Shift in Microbial Community Structure with temperature in Deulajhari Hot Spring Cluster, Odisha, india

Hot springs are the reservoirs of novel hyper thermophilic and often mesophilic bacteria that provide information about the prevailing community structure. Here we analyzed the metagenome profile based 16S rRNA amplicon sequencing of the three different springs from Deulajhari hot spring cluster, S1, S2 and S3, having a range of temperature (43°C to 65°C), pH (7.14 - 8.10) and variation in N, P, K, tOC, Salinity, COD and tDS. these thermal spring clusters are covered with the dense vegetation of Pandanus and continuously enriched by plant leaf debris, thus resulting in a high amount of total organic carbon (tOC). the number of phyla varied among the springs: 20 in S1 (43°C), 18 in S2 (55°C) and 24 in S3 (65°C) from the 16S rRNA data. Out of the reported phyla in each spring, the most abundant were Chloroflexi, Proteobacteria, Chlorobi and Acidobacteria, which correlated with the temperature gradient. Various metabolic pathways such as ABC transporters, two-component system, Purine metabolism were most abundantly present in the S2 sample compared to the other two. the CCA analysis revealed the correlation between physiochemical parameters and their functional annotation. the present study establishes the relation between the physiological parameters and the structural distribution of microbiota along the temperature gradient.


iNtRODuCtiON
Microorganisms play a vital role in various biogeochemical cycles 1 still there exists a scarcity of information about the considerate distribution of microbiota and their abundance with the variation in ecophysiological characteristics (geographic distance, temperature, pH and nutrients). 2,3 There has been an influence of the ecophysiological factors in governing the community structure. There have been various reports which link these ecophysiological parameters with the structural distribution of the microbiota. [3][4][5] Environmental heterogeneity (pH, temperature, salinity, C/N ratio and geographic distance) strongly drives the microbial taxa-area relationship. [6][7][8] Among the divergent environmental factors, the temperature is governing aspect that specifies the abundance and distribution of the microflora. In the previous studies, the inverse relationship between temperature and microbial diversity has been recognized, 5 which stipulates the effect of temperature on the composition of the ecosystem. Bison pool hot spring study unravels the predominance of chemotrophic and phototrophic bacteria at 92°C and 56°C respectively. 9 Thermophilic springs have been attracting researchers for their geological history, biogeography and physicochemical variation, as these provide a suitable habitat for thermophilic microbiota. 10 Diverse studies of the thermophilic microflora in various geothermal sites have provided detailed information on the thermophilic microbial diversity and shown the path to access the bioresources that have impending relevance in biotechnological applications. 11,12 Odisha, located in the eastern side of India, is bestowed with various geothermal springs, Atri, Taptapani, Tarabalo and Deulajhari, which have not been thoroughly investigated yet. Through the cultivation-based approach, few reports revealed the prevalence of thiooxydans, Chelatococcus sambhunathii, Thiomonas bhubaneswaensis and Gulbenkiania indica. 13,14 Deulajhari hot spring (Latitude 20.74 N, longitude 84.49 E) is located in the Angul district of Odisha. It has a varied temperature range (43°C to 65°C), thus providing the prospects of discovering the assorted microflora. Earlier studies over several decades on these thermophilic springs based on the cultivation-dependent approach could not present comprehensive information about the entire ecology. 10,15 Metagenomics, since the last few years 6,8,10 has revealed the 99% of the microflora which were earlier hard to detect through the cultivable approach. Next-generation sequencing (NGS) has enabled the study of the complete microbiome of the thermophilic springs and led to new biotechnological applications. 5 Despite the diverse temperature range and varied physiochemical parameters, significant work has not been reported from the Deulajhari hot spring except a recent report on community profiling of one cluster based on the pyrosequencing approach. 16 In this study, we examined the structural community of the three different springs of the Deulajhari hot spring cluster spread within a circle of 100 meters with varying temperatures. Thus, the differential microbial diversity and their relationship with the prevailing physiochemical conditions can be parceled out. Therefore, we aim to explore the abundance of the bacterial phylotypes along the 22°C temperature gradient within the Deulajhari hot spring cluster through the metagenomic 16S rRNA amplicon sequencing using the Illumina platform. This approach explains the role of temperature and the other physiochemical parameters in shaping the structural microbial ecology of the Deulajhari hot spring, which remains hidden to date.

