Proteomic Analysis of Virulence Proles in Clinical Strains of Shigella exneri in Manaus – Amazon State

Background: Shigella is a Gram-negative bacterium and belongs to Enterobacteriaceae family. These bacteria have been described as responsible for many diarrheic infections around the world and affects children as 5 years old. As a striking feature of these bacteria, we can say about the invasive capacity and the severe damage in the intestine of the host. Epidemiological studies conducted during 2007 to 2009 by our research group, Diagnosis and Control of Infectious Diseases of the Amazon - DCDIA, identied Shigella as the 4th most frequent bacterial pathogen in children with diarrhea treated in public hospitals in Manaus - AM. To understand the mechanisms of pathogenesis of these clinical strains, and to describe the mechanisms of cellular invasion, this study proposes the identication of the proteome of two isolates, through the mass spectrometry coupled to liquid chromatography. The clinical strains were submitted to experimental conditions that mimic the epithelial cellular contact in the host, using the inductor Congo Red, in order to investigate which proteins are being produced by this pathogen. Results: The proteomic prole of Shigella strain 201 reveals 386 intracellular proteins cultivated in LB medium and 189 intracellular proteins cultivated using the Congo Red inductor. For the M90T strain, a total of 470 intracellular proteins were detected in LB medium and 383 intracellular proteins cultivated with Congo Red. The ndings reveal that proteins exclusively induced by Congo Red in the clinical strain are related to virulence processes, such as IpaC and IpaD proteins, which have already been extensively investigated in the literature. Conclusions: However, new target proteins are pointed out, such as Hmp, YkfE, AepA, MobC, MetK, OsmY, LptA and LuxS which are classied as proteins predicted as pathogenic, based on our analyses. Although such proteins are involved in the virulence of enteric pathogens, their functions are still little explored or inexistent for the Shigella genus, mainly in the northern region of Brazil. We expected this work to reveal the mechanisms underlying the isolated clinical strains and elucidate new effectors and how they modulate the pathogenesis of these bacteria. the LuxS of the of activated methyl (Activated Methyl Cycle – In the methionine is converted to S-adenosylmethionine (SAM) by the enzyme SAM synthetase (MetK). The methyl group of S-adenosylmethionine to other acceptors results in S-adenosylomocysteine (SAH). This event has been reported as an important ag in the reactions involving the quorum sensing in bacteria, using an inductor (autoinducer2: AI-2)(17).

Background WHO reports that diarrheal diseases mainly affect children under 5 years of age that do not have access to basic sanitation and clean water. It is estimated that around 1.5 million deaths occur annually due to diarrheal diseases and are considered a disease of child morbidity and mortality (1)(2)(3). The successive episodes of diarrhea compromise the socioeconomic status of endemic areas and the physical and intellectual development of affected children. The consequences of long-term continuous infections are drastic and are manifested, for example, by de ciencies in growth and increased susceptibility to other pathogens and chronic diseases such as diabetes (4,5) Among the main pathogens causing diarrhea, Shigella is observed in patients with this infectious condition (6) .This genus has shown constant resistance to antimicrobials in different countries and a higher incidence of cases in low-income countries, in addition, the spread of strains with a high capacity to transfer virulence factors and capable of causing severe damage to the host have been worrying factors, since there is still no effective vaccine against Shigella spp. Although this context is global, in Brazil, more speci cally in the Northern region of the country, few studies aim to understanding the different aspects of this pathogen, either in epidemiological surveillance, in understanding mechanisms related to the pathogenesis or in strategies for its diagnosis, control and treatment.
Epidemiological studies conducted in 2007-2009 by our research group, Diagnosis and Control of Infectious Diseases of the Amazon -DCDIA, identi ed Shigella as the 4th most frequent bacterial pathogen in children with diarrhea treated in public hospitals in Manaus -AM. The group also veri ed the presence of virulence genes among the isolated samples, which were related to the symptoms presented by the patients as Shet1B and Shet2 genes. Besides these, other genes were also detected, such as IpaBCD, IpaH7.8, set-1A, set-1B, sen/ospD3, virF, e invE (7).
Based on previous researches of the group, among the evaluated isolates, we highlight the immunogenic potential of the strain S. exneri strain 201,which showed greater invasive capacity, greater lethality and morphological changes in the infected tissue such as hemorrhage, intense cell in ltration and with destruction of bronchial and alveolar epithelia, when compared to the reference strain in the study. In addition, this isolate presented a gene expression similar to the reference strain for the main effectors associated with the cell invasion process (8,9). However, since it is not yet known the totality of the effectors that are secreted and their respective functions , the whole mechanism of Shigella pathogenesis continues to be the target of new investigations, mainly through the proteomics approach (10).
Since not all the effectors that are secreted by this pathogen and that a part of a panel of proteins with the most different functions are known (11). Studies using in vivo and in vitro models, attenuated or mutant strains have contributed to the development of strategies aimed at the production of vaccines against shigellosis has provided a basis for the understanding of the pathogenesis of this microorganism (12,13).
In this study, we propose the identi cation of proteins expressed by these clinical strains with different virulence pro les during contact with Congo Red (CR) using the proteomic approach. We believe that the identi cation of the proteomic pro le has the potential to provide knowledge on how the genetic differences presented by the strains of the study re ect on the proteome, giving the basis for future research in understanding the mechanism of invasion by these clinical strains.

