Histoplasma Capsulatum Cyclophilin a Mediates Attachment to Dendritic Cell VLA-51

The Journal of Immunology

Histoplasma capsulatum Cyclophilin A Mediates Attachment to Dendritic Cell VLA-51

Francisco J. Gomez,2*† Robyn Pilcher-Roberts,*† Arash Alborzi,* and Simon L. Newman*

Histoplasma capsulatum (Hc) is a pathogenic fungus that replicates in macrophages (M␾). In dendritic cells (DC), Hc is killed and fungal Ags are processed and presented to T cells. DC recognize Hc yeasts via the VLA-5 receptor, whereas M␾ recognize yeasts via CD18. To identify ligand(s) on Hc recognized by DC, VLA-5 was used to probe a Far Western blot of a yeast freeze/thaw extract (F/TE) that inhibited Hc binding to DC. VLA-5 recognized a 20-kDa protein, identified as cyclophilin A (CypA), and CypA was present on the surface of Hc yeasts. rCypA inhibited the attachment of Hc to DC, but not to M␾. Silencing of Hc CypA by RNA interference reduced yeast binding to DC by 65–85%, but had no effect on binding to M␾. However, F/TE from CypA-silenced yeasts still inhibited binding of wild-type Hc to DC, and F/TE from wild-type yeasts depleted of CypA also inhibited yeast binding to DC. rCypA did not further inhibit the binding of CypA-silenced yeasts to DC. Polystyrene beads coated with rCypA or fibronectin bound to DC and M␾ and to Chinese hamster ovary cells transfected with VLA-5. Binding of rCypA-coated beads, but not fibronectin-coated beads, was inhibited by rCypA. These data demonstrate that CypA serves as a ligand for DC VLA-5, that binding of CypA to VLA-5 is at a site different from FN, and that there is at least one other ligand on the surface of Hc yeasts that mediates binding of Hc to DC. The Journal of Immunology, 2008, 181: 7106–7114.

istoplasma capsulatum (Hc)3 is a dimorphic fungal ally the blood stream, is thought to be the main mechanism for pathogen endemic to the Ohio and Mississippi River Hc dissemination into M␾-rich organs such as the spleen, liver, H Valleys. In its environmental form, Hc grows as a sa- and bone marrow (6). In immunocompetent hosts, infection probic mold with tropism for humid soils rich in nitrogen and with Hc is contained and eliminated by the development of organic material. Infection of the mammalian host is initiated by Hc-specific T cell-mediated immunity. CD4 T cells recognize inhalation of microconidia and small mycelial fragments which are Hc-derived antigenic peptides in the context of MHC class II deposited into the terminal bronchioles and alveoli of the lung. The molecules displayed in the surface of professional APCs. Ag- 37°C temperature of the host induces Hc transformation into a 2–5 specific T cells become activated, proliferate, and produce Th1 ␮M yeast, the form found in infected tissues, and in macrophages cytokines that in turn activate M␾ to exert fungistatic and fun- (M␾) and dendritic cells (DC) (1). gicidal activity (6). Histologically, cell-mediated immunity is Human monocyte-derived M␾ avidly bind to Hc yeasts and characterized by the development of granulomatous inflamma- conidia in an opsonin-independent fashion. Adhesion is medi- tion around foci infected with the fungus (7). ␤ ated by the interaction between 2 integrins, LFA-1 (CD11a/ DC, the hosts primary APCs, are abundantly present in lung CD18), complement receptor 3 (CD11b/CD18), and comple- tissues within the airway epithelium, the submucosa, and on the ment receptor 4 (CD11c/CD18) (2, 3) with the surface ligand alveolar surfaces (8). In vitro, DC avidly bind to and ingest Hc, heat shock protein 60 (HSP60) (4). As a consequence of this but do not require cytokine activation to exert phagolysosomal interaction, the fungus gains entrance to the M␾ endosome, fusion, intracellular killing, and degradation of the yeast (9, 10). which is a permissive environment for Hc survival and prolif- The molecules involved in DC-Hc adhesion are different from eration (5). M␾ migration into local lymph nodes, and eventu- the receptors used by M␾. Despite the presence of CD18 integrins on the surface of DC (11), Hc binds to the fibronectin receptor VLA-5 (9). Moreover, surface HSP60 plays no role in *Department of Internal Medicine, Division of Infectious Diseases, University of DC recognition of Hc (4). In the experiments described herein, Cincinnati College of Medicine, and †Veterans Administration Medical Center, Cin- we sought to identify the Hc surface ligand(s) involved in the cinnati, OH 45267 interaction between Hc and DC VLA-5. Freeze/thaw extracts Received for publication March 27, 2008. Accepted for publication September (F/TE) of Hc yeasts probed with purified human VLA-5 in a Far 9, 2008. Western blot identified a 20-kDa protein, cyclophilin A (CypA), The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance as a major ligand involved in DC-Hc recognition. We hypoth- with 18 U.S.C. Section 1734 solely to indicate this fact. esize that Hc CypA may play an important role in fungal-DC 1 This work was supported by U.S. Public Health Service Grants AI49358 and interaction, innate immune recognition, and the initiation of AI061298 from the National Institutes of Allergy and Infectious Diseases. cell-mediated immunity to Hc. 2 Address correspondence and reprint requests to Dr. Francisco Gomez, Division of Infectious Diseases, University of Cincinnati College of Medicine, P.O. Box 670560, Cincinnati, OH 45267. E-mail address: [email protected] Materials and Methods 3 Abbreviations used in this paper: Hc, Histoplasma capsulatum;M␾, macrophage; Yeasts DC, dendritic cell; CypA, cyclophilin A; F/TE, freeze/thaw extract; WT, wild type; FN, fibronectin; HK, heat killed; Sc, Saccharomyces cerevisiae; HSP60, heat shock Hc strain G217B was maintained in yeast form by serial passage on brain- protein 60; CHO, Chinese hamster ovary; 1D, one dimensional; 2D, two dimensional; heart infusion agar as described previously (3). Yeasts were grown in AI, attachment index; PPIase, peptidyl-prolyl cis-trans isomerase; Mip, M␾ infectiv- Histoplasma M␾ medium (12) at 37°C with orbital shaking at 150 rpm. For ity potentiator. binding assays, log-phase yeasts were heat-killed (HK) at 65°C for1hand www.jimmunol.org The Journal of Immunology 7107 stored at 4°C in PBS containing 0.05% sodium azide. HK yeasts were ufacturer’s instructions. rCypA differs from the native sequence by hav- labeled with FITC, washed, and resuspended in HBSS containing 20 mM ing an extra phenylalanine residue at its carboxyl terminus that was HEPES and 0.25% BSA (HBSA) as described previously (3). introduced to enhance intein-mediated cleavage.

