OUP user menu

Autotransporter proteins: novel targets at the bacterial cell surface

Timothy J. Wells, Jai J. Tree, Glen C. Ulett, Mark A. Schembri
DOI: http://dx.doi.org/10.1111/j.1574-6968.2007.00833.x 163-172 First published online: 1 September 2007


Autotransporter proteins constitute a family of outer membrane/secreted proteins that possess unique structural properties that facilitate their independent transport across the bacterial membrane system and final routing to the cell surface. Autotransporter proteins have been identified in a wide range of Gram-negative bacteria and are often associated with virulence functions such as adhesion, aggregation, invasion, biofilm formation and toxicity. The importance of autotransporter proteins is exemplified by the fact that they constitute an essential component of some human vaccines. Autotransporter proteins contain three structural motifs: a signal sequence, a passenger domain and a translocator domain. Here, the structural properties of the passenger and translocator domains of three type Va autotransporter proteins are compared and contrasted, namely pertactin from Bordetella pertussis, the adhesion and penetration protein (Hap) from Haemophilus influenzae and Antigen 43 (Ag43) from Escherichia coli. The Ag43 protein is described in detail to examine how its structure relates to functional properties such as cell adhesion, aggregation and biofilm formation. The widespread occurrence of autotransporter-encoding genes, their apparent uniform role in virulence and their ability to interact with host cells suggest that they may represent rational targets for the design of novel vaccines directed against Gram-negative pathogens.

  • autotransporter protein
  • antigen 43
  • pertactin
  • biofilm


The ability of bacteria to adhere to a diverse range of surfaces including host tissues and abiotic elements is essential for colonization, survival and persistence in many environments. The importance of adhesion to the lifestyle of bacteria is reflected by the large range of different adhesins found even in a single species. Adhesion to surfaces is generally mediated by proteins or structural organelles on the bacterial cell surface. The best characterized of these are fimbriae; long, filamentous structures that bind to a specific receptor target via a tip-located adhesin (Klemm & Schembri, 2000). Another group of proteins associated with adhesion are the autotransporter proteins. Autotransporter proteins are secreted via the type Va system (also referred to as the autotransporter pathway) in Gram-negative bacteria (Desvaux et al., 2004; Henderson et al., 2004) and are unique in that their primary sequence contains all the information needed to traverse the outer membrane. Here, pertactin (from Bordetella pertussis), the adhesion and penetration protein (Hap; from Haemophilus influenzae) and Antigen 43 (Ag43; from Escherichia coli) are compared as prototype autotransporter proteins involved in bacterial adherence; we focus on the structural domains associated with their synthesis, transport, and function. Ag43 is highlighted and discussed in relation to virulence-related properties including adhesion, aggregation and biofilm formation.

Conserved features of autotransporter proteins

The advent of genome sequencing has identified numerous genes predicted to encode for autotransporter proteins in Gram-negative bacteria (Henderson & Nataro, 2001; Henderson et al., 2004). Indeed, a recent report identified 507 bonafide Type Va autotransporter proteins on the NCBI database (Junker et al., 2006). Structurally, autotransporter proteins are characterized by the presence of three distinct domains: (1) an N-terminal signal sequence that mediates export of the protein across the cytoplasmic membrane, (2) a surface-localized ‘mature’ protein (termed the passenger or α-domain) and (3) a carboxy-terminal domain (termed the translocation or β-domain) that facilitates secretion of the passenger domain through the outer membrane (Fig. 1) (Jose et al., 1995). While the translocation domain is highly homologous among autotransporter proteins (conserving the mechanism of transport), the secreted passenger domain demonstrates considerable sequence variation. As a result, autotransporter protein passenger domains may confer many different virulence-related phenotypes including adhesion, autoaggregation, invasion, biofilm formation and cytotoxicity (Table 1) (Henderson & Nataro, 2001).

Figure 1

Structural domains of Hap from Haemophilus influenzae, Ag43 from Escherichia coli and Pertactin from Bordetella pertussis. The translocator (β) and passenger (α) domains are indicated with amino acid positions below. Signal sequences are depicted in red and extend from the first residue to the start of the passenger domain. Also indicated for each protein are the autotransporter domain (blue), Pertactin domain (green), RGD motif (pink) and autocatalytic cleavage site (dashed line). Numbers refer to amino acids in the primary protein sequence.

