Activation of Hypoxia-Inducible Factor-1 in Bacillary Angiomatosis
http://www.100md.com
循环学杂志 2005年第3期
the Institut für Medizinische Mikrobiologie und Hygiene, Eberhard-Karls Universit;t, Tübingen, Germany (V.A.J.K., M.L., J.W., U.E., J.F., S.M., G.G., I.B.A.)
Molecular/Cancer Biology Laboratory and Ludwig Institute for Cancer Research, Biomedicum Helsinki, University of Helsinki, Helsinki, Finland (K.A.)
Institute of Physiology, Tübingen, Germany (S.H.)
Universit;tsklinik für Kinderheilkunde und Jugendmedizin, Tübingen, Germany (B.S.)
Klinik für Dermatologie und Allergologie, Ludwig Maximilians Universit;t, Munich, Germany (C.A.S., A.S.Y.)
Medizinische Universit;tsklinik III, Tübingen, Germany (B.B.).
Abstract
Background— Bartonella species are the only known bacterial pathogens causing vasculoproliferative disorders in humans (bacillary angiomatosis [BA]). Cellular and bacterial pathogenetic mechanisms underlying the induction of BA are largely unknown.
Methods and Results— Activation of hypoxia-inducible factor-1 (HIF-1), the key transcription factor involved in angiogenesis, was detected in Bartonella henselae–infected host cells in vitro by immunofluorescence, Western blotting, electrophoretic mobility shift, and reporter gene assays and by immunohistochemistry in BA tissue lesions in vivo. Gene microarray analysis revealed that a B henselae infection resulted in the activation of genes typical for the cellular response to hypoxia. HIF-1 was essential for B henselae–induced expression of vascular endothelial growth factor as shown by inhibition with the use of HIF-1–specific short-interfering RNA. Moreover, infection with B henselae resulted in increased oxygen consumption, cellular hypoxia, and decreased ATP levels in host cells. Infection with a pilus-negative variant of B henselae did not lead to cellular hypoxia or activation of HIF-1 or vascular endothelial growth factor secretion, suggesting a crucial role of this bacterial surface protein in the angiogenic reprogramming of the host cells.
Conclusions— B henselae induces a proangiogenic host cell response via HIF-1. Our data provide for the first time evidence that HIF-1 may play a role in bacterial infections.
Key Words: angiomatosis, bacillary ; Bartonella henselae ; angiogenesis ; HIF-1 protein ; hypoxia
Introduction
Angiogenesis is a multistep process resulting in the formation of new blood vessels from preexisting vasculature. Newly formed vessels supply oxygen and nutrients to growing tumors and are necessary for tumor progression and metastasis.1 Angiogenesis is also a component of various cardiovascular and inflammatory diseases.2 Hypoxia-inducible factor-1 (HIF-1) is a key transcription factor for the induction of angiogenic growth factors that adjust the vascular oxygen supply to tissue metabolic demands.3,4 Of the many genes induced by HIF-1, vascular endothelial growth factor (VEGF) plays a critical role in triggering angiogenesis as the major hypoxia-inducible mitogen for endothelial cells.5
Interestingly, human herpesvirus-8 (HHV-8) and several Bartonella species induce angiogenesis in humans. HHV-8 causes the vasculoproliferative disorder Kaposi’s sarcoma,6 which has a high frequency among immunocompromised patients, such as those infected with HIV. HHV-8–infected cells express VEGF on HIF-1 activation,7 and this mechanism was implicated in endothelial cell proliferation in the lesions.8
Bartonella henselae and B quintana are the etiologic agents of bacillary angiomatosis (BA) and bacillary peliosis (BP), which are histologically characterized as lobulated proliferation of mainly capillary-sized vessels and predominantly affect HIV patients.9 These slow-growing bacteria are facultative intracellular pathogens that, like HHV-8, also induce VEGF in host cells in vitro and in BA or BP lesions of patients.10 Endothelial cells are one presumed habitat of Bartonella.11 Dissecting the angioproliferative strategies used by B henselae is highly important in terms of understanding bacterial pathogenicity, angiogenesis, and pathogen-triggered tumor formation in humans.
Methods
Bacteria and Cell Culture
B henselae Marseille, Legionella pneumophila (ATCC 33216), HeLa229, and human umbilical vein endothelial cells (HUVECs) were infected as described.10 SB202190 (p38 inhibitor), PD98059 (MEK1 inhibitor), and parthenolide (nuclear factor-B [NF-B] inhibitor12) were purchased from Calbiochem; tumor necrosis factor- (TNF-) was purchased from R&D Systems.
Immunostaining and Confocal Laser Scanning Microscopy
Differential staining of extracellular and intracellular B henselae and confocal laser scanning microscopy were performed as described.11
Transmission Electron Microscopy
Transmission electron microscopy was performed as described.10
Lipopolysaccharide Preparation
Lipopolysaccharide (LPS) was prepared as described.13
RNA Isolation and Microarray Analysis
Cells (2x106) were grown for 24 hours and infected with B henselae. Total RNA was extracted with the use of the RNeasy mini kit (Qiagen). For the Affymetrix microarray analysis, 4 independent sample preparations and hybridizations were performed with the use of RNA extracted from 4 lots of cells isolated from different samples. Total RNA (5 μg) was used for the synthesis of double-stranded (ds) cDNA with the use of the Custom SuperScript ds-cDNA Synthesis Kit (Invitrogen). Biotin-labeled cRNA was prepared with the use of the Enzo BioArray HighYield RNA Transcript Labeling Kit (Affymetrix). The hybridization, washing, and staining of Human Genome 95Av2 microarrays were performed according to the manufacturer’s instructions (Affymetrix, GeneChip Expression Analysis technical manual). The probe arrays were scanned by using an Agilent GeneArray Scanner, and the readings from the quantitative scanning were analyzed by the Affymetrix Microarray Suite version 5.0, Affymetrix MicroDB 3.0, and Data Mining Tool version 3.0.
Reverse Transcription of mRNA and Polymerase Chain Reactions
The reverse transcription (RT) reactions (HeLa cells, HUVECs) were performed with the use of the SuperScript cDNA synthesis kit (Invitrogen) and oligo-d(pT)18 primers. The adrenomedullin (ADM), insulinlike growth factor binding protein-3 (IGFBP-3), stanniocalcin-2 (STC2), hexokinase-2 (HK2), cyclin G2 (CCNG2), N-myc downstream regulated gene-1 (NDRG1), KIAA0742 protein (KIAA0742), Homo sapiens cDNA (FLJ22182fis), interleukin-1; (IL-1;), intercellular adhesion molecule-1 (ICAM-1), monocyte chemoattractant protein-1 (MCP-1), granulocyte-macrophage colony-stimulating factor (GM-CSF), VEGF receptor-2 (VEGFR-2), and ;-actin cDNAs were amplified by conventional RT-PCR (see the Data Supplement Table). Quantification of VEGF14 and interleukin-8 (IL-8) mRNA (LightCycler Primer Sets, Search-LC) was performed by real-time PCR with a LightCycler System (Roche).
Immunoassays and Western Blotting
VEGF, IL-8, IGFBP-3, and ADM concentrations were measured from cell culture supernatants as described.15 For Western blot analysis of HK2, polyclonal antibodies (SantaCruz) were used. Phosphorylation of p38 and p42/44 mitogen-activated protein (MAP) kinases was evaluated with the use of a phospho-MAP kinase p38 and p42/44 kit (Cell Signaling).
Detection of HIF-1 by Immunofluorescence and Western Blotting
For HIF-1 immunostaining, HeLa cells were seeded on coverslips and infected with B henselae or exposed to hypoxia.16 Immunostaining of HIF-1 was performed with the use of anti–HIF-1 (Novus) and TRITC-conjugated anti-IgG antibodies. Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI).
For HIF Western blotting, cellular proteins were extracted,17 subjected to the Bradford assay (BioRad), blotted onto polyvinylidene difluoride membranes, and incubated with HIF-1–specific (Becton Dickinson) or HIF-2–specific (Novus) antibodies.