Preparation of Hc F/TE Antibodies A F/TE enriched with Hc surface proteins was prepared as described pre- Rabbit polyclonal anti-Toxoplasma CypA was a gift from Dr. J. Aliberti viously (4). Hc yeasts were grown to log phase and harvested via centrif- (National Institute of Allergy and Infectious Diseases, Bethesda, MD). ugation. The pellets were frozen and thawed twice over 48 h. Yeasts were Rabbit polyclonal anti-Hc-CypA was prepared commercially by Antibod- removed by centrifugation, and the supernatant containing surface proteins ies by two immunization injections with 100 ␮g of rCypA. Specific anti- was sterile filtered and stored at 4°C. Protein concentration was determined CypA Abs were affinity purified from rabbit antiserum by passage through by the Bradford method (Bio-Rad). a column of rCypA immobilized on Sepharose, followed by elution of the bound protein at pH 3.0. The specificity of the polyclonal Ab was tested by Far Western blot analysis with purified VLA-5 probing a F/TE and total protein extract of Hc on Western blots. In both Five hundred micrograms of purified human VLA-5 (Chemicon Interna- cases, only a single band of 20 kDa was observed (data not shown). For tional) was biotin labeled using water-soluble sulfo-NHS-LC-biotin flow cytometry experiments, preimmune rabbit serum was used as a neg- (Pierce) following the manufacturer’s instructions. For one-dimensional ative control. The natural Abs present in rabbit serum against common (1D) analysis, 20 ␮g of F/TE was electrophoresed in a 12% SDS-PAGE fungal cell wall polysaccharides were removed by two absorptions with ϫ 6 gel. For two-dimensional (2D) analysis, 50 ␮g of F/TE was electrofocused 2 10 viable Saccharomyces cerevisiae (Sc) yeasts (4). As an additional on 7-cm Immoboline isoelectric focusing strips, with a linear pH 4–9 (Am- control, the anti-CypA Abs in immune rabbit serum were removed by ersham Pharmacia Biotech). After focusing was complete, the isoelectric absorption with rCypA immobilized on Sepharose. focusing strips were electrophoresed in a 8–16% precast SDS-PAGE gel (Bio-Rad). For both one-dimensional and two-dimensional gels, separated CypA- and fibronectin (FN)-coated polystyrene beads proteins and molecular mass standards were electrotransferred to a nitro- ϫ 9 ␮ cellulose membrane (Bio-Rad) and blocked for 30 min in HBSS containing One 10 of 2- m fluorescent polystyrene beads (Sigma-Aldrich) were 1% BSA and 0.1% Tween 20. Membranes then were incubated for2hat washed in pH 6.0 phosphate buffer and then incubated with 1 mg/ml rCypA room temperature with 20 ␮g/ml biotinylated VLA-5 dissolved in HBSS or FN (Sigma-Aldrich) in the same buffer for 12 h at room temperature. containing 1% BSA and 0.2% Tween 20. After extensive washing, mem- Approximately 30% of rCypA or FN was bound to the surface of the beads. branes were developed by incubation with a 1/1000 dilution of alkaline- Subsequently, the beads were blocked and washed five times in 10 mg/ml phosphatase conjugated anti-biotin mAb BN-34 (Sigma-Aldrich). Bands BSA in PBS. Control beads were coated with BSA using the same proce- then were visualized by incubation with chromogenic NBT-5-bromo-4- dure. The presence of surface-bound rCypA was confirmed by flow chloro-3-indolyl phosphate (BCIP). cytometry.

Identification, cloning, and expression of Hc CypA Preparation of CypA-depleted F/TE Fifty milligrams of F/TE was resolved by preparative isoelectrofocusing F/TE (500 ␮l) was incubated twice with 100 ␮l of rabbit anti-CypA im- using a Rotofor apparatus (Bio-Rad). Fractions were tested for the presence mobilized on Sepharose to deplete the CypA. As a control, F/TE was of VLA-5 ligand by running aliquots in 1D SDS-PAGE followed by Far incubated with rabbit IgG-Sepharose. The effectiveness of the depletion Western blot analysis. Fractions containing the VLA-5-reactive band were was ascertained by Western blot analysis of the FT/E supernatants. pooled and resolved again by preparative isoelectrofocusing using the small cell system of the Rotofor. Positive fractions were resolved in a Silencing of CypA expression in Hc preparative 12% SDS-PAGE gel and electroblotted onto polyvinylidene difluoride (PVDF) membranes (Immobilon-P; Millipore). Protein bands The uracil auxotrophic Hc strain UC7 was provided by Dr. G. Smulian were visualized with Coomassie brilliant blue, and the band with the ap- (University of Cincinnati, Cincinnati, OH). The CypA-silencing construct propriate MW was submitted for N-terminal Edman degradation sequenc- was constructed in the telomeric plasmid pCR83 provided by Dr. W. Gold- ing (Midwest Analytical). man (Washington University School of Medicine, St. Louis, MO). This To obtain internal sequence information, PVDF- immobilized protein plasmid contains telomeric sequences and the complementation marker was chemically cleaved with cyanogen bromide. Peptide fragments URA5 from Podospora anserina. A 600-bp inverted repeat encompassing were eluted from the membrane by organic extraction with acetonitrile, all of the CypA-coding region was placed in front of the P-CBP1 promoter and resolved in a 10–20% Tris-Tricine SDS-PAGE, followed by elec- and terminator. The plasmid then was introduced into UC7 by electropo- troblotting and Edman degradation sequencing as described above. ration and transformants were allowed to grow in minimal medium. A total Readable sequence was used in BLAST-X searches of GenBank and of 30 colonies was expanded and further analyzed. The effectiveness of the SwissProt databases and revealed homology to the amino-terminal se- silencing was verified by quantitative RT-PCR using primers specific for quences of CypA proteins from several fungal species. The typical mo- CypA sequences and by Western blot using the rabbit anti-CypA lecular mass of cyclophilins, 18–22 kDa, was consistent with the pro- polyclonal Ab. tein identified on Far Western blots. A search of the Hc genome project identified a putative-coding sequence in contig_11286 HCG217B. The Biotinylation and precipitation of surface proteins on Hc yeasts sequence was analyzed with the SoftBerry GeneFinder on-line server (http://www.softberry.com/berry.phtml) that correctly identified the Hc Log-phase Hc yeasts (2 ϫ 108) were suspended in 1 ml of PBS or PBS CypA coding sequence, the presence of three introns, and predicted a containing 10 mg/ml EZ-link sulfo-NHS-LC-biotin (Pierce) for2hat4°C. 181-aa protein with a molecular mass of 19 kDa and a pI of 7.83. The The yeasts then were washed with TBS to quench unreacted NHS-biotin. molecular identity of the original band recognized by VLA-5 was con- Surface proteins were extracted by two freeze thaw cycles as described firmed by in gel trypsin digestion and MALDI-TOF mass spectroscopy above. From the resulting F/TE, CypA was immunoprecipitated using 50 (Voyager DE-PRO; Applied Biosystems). ␮l of rabbit anti-CypA immobilized on Sepharose. F/TE and immunopre- CypA cDNA was prepared by RT-PCR using Pfu Ultra High Fidelity cipitated material then were electrophoresed on a 12% SDS-PAGE gel, DNA polymerase (Stratagene). The sense primer was GTGGTCATA electroblotted, and the presence of covalently bound biotin was detected by TGACGAGAACTTTCTTCGAG corresponding to the amino-terminal Western blot analysis using alkaline phosphatase-labeled anti-biotin mAb sequence of mature Hc CypA and introducing a NdeI site to facilitate BN-34 (Sigma-Aldrich). cloning. The antisense primer was GGTGGTTGCTCTTCCGCAAAA AACCTGACCACAGTTGACGATC, corresponding to the carboxyl Detection of CypA on the surface of Hc yeasts by flow cytometry terminus of the gene and introducing a SapI site to facilitate cloning. cDNA was cloned between the NdeI and SapI sites of the pTWIN vector Log-phase Hc yeasts (5 ϫ 105) were suspended in staining buffer (PBS (New England Biolabs). In this expression format, the CypA sequence containing 2% BSA and 0.1 M sodium azide) and incubated with equiv- is fused in frame with the Mxe GyrA intein, which catalyzes cleavage alent concentrations of the following Abs: preimmune rabbit IgG absorbed of the recombinant product in the presence of thiol reagents, followed with viable Sc, anti-CypA IgG absorbed with Sc, anti-CypA IgG absorbed by a chitin-binding protein, which is used for purification of the recom- with Sc and rCypA-Sepharose, and anti-CypA IgG absorbed with Sc and binant protein. The recombinant construct was expressed in Escherichia BSA-Sepharose. PBS was used as a negative control. After1hat4°C, the coli and affinity purified on chitin resin columns according to the man- yeasts were washed five times with PBS and then were incubated with 7108 Hc CypA ADHESION TO DC VLA-5

