Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M401316200 on March 17, 2004

J. Biol. Chem., Vol. 279, Issue 21, 22469-22476, May 21, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/21/22469    most recent
M401316200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Korres, H.
Right arrow Articles by Verma, N. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Korres, H.
Right arrow Articles by Verma, N. K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Topological Analysis of Glucosyltransferase GtrV of Shigella flexneri by a Dual Reporter System and Identification of a Unique Reentrant Loop*

Haralambos Korres and Naresh K. Verma{ddagger}

From the School of Biochemistry and Molecular Biology, Faculty of Science, The Australian National University, Canberra ACT 0200, Australia

Received for publication, February 6, 2004 , and in revised form, March 15, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Lipopolysaccharide, particularly the O-antigen component, is one of many virulence determinants necessary for Shigella flexneri pathogenesis. O-Antigen modification is mediated by glucosyltransferase genes (gtr) encoded by temperate serotype-converting bacteriophages. The gtrV gene encodes the GtrV glucosyltransferase, an integral membrane protein that catalyzes the transfer of a glucosyl residue via an {alpha}1,3 linkage to rhamnose II of the O-antigen unit. This mediates conversion of S. flexneri serotype Y to serotype 5a. Analysis of the GtrV amino acid sequence using computer prediction programs indicated that GtrV had 9–11 transmembrane segments. The computer prediction models were tested by genetically fusing C-terminal deletions of GtrV to a dual reporter system composed of alkaline phosphatase and {beta}-galactosidase. Sandwiched GtrV-PhoA/LacZ fusions were also constructed at predetermined positions. The enzyme activities of cells with the GtrV-PhoA/LacZ fusions and the particular location of the fusions in the gtrV indicated that GtrV has nine transmembrane segments and one large N-terminal periplasmic loop with the N and C termini located on the cytoplasmic and periplasmic sides of the membrane, respectively. The existence of a unique reentrant loop was discovered after transmembrane segment IV, a feature not documented in other bacterial glycosyltransferases. Its potential role in mediating serotype conversion in S. flexneri is discussed.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Bacillary dysentery or shigellosis is caused by the Shigella bacterium, which is of major concern in overcrowded areas of the developing world (1, 2). A major surface component of Shigella flexneri that contributes to pathogenesis is the lipopolysaccharide (3–7).

The lipopolysaccharide consists of three segments, lipid A, core polysaccharide, and the O-antigen (8). The O-antigen chain is composed of repeating sugar subunits that vary in their composition and so contribute to serotype diversity. For example, the different serotypes of S. flexneri arise from the addition of glucosyl or O-acetyl residues to the repeating sugar subunits. This specific attachment is thought to be mediated by glucosyl- and O-acetyltransferases (9–12). The glucosyltransferase genes gtrV, gtrI, gtrIV, and gtrX of S. flexneri serotypes 5a, 1a, 4a, and X, respectively, have been isolated and characterized (10, 13, 14). The only glucosyltransferase genes that exhibit some degree of homology at the amino acid and nucleotide level are gtrV and gtrX (35.7% identity and 61.3% similarity). The exact model of GtrV operation is not known, although two other proteins, GtrA and GtrB, are thought to interact with GtrV mediating in serotype conversion (13–15).

This paper deals with a glucosyltransferase, GtrV, that catalyzes the transfer of a glucosyl residue via an {alpha}1,3 linkage to rhamnose II of the O-antigen unit (10). According to hydropathy data, distribution of charged residues and possible turns, this protein consists of 9–11 hydrophobic segments, with the N terminus in the cytoplasm and the C terminus in the periplasm. The validity of the proposed model is examined by a genetic approach, which involves the construction of fusion proteins between reporter enzymes and GtrV. This study utilizes a newly developed dual reporter system consisting of alkaline phosphatase (phoA) which is in-frame with the {beta}-galactosidase {alpha} fragment (lacZ{alpha}) (16, 17). The signal sequence of PhoA1 can be replaced by export signals derived from other proteins such as GtrV, producing a chimeric protein that can be used to report cellular location according to the location of the protein domain under investigation (16–18). Alkaline phosphatase (AP) is active only when located in the periplasm. This is based on the assumption that in the periplasm the mature part of PhoA is oxidized. The cysteine residues form disulfide bridges that enable the correct folding of PhoA. The enzyme becomes active after dimer formation is complete. The process of folding, and assembly of PhoA occurs only after export to the periplasm because various factors prevent the formation of disulfide bonds in the cytoplasm (16–25). In contrast to AP, {beta}-galactosidase (BG) is active in the cytoplasm. The enzyme is inactive in the periplasmic space since its proper folding is prevented by becoming trapped in the membrane crossing to the periplasm. Also, {alpha}-complementation should occur only when the enzyme (lacZ{alpha}) is accessible to cytoplasmic located {omega}-fragment (16–19, 21, 24, 25). Thus fusions to periplasmic sites will be inactive whereas fusions to cytoplasmic domains will be active.

Alexeyev et al. (16) report that by using this novel approach of PhoA/LacZ dual reporters, BG and AP activities can be measured at the same point simultaneously without resorting to genetic recombination to switch fusions (17). The normalized activity ratios (NAR) can correct for variable expression. Thus, determination of protein synthesis rates is not necessary. The system provides easily interpretable information about subcellular localization of the reporter (16, 17). Furthermore, {alpha} complementation is based upon the availability of {omega}-fragment, which is solely confined to the cytoplasm and not the periplasm, rendering the enzyme inactive if located in the periplasm. In this way the formation of toxic aggregates in the periplasm is eliminated. Finally, the use of dual indicator plates in conjunction with these reporters discriminates between non-informative fusions (white), cytoplasmic fusions (red), and periplasmic fusions (blue or purple) (16, 17).