Electrophoretic Mobility Shift Assay
Nuclear extracts were prepared from HeLa cells as described,18 and protein concentrations were assessed via the Bradford assay (Biorad). Oligonucleotide probes (NF-B consensus [NF-Bc 5'-AGT TGA GGG GAC TTT CCC AGG C-3'; Santa Cruz], HIF-1 consensus [HIF-1c 5'-GCC CTA CGT GCT GTC TCA-3', Metabion]) were labeled with [32P]ATP (Amersham). In competition experiments, extracts were incubated with the labeled probes in the presence of a 100-fold excess of unlabeled oligonucleotides. Antibodies against p50 (Santa Cruz) were included in the binding reaction for supershift analysis.
DNA Constructs, Transient Transfection, and Determination of Reporter Gene Activity
For reporter gene assays, a VEGF promoter/luciferase reporter construct (pVEGF.4) was used.16 For normalization, cotransfections were performed with the use of pCMV ;-galactosidase (pCMV ;-gal; Clontech). HeLa cells were transiently transfected with the use of ExGen500 (Fermentas) and incubated for 24 hours at 37°C. Transfected cells were infected with B henselae or exposed to hypoxia. After 36 hours, cells were lysed for determination of Luc activity, protein determination, and measurement of ;-galactosidase with the use of a Luciferase Reporter Gene Assay (Roche). Levels of Luc expression were normalized to ;-galactosidase activity and total protein concentration. Every experiment was done in quadruplicate. The degree of induction was determined as the ratio of Luc activity of B henselae–infected or hypoxia-exposed cells to that of uninfected control cells.
Gene Silencing Using HIF-1 Short-Interfering RNA
Cells were seeded in 6-well plates, and the short-interfering RNA (siRNA) duplexes [HIF-1 sense r(CUGAUGACCAGCAACUUGA)d(TT), antisense r(UCAAGUUGCUGGUCAUCAG)d(TT)19; control sense r(UUCUCCGAACGUGUCACGU)d(TT), antisense r(ACGUGACACGUUCGGAGAA)d(TT)] were diluted at a final concentration of 200 nmol/L. RNAiFect (Qiagen) was used for transfection. Cell culture medium was changed, and the cells were incubated overnight before infection.
Immunohistochemistry
Sections of histologically confirmed paraffin-embedded specimens of BA were stained with the use of specific anti-HIF-1 antibodies (Novus) and the alkaline phosphatase/anti–alkaline phosphatase technique with hematoxylin counterstaining. Unaffected skin was used as control.
Detection of Cellular Hypoxia and Oxygen Consumption
Cellular hypoxia was detected by adding pimonidazole hydrochloride (200 mmol/L; Hypoxyprobe-1, Chemicon)20 to HeLa cells that were subsequently infected or exposed to hypoxia for 4 hours. Staining was performed according to the manufacturer’s instructions.
The dissolved O2 concentration was directly measured in the medium of B henselae–infected and control cell cultures (for details, see Data Supplement Figure I). An O2-tight closed chamber was built around the head of a Clark electrode (Conrad Electronics). HeLa cells grown on coverslips were infected for 4 hours with B henselae and placed in the chamber filled with RPMI 1640 medium saturated with CO2 (5%) and O2 (20.8%). The oxygen electrode was allowed to equilibrate for 1 minute at 37°C followed by a 10-minute period of oxygen measurement. O2 concentration (mg/L) was determined at 60-second intervals.
Measurement of Intracellular ATP
For quantification of cellular ATP, HeLa cells were infected with B henselae or exposed to hypoxia for 4 hours, and the intracellular ATP was determined with the use of the CLSII Assay Kit (Roche). ATP levels were normalized to the total protein concentration of the cellular extracts.
Statistical Analysis
All data shown are from representative experiments. Differences between mean values were analyzed with the Student t test. P<0.05 was considered statistically significant.
Results
B henselae Induces a Proangiogenic Gene Program
For analysis of the genomic response to B henselae infection, infected and uninfected HeLa cells were compared in RNA microarray analysis 6 hours after infection. Confocal laser scanning microscopy showed that at this time >99% of the cells contained bacteria (Data Supplement Figure II). Among 12 000 genes analyzed, 20 genes were found to be upregulated >2-fold on infection. As indicated in the Table, at least 7 of these genes (eg, IL-8, ADM, ephrin A1 [EFNA1], VEGF) are known to play a role in angiogenesis and vessel maturation, 3 are involved in glycolysis, and 7 are involved in signal transduction and cell growth. The function of 3 of the induced genes is currently unknown. Gene downregulation was not detected. Of the 20 genes, 14 of are known to be directly or indirectly regulated via HIF-1, suggesting that the major trigger of angiogenic gene induction on B henselae infection was HIF-1. IL-8 was an exception; this highly induced gene is strongly induced by NF-B but not by HIF-1.21 Quantitative real-time PCR or semiquantitative RT-PCR was used to confirm results of the gene array experiments (Figure 1, Data Supplement Figure III). As a test for possible false-negative results, selected NF-B–regulated genes (ICAM-1, MCP-1, IL-1;, GM-CSF) or the HIF-2–regulated VEGFR-2 gene were not affected on B henselae infection. Moreover, increased protein expression was shown for 5 of the induced genes (VEGF, IL-8, IGFBP-3, ADM, and HK2; Figure 1). Results were consistent with the microarray data indicating a significant change toward a proangiogenic gene program in host cells.
Upregulated Genes in HeLa Cells 6 Hours on B henselae Infection
B henselae Induces HIF-1, Cellular Hypoxia, and Decrease of ATP
The induced gene pattern suggested that HIF-1 activation has a crucial role in the host cell response to B henselae infection. We therefore investigated the activation of HIF-1 by B henselae in HeLa cells (Figure 2). Immunofluorescence analysis revealed the appearance of nuclear HIF-1 6 hours after B henselae infection, similar to hypoxia-treated cells (Figure 2a), and consistent results were obtained by Western blotting (Figure 2b). Transfection experiments with the use of a VEGF promoter/luciferase reporter that is specifically regulated by HIF-1 revealed that B henselae increases VEGF gene transcription 4-fold compared with 8-fold stimulation by hypoxia (Figure 2c). Nuclear HIF-1 activation was also demonstrated by electrophoretic mobility shift assay (EMSA), which indicated the enhanced capacity of the induced HIF-1 complex to bind to its target consensus DNA sequences (Figure 2d). Finally, activation of HIF-1 and subsequent induction of VEGF secretion after B henselae infection was completely inhibited when the cells were treated with HIF-1 siRNA before infection (Figure 2e, 2f). These results clearly indicate that HIF-1 is activated on a B henselae infection and that the increased VEGF secretion on B henselae infection is dependent on HIF-1.
Endothelial cells that were heavily infected by B henselae (Data Supplement Figure II) also upregulated VEGF, IL-8, ADM, and HK2 after B henselae infection (Figure 3). Moreover, VEGFR-2, known to be regulated via HIF-2,22 and the NF-B–regulated MCP-1 and GM-CSF genes were induced. Accordingly, HIF-1 and HIF-2 stabilization was observed in HUVECs (Figure 3b). Moreover, HIF-1 activation and VEGF secretion occurred in all cells investigated thus far (A549, MCF-7, HT29, ECC-1 [data not shown], EA.hy 926, MonoMac610,23). These experiments indicate that a B henselae infection generally results in HIF activation and the induction of a proangiogenic gene program.
