Journal of Applied Microbiology ISSN 1364-5072

ORIGINAL ARTICLE Phylogenetic analysis of intestinal in the Chinese mitten crab (Eriocheir sinensis) K. Li1, W. Guan2, G. Wei1, B. Liu1,J.Xu3, L. Zhao1 and Y. Zhang1

1 Laboratory of Molecular Microbial Ecology and Ecogenomics, College of Life Science and Biotechnology, Shanghai Jiaotong University, Shanghai, China 2 College of Marine Science, Shanghai Fishery University, Shanghai, China 3 College of Life Science and Biotechnology, Shanghai Jiaotong University, Shanghai, China

Keywords Abstract Chinese mitten crab, denaturing gradient gel electrophoresis, intestinal bacteria, real-time Aims: To identify the dominant intestinal bacteria in the Chinese mitten crab, quantitative PCR, 16S rRNA. and to investigate the differences in the intestinal bacteria between pond-raised and wild crabs. Correspondence Methods and Results: The diversity of intestinal bacteria in the Chinese mitten Yan Zhang, Laboratory of Molecular Microbial crabs was investigated by denaturing gradient gel electrophoresis (DGGE) fin- Ecology and Ecogenomics, College of Life gerprinting, 16S rRNA gene clone library analysis and real-time quantitative Science and Biotechnology, Shanghai Jiaotong University, 800 Dongchuan Road, Shanghai PCR. The principal component analysis of DGGE profiles indicated that sub- 200240, China. stantial intersubject variations existed in intestinal bacteria in pond-raised crab. E-mail: [email protected] The sequencing of 16S rRNA genes revealed that 90–95% of the phylotypes in the clone libraries were affiliated with and . Some 2006 ⁄ 1170: received 15 August 2006, revised genera were identified as unique in wild crabs and in pond-raised crabs, and accepted 5 December 2006 whereas Bacteroidetes was found to be common in all sampled crab groups. Real-time quantitative PCR indicated that the abundance of Bacteroides and doi:10.1111/j.1365-2672.2007.03295.x the total bacterial load were approximately four-to-10 times higher in pond- raised crabs than in wild crabs. A significant portion of the phylotypes shared low similarity with previously sequenced organisms, indicating that the bacteria in the gut of Chinese mitten crabs are yet to be described. Conclusions: The intestinal bacteria of pond-raised crabs showed higher inter- subject variation, total diversity and abundance than that observed in wild crabs. The high proportion of the clones of Proteobacteria and Bacteroidetes in the clone library is an indication that these bacteria may be the dominant pop- ulation in the gut of the Chinese mitten crab. Significance and Impact of the Study: This study demonstrated obvious differ- ences in the intestinal bacterial composition of pond-raised crabs and wild crabs. This knowledge will increase our understanding of the effects of aquacul- ture operations on bacterial community composition in the crab gut and pro- vide necessary data for the development of probiotic products for crab cultivation.

jiang River water system), and typically consume dead Introduction aquatic animals and decaying corpses. Most pond-raised The Chinese mitten crab (Eriocheir sinensis) is recognized Chinese mitten crabs are fed aquatic animals and haslets as a valuable aquatic protein resource in China and other rather than artificial food sources. Recently, probiotics, countries. Wild Chinese mitten crabs usually live in fresh- which are live microbial feed supplements, have enhanced water rivers and lakes (in China, mainly in the Chang- aquaculture and beneficially affected the host animal by

