The vital role of CovS in the establishment of Streptococcus equi subsp.zooepidemicus virulence

XU Bin ,MA Zhe ,ZHOU Hong ,LlN Hui-xing ,FAN Hong-jie

1 MOE Joint International Research Laboratory of Animal Health and Food Safety,College of Veterinary Medicine,Nanjing Agricultural University,Nanjing 210095,P.R.China

2 Key Laboratory of Veterinary Biological Engineering and Technology of Ministry of Agriculture and Rural Affairs,National Research Center of Veterinary Biologicals Engineering and Technology,Institute of Veterinary Medicine,Jiangsu Academy of Agricultural Sciences,Nanjing 210014,P.R.China

3 Jiangsu Co-innovation Center for the Prevention and Control of Important Animal Infectious Diseases and Zoonoses,Yangzhou University,Yangzhou 225009,P.R.China

Abstract Streptococcus equi subsp.zooepidemicus (SEZ) is an important zoonotic agent.Here,a virulence-attenuated strain M35246 derived from natural variation of wild-type SEZ ATCC35246 was found.M35246 showed a deletion of 25 contiguous genes as well as a loss-of-function mutation in covS.Subsequently,a 25-gene-deleted strain (ΔPI),a covSmutant strain (McovS),and relevant complementary strains were constructed and investigated.M35246 and McovS were significantly less encapsulated and exhibited poorer anti-phagocytic capacity compared to wild-type SEZ.McovS was significantly more sensitive to β-lactams,aminoglycosides,macrolides,and lincosamides than wild-type SEZ.M35246,McovS,and ΔPI exhibited an increase in median lethal dose (LD50) in mice by 105,105,and 5 times when compared to wild-type SEZ,respectively.Neither M35246 nor McovS were isolated from mice 48 h after being challenged with approximately 2 000 times the LD50 of wild-type SEZ.Transcriptome analysis showed that 668 significantly differentially expressed genes existed between McovS and wild-type SEZ.Numerous virulence factor-encoding genes and anabolicrelated genes in McovS that were involved in anti-phagocytosis,capsule formation,pathogenicity,and antibiotic resistance were downregulated significantly relative to the wild-type strain.This study revealed that the CovS plays a vital role in the establishment of SEZ virulence.

Keywords: Streptococcus equi subsp.zooepidemicus,covS,natural variation,virulence,regulation

1.lntroduction

Streptococcus equisubsp.zooepidemicus(SEZ),a group C streptococcus,is an important zoonotic pathogen that infects various animals,including horses,pigs,goats,chickens,dogs,and sometimes humans (Feng 1977;Kuusiet al.2006;Eyreet al.2010;Bisgaardet al.2012;Kerdsinet al.2021).In the equine and swine industries,infections caused by SEZ lead to severe health problems and economic losses (Feng 1977;Causeyet al.2010).However,despite the clinical relevance of SEZ,investigations on factors that contribute to the regulation of pathogenicity are limited.

Two-component systems (TCS) are mainly composed of a histidine kinase on the cell membrane that can sense external signals and a response regulator that has transcriptional regulatory activity in the cytoplasm.Bacteria sense changes in the external environment and regulate the transcription of target genes through the TCS,thus facilitates the survival and reproduction of bacteria and regulation of virulence(Watanabeet al.2008;Gotohet al.2010).Previous studies have shown that multiple TCS are involved in the regulation of streptococcal virulence.For example,Streptococcus suisutilizes various TCS such as SalK/SalR,CiaR/CiaH,Ihk/Ihr,and NisK/NisR for regulating virulence (Liet al.2008,2011;Hanet al.2012;Xuet al.2014).

Control of virulence regulator/sensor kinase (CovR/CovS) is the most important TCS inStreptococcus pyogenes(group A streptococcus,GAS),as it participates in the negative transcriptional regulation of various important virulence factor-encoding genes,such as thehasoperon,which is involved in the synthesis of the capsule;slo,which encodes type-O hemolysin;andnga,which encodes NAD glycoside hydrolase (Minamiet al.2010;Baoet al.2015).Natural mutations ofcovS,including frameshift mutations that lead to early termination of translation and mutations at key nucleotide sites result in loss of CovS function,are keys to the transformation ofS.pyogenesfrom a local infection to a systemic malignant infection,such as streptococcal toxic shock syndrome (Ikebeet al.2010;Tatsunoet al.2013;Baoet al.2015).Streptococcus mutansandS.suisdo not harbor acovSgene,and instead employcovRas an orphan response regulator,which is involved in the repression of transcriptional levels of virulence-associated genes,thus playing a role in the negative regulation of virulence (Panet al.2009;Dmitrievet al.2011).

In this study,for the first time a natural variation involvingcovSwas found in SEZ.Because natural variations involvingcovSare the key to enhancing the pathogenicity ofS.pyogenes,whether natural variation incovShas the same effect on SEZ virulence needs to be established.The results of this work provide crucial insight into the vital role ofcovSin the establishment of SEZ virulence,which is opposite to the situation inS.pyogenes.This study is helpful for the investigation of the pathogenic mechanism of SEZ,particularly involving the transcriptional regulation of virulence in SEZ.

