Molecular Microbiology {1992) 6(17), 2539-2548
Chlamydia trachomatis Mip-like protein Anker G. Lundemose,'"'^ Duncan A. Rouch,^ Svend Birkelund/ Gunna Christiansen^ and John H. Pearce^ institute of Medical Microbiology. Bartholin Building. University of Aarhus. DK-8000 Aarhus C. Denmark. ^Microbial Molecular Genetics and Cell Biology Group. University of Birmingham. Birmingham. UK.
Summary A 27 kDa Chlamydia trachomatis Mip-like protein with homology of a 175-amino-acid C-terminal fragment to the surface-ex posed LegioneUa pneumophila mipgene product has previously been described. In this paper the entire chlamydia Mip-like sequence of C. trachomatis serovar L2 (lymphogranuloma venereum (LGV) biovar) is presented. The sequence shows high similarity to the legionella Mip protein and its C-terminal region, like that of the legionella Mip, has high amino acid similarity to eukaryotic and prokaryotic FK506-binding proteins. The chlamydial mip-Wke gene was detected by polymerase chain reaction (PCR) in other C. trachomatis serovars and by sequencing of the mip-Wke genes of serovars B and E (trachoma biovar) was shown to be highly conserved within the two major biovars of C. trachomatis. Monocionai and polyclonal antibodies raised against the recombinant Mip-like protein failed to demonstrate surface-exposed epitopes on infectious elementary bodies or reproductive reticulate body forms either by immunofluorescence or immuno-gold electron microscopy. iHowever, a complement-dependent inhibition of up to 91% of infectivity for cell cultures was observed with antibodies to the N-terminal fragment of the Mip-like protein suggesting that antibodyaccessible epjtopes are present on infectious EBs.
Introduction The Chlamydia are obligate intracellular bacteria with a remarkable biphasic lifecycle. Following entry into host cells intracellular survival is believed to depend on avoidance of the lysosomal compartment and the infectious
Received 25 March, 1992: revised and accepted 27 May, 1992, •For correspondence, Tel, 86139711; Fax 86196128.
elementary body (EB) undergoes transformation to a noninfectious metabolicaily active reticulate body (RB) form that divides by binary fission. The life cycle is completed with differentiation of RBs to EBs and release of the latter for initiation of further cycles of infection. The species Chlamydia trachomatis is an important pathogen of man, with members of the trachoma biovar being responsible for the blinding eye disease, trachoma, and for sexually transmitted diseases; strains of the lymphogranuloma venereum (LGV) biovar are also sexually transmitted but are less common. Recent studies on initiation of infection implicate the major outer membrane protein both as a source of antigenic diversity and, via variable and conserved moieties, as an adhesin {Su et ai, 1990). Other proteins may also be involved in attachment (Joseph and Bose, 1991; Schmiel et ai. 1991), Entry into host cells appears to occur by both phagocytosis and pinocytosis {Ward and Murray, 1984; Hodinka et ai. 1986; Reynolds and Pearce, 1990; 1991), However, not all endocytosed organisms go on to multiply and events during entry appear to be pivotal for productive infection. The molecular determinants of the invasion stage of the cycle are unknown and. as for invasive enteric bacteria, are of intense interest, both biologically and for their role as possible vaccine antigens. We have previously described a 27 kOa protein of C. tracfiomatis serovar L2 (LGV biovar) and the 3'-end of the encoding gene (Lundemose et ai. 1991). The deduced amino acid sequence ot a 175-amino-acid C-terminal fragment showed high similarity to a LegioneUa pneumophila mip {macrophage infectivity potentiator)-gene product (Engleberg et ai. 1989) that is a virulence factor (Cianciatto et ai. 1990) known to play an important role in initiating the intracellular infection of L. pneumophila in human macrophages (Cianciotto etal. 1989) In this paper we present the cloning and analysis of the entire coding sequence of the Mip-like protein of the L2 serovar and investigation of its possible surface exposure. The data show similarity both with the Legionella Mip protein and with the eukaryotic and prokaryotic FK506-binding proteins (FKBPs). The chlamydial m/p-like gene is shown to be present in other C. trachomatis serovars and by sequence of the m/p-like genes of C, trachomatis serovars B and E (trachoma biovar) to be highly conserved within both biovars. Monoclonal and polyclonal antibodies against the recombinant Mip-like