In this paper we report a topology model of GtrV based on analysis of data from computer prediction models and fusion constructs consisting of gtrV-phoA/lacZ and gtrV-phoA/lacZ-gtrV sandwich fusions. The model suggests that GtrV consists of nine transmembrane segments and a large N-terminal periplasmic hydrophilic loop. The N terminus is located in the cytoplasm whereas the C terminus is located in the periplasm. The model also reveals that GtrV contains a unique reentrant loop after transmembrane segment IV. To the best of our knowledge, this is the first report showing the presence of a reentrant loop in bacterial glycosyl transferases.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Bacterial Strains and Growth Conditions—All bacterial strains used in this study are derivatives of Escherichia coli K-12. Their particular genotypes are described in Table I. The bacterial strains were grown aerobically at 37 °C in Luria broth (LB). LB agar plates were prepared as described previously (26). Ampicillin and chloramphenicol (Cm) were added to liquid and solid media at 100 and 30 µg/ml, respectively. The PhoA and LacZ colony phenotypes were identified on agar plates containing 1.5% Bacto-agar, 1% Bacto-tryptone, 0.5% yeast extract, 0.5% NaCl, 5-bromo-4-chloro-3-inolyl phosphate disodium salt (Sigma; X-phos 80 µg/ml), 6-chloro-3-indolyl-{beta}-D-galactoside (Research Organics Red-Gal; 100 µg/ml), 1 mM isopropyl-1-thio-{beta}-D-galactopyranoside, 80 mM K2HPO4 (pH 7.0), and 30 µg/ml Cm.


View this table:
[in this window]
[in a new window]
  		TABLE I
Bacterial strains and plasmids used in this study Plasmids carrying gtrV-phoA/lacZ fusions and gtrV-phoA/lacZ-gtrV sandwich fusions are shown in Table III. kb, kilobase(s).

 
Plasmids, Primers, PCR, and Sequencing—Plasmids used in this study are listed in Table I. Oligonucleotide primers used for PCR are listed in Table II. The primers were synthesized by Invitrogen. Double-stranded plasmid sequencing was performed at the Biomolecular Resources Facility, John Curtin School of Medical Research, Australian National University. Sequencing primer PHOSEQ specific for the 5' end of the pho-lac fusion (5'-TCACCCGTTAAACGGCGAGCACC-3') was used to determine the exact point of fusion (2). PCR was carried out using Pfu polymerase (Stratagene) as specified by the manufacturer.


View this table:
[in this window]
[in a new window]
  		TABLE II
Oligonucleotide primers used in this study to create sandwich fusions and amplification of gtrV fragment from pNV323

 
DNA Methodology—Restriction endonucleases and T4 DNA ligase were purchased from Amersham Biosciences and Promega, respectively, and used in accordance with the protocols supplied by manufacturers. Plasmids were maintained in strain JM109 and prepared using the Qiagen MiniPrep kit. Transformation of E. coli with plasmid DNA or ligation mixtures was performed using RbCl2 protocols (66).

Sequence Analysis—Eight computer programs available on the Internet were used to examine the GtrV protein sequence for the presence of hydrophobic regions. Programs used were DAS (27) (www.sbc.su.se/~miklos/DAS), HMMTOP (28) (enzim.hu/hmmtop), PHDhtm/PHDtopology (29, 30) (www.emblheidelberg.de/predictprotein), PSORT (31) (psort.nibb.ac.jp), SOSUI (32) (sosui.proteome.bio.tuat.ac.jp/sosuiframe0.html), TMHMM (33) (www.cbs.dtu.dk/services/TMHMM1.0), TMpred (34) (www.isrec.isbsib.ch/software/TMPRED_form.html), TopPred2 (35, 36) (www.sbc.su.se/~erikw/toppred2). Five of these programs (HMMTOP, PHDtopology, TMHMM, TMpred, and TopPred2) were used to predict the membrane topology of GtrV (37). Also hydrophobicity profiles were generated using the method of Kyte and Doolittle, implementing a sliding window of 19 residues.

Cloning of gtrV and Preparation for Reporter Gene Fusions—The wild type gtrV gene was amplified from plasmid pNV323 using the primers gtrVFSacI and gtrVRXbaI (Table II). Because these primers have the SacI and XbaI sites incorporated in them at the 5' end, it was possible to ligate the amplified fragment into pBCSK using the same sites that gave rise to pNV1077. The nested deletion method requires the presence of two unique restriction enzymes such as PstI and BamHI, allowing protection of the phoA/lacZ dual reporter and initiation of gtrV deletion by ExoIII, respectively. To use PstI as one of these enzymes, the PstI site in gtrV (1230 bp) had to be mutated so that only one PstI would be left between gtrV and the dual reporter. This was done by site-directed mutagenesis (Stratagene) changing the PstI site to a SphI site using primers gtrV1230FSphI and gtV1230RSphI (Table II) according to the manufacturer's protocol. Functionality of the gtrV gene was checked by introducing the new plasmid pNV1081 into SFL1444. In brief, plasmid pNV1081 was introduced by electroporation into SFL1444, which is a derivative of SFL124 except that it carries pNV1060 (gtrA and gtrB). Because the whole three-gene cluster has to be present for complete serotype conversion, this allows functional examination of gtrV with gtrA and gtrB, which are carried on another plasmid. Transformants were plated onto dual LB agar plates containing Cm (30 µg/ml) and ampicillin (100 µg/ml). Serotype conversion was tested by slide agglutination. This assay detects the expression of various S. flexneri epitopes. A glass slide was divided into two sections, and one drop of the test bacteria resuspended in saline (0.9% NaCl) was placed at each end of the slide. On the right section of the slide, one drop of S. flexneri serotype V (SEIKEN) antiserum was added above the bacterial suspension. On the left side of the slide one drop of saline was added in place of the specific antiserum. This served as a negative control. Using a sterile loop both antigen and serum or saline drops were mixed. The glass slide was rocked, and the mixtures were observed for agglutination. Only agglutination that occurred within 1 min was taken as positive. The functional gtrV carried by pNV1081 was cut with EcoRV to permit ligation of the dual indicator phoA/lacZ, which was isolated from pMA632 using the EcoRV and SmaI sites. This gave rise to pNV1090 (Fig. 1).