ª 2007 Shanghai Jiaotong University Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682 675 Intestinal bacteria in crabs K. Li et al. improving its intestinal microbial balance (Fuller 1989). Mini Kit (Qiagen, Hilden, Germany) and was eluted in Before the gut microbial balance in Chinese mitten crabs 100 ll of water. The kit was evaluated previously to yield can be manipulated using probiotics, however, the diver- good recovery of bacterial total DNA for both gram- sity of their intestinal bacteria must first be understood. negative and gram-positive species (McOrist et al. 2002; Molecular methods based on the recovery of microbial Li et al. 2003). community DNA followed by PCR and 16S rRNA gene sequence analysis have proven useful for studying natural DGGE analysis of the V3 regions microbial populations, and thus, are accepted methods in microbial ecology (Lane et al. 1985; Amann et al. 1995). The samples were analysed using PCR-DGGE fingerprint- Specifically, denaturing gradient gel electrophoresis ing technology. The V3 region of the 16S rRNA gene was (DGGE) of PCR-amplified 16S rRNA genes from total amplified using primers P2 (5¢-ATTACCGCGGCTGCTG- microbial DNA isolated from environmental samples pro- G-3¢) and P3 (5¢-GCclamp-CCTACGGGAGGCAGCAG- vides a DNA fingerprint of each sample and permits subse- 3¢), as described by Muyzer et al. (1993). PCR reactions ) quent identification of community members by sequence contained 2Æ5 llof10· buffer, 2 llof2Æ50 mmol l 1 analysis (Muyzer et al. 1993). Such molecular techniques dNTP mixture, 1-U Taq DNA polymerase (TaKaRa, have been used to characterize the dominant micro-organ- Shiga, Japan), 25 pmol of each primer, and 1 ll of ge- isms in haddock larvae (Griffith et al. 2001), farmed and nomic DNA in a total volume of 25 ll. The PCR reac- wild salmon (Holben et al. 2002), rainbow trout (Huber tions were cycled in a thermocycler PCR system (PCR et al. 2004) and abalone (Tanaka et al. 2004). Sprint; Thermo Electron, Waltham, MA, USA) with an Previous studies on the Chinese mitten crab have fo- initial denaturation at 94C for 3 min followed by 20 cussed primarily on physiology and disease (Silvestre cycles of denaturation for 1 min at 94C, primer anneal- et al. 2005; Sun et al. 2005). Few studies have been per- ing for 1 min at 55C, and primer extension for 1 min at formed to explore the diversity of the intestinal bacteria 72C. ‘Reconditioning PCR’ was performed as described of wild and pond-raised crab despite the potential impli- previously (Thompson et al. 2002). Sterile distilled water cations of the results. In this study, DGGE, 16S rRNA was used instead of chromosomal DNA isolated from the gene clone library analysis and real-time quantitative PCR microbial community for negative control PCR. (RTQ-PCR) were used to investigate the bacterial diver- DGGE was performed with a Bio-Rad DCodeTM muta- sity in the intestinal tract of wild and pond-raised Chi- tion detection system (Bio-Rad, Hercules, CA, USA) nese mitten crabs. according to the manufacturer’s instructions. Approxi- mately 300 ng of PCR product was deposited in each well of 8% (weight in volume, w ⁄ v) polyacrylamide gels con- Materials and methods taining a linear 25–55% denaturant gradient (100% denat- ) urant corresponds to 7-mol l 1 urea and 40% deionized Sampling of Chinese mitten crabs and DNA extraction formamide). Electrophoresis was performed for 160 min from the intestine at a constant voltage of 200 V at 60C. The gels were Chinese mitten crabs were collected from the Changjiang stained with SYBR-Green I solution (10000· diluted) and river estuary (P.R. China) and from two freshwater crab photographed (UVI, UVItec, Cambridge, England). farms in March 2005. The two crab farms were located on Using an image analysis system (ImageJ, National Insti- Chongmin Island in the estuary and the water used for tutes of Health) to analyse the DGGE band profiles, the aquaculture is pumped in from the estuary. The collected densities and migration patterns of the individual bands crabs were healthy and 1-year-old animals with a weight of were calculated. Principal component analysis (PCA), approximately 100 g. A total of 12 Chinese mitten crab based on the density and migration of the bands, was samples from the three sampling locations were analysed performed using Matlab (version 7Æ0, Mathwork Inc.). in this study. Samples 1–4 (group C1) were from wild Chi- nese mitten crabs obtained from the Changjiang river estu- 16S rRNA gene amplification and library construction ary (P.R. China), samples 5–8 (group C2) and 9–12 (group C3) were from the two crab farms (the samples in each The 16S rRNA gene was amplified from each sample by pond-raised group were picked from the same pond). using the universal bacterial primers P0 (5¢-GAGAGTTT- All animals were euthanized, and the whole intestinal GATCCTGGCTCAG-3¢) and P6 (5¢-CTACGGCTACCTT- tracts were immediately removed and clamped to prevent GTTACGA-3¢) (Weisburg et al. 1991). The PCR reaction loss of sample. Each whole gut sample was excised asepti- components and amplification parameters were as des- cally with sterile toothpicks. The DNA was isolated cribed (Di Cello et al. 1997). The PCR products of each according to the instructions of the QIAampR DNA stool sample were purified from the gels, and the DNA concen-

ª 2007 Shanghai Jiaotong University 676 Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682 K. Li et al. Intestinal bacteria in crabs tration was determined using a HoeferR DyNA QuantR phylotypes assigned to Bacteroidetes. Recombinant plas- 200 system (Amersham Pharmacia Biotech, San Francisco, mid DNA was isolated, purified and serially diluted in