2540
A. G. Lundemose eXa\.
1 TTTTTTACAGCATGTGATAAGGGTGGATGTCTTTCACCTAGGAAACGGTTTTCAAAAAAG RBS 61 CACATCTGAGAAAATCAGGTGCAACTGAGGAAGAAACGAATGAAGAATATATTAAGTTGG MetLysAsnIleLeuSerTrp
7
121 ATGCTTATGTTTGCAGTCGCTCTGCCTATCGTAGGATGTGATAACGGRGGCGGTTCGCAA MetLeuMetPheAlaValAlaLeuProIleValGlyCysAspAsnGlyGlyGlySerGln
27
181 ACATCGGCTACGGAGAAAAGCATGGTAGAAGACTCTGCATTGACAGACAATCAAAAGTTA ThrSerAlaThrGluLysSerMetValGluAspSerAlaLeuThrAspAsnGlnLysLea
47
2 41 TCAAGAACTTTTGGGCATTTATTGGCTCGTCAGTTGAGCCGAACGGAAGATTTTTCGTTA SerArgThrPheGlyHisLeuLei: M a irgGlnLeuSerArgThrGluAspPheSerLeu 67 Ser 301 GATCTTGTTGAAGTGATTAAAGGGATGCAATCTGAAATAGATGGACAGAGTGCTCCTTTA AspLeuValGluVallleLysGlyMetGlnSerGluIleAspGlyGlnSerAlaProLeu 97
Fig. 1. Sequence of Ihe C. trachomatis serowai L2 mip-like gene. The postulated ribosomal binding site is labelled RBS, The large arrow-head indicates the possible location of the cleavage site between the leader peptide and the mature protein; stem loop is indicated with arrows. Predicted molecular weight ot the polypeptide was 26 648 and moLwt of the mature protein was 24 531. The two boxes enclose the amino acid sequence differences between L2 and B/E sequences. In the sequence previously published (Lundemose etal.. 1991) bases no, 307-308 were cited TG, but should be GT, which is corrected in this sequence. The sequence is assigned to the EMBL database under accession numbers X66126(L2), X56127(E) and X66128(B),
3 61 ACAGATACAGAATATGAAAAACAAATGGCAGAAGTACAAAAAGCTAGTTTCGAAGCAAAA ThrAspTiirGluTyrGluLysGlnMetAlaGluValGlnLysAlaSerPheGluAlaLys 107 421 TGCTCGGAAAATTTAGCTTCTGCAGAAGAATTCTTAAAAGAAAATAAAGAGAAGGCTGGG CysSerGluAsnLeuAlaSerAlaGlL JIU PheLeuLysGluAsnLysGluLysAlaGly 127 Lys 481 GTTATTGAGTTAGAGCCTAATAAGTTACAGTACCGTGTTGTGAAAGAGGGTACAGGAGGG VallleGluLeuGluProAsnLysLeuGlnTyrArgValValLysGluGlyThrGlyArg 147 541 GTTCTTTCTGGGAAGCCTACAGCTTTGCTTCACTATACAGGGAGCTTCATCGATGGGAAG ValLeuSerGlyLysProThrAlaLeuLeuHisTyrThrGlySerPhelleAspGlyLys 167 601 GTTTTTGATTCTTCAGAGAAGAATAAAGAGCCCATTTTACTGCCTTTGACCAAAGTAATT ValPheAspSerSerGluLysAsnLysGluProIleLeuLeuProLeuThrLysVallle 137 661 CCTGGATTTTCCCAAGGTATGCAAGGGATGAAAGAAGGAGAGGTTCGAGTTCTTTACATA ProGlyPheSerGlnGlyMetGInGlyMetLysGluGiyGluValArgValLeuTyrlle 2 07 721 CATCCAGATTTAGCTTACGGAACAGCTGGACAATTACCTCCAAACTCTCTACTCATTTTT HisProAspLeuAlaTyrGlyThrAlaGlyGlnLeuProProAsnSerLeuLeuIlePhe 227 7 81 GAAGTGAAGTTAATTGAAGCAAACGACGATAATGTATCTGTTACAGAATAGTCGTTAGGA GluValLysLeuIleGluAlaAsnAspAspAsnValSerValThrGlu * 243 841 TTTTTTATGCGCATCGTTCTTCACAATCCAGATATCCCTCAGAATACAGGGAATATTGGT
protein failed to demonstrate surface exposure of the Miplike protein on purified EBs or RBs by immunocytochemical procedures. However, a complement-dependent inhibition of infectivity for ceil cultures was demonstrable, suggesting that the Mip-like protein has antibody-accessible epitopes on infectious EBs.
