Vet. Res. 38 (2007) 729–739 Available online at:c© INRA, EDP Sciences, 2007 www.vetres.orgDOI: 10.1051/vetres:2007028

Original article

Effect of ovotransferrin and lactoferrins onChlamydophila psittaci adhesion and invasion

in HD11 chicken macrophages

Delphine Sylvie Anne B*, Caroline M.A.D. VD, Bart J.A. D C, Patrick V O,

Daisy C.G. V

Department of Molecular Biotechnology, Faculty of Bioscience Engineering, Ghent University,Coupure Links 653, 9000 Gent, Belgium

(Received 22 December 2006; accepted 12 April 2007 )

Abstract – The effect of ovotransferrin (ovoTF), human lactoferrin (hLF) and bovine lactofer-rin (bLF) on the obligate intracellular pathogen Chlamydophila (Cp.) psittaci was evaluated usinga model of Buffalo Green Monkey kidney (BGM) cells and HD11 chicken macrophages as ar-tificial hosts. Firstly, the effect of transferrins on the infectivity of the bacteria was evaluated.Pre-incubation of Cp. psittaci with 0.5 to 5 mg/mL ovoTF prior to infecting BGM cells significantlylowered the infection rate (P < 0.05). For both lactoferrins, the infection rate could only be reducedwith 5 mg/mL, albeit not significantly as compared to the infection rate created by the untreatedbacteria. Secondly, transferrins were tested for their ability to influence bacterial adhesion and entryin HD11 cells. Maximal non-cytotoxic and non-bactericidal concentrations of 0.05 mg/mL ovoTFand 0.5 mg/mL hLF and bLF were used. Overall, ovoTF was more effective than human and bovineLF in inhibiting bacterial irreversible attachment and cell entry and the latter was accompanied bya dose-dependent reduction of actin recruitment at the bacterial entry site. However, once bacte-ria had entered HD11 cells, transferrins had apparently no effect on intracellular replication. Thepresent findings suggest a possible role for transferrins and especially ovoTF, in preventing avianCp. psittaci infections.

ovotransferrin / lactoferrin / Chlamydia / Chlamydophila psittaci

1. INTRODUCTION

As obligate intracellular Gram-negativebacteria, Chlamydiaceae share a uniquebiphasic developmental cycle. Infectiouselementary bodies (EB) enter the host celland transform into metabolic active retic-

* Corresponding author:Delphine.Beeckman@UGent.be

ular bodies (RB). These RB replicate inan envelope-like structure called an in-clusion. Approximately 50 h later, newlyformed EB escape from the cell to start anew infectious cycle. Chlamydophila (Cp.)psittaci infects birds, causing respiratoryinfections by colonization of mucosal epi-thelial cells and macrophages of the res-piratory tract, and spreading to various

Article available at http://www.vetres.org or http://dx.doi.org/10.1051/vetres:2007028

730 D.S.A. Beeckman et al.

organs. In man, Cp. psittaci causes psitta-cosis or parrot fever [2].

Lactoferrin (LF) is a member of thetransferrin family of iron-binding glyco-proteins [21] and is predominantly presentin milk, tears, saliva as well as in vaginalsecretions. This iron-binding protein is in-volved in host defense mechanisms relatedto the non-immune defense system againstpathogenic bacteria, fungi, and protozoa,both directly and through regulation of theinflammatory response [38]. Antimicrobialfunctions ascribed to this protein or itspeptides include iron sequestration [4],destabilization of the outer membrane ofGram-negative bacteria through bindingof bacterial lipopolysaccharides (LPS) [3,7, 12, 13, 25], selective permeation ofions [1], modulation of bacterial entry intohost cells through gene regulation [5] anddisrupting the bacterial Type III secre-tion system [23]. In birds, the extracellu-lar iron-binding glycoprotein ovotransfer-rin (ovoTF) or conalbumin belonging tothe transferrin family, is synthesized bythe liver and oviduct and is as a con-sequence predominantly present in serumand egg-white [14, 27]. Ovotransferrin isinvolved in iron transport and iron deliv-ery to cells [19,20], but is also proposed toexert antibacterial activity by permeatingouter membranes, reaching the inner mem-brane and causing selective permeation ofions and dissipation of electrical poten-tial [1]. A 92-amino acid ovotransferrinpeptide, OTAP-92, was also found to be ca-pable of killing Gram-negative bacteria bycrossing the bacterial outer membrane byself-promoted uptake, and damaging thecytoplasmic membrane [17, 18].

