P.O.K and T.R. contributed equally to this manuscript.
The head of Bartonella adhesin A is crucial for host cell interaction of Bartonella henselae
Article first published online: 10 JUL 2008
© 2008 The Authors. Journal compilation © 2008 Blackwell Publishing Ltd
Volume 10, Issue 11, pages 2223–2234, November 2008
How to Cite
Kaiser, P. O., Riess, T., Wagner, C. L., Linke, D., Lupas, A. N., Schwarz, H., Raddatz, G., Schäfer, A. and Kempf, V. A. J. (2008), The head of Bartonella adhesin A is crucial for host cell interaction of Bartonella henselae. Cellular Microbiology, 10: 2223–2234. doi: 10.1111/j.1462-5822.2008.01201.x
- Issue published online: 13 OCT 2008
- Article first published online: 10 JUL 2008
- Received 26 February, 2008; revised 30 June, 2008; accepted 1 July, 2008.
Human pathogenic Bartonella henselae cause cat scratch disease and vasculoproliferative disorders (e.g. bacillary angiomatosis). Expression of Bartonella adhesin A (BadA) is crucial for bacterial autoagglutination, adhesion to host cells, binding to extracellular matrix proteins and proangiogenic reprogramming via activation of hypoxia inducible factor (HIF)-1. Like the prototypic Yersinia adhesin A, BadA belongs to the class of trimeric autotransporter adhesins and is constructed modularly consisting of a head, a long and repetitive neck-stalk module and a membrane anchor. Until now, the exact biological role of these domains is not known. Here, we analysed the function of the BadA head by truncating the repetitive neck-stalk module of BadA (B. henselae badA–/pHN23). Like B. henselae Marseille wild type, B. henselae badA–/pHN23 showed autoagglutination, adhesion to collagen and endothelial cells and activation of HIF-1 in host cells. Remarkably, B. henselae badA–/pHN23 did not bind to fibronectin (Fn) suggesting a crucial role of the deleted stalk domain in Fn binding. Additionally, the recombinantly expressed BadA head adhered to human umbilical vein endothelial cells and to a lesser degree to epithelial (HeLa 229) cells. Our data suggest that the head represents the major functional domain of BadA responsible for host adhesion and angiogenic reprogramming.
Bartonella henselae is a facultative intracellular, fastidious and slow-growing bacterium causing cat scratch disease, an often self-limiting infection characterized by lymphadenopathy. In immunocompromised patients, B. henselae can cause tumorous proliferations of endothelial cells (ECs) in the skin and internal organs, referred to as bacillary angiomatosis and peliosis hepatis (Relman et al., 1990; Anderson and Neuman, 1997). In vitro and in vivo, infections with B. henselae result in the activation of hypoxia inducible factor (HIF)-1, the key transcription factor involved in angiogenesis (Pugh and Ratcliffe, 2003), and to the subsequent secretion of vasculoproliferative cytokines [e.g. vascular endothelial growth factor (VEGF)] (Kempf et al., 2001; 2005). Both HIF-1 activation and VEGF secretion are most likely responsible for the induction of pathological vessel growth by B. henselae.
One of the best characterized pathogenicity factors of B. henselae is Bartonella adhesin A (BadA) (Riess et al., 2004) that was originally described as a ‘type IV-like pilus’ (Batterman et al., 1995). Together with, e.g. the prototypic Yersinia adhesin A (YadA) of Yersinia enterocolitica, Neisseria adhesin A (NadA) of Neisseria meningitidis (Comanducci et al., 2002), ubiquitous surface proteins A1 and A2 (UspA1, UspA2) of Moraxella catarrhalis (Hoiczyk et al., 2000) and others, BadA belongs to the class of trimeric autotransporter adhesins (TAAs) that represent important virulence factors of Gram-negative pathogens. Remarkably, all TAAs share a similar modular architecture consisting of head-, neck-, stalk- and membrane anchor-domains (Hoiczyk et al., 2000; Linke et al., 2006). The C-terminal membrane anchor defines the TAA family and has been shown to form trimers (Meng et al., 2006; Wollmann et al., 2006). In the 328 kDa BadA (monomer), the N-terminal signal sequence is followed by a head region and a long, highly repetitive neck-stalk module (see Fig. 1). BadA is crucial for the adhesion of B. henselae to host cells and to extracellular matrix (ECM) proteins like fibronectin (Fn) or collagens. Moreover, expression of BadA correlates with the induction of a proangiogenic host cell response in a variety of host cells via activation of HIF-1 and the subsequent secretion of angiogenic cytokines [e.g. VEGF, interleukin (IL)-8] (Kempf et al., 2001; 2005; Riess et al., 2004). Other Bartonella species also harbour TAAs that are highly homologous to BadA including the variably expressed outer-membrane proteins (Vomps) of B. quintana (Zhang et al., 2004) and Bartonella repeat protein A (BrpA) of B. vinsonii (Gilmore et al., 2005).
The modular construction of BadA leads to the hypothesis that certain domains are responsible for defined biological functions during the infection of the host. Such a domain-function relationship has already been proposed for YadA featuring a prominent role of the YadA head in the infection process (Roggenkamp et al., 2003). The N-terminal part of the BadA head is homologous to the YadA head whereas the C-terminal part of the BadA head was originally described as a part of the stalk (Riess et al., 2004). Recently, the structure of this part of the protein was solved by X-ray crystallography (Szczesny et al., 2008) showing remarkable similarity to parts of the head of the TAA Hia from Haemophilus influenzae (Yeo et al., 2004). Sequence analysis of the heads from different B. henselae isolates revealed a high identity between these strains, while the length of the stalk domain varies significantly (Riess et al., 2007). TAAs like BadA represent an attractive opportunity to assign a defined biological function of a bacterial adhesin to a certain protein domain. However, apart from YadA, there is only limited knowledge about domain-function relationships in other TAAs.
