomes via fecal microbiota transplantation (FMT) holds potential utility. However, the utilization of mammalian model animals for FMT, which necessitates the colonization of donor microbiota, raises doubts and necessitates further validation. In future studies, the potential of developing a reptilian model organism holds great promise, while the exploration of bacterial isolation and culture techniques should also be duly acknowledged for future research endeavors.

Methods

Ethics statement

We have complied with all relevant ethical regulations for animal use. This study was approved by the Animal Care and Use Committee of Jiangxi Agricultural University.

Animals and experimental design

Oriental rat snakes in this study were bred in specially designed rooms, where bricks and wooden boards were used to create 4–5 layers of compartments as hiding and hibernation spaces, and the rooms were semi-open, allowing the full use of natural light. Thirty male oriental rat snakes (one year old) were randomly divided into three groups: Active group, Pre-hibernation group, and Hibernation group, 10 snakes per group were bred in one room. The Active group was sampled at September 2022, and the average temperature range within one month before sampling in snake house was 23–32 °C, and the food was provided regularly; the Pre-hibernation group was sampled at November 2022, when the average temperature range within one month before sampling in snake house was 15–18 °C, and no food was provided; the Hibernation group was sampled at January 2023, and the temperature within one month before sampling in snake house was controlled at about 8–12 °C, and no food was provided. On extremely hot and cold days, heating equipment and air conditioners were used to control the temperature within the above range. The food for these snakes was the culled newborn chicks bought from a chicken farm, while the snakes were fed frogs and mice two months before the start of the experiment. The experimental design was visualized in Fig. 1A.

Sampling scheme

Firstly, the blood sample was collected from the snake cloaca by using a syringe, and the blood was transferred to centrifuge tubes without anticoagulant, then the serum (about 4 mL) was collected after the blood was centrifuged. Then, the selected snakes were sacrificed, and the liver and intestine were isolated. Two liver samples (about 2 g) were collected from each snake and were used for transcriptome and Quantitative Real-time PCR (qRT-PCR) validation, respectively. Finally, luminal contents were collected from the whole intestinal segments (no less than 1 g) by extrusion of luminal contents into tubes using a sterile spatula. The luminal samples were collected in triplicate, which were used for microbial 16S amplicon sequencing and metabolome, respectively. All the samples were temporarily stored in liquid nitrogen and finally stored in the −80 °C refrigerator after transport to the laboratory. The information on sampled snakes and samples is exhibited in Supplementary Table 3. AB, AL, AF, and AM respectively represent blood, liver, and gut content samples collected from active snakes (F for microbiota analysis and M for metabolite analysis); PB, PL, PF, and PM respectively represent blood, liver, and gut content samples collected from pre-hibernation snakes; HB, HL, HF and HM respectively represent blood, liver, and gut content samples collected from hibernation snakes. Due to the lack of precise indicators to determine whether snakes have entered hibernation, the serum concentrations of glucose, lactate, lactate dehydrogenase, and lipase were measured to verify whether the hibernating snakes are in a state of low metabolism, and the result suggests that the snakes’ direct energy demand at hibernation is very low (Supplementary Table 4).

Serum biochemistry

Serum glucose, lactate, lactate dehydrogenase, and lipase concentrations were measured at a fully automated biochemical analyzer (BK-600, BIOBASE, Jinan, Shandong) by using specific kits (BIOBASE, Jinan, Shandong). Glucose was measured using the glucose kit (hexokinase method), lactate was measured using the lactate kit (enzymatic colorimetric method), lactate dehydrogenase was measured using the lactate dehydrogenase kit (DGKC recommended method), and lipase was measured using the lactate lipase kit (enzymatic colorimetric method). All procedures were carried out according to the official instructions.