Results Sequence analysis of the gene encoding the Mip-tike protein from C, trachomatis serovar L2 The sequence encoding the 175 amino acids of the C-terminal fragment of the m/p-like gene of C. trachomatis serovar L2 has been published (Lundemose etai, 1991). Cloning of the 68-amino-acid AAterminal fragment into high copy-number plasmid vectors was unsuccessful despite several attempts and the sequence was established only in the low copy-cosmid vector system pWE-
15. Primers corresponding to the flanking regions of the A/-terminal fragment were used to amplify the sequence followed by direct sequencing of the amplified product. The DNA sequence revealed one open reading frame of 729 bp encoding 243 amino acids (Fig. 1}, The 243 amino acids had a predicted molecular weight of 26648 and a predicted isoelectric point (pi) of pH 4.70, A potential ribosomal binding site AGGA was positioned eight bases upstream of the ATG codon. A possible W-terminal leader peptide consisting of 19 amino acids was predicted by computer analysis.The cleavage site resembles a signal peptidase II site suggesting that the Mip-like protein may be a lipoprotein. Hydrophobicity analysis (Engelman et ai, 1986) of the polypeptide, except for the signal sequence, indicates a relatively hydrophilic protein lacking strongly hydrophobic regions (stop-transfer sequences) that would allow the protein to be anchored in the inner membrane (data not shown).
Chlamydial Mip-like protein
L2 N
•730 bp
,390bp 340 bp
Fig. 2. A. The mip gene of C. tracfiomatis serovars B, D, E, G, H, I, and L2 was amplified and the products examined by gel electrophoresis. All serovars had distinct bands of 730 bp corresponding to the size of the m(p-like gene. Lambda phage digested with H/ndlll was used as size standard (marked S) negative control. B, Each amplified product was digested by restriction endonuclease Psft and the digested products examined by gei electrophoresis. All serovars digested with Pst\ showed distinct bands of 390 bp and 340 bp.
Anaiysis of the mip-//Ae genes from C. trachomatis serovars B. D. E. G. H, I The mip-like gene of C, trachomatis serovars B, D, E, G, H, I and L2 was amplified by polymerase chain reaction (PCR) and the products were examined by gel electrophoresis. All serovars examined had distinct bands of 730 bp corresponding to the size of the m/p-like gene of serovar L2 (Fig 2a) indicating its presence in both the trachoma and LGV biovars of C. trachomatis. No DNA products were observed when amplification was carried out on DNA from Chlamydia psittaci and Chlamydia pneumoniae (not shown). To assess the extent of conservation of m/plike genes in C. trachomatis, each amplified product was digested separately by restriction endonucleases BglW, Pst\ or Cla\ and the digested products were examined by gel electrophoresis. BglW, Pst\ and C/al were selected because ot their ability to cut the mip gene at only one site each. Bgl\ cuts towards the A/-terminal end (bp no, 203), Pst\ in the middle (bp no. 343) and C/al cuts towards the C-terminal end (bp no. 491). All serovars digested with Pst\ showed distinct bands of 390 bp and 340 bp (see Fig.
2541
2b). All serovars showed distinct bands of 490 bp and 240 bp when digested with C/al and of 200 bp and 530 bp when digested with BglW (not shown). The mip gene of serovars B and E was cloned and sequenced and very high DNA and amino acid similarity to serovar L2 was observed (Fig, 1), The DNA and amino acid identities were, respectively, 99.5 and 99.2% between serovars B and L2, 99,3 and 99,8% between serovars E and L2 and 99,5 and 100%, between serovars B and E. Between the protein sequences of B/E and L2 there were only two amino add sequence changes and these were conservative (Ser/Ala, Glu/Lys; see enclosed boxes in Fig. 1). The number of DNA sequence changes was slightly higher, at five, with three DNA base changes resulting in no amino acid changes. The data suggest that the Mip-like protein is well conserved among serovars of C. trachcmatis.
Sequence homology analysis of the Mip-like protein By computer search of the EMBL data base with the deduced amino acid sequence of the chlamydial Mip-like protein several proteins showed significant homology (Fig. 3). The chlamydial Mip-like protein had 35% amino acid identity to the mip gene product of L. pneumophila (Engleberg et ai, 1989) and the 110-amino-acid C-terminal region showed 4 1 % identity to human FK506-binding protein (FKBP) (Maki et ai, 1990) and 41,3% identity to the FKBP from Neisseria meningitidis (Sampson and Gotschlich, 1992), From the alignment (Fig, 3) it can be seen that the Mip sequences contain three regions: a leader peptide of 19 amino acids predicted by computer analysis, an W-terminal region of 114 amino acids extending from the leader region to the start of the FKBP sequences and a C-terminal region corresponding to the sequence homology with the FKBPs. Of the six residues making up the hydrophobic FK506'binding pocket of the human FKBP (Van Dyune et ai, 1991) four are identical to the chlamydial Mip sequence and the remaining two have conservative replacements (Fig.3). Five residues are involved in the FKBP-FK506 hydrogen-bond formation of the human FKBP (Van Dyune et ai. 1991), of which three are identical in the chlamydial Mip-like protein, one has a conserved replacement, and one is not conserved (Fig.3),
Polyclonal monospecific antibodies and monoclonal antibodies against the chlamydial Mip-like protein Antibodies raised to the Mip-like protein are shown in Table 1. Monoclonal and polyclonal monospecifio antibodies were produced against recombinant Mip-like protein and against fusion proteins covering the N- and C-terminal fragments of the Mip-like protein. For production of