Here, we analysed the function of the BadA head by performing in-frame deletion mutagenesis of badA leading to the expression of a truncated BadA (BadA HN23) lacking 21 of the 23 neck-stalk repeats that are present in B. henselae Marseille wild type BadA. Like wild type bacteria, B. henselae Marseille badA−/pHN23, expressing truncated BadA on the surface, showed autoagglutination, adhesion to collagen and to host cells and, moreover, induced the activation of HIF-1 as well as the subsequent secretion of angiogenic cytokines (VEGF, IL-8). In contrast to B. henselae Marseille wild type, badA−/pHN23 did not bind Fn indicating a crucial role of the neck-stalk domain in this process. Additionally, recombinantly expressed BadA head adhered to human umbilical vein endothelial cells (HUVECs) and to a lesser degree to epithelial (HeLa 229) cells. Summarized, our results demonstrate that the BadA head is the biologically active part of BadA responsible for host–microbe interaction of B. henselae.
Cloning and expression of BadA HN23 in B. henselae
To analyse the function of the BadA head in detail, in-frame deletion mutagenesis of badA was performed. The resulting truncated BadA (BadA HN23), which lacks 21 neck-stalk repeats, was expressed in B. henselae Marseille badA− (Riess et al., 2004) by plasmid pHN23 (Fig. 1). Mutagenesis was performed by overlap extension PCR creating first a 2.1 kb head fragment and a 1.0 kb anchor fragment including neck #23, followed by the fusion of these fragments creating the 3.1 kb badA HN23. The fusion construct was finally cloned into the broad host range vector pBBR1MCS-5 and electroporated in B. henselae badA− leading to B. henselae Marseille badA−/pHN23. Further details of cloning procedures are given in the Experimental procedures and in Tables 1 and 2.
|Strain||Characteristics||Reference or source|
|wild type||B. henselae Marseille, patient isolate, early passage, BadA-positive||Drancourt et al. (1996); Riess et al. (2004)|
|badA–||B. henselae Marseille, EZ::TN<KAN2>-transposon mutant, BadA-negative, Kmr||Riess et al. (2004)|
|badA–/pHN23||B. henselae badA– containing pHN23, Kmr, Gmr||This study|
|TOP 10||Host strain used for cloning||Invitrogen|
|DH5α||Host strain used for cloning||Invitrogen|
|XL1-Blue||Host strain used for protein expression||Stratagene|
|Plasmids||Characteristics||Reference or source|
|pBBR1MCS-5||Broad host range vector, Gmr||Kovach et al. (1995)|
|pQE-30 Xa||Expression vector, 6xHis tag, Apr||Qiagen|
|pCR-Blunt II TOPO||Blunt end topoisomerase vector, Kmr||Invitrogen|
|pTR14||pBluescript II KS containing B. henselae Marseille badA||Riess et al. (2004)|
|pPK3||pCR-Blunt II TOPO containing a 3.1 kb badA fragment carrying an in-frame deletion of aa 470–2850 (BadA HN23), Kmr||This study|
|pHN23||pBBR1MCS-5 containing the 3.1 kb badA fragment (BadA HN23) from pPK3, Gmr||This study|
|badA f8||TCGAATCTTGCGCTTACAGGAGC||This study|
|badA r8||TGATATCATGGATCCTTATGCTTTTAGCTGTGCa||This study|
|badA f10||CTGAATTTAGAGAGTGTAAGC||This study|
|badA r11||CTAACAGCTACGTAATTTTTTGACTACCAGCATTAATAC||This study|
|badA f14||GCTGGTAGTCAAAAACTTACGCATGTAGAGAATGG||This study|
|badA r15||TTTTTCGTAGAAACAAGAGACC||This study|
To prove the expression and surface localization of BadA HN23, badA−/pHN23 was analysed by immunofluorescence using BadA-specific antibodies (Abs) that target the C-terminal part of the BadA head and the first stalk repeat (Riess et al., 2004). Staining for BadA led to a strong immunofluorescent signal of badA−/pHN23 comparable to BadA-expressing B. henselae wild type whereas B. henselae badA− remained clearly negative. Moreover, immunofluorescence labelling revealed a surface expression of BadA HN23 appearing as a clearly visible halo phenomenon around the bacteria (see Fig. 2A).
To further elucidate the BadA HN23 surface expression, electron microscopy was performed (Fig. 2B). Here, B. henselae wild type expressed BadA in its natural length of ∼240 nm including a long stalk and a more electron-dense structure on the tip presumably representing the BadA head. B. henselae badA− showed no BadA expression and badA−/pHN23 expressed a strongly truncated BadA construct on its surface (length: ∼25 nm) which was clearly labelled in immunoelectronmicroscopic (IEM) studies by Abs directed against the C-terminal part of the BadA head (Fig. 2C).
Summarized, these data demonstrate that BadA HN23 is expressed on the surface of badA−/pHN23 suggesting that the autotransporter function of BadA is not affected by truncating the stalk region.
