Functional conservation of specialized ribosomes bearing genome-encoded variant rRNAs in Vibrio species

Heterogeneity of ribosomal RNA (rRNA) sequences has recently emerged as a mechanism that can lead to subpopulations of specialized ribosomes. Our previous study showed that ribosomes containing highly divergent rRNAs expressed from the rrnI operon (I-ribosomes) can preferentially translate a subset of mRNAs such as hspA and tpiA in the Vibrio vulnificus CMCP6 strain. Here, we explored the functional conservation of I-ribosomes across Vibrio species. Exogenous expression of the rrnI operon in another V. vulnificus strain, MO6-24/O, and in another Vibrio species, V. fischeri (strain MJ11), decreased heat shock susceptibility by upregulating HspA expression. In addition, we provide direct evidence for the preferential synthesis of HspA by I-ribosomes in the V. vulnificus MO6-24/O strain. Furthermore, exogenous expression of rrnI in V. vulnificus MO6-24/O cells led to higher mortality of infected mice when compared to the wild-type (WT) strain and a strain expressing exogenous rrnG, a redundant rRNA gene in the V. vulnificus CMCP6 strain. Our findings suggest that specialized ribosomes bearing heterogeneous rRNAs play a conserved role in translational regulation among Vibrio species. This study shows the functional importance of rRNA heterogeneity in gene expression control by preferential translation of specific mRNAs, providing another layer of specialized ribosome system.


Introduction
Recent growing evidence suggests that ribosomes are much more complex and heterogeneous macromolecules than previously thought [1][2][3][4][5].Ribosome heterogeneity, caused by compositional changes such as divergent rRNAs, ribosomal proteins, ribosome-associated factors, or modifications of these factors, is found within species across various organisms.These subpopulations of ribosome variants are proposed to have specialized functions in regulating many biological processes-such as stress response, differentiation, and development-through the preferential translation of a specific subset of mRNAs [6][7][8][9][10].Ribosomal heterogeneity due to variations in rRNA sequences has been observed in various organisms over several decades [11][12][13][14].These genome-encoded heterogeneous rRNAs have been implicated in cellular development and adaptation to environmental changes.For instance, Streptomyces coelicolor differentially expresses heterogeneous rRNAs during the morphological developmental stages [15,16], and heterogeneous 16S rRNAs in Haloarcula marismortui can be preferentially expressed under different growth temperatures [17].In Escherichia coli, among seven rRNA operons, rrnH is highly expressed in response to nutrient limitation [6].However, a critical question is whether the heterogeneity of rRNA sequences can lead to functional ribosome specialization.
Our previous experiments have shown that ribosome variants containing genome-encoded divergent rRNAs (I-ribosomes) can preferentially translate specific mRNAs, including hspA and tpiA mRNAs, in the opportunistic human pathogen Vibrio vulnificus CMCP6 [9].Here, we investigated whether the physiological functions of I-ribosomes seen in V. vulnificus CMCP6 are conserved in other Vibrio strains and species.

Ethics statement
All animal experiments were performed in accordance with the national guidelines for the use of animals in scientific research.The protocol was approved by Chung-Ang University Support Center (Approval No. CAU2012-0044).Per protocol, mice were sacrificed by using 100% CO 2 followed by cervical dislocation if mice developed tumors greater than 1.8 cm in diameter, or if the mice showed any signs of persistent morbidity such as loss in weight greater than 20%, lethargy, unwillingness to ambulate, hunched posturing and ruffled fur.No invasive procedures likely to produce moderate to severe pain were performed.All efforts were made to minimize possible pain and suffering for mice during irradiation, monitoring, and euthanasia.

Animals
7-week-old female ICR mice were kept under a 12:12 hours light/dark cycle, constant temperature of 25˚C, 50% humidity, and allowed free access to water and a standard diet.Mouse body weight was monitored twice a week.Mice were acclimatized to the new environment for a few weeks before any experimental procedure.