2.Materials and methods

2.1.Strains,plasmids,and cells

The bacterial strains and plasmids used in this study are listed in Table 1.SEZ was cultured in Todd-Hewitt(TH) medium at 37°C.Escherichia coliDH5α was used as host for the pSET2 (Takamatsuet al.2001a),or pSET4s (Takamatsuet al.2001b),and was cultured at 37°C in Luria-Bertani (LB) medium.For construction of the recombinant plasmids or the selection of mutants,spectinomycin was added to the medium at 100 μg mL-1for SEZ and 50 μg mL-1forE.coli.The human epithelial cell line HEp-2 (ATCC) and mouse macrophage cell line RAW264.7 (ATCC) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) high-glucose supplemented with 10% fetal bovine serum (FBS) at 37°C under 5% CO2.

Table 1 Bacterial strains and plasmids used in this study

2.2.Deletion,replacement,and functional complementation of genes

The primers used in this study are listed in Appendix A.The schematic diagram of genetic modification is shown in Appendix B.

The upstream and downstream flanking sequences of the DNA fragment to be deleted were amplified and integrated together by PCR.In order to mutate thecovSof the wild-type strain,the mutantcovSsequence and its upstream and downstream sequences were amplified by PCR using the M35246 genome as the template.After digestion with restriction endonucleases,the PCR products were ligated into pSET4s.Recombinant vectors were electroporated into corresponding strains.Transformants were cultured at 37°C on TH agar with spectinomycin.A single colony was subjected to at least five passages at 28°C in TH broth without spectinomycin.Subsequently,the bacterium solution was diluted and plated on TH agar.Spectinomycin-sensitive colonies were verified for deletion or replacement of a particular DNA fragment by PCR,while also being subject to product sequencing.

The DNA fragment containing a whole ORF of a gene and its promoter sequence was PCR amplified and cloned into pSET2.The verified recombinant vector was electroporated into the corresponding mutant to construct the complementation strain.

2.3.Extraction of bacterial total RNA,transcriptome analysis,and qPCR assays

The bacteria were cultivated to logarithmic phase(OD600=0.6).The total RNA from wild-type ATCC35246,M35246,and McovSwas extracted using an RNeasyMini Kit (QIAGEN,Hilden,Germany) according to the manufacturer’s protocol.Transcriptomic analysis was conducted at Beijing Novogene Bioinformatics Technology Co.,Ltd.,China.RNA integrity was assessed using an RNA 6000 Nano Kit and an Agilent Bioanalyzer 2100 System (Agilent Technologies,Santa Clara,US).mRNA was purified from the total RNA to remove rRNA using a Ribo-Zero rRNA Removal Kit (Illumina,San Diego,US).Resultant mRNA was then randomly disrupted in fragmentation buffer.The reverse transcript of mRNA was completed with random primers.The second strand of cDNA was synthesized.The double-stranded cDNA was purified,end-repaired,A-tailed,and ligated with sequencing adapters.Then USER enzyme (NEB,Ipswich,US) was added to degrade the second strand of cDNA containing U.Agencourt AMPure XP beads(Beckman Coulter,Beverly,USA) were used to enrich cDNA prior to PCR and purify the PCR product to obtain the sequencing library.Library quality was assessed using an Agilent Bioanalyzer 2100 System.Library sequencing was conducted on the Illumina NovaSeq platform (Illumina,San Diego,US) to generate paired-end reads.

Raw reads in the fastq format were processed using in-house perl scripts.Clean reads were ultimately obtained by removing reads containing adapters,reads containing N bases,and low-quality reads from the raw reads.All subsequent downstream analyses used the clean reads with high quality.The HTSeq Program was used to count read numbers mapped to each gene(Anderset al.2015).Expression levels were evaluated by counting the reads from the constructed library that uniquely fall into the coding regions of the genome and calculated as the expected number of fragments per kilobase of transcript sequence per millions base pairs sequenced (FPKM).Three biological replicates were used as surrogates for statistical testing based on the DESeq Package in R environments (Anders and Huber 2010).Negative binomial testing was used to assess theP-value of the differential levels of expression.The Benjamini-Hochberg multiple-test correction was applied to evaluate the false-discovery rate (FDR),which is referred to as the Q value (Ghosh 2012).When the Q value was less than 0.05,this was deemed as the screening standard of genes with significantly different transcription.The KOBAS v2.0 Program (http://bioinfo.org/kobas/) was used for identifying Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations (Xieet al.2011).The Rockhopper v1.2.1 Program was used to analyze the operon,transcription start site (TSS),and transcription termination site (TTS) according to distribution of reads in the genome (McClureet al.2013).Then,a 700-bp sequence upstream of the TSS of the gene was extracted,and promoter prediction was performed using the Time-Delay Neural Network (TDNN)method (Hampshire and Waibel 1990).

qPCR assays were used to confirm the results of transcriptomic analysis.Three independent extractions of total RNA from each strain that was grown to the log phase were used.cDNA synthesis was performed using the PrimeScriptTMRT Reagent Kit (TaKaRa Bio Inc.,Kusatsu,Japan),and mRNA levels were measured with the SYBR?Premix Ex TaqTMKit (TaKaRa Bio Inc.,Kusatsu,Japan) according to the manufacturer’s instructions.The primers used in this study are listed in Appendix C.The 16S rRNA gene was amplified as references (Yiet al.2013b;Xuet al.2016a).Relative changes in gene transcription were calculated using the comparative CTmethod (Livak and Schmittgen 2001).Each set of qPCR was repeated thrice with independent RNA preparations.