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 1.
Structure of the vector used to construct gtrV-phoA/lacZ fusions. pNV1090 is a derivative of pNV1081 that contains both the gtrV gene and the phoA/lacZ dual reporter separated by unique PstI and BamHI restriction endonuclease sites. Cleavage of pNV1090 with BamHI produced substrates for exonuclease III digestion. Cmr, chloramphenicol resistance. kb, kilobases.

 
Construction of gtrV-phoA/lacZ—The nested deletion method described by Sugiyama et al. (20) was used to produce a series of fusion plasmids in which various lengths of the 5' end of the gtrV gene were attached to the phoA/lacZ dual reporter (20). pNV1090 was linearized by cleavage of the unique PstI and BamHI sites located between the genes. This produced a 5' overhang just downstream of the gtrV gene and a 3' overhang positioned upstream of the phoA/lacZ gene. Exonuclease was used to progressively delete gtrV from its 3' end via the 5' overhang according to the method of Henikoff (38) as outlined in the instructions provided with the Promega kit. The reporter gene was protected from digestion since 3' overhangs are resistant to ExoIII digestion. After treatment with Klenow fragment and the four deoxynucleotide triphosphates, the ends were ligated. Circularized DNA carrying truncated versions of gtrV plus the full dual reporter was transformed into JM109 and plated onto dual indicator plates. The plasmid DNA was isolated from blue, purple, and red colonies (16, 17), and the exact site of the fusion point was determined by restriction digests and confirmed by DNA sequencing carried out on the ABI 3730 capillary sequence analyzer using the Big Dye Version 3.1 sequencing protocol and the PHOSEQ primer (2).

Construction of gtrV-phoA/lacZ-gtrV Sandwich Gene Fusions— Unique NruI restriction endonuclease sites were introduced throughout the gtrV gene present in pNV1077 by oligonucleotide-mediated site-directed mutagenesis (Stratagene) as specified in the manufacturer's protocol. Primers with an NruI site incorporated are listed in Table II. The presence of the desired mutation was checked by NruI digests and sequenced using gtrVFSacI and gtrV1780FSeq primers. The phoA/lacZ dual reporter was excised separately from pMA632 (2) using a combination of EcoRV-SmaI and StuI-NruI double digests. After ligation of the reporter insert to the desired construct carrying the NruI restriction site, the ligation mix was transformed in JM109 and then plated onto dual indicator plates. Colored colonies were picked and checked for inserts using digests and double-stranded sequencing to confirm the correct orientation, site of fusion, and maintenance of the correct reading frame.

Assays of AP and BG Activities—Overnight cultures of E. coli JM109 bearing fusion constructs and unfused plasmid (background control) were diluted 1:20 in fresh LB containing 30 µg/ml Cm grown to A600 ~ 0.5, at which point cultures were induced with 1 mM isopropyl-1-thio-{beta}-D-galactopyranoside for 1 h, and activities were assayed as described previously (38, 44). Background activities were subtracted from experimental data (16, 19, 39).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Topology of S. flexneri GtrV Based on Computational Analysis—Protein sequence analysis was performed initially by using computer programs as indicated under "Materials and Methods." Two of the most commonly used programs are shown in Fig. 2. This includes Kyte-Doolittle and TMHMM, which when applied to GtrV predict a topology that consists of nine putative transmembrane segments with a large N-terminal periplasmic loop and the N and C termini located in the cytoplasm and periplasm, respectively.



View larger version (50K):
[in this window]
[in a new window]
  		FIG. 2.
Hydropathy analysis of GtrV was determined by the algorithm of Kyte-Doolittle (A) using a sliding window of 19 amino acids. Numbered boxes indicate putative transmembrane segments. Panel B shows the TMHMM analysis algorithm applied to GtrV. Red regions indicated possible transmembrane segments with the relative probability of each indicated in the y axis. Pink and blue regions denote predicted extracellular and intracellular domains, respectively. Note that several other algorithms were also used in hydropathy analysis as described under "Materials and Methods" but are not presented in this figure.

