Printer Friendly

Investigation into the physiologies of Aeromonas veronii in vitro and inside the digestive tract of the medicinal leech using RNA-seq.


A key component to understanding host-microbe associations is the physiology of the microbial community because it reveals insight into the contribution of the microbiome to the host and the conditions the microbes experience inside the animal (Wier et al., 2010; Petersen et at, 2011). Researchers can reconstruct metabolic pathways and develop testable hypotheses about how the microbial population functions by identifying genes that are expressed during symbiosis. Several techniques can be used to examine gene expression, providing insight into the microbiome's physiology. Reverse-transcription PCR (RT-PCR) or quantitative reverse-transcription PCR (qRT-PCR) are used to investigate differentially expressed genes (Noda et at, 1999). These techniques are useful for examining a few targets; however, they are challenging for hundreds of targets and require a priori knowledge of the exact target sequences to which the primers anneal. These caveats make RT-PCR and qRT-PCR more useful in confirmatory experiments than in gene discovery applications. Clone libraries of mRNA-derived cDNA can also be used to gain access to the symbionts' physiologies and do not require sequenced genomes. For example, Tartar et at (2009) used clone libraries to uncover the digestive contributions of the protists from the hindgut microbiome in the termite Reticulitermes flavipes. Though valuable, construction of cDNA clone libraries for prokaryotes is more challenging than for eukaryotes due to the lack of a poly-A tail, and cloning can introduce bias into the resulting data sets. Compared to next-generation sequencing approaches, clone library sequencing is low-throughput and not quantitative.

Traditionally, researchers have used microarrays to assess global gene expression in microorganisms, and this technique has been successfully applied to many host-associated microbiomes such as the gut microbiota of gnotobiotic mice and Vibrio fischeri in the squid light organ (Sonnenburg et at, 2005; Martens et at, 2008; Wier a at, 2010). However, the application of microarrays is not as reliable for all host-associated microbiomes, particularly for field-caught animals whose symbiont communities can vary in composition. For instance, natural microbial communities can contain allelic variation, and transcripts from these variants may not be detected or may introduce an artificially lower expression level if the oligonucleotide probes do not match the transcript. The pangenome, which is the collective gene pool found within a particular species of prokaryotes, poses another challenge. Genome sequencing projects of strains belonging to the same bacterial species have shown for some species that as little as 40%-50% of a given genome is conserved between all strains examined. For example, a study that compared Escherichia coli genomes found that only about 2200 genes out of the approximately 4238 to 5589 genes encoded on each genome were conserved in the 17 genomes examined (Rasko et al., 2008). If different strains of the same species are present and the pangenome is unknown, entire genes or genomic islands can be missed using microarrays. This problem is magnified if not the species but the function of the species in the association or guild is preserved, which is often the case in the digestive-tract microbiome of mammals (Dethlefsen et al., 2007).

The advent of high-throughput sequencing in 2005 provided researchers with new opportunities to explore the metabolic potential of microorganisms by sequencing microbial genomes and metagenomes at a very high coverage (Margulies et at, 2005). The application of high-throughput sequencing to cDNA constructed from RNA transcripts, known as RNA-seq, facilitated investigations into physiologies of naturally occurring microbial communities such as those found in the ocean (Frias-Lopez et at, 2008; Gilbert et at, 2008) and in the soil (Urich et at, 2008). More recently, RNA-seq has been used to characterize the metatranscriptomes of host-associated microbial communities. Examples include the digestive-tract microbiomes of humanized mice using RNA from cecal contents (Turnbaugh et at, 2009), and cecal contents of pigs (Poroyko et at, 2010) and of humans using fecal samples as a proxy (Turnbaugh et at, 2010). RNA-seq has also been used to study in vivo gene expression of the intracellular symbionts of bivalves (Stewart et al., 2011). In contrast to microarrays, RNA-seq does not require a priori knowledge of community composition or genomes, though this information greatly aids in experimental design and data analysis. A challenge for this approach is the difficulty of removing rRNA prior to cDNA library preparation. rRNA often constitutes well over 90% of total RNA in a cell (Neidhardt, 1996), and the lack of a poly-A tail on mRNA in bacteria presents a challenge to enriching mRNA. However, mRNA enrichment methods that utilize oligonucleotides or enzymes that target rRNA are available (Yoder-Himes et al., 2009; Poretsky et al., 2009). In addition, the high-through-put and reduced sequencing costs of the Illumina platform (Luo et al., 2012) can overcome this difficulty.

We recently used RNA-seq to examine symbiont physiology within the digestive tract of the medicinal leech, Hirudo verbana (Bomar et al., 2011). The medicinal leech serves as a model for naturally occurring digestive-tract symbioses (Graf et at, 2006; Nelson and Graf, 2012). In contrast to most mammalian digestive tracts, the crop, the largest compartment of the leech gut, is colonized by two dominant symbionts. One of the symbionts, Aeromonas veronii, is a gram-negative, facultative anaerobe that is often cultured from aquatic environments and is also a human pathogen (Janda and Abbott, 2010). Interestingly, our initial RNA-seq study suggested that the A. veronii population in the crop was not homogeneous but exhibited at least two distinct physiologies at the same time: one of free-swimming or pelagic cells and one in mixed-species microcolonies with the second dominant symbiont, a recently cultured Rikenella-like bacterium (Worthen a al., 2006; Kikuchi and Graf, 2007; Bomar a al., 2011).

RNA-seq studies can also reveal the presence of previously unrecognized noncoding RNAs that can be highly abundant in bacterial cells (Shi et at, 2009). One such system is the CsrB family that is involved in global gene regulation. The CsrB family contains regulatory RNAs that bind to and sequester CsrA, an RNA binding protein that post-transcriptionally regulates gene expression (Babitzke and Romeo, 2007). Binding of CsrA by CsrB is achieved through a series of GGA repeats (Babitzke and Romeo, 2007). In various bacterial genera, the Csr system regulates carbon metabolism, biofilm formation, secondary metabolite production, quorum sensing, and stationary-phase gene expression (Babitzke and Romeo, 2007).

In the current study, our goal was to gain a better understanding of the differences in A. veronii's physiology in vitro compared to in vivo. We used RNA-seq to gain insight into A. veronii's physiology during exponential growth in a rich medium and compared it to the physiology inside the leech 42 h after ingestion of a blood meal. The data suggest that CsrB is highly expressed, Aeromonas degrades arginine, and bacteria experience extracytoplasmic stress.