2542
A. G. Lundemose et al.
•JGGGSQTSATEKSMVEE 5 T AiMAA1gp,T
MIPLp
KL
50 37
FKP3H FGHI
MIPLp
^QLSRTEDFSLSLVEVIKGMQ . KNQGIDVSPEAMJ KGMQ )AM
KQUtL
TE QQMKDVLNKFC K S
35 88
FKPBH MIPL2 MIPLp FKPBNm FKPBH
^ 147 / 139 ^ 15 GVQVETISECtt ^ 13
MIPL2 MrPLp FKPBNm FKPBH
polyclonal antibodies SDS-polyacrylamide gel electrophoresis {PAGE)-separated recombinant proteins were used as immunogens. Polyclonal monospecific antibodies (pAbs) 83 and 82 were raised against recombinant Mip overproduced in E, coli under control of a T7 promotor. pAbs 76 and 77 were raised against the C-terminal fragment of the Mip-like polypeptide using pCtX3-11 as immunogen, and pAbs 80 and 81 were raised against the A/-terminal fragment using pCtX3-10 as immunogen. pAbs 90 and 91 were raised against fusion protein (cro-(igalactosidase) lacking chlamydial protein elements in order to create a negative control serum. Despite four attempts it was not possible to raise specific antibodies against pCtX3-11a comprising the central fragment of the Mip-tike protein. However, pAb 83 reacted weakly with
Table 1. Characterization of antibodies against chlamydial Mip-like protein.
Fig. 3. Amino acid homology between the L2 chlamydial Mip-like protein and related proteins. The chtamydial Mip-like proiem has 35% ammo acid identity to the mip gene product of L. pneumophila. Black line indicates the leader peptide predicted by computer analysis; shaded pan represents the W terminal region of the Miplike protein from the leader peptide to the start of the homology with FKBP sequences; rest of the Mip-ltke protein corresponds to the C-lerminal region with homoiogy to FKBPs (41.3% identity to FK506 binding protein of N. meningitidis (FKBPNm) and 38% amino acid identity to human FK506 binding protein (FKBPH)), The six residues making up the FKBP-FK506 hydrophobic-binding pocket are marked A , The five residues involved in hydrogen bonding to FK506 are marked • An open reading frame from the P. aeruginosa genome encoding a Miplike protein has been reported (Kato etai. 1989), but IS not included here because it has not been reported to be expressed.
the fusion protein of pCtX3-11a in immunoblotting (not shown). Monoclonal antibodies (mAbs) 112.3 and 28.2 specific for the chlamydial Mip-like protein have previously been characterized (Lundemose ef ai. 1991), An additional five mAbs, 132, 139, 147, AL2 and AL54, all specific for the Mip-like protein, were raised using recombinant pCOSCt 6-3 or pCtX3-10 as immunogen. mAbs 112.3, 28.2, 132, 139 and 147 reacted with the C-terminal fragment of Mip (amino acid no. 115 to amino acid no, 243), whereas mAbs AL2 and AL54 reacted with the Nterminal fragment of Mip-like protein (amino acid no, 20 to amino acid no, 67), The specificity of the antibodies for SDS-PAGE-separated chlamydial antigens is shown in Fig, 4, Each Mip antibody reacted with the 27 kDa Miplike protein. The negative control serum pAb 90 did not
Antibody
Isotype
Antigen
pAb83 0Ab76 DAb80 pAb90 mAb 112,3 mAb 132 mAb 139 mAb 147 mAb AL2 mAbAL54 mAb 28,2
Polyclonal Polyclonal Polyclonal Polyclonal IgGI IgGI lgG2a IgGI IgGI IgGI lgG2a
pETMipiia pCtX3-11 pCtX3-10 pEX EBs pCosCt 6-3 pCosCt6 3 pCosCt6-3 pCtX3-10 pCtX3-10 EBs
Mip-like proteir"*
N-lermmus"
C-terminus"
+
+ ND ND ND +
+ ND ND ND
-
+
i -( ^ H 4 i i
a. *'- indicates reaction/no reaction wilh Mip-like protein determined by immunoblotting, b. +/-indicates reaction/no reaction with pCtX3-10 determined by immunoblotting. c. -•-/- indicates reaction/no reaction with pCtX3-11 determined by immunoblotting, ND Not determined due to reaction with cro-p-galactosidase.