Bacterial culture conditions
All bacterial strains and plasmids used in this study are listed in S1 Table .Vibrio strains were grown at 30-45˚C in LBS (Luria-Bertani [LB] medium containing NaCl at a final concentration of 2.5% [w/v]) under aerobic conditions.E. coli strains were grown in LB medium at 37˚C.To express rrnI or rrnG exogenously, the pRK415-rrnI or pRK415-rrnG plasmid was conjugated into the Vibrio strains [18,19].

Isolation of crude ribosomes
Total RNA and crude ribosomes were isolated as previously described [9].Briefly, cells grown to the mid-log phase were lysed using a French press, and the lysate was then subjected to loading onto a 30% sucrose cushion.Crude ribosomes were pelleted after centrifugation at 100,000g for 16 h at 4˚C.The pellet was resuspended and loaded onto a 15-40% sucrose gradient and centrifuged at 79,500g for 13.5 h at 4˚C.After centrifugation, 70S ribosomes were obtained by fractionation and measurement of the optical density (OD) at 260 nm.The quality and quantity of the extracted crude ribosomes were assessed using a Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific, MA, USA).

Quantification of I-rRNA in ribosomes
The proportion of ribosomes incorporating I-rRNAs was determined by allele-specific RT-PCR as previously described [9,20].Complementary DNA (cDNA) was synthesized using an iScript cDNA synthesis kit (Bio-Rad Laboratories, CA, USA) according to the manufacturer's instructions.Primers specific to the 23S rrnI or rrnG region were designed with a mismatch at a variable nucleotide position (Fig 1A).The PCR primers used for allele-specific RT-PCR are listed in S2 Table .All primers were purchased from BIONICS (Seoul, Korea).

Western blot analysis
Cell lysates were prepared and subjected to 12% SDS polyacrylamide gel electrophoresis (SDS-PAGE) for subsequent immunoblotting with the appropriate antibodies.Images of the western blots were obtained using an Amersham Imager 600 (GE Healthcare Life Sciences, Amersham, Buckinghamshire, UK) and quantified using Quantity One (Bio-Rad Laboratories).The ribosomal protein S1 was used as the control.The following antibodies were used: anti-HspA (generated in our laboratory; [9]), anti-RNAP-β (Abcam, Cambridge, UK), and anti-S1 (obtained from the laboratory of Dr. Stanley N. Cohen [9]).

Measurement of heat shock susceptibility
Heat shock susceptibility was measured as previously described [9].Briefly, Vibrio strains were grown in LBS medium containing 0.2 μg/ml tetracycline at 30˚C until reaching an OD 600 of 0.6, and then each culture was immediately incubated in a water bath at 45˚C for 180 min to induce heat shock.

Co-immunoprecipitation
Co-immunoprecipitation of nascent peptides of HspA or RNAP-β subunit bound ribosomes was performed as previously described [20].The presence of peptides in the eluted samples was confirmed by western blot, and total RNA was purified by ethanol precipitation.Purified RNA in the samples was quantified by allele-specific RT-PCR.

Mouse mortality test
The virulence of V. vulnificus was measured using a mouse mortality test as previously described [9].Briefly, the V. vulnificus MO6-24/O strains were grown in LBS medium to midlog phase.Before injection, pathogen-free, 7-week-old female ICR mice (n = 10 mice per treatment) were fasted for 24 h and water deprived for 4 h.Tetracycline (200 mg per kg of body weight) was administered to the mice orally, and iron dextran (80 μg per g body weight) was injected intraperitoneally.After 1 h, V. vulnificus MO6-24/O (1 × 10 6 ) cells were injected subcutaneously into the mice.Mice were monitored for 25 h, and dead mice were counted.

Measurement of viable bacterial counts in mouse organs
Viable bacterial counts of V. vulnificus in mouse organs were performed as described previously [9].Briefly, mice (n = 5 mice per treatment) were infected subcutaneously with V. vulnificus MO6-24/O.Mice were sacrificed 6 h post-infection, and spleens and livers were isolated.The organs were weighed, homogenized, and dissolved in LBS, and the homogenates were plated on LBS agar or LBS agar containing 0.2 μg/ml tetracycline and incubated overnight at 30˚C.Bacterial colonies were then counted and expressed as the number of colony-forming units (CFUs)/g.

