Identification of In Vivo -Induced Conserved Sequences from Yersinia pestis During Experimental Plague Infection in the Rabbit

Abstract

In an effort to identify the novel virulence determinants of Yersinia pestis, we applied the gene “discovery” methodology, in vivo-induced (IVI) antigen technology, to detect genes upregulated during infection in a laboratory rabbit model for bubonic plague. After screening over 70,000 Escherichia coli clones of Y. pestis DNA expression libraries, products from 25 loci were identified as being seroreactive to reductively adsorbed, pooled immune serum. Upon sequence analysis of the predicted IVI gene products, more frequently encountered conserved protein functional categories have emerged, to include type-V autotransporters and components of more complex secretion systems including types III and VI. The recombinant products from eight selected clones were subsequently immunoblotted against pooled immune serum from two naturally infected host species: the prairie dog, and a species refractory to lethal disease, the coyote. Immune prairie dog serum recognized 2–3 of the rabbit-reactive antigens, suggesting at least some overlap in the pathogen's in vivo survival mechanisms between these two hosts. Although the coyote serum failed to recognize most of the IVI antigens, LepA was universally reactive with all three host sera. Collectively, the profiles/patterns of IVI conserved sequences (IVICS) may represent immune “signatures” among different host species, possessing the potential for use as a diagnostic tool for plague. Further, the antigenic nature of IVICS makes them ideal for further evaluation as novel subunit vaccine candidates. The gathering of additional data and analysis of the intact IVI genes and the expressed IVICS products should provide insight into the unique biologic processes of Y. pestis during infection and reveal the genetic patterns of the pathogen's survival strategy in different hosts.

Introduction

M ost Yersinia pestis virulence effectors identified and characterized to date are encoded on one of three virulence plasmids. The 100-kb species, pFra, is most notably associated with expression of the F1 capsule. This highly antigenic protein has been associated with induction of protective immunity in numerous hosts and has been used as the basis for detection and diagnosis of plague for decades (Laboratory Manual of Plague Diagnostic Tests, CDC, 2000). Curiously, the lack of F1 expression fails to render the pathogen avirulent, and thus some strains deficient in capsule production remain virulent and are able to overcome F1-based immunity (Heath et al. 1998). Plague virulence determinants encoded by the medium-molecular-weight “low-calcium response” plasmid, pLcr, also have been characterized for their antigenicity in an infected host (Benner et al. 1999). The classic pLcr-encoded protein, LcrV, or simply the V antigen, has been found to be requisite for virulence (unlike F1), is highly immunogenic, has been employed as the major component of the next-generation, subunit plague vaccine for humans (Heath et al. 1998, Powell et al. 2005), and is currently in phase II human clinical trials. The low-molecular-weight plasmid, pPst, encodes plasminogen activator, necessary for spread of the microorganism from peripheral sites of infection (Welkos et al. 1997). Aside from this antigen, pPst also encodes the colicin-like protein, pesticin, whose role in virulence is largely unknown.

The chromosomally localized 100-kb pgm locus, associated with the “pigmentation” phenotype of Y. pestis, has been well characterized and linked to iron acquisition and storage in the pathogen (Buchrieser et al. 1999). Mutations in a number of genes within this region render Y. pestis attenuated by peripheral route of infection, yet pgm⁻ strains are still considered acceptable for use as a surrogate model for bubonic plague in laboratory animals when administered by intravenous route (Matson et al. 2005). Further, a laboratory-generated pgm⁻ derivative of wild-type strain CO92 was demonstrated to be virulent by respiratory route in nonhuman primates (Welkos et al. 2002). With the exception of pgm, chromosomally encoded loci associated with virulence have been understudied in this pathogen.

Genes and their products associated with Y. pestis infection that have been identified and characterized are typically isolated from laboratory-grown bacteria, and thus host factors that upregulate some virulence loci may or may not be present. One approach that overcomes this problem uses immune sera processed to specifically remove the antibodies to constitutively expressed antigenic bacterial proteins, while retaining those antibodies to bacterial antigens upregulated during infection. This technique, known as in vivo-induced antigen technology (IVIAT) (Handfield et al. 2000, Rollins et al. 2005) has been successfully employed to identify in vivo upregulated genes in many different bacterial pathogenic species (Deb et al. 2002, Hang et al. 2003, Jitsurong et al. 2003, Chang et al. 2005, Salim et al. 2005, Harris et al. 2006, Rollins et al. 2008). In this report, we describe the successful application of IVIAT on Y. pestis, which has led to the identification of genes upregulated in vivo and thus likely to be relevant to the survival of Y. pestis during infection in the mammalian host.