View this table:
Table 1

Structural and phenotypic features of selected AT proteins

ProteinOrganismPassenger domainAdherenceOther functionsOther information
Ag43Escherichia coliCleavedHeLa culture cellsAutoaggregation
Microcolony and biofilm formation
AIDA-IE. coliCleavedHeLa culture cellsAutoaggregation
Microcolony and biofilm formation
Tib AE. coliCleavedHCT8 ileocecal cells (human intestine epithelial cells) (Lindenthal & Elsinghorst, 2001)Invasion (Elsinghorst & Weitz, 1994)
PertactinBordetella pertussisCleavedHeLa culture cells
CHO culture cells
Used in many acellular vaccines
HapHaemophilus influenzaeCleaved or uncleavedFibronectin, collagen IV, laminin (Fink et al., 2002)
Respiratory epithelial cells
Microcolony formation (Fink et al., 2003)
Serine protease
Autoaggregation (Hendrixson & St Geme, 1998)
HiaH. influenzaeUncleavedChang conjunctival cells and variety of others (Laarmann et al., 2002)Trimeric AT
UspA1/UspA2Moraxella catarrhalisUncleavedChang conjunctival cells (Aebi et al., 1998)Serum resistance (Aebi et al., 1998)Trimeric AT
HagM. catarrhalisUncleavedType II alveolar epithelial cells (Forsgren et al., 2003; Holm et al., 2003)Biofilm interference (Pearson et al., 2006)
Hemagglutination (Pearson et al., 2002)
Autoaggregation (Pearson et al., 2002)
Trimeric AT
rOmpARickettsialesCleavedL-929 cells (Li & Walker, 1998)Used as immunogen
BabAHelicobacter pyloriHuman gastric epithelial cells (Ilver et al., 1998)Potential vaccine target
ShdASalmonella typhimuriumUncleavedMurine cecal mucosa and bovine fibronectin (Kingsley et al., 2002)
YadAYersinia enterocoliticaUncleavedEpithelial cells (Heesemann & Gruter, 1987)
Professional phagocytes (Roggenkamp et al., 1996)
Extracellular matrix proteins (Emody et al., 1989; Schulzekoops et al., 1992, 1993)
Protects against complement and defensin lysis (Balligand et al., 1985; Pilz et al., 1992; Flugel et al., 1994)
Autoaggregation (Skurnik et al., 1984)
Trimeric AT
  • AT, autotransporter protein.

Pertactin, Hap and Ag43 are all well-characterized cell-surface-located autotransporter proteins that mediate bacterial adhesion. Pertactin is the archetypal autotransporter protein and functions as a major virulence factor of B. pertussis (Brennan et al., 1988; Charles et al., 1989; Smith et al., 2001). Pertactin mediates bacterial binding to the lung epithelium of mammalian hosts (Everest et al., 1996), although the exact host receptor to which pertactin binds is unknown. Pertactin is an important component of most current acellular pertussis vaccines and thus plays a role in eliciting protective immunity. Hap is associated with attachment and entry into cultured epithelial cells (St Geme et al., 1994), attachment to extracellular matrix proteins (Fink et al., 2002; Fink & St Geme, 2003) and promotes bacterial aggregation and microcolony formation. Ag43 of E. coli possesses several functions including the promotion of bacterial adhesion, cell-to-cell aggregation and biofilm formation (Diderichsen, 1980; Owen et al., 1996; Kjaergaard et al., 2000b; Sherlock et al., 2006). Although these three autotransporter proteins are produced by pathogens that cause very different diseases, their conserved structural features imply that common therapeutic or preventative strategies may be used to inhibit their function and contribute to the control of bacterial infection.

