Microbiota of Foregut and Hindgut of Blunt Snout Bream (Megalobrama amblycephala)

Liangkun Pan1, Qun-Lan Zhou1*, Bo Liu1, Jiaxing Yan2 and Ajun Sun1 1Key Laboratory of Freshwater Fisheries and Germplasm Resources Utilization, Ministry of Agriculture, Freshwater Fisheries Research Center, Chinese Academy of Fishery Sciences, Wuxi 214081, China 2College of Information Science and Technology, Nanjing Agricultural University, Nanjing, Jiangsu, China Article Information Received 18 October 2019 Revised 11 January 2020 Accepted 04 March 2020 Available online 26 October 2020


INTRODUCTION
G ut microbiota has attracted increased attention in the past ten years after the use of 454 pyrotags to analyze the microbial profile began. It is accepted that the large and diverse bacterial community is very important to nutrient digestion and absorption, and to the health of the gut and the immunity of the host (Jiang et al., 2011;Desai et al., 2012;Wu et al., 2013;Ingerslev et al., 2014;Ye et al., 2014;Zhang et al., 2017). It has also been confirmed that microorganisms in the gut can improve the health of host gut cells by synthesizing important nutrients, such as short-chain fatty acids, peptide, amino acid etc. (Borsodi et al., 2017;Hao et al., 2017). It was well known that fish species  and the daily diet (He et al., 2013;Ingerslev et al., 2014; could be affected the composition and function of microbiota in the gut of fish (Reveco et al., 2014).

Rearing system
The trail was conducted in the cylindrical fiber glass tanks (300 L water per tank) that were part of a recirculating system, equipped with a sedimentation tank that contained activated carbon and corallite stones as a biological filter. All tanks were equipped with aeration. Water temperature ranged from 27 to 29 °C, which was controlled by a water temperature control system. pH was 7.4-7.8 and dissolved oxygen was approximately 6 mg L -1 throughout the trial.

Fish and feeding
A total of 150 juvenile blunt snout breams (initial weight 1.31 ± 0.45 g) were divided into 5 tanks with 30 fish in each tank, which were provided by the fish farm of Freshwater Fisheries Research Center, Chinese Academy of Fishery Sciences in China. Fish were fed with the commercial diet (330 g Kg -1 crude protein, 15.82 MJ Kg -1 ) (Tongwei Co., Ltd. Wuxi, China), which was formulated using the ingredients such as fishmeal, soybean meal, canola meal, cottonseed meal, wheat flour, rice bran etc. The feed trail lasted for 16 weeks. Fish were hand-fed four O n l i n e

F i r s t A r t i c l e
times a day at 8:00, 10:30, 13:30 and 16:00 until apparent satiation on the basis of visual observation. Pellets were distributed slowly, permitting all fish to eat. The final body weight of the fish was varied from 9.66 g to 14.67 g, with average final weight 12.02 ± 1.64 g.

Sample collection
After 16 weeks, fish were fasted for 24 h before collecting samples. Six fish from each tank were sampled and anaesthetized by MS-222 (tricainemethanesulfonate, 100mg L -1 , Sigma, USA) and then dissected to obtain the whole gut without digesta. After the complete intestine was removed from the fish, the contents of the intestine were scraped off with a spatula to retain the slimy intestinal sample. The gut was divided into two sections based on the length, the anterior third of the intestine was considered as the foregut and the remainder was considered as the hindgut. Individual gut samples were placed in Eppendorf tubes, immediately put into liquid nitrogen and stored at -80 o C until DNA extraction.

DNA extraction, PCR amplification and pyrosequencing
DNA was isolated from the gut content samples using QIAamp Fast DNA Stool Mini Kit (QIAGEN, Gemany), according to the manufacturer's instructions. DNA concentration was detected using Nanodrop (Thermo scientific, USA). 6 DNA samples from six different fish of the same tank obtained from the same section were mixed in equimolar amount to make a pooled sample.
For the 454 pyrosequencing, an amplicon library was prepared using eubacterial universal primers. Primers with special barcodes were used to amplify 450 bps of the 16S rRNA genes covering the V3 to V4 regions. Primers were as follows: 343F 5'-TACGGRAGGCAGCAG-3', 798R 5'-AGGGTATCTAATCCT-3'. PCR was carried out using polymerase (Trans Start Fastpfu DNA Polymerase, Transgen Biotech, Beijing, China) with an annealing temperature of 52 o C and 30 cycles to minimize PCR biases. Purified PCR products were submitted for pyrosequencing as described in the PE300 protocol using Illumina Miseq (Illumina, USA) in the Shanghai Majorbio Co. Ltd (Shanghai, China).

Sequence processing, OUT assignment, identification and classification
Sequences generated by pyrosequencing were filtered to remove the ambiguous sequences, homologous sequences and short length sequences from the resulting raw data set, provided by Illumina Miseq, using Trimmomatic software (Bolger et al., 2014), then edited using FLASH software (Reyon et al., 2012). Operational Taxonomic Units (OTUs, the equivalent of species) were picked using a closed-reference protocol at 97% similarity, as described by CD-HIT software (Li and Godzik, 2006). Reads that did not match any reference sequence with at least 97% identity were discarded. OTUs were classified and assigned taxonomic identities based on their best match with database Greengenes and Genebank of NCBI using the softwares RDP classifier and PyNAST.

