Development of Two Animal Models To Study the Function ofVibrio parahaemolyticus Type III Secretion Systems�
Pablo Pineyro, Xiaohui Zhou, Lisa H. Orfe, Patrick J. Friel, Kevin Lahmers, and Douglas R. Call*Department of Veterinary Microbiology and Pathology, Washington State University, Pullman, Washington 99164-7040
Received 4 May 2010/Returned for modification 8 June 2010/Accepted 27 August 2010
Vibrio parahaemolyticus is an emerging food- and waterborne pathogen that encodes two type III secretionsystems (T3SSs). Previous studies have linked type III secretion system 1 (T3SS1) to cytotoxicity and T3SS2to intestinal fluid accumulation, but animal challenge models needed to study these phenomena are limited. Inthis study we evaluated the roles of the T3SSs during infection using two novel animal models: a model in whichpiglets were inoculated orogastrically and a model in which mice were inoculated in their lungs (intrapulmo-narily). The bacterial strains employed in this study had equivalent growth rates and beta-hemolytic activitybased on in vitro assays. Inoculation of 48-h-old conventional piglets with 1011 CFU of the wild-type strain(NY-4) or T3SS1 deletion mutant strains resulted in acute, self-limiting diarrhea, whereas inoculation with aT3SS2 deletion mutant strain failed to produce any clinical symptoms. Intrapulmonary inoculation of C57BL/6mice with the wild-type strain and T3SS2 deletion mutant strains (5 � 105 CFU) induced mortality or amoribund state within 12 h (80 to 100% mortality), whereas inoculation with a T3SS1 deletion mutant or aT3SS1 T3SS2 double deletion mutant produced no mortality. Bacteria were recovered from multiple organsregardless of the strain used in the mouse model, indicating that the mice were capable of clearing the lunginfection in the absence of a functional T3SS1. Because all strains had a similar beta-hemolysin phenotype, wesurmise that thermostable direct hemolysin (TDH) plays a limited role in these models. The two modelsintroduced herein produce robust results and provide a means to determine how different T3SS1 and T3SS2effector proteins contribute to pathogenesis of V. parahaemolyticus infection.
Vibrio parahaemolyticus is a Gram-negative food- and water-borne pathogen that is recognized worldwide as a causativeagent of gastroenteritis associated with the consumption ofundercooked seafood (22). Gastrointestinal infection is char-acterized by acute self-limiting diarrhea, abdominal cramps,nausea, and vomiting. In approximately 5% of cases, V. para-haemolyticus gastrointestinal infection can progress to septice-mia and may be fatal for immunocompromised patients,including those with leukemia, liver disease, and patients in-fected by human immunodeficiency virus (14, 28).
The thermostable direct hemolysin (TDH) is a widely rec-ognized virulence factor of V. parahaemolyticus. This hemoly-sin is associated with the Kanagawa phenomenon, which is abeta-hemolysis reaction on a defined blood agar (10). TDH isencoded by 1 or 2 genes (tdhA and tdhS), and the proteinincreases permeability in human erythrocytes (19), increaseschloride secretion in an intestinal cell line (29), and is alsothought to be responsible for enterotoxigenicity in a rabbitsmall intestine model (5, 23). Early studies demonstrated thatdeletion of tdhS and tdhA will significantly reduce fluid accu-mulation in a rabbit ileal-loop model, but the phenotype is notcompletely abrogated, suggesting that other virulence factorscould be involved in pathogenicity (20, 26).
In addition to TDH, V. parahaemolyticus encodes two dis-tinct type III secretion systems (T3SSs) (21). T3SSs are used byGram-negative pathogens to secrete and translocate effector
proteins into the cytosol of eukaryotic cells (11, 16). In vitrostudies have shown that the type III secretion system 1 (T3SS1)(encoded on chromosome I) is required to induce cytotoxicityin several cell lines (13, 25, 33). This secretion system hasbeen associated with several phenotypic changes, includingactin rearrangement, autophagy, and oncosis (7, 33). More-over, T3SS2 can induce cytotoxicity in Caco-2 cells and alsoplays an important role in fluid secretion based on in vivomodels (18, 27).
Several animal challenge models have been used to study thepathogenesis of V. parahaemolyticus, including orogastric andperitoneal mouse infections (13, 15), rabbit ileal-loop ligations(4, 5), and oral infections in suckling rabbits and mice (8, 31).Some of this work demonstrated that V. parahaemolyticus cancause mouse mortality via oral or intraperitoneal injection (13,15). These models have provided clues about the pathophysi-ology of V. parahaemolyticus, but in most cases these modelshave not been used as comparative systems to study the distinctcontributions of T3SS1 and T3SS2 to pathogenesis. In the caseof other pathogens, newborn piglets have been used to studypathogenesis of Campylobacter jejuni and Escherichia coli in-fections; piglets offer an added advantage of having a higherdegree of similarity with human physiology (2, 3, 30). In thecase of Pseudomonas aeruginosa, an intrapulmonary mousemodel has been developed to assess the role of the T3SSduring lung infection (1).
The aim of the present study was to develop robust animalmodels that will allow investigators to determine the contribu-tion of V. parahaemolyticus effector proteins in the pathophys-iology of gastroenteritis and/or lethality. The presence of gas-trointestinal clinical signs was evaluated for infected piglets,and we found that deletion of T3SS2 attenuated acute diar-
* Corresponding author. Mailing address: Department of Veteri-nary Microbiology and Pathology, Washington State University, 402Bustad Hall, Pullman, WA 99164-7040. Phone: (509) 335-6313. Fax:(509) 335-8529. E-mail: [email protected].
rhea. A murine pulmonary model showed significant attenua-tion in mortality with deletion of a functional T3SS1. Findingsfrom these two robust models support the assertion that thetwo V. parahaemolyticus T3SSs contribute to distinct patho-genic mechanisms during infection.
MATERIALS AND METHODS
Bacterial strains. Wild-type V. parahaemolyticus NY-4 strain (serotype O3:K6)and �vcrD (T3SS1 knockout [KO]), �escV (T3SS2 KO), and �vcrD and �escV(T3SS1 and T3SS2 [T3SS1/2] KO) deletion mutants were generated previously(Table 1) (33). Bacterial strains were grown 6 to 8 h in Luria-Bertani (LB)medium supplemented with 2.5% NaCl (LB-S) and shaking (250 rpm) at 37°C.Strains were not induced to express T3SSs before use in animals. �exsD and�exsA deletion mutants (Table 1) were included in our analysis of hemolyticactivity. The �exsD deletion mutant expresses the T3SS1 constitutively, and the�exsA deletion mutant serves as an alternative knockout of the T3SS1. A tdhAdeletion mutant (�tdhA) was generated using previously described methods (26),except we employed a modified set of oligonucleotide primers (Table 1).
Comparison of growth rates. The growth rates of wild-type strain NY-4 (se-rotype O3:K6) and T3SS1 KO, T3SS2 KO, and T3SS1/2 KO mutant strains werecompared using a Bioscreen C plate reader (Oy Growth Curves AB Ltd., Hel-sinki, Finland). Briefly, 20 �l of overnight culture was inoculated into 180 �l offresh LB-S medium, and the optical density at 600 nm (OD600) was monitoredover 24 h. Growth curves were repeated three times with 10 independent wellsper replicate. After 24 h, the bacterial count (CFU) was determined by thedrop-plate method (9).
Hemolytic assay. The hemolytic activity of the supernatants was evaluated byfirst preparing overnight culture from isolated colonies grown from freezerstocks. The colonies were isolated twice on LB-S agar plates after which onecolony for each strain was selected from the second plate and dispersed sepa-rately in 10-ml LB-S broth tubes. Broth cultures were grown for approximately24 h at 37°C with shaking (250 rpm). A small aliquot (200 �l) was retrieved toquantify the OD and CFU per milliliter (final counts ranged between 2 � 109 and4.5 � 109 per ml). The remaining culture was pelleted by centrifugation (4,000 �g for 10 min) at room temperature, and the supernatant was transferred to 15-mlconcentrators (Amicon Ultra-15 with a PLBC Ultracel-PL membrane with amolecular size cutoff of 3 kDa). The supernatant was concentrated to �10% oforiginal volume (�1 ml). Human red blood cells or erythrocytes (HRBCs) (2%suspension in Alsever solution from a single donor; Innovative Research, Novi,MI) were pelleted, washed four times, and resuspended in an equal volume ofphosphate-buffered saline (PBS) (pH 7.4) before they were added to supernatantsamples (1:1). The mixtures were incubated at 37°C for 4 h after which sampleswere briefly centrifuged (600 � g for 3 min), and the hemolytic activity from eachsample was evaluated by measuring the optical absorbance from the supernatant(541 nm) with a Multiskan MCC/340 spectrophotometer (Thermo LabSystems,Helsinki, Finland). For a positive control, PBS (150 �l) was mixed with 150 �l ofHRBCs and incubated for 4 h at 37°C. After incubation, 150 �l of a 10% solution(in PBS) of Triton X-100 (Triton N-101; Sigma Chemical Co., St. Louis, MO)was added. Three biological replicates (including three technical replicates each)were used in the analysis, and hemolytic activity of each strain was then normal-ized against the log concentration from the original culture (CFU/ml) or usingthe total protein from the supernatant (quantified using a MicroBCA proteinassay; Thermo Fisher Scientific, Rockford, IL).
Piglet model. The piglets were taken from the dam when they were approxi-mately 6 h old and were raised until they were 48 h old. The piglets were fedthree times a day with infant milk formula (Enfamil Lipil milk-based infantformula with iron) using individual feeders and with increasing volume from 80ml to 120 ml until they were 5 days old. Both males and females were used inequal frequency, and the average body weight at infection was 1.4 kg (standarddeviation [SD], 0.15 kg).
(i) Dose determination. To determine the dose, 20-ml samples of 6-h-oldcultures of bacteria with an OD600 of 2.5 to 2.7 were centrifuged at 4,000 � g for10 min at room temperature and were resuspended in 10 ml of sterile PBS (pH7.4). Serial 10-fold dilutions were generated corresponding to 106 to 1012 CFU/ml. The inoculum was mixed with 10 ml of infant milk formula and placed inclean and disinfected feeders, allowing each pig to drink the complete dose. Thefeeders were refilled several times with a small amount of infant formula (5 ml)to ensure that the bulk of the inoculum was ingested.
(ii) Experimental design. After dose determination trials, we performed threeindependent experiments consisting of two piglets per bacterial strain (treat-ment), for a total of 6 piglets per treatment. Treatments consisted of a dose of1011 CFU of either the wild-type strain NY-4 (serotype O3:K6) or T3SS1 KO orT3SS2 KO mutant strain or PBS (pH 7.4) as a vehicle control.
(iii) Clinical observations. The body temperature of each piglet was evaluatedrectally, and individual rectal swabs were taken and placed immediately in 10 mlsterile PBS prior to infection. All piglets were observed every 2 h postinoculation,and clinical signs were evaluated (diarrhea, vomiting, feed consumption, andalertness). In addition, body temperature and rectal swabs were taken to assesspyrexia and bacterial shedding, respectively. The occurrence of diarrhea wasevaluated by a subjective score as follows: presence of soft to liquid feces on swab(�), presence of soft to liquid feces on the swab and stained perineum (��), andpresence of soft to liquid feces on the swab, stained perineum, and presence ofa large amount of loose fecal discharge in the cage (���). Vomiting wasrecorded as present or absent. At 8 h and 24 h after inoculation, one piglet fromeach group was euthanized with an intravenous overdose of 2 ml of pentobar-bital. A full necropsy was performed immediately after death, and gross changeswere recorded. Samples for bacteriology (lung, liver, spleen, kidney, stomach,duodenum, jejunum, ileum, cecum, colon, and rectum) were immediately placedin 10 ml sterile PBS. Samples for histological analysis were placed in 10%formaldehyde for fixation. All tissues were processed and stained with hematox-ylin and eosin for histological evaluation.
(iv) Bacteriology. Fecal samples obtained immediately before inoculation,follow-up rectal swabs, and necropsy tissues were cultured on thiosulfate-citrate-bile salts-sucrose agar (TCBS; Difco Laboratories, Detroit, MI) containing 100�g/ml ampicillin (NY-4 and derivative strains are resistant to ampicillin). Theplates were incubated at 37°C overnight and were then examined for the pres-ence of smooth translucent green-blue colonies. Isolation of V. parahaemolyticusfrom rectal swabs was recorded as present or absent. The tissue samples wereprocessed with a stomacher 80 lab blender (Seward Medical London, UnitedKingdom), and when possible, the number of CFU per gram of tissue wasdetermined.
Intrapulmonary mouse model. Mice were housed in microisolator cages at theAnimal Science Experimental Animal Laboratory Building (Pullman, WA) andallowed ad libitum access to food and water. The age and body weight of infectedmice ranged from 4 to 5 months old and 30 to 40 g, respectively, and males andfemales were used at similar frequencies. Mice were injected intraperitoneallywith a combination of 60 to 70 mg of ketamine per kg of body weight and 12 to14 mg of xylazine per kg of body weight (24). Upon deep sedation, each mouse
TABLE 1. Bacterial strains used in this study
Strain Original straindesignation Source In-frame gene
deletion Characteristic or description Reference
NY-4 (wild type) VP208 Clinical isolate Serotype O3:K6 Yeung and Boor (31a)�vcrD mutant NY4-v1 NY-4 vcrD (vp1662) T3SS1 KO mutant Zhou et al. (33)�escV mutant NY4-e1 NY-4 escV (vp1354) T3SS2 KO mutant Zhou et al. (33)�vcrD �escV mutant NY4-ev1 NY-4 vcrD and escV T3SS1/2 KO mutant Zhou et al. (33)�exsD mutant NY-4 exsD (vp1698) T3SS1 negative regulator Zhou et al. (34)�exsA mutant NY-4 exsA (vp1699) T3SS1 positive regulator Zhou et al. (34)�tdhA mutanta NY-4 tdhA (vp1314) Thermostable direct hemolysin A This study
a This deletion mutant was constructed by the method of Ono et al. (25) but with the following primers: A1 (5�-ACGAATCGGAGCCAAAATTCT-3�), A2(5�-TACATTAACAAAATACAATATCTCATCAGA-3�), A3 (5�-TCTGATGAGATATTGTATTTTGTTAATGTA-3�), and A4 (5�-TGTAACCATCAGAAGCGAATA-3�).
was placed on a board with its mouth kept open by holding the upper and lowerincisors with plastic forceps. The tongue was then extended with forceps to oneside, and a microsyringe was used to deposit 50 �l of inoculum onto the back ofthe tongue and proximal to the larynx. The inoculum was inhaled into the lungsby normal breathing. The mouse was then held in an upright position for 10 minand allowed to recover (32). The presence of clinical signs in mice was observedevery hour during the first 12 h and every 3 h thereafter. Those animals thatsurvived up to 24 h postinoculation were euthanized in a CO2 chamber followedby cervical dislocation; mice judged to be moribund were euthanized immedi-ately. A full necropsy was performed immediately upon death, and gross changeswere recorded. Samples for bacteriology (lung, liver, spleen, and gastrointestinaltract) were immediately placed in sterile PBS. The histological samples wereobtained from the same tissues as described above with lung and several sectionsof the small intestine insufflated and placed in 10% formaldehyde until they wereprocessed.
(i) Dose determination. To establish a suitable dose for the wild-type strain(NY-4), 10-fold serial dilutions (104 to 1010) of bacteria were inoculated into 28adult C57BL/6 mice. The number of CFU per milliliter was calculated using theOD600 measurement of overnight cultures and then verified through bacterialenumeration using a drop-plate technique (9).
(ii) Mortality studies. The same methods for preparing inoculum were usedfor comparative studies. To achieve a mortality rate between 80 and 100%, weused a dose of 5 � 105 CFU (see Results). Four groups of 10 adult mice(C57BL/6) were infected with wild-type strain NY-4 or T3SS1 KO, T3SS2 KO, orT3SS1/2 KO mutant strain. Two replicate studies were completed. Vehicle con-trols were used in one trial (n � 5 mice) to confirm that our procedures did notconfound the experimental outcome.
(iii) Bacteriology. All tissue samples were first weighed and then placed eitherin 2 ml PBS (spleen and lung) or in 10 ml PBS (liver and gastrointestinal tract).All samples were plated on TCBS and incubated at 37°C overnight. Whenpossible, the number of bacterial CFU per gram of tissue was estimated.
The experimental protocols described herein were approved by the Washing-ton State University Institutional Animal Use and Care Committee (ASAF 3841and 3638).
Statistical analysis. Analysis of variance (ANOVA) was used to compareculture growth and hemolytic activity, and a P value of �0.05 was consideredstatistically significant. For the hemolytic assay results, we employed a Dunnett’supper one-sided multiple-comparison test with either the PBS-only or �tdhAresults used as the “control” for comparison purposes. The clinical score wasanalyzed using repeated-measures ANOVA, and pair-wise comparisons wereperformed using a Tukey-Kramer test. These analyses were conducted usingNCSS 2004 (Number Cruncher Statistical System, Kaysville, UT). Analysis of themouse survival ratio was performed with the Kaplan-Meier and log-rank testusing Graph Pad Prism 5.01 (GraphPad Software Inc., San Diego, CA).
RESULTS
Growth rates and hemolytic activities are equivalent for thestrains employed in challenge experiments. The growth ratesand hemolytic activities of the wild-type and mutant strainsused in this study were evaluated to determine whether mutantstrains with deletion mutations were compromised for thesetraits. There was no difference in the shape of the growthcurves or the final OD600 at 24 h (P 0.05) (Fig. 1A). Allstrains except the �tdhA strain retained hemolytic activityagainst HRBCs relative to negative control (P 0.05) (Fig.1B). We included �exsA and �exsD strains in this analysis,which serve as positive and negative regulators of the T3SS1expression, respectively (34, 35). Hemolytic activity was notdifferent for either strain (Fig. 1B). The inclusion of �tdhAdeletion mutant confirmed that hemolysis in this assay wasresponsive to the absence of TDH (P � 0.0001) (Fig. 1B), andthis result was similar when raw optical density data were usedor when these data were normalized by CFU/ml (data notshown). We have shown in a separate experiment that normal-izing by total protein concentration does not change the con-clusion that only the �tdhA strain is attenuated for hemolyticactivity (data not shown). From these data, we concluded that
the results from our challenge experiments are not likely to beconfounded by inadvertent differences in growth rates or bydifferences in TDH activity.
T3SS2 is required for gastrointestinal disease in 2-day-oldpiglets. Two-day-old piglets were inoculated with 106 to 1012
CFU of wild-type strain NY-4 (n � 2 animals per dose), andthe 50% infective dose (ID50) from this experiment was esti-mated to be 1011 CFU when the primary response variable wasthe presence or absence of diarrhea. We used 2-day-old pigletsin these experiments, because they were readily available, butother ages may work as well, although ID50 experiments mayneed to be repeated. Upon infection with the wild-type NY-4strain, piglets developed vomiting and yellow watery diarrheawithin 2 h postinoculation. The clinical scores among the dif-ferent infected groups and controls during the time course ofthe experiment were clearly different. Wild-type- and T3SS1KO-inoculated piglets produced obvious clinical symptoms,and knocking out T3SS2 abrogated this effect on the piglets
FIG. 1. (A) The growth rates of V. parahaemolyticus strains (wild-type strain NY-4 [serotype O3:K6] and T3SS1 knockout KO, T3SS2KO, and T3SS1/2 KO mutant strains) were compared over 24 h inbroth culture. The value at each hour is an average of three biologicalreplicates (variance not shown). The final OD600 was not different forthe different strains (P 0.05). Ctrl, control. (B) The hemolytic activityof the supernatants was evaluated by measuring the optical absorbanceat 541 nm 4 h after mixing culture supernatant with human red bloodcells. The values (bars) are the means plus standard errors of themeans (error bars) for three biological replicates. For simplicity, thevalues shown here were normalized by dividing the values by the valuefor the lysis positive control. The �tdhA strain and the negative control(Neg Ctrl) had significantly lower hemolytic activity (P � 0.05) thanthe other strains (indicated by an asterisk). The outcome of the anal-ysis did not change when the values were normalized by CFU/ml(except the negative control was not included in the latter analysis).
VOL. 78, 2010 V. PARAHAEMOLYTICUS CHALLENGE MODELS 4553
(Fig. 2A). Vomiting, ranging from mild to severe, was observedfor all piglets inoculated with wild-type strain NY-4 and in 50%of the piglets inoculated with the T3SS1 KO strain, but vom-iting was not observed in the piglets inoculated with the T3SS2KO strain or PBS. None of the inoculated piglets displayedsigns of dehydration or anorexia. We observed no significantdifferences in body temperature among the three infectedgroups and controls during the time course of the infection.Fecal shedding of V. parahaemolyticus was observed only in thepiglets infected with NY-4 and T3SS1 KO strains and onlyduring the first 12 h postinoculation (Fig. 2B).
We observed no gross lesions related with V. parahaemolyti-cus infection in any of the piglets evaluated. At 8 h postinfec-tion, piglets treated with wild-type strain NY-4 or T3SS1 KOstrain had a mildly distended jejunum and ileum with a mod-erate amount of yellow liquid digesta and gas. In addition, thececum and spiral colon were filled with moderate to abundantamounts of soft pasty to liquid green digest. Histological ex-amination showed that the cecal submucosa and serosa weremoderately expanded by edema (Fig. 3). No other histologicalchanges were observed in all the tissues evaluated. Neither
gross lesions nor histological changes were observed in pigletsinoculated with the T3SS2 KO strain or PBS.
Bacteria were consistently recovered at 8 h and 24 h postin-fection from the gastrointestinal tracts of piglets infected withwild-type strain NY-4 or T3SS1 KO strain. Only one pigletinfected with T3SS2 KO strain had recoverable bacteria fromthe gastrointestinal samples at 8 h postinfection (Table 2). NoV. parahaemolyticus was recovered in control animals or fromthe other tissue samples, indicating that cross-contaminationwas unlikely to be a confounding factor in these experiments.Collectively, these data indicate that T3SS2 is required tocause gastrointestinal disease (acute diarrhea and vomiting) inneonatal pigs, and deletion of the functional T3SS2 results inan absence of clinical symptoms characteristic of V. parahae-molyticus infection.
T3SS1 is necessary for mortality of infected mice. We esti-mated the ID50 for the intrapulmonary mouse model to be1.4 � 105 CFU. Consequently, we adopted a dose of 5 � 105
CFU for these experiments to ensure a mortality rate between80% and 100% after infection with the wild-type strain. Weused 30- to 40-g mice in these experiments, because they were
FIG. 2. (A) The presence or absence of diarrhea was evaluated every 2 h during the first 12 h postinoculation and every 3 h thereafter. Theaverage diarrhea severity score (see Materials and Methods) was estimated for six piglets in each treatment group. Piglets infected with wild-typestrain NY-4 or T3SS1 KO mutant strain had significantly higher clinical scores (P � 0.003) than piglets infected with T3SS2 KO strain or pigletsin the PBS control group at 2 and 4 h postinoculation (p.i) (indicated by three asterisks). N.S, not significant. (B) Rectal swabs were cultured every2 h during the first 12 h postinoculation and every 3 h thereafter. Bars show the percentages of piglets shedding V. parahaemolyticus. Numericvalues below bars indicate hours p.i.