PE-labeled goat anti-rabbit IgG (BD Biosciences Pharmingen) for an additional hour. After further washing, the yeasts were fixed in 1% paraformaldehyde. All samples were analyzed by flow cytometry on a FACSCali- bur flow cytometer (BD Biosciences) and the acquired data were analyzed with FCS Express (De Novo Software). Preparation of human M␾ and DC Human monocytes were purified from blood obtained from the Hox- worth Blood Center (Cincinnati, OH). After partial purification via dex- tran sedimentation and Ficoll-Hypaque centrifugation, monocytes were separated from lymphocytes by positive selection with anti-CD14 and EasySep magnetic nanoparticles (StemCell Technologies) according to the manufacturer’s instructions. M␾ were obtained by culture of monocytes at 1 ϫ 106/ml in Teflon beakers in RPMI 1640 containing 15% human serum, 100 U/ml penicillin, and 100 ␮g/ml streptomycin (Sig- ma-Aldrich) for 5–7 days (3). DC were obtained by culture of the monocytes in 6-well tissue culture plates (Corning-Costar) at 6.5 ϫ 105/ml in RPMI 1640 containing 200 mM glutamine, 50 ␮M 2-ME (Sigma-Al- drich), 10% heat-inactivated FCS (Life Technologies), 50 ng/ml kana- mycin (Sigma-Aldrich), 1% nonessential amino acids (BioWhittaker), FIGURE 1. Far Western blots of VLA-5 binding to a F/TE prepared and 1% pyruvate (BioWhittaker). Human rGM-CSF (115 ng/ml; Pep- from Hc yeasts. Twenty micrograms of Hc F/TE were separated in a roTech) and human rIL-4 (50 ng/ml; PeproTech) also were added to 8–16% SDS-PAGE gel (1D; panel A). For 2D analysis, 50 ␮g of F/TE each well and DC were studied after 6–8 days of culture (9). The was resolved by isoelectric focusing in a pH 4–9 gradient strip, fol- acquisition and use of human blood cells for these studies has been lowed by SDS-PAGE in a 8–16% gel (2D; panel B). Proteins were approved by the Internal Review Board of the University of Cincinnati transferred to PVDF membranes, blocked with 2% BSA in TBS con- College of Medicine. taining 0.2% Tween 20, 5 mM Ca2ϩ,and5mMMg2ϩ (TBS-BSA) and Chinese hamster ovary cells (CHO) transfected with VLA-5 then incubated with 25 ␮g/ml biotinylated VLA-5. The blots then were (CHO VLA-5) incubated with alkaline phosphatase-conjugated anti-biotin and developed with NBT-BCIP. CHO VLA-5 cells and nontransfected CHO cells (a gift from Dr. R. Juliano, University of North Carolina, Chapel Hill, NC) were cultured in ␣MEM medium (Life Technologies) containing 10% FBS and 10 ␮g/ml gentamicin. Five hundred micrograms of geneticin/ml (G418; Life Tech- To determine the identity of the 20-kDa ligand, the F/TE was nologies) was included with the CHO VLA-5 cells to maintain the trans- run in 2D SDS-PAGE gels, the appropriate spot cut out, and Ed- fection. Cells were harvested by scraping after reaching 80% confluence man degradation protein sequencing performed on the extracted and were passaged or used in binding experiments with Hc yeasts or polystyrene beads. protein. The amino terminus and one of the cyanogen bromide- generated internal fragments produced the sequence: TRTFFE ␾ Binding of Hc yeasts and polystyrene beads to DC, M , and VEYA. Two of the other internal fragments produced the sequence CHO VLA-5 cells ANAGPNTNGS. BLAST-X searches of translated GenBank and DC, M␾, or CHO VLA-5 cells were suspended to 2.5 ϫ 105 cells/ml in SwissProt databases revealed homology to the sequences of CypA HBSA containing 2% aprotinin and 5 ␮l of cells were allowed to adhere in proteins from several fungal species. The typical molecular mass the wells of a Terasaki culture plate (Miles) that had been coated with 1% of CypA, 18–22 kDa, was consistent with the protein identified on human serum albumin. Nonadherent cells were removed by washing and 5 ␮l of test proteins or HBSA as a control then were added to each well. Far Western blots. After 30 min of incubation at 37°C, 5 ␮l of FITC-labeled HK Hc yeasts A search of the Hc genome project identified a putative-cod- (5 ϫ 106/ml) or 5 ␮l of protein-coated polystyrene beads (2 ϫ 107/ml) in ing sequence in contig_11286 HCG217B. The sequence HBSA were added to the wells and incubated at 37°C for 30 min. Non- ANAGPNTNGS also was contained in the coding sequence, but adherent yeasts or beads were removed by washing with HBSS and the is interrupted by an intron. The contig was analyzed using the monolayers were fixed with 1% paraformaldehyde overnight. Cell-bound yeasts or beads were quantified by phase-contrast and fluorescent micros- SoftBerry GeneFinder on-line server (http://www.softberry. copy on an inverted microscope (Diaphot; Nikon) (4). The results are ex- com/berry.phtml), which correctly identified the open reading pressed as the attachment index (AI), the total number of yeasts or beads frame, the presence of three introns, and a 181-aa protein with bound to 100 cells, or the percent inhibition of binding (1 Ϫ experimental ϫ a molecular mass of 19 kDa and a pI of 7.83 (supplemental Fig. AI/control AI 100). All experiments were performed in duplicate or 4 triplicate. 1). The location of the introns and the transcription start site of the gene were confirmed by RT- PCR and 5Ј RACE analysis Results (data not shown). DC VLA-5 recognizes Hc CypA The identity of the 20-kDa band as CypA was confirmed by Previously, we demonstrated that adherence of Hc yeasts to DC two additional methods. Given that human CypA has 61% iden- is inhibited by preincubation of the DC with a Hc F/TE or a Hc tity with the Hc protein, aliquots of F/TE were run in 1D SDS- F/TE immunodepleted of HSP60 (4). These findings indicated PAGE gels and probed with either VLA-5 by Far Western blot that the F/TE contains one or more moieties that are involved in or with a commercial polyclonal rabbit anti-human CypA (US the binding of Hc yeasts to DC VLA-5 that is distinct from Biologicals). Fig. 2 shows that both probes recognize a band of HSP60. To identify these ligands, we applied the same ap- identical molecular mass in the F/TE. We performed in silico proach that we had used successfully to identify HSP60 as a Hc digestion of the predicted Hc CypA using Protein Prospector ligand for M␾ CD18 (4). VLA-5 was covalently tagged with MS-Digest (http://prospector.ucsf.edu/ucsfhtml4.0/msdigest. biotin and used to probe 1D and 2D Far Western blots of Hc htm) and compared the predicted masses with the spectrum ob- F/TE. Fig. 1 shows a 20-kDa protein with a pI of ϳ8.0 as a tained by trypsin digestion/MALDI-TOF spectroscopy. Se- major VLA-5-binding protein present in F/TE. The presence of quence coverage of 25% was observed, indicating a positive 5 mM EDTA appropriately abrogated recognition of the band, identification (data not shown). as the binding of Hc yeasts to DC VLA-5 is calcium dependent (9) (data not shown). 4 The online version of this article contains supplemental material. The Journal of Immunology 7109