Bacterial strain and Growth Condition
Shigella exneri 5a M90T used as reference strain and Shigella exneri strain 201 isolated from the patient as previously described (7), obtained from Hospital in Manaus. The isolates were kept frozen in 60% glycerol at -80ºC before to the experiments. They were subcultivated in Luria-Bertani (LB) medium grown, pH 7.0, at 37ºC with constant agitation for 24 hours. Before each experiment, the colonies were inoculated in fresh nutrient media (50 mL) with a 1,5x10 8 CFU of the overnight culture, with or without the Congo Red (CR) stain and grown at 37 ºC with vigorous agitation.

Bacterial Growth curve
The growth (measured in log 10 CFU/mL) of Shigella spp. strains were performed in triplicate in LB medium broth. The growth curve of the Shigella exneri 5a M90T control was compared with those Shigella exneri 201 strains, both with (0,01%) and without CR staining. OD 600nm was measured at 1-hour intervals until the decline phase.
Protein extraction, precipitation and quanti cation For intracellular proteins: the cells were pelleted and washed with 3 mL (Tris-HCl 50mM, pH 7.5), centrifuged for 15 min, 42727 xg at 4 ºC. Then, the pellet was resuspended in 1 mL of lysis buffer (7M Urea, 2M thiourea, 4%CHAPS, 50 mM DTT -Dithiothreitol, 1 mM PMSF -Phenylmethylsulfonyl uoride) and sonicated in three cycles of 30 s, at 4ºC. The samples were incubated for 1 hour in ice, then centrifuged for 30 min, 114462 xg at 4ºC. Each 100 µL of supernatant was redistributed into microtubes and precipitated with 5V of methanol. The pellet was resuspended in 1 mL methanol, centrifuged for 15 min, 12.000 rpm at 4ºC twice. After washes, the pellet was dried in room temperature and stored at -20ºC. Three samples were collected to perform quanti cation using 2D Quant Kit (GE Healthcare), as manufacturer´s instructions. In-gel digestion of intracellular proteins

1-DE gel electrophoresis and analysis
The gel slices were submitted to in-gel digestion protocol (14) with modi cations. Brie y, the gel slices were destained in 500 µL of solution 1 (50% methanol, 2,5% acetic acid) for 3 hours at room temperature and dehydrated with 200 µL of acetonitrile for 5 minutes. The gel slices were incubated in 50 µL DTT 10 mM at room temperature for 30 min, then 50 µL Iodoacetamide (IAA) 50 mM for same time and temperature, in the dark. The gel slices were washed with ammonium bicarbonate 100 mM for 10 min. The solution was removed, and the gel slices were dehydrated in 200 µL of acetonitrile for 10 minutes and rehydrated with ammonium bicarbonate 100 mM for 10 minutes. This step was repeated once and after dehydrating the samples with acetonitrile, the residual volume was evaporated in freeze drier. A trypsin solution (20ng/µL) was added in each microtube and the slices were rehydrated for 30 minutes in ice. The solution excess was removed and for in-gel digestion, the samples were incubated overnight with 20 µL ammonium bicarbonate 50 mM at 37ºC. In the other day, the slices were submitted to peptides extraction. The slices were incubated in 10 µL formic acid 5% solution for 10 minutes and the supernatant transferred to another microtube. The second step was incubated the gel slices in 12 µL 5% formic acid/50% acetonitrile solution for 10 minutes and the supernatant transferred to another microtube, with previously supernatant. The samples were dried and stored at -20ºC until mass spectrometry analysis.