2.4.Adherence and anti-phagocytosis assays

Adherence and anti-phagocytosis assays were performed as reported previously with some modifications (Meijerinket al.2012;Redlichet al.2012;Liet al.2018).The HEp-2 or RAW264.7 cells were cultured to approximately 5.0×105cells per well in 24-well plates and used in the assays.Immortalized cells were counted using Countess 3 Automated Cell Counters (Thermo Fisher Scientific,Waltham,US).Bacteria in the logarithmic growth phase were harvested by centrifugation (5 000×g,5 min) and then washed with DMEM.The bacterial concentrations were determined based on the corresponding relationships between the OD600readings of the bacterial suspensions and their plate counts.The bacteria were added to HEp-2 or RAW264.7 cells at a multiplicity of infection (MOI) of 1 or 10,respectively.After 2 h of incubation at 37°C under 5% CO2,the cells were washed with sterile 0.1 mol L-1PBS (pH=7.4).For the adherence assays,HEp-2 cells were added to double-distilled water to lyse the cells,and cell-associated bacteria (including adherent and invasive bacteria) were recovered.For the anti-phagocytosis assays,DMEM containing 10 μg mL-1penicillin and 200 μg mL-1gentamicin was added to RAW264.7 cells and incubated for an additional hour to kill extracellular bacteria.Then,after another three washes with PBS,the cells were lysed with double-distilled water to recover intracellular bacteria.Recovered bacteria were calculated using the plate count method.The assays were repeated as four independent experiments.

2.5.Capsule measurements

Relative quantification of capsule production by SEZ was performed as described previously (Blumenkrantz and Asboe-Hansen 1973;Xuet al.2016a).Briefly,bacteria grown in TH broth to an OD600of about 0.5 were collected by centrifugation.Every bacterial pellet from 4 mL of broth was washed once with 150 mmol L-1Tris-HCl (pH 7.0) and resuspended in 200 μL of the same buffer.After mixed with 1.2 mL of 12.5 mmol L-1tetraborate in 96% sulfuric acid,the suspension was incubated at 99°C for 5 min.Subsequently,the sample was mixed with 20 μL of 0.5% NaOH containing 0.15% m-hydroxydiphenyl,or without m-hydroxydiphenyl as a blank control for subtraction.Changes in color were measured spectrophotometrically at a wavelength of 520 nm.Experiments were repeated four times.

2.6.Antibacterial sensitivity tests

Antibacterial sensitivity was assessed using the paper diffusion method according to the Clinical and Laboratory Standards Institute (CLSI) antibacterial sensitivity test standards M02-A11 and M07-A9.5.0×107CFUs of bacteria were evenly coated on MHA plates containing 5%sheep blood and cultured in a 37°C,5% CO2for 20 h,and the inhibition zone radius was measured and counted.Each test was repeated thrice.

2.7.Mouse infection assays

Pathogenicity of different strains was analyzed in mice by measuring the median lethal dose (LD50) as reported earlier with some modifications (Fanet al.2009;Zhenget al.2013).SEZ was cultured in TH broth to an OD600=0.6.The bacterial pellet was collected by centrifugation and then washed with and resuspended in sterilized 0.1 mol L-1PBS (pH=7.4).The bacterial concentration was adjusted to 1×109CFU mL-1and serially 10-fold diluted until the lowest concentration of 1×102CFU mL-1was reached.A total of 100 μL of the bacterial suspensions were intraperitoneally injected into 4-wk-old ICR mouse.Five mice as a group were injected with the same concentration.Injection concentrations for each group ranged from 1×101CFUs per mouse to 1×108CFUs per mouse.After 7 d of challenge,the survival rates were recorded.LD50values were subsequently calculated using the Bliss method (Lieberman 1983).

Viable bacteria within organs were measured using a previously described method (Yiet al.2013a).Fourweek-old ICR mice were assigned to four groups,with six mice per group.SEZ in the logarithmic growth phase was washed with and resuspended in sterilized 0.1 mol L-1PBS (pH=7.4).Every group of mice was intraperitoneally injected with 5×105CFUs of SEZ suspensions of different strains,respectively.After 48 h of challenge,the viable bacteria in blood,heart,liver,spleen,lung,kidney,and brain of the infected mice were collected and homogenized in 0.1 mol L-1PBS (pH=7.4) at a mass to volume ratio of 1:10.Finally,the blood and six organ homogenates were diluted with 0.1 mol L-1PBS (pH=7.4)and plated on TH agar.Colony counts were subsequently performed after overnight incubation of plates.