 
Topology of S. flexneri GtrV:gtrV-phoA/lacZ Fusions—The nested deletion system as described by Sugiyama et al. (20) was used to create different length hybrid protein fusions between gtrV and the dual reporter phoA/LacZ (20). In particular, exonuclease III was used to progressively delete gtrV from the 3' end before ligating it back to the full dual reporter. Transformed cells were grown on media containing the chromogenic substrates Red-Gal (6-chloro-3-indolyl-{beta}-D-galactoside) and X-phos (5-bromo-4-chloro-3-inolyl phosphate disodium salt) used by LacZ and PhoA, respectively. Coloration was produced due to in-frame fusion of the dual reporter to the truncated gtrV series of deletions. Red was produced due to LacZ{alpha} being active in the cytoplasm, and blue was due to PhoA being active in the periplasm (16, 17). Purple colonies were also produced when fusions were adjacent to transmembrane segments or periplasmic domains (16, 17). Fusions not carrying the dual reporter produced clear colonies on dual indicator plates, whereas no in-frame reporter fusions appeared clear, but prolonged incubation after 14–30 h produced red colonies, probably a result of translational reinitiation in the vicinity of lacZ{alpha} (16, 17). Plasmid DNA from 200 such transformants was analyzed by restriction digestion. Sixty-seven fusions falling within the coding region of gtrV were sequenced to determine the exact fusion point. A total of 34 unique in-frame gtrV-phoA/lacZ gene fusions were isolated. Alkaline AP and BG for each of the constructs harboring the dual reporter were measured, and NARs were determined (NAR + (AP/highest AP) - (BG/highest BG)). The NAR values of the truncated fusions between gtrV and the dual reporter at cytoplasmic, periplasmic, and transmembrane segments are shown in Table III. Those values with NARs (AP:BG) of greater than 2:1 (in the case of AP) or less than 1:2 (in the case of BG) indicate that 67% or more of the reporter activity is properly localized (16, 17). Fig. 3A shows the topological model of GtrV as determined by gtrV-phoA/lacZ fusions and hydrophobicity data. Both support a nine-transmembrane protein with the N terminus in the cytoplasm and the C terminus in the periplasm. A characteristic extended N-terminal periplasmic loop is also apparent as observed in other S. flexneri glucosyltransferases (data not shown). NAR values of only two of them conflict with the model suggested by the remainder of the fusions as depicted in Table III and Fig. 3A. One is in the big N-terminal periplasmic loop number 2 (A40), and one in the putative cytoplasmic loop number 3 (N140). The contradicting NAR values of C-terminal fusions A40 and N140 support a cytoplasmic and periplasmic localization, respectively. Five fusions were located within the transmembrane segments (arrowheads); they predicted the correct topology of the nearest membrane face. Periplasmic loop number 4 and cytoplasmic loop number 5 were not targeted by C-terminal fusions, indicating that this particular region might be highly unstable in its structure, especially when fusions to the phoA/lacZ reporter are desired. Loops 2, 3, 4, 5, and 6 were targeted with sandwich fusions (see below), which are believed to be more accurate indicators of topology and are used for interpretation of controversial data (16, 24, 40, 41). The median NAR for this set of fusions was 25 (or 1/25), indicating the accuracy of the data. In general, the C-terminal replacement fusions strongly support a nine-transmembrane domain model for the topology of the S. flexneri GtrV glucosyltransferase as initially determined by the hydropathy data and charge distribution.


View this table:
[in this window]
[in a new window]
  		TABLE III
Analysis of GtrV topology, C-terminal fusions gtrV-phoA/lacZ and Sandwich fusions gtrV-phoA/lacZ-gtrV

 



View larger version (30K):
[in this window]
[in a new window]
  		FIG. 3.
A, topological model of S. flexneri GtrV glucosyltransferase based on truncated fusions created by ExoIII. Circles indicate fusions in hydrophilic loops established by ExoIII digestion, where NAR values support the proposed topological model (vertical-lined circles, periplasm; open circles, cytoplasm) and in conflict with the model (dotted circles). Arrowheads in transmembrane segments pointing to cytoplasm or periplasm indicate fusions whose NAR values are characteristic of cytoplasmic or periplasmic fusions, respectively. Panel B depicts the final topology of S. flexneri GtrV glucosyltransferase identified from C-terminal fusion points and characteristic sandwich fusions in the gtrV gene. Transmembrane segment number 4 from a previous model (A) is shown as a reentrant loop. Sandwich fusions are indicated for its targeted hydrophilic loop (filled black) followed by relevant restriction sites introduced (boxes).

 
Construction of gtrV-phoA/lacZ-gtrV Sandwich Fusions—To resolve inconsistencies surrounding periplasmic loop number 2, cytoplasmic loop number 3, periplasmic loop number 4, and cytoplasmic loop number 5, a total of 5 sandwich fusions were constructed by introducing restriction endonuclease sites at predetermined points along the gtrV gene using site-directed mutagenesis followed by the introduction of the dual reporter phoA/lacZ. The AP and BG activities and calculated NAR values are shown in Table III. The overall activity of the sandwich fusions in these regions was noticeably lower in comparison with the truncated versions of gtrV fused to the dual reporter. Therefore, this set of data is presented in a separate table, and a different reference fusion (highest expression level in analyzed data) was used to normalize the AP and BG activities as observed by Alexeyev and Winkler (16). The uncertainty arising in loop 2 by the C-terminal fusion A40 was alleviated by sandwich fusion H47/NruI, with a NAR value of 33:1. This clearly indicated that this loop is in the periplasm, as supported by C-terminal fusions R34, A43, W46, N53, and T57. However, the two sandwich fusions H149/NruI and Q178/NruI are inconsistent with the proposed model (Fig. 3A) as indicated by their NAR values of 6:1 and 1:50, respectively. Sandwich fusion H149/NruI is in agreement with the C-terminal fusion N140, indicating that this loop is part of the periplasm. Sandwich fusions R212/NruI and S227/NruI are in agreement with the proposed model as supported by their NAR values. Sandwich fusion S227/NruI is further supported by C-terminal fusion E228 as shown in Fig. 3A and Table III. Results from sandwich fusions H149/NruI and Q178/NruI indicate that a transmembrane segment has to precede sandwich H149/NruI. This is weakly indicated by hydrophobicity data (Fig. 2). After this and according to sandwich fusion Q178/NruI and charge distribution analysis, transmembrane segment IV (Fig. 3A) can be visualized as a reentrant loop (see "Discussion"). Most residues in this segment are hydrophobic except residues in loops targeted by sandwich fusions Q178/NruI and R212/NruI. Fig. 3B depicts the new topological model of GtrV, which accounts for transmembrane segment IV from the previous model as a reentrant loop. The model introduces GtrV as a nine-transmembrane protein with the N and C terminus in the periplasm and cytoplasm, respectively, followed by a reentrant loop after segment IV, a novel feature in bacterial glucosyltransferases (Fig. 3B).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Elucidation of the structure of membrane proteins such as GtrV and their orientation across the membrane are required to make predictions about their mechanisms of action. The topology of GtrV was determined experimentally using a genetic approach involving construction of in-frame fusions between the gtrV gene and the dual reporter genes of lacZ and phoA. Although the technique employing two reporters individually has been used in the past to elucidate the topology of a number of membrane proteins, the use of a dual reporter system is novel (16, 17, 19, 20, 22, 23, 25, 40, 42–50).