Mass spectrometry and data analysis
For protein analysis, peptides (4.5 µL) were separated by C18 (100 µm x 100 mm) RP-nanoUPLC (nanoAcquity, Waters) coupled with a Q-Tof Premier mass spectrometer (Waters) with nanoelectrospray source at a ow rate of 0.6 µL/min. The gradient was 2-90% acetonitrile in 0.1% formic acid over 45 min. The nanoelectrospray voltage was set to 3.5 kV, a cone voltage of 30 V and the source temperature was 100ºC. The instrument was operated in the 'top three' mode, in which one MS spectrum is acquired followed by MS/MS of the top three most-intense peaks detected. After MS/MS fragmentation, the ion was placed on exclusion list for 60 s and for the analysis of endogenous cleavage peptides, a real time exclusion was used. The spectra were acquired using software MassLynx v.4.1 and the raw data les were converted to a peak list format (mgf) without summing the scans by the software Mascot Distiller Protein interaction prediction was made using the computational tool Search Tool for Retrieval of Interacting Genes (STRING), available at <https://string-db.org/>. Identi cation of proteins was performed using the Panther Data Base Classi cation System (Panther DB), available at http://www.pantherdb.org/ and Cello2go (16). Potentially pathogenic proteins were predicted using Predict Pathogenic Proteins in Metagenomic Datasets (MP3) software, following the following parameters: threshold score: -0.2, which sets the sensitivity of 82.53%, speci city of 86.97% and precision of 86.02%.

Results
Growth curves of Shigella spp. wild type (M90T) and clinical strain 201. Figure 1a shows the growth curve of Shigella spp. wild-type strains (M90T), and Figure 1b, the growth curve of the strain S. exneri 201 in liquid culture medium LB. The sequential sampling data were used from the assays done every 1 hour for 35 hours. It can be observed in both Figures (1a and 1b) that S. exneri 201 strain and S. exneri M90T strain grow similarly until 10 hours. In both conditions, it can be observed that the decline phase is about 30 hours, but in Figure 1a, S. exneri 201 strain in CR condition showed a signi cant rate of decline in relation to LB condition, with a p-value of 0.0052. For protein extraction, the 10-hour growth culture was collected, extracted and analyzed in one-dimensional gel electrophoresis, as previously described in Methods.
Proteomic pro les of Shigella spp. strain 201 and M90T Figure 2 shows the one-unidimensional gel pro les of M90T and strain 201, and the selected lanes to excised and submitted to tryptic digestion. In total, twelve lanes were excised for LB condition in M90T and eleven lanes, for the CR condition. For the 201 strain, thirteen lanes were excised for both LB and CR condition. The total proteins identi ed in mass spectrometry experiments are shown in Table 01. After the identi cation of the proteins, the results were analyzed and compared, and the proteins that appears in two independent experiments were considered for analyses. Figure 3 shows the comparative proteins between the conditions and the bacterial strain. This study aims to understand the biological mechanisms by which the S. exneri 201 strain has more virulence, and which are the possible metabolic pathways presented in CR conditions. All proteins compared were analyzed by the Veen diagram shown in Figure 3 (C) and the gene ontologies were researched with proteins from both strains and conditions.
Biological processes and molecular functions in both bactéria strain cultivated in LB and CR medium Using the STRING program, biological processes, molecular functions and cellular components were mapped for each experimental condition. For the M90T condition grown in the LB medium, 470 proteins with Uniprot entries were identi ed, 299 proteins with an annotated gene name and of these, 110 proteins mapped in the STRING platform ( Figure 4). Of these 149 proteins mapped, 125 biological processes related to the identi ed proteins were observed. For the M90T condition grown in the environment with the presence of the inductor VC, it was identi ed 383 proteins with entries in Uniprot, 249 mapped with the gene name and only 134 in the STRING platform. As a result, 149 biological processes were described for the M90T-CR condition.
In  Table S1). The result for this condition was the identi cation of 63 associated biological processes.
From these results, the exclusive proteins of S. exneri 201 grown in CR were evaluated together with the biological processes mapped by STRING and Cello2GO in search of a more holistic understanding of the metabolism of this bacterium and how it differs from the wild strain in relation to its virulence capacity ( Figure 5). As observed, in both strains, in molecular function categories we can observe the response to stress and pathogenesis categories.The following results discuss some proteins with implications on the virulence of the bacterium, host infection and adaptation in aerobic and anaerobic conditions. Using the Predict Pathogenic Proteins in Metagenomic Datasets (MP3), Figure 6 indicates the main proteins predicted to be involved with virulence of the S. exneri 201 strain grown in CR medium.

