* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download A secondary structure motif predictive of protein localization to the
Survey
Document related concepts
Cell membrane wikipedia , lookup
G protein–coupled receptor wikipedia , lookup
Protein phosphorylation wikipedia , lookup
Nuclear magnetic resonance spectroscopy of proteins wikipedia , lookup
Magnesium transporter wikipedia , lookup
Protein structure prediction wikipedia , lookup
Endomembrane system wikipedia , lookup
Signal transduction wikipedia , lookup
Protein moonlighting wikipedia , lookup
Intrinsically disordered proteins wikipedia , lookup
Protein–protein interaction wikipedia , lookup
List of types of proteins wikipedia , lookup
Transcript
Cellular Microbiology (2000) 2(1), 35±47 A secondary structure motif predictive of protein localization to the chlamydial inclusion membrane J. P. Bannantine,² R. S. Grif®ths, W. Viratyosin, W. J. Brown and D. D. Rockey* Department of Microbiology, Oregon State University, Corvallis, OR 97331-3804, USA. Summary Chlamydiae are obligate intracellular pathogens that spend their entire growth phase sequestered in a membrane-bound vacuole called an inclusion. A set of chlamydial proteins, labelled Inc proteins, has been identi®ed in the inclusion membrane (IM). The predicted IncA, IncB and IncC amino acid sequences share very limited similarity, but a common hydrophobicity motif is present within each Inc protein. In an effort to identify a relatively complete catalogue of Chlamydia trachomatis proteins present in the IM of infected cells, we have screened the genome for open reading frames encoding this structural motif. Hydropathy plot analysis was used to screen each translated open reading frame in the C. trachomatis genome database. Forty-six candidate IM proteins (C-Incs) that satis®ed the criteria of containing a bilobed hydrophobic domain of at least 50 amino acids were identi®ed. The genome of Chlamydia pneumoniae encodes a larger collection of C-Inc proteins, and only approximately half of the C-Incs are encoded within both genomes. In order to con®rm the hydropathy plot screening method as a valid predictor of C-Incs, antisera and/or monoclonal antibodies were prepared against six of the C. trachomatis C-Incs. Immuno¯uorescence microscopy of C. trachomatis-infected cells probed with these antibodies showed that ®ve out of six C-Incs are present in the chlamydial IM. Antisera were also produced against C. pneumoniae p186, a protein sharing identity with Chlamydia psittaci IncA and carrying a similar bilobed hydrophobic domain. These antisera labelled the inclusion membrane in C. pneumoniaeinfected cells, con®rming that proteins sharing the unique secondary structural characteristic also localize to the inclusion membrane of C. pneumoniae. Sera from patients with high-titre antibodies to C. trachomatis Received 6 August, 1999; revised 2 November, 1999; accepted 8 November, 1999. ²Present address: NADC, 2300 Dayton Ave, Ames, IA 50010, USA. *For correspondence. E-mail rockeyd@ucs. orst.edu; Tel. (1) 541 737 1848; Fax (1) 541 737 0496. Q 2000 Blackwell Science Ltd were examined for reactivity with each tested C-Inc protein. Three out of six tested C-Incs were recognized by a majority of these patient sera. Collectively, these studies identify and characterize novel proteins localized to the chlamydial IM and demonstrate the existence of a potential secondary structural targeting motif for localization of chlamydial proteins to this unique intracellular environment. Introduction Chlamydiae are obligate intracellular pathogens that cause a variety of diseases in many different animal species. In the human system, Chlamydia trachomatis causes trachoma, the leading cause of preventable blindness worldwide (Thylefors et al., 1995), and a collection of sexually transmitted conditions (Washington et al., 1987). Chlamydia pneumoniae infection initiates in the respiratory tract but has recently been implicated in a variety of chronic diseases, including cardiovascular disease and arthritis (Saario and Toivanen, 1993; Moling et al., 1996; Danesh et al., 1997; Normann et al., 1998). Although the diseases caused by chlamydiae are quite diverse, the four chlamydial species share many common features. All chlamydiae are obligate intracellular bacteria that develop within a membranebound vacuole called an inclusion. Infection of host cells is mediated by the metabolically inactive elementary body (EB). After infection, the EB differentiates into the metabolically active but non-infectious reticulate body (RB). Concurrent with RB growth and multiplication, the inclusion enlarges to accommodate the increasing number of developmental forms. Reticulate bodies are tightly associated with the inclusion membrane (IM) throughout their development, possibly using pores in the membrane to access directly the cytoplasm for nutrients. Presently, unidenti®ed signals are then sensed by the bacterium and RBs begin to differentiate back to EBs. By 48±72 h post infection (PI), the chlamydiae are released from both the vacuole and the host cell, and EBs are then free to infect another cell. Many of the aspects of the chlamydial developmental cycle have not been elucidated, including the nature of the chlamydial inclusion. Intravacuolar pathogens such as chlamydiae, Legionella, Mycobacterium and Toxoplasma modify the vacuoles they inhabit to prevent maturation to secondary lysosomes. Much has been learned about 36 J. P. Bannantine et al. some of these pathogens by examining host cell markers that co-localize with the vacuole or vacuolar membrane. The legionellae, for example, develop within a vacuole that displays properties of the endoplasmic reticulum (Vogel and Isberg, 1999). Mycobacteria occupy a vacuole that appears similar to recycling endosomes (Russell et al., 1996). Although co-localization experiments have been a useful technique for characterizing the parasitophorous vacuole of these species, this approach has been less successful in the characterization of the chlamydial inclusion. Despite a vigorous effort to identify host cell proteins that localize to the chlamydial inclusion, no such protein has been identi®ed in that environment (Heinzen et al., 1996). In contrast, two lipid markers can be shown to localize to the IM. The ¯uorescent lipid NBD-ceramide, a marker of traf®cking within the Golgi apparatus, is found transiently in the IM and is stably localized to the chlamydiae within