FIGURE 2. Purified VLA-5 and Ab to CypA recognize the same molecular mass moiety in Hc F/TE. Several lanes of a 12% SDS-PAGE gel FIGURE 3. CypA is on the surface of Hc yeasts. Surface molecules of were loaded with prestained molecular mass markers or 20 ␮gofHc log-phase Hc yeast were covalently labeled by incubation with EZ-link F/TE. After electrophoresis and electroblotting to PVDF membranes, sulfo-NHS-LC-biotin. Proteins from biotin-labeled and unlabeled yeasts the blot was cut in half. Blot A was blocked, incubated with TBS-BSA were released by freeze-thaw cycles as described in Materials and Meth- containing 25␮g/ml biotinylated VLA-5, and developed with alkaline ods. Aliquots of 10 ␮g of unlabeled and labeled F/TE were loaded in lanes phosphatase-conjugated anti-biotin. Blot B was incubated with a poly- A and C, respectively. To specifically detect CypA among the labeled pro- clonal rabbit anti-human CypA Ab, followed by alkaline phosphatase- teins, labeled and unlabeled extracts were incubated with 50 ␮l of rabbit conjugated rat anti-rabbit IgG. Bands were visualized with NBT-BCIP. anti-CypA-conjugated Sepharose. After extensive washing with PBS con- The blots were aligned using the prestained molecular mass markers as taining 1% Triton X-100, the Sepharose beads were boiled with an equal reference. volume of SDS-PAGE sample buffer. The immunoprecipitation supernatants of unlabeled or labeled material were loaded in lanes B and D, respectively. After electrophoresis and electroblotting, biotinylated moieties CypA localization to the surface of Hc yeasts were visualized by incubation with alkaline phosphatase-conjugated anti- Although identification of CypA in F/TE of Hc yeasts suggests that biotin and NBT-BCIP. this molecule must be present on the yeasts’ surface, we sought to confirm this localization in two additional experiments. In the experiment shown in Fig. 3, Hc surface proteins were labeled with of rCypA. F/TE was used as a positive control and rHSP60 was NHS-biotin that is impermeant to cell membranes. The labeled used as a negative control. FITC-labeled HK Hc yeasts then were proteins were released by two freeze-thaw cycles and visualized incubated with the DC for an additional 30 min. As shown in Fig. with alkaline phosphatase-labeled anti-biotin mAb. Nonbiotiny- 5, rCypA inhibited the binding of Hc to DC in a concentration- lated Hc yeasts were used as a negative control (Fig. 3, lane A: dependent manner. In contrast, rCypA did not block the binding of nonbiotinylated F/TE extract; lane B: nonbiotinylated F/TE extract immunoprecipitated with anti-CypA), demonstrating that no protein in Hc F/TE is naturally biotinylated. A complex mixture of Hc surface proteins was revealed on the biotinylated yeasts (Fig. 3, lane C). CypA then was specifically immunoprecipitated from the mixture and the biotin tag was visualized by colorimetric assay (Fig. 3, lane D). In a second experiment, CypA was detected on the surface of Hc yeasts by flow cytometry. Because even preimmune rabbit serum contains natural Abs reactive to Hc yeasts, both preimmune serum and immune serum were absorbed with Sc yeasts to reduce background binding (4). This treatment effectively eliminated the non- specific Hc surface reactivity present in preimmune rabbit serum, but not the anti-CypA immune serum (Fig. 4, lines A and B, respectively). To further confirm the specificity of the anti-CypA serum, an aliquot of the serum was absorbed with rCypA–conjugated Sepharose. This treatment effectively abrogated the recognition of CypA in a Western blot (data not shown) and by flow cytometry (Fig. 4, line C). Absorption of the immune serum with BSA-conjugated Sepharose did not abrogate recognition of CypA FIGURE 4. Analysis of CypA on Hc yeasts by flow cytometry. Aliquots of 5 ϫ 105 Hc yeasts were incubated with saturating concentrations of (data not shown). Sc-adsorbed preimmune rabbit serum (line A) or Sc-adsorbed anti-CypA rCypA inhibits Hc yeast binding to DC (line B). The yeasts then were incubated with PE-labeled goat anti-rabbit IgG. To confirm the specificity of the anti-CypA Ab, it was preincubated To determine whether rCypA inhibited the binding of Hc yeasts to with rCypA-conjugated Sepharose (line C). The fluorescence of Hc yeasts DC, DC were preincubated for 30 min with varying concentrations was quantified with a FACSCalibur flow cytometer. 7110 Hc CypA ADHESION TO DC VLA-5

FIGURE 5. Inhibition of the binding of Hc yeasts to DC by rCypA. DC were harvested and adhered in Terasaki plates for2hat37°C. After washing, the monolayers were preincubated for 30 min. with HBSA (CO), 5 ␮l of F/TE, 10 ␮g of HSP60, or varying concentrations of rCypA. Hc yeasts were added for 30 min and the monolayers were washed to remove nonadherent yeasts. After overnight fixation in paraformaldehyde at 4°C, bound yeasts were quantified by phase-contrast and fluorescent microscopy as described in Materials and Methods. The data are presented as the mean Ϯ SEM (n ϭ 2–5) of the AI. CO, Control.