• •••
4-
+ -
Chlamydial Mip-like protein 2
3
4
5
6
7
8
9
10
t
Chlamydia trachomatis Mip-like protein Anker G. Lundemose,'"'^ Duncan A. Rouch,^ Svend Birkelund/ Gunna Christiansen^ and John H. Pearce^ institute of Medical Microbiology. Bartholin Building. University of Aarhus. DK-8000 Aarhus C. Denmark. ^Microbial Molecular Genetics and Cell Biology Group. University of Birmingham. Birmingham. UK.
Summary A 27 kDa Chlamydia trachomatis Mip-like protein with homology of a 175-amino-acid C-terminal fragment to the surface-ex posed LegioneUa pneumophila mipgene product has previously been described. In this paper the entire chlamydia Mip-like sequence of C. trachomatis serovar L2 (lymphogranuloma venereum (LGV) biovar) is presented. The sequence shows high similarity to the legionella Mip protein and its C-terminal region, like that of the legionella Mip, has high amino acid similarity to eukaryotic and prokaryotic FK506-binding proteins. The chlamydial mip-Wke gene was detected by polymerase chain reaction (PCR) in other C. trachomatis serovars and by sequencing of the mip-Wke genes of serovars B and E (trachoma biovar) was shown to be highly conserved within the two major biovars of C. trachomatis. Monocionai and polyclonal antibodies raised against the recombinant Mip-like protein failed to demonstrate surface-exposed epitopes on infectious elementary bodies or reproductive reticulate body forms either by immunofluorescence or immuno-gold electron microscopy. iHowever, a complement-dependent inhibition of up to 91% of infectivity for cell cultures was observed with antibodies to the N-terminal fragment of the Mip-like protein suggesting that antibodyaccessible epjtopes are present on infectious EBs.
Introduction The Chlamydia are obligate intracellular bacteria with a remarkable biphasic lifecycle. Following entry into host cells intracellular survival is believed to depend on avoidance of the lysosomal compartment and the infectious
Received 25 March, 1992: revised and accepted 27 May, 1992, •For correspondence, Tel, 86139711; Fax 86196128.
elementary body (EB) undergoes transformation to a noninfectious metabolicaily active reticulate body (RB) form that divides by binary fission. The life cycle is completed with differentiation of RBs to EBs and release of the latter for initiation of further cycles of infection. The species Chlamydia trachomatis is an important pathogen of man, with members of the trachoma biovar being responsible for the blinding eye disease, trachoma, and for sexually transmitted diseases; strains of the lymphogranuloma venereum (LGV) biovar are also sexually transmitted but are less common. Recent studies on initiation of infection implicate the major outer membrane protein both as a source of antigenic diversity and, via variable and conserved moieties, as an adhesin {Su et ai, 1990). Other proteins may also be involved in attachment (Joseph and Bose, 1991; Schmiel et ai. 1991), Entry into host cells appears to occur by both phagocytosis and pinocytosis {Ward and Murray, 1984; Hodinka et ai. 1986; Reynolds and Pearce, 1990; 1991), However, not all endocytosed organisms go on to multiply and events during entry appear to be pivotal for productive infection. The molecular determinants of the invasion stage of the cycle are unknown and. as for invasive enteric bacteria, are of intense interest, both biologically and for their role as possible vaccine antigens. We have previously described a 27 kOa protein of C. tracfiomatis serovar L2 (LGV biovar) and the 3'-end of the encoding gene (Lundemose et ai. 1991). The deduced amino acid sequence ot a 175-amino-acid C-terminal fragment showed high similarity to a LegioneUa pneumophila mip {macrophage infectivity potentiator)-gene product (Engleberg et ai. 1989) that is a virulence factor (Cianciatto et ai. 1990) known to play an important role in initiating the intracellular infection of L. pneumophila in human macrophages (Cianciotto etal. 1989) In this paper we present the cloning and analysis of the entire coding sequence of the Mip-like protein of the L2 serovar and investigation of its possible surface exposure. The data show similarity both with the Legionella Mip protein and with the eukaryotic and prokaryotic FK506-binding proteins (FKBPs). The chlamydial m/p-like gene is shown to be present in other C. trachomatis serovars and by sequence of the m/p-like genes of C, trachomatis serovars B and E (trachoma biovar) to be highly conserved within both biovars. Monoclonal and polyclonal antibodies against the recombinant Mip-like
