Gut bacteria associated with different diets in reared Nephrops norvegicus

Author's personal copy

Systematic and Applied Microbiology 35 (2012) 473– 482

Contents lists available at SciVerse ScienceDirect

Systematic and Applied Microbiology

jo u rn al homepage: www.elsev ier .de /syapm

Gut bacteria associated with different diets in reared Nephrops norvegicus

Alexandra Meziti, Eleni Mente, Konstantinos Ar. Kormas ∗

Department of Ichthyology & Aquatic Environment, School of Agricultural Sciences, University of Thessaly, 384 46 Volos, Greece

a r t i c l e i n f o

Article history:Received 7 June 2012Received in revised form 26 July 2012Accepted 27 July 2012

Keywords:BacteriaGutDietNephrops norvegicusAquaculture

a b s t r a c t

The impact of different diets on the gut microbiota of reared Nephrops norvegicus was investigated basedon bacterial 16S rRNA gene diversity. Specimens were collected from Pagasitikos Gulf (Greece) and keptin experimental rearing tanks, under in situ conditions, for 6 months. Treatments included three diets:frozen natural (mussel) food (M), dry formulated pellet (P) and starvation (S). Gut samples were collectedat the initiation of the experiment, and after 3 and 6 months. Tank water and diet samples were alsoanalyzed for bacterial 16S rRNA gene diversity. Statistical analysis separated the two groups fed or starved(M and P vs. S samples). Most gut bacteria were not related to the water or diet bacteria, while bacterialdiversity was higher in the starvation samples. M and P samples were dominated by Gammaproteobacteria,Epsilonproteobacteria and Tenericutes. Phylotypes clustering in Photobacterium leiognathi, Shewanella sp.and Entomoplasmatales had high frequencies in the M and P samples but low sequence frequencies in Ssamples. The study showed that feeding resulted in the selection of specific species, which also occurs inthe natural population, and might be associated with the animal’s nutrition.

© 2012 Elsevier GmbH. All rights reserved.

Introduction

The Norway lobster Nephrops norvegicus is a decapod crustaceanliving at depths of 20–800 m in the Mediterranean, the North Seaand the North East Atlantic Ocean. It is a commercially importantspecies [4], but overexploitation and inappropriate managementstrategies have led to possible depletion of the existing stocks. Sev-eral studies of N. norvegicus have been performed in the past ontheir biology [4], population abundance and structure [1], moltingand growth [17], feeding ecology and behavior [9] and reproduc-tive biology [35,46,47]. However, the lack of knowledge concerningtheir nutritional requirements and rearing under laboratory condi-tions [45] is the constraint limiting the successful culture of theNorway lobster, either for restocking or for commercial size pro-duction. Recent studies [34] have provided valuable data on thesurvival, growth and feeding behavior of N. norvegicus kept in con-trolled laboratory conditions, and they have shown, among others,the need for high-quality dry food.

N. norvegicus mainly feeds on fish, mollusks, crustaceans, poly-chaetes, echinoderms and foraminifers [9]. However, the syntheticfeed provided to N. norvegicus in past studies [34] was mostly pel-lets used in fish aquaculture that consisted mainly of fishmeal andsoy. The use of probiotics, although very important in the aquacul-ture of crustaceans [15], has never been attempted in the rearing ofN. norvegicus. The probiotics used in crustacean aquaculture have

∗ Corresponding author. Tel.: +30 242 109 3082; fax: +30 242 109 3157.E-mail address: [email protected] (K.Ar. Kormas).

been shown to increase the growth and survival of the species ([15]and references therein) without the use of antibiotics. The probio-tics used for each species are determined by several factors, such asthe non-pathogenicity and non-toxicity of the strain used, as well asits ability to survive in and adhere to the gut [15]. Additionally, pro-biotics should benefit their host in certain ways, such as promotinggrowth or protecting against pathogens [2,25].

Recent studies have shown that the gut microbial communi-ties of several animals are influenced by the nutritional habits oftheir hosts [29] and, at the same time, they metabolize part of theingested food and provide the host with important nutrients (e.g.cellulose digestion). Resident gut microbes that are able to metab-olize complex compounds are very good candidates as probioticssince they fulfill all the above-mentioned criteria. In the case of thegut microorganisms of N. norvegicus, the only study that has beenperformed in wild specimens showed that a seasonal variation ofmid-gut bacterial communities was mostly related to differencesin food supply from the overlying water column [36].

During the last 20 years, N. norvegicus catches per unit of efforthave steadily declined within the Mediterranean Sea and, thus, sup-port the need for measures to conserve this crustacean species.Aquaculture has been a very useful tool for restocking and stockenhancement programs for a number of fish and shellfish species.Although commercial cultivation of Norway lobsters might be hin-dered by the slow growth rate exhibited by juveniles of this speciesin nature [38] and in captivity [10], there is a lot of interest in theintensive cultivation of these animals due to their high nutritionaland commercial value. Although there are studies on husbandryand rearing conditions for the Norway lobster [34,45], knowledge

0723-2020/$ – see front matter © 2012 Elsevier GmbH. All rights reserved.http://dx.doi.org/10.1016/j.syapm.2012.07.004


474 A. Meziti et al. / Systematic and Applied Microbiology 35 (2012) 473– 482

Table 1Codes of samples collected and analyzed. Samples from the 2nd batch (April 2009)are indicated in boldface.

Samples t0 (start) t1 (3 months) t2 (6 months)

Natural population Nat1/Nat2Starvation S3m1 S6m1/S6m2Mussels M3m1 M6m1/M6m2Pellets P6m1/P6m2

on the nutritional physiology of the species remains limited. There-fore, the present study contributes to our knowledge of the possiblerole of its gut microbiota in nutrition. Such information can providethe basis for diet formulation in commercial scale culturing of theNorway lobster. In particular, this study aimed at investigating thefactors that affected the dominant gut bacterial communities of N.norvegicus grown under stable laboratory conditions. For this pur-pose, N. norvegicus individuals were reared in three different groupswhere mussels (natural feed) and pellets were provided, respec-tively, while the third group was starved. Mid-gut samples werecollected at the beginning of the experiment and 3 and 6 monthsafterwards, and they were investigated by 16S rRNA gene diversityanalysis.