USA). The PCR products from four samples in each double-distilled water (dd H2O) to a final concentration ) group were mixed in an equal ratio. Pooled PCR ranging from 6Æ83 · 102 to 6Æ83 · 107 copies ll 1 as des- products were cloned into the p-GEM T Easy vector cribed previously (Liu et al. 2005). One-microlitre aliqu- (Promega, Madison, WI, USA) according to the manufac- ots of each dilution were used for RTQ-PCR to generate turer’s instructions. Competent Escherichia coli DH5a cells the standard curve, and used as quantification standards were transformed and screened for plasmid insertions for crab gut samples. The DNA isolated from four sam- according to the manufacturer’s instructions. ples in each group was mixed with an equal volume, and was run using a DNA Engine Opticon 2 system (MJ Research). The data were recorded, and analysed with the Sequencing and phylogenetic analysis corresponding Monitor software (version 3Æ1). One hundred and twenty-one clones from C1, 100 clones Primers BfrF (5¢-CTGAACCAGCCAAGTAGCG-3¢) and from C2, and 108 clones from C3 were chosen randomly BfrR (5¢-CCGCAAACTTTCACAACTGACTTA-3¢) were for further analysis. The coverage index (Good 1953) was used to quantify the Bacteroidetes group of organisms calculated by the formula [1 ) (n ⁄ N)] · 100%, where n is (Liu et al. 2003), whereas a set of universal primers (the the number of phylotypes represented by one clone and N forward primer, 5¢-TCCTACGGGAGGCAGCAGT-3¢ and is the total number of clones. Two other statistical indices, the reverse primer, 5¢-GGACTACCAGGGTATCTAATCC- the Shannon–Wiener index (H) (Krebs 1989) and the TGTT-3¢) were used for the amplification of the 16S Simpson’s index (D) (Simpson 1949), also were calculated rRNA gene from the Bacteria to estimate the for the three libraries. All insertions were amplified using total bacterial load (Nadkarni et al. 2002). The 25-ll primers specific for the T7 (5¢-GGCCGCGGGAATTC- amplification mixture contained 1Æ5-U Taq DNA GATT-3¢) and SP6 (5¢-GCGAATTCACTAGTGATT-3¢) polymerase (TaKaRa Co.) and 2Æ5 ll10· buffer, 2 llof ) sequences of the pGEM-T Easy vector. The products were 2Æ5-mmol l 1 dNTP mixture, 25 pmol of each primer, characterized by digestion with Csp6 I and Hinf I restric- 0Æ1 ll of genomic DNA and 1·SYBR Green I (Amresco). tion enzymes (Promega) to identify the different recom- RTQ-PCR was carried out for 40 cycles. Each cycle con- binant clones (phylotypes). Bidirectional sequencing of sisted of 95C for 20 s for denaturation, annealing for representative clones for each phylotype was performed 1 min at 52C to quantify the Bacteroidetes group or using the primers T7 and SP6 (Invitrogen, Shanghai, 55C to estimate total bacterial load; extension was per- China) to obtain partial sequences with a length of formed at 72C for 30 s. A final extension of 72C for approximately 1600 bp. All sequences were analysed for 5 min was added. At the end of each cycle (at 82C), the chimera formation with the Chimera Check feature of fluorescence signal was measured. RDP-II (Ribosomal Database Project II, http://rdp.cme.m- su.edu). All nonchimera sequences were compared with Nucleotide sequence accession numbers current accessible sequences to determine the closest known neighbour in GenBank at National Center of Bio- The 16S rRNA gene sequences obtained from the intesti- technology Information (NCBI) employing the Sequence nal bacteria in Chinese mitten crabs have been submitted Match feature (with KNN matches: 20) of RDP-II. The to GenBank under the accession numbers DQ856498 MEGA2 program was used to calculate the degree of simi- through DQ856562. larity between sequences. The classifier analysis tool (with default confidence threshold of 80%) of RDP-II was used Results to assign 16S rRNA sequences to the taxonomical hier- archy. These sequences were aligned using ClustalX ver- DGGE analysis of V3 regions sion 1Æ81 (Thompson et al. 1997). The phylogenetic trees were constructed using the neighbor-joining method by Approximately 300 ng to 4 lg of genomic DNA was iso- MEGA2 (Kumar et al. 2001). The bootstrap analysis of lated from each sample; the A260 ⁄ A280 ratios ranged 100 replicates also was performed using the same software. from 1Æ8to1Æ9. A 200-bp region of the 16S rRNA gene was amplified by PCR from DNA purified from the intes- tinal bacteria of Chinese mitten crabs. The DGGE analysis Real-time quantitative PCR was performed for each sample (Fig. 1). Approximately To establish the standard curve that was included in each 20 DGGE bands were observed for each profile. Each RTQ-PCR run, the representative clone from the phylo- sample profile displayed a unique banding pattern. The type C3B was used. C3B is the most dominant of the PCA of the DGGE band profiles showed that samples 5

ª 2007 Shanghai Jiaotong University Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682 677 Intestinal bacteria in crabs K. Li et al.