Materials and Methods

DNA expression library construction

Y. pestis CO92 and its pLCR⁻ pFra⁺ isogenic derivative were grown at 37°C for 48 h in trypticase soy broth from single colonies obtained from blood agar working stock cultures. Genomic DNA was extracted using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI), and partial Sau3A restriction enzyme digests were generated. Size-fractionated DNA fragments were then ligated into the pET-30abc system (EMD Biosciences, La Jolla, CA). These series of plasmids drive gene expression via the T7/lac promoter and facilitate allele cloning in each of the three reading frames. The expression library was subsequently transformed into Escherichia coli BL-21 cells (Stratagene, La Jolla, CA) for library plasmid amplification, and the DNA was extracted, pooled, and retransformed into BL-21[DE3] cells for isopropyl β-D-1-thiogalactopyranoside (IPTG)-induced expression.

Immune serum processing

Two sources of immune rabbit serum were used for screening of the DNA expression libraries for IVI gene products. The first was a commercially made lyophil, representing a pool of immune serum from over 20 rabbits infected by intravenous route with Y. pestis strain A1122 (pgm⁻ ). This serum was originally intended as a human immunotherapeutic for plague infection (lot no. 7-1036-13H, 1952; Lederle Laboratories, New York, NY). The second serum pool was generated in-house from a group of five New Zealand White rabbits subcutaneously infected with a sublethal dose of wild-type Y. pestis CO92. All serum pools were filtered through a 0.2-mm low protein-binding syringe filter before mixing with in vitro grown (37°C) whole cells of wild-type Y. pestis CO92 for treatment of in-house serum, and pgm⁻ CO92 for treatment of the Lederle serum. Additionally, both serum pools were adsorbed with virulent F1⁻ Y. pestis C12 and pgm⁻ C12, respectively, to preclude the possibility that F1 capsule would block adsorption of antibody to certain constitutively expressed outer membrane proteins. Adsorptions were also performed with E. coli BL-21 whole cells and lysates to remove nonspecific antibodies, as previously described (Deb et al. 2002, Hang et al. 2003, Jitsurong et al. 2003, Chang et al. 2005, Salim et al. 2005, Harris et al. 2006). The adsorption process was repeated 3–4 times with the in vitro grown pathogen, and reactivity was monitored by immunoblot to ensure removal of antibody to constitutively expressed antigens.

Immunoscreening

After 5 h of incubation at 37°C, E. coli BL-21[DE3] colonies containing Y. pestis DNA were lifted from the plates using Protran nitrocellulose membrane disks (Whatman, Kent, UK), inverted and placed on fresh LB agar supplemented with antibiotic and 1 mM IPTG. After incubation at 30°C overnight, the membranes were exposed to chloroform vapor for 15 min to lyse the cells. Once dry, the membranes were treated with 5% skim milk + 0.05% Tween-20 for 1 h, rinsed for 5 min with phosphate-buffered saline (PBS)–Tween, and incubated with absorbed sera at a 1:1000 dilution in skim milk for 2 h. The immune-labeled membranes were then washed 3–5 times with PBS and incubated with a 1:5000 dilution of alkaline phosphatase-conjugated Protein G (Rockland, Gilbertsville, PA) in skim milk for 1 h. Following secondary labeling, the membrane was washed three times for 5 min with PBS and incubated for 5 min with stabilized substrate (Promega). After color development, the reaction was stopped by rinsing the membranes with deionized water for 5 min. Reactivity was assessed against E. coli possessing empty vector, as well as duplicate membranes probed with secondary antibody alone.

Gene/antigen identification

Following rinsing and drying steps, the membranes were matched to the master plate and appropriate colonies selected and isolated. Suspect reactive colonies were grown in broth culture and serial diluted onto new plates. Screening with the processed immune serum was repeated again and plasmid DNA was extracted from confirmed positive clones using the QIAprep Spin Miniprep kit (Qiagen, Valencia, CA). Sequences of the Y. pestis insert DNA were determined using vector sequence-specific primers (University of Wyoming Nucleic Acid Exploration Facility). Sequences were then compared against the NCBI Entrez database's Y. pestis CO92 genome for DNA/protein alignments. The most significant matches were designated as IVI genes. Subcellular localization of putative IVI products was predicted using the PSORTb v 2.0 algorithm (Gardy et al. 2005).

Serologic analysis in other hosts

Serologic cross-reactivity of selected “rabbit-reactive” IVI clones was examined using pooled immune serum samples from two additional hosts: the black-tailed prairie dog (Cynomys ludovicianus) and the coyote (Canis latrans). Pooled serum from the former represented 11 and 9 survivors of colonies from two distinct geographic locations (Colorado and Texas), confirmed seropositive for Y. pestis infection. Serum from the latter was obtained from the Wyoming Game and Fish archives and represented a pool from eight plague-seropositive animals.