2540
A. G. Lundemose eXa\.
1 TTTTTTACAGCATGTGATAAGGGTGGATGTCTTTCACCTAGGAAACGGTTTTCAAAAAAG RBS 61 CACATCTGAGAAAATCAGGTGCAACTGAGGAAGAAACGAATGAAGAATATATTAAGTTGG MetLysAsnIleLeuSerTrp
7
121 ATGCTTATGTTTGCAGTCGCTCTGCCTATCGTAGGATGTGATAACGGRGGCGGTTCGCAA MetLeuMetPheAlaValAlaLeuProIleValGlyCysAspAsnGlyGlyGlySerGln
27
181 ACATCGGCTACGGAGAAAAGCATGGTAGAAGACTCTGCATTGACAGACAATCAAAAGTTA ThrSerAlaThrGluLysSerMetValGluAspSerAlaLeuThrAspAsnGlnLysLea
47
2 41 TCAAGAACTTTTGGGCATTTATTGGCTCGTCAGTTGAGCCGAACGGAAGATTTTTCGTTA SerArgThrPheGlyHisLeuLei: M a irgGlnLeuSerArgThrGluAspPheSerLeu 67 Ser 301 GATCTTGTTGAAGTGATTAAAGGGATGCAATCTGAAATAGATGGACAGAGTGCTCCTTTA AspLeuValGluVallleLysGlyMetGlnSerGluIleAspGlyGlnSerAlaProLeu 97
Fig. 1. Sequence of Ihe C. trachomatis serowai L2 mip-like gene. The postulated ribosomal binding site is labelled RBS, The large arrow-head indicates the possible location of the cleavage site between the leader peptide and the mature protein; stem loop is indicated with arrows. Predicted molecular weight ot the polypeptide was 26 648 and moLwt of the mature protein was 24 531. The two boxes enclose the amino acid sequence differences between L2 and B/E sequences. In the sequence previously published (Lundemose etal.. 1991) bases no, 307-308 were cited TG, but should be GT, which is corrected in this sequence. The sequence is assigned to the EMBL database under accession numbers X66126(L2), X56127(E) and X66128(B),
3 61 ACAGATACAGAATATGAAAAACAAATGGCAGAAGTACAAAAAGCTAGTTTCGAAGCAAAA ThrAspTiirGluTyrGluLysGlnMetAlaGluValGlnLysAlaSerPheGluAlaLys 107 421 TGCTCGGAAAATTTAGCTTCTGCAGAAGAATTCTTAAAAGAAAATAAAGAGAAGGCTGGG CysSerGluAsnLeuAlaSerAlaGlL JIU PheLeuLysGluAsnLysGluLysAlaGly 127 Lys 481 GTTATTGAGTTAGAGCCTAATAAGTTACAGTACCGTGTTGTGAAAGAGGGTACAGGAGGG VallleGluLeuGluProAsnLysLeuGlnTyrArgValValLysGluGlyThrGlyArg 147 541 GTTCTTTCTGGGAAGCCTACAGCTTTGCTTCACTATACAGGGAGCTTCATCGATGGGAAG ValLeuSerGlyLysProThrAlaLeuLeuHisTyrThrGlySerPhelleAspGlyLys 167 601 GTTTTTGATTCTTCAGAGAAGAATAAAGAGCCCATTTTACTGCCTTTGACCAAAGTAATT ValPheAspSerSerGluLysAsnLysGluProIleLeuLeuProLeuThrLysVallle 137 661 CCTGGATTTTCCCAAGGTATGCAAGGGATGAAAGAAGGAGAGGTTCGAGTTCTTTACATA ProGlyPheSerGlnGlyMetGInGlyMetLysGluGiyGluValArgValLeuTyrlle 2 07 721 CATCCAGATTTAGCTTACGGAACAGCTGGACAATTACCTCCAAACTCTCTACTCATTTTT HisProAspLeuAlaTyrGlyThrAlaGlyGlnLeuProProAsnSerLeuLeuIlePhe 227 7 81 GAAGTGAAGTTAATTGAAGCAAACGACGATAATGTATCTGTTACAGAATAGTCGTTAGGA GluValLysLeuIleGluAlaAsnAspAspAsnValSerValThrGlu * 243 841 TTTTTTATGCGCATCGTTCTTCACAATCCAGATATCCCTCAGAATACAGGGAATATTGGT
protein failed to demonstrate surface exposure of the Miplike protein on purified EBs or RBs by immunocytochemical procedures. However, a complement-dependent inhibition of infectivity for ceil cultures was demonstrable, suggesting that the Mip-like protein has antibody-accessible epitopes on infectious EBs.