the inclusion (Hackstadt et al., 1995). Fluorescent cholesterol analogues also localize to the inclusion, and other lipids found within the developing chlamydiae are probably of host cell origin (Carabeo et al., 1999). Collectively, these studies suggest that the inclusion probably acquires lipids from several host cell sources, including exocytic traf®cking pathways from the Golgi. Although there are no known host cell proteins localized to the IM, there are chlamydial proteins found in that environment. The chlamydiae produce and release a collection of proteins, termed Inc proteins, which are localized to the IM. Inc proteins were ®rst identi®ed in Chlamydia psittaci and have recently been identi®ed in C. trachomatis (Rockey et al., 1995; Bannantine et al., 1998a,b; Scidmore-Carlson et al., 1999). The Inc proteins are the only known bacterial proteins that reside in the membrane of a parasitophorous vacuole, but their function is unknown. Many models can be proposed addressing possible Inc protein function, including participating in inclusion development, escape from the lysosomal pathway and nutrient acquisition. One Inc protein, C. psittaci IncA, is exposed at the surface of the inclusion and is phosphorylated in cells in the absence of a chlamydial background (Rockey et al., 1997), traits consistent with possible involvement with host cell vesicular traf®cking events surrounding the development of the inclusion. A similarly localized set of proteins is produced by Toxoplasma gondii, a protozoan parasite that occupies a similarly non-acidi®ed intravacuolar environment (Lecordier et al., 1993; Beckers et al., 1994). To date, eight different Inc proteins have been shown to reside in the IM. This includes IncA from C. psittaci and C. trachomatis, IncB and C from C. psittaci and Inc D± G of C. trachomatis. The genes encoding IncA from each species share moderate identity, but within a species IncA, IncB and IncC share minimal similarity. The only obvious structural trait common among the four proteins is a bilobed hydrophobic domain of approximately 60 amino acids. This secondary structure pattern is striking, primarily in the light of low primary sequence identity either surrounding or within the hydrophobic domain. The completion of the C. trachomatis and C. pneumoniae genome sequences has created opportunities for experimentation that have been previously unavailable in the chlamydial system. In the absence of genetic techniques for manipulation of the chlamydial genome, it has been dif®cult to conduct detailed analysis of protein function. The availability of the genome sequence allowed a search for proteins that localized to the IM, based upon the presence of this characteristic hydrophobicity motif. We investigated whether open reading frames (ORFs) in the genome encode proteins containing this unique secondary structure and whether those proteins are be targeted to the IM. We therefore performed a screen of all ORFs identi®ed from the C. trachomatis genome project and identi®ed over 40 candidate inc genes (c-incs) that encode candidate Inc proteins (C-Incs). In this report, we describe a preliminary characterization of these proteins and show that ®ve out of six tested c-incs encoded proteins localized to the chlamydial IM. Additionally, we show that C. pneumoniae also encodes C-Incs and that one of these is localized to the C. pneumoniae IM. Results Hydropathy plot screen and bioinformatics analyses The hydropathy plots of several known IM proteins, for both C. trachomatis and C. psittaci, are shown in Fig. 1. Note the large bi-lobed hydrophobic domain spanning at least 50 amino acids in each. Hydropathy plot analysis was performed on all ORFs present in the C. trachomatis genome. Translated sequences that contain a bi-lobed hydrophobic stretch of at least 50 amino acids were identi®ed as C-Incs. Based on this selection criterion, 45 out of 894 ORFs were selected (Table 1). All C-Incs were subjected to gap BLAST analysis to identify similarities with known proteins in the database. No C-Inc shared obvious sequence similarity with known gene products. ORFs encoding the C-Inc proteins appear to be randomly distributed throughout the genome, although two regions contained clusters of four and seven c-inc ORFs. A four c-inc cluster (D115 to D118) is located immediately downstream of IncA (Fig. 2A). These ORFs encode proteins that have recently been shown by Scidmore-Carlson et al. (1999) to be localized to the IM. A seven c-inc cluster (D223 to D229) is adjacent to genes encoding amino acid- and sodium-dependent transporters (Fig. 2A and B). Immediately downstream of these transporter genes are the C. trachomatis genes with highest sequence similarity to C. psittaci incB and incC, genes encoding proteins known to localize to the C. psittaci IM. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 Chlamydial hydrophobicity motif Fig. 1. Comparison of hydropathy pro®les for the predicted amino acid sequences of known inclusion membrane proteins. Each of these proteins has been localized to the chlamydial inclusion membrane using ¯uorescence microscopy with speci®c antibody either previously (Rockey et al., 1995; Bannantine et al., 1998a, 1998b) or in this report (C. pneumoniae IncA). Pro®les were determined using the algorithm developed by Kyte and Doolittle (1982) with a window size of seven amino acids. The vertical axis displays relative hydrophilicity with negative scores indicating relative hydrophobicity. The horizontal axis indicates amino acid number. Note that although the size of each graph is similar, the length of the proteins are distinct. Hydrophobic regions of similar size and shape are circled in each plot. The recently available C. pneumoniae genome was also searched for c-incs. A similar number of predicted coding sequences was also identi®ed in this species (Table 1). A comparison of the c-incs from C. pneumoniae and C. trachomatis demonstrated that 22 of these ORFs have apparently homologous sequences within the other genome, whereas the remaining c-incs shared no such identity. This is consistent with BLAST data provided at the Chlamydia Genome Project website. In C. trachomatis, clusters of c-incs were tightly linked to both the incA and the incB/incC homologues, but no such arrangement was seen in C. pneumoniae. The gene order is also not Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 37 conserved in the region surrounding C. psittaci incA relative