Results

After pooling and extensively processing the immune rabbit serum, we screened over 70,000 transformants of Y. pestis DNA from E. coli expression libraries. This effort has led to the identification of 25 IVI genes. Despite the potential presence of plasmid DNA in the libraries, all of the IVI loci detected were mapped to the Y. pestis chromosome. All clones positive with the pooled, adsorbed Lederle serum were conversely reactive when cross-tested with adsorbed serum from rabbits infected with wild-type CO92, suggesting that the in vivo expression profiles of the antigens identified in this study are the same between wild-type and pgm ⁻ Y. pestis. Sequencing of the insert DNA revealed that all of the predicted gene products possessed one or a number of highly conserved domains (cd) which fell within either the (Clusters of Orthologous Groups [COG] of proteins), or cd NCBI databases (Tables 1 –4). Further analysis using the pSORTb algorithm for protein localization allowed us to group the IVI gene products according to predicted cellular localization. As shown in Table 1, at least four of the IVI proteins identified (16%) are predicted to be secreted to the outer membrane and/or extracellularly. Three of the four IVI loci encode putative type-V antigens, containing a conserved “passenger” domain which specifies an autotransporting function (Junker et al. 2006). The fourth IVI gene encodes IutA, an iron siderophore receptor, which does not appear to be linked to the pgm locus (Buchrieser et al. 1999). Orthologs of this protein exist in at least 16 additional species within the Enterobacteriaceae, as well as 23 species across genera within the class Gammaproteobacteria.

Table 1.

Yersinia pestis In Vivo-Induced Gene Products Predicted to be Extracellular

Clone ID ^a	Homology ^b	COG/cdd Id ^c	Category ^d	Function ^e
3b-4/26 YPO1672 (YapK)	Hypothetical exported protein	3468	U	Type-V passenger domain; autosecreted protein
3R919 YPO1004 (YapH)	Hypothetical exported protein	3468	U	Type-V secretion“passenger” domain
RAYPA17R YPO0587 (YapG)	Hypothetical exported protein	3468	H	Type-V secretion “passenger” domain
RAYPA11 YPO0994 (IutA)	Ferric siderophore receptor	73259 (01347)	P	Ligand-gated channel; TonB-dependent N-terminal “plug”

Yersinia pestis expression libraries were constructed in Escherichia coli and screened using pooled, reductively adsorbed immune rabbit serum. Insert DNA was isolated, sequenced, and aligned with sequences in the NCBI Entrez Protein database. In vivo-induced (IVI) gene products were grouped according to predicted cellular localization using the PSORTb, version 2.0.4 algorithm.

Y. pestis open reading frame (orf ) identifications.

Clone homologies to the NCBI Conserved Domain (cd) and Cluster of Orthologous Groups (of proteins) Databases.

Cluster of Orthologous Groups (COG [of proteins]) or Conserved Domain Database (cdd) identifier number.

COG functional categories: H, coenzyme transport and metabolism; P, inorganic ion transport; U, secretion.

Predicted gene function of IVI orf(s).

Table 2.

Yersinia pestis In Vivo-Induced Gene Products Predicted to Localize in the Periplasm

Clone ID ^a	Homology ^b	COG/cdd ID ^c	Category ^d	Function ^e
RAYPA2 YPO2758	Resistance protein	1794	H	Aspartate racemase
RAYPA6 YPO1846	Cystine-binding protein	0834	E	Amino acid ABC-transporter (FliY)
RAYPA8 YPO1317	Ion transporter	1464	P	Inorganic ion ABC-transporter (NlpA)
RAYPA13 YPO2774	Histidine-binding protein	0834	E	Amino acid ABC-transporter (HisJ)

Methods were followed as described in Table 1.

Y. pestis open reading frame (orf ) identifications.

Clone homologies to the NCBI Conserved Domain (cd) and Cluster of Orthologous Groups (of proteins) Databases.

Clusters of Orthologous Groups (COG) or cdd identifier number.

COG functional categories: E, amino acid transport and metabolism; H, coenzyme transport and metabolism; P, inorganic ion transport.

Predicted gene function of IVI orf(s).

Table 3.