BadA HN23 expression is sufficient for autoagglutination of B. henselae
A typical feature of TAAs is their ability to mediate autoagglutination (Linke et al., 2006) as shown for BadA (Riess et al., 2007) and for the Vomps of B. quintana (Zhang et al., 2004). Therefore, the ability of BadA HN23 to mediate autoagglutination of B. henselae was investigated. Confocal laser scanning microscopy (CLSM) revealed that expression of BadA HN23 in B. henselae was sufficient to cause autoagglutination similar to B. henselae wild type whereas this phenomenon was not observed in B. henselae badA− (Fig. 3A). Moreover, BadA head-dependent autoagglutination was also demonstrated by the clearance of bacterial suspensions of the different B. henselae strains. Here, only suspensions of B. henselae wild type and badA−/pHN23 showed clearance whereas the B. henselae badA− suspension remained turbid (Fig. 3B). These data suggest that the BadA head plays a crucial role in the BadA-dependent autoagglutination of B. henselae.
BadA HN23 expression is sufficient for adhesion of B. henselae to collagen and HUVECs
Expression of BadA is crucial for the interaction of B. henselae with ECM proteins and HUVECs (Riess et al., 2004). To investigate whether this phenomenon is specifically linked to the BadA head, the binding of badA−/pHN23 to ECM proteins (collagen, Fn) was assessed. For this purpose, B. henselae was exposed to collagen-coated coverslips and the adherent bacteria were visualized by 4',6-diamidino-2-phenylindole (DAPI) staining and CLSM. Data revealed that both B. henselae wild type and badA−/pHN23 bound to a significantly higher amount to collagen-coated coverslips than B. henselae badA− did (see Fig. 4).
Bacteria-bound Fn [originating from Columbia blood agar (CBA) on which B. henselae were grown] was detected by immunoblotting using whole bacterial lysates and anti-Fn Abs. Interestingly, these experiments revealed that only B. henselae wild type (expressing BadA with its highly repetitive stalk region) bound Fn whereas B. henselae badA− or badA−/pHN23 did not (see Fig. 5A), indicating a crucial role of the repetitive neck-stalk module in this process and suggesting different mechanisms for collagen and Fn binding by BadA. For internal control, expression of BadA and BadA HN23 was analysed in parallel by immunoblotting using anti-BadA Abs. As described before (Riess et al., 2004; 2007), multiple bands of BadA were present in wild type bacteria representing presumably degradation products of this high molecular weight adhesin. These bands are absent in B. henselae badA− whereas in lysates of badA−/pHN23 monomeric (∼74 kDa) and trimeric (∼220 kDa) BadA HN23 was detected (see Fig. 5B).
To investigate the role of the BadA head in BadA-dependent binding of B. henselae to host cells, HUVECs were infected with B. henselae wild type, badA− and badA−/pHN23 and adherence was analysed 30 min after infection by CLSM. In contrast to B. henselae badA−, high numbers of B. henselae wild type and badA−/pHN23 adhered to HUVECs (see Fig. 6) indicating that the expression of BadA HN23 is sufficient to mediate binding of B. henselae to host cells. Taken together, these data suggest that the head of BadA seems to be most important for the adhesion of BadA to collagen and host cells whereas the disability of badA−/pHN23 to bind to Fn indicates a crucial role of the neck-stalk repeats in this process.
BadA HN23 expression is sufficient for the induction of a proangiogenic host cell response
Angiogenic reprogramming of host cells by B. henselae underlies a BadA-dependent activation of HIF-1 (Riess et al., 2004; Kempf et al., 2005), the key transcription factor involved in angiogenesis regulating, e.g. the secretion of VEGF. To verify whether this reprogramming is also mediated by a truncated BadA, epithelial HeLa 229 cells were infected with B. henselae wild type, badA− and badA−/pHN23 and HIF-1 activation was detected by immunoblotting of whole cell lysates using anti-HIF-1α Abs. Data revealed that both B. henselae wild type and badA−/pHN23 induced a robust HIF-1 activation whereas B. henselae badA− did not (see Fig. 7A).
Next, the ability of badA−/pHN23 to induce the secretion of proangiogenic cytokines (VEGF, IL-8) by infected host cells was assessed. HeLa 229 cells were infected with B. henselae wild type, B. henselae badA− and badA−/pHN23 and VEGF as well as IL-8 levels were quantified from cell culture supernatants by ELISA. Infection with both B. henselae wild type and badA−/pHN23 triggered the secretion of VEGF and IL-8 whereas B. henselae badA− did not (see Fig. 7B and C).
Summarized, these data indicate that the expression of the BadA head by B. henselae is sufficient for the induction of an angiogenic host cell response (HIF-1 activation, VEGF secretion, IL-8 secretion) and suggest that angiogenic reprogramming is independent from the presence of the long repetitive neck-stalk module of BadA.
Recombinantly expressed BadA head binds directly to host cells
To analyse the biological function of the N-terminal part of the BadA head, which is homologous to the YadA head, this 42 kDa subdomain [BadAhead: amino acids (aa) 48–376] was expressed recombinantly in Escherichia coli. After purification of the recombinant protein, the interaction of the BadAhead with host cells was investigated. For this purpose, total host cell-bound BadAhead was detected by immunoblotting of whole cell lysates from BadAhead-exposed HUVECs and HeLa 229 cells. Data revealed that purified BadAhead adhered to both cell types (see Fig. 8A). Remarkably, BadAhead adhered much stronger to HUVECs than to HeLa cells, and this finding is in accordance with data from adhesion experiments using BadA-expressing B. henselae wild type. Here, the number of cell-adherent bacteria was significantly higher when HUVECs were infected (18.5% of inoculum) compared with HeLa 229 cells (9.0%, see Fig. 8B). Therefore, these data suggest a crucial role of the BadA head in binding to host cells and, moreover, might possibly explain the endothelial tropism of B. henselae on a molecular basis.