Phylogenetic analysis of I-rRNA based on rRNA gene sequences
For the phylogenetic analysis of Vibrio species, 16S or 23S rRNA sequences were aligned using Silva (https://www.arb-silva.de/).Evolutionary distances were calculated using the Kimura two-parameter model [21].The phylogenetic trees were constructed using the neighbor-joining method [22] in the MEGA7 program [23] with bootstrap values based on 1,000 replicates [24][25][26].

Statistical analysis
All statistical details of experiments are included in the figure legends.Data are presented as the mean ± standard error of the mean (SEM).Statistical analyses were performed using Graph Pad Prism 9 software (version 9.4.1;Graph Pad Software Inc.), and P-values were obtained from a two-tailed unpaired t-test or one-way analysis of variance (ANOVA) followed by Dunnett's multiple comparison test.

Effect of rrnI expression on HspA levels in V. vulnificus MO6-24/O
To determine the effect of rrnI expression on HspA levels in a V. vulnificus strain that does not express highly variant rRNA genes, V. vulnificus MO6-24/O strains exogenously expressing rrnI (MO6+rrnI) or rrnG (MO6+rrnG) were constructed.The rrnG operon was used as a negative control to evaluate the effect of rRNA gene copy number variation.It is one of the redundant rRNA genes in the V. vulnificus CMCP6 strain [9].The incorporation of exogenously expressed rRNA from the rrnI or rrnG operon (I-rRNAs or G-rRNAs) into ribosomes was assessed by allele-specific RT-PCR using primers specific to bind to 23S I-rRNA and G-rRNA (Fig 1A).These exogenously expressed rRNAs represented ~15% of the total rRNAs in these strains (Fig 1B).Similar results were obtained using the 23S I-rRNA-specific primers in allelespecific RT-PCR (S1 Fig) .Next, we investigated whether the HspA protein level was increased in MO6+rrnI cells.The sequences of the hspA gene were 99% identical between the V. vulnificus MO6-24/O and CMCP6 strains (Fig 1C ), generating virtually the same secondary structure of hspA mRNA (Fig 1D).Expression of HspA was increased by approximately 1.5-fold in MO6+rrnI cells compared with the levels seen in MO6 WT and MO6+rrnG cells (Fig 1E).Consistent with previous results showing a strong correlation between heat shock susceptibility and the rrnI expression-dependent changes in HspA expression in the V. vulnificus CMCP6 strain [9], heat shock at 45˚C for 60 min resulted in a ~30% increase in CFUs/g of the MO6+rrnI strain compared to those of the MO6 WT and MO6+rrnG strains (Fig 1F).These results indicate that the increased heat shock susceptibility of MO6+rrnI cells is a direct result of the overexpression of HspA in these cells.

Preferential association of the I-ribosome to nascent HspA peptides in V. vulnificus MO6-24/O
To explore whether hspA mRNA is preferentially translated by I-ribosomes in V. vulnificus MO6-24/O cells, co-immunoprecipitation (co-IP) analysis using anti-HspA antibodies was performed in MO6 WT, MO6+rrnI, and MO6+rrnG cells [20]; Fig 2A).Immunoprecipitation of RNAP-β, levels of which are not rrnI expression dependent, was used as a control.
Total rRNA was purified from the co-immunoprecipitated ribosomes and relative amounts of I-rRNA in the total rRNA populations were measured by allele-specific RT-PCR.In MO6 +rrnI cells, I-rRNA accounted for ~50% of the total rRNA from HspA-bound ribosomes, whereas it represented ~15% of the total rRNAs from crude ribosome samples or RNAP-βbound ribosomes (Fig 2B).This suggests that hspA mRNA can be preferentially translated by I-ribosomes in the V. vulnificus MO6-24/O strain.