readily available, but other sizes may work, although the 50%lethal dose (LD50) may need to be recalculated and a smallerinoculum volume (�50 �l) may be needed for smaller mice. Todetermine whether the presence of T3SS1 or T3SS2 is neces-
sary for mortality in the intrapulmonary model, the mortalityrates of mice infected with wild-type strain NY-4 or T3SS1 KO,T3SS2 KO, or T3SS1/2 KO mutant strain were compared (Fig.4). High mortality was observed for mice infected with NY-4 orT3SS2 KO strain (40 to 80% mortality by 8 h postinoculationand 80 to 100% by 24 h). No mortality was observed in miceduring the 24-h period after challenge with T3SS1 KO orT3SS1/2 KO strain. All mice infected with wild-type and T3SS2KO strains showed diffuse and severe pulmonary hemorrhage.No gross lesions were observed in mice infected with theT3SS1 KO and T3SS1/2 KO strains. Histological evaluation ofthe lung in all mice showed severe and diffuse capillary con-gestion with perivascular edema and large areas of alveolarhemorrhage. The alveolar lumen was severely distended bylarge aggregates of neutrophils intermixed with some foamymacrophages, abundant edema, and cellular debris. The inter-stitium was randomly infiltrated by small aggregates of neutro-phils (Fig. 5). To better assess how V. parahaemolyticus strainswere impacting mice at death, we repeated the challenge trialtwice and collected bacterial counts when mice died or wereeuthanized between 8 and 12 h postinoculation (p.i.). Impor-tantly, during the latter experiment when one mouse died orwas moribund and euthanized, one mouse from each of theremaining treatments was simultaneously euthanized (regard-less of condition) to obtain time-matched bacteriology data;the samples were processed immediately to limit changes inmicrobial population density after death. No differences wereobserved in the numbers of bacteria recovered from the lung,gastrointestinal tract, or liver regardless of the strain used (Fig.6). Mice inoculated with NY-4 and T3SS2 KO strains becamebacteremic (70% and 30%, respectively) with similar bloodCFU levels (0.7 0.6 and 0.4 0.1 log10 CFU SD, respec-tively). Spleen samples from bacteremic animals were alsopositive in both animals inoculated with wild-type strain NY-4(45%) and animals inoculated with T3SS2 KO mutant strain(100%). Mice inoculated with T3SS1 KO strain remained cul-ture positive for V. parahaemolyticus in the lungs, gastrointes-tinal tract, and liver after 24, but only one animal had a culture-positive spleen at 24 h. These data indicate that T3SS1 isnecessary to produce mortality following pulmonary challengeand that T3SS1 is strongly associated with systemic V. parah-aemolyticus infection.
FIG. 3. (A) Piglets infected with wild-type strain NY-4 and T3SS1KO mutant strain showed a mild to severe colonic submucosal andserosal edema (arrows). (B) No histological lesions were observed inthe piglets infected with the T3SS2 KO strain. The sections werestained with hematoxylin and eosin. Magnification, �100.
TABLE 2. V. parahaemolyticus recovered from each section of the gastrointestinal tract at 8 h and 24 h postinoculationa
Section of gastrointestinaltract
No. of V. parahaemolyticus recovered from section of the gastrointestinal tract from piglets infected with the following strain atthe indicated timeb:
NY-4 (wild type) T3SS1 KO mutant T3SS2 KO mutant
8 h p.i. (3/3)c 24 h p.i. (3/3) 8 h p.i. (3/3) 24 h p.i. (3/3) 8 h p.i. (1/3) 24 h p.i. (0/3)
Stomach TNTC NR 4.3 (0.5) NR NR NRDuodenum TNTC NR 4.8 (0.5) NR NR NRJejunum TNTC NR 5.0 (0.7) NR 3.4 NRIleum 4.0 (0.8) TNTC 5.1 (0.9) NR 5.6 NRCecum 3.5 (1.0) 3.4 (0.5) 5.3 (1.0) TNTC 4.2 NRSpiral colon 4.6 (1.0) 4.9 (1.2) 4.3 (0.8) TNTC 5.0 NRRectum 4.2 (1.0) 3.4 (0.7) 5.1 (0.9) TNTC NR NR
a No V. parahaemolyticus was recovered from PBS controls.b The number of V. parahaemolyticus recovered is shown as the average log10 CFU/g of tissue (standard deviation). TNTC, too numerous to count; NR, none
recovered.c The number of piglets with bacteria recovered to the number of piglets tested is shown in parentheses after the time postinoculation.
VOL. 78, 2010 V. PARAHAEMOLYTICUS CHALLENGE MODELS 4555
In this study we developed two different animal models thatare suitable to assess the differential contributions of T3SS1(mouse) and T3SS2 (piglet) to V. parahaemolyticus virulence.We found no differences for in vitro growth rates between thestrains used in this study, so this is unlikely to be a confoundingvariable in the analysis. Similarly, we found no loss of beta-hemolytic activity between these strains, indicating that dele-tion mutations were not affected for this important phenotypethat is usually associated with thermostable direct hemolysin(TDH) (12). TDH-dependent lysis in human red blood cells(HRBCs) is caused by formation of pores in the cell membraneand activation of cation channels (19). While T3SSs are in-
FIG. 5. Histopathological examination of the mouse lung speci-mens. (A) At 8 h postinoculation for all infected groups (wild-typestrain NY-4 and T3SS2 KO, T3SS1 KO, and T3SS1/2 KO mutantstrains), the mice presented diffuse capillary congestion with perivas-cular edema and large areas of alveolar hemorrhage. The alveolarlumen is severely distended by large aggregates of neutrophils inter-mixed with a few foamy macrophages and cellular debris. (B) At 12 hpostinfection (NY-4, T3SS1 KO, and T3SS1/2 KO strains), there wassevere neutrophilic infiltration. The bronchi were completely pluggedby abundant suppurative exudates and infiltrated by large peribron-chiolar cuff of neutrophils and macrophages. (C) Section in controlanimals (PBS only) are characterized by diffuse capillary hyperemia.The sections were stained with hematoxylin and eosin. Magnification,�200.
FIG. 4. Four groups of C57BL/6 adult mice (10 mice in each group)were inoculated with different strains of V. parahaemolyticus. The sur-vival rate was evaluated for two independent replicates shown in pan-els A and B by Kaplan-Meier and log-rank tests. Values that aresignificantly different (P � 0.0001) are indicated by three asterisks.Values that are not significantly different (N.S) are indicated.
volved in pore formation in the host cell membrane and trans-location of effector proteins directly into the target cells (11),there is no information regarding the contribution of thesesystems to hemolytic activity against HRBCs. Our results dem-onstrate that the soluble products from V. parahaemolyticusculture are sufficient to induce HRBC lysis, which is consistentwith TDH activity, and that the HRBC lysis phenotype is
curtailed significantly when one of the two tdh genes is deleted(tdhA). Loss of function from either T3SS1 or T3SS2, however,made no impact on the in vitro HRBC hemolysis phenotype,and thus, differences between these strains in the animal chal-lenges are not attributed to differences in TDH activity of thestrains.