The important pathogenicity factor BadA of B. henselae belongs to the class of TAAs (Linke et al., 2006) that are characterized by a conserved domain organization (head, neck-stalk, membrane anchor). The prototypical TAA YadA (Hoiczyk et al., 2000) mediates binding of Y. enterocolitica to ECM proteins and host cells and is also essential for autoagglutination (reviewed in el Tahir and Skurnik, 2001). Moreover, YadA expression in Y. enterocolitica triggers IL-8 secretion from host cells resulting in the recruitment of polymorphonuclear leucocytes to the intestinal mucosa (Schmid et al., 2004).
For YadA, some detailed knowledge exists about the biological functions of different domains. The YadA head domain is responsible for binding to collagen and epithelial cells and for autoagglutination independent of the length of the stalk (Roggenkamp et al., 2003). In the head domain of YadA from Y. pseudotuberculosis, a short N-terminal sequence not present in YadA from Y. enterocolitica is responsible for host cell invasion and Fn binding. Additionally, analysis of YadA mutants indicated that the binding regions for collagen and laminin may be linked but are distinct from the Fn-binding domain (Heise and Dersch, 2006). The so-called neck-domain seems to be necessary for correct folding of the YadA trimer (Roggenkamp et al., 2003) whereas the membrane anchor and the linking region are indispensable for autotransport and stable expression of YadA on the bacterial surface (Roggenkamp et al., 2003; Grosskinsky et al., 2007).
In the genus Bartonella, several close relatives of BadA exist: the Vomps (VompA–D) of B. quintana which are important for collagen binding and autoagglutination (Zhang et al., 2004) and BrpA of B. vinsonii with unclear function (Gilmore et al., 2005). BadA is also present in the rat pathogenic B. tribocorum strain (Saenz et al., 2007). All TAAs of Bartonella spp. are characterized by a highly repetitive neck-stalk module consisting, e.g. of 23 repeats in case of B. henselae Marseille responsible for the enormous length of BadA (Riess et al., 2007).
The data presented herein reveal that BadA HN23 is surface-expressed in B. henselae, meaning that a long and repetitive neck-stalk module is not necessary for autotransport through the membrane-anchor-built pore. Additionally, the BadA head is obviously responsible for autoagglutination and the adherence of B. henselae to collagen and HUVECs (Figs 3, 4 and 6). Moreover, expression of BadA HN23 is sufficient to induce angiogenic reprogramming of infected HeLa 229 cells (HIF-1 activation and the secretion of angiogenic cytokines, see Fig. 7). These data are consistent with experiments in which recombinantly expressed BadAhead (meaning the N-terminal region homologous to the YadA head) was used: here, BadAhead binds more efficiently to HUVECs than to HeLa 229 cells and this observation correlates with the ability of host-cell binding of viable B. henselae (see Fig. 8). Therefore, these binding assays might explain the molecular basis for the endothelial tropism of B. henselae. Taken together, these data underline that the head indeed represents the most crucial part of BadA for host cell interaction.
Our data also suggest that the C-terminal part of the head (missing in recombinantly expressed BadAhead, see Fig. 1) is not necessarily involved in host cell adhesion. In fact, the exact role of this subdomain is not clear right now. Recent data show that this part of the BadA head is homologous to parts of the head of H. influenzae Hia (Szczesny et al., 2008), but an unambiguous binding specificity to host matrix proteins could not be established yet. It is notable that this subdomain is not present in the Vomps A, B, C that are expressed in B. quintana, but only in VompD that is not expressed under standard growth conditions (Zhang et al., 2004; Schulte et al., 2006). As B. quintana does not bind Fn, this led to the original hypothesis that the C-terminal part of the BadA/VompD head might be the crucial part for interactions with Fn (Schulte et al., 2006).
Interestingly, B. henselae expressing BadA HN23 does not bind Fn (Fig. 5). This observation suggests that the Fn-binding region of BadA is not located in the head but probably in the neck-stalk module. Other possible explanations might be that Fn binding sites located both in the head and in the stalk are involved in this interaction or that a binding site in the head region is not accessible when the stalk domain, which might function as a spacer, is too short. We attempted to produce B. henselae expressing BadA with neck-stalk modules of intermediate length to clarify this question. However, we have not been successful in generating such BadA-constructs until now for unclear reasons. Alternatively, cocrystallization of BadA domains together with Fn might help to elucidate the exact mode of Fn binding by BadA. It is obvious that functional analysis of further B. henselae-BadA mutants (deletion of the head, deletion of different numbers of neck-stalk repeats, etc.) would help to elucidate the biological role of particular BadA domains as this has been done for, e.g. YadA (Roggenkamp et al., 2003). However, it must be realized that genetic manipulation of badA is difficult: because of its enormous size (badA: 9.3 kb) and the number of highly repetitive sequences, many genetic approaches like PCR-based cloning or site-directed mutagenesis are difficult to perform. Moreover, B. henselae is growing very slowly. Using a newly described liquid medium (Riess et al., 2008), performing such modifications might be easier in future.
Based on infection experiments using β1-integrin deficient cell lines, we have formerly hypothesized that B. henselae binds to β1-integrins expressed on the surface of ECs and that this interaction is mediated via a BadA–Fn bridge (Riess et al., 2004). Accordingly, it has been speculated that a similar mechanism underlies YadA-mediated host cell interaction of Yersinia (Bliska et al., 1993). Surprisingly, although not binding to Fn, badA−/pHN23 still adheres to HUVECs and this finding is in accordance with the data obtained using recombinant BadAhead (Fig. 8). In light of our current results, our original hypothesis at least needs to be modified in that way that Fn bridging might not be exclusively necessary for BadA-mediated bacterial adhesion to ECs. It is known that certain members of the β1-integrin family (e.g. α5β1-integrins) bind Fn whereas other heterodimers (e.g. α1β1-, α2β1-, α6β1- and α7β1-integrins) bind collagen and/or laminin (Hynes, 1992). Host-cell binding might therefore also be mediated via collagens or laminin and this would be in accordance with the unaffected ability of badA-/pHN23 to bind collagen (see Fig. 4). Other possibilities include direct binding of BadA to β1-integrins or the involvement of further host-cell receptors like carcinoembryonic antigen-related cell adhesion molecules which have been shown to be crucial for the infection process with other human pathogenic bacteria [e.g. M. catarrhalis (N'Guessan et al., 2007), H. influenzae (Bookwalter et al., 2008) or N. meningitidis (Griffiths et al. 2007)].
Until now it is still unclear (i) whether BadA represents just an adhesin mediating the first step in the infection process necessary for further ‘angiogenic’ interactions of the pathogen with host cells [e.g. via the VirB type IV secretion system (T4SS)]; or (ii) whether BadA itself mediates angiogenic reprogramming. Unfortunately, the analysis of HIF-1 activation in HeLa 229 cells exposed to BadAhead led to contradictory results (data not shown) which might be due to, e.g. instability of the refolded protein or lipopolysaccharide contaminations of the protein preparation leading to a camouflage of a specific HIF-1 activation by BadAhead (Frede et al., 2006). Such problems might be overcome when using B. henselae mutants expressing BadA with a deletion of the head. However, the generation of such mutants is hampered by the long size of badA and the high number of repetitive sequences in the stalk region. Moreover, it can be speculated that Bartonella-translocated effector proteins (Beps), which are translocated by the VirB4 T4SS of B. henselae might also contribute to the induction of angiogenesis, e.g. by inhibiting the apoptosis of ECs (Schülein et al., 2005). The detailed analysis of the interaction of BadA and the VirB4 T4SS in the infection process of the host might provide even a more precise knowledge about the induction of angiogenesis by B. henselae.
Trimeric autotransporter adhesins represent important pathogenicity factors of Gram-negative pathogens. For instance, expression of YadA and NadA is essential for establishing enteritis or neonatal infections (Comanducci et al., 2002; Tamm et al., 1993). For Bartonella spp. it was shown that the expression of a BadA-homologue by B. tribocorum or Vomps by B. quintana is necessary for establishing persistent bacteremia in animal models (Saenz et al., 2007; Mackichan et al., 2008). Until today, unfortunately, no animal models for the analysis of vasculoproliferative disorders caused by B. henselae are described. Our data demonstrate that the head of BadA holds most of the biological activity of BadA. Obviously, the existence of an appropriate animal model might help to elucidate the in vivo relevance of our findings. It is also clear that a systematic approach analysing the biological functions of TAA-domains from a broader repertoire of bacteria would allow to understand how host-adapted interaction has developed in human pathogenic Gram-negative bacteria. For instance, such experiments might elucidate the specificity of epithelial cell adherence by respiratory pathogens (M. catarrhalis, H. influenzae, etc.) versus the well-described EC tropsim of B. henselae and, moreover, might reveal a potential linkage between certain TAA domains and the induction of an inflammatory host response (e.g. depending on the expression of YadA by Y. enterocolitica) versus the activation of angiogenic signalling cascades in infections with BadA-expressing B. henselae.
Bacterial strains and growth conditions
Bacteria used in this study are summarized in Table 1. B. henselae was grown on CBA in a humidified atmosphere at 37°C and 5% CO2. E. coli was grown in Luria–Bertani (LB) broth at 37°C. Antibiotics were used at the following concentrations: kanamycin 30 μg ml−1, gentamicin 10 μg ml−1. For production of bacterial stock suspensions, bacteria were harvested from agar plates after 5 days of culture and frozen at −80°C in LB broth containing 20% glycerol.
Construction and cloning of badA HN23
Plasmids and primers used in this study are listed in Table 2. DNA manipulations were performed according to standard protocols.
In-frame deletion within badA was performed by overlap extension PCR. A 2127 bp fragment including the badA promoter and the region coding for amino acid residues (aa) 1–469 (head region; primers: badA f10 and badA r11) and a 1008 bp fragment of badA including the region coding for aa 2851–3082 (stalk region beginning with neck #23 and membrane anchor; primers: badA f14 and badA r15) were amplified by PCR using Phusion polymerase (Finnzymes, Espoo, Finland). pTR14 was used as template DNA. The primer badA r11 contained a short 5′-overhang complementary to the 5′-end sequence of the neck 23 fragment. badA f14, respectively, contained a short 5′-overhang complementary to the 3′-end sequence of the head fragment. For construction of BadA HN23 (Head Neck-23, Fig. 1), these two fragments were fused in a third PCR by using the primers badA f10 and badA r15. The 3135 bp fusion-construct badA HN23 was ligated into pCR-Blunt II TOPO (resulting in the plasmid pPK3) and electroporated into E. coli TOP 10. For expression in B. henselae, pPK3 was digested with EcoRI and the insert was ligated into the broad host range vector pBBR1MCS-5. The resulting plasmid pHN23 was amplified in E. coli DH5α and electroporated in B. henselae badA− resulting in badA−/pHN23 which expresses BadA HN23 by pHN23.