Anti-chlamydial activities of lactoferrinor ovotransferrin in mammalian or avianhosts have not been investigated. In or-der to establish whether these transferrinsare capable of inhibiting chlamydial infec-tions, we tested both human and bovinelactoferrin, as well as ovotransferrin inour model of Buffalo Green Monkey kid-

ney (BGM) cells and chicken macrophages(HD11 cells).

2. MATERIALS AND METHODS

2.1. Organism and cell culture

Cp. psittaci genotype D strain92/1293 [36], a well-characterizedvirulent strain, was used in this study. Theeffect of transferrins on Cp. psittaci wasexamined by use of Buffalo Green Mon-key kidney (BGM) cells and the chickenmacrophage-like cell line HD11 [6]. BGMcells were cultured in Eagle’s minimalessential medium (MEM) supplementedwith 10% heat-inactivated fetal bovineserum (Invitrogen, Merelbeke, Belgium),2 mM L-glutamine (Invitrogen), 1%vitamins for MEM (Invitrogen), 10 µL/mLof streptomycin sulfate (1% w/v; Sigma,Antwerp, Belgium) and 20 µL/mL ofvancomycin (0.5% w/v; Eli Lilly, Brus-sels, Belgium). HD11 cells were culturedin Dulbecco’s modified Eagle’s minimalessential medium (DMEM) supplementedwith 5% heat-inactivated chicken serum,1% sodium pyruvate, 1% L-glutamineand 0.5% gentamycin (all products fromInvitrogen). Cultures were incubated at5% CO2 and 37 ◦C.

2.2. Transferrins

Ovotransferrin (ovoTF), human lacto-ferrin (hLF) and bovine lactoferrin (bLF)purified from chicken egg white, hu-man milk and bovine colostrum respec-tively, were purchased from Sigma andiron saturated essentially as previously de-scribed [25]. Briefly, the proteins weredissolved in 10 mM sodium bicarbon-ate and subsequently supplemented withsodium citrate and Fe (III) chloride (finalconcentration 5 mM) (all products fromAcros Organics, Geel, Belgium). For each

Anti-Cp. psittaci effect of transferrins 731

protein solution, pH was adjusted to 8with dissolved sodium bicarbonate. Afterstirring the solutions for 12 h at 4 ◦C,dialysis was performed against 10 mMsodium bicarbonate for 48 h at 4 ◦C.Subsequently, proteins were concentratedusing Centricon� Plus-70 Centrifugal Fil-ter Units (Millipore, Brussels, Belgium)as described by the manufacturer. Pro-tein concentrations were determined spec-trophotometrically (260 nm) as well as ironsaturation (468 nm, 1% solution). Concen-trated protein solutions were stored at 4 ◦Cfor no more than 24 h.

2.3. Transferrin cytotoxicity assays

To check the putative cytotoxic ef-fects of transferrins, HD11 cells wereseeded in 96-well plates (Greiner Bio-One, Wemmel, Belgium) at a concentra-tion of 300 000 cells/mL and the follow-ing day were exposed to concentrationsof 0.5, 1, 2.5, 5 and 10 mg/mL of trans-ferrins in culture medium. Incubations for24, 36 and 48 h were performed in du-plicate. At these time points, cytotoxic-ity in a dose-dependent manner was as-sessed by both viable cell counts (NikonEclipse TS100, 100×) after nigrosin stain-ing and by the 3-(4,5-dimethylthiazol-2-yl-2,5-diphenyltetrazolium bromide) MTTassay, actually measuring mitochondrialactivity [22]. The tetrazolium salt MTT isreduced by viable cells to a colored water-insoluble formazan salt. After it is solu-bilized, formazan can be quantified spec-trophotometrically at 585 nm. The MTTassay was performed as follows. Ten µLMTT (5 mg/mL, Sigma) in Hanks bal-anced salt solution (Invitrogen) was addedto each well and after 3.5 h of incubationat 37 ◦C, the MTT solution was replacedby 200 µL DMSO in ethanol (1/1 v/v).The plates were agitated for 15 min ona platform shaker (450 RPM) to dissolvethe formazan crystals and subsequentlyanalyzed spectrophotometrically at both

585 nm (OD1) and 620 nm (OD2). Thelatter wavelength was used to correct forcell debris and well imperfections. Finaloptical densities obtained from formazanformation were presented as OD1 minusOD2.

2.4. Effect on infectivity of Cp. psittacifor BGM cells

Incubation of Cp. psittaci in the pres-ence of transferrins prior to inoculationallowed us to examine the direct anti-chlamydial effect on extracellular bacte-ria and to determine the maximal non-bactericidal concentration of transferrinswhich was going to be used in subse-quent adhesion and invasion assays. Bac-teria (108 TCID50/mL) were incubated at37 ◦C for 1 h in BGM culture mediumsupplemented with transferrin concentra-tions of 0.005, 0.05, 0.5 and 5.0 mg/mL.Subsequently, bacteria were washed twice(45 000× g, 45 min, 4 ◦C) in PBS and in-oculated in BGM cells which are highlysensitive artificial host cells routinely usedfor Cp. psittaci diagnosis [35]. At 36 hpost inoculation (p.i.), bacterial replicationwas quantified by an immunofluorescenceassay.

2.5. Chlamydial adhesion and invasionassay

Protein stock solutions were diluted inHD11 culture medium supplemented with5.5 mg/mL glucose (Sigma) to the max-imal non-cytotoxic and non-bactericidalconcentrations of 0.05 mg/mL ovoTF and0.5 mg/mL hLF and bLF. Additionally,0.005 mg/mL ovoTF was tested. All mediawere filter sterilized (0.22 µm, Millipore).HD11 cells, seeded at 300 000 cells/mL,were grown on sterile glass coverslips(13 mm) at the bottom of Chlamydia Tracbottles (Bibby Sterilin Ltd., Stone, UK)for 24 h at 37 ◦C and 5% CO2. The cellswere inoculated with 20 µL of Cp. psittaci

732 D.S.A. Beeckman et al.

strain 92/1293 (108 TCID50/mL) follow-ing standard procedures [35], only with-out using cycloheximide, since it couldinhibit actin polymerization during the en-try phase of the bacteria, and inoculationat 4 ◦C for 15 min (1000× g) insteadof 37 ◦C to synchronize irreversible at-tachment and entry [29]. Monolayers werewashed to remove unattached bacteria andculture medium with the appropriate con-centrations of transferrins was added. Cellsinoculated with PBS served as uninfected,untreated controls. All transferrin concen-trations were tested in duplicate and theexperiments were performed twice. Inocu-lated cells were further incubated at 37 ◦Cand 5% CO2 till the bacteria and the actinskeleton were stained at 15 min (optimaltimepoint for visualization of actin poly-merization as observed in infected non-treated controls) and at 1, 6 and 48 h p.i.by an immunofluorescence assay.

2.6. Effect on intracellular replicatingCp. psittaci

Non-cytotoxic transferrin concentra-tions of 0.005 to 5.0 mg/mL ovoTF and0.005 to 1 mg/mL for both hLF and bLFwere added to the HD11 culture mediumat 3 h p.i., when all bacteria had enteredthe cells and binary fission was about tostart. The outcome of the infection at 48 hp.i. was analyzed by immunofluorescencestaining using a scoring system from 0 to 5as previously described [34].

2.7. Immunofluorescence assays

The bactericidal effect of different trans-ferrins as well as their effect on attachment,invasion and intracellular bacterial replica-tion was examined by immunofluorescencestaining. All stainings were performed asfollows. The cells were washed twice withPBS and then fixed with 4% paraformalde-hyde (Merck, VWR, Leuven, Belgium)

and 120 mM sucrose (Merck) in PBS for30 min at room temperature. After wash-ing once more, the cells were incubatedfor 10 min with 50 mM NH4Cl in PBSat room temperature and permeabilizedin 0.05% saponin (Sigma) and 1 mg/mLBSA (Sigma) in PBS. Cp. psittaci was de-tected using a rabbit polyclonal antibodyagainst purified Cp. psittaci EB, followedby an AlexaFluor 546 labeled goat anti-rabbit conjugate (Molecular Probes) bothdiluted 1 in 200 in PBS with 1% BSA.At 15 min, 1 h and 6 h p.i., the actinskeleton was visualized using AlexaFluor488 coupled phalloidin (Molecular Probes,Invitrogen) as described by the manu-facturer. Coverslips were mounted usingMowiol (Calbiochem, VWR) with 0.01%p-phenylenediamine (PPD, Sigma) [33].

At all time points, fluorescently labeledbacteria were quantified using an EclipseTE300 fluorescence microscope (Nikon)provided with a CCD camera (Spot 3.0.1,Diagnostic Instruments Inc.). Briefly, wecounted the number of chlamydiae at15 min p.i. and/or chlamydial inclusions(from 1 h p.i. to 48 h p.i.) in 81 im-ages taken from a single slide and deter-mined the mean. This parameter was desig-nated ‘Mean Number of Inclusion formingunits’, further referred to as MNI. TheTukey HSD test was used for comparisonof MNI values over time in the presence orabsence of different transferrin concentra-tions.

Additionally, images of chlamydial at-tachment and invasion as well as of actinrecruitment were acquired using confocallaser scanning microscopy (CLSM) (Radi-ance 2000, Bio-Rad; 1000×).

3. RESULTS

3.1. Transferrin cytotoxicity assays

A preliminary experiment was car-ried out to determine the maximal non-cytotoxic concentration of ovoTF, hLF and

Anti-Cp. psittaci effect of transferrins 733

0

10

20

30

40

50

60

70

80

90

0 1 2 3 4 5 6

Time (h)

MNI

Control 0.005 ovoTF 0.05 ovoTF 0.5 hLF 0.5 bLF

Figure 1. Chlamydial adhesion and invasion assay. Mean Number of Inclusion forming units (MNI± standard error of the mean) shown from 15 min to 6 h p.i. As expected, at 1 h p.i. more bacteriahave irreversibly attached to HD11 cells and subsequently internalized the host cell, when no proteinwas applied. Treatment with either ovoTF or LF significantly inhibited this process.

bLF for HD11 chicken macrophages. Up to5 mg/mL ovoTF and up to 1 mg/mL hLFand bLF did not affect any of the cytotoxicparameters at all three time points exam-ined. However, at 36 h p.i., the MTT assayrevealed a 25% drop in cellular metabolismfor HD11 cells treated with 0.5 mg/mLbLF, increasing to 50% at 48 h p.i.

3.2. Effect on infectivity of Cp. psittacifor BGM cells

To determine the maximal non-bactericidal concentration of transferrinson Cp. psittaci, a pre-incubation experi-ment was set up exposing Cp. psittaci92/1293 to ten-fold dilutions of ovoTF,hLF and bLF prior to inoculating themon BGM cells. Pre-incubation with 0.5to 5 mg/mL ovoTF significantly loweredthe outcome of the infection at 36 h p.i.(data not shown). For both lactoferrins,5 mg/mL reduced the infection outcome,albeit not significantly as compared to the

controls. Therefore, the non-bactericidalconcentrations used in the followingadhesion and invasion assays were 0.005and 0.05 mg/mL ovoTF and 0.5 mg/mLhuman or bovine LF. Thus, the bactericidalactivity of ovoTF was higher than forhuman and bovine LF.

3.3. Chlamydial adhesion and invasionassay

The effect on bacterial irreversibleadhesion and invasion was investigatedby addition of non-cytotoxic and non-bactericidal concentrations of ovotransfer-rin and both lactoferrins at the time ofinoculation of chicken macrophages. At15 min p.i., all transferrin concentrations,except 0.005 mg/mL ovoTF significantlylowered MNI values as compared to un-treated bacterial cultures (Fig. 1). At 1 hp.i., all transferrins significantly (P < 0.05)inhibited bacterial irreversible attachmentand entry and for ovoTF a dose-dependent

734 D.S.A. Beeckman et al.

inhibition could be demonstrated. At theend of the early stage of the developmen-tal cycle, at 6 h p.i., just before chlamydialreplication was about to start, MNI val-ues were significantly (P < 0.05) loweredfor all transferrins except for 0.005 mg/mLovoTF, as compared to the untreated con-trol. The inhibitory effects of 0.05 mg/mLovoTF and 0.5 mg/mL of both LF wereequal.

Confocal images, taken at 15 min(Fig. 2b) clearly demonstrated host cellactin polymerization at the bacterial en-try site. Inoculation in the presence ofnon-cytotoxic concentrations (0.005 or0.05 mg/mL, Fig. 2c) ovoTF significantlyinhibited actin polymerization in a dose-dependent manner (Fig. 1). The same wasobserved at 1 h p.i. (data not shown).Interestingly, addition of bactericidal con-centrations of transferrins resulted in theabsence of actin polymerization at the bac-terial entry site, although the cells weredefinitely still able to polymerize actin(Fig. 2d).

Focusing on 48 h p.i., both ovoTF con-centrations used and 0.5 mg/mL LF sig-nificantly diminished the outcome of theinfection (P < 0.05) in HD11 cells whencompared to the control (Tab. I).

Overall, ovoTF was more effective thanhuman and bovine LF in inhibiting HD11infection since both attachment and in-vasion were inhibited at a 10-fold lowerconcentration.

3.4. Effect on intracellular replicatingCp. psittaci

The addition of non-cytotoxic concen-trations of transferrins at 3 h p.i. apparentlyhad no effect on subsequent intracellu-lar chlamydial replication, since scores forall treatments ranged from 4 to 5 (datanot shown), as well as for the control(score 5).

4. DISCUSSION

Bactericidal effects have previouslybeen described for both ovotransfer-rin [17, 31] and lactoferrin (reviewedin [32]), but their effect on chlamydial bac-teria has not yet been investigated. Thebactericidal effect of ovoTF and humanand bovine LF on the avian pathogenCp. psittaci was evaluated by incubatingCp. psittaci with different concentrationsof these natural anti-microbial proteinsprior to infecting BGM cells. Therefore,all proteins were iron-saturated to avoidthe occurrence of an anti-chlamydial ef-fect by iron-sequestration. All transferrinsexhibited anti-chlamydial activity directlytowards extracellular organisms, resultingin a lower infection rate of BGM cells ascompared to the infection rate establishedby untreated control bacteria. However, thebactericidal activity of ovoTF was higherthan for human and bovine LF. Concen-trations of 0.5 mg/mL bLF and hLF weretoo low to significantly lower the infec-tion rate of BGM cells, while the sameamount of ovoTF could do so. There isno obvious explanation for the differencesin anti-chlamydial activity. Maybe ovoTF,exclusively found in birds, is simply moreadapted to destroy avian pathogens.

Chlamydiaceae are believed to medi-ate their attachment through ligands suchas MOMP, hsp70 or OmcB (reviewedin [16]). Considerable evidence suggeststhat electrostatic interactions are involvedin an initial, reversible interaction withthe eukaryotic host cell for many, but notall strains and species of chlamydiae [26,28, 30]. Initial reversible attachment toheparan sulfate-like glycosaminoglycans isfollowed by irreversible binding to an un-known secondary receptor and this leads tothe recruitment of actin to the attachmentsite, formation of an actin-rich, pedestal-like structure, and finally internalization ofthe bacteria [10]. Indeed, confocal imagestaken at 15 min p.i. of HD11 cells in

Anti-Cp. psittaci effect of transferrins 735

10 µm

A

B

C

D

Figure 2. Confocal images of infected HD11 chicken macrophages. Chlamydophila psittaci92/1293 was allowed to attach to HD11 cells by centrifugation at 4 ◦C and whereafter the tempera-ture was shifted to 37 ◦C for 15 min to allow chlamydial induction of host cell actin polymerizationin the presence of lactoferrin or ovotransferrin. Representative confocal images are shown: (a) un-infected control, (b) infected, but untreated control, (c) infection in the presence of 0.05 mg/mLovotransferrin, (d) infection in the presence of 5 mg/mL ovotransferrin. From left to right: actinstaining (green), Cp. psittaci staining (red), merged. Arrows indicate zones of decreased actin re-cruitment when treated with ovotransferrin. (Please consult www.vetres.org for a color version ofthis figure.)

736 D.S.A. Beeckman et al.

Table I. Mean Number of Inclusion formingunits (MNI, ± standard deviation) at 48 h p.i. inthe absence (control) or presence of lactoferrinor ovotransferrin.

Treatment Mean Number ofInclusion forming unitsa

Control 372 ± 75

0.005 mg/mL ovoTF 63a ± 27

0.05 mg/mL ovoTF 87a ± 34

0.5 mg/mL hLF 154a ± 42

0.5 mg/mL bLF 124a ± 52a Significantly different (P < 0.05) from value foruntreated control as calculated by the Tukey HSDtest.

the absence of transferrins clearly showedco-localization of attached Cp. psittaciand underlying actin recruitment, indicat-ing that, as observed earlier for Chlamy-dia (C.) trachomatis, C. muridarum, Cp.caviae and Cp. pneumoniae [8–10,29], Cp.psittaci also induces a localized host cellactin recruitment upon irreversible attach-ment. However, as observed for C. tra-chomatis attachment to HeLa cells [29],irreversible attachment of Cp. psittaci toHD11 cells was not always accompaniedby actin recruitment. The latter can beexplained as follows: (1) although we at-tempted to synchronize irreversible bacte-rial attachment by centrifugation at 4 ◦C,synchronized entry could probably not beguaranteed completely, (2) perhaps not allfluorescently stained bacteria were still ac-tive and infectious, (3) maybe, as sug-gested earlier by Subtil et al. [29], actinrecruitment is transient, rendering it im-possible to observe the induction of actinrecruitment for all the invasive bacteria bythe presently used immunofluorescence as-say staining dead cells and finally, (4) Cp.psittaci also uses a microfilament indepen-dent entry mechanism, involving clathrincoated pits [37].

The effect of non-cytotoxic andnon-bactericidal ovoTF, hLF and bLFconcentrations on Cp. psittaci irreversibleadhesion and subsequent entry in HD11chicken macrophages was evaluated.Based on a former electron microscopicstudy on the intracellular developmen-tal cycle of strain 92/1293, time pointschosen to calculate the MNI per imagerepresented the attachment phase (15 minp.i.), the attachment and entry phase (1 hp.i.), the end of the internalization phase(6 h p.i.) and the outcome of the infection(48 h p.i.). Overall, non-cytotoxic andnon-bactericidal concentrations of ovoTFwere more efficient than human and bovineLF in inhibiting Cp. psittaci attachmentand entry in a homologous cell system,as also shown by confocal images takenof the chlamydial entry 15 min afterinoculation. Thus, in this experimentalavian Cp. psittaci system, ovoTF was moreactive than bLF or hLF. Analogous resultswere obtained by Giansanti et al. [14]evaluating the anti-viral activity of ovoTF,hLF and bLF on inhibiting avian Marek’sdisease virus infection in chicken embryofibroblast cultures.

Recently, a functional Type III secretionsystem (T3SS) was described in Chlamy-diaceae, including Cp. psittaci1, and inC. trachomatis and Cp. caviae, the translo-cated actin-recruiting phosphoprotein Tarpwas shown to be essential for actin re-cruitment that coincides with bacterialendocytosis [11]. In the present study, in-hibition of Cp. psittaci entry by transfer-rins was accompanied by the reduction

1 Beeckman D., Geens T., Vanrompay D., Iden-tification and characterization of a type III se-cretion system in Chlamydophila psittaci, in:Longbottom D., Rocchi M. (Eds.), Diagnosis,pathogenesis and control of animal chlamydo-sis, Proc. Fourth Annual Workshop of COSTAction 855, Animal chlamydiosis and zoonoticimplications, Edinburgh, Midlothian, Scotland,UK, 2006, p. 47.

Anti-Cp. psittaci effect of transferrins 737

or even absence of actin recruitment atthe bacterial entry site, although the cellwas still fully capable to polymerize actinas demonstrated by the presence of cellspreading (Fig. 2d). Since lactoferrin bindsLPS in the outer membrane of Gram-negative bacteria, giving the membranea more rigid character in this way [7],the stability of the T3SS and secretion ofType III secretion effector proteins such asTarp might be affected in a negative way.Moreover, Ochoa et al. [23, 24] showedthat lactoferrin inhibits the attachment ofenteropathogenic E. coli (EPEC) to hostcells as well as actin polymerization bydisruption of the EPEC T3SS transloconproteins EspA, B and D. Similarly, treat-ment of Shigella flexneri with lactofer-rin diminished the infection of host cellsby releasing and subsequently degradingthe T3SS translocon proteins IpaB andIpaC, both homologues to the chlamydialCopB and CopD [15]. Based on the se-quence and structural homology betweenhuman lactoferrin and ovoTF, a destabi-lizing effect on the T3SS or even a prote-olytic effect on chlamydial Type III secre-tion translocon components might explainthe present observations. Currently, exper-iments addressing the effect of transferrinson recombinantly produced T3SS translo-con proteins are in progress.

Focusing on the outcome of the infec-tion at 48 h p.i., both ovoTF concentra-tions used and 0.5 mg/mL LF significantlydiminished the outcome of the infectionin HD11 macrophages as could be ex-pected since bacterial attachment and entrywas significantly reduced. Moreover, thisinhibition of bacterial proliferation wasstronger in the presence of ovoTF, therebyindicating a species-specific effect exertedby the proteins studied.

In conclusion, we provided evidencefor an anti-chlamydial effect exerted bylactoferrin and its avian homologue ovo-transferrin. Our results clearly showeda bactericidal effect of both ovotrans-

ferrin and lactoferrins on extracellu-lar Cp. psittaci. However, ovotransferrinwas apparently more effective in killingCp. psittaci. Moreover, we demonstratedthat Cp. psittaci internalization in HD11chicken macrophage cultures is more ef-ficiently inhibited by ovotransferrin thanlactoferrins. The inhibition may be dueto the blocking of actin recruitment to-wards the bacterial entry site as a resultof destabilizing the bacterial T3SS or ofproteolytic degradation of Type III secre-tion effector proteins. Based on the presentfindings, ovotransferrin could be used asa natural anti-chlamydial agent. However,further research towards the practical ap-plication in preventing Cp. psittaci infec-tions in birds is needed.

ACKNOWLEDGEMENTS

Delphine Beeckman is a research assistantof the Research Foundation – Flanders (FWO –Vlaanderen) and this institution is acknowl-edged for providing a Ph.D. grant. The workwas supported by a grant of the Belgian Min-istry of Health (RT-05/08 MINSPEC). The au-thors gratefully acknowledge Geert Meesen andWinnok De Vos for assisting with digital titra-tion and confocal laser scanning microscopy,respectively.

REFERENCES

[1] Aguilera O., Quiros L.M., Fierro J.F.,Transferrins selectively cause ion effluxthrough bacterial and artificial membranes,FEBS Lett. (2003) 548:5–10.

[2] Andersen A.A., Vanrompay D., Avianchlamydiosis, Rev. Sci. Tech. (2000)19:396–404.

[3] Appelmelk B.J., An Y.Q., Geerts M., ThijsB.G., de Boer H.A., MacLaren D.M., deGraaff J., Nuijens J.H., Lactoferrin is a lipidA-binding protein, Infect. Immun. (1994)62:2628–2632.

[4] Arnold R.R., Cole M.F., McGhee J.R., Abactericidal effect for human lactoferrin,Science (1977) 197:263–265.

738 D.S.A. Beeckman et al.

[5] Ashida K., Sasaki H., Suzuki Y.A.,Lonnerdal B., Cellular internalizationof lactoferrin in intestinal epithelial cells,Biometals (2004) 17:311–315.

[6] Beug H., von Kirchbach A., DoderleinG., Conscience J.F., Graf T., Chickenhematopoietic cells transformed by sevenstrains of defective avian leukemia virusesdisplay three distinct phenotypes of differen-tiation, Cell (1979) 18:375–390.

[7] Brandenburg K., Jurgens G., Muller M.,Fukuoka S., Koch M.H., Biophysical char-acterization of lipopolysaccharide and lipidA inactivation by lactoferrin, Biol. Chem.(2001) 382:1215–1225.

[8] Carabeo R.A., Grieshaber S.S., Fischer E.,Hackstadt T., Chlamydia trachomatis in-duces remodeling of the actin cytoskeletonduring attachment and entry into HeLa cells,Infect. Immun. (2002) 70:3793–3803.

[9] Carabeo R.A., Grieshaber S.S., HasenkrugA., Dooley C., Hackstadt T., Requirementfor the Rac GTPase in Chlamydia trachoma-tis invasion of non-phagocytic cells, Traffic(2004) 5:418–425.

[10] Clifton D.R., Fields K.A., Grieshaber S.S.,Dooley C.A., Fischer E.R., Mead D.J.,Carabeo R.A., Hackstadt T., A chlamydialtype III translocated protein is tyrosine-phosphorylated at the site of entry and asso-ciated with recruitment of actin, Proc. Natl.Acad. Sci. USA (2004) 101:10166–10171.

[11] Clifton D.R., Dooley C.A., Grieshaber S.S.,Carabeo R.A., Fields K.A., Hackstadt T.,Tyrosine phosphorylation of the chlamydialeffector protein Tarp is species specific andnot required for recruitment of actin, Infect.Immun. (2005) 73:3860–3868.

[12] Ellison R.T. III, Giehl T.J., LaForce F.M.,Damage of the outer membrane of entericgram-negative bacteria by lactoferrin andtransferrin, Infect. Immun. (1988) 56:2774–2781.

[13] Ellison R.T. III, LaForce F.M., Giehl T.J.,Boose D.S., Dunn B.E., Lactoferrin andtransferrin damage of the gram-negativeouter membrane is modulated by Ca2+ andMg2+, J. Gen. Microbiol. (1990) 136:1437–1446.

[14] Giansanti F., Rossi P., Massucci M.T., BottiD., Antonini G., Valenti P., Seganti L.,Antiviral activity of ovotransferrin disclosesan evolutionary strategy for the defensive ac-

tivities of lactoferrin, Biochem. Cell Biol.(2002) 80:125–130.

[15] Gomez H.F., Ochoa T.J., Carlin L.G., ClearyT.G., Human lactoferrin impairs virulenceof Shigella flexneri, J. Infect. Dis. (2003)187:87–95.

[16] Hackstadt T., Cell Biology, in: StephensR.S. (Ed.), Chlamydia: Intracellular Biology,Pathogenesis, and Immunity, ASM Press,Washington, DC, 1999, pp. 101–138.

[17] Ibrahim H.R., Iwamori E., Sugimoto Y.,Aoki T., Identification of a distinct antibac-terial domain within the N-lobe of ovo-transferrin, Biochim. Biophys. Acta (1998)1401:289–303.

[18] Ibrahim H.R., Sugimoto Y., Aoki T.,Ovotransferrin antimicrobial peptide(OTAP-92) kills bacteria through a mem-brane damage mechanism, Biochim.Biophys. Acta (2000) 1523:196–205.