2.8.Statistical analysis

Data were analyzed using GraphPad Prism version 7.0(GraphPad Software).Data are presented as means with standard deviations or standard errors.Statistical analyses for all pairwise comparisons were assessed using the unpairedt-test andP＜0.05 was considered significant.

3.Results

3.1.Discovery and identification of natural variation of covS in SEZ

In a preliminary investigation of the pathogenicity of SEZ,wherein the zeta-toxin-encoding gene (SeseC_01875) of ATCC35246 was deleted,a strain with a loss of upstream and downstream sequences including SeseC_01875 itself was unexpectedly found.The strain was referred to as M35246.DNA sequencing of corresponding fragments after PCR amplification indicated that when M35246 was compared to wild-type SEZ ATCC35246,25 contiguous genes from SeseC_01867 to SeseC_01898 including SeseC_01875 (29 838 bp;Appendix D) were lost.Meanwhile,the artificial deletion of gene SeseC_01875 was conducted,leading to construction of a strain termed ΔSeseC_01875.ΔSeseC_01875 did not lose other 24 genes.

RNA sequencing in transcriptome analysis of ATCC35246 and M35246 revealed 87 insertion/deletion mutations (InDels) throughout the transcript when M35246 was compared to ATCC35246 at the mRNA level(Appendix E).Among these InDels,an insertion mutation in the coding region ofcovSwas identified in M35246.PCR amplification and DNA sequencing ofcovS-coding sequence as well as its upstream and downstream sequences identified four single nucleotide substitution mutations and one single nucleotide insertion mutation in the coding sequence ofcovSwhen comparing M35246 to wild-type SEZ ATCC35246 (Fig.1-A).Only the first six amino acid sequences encoded by the wild-typecovSand the mutantcovSwere identical.Due to the frameshift mutation,the mutantcovSencountered the stop codon TAA after translation of the mRNA encoding 54 amino acids,resulting in premature termination of translation.Thus,the mutantcovScompletely lost the original function of the gene.

Fig.1 Schematic representation of natural variation of covS sequence.A,the regions that differ in the DNA sequences of wild-type and mutant covS.B,mRNA sequence of mutant covS that forms a terminator structure due to four single nucleotide substitution mutations and one single nucleotide insertion frameshift mutation.Nucleotides in green are single nucleotide insertion frameshift mutations of covS;nucleotides in red are nucleotide substitution mutations of covS.

The transcript levels ofcovSin M35246 and McovSwere both significantly lower than in wild-type ATCC35246 by about 27-fold (Q value=5.43×10-147) and 22-fold (Q value=1.88×10-23),respectively.There was no significant difference between M35246 and McovSin the transcript levels ofcovS(Appendix F).The ARNold Program(http://rssf.i2bc.paris-saclay.fr/toolbox/arnold/index.php)was used to predict transcription terminator structure,revealing that the mRNA of the mutantcovSform a hairpin structure,while that of the wild-typecovSdid not (Fig.1-B).The hairpin structure was followed by a poly(U) sequence,such that the two components together formed a terminator structure.This could explain why the transcription levels ofcovSin M35246 and McovSwere greatly reduced when compared to transcription within the wild-type ATCC35246.

Subsequently,after the 25-gene-deleted strain (ΔPI,abbreviation for Pathogenicity Island;Table 1) was artificially constructed,thecovSin ΔPI was not mutated.In addition,thecovSin the strain ΔSeseC_01875 was wild-type.Correspondingly,McovSwas constructed after artificially mutatingcovSto its natural mutate form observed in M35246 (Table 1).The 25 genes from SeseC_01867 to SeseC_01898 in McovSremained intact.These results indicated that there was no necessary connection between the 25 genes lost and the mutation ofcovS.Furthermore,the single 25-geneslost strain and the singlecovSmutated strain by natural mutation of wild-type ATCC35246 were not observed.The cause ofcovSnatural variation has not yet been established.

3.2.Mutation of covS affected SEZ adhesion to epithelial cells and resistance to macrophage phagocytosis

To determine whether the characteristics of M35246 are due to the loss of the 25 contiguous genes and/or the mutation ofcovS,a series of mutant strains that mentioned above and complementary strains including M35246 CcovSand CMcovSwere constructed.These strains are listed and described in Table 1.

Cell experiments showed that M35246 and McovSexhibited a significant increase in adhesion to HEp-2 cells relative to the wild-type ATCC35246,which increased by 1.84-and 1.14-fold,respectively (P＜0.05;Fig.2-A).In contrast,no statistical difference in adhesion to HEp-2 between wild-type ATCC35246 and M35246 CcovS,CMcovS,or ΔPI was observed (Fig.2-A).Compared to wild-type ATCC35246,the CFUs of M35246 and McovSrecovered from phagocytosis of RAW264.7 significantly increased by 11.03-and 12.72-fold,respectively (P＜0.001;Fig.2-B).No significant difference in anti-phagocytic ability was observed between M35246 and McovS.The CFUs of M35246 CcovSrecovered from phagocytosis of RAW264.7 significantly decreased compared to M35246 (P＜0.01).No significant difference in the antiphagocytic ability to Raw264.7 was observed between wild-type ATCC35246 and CMcovS,or ΔPI.These results indicated that the significant differences in adhesion and anti-phagocytosis to relevant cells between M35246 and wild-type ATCC35246 were mainly due to the mutation of thecovSand had nothing to do with the loss of the associated 25 genes.