Discussion
The Works indicate that Gram-Negative bacteria secrete three types of signaling molecules, called selfinducers. When the accumulation of these autoinducers occurs as a function of cell density (Quorum sensing), a cascade of signaling is activated leading to modi cations in the patterns of gene expression and adaptive responses (18)(19)(20)(21)(22)(23)(24); .Another important fact demonstrates that the inhibition of an enzyme related to this process (Methylthioadenosine/S-adenosylhomocysteine nucleosidase) limits the synthesis of self-inducers, decreasing bio lm formation and attenuating virulence. The presence of two proteins associated with this metabolic pathway may suggest their participation in the virulence of the clinical strain 201. Reviews of (25) report that strains of Neisseria meningitidis have similar cellular machinery in colonizing their hosts, but the difference in expression of some of their genes and their regulation determines their virulence capacity. The work suggests that genes involved in responses to oxidative stress and glutathione metabolism during infection may be key in determining strain virulence. Although not exclusive to the CR condition, proteins such as Thiol peroxidase (TPX; A0A090NWB6) and Thioredoxin/glutathione peroxidase (BtuE; A0A127GKP1) were identi ed in the present proteome and are proteins associated with oxidative stress and glutathione metabolism.
Additional experiments to validate the expression of these proteins in the CR condition should guide future discussions if the clinical strain 201 has pathogenicity mechanisms like those described for N. meningitidis.
Associated with these factors, the authors discuss that metabolic adaptations of the strains are fundamental for the bacterium to infect its host e ciently, reporting previous ndings that associate with the term "nutritional virulence", i.e., they are speci c mechanisms that the bacteria perform.

The protein Flavohemoprotein expressed in the strain S. exneri 201 CR is responsible for detoxi cation of nitric oxide and is associated with virulence in bacteria
Among the exclusive proteins of the condition of the clinical strain with the CR inductor, the expression of the avorhemoprotein protein (Hmp; A0A1S9KCV2) was observed. In aerobic conditions, this protein is associated to the detoxi cation of nitric oxide (ON), protecting the bacteria from several nitrogenous compounds, and playing a central role in the bacterial response to nitrosative stress. This protein can reduce NO to N2O (nitrous oxide) under anaerobic conditions (26). The literature reports that the body's defense cells are able to produce nitric oxide to prevent microbial infection. Some bacteria have already been described as capable of overlapping this host defense mechanism, producing proteins related to detoxi cation to ON and this is related to the virulence capacity of these bacteria, as already observed in Mycobacterium tuberculosis, Neisseria meningitides, Vibrio cholerae, Salmonella enterica serovar Typhimurium, Pseudomonas aeruginosa, and enterohemorrhagic Escherichia coli (EHEC) (27).
The lysozyme type C inhibitor (YkfE) overlaps the host's defense mechanisms and is expressed exclusively by S. exneri 201 CR.
One of the mechanisms studied to combat bacterial invasion is the expression of lysozyme by the host, which results in the hydrolysis of the peptidoglican bacterial wall. In contrast, bacteria have developed systems to superimpose the host's defense mechanisms, and one of these mechanisms lies in the expression of lysozyme inhibitors. There are 3 bacterial defense systems: modi cation of the peptidoglican wall to resist hydrolysis by lysozyme, modi cations in the load and integrity of the envelope and expression of inhibitors of this enzyme (28). In the present work, we identi ed the expression of the protein C-lysozyme inhibitor (YkfE or Ivy; A0A0F6MA48) exclusively in the clinical strain grown in the medium with the VC inductor.
This protein has already been described in other bacteria as Pseudomonas aeruginosa and has been reported as a virulence factor for Gram-negative bacteria. Inhibition occurs through a loop protrusion in Ivy protein that occludes the active site of lysozyme via a lock-and-key mechanism (29). Its presence exclusively in the condition of the clinical strain with the CR inductor suggests that this may be a mechanism that potentiates the virulence of this strain.
The expression of the exoenzyme AepA is only present in condition 201 CR and is associated with the virulence of bacteria.
Our proteomics studies demonstrated that the strain 201 grown in medium with the CR inductor is capable of expressing a protein called Exoenzymes regulatory protein (AepA; D6BD12). In the literature there are few reports of AepA protein, but some publications discuss that AepA protein has been described as a transcriptional activator of the enzymes pectato liasase, cellulase, polygalacturonase and protease in Erwinia carotovora subsp. carotovora (30,31). Reviews indicate that the synthesis of exoenzymes and enzyme secretion is related to virulence aspects in bacteria. In E. carotovora, it was reported that the expression of exoenzymes is closely related to certain compounds of the plant that it infects and that this response is dependent on the presence of the genes aepA and aepB (32).

The MobC protein facilitates the transmission of virulence factors and is expressed exclusively by S. exneri 201 CR
This protein is part of the transfer system called MOB, which is composed of six families, including MobC. This system acts in the conjugation process, transferring moving elements, such as plasmids to a donor cell. In order to have the transference, the DNA is cleaved by the relaxases, which are proteins of multiple domains, allowing the mobilization of the genetic elements, called conjugative plasmids.
Several plasmids of MobC are found in different classes of bacteria, as in gamaproteobacteria, rmicutes and tenricutes.Clinical Isolates of E.coli, Yersinia enterocolítica and Yersinia pseudotuberculosis share such genetic elements as well as other enterobacteria, thus suggesting that such mechanism facilitates the transmission of virulence factors.
Furthermore, MobC counterparts are identi ed in E.coli and Klebsiella pneumonie, for example, as well as other bacteria that share plasmids, which contain relaxases, involved in antibiotic resistance (33).
Genetic studies indicate that bacteria such as Yersinia, Salmonella and Klebsiella share similar sequences of transfer elements, such as MobC, indicating genomic ow between enterobacteria (34).
A recent study identi ed the MobC gene present in a plasmid of virulence in Shigella exneri, such plasmid confers resistance to antibiotics of clinical importance, however, the study did not determine the ability of this plasmid to be transferred by conjugating (35). Studies relating MobC to virulence mechanisms are still scarce, especially in bacteria that cause damage to the intestinal microbiota, since little research is directed to the identi cation and characterization of conjugative systems, as well as the low frequency of the MobC family in relation to other components of the MOB family (36).
Although this family contributes to a greater mobilization of plasmids, many proteins of this family are not recorded (37).
S. exneri 201 CR also expresses lipoprotein OsmY related to carrier biogenesis related to bacterial virulence.
The protein OsmY (Osmotically inducible protein Y) is a lipoprotein found in Gram-negative and contributes to adaptation to the gastrointestinal osmotic stress environment, acting as a membrane stabilizer (38,39). Another recently discovered role is inhibition of protein aggregation, although studies have been proven only in vitro to date (40).
Although this protein is conserved and identi ed and predicted as virulent by our study, the literature lacks information related to bacterial virulence, especially for the genus of Shigella sp.
However, there is evidence that OsmY is involved in the biogenesis of auto carriers, such as type AIDA -I (adhesins involved in diffuse adherence), at E.coli. The AIDA -I protein is involved in bacterial virulence, as in bio lm formation, and in other virulence phenotypes as adhesion (41,42). The literature states that AIDA-I, already cited above, is a counterpart to the Flu protein (A0A0H2V1S1), found exclusively in the CR condition of the strain S. exneri 201. Flu protein, also called ag43, is an external membrane protein belonging to the autotransport family and is mainly identi ed in uropathogenic strains of E.coli and S. exneri. Being AIDA-I homologous to Flu protein, its role in bio lm formation and secretion and translocation processes is reinforced, through in vitro assays (43). Just as OsmY is involved in the biogenesis of AIDA-I, it is also in ag43, since de cient mutants of OsmY tend to be more sensitive to proteolytic action, for example (42).
Variants of ag43 present in uropathogenic strains of E.coli indicate that it contributes to the persistence of the species in the urinary tract in patients with urinary infection. However, the mechanisms that contribute to such persistence, as well as its direct involvement in the virulence of other pathogens, such as Shigella spp. are not yet fully clari ed (44).
1. exneri 201 CR expresses the LptA protein and this is associated with resistance to antibiotics.
Another potential target for therapeutic action against Gram-negatives is the Lpt translocation system, more speci cally one of the subunits that make up this system, which is the Lipopolysaccharide export system protein (LptA; A0A090NL57), also known as YhbN.
This system works on the transport of Lipopolysaccharide (LPS) to the external membrane, which besides composing the bacterial cell structure, contributes to resistance to attacks by external agents, such as antibiotics (45).
Thus, biosynthesis and transport of lipopolysaccharides have been targeted for new drugs, as in E.coli.
Compounds that interact speci cally as LptA can damage the LPS transport system, consequently affecting growth and altering bacterial morphology. It is suggested that this system may also be harmed by other Gram-negative bacteria, such as Shigella spp. (45).
LptA was identi ed to interact with another set of proteins also classi ed as pathogenic, such as YdgH (A0A090NGX6), the latter identi ed in the strains S. exneri M90T LB and S. exneri 201 CR. However, there is a scarcity of data on this protein, although its domain is preserved among the enterobacteria (46).

