Changes in intestinal microbiota across an altitudinal gradient in the lizard Phrynocephalus vlangalii

Abstract High altitude is an important driving force in animal evolution. However, the effect of altitude on gut microbial communities in reptiles has not been examined in detail. Here, we investigated the intestinal microbiota of three populations of the lizard Phrynocephalus vlangalii living at different altitudes using 16S rRNA gene sequencing. Bacteroidetes, Firmicutes, and Proteobacteria were the most abundant phyla. Bacteroides, Odoribacter, and Parabacteroides were the most abundant genera. Significant differences in the intestinal microbiota composition were found among the three populations from different altitudes. The proportions of Verrucomicrobia and Akkermansia decreased, whereas Bacteroides increased significantly with altitude. Greater abundance of Bacteroides at higher altitude led to the fractional increase in the phylum Bacteroides relative to other phyla. Hypoxia may be the main factor that caused intestinal microbiota variation in P. vlangalii along the altitude gradient. Overall, our study suggested that the community composition and structure of intestinal microbiota of the lizard P. vlangalii varied along altitudes, and such differences likely play a certain role in highland adaptation. Our findings warrant a further study that would determine whether ambient and body temperatures play a key role in the modulation of intestinal microbiota in reptiles.

inhabiting lowlands. On the other hand, highland ectotherms also increase their energy production. At low temperature and under hypoxic conditions, highland ectotherms can provide a steady flow of energy for life by increasing their heart rate ), blood oxygen affinity (He, Xiu, Tang, Yue, et al. 2013;Lu et al., 2015), lung and/or heart mass (Han et al., 2016;He, Xiu, Tang, Yue, et al. 2013) as well as by decreasing aerobic respiration .
Gut microbiota, which assists food digestion and absorption, plays a key role in the energy budget of its host (Pi, Gao, & Zhu, 2017;Tremaroli & Backhed, 2012). The intestinal tract mass of highland animals is always directly proportional to the altitude (Han et al., 2016), which suggests that digestive and absorptive functions are under selection. Studies that utilized acute hypoxic exposure indicated that air pressure is an important exogenous factor that strongly modulates the composition of intestinal microbiota (Maity, Adak, Ghosh, Pait, & Mondal, 2013;Maity, Adak, Pathak, Pati, & Mondal, 2012).
Comparative studies in pikas and humans living at different altitudes also confirmed that gut microbiota composition is influenced by altitude (Li & Zhao, 2015;Li, Gesang, et al. 2016).
However, the response of intestinal microbiota during highland adaptation in reptiles has been rarely reported.

| DNA extraction and amplification
Stomach was removed before procedure, and large contents in the intestine were carefully removed to avoid peeling off gut tissue. Afterward, the full intestinal tract was dissected and collected into tubes in sterile conditions. Total genomic DNA extraction was performed using a Fecal DNA Extraction Kit from Sangon (DP328, Sangon, Shanghai, China). Quantification of DNA was carried out using Qubit@ 2.0 Fluorometer (Thermo Scientific, USA). DNA purity was monitored on 1% agarose gels. DNA was diluted to 1 ng/μl using sterile water.
16S rRNA genes were amplified using the specific primer with the barcode. All PCR reactions were carried out in a total volume of 30 μl reaction system, made up with 15 μl of Phusion ® High-Fidelity PCR Master Mix (E0553L, Biolabs, New England), 0.2 μmol/L solutions of forward and reverse primers, and 10 ng of template DNA. Thermal cycling consisted of the initial denaturation at 98°C for 1 min, followed by 30 cycles of denaturation at 98°C for 10 s, annealing at 50°C for 30 s, and elongation at 72°C for 60 s. Finally, the PCR system was held at 72°C for 5 min. PCR products were detected by electrophoresis on a 2% agarose gel and the bands between 400-450 bp were purified with a GeneJET Gel Extraction Kit (Thermo Scientific, USA) for further experiments.

| Library preparation and sequencing
Sequencing libraries were generated using a TruSeq ® DNA PCR-Free Sample Preparation Kit (Illumina, USA) following manufacturer's recommendations and index codes were added. The library quality was assessed using Qubit@ 2.0 Fluorometer (Thermo Scientific, USA) and Agilent Bioanalyzer 2100 system. Finally, the library was sequenced on an Illumina HiSeq 2500 platform and 250 bp paired-end reads were generated.
Then, GreenGene database (GG 13.5) based on the RDP classifier algorithm (v2.2; Wang, Garrity, Tiedje, & Cole, 2007) was used to annotate taxonomic information. OTU abundance information was normalized using a standard sequence number corresponding to the sample with the least number of sequences. Subsequent analyses of alpha diversity and beta diversity were all performed based on this normalized data output.
To evaluate the differences in intestinal microbial composition and structure along an altitude gradient, the principal coordinate analysis (PCoA) was performed based on the weighted UniFrac metric by WGCNA package as well as by stats and ggplot2 packages in R software. To identify differences in microbial communities between the two groups, analysis of similarities (ANOSIM; Clarke, 1993) was performed based on the Bray-Curtis dissimilarity distance matrices by vegan package in R software. The linear discriminant analysis (LDA), as implemented in LEfSe software (Segata et al., 2011), was used to search for the taxon for which the relative abundance was significantly different among the various populations. In addition, the redundancy analysis (RDA) of principal coordinates was used to evaluate the significance of altitude, temperature, or hypoxia using vegan package in R software.
The variations in microbial count, alpha index, and beta index were examined by the Kruskal-Wallis one-way analysis of variance. The Spearman's rank correlation test was used to determine whether the relative abundance of intestinal microbiota was significantly related to environmental factors (temperature and air pressure). Statistical analysis was conducted with SPSS 19.0 for Windows. Values are expressed as the means ± standard deviation (SD). All statistical analyses were conducted with a significance level of α = 0.05 (p < .05).