C1 C2 C3 Phylogenetic analysis of DNA sequences 123456789101112 The 16S rRNA gene was amplified from each DNA pool and was used to construct clone libraries (Table 1). The sequences of a length of approximately 1500 bp for clones representing individual phylotypes were obtained and used to construct two phylogenetic trees (Figs 3 and 4). For the three constructed libraries, the coverage index (C) ranged from 0Æ926 to 0Æ943, indicating that the libraries were large enough to sample the major biodiversity pre- sent in the samples. The Shannon–Wiener index (H) and the Simpson’s index (D) also were calculated (Table 1). These data suggested that the phylotype diversity was the lowest in the C1 library and highest in the C2 library. This result was consistent with the PCA of the DGGE profiles (Fig. 2). Thus, the bacterial diversity was higher Figure 1 Denaturing gradient gel electrophoresis banding profiles of in the guts of pond-raised crabs than in those of wild 16S rRNA gene V3 regions amplified from community DNA extracted crabs. from 12 intestinal microbiota samples from Chinese mitten crabs. The The 16S rRNA gene phylogenetic analysis revealed that lanes labelled 1 to 12 represent 12 individual crab samples. C1 repre- most phylotypes could be assigned to two phyla, Proteo- sents wild crab and C2 and C3 represent pond-raised crab from two bacteria and Bacteroidetes (Figs 3 and 4). Approximately farms. 43–50% of the phylotypes were phylogenetically assigned to the phylum Proteobacteria, whereas 40–52% of the 400 phylotypes were assigned to the phylum Bacteroidetes. 2 These data indicated that Proteobacteria and Bacteroidetes 4 3 C1 may constitute the dominant components of the intestinal 200 bacteria in the Chinese mitten crab. Variations in the 1 proportion of each phylotype were observed in the clone 5 libraries. A higher number of clones were assigned to Pro- C2 teobacteria (and ⁄ or Gammaproteobacteria) in library C1 0 6 7 (wild crabs) compared with the other libraries, whereas 8 more clones were assigned to Bacteroidetes in the C2 and 9 PC2 (27%) C3 libraries (pond-raised crabs). –200 11 Approximately 25% of the phylotypes from C1, 48% of 10 C3 the phylotypes from C2, and 57% of the phylotypes from C3 could not be identified at the level using the –400 12 classifier analysis tool of RDP-II. In those identified phyl- otypes, Bacteroides was common in all the three libraries. –200 0 200 400 600 Furthermore, some genera, Roseibium, Sphingobacterium, PC1 (30%) Thermomonas and Pseudomonas were identified as unique in wild crabs, whereas Chryseobacterium, Sejongia, Sulfi- Figure 2 Principal component analysis ordination showing the simi- larity of bacterial communities in the gut of Chinese mitten crabs. tobacter, Porohyrobacter, Propionivibrio, Rhodoferax, Vibrio, Each symbol represents the bacterial community from one sample, Anaerophaga, Hongiella, Haliscomenobacter and Verrucom- and the distance between symbols is equivalent to the dissimilarity of icrobium were identified as unique in pond-raised crabs the samples. Other details are as in the legend to Fig. 1. (Figs 3 and 4). and 6 were distinct from the other samples (Fig. 2), indi- Real-time quantitative PCR cating that substantial intersubject variation existed in intestinal bacteria in pond-raised crabs. Likewise, the bac- RTQ-PCR results indicated that the 16S rRNA gene copy terial communities gradually changed from group C3 to number of the genera Bacteroidetes increased by a magni- C1 along the axis of PC2 (Fig. 2). This result indicated tude of approximately four times and the total bacterial that intergroup variation also existed among the intestinal load was increased by four to 10 times in pond-raised bacteria. crabs as compared with wild crabs. The corresponding

ª 2007 Shanghai Jiaotong University 678 Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682 K. Li et al. Intestinal bacteria in crabs

Table 1 Comparison of coverage and diversity indices of the three 16S rRNA Number of Number of Coverage Shannon-Wiener Simpson’s gene clone libraries Clone library clones phylotypes (C) (%) index (H) index (D) Library C1 118 20 94Æ12Æ50 0Æ107 Library C2 95 23 92Æ62Æ81 0Æ074 Library C3 106 21 94Æ32Æ62 0Æ095