Structure of the passenger domain

Pertactin was the first autotransporter protein to have the structure of the passenger domain determined by X-ray diffraction. The structure was solved at a resolution of 2.5 Å, revealing a right-handed parallel β-helix (Emsley et al., 1994, 1996). β-Helical structures typically contain three β-sheets separated by three turns, giving the protein a ‘V’ shape in cross-section (Jenkins & Pickersgill, 2001). A complete turn of the β-helix is known as a coil, with pertactin having 16 coils in total. Pertactin demonstrates extensive ‘stacking’ across its coils, whereby similar aliphatic residues occupy equivalent positions in neighbouring β-sheets, leading to ridges of aliphatic residues across the coils (Jenkins & Pickersgill, 2001).

The right handed β-helical structure of pertactin is conserved among >97% of autotransporter protein passenger domains despite a large diversity in its sequence, length and function. The passenger domain adopts a predominantly unfolded conformation during its passage through the outer membrane, a process that occurs independent of ATP and proton gradients (Junker et al., 2006). The conserved β-helical structure may contribute to protein folding after transport through the translocation domain and may also play a role in presenting an adhesive functional tip away from the cell surface (Klemm et al., 2004; Junker et al., 2006). The C-terminal region of the pertactin passenger domain consists of a pertactin motif that is characterized by the β-helix turning into a β-sandwich that is eventually capped by a β-hairpin (Jenkins & Pickersgill, 2001). The β-sandwich (labelled as a pertactin motif in Pfam) is conserved in many autotransporter proteins including Ag43 and Hap (Fig. 1). The pertactin motif is also critical for the stability and folding of the B. pertussis BrkA autotransporter protein (Oliver et al., 2003).

The β-helical structure of the pertactin passenger domain allows functional groups to be inserted within turns that protrude from the β-helix without disturbing its structural integrity (Emsley et al., 1996). Pertactin has two motifs inserted within turns, which suggest protein–protein interactions and cell binding. The first is the tripeptide Arg–Gly–Asp (RGD), which is the attachment site for many mammalian adhesion proteins, including fibronectin, vitronectin and fibrinogen (Hynes, 1987). The pertactin RGD sequence also mediates adhesion to CHO cells, suggesting that it can bind integrin (Leininger et al., 1991). This motif is also seen in Ag43 (Klemm et al., 2004) as well as multiple other autotransporter adhesins including BrkA (Fernandez & Weiss, 1994) and Tef (Finn & Stevens, 1995). The second functional motif is two proline-rich regions; one lying immediately adjacent to the RGD motif and the other within the C-terminal region of the protein (Emsley et al., 1996). These sites may provide rapid (though weak), nonstoichiometric binding sites that are functionally important (Williamson, 1994). Other autotransporter proteins also possess loops extending from the predicted β-helix structure, which may incorporate uncharacterized functional motifs.

A structural model of the passenger domain of Ag43 has been proposed based on the tertiary structure of P.69 pertactin. These authors suggest that the Ag43 passenger domain adopts a β-helix conformation that consists of 18 internal repeats made up of c. 19 residues between amino acids 53 and 450 of the mature protein (Klemm et al., 2004). Akin to pertactin, each repeat is predicted to form a β-helical rung consisting of three β-strands and three turns, with the cylindrical core predominantly composed of hydrophobic side chains. The bottom of this long helix is predicted to fit into the β-barrel formed by the translocator domain, allowing the tip to protrude ∼10 nm from the cell surface where it is free to interact with Ag43 from neighbouring cells (Fig. 2) (Klemm et al., 2004). This model is consistent with the ability of larger cell surface structures such as fimbriae, flagella and the capsule to prevent Ag43-mediated aggregation (Hasman et al., 1999; Schembri et al., 2004; Beloin et al., 2006; Ulett et al., 2006).

Figure 2

Model of Ag43 secretion based on the mechanism proposed by Pohlner (1987). (i) The signal sequence directs translocation of the Ag43 precursor protein across the inner membrane via the Sec-dependent secretory apparatus. (ii) The translocator domain associates with the outer membrane and facilitates translocation of the unfolded passenger domain to the outside of the cell. (iii) The passenger domain folds into an active conformation. In the case of Ag43, this is predicted to be a β-helix structure composed of 19 loops. The Ag43 passenger domain is also processed by autocatalytic cleavage but remains in contact with the translocator domain via noncovalent interactions.