Results Sequence analysis of the gene encoding the Mip-tike protein from C, trachomatis serovar L2 The sequence encoding the 175 amino acids of the C-terminal fragment of the m/p-like gene of C. trachomatis serovar L2 has been published (Lundemose etai, 1991). Cloning of the 68-amino-acid AAterminal fragment into high copy-number plasmid vectors was unsuccessful despite several attempts and the sequence was established only in the low copy-cosmid vector system pWE-
15. Primers corresponding to the flanking regions of the A/-terminal fragment were used to amplify the sequence followed by direct sequencing of the amplified product. The DNA sequence revealed one open reading frame of 729 bp encoding 243 amino acids (Fig. 1}, The 243 amino acids had a predicted molecular weight of 26648 and a predicted isoelectric point (pi) of pH 4.70, A potential ribosomal binding site AGGA was positioned eight bases upstream of the ATG codon. A possible W-terminal leader peptide consisting of 19 amino acids was predicted by computer analysis.The cleavage site resembles a signal peptidase II site suggesting that the Mip-like protein may be a lipoprotein. Hydrophobicity analysis (Engelman et ai, 1986) of the polypeptide, except for the signal sequence, indicates a relatively hydrophilic protein lacking strongly hydrophobic regions (stop-transfer sequences) that would allow the protein to be anchored in the inner membrane (data not shown).
Chlamydial Mip-like protein
L2 N
•730 bp
,390bp 340 bp
Fig. 2. A. The mip gene of C. tracfiomatis serovars B, D, E, G, H, I, and L2 was amplified and the products examined by gel electrophoresis. All serovars had distinct bands of 730 bp corresponding to the size of the m(p-like gene. Lambda phage digested with H/ndlll was used as size standard (marked S) negative control. B, Each amplified product was digested by restriction endonuclease Psft and the digested products examined by gei electrophoresis. All serovars digested with Pst\ showed distinct bands of 390 bp and 340 bp.
Anaiysis of the mip-//Ae genes from C. trachomatis serovars B. D. E. G. H, I The mip-like gene of C, trachomatis serovars B, D, E, G, H, I and L2 was amplified by polymerase chain reaction (PCR) and the products were examined by gel electrophoresis. All serovars examined had distinct bands of 730 bp corresponding to the size of the m/p-like gene of serovar L2 (Fig 2a) indicating its presence in both the trachoma and LGV biovars of C. trachomatis. No DNA products were observed when amplification was carried out on DNA from Chlamydia psittaci and Chlamydia pneumoniae (not shown). To assess the extent of conservation of m/plike genes in C. trachomatis, each amplified product was digested separately by restriction endonucleases BglW, Pst\ or Cla\ and the digested products were examined by gel electrophoresis. BglW, Pst\ and C/al were selected because ot their ability to cut the mip gene at only one site each. Bgl\ cuts towards the A/-terminal end (bp no, 203), Pst\ in the middle (bp no. 343) and C/al cuts towards the C-terminal end (bp no. 491). All serovars digested with Pst\ showed distinct bands of 390 bp and 340 bp (see Fig.
2541
2b). All serovars showed distinct bands of 490 bp and 240 bp when digested with C/al and of 200 bp and 530 bp when digested with BglW (not shown). The mip gene of serovars B and E was cloned and sequenced and very high DNA and amino acid similarity to serovar L2 was observed (Fig, 1), The DNA and amino acid identities were, respectively, 99.5 and 99.2% between serovars B and L2, 99,3 and 99,8% between serovars E and L2 and 99,5 and 100%, between serovars B and E. Between the protein sequences of B/E and L2 there were only two amino add sequence changes and these were conservative (Ser/Ala, Glu/Lys; see enclosed boxes in Fig. 1). The number of DNA sequence changes was slightly higher, at five, with three DNA base changes resulting in no amino acid changes. The data suggest that the Mip-like protein is well conserved among serovars of C. trachcmatis.
Sequence homology analysis of the Mip-like protein By computer search of the EMBL data base with the deduced amino acid sequence of the chlamydial Mip-like protein several proteins showed significant homology (Fig. 3). The chlamydial Mip-like protein had 35% amino acid identity to the mip gene product of L. pneumophila (Engleberg et ai, 1989) and the 110-amino-acid C-terminal region showed 4 1 % identity to human FK506-binding protein (FKBP) (Maki et ai, 1990) and 41,3% identity to the FKBP from Neisseria meningitidis (Sampson and Gotschlich, 1992), From the alignment (Fig, 3) it can be seen that the Mip sequences contain three regions: a leader peptide of 19 amino acids predicted by computer analysis, an W-terminal region of 114 amino acids extending from the leader region to the start of the FKBP sequences and a C-terminal region corresponding to the sequence homology with the FKBPs. Of the six residues making up the hydrophobic FK506'binding pocket of the human FKBP (Van Dyune et ai, 1991) four are identical to the chlamydial Mip sequence and the remaining two have conservative replacements (Fig.3). Five residues are involved in the FKBP-FK506 hydrogen-bond formation of the human FKBP (Van Dyune et ai. 1991), of which three are identical in the chlamydial Mip-like protein, one has a conserved replacement, and one is not conserved (Fig.3),
Polyclonal monospecific antibodies and monoclonal antibodies against the chlamydial Mip-like protein Antibodies raised to the Mip-like protein are shown in Table 1. Monoclonal and polyclonal monospecifio antibodies were produced against recombinant Mip-like protein and against fusion proteins covering the N- and C-terminal fragments of the Mip-like protein. For production of