Sample Collection and Analysis
Sediment and water samples were collected from the three different springs (S1, S2, and S3) of the Deulajhari hot spring cluster (Latitude 20.74 N, longitude 84.49 E) located in Odisha, India ( Figure 1). The three springs (S1, S2, and S3) were selected from the widely spread Deulajhari hot spring cluster. Four sediment and water samples from each site were collected and pooled together to make it one representative sample of that particular site. The water temperature of the three springs was recorded in situ using the Enviro-safe thermometer (Sigma, USA) and the pH of both the water and sediment sample was measured using the portable pH meter (Sigma, USA). All sediment and water samples were collected in sterile plastic bags, sealed and transported on ice (4°C) to the laboratory and stored at −80°C for further investigation. Various physiochemical parameters of the sediment samples such as electro conductivity (EC), total organic carbon (TOC), nitrogen (N), phosphorus (P), potassium (K), chloride (Cl-), chemical oxygen demand (COD) and ammonia (NH 3 ) was estimated as described by Jackson (1962) at Department of Soil science, Orissa University of Agriculture and Technology, Bhubaneswar, Odisha. Physiochemical parameters of the water such as salinity, oxidationreduction potential (ORP), dissolved oxygen (DO), total dissolved solutes (TDS), nitrate (NO 3-) and ammonia (NH 3 ) carried out at the department of chemistry, Institute of Technical Education and Research (ITER), Siksha 'O' Anusandhan (Deemed to be University), Bhubaneswar, Odisha, India.

DNA extraction from the Hot Spring Sediment
Hot spring sediment samples (5 gm) from each spring S1, S2 and S3 were mixed with 10 ml lysis buffer [100 mM Tris -HCl (pH 8.0), 10mM sodium EDTA (pH 8.0) and 100 µl of lysozyme (100mg/ml)] by horizontal shaking at 200 rpm for overnight at 37°C. First, 100 µl of proteinase K (10 mg/ml) was added after the overnight incubation and incubated for one hour (hr) with gentle endto-end inversions every 15 min. Next, the samples were incubated at 37°C for one hr with gentle endto-end inversions every 15 min after adding 1 ml of SDS (10% w/v) to each. After completing one hr incubation, samples were again incubated at 90°C for one hr with gentle end-to-end inversion every 15 min. The extraction was prepared with an equal volume of phenol: chloroform: isoamyl alcohol (25:24:1, v/v/v). After centrifugation at 10,000 rpm for 10 min at 25°C, the aqueous layer was collected cautiously. Precipitation of the aqueous phase was done with 0.3 M sodium acetate and two volumes of chilled ethanol at -20°C for one hr. Then the centrifugation was carried out at 10,000 rpm for 10 min at 25°C; the pellet obtained was washed with chilled 70% ethanol and resuspended in sterile deionized water.

library Construction, Sequencing and Data Generation
Amplification of the V3-V4 region o f t h e 1 6 S r R N A u s i n g p r i m e rs 3 4 1 F, 5′-CCTACGGGAGGCAGCAG-3′ and 518R, 5′-ATTACCGCGGCTGCTGG-3′ carried out using the 50 ng of DNA (Bartram et al. 2011). PCR conditions were an initial denaturation at 98°C for 9 min, followed by 20 cycles at 95°C for 30 sec, 60°C for 30 sec, 72°C for 30 sec and ended with a final extension step at 72°C for 5 min. Min Elute gel extraction kit (Qiagen, Germany) was used to purify the PCR amplified product. The amplified V3-V4 amplicons (100 ng) as a template were cleaned using AMPure XP beads (Beckman Coulter, CA, USA) to filtred out nonspecific fragments. 17 The cleaned amplicons were further subjected to library preparation using NEBNext Ultra DNA Library Prep Kit for Illumin's GAIIx sequencer (New England Biolab, UK). The prepared libraries were finally subjected to the 2x150 bp paired-end multiplex sequencing using Illumina GAIIx sequencer at Genotypic Technology Pvt. Ltd (Bangalore, India). The raw sequence reads of each sample were further quality checked using the FastQC v0.11.3. 18 The adapter and low-quality ends from threads were trimmed using Trimmomatic v0.32 (https://github.com/ usadellab/Trimmomatic).