Although no crystal structure for Hap has been determined, the functional properties of Hap have been localized to the C-terminal 311 amino acids of the passenger domain (Fink et al., 2003). Hap also has a pertactin motif that forms a β-sandwich thought to be involved in the stability and processing of the protein.

Structure of the translocation domain

The translocation domain of autotransporter proteins is highly conserved and predicted to consist of β-pleated sheets in the form of a β-barrel (Loveless & Saier, 1997). Although diverse in sequence, all translocator domains generally consist of 250–300 amino acid residues. Structural algorithms predict that most translocator domains contain 14 antiparallel strands consisting of nine to 12 residues (Loveless & Saier, 1997; Yen et al., 2002). For example, the translocation domain of Hap has a β-barrel structure that contains a membrane-spanning α-helix and 14 antiparallel, membrane-spanning β-strands (Hendrixson et al., 1997).

Autotransporter translocation domains also share a consensus amino acid motif at the C-terminus (Struyve et al., 1991; Jose et al., 1995; Loveless & Saier, 1997). The C-terminal nine amino acids are generally alternating aromatic/hydrophobic and charged/hydrophilic, with the last residue being a tryptophan or phenylalanine. This sequence (in particular the C-terminal three residues) is predicted to play a role in outer membrane localization and/or stability of autotransporter proteins. Deletion of the C-terminal three residues of Hap (i.e. YSF) results in the absence of Hap from the cell outer membrane (Hendrixson et al., 1997).

Processing of the passenger domain

The fate of the passenger domain of autotransporter proteins is dependent on the extent of processing that occurs at the bacterial cell surface. The passenger domain may remain attached to the translocator domain either as an intact or processed unit, or may be processed and secreted into the extracellular milieu. In the case of pertactin and Ag43, the peptide bond linking the passenger and translocator domain is cleaved, either during or subsequent to the transfer of the passenger domain to the cell surface. The protease responsible for this process has not been identified for either protein; current evidence suggests that cleavage may occur via an autocatalytic mechanism (Charles et al., 1994; Henderson & Owen, 1999). Thus, pertactin is produced as a 93 kDa precursor in B. pertussis that is processed to yield a 60.3 kDa passenger domain, while Ag43 is produced as a 107 kDa precursor in E. coli that is processed to yield a 49.8 kDa passenger domain (Charles et al., 1994; Henderson & Owen, 1999). Despite this processing event, the passenger domain for pertactin and Ag43 remains attached to the bacterial cell surface via noncovalent interactions with its respective translocator subunit (Charles et al., 1989; Henderson et al., 2004). Hap is distinct from pertactin and Ag43 because although the passenger domain can be cleaved and secreted, only the uncleaved, cell-associated protein has a role in adherence (Hendrixson & St Geme, 1998).

Processing of the passenger domain also occurs for other autotransporter proteins such as AIDA-I, an E. coli outer membrane protein involved with diffuse adherence (Benz & Schmidt, 1992). Like pertactin and Ag43, AIDA-I processing is thought to occur via autocatalytic action. In contrast, TibA, an autotransporter protein that shares strong similarity to Ag43 and AIDA-I, is not proteolytically processed and its passenger and translocator domains remain covalently linked. Another exception is the IcsA autotransporter protein from Shigella flexneri, which is processed by an independent membrane-bound protease (Shere et al., 1997). Recently, it was shown that cleavage of the passenger domain of the E. coli EspP autotransporter protein is due to a novel autoproteolytic reaction that occurs inside the β-barrel (Dautin et al., 2007). Two conserved residues, an aspartate within the translocator domain (Asp1120) and an asparagine within the cleavage junction (Asn1023), are critical for processing and are predicted to form an unusual catalytic dyad that mediates self-cleavage through the cyclization of the asparagine.