copy numbers of the genera Bacteroidetes were 5Æ81 · 105, 100 C1T [1, Rhodobacter] 2Æ67 · 106 and 2Æ07 · 106 copies per microlitre of DNA 90 Uncultured Rhodobacteraceae (DQ234188) 47 C2N [3, Rhodobacter] from C1 to C3, whereas the copy numbers of total bacter- 52 C2J [3, Rhodobacter] 6 7 7 87 Æ · Æ · Æ · Rhodobacter sp. (DQ413163) ial load were 3 62 10 ,144 10 and 3 89 10 copies 87 Rhodobacter sp. (AY244771) per microlitre of DNA from C1 to C3. The data are pre- C3Y [1] sented as the means of triplicate determinations. The 41 C1H [3, Rhodobacter] 96 100 Pseudorhodobacter incheonensis (DQ001322) standard deviation of means varied by £20%. 94 C3N [6, Paracoccus] 100 Paracoccus sp. (AM231050) C1I [6, Paracoccus] 100 73 100 C2L [6, Sulfitobacter] Discussion Uncultured Sulfitobacter sp. (AY697912) 88 100 C3P [5] 99 Roseobacter sp. (AY167339) The purpose of this study was to describe the diversity of C2P [4] intestinal bacteria found in Chinese mitten crabs from 76 C1V [19, Roseibium] 100 Bacterium K2-29 (AY345425) three sources (one wild and two crab farms). These 92 Mesorhizobium amorphae (DQ022832) results of this study were to promote more systematic 97 C2U [1] Sphingomonas sp. (AB074191) knowledge about the natural intestinal bacterial commu- 100 C2F [3, Porphyrobacter] Alpha-proteobacterium 60 100 Alpha proteobacterium (AF235995) nities in crabs and increase our understanding of the 59 Sphingomonas sp. (AB219359) effects of aquaculture operations. Furthermore, the data 33 C3X [1] C3M [2] will promote the development of optimal probiotic prod- 100 C1B [19] ucts to increase feeding efficiencies by improving the 57 C2E [10] C3K [17] intestinal microbial balance. Here, we report the differ- 100 C2Q [7] 100 C3E [2] ences and similarities present between these crab popula- 100 C2T [1, Propionivibrio] tions in regard to the gut populations of bacteria. 100 Rhodocyclus sp. (AY691423) C2M [1, Rhodoferax] This study is the first study to examine intestinal bac- 78 Uncultured beta proteobacterium Beta- teria in the Chinese mitten crab using molecular biology C2G [2] (AY622248)

100 C1Y [1, Thermomonas] proteobacterium techniques, such as DGGE, clone library analysis and 98 Gamma proteobacterium Gsoil (AB245359) RTQ-PCR. These PCR-based approaches, using 16S rRNA 100 100 C3S [1, Vibrio] Vibrio logei (AY292934) genes as markers, could have possible biases, and the reli- 100 C1L [1, Pseudomonas] 100 Pseudomonas sp. (DQ227349) ability is dependent mainly upon the efficiency of DNA 76 C1S [1, Pseudomonas] extraction and PCR bias (von Wintzingerode et al. 1997; 100 Pseudomonas migulae (AY047218) 79 84 C1A [15, Acinetobacter] Qiu et al. 2001). In this study, all samples were treated Alkanindiges hongkongensis (AY251391) identically throughout the DNA extraction procedure and 100 C3Q [1, Acinetobacter]

100 C1J [5, Acinetobacter] Gammaproteobacterium PCR amplification for DGGE and clone library analysis. 100 Acinetobacter johnsonii (AB099655) Furthermore, the effects of PCR bias should be minimized 0·02 because four different primer pairs were used in this

Figure 3 Dendrogram of 16S rRNA gene sequences showing the study. Such a combined approach should yield a more phylogenetic affiliation of intestinal bacteria in Chinese mitten crabs. reliable estimation of biological variability within and The neighbor-joining tree was constructed from 16S rRNA gene among community structures. Therefore, the samples sequences of representative clones of each phylotype assigned to Pro- could be compared. teobacteria and from sequences retrieved from the GenBank data- Substantial differences in the intestinal bacterial com- base. The phylotype name is shown in black typeface, and the munity composition and diversity have been observed number of clones belonging to that phylotype and taxonomic affili- previously in some cultural aquatic animals isolated from ation on the genus level identified with the classifier analysis tool of RDP-II are shown in square brackets. The scale bar represents five nuc- different sites; however, intersubject variation between leotide substitutions per 100 nucleotides. Bootstrap values (100 repli- samples from the same site were reported to be relatively cates) are shown at branch nodes. C1 represents wild crab and C2 slight. For example, the DGGE profiles of PCR-ampli- and C3 represent pond-raised crabs from two nearby farms. fied segments of the bacterial 16S rRNA gene from three

ª 2007 Shanghai Jiaotong University Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682 679 Intestinal bacteria in crabs K. Li et al.