Materials and Methods

Sample collection for harvesting RNA

Leech crop samples. Two to four Hirudo verbana specimens (Leeches USA, Westbury, NY) per time point were fed heparinized sheep blood (Quad Five, Ryegate, MT) and maintained at room temperature for 2 h, 4 h, 8 h, 24 h, 42 h, 96 h, or 14 d. Leeches were dipped in 70% ethanol and sacrificed; 50-[micro]l aliquots of the intraluminal fluid (ILF) were then harvested and flash-frozen in liquid nitrogen. In vitro grown A. veronii Hm21, an A. veronii strain isolated from the medicinal leech, was cultivated in lysogeny broth, LB, at 30 [degrees]C at 200 rpm (Sambrook and Russell, 2001); Aliquots of cells representing different phases of growth were collected and flash-frozen in liquid nitrogen. Samples were stored at -80 [degrees]C until further use.

In vivo growth curve of native Aeromonas

Three to eight leeches per time point were fed heparinized sheep blood and maintained at room temperature for 2 h, 4 h, 8 h, 24 h, 42 h, or 96 h. Leeches were sacrificed as described above, and the ILF was harvested. ILF was serially diluted in 0.85% (wt/vol) NaCl and plated on LB agar. Plates were incubated at 30 [degrees]C overnight, and colonies were enumerated to determine the level of native Aeromonas colonization in the crop.

RNA extraction

Total nucleic acid was extracted from ILF or in vitro grown Hm21 using the MasterPure RNA purification kit (Epicentre, Madison, WI). For the ILF samples, the whole-blood protocol was followed except that samples were treated with 50 [micro]g of proteinase K in 300 [micro]l of tissue and cell lysis solution prior to nucleic acid precipitation. For the in vitro samples, the protocol for cell samples was followed. DNA was removed using the Turbo DNA free kit (Ambion, Austin, TX), and DNA contamination was tested for using Aeromonas-specific primers RpoBF1/RpoBR1 and the DNAse-treated RNA as template, as described previously (Bomar et al., 2011).

cDNA library preparation and sequencing

Total RNA was quantified with the NanoDrop ND-1000 (Thermo Scientific, Wilmington, DE), and integrity was analyzed using an RNA FlashGel (Lonza, Basel, Switzerland). Ten micrograms of equally pooled, total RNA from the leech and 10 [micro]g of total RNA from the in vitro samples were enriched for mRNA using mRNA-only (Epicentre, Madison, WI). Enriched RNA was quantified using Ri-boGreen (Invitrogen, Carlsbad, CA). One hundred nanograms each of the in vitro and in vivo enriched RNA were used for library preparations using SuperScript II (Invitrogen, Carlsbad, CA) and an mRNA-seq kit (Illumina, San Diego, CA) following the manufacturer's instructions, except that the poly(T) bead enrichment step was omitted. The cDNA libraries were submitted to the Translational Genomics Core at the University of Connecticut Health Center for cluster generation and 1 X 76 by sequencing.

Data analysis

cDNA reads were mapped against the draft genome of A. veronii strain HM21 using CLC Genomics Workbench (ver. 4.0.2; Aarhus, Denmark). Reads that mapped to more than one region of the draft genome or mapped with more than two mismatches to the draft genome were excluded from the analysis. An expression value (EV) was calculated for each gene using the following formula: EV = [(number of reads mapped to the gene)/(length of the gene in kilobases) x (total number of reads mapped in millions)].

Nucleotide sequence alignment of CsrB sequences

Nucleotide sequences representative of the CsrB family were aligned in Geneious Pro (ver. 5.5.6; Auckland, New Zealand) using MUSCLE (Edgar, 2004).

Colony PCR screen for csrB and csrC

The primer pairs CsrBF1/CsrBR1 and CsrCF1/CsrCR1 were used for the amplification of csrB and csrC, respectively. Primer sequences are listed in Table 1 and the following amplification protocol was used: (i) 120 s at 95 [degrees]C; (ii) 30 cycles of 30 s at 95 [degrees]C, 30 s at 50 [degrees]C, 30 s at 72 [degrees]C; (iii) 120 s at 72 [degrees]C. Reactions contained 1X GoTaq Green MasterMix (Promega, Madison, WI), 1 [micro]mol [1.sup.-1] each of the appropriate forward and reverse primers, and 1 [micro]l of a 1:20 colony dilution in a final reaction volume of 20 [micro]l. The specificity of the primers was confirmed by sequencing the PCR amplicons generated using A. veronii wildtype genomic DNA as template, as previously described (Silver et at, 2007).

Table 1 Primers used in this study

Primer   Sequence (5' to 3')
















Quantitative, reverse-transcription PCR

Double-stranded cDNA synthesis was carried out as described above using reagents from the Illumina mRNA-seq kit or New England Biolabs (Beverly, MA). Copy number of target genes was quantified using 1 x SsoFast EvaGreen Supermix (Biorad, Hercules, CA), 500 nmol [1.sup.-1] each of the appropriate forward and reverse primers, and 1 [micro]l of a 10-fold dilution of cDNA in a final reaction volume of 12.5 [micro]l. The following primer pairs were used for amplification of qPCR targets: CsrBF2/CsrBR1, CsrCF1/R1, RpoDF1/ R1, RpoEF1/R1, PspA2F1/R1, ArcAF1/ArcAR1, and ArcBF1/ArcBR1. All primer sequences are listed in Table 1. The following amplification protocol was used for all primer pairs: (i) 30 s at 95 [degrees]C; (ii) 35 cycles of 5 s at 95 [degrees]C, 15 s at 50 [degrees]C. Primer specificity was confirmed by sequencing either amplicons generated in standard PCR using A. veronii genomic DNA as template or cloned amplicons. A melting curve analysis was also done where the dwell temperature increased from 50 [degrees]C to 95 [degrees]C in 0.5 [degrees]C increments every 10 s. The [2.sup.[DELTA][DELTA].C]T method was used to calculate relative gene expression. Primer efficiency was evaluated using the standard curve method and ranged between 98% and 101%.

RNA-seq data accession numbers

Raw and processed RNA-seq data files were deposited in Gene Expression Omnibus at the NCBI under accession number GSE38068.