Functional case study: Ag43

In this mini-review, we focus on Ag43 as a prototype autotransporter protein. The hallmarks of Ag43 are its phase-variable expression, its classic three-domain autotransporter structure and its functional properties linked to adhesion, aggregation and biofilm formation (Henderson et al., 1997; Hasman et al., 1999; Danese et al., 2000; Klemm et al., 2004; Lim & van Oudenaarden, 2007). Ag43 is a self-recognizing autotransporter adhesin that is produced as a 1039 amino acid preprotein. An N-terminal signal peptide (52 amino acids) directs translocation across the cytoplasmic membrane to the periplasm. Ag43 possesses a classical passenger-translocator domain structure, with the two domains comprising 499 and 488 amino acids, respectively. The translocator subunit encodes an outer membrane pore-forming component that facilitates transport of the passenger domain to the bacterial surface. The nearest homologue to the Ag43 passenger domain with a known tertiary structure is P.69 pertactin (Emsley et al., 1996), which exhibits 20% amino acid sequence identity and 35% similarity to the Ag43 passenger domain.

The expression of Ag43 is phase variable, with switching rates of ∼10−3 per cell per generation due to the concerted actions of Dam-methylase (positive regulation) and OxyR (negative regulation) (Henderson & Owen, 1999; Schembri & Klemm, 2001; Waldron et al., 2002; Schembri et al., 2003a; Lim & van Oudenaarden, 2007). Ag43 can be expressed on the E. coli cell surface in very high numbers (up to 50 000 copies per cell) and in K-12 strains this results in a characteristic frizzy colony morphology (Owen, 1992; Henderson & Owen, 1999; Hasman et al., 2000). Ag43-mediated aggregation is a distinct phenotype that can be visualized macroscopically as flocculation and settling of cells in static liquid suspensions. Ag43 exhibits a high degree of homology within its translocation domain to the corresponding region from several other autotransporter proteins including AIDA-I and TibA (Sherlock et al., 2004, 2005). In E. coli K-12, Ag43 is distributed evenly over the entire surface of the cell (Henderson et al., 1997). Recently, it was shown that the AIDA-I autotransporter protein is only located at the poles of the cell in E. coli strains that produce a complete LPS structure (Jain et al., 2006). These authors also demonstrated that several additional autotransporter proteins including IcsA and SepA of S. flexneri and BrkA of B. pertussis were localized to the bacterial pole in the presence of a complete LPS. Whether or not this specific translocation to the cell pole also occurs for Ag43 and other autotransporter proteins in the presence of a complete LPS remains to be elucidated.

Ag43 and adhesion

Ag43-mediated aggregation is a self-recognition process mediated by receptor–ligand reactions between Ag43 molecules on adjacent cells (Hasman et al., 1999; Kjaergaard et al., 2000a). The aggregation phenotype of Ag43 has been localized to the passenger domain (Klemm et al., 2004). Evidence for this is based on the following observations: (1) cells devoid of the Ag43 passenger subunit but not the translocator subunit do not autoaggregate, (2) the addition of crude preparations of the Ag43 passenger subunit to Ag43-expressing cells significantly reduces autoaggregation and (3) cells treated with antiserum specific for the Ag43 passenger subunit do not autoaggregate (Klemm et al., 2004). The autoaggregation property of Ag43 is also characteristic of several other autotransporter proteins including AIDA-I, TibA and Hap (Hendrixson & St Geme, 1998; Sherlock et al., 2004; Sherlock et al., 2005).

The presence of the agn43 gene (also known as the flu gene) appears to be highly conserved in E. coli and many wild-type clinical isolates contain multiple copies of this allele (Roche et al., 2001; Ulett et al., 2007). Sequence variation within the passenger-encoding domain of Ag43 has been correlated with altered characteristics of the protein (Klemm et al., 2004). For example, in a recent study on wild-type variants of Ag43 that were expressed in a defined K-12 background three out of nine Ag43 variants did not confer aggregation (Klemm et al., 2004). These variants of Ag43 were expressed in normal quantities on the cell surface and gave rise to frizzy colonies, suggesting that cell aggregation may be distinct from the characteristic frizzy phenotype commonly associated with Ag43. These studies also localized the region responsible for autoaggregation to the N-terminal one-third of the Ag43 passenger domain (Klemm et al., 2004). Current evidence indicates that ionic interactions between charged basic and acidic side chains play a role in Ag43–Ag43 self-recognition. In support of this, the N-terminal segment of the Ag43 passenger domain contains numerous basic and acidic amino acids. Ag43-mediated autoaggregation is inhibited by high NaCl concentrations and is also strongly affected by pH; autoaggregation is optimal at neutral and weakly acidic pH but is abolished at pH<3 and pH>10 (Klemm et al., 2004).