Materials and methods

Collection of N. norvegicus individuals

Individuals were collected from Pagasitikos Gulf (Greece) inMarch (1st batch) and April 2009 (2nd batch). Sampling depthvaried from 60 to 88 m and only male individuals were kept forfurther analysis, since it has been shown from previous studiesthat males have better survival rates than females under rearingconditions [41], and that sex is not a significant factor for gut bac-terial diversity [36]. After collection, N. norvegicus individuals wereimmediately transferred to the laboratory in aerated seawater, andanimal weight and carapace length were measured. One individualfrom each batch (Nat1–Nat2) was immediately sacrificed, while therest were kept for rearing (Table 1).

Rearing procedure

Individuals for rearing were placed in 100 L glass tanks in thelaboratory. Water from Pagasitikos Gulf was transported in thetanks before N. norvegicus transport. In order to establish the bac-terial nitrifying communities on the filters of the tanks, bacteriafrom commercial solutions (Stability, Seachem Inc., USA) wereadded to the tanks. Water was constantly recycled through carbonand biological filters. Water temperature, salinity and photoperiodwere maintained at 11.9 ± 0.8 ◦C, 374 ± 0.2 ppm and 24 h darkness,respectively, reflecting in situ conditions.

After a 15-day acclimatization period, the animals were dividedinto three groups (M, P, S), and each animal was placed in a sepa-rate compartment in tanks made from Plexiglas and plastic netting.Groups M and P were supplied with 1–4 g frozen mussels (naturalfeed) and approximately 1 g fish pellets (synthetic feed), respec-tively, three times per week, while group S was starved. The diet’schemical composition was 69% protein, 7.5% lipid and 23.5% car-bohydrate for dry mussels and 42% protein, 11.1% lipid and 46.9%carbohydrate for pellets (Rotllant, unpublished data).

Collection of samples

Three and 6 months after the initiation of the experiment,animals from each group (Table 1) were sacrificed and their mor-phometric characteristics were measured. All animals were healthy

when sacrificed. During the rearing period, molting had beenobserved only in one individual from group P (P6m2) 2 days beforethe animal was sacrificed (post-molt stage).

Water samples were collected from the tanks 2–3 days beforethe sampling of the animals. Water samples were collected insterile 1 L bottles and 1 L, 500 mL and 800 mL were filtered fromsamples wt2 (t1), wt3 (t2, 1st batch) and wt4 (t2, 2nd batch), respec-tively. Filtering was performed under vacuum using 0.2 �m filters(GTTP, Millipore, USA) and filters were kept at −20 ◦C until furtherprocessing. Samples from mussels and pellets from the batchesprovided for feeding were also kept.

Mid-gut isolation

The animals were dissected using sterile lancets and the intes-tine was extracted using sterile forceps, as described in Mezitiet al. [36]. Since the bacterial communities established on or in thegut tissue were required, the intestine was emptied by applyingmechanical force and by rinsing three times in autoclaved particle-free seawater in order to remove all gut content. The posterior partof the mid-gut was used for further analysis. All dissecting toolswere alcohol-flame sterilized between each individual sample.

DNA extraction

DNA extraction was performed on 10 N. norvegicus gut tissuesusing the QIAamp DNA Mini Kit (Qiagen Inc., USA) and followingthe manufacturer’s standard protocol. At the final step, DNA wasdiluted in 100 �L of the elution buffer provided with the kit and itwas stored at −20 ◦C.

DNA extraction from the water samples was performed usingthe UltraClean Soil DNA Kit (MoBio Laboratories Inc. USA), follow-ing the manufacturer’s protocol. DNA was finally eluted in the 50 �Lelution buffer provided by the manufacturer.

DNA was extracted from pooled foot and mantle tissues of threemussels from the batch used for feeding (frozen Mytilus edulis)following the same procedure as for the mid-gut samples. Forthe pellets, DNA was extracted from 10 pooled pellets using theUltraClean Soil DNA kit (MoBio Laboratories Inc.), following themanufacturer’s standard protocol.

Cloning and sequencing of 16S rRNA genes

Part of the bacterial 16S rRNA gene was amplified from allmid-gut, water, mussel and pellet samples using the primers27f BAC (5′-AGAGTTTGATCMTGGCTCAG-3′) [27] and 907R (5′-CCGTCAATTCCTTTRAGTTT-3′) [37]. PCR conditions were 5 min at94 ◦C followed by 22–27 cycles of 1 min at 94 ◦C, 1 min at 52.5 ◦Cand 1 min at 72 ◦C and a final step of 7 min at 72 ◦C. The num-ber of PCR cycles was adjusted when needed in order to decreasenon-specific products. The total number of cycles for all sam-ples varied from 23 (sample S6m1) to 27 cycles (sample P6m1).PCR products were purified with the Montage Purification Kit(Millipore, USA) and were cloned directly using the TOPO TA Kitfor sequencing (Invitrogen Inc., USA) with electrocompetent cells.The insert size was checked using PCR with M13f–M13R vector-binding primers. Positive clones were grown overnight in 1.5 mLof Luria–Bertani medium containing kanamycin (50 �g mL−1), andplasmids were purified from the pelleted cells using the Nucle-ospin Plasmid QuickPure Kit (Macherey-Nagel GmbH and Co., KG,Germany). Plasmids were partially sequenced with primer M13f(5′-GTAAAACGACGGCCAG-3′). After alignment with ClustalW [28],manual correction, elimination of chimeras using the Pintail soft-ware [3] and visual examination of the alignments, clones weregrouped based on a 16S rRNA similarity cut-off level of 98% and rep-resentatives from each group were sequenced using primer M13R


A. Meziti et al. / Systematic and Applied Microbiology 35 (2012) 473– 482 475

Table 2Dominant phylotypes (>15%) in the studied samples.