2542
A. G. Lundemose et al.
•JGGGSQTSATEKSMVEE 5 T AiMAA1gp,T
MIPLp
KL
50 37
FKP3H FGHI
MIPLp
^QLSRTEDFSLSLVEVIKGMQ . KNQGIDVSPEAMJ KGMQ )AM
KQUtL
TE QQMKDVLNKFC K S
35 88
FKPBH MIPL2 MIPLp FKPBNm FKPBH
^ 147 / 139 ^ 15 GVQVETISECtt ^ 13
MIPL2 MrPLp FKPBNm FKPBH
polyclonal antibodies SDS-polyacrylamide gel electrophoresis {PAGE)-separated recombinant proteins were used as immunogens. Polyclonal monospecific antibodies (pAbs) 83 and 82 were raised against recombinant Mip overproduced in E, coli under control of a T7 promotor. pAbs 76 and 77 were raised against the C-terminal fragment of the Mip-like polypeptide using pCtX3-11 as immunogen, and pAbs 80 and 81 were raised against the A/-terminal fragment using pCtX3-10 as immunogen. pAbs 90 and 91 were raised against fusion protein (cro-(igalactosidase) lacking chlamydial protein elements in order to create a negative control serum. Despite four attempts it was not possible to raise specific antibodies against pCtX3-11a comprising the central fragment of the Mip-tike protein. However, pAb 83 reacted weakly with
Table 1. Characterization of antibodies against chlamydial Mip-like protein.
Fig. 3. Amino acid homology between the L2 chlamydial Mip-like protein and related proteins. The chtamydial Mip-like proiem has 35% ammo acid identity to the mip gene product of L. pneumophila. Black line indicates the leader peptide predicted by computer analysis; shaded pan represents the W terminal region of the Miplike protein from the leader peptide to the start of the homology with FKBP sequences; rest of the Mip-ltke protein corresponds to the C-lerminal region with homoiogy to FKBPs (41.3% identity to FK506 binding protein of N. meningitidis (FKBPNm) and 38% amino acid identity to human FK506 binding protein (FKBPH)), The six residues making up the FKBP-FK506 hydrophobic-binding pocket are marked A , The five residues involved in hydrogen bonding to FK506 are marked • An open reading frame from the P. aeruginosa genome encoding a Miplike protein has been reported (Kato etai. 1989), but IS not included here because it has not been reported to be expressed.
the fusion protein of pCtX3-11a in immunoblotting (not shown). Monoclonal antibodies (mAbs) 112.3 and 28.2 specific for the chlamydial Mip-like protein have previously been characterized (Lundemose ef ai. 1991), An additional five mAbs, 132, 139, 147, AL2 and AL54, all specific for the Mip-like protein, were raised using recombinant pCOSCt 6-3 or pCtX3-10 as immunogen. mAbs 112.3, 28.2, 132, 139 and 147 reacted with the C-terminal fragment of Mip (amino acid no. 115 to amino acid no, 243), whereas mAbs AL2 and AL54 reacted with the Nterminal fragment of Mip-like protein (amino acid no, 20 to amino acid no, 67), The specificity of the antibodies for SDS-PAGE-separated chlamydial antigens is shown in Fig, 4, Each Mip antibody reacted with the 27 kDa Miplike protein. The negative control serum pAb 90 did not
Antibody
Isotype
Antigen
pAb83 0Ab76 DAb80 pAb90 mAb 112,3 mAb 132 mAb 139 mAb 147 mAb AL2 mAbAL54 mAb 28,2
Polyclonal Polyclonal Polyclonal Polyclonal IgGI IgGI lgG2a IgGI IgGI IgGI lgG2a
pETMipiia pCtX3-11 pCtX3-10 pEX EBs pCosCt 6-3 pCosCt6 3 pCosCt6-3 pCtX3-10 pCtX3-10 EBs
Mip-like proteir"*
N-lermmus"
C-terminus"
+
+ ND ND ND +
+ ND ND ND
-
+
i -( ^ H 4 i i
a. *'- indicates reaction/no reaction wilh Mip-like protein determined by immunoblotting, b. +/-indicates reaction/no reaction with pCtX3-10 determined by immunoblotting. c. -•-/- indicates reaction/no reaction with pCtX3-11 determined by immunoblotting, ND Not determined due to reaction with cro-p-galactosidase.
• •••
4-
+ -
Chlamydial Mip-like protein 2
3
4
5
6
7
8
9
10
t