Ag43 confers weak binding to some human cell lines including Hep-2 cells (Henderson & Owen, 1999). However, this binding may be associated with amino acid variations in the passenger domain as the calcium-binding Ag43 homologue (Cah) from E. coli O157:H7 strain EDL933 did not display an affinity towards HeLa cells (Torres et al., 2002). Ag43 can be glycosylated by the addition of heptose residues at several positions in its passenger domain (Sherlock et al., 2006). While glycosylation of Ag43 does not occur in E. coli K-12, some uropathogenic E. coli isolates possess a glycosylated derivative of Ag43 (Sherlock et al., 2006). The genome-sequenced uropathogenic strain 536 encodes Ag43 variants glycosylated with heptosyl residues and this modification results in enhanced adhesion to human epithelial cells. Glycosylation of Ag43 does not affect its capacity to mediate cell–cell aggregation (Sherlock et al., 2006), a characteristic also common to AIDA-I and TibA (Sherlock et al., 2004, 2005).

Ag43 and biofilm formation

The attachment of bacteria to a surface often results in proliferation into more complex microcolony structures. Indeed, bacterial aggregation and microcolony formation can be seen as a prelude to biofilm formation. Ag43 dramatically enhances biofilm formation on abiotic surfaces (Danese et al., 2000; Kjaergaard et al., 2000a, b; Torres et al., 2002; Reisner et al., 2003) and is specifically correlated with the biofilm mode of growth (Schembri et al., 2003b). Recent work has demonstrated that Ag43 expression is correlated with biofilm formation by uropathogenic E. coli (UPEC) during infection of bladder cells (Anderson et al., 2003) and in enteropathogenic E. coli (Torres et al., 2002). Global gene expression profiling of E. coli during biofilm growth demonstrated that agn43 expression is specifically up-regulated when compared with both exponential and stationary phase planktonic cultures (Schembri et al., 2003b). Further, a comparison of the biofilm-forming capacity of Ag43 variants from different pathogenic E. coli strains demonstrated a conserved ability of Ag43 to enhance biofilm growth, albeit with different efficiency (Klemm et al., 2004). Ag43-mediated aggregation also protects bacteria against hydrogen peroxide, a phenomenon that may enhance the resistance of biofilm cells to antimicrobial agents (Schembri et al., 2003a). The simple translocation mechanism of Ag43 across the E. coli double membrane has been exploited to demonstrate that Ag43 can be expressed on the cell surface of other Gram-negative bacteria such as Pseudomonas fluorescens in a fully functional form (Kjaergaard et al., 2000a, b). This work also showed that Ag43 can be used to generate mixed bacterial biofilms (Kjaergaard et al., 2000a, b).

Ag43 and pathogenesis

The ability of bacteria to autoaggregate results in the formation of tight communities of cells that contribute to the colonization, survival and persistence of pathogenic bacteria during infection. Aggregation enables bacteria to resist host defences such as complement attack and phagocytosis more efficiently than nonaggregative bacteria (Ochiai et al., 1993; Berge et al., 1997). The ability of Ag43 to enhance bacterial aggregation and biofilm formation are phenotypes that enhance virulence. In UPEC, Ag43 is expressed in vivo during the formation of intracellular bacterial aggregates or pods in the bladder (Anderson et al., 2003). Here, it is suggested that the bacteria reside in a quiescent state in these aggregates, thereby avoiding the immune defences of the host (Justice et al., 2004; Mysorekar & Hultgren, 2006). Recently, a role for Ag43 in long-term persistence of UPEC in the urinary tract was described (Ulett et al., 2007). Of two Ag43-encoding variant genes present in the UPEC prototype strain CFT073, one was shown to mediate strong biofilm formation and promote long-term colonization in the mouse urinary bladder. Thus, sequence variation in the passenger domain of Ag43 may be associated with different functional characteristics acquired through the acquisition of pathogenicity-adapted mutations as seen in other cell surface adhesins such as FimH from type 1 fimbriae (Sokurenko et al., 1998).