Bacterial Diversity and Functional Prediction
Bacterial census data derived from nextgeneration sequencing of the three different springs S1, S2, S3 for 16S amplicon analysis using the QIIME2 pipeline. 19 Briefly, the fastq-Join tool converted the overlapping paired-end reads into a consensus sequence of the V3-V4 region. Chimeric sequences were filtered out using the uchimedenovo with default settings. UCLUST algorithm was used for denovo OTU picking, taxonomic assignment and phylogenetic at sequence similarity of ≥97%. Finally, an average of 300 OTUs was identified at the species level. The taxonomicto-phenotypic mapping was implemented using METAGENassist (http://www.metagenassist.ca/ METAGENassist/) to predict seventeen potential energy sources at varying abundance. The functional capabilities was predicted by PICRUSt (Phylogenetic investigation of communities by reconstruction of unobserved states) 20 and Tax4fun software 21 based on GreenGenes v13.8 and SilvaSSU v123 databases respectively.

Statistical Analysis
The non-parametric richness estimates, Dominance (D), Chao1, Shannon-Weaver diversity indices (H) and Simpson index was determined by PAleontological STatistics (PAST) v3.02. 22 The multivariate Canonical Component Analysis (CCA) method in PAST was used to elucidate the relationship between physiochemical parameters and taxonomic abundance as well as predicted functional capabilities. Further, to assess the community richness of the samples, the rarefaction curve was constructed using the same software.

Differential Physiochemical Characteristics of Springs of Deulajhari
Deulajhari hot spring cluster being the hub of variant springs with variable thermal conditions harbors the diverse microbial community (Figure  1). The pH of the three springs, S1, S2 and S3, was in the range of 7.14-8.10. These springs have a variant temperature range, highest in S3 (65°C) followed by S2 (55°C) and S1 (43°C), thus presenting the temperature gradient of 22°C among the Deulajhari hot spring cluster ( Figure  1). The sediment analysis of the three different springs, S1, S2 and S3 showed variation in several parameters viz; electroconductivity, total organic carbon (TOC), nitrogen, phosphorus, potassium, chloride, chemical oxygen demand (COD), and ammonia. While the aqueous concentration of the phosphate, fluoride, dissolved oxygen (DO), total dissolved solutes (TDS), nitrate and ammonia also varied in the three springs of the Deulajhari hot spring cluster (Table 1). On the whole, the spring S3 had the highest concentration of ORP (194.6 m/V), temperature (65°C) and potassium (0.692 mg/l), while the spring S2 have the highest concentration of ammonia (568.1 mg/l), chloride (0.229 mg/l) and COD (29440 mg/l), whereas the spring S1 showed the dominance in the concentration of nitrogen (187.5 mg/l), TOC (3.80 H), nitrate (312 mg/l), pH (7.95), TDS (581 mg/l) and fluoride (4.06 mg/l) (Supplementary Table S1). Thus, due to the possible difference in the physiochemical parameters, these three springs of the Deulajhari cluster finds the probability of diverse ecology.
Although all three springs of the Deulajhari cluster shares several taxonomic groups both at the phylum and the class level, there is a significant difference in overall phylotypic composition and diversity (Table 2). Furthermore, the microbial diversity of the three springs at different taxa levels are found to be diverse irrespective of the phylum level. The above significant diversity in the phylotypic composition may be due to the three spring's differential physiochemical variables within one geographical location. Thus, the Illumina sequencing gave us a brief outlook over the taxonomic composition of these three diverse springs, S1, S2 and S3, which shows some taxonomic diversity concerning the temperature and other physiochemical variations. In addition, this study observed that the reservoirs of diverse novel phylotypes persist in these springs of Deulajhari hot spring, which are yet to be explored by the scientific community.