Yersinia pestis In Vivo-Induced Gene Products Predicted to be Intracellular (Cytoplasmic/Inner Membrane)

Clone ID ^a	Homology ^b	COG/Cdd ID ^c	Category ^d	Function ^e
10aR919 YPO2068/69	Integral membrane protein	2076	P	Transporter (pump); similar to YdgF; inner membrane
RAYPA14-1 YPO3608	Hypothetical protein	3209	U	Rhs-family protein; similar to E. coli CorE; inner membrane
RAYPA17f YPO0038	Stringent factor	0317	H	Guanosine pyrophophohydrolase (SpoT); cytoplasmic
RAYPA10 YPO2502	Zinc-binding protein	1062/63	P	Zinc-dependent dehydrogenase (GutB) cytoplasmic
RAYPA5 YPO1106	ATP-NAD kinase	0061	H	NADP biosynthesis (PpnK)
RAYPB2 YPO1316	Hypothetical protein	86262	R	Iron/ascorbate oxidoreductase; cytoplasmic
2aR919 YPO3212	Recombination-associated protein	32793	L	Similar to E. coli RdgC; cytoplasmic
2a-dot YPO3569	UDP-n-acetylglucosamine-1-carboxyvinaltransferase	30128	M	Similar to Enterobacter MurA; cytoplasmic
ypo4008 (gene)	Two-component sensor kinase	—	PG^f	“Sensor” protein; inner membrane
10br919 ypo1835 (gene)	Putative chemotaxis protein	—	PG^g	Inner membrane protein

Methods were followed as described in Table 1.

Y. pestis open reading frame (orf ) identifications.

Clone homologies to the NCBI Conserved Domain (cd) and Cluster of Orthologous Groups (of proteins) Databases.

Clusters of Orthologous Groups (COG) or cdd identifier number.

COG functional categories: H, coenzyme transport and metabolism; L, replication, recombination, and repair; M, cell wall biogenesis; P, inorganic ion transport; R, oxygenase, general function prediction only; U, secretion.

Predicted gene function of IVI orf(s).

Pseudogene IS100 at 266 of 1154 residues.

Pseudogene FS at 631 of 992 residues.

Table 4.

Yersinia pestis In Vivo-Induced Gene Products of Unknown Localization

Clone ID ^a	Homology ^b	COG/cdd ^c	Category ^d	Function ^e
RAYPA14-2 YPO0879	Hypothetical protein	4679	S	Similar to phage-related protein (GP49)
RAYPA18 YPO2888	Hypothetical protein	83897	U	Similar to Salmonella SseB (TTSS filament protein)
RAYPA3 YPO2330	Putative heat-shock protein	3187	O	Similar to E. coli Hs1J
1a-lrg YPO0262	Type-III secretion component	83070	U	Similar to Salmonella SsaH/I (TTSS component)
RAYPA1 YPO1559	Hypothetical protein	0451	S	Similar to E. coli hemolysin, HlyF
3c-4/26 YPO3705	Hypothetical protein	3518	S (U)	Predicted type-VI secretion component
2r721 YPO2716	GTP-binding elongation factor	133290	O	LepA

Methods were followed as described in Table 1.

Y. pestis open reading frame (orf ) identifications.

Clone homologies to the NCBI Conserved Domain (cd) and Cluster of Orthologous Groups (of proteins) Databases.

Clusters of Orthologous Groups (COG) or cdd identifier number.

COG functional categories: O, protein modification, turnover, and/or trafficking; S, function unknown; U, secretion.

Predicted gene function of IVI orf(s).

GTP, guanosine triphosphate; TTSS, type-III secretion system.

The second pSORTb category represents an additional 16% of the reactive clones in the IVI gene set and contains loci that encode proteins secreted beyond the inner membrane into the periplasm (Table 2). Three of the four IVI genes in this category were found to encode periplasmic components of small-molecule ATP-binding cassette (ABC) transport systems, which possess orthologs across three classes of Proteobacteria and four species across three additional phyla. The other IVI antigen in this category was an aspartate racemase. In addition to the presence of sequence encoding this protein in 20 different species within the Enterobacteriaceae, we have also identified orthologs in the Alpha- and Betaproteobacteria, as well as cross-phyla.

The third pSORTb category is listed in Table 3 and represents 40% of the IVI antigen set, which consists of proteins predicted to localize in the inner membrane or cytoplasm of the pathogen. Most of these antigens are metabolic enzymes, involved in processes ranging from murein biosynthesis (YPO3569) to stringent response (YPO0038). A recombination-associated protein, YPO3212, is highly conserved among species in yersiniae, in addition to the presence of orthologs in 13 other species among the Gammaproteobacteria. The most taxonomically broad IVI sequence (YPO1316) in this group encodes an iron/ascorbate-dependent oxidoreductase, with orthologs present cross-phyla, in addition to the presence of over two dozen eukaryotic homologs. Interestingly, two truncated antigenic gene products were identified from coding sequences interrupted by either an IS element (ypo4008), or a reading frame shift mutation (ypo1835), both involving sensor/transducer regulatory mechanisms within the cell. Despite the open reading frame interruptions, the intact N-terminal sequences are predicted to encode 266 and 631 amino acid residue products, respectively. Unlike other IVI proteins in this category, orthologs of these products appear less conserved across taxonomic families. Similarly, YPO2068 and YPO3608, both inner membrane proteins, are primarily conserved within the Enterobacteriaceae, with the latter showing significant divergence outside of the yersiniae. Interestingly, YPO2068 bears similarity to the ubiquitous inner membrane small-molecule transport protein, YdgF, present among gram-negative and gram-positive species alike.

The final pSORTb group represents proteins of unknown localization (Table 4) and comprises 28% of the IVI antigen set. Conserved functional categories include secretion, protein modification/trafficking, and products of unknown function. One such product of this last category is a phage-related protein, YPO0879, with similarity to a structural component (GP-49) of the Pseudomonas bacteriophage, LUZ24. This protein shows greater 30–40% sequence identity across several taxonomic classes of bacteria. Another IVI antigen of unknown function, YPO1559, was identified as possessing similarity to the E. coli hemolysin, HlyF. Most interestingly, YPO2716 was identified as the guanosine triphosphate (GTP)-dependent elongation factor, LepA (Evans et al. 2008). LepA orthologs were identified in all the yersiniae, as well as many genera of the Enterobacteriaceae and two other taxonomic orders.

Three IVI proteins representing components of more complex secretions also fell into this fourth category. Two type-III secretion accessory proteins, YPO0262 and YPO2888, were identified, the latter of which possesses similarity to the type-III secretion system (TTSS) filament component, SseB, of Salmonella (Ruiz-Albert et al. 2003). Orthologs of both these proteins were identified across species within the yersiniae, but were divergent across taxonomic classes. A component of the more recently described type-VI secretion system was also found, YPO3705. Orthologs of this protein appear restricted to the Enterobacteriaceae.

Finally, a heat-shock lipoprotein, YPO2330, was identified. Orthologs were found to be present among members of the yersiniae, with homologous sequences spanning cross-phyla. Similarity to the E. coli and Vibrio lipoprotein HslJ was also noted.

To assess if any of the IVI antigens identified through rabbit antiplague screening were expressed and immunogenic in other hosts naturally infected, we cross-tested eight selected IVI clones by colony immunoblot, which were sampled from all four pSORTb categories, with pooled immune serum from two prairie dog colonies and coyotes with confirmed exposure to Y. pestis. As shown in Table 5, at least two of the autosecreted antigens were reactive with two pooled serum samples from immune prairie dogs, and another IVI protein, LepA, was reactive to antibody in this serum. Similarly, antibody to LepA was detected in pooled immune coyote serum, whereas no other antigen from the remaining seven in the panel was reactive with immune serum from this host.

Table 5.

Comparison of Plague In Vivo-Induced Conserved Sequence Immune Signatures in Two Different Naturally Infected Hosts

	Sera ^a
In vivo-induced product reactive with rabbit serum	Prairie Dog (CO)	Prairie Dog (TX)	Coyote
YPO1672: Type-V autosecreted, YapK	+	+	−
YPO3705: Type-VI secretion component	−	+	−
YPO1004: Type-V autosecreted, YapH	+	+	−
YPO2758: Aspartate racemase	−	−	−
YPO3212: Recombination-associated protein (RdgC)	−	−	−
YPO2068: Inner membrane protein	−	−	−
YPO2716: GTP-binding elongation factor (LepA)	+	+	+
YPO1835: Chemotaxis protein (truncate)	−	−	−

Expression clones of eight IVI sequences across all four PSORTb categories were patch purified and retested for seroreactivity with the pooled, adsorbed immune rabbit serum by colony immunoblot. Pooled, adsorbed prairie dog and coyote sera were subsequently tested against the clones by a similar method, and scored as either positive or negative for reactivity, compared with immune rabbit serum and E. coli containing “empty” vector, respectively.

CO, Colorado; TX, Texas.

Discussion

Through the use of the gene discovery methodology, IVIAT, we have identified over two dozen gene products upregulated during plague infection in rabbits. These loci and their products have been subsequently designated as “IVI conserved sequences” (IVICS), because the taxonomic distribution of these genes in general has been found to be broad, with most possessing large domains falling into either the COG (of proteins) or the cd database. Although findings among gene products from more recently evolved homologs, such as Yersinia LcrV and Pseudomonas PcrV, suggest sequence divergence with conserved functionality, the very nature of selection pressure on IVICS dictates highly conserved function that is consistent with the ancient conserved integrity of the protein's primary/secondary structure. With respect to pathogenicity, IVICS may represent common antigenic proteins with similar fundamental in vivo functions among numerous bacterial species. Studies by others who have employed IVIAT on different pathogens are consistent with our finding that the majority of IVI products fall within COG and/or cd database common functional categories (Deb et al. 2002, Hang et al. 2003, Salim et al. 2005). Interestingly, clones of more frequently encountered functional categories have emerged from our analyses. These categories consist of metabolic enzymes, ATP-utilizing small-molecule transport components, particularly autotransported proteins (type-V secretion), and accessory proteins associated with more complex secretion systems. In fact, our recent application of IVIAT on Brucella abortus has also revealed IVICS within these same functional groups (Andrews, unpublished data).