to either sequenced genome (Bannantine et al., 1998a; Stephens et al., 1998). The predicted sequences of both C. trachomatis and C. pneumoniae IncA are more similar to C. psittaci IncA than to each other. To examine the possibility that the characteristic hydrophobicity pro®le was generally present within bacterial genomes, a similar search was undertaken on the 587 unknown ORFs contained within the Mycobacterium tuberculosis genome (Cole et al., 1998). Some of these predicted proteins shared moderate similarity with the C-Inc hydrophobicity motif, but this was commonly found in proteins with a largely hydrophobic overall structure (examples include RV010c, Rv0093c, RV0996 and Rv1233c). None of these sequences shares a clear structural relationship with the C-Inc hydrophobicity motif. Although C-Incs collectively share a very distinctive secondary structural characteristic, as a group the actual primary sequence identity is minimal. Phylogenetic analyses, using programs available through the GCG software package (Wisconsin Package, SEQLAB, SEQWEB) were universally unsuccessful at assessing the global genetic relationships of the C-Incs. This was true for analyses of complete protein sequences, analyses of only the predicted hydrophobic motifs or analyses of the genes remaining after removal of the bilobed hydrophobic region. In some cases, however, there are clear examples of relationships among individual C-Incs between and/or within a species (Table 1). For example, C. trachomatis ORF D484 encodes a protein with high similarity to C. pneumoniae p602. Clustal analysis of these proteins demonstrates 67% identity and 81% similarity over the length of the proteins. A representative example of C-Inc similarity between the two species, as well as evidence of c-inc gene duplication, is provided by C. trachomatis p058 and proteins encoded by C. pneumoniae ORFs P107, P367, P369 and P370. Although the identity is not as high as that shared by D484/P602, these proteins are quite similar, especially at their carboxy termini (Fig. 3B). One of these proteins (C. pneumoniae p107) is shorter than the others and lacks the characteristic hydrophobicity motif (Fig. 3A). A ®nal example of c-Inc duplication is seen in C. pneumoniae ORFs P007±P0011, P1054 and P1055. Each of these orfs encodes highly similar gene products, and one (p009) lacks the hydrophobicity motif. This group of genes is not represented in C. trachomatis. Collectively, these analyses demonstrate that selected C-Inc proteins share identity between species, and that this similarity can occur between both C-Inc and non-CInc sequences. Selection of C-Incs for production of antisera In order to determine whether the hydrophobicity motif was a predictor of localization to the IM, antisera against 38 J. P. Bannantine et al. Table 1. Candidate inc genes identi®ed by hydrophilicity pro®le within each sequenced chlamydial genome. Candidate ORFs showing conservation between the species C. trachomatis 484 728 195 565 005 383 288 058 006 850 101 232 058 618 483 233 442 058 058 440 449 345 119 (IncA) C. pneumoniae 602 869 288 822 443 480 065 369 442 1008 312 291 370 753 601 292 556 524 367 554 565 026 186 Candidate ORFs unique to each species e-value 85 5.2 e 1.8 e 50 2.7 e 42 2.2 e 30 2 e 26 2.5 e 20 1.5 e 15 1.7 e 13 7.4 e 12 2.9 e 12 1.4 e 12 4.2 e 10 2 e 10 2 e 10 4.1 e 08 1.5 e 07 1.2 e 07 7 e 06 2 e 05 1.2 e 06 0.00045 0.0021 0.002 C. trachomatis C. pneumoniae 036 115 116 117 118 134 135 164 192 196 214 223 224 225 226 227 228 229 300 357 358 789 813 007 008 010 011 041 043 045 067 124 126 130 131 132 146 147 164 166 169 174 215 216 225 240 266 267 277 284 285 352 354 355 357 365 366 371 372 375 431 432 439 440 585 829 830 1054 1055 241 Boldface identifies ORFs examined in this work. six C-Incs were produced. Hydrophilicity pro®les for each selected protein are shown in Fig. 2B and C. Open reading frame D233 encodes the C. trachomatis gene with highest similarity to C. psittaci IncC and was selected based on this similarity. Open reading frame D442 was selected because this protein has previously been identi®ed as a cysteine- or sulphur-rich protein (crp/srp), but this designation and its role in chlamydial biology are unclear (Clarke et al., 1988; Lambden et al., 1990; de la Maza et al., 1991). The remaining four ORFs were chosen at random, with an attempt to provide a cross-section of the C-Inc proteins. ORFs D223 and D229 are c-incs that ¯ank the seven-gene cluster shown in Fig. 2A. Open reading frame D288 encodes a large protein that contains two independent hydrophobic domains and has structural characteristics of two tandem IncA-like proteins joined tail to head. Open reading frame D484 also encodes a protein structurally similar to IncA, but the predicted protein Fig. 2. Characterization of two c-inc clusters present in the C. trachomatis genome. A. ORF maps of two 10 kb segments of the C. trachomatis genome labelled contig 2.3 and contig 3.5. Open reading frames encoding C-Incs include D115 to D118 from contig 2.3 and D223 to D229 from contig 3.5. Note the close proximity of IncA, B and C to each of the respective C-Inc clusters. B. Hydropathy plots of C. trachomatis C-Incs used for production of speci®c antibody. Note that each protein contains a large bi-lobed hydrophobic domain similar to that seen for the proteins shown in Fig. 1. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 Chlamydial hydrophobicity motif 39 Fig. 3. Comparisons of predicted proteins encoded by C. trachomatis D058 and C. pneumoniae P107, P367, P369 and P370. A. Hydrophilicity pro®les for three of these predicted proteins. All but C. pneumoniae p107 have predicted sequences that contain the characteristic C-Inc hydrophobicity motif. Note the scale differences for each plot. B. Clustal similarity comparisons using truncated sequences of each protein. Protein identity and amino acid position are indicated to the left of each line. possesses a more hydrophobic pro®le over the entire protein sequence. Polyclonal antibodies against each of these six proteins were produced in rabbits, mice or guinea pigs. These sera were each checked for reactivity with the appropriate protein expressed in Escherichia coli (not shown). The sera were then used as a probe of C. trachomatis-infected HeLa cells ®xed with either methanol or paraformaldehyde. Antisera against C. trachomatis IncA and monoclonal antibody (mAb) against the chlamydial HSP60 protein were used