Cloning, expression and purification of the BadA head (BadAhead)
For cloning and expression of the N-terminal part of the BadA head, a 987 bp fragment coding for aa 48–376 (nucleotide position 142–1128) was amplified by PCR with primers badA f8 and badA r8 and ligated into the pQE-30 Xa vector (Qiagen, Hilden, Germany) as recently described (Wagner et al., 2008). In brief, the expression of the head subunit in E. coli XL1-Blue (Stratagene, La Jolla, CA, USA) was induced with 1 mM IPTG (Fermentas, St Leon-Rot, Germany) for 4 h. Cells were harvested, resuspended and lysed using a French press (Thermo Spectronic; Rochester, NY, USA). After the first pass through the French pressure cell, the buffer was supplemented with 10 μg ml−1 of DNase, 10 mM MgCl2 and 10 mM MnCl2. After three passes, inclusion bodies were harvested by centrifugation (10 min, 2500 g) and the resulting pellet was washed. The pellet was resuspended in 6 M guanidine-HCl supplemented with 10% glycerol, 500 mM NaCl and 50 mM Tris-HCl (pH 8). The His-tagged BadA fragment was purified under denaturing conditions using a linear gradient from 0 to 500 mM imidazole with a NiNTA column on an Äkta Purifier system (GE Healthcare, Munich, Germany). For refolding, the peak fractions from the chromatography were diluted into a cold Tris/Arginine buffer (50 mM Tris-HCl pH 8, 100 mM NaCl, 1 M arginine) and the solution was dialysed against 50 mM Tris-HCl pH 8. Precipitated protein was removed by centrifugation and the concentration of the purified soluble protein was measured. Size exclusion chromatography showed that the protein solution contained mostly trimers; only a small peak was observed for the monomeric form (data not shown).
Generation of polyclonal anti-BadAhead antibodies
A mouse was immunized intraperitoneally with 25 μg of purified BadAhead protein (dissolved in 100 μl) mixed with 100 μl of incomplete Freud's adjuvant (days 1, 19 and 29). Serum was taken on day 39 and BadA reactivity was assessed via immunoblotting of whole-cell lysates from B. henselae Marseille wild type (data not shown).
Generation of polyclonal anti-BadA antibodies
BadA-specific Abs were generated as described (Riess et al., 2004). Briefly, a rabbit was immunized with a purified BadA-fragment comprising the C-terminal part of the head and the beginning of the stalk (aa 375–536) followed by affinity chromatography of the serum.
Immunostaining and CLSM
For microscopic analysis of BadA expression, the respective B. henselae strains were resuspended in PBS, air-dried on glass slides and fixed in 3.75% PBS-buffered paraformaldehyde (PFA). Briefly, PFA-fixed bacteria were washed three times with PBS at the beginning and after each incubation step. Bacteria were incubated with anti-BadA-specific rabbit IgG Abs (Riess et al., 2004) for 45 min followed by incubation with carbocyanine (Cy2)-conjugated secondary anti-IgG Abs (Dianova, Hamburg, Germany). Finally, bacteria were stained with 1 μg ml−1 DAPI for 5 min and the slides were analysed with a Leica DMRE fluorescent microscope equipped with a Spot RT monochrome digital camera and the related Spot advanced software (Visitron, Puchheim, Germany).
For determination of bacterial adhesion onto host cells, HUVECs (see below) were seeded onto collagen G (Biochrom, Berlin, Germany) coated coverslips (1.0 × 105 cells), infected with the respective B. henselae strains for 30 min, washed gently with pre-warmed cell culture medium and infection was stopped by adding 3.75% PFA. Bacteria were stained with mouse polyclonal Abs raised against B. henselae for 45 min followed by incubation with a FITC-conjugated secondary anti-IgG Ab (Dako, Hamburg, Germany). Then, cells were permeabilized by incubation with 0.1% Triton X-100 (Sigma, Deisenhofen, Germany) in PBS for 15 min, washed and incubated with tetramethyl rhodamine iso-thiocyanate (TRITC)-labelled phalloidin (Sigma) for 1 h. Cellular fluorescence was evaluated using a Leica DM IRE 2 CLSM. Three different fluorochromes were detected representing the green (FITC), red (TRITC) and blue (DAPI) channels. Images were digitally processed with Photoshop 7.0 (Adobe Systems, Mountain View, CA, USA).
Determination of autoagglutination
For microscopic analysis of autoagglutination, B. henselae strains were harvested from agar plates and extensively resuspended in PBS. Bacterial suspensions were adjusted to a concentration of 2.0 × 108 bacteria ml−1[optical density (OD) 600 nm 1.0:5.0 × 108 bacteria ml−1]. Three microlitres were added in a plastic tube and incubated at 37°C and 5% CO2 in a humidified atmosphere. After 1 h of incubation, 10 μl of the suspension were taken from the middle of the tube and transferred to a glass slide, air-dried and fixed with 3.75% PFA. After washing with PBS, bacteria were stained with DAPI and autoagglutination was evaluated by CLSM.
Macroscopic autoagglutination was assayed by a modification of the method described by Laird and Cavanaugh (Laird and Cavanaugh, 1980). Three microlitres of the bacterial suspension (1.0 × 109 bacteria ml−1) were incubated for 24 h in a plastic tube at 37°C and 5% CO2 in a humidified atmosphere. Clearance of the suspension was recorded with a digital camera.
Determination of collagen binding of B. henselae
To assess the binding of B. henselae to collagen, bacteria were adjusted to a concentration of 1.0 × 107 bacteria ml−1 and sedimented onto collagen G-coated coverslips by centrifugation for 5 min at 300 g. After 30 min, coverslips were washed extensively with PBS, fixed with 3.75% PFA, and bacteria were stained with DAPI. Adherence was determined via CLSM.