3.3.Mutation of covS affected the growth rate,capsular formation,and antibiotic resistance of SEZ

Growth curve analysis showed that the growth rate of M35246 was slower than that of wild-type ATCC35246,but the concentration of bacterium during the stationary period was higher than that of wild-type ATCC35246.Compared to the wild-type strain,the growth rate of McovSwas markedly slower,and bacterium concentration during the stable period was also significantly lower.No difference in growth curves between wild-type ATCC35246 and CMcovS,M35246 CcovS,or ΔPI was observed (Fig.2-C).Relative quantitative determination of capsule formation showed that the content of capsular polysaccharide in M35246 and McovSsignificantly decreased compared to wild-type ATCC35246 (P＜0.001),whereas that in M35246 CcovS,CMcovS,and ΔPI did not significantly change.No significant difference in capsule content between M35246 and McovSwas observed (Fig.2-D).These results indicated that the significant difference in capsule production between M35246 and wild-type ATCC35246 was linked to the mutation incovS,and was not associated with the loss of the 25 gene sequence.Our previous research showed that the relative OD520values of the wild-type ATCC35246 were 0.2130±0.0003,while the relative OD520values of the capsular polysaccharide-deleted mutant derived from wild-type ATCC35246 were 0.0048±0.0012 (Xuet al.2016a).These showed that the capsule content in M35246 and McovSis only significantly reduced,but not completely lost.

Fig.2 A series of assays of wild-type ATCC35246,its mutants and complementation strains.A and B,the numbers of CFU of wild-type ATCC35246,M35246,M35246 CcovS,McovS,CMcovS,or ΔPI per 105 mammalian cells recovered from adherence of HEp-2 and phagocytosis of Raw264.7.C,growth kinetics of wild-type ATCC35246,its mutations,and complementation strains cultured at 37°C and 180 r min-1 aerobically in TH broth were shown.The OD600 of the culture was measured each hour.D,the quantification of capsule (CPS) production made by wild-type ATCC35246,its mutations,and complementation strains.For A,B,and D,data are mean and standard error of four independent experiments.For C,data are mean and standard deviation for the results of three independent experiments.Statistical significance was assessed using the unpaired t-test (ns,not significant;＊,P＜0.05;＊＊,P＜0.01;＊＊＊,P＜0.001).

Antibiotic susceptibility testing showed that the inhibition zone radii of kanamycin,clindamycin,azithromycin,clarithromycin,ampicillin,cefotaxime,cefepime,and gentamicin significantly increased with M35246 or McovScompared to wild-type ATCC35246(P＜0.05).The inhibition zone radii of linezolidamine and erythromycin were significantly larger with M35246 compared to wild-type ATCC35246 (P＜0.05).The inhibition zone radii of ceftriaxone,cefazolin,penicillin,and streptomycin significantly increased with McovScompared to wild-type ATCC35246 (P＜0.05).The inhibition zone radii of doxycycline and minocycline were significantly smaller with ΔPI compared to wild-type ATCC35246 (P＜0.05) (Table 2).

3.4.M35246 and McovS exhibited substantial loss of pathogenicity in mice

The differences in lethality of ATCC35246 with respect to M35246,McovS,and ΔPI in mice were determined.The results showed the LD50of M35246 and McovSin mice increased by about 105-fold relative to that of the wild-type ATCC35246,while ΔPI exhibited a 5-fold increase relative to the wild-type strain (Table 3).Further investigation indicated that M35246 or McovScould not be successfully isolated from the heart,liver,spleen,lung,kidney,brain,and blood of infected mice after 48 h of being challenged.Meanwhile,there was no statistical difference in viable bacteria populations in any organ or blood of infected mice between the ΔPI infection group and the wild-type strain infection group (Fig.3).These results indicated that the large attenuation of virulence in M35246 was mainly due to the mutation of thecovS.

Table 2 Antimicrobial susceptibility testing

Table 3 Calculation of median lethal dose (LD50) for Streptococcus equi subsp.zooepidemicus (SEZ) in mouse

3.5.Mutation of covS significantly downregulated the transcription levels of numerous SEZ virulence factor-encoding genes and anabolic-related genes

Based on the genome-wide annotation results of ATCC35246 (NCBI GenBank accession number CP002904.1),the transcriptome of wild-type ATCC35246,M35246,and McovSwere analyzed.The numbers and details of differential genes when compared every two strains are shown in the Fig.4-A-D and Appendix F.

Some virulence factor-encoding genes with important biological functions and significant differences in transcriptome analysis were selected for qPCR validation.The results showed that the differences in the transcription levels of the ATCC35246,M35246,and McovSgenes coincided with those of transcriptome analysis,indicating the reliability of the transcriptome analysis results presented here (Fig.4-E).