Effect of rrnI on V. vulnificus MO6-24/O virulence in mice
In our previous study, the survival rates and duration of survival of mice infected with rrnIdeleted V. vulnificus CMCP6 cells were increased compared to that of mice infected with the CMCP6 WT strain [9].To confirm the physiological function of I-ribosomes in the virulence of V. vulnificus MO6-24/O, the MO6 WT, MO6+rrnG, or MO6+rrnI strains were intraperitoneally injected into BALB/c mice, and the survival rate and duration of infected mice were examined.In addition, since plasmid vectors require antibiotics for maintenance in V. vulnificus cells after being infected into mice [27], we first confirmed the plasmid loss rate of WT V. vulnificus MO6-24/O by comparing the number of CFUs isolated from representative mice organs (spleen and liver) after administration of tetracycline.As shown in Fig 3A, the plasmid loss rate was ~82% and ~49% after the first administration of antibiotics in the spleen and liver, respectively (Fig 3A).After the second administration of antibiotics, the plasmid loss rate was reduced in both organs (Fig 3A ), suggesting that two consecutive administrations of antibiotics are needed to maintain plasmid vectors in V. vulnificus cells in infected mice.Thus, we used two consecutive injections of tetracycline for vector maintenance in the mouse mortality experiment.Lower survival rates were observed in MO6+rrnI-infected mice than in mice infected with the MO6 WT or MO6+rrnG strains (Fig 3B ), indicating that I-rRNA influences the virulence of V. vulnificus MO6-24/O.
Next, we measured the number of viable bacterial cells in the spleen and liver of mice infected with the MO6 WT, MO6+rrnG, or MO6+rrnI strains to further explore the effect of https://doi.org/10.1371/journal.pone.0289072.g002rrnI expression on bacterial survival in the host environment.We administered tetracycline twice and sacrificed the mice 6 h after bacterial infection, a time chosen based on the mouse survival duration.A ~3-fold increase in colonization was observed in the organs of the MO6 +rrnI strain-infected mice compared to the MO6 WT strain-infected mice (Fig 3C).By contrast, no significant changes in colonization were detected between mice infected by the MO6 +rrnG strain and the MO6 WT strain (Fig 3C ), suggesting that MO6+rrnI cells can grow more rapidly than the other strains in mouse organs.These results indicate that expression of I-rRNA makes V. vulnificus more virulent in mice.

Effect of rrnI expression on HspA levels in V. fischeri MJ11
To determine the effect of rrnI expression on HspA expression in another Vibrio species that does not express highly variant rRNA genes, V. fischeri MJ11 strains exogenously expressing rrnI We also observed enhanced expression of HspA and heat shock tolerance in the MJ11+rrnI strain compared to those in the MJ11 WT and MJ11+rrnG strains (Fig 4C and 4D).These results suggest that exogenous expression of rrnI enhances HspA expression, leading to the increased heat shock resistance of V. fischeri cells.

Discussion
Emerging studies have suggested that heterogeneous rRNAs have specialized roles in protein synthesis.However, a longstanding question is how ribosome heterogeneity at the level of rRNAs can contribute to specialized ribosome function.Our previous studies showed that ribosomes containing the most divergent rRNAs preferentially bind and translate specific V. vulnificus strains that contain rRNA genes similar to the rrnI operon were selected using NCBI nucleotide blast (https://blast.ncbi.nlm.nih.gov) and the rRNA gene sequences of these strains were obtained from the NCBI database.The rRNA genes of each strain were named starting with "A" according to the nucleotide position in the genome.The stress-response mRNAs in V. vulnificus CMCP6 cells under stress conditions [9,20].Our present study highlights the functional conservation of these divergent rRNAs in other Vibrio species that do not express highly variant rRNAs.Specifically, exogenous expression of I-rRNAs in the V. vulnificus MO6-24/O strain resulted in enhanced HspA expression, heat shock tolerance, and virulence (Figs 1, 3 and S1 Fig) .In addition, the co-immunoprecipitation analysis shows that the enhanced HspA expression is a direct consequence of I-ribosome-mediated preferential translation of hspA mRNA (Fig 2).As previously proposed, the target mRNAs of the I-ribosome showed a common feature [9].In the case of hspA mRNA in V. fischeri, the putative Shine-Dalgarno (SD) sequence is also embedded in the secondary structure (Fig 4B ), indicating that I-ribosome mediated mRNA selection does not require a strong SD/anti-SD interaction.
A comparative analysis of the 16S and 23S rRNA gene sequences of Vibrio strains indicates that rrnI operon from V. vulnificus CMCP6 is not closely related to rRNA operons of the other Vibrio species (Fig 5A and 5B), which suggests that the rrnI operon might have evolved recently in the V. vulnificus CMCP6 strain.
Taken together, our findings indicate that heterogeneous ribosomes differing in their rRNA sequences have evolved to have a unique and conserved function in selectively translating specific mRNAs in Vibrio species (Fig 5C ).Further studies are needed to unveil why some bacterial species have retained an evolutionarily unfavorable pathway of maintaining divergent rRNA genes to control protein synthesis from specific mRNAs.