effect of environmental Variables in the Composition of Microbial Community Structure
Environmental factors play a leading role in governing the microbial structure of hot springs. The three springs of the Deulajhari cluster exhibit the differential temperature viz. 43°C to 65°C, thus developing the temperature gradient of 22°C. The differential environmental gradients are responsible for the divergent, phenotypic characteristics. 23 From the sequencing analysis, it revealed that the taxonomic richness of the microbiota in spring S1 (43°C) was more (20 phyla and 40 class) as compared to the remaining two springs S2 (55°C) with 18 phyla and 32 class and spring S3 (65°C) with 24 phyla and 46 class. Thus, this observation confirms that with the increase in the habitat temperature, the current microbiota's taxonomic richness and structural diversity decreases. 5,16,23 The taxa distribution in all the springs concerning variation in the temperature is depicted in Figure 2A. From the Figure , it is clear that there is a drift in each taxon's distribution with the deviation in temperature. Proteobacteria was found to be dominant in the spring S1 and shows the decrease in the population with the increase in temperature (S1> S2 > S3, i.e. 43°C > 55°C > 65°C with the phyla abundance of 0.27, 0.19 and 0.15 respectively). Besides the differential temperature, carbon metabolism may also be the second key factor governing their abundance. 10,16 Thus the specific distribution of these taxa indicates the influence of the geochemical variables in the different springs of the Deulajhari cluster. Based on the CCA analysis ( Figure 2C), the significant taxa present across the samples and their association with the fifteen different physiochemical parameters are studied. The CCA analysis ( Figure 2C) depicts the abundance of microflora Aquificae, OS-K and Clostridia highly correlated with potassium and ammonia. Similarly, the microbial diversity representing the Brachyspirae, Gammaproteobacteria, Chthonomonadetes and Chloroflexi seems to be strongly impacted by the abundance of floride, TDS, phosphorus and pH ( Figure 2C). While the microbial diversity representing Chlorobia, Ignavibacteria, Spirochaetes, Deinococci, OPB56 and OM190 seems to be influenced by the abundance of the chloride, COD, DO, ORP and temperature. The aqueous concentration of TOC, nitrate, potassium and ammonia seem to affect the population of Planctomycetia ( Figure 2C).

Alpha Diversity Study of the Deulajhari Hot spring
Various diversity indices viz, Dominance (D), Chao I, Shannon and Simpson are calculated for the three springs S1, S2 and S3 (Table 2). From the statistical approach, it predicts that the spring S2 has the most dominant diversity as compared to S1 (second most diverse) and S3 (most minor various); the above data approved by the abundance of the observed species, which appears to be highest in the spring S2 (858) followed by S1 (739) and S3 (346) respectively ( Table 2). Out of The Chao I, Shannon and Simpson's indices calculated ( Table 2) the Shannon and Simpson indices were observed to be positively correlated with each other (the alpha diversity of both indexes appears to be S3 > S1 > S2) in the prediction of the alpha diversity of the Deulajhari hot spring cluster.