The 39 unique NAR values were used to correct for variable expression of the different fusions to provide interpretable data about reporter membrane localization. This allowed determination of the topology of GtrV, consisting of nine transmembrane segments, a large N-terminal periplasmic loop, a unique reentrant loop after segment IV, followed by the N-terminal and C-terminal protein ends in the cytoplasm and periplasm, respectively. The quality gained using NAR values is directly related to the size and diversity of the set of fusions examined. In this analysis choice of the second best reference point in calculating NAR values (data not shown) did not interfere with the true localization of the reporter, thus indicating the stability and reliability of the data (16, 17). For example, even fusions with NAR values greater than 3:1 and lower than 1:3 were used to show the correct localization of the reporter even when the next best reference point was used (16, 17). In the case of sandwich fusions, a different reference point within this dataset had to be chosen since the activities of BG were remarkably lower than the activities recorded in the set of C-terminal fusions. This could relate to steric problems in which short {alpha}-fragments tethered by AP and the C-terminal part of the protein have to cause association of four bulky {omega}-fragments (16, 17).

The first model (Fig. 3A) constructed by charge distribution data and hydropathy plots was in general agreement with the C-terminal fusion generated, although a couple of fusions were in conflict with the model. Fusion N140 displays a NAR value of 3:1, clearly indicating that this fusion belongs to the periplasm. The absence of fusions in loops 4 and 5 and obscurity of hydropathy data in that region did not allow proper interpretation and manipulation of the model. The absence of C-terminal fusions in these segments may indicate that this region of the protein is unstable when such fusions are attempted. This could be because transmembrane segment IV does not exist as a real transmembrane region and highlights a region that is perhaps structurally and conformationally unstable. This could affect the proper folding of AP and BG, which occurs in the periplasm and cytoplasm, respectively (16–19, 22, 42–44).

Sandwich fusions H149, Q178, R212, and S227/NruI indicated that the C-terminal fusion N140 did indeed belong in the periplasm, pushing loop 4 into the cytoplasm (Fig. 3A). For the hydrophilic portion of loop 3 surrounding the N140 and H149/NruI sandwich fusions to be pushed in the periplasm, a transmembrane segment has to follow. This segment was detected by hydrophobicity and charge distribution data, but it was excluded from the original model since the transmembrane segment comprised only 17 amino acids. This type of rearrangement would mean that other loops have to swap positions in the membrane. This would not be in agreement with the rest of the downstream C-terminal fusions and sandwich fusion S227/NruI in loop 6, which supports localization of this domain to the periplasm, also in agreement with fusion E228. Sandwich fusion R212/NruI clearly supported the localization of loop 5 to the cytoplasm, further supporting the periplasmic localization of loop 6. Sandwich fusion Q178/NruI showed localization of loop 4 (NAR, 1:50) to the cytoplasm. This is now in conflict with loop 5, which also shows localization to the cytoplasm. As a result, both loop 4 and loop 5 have to be located in the cytoplasm, suggesting that transmembrane IV (Fig. 3A) is not a real transmembrane segment but, rather, a putative loop. Although this loop can be pictured as it exists in the cytoplasm, its hydrophobic nature would be consistent with a portion of it spanning the membrane. This kind of arrangement is further supported by the fact that the hydrophobic regions spanning this loop are not long enough for a real transmembrane segment to be formed since only nine amino acids are present. According to N-terminal fusion Y7 loop 1 and C-terminal fusions in loop 10, which show localization to the cytoplasm and periplasm, respectively, there should be an odd number of transmembrane segments, meaning that the protein could have either 7, 9, or 11 segments. The only model that supports all fusions constructed here is the model comprising of nine transmembrane segments with transmembrane segment IV (Fig. 3A) as a reentrant loop, as depicted in Fig. 3B.

This kind of arrangement of a reentrant loop is not documented in any of the glucosyltransferase families; here we report for the first time its unique existence in a serotype-converting glucosyltransferase of S. flexneri. The existence of a reentrant loop has previously been reported in eukaryotic glutamate transporters and bacterial potassium channel KcsA and also in glycerol and water channels (9, 14). In the latter two cases, the two loops depicted in these channels have an opposite orientation, playing a major role in permeation. Furthermore, the relatively rigid structures serve as selectivity filters. In contrast to the pore loops of KcsA and the aquaporins, the reentrant loop appears not to be fixed but to undergo conformational changes dependent on the presence of sodium. It is thought that these differences in pore loops of channels and glutamate transporters relate to differences in the permeation of the two kinds (51–57). Hydrophilic residues SGP located in the middle of this reentrant loop could make contact with the periplasmic side as shown in Fig. 3B. Furthermore, computer flexibility data (not shown) suggest that this region of the reentrant loop is quite flexible. A reentrant loop in GtrX (glucosyltransferase for serotype X) at a similar location to GtrV has also been identified.2 This means that this region of GtrV might be functionally important in enabling the protein to change conformation after binding to undecaprenol phosphate-{beta}-glucose. The flexibility observed can allow conformational change and might play a role in specifically attaching the glucosyl residue to the growing O-antigen by allowing the enzyme to come in contact with the correct rhamnose II of the O-antigen. Both GtrV and GtrX mediate the formation of the common {alpha}1,3 linkage to the specific rhamnose. This phenomenon is also observed in the O-mannosyltransferase, ScPmt1p, where certain conformational changes in the protein structure bring specific loops into close proximity to form a functional catalytic unit (58). In another example, MurG:UDP-GlcNAc, a glucosyltransferase responsible for forming the glycosidic linkage between N-acetyl muramyl pentapeptide and N-acetyl glucosamine in the biosynthesis of the bacterial cell wall, exhibits a conformational change that results mostly from a rigid body domain movement in which the entire C-terminal body rotates ~10° relative to its position when the UDP-GlcNaAc is bound (59).