Hc yeasts to M␾ (data not shown). As expected, rHSP60 did not inhibit Hc binding to DC.

Binding of CypA-silenced yeasts to DC and M␾ Although VLA-5 only recognized a major single band on Far Western blots, this observation does not prove that CypA is the only possible Hc ligand for VLA-5. Therefore, to determine whether other surface moieties on Hc yeasts might mediate binding to DC, we used RNA interference to silence CypA gene expression. Silencing of CypA in several individual strains was confirmed via Western blotting of F/TE for the expressed protein (Fig. 6A). As an additional control, we also tested the F/TE from WT FIGURE 6. Binding of CypA-silenced Hc yeasts to DC and M␾. A, Hc and CypA-silenced strains for the expression of HSP60. The in- strains containing a CypA-silencing construct or parent strains were grown tensity of the HSP60 bands was comparable between WT and si- to log phase. Twenty-five micrograms of F/TE from parent strains (G217B lenced strains (supplemental Fig. 2). and UC7), silenced strains, or 200 ng of rCypA were separated in a 12% We then tested the ability of the CypA-silenced Hc strains to SDS-PAGE, electroblotted, and the presence of CypA was detected with bind to DC and M␾ compared with the wild-type (WT) strain of anti-CypA. Six individual clones with varying degrees of CypA silencing G217B and to the G217B ura5 (UC7), the progenitor strain for the are shown. WT, UC7 control yeasts, or five strains of CypA- silenced yeasts were incubated with monolayers of DC (B)orM␾ (C) for 30 min at silencing experiments. All five silenced strains tested showed sig- 37°C. (Note that Hc strain UC7–31 was not tested because it had consid- nificantly reduced binding (65–85%) to DC (Fig. 6B), but bound erably more residual CypA than the other silenced strains.) Bound yeasts ␾ normally to M (Fig. 6C). These data suggest that at least another were quantified by phase- contrast and fluorescent microscopy as described moiety on the surface of Hc yeasts is present that mediates binding in Materials and Methods. The data are presented as the mean Ϯ SEM (n ϭ .p Ͻ 0.001, t test ,ء .to DC. 5 for DC; n ϭ 3 for M␾) of the AI

F/TE from CypA-silenced Hc yeasts or from WT yeasts depleted of CypA inhibit the binding of yeasts to DC To confirm the idea that additional ligands play a role in the at- F/TE was confirmed by Western blot (Fig. 8A). Remarkably, the tachment of Hc yeasts to DC, two additional experiments were CypA-depleted F/TE inhibited the binding of Hc yeasts to DC as performed. In the first experiment, F/TE was prepared from three well as the control F/TE or the mock-depleted F/TE (Fig. 8B). of the CypA-silenced strains and then tested for their capacity to inhibit the binding of WT yeasts to DC. As shown in Fig. 7, F/TE rCypA does not inhibit the binding of CypA-silenced Hc yeasts from WT and UC7 yeasts inhibited binding of yeasts to DC by to DC ϳ80%. The F/TE from the three CypA-silenced strains inhibited Finally, we sought to determine whether rCypA would further in- the binding of yeasts to DC by 50%. hibit the binding of CypA-silenced yeasts to DC. DC were prein- In the second experiment, DC monolayers were preincubated cubated with HBSA or rCypA for 30 min and then incubated for with either F/TE, CypA-depleted F/TE, or mock- depleted FTE an additional 30 min with either WT yeasts or one of the CypA- and then incubated with Hc yeasts. Depletion of CypA from the silenced strains. In two independent experiments, rCypA inhibited The Journal of Immunology 7111

Table I. rCypA does not inhibit the binding of CypA-silenced Hc yeasts to DC

Expt. 1a Expt. 2a

AI AI WT Hc 1084 818 WT Hc ϩ rCypA 460 243 CypA-silenced Hc 396 218 CypA-silenced Hc ϩ rCypA 330 217

a DC were adhered in Terasaki culture dishes and preincubated with rCypA or buffer for 30 min at 37°C. FITC-labeled WT or CypA-silenced Hc yeasts then were added for an additional 30 min. After washing the monolayers to remove unbound yeasts, the DC were ﬁxed in 1% paraformaldehyde. The data from each experiment is the mean AI from two wells.

with FN (another ligand for VLA-5) and quantified their attach- FIGURE 7. Inhibition of binding of Hc yeasts to DC by F/TE from ment to DC, M␾, and CHO VLA-5 cells. Based on the amount of CypA-silenced yeasts. Monolayers of DC were preincubated for 30 min at protein bound to the beads, the beads contained an average 1.0 ϫ ␮ 37°C with 5 l of F/TE from WT and UC7 yeasts and from CypA-silenced 107 molecules of CypA/bead and 4.1 ϫ 105 molecules of FN/bead. yeasts UC7-1, UC7-5, and UC7-9. WT Hc yeasts then were added for 30 DC and M␾ were incubated in HBSA or with rCypA for 30 min min. Bound yeasts were quantified by phase-contrast and fluorescent mi- and then incubated with rCypA- or FN-coated beads for an addi- croscopy as described in Materials and Methods. The data are presented as p ϭ 0.002, t test. tional 30 min. Both CypA- and FN-coated beads bound avidly to ,ءء p Ͻ 0.001 and ,ء .the mean Ϯ SEM (n ϭ 3) of the AI DC and M␾, and binding was significantly greater than the attachment of BSA-coated beads that served as a control for background the binding of WT yeasts to the same level of binding as the CypA- binding (Fig. 9). Furthermore, rCypA inhibited the binding of silenced yeasts, but rCypA did not further reduce the adherence of rCypA-coated beads, but not FN-coated coated beads, to DC ␾ CypA-silenced Hc yeasts (Table I). and M .