Phylotype Frequency (%) Closest relative (GenBank No) (% similarity) Phylogeny Publication

Nat1-12 66.6 Uncultured bacterium clone My46-424 (GQ866072) (99) Gammaproteobacteria [36]Nat2-8 48.0 Uncultured clone Ag31-3 (GQ866101) (99) Gammaproteobacteria [36]S3m1-3 15.0 Marine sponge bacterium plate OTU18 (EU346505) (99) Alphaproteobacteria [52]S3m1-6 15.0 Marinicella litoralis (AB500095) (96) Gammaproteobacteria [42]S3m1-7 15.0 Litoreibacter albidus (AB518881) (99) Alphaproteobacteria [43]M3m1-15 15.8 Uncultured bacterium clone C2E (DQ856531) (96) Alphaproteobacteria [30]M3m1-41 15.8 Epsilonproteobacterium Oy-M7 clone 465.4 (DQ357825) (96) Epsilonproteobacteria [49]M6m1-4 20.5 Uncultured clone Ag31-3 (GQ866101) (99) Gammaproteobacteria [36]M6m1-56 17.9 Uncultured bacterium clone TIGU1075 (HM558927) (99) Gammaproteobacteria [54]M6m2-6 28.9 Uncultured clone Ag31-3 (GQ866101) (99) Gammaproteobacteria [36]M6m2-3 15.8 Uncultured Mycoplasmataceae clone Lo Hep1.15 (EU646198) (88) Tenericutes [16]P6m1-12 33.3 Uncultured bacterium clone D1-674 (GQ866083) (99) Gammaproteobacteria [36]P6m1-11 20.8 Uncultured clone Ag31-3 (GQ866101) (99) Gammaproteobacteria [36]P6m2-2 73.5 Uncultured clone Ag31-3 (GQ866101) (99) Gammaproteobacteria [36]

(5′-CAGGAAACAGCTATGAC-3′). Sequence data were obtained bycapillary electrophoresis (Macrogen Inc., Korea) using the Big DyeTerminator Kit (Applied Biosystems Inc., USA). Sequences werechecked for closest relatives using the BLAST application and allsequences were checked for chimeras using Pintail. 16S rRNAsequences were aligned using the ARB software [31] and the SILVAaligner application [40]. 16S rRNA distance matrices were cal-culated with the Jukes–Cantor formula and they were clusteredwith the neighbor-joining method. Bootstrap values were obtainedfrom 1000 replicates using similar parameters. All 16S rRNAsequences from this study were deposited in GenBank under num-bers JN092133–JN092292 (mid-gut samples), JN639288–JN639332(water samples) and JN858926–JN858954 (mussel and pelletsamples).

NMDS analysis

Unconstrained ordinations, based on the frequencies of thephylotypes, were performed in order to illustrate the rela-tionships between gut and water samples graphically usingthree-dimensional non-metric multidimensional scaling (NMDS)[26], implemented in R (version 2.9.1). NMDS ordination attemptsto place all samples in a three-dimensional space such thattheir ordering relationships (here based on a Bray–Curtis simi-larity matrix) can be preserved. Hence, the closer the samplesare in the resulting ordination, the more similar the bacterialcommunities are. Kruskal’s stress value reflects the difficultyinvolved in fitting the relationships of the samples into a three-dimensional ordination space. The hypothesis that gut microbialcommunities differed depending on whether food was providedor not was tested with the use of the non-parametric analysisof similarities (ANOSIM) [6]. ANOSIM generates a test statistic,R, that ranges from −1 to 1. The magnitude of R is indicativeof the degree of separation between groups, with a score of 1indicating complete separation and 0 indicating no separation[5].

Diversity and similarity analysis

The indices of Shannon–Wiener (H) [50], Simpson (D) [51] andMargalef [33] were used for diversity estimates and were cal-culated using the PAST software. Morisita similarity indices onthe phylotypes of the samples were calculated with the SPADEsoftware (http://chao.stat.nthu.edu.tw/softwareCE.html). Clusteranalysis was applied to Morisita similarity indices using the PASTprogram [19]. ANOSIM between the phylotype frequencies ofthe potential groups was also performed using the PAST pro-gram.

Results

Phylogenetic analysis

A total of 520 partial 16S rRNA sequences were analyzedfor samples Nat1 (30), Nat2 (25), S6m1 (39), S6m2 (43), M6m1(39), M6m2 (38), P6m1 (24), P6m2 (34), M3m1 (38) and S3m1(40), wt2 (34), wt3 (44), wt4 (34), Mus (30) and Pl (28). Eachclone library had 6–30 different phylotypes based on a 98% cut-off similarity. Good’s coverage was calculated using the formulaC = 1 − (ni/N), where ni is the number of singleton phylotypesand N is the total number of clones analyzed. It ranged from52% (S6m2) to 91% (P6m2) (Table S1), showing that at least50% of the total mid-gut bacterial species richness was revealedin all samples. The slope of the collector’s curves, plotting thenumber of total clones sampled against the number of differ-ent phylotypes showed that sampling was incomplete and rarespecies probably remained undetected (Fig. S1). However, almostall libraries from the rearing samples (apart from the starvedones) had one to two dominant phylotypes (sequence frequen-cies >15%) implying that the dominant phylotypes were detected(Table 2).

Mid-gut samples

In five samples (Nat2, M6m1, M6m2, P6m1 and P6m2)the dominant phylotypes (Table 2) were 98–99% similar andwere closely related (98–99%) to Photobacterium leiognathi strainRM1 (AY292947) [39] and to phylotypes Jl1-1 (GQ866087), O2-1 (GQ866108) and Ag31-3 (GQ866101), previously detected inthe gut of the Pagasitikos Gulf Norway lobster population [36](Figs. 1 and 2).

In sample P6m1, the dominant phylotype (33.3%) was 99%similar to Shewanella sp. E5050-7 (FJ169983), a protease produc-ing bacterium from the South China Sea [59], and to phylotypeD1-674 (GQ866083) detected in the gut of wild N. norvegi-cus [36]. Closely related (98%) phylotypes were also found inlower frequencies (Figs. 1 and 2) in samples M6m2, S6m1 andS6m2.

Phylotypes showing high frequencies in the 6-month mussel-fed samples (M6m1-2, 12.8% and M6m2-3, 15.7%) clustered withinthe order of Entomoplasmatales (Fig. 3) but they were distantlyrelated (91%) to all other members of the order. Their closestrelatives were the uncultured Candidatus Hepatoplasma clones Ty-Hep1.19 and Lo-Hep 11.5 detected in the hepatopancreas of theisopods Tylos europaeus and Ligia oceanica, respectively [16]. Phy-lotypes clustering in the same group were also detected in othersamples (S6m2 and Nat2) but in lower frequencies (Figs. 2 and 3).