Statistical analyses
The relative abundance of micro community was assessed at phylum and genus level were calculated based on the OUT number using software QIIME. The dominant microorganism was identified based on its relative abundance among all other microorganisms. Alpha diversity indices were determined using the Shannon-Wiener index, Simpson index for diversity and the Chao1 index for species richness, which were calculated using software QIIME. P value were calculated using SPSS 19.0 T test. A heatmap with cluster at genus level was draw based on the abundance of gut microbiota using the software Fast Tree. The dominant microbiota was selected based on the relative abundance. A boxplot was drawn based on the relative abundance of dominant microbiota using software SPSS 19.0, to compare the different distributions between the foregut and hindgut.

RESULTS
Based on the alpha-diversity analysis, the Chao1 index, Shannon Wiener index and Simpson index were all higher in the hindgut than the foregut, but the difference was not significant. and Pseudomonas (7.34%). Based on the cluster of all samples, they could be clustered into two broad categories of foregut and hindgut samples except S1H, which was clustered with foregut. The dominant phylum was different between the foregut and hindgut, which was Firmicutes (25.88%) in the foregut and Fusobacteria (33.52%) in the hindgut. As shown in Table II and Figure 2, the percentage of core microbiota at the genus level varied greatly between the foregut and the hindgut. Lactococcus (22.62%) preferred the foregut, while Cetobacterium (33.42%), CK-1C4-19 (14.52%) and Flavobacterium (7.82%) were usually found in the hindgut.

DISCUSSION
The alpha-diversity was higher in the hindgut than the foregut. That means not only the microbiota diversity but also the species richness was high in the hindgut than in the foregut. This may be attributed to the different functions of foregut and hindgut and the differences in the chemical environment of these two sections (Sullam et al., 2012). Meanwhile it is well known that high diversity of the microbiome could make this system more stable, this may be the reason why the alpha-diversity between the foregut and hindgut was not significant. The cluster results were showed that the microbiota of foregut and hindgut were different too. This may mean we should choose the part of gut to study based on the research purpose.  The dominant microbiota at the phylum level of blunt snout bream was similar to previous studies carried out on the larvae of blunt snout from a commercial fish farm  and on the blunt snout bream from Wuhu Lake (Li et al., 2014). However, performing the analysis O n l i n e

F i r s t A r t i c l e
at genus level, results were different. Only Cetobacterium was considered as one of the dominant genera of blunt snout bream in the previous study . Cetobacterium was also the dominant genera in several other fish species, including the herbivorous grass carp, the filter-feeding big head carp, omnivorous common carp (Cyprinus carpio), Japanese white crucian carp (Carassius cuvieri), bluegill (Lepomis macrochirus), the carnivorous channel catfish (Ictalurus punctatus) and largemouth bass (Micropterus salmoides) (van Kessel et al., 2011;Roeselers et al., 2011;Larsen et al., 2014;Hao et al., 2017). Lactoccus was one of the dominant bacterial genera found in the grass carp too . Similar to the report on grass carp, Pseudomonas was another genus of core microbiota found in blunt snout bream . Previous studies have indicated that Pseudomonas is an important biological control agent in aquaculture that can degrade certain toxic compounds and remove harmful residual materials from fish intestines (Nam et al., 2003). Based on the cluster result, microbiota in the foregut and hindgut were divided into two groups. But the microbiota in the hindgut of Sample 1 was clustered with foregut group. This might because the dominant genus of Cetobacterium in the S1H was as low as that in the foregut. On the other hand, it suggested that gut microbes vary greatly between individuals.
It was found in this study that Lactococcus, Rhodobacter and Pseudomonas preferred the foregut, while Cetobacterium and Flavobacterium usually found in the hindgut. Lactococcus is a notable probiotic bacteria, which can not only help the host to maintain a healthy gut environment by producing antibacterial substances that inhibit the spread of harmful intestinal bacteria and suppress growth of competing bacteria (Dawood et al., 2016), but it can also ferment α-galactosides, such as melibiose and raffinose (Boucher et al., 2003). Meanwhile, it preferred the foregut maybe decided by the chemical environment of gut, for it is a facultative anaerobe.
Cetobacterium was considered to relate to protein digestion, for it could ferment peptone to produce acetic and propionic acids (Tsuchiya et al., 2008). Acetic and propionic acids are short chain fatty acids that can improve mucosal morphology, preserve enterocyte ultrastructure, and positively impact the host's health via prevention of certain diseases (Hao et al., 2017). It was preferred the hindgut, which was in agreement with the report on grass carp (Hao et al., 2017). The preference of Cetobacterium to the hindgut can be attributed to its function in fermenting peptone to produce acetic and propionic acids (Tsuchiya et al., 2008).
In conclusion, this study provides an insight into the gut microbiota of the blunt snout bream. The results reveal that the diversity of the foregut and hindgut were similar, but the microbiota was different. Some dominant bacteria preferred to plant in different sections of gut, which decided by its function and physiological characteristics. More studies are still needed to identify the function of certain dominant genera in freshwater fish such as Cetobacterium and Lactococus.