The protective nature of Ag43-mediated aggregation against adverse environmental conditions such as oxidative stress and host defences also highlights the role of Ag43 in virulence. The tight packing of bacteria in cellular aggregates via Ag43 provides a mechanism for reducing local oxygen concentrations, thereby protecting the cells from damaging oxidizing agents including H2O2 (Schembri et al., 2003a). Furthermore, Ag43-expressing E. coli cells are efficiently taken up by polymorphonuclear neutrophils (PMNs) (Fexby et al., 2007). However, rather than being sacrificed, these intracellular bacteria were found to reside as tight aggregates within PMNs. Ag43-mediated uptake and survival in PMNs may therefore constitute a mechanism of survival against a primary defence mechanism associated with human innate immunity.

Autotransporter proteins as novel vaccine targets

The primary example of an autotransporter protein used as a vaccine is pertactin. For the last 50 years, efficacious whole-cell vaccines against whooping cough caused by B. pertussis have been available. Acellular pertussis vaccines have also been introduced that comprise from one to five proteins derived from B. pertussis, namely filamentous haemagglutinin, serotype 2 and serotype 3 fimbriae, pertussis toxin and the autotransporter pertactin. Of these proteins, high levels of antibody to pertactin have the highest correlation to a decreased likelihood of acquiring pertussis (Cherry et al., 1998; Hewlett & Halperin, 1998; Storsaeter et al., 1998).

Hap is also a potential vaccine target against nontypeable H. influenzae. The purified passenger domain of Hap (known as HapS) is immunogenic in mice, elicits significant anti-HapS antibody titres and protects preimmunized mice from nasopharyngeal colonization (Cutter et al., 2002). HapS self-associates, making it difficult to purify large quantities of the protein. This has been overcome by producing a truncated protein that comprises only the C-terminal 311 amino acids of HapS, i.e. the region that contributes to bacterial adherence to epithelial cells (Fink et al., 2003). This smaller fragment of HapS still reduces nasopharyngeal colonization, suggesting that Hap could be used as a vaccine antigen against nontypeable H. influenzae diseases.

Genomic sequences have greatly expanded the autotransporter protein family, making this the largest family of secreted proteins in Gram-negative bacteria. A search for the Pfam:autotransporter-β domain in selected completed genomes of Gram-negative human pathogens revealed multiple ORFs that encode putative proteins that have this autotransporter domain. These include Bordetella bronchisepticus RB50 (21 ORFs with autotransporter domains), Chlamydia pneumoniae TW-138 (22), E. coli O157:H7 EDL933 (9), Fusobacterium nucleatum ATCC25586 (13), Helicobacter pylori 26695 (4), Neisseria meningitidis serogroup A Z2491 (5), P. fluorescens PfO-1 (9), Rickettsia conorii Malish7 (4), Salmonella typhimurium LT2 (5), Shigella dysenteriae Sd197 (5), S. flexneri 2a 301 (8), Yersinia pestis KIM5 (13) and Yersinia pseudotuberculosis IP 32953 (13). The majority of these putative autotransporter proteins are uncharacterized in relation to persistence and virulence in the host. Future research will determine whether targeting autotransporter proteins is a viable approach to disease prevention.


Cell-surface-associated type Va autotransporter proteins have diverse functions, including adhesion, aggregation, biofilm formation and resistance to host defences. Despite the fact that only a limited number of these proteins have been characterized in detail at the molecular level, their function is associated with virulence. Future research will provide a more thorough understanding of the virulence-related properties of this group of autotransporter proteins and elucidate their role in host interactions. The common features associated with the secretion of autotransporter proteins, in addition to their role in virulence and their widespread occurrence in bacteria, suggest that they may represent novel vaccine targets for the prevention of infections caused by Gram-negative pathogens.


This work was supported by grants from the Australian Research Council (DP0557615) and the University of Queensland.


  • Editor: Ian Henderson


View Abstract