Because of the strong IL-8 induction, we also confirmed NF-B activation by EMSA (Figure 4a). To dissect the NF-B–regulated IL-8 from the HIF-1–regulated proangiogenic response, the infection process was modulated by the NF-B inhibitor parthenolide12 (Figure 4b). This inhibited IL-8 secretion from B henselae–infected cells completely but had only little effect on VEGF secretion. This and the data obtained with HIF-1 siRNA strongly suggest that NF-B is responsible for IL-8 induction, whereas HIF-1 triggers VEGF induction. The MAP kinases p38 and p42/44, which were rapidly and persistently phosphorylated in B henselae–infected cells (Figure 4c), contribute only little to the B henselae–induced VEGF secretion because preincubation with PD98059 (p42/44 inhibitor) or SB202190 (p38 inhibitor) caused only a slight reduction of the VEGF levels (Figure 4d). These results suggest that activation of HIF-1 on a Bartonella infection occurs via hypoxic activation, presumably involving inhibition of prolyl and asparaginyl hydroxylases.4
Because hypoxia is known to be a potent activator of HIF-1,4 we investigated whether infection with B henselae leads to an increased O2 consumption and cellular hypoxia. Experiments quantifying the dissolved O2 concentration in a closed cell culture chamber and with the use of a hypoxia-sensitive dye indicated that both processes indeed occurred (Figure 5a to 5c). O2 consumption was increased 9-fold in the B henselae–infected cells. Cellular hypoxia did not result from bacterial overgrowth because the pH of cell culture supernatants was unaffected in the same experiments (control 7.52, B henselae 7.51, hypoxia 7.28). In addition, infection resulted in decreased ATP levels in host cells (Figure 5d).
Expression of B henselae Pili Is Crucial for Activation of HIF-1
To analyze whether pilus expression determines the activation of HIF-1, HeLa cells were infected with wild-type (wt) or pilus-negative (Pil–) B henselae (Figure 6a). Analysis of LPS composition disclosed no differences between both strains (Figure 6b). Pilus expression was crucial for activation of HIF-1 (Figure 6c), VEGF expression (Figure 6d), and cellular hypoxia (Figure 6e), suggesting that pili are necessary for inducing an angiogenic host cell response by B henselae. Moreover, HIF-1 activation was not observed when cells were exposed to B henselae LPS or when infected with L pneumophila (Data Supplement Figures IV and V).
HIF-1 Is Activated in BA Lesions From Patients
To investigate whether HIF-1 activation also occurs in vivo in Bartonella-infected tissues, sections from BA lesions were stained for HIF-1 (Figure 7). Nuclear HIF-1 was stained strongly in cells identified as histiocytes or macrophages infiltrating BA lesions. These observations are consistent with previously published results of HIF-1 staining in hypoxic tumor tissues17 and with recent data showing that HIF-1 is crucial in myeloid cell functions associated with inflammation and bacterial defense functions.24 From this data we conclude that HIF-1 induction occurs in B henselae infections both in vitro and in vivo.
Discussion
Our present studies reveal a novel mechanism of angiogenesis associated with a bacterial infection. We show here that infection with B henselae, the causative agent of angiogenic proliferation of capillary vessels in BA, results in HIF-1 activation and leads to a gene expression pattern typical for the cellular response to hypoxia. In accordance, nuclear HIF-1 was detected in Bartonella-infected cells in vitro and in vivo in BA patient samples. Our results suggest that activation of a HIF-1–dependent cell response on a Bartonella infection causes the vasculoproliferative disorder in BA and BP.
Angiogenesis research has been focused mainly on cancer and cardiovascular diseases.1 The HIF-1–induced angiogenic gene reprogramming would indeed be an excellent target for therapeutic intervention in solid tumors, including hemangioblastomas, and also in Kaposi’s sarcoma.4,7 On the other hand, activation of the HIF pathway could be of therapeutic value in tissue ischemia associated with cardiovascular diseases.1 However, HIF activation in bacterial infections has not been described previously, although genetic deletion of HIF-1 in myeloid cells compromises inflammatory responses.24 Bartonella species are the only known bacterial pathogens causing vasculoproliferative disorders in humans.9 VEGF induction was described in B henselae infections in vivo and in vitro,10 and a role has also been suggested for macrophage-secreted angiogenic agents.25
Only 20 genes were found to be significantly upregulated on B henselae infection in our microarray analysis. Of these, 14 are known HIF targets.3 Several of these were confirmed by independent assays, including VEGF, which is known to be a key factor in the triggering of angiogenic processes,5 and STC2, ADM, IGFBP-3, and HK2, which have also been implicated in hypoxic cellular responses or angiogenic processes.26–30 IL-8, which is apparently the only NF-B–regulated gene in our B henselae infection model of HeLa cells, is also known to be a strong vasculoproliferative agent.31 We hypothesize that the induced growth factors and cytokines secreted by the infected host cells contribute to an angiogenic cocktail responsible for endothelial proliferation in BA or BP, whereas the induction of HIF-regulated glycolytic genes may be linked to the metabolic demands of the B henselae–infected host cells (Figure 8).
Interestingly, some of the induced genes have been reported to be upregulated via HIF-1 or hypoxia in various types of human tumors (eg, enolase-2 [ENO2], NDRG-1, musculoaponeurotic fibrosarcoma oncogene homolog f [MAFF], nuclear factor IL-3 regulated [NFIL3], cyclin G2 [CCNG2], p21-activated kinase-2 [PAK2], and the hypothetical protein FLJ20500A HREF="#R27-160629">27,30,32). The gene expression pattern in Bartonella-infected cells thus shows striking similarities with that in other angiogenesis models.26,27,30 Of interest is the fact that besides the angiogenic cytokine IL-8, the response in HeLa cells lacked the typical proinflammatory genes that are found to be induced via NF-B in other bacterial infections.33 However, the induction of a broad spectrum of proinflammatory genes and cytokines could be deleterious for the facultative intracellular bacteria.11 In accordance, we did not find a broad activation of NF-B–regulated proinflammatory cytokines. These cytokines are commonly induced on bacterial infection of endothelial cells, possibly because of LPS-mediated signals. However, HeLa cells, in contrast to endothelial cells, do not express toll-like receptor-4, which is required for these signals.34 Finally, the HIF-2 activation in HUVECs is a cell type–specific response that upregulates VEGFR-2 and may amplify the VEGF-driven angiogenic loop by making endothelial cells more sensitive to the secreted VEGF.
HIF-1 activation was observed in vitro and in patient samples from BA. This is consistent with our previous observations regarding increased VEGF expression in BA and BP lesions.10 Macrophages that infiltrate BA lesions were highly positive for HIF-1 and they are known to secrete VEGF on Bartonella infection in vitro.25 How HIF-1 is activated on a B henselae infection is not yet clear. The intracellular bacteria should consume a lot of oxygen, and the resulting hypoxia is known to inhibit the prolyl hydroxylases that target HIF-1 to ubiquitin-mediated degradation.4 In fact, infection with B henselae resulted in increased O2 consumption and hypoxia in the host cells and was accompanied by an increased consumption of cellular ATP. Furthermore, the MAP kinases p42/44 and p38 were phosphorylated on B henselae infection. However, although these kinases have been reported to enhance the transcriptional activity of HIF-1 by phosphorylation (p42/44)35 and to increase VEGF mRNA stability (p38),36 these specific MAP kinase inhibitors induced only a small decrease of the B henselae–triggered VEGF secretion in our model. Stabilization of HIF-1 on a B henselae infection could thus occur by a mechanism similar to that in cells exposed to hypoxia.37
Interestingly, the B henselae Pil– variant did not induce cellular hypoxia, activation of HIF-1, or VEGF expression (Figure 6). Recently, we identified this "pilus" as the nonfimbrial adhesin BadA.15 When one considers this and the observation that HIF-1 activation does not occur in LPS-exposed cells, it is highly unlikely that LPS is involved in the activation of HIF-1 by B henselae. Additionally, HIF-1 activation did not occur when cells were infected with L pneumophila. This is not surprising because L pneumophila infections do not result in vasculoproliferative disorders but instead in inflammatory disease. Our results suggest that Bartonella-specific bacterial mechanisms are necessary for the induction of the angiogenic response. The observation that Bartonella induces an HIF-1–driven cellular phenotype is exciting and provides a new aspect to angiogenesis research. Whether similar strategies are shared by other bacterial pathogens needs to be elucidated further.