as controls to demonstrate staining of the IM or the intracellular developmental forms respectively (Fig. 4A and B). Using this assay, ®ve out of the six tested C-Inc proteins were shown to reside in the IM. These include the C. trachomatis IncC homologue p233, the crp protein (p442) and three previously unexamined proteins encoded by ORFs D223, D229 and D288. Two of these Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 proteins, IncC and p288, could not be localized to the IM after ®xation with methanol. Fixation with paraformaldehyde followed by permeablization with Triton X-100 was used to show their localization to the IM. Under these conditions, both IncC (p233) and p288 were shown to localize to discrete patches within the IM (Fig. 4E and F). Antisera to p223 also labelled C. trachomatis L2 inclusions in a unique pattern (Fig. 4D). The distinction between IncA and p223 localization is evident in cells doubly labelled with anti-IncA and anti-p223 (Fig. 5). Antibody directed at p223 revealed a distinct patchy staining pattern, relative to the uniform pattern observed for IncA. The unique labelling pattern observed with anti-p223 antisera was examined in more detail, using mAb anti-p223 and anti-HSP60 in double-label ¯uorescence microscopy. Each of two anti-p223 mAbs labelled the IM similarly to the polyclonal anti-p223. Laser confocal microscopic analysis 40 J. P. Bannantine et al. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 Chlamydial hydrophobicity motif 41 Fig. 5. C. trachomatis-infected cell ®xed with methanol and doubly labelled with anti-IncA (A) and anti-p223 (B) antisera. This image highlights the distinctions evident between the staining with anti-IncA and anti-p223. of C. trachomatis-infected cells probed with these antibodies showed that the patches in the IM occupied by p223 were largely devoid of contact with chlamydial developmental forms (Fig. 6). This was true of cells ®xed with both paraformaldehyde and methanol. An analysis of p223 localization throughout chlamydial development was also undertaken. This protein began to accumulate on the membrane of inclusions containing approximately eight or more RB (< 8±10 h post infection; not shown). Antigen localization observed at these early time points was similar to that shown in Fig. 4. Antisera to one C-Inc, p484, did not label the IM of either early or late inclusions ®xed with either paraformaldehyde or methanol. The labelling of this antigen showed distribution in the developmental forms within the inclusion (Fig. 4H). Of all the C-Incs, p484 shows the highest similarity with a corresponding protein in the C. pneumoniae genome sequence. Because of this similarity, anti-p484 was examined for reactivity with C. pneumoniae-infected cells. Cross-reactivity was not strong, but the antisera weakly localized at or around the inclusion membrane in this species (not shown). The sequencing of the C. pneumoniae genome also identi®ed a gene encoding a protein (p186) sharing < 21±23% identity with both C. trachomatis and C. psittaci IncA. Each previously identi®ed IncA has been shown to reside within the inclusion membrane of infected cells; (Rockey et al., 1995; Bannantine et al., 1998a). As can be seen in Table 1, C. pneumoniae p186 was also identi®ed in the initial hydrophobicity screening process, as it possesses the characteristic hydrophobicity motif present in each known Inc protein. Antisera were produced against p186 and used in ¯uorescent microscopic analysis of C. pneumoniae-infected cells ®xed with methanol 44 h post infection. The antisera clearly label the inclusion membrane in these cells, in contrast to anti-HSP60, which labelled the developmental forms within the inclusion (Fig. 7). Additionally, antigenic ®bres were observed extending away from the inclusion in these cells (not shown), consistent with results seen with both C. psittaci and C. trachomatis IncA. These data demonstrated that proteins of chlamydial origin are found in the C. pneumoniae inclusion membrane, and provide more evidence that the hydrophobicity domain is characteristic of proteins in the chlamydial IM. Recognition of C-Incs by human sera To examine the immunogenicity of the selected C-Incs in patients diagnosed and treated for a sexually transmitted C. trachomatis infection, immunoblots were produced and probed with sera collected from Chlamydia-infected patients. Recognition of the C-Inc fusion proteins was different for each antigen. IncA, as has been shown previously (Bannantine et al., 1998a), was recognized by several of these patients. P223, p229 and p442 (crp) were also recognized by over a half of these patients. The remaining Fig. 4. Immuno¯uorescence microscopy of C. trachomatis-infected cells probed with antisera against each tested C-Inc. Panels A and B represent a cell doubly labelled with anti-IncA (A) and HSP 60 (B) and demonstrate the characteristic labelling evident for known Inc proteins. Each subsequent panel represents cells singly labelled with each anti-C-Inc antisera. C, p229; D, p223; E, p228; F, p233; G, p442; H, p484. Arrows indicate the inclusion membrane in E, F and G. Bar in H represents 10 microns for each image. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 42 J. P. Bannantine et al. Fig. 6 Laser confocal microscopic images of methanol-®xed cells infected with C. trachomatis. Cells were doubly labelled with anti-p223 (red) and anti-HSP60 (green). Notice the localization of p223 to areas of the inclusion apparently devoid of contact with chlamydial developmental forms. antigens were not consistently recognized by antibodies within these sera and/or control sera were positive with the antigen. High-titre antisera from two patients who had recovered from C. pneumoniae infection were also negative against the C-Inc proteins (Fig. 8). Therefore, similarly to the situation observed with guinea pigs experimentally infected with C. psittaci (Bannantine et al., 1998a), some Inc proteins are antigenic in the context of infection whereas others are not. Discussion Many aspects of the molecular and cellular biology surrounding the interaction of chlamydiae with their host cell are currently unclear. A major area of interest surrounds the interaction between the chlamydial inclusion and the host intracellular environment. The inclusion interacts with the Golgi apparatus as the transfer of ¯uorescent lipid analogues is traf®cked from the Golgi to the inclusion. This transfer is energy dependent