Fig. 1. Neighbor-joining tree based on gammaproteobacterial 16S rRNA gene sequences from wild and reared Nephrops norvegicus. Arcobacter nitrofigilis was set as anoutgroup. The bar corresponds to a 10% nucleotide difference and bootstrap values were calculated from 1000 replicate trees. Phylotypes from previous N. norvegicus studiesare designated with *. Frequencies of retrieved phylotypes in each clone library are shown in parentheses.

Similarly, the dominant phylotype in the 3-month mussel-fedsample (M3m1-41, 15.8%), was closely related (99%) to phylotypesfrom the rest of the mussel-fed samples and from sample P6m1(Figs. 2 and 3). These phylotypes were affiliated to the unculturedbacterium Oy-M7 clone 465.4 (DQ357825) from oyster hatcheries[49], clustering within the genus Arcobacter according to the SILVAdatabase 108 [40]. The other dominant phylotype from the 3-month mussel-fed sample (M3m1-15, 15.8%) was affiliated (96%)to the alphaproteobacterial clone C2E (DQ856531) detected in theintestine of the Chinese mitten crab Eriocheir sinensis [30], whichclusters in the genus Defluviicoccus according to the SILVA database.

The rest of the phylotypes clustered within the Alpha-, Beta-and Gammaproteobacteria, in the phyla Bacteroidetes, Fibrobacteres,Firmicutes and Actinobacteria, and in the candidate divisions of OD1,OP11 and TM6 (Table S2).

Water samples

The dominant phylotype (23.5%) in sample wt2 was clas-sified as Flavobacteria (Fig. 4). Its closest relative was strainKordia algicida OT-1 isolated from the marine environment and itwas able to decompose the diatom Skeletonema costatum, whichis responsible for the formation of red tides [53]. In sample wt3,the dominant phylotype (47.7%) was classified in Rhodobacter-aceae (Fig. 3). Its closest relative was the marine bacterium strainATAM407 56 (AF359535) belonging to the genus Phaeobacter, thathas been isolated from cultures of the slightly toxic dinoflagel-late Alexandrium affine NEPCC 607 [22]. Similar phylotypes (>98%)were detected in samples S6m1 and M6m1 (Fig. 2). The domi-nant phylotype (47%) of sample wt4 was grouped in the familyRhodobacteraceae of Alphaproteobacteria. Its closest relative was



Fig. 2. Dominant phylotypes in the mid-gut of Nephrops norvegicus-reared populations. Different colors correspond to different groups of phylotypes, named by the genusof their closest relative. Grouping is based on a cut-off similarity of 97%, except for the relatives of Candidatus Hepatoplasma (<97%). (For interpretation of the references tocolor in this figure legend, the reader is referred to the web version of this article.)

strain Marivita cryptomonadis CL-SK44 (EU512919) that has beenisolated from the marine species Cryptophyta sp. CR-MAL01 [24]. Asimilar phylotype was detected in sample S6m1. Generally, severalphylotypes from the water samples were closely related to phylo-types detected in the gut samples and mostly in the ones from thestarvation group (Fig. 4).

Mussel and pellet samples

The phylotypes detected in the mussels clustered within theAlpha-, Gamma-, Epsilonproteobacteria, Bacteroidetes and Fibrobac-teres, while the ones detected in the pellets clustered mostly withinthe Firmicutes (data not shown). No common phylotypes weredetected between the samples of the feed provided and the mid-gutsamples.

NMDS analysis

The three-dimensional NMDS analysis performed on all waterand mid-gut samples revealed the grouping of all mid-gut samplesthat had been fed with mussels or pellets for 6 months with onesample from the natural populations (Nat2), while samples fromstarved animals and from the animal that had been fed only for3 months (M3m1) were grouped with the water samples (Fig. 5).This grouping was statistically significant, as shown by the ANOSIManalysis (R = 0.613, p = 0.004).

Diversity and similarity of bacterial communities

The three diversity indices gave similar results regardingthe bacterial diversity of the gut samples (Table S3). They allshowed their lowest values in sample P6m2 and the highest



Fig. 3. Neighbor-joining tree based on 16S rRNA gene sequences from wild and reared Nephrops norvegicus. Aquifex pyrophilus was set as an outgroup. The bar corresponds to a10% nucleotide difference and bootstrap values were calculated from 1000 replicate trees. Frequencies of retrieved phylotypes in each clone library are shown in parentheses.

values in sample S6m2. Overall, the indices indicated higherdiversity in the starvation and the M3m1 samples. Neighbor-joining analysis of the samples based on the frequencies of thephylotypes showed similar results to the Morisita similaritiescluster analysis and NMDS analysis (Fig. S2). Neighbor-joininganalysis exhibited a differentiation between Group I (G1: M6m1,M6m2, P6m1, P6m2 and Nat2) and Group II (G2: M3m1, S3m1,S6m1 and S6m2) with Nat1 as an outgroup. The Morisita sim-ilarities cluster analysis exhibited similarities of >0.7 betweenthe members of G1, while the members of G2 were highlydifferentiated with similarities of <0.4 (Fig. S2). This groupingalso proved to be statistically significant after ANOSIM analysis(R = 0.665, p < 0.001).

Discussion

This study analyzed the differences of the gut bacterial com-munities in experimentally reared N. norvegicus individuals whendifferent food sources were provided. Statistical analysis of the gutbacterial diversity showed the presence of two groups dependingprimarily on whether food was provided or not. The G1 samples(Nat2, M6m1, M6m2, P6m1 and P6m2) had lower bacterial diver-sity than the G2 samples (S3m1, S6m1, S6m2 and M3m) (Table S3).All starvation samples were grouped together with M3m1 whilesome of the phylotypes detected in G2 were similar to phylotypesdetected in the water of the tanks (Fig. 4). After 3 months feeding,the bacterial diversity in mussel-fed animals was higher than at thebeginning of the experiment, although specific bacterial commu-nities had still not been established and resembled the starvationsamples more.

Bacterial diversity was lower in the members of G1 and thisdifference was attributed to food provision that helped the estab-lishment of more stable gut bacterial communities. The high gutbacterial diversity in the members of G2 could be attributed eitherto starvation or to the short time (3 months) between the initia-tion of the experiment and the first sampling. An increase in gutbacterial diversity in starved animals has also been observed in thelocust Schistocerca gregaria [12]. However, the exact reasons for thisincrease in bacterial diversity have not been studied in either of thestudies.

Apart from samples Nat1, Nat2 and P6m2, frequencies of specificbacterial phylotypes never exceeded 33.3% in reared populations(P6m1-12). This is different to previous results [36] where gutbacterial communities in N. norvegicus natural populations weredominated (≥58%) by a single bacterial phylotype. Thus, the gutbacterial communities of reared samples after 3 and 6 monthsshowed higher diversity than the communities of the wild ones(Table S3). In the case of the Norway lobster, dominant bacterialdiversities are considered to indicate the performance of specificdigestive functions that assist the host. The decrease of the bacterialdiversity in reared animals could be attributed to the slow estab-lishment of dominant bacterial communities resulting from the lowfood consumption (0.049–0.069 mg/gdry body weight/day) of musselsand pellets (Mente and Karapanagiotidis, unpublished data) com-pared to previous studies (0.025 mg/gbody weight/day) [47].

The phylotypes related to P. leiognathi were practically identicalto those previously detected in the gut of wild Norway lobsters[36]. They were present in all the gut bacterial communities ofG1 samples where mussels and pellets were consumed. Micro-biological studies have proved that most P. leiognathi strains



Fig. 4. Neighbor-joining tree based on 16S rRNA gene sequences from the water of the rearing tanks. Aquifex pyrophilus was set as an outgroup. The bar corresponds to a 10%nucleotide difference and bootstrap values were calculated from 1000 replicate trees. Frequencies of retrieved phylotypes in each clone library are shown in parentheses.

are chitinolytic (96%) and lipolytic (82%) [14]. Recently, the firstcompleted genomic study of the species (NZ BACE00000000),showed the presence of multiple genes coding for lipases,proteases and chitinases (Microbial Genome Resources, 2011;http://www.ncbi.nlm.nih.gov/genomes/MICROBES/microbial taxtree.html). Although no lipase, chitinase or protease tests were

performed in this study, data showing the high similaritiesbetween phylotypes, the dominance of the phylotypes in all G1and wild samples (from this and from previous studies), and theresults from previous microbiologic and genomic studies on thespecies indicated the presence of a specific bacterial communitywith potential positive effects on N. norvegicus digestive function.



Similar P. leiognathi clustering phylotypes were dominant (73%)in the P6m2 sample that had molted 2–3 days before sampling.During molting, the chitinase lining of the hind-gut cuticle is shedtogether with the rest of the exoskeleton [7] and, as a result, thebacterial abundance of the hind-gut decreases [8]. However, boththe role of gut bacteria in the molting procedure and their fateafter molting is as yet unknown. Studies have shown that chitinaseactivity in the gut is increased during the post-molt period (thefirst 2–3 days after molting) [18], since this has been related to therelease of the molt [32,56]. In this case, the inferred chitinolyticactivity of P. leiognathi-related phylotypes could explain theirdominance in the post-molt sample P6m2.

Arcobacter-related phylotypes were frequent and recurringin the gut samples and mostly in the ones fed with mussels(Figs. 2 and 3). Their closest relative was isolated from oyster mantle[49] clustering in the genus Arcobacter. The genus Arcobacter oxi-dizes hydrogen sulfide and produces sulfur [55], while some of thestrains of this species are capable of denitrification [21]. Arcobacter-like bacteria have been found in the intestinal tract of humans andanimals [55], the deep sea [23], lake water columns [48], activatedsludge [21] and marine oysters [44,51]. In this study, the pres-ence of Arcobacter-like phylotypes was mostly associated with theprotein-rich mussels that were used for feeding, and they might beconnected to nitrogen metabolism and the sulfur cycle. Similarly,high rates of denitrification have been detected in previous studies[20] in the gut of the aquacultured shrimp Litopenaeus vannameiand were mainly attributed to the activity of denitrifying bacteria.

The Shewanella-like phylotypes (Fig. 1 and Table 2) were closelyrelated to a protease producing strain. Apart from the potential pro-teolytic function in the gut of the Norway lobster these phylotypeswere similar to the one previously detected (D1-674) in the gut ofN. norvegicus [36] and, thus, may belong to the resident bacterialcommunity of the gut. Their presence is reinforced by their abil-ity to grow solely on leucine, which was abundant on the pelletsprovided as synthetic feed in this experiment [34].

Entomoplasmatales phylotypes were dominant in the 6-monthmussel-fed samples (M6m) and appeared with lower frequencies inother samples (Figs. 2 and 3). Phylotypes from the same cluster haveappeared in the intestine [11,12,13,58] and the hepatopancreas[16] of other invertebrates with functions that are still unknown.In the case of the uncultured Candidatus Hepatoplasma clones,which were detected as closest relatives in our study, they hadbeen related to higher survival rates of their hosts when low qual-ity food was provided [16]. The phylotypes detected in the gut of N.norvegicus were distantly related to the Candidatus Hepatoplasmaclones (<93%) but clearly clustered in different genera, as has beendescribed recently from Yarza et al. [57] when setting genus bound-aries at 94.5% SSU similarities. Thus, a similar relationship betweenhigher survival rates and food quality cannot be assumed.

Regarding the phylotypes detected in the starvation samples,there was no pattern that was present in all of them, accord-ing to sampling time or to specific bacterial communities. Thestarvation phylotypes mostly clustered in the Alphaproteobacteriaand the Bacteroidetes (Figs. 3 and 4, Table S2) and some of themwere closely related to phylotypes detected in the water samples(Fig. 4), suggesting an influence of the gut bacterial communitiesfrom the surrounding environment. The reverse hypothesis hasbeen excluded since all the dominant common phylotypes wererelated to bacteria detected in the marine environment that werenot associated with the digestive tract. Apart from that, some rareS phylotypes (Figs. 1 and 2) were closely related to the dominantPhotobacterium-like and Shewanella-like phylotypes detected in theM6m and P6m samples. However, their occurrence in S sampleswas low (<7%) and occasional.

Although no clear differences between mussel- and pellet-fedanimals were observed, the presence of food seemed to fuel the

Fig. 5. NMDS ordination plot (Bray–Curtis distance matrix) of the phylotype fre-quencies from the Nephrops norvegicus samples (ordination stress = 0.04). Each gutsample is indicated by a dot with different colors (light red: wild populations t0;dark-red: pellet-fed samples; orange: mussel-fed samples; blue: starvation sam-ples; cyan: tank water samples). (For interpretation of the references to color in thisfigure legend, the reader is referred to the web version of this article.)

establishment of dominant microbial communities. Bacterial com-munities seemed to be agitated after transport in the rearing tanks,since starvation samples and the 3-month mussel-fed sample weremore diverse compared to that known concerning gut bacterialdiversity of wild Nephrops from past studies [36] and from thisstudy (Table S3). The dominance of Photobacterium sp., Shewanellasp. and Mycoplasmataceae clustering phylotypes in G1 samples(6 months) seemed to be a result of feeding, since these phylotypes,although potentially resident (as assumed from their concurrencein wild samples), showed a low and random presence in G2 samplesand formed more stable communities after 6 months feeding.

From our findings, combined with previous studies [36], it seemsthat among all the dominant phylotypes detected in this study, P.leiognathi fulfills most of the known criteria [15] for the selectionof a probiotic microorganism. It is non-pathogenic for N. norvegi-cus, forms dominant communities in the gut of wild and rearedN. norvegicus populations, and has potentially positive effects on N.norvegicus digestive function. Thus, it is a very promising candidatefor future use as a probiotic.

This is the first study of gut bacterial communities in rearedN. norvegicus. By analyzing the community changes under differ-ent diets for a rearing period of 6 months, it was shown that foodintake promoted the establishment of specific bacterial commu-nities, which were dominated by species that have been found tooccur previously in natural populations of N. norvegicus, renderingthem possible resident symbionts. From these bacterial species, P.leiognathi appeared as the most promising candidate to be used asa probiotic in future N. norvegicus rearing efforts.

Acknowledgments

We thank Ioannis Karapanagiotidis for his contribution to therearing experiments and Zisis Petmezas, Konstantinos Kroupis,Vaso Kefeke and Maria Sakkomitrou for helping in the rearingexperiments, as well as the histological and physiological anal-yses. The two fishermen from Volos are fully acknowledged fortheir assistance in the collection of the lobsters. AM would also liketo thank the International Max Planck Research School of Marine



Microbiology (MarMic), Bremen, Germany program for supportingpart of this work, and Elmar Pruesse and Tryfonas Farmakakis fortheir assistance in the bioinformatics part of this work.

Appendix A. Supplementary data

Supplementary data associated with this article can befound, in the online version, at http://dx.doi.org/10.1016/j.syapm.2012.07.004.

References

[1] Abello, P., Abella, A., Adamidou, A., Jukic-Peladic, S., Majorano, P., Spedicato,M.T. (2002) Geographical patterns in abundance and population structure ofNephrops norvegicus and Parapenaeus longirostris (Crustacea: Decapoda) alongthe European Mediterranean coasts. Sci. Mar. 66, 125–141.

[2] Ali, A. (2000) Probiotics in fish farming. Evaluation of a bacterial mixture. Thesis,Swedish University of Agricultural Sciences, Umeå, Sweden.

[3] Ashelford, K.E., Chuzhanova, N.A., Fry, J.C., Jones, A.J., Weightman, A.J. (2005)At least 1 in 20 16S rRNA sequence records currently held in public reposito-ries is estimated to contain substantial anomalies. Appl. Environ. Microbiol. 71,7724–7736.

[4] Bell, M., Redant, F., Tuck, I. (2006) Nephrops species. In: Phillips, B. (Ed.), Lob-sters: Biology, Management, Aquaculture and Fisheries, Blackwell, Oxford, pp.412–469.

[5] Clarke, K.R. (1993) Non-parametric multivariate analyses of changes in com-munity structure. Aust. J. Ecol. 18, 117–143.

[6] Clarke, K.R., Green, R.H. (1988) Statistical design and analysis for a biologicaleffects study. Mar. Ecol.-Prog. Ser. 46, 213–226.

[7] Conklin, D.E. (1995) Digestive physiology and nutrition. In: Factor, I.R. (Ed.),Biology of the Lobster Homarus americanus, Academic Press, USA, pp. 441–458.

[8] Crawford, G.S., Minion, G.P., Boyers, M.D. (1983) Intima morphology, bacterialmorphotypes, and effects of annual molt on microflora in the hindgut of thedesert millipede, Orthopus ornatus (Girard) (Diplopoda: Spirostreptidae). Int. J.Insect. Morphol. Embryol. 12, 301–312.

[9] Cristo, M., Cartes, J.E. (1998) A comparative study of the feeding ecology ofNephrops norvegicus (L.), (Decapoda: Nephropidae) in the bathyal Mediter-ranean and the adjacent Atlantic. Sci. Mar. 62, 81–90.

[10] de Figueiredo, M.J., Vilela, M.H. (1972) On the artificial culture of Nephropsnorvegicus reared from the egg. Aquaculture 1 (C), 173–180.

[11] Demiri, A., Meziti, A., Papaspyrou, S., Thessalou-Legaki, M., Kormas, K.A. (2009)Abdominal setae and midgut bacteria of the mud shrimp Pestarella tyrrhena.Cent. Eur. J. Biol. 4, 558–566.

[12] Dillon, R.J., Webster, G., Weightman, A.J., Charnley, A.K. (2010) Diversity of gutmicrobiota increases with aging and starvation in the desert locust. Anton.Leeuw. Int. J. Gen. Mol. Microbiol. 97, 69–77.

[13] Durand, L., Zbinden, M., Cueff-Gauchard, V., Duperron, S., Roussel, E.G., Shillito,B., Cambon-Bonavita, M.A. (2010) Microbial diversity associated with thehydrothermal shrimp Rimicaris exoculata gut and occurrence of a residentmicrobial community. FEMS Microbiol. Ecol. 71, 291–303.

[14] Farmer, J.J., III, Hickman-Brenner, F.W. (2006) The genera Vibrio and Pho-tobacterium. In: Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K.-H.,Stackenbrandt, E. (Eds.), The Prokaryotes, Springer, New York, pp. 508–563.

[15] Farzanfar, A. (2006) The use of probiotics in shrimp aquaculture. FEMSImmunol. Med. Microbiol. 48, 149–158.

[16] Fraune, S., Zimmer, M. (2008) Host-specificity of environmentally transmittedMycoplasma-like isopod symbionts. Environ. Microbiol. 10, 2497–2504.

[17] Grammito, M.E. (1998) Molt pattern identification through gastrolith exam-ination on Nephrops norvegicus (L.) in the Mediterranean Sea. Sci. Mar. 62,17–23.

[18] Gulmann, L.K. (2004) Gut associated microbial symbionts of the marsh fiddlercrab, Uca pugnax. Ph.D. Thesis, Massachusetts.

[19] Hammer, Ø., Harper, D.A.T., Ryan, P.D. (2001) Past: paleontological statisticssoftware package for education and data analysis. Palaeontol. Electron. 4, 1–9.

[20] Heisterkamp, I.M., Schramm, A., de Beer, D., Stief, P. (2010) Nitrous oxide pro-duction associated with coastal marine invertebrates. Mar. Ecol. Prog. Ser. 415,1–9.

[21] Heylen, K., Vanprays, B., Wittebolle, L., Verstraete, W., Boon, N., de Vos, P. (2006)Cultivation of denitrifying bacteria: optimization of isolation conditions anddiversity study. Appl. Environ. Microbiol. 72, 2637–2643.

[22] Hold, G.L., Smith, E.A., Rappe, M.S., Maas, E.W., Moore, E.R.B., Stroempl, C.,Stephen, J.R., Prosser, J.I., Birkbeck, T.H., Gallacher, S. (2001) Characterizationof bacterial communities associated with toxic and non-toxic dinoflagellates:Alexandrium spp. and Scrippsiella trochoidea. FEMS Microbiol. Ecol. 37, 161–173.

[23] Huber, A.J., Johnson, H.P., Butterfield, D.A., Baross, J.A. (2006) Microbial life inridge flank crystal fluids. Environ. Microbiol. 8, 88–99.

[24] Hwang, C.Y., Bae, G.D., Yih, W., Cho, B.C. (2009) Marivita cryptomonadis gen. nov.,sp. nov. and Marivita litorea sp. nov., of the family Rhodobacteraceae, isolatedfrom marine habitats. Int. J. Syst. Evol. Microbiol. 59, 1568–1575.

[25] Irianto, A., Austin, B. (2003) Use of dead probiotic cells to control furunculosisin rainbow trout, Onchorhynchus mykiss. J. Fish Dis. 26, 59–62.

[26] Kruskal, J.B. (1964) Multidimensional scaling by optimizing a goodness of fit toa non-metric hypothesis. Psychometrics 29, 1–28.

[27] Lane, D.J. (1991) 16S/23S rRNA sequencing. In: Stackenbrandt, E., Goodfellow,M. (Eds.), Nucleic Acid Techniques in Bacterial Systematics, John Wiley & Sons,Chichester, pp. 115–175.

[28] Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A.,McWilliam, H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D.,Gibson, T.J., Higgins, D.G. (2007) Clustal W and Clustal X version 2.0. Bioinfor-matics 23, 2947–2948.

[29] Ley, R.E., Lozupone, C.A., Hamady, M., Knight, R., Gordon, J.I. (2008) Worldswithin worlds: evolution of the vertebrate gut microbiota. Nat. Rev. Microbiol.6, 776–788.

[30] Li, K., Guan, W., Wei, G., Liu, B., Xu, J., Zhao, L., Zhang, Y. (2007) Phylogeneticanalysis of intestinal bacteria in the Chinese mitten crab (Eriocheir sinensis). J.Appl. Microbiol. 103, 675–682.

[31] Ludwig, W., Strunk, O., Westram, R., Richter, L., Meier, H., Yadhukumar,Buchner, A., Lai, T., Steppi, S., Jobb, G., Förster, W., Brettske, I., Gerber, S.,Ginhart, A.W., Gross, O., Grumann, S., Hermann, S., Jost, R., König, A., Liss,T., Lüssmann, R., May, M., Nonhoff, B., Reichel, B., Strehlow, R., Stamatakis,A., Stuckmann, N., Vilbig, A., Lenke, M., Ludwig, T., Bode, A., Schleifer, K.H.(2004) ARB: a software environment for sequence data. Nucleic Acids Res. 32,1363–1371.

[32] Lustigman, S., McKerrow, J.H., Shah, K., Lui, J., Huima, T., Hough, M., Brotman,B. (1996) Cloning of a cysteine protease required for the molting of Onchocercavolvulus third stage larvae. J. Biol. Chem. 271, 30181–30189.

[33] Margalef, R. 1958 Temporal Succession and Spatial Heterogeneity in Phy-toplankton (Perspectives in Marine Biology), Univ. Calif. Press, Berkeley/LosAngeles, pp. 323–350.

[34] Mente, E. (2010) Survival, food consumption and growth of Norway lob-ster (Nephrops norvegicus) kept in laboratory conditions. Integr. Zool. 5, 256–263.

[35] Mente, E., Karapanagiotidis, I.T., Logothetis, P., Vafidis, D., Malandrakis, E.,Neofitou, N., Exadactylos, A., Stratakos, A. (2009) The reproductive cycle ofNorway lobster. J. Zool. 278, 324–332.

[36] Meziti, A., Ramette, A., Mente, E., Kormas, K.A. (2010) Temporal shifts of theNorway lobster (Nephrops norvegicus) gut bacterial communities. FEMS Micro-biol. Ecol. 74, 472–484.

[37] Muyzer, G., Teske, A., Wirsen, C.O., Jannasch, H.W. (1995) Phylogenetic-relationships of Thiomicrospira species and their identification in deep-seahydrothermal vent samples by denaturing gradient gel-electrophoresis of 16srDNA fragments. Arch. Microbiol. 164, 165–172.

[38] Mytilineou, C., Sardà, F. (1995) Age and growth of Nephrops norvegicusin the Catalan Sea, using length-frequency analysis. Fish. Res. 23, 283–299.

[39] Nishiguchi, M.K., Nair, V.S. (2003) Evolution of symbiosis in the Vibrionaceae: acombined approach using molecules and physiology. Int. J. Syst. Evol. Microbiol.53, 2019–2026.

[40] Pruesse, E., Quast, C., Knittel, K., Fuchs, B.M., Ludwig, W., Peplies, J., Gloeckner,F.O. (2007) SILVA: a comprehensive online resource for quality checked andaligned ribosomal RNA sequence data compatible with ARB. Nucleic Acids Res.35, 7188–7196.

[41] Ritar, A.J., Smith, G.G. (2008) Reproductive biology and growth ofmarine lobsters. In: Mente, E. (Ed.), Reproductive Biology of Crus-taceans. Case Studies of Decapod Crustaceans, Science Publishers, Enfield,pp. 391–425.

[42] Romanenko, L.A., Tanaka, N., Frolova, G.M., Mikhailov, V.V. (2010) Marinicellalitoralis gen. nov., sp. nov., a gammaproteobacterium isolated from coastal sea-water. Int. J. Syst. Evol. Microbiol. 60, 1613–1619.

[43] Romanenko, L.A., Tanaka, N., Frolova, G.M., Svetashev, V.I., Mikhailov, V.V.(2011) Litoreibacter albidus gen. nov., sp. nov. and Litoreibacter janthinus sp.nov., members of the class Alphaproteobacteria isolated from the seashore. Int.J. Syst. Evol. Microbiol. 61, 148–154.

[44] Romero, J., Garcia-Varela, M., Laclette, J.P., Espejo, R.T. (2002) Bacterial 16S geneanalysis revealed that bacteria related to Arcobacter spp. constitute an abundantand common component of the oyster microbiota (Tiostrea chilensis). Microb.Ecol. 4, 365–371.

[45] Rotllant, G., Charmantier-Daures, M., Anger, K., Sarda, F. (2001) Effects of dieton Nephrops norvegicus (L.) larval and postlarval development, growth andelemental composition. J. Shellfish Res. 20, 347–352.

[46] Rotllant, G., Ribes, E., Company, J.B., Durfort, M. (2005) The ovarian maturationcycle of the Norway lobster Nephrops norvegicus (Linnaeus, 1758) (Crustacea,Decapoda) from the western Mediterranean Sea. Invertebr. Reprod. Dev. 48,161–169.

[47] Sarda, F., Valladares, F.J. (1990) Gastric evacuation of different foods byNephrops norvegicus (Crustacea, Decapoda) and estimation of soft-tissueingested, maximum food-intake and cannibalism in captivity. Mar. Biol. 104,25–30.

[48] Schmidtova, J., Hallam, S.J., Baldwin, S.A. (2009) Phylogenetic diversity of tran-sition and anoxic zone bacterial communities within a near-shore anoxic basin:Nitinat Lake. Environ. Microbiol. 11, 3233–3251.

[49] Schulze, A.D., Alabi, A.O., Tattersall-Sheldrake, A.R., Miller, K.M. (2006) Bacterialdiversity in a marine hatchery: balance between pathogenic and potentiallyprobiotic bacterial strains. Aquaculture 256, 50–73.

[50] Shannon, C.E., Weaver, W. 1949 The Mathematical Theory of Communication,University of Illinois Press, Urbana.

[51] Simpson, E.H. (1949) Measurement of diversity. Nature 163, 688.



[52] Sipkema, D., Schippers, K., Maalcke, W.J., Yang, Y., Salim, S., Blanch, H.W.(2011) Multiple approaches to enhance the cultivability of bacteria associ-ated with the marine sponge Haliclona (gellius) sp. Appl. Environ. Microbiol. 77,2130–2140.

[53] Sohn, J.H., Lee, J.-H., Yi, H., Chun, J., Bae, K.S., Ahn, T.-Y., Kim, S.-J. (2004) Kordiaalgicida gen. nov., sp. nov., an algicidal bacterium isolated from red tide. Int. J.Syst. Evol. Microbiol. 54, 675–680.

[54] Suen, G., Scott, J.J., Aylward, F.O., Adams, S.M., Tringe, S.G., Pinto-Tomas, A.A.,Foster, C.E., Pauly, M., Weimer, P.J., Barry, K.W., Goodwin, L.A., Bouffard, P., Li,L., Osterberger, J., Harkins, T.T., Slater, S.C., Donohue, T.J., Currie, C.R. (2010)An insect herbivore microbiome with high plant biomass-degrading capacity.PLoS Genet. 6, e1001129.

[55] Vandamme, P., Dewhirst, F.E., Paster, B.J., On, S.L.W. (2004) Genus II. Arcobacter,in: Garrity, G.M. (Ed.), Bergey’s Manual of Bacteriology, vol. 2, Springer, MI, USA,pp. 1161–1165.

[56] Vega-Villasante, F., Fernandez, I., Preciado, R., Oliva, M., Torvar, D. (1999) Theactivity of digestive enzymes during the molting stages of the arched swim-ming Callinectes arcuatus, Ordway, 1963 (Crustacea: Decapoda: Portunidae).Bull. Mar. Sci. 65, 1–9.

[57] Yarza, P., Ludwig, W., Euzéby, J., Amann, R., Schleifer, K.-H., Glockner, F.O.,Rosselló-Móra, R. (2010) Update of the All-Species Living Tree Project basedon 16S and 23S rRNA sequence analyses. Syst. Appl. Microbiol. 33, 291–299.

[58] Zbinden, M., Cambon-Bonavita, M.A. (2003) Occurrence of Deferribacterales andEntomoplasmatales in the deep-sea Alvinocarid shrimp Rimicaris exoculata gut.FEMS Microbiol. Ecol. 46, 23–30.

[59] Zhou, M.-Y., Chen, X.-L., Zhao, H.-L., Dang, H.-Y., Luan, X.-W., Zhang, X.-Y., He,H.-L., Zhou, B.-C., Zhang, Y.-Z. (2009) Diversity of both cultivable protease-producing bacteria and their extracellular proteases in the sediments of theSouth China Sea. Microb. Ecol. 58, 582–590.

Gut bacteria associated with different diets in reared Nephrops norvegicus

Documents