Acknowledgments
This work was supported by grants from the Deutsche Forschungsgemeinschaft and from the University of Tübingen (Fortüne Programm). The authors thank A. Schaefer for excellent technical assistance, L. Poellinger for kind advice, and U. Z;hringer for providing Bartonella LPS.
Footnotes
The first 3 authors contributed equally to this work.
The online-only Data Supplement, which contains Figures I through V and a table, can be found with this article at http://www.circulationaha.org.
References
Ferrara N, Alitalo K. Clinical applications of angiogenic growth factors and their inhibitors. Nat Med. 1999; 5: 1359–1364.
Risau W. Mechanisms of angiogenesis. Nature. 1997; 386: 671–674.
Semenza GL. HIF-1, O(2), and the 3 PHDs: how animal cells signal hypoxia to the nucleus. Cell. 2001; 107: 1–3.
Pugh CW, Ratcliffe PJ. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med. 2003; 9: 677–684.
Leung DW, Cachianes G, Kuang WJ, et al. Vascular endothelial growth factor is a secreted angiogenic mitogen. Science. 1989; 246: 1306–1309.
Chang Y, Cesarman E, Pessin MS, et al. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi’s sarcoma. Science. 1994; 266: 1865–1869.
Sodhi A, Montaner S, Patel V, et al. The Kaposi’s sarcoma-associated herpes virus G protein–coupled receptor up-regulates vascular endothelial growth factor expression and secretion through mitogen-activated protein kinase and p38 pathways acting on hypoxia-inducible factor 1alpha. Cancer Res. 2000; 60: 4873–4880.
Flore O, Rafii S, Ely S, et al. Transformation of primary human endothelial cells by Kaposi’s sarcoma-associated herpesvirus. Nature. 1998; 394: 588–592.
Relman DA, Falkow S, LeBoit PE, et al. The organism causing bacillary angiomatosis, peliosis hepatis, and fever and bacteremia in immunocompromised patients. N Engl J Med. 1991; 324: 1514.
Kempf VA, Volkmann B, Schaller M, et al. Evidence of a leading role for VEGF in Bartonella henselae–induced endothelial cell proliferations. Cell Microbiol. 2001; 3: 623–632.
Kempf VA, Schaller M, Behrendt S, et al. Interaction of Bartonella henselae with endothelial cells results in rapid bacterial rRNA synthesis and replication. Cell Microbiol. 2000; 2: 431–441.
Hehner SP, Heinrich M, Bork PM, et al. Sesquiterpene lactones specifically inhibit activation of NF-kappa B by preventing the degradation of I kappa B-alpha and I kappa B-beta. J Biol Chem. 1998; 273: 1288–1297.
Zahringer U, Lindner B, Knirel YA, et al. Structure and biological activity of the short-chain lipopolysaccharide from Bartonella henselae ATCC 49882T. J Biol Chem. 2004; 279: 21046–21054.
Wellmann S, Taube T, Paal K, et al. Specific reverse transcription-PCR quantification of vascular endothelial growth factor (VEGF) splice variants by LightCycler technology. Clin Chem. 2001; 47: 654–660.
Riess T, Andersson SG, Lupas A, et al. Bartonella adhesin A mediates a proangiogenic host cell response. J Exp Med. 2004; 200: 1267–1278.
Ikeda E, Achen MG, Breier G, et al. Hypoxia-induced transcriptional activation and increased mRNA stability of vascular endothelial growth factor in C6 glioma cells. J Biol Chem. 1995; 270: 19761–19766.
Wiesener MS, Munchenhagen PM, Berger I, et al. Constitutive activation of hypoxia-inducible genes related to overexpression of hypoxia-inducible factor-1alpha in clear cell renal carcinomas. Cancer Res. 2001; 61: 5215–5222.
Camenisch G, Wenger RH, Gassmann M. DNA-binding activity of hypoxia-inducible factors (HIFs). Methods Mol Biol. 2002; 196: 117–129.
Sowter HM, Raval RR, Moore JW, et al. Predominant role of hypoxia-inducible transcription factor (HIF)-1alpha versus HIF-2alpha in regulation of the transcriptional response to hypoxia. Cancer Res. 2003; 63: 6130–6134.
Raleigh JA, Calkins-Adams DP, Rinker LH, et al. Hypoxia and vascular endothelial growth factor expression in human squamous cell carcinomas using pimonidazole as a hypoxia marker. Cancer Res. 1998; 58: 3765–3768.
Kunsch C, Rosen CA. NF-kappa B subunit-specific regulation of the interleukin-8 promoter. Mol Cell Biol. 1993; 13: 6137–6146.
Elvert G, Kappel A, Heidenreich R, et al. Cooperative interaction of hypoxia-inducible factor-2alpha (HIF-2alpha) and Ets-1 in the transcriptional activation of vascular endothelial growth factor receptor-2 (Flk-1). J Biol Chem. 2003; 278: 7520–3750.
Kempf VA, Schairer A, Neumann D, et al. Bartonella henselae inhibits apoptosis in Mono Mac 6 cells. Cell Microbiol. 2005; 7: 91–104.
Cramer T, Yamanishi Y, Clausen BE, et al. HIF-1alpha is essential for myeloid cell-mediated inflammation. Cell. 2003; 112: 645–657.
Resto-Ruiz SI, Schmiederer M, Sweger D, et al. Induction of a potential paracrine angiogenic loop between human THP-1 macrophages and human microvascular endothelial cells during Bartonella henselae infection. Infect Immun. 2002; 70: 4564–4570.
Kahn J, Mehraban F, Ingle G, et al. Gene expression profiling in an in vitro model of angiogenesis. Am J Pathol. 2000; 156: 1887–1900.
Leonard MO, Cottell DC, Godson C, et al. The role of HIF-1alpha in transcriptional regulation of the proximal tubular epithelial cell response to hypoxia. J Biol Chem. 2003; 278: 40296–40304.
Oehler MK, Hague S, Rees MC, et al. Adrenomedullin promotes formation of xenografted endometrial tumors by stimulation of autocrine growth and angiogenesis. Oncogene. 2002; 21: 2815–2821.
Schmid MC, Bisoffi M, Wetterwald A, et al. Insulin-like growth factor binding protein-3 is overexpressed in endothelial cells of mouse breast tumor vessels. Int J Cancer. 2003; 103: 577–586.
Lal A, Peters H, St Croix B, et al. Transcriptional response to hypoxia in human tumors. J Natl Cancer Inst. 2001; 93: 1337–1343.
Koch AE, Polverini PJ, Kunkel SL, et al. Interleukin-8 as a macrophage-derived mediator of angiogenesis. Science. 1992; 258: 1798–1801.
Feldenzer JA, McKeever PE. Selective localization of gamma-enolase in stromal cells of cerebellar hemangioblastomas. Acta Neuropathol (Berl). 1987; 72: 281–285.
Bohn E, Muller S, Lauber J, et al. Gene expression patterns of epithelial cells modulated by pathogenicity factors of Yersinia enterocolitica. Cell Microbiol. 2004; 6: 129–141.
Pridmore AC, Wyllie DH, Abdillahi F, et al. A lipopolysaccharide-deficient mutant of Neisseria meningitidis elicits attenuated cytokine release by human macrophages and signals via toll-like receptor (TLR) 2 but not via TLR4/MD2. J Infect Dis. 2001; 183: 89–96.
Richard DE, Berra E, Gothie E, et al. p42/p44 mitogen-activated protein kinases phosphorylate hypoxia-inducible factor 1alpha (HIF-1alpha) and enhance the transcriptional activity of HIF-1. J Biol Chem. 1999; 274: 32631–32637.
Pages G, Berra E, Milanini J, et al. Stress-activated protein kinases (JNK and p38/HOG) are essential for vascular endothelial growth factor mRNA stability. J Biol Chem. 2000; 275: 26484–26491.