Sequencing Aeromonas veronii's transcriptome in culture and inside the leech gut

Genes that were expressed by A. veronii during exponential growth in a rich medium and inside the medicinal leech digestive tract were identified using 11lumina RNA-seq. For in vitro growth of A. veronii we cultivated a medicinal leech isolate, Hm21 (Graf, 1999) under typical laboratory conditions in a rich medium (200 rpm at 30 [degrees]C in LB). Under these conditions, A. veronii doubles approximately every 30 min (data not shown). A. veronii was sampled in mid-log phase. For in vivo growth we sampled the leech intraluminal fluid 42 h after the animal had ingested a blood meal. At this time, A. veronii is still actively proliferating, although a portion of the population is entering stationary phase (Kikuchi and Graf, 2007; Bomar et at, 2011). By this time the host has modified the blood meal by the removal of water and osmolytes (Sawyer, 1986), and hemocyte-like cells are patrolling the crop, phagocytosing sensitive bacteria (Silver et al., 2007). We have previously examined only the leech crop microbiome 42 h after feeding using RNA-seq (Bomar et at, 2011), using a different method to remove ribosomal RNA before library preparation. In the previous study, we utilized an oligonucleotide-based approach that was not very efficient (Table 2), and in the present study we utilized an enzymatic mRNA-enrichment method in an effort to increase our coverage of non-ribosomal transcripts.

Table 2 Summary of RNA-seq datasets and mRNA enrichment efficiency

Sample    Enrichment         No. reads  Total no.  No. reads  No. reads
          method                            reads  mapped to  mapped to
                                           mapped    23S (%)    16S (%)

Crop      Oligonucleotide   15,108,840  6,934,692   4,6419O9  2,045.615
                                                        (67)       (30)

Crop      Enzymatic         11,107,024  3,449,889  2,354,3O4  1,043,099
                                                        (68)       (30)

Lysogeny  Enzymatic          5,500,677  3,970,130  3,054,529    711,176
broth                                                   (77)       (18)

Sample       No.  No. reads   Reference
           reads  mapped to
          mapped   non-rRNA
           to 5S

Crop      25,310    246,388   Bornar et
           (0.4)        (a)   al., 2011

Crop         780     52,512  This study

Lysogeny   2,597    201,558  This study
broth     (0.07)

(a) This value is different than the value previously reported (Bomar
et al., 2011): in the previous study we reported the number of reads
that mapped to coding regions, hut here we report the number of reads
that map to noncoding regions as well.

The cDNA libraries were sequenced using an Illumina platform, generating 5,500,677 and 11,107,024 reads for the in vitro and in vivo libraries, respectively (Table 2). Approximately 201,558 of the in vitro library reads and 52,514 of the in vivo library reads mapped to non-ribosomal regions of Hm21's genome. For the in vivo library, the lower percentage of the reads mapping to the genome of Hm21 is largely due to reads mapping to the genome of the other dominant symbiont, the Rikenella-like bacterium. Overall, 95% and 98% of the reads were ribosomal for the in vitro and in vivo libraries, respectively, indicating that the mRNA-enrichment method we used was not very efficient. However, the sequencing depth of Illumina enabled us to overcome this issue because we were still able to detect cDNA reads that originated from protein-coding transcripts.

Carbon storage regulator (Csr) system

A powerful aspect of having both in vitro and in vivo data sets for A. veronii is the ability to identify genes that are differentially expressed between the two conditions. The most highly expressed gene in vivo, csrB, is expressed 65-fold lower in vitro and was originally annotated as a hypothetical protein in Hm21's draft genome (Table 3). Prior to discovering the identity of this gene, we queried the nucleotide sequence against the non-redundant GenBank database using BLASTN (Altschul et al., 1990), and the closest match corresponded to a genomic region in A. hydrophila PPD134/91 (83% identity, 5 X [10.sup.-114] expected value, accession number AY378289). This region was identified during a genomic subtraction screen for virulence loci in A. hydrophila strains that used a non-virulent A. hydrophila strain to remove DNA of high similarity from the fish pathogen (Yu et aL, 2005). Our gene of interest, which is 469 by in length, has 203 by that overlap with an open reading frame (ORF) predicted to encode a 115 amino acid protein of unknown function. Because of its high expression and its implication in virulence, we investigated the highly expressed gene further.

Non-coding RNAs (ncRNAs), which are RNAs that do not encode for proteins, are often short in length and can be highly abundant (Shi et at, 2009), similar to the highly expressed gene that we observed. Therefore, we queried the nucleotide sequence of the gene against the RFam database (Gardner et al., 2011). On the basis of the results, this gene is predicted to be a CsrB family ncRNA (1.8 X [10.sup.-13] expected value), and thus we refer to this ncRNA as CsrB. In A. veronii's CsrB we found 40 GGA repeats, 31 of which are preceded by an A residue or AN residues. This sequence structure is similar to that found in other CsrB family members. Bacterial genomes can possess functionally redundant CsrBs (Lenz et at, 2005), so we queried Hm21's genome with CsrB's nucleotide sequence using BLASTN. This analysis revealed a 20-nucleotide stretch that was 100% identical to the query sequence. Further examination of the sequence flanking the alignment revealed a series of 16 GGA repeats, 13 of which are preceded by an A residue or AN residues. Based on the alignment with CsrB and other CsrB family members from the Rfam database and from A. salmonieida (Fig. 1), we annotated this newly identified, second ncRNA of the draft genome as CsrC. Similar to CsrB, CsrC expression is 50-fold higher during symbiosis than in vitro (Table 3). A BLASTN search of Hm21's genome using the nucleotide sequence of CsrC did not reveal additional CsrB family members, nor did querying the Rfam database (version 10.1) with the Hm21 genome. As this is only a draft genome, we cannot exclude the possibility that another CsrB homolog is present in the genome.

Table 3 Invitro and in vivo expression values for select genes

Predicted       Accession no.  Expected value  Expected
product                   (a)             (b)     value

                                                Culture       Leech


RpoD                YP 855378               0       728         628

RpoB            YP_0043943791               0      3472         595


CsrA             YP_004394252          6.83 x       471         334

CsrB (f)                  --            --     9875       6.5 x

CsrC (f)                  --            --     7947       4.0 x


ArcA                YP_858517               0         0        2478

ArcB                YP_858515               0         0        1160

ArcC            YP_0043 94442               0         7        1632

ArcD             YP_001140165               0         5        1031


RpoE             YP_004391428          8.00 x       456        3954

RpoA1            YP_004392261          5.88 x         0        3303

PspA2            YP_004391182          5.64 x         0        4421

PspB             YP_004392262          4.90 x        27        1596

PspC             YP_004392263          4.00 x        16        1634

Predicted       Fold change
product                 (d)


RpoD                   -1.2

RpoB                   -5.8


CsrA                   -1.4

CsrB (f)               65.8

CsrC (f)               50.1


ArcA             [infinity]

ArcB             [infinity]

ArcC                  233.1

ArcD                  206.2


RpoE                    7.2

RpoA1            [infinity]

PspA2            [infinity]

PspB                   59.1

PspC                  102.1

(a) GenBank accession number for the top informative BLASTX hit at the
(b) Expected value based on querying the nonredundant database at the
(c) Expression value calculated as (number of reads mapped to the
gene)/(length of the gene in kilobases) X (total number of reads
mapped in millions).
(d) The change in gene expression from in vitro to in vivo.
Calculated as (Leech EV/Culture EV). If the value is less than one,
the negative inverse of the value is reported.
(e) Housekeeping genes were included in the table for comparison.
(f) Dash (--) in a column indicates that the gene of interest
encodes an ncRNA and that BLASTX information is not relevant.