Detection of fibronectin binding of B. henselae and BadA immunoblotting
Fibronectin binding of B. henselae and BadA expression was assessed by immunoblotting. Bacteria were harvested in PBS from CBA (containing Fn), adjusted to an OD600 of 1.0 and lysed in SDS sample buffer and separated by 8% SDS-PAGE. Membranes were incubated with monoclonal anti-Fn Abs (Becton Dickinson, Heidelberg, Germany) or with anti-BadA-specific rabbit IgG Abs as described (Riess et al., 2004).
Culture and infection of HUVECs and HeLa cells
Human umbilical vein ECs were cultured in EC growth medium (PromoCell, Mannheim, Germany) as described (Kempf et al., 2000). For infection experiments 1.0 × 105 cells were seeded onto collagen G-coated coverslips the day before the experiment. HeLa 229 cells were grown in RPMI 1640 supplemented with 10% fetal calf serum (FCS, Sigma). Infection experiments were performed in cell culture media without antibiotics and FCS to avoid unspecific HIF-1 activation or cytokine secretion. For the determination of HIF-1 activation, 5.0 × 105 cells were seeded in 12-well plates, for adhesion assays and the determination of secreted cytokines, 1.0 × 105 cells were seeded in 24-well plates. B. henselae was used at a multiplicity of infection (moi) of 100 for adhesion experiments and at an moi of 500 for HIF-1 activation and VEGF secretion and IL-8 secretion. Bacteria were centrifuged onto cultured cells for 5 min at 300 g. Uninfected cells were used for negative control; desferrioxamine (DFO, 200 μM, Sigma) was used as positive control. Bacterial adhesion was determined 30 min after infection with B. henselae (moi 100) as described (Riess et al., 2004). For adhesion of recombinantly expressed BadAhead to host cells, protein was added to host cells at a concentration of 10 μg ml−1 for 30 min. For internal control, BadAhead was digested with 200 μg ml−1 proteinase K (Qiagen) for 45 min at 37°C followed by heat inactivation for 10 min at 98°C. Adhesion was determined by immunoblotting of whole cell lysates using polyclonal anti-BadAhead Abs.
Hypoxia inducible factor-1α immunoblotting
Proteins from cell cultures were extracted as described (Kempf et al., 2005) and blotted onto polyvinylidene fluoride (PVDF) membranes (Millipore, Schwalbach, Germany). Mouse anti-HIF-1α Abs (Becton Dickinson) were used as primary Abs and horseradish peroxidase-conjugated goat anti-mouse IgG (DAKO, Glostrup, Denmark) were used as secondary Abs. Signals were visualized with the enhanced chemiluminescent reagent (Amersham-Pharmacia). As an internal loading control, actin-specific Abs (Sigma) were used.
Quantification of VEGF and IL-8 secretion
Vascular endothelial growth factor concentrations were measured from cell culture supernatants by ELISA (R&D systems, Wiesbaden, Germany) according to the manufacturer's instructions. Secreted IL-8 was quantified by an in-house ELISA as described (Riess et al., 2004).
Transmission electronmicroscopy and immunoelectronmicroscopy
For ultrastructural studies, bacterial colonies grown on CBA were fixed with 2.5% glutaraldehyde in PBS for 20 min at ambient temperature and kept for 20 h at 4°C. Fixed cells were covered with 2% agarose and blocks containing single colonies were cut out. After post-fixation with 1% osmium tetroxide in 100 mM phosphate buffer pH 7.2 for 1 h on ice, these blocks were rinsed with aqua bidest, treated with 1% aqueous uranyl acetate for 1 h at 4°C, dehydrated through a graded series of ethanol and embedded in epon.
For on-section immunolabelling, cells were fixed with 2.5% glutaraldehyde in PBS, dehydrated in a graded series of ethanol at progressive lower temperature from 0°C down to −40°C, infiltrated with Lowicryl HM20 and UV-polymerized at −40°C. Unspecific binding sites on ultrathin sections were blocked with 0.5% bovine serum albumin and 0.2% gelatine in PBS. Ultrathin sections were then incubated with a BadA-specific rabbit IgG Ab [10 μg ml−1; raised against the C-terminal part of the BadA head (Riess et al., 2004)] followed by protein A – 10 nm gold conjugates (gift from Dr Y. Stierhof, Tübingen). Sections were stained with 1% aqueous uranyl acetate and lead citrate and analysed in a Philips CM10 electron microscope at 60 kV using a 30 μm objective aperture.
All experiments were performed at least three times and revealed comparable results. Differences between mean values of experimental and control groups were analysed by Student's t-test. A value of P < 0.05 was considered statistically significant.
This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG-SFB 766) and from the University of Tübingen (Center for Interdisciplinary Clinical Research, junior group program) to V. K.
- 1997) Bartonella spp. as emerging human pathogens. Clin Microbiol Rev 10: 203–219. , and (
- 1995) Bartonella henselae and Bartonella quintana adherence to and entry into cultured human epithelial cells. Infect Immun 63: 4553–4556. , , , , and (
- 1993) The Yersinia pseudotuberculosis adhesin YadA mediates intimate bacterial attachment to and entry into HEp-2 cells. Infect Immun 61: 3914–3921. , , and (
- 2008) A carcinoembryonic antigen-related cell adhesion molecule 1 homologue plays a pivotal role in nontypeable Haemophilus influenzae colonization of the chinchilla nasopharynx via the outer membrane protein P5-homologous adhesin. Infect Immun 76: 48–55. , , , , , and (
- 2002) NadA, a novel vaccine candidate of Neisseria meningitidis. J Exp Med 195: 1445–1454. , , , , , , et al. (
- 1996) New serotype of Bartonella henselae in endocarditis and cat-scratch disease. Lancet 347: 441–443. , , , , , and (
- 2006) Bacterial lipopolysaccharides induce HIF-1 activation in human monocytes via p44/42 MAPK and NF-kappaB. Biochem J 396: 517–527. , , , and (
- 2005) The Bartonella vinsonii subsp. arupensis immunodominant surface antigen BrpA gene, encoding a 382-kilodalton protein composed of repetitive sequences, is a member of a multigene family conserved among Bartonella species. Infect Immun 73: 3128–3136. , Jr, , , and (
- 2007) IFN-gamma amplifies NFkappaB-dependent Neisseria meningitidis invasion of epithelial cells via specific upregulation of CEA-related cell adhesion molecule 1. Cell Microbiol 9: 2968–2983. , , , and (
- 2007) A conserved glycine residue of trimeric autotransporter domains plays a key role in Yersinia adhesin A autotransport. J Bacteriol 189: 9011–9019. , , , , , , et al. (
- 2006) Identification of a domain in Yersinia virulence factor YadA that is crucial for extracellular matrix-specific cell adhesion and uptake. Proc Natl Acad Sci USA 103: 3375–3380. , and (
- 2000) Structure and sequence analysis of Yersinia YadA and Moraxella UspAs reveal a novel class of adhesins. EMBO J 19: 5989–5999. , , , , and (
- 1992) Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69: 11–25. (
- 2000) Interaction of Bartonella henselae with endothelial cells results in rapid bacterial rRNA synthesis and replication. Cell Microbiol 2: 431–441. , , , , , , et al. (
- 2001) Evidence of a leading role for VEGF in Bartonella henselae-induced endothelial cell proliferations. Cell Microbiol 3: 623–632. , , , , , , et al. (
- 2005) Activation of hypoxia-inducible factor-1 in bacillary angiomatosis: evidence for a role of hypoxia-inducible factor-1 in bacterial infections. Circulation 111: 1054–1062. , , , , , , et al. (
- 1995) Four new derivatives of the broad-host-range cloning vector pBBR1MCS, carrying different antibiotic-resistance cassettes. Gene 166: 175–176. , , , , , , et al. (
- 1980) Correlation of autoagglutination and virulence of Yersiniae. J Clin Microbiol 11: 430–432. , and (
- 2006) Trimeric autotransporter adhesins: variable structure, common function. Trends Microbiol 14: 264–270. , , , , and (
- 2008) A SacB mutagenesis strategy reveals that the Bartonella quintana variably expressed outer membrane proteins are required for bloodstream infection of the host. Infect Immun 76: 788–795. , , , , and (
- 2006) Structure of the outer membrane translocator domain of the Haemophilus influenzae Hia trimeric autotransporter. EMBO J 25: 2297–2304. , , and (
- 2007) The UspA1 protein of Moraxella catarrhalis induces CEACAM-1-dependent apoptosis in alveolar epithelial cells. J Infect Dis 195: 1651–1660. , , , , , , et al. (
- 2003) Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med 9: 677–684. , and (
- 1990) The agent of bacillary angiomatosis. An approach to the identification of uncultured pathogens. N Engl J Med 323: 1573–1580. , , , , and (
- 2004) Bartonella adhesin A mediates a proangiogenic host cell response. J Exp Med 200: 1267–1278. , , , , , , et al. (
- 2007) Analysis of Bartonella adhesin A expression reveals differences between various B. henselae strains. Infect Immun 75: 35–43. , , , , and (
- 2008) Analysis of a novel insect-cell culture medium-based growth medium for Bartonella species. Appl Environ Microbiol (in press). doi: 10.1128/AEM.00621-08. , , , , , , et al. (
- 2003) Molecular analysis of transport and oligomerization of the Yersinia enterocolitica adhesin YadA. J Bacteriol 185: 3735–3744. , , , , , and (
- 2007) Genomic analysis of Bartonella identifies type IV secretion systems as host adaptability factors. Nat Genet 39: 1469–1476. , , , , , , et al. (
- 2004) Yersinia enterocolitica adhesin A induces production of interleukin-8 in epithelial cells. Infect Immun 72: 6780–6789. , , , , , and (
- 2005) A bipartite signal mediates the transfer of type IV secretion substrates of Bartonella henselae into human cells. Proc Natl Acad Sci USA 102: 856–861. , , , , , , et al. (
- 2006) Bartonella quintana variably expressed outer membrane proteins mediate vascular endothelial growth factor secretion but not host cell adherence. Infect Immun 74: 5003–5013. , , , , , , et al. (
- 2008) Structure of the head of the Bartonella adhesin BadA. Plos Pathog (in press). , , , , , , et al. (
- 2001) YadA, the multifaceted Yersinia adhesin. Int J Med Microbiol 291: 209–218. , and (
- 1993) Hydrophobic domains affect the collagen-binding specificity and surface polymerization as well as the virulence potential of the YadA protein of Yersinia enterocolitica. Mol Microbiol 10: 995–1011. , , , , , and (
- 2008) Use of Bartonella adhesin A (BadA) immunoblotting in the serodiagnosis of Bartonella henselae infections. Int J Med Microbiol (in press). doi: 10.1016/j.ijmm.2008.01.013. , , , , , , et al. (
- 2006) Purification of the YadA membrane anchor for secondary structure analysis and crystallization. Int J Biol Macromol 39: 3–9. , , , and (
- 2004) Structural basis for host recognition by the Haemophilus influenzae Hia autotransporter. EMBO J 23: 1245–1256. , , , , III and (
- 2004) A family of variably expressed outer-membrane proteins (Vomp) mediates adhesion and autoaggregation in Bartonella quintana. Proc Natl Acad Sci USA 101: 13630–13635. , , , , , , et al. (