Berra E, Benizri E, Ginouves A, et al. HIF prolyl-hydroxylase 2 is the key oxygen sensor setting low steady-state levels of HIF-1alpha in normoxia. EMBO J. 2003; 22: 4082–4090.(Volkhard A.J. Kempf, MD; )
Molecular/Cancer Biology Laboratory and Ludwig Institute for Cancer Research, Biomedicum Helsinki, University of Helsinki, Helsinki, Finland (K.A.)
Institute of Physiology, Tübingen, Germany (S.H.)
Universit;tsklinik für Kinderheilkunde und Jugendmedizin, Tübingen, Germany (B.S.)
Klinik für Dermatologie und Allergologie, Ludwig Maximilians Universit;t, Munich, Germany (C.A.S., A.S.Y.)
Medizinische Universit;tsklinik III, Tübingen, Germany (B.B.).
Abstract
Background— Bartonella species are the only known bacterial pathogens causing vasculoproliferative disorders in humans (bacillary angiomatosis [BA]). Cellular and bacterial pathogenetic mechanisms underlying the induction of BA are largely unknown.
Methods and Results— Activation of hypoxia-inducible factor-1 (HIF-1), the key transcription factor involved in angiogenesis, was detected in Bartonella henselae–infected host cells in vitro by immunofluorescence, Western blotting, electrophoretic mobility shift, and reporter gene assays and by immunohistochemistry in BA tissue lesions in vivo. Gene microarray analysis revealed that a B henselae infection resulted in the activation of genes typical for the cellular response to hypoxia. HIF-1 was essential for B henselae–induced expression of vascular endothelial growth factor as shown by inhibition with the use of HIF-1–specific short-interfering RNA. Moreover, infection with B henselae resulted in increased oxygen consumption, cellular hypoxia, and decreased ATP levels in host cells. Infection with a pilus-negative variant of B henselae did not lead to cellular hypoxia or activation of HIF-1 or vascular endothelial growth factor secretion, suggesting a crucial role of this bacterial surface protein in the angiogenic reprogramming of the host cells.
Conclusions— B henselae induces a proangiogenic host cell response via HIF-1. Our data provide for the first time evidence that HIF-1 may play a role in bacterial infections.
Key Words: angiomatosis, bacillary ; Bartonella henselae ; angiogenesis ; HIF-1 protein ; hypoxia
Introduction
Angiogenesis is a multistep process resulting in the formation of new blood vessels from preexisting vasculature. Newly formed vessels supply oxygen and nutrients to growing tumors and are necessary for tumor progression and metastasis.1 Angiogenesis is also a component of various cardiovascular and inflammatory diseases.2 Hypoxia-inducible factor-1 (HIF-1) is a key transcription factor for the induction of angiogenic growth factors that adjust the vascular oxygen supply to tissue metabolic demands.3,4 Of the many genes induced by HIF-1, vascular endothelial growth factor (VEGF) plays a critical role in triggering angiogenesis as the major hypoxia-inducible mitogen for endothelial cells.5
Interestingly, human herpesvirus-8 (HHV-8) and several Bartonella species induce angiogenesis in humans. HHV-8 causes the vasculoproliferative disorder Kaposi’s sarcoma,6 which has a high frequency among immunocompromised patients, such as those infected with HIV. HHV-8–infected cells express VEGF on HIF-1 activation,7 and this mechanism was implicated in endothelial cell proliferation in the lesions.8
Bartonella henselae and B quintana are the etiologic agents of bacillary angiomatosis (BA) and bacillary peliosis (BP), which are histologically characterized as lobulated proliferation of mainly capillary-sized vessels and predominantly affect HIV patients.9 These slow-growing bacteria are facultative intracellular pathogens that, like HHV-8, also induce VEGF in host cells in vitro and in BA or BP lesions of patients.10 Endothelial cells are one presumed habitat of Bartonella.11 Dissecting the angioproliferative strategies used by B henselae is highly important in terms of understanding bacterial pathogenicity, angiogenesis, and pathogen-triggered tumor formation in humans.
Methods
Bacteria and Cell Culture
B henselae Marseille, Legionella pneumophila (ATCC 33216), HeLa229, and human umbilical vein endothelial cells (HUVECs) were infected as described.10 SB202190 (p38 inhibitor), PD98059 (MEK1 inhibitor), and parthenolide (nuclear factor-B [NF-B] inhibitor12) were purchased from Calbiochem; tumor necrosis factor- (TNF-) was purchased from R&D Systems.
Immunostaining and Confocal Laser Scanning Microscopy
Differential staining of extracellular and intracellular B henselae and confocal laser scanning microscopy were performed as described.11
Transmission Electron Microscopy
Transmission electron microscopy was performed as described.10
Lipopolysaccharide Preparation
Lipopolysaccharide (LPS) was prepared as described.13
RNA Isolation and Microarray Analysis
Cells (2x106) were grown for 24 hours and infected with B henselae. Total RNA was extracted with the use of the RNeasy mini kit (Qiagen). For the Affymetrix microarray analysis, 4 independent sample preparations and hybridizations were performed with the use of RNA extracted from 4 lots of cells isolated from different samples. Total RNA (5 μg) was used for the synthesis of double-stranded (ds) cDNA with the use of the Custom SuperScript ds-cDNA Synthesis Kit (Invitrogen). Biotin-labeled cRNA was prepared with the use of the Enzo BioArray HighYield RNA Transcript Labeling Kit (Affymetrix). The hybridization, washing, and staining of Human Genome 95Av2 microarrays were performed according to the manufacturer’s instructions (Affymetrix, GeneChip Expression Analysis technical manual). The probe arrays were scanned by using an Agilent GeneArray Scanner, and the readings from the quantitative scanning were analyzed by the Affymetrix Microarray Suite version 5.0, Affymetrix MicroDB 3.0, and Data Mining Tool version 3.0.
Reverse Transcription of mRNA and Polymerase Chain Reactions
The reverse transcription (RT) reactions (HeLa cells, HUVECs) were performed with the use of the SuperScript cDNA synthesis kit (Invitrogen) and oligo-d(pT)18 primers. The adrenomedullin (ADM), insulinlike growth factor binding protein-3 (IGFBP-3), stanniocalcin-2 (STC2), hexokinase-2 (HK2), cyclin G2 (CCNG2), N-myc downstream regulated gene-1 (NDRG1), KIAA0742 protein (KIAA0742), Homo sapiens cDNA (FLJ22182fis), interleukin-1; (IL-1;), intercellular adhesion molecule-1 (ICAM-1), monocyte chemoattractant protein-1 (MCP-1), granulocyte-macrophage colony-stimulating factor (GM-CSF), VEGF receptor-2 (VEGFR-2), and ;-actin cDNAs were amplified by conventional RT-PCR (see the Data Supplement Table). Quantification of VEGF14 and interleukin-8 (IL-8) mRNA (LightCycler Primer Sets, Search-LC) was performed by real-time PCR with a LightCycler System (Roche).
Immunoassays and Western Blotting
VEGF, IL-8, IGFBP-3, and ADM concentrations were measured from cell culture supernatants as described.15 For Western blot analysis of HK2, polyclonal antibodies (SantaCruz) were used. Phosphorylation of p38 and p42/44 mitogen-activated protein (MAP) kinases was evaluated with the use of a phospho-MAP kinase p38 and p42/44 kit (Cell Signaling).
Detection of HIF-1 by Immunofluorescence and Western Blotting
For HIF-1 immunostaining, HeLa cells were seeded on coverslips and infected with B henselae or exposed to hypoxia.16 Immunostaining of HIF-1 was performed with the use of anti–HIF-1 (Novus) and TRITC-conjugated anti-IgG antibodies. Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI).
For HIF Western blotting, cellular proteins were extracted,17 subjected to the Bradford assay (BioRad), blotted onto polyvinylidene difluoride membranes, and incubated with HIF-1–specific (Becton Dickinson) or HIF-2–specific (Novus) antibodies.
Electrophoretic Mobility Shift Assay
Nuclear extracts were prepared from HeLa cells as described,18 and protein concentrations were assessed via the Bradford assay (Biorad). Oligonucleotide probes (NF-B consensus [NF-Bc 5'-AGT TGA GGG GAC TTT CCC AGG C-3'; Santa Cruz], HIF-1 consensus [HIF-1c 5'-GCC CTA CGT GCT GTC TCA-3', Metabion]) were labeled with [32P]ATP (Amersham). In competition experiments, extracts were incubated with the labeled probes in the presence of a 100-fold excess of unlabeled oligonucleotides. Antibodies against p50 (Santa Cruz) were included in the binding reaction for supershift analysis.
DNA Constructs, Transient Transfection, and Determination of Reporter Gene Activity
For reporter gene assays, a VEGF promoter/luciferase reporter construct (pVEGF.4) was used.16 For normalization, cotransfections were performed with the use of pCMV ;-galactosidase (pCMV ;-gal; Clontech). HeLa cells were transiently transfected with the use of ExGen500 (Fermentas) and incubated for 24 hours at 37°C. Transfected cells were infected with B henselae or exposed to hypoxia. After 36 hours, cells were lysed for determination of Luc activity, protein determination, and measurement of ;-galactosidase with the use of a Luciferase Reporter Gene Assay (Roche). Levels of Luc expression were normalized to ;-galactosidase activity and total protein concentration. Every experiment was done in quadruplicate. The degree of induction was determined as the ratio of Luc activity of B henselae–infected or hypoxia-exposed cells to that of uninfected control cells.
Gene Silencing Using HIF-1 Short-Interfering RNA
Cells were seeded in 6-well plates, and the short-interfering RNA (siRNA) duplexes [HIF-1 sense r(CUGAUGACCAGCAACUUGA)d(TT), antisense r(UCAAGUUGCUGGUCAUCAG)d(TT)19; control sense r(UUCUCCGAACGUGUCACGU)d(TT), antisense r(ACGUGACACGUUCGGAGAA)d(TT)] were diluted at a final concentration of 200 nmol/L. RNAiFect (Qiagen) was used for transfection. Cell culture medium was changed, and the cells were incubated overnight before infection.
Immunohistochemistry
Sections of histologically confirmed paraffin-embedded specimens of BA were stained with the use of specific anti-HIF-1 antibodies (Novus) and the alkaline phosphatase/anti–alkaline phosphatase technique with hematoxylin counterstaining. Unaffected skin was used as control.
Detection of Cellular Hypoxia and Oxygen Consumption
Cellular hypoxia was detected by adding pimonidazole hydrochloride (200 mmol/L; Hypoxyprobe-1, Chemicon)20 to HeLa cells that were subsequently infected or exposed to hypoxia for 4 hours. Staining was performed according to the manufacturer’s instructions.
The dissolved O2 concentration was directly measured in the medium of B henselae–infected and control cell cultures (for details, see Data Supplement Figure I). An O2-tight closed chamber was built around the head of a Clark electrode (Conrad Electronics). HeLa cells grown on coverslips were infected for 4 hours with B henselae and placed in the chamber filled with RPMI 1640 medium saturated with CO2 (5%) and O2 (20.8%). The oxygen electrode was allowed to equilibrate for 1 minute at 37°C followed by a 10-minute period of oxygen measurement. O2 concentration (mg/L) was determined at 60-second intervals.
Measurement of Intracellular ATP
For quantification of cellular ATP, HeLa cells were infected with B henselae or exposed to hypoxia for 4 hours, and the intracellular ATP was determined with the use of the CLSII Assay Kit (Roche). ATP levels were normalized to the total protein concentration of the cellular extracts.
Statistical Analysis
All data shown are from representative experiments. Differences between mean values were analyzed with the Student t test. P<0.05 was considered statistically significant.
Results
B henselae Induces a Proangiogenic Gene Program
For analysis of the genomic response to B henselae infection, infected and uninfected HeLa cells were compared in RNA microarray analysis 6 hours after infection. Confocal laser scanning microscopy showed that at this time >99% of the cells contained bacteria (Data Supplement Figure II). Among 12 000 genes analyzed, 20 genes were found to be upregulated >2-fold on infection. As indicated in the Table, at least 7 of these genes (eg, IL-8, ADM, ephrin A1 [EFNA1], VEGF) are known to play a role in angiogenesis and vessel maturation, 3 are involved in glycolysis, and 7 are involved in signal transduction and cell growth. The function of 3 of the induced genes is currently unknown. Gene downregulation was not detected. Of the 20 genes, 14 of are known to be directly or indirectly regulated via HIF-1, suggesting that the major trigger of angiogenic gene induction on B henselae infection was HIF-1. IL-8 was an exception; this highly induced gene is strongly induced by NF-B but not by HIF-1.21 Quantitative real-time PCR or semiquantitative RT-PCR was used to confirm results of the gene array experiments (Figure 1, Data Supplement Figure III). As a test for possible false-negative results, selected NF-B–regulated genes (ICAM-1, MCP-1, IL-1;, GM-CSF) or the HIF-2–regulated VEGFR-2 gene were not affected on B henselae infection. Moreover, increased protein expression was shown for 5 of the induced genes (VEGF, IL-8, IGFBP-3, ADM, and HK2; Figure 1). Results were consistent with the microarray data indicating a significant change toward a proangiogenic gene program in host cells.
Upregulated Genes in HeLa Cells 6 Hours on B henselae Infection
B henselae Induces HIF-1, Cellular Hypoxia, and Decrease of ATP
The induced gene pattern suggested that HIF-1 activation has a crucial role in the host cell response to B henselae infection. We therefore investigated the activation of HIF-1 by B henselae in HeLa cells (Figure 2). Immunofluorescence analysis revealed the appearance of nuclear HIF-1 6 hours after B henselae infection, similar to hypoxia-treated cells (Figure 2a), and consistent results were obtained by Western blotting (Figure 2b). Transfection experiments with the use of a VEGF promoter/luciferase reporter that is specifically regulated by HIF-1 revealed that B henselae increases VEGF gene transcription 4-fold compared with 8-fold stimulation by hypoxia (Figure 2c). Nuclear HIF-1 activation was also demonstrated by electrophoretic mobility shift assay (EMSA), which indicated the enhanced capacity of the induced HIF-1 complex to bind to its target consensus DNA sequences (Figure 2d). Finally, activation of HIF-1 and subsequent induction of VEGF secretion after B henselae infection was completely inhibited when the cells were treated with HIF-1 siRNA before infection (Figure 2e, 2f). These results clearly indicate that HIF-1 is activated on a B henselae infection and that the increased VEGF secretion on B henselae infection is dependent on HIF-1.
Endothelial cells that were heavily infected by B henselae (Data Supplement Figure II) also upregulated VEGF, IL-8, ADM, and HK2 after B henselae infection (Figure 3). Moreover, VEGFR-2, known to be regulated via HIF-2,22 and the NF-B–regulated MCP-1 and GM-CSF genes were induced. Accordingly, HIF-1 and HIF-2 stabilization was observed in HUVECs (Figure 3b). Moreover, HIF-1 activation and VEGF secretion occurred in all cells investigated thus far (A549, MCF-7, HT29, ECC-1 [data not shown], EA.hy 926, MonoMac610,23). These experiments indicate that a B henselae infection generally results in HIF activation and the induction of a proangiogenic gene program.
Because of the strong IL-8 induction, we also confirmed NF-B activation by EMSA (Figure 4a). To dissect the NF-B–regulated IL-8 from the HIF-1–regulated proangiogenic response, the infection process was modulated by the NF-B inhibitor parthenolide12 (Figure 4b). This inhibited IL-8 secretion from B henselae–infected cells completely but had only little effect on VEGF secretion. This and the data obtained with HIF-1 siRNA strongly suggest that NF-B is responsible for IL-8 induction, whereas HIF-1 triggers VEGF induction. The MAP kinases p38 and p42/44, which were rapidly and persistently phosphorylated in B henselae–infected cells (Figure 4c), contribute only little to the B henselae–induced VEGF secretion because preincubation with PD98059 (p42/44 inhibitor) or SB202190 (p38 inhibitor) caused only a slight reduction of the VEGF levels (Figure 4d). These results suggest that activation of HIF-1 on a Bartonella infection occurs via hypoxic activation, presumably involving inhibition of prolyl and asparaginyl hydroxylases.4
Because hypoxia is known to be a potent activator of HIF-1,4 we investigated whether infection with B henselae leads to an increased O2 consumption and cellular hypoxia. Experiments quantifying the dissolved O2 concentration in a closed cell culture chamber and with the use of a hypoxia-sensitive dye indicated that both processes indeed occurred (Figure 5a to 5c). O2 consumption was increased 9-fold in the B henselae–infected cells. Cellular hypoxia did not result from bacterial overgrowth because the pH of cell culture supernatants was unaffected in the same experiments (control 7.52, B henselae 7.51, hypoxia 7.28). In addition, infection resulted in decreased ATP levels in host cells (Figure 5d).
Expression of B henselae Pili Is Crucial for Activation of HIF-1
To analyze whether pilus expression determines the activation of HIF-1, HeLa cells were infected with wild-type (wt) or pilus-negative (Pil–) B henselae (Figure 6a). Analysis of LPS composition disclosed no differences between both strains (Figure 6b). Pilus expression was crucial for activation of HIF-1 (Figure 6c), VEGF expression (Figure 6d), and cellular hypoxia (Figure 6e), suggesting that pili are necessary for inducing an angiogenic host cell response by B henselae. Moreover, HIF-1 activation was not observed when cells were exposed to B henselae LPS or when infected with L pneumophila (Data Supplement Figures IV and V).
HIF-1 Is Activated in BA Lesions From Patients
To investigate whether HIF-1 activation also occurs in vivo in Bartonella-infected tissues, sections from BA lesions were stained for HIF-1 (Figure 7). Nuclear HIF-1 was stained strongly in cells identified as histiocytes or macrophages infiltrating BA lesions. These observations are consistent with previously published results of HIF-1 staining in hypoxic tumor tissues17 and with recent data showing that HIF-1 is crucial in myeloid cell functions associated with inflammation and bacterial defense functions.24 From this data we conclude that HIF-1 induction occurs in B henselae infections both in vitro and in vivo.
Discussion
Our present studies reveal a novel mechanism of angiogenesis associated with a bacterial infection. We show here that infection with B henselae, the causative agent of angiogenic proliferation of capillary vessels in BA, results in HIF-1 activation and leads to a gene expression pattern typical for the cellular response to hypoxia. In accordance, nuclear HIF-1 was detected in Bartonella-infected cells in vitro and in vivo in BA patient samples. Our results suggest that activation of a HIF-1–dependent cell response on a Bartonella infection causes the vasculoproliferative disorder in BA and BP.
Angiogenesis research has been focused mainly on cancer and cardiovascular diseases.1 The HIF-1–induced angiogenic gene reprogramming would indeed be an excellent target for therapeutic intervention in solid tumors, including hemangioblastomas, and also in Kaposi’s sarcoma.4,7 On the other hand, activation of the HIF pathway could be of therapeutic value in tissue ischemia associated with cardiovascular diseases.1 However, HIF activation in bacterial infections has not been described previously, although genetic deletion of HIF-1 in myeloid cells compromises inflammatory responses.24 Bartonella species are the only known bacterial pathogens causing vasculoproliferative disorders in humans.9 VEGF induction was described in B henselae infections in vivo and in vitro,10 and a role has also been suggested for macrophage-secreted angiogenic agents.25
Only 20 genes were found to be significantly upregulated on B henselae infection in our microarray analysis. Of these, 14 are known HIF targets.3 Several of these were confirmed by independent assays, including VEGF, which is known to be a key factor in the triggering of angiogenic processes,5 and STC2, ADM, IGFBP-3, and HK2, which have also been implicated in hypoxic cellular responses or angiogenic processes.26–30 IL-8, which is apparently the only NF-B–regulated gene in our B henselae infection model of HeLa cells, is also known to be a strong vasculoproliferative agent.31 We hypothesize that the induced growth factors and cytokines secreted by the infected host cells contribute to an angiogenic cocktail responsible for endothelial proliferation in BA or BP, whereas the induction of HIF-regulated glycolytic genes may be linked to the metabolic demands of the B henselae–infected host cells (Figure 8).
Interestingly, some of the induced genes have been reported to be upregulated via HIF-1 or hypoxia in various types of human tumors (eg, enolase-2 [ENO2], NDRG-1, musculoaponeurotic fibrosarcoma oncogene homolog f [MAFF], nuclear factor IL-3 regulated [NFIL3], cyclin G2 [CCNG2], p21-activated kinase-2 [PAK2], and the hypothetical protein FLJ20500A HREF="#R27-160629">27,30,32). The gene expression pattern in Bartonella-infected cells thus shows striking similarities with that in other angiogenesis models.26,27,30 Of interest is the fact that besides the angiogenic cytokine IL-8, the response in HeLa cells lacked the typical proinflammatory genes that are found to be induced via NF-B in other bacterial infections.33 However, the induction of a broad spectrum of proinflammatory genes and cytokines could be deleterious for the facultative intracellular bacteria.11 In accordance, we did not find a broad activation of NF-B–regulated proinflammatory cytokines. These cytokines are commonly induced on bacterial infection of endothelial cells, possibly because of LPS-mediated signals. However, HeLa cells, in contrast to endothelial cells, do not express toll-like receptor-4, which is required for these signals.34 Finally, the HIF-2 activation in HUVECs is a cell type–specific response that upregulates VEGFR-2 and may amplify the VEGF-driven angiogenic loop by making endothelial cells more sensitive to the secreted VEGF.
HIF-1 activation was observed in vitro and in patient samples from BA. This is consistent with our previous observations regarding increased VEGF expression in BA and BP lesions.10 Macrophages that infiltrate BA lesions were highly positive for HIF-1 and they are known to secrete VEGF on Bartonella infection in vitro.25 How HIF-1 is activated on a B henselae infection is not yet clear. The intracellular bacteria should consume a lot of oxygen, and the resulting hypoxia is known to inhibit the prolyl hydroxylases that target HIF-1 to ubiquitin-mediated degradation.4 In fact, infection with B henselae resulted in increased O2 consumption and hypoxia in the host cells and was accompanied by an increased consumption of cellular ATP. Furthermore, the MAP kinases p42/44 and p38 were phosphorylated on B henselae infection. However, although these kinases have been reported to enhance the transcriptional activity of HIF-1 by phosphorylation (p42/44)35 and to increase VEGF mRNA stability (p38),36 these specific MAP kinase inhibitors induced only a small decrease of the B henselae–triggered VEGF secretion in our model. Stabilization of HIF-1 on a B henselae infection could thus occur by a mechanism similar to that in cells exposed to hypoxia.37
Interestingly, the B henselae Pil– variant did not induce cellular hypoxia, activation of HIF-1, or VEGF expression (Figure 6). Recently, we identified this "pilus" as the nonfimbrial adhesin BadA.15 When one considers this and the observation that HIF-1 activation does not occur in LPS-exposed cells, it is highly unlikely that LPS is involved in the activation of HIF-1 by B henselae. Additionally, HIF-1 activation did not occur when cells were infected with L pneumophila. This is not surprising because L pneumophila infections do not result in vasculoproliferative disorders but instead in inflammatory disease. Our results suggest that Bartonella-specific bacterial mechanisms are necessary for the induction of the angiogenic response. The observation that Bartonella induces an HIF-1–driven cellular phenotype is exciting and provides a new aspect to angiogenesis research. Whether similar strategies are shared by other bacterial pathogens needs to be elucidated further.
Acknowledgments
This work was supported by grants from the Deutsche Forschungsgemeinschaft and from the University of Tübingen (Fortüne Programm). The authors thank A. Schaefer for excellent technical assistance, L. Poellinger for kind advice, and U. Z;hringer for providing Bartonella LPS.