Conclusions
This study provides evidence for the functional conservation of ribosomes carrying heterogeneous rRNAs among Vibrio species.We observed that the exogenous expression of the rrnI operon from V. vulnificus CMCP6 led to enhanced expression of HspA by preferential binding of I-ribosomes to hspA mRNA, and, consequently, decreased heat shock susceptibility in V. vulnificus MO6-24/O strain and V. fischeri.This study highlights functional conservation of specialized ribosomes in preferential translation of specific mRNAs among Vibrio species and provides another layer of regulation of gene expression at the ribosome level.

Fig 1 .
Fig 1. rrnI-dependent expression of HspA in V. vulnificus MO6-24/O strains.(A) Schematic representation of the allele-specific RT-PCR analysis analyzing the relative amounts of I-rRNA or G-rRNA.(B) The number of I-rRNA or G-rRNA amplicons and other rRNAs amplified from the cDNA of the MO6 WT, MO6+rrnG, and MO6+rrnI strains was determined by PCR using common and allele-specific primers.The cDNA was synthesized from rRNAs purified from crude ribosomes of these strains.PCR products were resolved on a 2% agarose gel.Data are presented as the mean ± SEM of three independent experiments.Statistical significance was determined using one-way ANOVA followed by Dunnett's multiple comparison test (ns, not significant; ****, P < 0.0001).(C) Sequence alignment and secondary structure prediction of hspA mRNAs in the V. vulnificus strains MO6-24/O and CMCP6.Sequences of the hspA genes from V. vulnificus strains MO6-24/O and CMCP6 were obtained from the NCBI database (NC_014965.1 and NC_004459.2,respectively) and aligned using ClustalW (https://www.genome.jp/tools-bin/clustalw).The putative transcriptional start site (TSS) and the first nucleotide of the hspA gene are indicated in yellow and red, respectively.(D) Predicted secondary structures of hspA mRNAs from V. vulnificus CMCP6 (left) or V. vulnificus MO6-24/O (right).Secondary structures were obtained using the M-fold program (http://unafold.rna.Albany.edu).(E) rrnI-dependent expression of HspA.Vibrio vulnificus MO6-24/O strains (WT, +rrnG, and +rrnI) were grown in LBS at 30˚C until mid-log phase and harvested for western blot analysis of HspA proteins using polyclonal antibodies against HspA.(F) Effect of rrnI expression on heat shock susceptibility of V. vulnificus MO6-24/O strains (WT, +rrnG, and +rrnI, and ΔhspA).The number of CFUs of each V. vulnificus MO6-24/O strain grown at 30˚C or transiently grown at 45˚C for 180 min was measured.Data are presented as the mean ± SEM of three independent experiments.Statistical significance was determined using one-way ANOVA followed by Dunnett's multiple comparison test (ns, not significant; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).https://doi.org/10.1371/journal.pone.0289072.g001