Because a genomic region containing csrB was detected during a screen for virulence loci in A. hydrophila, we were interested in the prevalence of csrB and csrC in other aeromonads. We screened 17 strains of A. veronii, including clinical and leech isolates, and 9 additional Aeromonas species for csrB and csrC using colony PCR (Table 4). The screen revealed the presence of PCR products for both genes in all the A. veronii isolates tested, including those that cannot colonize the medicinal leech to normal levels, suggesting that these genes do not account for the observed competitive dominance of leech isolates over strains isolated from other sources (Silver et al., 2011). In our screen for csrC, amplicons of the expected size were detected in all of the nine Aeromonas species tested. csrB amplicons of the expected size were also detected for all nine Aeromonas species tested; however, amplification of A. hydrophila and A. caviae template produced an additional amplicon of slightly larger size. The PCR primers used for this screen were optimized using Hm21 DNA as template, and the additional amplicons could be due to nonspecific amplification of A. hydrophila and A. caviae template DNA. Over-all, these particular ncRNAs appear to be widely conserved in the Aeromonas genus.

Table 4 Aeromonas strains used for csrB and csrC screen and PCR results

Species               Strain      Isolation       PCR result
source                                            (csrB/csrC)

Aeromonas veronii     AER397      Blood           +/+

A. veronii            AMC34       Human feces     +/+

A. veronii            AMC35       Wound           +/+

A. veronii            CDC0437-84  Fish            +/+

A. veronii            AER28       Feces           +/+

A. veronii            AER39       Blood           +/+

A. veronii            AMC22       Feces           +/+

A. veronii            AMC23       Finger wound    +/+

A. veronii            AMC24       Feces           +/+

A. veronii            AMC25       Duck            +/+

A. veronii            AMC26       Foot wound      +/+

A. veronii            HV221       Hirudo verbana  +/+

A. veronii            HV231       H. verbana      +/+

A. veronii            HV241       H. verbana      +/+

A. veronii            HM21        H. verbana      +/+

A. veronii            Ho635       H. orientalis   +/+

A. veronii            LMG13695    Feces           +/+

A. allosaccharophila  LMG140549T  Eel             +/+

A. hydrophila         ATCC7966T   Canned milk     # (a)/+

A. bestiarum          ATCC14715   Silver salmon   +/+

A. salmonicida        CDC434-84   Fresh water     +/+

A. caviae             ATCCI5468   Guinea pig      # (a)/+

A. sobria             CIP7433     Fish            +/+

A. jandaei            ATCC49568T  Feces           +/+

A. encheleia          LMG16330T   Eel             +/+

A. trota              ATCC49657T  Feces           +/+

(a) Indicates that an additional PCR amplicon was generated using
this template.

We validated the RNA-seq expression levels using qRT-PCR to quantify csrB and csrC expression over time in vitro and in vivo. We examined five time points in vitro that ranged from early exponential growth ([OD.sub.600nm], 0.158) to late stationary phase ([0D.sub.600nm] 5.34) (Fig. 2B). From early exponential growth to late stationary phase, csrB and csrC expression increased approximately 29-fold and 52-fold, respectively (Fig. 2A).

The expression of these genes was also measured in vivo. Because csrB and csrC expression has been reported to change depending on an organism's growth phase, it was important to determine the growth rates of the native Aeromonas community at the time points we sampled: 2 h, 4 h, 8 h, 24 h, 42 h, 96 h, and 14 d after feeding. Figure 2C shows a growth curve for native Aeromonas colonizing the leech gut at some of the corresponding sampling times. The data suggest that A. veronii is rapidly proliferating in the crop, perhaps doubling as fast as every 60 min at the earliest time points and slowing down to a doubling time of more than 250 min between 8 and 24 h after feeding. There are two important caveats to these measurements: the starting number of A. veronii in each animal is unknown, and the animal removes water from the crop to concentrate the blood, which accounts for about a 50% weight loss during the first 48 h after consuming a blood meal. These data are consistent with a doubling time of 70 min during the first 12 h that we determined previously with an A. veronii strain that was introduced with a blood meal (Graf, 1999). The expression of csrB and csrC is below the limit of detection 2 h after feeding, but expression is detected at 4 h after feeding (Fig. 2A). Surprisingly, csrB and csrC expression is high 8 h after feeding (Fig. 2A), a time when the bacteria are still increasing in cell number (Fig. 2C). The observed expression levels are comparable to what was measured in late stationary phase under laboratory conditions when A. veronii is no longer proliferating. From 8 h onward, the expression of these two genes remains high.

Aeromonas veronii utilizes arginine as a nutrient source within the crop

An important question in all host-microbe associations is "what is the nutrient source for the microbiome?" In our previous report, we provided evidence for A. veronii's in vivo catabolism of acetate and fatty acids via the glyoxylate shunt of the citric acid cycle and for the utilization of carbohydrates via the Embden-Meyerhoff Parnas pathway (Bomar et al., 2011). Here, we present evidence that A. veronii also utilizes arginine in vivo. This hypothesis is supported by expression of arcA, arcB, arcC, and arcD, genes predicted to encode an arginine deiminase, an ornithine carbamoyltransferase, a carbamate kinase, and an arginine/ornithine antiporter, respectively (Table 3). qRT-PCR was used to verify expression of select arginine catabolism-related genes in mid-log phase ([OD.sub.600nm], = 0.301) and in the crop at 42 h after feeding (Fig 3). As expected, the expression of arcA and arcB was elevated in the crop relative to in vitro, suggesting that arginine is utilized as a nutrient source in vivo.