We have identified three type-V autosecreting gene products in Y. pestis, upregulated in vivo. The type-V proteins represent extracellular molecules that are highly immunogenic and may thus make excellent candidates for induction of protective immunity or can be utilized as robust serologic markers for Y. pestis infection. Indeed, at least five of the type-V proteins have been demonstrated in vitro to be implicated in potential virulence functions (Yen et al. 2007, Felek 2008). More recently, YapE has been shown to be required for full virulence of Y. pestis and appears to promote cell type-specific adherence (Lawrenz 2009). Our findings that at least three of these proteins are antigenic and upregulated in vivo corroborate the importance of this category of IVICS to the survival of Y. pestis during infection.

The finding that two IVI antigens were components of a chromosomal TTSS was particularly remarkable. Previous studies have reported chromosomal TTSSs in all three pathogenic yersiniae (Haller et al. 2000, Parkill et al. 2001, Foultier et al. 2002, Pujol and Bliska 2003) and have likely evolved from systems in plant pathogens (Foultier et al. 2002). Similarly, we found the sequence of YPO2888 and a cd in YPO0262 homologous to such loci. The role in virulence of the chromosomal TTSS in the yersiniae remains unclear. Curiously, genes of the Yersinia enterocolitica system were found to be upregulated at room temperature (Foultier et al. 2002), and mutants were only slightly attenuated when examined in a laboratory mouse model (Haller et al. 2000). Our finding that two chromosomal TTSS components are expressed during infection suggests that the system in Y. pestis may be more relevant to its survival in the mammalian host.

Another IVI antigen associated with protein translocation was identified as a putative component of a type-VI secretion system, YPO3705. This system was first identified in the extracellular pathogen, Vibrio cholera (Pukatzki et al. 2005), and demonstrated to be upregulated in vivo in Burkholderia pseudomallei (Shalom et al. 2007).

Although the majority of plague IVI genes identified in this study specify products secreted beyond the inner membrane, one pSORTb category is represented by loci predicted to encode intracellular gene products. The antigenic nature of these proteins strongly suggests that a subpopulation of bacteria are killed through host lytic mechanisms and subsequently processed by the adaptive immune system. The relevance of a humoral immune response against these products to the clearance of Y. pestis infection is not intuitively obvious, but we have not ruled out the possibility that one or a number of these intracellular proteins may have immune-modulatory properties as observed in other host–pathogen systems (Montes et al. 2002, Chamond et al. 2005, Spera et al. 2006).

In addition to facilitating a better understanding of the pathogenesis of plague in different hosts, collectively, IVICS may represent an approach to the diagnosis of exposure/infection. In this regard, we found at least three IVI antigens commonly reactive in a naturally infected host sensitive to disease, the prairie dog. The pattern of reactivity, however, was not identical to that of our model system, which suggests differences in how the pathogen is responding to in vivo cues within the different host, and/or how the host's immune system is responding to infection. Interestingly, the rabbit has been described as a host with “intermediate” sensitivity to infection (Adamovicz, unpublished data). Further, a species refractory to disease but not infection, the coyote, failed to show reactivity to any of the selected rabbit IVI antigens, with the exception of LepA (YPO2716), which was reactive across all three host species. In this regard, LepA may represent a universal marker for Y. pestis infection, regardless of host sensitivity to disease. The role of LepA in Yersinia pathogenesis was supported by an earlier study in which the gene was identified as being transcriptionally upregulated in vivo in Y. enterocolita by a promoter trap methodology (in vivo expression technology; IVET) on infected mice (Gort and Miller 2000). Further, an ortholog of the protein in Legionella has been characterized and found to be required for full virulence (Chen et al. 2007).

Finally, although the efficacy of the new recombinant F1-V subunit vaccine has been well demonstrated over the last decade (Heath et al. 1998, Anderson et al. 1998, Goodin et al. 2007, Powell et al. 2005), additional protective components may not only augment immunity to infection, but also preclude resistance to immunity by either natural or deliberate genetic alteration of the pathogen (Adamovicz and Andrews 2003). Although other Y. pestis proteins have been reported to induce protective immunity to plague (Andrews et al. 1999, Swietnicki et al. 2005), it is possible that several of the gene products identified in this study, employed as subunit antigens, may contribute to long-term protection against disease. Further, because of the conserved nature of IVI antigens, some may be cross-protective against multiple bacterial pathogens and/or suprainfection. We are currently exploring these possibilities through the evaluation of selected purified IVI recombinant proteins as protective antigens against virulent challenge in a laboratory animal model for plague.