The relationship between this glucosyltransferase and the other glucosyltransferases that use sugar-nucleotide donors (GT) is insignificant (data not shown). The GT-A superfamily is characterized by the presence of a DXD motif and depends on metal ions for activity. This motif is not seen in GtrV or its relatives (60, 61). The association of GtrV with a metal ion is limited. Another superfamily, GT-B, which does not require metal ions (59, 62–64) for enzymatic activity, also shows weak similarity to GtrV, indicating its unique function and mode of action. By comparison, the enzymes from both GT-A and B superfamilies are quite diverse from the multitransmembrane O-antigen-modifying glucosyltransferases of S. flexneri. Because GtrV uses a lipid-linked donor such as undecaprenol phosphate-{beta}-glucose, it can be compared with such other proteins from a range of organisms. For example, glycosyltransferases that use dolichyl phosphate (Dol-P)-linked monosaccharides (Dol-P-mannose and Dol-P-Glucose) as the donor substrate show similar topological arrangements. Apart from the fact that these glycosyltransferases consist of at least 6–12 transmembrane segments, they all share a highly conserved N-terminal loop after transmembrane segment I (65–67). This arrangement is also evident in the topological structure of GtrV (Fig. 3).

Fusions A40 and N140 were resolved by constructing sandwich fusions (Fig. 3A). Fusion A40 displayed a NAR of 1:27, whereas the sandwich at position H47/NruI displayed a NAR of 33:1. C-terminal fusions after this sandwich fusion further supported the existence of this loop in the periplasm. The introduction of sandwich fusion H149/NruI downstream of the C-terminal fusion N140 indicated that loop number 3 (Fig. 3A) belongs in the periplasm, as later shown in the model depicted in Fig. 3B.

Because these fusions are generated by truncating the protein and fusing the dual reporter after it, as in the case of A40, an unbalanced protein charge is created. It is well documented that positively charged amino acids function as determinants of cytoplasmic localization, following the "positive inside" rule, and that the orientation of the transmembrane segment can be reversed by manipulating these residues (68). The distribution of charged amino acid residues of the GtrV protein is consistent with the positive inside rule. Residues located before fusion A40 have a more positive net charge than residues in loop 1. The net charge of loop number 2 in the full protein is less positive than the charge depicted in loop number 1, thus enabling the protein to be inserted in the correct orientation. A more positive periplasmic charge would mean the transmembrane segment would flip, showing incorrect cytoplasmic localization. On the other hand, fusion R34, which is closer to the transmembrane, shows the correct localization since the positively charged residues after this fusion are not present and the segment is inserted correctly. Also fusion A43 with a NAR of 1:>100 clearly indicates the periplasmic localization of this loop. This particular C-terminal fusion includes negatively charged residues E41 and D42 (not shown in model), which means that net positive charge generated is to an extent less than fusion A40, which lacks residues E41 and D42, ensuring the correct insertion of the transmembrane.

Apart from the reentrant loop, which appears to be functionally important, periplasmic loops 2 and 10 of GtrV can be assumed to have some significance in function. This is not only due to their relative larger size but also because O-antigen modification is thought to take place in the periplasm. A similar prediction has also been made in Dol-P-linked glycosyltransferases, in which the first large and highly conserved loop is thought to be involved in the formation of the active site (65). The topology analysis and identification of a novel reentrant loop in GtrV will give further insight into how glucosyltransferases transfer glucose from undecaprenol phosphate-{beta}-glucose to specific rhamnose units of the O-antigen mediating serotype conversion in S. flexneri. This study provides the basis for the identification and characterization of functional domains as well as the localization of the active site.


   FOOTNOTES
 
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} To whom correspondence should be addressed. Tel.: 61-2-61252666; Fax: 61-2-61250313; E-mail: Naresh.Verma{at}anu.edu.au.

1 The abbreviations used are: PhoA, alkaline phosphatase; LacZ, {beta}-galactosidase; AP, alkaline phosphatase; BG, {beta}-galactosidase; NAR, normalized activity ratio; Cm, chloramphenicol; GT, glycosyltransferase. Back