A. veronii experiences extracytoplasmic stress inside the leech digestive tract

Another interesting feature revealed by the transcriptome data is that several stress-response-related genes are highly expressed during symbiosis and are expressed at a lower level or not detected during growth in vitro (Table 3). Most notable is the elevated expression of the alternative sigma factor, rpoE, in vivo. Bacteria use alternative sigma factors to direct the RNA polymerase to bind different sets of promoters, allowing them to adapt to changing environmental conditions. In bacteria, rpoE expression is induced when cells experience extracytoplasmic stress. This stress can be the result of many factors including changes in osmolarity, high temperatures, or exposure to antimicrobial compounds or reactive oxygen species (Rowley et al., 2006). In vivo, rpoE expression is 7.2-fold higher than in vitro. In addition, in vivo rpoE expression is 6.3-fold greater than the expression of the housekeeping gene sigma factor rpoD in vivo. For comparison, in vitro rpoE expression is about 1.3-fold lower than in vitro rpoD expression. Phage shock genes (pspA, pspB, and pspC) are also highly expressed in the crop relative to in vitro (Table 3). Similar to rpoE, the psp genes have been implicated in the extracytoplasmic stress response. Depending on the microorganism and the environment, pspA gene expression can be induced by a number of stimuli including osmotic stress, high temperatures, loss of proton motive force, exposure to organic solvents, stationary phase growth, and filamentous phage infection (Rowley et al., 2006). Again, we used qRT-PCR to verify the expression of select stress-response-related genes, pspA2 and rpoE (Fig. 3). pspA2 and rpoE were expressed about 7087-fold and 62-fold higher in vivo, validating our transcriptome findings that these genes are upregulated in the leech gut.

A comparison of in vivo data sets

We compared the current in vivo data set to a previously published data set from the same time point, in which an alternative method for mRNA enrichment was used (Bomar et al., 2011). For the comparison, we calculated gene expression values either including or excluding the highly expressed ncRNAs CsrB and CsrC (Table 5). The reason for this was that between 24.5% and 38.5% of reads mapped to the csrB and csrC genes, and a small difference in the number of reads mapping to those genes between the two samples would disproportionally affect the expression values of all other genes in the data set because the expression value is normalized to the total number of reads that mapped to the reference genome (Table 5). In the data set generated using the oligonucleotide method for rRNA removal, csrB and csrC accounted for a higher proportion of the reads; therefore the gene expression values in this data set are typically lower, potentially skewing the data set comparisons. In both analyses the expression values for most of the genes differ in expression by less than 2-fold (Table 5). The discrepancies that do exist between the two data sets could be due to animal-animal variation, differences in library coverage, or the different methods used to remove ribosomal RNA. Overall, the comparison of the two data sets suggests that the gene expression values obtained are consistent and independent of the rRNA removal method.

Table 5 Comparison of gene expression values from in vivo datasets
generated using different mRNA enrichment methods

Predicted product            Analysis including ncRNAs      Analysis

                       EV (a) Enzyme    EV Oligo    Fold     EV      EV
                                                 change  Enzyme   Oligo


RpoD                            628         311    -1.9   1,056     708

RpoB                            595         363    -1.6   1,000     826

Rps19                         6,214       3,045    -2.0  10,449   6,926


CsrA *                          334         874     2.6     561   1,988

CsrB                          6.5 x       8.1 x     1.3     --     --
                         [10.sup.5]  [10.sup.5]

CsrC                          4.0 x       7.0 x     1.8     --     --
                         [10.sup.5]  [10.sup.5]

Nutrient utilization

ArcA                          2,478       1,391    -1.8   4,167   3,165

ArcB                          1,160         998    -1.2   1,951   2,270

ArcC                          1,632         535    -3.1   2,744   1,216

ArcD                          1,031         373    -2.8    1733     849

AceA                            828         385    -2.1   1,393     876

AceB *                          137          66    -2.1     231     150

Pfk *                           195          61    -3.2     327     138

Fbp *                           152         154     1.0     256     350

SucB *                          211         123    -1.7     356     282

CcoN *                          332         132    -2.5     558     301


RpoE                          3,954       2,939    -1.3   6,648   6,686

PspAl                         3,303       2,542    -1.3   5,555   5,781

PspA2                         4,421       3,959    -1.1   7,435   9,006

PspB                          1,596       1,319    -1.2   2,684   3,001

PspC                          1,634       1,238    -1.3   2,748   2,816

RpoS                            769         887     1.2   1,293   2,018

Stationary phase             13,834      17,861     1.3  23,262  40,628



RpoD                    -1.5

RpoB                    -1.2

Rps19                   -1.5


CsrA *                   3.5

CsrB                      --

CsrC                      --

Nutrient utilization

ArcA                    -1.3

ArcB                     1.2

ArcC                    -2.3

ArcD                    -2.0

AceA                    -1.6

AceB *                  -1.5

Pfk *                   -2.4

Fbp *                    1.4

SucB *                  -1.2

CcoN *                  -1.9


RpoE                     1.0

PspAl                    1.0

PspA2                    1.2

PspB                     1.1

PspC                     1.0

RpoS                     1.6

Stationary phase

Rmf                      1.7

(a) EV, expression value calculated as [(number of reads mapped to the
gene)/(length of the gene in kilobases) x (total number of reads mapped
in millions)].
(b) The change in gene expression between the enzymatically
treated RNA-seq data set and the oligonucleotide-based treatment of the
previously published RNA-seq data set.
Calculated as (Crop EV Oligo/Crop EV Enzyme). if the value is less
than one, the negative inverse of the value is reported.

* Indicates there is less than 1X coverage of this gene in one or both
of the RNA-seq data sets.


In this study, we used high-throughput sequencing of RNA transcripts to characterize the transcriptomes of Aeromonas veronii during exponential growth in a rich medium and inside the digestive tract of the medicinal leech. Our goal was to gain further insight into the physiology of A. veronii within the host by identifying genes that are upregulated during symbiosis. The availability of both in vitro and in vivo data sets helped us determine the microbe's physiological state inside the host.

Utilizing both data sets, we identified two previously unannotated ncRNAs, csrB and csrC, that are highly expressed during symbiosis and appear to be conserved among aeromonads. We verified the expression of these genes in vitro and in vivo over time and found that their expression increases as cells approach stationary phase. This is consistent with previous reports that demonstrate that csrB and csrC expression is induced as cells approach stationary phase (Babitze and Romeo, 2007). Because of the large number of genes that have been reported to be regulated by the Csr system, it is difficult to predict the functional role of these ncRNAs in A. veronii during gut colonization. However, the Csr system has been reported to play a role in host colonization in a range of microorganisms (Lucchetti-Miganeh et al., 2008), and an interesting future direction will be to investigate the role of that system in A. veronii colonization of the medicinal leech gut by constructing mutants.

The transcriptome data also provided new insight into the microorganism's nutrient sources within the crop. Data indicated that arginine is being utilized in vivo. We verified the expression of select genes involved in the catabolism of arginine and found that expression of arcA and arcB was elevated in the crop. Surprisingly, we detected in vitro expression of arcA and arcB, though the levels were close to our limit of detection. This was in contrast to the transcriptome data in which no expression was detected, suggesting that qRT-PCR with gene-specific primers is more sensitive than our transcriptome data. This difference could be due to the different priming method (random hexamers) used in the cDNA synthesis reactions for the transcriptome libraries or to an insufficient sequencing depth. The degradation of arginine through the combined activities of ArcA, ArcB, and ArcC is known as the arginine deiminase pathway and results in the production of ATP, ammonia, carbon dioxide, and ornithine, which is excreted through AreD down a concentration gradient (Cunin et al., 1986). Arginine is a poor nutrient source, and its catabolism results in one net mole of ATP per mole of arginine catabolized. The utilization of this nutrient source suggests that the A. veronii population has depleted more energetically favorable nutrients and could explain why the population begins to level off (Kikuchi and Graf 2007; Bomar et al., 2011). In addition, degradation of arginine is induced during anaerobiosis (Cunin et al., 1986), suggesting that there are anoxic conditions in the crop. The release of ammonia could also raise the pH of the intraluminal fluid, counteracting the acidification during fermentation. Several microbes have been found to use arginine as a growth substrate in the laboratory, and there is evidence for microbes utilizing arginine during host association (D'Hooghe et al., 1997; Fulde et al., 2011). Given the nutrient complexity of the blood meal and the provision of mucin by the leech epithelium, it is not surprising that A. veronii is able to use a variety of nutrient sources in the leech gut for growth or maintenance. One would expect the symbiont to have redundant mechanisms for nutrient utilization. A temporal analysis of nutrient utilization inside the animal would provide interesting insights into the nutritional preferences in vivo.

A striking feature of the transcriptome data was the upregulation of stress-response-related genes in vivo, particularly those indicative of extracytoplasmic stress, such as rpoE and the psp genes. We verified the expression of pspA2 and rpoE in both in vitro and in vivo samples and found that expression of both genes was elevated in the crop. In contrast to the transcriptome data, these data showed in vitro expression of pspA2, suggesting that qRT-PCR with gene-specific primers can detect expression of genes missed in the RNA-seq data. The importance of rpoE for host colonization has been reported in other animal models. For example, Xenorhabdus nematophila rpoE mutants have a decreased ability to colonize the intestinal vesicle of the entomopathogenic nematode (Heungens et at, 2002). rpoE is also required for pathogenic microbes such as Vibrio cholerae (Kovacikova and Skorupski, 2002) and Salmonella typhinzurium to colonize mice (Humphreys et al., 1999). The extracytoplasmic stress experienced by A. veronii could be due to a number of factors including osmotic stress. The osmotic pressure of mammalian blood is approximately 300 mosm/1 (Sawyer, 1986). Already during feeding, the leech removes water and osmolytes from the blood meal. By 2 d after feeding, the ingested blood meal is isosmotic with the hemolymph, which is about 200 mosm/1 (Sawyer, 1986; Zebe et al., 1986). This modification would expose the resident crop bacteria to changes in osmolarity, which likely causes osmotic stress for the bacteria. Another possible cause for the extracytoplasmic stress could be host antimicrobial peptides or antimicrobial proteins such as bactericidal permeability-increasing proteins, which often function by disrupting the bacterial cell membrane. The role of rpoE in resistance to antimicrobial peptides, such as polytnixin B, has been demonstrated in S. typhimurium in vitro (Humphreys et at, 1999). Antimicrobial peptides have been detected in the medicinal leech nervous system (Schikorski et at, 2008) and the intestinal epithelium of a distantly related leech species, Theromyzon tessulatum (Tasiemski et at, 2004). It seems likely that there are also antimicrobials in the medicinal leech gut, and that these could induce extracytoplasmic stress in A. veronii. A temporal transcriptome study of host and symbiont genes would provide important insight into these interactions.

Direct RNA sequencing techniques, such as RNA-seq, have been slow to be used in bacteria as compared to in eukaryotes because of the lack of a poly-A tail, which greatly hinders the enrichment of mRNA because it accounts for only a small percentage of the total RNA pool, with rRNA making up over 90% of the pool (Neidhardt, 1996). While there are enzymatic- and oligonucleotide-based approaches that remove rRNA, they are not always very efficient, and this can depend on the species being used and also the source of the sample. Some reports observed varying enrichment efficiencies between replicate samples (Yoder-Himes et al., 2009; Stewart et at, 2010). However, the increased throughput of next-generation sequencing platforms such as Illumina, which currently generates hundreds of a millions of reads in a single run, enables researchers to sequence through the ribosomal RNA and detect the more informative mRNA transcripts.

In addition, Illumina RNA-seq makes it feasible for researchers to study microbiomes that are in close association with a host. While there have been only a few published reports of RNA-seq being used to investigate host-associated microbiomes (Turnbaugh et al., 2009, 2010; Stewart et al., 2011; Bomar et al., 2011), we anticipate that the recent advancements in RNA-seq will facilitate a greater use of this technique for symbioses. For instance, the high throughput enables the detection of transcripts from the microbiome despite contaminating host RNA and the presence of a mixed-species community. In addition, the ability to multiplex samples, where multiple samples are sequenced on the same flow cell lane, makes RNA-seq a cost-effective tool for transcriptomics. The application of this technology to in vivo and in vitro microbiomes, when possible, allows researchers to identify the physiology of a microorganism that is unique to its symbiotic state and can aid in determining the contribution of the microbiome to its host.


We thank M. Nelson, A. Tomasi, and E. McClure for helpful comments on the manuscript. This research was supported by NSF Career Award MCB 0448052, NIH ROI GM095390, and UCONN Research Foundation award J980 to J.G.

Literature Cited

Altschul, S., W. Gish, W. Miller, E. Myers, and D. Lipman. 1990. Basic local alignment search tool. J. Mot Biol. 215: 403-410.

Babitzke, P., and T. Romeo. 2007. CsrB sRNA family: sequestration of RNA-binding regulatory proteins. Curr. Opin. Microbiol. 10: 156-163.

Bomar, L., M. Maltz, S. Colston, and J. Graf. 2011. Directed culturing of microorganisms using metatranscriptomics. mBio 2: e00012-11.

Cunin, R., N. Glansdorff, A. Pierard, and V. Stalon. 1986. Biosynthesis and metabolism of arginine in bacteria. Microbiol. Mol. Biol. Rev. 50: 314-352.

Dethlefsen, L., M. McFall-Ngai, and D. A. Reiman. 2007. An ecological and evolutionary perspective on human-microbe mutualism and disease. Nature 449: 811-818.

D'Hooghe, I., C. VanderWauven, J. Michiels, C. Tricot, P. de Wilde, J. Vanderleyden, and V. Stalon. 1997. The arginine deiminase pathway in Rhizobium etli: DNA sequence analysis and functional study of the arcABC genes. J. Bacteriol. 179: 7403-7409.

Edgar, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 32: 1792-1797.

Frias-Lopez, J., Y. Shi, G. W. Tyson, M. L. Coleman, S. C. Schuster, S. W. Chisholm, and E. F. DeLong. 2008. Microbial community gene expression in ocean surface waters. Proc. Natl. Acad. Sci. USA 105: 3805-3810.

Fulde, M., J. Willenborg, A. de Greeff, L. Benga, H. E. Smith, P. Valentin-Weigand, and R. Goethe. 2011. ArgR is an essential local transcriptional regulator of the arcABC operon in Streptococcus suis and is crucial for biological fitness in an acidic environment. Microbiology 157: 572-582.

Gardner, P. P., J. Daub, J. Tate, B. L. Moore, I. H. Osuch, S. Griffiths-Jones, R. D. Finn, E. P. Nawrocki, D. L. Kolbe, S. R. Eddy et at 2011. Rfam: Wikipedia, clans and the "decimal" release. Nucleic Acids Res. 39 (Suppl 1): D141-145. doi: 10.1093/nar/gkq1129.

Gilbert, J. A., D. Field, Y. Huang, R. Edwards, W. Li, P. Gilna, and I. Joint. 2008. Detection of large numbers of novel sequences in the metatranscriptomes of complex marine microbial communities. PloS One 3: e3042.

Graf, J. 1999. Symbiosis of Aeromonas veronii biovar sobria and Hirudo medicinalis, the medicinal leech: a novel model for digestive-tract associations. Infect. Immun. 67: 1-7.

Graf, J., Y. Kikuchi, and R. V. M. Rio. 2006. Leeches and their microbiota: naturally simple symbiosis models. Trends Microbiol. 14: 365-371.

Heungens, K., C. E. Cowles, and H. Goodrich-Blair. 2002. Identification of Xenorhabdus nematophila genes required for mutualistic colonization of Steinernerna carpocapsae nematodes. Mol. Microbiol. 45: 1337-1353.

Humphreys, S., A. Stevenson, A. Bacon, A. B. Weinhardt, and M. Roberts. 1999. The alternative sigma factor, of, is critically important for the virulence of Salmonella typhimurium. Infect. Immun. 67: 1560-1568.

Janda, J. M., and S. L. Abbott. 2010. The genus Aeromonas: taxonomy, pathogenicity, and infection. Clin. Microbiol. Rev. 23: 35-73.

Kikuchi, Y., and J. Graf. 2007. Spatial and temporal population dynamics of a naturally occurring two-species microbial community inside the digestive tract of the medicinal leech. Appl. Environ. Microbiol. 73: 1984-1991.

Kovacikova, G., and K. Skorupski. 2002. The alternative sigma factor sigma(E) plays an important role in intestinal survival and virulence in Vibrio cholerae. Infect. Immun. 70: 5355-5362.

Lenz, D. H., M. B. Miller, J. Zhu, R. V. Kulkarni, and B. L. Sassier. 2005. CsrA and three redundant small RNAs regulate quorum sensing in Vibrio cholerae. Mol. Microbiol. 58: 1186-1202.

Lucchetti-Miganeh, C., E. Burrowes, C. Baysse, and G. Ermel. 2008. The post-transcriptional regulator CsrA plays a central role in the adaptation of bacterial pathogens to different stages of infection in animal hosts. Microbiology 154: 16-29.

Luo, C., D. Tsementzi, N. Kyrpides, T. Read, and K. T. Konstantinidis. 2012. Direct comparisons of Illumina vs. Roche 454 Sequencing Technologies on the same microbial community DNA sample. PloS One 7(2): e30087.

Margulies, M., M. Egholm, W. E. Altman, S. Attiya, J. S. Bader, L. A. Bemben, J. Berka, M. S. Braverman, Y. Chen, Z. Chen et al 2005. Genome sequencing in microfabricated high-density picolitre reactors. Nature 437: 376-380.

Martens, E. C., H. C. Chiang, and J. I. Gordon. 2008. Mucosal glycan foraging enhances fitness and transmission of a saccharolytic human gut bacterial symbiont. Cell Host Microbe 4: 447-457.

Neidhardt, F. C., ed. 1996. Escherichia soli and Salmonella. ASM Press, Washington, DC.

Nelson, M. C., and J. Graf. 2012. Bacterial symbioses of the medicinal leech Hirudo verbana. Gut Microbes 3(4): doi:10.416 Ugmic.20227.

Noda, S., M. Ohkuma, R. Usami, K. Horikoshi, and T. Kudo. 1999. Culture-independent characterization of a gene responsible for nitrogen fixation in the symbiotic microbial community in the gut of the termite Neotermes koshunensis. Appl. Environ. Microbial. 65: 4935-4942.

Petersen, J. M., F. U. Zielinski, T. Pape, R. Seifert, C. Moraru, R. Amann, S. Hourdez, P. R. Girguis, S. D. Wankel, V. Barbe et al. 2011. Hydrogen is an energy source for hydrothermal vent symbioses. Nature 476: 176-180.

Poretsky, R. S., I. Hewson, S. Sun, A. E. Allen, J. P. Zehr, and M.A. Moran. 2009. Comparative day/night metatranscriptomic analysis of microbial communities in the north pacific subtropical gyre. Environ. Microbial. 11: 1358-1375.

Poroyko, V., J. R. White, M. Wang, S. Donovan, J. Alverdy, D. C. Liu, and M. J. Morowitz. 2010. Gut microbial gene expression in mother-fed and formula-fed piglets. PloS One 5: e12459.