[19] Keung W.M., Azari P., Structure and func-tion of ovotransferrin. II. Iron-transferringactivity of iron-binding fragments of ovo-transferrin with chicken embryo red cells, J.Biol. Chem. (1982) 257:1184–1188.

[20] Martin-Mateo M.C., Planas J., Conalbuminand serum iron transport in birds. I. Study inthe chicken (Gallus domesticus), Rev. Esp.Fisiol. (1965) 21:1–7.

[21] Metz-Boutigue M.H., Jolles J., MazurierJ., Schoentgen F., Legrand D., Spik G.,Montreuil J., Jolles P., Human lactotrans-ferrin: amino acid sequence and structuralcomparisons with other transferrins, Eur. J.Biochem. (1984) 145:659–676.

[22] Mosmann T., Rapid colorimetric assay forcellular growth and survival: applicationto proliferation and cytotoxicity assays, J.Immunol. Methods (1983) 65:55–63.

[23] Ochoa T.J., Noguera-Obenza M., Ebel F.,Guzman C.A., Gomez H.F., Cleary T.G.,Lactoferrin impairs type III secretory systemfunction in enteropathogenic Escherichiacoli, Infect. Immun. (2003) 71:5149–5155.

[24] Ochoa T.J., Clearly T.G., Lactoferrin disrup-tion of bacterial type III secretion systems,Biometals (2004) 17:257–260.

[25] Rossi P., Giansanti F., Boffi A., Ajello M.,Valenti P., Chiancone E., Antonini G., Ca2+

binding to bovine lactoferrin enhances pro-tein stability and influences the release of

Anti-Cp. psittaci effect of transferrins 739

bacterial lipopolysaccharide, Biochem. CellBiol. (2002) 80:41–48.

[26] Stephens R.S., Koshiyama K., Lewis E.,Kubo A., Heparin-binding outer membraneprotein of chlamydiae, Mol. Microbiol.(2001) 40:691–699.

[27] Stevens L., Egg white proteins, Comp.Biochem. Physiol. B (1991) 100:1–9.

[28] Su H., Raymond L., Rockey D.D., FischerE., Hackstadt T., Caldwell H.D., A recom-binant Chlamydia trachomatis major outermembrane protein binds to heparan sul-fate receptors on epithelial cells, Proc. Natl.Acad. Sci. USA (1996) 93:11143–11148.

[29] Subtil A., Wyplosz B., Balana M.E., Dautry-Varsat A., Analysis of Chlamydia caviaeentry sites and involvement of Cdc42 andRac activity, J. Cell Sci. (2004) 117:3923–3933.

[30] Taraktchoglou M., Pacey A.A., Turnbull J.E.,Eley A., Infectivity of Chlamydia trachoma-tis serovar LGV but not E is dependenton host cell heparan sulfate, Infect. Immun.(2001) 69:968–976.

[31] Valenti P., Antonini G., Von Hunolstein C.,Visca P., Orsi N., Antonini E., Studies of theantimicrobial activity of ovotransferrin, Int.J. Tissue React. (1983) 5:97–105.

[32] Valenti P., Antonini G., Lactoferrin: animportant host defence against microbial and

viral attack, Cell. Mol. Life Sci. (2005)62:2576–2587.

[33] Valnes K., Brandtzaeg P., Retardation ofimmunofluorescence fading during mi-croscopy, J. Histochem. Cytochem. (1985)33:755–761.

[34] Van Loock M., Loots K., Van HeerdenM., Vanrompay D., Goddeeris B.M.,Exacerbation of Chlamydophila psittacipathogenicity in turkeys superinfectedby Escherichia coli, Vet. Res. (2006)37:745–755.

[35] Vanrompay D., Ducatelle R., HaesebrouckF., Diagnosis of avian chlamydiosis: speci-ficity of the modified Gimenez staining onsmears and comparison of the sensitivityof isolation in eggs and three different cellcultures, Zentralbl. Veterinarmed. B (1992)39:105–112.

[36] Vanrompay D., Andersen A.A., DucatelleR., Haesebrouck F., Serotyping of Europeanisolates of Chlamydia psittaci from poultryand other birds, J. Clin. Microbiol. (1993)31:134–137.

[37] Vanrompay D., Charlier G., Ducatelle R.,Haesebrouck F., Ultrastructural changes inavian Chlamydia psittaci serovar A-, B-,and D-infected Buffalo Green Monkey cells,Infect. Immun. (1996) 64:1265–1271.

[38] Vorland L.H., Lactoferrin: a multifunctionalglycoprotein, APMIS (1999) 107:971–981.