and requires chlamydial protein synthesis (Hackstadt et al., 1995, 1996), but the molecules that participate in these interactions are unknown. It is likely that proteins at the surface of the inclusion mediate some form of vesicular transport to the inclusion. The Inc proteins of C. psittaci and C. trachomatis are candidates for these processes. A valuable resource for chlamydiologists has been the recent completion of the genome sequences for C. trachomatis and C. pneumoniae. The genome project is critical to our understanding of chlamydial biology and has opened avenues of research that were previously unavailable. The present study is an example of the utility of the databases. Without their availability the identi®cation of the large number of c-incs from each species would not have been possible. Within each species, the C-Incs occupy a signi®cant fraction of the genome ± Fig. 7. Double-label ¯uorescence microscopy of C. pneumoniae-infected HeLa cells ®xed with methanol 44 h PI. Cells were labelled with anti-HSP60 (A) and anti-C. pneumoniae IncA (B). Note that IncA is distributed to the margins of the chlamydial inclusion, similarly to as seen with C. psittaci strain GPIC (Rockey et al., 1995) and C. trachomatis (Fig. 4, A). Bar indicates 10 microns. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 Chlamydial hydrophobicity motif Fig. 8. Immunoblot analysis of puri®ed C-Inc-MBP fusion proteins probed with a collection of sera from 13 Chlamydia-infected patients, two C. pneumoniae-infected patients, and two negative control patients. The results for IncA have previously been reported (Bannantine et al., 1998a). 2.7% for C. trachomatis and 5.5% for C. pneumoniae. These genes represent a much higher percentage of the chlamydial unidenti®ed predicted protein sequences. Analysis of the global sequence databases also indicates that the C-Incs are unique ± they appear to be chlamydia speci®c in their distribution. The most obvious trait shared by the known Inc proteins is a bilobed hydrophobic domain of < 60 amino acids. This motif is perplexing because of its apparent lack of a primary amino acid sequence signature. This lack of similarity was emphasized in the present study. Statistical phylogenetic analyses failed to identify signi®cant relationships among the different C-Inc protein sequences, including both those proteins shown to reside in the IM and those that have not yet been tested. Even in the case of Incs that are otherwise quite similar, the identity within the hydrophobic domains is often not strong. This lack of similarity at the primary sequence level suggests that secondary structure constraints may be the driving force behind retention of this region; the primary sequence may serve only as a means to frame the hydrophobic domains. The large group of c-incs are not the only unique gene family identi®ed in the chlamydiae. A collection of clustered genes encoding several large putative membrane proteins (Pmps) has also been identi®ed, by both screening conventional expression libraries and using a genomics analysis (Grimwood et al., 1998; Longbottom et al., 1998; Stephens et al., 1998). Chlamydia trachomatis has a total of nine genes encoding Pmps scattered throughout the genome, whereas the C. pneumoniae genome encodes 20 Pmps. Although there are apparent examples of highly related Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 43 Pmp sequences within each chlamydial species, as a family they share relatively limited primary sequence identity. These proteins are instead grouped on the basis of size, gene clustering and by the presence of several copies each of distinctive four amino acid repeat sequences. The data surrounding c-inc and pmp sequences indicate that large gene `families' exist in the chlamydiae, even under the constraints of a small genome. In the case of both gene groups, the similarity is not based on high primary sequence identity, but instead on more cryptic patterns associated with short amino acid repeats or secondary structure characteristics. This study identi®ed over 40 C-Inc proteins from each chlamydial species. When six C. trachomatis C-Inc sequences were chosen for analysis, ®ve were shown to reside in the IM. The tested C-Incs lack signi®cant primary identity, but each contains the unique bilobed hydrophobic domain. Based on these results, we propose that this hydrophilicity motif can be used as a predictor of localization to the IM. However, the distribution of each C-Inc within the inclusion was not uniform. p223 was shown to be present in discrete patches at the surface of the inclusion, whereas IncA in the same cells was shown to be uniformly distributed around the inclusion surface. Two other tested C-Incs, p233 and p288, also localized uniquely to patches in the IM. Although efforts were made to minimize the possibility, differences in localization in the IM may be artifactual and introduced by the ®xation process. One result that is apparently not an artifact is the lack of p484 localization to the IM. This protein was localized to developmental forms in C. trachomatis inclusions, regardless of the ®xation used. This fact demonstrates that the hydrophilicity motif is not necessarily an absolute predictor of IM localization and that other unidenti®ed structural constraints may come into play as well. D484 encodes a protein with a generally more hydrophobic nature and it is possible this may affect its localization. Comparative analysis of the two sequenced genomes allowed the identi®cation of C-Incs sharing identity between the two species. Some C-Incs were highly similar, such as C. trachomatis p484 and C. pneumoniae p602. Within C. trachomatis, p484 could not be localized to the inclusion membrane and the situation with p565 is not clear. These analyses have also demonstrated that the family of c-inc sequences has been expanded in C. pneumoniae, similarly to what has been seen for the pmp genes. The data in Table 1 show that there are at least 60 c-incs in C. pneumoniae, over 20 more than are found in C. trachomatis. In some cases, this expansion was apparently facilitated by multiple gene duplications (p367±370; p007±p0011). Analysis of the C. pneumoniae genome sequence also facilitated the identi®cation of C. pneumoniae IncA (p186), and we used this sequence to produce anti-C. pneumoniae IncAantisera. These antisera uniformly labelled the inclusion 44 J. P. Bannantine et al. membrane in ®xed, C. pneumoniae-infected