Fig.4 Transcriptome analysis and qPCR assays.A-C,Volcano plots of significantly differentially expressed genes.Genes showing significant differential expression are indicated by red dots (upregulated,UP) and green dots (downregulated,DOWN).Genes with no significant differential expression were represented by blue dots (no significant,NO).Abscissas represent fold change of genes in different samples.Ordinates represent genes statistical significance of differences in expression changes.D,Venn diagram of significantly differentially expressed genes.The sum of the numbers in each large circle represents the total number of differential genes in the comparison combination,and the overlapping portions of the circles represent the differential genes shared between the combinations.E,the relative mRNA levels of test genes in wild-type ATCC35246,M35246,and McovS by qPCR assays.Data represent mean and standard error of three independent experiments.Statistical significance of comparisons between wild-type ATCC35246 and M35246 or McovS were determined using the unpaired t-test (＊,P＜0.05;＊＊,P＜0.01;＊＊＊,P＜0.001).

The results of transcriptome analysis of the genes encoding the important virulence-associated factors of SEZ are summarized in Table 4.The results showed that the mutation of SEZcovSled to a significant decrease in the transcriptional levels of a large number of virulence factor-encoding genes such as thehasoperon,szm,fbpZ,srtA,srtC,prtS,5nuc,scpZ,sse,andarcoperon.Conversely,the transcription levels of some other virulence factor-encoding genes such asszp,enuc,cspZ,andslaawere significantly increased in McovScompared to wild-type ATCC35246.Meanwhile,significant reductions in transcript levels of some virulence factor-encoding genes,such asskc,sfs,shr,endoS,andhtpZ,in M35426 relative to ATCC35246 were found to be independent ofcovS,as there were no significant differences in transcript levels of these genes in McovScompared to wild-type ATCC35246.In addition,the significant differences in transcriptional levels of the common virulence-related factors inS.pyogenes covSmutant relative to the wild-type strain were summarized based on the other studies (Baoet al.2015).It was found that there were numerous differences in the regulation of the same virulence factors by CovS between the twoStreptococcusstrains.

Table 4 Fold changes of significantly differentially expressed chromosomal virulence genes1)

KEGG functional annotation and clustering analysis was subsequently conducted for significantly differentially expressed genes.Compared to the wild-type strains,M35246 exhibits significantly lower transcript levels of genes belonging to 55 KEGG pathways,such as pathways of ribosome and RNA polymerase synthesis,phosphotransferase system (PTS) pathway,ABC transporters pathway,beta-Lactam resistance pathway,multiple carbohydrate and amino acid metabolism pathways,and purine and pyrimidine metabolism pathways (Fig.5-A;Appendix G),as well as significantly higher transcript levels of genes belonging to 54 KEGG pathways,such as fatty acid biosynthesis and metabolism pathways,multiple antibiotic-resistancerelated pathways,terpenoid backbone biosynthesis pathway,peptidoglycan biosynthesis pathway,biotin metabolism pathway,and gene repair pathways (Fig.5-B;Appendix G).Relative to wild-type ATCC35246,McovSshowed significantly lower transcript levels of genes belonging to 54 KEGG pathways,such as ribosome and RNA polymerase synthesis pathways,beta-Lactam resistance pathway,oxidative phosphorylation pathway,pentose phosphate pathway,folate biosynthesis pathway,ascorbate and aldarate metabolism pathway,purine and pyrimidine metabolism pathways,and metabolic pathways (Fig.5-C;Appendix G),as well as significantly higher levels of transcription of genes belonging to 34 KEGG pathways,such as valine,leucine,and isoleucine degradation pathways,and base excision repair pathway(Fig.5-D;Appendix G).Compared to M35246,McovSexhibits significantly lower transcription levels of genes belonging to 55 KEGG pathways,such as ribosome and RNA polymerase synthesis pathways,oxidative phosphorylation pathway,pentose phosphate pathway,multiple group B vitamins metabolic pathways,purine and pyrimidine metabolism pathways,and generally metabolic pathways (Fig.5-E;Appendix G),as well as significantly higher transcription levels of genes belonging to 29 KEGG pathways,such as base excision repair pathway,and ABC transporters pathways (Fig.5-F;Appendix G).These results suggested impacts of other non-covSmutations,including the loss of 25 genes,on the regulation of some of the genes indicated above.

Fig.5 Differentially expressed gene-enriched KEGG pathway scatterplots.The scatterplots showed KEGG pathways enriched with downregulated or upregulated differentially expressed genes when two strains are compared.The ordinate indicates the pathway names,and the abscissa represents the Rich factor.Rich factor refers to the ratios of the number of difference genes enriched in the pathway to the number of annotation genes.The larger the Rich factor,the greater the degree of enrichment.The size of the dot indicates the number of differentially expressed genes in the pathway.

When McovSwas compared to wild-type strain or M35246,there was 20 or 26 more KEGG pathway containing genes with significantly lower expression levels than pathways containing genes with significantly higher expression levels.Compared to the wild-type strains,there were numerous downregulated pathways in M35246 and McovSin common,whereas there were many differences in the upregulated pathways.ThecovSis more involved in the positive transcriptional regulation of genome genes in SEZ.