Natural infection with V. parahaemolyticus causes acute gas-trointestinal clinical signs, such as diarrhea, vomiting, and ab-dominal cramping (28). An experimental rabbit ileal-loopmodel showed that V. parahaemolyticus infection results influid accumulation and that this phenotype is associated withTDH production (5, 23). Nevertheless, it has also been shownthat mutant strains lacking tdhS and tdhA genes are still able tocause intestinal fluid accumulation in the same model (26).There is also evidence indicating that T3SS2 is an importantfactor in fluid secretion for ileal-loop models (17, 18, 27). Inthe present study, the clinical signs of infected piglets resem-bled the gastrointestinal presentation of the human disease.Challenge with the wild-type strain produced acute diarrheawith mild to moderate gross or histological lesions withoutsigns of bacteremia, and these signs and lesions required afunctional T3SS2. Although T3SS2 has been previously asso-ciated with cell cytotoxicity in vitro and enterotoxicity in vivo(18), we found no signs of inflammation in the gastrointestinaltract and there were no obvious morphological changes in theintestinal epithelium.
Our study was originally limited to deciphering the contri-bution of T3SS1 and T3SS2 to acute clinical symptoms, and itis apparent that T3SS2 is involved. Nevertheless, becauseT3SS2 is encoded by genes carried on an 80-kbp pathogenicityisland and the genes encoding T3SS2 are flanked by tdhA andtdhS in the pandemic serovar (17), it is possible that the mu-tation of a T3SS2 structural gene could have polar effects onthe synthesis of TDH. This possibility is enhanced on the basisof the fact that we have been unable to complement either ofour T3SS1 or T3SS2 deletion mutant strains, probably owing topolar effects within the islands carrying the genes encoding theT3SSs. Consequently, it is important to emphasize that wedetect a significant reduction in hemolytic activity only whenthe tdhA gene itself is deleted. The presence of a functional ordysfunctional T3SS2 did not reduce the apparent expressionand functional activity of TDH in vitro, and therefore, TDH isnot a confounding variable in the analysis (Fig. 1B).
V. parahaemolyticus also causes wound infections and septi-cemia (from both the wound infections and, to a lesser extent,gastrointestinal infections) with immunocompromised patientsbeing most at risk for possible mortality (14). The pathogenicmechanisms involved in these clinical presentations are notclear. Previous studies have demonstrated that intraperitonealinfection in mice induces mortality that is not associated withthe presence of TDH, implying that another pathogenic mech-anism is involved (15). Recently, studies have demonstratedthat T3SS1 is implicated in cellular oncosis and autophagy invitro, and this may be a major contributing factor to mortalityin mice (13, 28, 33). Our results demonstrate that T3SS1 isnecessary to cause mortality in mice inoculated intrapulmonar-ily, and we surmise that alternative delivery strategies (e.g.,intranasal) would produce similar results. The mortality rateand time course of the clinical signs agree with those obtainedby intraperitoneal infection in other studies (13, 15). While
FIG. 6. Average CFU recovered from lung (A), gastrointestinaltract (B), and liver (C) for mice from which bacteria were recovered.Mice infected with wild-type strain NY-4 and T3SS2 KO mutant straindied or were euthanized between 8 and 12 h postinoculation. The miceinoculated with the T3SS1 KO mutant strain were euthanized at thesame time to have matched bacterial counts over time. The number ofanimals with recoverable bacteria/number of inoculated mice (n/n°) isshown on the x axes.
VOL. 78, 2010 V. PARAHAEMOLYTICUS CHALLENGE MODELS 4557
T3SS1 has been described as causing cytotoxicity in various celllines (25, 33), no histological difference were observed relativeto either severity or distribution of pulmonary inflammatoryresponse among the bacterial strains used in this study.
While the group inoculated with the T3SS1 KO mutantstrain showed the same inflammatory pattern as those inocu-lated with the wild-type strain and T3SS2 KO mutant strainand the bacterial recovery rate was similar between 8 and 12 hof infection, the high mortality rate requires a functionalT3SS1. Moreover, in those mice inoculated with the T3SS1 KOstrain, the bacterial recovery rate after 24 h decreases dramat-ically (mice inoculated with the wild-type strain or T3SS2 KOmutant strain rarely survived 12 h). Consequently, whenT3SS1 is dysfunctional, it is very likely that the host will clearthe bacteria from the lung.
Interestingly, even when mice were infected with the T3SS1KO mutant strain, in some cases bacteria could be recoveredfrom multiple organs and blood. Given that bacteremia occurswith the T3SS1 KO strain, bacteria were recovered in similarnumbers from lungs and there was no evidence of growth orTDH defects in this strain, it is likely that T3SS1-dependentmortality observed in this study is due to mechanisms that havebeen characterized in vitro, including autophagy (7) and onco-sis (33). It is also possible that an unrecognized T3SS1-depen-dent exotoxigenic factor is involved in lethality. Experimentsthat include a vopQ knockout (6) would test the contribution ofautophagy to the model outcome. Experiments that include avp1659 gene knockout will allow a test for exotoxigenic factors,because deletion of this gene produces a strain that will activelysecrete T3SS1 proteins but that is unable to efficiently trans-locate these proteins into host cells (36). At present, the ge-netic trait required to induce oncosis in vitro has not beenidentified.
It is worth noting that we originally intended to adopt apublished Vibrio cholerae infection model (24) to study mouseinfection following orogastric inoculation. We used a canula todeliver inoculum to the stomachs of anesthetized mice, andthis produced lung infections in a number of mice. The lunginfections suggested the possibility that we were contaminatingthe larynx region when we withdrew the canula. We then useda double canula to deliver inoculum and did not observe lunginfection (or mortality). On the basis of these data, we alteredour strategy to use the intrapulmonary model, which producesvery consistent results. These results raise the possibility thatpreviously published oral inoculation models (e.g., in rabbitand mouse) that produced septicemia may have been inadver-tent pulmonary models.
In this study we have demonstrated that T3SS1 is necessaryto cause mortality in mice using an intrapulmonary model andthat T3SS2 is necessary to cause gastrointestinal clinical signsin newborn piglets. The oral inoculation of the piglets is rep-resentative of the natural route of infection in humans, withthe clinical signs observed in the piglet model being very sim-ilar to those observed in human infection. In the mouse model,the pathogen has to surmount one epithelial boundary prior tocausing disease, and mortality requires a functional T3SS1. It ispossible that during human infection T3SS2 contributes toepithelial layer damage in immunocompromised individualsand T3SS1 contributes to subsequent sepsis and septic shock;otherwise, it does not appear that these two traits act syner-
gistically during conventional food-borne vibriosis. Impor-tantly, both of the animal models described herein provideunconfounded and unique opportunities for understanding theroles of the two distinct T3SSs and their effector proteins in thepathogenesis of V. parahaemolyticus-associated disease. Theyalso afford opportunities to consider the host immune responseto V. parahaemolyticus infection.
ACKNOWLEDGMENTS
We thank Amelia Lanier for technical assistance in construction ofthe V. parahaemolyticus tdhA KO strain. Smriti Shringi provided assis-tance with piglet inoculations and handling. Michael Konkel andThomas Besser provided technical review at different stages of thiswork. Dan Erwin and Seth Nydam provided helpful discussions andassistance.
This project has been funded in part with federal funds from theNational Institute of Allergy and Infectious Diseases, National Insti-tutes of Health, Department of Health and Human Services, undercontract number NO1-AI-30055 and by the Agricultural AnimalHealth Program, College of Veterinary Medicine, Washington StateUniversity.
REFERENCES
1. Ader, F., R. Le Berre, K. Faure, P. Gosset, O. Epaulard, B. Toussaint, B.Polack, E. Nowak, N. B. Viget, E. Kipnis, and B. P. Guery. 2005. Alveolarresponse to Pseudomonas aeruginosa: role of the type III secretion system.Infect. Immun. 73:4263–4271.
2. Akopyants, N. S., K. A. Eaton, and D. E. Berg. 1995. Adaptive mutation andcocolonization during Helicobacter pylori infection of gnotobiotic piglets.Infect. Immun. 63:116–121.
3. Babakhani, F. K., G. A. Bradley, and L. A. Joens. 1993. Newborn pigletmodel for campylobacteriosis. Infect. Immun. 61:3466–3475.
4. Boutin, B. K., S. F. Townsend, P. V. Scarpino, and R. M. Twedt. 1979.Demonstration of invasiveness of Vibrio parahaemolyticus in adult rabbits byimmunofluorescence. Appl. Environ. Microbiol. 37:647–653.
5. Brown, D. F., P. L. Spaulding, and R. M. Twedt. 1977. Enteropathogenicityof Vibrio parahaemolyticus in the ligated rabbit ileum. Appl. Environ. Mi-crobiol. 33:10–14.
6. Burdette, D. L., J. Seemann, and K. Orth. 2009. Vibrio VopQ inducesPI3-kinase-independent autophagy and antagonizes phagocytosis. Mol. Mi-crobiol. 73:639–649.
7. Burdette, D. L., M. L. Yarbrough, A. Orvedahl, C. J. Gilpin, and K. Orth.2008. Vibrio parahaemolyticus orchestrates a multifaceted host cell infectionby induction of autophagy, cell rounding, and then cell lysis. Proc. Natl.Acad. Sci. U. S. A. 105:12497–12502.
8. Calia, F. M., and D. E. Johnson. 1975. Bacteremia in suckling rabbits afteroral challenge with Vibrio parahaemolyticus. Infect. Immun. 11:1222–1225.
9. Chen, C. Y., G. W. Nace, and P. L. Irwin. 2003. A 6 � 6 drop plate methodfor simultaneous colony counting and MPN enumeration of Campylobacterjejuni, Listeria monocytogenes, and Escherichia coli. J. Microbiol. Methods55:475–479.
10. Chun, D., J. K. Chung, R. Tak, and S. Y. Seol. 1975. Nature of the Kanagawaphenomenon of Vibrio parahaemolyticus. Infect. Immun. 12:81–87.
11. Cornelis, G. R. 2006. The type III secretion injectisome. Nat. Rev. Microbiol.4:811–825.
12. Fujino, T., T. Miwatani, Y. Takeda, and A. Tomaru. 1969. A thermolabiledirect hemolysin of Vibrio parahaemolyticus. Biken J. 12:145–148.
13. Hiyoshi, H., T. Kodama, T. Iida, and T. Honda. 2010. Contribution of Vibrioparahaemolyticus virulence factors to cytotoxicity, enterotoxicity, and lethal-ity in mice. Infect. Immun. 78:1772–1780.
14. Hlady, W. G., and K. C. Klontz. 1996. The epidemiology of Vibrio infectionsin Florida, 1981–1993. J. Infect. Dis. 173:1176–1183.
15. Hoashi, K., K. Ogata, H. Taniguchi, H. Yamashita, K. Tsuji, Y. Mizuguchi,and N. Ohtomo. 1990. Pathogenesis of Vibrio parahaemolyticus: intraperito-neal and orogastric challenge experiments in mice. Microbiol. Immunol.34:355–366.
16. Hueck, C. J. 1998. Type III protein secretion systems in bacterial pathogensof animals and plants. Microbiol. Mol. Biol. Rev. 62:379–433.
17. Kodama, T., K. Gotoh, H. Hiyoshi, M. Morita, K. Izutsu, Y. Akeda, K.-S.Park, V. V. Cantarelli, R. Dryselius, T. Iida, and T. Honda. 2010. Tworegulators of Vibrio parahaemolyticus play important roles in enterotoxicityby controlling the expression of genes in the Vp-PAI region. PLoS One5:e8678.
18. Kodama, T., H. Hiyoshi, K. Gotoh, Y. Akeda, S. Matsuda, K. S. Park, V. V.Cantarelli, T. Iida, and T. Honda. 2008. Identification of two transloconproteins of Vibrio parahaemolyticus type III secretion system 2. Infect. Im-mun. 76:4282–4289.