cells, and demonstrated conclusively that this species also encodes proteins localized to the inclusion membrane in a similar fashion to that seen in the other chlamydial species. Protein p186 also contains the characteristic bi-lobed hydrophobic domain, supporting the conclusion that this trait may be a predictor of localization to the IM. Chlamydia psittaci IncA and IncC were initially detected by sera from C. psittaci-infected guinea pigs (Rockey et al., 1995; Bannantine et al., 1998a). Chlamydia trachomatis IncA is also recognized by a majority of sera from C. trachomatis-infected humans as well as experimentally infected primates (Bannantine et al., 1998b). Antibodies to linear Inc protein epitopes are not, however, universal. C. psittaci IncB was identi®ed as an IM-localized protein because it is encoded immediately adjacent to IncC and because it possesses a similar hydrophobic domain (Bannantine et al., 1998a). The fact that three out of four known Inc proteins were immunogenic in the context of infection led us to determine whether the C-Incs tested in this work were immunogenic in humans after natural C. trachomatis infection. The results paralleled those observed in the guinea pigs. Linear epitopes of three out of ®ve identi®ed Inc proteins (p223, p229 and p442/crp) were recognized by a majority of these antisera whereas the other Incs (p233/IncC and p288) were not. Therefore, different Incs are recognized differently by both animals and humans after experimental or natural infection. Although several Inc proteins are localized to the IM, virtually nothing is known of their function. C. psittaci IncA is a serine/threonine phosphoprotein that contacts the cytosol of the host cell (Rockey et al., 1997), and nothing is known of other Incs except that they localize to the IM. The mechanisms used by chlamydiae to target the Inc proteins to the IM are also uninvestigated. None of the demonstrated Incs has classical amino-terminal secretory signals and therefore it is likely that they are secreted by other means. Selected genes of type III secretory processes have homologues in the chlamydial genome (Hsia et al., 1997) and therefore it is possible that Inc proteins may be effector proteins transported out of the chlamydial RB via these means. There is evidence that at least one Inc protein (C. trachomatis IncA) may have a function in chlamydial development. A collection of naturally occurring C. trachomatis isolates encoding mutant Inc A proteins (Suchland et al., 1999). IncA was not detected in these mutants, either by immunoblot or by ¯uorescence microscopy. Each mutant was identi®ed by the identi®cation of a defect in the fusion of C. trachomatis inclusions with one another. The genes encoding two mutant Inc proteins have been sequenced, and each encodes proteins with lesions within the ®rst lobe of the hydrophobic domain. This suggests the domain is important to IncA function but leaves open several possibilities for its role in chlamydial biology. The large number of C-Inc proteins allows for the possibility that the chlamydiae have an extraordinary repertoire of unique proteins in contact with the cytosol of the infected cell, but their function remains elusive. Inclusion membrane proteins are probably important in some aspect of communication with the host cell cytoplasmic environment. This may include aspects of intracellular vesicular traf®cking, establishment or maintenance of the inclusion and/or nutrient acquisition. The size and shape of the bilobed domain present in all C-Incs, combined with the work of Suchland et al. (1999) described above, leads to speculation regarding the function of the domain. It is possible that the hydrophobicity motif is associated with Inc protein localization; one lobe of the motif may be embedded in the chlamydial cell wall whereas the other may be in the IM. These proteins may then serve to anchor the pathogen to the IM, facilitating direct contact between the reticulate body and the IM as well as providing molecular contact between the RB and the cytosol. Further experimentation will be directed at biochemical and, possibly, genetic experiments designed to elucidate Inc protein transport/ localization within the chlamydial inclusion. We also anticipate that these studies will address possible functions of this large variety of proteins localized to the chlamydial inclusion membrane. Experimental procedures Chlamydiae and culture conditions Chlamydia trachomatis strain LGV 434 (serotype L2) and C. pneumoniae strain TW-183 were cultured in HeLa cells grown in minimal essential medium (MEM) plus 10% fetal calf serum and 1 mg ml 1 gentamicin (MEM-10), unless otherwise indicated. Puri®ed chlamydiae were prepared as has been described by Caldwell et al. (1981). All chlamydiae were stored at 808C in SPG buffer (0.25 M sucrose, 10 mM sodium phosphate, 5 mM L-glutamic acid pH 7.2) prior to use. Infected cells to be used for microscopy were produced by inoculating chlamydiae at a multiplicity of infection (MOI) of 0.1 onto semicon¯uent cells cultured on sterile glass coverslips. Chlamydiae diluted in Hanks' balanced salt solution (HBSS) were added to cells and incubated for 1 h at room temperature with gentle rocking. Inoculae were removed and cells cultured in MEM-10 for between 30 and 40 h. Monolayers were then ®xed with absolute methanol for 10 min or 4% paraformaldehyde/PBS for 1 h and subsequently stored at 48C in phosphate-buffered saline (PBS) prior to use. Cells ®xed with paraformaldehyde were permeablized with 0.1% Triton X100 (Pierce Chemical Co.) in PBS for 2 min and washed twice with PBS prior to blocking and incubating with antibody. Sequence analysis/bioinformatics analysis Sequence analysis was performed on ORFs identi®ed from the C. trachomatis and C. pneumoniae genome databases (Stephens et al., 1998; available at http://chlamydia-www. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 Chlamydial hydrophobicity motif berkeley.edu:4231/). All ORF designations are consistent with those de®ned in these databases. Proteins encoded by each c-inc are indicated with a lower case `p' designation (i.e. ORF D223 encodes p233). Hydropathy plot and Clustal analyses were conducted using MACVECTORTM 6.0 (Oxford Molecular). Hydropathy was calculated using KYTE-DOOLITLE algorithms and a window of seven (Kyte and Doolittle, 1982). Repetitive examination of individual ORFs was facilitated using QUICKEYS software (CESoftware). Homology searches of Inc proteins were performed using the gap BLAST algorithm (available at http://www.ncbi.nlm.nih. gov/ BLAST; Altschul et al., 1997). The BLAST 2 sequences program was used to evaluate the similarity between C. trachomatis and C. pneumoniae IncA (Tatusova and Madden, 1999). Antigens, antisera and mAbs Candidate inc genes were ampli®ed using primers indicated in Table 2. In some cases (indicated in Table 2), the 58 end of the gene was deleted because full-length constructs were apparently toxic to E. coli. The ampli®cation products were cloned into pMAL-c2 (New England Biolabs) and overexpressed in E. coli. In each case, the puri®ed fusion product, including the maltose-binding protein fusion partner, was used as the antigen for immunization. Controls included antisera raised against puri®ed maltose-binding protein. The immunization of rabbits, guinea pigs or mice was performed as has been previously described (Rockey et al., 1995; Bannantine et al., 1998a). Monoclonal antibodies against one C-Inc protein, p223, were produced with the assistance of the Monoclonal Antibody Facility of the University of Oregon. Mice immunized with p223 were sacri®ced and splenic lymphocytes collected. These cells were fused to SP/2 myeloma cells, using a modi®cation of standard hybridoma production techniques (Marusich, 1988). Because of solubility problems associated with the MBP/p223 fusion protein, each well was screened by immunoblot, using MBP/ p223 as the antigen. Positive cell lines were cloned and mAbs harvested from hybridoma culture supernatants. Two clones, 12 E7 (IgG1,k) and 20 F12 (IgG2a,k), were used in these studies. Hybridoma supernatant from clone A57 B9 (IgG1,k), a clone producing a mAb directed at the chlamydial HSP 60 protein, was produced and collected using methods described by Yuan et al. (1992). Appropriate secondary antibodies were purchased from Pierce or acquired from the Southern Biotechnology Associates. Methods used for production of antisera were examined and approved by the Oregon State University Animal Care and Use Committee. 45 Human sera that demonstrated high titres to C. trachomatis or C. pneumoniae using a microimmuno¯uorescence assay were selected from stored frozen specimens at the University of Washington. Serum samples with high titres against C. pneumoniae did not have detectable antibody directed at C. trachomatis-speci®c antigens. Using microimmuno¯uorescence, negative control antisera were chosen based on nonreactivity with chlamydiae. Immuno¯uorescence microscopy and immunoblotting Chlamydia-infected cells, cultured and ®xed as described above, were immunostained for localization of speci®c antigens within infected cells. For these assays, all polyclonal antisera were diluted 1:100 in PBS containing 2% (w/v) BSA (PBS-BSA), followed by ¯uorescein- or rhodamine-conjugated secondary antibodies diluted in PBS-BSA. Staining was visualized using either a Leitz laser scanning confocal microscope or a Zeiss epi¯uorescence microscope. Hybridoma supernatants were diluted 1:10 in BSA-PBS prior to incubation on cells. Digital images collected on the confocal microscope were processed through Adobe PHOTOSHOP and CANVAS graphics software (Deneba). Images were collected on the standard microscope using TMAX ®lm (Kodak). Electrophoresis and immunoblotting were conducted using described procedures (Rockey et al., 1995), with the following modi®cations. Comparisons of the reactivity of human sera with a single antigen were conducted using a slot-blotting apparatus (Bio-Rad). Lysates of puri®ed fusion proteins were electrophoresed through preparative 12% polyacrylamide gels and transferred to nitrocellulose. These ®lters were placed into a blocking solution (PBS-BSA plus 0.1% Tween-20; TPBS-BSA) for 1 h. After blocking, the ®lters were placed into the slot-blot device. Puri®ed MBP was used as a control in these immunoblots. Each human serum sample was diluted 1:10 in TPBS-BSA and added to independent slots. After 2 h of incubation with gentle rocking, blots were washed three times with TPBS and incubated in 35S protein A (100±150 nCi ml 1; Amersham) for 1 h. Blots were washed three times in TPBS, dried and placed on autoradiography ®lm at room temperature for 16 h. Films were scanned into Adobe PHOTOSHOP for image processing. Acknowledgements We thank Walter Stamm, Bob Suchland and Linda Cles of the Table 2. Olignucleotides used for preparation of c-inc/malE hybrid genes for production of fusion proteins. Target ORF 58-Oligonucleotide 38-Oligonucleotide Gene region C. trachomatis D119 D223 D442 D229 D233 D288 D484 AGCCATAGGATCTGGTTTCAGCGA ATGGTGAGTTTAGCATTAGGGA ATGAGCACTGTACCCGTTGTT ATGAGCTGTTCTAATATTAAT ATGACGTACTCTATATCCGAT CGAAAACAAAAACAACTAGAAGAG ATAGTGCCTCCCTACGACTCTATC GCGCGGATCCTAGGAGCTTTTTGTAGAGGGTGA GCGCGGGATCCTACACCCGAGAGCCGTAATTGAA GCGCGGATCCTCATTGGGTCTGATCCACCAG GCGCGGATCCTTATTTTTTACGACGGGATGCCTG GCGCAAGCTTAGCTTACATATAAAGTTTG GCACAAGCTTAATGCACTCTCCACAAATCTTCTAA GCGCGCAAACTTCTATTTAGCTACTACTACCCCGTAGAA Full length Full length Full length Full length Full length 316±1452 433±1014 GCGCGGATCCTTACTGACCTCCTG Full length C. pneumoniae P186 GCGCGGAATTCATAAGTGGAGGAGAAG Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 46 J. P. Bannantine et al. University of Washington for productive discussions and technical assistance during the course of this project. Discussions with Steve Giovannoni of Oregon State University were valuable in the genomics analyses. This work was supported by PHS grant AI42869-01 to D.D.R. and the O.S.U. Department of Microbiology Tartar scholarship Fund. Oregon Agricultural Experiment Station Technical Paper # 11544. References Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.J. (1997) Gapped BLAST and PSIBLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389±3402. Bannantine, J.P., and Rockey, D.D. (1999) Use of a primate model system to identify Chlamydia trachomatis protein antigens recognized uniquely in the context of infection. Microbiology 145: 2077±2085. Bannantine, J.P., Stamm, W.E., Suchland, R.J., and Rockey, D.D. (1998b) Chlamydia trachomatis IncA is localized to the inclusion membrane and is recognized by antisera from infected humans and primates. Infect Immun 66: 6017±6021. Beckers, C.J., Dubremetz, J.F., Mercereau-Puijalon, O., and Joiner, K.A. (1994) The Toxoplasma gondii rhoptry protein ROP 2 is inserted into the parasitophorous vacuole membrane, surrounding the intracellular parasite, and is exposed