4.Discussion

CovR/CovS is closely related to disease phenotype and pathogenic ability ofS.pyogenes.Numerous studies have shown that thecovSofS.pyogenesafter serially passage in experimental animals or isolated from severely infected individuals tends to undergo a lossof-function mutation,indicating thatcovSmutations are key to enhancing virulence inS.pyogenes(Garciaet al.2010;Tatsunoet al.2013;Masunoet al.2014).The CovR/CovS ofS.pyogenesparticipates in the transcriptional regulation of approximately 15-18% of whole chromosome genes,including thehasoperon,M protein-encoding genes,prtS,scpA,ska,sfb1,ideS,and many other important virulence factors (Grahamet al.2002;Baoet al.2015).In terms of substrate utilization,the mutation involvingcovSinS.pyogenesresulted in the upregulation of genes that are related to ingestion and metabolism of polysaccharides,glycoproteins,fatty acids,and co-enzyme factors,as well as those involved in ingestion of heme and participate in the synthesis of purine and glutamate (Baoet al.2015).

SEZ belongs to group C streptococcus;although it shows 70-80% homology withS.pyogenes(Holdenet al.2009;Maet al.2013).The virulence factor-related genes found in SEZ are almost entirely present inS.pyogenes(Table 4).These two kinds of streptococci both can cause suppurative local infections,septicemia,meningitis,streptococcus shock syndrome,and a series of similar symptoms in the host (Eyreet al.2010;Ikebeet al.2010).To some extent,these findings indicate that these two kinds of streptococci have many similarities in pathogenicity.CovR/CovS is also present in SEZ.Details on the regulation of the CovR/CovS in this strain are unknown.No reports on spontaneous mutations or genetic engineering of CovR/CovS genes of SEZ are currently available.ThecovSin M35246 exhibited mutations at the beginning of the open reading frame resulting in a frameshift mutation and early translational termination,in addition to the formation of a terminator structure that inhibited transcription.M35246 exhibited greatly reduced the transcription level ofcovS.Although somecovSmRNA was present,it could not be translated into normal CovS.Therefore,there is no doubt that thecovSin M35246 is completely inactivated.Therefore,this study for the first time found a SEZ strain presented a natural variation in thecovS,which caused it to completely lose the function of CovS.

In this study,we constructed the McovS,which exhibited the same mutation ascovSin M35246 and complete loss of CovS function.Just like theS.pyogenes covS-deleted strain and its corresponding wild-type strain(Baoet al.2015),McovSshowed a marked decrease in growth rate in TH broth relative to the wild-type SEZ strain.However,compared to wild-type ATCC35246,the pathogenicity of M35246 and McovSin mice was drastically reduced,and rapidly cleared from the mouse body after intraperitoneal injection,and it significantly reduced the anti-phagocytic ability and capsule production of bacteria,which is contrary to the change in traits caused byS.pyogenes covSmutations (Tatsunoet al.2013;Baoet al.2015).These results suggest that the regulatory mechanism of CovS in pathogenicity differs between the two strains,as the CovS of SEZ acts as a virulence activator,while the CovS ofS.pyogenesacts as a virulence repressor.

Transcriptome analysis showed that the SEZcovSis essential to maintain the transcription levels of multiple virulence-factor encoding genes,such as thehasoperon,szm,fbpZ,srtA,srtC,fszE,fszF,prtS,5nuc,scpZ,sse,andarcoperon.In these virulence factors,S.pyogeneshas been reported to upregulate the transcript levels of thehasoperon,szm,prtS,andscpZdue to the loss of function ofcovS,which was diametrically opposed relative to SEZcovS(Table 4).Many other virulence-factor coding genes,such asideZ,thesagoperon,skc,shr,eno,endoS,andhtpZin whichS.pyogenesCovS participates in regulation were not regulated by CovS in SEZ.In SEZ,enuc,5nuc,sse,slaa,and thearcoperon,in which CovS participates in regulation,are also not reported to be regulated by CovS inS.pyogenes.Besides,in thecovSmutant strain ofS.pyogenes,the transcription levels of genes related to purine synthesis (purC,purL,purF,purM,purN,purH,ami,purD,purE,purK,andpurB) (Jenkinset al.2011;Baoet al.2015) and heme ingestion (Shr-Shp-HtsABC system) (Nygaardet al.2006a,b) have been reported to be upregulated to adapt to internal environments with low purine and iron ion contents in the host (Anzaldi and Skaar 2010;Lobelet al.2012;Choby and Skaar 2016).However,in the present study,transcriptome analysis showed that except forpurB,which is positively regulated by SEZ CovS,the rest of the genes that are related to purine synthesis and heme ingestion are not regulated by CovS in SEZ (Appendix F).Moreover,the participation of SEZ CovS in the positive regulation of thearcoperon helps SEZ in surviving in acidic environments(Xuet al.2016b).Therefore,the differential expression of genes that are related to virulence factors and important metabolism pathways can basically explain the positive regulatory mechanism for SEZ pathogenicity of CovS,which is opposite to that ofS.pyogenes.This suggests that even highly homologous proteins from closely related bacteria of the same genus may exhibit very different functions.These functional differences may be the reason why different bacteria prefer to infect different hosts and cause different symptoms.