| RE SULTS
In total, 687,335 high-quality (>Q30) reads were filtered from 713,928 raw reads obtained from 13 intestinal content samples. The total number of OTUs at the 97% similarity level was 9,025 ( Table 2).
The Shannon-Wiener curve of all samples suggested that enough OTUs have been detected as the rarefaction has levelled out, and thus did not limit their interpretation ( Figure S1). Estimators of the Good's coverage, community richness (Chao), and diversity (Shannon index) are summarized in Table 2.

| Intestinal microbiota composition
Overall, representatives of 29 phyla and 373 genera were found in intestinal samples of P. vlangalii. The composition of intestinal microbiota varied among the three populations. In similarity with the data from other vertebrates, the dominant phyla in all populations were Bacteroidetes, Firmicutes, and Proteobacteria ( Figure 3a).
Interestingly, there was no significant difference in the relative abundance of these three main microbial phyla (Kruskal-Wallis ANOVA, p > .051 in all cases), despite Bacteroidetes exhibited a tendency to be more predominant at higher altitude, whereas Firmicutes and Proteobacteria demonstrated an opposite trend ( Figure 3a; Table   S1). Notably, the abundance of Verrucomicrobia and Deferribacteres, which represented a very small proportion of microbiota (1.84% and 0.19% on average across all samples, respectively), decreased drastically with altitude (p = .009 and .022, respectively; Figure S2). Anaerotruncus, and Hungatella ( Figure 3b; Table S1). Among these nine genera, only relative abundance of Akkermansia was significantly affected by altitude (p = .006, Table S1). In other rare (<0.5%) genera, only Oscillospira, Mucispirillum, and Intestinimonas varied significantly among P. vlangalii populations at different altitudes (p = .047, .022, and .038, respectively; Table S1). The relative abundances of these genera were inversely proportional to altitude (Table S1; Figure S3).   All indices were calculated at the 97% similarity level. The number in the parentheses represents the sample size. DLH = Delingha, DL = Dulan, MD = Maduo.

| Diversity of intestinal microbiota in lizards at different altitudes
For beta diversity analysis, we performed weighted UniFrac PCoA and found that there was an obvious separation of MD samples from other two groups. Pco1 and Pco2 accounted for 59.59% and 12.46% of the total variation, respectively ( Figure 6). Analysis of similarities showed that intestinal microbiota was significantly different in lizards from DLH and MD locations (r = .375, p = .023), but was similar both between DLH and DL (r = .043, p = .326) and between DL and MD (r = .239, p = .057).

| Correlation between intestinal microbiota and environmental factors
Redundancy analysis (RDA) showed that intestinal microbiota of MD population of P. vlangalii was mainly affected by air pressure, whereas DLH population of lizards was mainly influenced by elevation ( Figure 7a). RDA also showed that some bacterial genera were associated with certain environmental factors, such as