to the host cell cytoplasm. J Cell Biol 127: 947±961. Caldwell, H.D., Kromhout, J., and Schachter, J. (1981) Puri®cation and partial characterization of the major outer membrane protein of Chlamydia trachomatis. Infect Imuun 31: 1161±1176. Carabeo, R.A., Mead, D.J., and Hackstadt, T. (1999) Traf®cking of cholestrol to the chlamydial inclusion in HeLa cells. Abstr. from A Cell Biology Approach to Microbial Pathogenesis. April 25±28, Portland, OR. Clarke, I.N., Ward, M.E., and Lambden, P.R. (1988) Molecular cloning and sequence analysis of a developmentally regulated cysteine-rich outer membrane protein from Chlamydia trachomatis. Gene 71: 307±314. Cole, S.T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris, D.. et al. (1998) Deciphering the biology of Mycobacterium tuberculosis from the complete genome sequence. Nature 393: 537±544. Danesh, J., Collins, R., and Peto, R. (1997) Chronic infections and coronary heart disease: is there a link? Lancet 350: 430±436. Grimwood, J., Mitchell, W., Stephens, R.S. (1998) Phylogenetic analysis of a multigene family conserved between Chlamydia trachomatis and Chlamydia pneumoniae. In: Chlamydial Infections: Proceedings of the Ninth International Symposium on Human Chlamydial Infection. R.S. Stephens, G.I. Byrne, G. Christiansen, I.N. Clarke, J.T. Grayston, R.G. Rank, G.L. Ridgway et al., (eds). San Francisco, CA: pp. 263±266. Hackstadt, T., Scidmore, M.A., and Rockey, D.D. (1995) Lipid metabolism in Chlamydia trachomatis-infected cells: directed traf®cking of Golgi-derived sphingolipids to the chlamydial inclusion. Proc Natl Acad Sci U A 92: 4877±4881. Hackstadt, T., Rockey, D.D., Heinzen, R.A., and Scidmore, M.A. (1996) Chlamydia trachomatis interrupts an exocytic pathway to acquire endogenously synthesized sphingomyelin in transit from the Golgi apparatus to the plasma membrane. EMBO J 15: 964±977. Heinzen, R.A., Scidmore, M.A., Rockey, D.D., and Hackstadt, T. (1996) Differential interaction with endocytic and exocytic pathways distinguish parasitophorous vacuoles of Coxiella burnetii and Chlamydia trachomatis. Infect Immun 64: 796±809. Hsia, R.C., Pannekoek, Y., Ingerowski, E., and Bavoil, P.M. (1997) Type III secretion genes identify a putative virulence locus of Chlamydia. Mol Microbiol 25: 351±359. Kyte, J., and Doolittle, R.F. (1982) A simple method for displaying the hydropathic character of a protein. J Mol Biol 157: 105±132. Lambden, P.R., Everson, J.S., Ward, M.E., and Clarke, I.N. (1990) Sulfur-rich proteins of Chlamydia trachomatis: developmentally regulated transcription of polycistronic mRNA from tandem promoters. Gene 87: 105±112. Lecordier, L., Mercier, C., Torpier, G., Tourvieille, B., Darcy, F., Liu, J.L., et al. (1993) Molecular structure of a Toxoplasma gondii dense granule antigen (GRA 5) associated with the parasitophorous vacuole membrane. Mol Biochem Parasitol 59: 143±153. Longbottom, D., Russell, M., Dunbar, S.M., Jones, G.E., and Herring, A.J. (1998) Molecular cloning and characterization of the genes coding for the highly immunogenic cluster of 90-kilodalton envelope proteins from the Chlamydia psittaci subtype that causes abortion in sheep. Infect Immun 66: 1317±1324. Marusich, M.F. (1988) Ef®cient hybridoma production using previously frozen splenocytes. J Immunol Methods 114: 155±159. de la Maza, L.M., Fielder, T.J., Carlson, E.J., Markoff, B.A., and Peterson, E.M. (1991) Sequence diversity of the 60-kilodalton protein and of a putative 15-kilodalton protein between the trachoma and lymphogranuloma venereum biovars of Chlamydia trachomatis. Infect Immun 59: 1196±1201. Moling, O., Pegoretti, S., Rielli, M., Rimenti, G., Vedovelli, C., Pristera, R., and Mian, P. (1996) Chlamydia pneumoniae ± reactive arthritis and persistent infection. Br J Rheumatol 35: 1189±1190. Normann, E., Gnarpe, J., Gnarpe, H., and Wettergren, B. (1998) Chlamydia pneumoniae in children with acute respiratory tract infections. Acta Paediatr 87: 23±27. Rockey, D.D., Heinzen, R.A., and Hackstadt, T. (1995) Cloning and characterization of a Chlamydia psittaci gene coding for a protein localized in the inclusion membrane of infected cells. Mol Microbiol 15: 617±626. Rockey, D.D., Grosenbach, D., Hruby, D.E., Peacock, M.G., Heinzen, R.A., and Hackstadt, T. (1997) Chlamydia psittaci IncA is phosphorylated by the host cell and is exposed on the cytoplasmic face of the developing inclusion. Mol Microbiol 24: 217±228. Russell, D.G., Dant, J., and Sturgill-Koszycki, S. (1996) Mycobacterium avium- and Mycobacterium tuberculosis-containing vacuoles are dynamic, fusion-competent vesicles that are accessible to glycosphingolipids from the host cell plasmalemma. J Immunol 156: 4764±4773. Saario, R., and Toivanen, A. (1993) Chlamydia pneumoniae as a cause of reactive arthritis. Br J Rheumatol 32: 1112. Scidmore-Carlson, M.A., Shaw, E.I., Dooley, C.A., Fischer, E.R., and Hackstadt, T. (1999) Identi®cation and characterization of a Chlamydia trachomatis early operon encoding four novel inclusion membrane proteins. Mol Microbiol 33: 753±765. Stephens, R.S., Kalman, S., Lammel, C., Fan, J., Marathe, R., Aravind, L., et al. (1998) Genome sequence of an obligate intracellular pathogen of humans: Chlamydia trachomatis. Science Suchland, R.J., Rockey, D.D., Bannantine, J.P., and Stamm, W.E. (2000) Clinical isolates of Chlamydia trachomatis that occupy nonfusogenic inclusions lack IncA, a protein localized to the inclusion membrane. Infect Immun 68: issue 1. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 Chlamydial hydrophobicity motif Tatusova, T.A., and Madden, T.L. (1999) BLAST 2 Sequences, a new tool for comparing protein and nucleotide sequences. FEMS Microbiol Lett 174: 247±250. Thylefors, B., Negrel, A.D., Pararajasegaram, R., and Dadzie, K.Y. (1995) Global data on blindness. Bull World Health Organ 73: 115±121. Vogel, J.P., and Isberg, R.R. (1999) Cell biology of Legionella pneumophila. Curr Opin Microbiol 2: 30±34. Q 2000 Blackwell Science Ltd, Cellular Microbiology, 2, 35±47 47 Washington, A.E., Johnson, R.E., and Sanders, Jr, L.L. (1987) Chlamydia trachomatis infections in the United States. What are they costing us? JAMA 257: 2070±2072. Yuan, Y., Lyng, K., Zhang, Y.X., Rockey, D.D., and Morrison, R.P. (1992) Monoclonal antibodies de®ne genus-speci®c, species-speci®c, and cross-reactive epitopes of the chlamydial 60 kilodalton heat shock protein (hsp60): speci®c immunodetection and puri®cation of chlamydial hsp60. Infect Immun 60: 2288±2296.