Footnotes

Acknowledgments

The authors thank Steven Little and Amy Permenter for their excellent technical help, as well as Kim Andrews for her critical review of this manuscript. This study was supported by the U.S. Army Medical Research Institute of Infectious Diseases Cooperative Research and Development Agreement (USAMRIID CRDA; no. W81XWH-05-0047).

Disclosure Statement

No competing financial interests exist. The opinions and assertions made in this manuscript are solely those of the authors and do not represent those of USAMRIID, the U.S. Army, USGS, or Midwest Research Institute.

References

Adamovicz

, Andrews

. Plague vaccines: retrospective analysis and future developments. Biological Weapons Defense—Principles and Mechanisms for Infectious Diseases Counter-Bioterrorism. Lindler

, Lebedo

, Korch

. Totowa, NJ: Humana Press, 2003.

Anderson

Jr. , Heath

, Bolt

, Welkos

et al. Short- and long-term efficacy of single-dose subunit vaccines against Yersinia pestis in mice. Am J Trop Med Hyg, 1998; 58:793–799.

Andrews

, Strachan

, Benner

, Sample

et al. Protective efficacy of recombinant Yersinia outer proteins (Yops) against bubonic plague caused by encapsulated and non-encapsulated Yersinia pestis. Infect Immun, 1999; 67:1533–1537.

Benner

, Andrews

, Byrne

, Strachan

et al. Immune response to Yersinia outer proteins and other Y. pestis antigens after experimental plague infection in mice. Infect Immun, 1999; 67:1922–1928.

Buchrieser

, Rusniok

, Frangeul

, Couve

et al. The 102-kilobase pgm locus of Yersinia pestis: sequence analysis and comparison of selected regions among different Yersinia pestis and Yersinia pseudotuberculosis strains. Infect Immun, 1999; 67:4851–4861.

Chamond

, Goytia

, Coatnoan

, Barale

, Cosson

, Degrave

, Minoprio

. Trypanosoma cruzi proline racemases are involved in parasite differentiation and infectivity. Mol Microbiol, 2005; 58:46–60.

Chang

, Kura

, Amemura-Maekawa

, Koizumi

et al. Identification of a novel adhesion molecule involved in the virulence of Legionella pneumophila. Infect Immun, 2005; 73:4272–4280.

Chen

, Reyes

, Clarke

, Shuman

. Host cell-dependent secretion and translocation of the LepA and LepB effectors of Legionella pneumophila. Cell Microbiol, 2007; 9:1660–1671.

Deb

, Dahiya

, Srivastava

et al. Selective identification of new therapeutic targets of Mycobacterium tuberculosis by IVIAT approach. Tuberculosis, 2002; 82:175–182.

10.

Evans

, Blaha

, Bailey

, Steitz

. The structure of LepA, the ribosomal back translocase. Proc Natl Acad Sci USA, 2008; 105:4673–4678.

11.

Felek

, Lawrenz

, Krukonis

. The Yersinia pestis auto-transporter YapC mediates host cell binding, autoaggregation and biofilm formation. Microbiology, 2008; 154:1802–1812.

12.

Foultier

, Troisfontaines

, Muller

, Opperdoes

et al. Characterization of the ysa pathogenicity locus in the chromosome of Yersinia enterocolitica and the phylogeny analysis of Type-III secretion systems. J Mol Evol, 2002; 55:37–51.

13.

Gardy

, Laird

, Chen

, Rey

et al. PSORTb v.2.0: expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis. Bioinformatics, 2005; 21:617–623.

14.

Goodin

, Nellis

, Powell

, Vyas

et al. Purification and protective efficacy of monomeric and modified Yersinia pestis capsular F1-V antigen fusion proteins for vaccination against plague. Protein Expr Purif, 2007; 53:63–79.

15.

Gort

, Miller

. Identification and characterization of Yersinia enterocolitica genes induced during systemic infection. Infect Immun, 2000; 68:6633–6642.

16.

Haller

, Carlson

, Pederson

, Pierson

. A chromosomally encoded Type-III secretion pathway in Yersinia enterocolitica is important in virulence. Mol Microbiol, 2000; 36:1436–1446.

17.

Handfield

, Brady

, Progulske-Fox

, Hillman

. IVIAT: a novel method to identify microbial genes expressed specifically during human infections. Trends Microbiol, 2000; 8:336–339.

18.

Hang

, John

, Asaduzzaman

, Bridges

et al. Use of in vivo-induced antigen technology (IVIAT) to identify genes uniquely expressed during human infection with Vibrio cholerae. Proc Natl Acad Sci USA, 2003; 100:8508–8513.