48 C1F [8, Bacteroides] 53 C2B [7, Bacteroides] 100 C2X [3, Bacteroides] 100 C2B [11, Bacteroides] 100 Bacteroides sp. (AY082449) 100 C1P [3, Dysgonomonas] C2I [1, Dysgonomonas] 99 D. capnocytophagoides (U41355) 63 85 C1O [1, Dysgonomonas] 100 Uncultured bacterium (AB192157) Uncultured bacterium (AY082469) 100 C3I [5] 100 C3V [1] 68 100 C2C [8] C2O [1] 99 C1G [18] 100 C2D [8] 97 C3C [11] 80 Uncultured Cytophaga sp. (AB015260) C2V [1, Anaerophaga] 100 Bacteroidetes 99 C2H [14] 100 C3F [18] C1M [1, Flavobacterium] C1R [1, Flavobacterium] 91 100 57 C3A [3, Flavobacterium] 88 C3L [2, Flavobacterium] 99 Flavobacterium sp. (AM177626) 67 Flavobacterium sp. (AM177635) 75 C2K [1] 100 C2S [6, Chryseobacterium] 100 Chryseobacterium sp. (AB164636) 85 C3D [8, Sejongia] 46 100 Chryseobacterium sp. (AJ495802) C1C [6, Sphingobacterium] 100 Sphingobacterium faecium (AJ438176) 100 C3G [3] 100 Kaistomonas ginsengisoli (AB245370) 100 C3U [1, Hongiella] 100 Chimaereice alkaliphila (AJ717393) C1Q [1] 93 C3H [2, Haliscomenobacter] 100 Uncultured bacterium (AJ318142) 100 C2A [2, Verrucomicrobium] 99 Uncultured bacterium (AJ318142) C2R [2] Unaffiliated 100 Uncultured bacterium (AF368181) 98 Lactovum miscens (AJ439543) 99 C1N [4] 100 Uncultured bacterium (AB234512) Uncultured Lactococcus sp. (AB198466) Figure 4 Dendrogram of 16S rRNA gene 96 100 C1X [4] sequences showing the phylogenetic affili- C3T [5] ation of intestinal bacteria in Chinese mitten 100 Uncultured bacterium (AB088998) crabs, except those assigned to Proteo- 99 Uncultured Clostridiales sp. (AB234484) bacteria. Other details are as in the legend to 0·05 Fig. 3. rainbow trout from the same farm were similar (Huber Even though intersubject variation existed within each et al. 2004). Similarly, little variation in the abundance of group, the PCA analysis of DGGE fingerprints of all indi- some phylogenetic groups in the gut bacteria community vidual crabs revealed that the three crab groups were sep- of five abalone fed artificial diets containing brown algae arated from each other. This observation was a good were observed (Tanaka et al. 2004). Using similar indication that each group of crab still shares similarities molecular techniques, we revealed that variations in the in gut community structure so that the samples from gut bacteria community among pond-raised crabs (especi- each group could be pooled to get an overall description ally group C2) were more obvious than those for wild of each group’s gut bacterial structure. In the three 16S crabs. We revealed the large spread in group C2; however, rRNA gene clone libraries, the majority of phylotypes more data on aquaculture operations may be needed to were assigned to either Proteobacteria or Bacteroidetes. reveal the causes. In this study, we focussed mainly on Few phylotypes affiliated with gram-positive bacteria were intergroup variations. found in crab guts, although a high proportion of gram-