2 H. Korres and N. K. Verma, unpublished results. Back


   ACKNOWLEDGMENTS
 
We thank Dr. Herbert H. Winkler for providing vectors carrying the dual reporters. Also, our thanks go to Tan Hua for helping with assays for AP and BG activity. A special thanks to Helen O'Neill for reading and editing the manuscript.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. WHO (1997) World Health Forum 18, 1-8[Medline] [Order article via Infotrieve]
  2. Kotloff, K. L., Winickoff, J. P., Ivanoff, B., Clemens, J. D., Swerdlow, D. L., Sansonetti, P. J., Adak, G. K., and Levine, M. M. (1999) Bull. W. H. O. 77, 651-666[Medline] [Order article via Infotrieve]
  3. Okada, N., Sasakawa, C., Tobe, T., Yamada, M., Nagai, S., Talukder, K. A., Komatsu, K., Kanegasaki, and Yoshikawa, M. (1991) Mol. Microbiol. 5, 187-195[CrossRef][Medline] [Order article via Infotrieve]
  4. Brahmbhatt, H. N., Lindberg, A. A., and Timmis, K. N. (1992) in Pathogenesis of Shigellosis (Sansonetti, P. J., ed) pp. 45-64, Springer-Verlag, Berlin
  5. Sandlin, R. C., Goldberg, M. B., and Maurelli, A. T. (1996) Mol. Microbiol. 22, 63-73[CrossRef][Medline] [Order article via Infotrieve]
  6. Sansonetti, P. J. (1998) Jpn. J. Med. Sci. Biol. 51, 69-80
  7. Sansonetti, P. J. (2001) FEMS Microbiol. Lett. 25, 3-14
  8. Makela, P. H., and Stocker, B. A. D. (1984) in Chemistry of Endotoxin (Rietschel, E. T., ed) Vol. 1, p. 419, Elsevier Science Publishers B. V., Amsterdam
  9. Huan, P. T., Bastin, D. A., Whittle, B. L., Lindberg, A. A., and Verma, N. K. (1997A) Gene 195, 217-227[CrossRef][Medline] [Order article via Infotrieve]
  10. Allison, G. E., and Verma, N. K. (2000) Trends Microbiol. 8, 17-23[CrossRef][Medline] [Order article via Infotrieve]
  11. Kenne, L., Lindberg, B., Petesson, K., Katzenellenbogen, E., and Romanowska, E. (1978) Eur. J. Biochem. 91, 279-284[Medline] [Order article via Infotrieve]
  12. Verma, N. K., Verma, D. K., Huan, P. T., and Lindberg, A. A. (1993) Gene (Amst.) 129, 99-101[CrossRef][Medline] [Order article via Infotrieve]
  13. Guan, S., Bastin, D. A., and Verma, N. K. (1999) Microbiology 145, 1263-1273[Abstract/Free Full Text]
  14. Mavris, M., Manning, P. A., and Morona, R. (1997) Mol. Microbiol. 26, 939-950[CrossRef][Medline] [Order article via Infotrieve]
  15. Adhikari, P., Allison, G., Whittle, B., and Verma, N. K. (1999) J. Bacteriol. 181, 4711-4718[Abstract/Free Full Text]
  16. Alexeyev, M. F., and Winkler, H. H. (1999) J. Mol. Biol. 285, 1503-1513[CrossRef][Medline] [Order article via Infotrieve]
  17. Alexeyev, M. F., and Winkler, H. H. (2002) Biochemistry 41, 406-414[CrossRef][Medline] [Order article via Infotrieve]
  18. van Geest, M., and Lolkema, J. S. (2000) Microbiol. Mol. Biol. Rev. 64, 13-33[Abstract/Free Full Text]
  19. Manoil, C. (1991) Methods Cell Biol. 34, 61-75[Medline] [Order article via Infotrieve]
  20. Sugiyama, J. E., Mahmoodian, S., and Jacobson, G. R. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 9603-9607[Abstract/Free Full Text]
  21. Boyd, D. (1994) in Membrane Protein Structure: Experimental Approaches (White, S. H., ed) pp. 144-163, Oxford University Press, Oxford
  22. Manoil, C. (1990) J. Bacteriol. 172, 1035-1042[Abstract/Free Full Text]
  23. Manoil, C., and Beckwith, J. (1986) Science 233, 1403-1408[Abstract/Free Full Text]
  24. Sarsero, J. P., and Pittard, A. J. (1995) J. Bacteriol. 177, 297-306[Abstract/Free Full Text]
  25. Daniels, C., Vindurampulle, C., and Morona, R. (1998) Mol. Microbiol. 28, 1211-1222[CrossRef][Medline] [Order article via Infotrieve]
  26. Sambrook, J., and Russell, W. D. (2001) Molecular Cloning: A Laboratory Manual, 3rd Ed., pp. A2.2-A2.5, Cold Spring Harbor Laboratory Press, Cold Spring Harbor Laboratory, NY
  27. Cserzo, M., Wallin, E., Simon, I., von Heijne, G., and Elofsson, A. (1997) Protein Eng. 10, 673-676[Abstract/Free Full Text]
  28. Tusnady, G. E., and Simon, I. (1998) J. Mol. Biol. 283, 489-506[CrossRef][Medline] [Order article via Infotrieve]
  29. Rost, B., Casadio, R., Fariselli, P., and Sander, C. (1995) Protein Sci. 4, 521-533[Medline] [Order article via Infotrieve]
  30. Rost, B., Fariselli, P., and Casadio, R. (1996) Protein Sci. 5, 1704-1718[Medline] [Order article via Infotrieve]
  31. Nakai, K., and Kanehisa, M. (1991) Proteins 11, 95-110[CrossRef][Medline] [Order article via Infotrieve]
  32. Hirokawa, T., Boon-Chieng, S., and Mitaku, S. (1998) Bioinformatics 14, 378-379[Abstract/Free Full Text]
  33. Sonnhammer, E. L., von Heijne, G., and Krogh, A. (1998) Proc. Int. Conf. Intell. Syst. Mol. Biol. 6, 175-182[Medline] [Order article via Infotrieve]
  34. Hofmann, K., and Stoffel, W. (1993) Biol. Chem. Hoppe-Seyler 374, 166-168
  35. Claros, M. G., and von Heijne, G. (1994) Comput. Appl. Biosci. 10, 685-686[Free Full Text]
  36. von Heijne, G. (1992) J. Mol. Biol. 225, 487-494[CrossRef][Medline] [Order article via Infotrieve]
  37. Nilsson, J., Persson, B., and von Heijne, G. (2000) FEBS Lett. 486, 267-269[CrossRef][Medline] [Order article via Infotrieve]
  38. Henikoff, S. (1987) Methods Enzymol. 155, 156-165[Medline] [Order article via Infotrieve]
  39. Miller, J. H. (1992) A Short Course in Bacterial Genetics, pp. 72-74, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY
  40. Ehrmann, M., Boyd, D., and Beckwith, J. (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 7574-7578[Abstract/Free Full Text]
  41. Lacatena, R. M., Cellini, A., Scavizzi, F., and Tocchini-Valentini, G. P. (1994) Proc. Natl. Acad. Sci. U. S. A. 91, 10521-10525[Abstract/Free Full Text]
  42. Manoil, C., Mekalanos, J. J., and Beckwith, J. (1990) J. Bacteriol. 172, 515-518[Abstract/Free Full Text]
  43. McGovern, K., Ehrmann, M., and Beckwith, J. (1991) EMBO J. 10, 2773-2782[Medline] [Order article via Infotrieve]
  44. Traxler, B., Boyd, D., and Beckwith, J. (1993) J. Membr. Biol. 132, 1-11[Medline] [Order article via Infotrieve]
  45. Caldwell, A. M., and Smith, R. L. (2003) J. Bacteriol. 185, 374-376[Abstract/Free Full Text]
  46. Gerard, P., Vernet, T., and Zapun, A. (2002) J. Bacteriol. 184, 1925-1931[Abstract/Free Full Text]
  47. Haardt, M., and Bremer, E. (1996) J. Bacteriol. 178, 5370-5381[Abstract/Free Full Text]
  48. Bibi, E., and Beja, O. (1994) J. Biol. Chem. 269, 19910-19915[Abstract/Free Full Text]
  49. Wong, K. K., and Hancock, R. E. (2000) J. Bacteriol. 182, 2402-2410[Abstract/Free Full Text]
  50. Ginn, S. L., Brown, M. H., and Skurray, R. A. (1997) J. Bacteriol. 179, 3786-3789[Abstract/Free Full Text]
  51. Seal, R. P., Leighton, B. H., and Amara, S. G. (2000) Neuron 25, 695-706[CrossRef][Medline] [Order article via Infotrieve]
  52. Lester, H. A., Mager, S., Quick, M. W., and Corey, J. L. (1994) Annu. Rev. Pharmacol. Toxicol. 34, 219-249[CrossRef][Medline] [Order article via Infotrieve]
  53. DeFelice, L. J., and Blakely, R. D. (1996) Biophys. J. 70, 579-580[Medline] [Order article via Infotrieve]
  54. Sonders, M. S., and Amara, S. G. (1996) Curr. Opin. Neurobiol. 6, 294-302[CrossRef][Medline] [Order article via Infotrieve]
  55. MacKinnon, R. (1995) Neuron 14, 889-892[CrossRef][Medline] [Order article via Infotrieve]
  56. Grunewald, M., Menaker, D., and Kanner, B. I. (2002) J. Biol. Chem. 277, 26074-26080[Abstract/Free Full Text]
  57. Brocke, L., Bendahan, A., Grunewald, M., and Kanner, B. I. (2002) J. Biol. Chem. 277, 3985-3992[Abstract/Free Full Text]
  58. Girrbach, V., Zeller, T., Priesmeier, M., Strahl-Bolsinger, S., Bartsevich, V. V., and Pakrasi, H. B. (2000) J. Biol. Chem. 275, 19288-19296[Abstract/Free Full Text]
  59. Hu, Y., Chen, L., Ha, S., Gross, B., Falcone, B., Walker, D., Mokhtarzadeh, M., and Walker, S. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 845-849[Abstract/Free Full Text]
  60. Bourne, Y., and Henrissat, B. (2001) Curr. Opin. Struct. Biol. 11, 593-600[CrossRef][Medline] [Order article via Infotrieve]
  61. Unligil, U. M., and Rini, J. M. (2000) Curr. Opin. Struct. Biol. 10, 510-517[CrossRef][Medline] [Order article via Infotrieve]
  62. Morera, S., Imberty, A., Aschke-Sonnenborn, U., Ruger, W., and Freemont, P. S. (1999) J. Mol. Biol. 292, 717-730[CrossRef][Medline] [Order article via Infotrieve]
  63. Mulichak, A. M., Losey, H. C., Walsh, C. T., and Garavito, R. M. (2001) Structure (Camb) 9, 547-557[Medline] [Order article via Infotrieve]
  64. Ha, S., Gross, B., and Walker, S. (2001) Curr. Drug Targets Infect. Disord. 1, 201-213[CrossRef][Medline] [Order article via Infotrieve]
  65. Oriol, R., Martinez-Duncker, I., Chantret, I., Mollicone, R., and Codogno, P. (2002) Mol. Biol. Evol. 19, 1451-1463[Abstract/Free Full Text]
  66. Strahl-Bolsinger, S., Scheinost, A., Baulard, A. R., Gurcha, S. S., Engohang-Ndong, J., Gouffi, K., Locht, C., Besra, G. S., Girrbach, V., Zeller, T., Priesmeier, M., Bartsevich, V. V., and Pakrasi, H. B. (1999) J. Biol. Chem. 274, 9068-9075[Abstract/Free Full Text]
  67. Baulard, A. R., Gurcha, S. S., Engohang-Ndong, J., Gouffi, K., Locht, C., Besra, G. S., Girrbach, V., Zeller, T., Priesmeier, M., Strahl-Bolsinger, S., Bartsevich, V. V., and Pakrasi, H. B. (2003) J. Biol. Chem. 278, 2242-2248[Abstract/Free Full Text]
  68. Nilsson, I., and von Heijne, G. (1990) Cell 62, 1135-1141[CrossRef][Medline] [Order article via Infotrieve]
  69. Hanahan, D. (1983) J.Mol. Biol. 166, 557-580[Medline] [Order article via Infotrieve]
  70. Yanisch-Perron, C., Vieira, J., and Messing, J. (1985) Gene (Amst.) 33, 103-119[CrossRef][Medline] [Order article via Infotrieve]
  71. Jerpseth, B., Greener, A., Short, J. M., Viola, J., and Kretz, P. L. (1992) in Strategies Newsletter, Vol. 5, pp. 81-83, Stratagene, La Jolla, CA
  72. Lindberg, A. A., Karnell, A., Stocker, B. A., Katakura, S., Sweiha, H., and Reinholt, F. P. (1988) Vaccine 6, 146-150[CrossRef][Medline] [Order article via Infotrieve]
  73. Bachmann, B. J. (1990) Microbiol. Rev. 54, 130-197[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?



This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/21/22469    most recent
M401316200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Korres, H.
Right arrow Articles by Verma, N. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Korres, H.
Right arrow Articles by Verma, N. K.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2004 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement