Chlamydia pneumoniae directly interferes with HIF-1astabilization in human host cells

Jan Rupp,1* Jens Gieffers,1 Matthias Klinger,2

Ger van Zandbergen,1 Robert Wrase,3

Matthias Maass,4 Werner Solbach,1 Joerg Deiwick3

and Thomas Hellwig-Burgel5

Institutes of 1Medical Microbiology and Hygiene,2Anatomy, 3Biochemistry, and 5Physiology, Center forStructural and Cell Biology in Medicine, University ofLuebeck, 23538 Luebeck, Germany.4Institute of Medical Microbiology, Hygiene andInfectious Diseases, University Hospital Salzburg,Austria.

Summary

Chlamydiaceae are obligate intracellular bacteriathat cause endemic trachoma, sexually transmitteddiseases and respiratory infections. The course ofthe diseases is determined by local inflammatoryimmune responses and the propensity of the patho-gen to replicate within infected host cells. Both fea-tures require energy which is inseparably coupled tooxygen availability in the microenvironment. Hypoxia-inducible factor-1 (HIF-1) regulates crucial genesinvolved in the adaptation to low oxygen concentra-tions, cell metabolism and the innate immuneresponse. Here we report that Chlamydia pneumoniaedirectly interferes with host cell HIF-1a regulation in abiphasic manner. In hypoxia, C. pneumoniae infectionhad an additive effect on HIF-1a stabilization resultingin enhanced glucose uptake during the early phaseof infection. During the late phase of intracellularchlamydial replication, host cell adaptation tohypoxia was actively silenced by pathogen-inducedHIF-1a degradation. HIF-1a was targeted by thechlamydial protease-like activity factor, which wassecreted into the cytoplasm of infected cells. Directinterference with HIF-1a stabilization was essentialfor efficient C. pneumoniae replication in hypoxia andhighlights a novel strategy of adaptive pathogen–hostinteraction in chlamydial diseases.

Introduction

Chlamydiales are an evolutionary distinct group of humanpathogenic bacteria and chlamydia-related symbiontssharing a biphasic developmental cycle in which theintracellular phase tethers the pathogen to its specifichost (Horn et al., 2004). Chlamydia pneumoniae,C. trachomatis and C. psittaci which are known patho-genic species for humans, take advantage of metabolicand biosynthetic pathways of infected host cells anddirectly interfere with cell-death signalling mechanism(Caldwell et al., 2003; Fischer et al., 2004). Deprivation ofcertain amino acids or glucose prevents the implementa-tion of replicative chlamydial infection and may result inchlamydial persistence, characterized by reduced infec-tivity and metabolism (Harper et al., 2000; Nelson et al.,2005). Nutrient supply and glucose uptake of potentialhost cells in humans is connected to oxygen saturation inorgan tissues and sensitive to local changes in haemody-namic and oxygen transport parameters (Gleadle andRatcliffe, 1997; Iyer et al., 1998). Inflammatory reactionsto bacterial infections usually reduce tissue oxygenationand induce cellular responses to hypoxia in diseasedorgans (L’Her and Sebert, 2001; Koury et al., 2004). Cel-lular responses to low oxygen concentrations are centrallyregulated via the hypoxia-inducible factor-1 (HIF-1).

The hypoxia-inducible factor-1 is a ubiquitously andconstitutively produced transcription factor composed ofan oxygen-labile a subunit and an oxygen-resistent bsubunit, with the latter being identical with the aryl hydro-carbon nuclear translocator (Gradin et al., 1996; Woodet al., 1996). HIF-1a possesses two central oxygen-dependent degradation domains (ODDs) and two trans-activation domains. In the presence of O2, human HIF-1ais hydroxylated at the proline residues Pro-402 and Pro-564 within the ODDs (Ivan et al., 2001; Jaakkola et al.,2001; Masson et al., 2001). This reaction is catalysed byspecific HIF-1 prolyl hydroxylase domain-containingenzymes that are members of the 2-oxoglutarate-dependent and Fe2+-dependent dioxygenases (Epsteinet al., 2001). Prolyl-hydroxylated HIF-1 is captured imme-diately by the von Hippel–Lindau protein (Ivan et al.,2001; Jaakkola et al., 2001), which is the substrate-recognizing subunit of an E3 ubiquitin ligase (Maxwellet al., 1999; Ohh et al., 2000). Once polyubiquitinated,HIF-1a is degraded by the 26S proteasome (Salceda and

Received 1 December, 2006; revised 13 March, 2007; accepted 29March, 2007. *For correspondence. E-mail jan.rupp@uk-sh.de;Tel. (+49) 451 500 4409; Fax (+49) 451 500 2808.

Cellular Microbiology (2007) 9(9), 2181–2191 doi:10.1111/j.1462-5822.2007.00948.xFirst published online 8 May 2007

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd

Caro, 1997; Huang et al., 1998). Here we report thatC. pneumoniae directly interferes with the central oxygen-sensing mechanisms in infected host cells by differentialmodulation of HIF-1a during intracellular replication inhypoxia. C. pneumoniae-induced HIF-1a stabilization inthe early phase of infection was a prerequisite for efficientintracellular replication, whereas HIF-1a was selectivelydegraded by the chlamydial protease-like activity factor(CPAF) during later stages of the chlamydial developmen-tal cycle.

Results

Growth of C. pneumoniae in hypoxia

Survival mechanisms of obligate intracellular bacteria inhuman host cells have not been investigated in hypoxia.We initially analysed the potential of C. pneumoniae toreplicate under hypoxia (3% O2) in the epidermoid carci-noma cell line HEp-2 which is conventionally used forchlamydial propagation. Effectivity of infection (> 90%),differentiation to reticulate bodies (RBs) and retransforma-tion to infectious elementary bodies (EBs) did not differbetween hypoxic and normoxic conditions. Thus, the per-centage of the chlamydial inclusions to the total size ofinfected cells was 24.3% versus 17.8% (normoxia versushypoxia) after 48 h and 37.5% versus 38.5% after 72 hrespectively (data not shown). Quantification of chlamydialdevelopmental forms in intracellular replication revealedno significant changes between cells incubated in hypoxiacompared with normoxia (Table 1). However, the morpho-logical appearance of intracellular inclusions differedmarkedly between both conditions (Fig. 1A and B). Thus,multiple inclusions were observed in 26 � 7% of normoxicbut only 2 � 2% of hypoxic cells within 48 h after C. pneu-moniae infection (Fig. 1C). Individual chlamydial inclusionswere drastically enlarged in hypoxia when compared withthe same time points in normoxia (Fig. 1A and B).Differences in morphology did not alter chlamydialfitness as recovery rates of chlamydial infectious EBswere 1003 � 204 ifu ml-1 versus 5080 � 1581 ifu ml-1

(ifu, inclusion-forming units; normoxia versus hypoxia)

after 48 h and 63667 � 8875 ifu ml-1 versus 73633 �

26366 ifu ml-1 after 72 h respectively (Fig. 1D). As cellviability of C. pneumoniae-infected cells was 96 � 3%under hypoxia and 94 � 4% under normoxia after 48 h(data not shown), we concluded that energy supply wasboth sufficient for chlamydial replication and for host cellintegrity under conditions of reduced oxygen supply.

C. pneumoniae stabilizes HIF-1a in the early phase ofinfection

We wondered about the mechanisms involved in the fine-balanced regulation of pathogen and host cell metabolismin hypoxia and started to analyse the regulation of HIF-1ain chlamydial infection. Within 24 h after stimulation,increased stabilization of HIF-1a was observed inC. pneumoniae-infected cells but also in cells treated withheat-inactivated chlamydiae under hypoxic conditions(Fig. 2A). C. pneumoniae-induced HIF-1a stabilization inthe early phase of infection was essential for effectiveintracellular replication in hypoxia but not in normoxia.Thus, pretreatment of HEp-2 cells with small interferingRNA (siRNA) targeted against HIF-1a significantly abol-ished formation of chlamydial inclusions in hypoxic condi-tions (Fig. 2B). Development of intracellular chlamydialinclusions in hypoxia was reduced from 91 � 1% to46 � 1% (P < 0.01) by siRNA-dependent inhibition of HIF-1a stabilization upon C. pneumoniae infection (Fig. 2C).

C. pneumoniae degrades HIF-1a in the mid- to latephase of infection

In the mid and late phases (48–72 h) of intracellularchlamydial replication in hypoxia, HIF-1a was reducedin infected cells compared with non-infected cells andcells treated with heat-inactivated chlamydiae (Fig. 3A).Looking for a genus-specific chlamydial factor with pro-teolytic activity we focused on the CPAF which has beenrecently described by Zhong et al. (2001). We analysedthe expression kinetics of CPAF in C. pneumoniae-infected HEp-2 cells in hypoxia and normoxia (Fig. 3B–D).Hypoxic CPAF expression in C. pneumoniae-infectedhost cells occurred within 48 h (Fig. 3B and C). Comparedwith normoxia, CPAF protein expression was enhanced incells 72 h after C. pneumoniae infection in hypoxia(Fig. 3C). This was due to enhanced CPAF transcriptionsuggesting an adaptive pathogen response to an alteredhost environment (Fig. 3D).

Chlamydial protease-like activity factor directly interactswith eukaryotic HIF-1a

Based on the findings of decreased HIF-1a stabilizationand increased CPAF expression in C. pneumoniae-infected host cells under hypoxia, we investigated direct

Table 1. Quantification of the chlamydial developmental stages inC. pneumoniae-infected HEp-2 cells in normoxia (20% O2) andhypoxia (3% O2).

Post-infection RBs IBs EBs

20% O2 48 h 90% 7% 3%72 h 43% 12% 45%

3% O2 48 h 92% 3% 5%72 h 39% 11% 50%

More than 30 cells were counted under each condition to assess thedistribution of reticulate bodies (RBs), intermediate bodies (IBs) andelementary bodies (EBs) in chlamydial inclusion 48 and 72 h afterinfection (data are indicated as mean).

2182 J. Rupp et al.

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

A

B

C D

ifu

ul–1

Fig. 1. Fluorescence and electron microscopy of C. pneumoniae-infected HEp-2 cells after 72 h in a normoxic and hypoxic atmosphere.A. FITC-labelled anti-chlamydia antibodies visualized enlarged intracellular chlamydial inclusions in hypoxia.B. Electron microscopy of infected HEp-2 cells proved chlamydial development and differentiation under normoxia and hypoxia. Scale bars:1 mm.C. Occurrence of multiple inclusions were significantly inhibited in hypoxia after 48 h.D. Recovery rates of infectious chlamydial EBs (ifu ml-1) did not differ significantly after 48 and 72 h under normoxic or hypoxic conditions.

Chlamydiae alter host cell HIF-1a stabilization 2183

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

interactions between prokaryotic CPAF and eukaryoticHIF-1a. Using cell-free degradation assays it becameclearly evident that 35S-Met-labelled HIF-1a was com-pletely cleaved by cytoplasmic extracts from cellsinfected with C. pneumoniae (Fig. 4A). The effect wasstrongly dependent on the developmental stage andviability of intracellular chlamydiae. Biological activity ofCPAF in extracts from C. pneumoniae-infected cells wasconfirmed by degradation of the regulatory factor X5(RFX5, Fig. 4B) known to be cleaved by CPAF (Zhonget al., 2001). To prove that CPAF is the HIF-1a degrad-ing agent, degradation assays were repeated afterCPAF was removed from cytoplasmic extracts ofinfected cells via a pull-down assay (Fig. 4C). Pre-incubation of cytoplasmic extracts of C. pneumoniae-infected cells with monoclonal anti-CPAF antibody butnot with mouse control serum completely protectedHIF-1a from degradation (Fig. 4C). To further tighten theline of evidence, CPAF was expressed in Escherichiacoli as glutathione S-transferase (GST)-fusion protein,solubilized and purified. Used in the HIF-1a degradationassays purified GST–CPAF retained its HIF-1a degrad-ing activity (Fig. 4D).

CPAF-induced HIF-1a degradation determines host cellmetabolism and survival

To assess the functional relevance of increased HIF-1astabilization in the early and HIF-1a degradation in themid to late phase of chlamydial infection, we analysedthe metabolic activity of infected and non-infected hostcells under hypoxic conditions. Considering Glut-1 asHIF-1 target gene, glucose uptake was enhanced inC. pneumoniae-infected cells when HIF-1a was upregu-lated by chlamydiae (Fig. 5A). During the later course ofthe chlamydial developmental cycle, glucose uptake inC. pneumoniae-infected cells was dramatically decreasedcompared with non-infected cells and further declinedsubsequent to HIF-1a degradation (Fig. 5A). Simulta-neously, hypoxic HEp-2 cells infected with C. pneumoniaefor 48 h showed enhanced resistance against pro-apoptotic stimuli compared with non-infected controls. Inhypoxia, C. pneumoniae infection reduced chromatin con-densation and DNA fragmentation in UV-treated host cellsas an indication that pro-apoptotic HIF-1a activity wassuppressed during the mid phase of intracellular chlamy-dial replication (Fig. 5B and C).

Discussion

Chlamydiae have evolved strategies to directly influenceimmune recognition and viability of host cells to imple-ment their developmental cycle, resulting in the release ofeither infectious pathogens or intracellular chlamydial

A

B

C

Fig. 2. Western blot analysis (A) showed induction of HIF-1astabilization in early (after 24 h) C. pneumoniae-infected HEp-2cells in normoxia (20% O2) and hypoxia (3% O2). Increased HIF-1astabilization in the early phase of C. pneumoniae infection wascrucial for intracellular chlamydial replication in hypoxia.Pretreatment of HEp-2 cells with siRNA against HIF-1a but not withscRNA (scrambled RNA, control) significantly blocked chlamydialreplication in hypoxia but not in normoxia (B, C).

2184 J. Rupp et al.

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

Fig. 3. A. Western blot analysis of HIF-1ashowed significant reduction inC. pneumoniae-infected cells in hypoxia after48 and 72 h.B and C. Analysis of the CPAF expression inC. pneumoniae-infected HEp-2 cells showed atime-dependent increase in normoxia andhypoxia by immunohistochemistry (B, 3% O2)and Western blot analysis (C). Scale bars: B,10 mm.D. Relative quantification of CPAF mRNAamounts (DDCT- method, respective normoxicsamples were set to 1.0) showed increasedlevels of CPAF mRNA in hypoxia comparedwith normoxia.

A

B

C

D

Chlamydiae alter host cell HIF-1a stabilization 2185

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

persistence (Zhong et al., 2001; Hogan et al., 2004). Lessis known about the chlamydial capacity to cope with envi-ronmental conditions of reduced oxygenation which canbe found physiologically in different human tissues and asa consequence of inflammation in diseased organs. Wecould show for the first time that chlamydiae are ableto replicate within infected host cells and to maintaininfectivity under hypoxic conditions (3% O2). To directlymanipulate host cell metabolism and function in hypoxia,chlamydiae thereby take control of the central oxygen-sensing mechanism HIF-1a. Thus, C. pneumoniae-induced HIF-1a stabilization in the early phase of infection

was associated with an increased glucose uptake ininfected host cells, which may be needed for efficientintracellular chlamydial replication and differentiation(Yaraei et al., 2005). Although cellular HIF-1a stabilizationupon bacterial challenge does not seem to be achlamydia-specific mechanism (Kempf et al., 2005; Fredeet al., 2006), increased HIF-1a stabilization in the earlyphase of chlamydial infection was crucial for intracellularreplication in an environment of reduced oxygenation.Inhibition of the C. pneumoniae-induced HIF-1a responseby siRNA significantly reduced chlamydial infectivity inhypoxia.

A B

C

D

Fig. 4. Kinetics of in vitro transcribed/translated 35S-Met-labelled HIF-1a degradation by cytoplasmic extracts from C. pneumoniae-infectedHEp-2 cells.A. Degradation of 35S-Met-labelled HIF-1a was observed by incubation with cytoplasmic extracts from HEp-2 cells infected for 48 and 72 hwith viable C. pneumoniae (CV6) in normoxia and hypoxia but not with heat-inactivated chlamydiae.B. CPAF activity in cytoplasmic extracts from C. pneumoniae-infected cells was confirmed by a degradation assay with 35S-Met-labelled RFX5.C. Degradation of HIF-1a was completely blocked by immunoprecipitation (IP) of the cellular extract from C. pneumoniae-infected cells withmouse monoclonal EB3.1 antibody against CPAF but not with mouse control IgG alone.D. Dose-dependent degradation of purified GST-HIF-1a by CPAF was observed in titration experiments using a GST–CPAF fusion protein andquantified by using PCBAS software and were given as relative radioactivity per mm2.

2186 J. Rupp et al.

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

A

B

C

Fig. 5. Analysis of glucose uptake in C. pneumoniae-infected and non-infected HEp-2 cells in hypoxia (3% O2).A. Early C. pneumoniae infection (24 h) induced glucose uptake into HEp-2 cells resulting in diminished glucose levels in the supernatantscompared with non-infected cells. During the mid to late phase (48–72 h) of intracellular chlamydial infection in hypoxia glucose consumptionof C. pneumoniae-infected cells was reduced compared with non-infected cells.B and C. UV-induced apoptosis of HEp-2 cells (500 mJ cm-2, 5 h) in hypoxia (Hox) was diminished in HEp-2 cells 48 h after infection withC. pneumoniae. Apoptotic cells were determined and quantified by analysis of the nuclear morphology of infected and non-infected cells (B,live/dead staining with Syto16 and Et.Br.) and the TUNEL assay as a marker for DNA fragmentation (C).B. The arrows indicate condensed nuclei. C. The arrows indicate TUNEL positive cells.

Chlamydiae alter host cell HIF-1a stabilization 2187

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

In contrast, during mid to late phase of the infection,HIF-1a was actively degraded by CPAF, a chlamydialprotease secreted into the cytoplasm of the cells within 48to 72 h after infection. It is known that in normoxia,chlamydiae exhibit anti-apoptotic activities to stabilize theinfected host cell during mid to late stages in the devel-opmental cycle (Miyairi and Byrne, 2006). Hypoxia hasbeen shown to induce apoptosis of epithelial cells throughactivation of HIF-1a and the pro-apoptotic protein Bnip3L(Krick et al., 2005). Cell adaptation to hypoxia and cellapoptosis is thereby controlled by a network involvingHIF-1a, p53 and members of the bcl-2 protein family (Kimet al., 2004; Krick et al., 2005; Tanaka et al., 2005; Liuet al., 2006). Our results indicate that chlamydiae directlytarget HIF-1a to prevent hypoxic cell death in order tocomplete intracellular development and to maintain infec-tivity in hypoxia. However, the ability of chlamydiae tosurvive and to replicate in an environment of reducedoxygen supply is limited as O2- levels below 1% resultedin almost complete disruption of the intracellular develop-mental cycle (data not shown). The mechanisms thatdetermine the fate of chlamydial replication in conditionsnear anoxia (between 1% and 3% O2) are currentlyinvestigated.

In summary, we have shown on the one hand thathypoxia alters intracellular chlamydial morphology andelicits enhanced transcriptional activity of the bacterialCPAF. On the other hand, HIF-1a is directly and differen-tially targeted by chlamydiae to control host cell metabo-lism and integrity during the intracellular developmentalcycle. Thus, C. pneumoniae-induced HIF-1a stabilizationand subsequent glucose uptake in the early phase ofinfection was crucial for chlamydial replication in hypoxia.Furthermore, CPAF-dependent degradation of HIF-1aduring the mid to late phase of chlamydial infection seemsto prevent pro-apoptotic host cell signalling and allowedthe pathogen to remain infectious in a hypoxicenvironment. Further studies will pinpoint the potency ofCPAF, as an effective bacteria-derived protease forHIF-1a degradation, to modulate cellular and systemichomeostatic responses to hypoxia in vivo.

Experimental procedures

Cell cultures

Culture of C. pneumoniae was propagated in HEp-2 cells (ATCCCCL-23) over serial passages in tissue culture plates (Greiner)with growth medium containing Eagle’s minimum essentialmedium (Sigma), 10% fetal calf serum (Biochrom), L-glutamine(2 mM, Invitrogen), non-essential amino acids, gentamicin(10 mg l-1), vancomycin (50 mg l-1), and amphotericin B(2 mg l-1). Confluent monolayers were infected with the vascularchlamydial strain CV-6 or the CWL strain (ATCC VR-1310). CV-6was isolated by us from an atheromatous lesion of a restenotic

coronary bypass graft (Maass et al., 1998). In the experimentalsetting, cells were infected with 4 ifu per cell of freshly isolatedchlamydiae (Maass et al., 1998). In preliminary experiments,mock-treated cells showed no difference to cells treated withmedium alone. All experiments were done in triplicate.

Oxygen deprivation

HEp-2 cells (2.5 ¥ 105) were seeded in 9.6 cm2 cell culture dishesto reach 70% confluency. In the experimental setting, cells wereinfected with 4 ifu per cell of freshly isolated chlamydiae andplaced at 37°C in a humidified atmosphere with 3% O2

(21 mmHg), 5% CO2 and balanced N2 for hypoxic incubation or in20% O2, 5% CO2 and balanced N2 for normoxia. Heat-inactivatedchlamydiae (95°C, 30 min) served as control and were includedin every single experiment. Infected and non-infected cells wereharvested after 24, 48 and 72 h, and immediately lysed in respec-tive buffers for mRNA extraction, protein analysis or isolation ofcytoplasmic extracts.

Fluorescence microscopy, immunohistochemistry andelectron microscopy

Cells were grown to 70% confluency on coverslips for immuno-fluorescence staining and immunohistochemistry (IHC). FITC-labelled monoclonal anti-chlamydia-LPS antibodies (Dako) onmethanol-fixed slides showed chlamydial development stagesafter 24, 48 and 72 h, whereas coincubation with heat-inactivatedC. pneumoniae did not result in replicative infection. Infectionrates and the proportion of cells with multiple inclusions areindicated as mean � SEM. IHC for CPAF was performed withmouse monoclonal anti-CPAF antibody (IgG3, clone EB3.1; kindlyprovided by Dr Guangming Zhong, University of Texas HealthScience Center) (Fan et al., 2002). Visualization was performedby a DAB substrate kit (VectorLabs). Negative controls com-prised omission of the antibody. For transmission electronmicroscopy, cells grown on Thermanox slides were fixed with 5%glutaraldehyde in PBS (pH 7.4) for 1 h at 4°C. Post-fixation wasperformed with 1% OsO4 for 2 h; samples were dehydrated in agraded ethanol series and embedded in Araldite (Fluka). Ultrathinsections were contrasted with uranyl acetate and lead citrate andwere examined with a Philips EM 400. To assess inclusion mor-phology and development stages of chlamydial inclusions, morethan 30 cells were counted per sample.

Western blot analysis

For determination of immunoreactive HIF-1a protein in the nucleiof cells, nuclear extracts were prepared as already described andsubjected to Western blot analysis (Hellwig-Burgel et al., 1999).Total cell extracts were used to determine CPAF expression inC. pneumoniae-infected cells. Samples were run on SDS/7.5%polyacrylamide gels and transferred to nitrocellulose membranes(Amersham). Equal loading and blotting efficiency were verifiedby staining with 2% Ponceau S or by reprobing with an anti-histone H1 (Santa Cruz) or anti-b actin antibody (Cell Signaling).Membranes were blocked with PBS/5% fat-free skimmed milkand then incubated with a monoclonal mouse antibody raisedagainst human HIF-1a (BD Biosciences; 1:1000 overnight at

2188 J. Rupp et al.

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

4°C) or the mouse monoclonal antibody EB3.1 against CPAF(1:200 overnight at 4°C). For detection, a horseradishperoxidase-linked anti-mouse IgG antibody (1:2000, 1 h at roomtemperature; Dako) and enhanced chemiluminescence substrate(Amersham) were used.

In vitro transcription/translation and cell-free degradationassays

HIF-1a and RFX-5 proteins were produced using expressionplasmids containing the T7 (for HIF-1a) or the SP-6 (for RFX5)binding sites upstream of the translational start codon in coupledtranscription/translation reactions (T7-reticulocyte lysate IVTT Kitor SP6-wheat germ IVTT Kit, both from Promega) in the presenceof 35S-labelled methionine according to the instructions given bythe manufacturer. Cell-free degradation assays were performedin 1 mM Tris (pH 7.4), 15 mM NaCl, 140 mM KCl, 0.1 mM MgCl2,with 0.5 ml of IVTT-products and 6 mg of cytoplasmic extracts or10 ml of purified GST–CPAF in a total volume of 25 ml. Sampleswere incubated at 37°C for 1 h, stopped by the addition of SDSsample buffer and heating to 99°C for 10 min. Samples wereresolved by SDS-PAGE in 7.5% polyacrylamide gels. After gelswere dried the signals were visualized by phosphoimaging (BAS1000, Fuji) and signals were quantified using PCBAS software.Quantitative data of 35S-Met-HIF-1a-specific signals were givenas relative radioactivity per mm2.

Cloning and purification

The CPAF gene was amplified from C. pneumoniae AR39genomic DNA by PCR using the primers 5′-GCGGGATCCATGAAAAAAGGGAAATTAGGAGC-3′ and 5′-GCGCTCGAGTTACAAAGCAGAAGTCGTTGTG-3′. The product was confirmed byDNA sequencing and cloned via BamHI/XhoI into pGEX-6P-1(GE Healthcare). Expression was carried out in E. coli (DE3)Tuner pLacI (Novagen/Merck).

The 96 kDa GST–CPAF fusion protein was purified from bac-terial lysates on glutathione-sepharose four columns (GE Health-care) according to manufacturer’s instructions and dialysedagainst PBS buffer to a final concentration of 0.1 mg ml-1.

Glucose assay

The amount of glucose in the supernatants of cell cultures wasdetermined by a commercially available enzymatic assay (RocheApplied Science). Pre-conditioned culture medium under hypoxiawas replaced every 24 h to analyse glucose consumption ofinfected and non-infected cells in regular intervals of 24 h. Highglucose consumption of cells thereby corresponded to lowglucose levels in the supernatants. The amount of NADPH as amarker for the conversion from glucose to gluconate 6-phosphatewas measured by means of its light absorbance at 340 nm andglucose concentration was calculated in g l-1.

Chlamydial recovery

To determine the burden of infectious C. pneumoniae elementarybodies after intracellular development under normoxic andhypoxic conditions, we performed titration experiments. Infected

HEp-2 monolayers were mechanically detached with a cellscraper, re-suspended in fresh growth medium and inoculated inserial dilutions on confluent HEp-2 monolayers. Development ofchlamydial inclusions was analysed on methanol-fixed slidesusing FITC-labelled monoclonal anti-chlamydia-LPS antibodies(Dako) after 48 h and calculated as ifu ml-1 by observation of 10high power fields (400¥ magnification).

Gene silencing

RNA interference (RNAi) experiments were performed either in24-well or 6-well culture dishes using 40 pmol or 120 pmol siRNA(HIF-1a Stealth RNAi and Stealth RNAi negative control Low GC,Invitrogen) respectively, and Lipofectamine Transfection Reagent(Invitrogen). After transfection cells were allowed to grow for12 h before infection and incubation under the experimentalconditions. Protein knock-down efficiency was controlled byWestern blot analysis for HIF-1a (data not shown).

Apoptosis assays

Apoptosis in C. pneumoniae-infected and non-infected HEp-2cells was induced by 5 h exposure to UV-light (500 mJ/cm2).Apoptotic cells were determined by analysis of nuclear morphol-ogy using live/dead staining and the TUNEL assay detectingapoptotic cell death by enzymatic labelling of DNA strand breakswith fluorescein-dUTP and terminal deoxy-nucleotidyl trans-ferase (TdT) (Gorczyca et al., 1993). Live/dead staining wasperformed with a combination of SYTO-16 (Molecular Probes) todetect intact (round shaped) or condensed (fragmented) nucleiand ethidium bromide (Et.Br.) to detect dead cells with lost mem-brane integrity. Cell morphology was examined under oil immer-sion light microscopy and a minimum of 200 cells/slide wereexamined and graded as apoptotic/non-apoptotic using a ZeissAxioskop-2® fluorescent microscope fitted with a HRS-Axiocam®

and Axiovision software 4.5®.

Statistical analysis

Data are indicated as mean � SEM. Statistical analysis wasperformed with a two-tailed, unpaired Student t-test.

Acknowledgements

Special thanks to Dr Guangming Zhong (University of TexasHealth Science Center) for his generous supply with monoclonalantibodies against CPAF. We are grateful to A. Gravenhorst,A. Hellberg, T. Luedemann and G. Huck for excellent technicalassistance.

References

Caldwell, H.D., Wood, H., Crane, D., Bailey, R., Jones, R.B.,Mabey, D., et al. (2003) Polymorphisms in Chlamydia tra-chomatis tryptophan synthase genes differentiate betweengenital and ocular isolates. J Clin Invest 111: 1757–1769.

Epstein, A.C., Gleadle, J.M., McNeill, L.A., Hewitson, K.S.,O’Rourke, J., Mole, D.R., et al. (2001) C. elegans EGL-9

Chlamydiae alter host cell HIF-1a stabilization 2189

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

and mammalian homologs define a family of dioxygenasesthat regulate HIF by prolyl hydroxylation. Cell 107: 43–54.

Fan, P., Dong, F., Huang, Y., and Zhong, G. (2002) Chlamy-dia pneumoniae secretion of a protease-like activity factorfor degrading host cell transcription factors required for[correction of factors is required for] major histocompatibil-ity complex antigen expression. Infect Immun 70: 345–349.

Fischer, S.F., Vier, J., Kirschnek, S., Klos, A., Hess, S., Ying,S., et al. (2004) Chlamydia inhibit host cell apoptosis bydegradation of proapoptotic BH3-only proteins. J Exp Med200: 905–916.

Frede, S., Stockmann, C., Freitag, P., and Fandrey, J. (2006)Bacterial lipopolysaccharide induces HIF-1 activation inhuman monocytes via p44/42 MAPK and NF-kappaB.Biochem J 396: 517–527.

Gleadle, J.M., and Ratcliffe, P.J. (1997) Induction of hypoxia-inducible factor-1, erythropoietin, vascular endothelialgrowth factor, and glucose transporter-1 by hypoxia: evi-dence against a regulatory role for Src kinase. Blood 89:503–509.

Gorczyca, W., Gong, J., and Darzynkiewicz, Z. (1993) Detec-tion of DNA strand breaks in individual apoptotic cells bythe in situ terminal deoxynucleotidyl transferase and nicktranslation assays. Cancer Res 53: 1945–1951.

Gradin, K., McGuire, J., Wenger, R.H., Kvietikova, I., Fhite-law, M.L., Toftgard, R., et al. (1996) Functional interferencebetween hypoxia and dioxin signal transduction pathways:competition for recruitment of the ARNT transcriptionfactor. Mol Cell Biol 16: 5221–5231.

Harper, A., Pogson, C.I., Jones, M.L., and Pearce, J.H.(2000) Chlamydial development is adversely affected byminor changes in amino acid supply, blood plasma aminoacid levels, and glucose deprivation. Infect Immun 68:1457–1464.

Hellwig-Burgel, T., Rutkowski, K., Metzen, E., Fandrey, J.,and Jelkmann, W. (1999) Interleukin-1beta and tumornecrosis factor-alpha stimulate DNA binding of hypoxia-inducible factor-1. Blood 94: 1561–1567.

Hogan, R.J., Mathews, S.A., Mukhopadhyay, S., Summers-gill, J.T., and Timms, P. (2004) Chlamydial persistence:beyond the biphasic paradigm. Infect Immun 72: 1843–1855.

Horn, M., Collingro, A., Schmitz-Esser, S., Beier, C.L.,Purkhold, U., Fartmann, B., et al. (2004) Illuminating theevolutionary history of chlamydiae. Science 304: 728–730.

Huang, L.E., Gu, J., Schau, M., and Bunn, H.F. (1998) Regu-lation of hypoxia-inducible factor 1alpha is mediated byan O2-dependent degradation domain via the ubiquitin-proteasome pathway. Proc Natl Acad Sci USA 95: 7987–7992.

Ivan, M., Kondo, K., Yang, H., Kim, W., Valiando, J., Ohh, M.,et al. (2001) HIFalpha targeted for VHL-mediated destruc-tion by proline hydroxylation: implications for O2 sensing.Science 292: 464–468.

Iyer, N.V., Kotch, L.E., Agani, F., Leung, S.W., Laughner, E.,Wenger, R.H., et al. (1998) Cellular and developmentalcontrol of O2 homeostasis by hypoxia-inducible factor 1alpha. Genes Dev 12: 149–162.

Jaakkola, P., Mole, D.R., Tian, Y.M., Wilson, M.I., Gielbert, J.,Gaskell, S.J., et al. (2001) Targeting of HIF-alpha to thevon Hippel-Lindau ubiquitylation complex by O2-regulatedprolyl hydroxylation. Science 292: 468–472.

Kempf, V.A., Lebiedziejewski, M., Alitalo, K., Walzlein, J.H.,Ehehalt, U., Fiebig, J., et al. (2005) Activation of hypoxia-inducible factor-1 in bacillary angiomatosis: evidence for arole of hypoxia-inducible factor-1 in bacterial infections.Circulation 111: 1054–1062.

Kim, J.Y., Ahn, H.J., Ryu, J.H., Suk, K., and Park, J.H. (2004)BH3-only protein Noxa is a mediator of hypoxic cell deathinduced by hypoxia-inducible factor 1alpha. J Exp Med199: 113–124.

Koury, J., Deitch, E.A., Homma, H., Abungu, B., Gangurde,P., Condon, M.R., et al. (2004) Persistent HIF-1alpha acti-vation in gut ischemia/reperfusion injury: potential role ofbacteria and lipopolysaccharide. Shock 22: 270–277.

Krick, S., Eul, B.G., Hanze, J., Savai, R., Grimminger, F.,Seeger, W., et al. (2005) Role of hypoxia-inducible factor-1alpha in hypoxia-induced apoptosis of primary alveolarepithelial type II cells. Am J Respir Cell Mol Biol 32: 395–403.

L’Her, E., and Sebert, P. (2001) A global approach to energymetabolism in an experimental model of sepsis. Am JRespir Crit Care Med 164: 1444–1447.

Liu, X.H., Yu, E.Z., Li, Y.Y., and Kagan, E. (2006) HIF-1alphahas an anti-apoptotic effect in human airway epitheliumthat is mediated via Mcl-1 gene expression. J Cell Biochem97: 755–765.

Maass, M., Bartels, C., Engel, P.M., Mamat, U., and Sievers,H.H. (1998) Endovascular presence of viable Chlamydiapneumoniae is a common phenomenon in coronary arterydisease. J Am Coll Cardiol 31: 827–832.

Masson, N., Willam, C., Maxwell, P.H., Pugh, C.W., andRatcliffe, P.J. (2001) Independent function of two destruc-tion domains in hypoxia-inducible factor-alpha chainsactivated by prolyl hydroxylation. EMBO J 20: 5197–5206.

Maxwell, P.H., Wiesener, M.S., Chang, G.W., Clifford, S.C.,Vaux, E.C., Cockman, M.E., et al. (1999) The tumour sup-pressor protein VHL targets hypoxia-inducible factors foroxygen-dependent proteolysis. Nature 399: 271–275.

Miyairi, I., and Byrne, G.I. (2006) Chlamydia and pro-grammed cell death. Curr Opin Microbiol 9: 102–108.

Nelson, D.E., Virok, D.P., Wood, H., Roshick, C., Johnson,R.M., Whitmire, W.M., et al. (2005) Chlamydial IFN-gamma immune evasion is linked to host infection tropism.Proc Natl Acad Sci USA 102: 10658–10663.

Ohh, M., Park, C.W., Ivan, M., Hoffman, M.A., Kim, T.Y.,Huang, L.E., et al. (2000) Ubiquitination of hypoxia-inducible factor requires direct binding to the beta-domainof the von Hippel–Lindau protein. Nat Cell Biol 2: 423–427.

Salceda, S., and Caro, J. (1997) Hypoxia-inducible factor1alpha (HIF-1alpha) protein is rapidly degraded by theubiquitin-proteasome system under normoxic conditions.Its stabilization by hypoxia depends on redox-inducedchanges. J Biol Chem 272: 22642–22647.

Tanaka, T., Kojima, I., Ohse, T., Inagi, R., Miyata, T., Ingel-finger, J.R., et al. (2005) Hypoxia-inducible factor modu-lates tubular cell survival in cisplatin nephrotoxicity. Am JPhysiol Renal Physiol 289: F1123–F1133.

2190 J. Rupp et al.

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191

Wood, S.M., Gleadle, J.M., Pugh, C.W., Hankinson, O., andRatcliffe, P.J. (1996) The role of the aryl hydrocarbonreceptor nuclear translocator (ARNT) in hypoxic inductionof gene expression. Studies in ARNT-deficient cells. J BiolChem 271: 15117–15123.

Yaraei, K., Campbell, L.A., Zhu, X., Liles, W.C., Kuo, C.C.,and Rosenfeld, M.E. (2005) Effect of Chlamydia pneumo-

niae on cellular ATP content in mouse macrophages:role of Toll-like receptor 2. Infect Immun 73: 4323–4326.

Zhong, G., Fan, P., Ji, H., Dong, F., and Huang, Y. (2001)Identification of a chlamydial protease-like activity factorresponsible for the degradation of host transcriptionfactors. J Exp Med 193: 935–942.

Chlamydiae alter host cell HIF-1a stabilization 2191

© 2007 The AuthorsJournal compilation © 2007 Blackwell Publishing Ltd, Cellular Microbiology, 9, 2181–2191