Conclusions
The identi cation of proteins in bacteria of clinical importance has been essential, especially in recent years, for the understanding of various molecular mechanisms, especially in the pathogenesis of diseases.The protein pro le identi ed in this study provides unpublished data about the strains studied, thus adding new ndings that contribute to previous research done by our group.
This lack of data is related to the fact that studies of this size occur with known standards. Studies that carry out the genetic and proteomic characterization of clinical isolates are rare, but they are of enormous importance since they help us to elucidate the pathogenicity of a microorganism in practice. The Shigella exneri 201 CR of this study showed the secretion of several proteins related to pathogenicity, quorum sensing, adhesion, resistance to antibiotics and evasion of the immune system. Although many of these proteins are characterized for other enterobacteria, little is known about their role in the pathogenicity of Shigella and some of the proteins highlighted in this work are still little studied in the literature, mainly in processes related to virulence, ie Bacteria are microorganisms fast and dynamic evolution. We believe that these ndings with clinical strains can help in regional vaccine study, since there is a strong relationship between phenotype and genotype and expression of the same.
Some of the proteins highlighted in this work are still little studied in the literature, mainly in processes related to virulence, ie, there is a lack of information, especially on the pathogenesis of Shigella spp., Which limits the deepening of discussions on how such proteins are promoting increased virulence mechanisms, for example. In this sense, we believe that in silico prediction of pathogenic proteins presented here, opens ways for new investigations, not only for Shigella species, but for other enterobacteria, thus contributing to the elucidation of virulence of this genus, which is still responsible for the death of thousands of people. Supporting institutions did not participate in the writing of this paper, nor did it read or approve it.

Availability of data and materials
The datasets used and analyzed in the study are available from the corresponding author on reasonable request.

Consent for publication
Not applicable.

Competing interests
The author(s) declare that they have no competing interests.    Table (Additional le 1: Table S1).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.