19.

Harris

, Baresch-Bernal

, Rollins

, Alam

et al. Identification of in vivo-induced bacterial protein antigens during human infection with Salmonella enterica serovar Typhi. Infect Immun, 2006; 74:5161–5168.

20.

Heath

, Anderson

Jr. , Mauro

, Welkos

et al. Protection against experimental bubonic and pnuemonic plague by a recombinant capsular F1-V antigen fusion protein vaccine. Vaccine, 1998; 16:1131–1137.

21.

Jitsurong

, Chirathaworn

, Brown

, Beachum

et al. Searching for virulence Burkholderia pseudomallei genes by immunoscreening the lambda ZAPII expressed genomic library. Southeast Asian J Trop Med Public Health, 2003; 34:810–821.

22.

Junker

, Schuster

, McDonnell

, Sorg

et al. Pertactin B-helix folding mechanism suggests common themes for the secretion and folding of autotransporter proteins. Proc Natl Acad Sci USA, 2006; 103:4918–4923.

23.

Lawrenz

, Lenz

, Miller

. A novel autotransporter adhesin is required for efficient colonization during bubonic plague. Infect Immun, 2009; 77:317–326.

24.

Matson

, Durick

, Bradley

, Nilles

. Immunization of mice with YscF provides protection from Yersinia pestis infections. BMC Microbiol, 2005; 5:38.

25.

Montes

, Zuniga

, Vazquez

, Arce

, Gruppi

. Trypanosoma cruzi mitochondrial malate dehydrogenase triggers polyclonal B-cell activation. Clin Exp Immunol, 2002; 127:27–36.

26.

Parkill

, Wren

, Thomson

, Titball

et al. Genomic sequence of Yersinia pestis, the causative agent of plague. Nature, 2001; 413:523–527.

27.

Powell

, Andrews

, Enama

, Jendrek

et al. Design and testing for a Non-tagged F1-V fusion protein as vaccine antigen against bubonic and pneumonic plague. Biotechnol Prog, 2005; 21:1490–1510.

28.

Pukatzki

, Ma

, Sturtevant

, Krastins

et al. Identification of a conserved bacterial protein secretion system in Vibrio cholerae using the Dictyostelium host model system. Proc Natl Acad Sci USA, 2005; 103:1528–1533.

29.

Pujol

, Bliska

. The ability to replicate in macrophages is conserved between Yersinia pestis and Yersinia pseudotuberculosis. Infect Immun, 2003; 71:5892–5899.

30.

Rollins

, Peppercorn

, Hang

, Hillman

et al.

In vivo induced antigen technology (IVIAT)

Cell Microbiol, 2005; 7:1–9.

31.

Rollins

, Peppercorn

, Young

, Drysdale

et al. Application of in vivo induced antigen technology (IVIAT) to Bacillus anthracis. PLoS ONE, 2008; 3:e1824.

32.

Ruiz-Albert

, Mundy

, Yu

, Beuzón

et al. SseA is a chaperone for the SseB and SseD translocon components of the Salmonella pathogenicity-island-2-encoded type III secretion system. Microbiology, 2003; 149:1103–1111.

33.

Salim

, Dennis

, Cvitkovitch

, Chang

et al. Identification of group A Streptococcus antigenic determinants up-regulated in vivo. Infect Immun, 2005; 73:6026–6038.

34.

Shalom

, Shaw

, Thomas

. In vivo expression technology identifies a type VI secretion system locus in Burkholderia pseudomallei that is induced upon invasion of macrophages. Microbiology, 2007; 153:2689–2699.

35.

Spera

, Ugalde

, Mucci

, Comerci

, Ugalde

. A B lymphocyte mitogen is a Brucella abortus virulence factor required for persistent infection. Proc Natl Acad Sci USA, 2006; 103:16514–16519.

36.

Swietnicki

, Powell

, Goodin

. Yersinia pestis Yop secretion protein F: purification, characterization, and protective efficacy against bubonic plague. Protein Expr Purif, 2005; 42:166–172.

37.

Welkos

, Friedlander

, Davis

. Studies on the role of plasminogen activator in systemic infection by virulent Yersinia pestis strain C092. Microbiol Pathog, 1997; 23:211–223.

38.

Welkos

, Pitt

MLM

, Martinez

, Friedlander

et al. Determination of the virulence of the pigmentation-deficient and pigmentation-/plasminogen activator-deficient strains of Yersinia pestis in non-human primate and mouse models of pneumonic plague. Vaccine, 2002; 20:2206–2214.

39.

Yen

, Karkal

, Bhattacharya

, Fernandez

et al. Identification and characterization of autotransporter proteins of Yersinia pestis KIM. Mol Membr Biol, 2007; 24:28–40.