ª 2007 Shanghai Jiaotong University 680 Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682 K. Li et al. Intestinal bacteria in crabs positive bacteria in the water environment of the Chang- and diversity of the bacterial community composition in jiang river estuary was reported (Sekiguchi et al. 2002). the gut within sampling groups may be a direct result of Similarly, previous studies have shown that Proteobacteria different nutrients present in their relatively unique diets. also dominated in the guts of aquatic animals mentioned Specifically, numerous natural diets are used on crab earlier, whereas only low numbers of Bacteroidetes and farms, and these diets change much more frequently than gram-positive bacteria were found. in the wild. The enrichment of nutrition in the gut of Several genera, Bacteroides, Acinetobacter, Flavobacteri- pond-raised crabs, and therefore, the higher intersubject um, Chryseobacterium and Porphyrobacter, were identified variation, total diversity and abundance of the intestinal in crab guts, and some species belonging to these genera bacteria in pond-raised crabs likely is the effect of these ordinarily are related to some disease. For example, some feeding methods. Such variations could lead to differences Bacteroides species are opportunistic pathogens owing to in digestive and developmental efficiencies among the their association with a variety of soft tissue and other crab populations. Artificial diets and the use of probiotics infections (Liu et al. 2003); yet Bacteroides were identified may be important to increase the feeding efficiencies of in crab guts from all sampling sites. the crabs by improving the balance of their intestinal Besides Bacteroides, some genera, Rhodabacter, Paracoc- microbes. cus, Dysgonomonas and Flavobacterium were found to be Finally, the combination of DGGE analysis, 16S rRNA common in wild crabs and pond-raised crabs. However, gene sequencing and RTQ-PCR proved to be a powerful some genera were identified as unique in wild or pond- tool for gaining detailed insight into the bacterial diver- raised crabs. For example, Chryseobacterium and Sejongia sity of the intestinal system in crabs. Future studies may were identified in C2 and C3, respectively. Sejongia was benefit these molecular approaches to determine the effect reported as a novel genus and showed moderate relation- of artificial diets and probiotics on the stability of the ships to the genera Chryseobacterium (92Æ5–95Æ3%) (Hana intestinal bacteria and on the digestive and developmental et al. 2005). The members of Chryseobacterium are widely efficiency of the crabs. distributed in soil, water and clinical sources (Gallego et al. 2006). Some species of Chryseobacterium are patho- References gens causing infection (Padmaja et al. 2006). Using the classifier analysis tool of RDP-II, some phyl- Amann, R.I., Ludwig, W. and Schleifer, K.H. (1995) Phylo- otypes in the clone libraries were identified at the genus genetic identification and in situ detection of individual level. The classification was illuminated based on the microbial cells without cultivation. Microbiol Rev 59, 143– known type strain 16S rRNA sequences and often not 169. well supported for less-well-studied bacteria. In this study, Di Cello, F., Bevivino, A., Chiarini, L., Fani, R., Paffetti, D., more phylotypes in the libraries from pond-raised crabs Tabacchioni, S. and Dalmastri, C. (1997) Biodiversity of a could not be identified at the genus level than those from Burkholderia cepacia population isolated from the maize wild crabs using the same analysis tool. This finding indi- rhizosphere at different plant growth stages. Appl Environ cated that the intestinal bacterial communities in the guts Microbiol 63, 4485–4493. of Chinese mitten crabs, especially pond-raised crabs, Fuller, R. (1989) Probiotics in man and animals. J Appl Bacte- riol 66, 365–378. remain undefined and require further study. Gallego, V., Garcia, M.T. and Ventosa, A. (2006) Chryseobacte- RTQ-PCR was performed to quantify the genus Bacter- rium hispanicum sp. nov., isolated from the drinking water oides and total bacterial load in the crab gut samples. An distribution system of Sevilla, Spain. Int J Syst Evol Micro- issue that needs to be considered is that the 16S rRNA biol 56, 1589–1592. gene copy number generated from the RTQ-PCR data Good, I.J. (1953) The population frequencies of species and cannot be converted directly into cell counts, as the num- the estimation of population parameters. Biometrica 40, ber of rRNA gene operons per bacterial genome can vary 237–264. from one to as many as 15 (Klappenbach et al. 2001). Griffith, S., Melville, K., Cook, M., Vincent, S., St. Pierre, M. Nevertheless, if it is assumed that the average number of and Lanteigne, C. (2001) Profiling of bacterial species copies of the 16S rRNA gene per cell was the same for associated with haddock larviculture by PCR amplification the three groups (Nadkarni et al. 2002; Liu et al. 2005), of 16S rDNA and denaturing gradient gel electrophoresis. our results still indicated that the abundance of Bactero- J Aquatic Animal Health 13, 355–471. ides and the total bacterial load were approximately four Hana, Y., Yoon, H.I. and Jongsik, C. (2005) Sejongia antarctica to 10 times higher in pond-raised crabs than in wild gen. nov., sp. nov. and Sejongia jeonii sp. nov., isolated crabs. from the Antarctic. Int J Syst Evol Microbiol 55, 409–416. As wild and pond-raised Chinese mitten crabs typically Holben, W.E., Williams, P., Gilbert, M.A., Saarinen, M., consume dead aquatic animals, the intersubject variation Sarkilahti, L.K. and Apajalahti, J.H. (2002) Phylogenetic

ª 2007 Shanghai Jiaotong University Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682 681 Intestinal bacteria in crabs K. Li et al.

analysis of intestinal microflora indicates a novel Mycoplas- Padmaja, P., Verghese, S., Bhirmanandham, C.V., Ajith, ma phylotype in farmed and wild salmon. Microb Ecol 44, Thirugnanasambandham, S. and Rameshm, S. (2006) 175–185. Chryseobacterium meningosepticum – an uncommon patho- Huber, I., Spanggaard, B., Appel, K.F., Rossen, L., Nielsen, T. gen causing adult bacterial meningitis. Indian J Pathol and Gram, L. (2004) Phylogenetic analysis and in situ Microbiol 49, 293–295. identification of the intestinal microbial community of Qiu, X., Wu, L., Huang, H., McDonel, P.E., Palumbo, A.V., rainbow trout (Oncorhynchus mykiss, Walbaum). J Appl Tiedje, J.M. and Zhou, J (2001) Evaluation of PCR-gener- Microbiol 96, 117–132. ated chimeras, mutations, and heteroduplexes with 16S Klappenbach, J.A., Saxman, P.R., Cole, J.R. and Schmidt, T.M. rRNA gene-based cloning. Appl Environ Microbiol 67, 880– (2001) rrndb: the ribosomal RNA operon copy number 887. database. Nucleic Acids Res 29, 181–184. Sekiguchi, H., Watanabe, M., Nakahara, T., Xu, B. and Uchiy- Krebs, C.J. (1989) Ecological Methodology. New York: Harper ama, H. (2002) Succession of bacterial community struc- Collins. ture along the Changjiang River determined by denaturing Kumar, S., Tamura, K., Jakobsen, I.B. and Nei, M. (2001) gradient gel electrophoresis and clone library analysis. Appl MEGA2: molecular evolutionary genetics analysis software. Environ Microbiol 68, 5142–5150. Bioinformatics 17, 1244–1245. Silvestre, F., Trausch, G. and Devos, P. (2005) Hyper-osmo- Lane, D.J., Pace, B., Olsen, G.J., Stahl, D.A., Sogin, M.L. and regulatory capacity of the Chinese mitten crab (Eriocheir Pace, N.R. (1985) Rapid determination of 16S ribosomal sinensis) exposed to cadmium; acclimation during chronic RNA sequences for phylogenetic analyses. Proc Natl Acad exposure. Comp Biochem Physiol C Toxicol Pharmacol 140, Sci U S A 82, 6955–6959. 29–37. Li, M., Gong, J., Cottrill, M., Yu, H., De Lange, C., Burton, J. Simpson, E.H. (1949) Measurement of diversity. Nature 163, and Topp, E. (2003) Evaluation of QIAamp DNA Stool 688. Mini Kit for ecological studies of gut microbiota. J Micro- Sun, H., Zhou, K. and Song, D. (2005) Mitochondrial genome biol Methods 54, 13–20. of the Chinese mitten crab Eriocheir japonica sinenesis Liu, C., Song, Y., McTeague, M., Vu, A.W., Wexler, H. and (Brachyura: Thoracotremata: Grapsoidea) reveals a novel Finegold, S.M. (2003) Rapid identification of the species of gene order and two target regions of gene rearrangements. the Bacteroides fragilis group by multiplex PCR assays Gene 349, 207–217. using group- and species-specific primers. FEMS Microbiol Tanaka, R., Ootsubo, M., Sawabe, T., Ezura, Y. and Tajima, K. Lett 222, 9–16. (2004) Biodiversity and in situ abundance of gut microb- Liu, B., Zhang, F., Feng, X., Liu, Y., Yan, X., Zhang, X., Wang, iota of abalone (Haliotis discus hannai) determined by cul- L. and Zhao, L. (2005) Thauera and Azoarcus as function- ture-independent techniques. Aquaculture 241, 453–463. ally important generaina denitrifying quinoline-removal Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. bioreactoras revealed by microbial community structure and Higgins, D.G. (1997) The CLUSTAL_X windows comparison. FEMS Microbiol Ecol 55, 274–286. interface: flexible strategies for multiple sequence align- McOrist, A.L., Jackson, M. and Bird, A.R. (2002) A compar- ment aided by quality analysis tools. Nucleic Acids Res 25, ison of five methods for extraction of bacterial DNA 4876–4882. from human faecal samples. J Microbiol Methods 50, Thompson, J.R., Marcelino, L.A. and Polz, M.F. (2002) Het- 131–139. eroduplexes in mixed-template amplifications: formation, Muyzer, G., de Waal, E.C. and Uitterlinden, A.G. (1993) Pro- consequence and elimination by ‘reconditioning PCR’. filing of complex microbial populations by denaturing gra- Nucleic Acids Res 30, 2083–2088. dient gel electrophoresis analysis of polymerase chain Von Wintzingerode, F.V., Gobel, U.B. and Stackebrandt, E. reaction-amplified genes coding for 16S rRNA. Appl Envi- (1997) Determination of microbial diversity in environ- ron Microbiol 59, 695–700. mental samples: pitfalls of PCR-based rRNA analysis. Nadkarni, M.A., Martin, F.E., Jacques, N.A. and Hunter, N. FEMS Microbiol Rev 21, 213–229. (2002) Determination of bacterial load by real-time PCR Weisburg, W.G., Barns, S.M., Pelletier, D.A. and Lane, D.J. using a broad-range (universal) probe and primers set. (1991) 16S ribosomal DNA amplification for phylogenetic Microbiology 148, 257–266. study. J Bacteriol 173, 697–703.

ª 2007 Shanghai Jiaotong University 682 Journal compilation ª 2007 The Society for Applied Microbiology, Journal of Applied Microbiology 103 (2007) 675–682