Predicted Functional Analysis of the S1, S2 and S3 Deulajhari Clusters
PICRUSt and Tax4Fun software were used to predict the functional analysis of the three springs of the Deulajhari hot spring cluster. Depending upon the 16S rRNA gene frequency, the functional profiling of the three springs showed that cellular processes were abundant in spring S3 (4.9), followed by S1 (4.19) and S2 (4.15). The membrane transport was the most dominant among the various cellular processes with the abundance level of 15% in S2, 14.2% in S1 and 13.09% in S3. Among the ABC transport were highly abundant in S2 (7.61%), followed by S1 (6.94%) and S3 (5.70%) and two-component systems were highly represented in S3 (6.6%), followed by S1 (6.32%), and S3 (6.14%). Other membrane transports like bacterial secretion system, phosphotransferase system (PTS) and secretion system also reported to a shallow extent. Signal transduction among the cellular process was observed to be 2.45% (S1), 2.40% (S3) and 2.14% (S2). Among the signal transduction, various signaling pathways were reported in which the most abundant one was two-component systems (2.29% in S1, 2.24% in S3 and 1.99% in S2). The other pathways, like the calcium signaling pathway, mTOR signaling pathway, were reportedly absent in all three springs ( Figure 3A).
Thus from the functional analysis, it was observed that various vital pathways and genes involved in overall metabolism in the three springs of Deulajhari hot spring were found to be similar, but their profusion expressing the functionality with respect to individual subsystem were found to be different probably because of the variation in the eco-physiochemical characteristics existing among these springs, signifying the occurrence of unique metabolism in between the ecological niche of the three other springs of the Deulajhari hot spring cluster thus encouraging discrepancy in the phylotypic abundance. It helps us in predicting the functional analysis of the metagenomics data of the Deulajhari hot spring cluster. The most common subsystem in all the levels seemed to support resolving the complexity in the prevailing metabolic system in their ecological niche. The PICRUSt mediated interpretation showed the key pathways responsible for the overall functional metabolism. Although the difference in abundance at all three levels 1, 2, 3 may be the outcome of differential ecophysiology of the three springs S1, S2, and S3 of the Deulajhari. This differential parameter supports the distinct metabolism in the ecophysiological niche of all the three springs, S1, S2, and S3, supporting the diverse microbial community composition and structure ( Figure 3A).

Phenotypic enrichment Study from the Springs (S1, S2 And S3) of the Deulajhari Hot Spring Cluster
The METAGENassist web ser ver provided a detailed study concerning phenotypic enrichment 24 of the three springs S1, S2 and S3. Phylotypic symbiotic and free-living abundance data deduced from the three springs led to the prediction of the highly abundant free-living microbial diversity compared to the symbiotic abundance ( Figure 4C). Among the three springs (S1, S2 and S3), the free-living is abundant in the spring S3 followed by S2 and S1; similarly, the symbiotic abundance is high in the order S3 > S2 > S1 ( Figure 4C). Differential energy sources of the microbial community in all the three springs, S1, S2 and S3, predict that autotrophs are equally present in S3 and S2 with a slight decrease in S1 ( Figure 4A). Whereas organotrophs seem to be dominant in S1, followed by S2 and S3 respectively. Phototrophs shows the almost similar abundance with the autotroph's population (i.e., S3 > S2 > S1). Chemoorgantotrophs and methylotrophs are highly present in the spring S3, followed by S1 and S2, respectively. Heterotrophs seem to be highest in S2 followed by S3 and S1, respectively ( Figure 4A). Lithoautotrophs were highest in S3 and absent in S2 and very few were present in S1. Chemoheterotrophs are highly abundant in the spring S3, while chemolithotrophs are highly abundant in S1 followed by S2 and absent in S3. Oligotrophs, lithoheterotrophs, diazotrophs, lithotrophs, methanotrophs, photosynthetic and photoheterotrophs present with the variation in the abundance ( Figure 4A). At the different temperature levels viz. S3 (65°C), S2 (55°C) and S1 (43°C) diverse microbial communities like mesophilic, thermophilic, psychrophilic and hyperthermophilic are present. Spring S3 showed the abundance of hyperthermophilic, thermophilic and mesophilic microflora. Pscryophilic diversity of the microbiota appears to be absent from the spring S3. Spring S2 showed the second most richness of the mesophilic and thermophilic microbiota of the spring S3. Psychrophilic microflora reported to be highest in spring S2 ( Figure 4D). Spring S1 has the lowest abundance of mesophilic and thermophilic microbial diversity. Some psychrophilic microflorae were also reported in the spring S1 ( Figure 4D).
taxonomic Correlation between the Bacteriodetes and Chlorobi among the three Springs S1, S2 and S3 Various studies suggest that the phylogenetically cultured representative of the phyla Chlorobi shows the common root with the Bacteriodetes. [25][26][27] In our study, the differential population of Chlorobi was also positively correlated with the abundance of Bacteriodetes in all three springs, which supports the above studies. The phyla Bacteriodetes and Chlorobi are also observed to be linked with the temperature ( Figure  4B). The population of Chlorobi is highest in spring S2 (55°C) with an abundance of 10.82% followed by 9.52% in S3 (65°C) and observed to be lowest in the S1 (43°C) with the population frequency of 8.32%. Bacteriodetes, on the other hand, follow the same trend of abundance shown by Chlorobi, i.e., S2> S3> S1, with the plenty of 4.65%, 3.10% and 1.51%, respectively. The two phyla Chlorobi and Bacteriodetes ( Figure 4B) also relate to each other in conserved proteins and indels, as evident from the previous research. 23,26 Thus, the presence of various conserved indels in numerous proteins like FtsK, UvrB and ATP synthase alpha subunit also justifies the typical ancestral relationship between Chlorobi and Bacteriodetes. From our study, since the Bacteriodetes feeds on the waste produced by the Chlorobi and are phylogenetically related 23,26 both the population are abundant at the temperature of 55°C, it concluded that there exists the taxonomic correlation between the two taxa ( Figure 4B).