This study found that the transcription levels ofcovRwere significantly upregulated in M35246 and McovSrelative to the wild-type ATCC35246.Researches on the CovR/CovS two-component system ofS.pyogenesshowed that the response of CovS to external signals and the phosphorylation of CovS on CovR are important for CovR to exert its transcriptional regulation (Horstmannet al.2015).It is suggested that in view of inactivation of CovS,CovR cannot perform its original function due to the loss of phosphorylation of CovS,which causes the SEZ to increase the transcription ofcovRthrough some transcriptional regulation pathway accordingly.Meanwhile,lacking of phosphorylation modification of CovS,the increase in the expression level of CovR cannot make CovR fully play its due role.

Some genes influencing the anti-phagocytosis phenotype of SEZ were reported in one of our previous studies (Xuet al.2019),such ashasA,szm,adhP(SeseC_00054,encoding an alcohol dehydrogenase),guaA(SeseC_01336,encoding a GMP synthase),SeseC_00088 (encoding a membrane protein),lacI(SeseC_00423,encoding the sucrose operon repressor),and a polycistron includingfbpZ,srtA,srtC,fszE,andfszF(SeseC_02469-SeseC_02473),significantly decreased in the transcription levels in McovS,compared to the wild-type ATCC35246 (Appendix F).Thehasoperonexpressed proteins involved in production of capsular polysaccharide and an M-like protein encoded byszmhave also been reported as important virulence factors for SEZ resistance to phagocytosis in other studies(Boschwitz and Timoney 1994;Weiet al.2012;Velineni and Timoney 2013).These results described above can explain the significant decrease in anti-phagocytic ability of McovSand M35246,when compared to the wild-type strain,which indicates that the normal expression ofcovSis important for the SEZ resistance to phagocytosis.Furthermore,no significant differences in the transcription levels ofhasoperon genes between McovSand M35246 indicates thatcovS,rather than the 25 gene deletion,was responsible for the transcriptional regulation of thehasoperon.Additionally,the decrease of capsule production caused by the lower transcript levels ofhasoperon genes due to the mutation ofcovScan explain the enhanced adherence abilities to barrier cells of McovSand M35246 relative to wild-type ATCC35246,as described in a previous study (Xuet al.2016a).

In terms of antibiotic resistance,the results of antibiotic sensitivity experiments showed that compared to the wild-type strain,the radius of inhibition zone of McovSto β-lactam antibiotics such as penicillin,ampicillin,ceftriaxone,cefazolin,cefotaxime,and cefepime significantly increased.Transcriptome analysis showed that the transcription levels of 12/17 genes involved in β-lactam resistance in McovSsignificantly decreased relative to ATCC35246 (Fig.5-C;Appendix G) which coincides with the results of our antibiotic sensitivity experiments.Furthermore,the deficiency ofcovSin SEZ causes a significant reduction in transcription levels of 43/65 ribosome-encoding genes,4/5 RNA polymeraseencoding genes,genes related to synthesis and metabolism of 16 kinds of amino acids,20/82 aminoacyltRNA synthesis-encoding genes,and seven F0F1ATP synthetase-encoding genes (Fig.5-C;Appendix G).As a result,the overall protein synthesis and amino acid metabolism levels of bacteria were low,which is the main reason for the markedly slow growth rate of McovS,and can explain to a large extent why McovSis more sensitive to various antibiotics that affect protein synthesis such as aminoglycosides (streptomycin,gentamicin,and kanamycin),macrolides (azithromycin and clarithromycin),and lincosamides (clindamycin) compared to wild-type ATCC35246.

5.Conclusion

This paper systematically studied the function of SEZ CovS in bacterial anti-phagocytosis,capsule formation,pathogenicity,antibiotic resistance,and transcriptional regulation of various important virulence factors and key metabolic systems.Furthermore,transcriptome analysis was used to demonstrate the regulatory mechanism of CovS in anti-phagocytosis,capsule formation,pathogenicity,and antibiotic resistance.This work filled the gap on the regulation of pathogenicity and antibiotic resistance factors involving SEZ,thereby revealing the diversity of the TCS in different bacteria,and indicating that the CovS plays a vital role in the establishment of SEZ virulence.

Acknowledgements

This study was supported by the National Key Research and Development Program of China (2021YFD1800400),the National Natural Science Foundation of China(31872480),the Jiangsu Agriculture Science and Technology Innovation Fund,China (CX(19)2020),and the Priority Academic Program Development of Jiangsu Higher Education Institutions,China (PAPD).

Declaration of competing interest

The authors declare that they have no conflict of interest.

Ethical approval

All animal procedures and experiments were performed according to protocols approved by the Ethical Committee for Animal Experiments of the Nanjing Agricultural University,China (Protocol approval number: PZ2019192)and performed in accordance with the National Institute of Health Guidelines.

Appendicesassociated with this paper are available on http://www.ChinaAgriSci.com/V2/En/appendix.htm