Footnotes
The first 3 authors contributed equally to this work.
The online-only Data Supplement, which contains Figures I through V and a table, can be found with this article at http://www.circulationaha.org.
References
Ferrara N, Alitalo K. Clinical applications of angiogenic growth factors and their inhibitors. Nat Med. 1999; 5: 1359–1364.
Risau W. Mechanisms of angiogenesis. Nature. 1997; 386: 671–674.
Semenza GL. HIF-1, O(2), and the 3 PHDs: how animal cells signal hypoxia to the nucleus. Cell. 2001; 107: 1–3.
Pugh CW, Ratcliffe PJ. Regulation of angiogenesis by hypoxia: role of the HIF system. Nat Med. 2003; 9: 677–684.
Leung DW, Cachianes G, Kuang WJ, et al. Vascular endothelial growth factor is a secreted angiogenic mitogen. Science. 1989; 246: 1306–1309.
Chang Y, Cesarman E, Pessin MS, et al. Identification of herpesvirus-like DNA sequences in AIDS-associated Kaposi’s sarcoma. Science. 1994; 266: 1865–1869.
Sodhi A, Montaner S, Patel V, et al. The Kaposi’s sarcoma-associated herpes virus G protein–coupled receptor up-regulates vascular endothelial growth factor expression and secretion through mitogen-activated protein kinase and p38 pathways acting on hypoxia-inducible factor 1alpha. Cancer Res. 2000; 60: 4873–4880.
Flore O, Rafii S, Ely S, et al. Transformation of primary human endothelial cells by Kaposi’s sarcoma-associated herpesvirus. Nature. 1998; 394: 588–592.
Relman DA, Falkow S, LeBoit PE, et al. The organism causing bacillary angiomatosis, peliosis hepatis, and fever and bacteremia in immunocompromised patients. N Engl J Med. 1991; 324: 1514.
Kempf VA, Volkmann B, Schaller M, et al. Evidence of a leading role for VEGF in Bartonella henselae–induced endothelial cell proliferations. Cell Microbiol. 2001; 3: 623–632.
Kempf VA, Schaller M, Behrendt S, et al. Interaction of Bartonella henselae with endothelial cells results in rapid bacterial rRNA synthesis and replication. Cell Microbiol. 2000; 2: 431–441.
Hehner SP, Heinrich M, Bork PM, et al. Sesquiterpene lactones specifically inhibit activation of NF-kappa B by preventing the degradation of I kappa B-alpha and I kappa B-beta. J Biol Chem. 1998; 273: 1288–1297.
Zahringer U, Lindner B, Knirel YA, et al. Structure and biological activity of the short-chain lipopolysaccharide from Bartonella henselae ATCC 49882T. J Biol Chem. 2004; 279: 21046–21054.
Wellmann S, Taube T, Paal K, et al. Specific reverse transcription-PCR quantification of vascular endothelial growth factor (VEGF) splice variants by LightCycler technology. Clin Chem. 2001; 47: 654–660.
Riess T, Andersson SG, Lupas A, et al. Bartonella adhesin A mediates a proangiogenic host cell response. J Exp Med. 2004; 200: 1267–1278.
Ikeda E, Achen MG, Breier G, et al. Hypoxia-induced transcriptional activation and increased mRNA stability of vascular endothelial growth factor in C6 glioma cells. J Biol Chem. 1995; 270: 19761–19766.
Wiesener MS, Munchenhagen PM, Berger I, et al. Constitutive activation of hypoxia-inducible genes related to overexpression of hypoxia-inducible factor-1alpha in clear cell renal carcinomas. Cancer Res. 2001; 61: 5215–5222.
Camenisch G, Wenger RH, Gassmann M. DNA-binding activity of hypoxia-inducible factors (HIFs). Methods Mol Biol. 2002; 196: 117–129.
Sowter HM, Raval RR, Moore JW, et al. Predominant role of hypoxia-inducible transcription factor (HIF)-1alpha versus HIF-2alpha in regulation of the transcriptional response to hypoxia. Cancer Res. 2003; 63: 6130–6134.
Raleigh JA, Calkins-Adams DP, Rinker LH, et al. Hypoxia and vascular endothelial growth factor expression in human squamous cell carcinomas using pimonidazole as a hypoxia marker. Cancer Res. 1998; 58: 3765–3768.
Kunsch C, Rosen CA. NF-kappa B subunit-specific regulation of the interleukin-8 promoter. Mol Cell Biol. 1993; 13: 6137–6146.
Elvert G, Kappel A, Heidenreich R, et al. Cooperative interaction of hypoxia-inducible factor-2alpha (HIF-2alpha) and Ets-1 in the transcriptional activation of vascular endothelial growth factor receptor-2 (Flk-1). J Biol Chem. 2003; 278: 7520–3750.
Kempf VA, Schairer A, Neumann D, et al. Bartonella henselae inhibits apoptosis in Mono Mac 6 cells. Cell Microbiol. 2005; 7: 91–104.
Cramer T, Yamanishi Y, Clausen BE, et al. HIF-1alpha is essential for myeloid cell-mediated inflammation. Cell. 2003; 112: 645–657.
Resto-Ruiz SI, Schmiederer M, Sweger D, et al. Induction of a potential paracrine angiogenic loop between human THP-1 macrophages and human microvascular endothelial cells during Bartonella henselae infection. Infect Immun. 2002; 70: 4564–4570.
Kahn J, Mehraban F, Ingle G, et al. Gene expression profiling in an in vitro model of angiogenesis. Am J Pathol. 2000; 156: 1887–1900.
Leonard MO, Cottell DC, Godson C, et al. The role of HIF-1alpha in transcriptional regulation of the proximal tubular epithelial cell response to hypoxia. J Biol Chem. 2003; 278: 40296–40304.
Oehler MK, Hague S, Rees MC, et al. Adrenomedullin promotes formation of xenografted endometrial tumors by stimulation of autocrine growth and angiogenesis. Oncogene. 2002; 21: 2815–2821.
Schmid MC, Bisoffi M, Wetterwald A, et al. Insulin-like growth factor binding protein-3 is overexpressed in endothelial cells of mouse breast tumor vessels. Int J Cancer. 2003; 103: 577–586.
Lal A, Peters H, St Croix B, et al. Transcriptional response to hypoxia in human tumors. J Natl Cancer Inst. 2001; 93: 1337–1343.
Koch AE, Polverini PJ, Kunkel SL, et al. Interleukin-8 as a macrophage-derived mediator of angiogenesis. Science. 1992; 258: 1798–1801.
Feldenzer JA, McKeever PE. Selective localization of gamma-enolase in stromal cells of cerebellar hemangioblastomas. Acta Neuropathol (Berl). 1987; 72: 281–285.
Bohn E, Muller S, Lauber J, et al. Gene expression patterns of epithelial cells modulated by pathogenicity factors of Yersinia enterocolitica. Cell Microbiol. 2004; 6: 129–141.
Pridmore AC, Wyllie DH, Abdillahi F, et al. A lipopolysaccharide-deficient mutant of Neisseria meningitidis elicits attenuated cytokine release by human macrophages and signals via toll-like receptor (TLR) 2 but not via TLR4/MD2. J Infect Dis. 2001; 183: 89–96.
Richard DE, Berra E, Gothie E, et al. p42/p44 mitogen-activated protein kinases phosphorylate hypoxia-inducible factor 1alpha (HIF-1alpha) and enhance the transcriptional activity of HIF-1. J Biol Chem. 1999; 274: 32631–32637.
Pages G, Berra E, Milanini J, et al. Stress-activated protein kinases (JNK and p38/HOG) are essential for vascular endothelial growth factor mRNA stability. J Biol Chem. 2000; 275: 26484–26491.
Berra E, Benizri E, Ginouves A, et al. HIF prolyl-hydroxylase 2 is the key oxygen sensor setting low steady-state levels of HIF-1alpha in normoxia. EMBO J. 2003; 22: 4082–4090.(Volkhard A.J. Kempf, MD; )