Analysis of the Functional Characteristics of Microbial Structure Concerning the Physiochemical Parameters
T h e e x i s t i n g p h y s i o c h e m i c a l environmental conditions have an extensive impact on the functional composition of the microbial constitution. 15,23,28 The triplot generated for canonical correspondence analysis (CCA) represents the direct link between the physiochemical parameters and the relative abundance of the metabolic pathways. The starch and sucrose metabolism were found positively correlated with the pH, fluoride, TDS and potassium. In contrast, fructose and mannose metabolism, transporter, ABC transporters, glycine, serine and threonine, arginine and proline metabolism are positively associated with physiological factors like ammonia, COD and phosphorus ( Figure  3B). The Peptidoglycan biosynthesis, pyrimidine metabolism, purine metabolism two-component system oxidative phosphorylation linked with temperature, ORP, DO and chlorine, while the nitrogen metabolism, carbon fixation pathway and chlorophyll metabolism and methane metabolism correlated with TOC, nitrate, and ammonia ( Figure 3B). Thus, the physiochemical parameters (ammonia, potassium, TOC, nitrate, nitrogen, COD, chloride, pH, TDS, phosphorus, DO, fluoride and ORP) govern the microbial structures. They also play a leading role in the community's diffusion of organic function and adaptive retort in the Deulajhari hot spring niche.

CONCluSiON
In the present investigation, we analyzed the unexplored diversity with their predicted function of the three springs (S1, S2 and S3) of the Deulajhari hot spring cluster through the 16S rRNA amplicon profiling. The physiochemical parameters analysis of collected sediment samples indicated the changes in BOD, COD, pH, phosphorus properties etc. Community structure among the bacterial population indicated dominance of Chloroflexi and Proteobacteria (at phylum level) and Dehalococcoidetes, Alphaproteobacteria, Deltaproteobacteria and Chloroflexi (at class level). The OTUs with no BLAST hits with known 16S rRNA in the database could be novel microbes in these habitats which needs to be further established. ABC transporter, two-component system and other transporter were expected to be the most prevalent microbial processes in sediment, while organotroph, autotroph and free-living organism were expected to occur highest in all the hot spring sediments sample. Furthermore, the co-occurrence of the Bacteriodetes and Chlorobi, along with the variant temperature, was also successfully established. Our results provide valuable insights into the putative microbial diversity and the function with their physiochemical correlation by sediment microbial communities in the hot springs sample. Further studies on in vitro culture followed by physiological analysis of these important microbes would be required to determine their precise functional roles within these communities.