| D ISCUSS I ON
The complex and diverse microbiota in the gastrointestinal tract plays an important role in maintaining host biological functions, including nutrient metabolism (Sekirov, Russell, Antunes, & Finlay, 2010;Tremaroli & Backhed, 2012). A growing number of studies have indicated that animals inhabiting high-altitude environments possess distinct gastrointestinal microbiota compared to that of representatives of the same species that live at low altitude (Li & Zhao, 2015;Li, Gesang, et al. 2016). In the present study, we also found that P. vlangalii lizards inhabiting different altitudes have relatively different intestinal microbiota and confirmed that geography and climate are significant factors regulating gut microbiota of the reptiles on the Tibetan Plateau.
Similar to the results of other studies in reptiles (Colston, Noonan, & Jackson, 2015;Kohl et al., 2016), we found that Bacteroidetes and in P. vlangalii, which was a different pattern from that reported in previous studies of herbivorous lizards (Hong, Wheeler, Cann, & Mackie, 2011;Kohl et al., 2016). As diet plays a key role in determining the composition of gastrointestinal microbial communities, this difference may be explained by the fact that P. vlangalii is a carnivore (Bao, Zeng, Ma, & Yuan, 1998). Considering that gut Bacteroides spp.
also contribute to the degradation of ingested heterologous polysaccharides in plant material (Patrick, 2015), this difference may also depend on some other unknown factors, such as host phylogeny or geographical location (Lankau, Hong, & Mackie, 2012).
The relative abundance of the phyla Bacteroidetes in the intestines of P. vlangalii lizards tended to be higher with altitude, which was mainly caused by a significantly increasing fraction of the genus Bacteroides. In contrast, the relative abundance of the phyla Firmicutes was inversely proportional to the altitude. Our results are distinctly different from the observations in humans (Li & Zhao, 2015;Li, Gesang, et al. 2016), but are similar to the findings in pikas (Li et al., 2016a). Despite observations in these different species are dissimilar, it is possible that the Firmicutes/Bacteroidetes (F/B) ratio is determined by the diet. For example, the F/B ratio, which has been used as specific index of gut microbiota, is affected by the caloric value of the diet (Murphy et al., 2010). Different F/B ratios in Tibetans and Chinese Han may be explained by distinct dietary habits (Li & Zhao, 2015). Whereas the increase in the proportion of Bacteroidetes and relative decrease in Firmicutes may help pikas to digest plant cellulose more efficiently ), P. vlangalii seems to consume exclusively insects (Bao et al., 1998), and the information about their diet variation is lacking. In addition, soil bacteria and the microbes injected with invertebrate diet have been proved to contribute little to the gut microbiota of lizards (Kohl et al., 2016). Therefore, whether the lower F/B ratio at high altitude, caused by the higher proportion of Bacteroidetes and lower fraction of Firmicutes, was induced by diet variation or by other unknown factors requires a further study.
As it had been observed in other reptiles (Colston & Jackson, 2016;Kohl et al., 2016), produce vitamins (Colston & Jackson, 2016). In addition, many studies indicated that the relative abundance of Akkermansia increases under caloric restriction conditions (Derrien, Belzer, & de Vos, 2016). These variations may imply P. vlangalii living at high altitude had abundant food supply. Actually, studies of insect richness on the Qinghai-Tibet Plateau demonstrated that the numbers of Hymenoptera insects increase as altitude goes up (Liu, Gong, Shi, Yang, & Feng, 2007).
Notably, the microbial community structure and species richness of intestinal microbiota were also affected by altitude gradient. Both Kruskal-Wallis ANOVA and LEfSe analysis indicated that almost all significant variations in proportions were found between the lowest and highest populations of P. vlangalii ( Figure 5; Table S1). PCoA also suggested that the intestinal microbiota community structure was not different between DLH (2,900 m) and DL (3,338 m) sites, or between DL (3,338 m) and MD (4,250 m) sites. These findings indicated that a small altitude gradient may not be sufficient to cause a significant variation in P. vlangalii intestinal microbiota. Nonetheless, it should be noted that studies in pika found a significant variation in gut microbiota along a relative small altitude gradient (between 3,694 m and 4,331 m; . It is slightly counterintuitive, because considering that P. vlangalii is an ectothermic species, its intestinal microbiota should be more sensitive to the environmental variation than that of pika. Thus, further studies are needed to examine such potentially higher sensitivity.
Hyperbaric pressure, which is a typical characteristic of high altitude, has been proved as an important factor that strongly modulates the composition of gut microflora in mammals (Adak, Maity, Ghosh, Pati, & Mondal, 2013;Maity et al., 2012Maity et al., , 2013. Our results also confirmed that hypoxia may affect intestinal microbial composition in P. vlangalii. Hypoxia may be the main factor that caused the increase in the fraction of Bacteroides and decrease in the proportion of Akkermansia along the altitude gradient. Low temperature, which significantly and positively correlates with air pressure, is another typical feature of high altitude. It has been proposed that water temperature may influence gut microbiota of fishes (Neuman et al., 2016). Despite body temperature of lizards highly depends on the ambient temperature, they can maintain a relatively stable body The white circles represent samples from DLH, the black boxes represent samples from DL, and the black triangles represent samples from MD (a). The black boxes represent the intestinal microbiota at genus level and only those genera with relative abundance greater than 0.5% were shown (b) temperature through behavioral thermoregulation (Gvoždík & Castilla, 2001;Méndez-Galeano & Calderón-Espinosa, 2016). Therefore, the effect of low ambient temperature on intestinal microbiota may be buffered by thermoregulation, and our present results may reflect the effect of hypoxia. A further study that would specifically examine the variation in body temperature of P. vlangalii living at different altitudes will contribute to our understanding of this issue.
In conclusion, our study revealed that the community composition and structure of intestinal microbiota of the lizard P. vlangalii analyzed the data. Z.W. and Z.W.Y. drafted the manuscript.

CO N FLI C T O F I NTE R E S T
None declared.