19. Lang, P. A., S. Kaiser, S. Myssina, C. Birka, C. Weinstock, H. Northoff, T.Wieder, F. Lang, and S. M. Huber. 2004. Effect of Vibrio parahaemolyticushaemolysin on human erythrocytes. Cell. Microbiol. 6:391–400.
20. Lynch, T., S. Livingstone, E. Buenaventura, E. Lutter, J. Fedwick, A. G.Buret, D. Graham, and R. DeVinney. 2005. Vibrio parahaemolyticus disrup-tion of epithelial cell tight junctions occurs independently of toxin produc-tion. Infect. Immun. 73:1275–1283.
21. Makino, K., K. Oshima, K. Kurokawa, K. Yokoyama, T. Uda, K. Tagomori,Y. Iijima, M. Najima, M. Nakano, A. Yamashita, Y. Kubota, S. Kimura, T.Yasunaga, T. Honda, H. Shinagawa, M. Hattori, and T. Iida. 2003. Genomesequence of Vibrio parahaemolyticus: a pathogenic mechanism distinct fromthat of V cholerae. Lancet 361:743–749.
22. Nair, G. B., T. Ramamurthy, S. K. Bhattacharya, B. Dutta, Y. Takeda, andD. A. Sack. 2007. Global dissemination of Vibrio parahaemolyticus serotypeO3:K6 and its serovariants. Clin. Microbiol. Rev. 20:39–48.
23. Nishibuchi, M., A. Fasano, R. G. Russell, and J. B. Kaper. 1992. Entero-toxigenicity of Vibrio parahaemolyticus with and without genes encodingthermostable direct hemolysin. Infect. Immun. 60:3539–3545.
24. Olivier, V., J. Queen, and K. J. Satchell. 2009. Successful small intestinecolonization of adult mice by Vibrio cholerae requires ketamine anesthesiaand accessory toxins. PLoS One 4:e7352.
25. Ono, T., K. S. Park, M. Ueta, T. Iida, and T. Honda. 2006. Identification ofproteins secreted via Vibrio parahaemolyticus type III secretion system 1.Infect. Immun. 74:1032–1042.
26. Park, K. S., T. Ono, M. Rokuda, M. H. Jang, T. Iida, and T. Honda. 2004.Cytotoxicity and enterotoxicity of the thermostable direct hemolysin-dele-tion mutants of Vibrio parahaemolyticus. Microbiol. Immunol. 48:313–318.
27. Park, K. S., T. Ono, M. Rokuda, M. H. Jang, K. Okada, T. Iida, and T.Honda. 2004. Functional characterization of two type III secretion systems ofVibrio parahaemolyticus. Infect. Immun. 72:6659–6665.
28. Qadri, F., M. S. Alam, M. Nishibuchi, T. Rahman, N. H. Alam, J. Chisti, S.Kondo, J. Sugiyama, N. A. Bhuiyan, M. M. Mathan, D. A. Sack, and G. B.
Nair. 2003. Adaptive and inflammatory immune responses in patients in-fected with strains of Vibrio parahaemolyticus. J. Infect. Dis. 187:1085–1096.
29. Takahashi, A., N. Kenjyo, K. Imura, Y. Myonsun, and T. Honda. 2000. Cl�
secretion in colonic epithelial cells induced by the Vibrio parahaemolyticushemolytic toxin related to thermostable direct hemolysin. Infect. Immun.68:5435–5438.
30. Tzipori, S., F. Gunzer, M. S. Donnenberg, L. de Montigny, J. B. Kaper, andA. Donohue-Rolfe. 1995. The role of the eaeA gene in diarrhea and neuro-logical complications in a gnotobiotic piglet model of enterohemorrhagicEscherichia coli infection. Infect. Immun. 63:3621–3627.
31. Vongxay, K., S. Wang, X. Zhang, B. Wu, H. Hu, Z. Pan, S. Chen, and W.Fang. 2008. Pathogenetic characterization of Vibrio parahaemolyticus isolatesfrom clinical and seafood sources. Int. J. Food Microbiol. 126:71–75.
31a.Yeung, P. S., and K. J. Boor. 2004. Epidemiology, pathogenesis, and pre-vention of foodborne Vibrio parahaemolyticus infections. Foodborne Pathog.Dis. 1:74–88.
32. Yu, C. K., C. M. Shieh, and H. Y. Lei. 1999. Repeated intratracheal inocu-lation of house dust mite (Dermatophagoides farinae) induces pulmonaryeosinophilic inflammation and IgE antibody production in mice. J. AllergyClin. Immunol. 104:228–236.
33. Zhou, X., M. E. Konkel, and D. R. Call. 2009. Type III secretion system 1 ofVibrio parahaemolyticus induces oncosis in both epithelial and monocytic celllines. Microbiology 155:837–851.
34. Zhou, X., D. H. Shah, M. E. Konkel, and D. R. Call. 2008. Type III secretionsystem 1 genes in Vibrio parahaemolyticus are positively regulated by ExsAand negatively regulated by ExsD. Mol. Microbiol. 69:747–764.
35. Zhou, X., M. E. Konkel, and D. R. Call. 2010. Regulation of type IIIsecretion system 1 gene expression in Vibrio parahaemolyticus is dependenton interactions between ExsA, ExsC, and ExsD. Virulence 1:260–272.
36. Zhou, X., M. E. Konkel, and D. R. Call. 2010. Vp1659 is a Vibrio parahae-molyticus type III secretion system 1 protein that contributes to translocationof effector proteins needed to induce cytolysis, autophagy and disruption ofactin structure in HeLa cells. J. Bacteriol. 192:3491–3502.
Editor: S. M. Payne
VOL. 78, 2010 V. PARAHAEMOLYTICUS CHALLENGE MODELS 4559
![Shuangjian Guo1, Xiaohui Zhang2, Shengxiang Wang3 arXiv ...arXiv:2002.06017v1 [math.RA] 13 Feb 2020 OnsplitregularHom-Leibniz-Rinehartalgebras Shuangjian Guo1, Xiaohui Zhang2, Shengxiang](https://static.documents.pub/doc/80x56/5ffa640993124b7945333343/shuangjian-guo1-xiaohui-zhang2-shengxiang-wang3-arxiv-arxiv200206017v1-mathra.jpg)