Fig 2 .
Fig 2. Co-immunoprecipitation (co-IP) showing the predominant association of I-ribosomes with HspA nascent peptides in V. vulnificus MO6-24/O strains.(A) Identification of HspA and the RNAP-β subunit after co-IP.The precipitate was subjected to western blotting with antibodies against HspA or RNAP-β.S1 protein was used as a loading control.(B) Characterization of the expression and assembly of I-rRNAs after co-IP.The number of amplicons of 23S I-rRNAs and other 23S rRNAs amplified from the cDNAs of the MO6 WT, MO6+rrnG, and MO6+rrnI strains was determined by RT-PCR using common and allele-specific primers.The cDNA was synthesized from rRNAs purified from immunoprecipitated samples or crude ribosome samples, which was used as controls.PCR products were resolved on a 2% agarose gel.Data are presented as the mean ± SEM of three independent experiments.Statistical significance was determined using one-way ANOVA followed by Dunnett's multiple comparison test (ns, not significant; ****, P < 0.0001).

Fig 3 .
Fig 3. Effect of rrnI expression on the virulence of V. vulnificus MO6-24/O in mice.(A) Plasmid loss rate of V. vulnificus MO6-24/O in mice.Pathogen-free 7-week-old female ICR mice that were pretreated with tetracycline and iron dextran (n = 5 mice per group) were subcutaneously injected with 1 × 10 6 cells of V. vulnificus MO6-24/O (MO6 WT).Mice with tetracycline administered orally before infection and those treated once more 6 h after the initial treatment were sacrificed 6 h and 12 h after infection, respectively.Data are presented as the mean ± SEM of three independent experiments.Statistical significance was determined using a two-tailed unpaired Student's t-test (ns, not significant; ****, P < 0.0001).(B) Survival rates and the survival time of mice infected with V. vulnificus MO6-24/O strains (MO6 WT, MO6+rrnG, and MO6+rrnI).Pathogen-free 7-week-old female ICR mice pretreated with iron dextran (n = 10 mice per group) were intraperitoneally injected with 8 × 10 2 cells of the V. vulnificus MO6-24/O strains.Survival rate was monitored for 25 h.Data are presented as the mean ± SEM of three independent experiments.Two-tailed unpaired Student's t-tests were used to assess significant differences: ** denotes P < 0.01 for comparisons of MO6+rrnI-infected mice versus WT-or MO6+rrnG-infected mice.(C) Viable bacterial counts in the spleen and liver of mice infected with the V. vulnificus MO6-24/O strains (MO6 WT, MO6+rrnG, and MO6+rrnI).Six hours after bacterial infection, the mice (n = 5 mice per group) were sacrificed.Results are expressed as numbers of CFU/g of each organ.Data are presented as the mean ± SEM of three independent experiments.Statistical significance was determined using one-way ANOVA followed by Dunnett's multiple comparison test (ns, not significant; ****, P < 0.0001).https://doi.org/10.1371/journal.pone.0289072.g003

Fig 4 .Fig 5 .
Fig 4. rrnI-dependent expression of HspA in V. fischeri MJ11 strains.(A) Sequences of hspA genes in V. fischeri MJ11 and V. vulnificus CMCP6 were obtained from the NCBI database (NC_011184.1 and NC_004459.2,respectively) and were aligned using ClustalW.The putative transcriptional start site of V. vulnificus CMCP6 and the first nucleotide of the hspA sequence are indicated in yellow and red, respectively.(B) Prediction of the secondary structures of V. fischeri MJ11 hspA mRNAs.Secondary structures were obtained using the M-fold program.(C) V. fischeri MJ11 strains (WT, +rrnG, and +rrnI) were grown and harvested for a western blot analysis of HspA proteins using polyclonal antibodies against HspA (see the Fig 1E legend for details).(D) Effect of rrnI expression on heat shock susceptibility of V. fischeri MJ11 strains (WT, +rrnG, and +rrnI).The number of CFUs of each V. fischeri MJ11 strain was measured as described in the legend of Fig 1F.The expression levels of HspA and the number of CFUs were compared by setting those of the WT to 1. Data are presented as the mean ± SEM of three independent experiments.Statistical significance was determined using one-way ANOVA followed by Dunnett's multiple comparison test (ns, not significant; *, P < 0.05 and ***, P < 0.001).https://doi.org/10.1371/journal.pone.0289072.g004