Binding of rCypA-coated and FN-coated latex beads to DC, M␾, and CHO VLA-5 cells M␾ and DC possess both CD18 and VLA-5, and yet Hc yeasts bind to CD18 on M␾ (2) and to VLA-5 on DC (9). It also is known that CD18 receptor mobility is necessary for the attachment of Hc yeasts to M␾ (3). Therefore, it is possible that differences in receptor mobility on DC and M␾ account for the speciﬁc receptor usage by Hc yeasts. A second possible explanation for differential receptor usage is that speciﬁc receptor engagement is a function of ligand density on the surface of the yeasts. To explore the latter possibility, we prepared latex beads coated either with rCypA or

FIGURE 8. F/TE-depleted of CypA still inhibits the binding of Hc yeasts to DC. A, F/TE from WT Hc yeasts was adsorbed on rabbit IgG- Sepharose (lane 1) or rabbit anti-CypA-Sepharose (lane 2). Ten micrograms of protein was run on a 12% SDS-PAGE gel, electroblotted, and the presence of CypA was revealed by incubation with anti-CypA Ab followed by alkaline phosphatase-conjugated goat anti-rabbit IgG and chromogenic FIGURE 9. CypA and FN bind to different sites on DC (A) and M␾ (B). NBT-BCIP. B, Monolayers of DC were preincubated for 30 min at 37°C DC were preincubated with HBSA as a control or 13 ␮g of rCypA for 30 with F/TE, F/TE absorbed with anti-CypA, and F/TE absorbed with rabbit min at 37°C and then incubated with BSA-, FN-, or CypA-coated latex IgG (Mock Abs). Hc yeasts then were added for an additional 30 min. beads for an additional 30 min. Bound yeasts were quantified by phase- Bound yeasts were quantified by phase-contrast and fluorescent micros- contrast and fluorescent microscopy as described in Materials and Meth- copy as described in Materials and Methods. The data are presented as the ods. The data are presented as the mean Ϯ SEM (n ϭ 8 for DC; n ϭ 7 for .p Ͻ 0.05, t test ,ء .p Ͻ 0.001, t test. Co, Control. M␾) of the AI ,ء .mean Ϯ SEM (n ϭ 3) of the AI 7112 Hc CypA ADHESION TO DC VLA-5

isomerases (PPIases). Its purported function is to catalyze peptide chain conformation changes around proline residues. CypA also is the molecular target of the immunosuppressive antibiotic cyclo- sporin A, which inhibits its enzymatic activity (14). A second group of PPIases known as FK506-binding proteins, and generally termed immunophilins, are the receptors for the immunosuppressive drugs FK506 and rapamycin (15). Despite their ubiquitous presence in evolution, from bacteria to mammals, PPIases are dis- pensable for survival. For example, CypA-deficient mice are viable and immunocompetent, but are resistant to the immunosuppressive effects of cyclosporine (16). It is noteworthy that rCypA inhibited the binding of Hc yeasts to DC but not to M␾, but that CypA-coated polystyrene beads bound to both DC and M␾, in sync with the fact that both of these innate immune cells contain equivalent amounts of VLA-5 on their surface (L. Gildea, unpublished observations). Furthermore, rCypA inhibited the binding of CypA-coated beads, but not FN-coated FIGURE 10. CypA and FN bind to different sites on VLA-5. CHO cells beads, to DC, M␾, and CHO-VLA-5 cells, demonstrating that transfected with VLA-5 were preincubated with HBSA as a control or 13 CypA and FN bind to different sites on the VLA-5 molecule. Pro- ␮ g of rCypA for 30 min at 37°C and then incubated with BSA-, FN-, or vocatively, Hc yeasts did not bind to the CHO-VLA-5 cells, even CypA-coated latex beads for an additional 30 min. Bound yeasts were after coculture for up to 2 h. This result may be explained by the quantified by phase-contrast and fluorescent microscopy as described in Materials and Methods. The data are presented as the mean Ϯ SEM (n ϭ following two facts. First, the amount of CypA on the surface of p Ͻ 0.01, t test. Hc yeasts is considerably less than the amount of CypA that was ,ء .of the AI (5 bound to the polystyrene beads. Second, FACS analysis demonstrated that the CHO-VLA-5 cells had a half log less VLA-5 on To further demonstrate that binding of the beads was to VLA-5, their surface than DC (supplemental Fig. 3). an identical experiment was performed with CHO VLA-5 cells. Since DC and M␾ have equivalent amounts of VLA-5 on their Again, both CypA- and FN-coated beads bound avidly to the CHO surface, the basis for the differential receptor usage by DC and M␾ VLA-5 cells, whereas there was minimal binding of BSA-coated to bind Hc yeasts is unclear. One hypothesis supported by the data beads (Fig. 10). As was found with DC and M␾, rCypA inhibited is that the differential receptor usage is due to reciprocal receptor the binding of the rCypA-coated beads but not FN-coated beads. mobility. Only minimal binding of the beads was observed with nontrans- Another possible explanation for the differential receptor usage fected CHO cells (data not shown). Most interesting was the fact is the phenomenon “inside-out” regulation of integrin avidity for that Hc yeasts did not bind to the CHO VLA-5 cells, even after ligands. Intracellular signals in the phagocyte cytoplasm may coculture for up to 2 h (data not shown). change the phosphorylation status in the cytoplasmic tails of CD11/CD18, VLA-5 and other integrins, effectively switching Discussion them from a high-affinity state to a low-affinity state (or vice versa) Attachment to host cell membranes is a critical step in the patho- for cognate ligands (17, 18). Thus, it is possible that M␾ and DC genesis of intracellular pathogens such as Hc. After inhalation of have the ability to choose which integrin to utilize in binding and airborne Hc conidia into the alveolar spaces in the lung, the conidia ingesting a specific parasite, based on other as yet unidentified convert into the pathogenic yeast phase and are taken up by DC signals. and M␾, presumably in the absence of serum opsonins (1). The The demonstration that CypA plays a role in the pathogenicity identity of the host cell receptors and corresponding fungal ligands of Hc is not unique. Legionella pneumophila expresses a surface involved in this interaction is pivotal to understanding the patho- virulence factor, M␾ infectivity potentiator (Mip), that is an 18- genesis of Hc and the generation of the adaptive immune response. kDa immunophilin with PPIase activity that has a high affinity for Original studies demonstrated that Hc yeasts are recognized by the the immunosuppressive drug FK506 (19). Legionella also possess ␤ ␾ CD11/CD18 2 integrin receptors on human M (2) and that these a second cytoplasmic cyclophilin, cyclophilin 18. Mip protein is integrins recognize HSP60 on the surface of Hc yeasts (4). At- not necessary for the extracellular survival of Legionella, but dis- tachment of Hc yeasts to the M␾ results in rapid ingestion and ruption mutants are 10 times less invasive of Acanthamoeba cas- localization in a phagosome (3). Phagosome maturation is pre- tellani (20). Interestingly, the Hc CypA sequence is not homolo- vented by the fungus. permitting the yeasts to survive and prolif- gous to L. pneumophila Mip, but shows a high degree of erate (13). conservation with the cytoplasmic form cyclophilin 18. In exten- In contrast to M␾, Hc engulfed by human DC are killed and sive searches through the Hc genome, we did not find a Mip ho- degraded, and then Hc-derived Ags are presented to T cells induc- molog or other sequence homologs to Hc CypA. ing them to proliferate (9). Thus, phagocytosis of Hc yeasts by DC Trypanosoma cruzi secretes a Mip-like protein involved in cell is a crucial step in the development of T cell-mediated immunity. invasion and virulence and its activity as an invasion factor is Although human DC contain CD18 integrins on their surface, Hc inhibited by FK506 (21). Several species of Plasmodia secrete low ␣ ␤ binds to the 5 1 integrin, VLA-5, a FN receptor (9). Because molecular mass FK506-binding PPIases that are necessary for in- rHSP60 does not inhibit the attachment of Hc yeasts to human DC, tracellular invasion, and their inhibition by cyclosporine and at least one other molecule must be involved in DC-Hc interaction. FK506 explain the antiparasitic properties of these drugs (22, 23). The experiments described in this report provide several lines of Finally, Toxoplasma gondii secretes an 18-kDa protein, cyclophi- evidence that one VLA-5 cognate ligand is Hc CypA. lin 18, that activates human DC through the chemokine receptor CypA is a highly conserved, low molecular mass chaperone, CCR5 and induces the secretion of IL-12 (24). Thus, PPIases from which belongs to the general group of peptidyl-prolyl cis-trans several microorganisms play a role in their pathogenesis. The Journal of Immunology 7113

␤ VLA-5 is a heterodimeric 1 integrin receptor composed of the References ␣ ␤ 5 chain (CD49e) and the 1 chain (CD29) (25) and is one of 1. Newman, S. L. 2006. Intracellular pathogenesis of Histoplasma capsulatum. In several integrin receptors for FN (26). It is expressed on a wide Molecular Principles of Fungal Pathogenesis. J. Heitman, S. G. Filler, ␾ J. E. Edwards, and A. P. Mitchell, eds. ASM Press, Washington, DC, pp. variety of cells including T cells, monocytes, M and DC (27). 527–536. VLA-5 and VLA-4 both play significant roles in the transendothe- 2. Bullock, W. E., and S. D. Wright. 1987. Role of the adherence-promoting re- lial migration of T cells, M␾, and DC (28–30). ceptors, CR3, LFA-1, and p150,95, in binding of Histoplasma capsulatum by ␤ human macrophages. J. Exp. Med. 165: 195–210. With regard to pathogen recognition, 1 integrins are utilized by 3. Newman, S. L., C. Bucher, J. Rhodes, and W. E. Bullock. 1990. Phagocytosis of T. cruzi to gain entrance into human M␾ (31), and the Yersinia Histoplasma capsulatum yeasts and microconidia by human cultured macrophages and alveolar macrophages: cellular cytoskeleton requirement for attach- membrane protein invasin mediates adhesion to human T cells via ment and ingestion. J. Clin. Invest. 85: 223–230. VLA-4 (32). In an apparent two step process, binding and cross- 4. Long, K. H., F. J. Gomez, R. E. Morris, and S. L. Newman. 2003. Identification linking of monocyte VLA-5 by Bordetella pertussis enhances the of heat shock protein 60 as the ligand on Histoplasma capsulatum that mediates binding to CD18 receptors on human macrophages. J. Immunol. 170: 487–494. binding activity of CD11b/CD18 (CR3), which, in turn, facilitates 5. Newman, S. L. 1999. Macrophages in host defense against Histoplasma capsu- phagocytosis of the bacteria (33). Other intracellular pathogens, latum. Trends Microbiol. 7: 67–71. 6. Newman, S. L. 2001. Cell-mediated immunity to Histoplasma capsulatum. Se- including L. pneumophila (34), Mycobacterium tuberculosis (35), min. Respir. Infect. 16: 102–108. and Leishmania major (36) utilize CD18 receptors to gain entrance 7. Deepe, G. S., and W. E. Bullock. 1992. Histoplasmosis: a granulomatous inflam- into M␾. Perhaps uniquely, human M␾ and DC use different re- matory response. In Inflammation: Basic Principles and Clinical Correlates. J. I. Gallin, I. M. Goldstein, and R. Snyderman, eds. Raven Press, New York, pp. ceptors, CD18 and VLA-5, respectively, to mediate attachment of 943–958. the same microorganism, Hc. Conversely, different Hc ligands 8. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of immunity. Nature 392: 245–252. are recognized by these receptors. This situation occurs despite the 9. Gildea, L. A., R. E. Morris, and S. L. Newman. 2001. Histoplasma capsulatum fact that M␾ and DC have both CD18 and VLA-5 receptors. We yeasts are phagocytosed via very late antigen-5, killed, and processed for antigen hypothesize that the specific receptor-ligand interaction of Hc presentation by human dendritic cells. J. Immunol. 166: 1049–1056. ␾ 10. Gildea, L. A., G. M. Ciraolo, R. E. Morris, and S. L. Newman. 2005. Human yeasts with M vs DC leads to the induction of unique signaling dendritic cell activity against Histoplasma capsulatum is mediated via phagoly- pathways that allow the yeasts to survive in M␾, but leads to their sosomal fusion. Infect. Immun. 73: 6803–6811. 11. Banchereau, J., F. Briere, C. Caux, J. Davoust, S. Lebecque, Y. J. Liu, demise in DC. B. Pulendran, and K. Palucka. 2000. Immunobiology of dendritic cells. Annu. This mechanism of Hc attachment and invasion of mammalian Rev. Immunol. 18: 767–811. host cells is shared by a wide array of prokaryotic and eukaryotic 12. Worsham, P. L., and W. E. Goldman. 1988. Quantitative plating of Histoplasma capsulatum without addition of conditioned medium or siderophores. J. Med. Vet. intracellular parasites, suggesting either a conserved ancestral Mycol. 26: 137–143. mechanism of cell-cell interaction or a phenomenon of convergent 13. Newman, S. L., L. Gootee, C. Bucher, and W. E. Bullock. 1991. Inhibition of evolution. Hc choice of ligands, HSP60 and CypA, bear remark- intracellular growth of Histoplasma capsulatum yeast cells by cytokine-activated human monocytes and macrophages. Infect. Immun. 59: 737–741. able parallels: Both molecules are evolutionary conserved molec- 14. Arevalo-Rodriguez, M., X. Wu, S. D. Hanes, and J. Heitman. 2004. Prolyl ular chaperones. When found in the cytoplasm, they are involved isomerases in yeast. Front. Biosci. 9: 2420–2446. 15. Soldin, S. J., J. N. Murthy, J. G. Donnelly, Y. Chen, and N. Goodyear. 1993. in the conformation and renaturation of nascent peptides. When Immunophilin receptors for immunosuppressive drugs. Ther. Drug Monit. 15: found in the extracellular milieu, both molecules have stimulatory 468–471. activity for the innate immune system and, in the case of HP60, an 16. Colgan, J., M. Asmal, B. Yu, and J. Luban. 2005. Cyclophilin A-deficient mice are resistant to immunosuppression by cyclosporine. J. Immunol. 174: adaptive immunity inducer and a protective vaccine (37). This 6030–6038. “moonlighting” multiplicity of functions attributable to a small 17. Takagi, J., H. P. Erickson, and T. A. Springer. 2001. C-terminal opening mimics ␤ “inside-out” activation of integrin alpha5 1. Nat. Struct. Biol. 8: 412–416. number of proteins can, in part, be explained by these proteins’ 18. Patcha, V., J. Wigren, M. E. Winberg, B. Rasmusson, J. Li, and E. Sarndahl. ␤ promiscuous interaction with a wide variety of partners. In con- 2004. Differential inside-out activation of 2-integrins by leukotriene B4 and trast, in their capacity as microbial ligands, each of these proteins fMLP in human neutrophils. Exp. Cell Res. 300: 308–319. 19. Fischer, G., H. Bang, B. Ludwig, K. Mann, and J. Hacker. 1992. Mip protein of apparently play quite specific roles in Hc pathogenesis: HSP60 Legionella pneumophila exhibits peptidyl-prolyl-cis/trans isomerase (PPlase) ac- involved in intracellular survival, proliferation, and dissemination, tivity. Mol. Microbiol. 6: 1375–1383. 20. Schmidt, B., T. Tradler, J. U. Rahfeld, B. Ludwig, B. Jain, K. Mann, while CypA is involved in initiation of adaptive immune K. P. Rucknagel, B. Janowski, A. Schierhorn, G. Kullertz, et al. 1996. A cyclo- responses. philin-like peptidyl-prolyl cis/trans isomerase from Legionella pneumophila: Finally, the data clearly demonstrate that CypA is not the only characterization, molecular cloning, and overexpression. Mol. Microbiol. 21: 1147–1160. ligand on Hc yeasts that promotes binding to DC. Since rCypA 21. Pereira, P. J., M. C. Vega, E. Gonzalez-Rey, R. Fernandez-Carazo, almost completely blocks binding of yeasts to DC, the other li- S. Macedo-Ribeiro, F. X. Gomis-Ruth, A. Gonzalez, and M. Coll. 2002. Trypano- soma cruzi macrophage infectivity potentiator has a rotamase core and a highly gand(s) probably is in close proximity to CypA on the yeast’s exposed ␣-helix. EMBO Rep. 3: 88–94. surface, and, therefore, is unable to promote binding to DC under 22. Kumar, R., B. Adams, A. Musiyenko, O. Shulyayeva, and S. Barik. 2005. The these circumstances. However, if CypA is removed from the sur- FK506-binding protein of the malaria parasite, Plasmodium falciparum,isa FK506-sensitive chaperone with FK506-independent calcineurin-inhibitory ac- face of the yeasts, the yeasts still bind to DC. Furthermore, if CypA tivity. Mol. Biochem. Parasitol. 141: 163–173. is removed from F/TE, the depleted F/TE still inhibits the binding 23. Monaghan, P., and A. Bell. 2005. A Plasmodium falciparum FK506-binding of yeasts to DC. This interpretation of the data is supported by the protein with peptidyl-prolyl cis-trans isomerase and chaperone activities. Mol. Biochem. Parasitol. 139: 185–195. fact that the 6.5 ␮g of rCypA that optimally inhibits Hc yeast 24. Aliberti, J., J. G. Valenzuela, V. B. Carruthers, S. Hieny, J. Andersen, H. Charest, attachment to DC is three to five times more than the amount of C. Reis e Sousa, A. Fairlamb, J. M. Ribeiro, and A. Sher. 2003. Molecular ␮ mimicry of a CCR5 binding-domain in the microbial activation of dendritic cells. CypA that is in 5 l of F/TE that also optimally inhibits Hc bind- Nat. Immunol. 4: 485–490. ing to DC (supplemental Fig. 4). Finally, in reviewing our initial 25. Hemler, M. E., C. Huang, and L. Schwarz. 1987. The VLA protein family: char- Far Western blots, a second very feint band can be seen that binds acterization of five distinct cell surface heterodimers each with a common 130,000 molecular weight ␤ subunit. J. Biol. Chem. 262: 3300–3309. VLA-5. Current efforts are directed to identifying this second li- 26. Pytela, R., M. D. Pierschbacher, and E. Ruoslahti. 1985. Identification and iso- gand for DC VLA-5. lation of a 140 kd cell surface glycoprotein with properties expected of a fibronectin receptor. Cell 40: 191–198. ␤ 27. Hemler, M. E., and R. R. Lobb. 1995. The leukocyte 1 integrins. Curr. Opin. Hematol. 2: 61–67. 28. Yabkowitz, R., V. M. Dixit, N. Guo, D. D. Roberts, and Y. Shimizu. 1993. Disclosures ␣ ␤ Activated T-cell adhesion to thrombospondin is mediated by the 4 1 (VLA-4) ␣ ␤ The authors have no financial conflict of interest. and 5 1 (VLA-5) integrins. J. Immunol. 151: 149–158. 7114 Hc CypA ADHESION TO DC VLA-5

29. Garcia-Vicuna, R., A. Humbria, A. A. Postigo, C. Lopez-Elzaurdia, action between Bordetella pertussis and human monocytes. J. Immunol. 151: M. O. de Landazuri, F. Sanchez-Madrid, and A. Laffon. 1992. VLA family in 6274–6282. rheumatoid arthritis: evidence for in vivo regulated adhesion of synovial ﬂuid 34. Payne, N. R., and M. A. Horwitz. 1987. Phagocytosis of Legionella pneumophila T cells to ﬁbronectin through VLA-5 integrin. Clin. Exp. Immunol. 88: is mediated by human monocyte complement receptors. J. Exp. Med. 166: 435–441. 1377–1389. 30. D’Amico, G., G. Bianchi, S. Bernasconi, L. Bersani, L. Piemonti, S. Sozzani, 35. Melo, M. D., I. R. Catchpole, G. Haggar, and R. W. Stokes. 2000. Utilization of A. Mantovani, and P. Allavena. 1998. Adhesion, transendothelial migration, and CD11b knockout mice to characterize the role of complement receptor 3 (CR3, reverse transmigration of in vitro cultured dendritic cells. Blood 92: 207–214. CD11b/CD18) in the growth of Mycobacterium tuberculosis in macrophages. 31. Fernandez, M. A., M. A. Munoz-Fernandez, and M. Fresno. 1993. Involvement Cell. Immunol. 205: 13–23. ␤ of 1 integrins in the binding and entry of Trypanosoma cruzi into human mac- 36. Rosenthal, L. A., F. S. Sutterwala, M. E. Kehrli, and D. M. Mosser. 1996. Leish- rophages. Eur. J. Immunol. 23: 552–557. mania major-human macrophage interactions: cooperation between Mac- 1 32. Ennis, E., R. R. Isberg, and Y. Shimizu. 1993. Very late antigen 4-dependent (CD11b/CD18) and complement receptor type 1 (CD35) in promastigote adhe- ␤ adhesion and costimulation of resting human T cells by the bacterial 1 integrin sion. Infect. Immun. 64: 2206–2215. ligand invasin. J. Exp. Med. 177: 207–212. 37. Gomez, F. J., R. Allendoerfer, and G. S. Deepe, Jr. 1995. Vaccination with 33. Hazenbos, W. L., B. M. van den Berg, and R. van Furth. 1993. Very late recombinant heat shock protein 60 from Histoplasma capsulatum protects mice antigen-5 and complement receptor type 3 cooperatively mediate the inter- against pulmonary histoplasmosis. Infect. Immun. 63: 2587–2595.