Rasko, D. A., NI. J. Rosovitz, G. S. A. Myers, E. F. Mongodin, W. F. Fricke, P. Gajer, J. Crabtree, M. Sebaihia, N. R. Thomson, and R. Chaudhuri et al. 2008. The pangenome structure of Escherichia coli: comparative genomic analysis of Escherichia coli commensal and pathogenic isolates. J. BacterioL 190: 6881-6893.

Rowley, G., M. Spector, J. Kormanec, and M. Roberts. 2006. Pushing the envelope: extracytoplasmic stress responses in bacterial pathogens. Nat. Rev. Microbiol. 4: 383-394.

Sambrook, J., and a W. Russell. 2001. Molecular cloning: A laboratory manual. Cold Spring Harbor Press, Cold Spring Harbor, New York.

Sawyer, R. 1986. Leech Biology and Behavior. Oxford University Press, Oxford.

Schikorski, D., V. Cuvillier-Hot, M. Leippe, C. Boidin-Wichlacz, C. Slomianny, E. Macagno, M. Salzet, and A. Tasiemski. 2008. Microbial challenge promotes the regenerative process of the injured central nervous system of the medicinal leech by inducing the synthesis of antimicrobial peptides in neurons and microglia. J. ImmunoL 181: 1083-1095.

Shi, Y., G. W. Tyson, and E. F. DeLong. 2009. Metatranscriptomics reveals unique microbial small RNAs in the ocean's water column. Nature 459: 266-269.

Silver, A. C., Y. Kikuchi, A. A. Fadl, J. Sha, A. K. Chopra, and J. Graf. 2007. Interaction between innate immune cells and a bacterial type III secretion system in mutualistic and pathogenic associations. Proc. Nett Acad. Sci. USA 104: 9481-9486.

Silver, A. C., D. Williams, J. Faucher, A. J. Horneman, J. P. Gogarten, and J. Graf. 2011. Complex evolutionary history of the Aeromonas veronii group revealed by host interaction and DNA sequence data. PloS One 6: e16751.

Sonnenburg, J. L., J. Xu, D. D. Leip, C. Chen, B. P. Westover, J. Weatherford, J. D. Buhler, and J. I. Gordon. 2005. Glycan foraging in vivo by an intestine-adapted bacterial symbiont. Science 307: 1955-1959.

Stewart, F. J., E A. Ottesen, and E. F. DeLong. 2010. Development and quantitative analyses of a universal rRNA-subtraction protocol for microbial metatranscriptomics. ISMS J. 4: 896-907.

Stewart, F. J., 0. Dmytrenko, E. F. DeLong, and C. M. Cavanaugh. 2011. Metatranscriptomic analysis of sulfur oxidation genes in the endosymbiont of Solemya velum. Front. Microbiol 2: 134.

Tartar, A., M. M. Wheeler, X. Zhou, M. R. Coy, D. G. Boucias, and M. E Scharf. 2009. Parallel metatranscriptome analyses of host and symbiont gene expression in the gut of the termite Reticulitermes flavipes. Biotechnol. Biofuels 2: 25.

Tasiemski, A., F. Vandenbulcke, G. Mitta, J. Lemoine, C. Lefebvre, P. Sautiere, and M. Salzet. 2004. Molecular characterization of two novel antibacterial peptides inducible upon bacterial challenge in an annelid: the leech Theromyzon tessulatum. J. Biol. Chem. 279: 3097330982.

Turnbaugh, P. J., V. K. Ridaura, J. J. Faith, F. E. Rey, R. Knight, and J. I. Gordon. 2009. The effect of diet on the human gut microbiome: a metagenomic analysis in humanized gnotobiotic mice. Sci. Transl. Med. 1: 6ra14.

Turnbaugh, P. J., C. Quince, J. J. Faith, A. C. McHardy, T. Yatsunenko, F. Niazi, J. Affourtit, M. Egholm, B. Henrissat, R. Knight et at 2010. Organismal, genetic, and transcriptional variation in the deeply sequenced gut microbiomes of identical twins. Proc. Natl. Acad. Sci. USA 107: 7503-7508.

Urich, T., A. Lanzen, J. Qi, D. H. Huson, C. Schleper, and S. C. Schuster. 2008. Simultaneous assessment of soil microbial community structure and function through analysis of the meta-transcriptome. PloS One 3: e2527.

Wier, A. M., S. V. Nyholm, M. J. Mandel, R. Prisca Massengo-Tiasse, A. I,. Schaefer, I. Koroleva, S. Splinter-Bondurant, B. Brown, L. Manzella, E. Snir et at 2010. Transcriptional patterns in both host and bacterium underlie a daily rhythm of anatomical and metabolic change in a beneficial symbiosis. Proc. Natl. Acad. Sci. USA 107: 2259-2264.

Worthen, P. L., C. J. Gode, and J. Graf. 2006. Culture-independent characterization of the digestive-tract microbiota of the medicinal leech reveals a tripartite symbiosis. Appl. Environ. Microbial. 72: 4775-4781.

Yoder-Himes, D. R., P. S. G. Chain, Y. Zhu, 0. Wurtzel, E. M. Rubin, J. M. Tiedje, and R. Sorek. 2009. Mapping the Burkholderia cenocepacia niche response via high-throughput sequencing. Proc. Natl. Acad. Sci. USA 106: 3976-3981.

Yu, H. B., Y. L. Zhang, Y. L. Lau, F. Yao, S. Vilches, S. Merino, J. M. Tomas, S. P. Howard, and K. Y. Leung. 2005. Identification and characterization of putative virulence genes and gene clusters in Aeromonas hydrophila PPD134/91. Appl. Environ. Microbial. 71: 4469-4477.

Zebe, E., F.-J. Roters, and B. Kaiping. 1986. Metabolic changes in the medicinal leech Hirudo methcinalis following feeding. Comp. Biochem. Physiol. 84A: 49-55.

* To whom correspondence should be addressed. E-mail:

Received 22 January 2012; accepted 5 July 2012.


Department of Molecular and Cell Biology, University of Connecticut, Storrs, Connecticut 06269
COPYRIGHT 2012 University of Chicago Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Bomar, Lindsey; Graf, Joerg
Publication:The Biological Bulletin
Article Type:Report
Geographic Code:1USA
Date:Aug 1, 2012
Previous Article:16S rDNA-based metagenomic analysis of bacterial diversity associated with two populations of the Kleptoplastic sea slug Elysia chlorotica and its...
Next Article:Community ecology and the evolution of molecules of keystone significance.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters