Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Inhibition of Type 1 Cytokine–mediated Inflammation by a Soluble CD30 Homologue Encoded by Ectromelia (Mousepox) Virus

Inhibition of Type 1 Cytokine–mediated Inflammation by a Soluble CD30 Homologue Encoded by... CD30 is up-regulated in several human diseases and viral infections but its role in immune reg- ulation is poorly understood. Here, we report the expression of a functional soluble CD30 ho- mologue, viral CD30 (vCD30), encoded by ectromelia (mousepox) virus, a poxvirus that causes a severe disease related to human smallpox. We show that vCD30 is a 12-kD secreted protein that not only binds CD30L with high affinity and prevents its interaction with CD30, but it also induces reverse signaling in cells expressing CD30L. vCD30 blocked the generation of interferon –producing cells in vitro and was a potent inhibitor of T helper cell (Th)1- but not Th2-mediated inflammation in vivo. The finding of a CD30 homologue encoded by ec- tromelia virus suggests a role for CD30 in antiviral defense. Characterization of the immuno- logical properties of vCD30 has uncovered a role of CD30–CD30L interactions in the genera- tion of inflammatory responses. Key words: poxviruses • immunomodulation • cytokine • TNFR superfamily • Th-1 Introduction CD30 and CD30L (CD153) are members of the TNFR cells, activated T cells and macrophages, and in neoplasic and TNF superfamilies, respectively. CD30 was identified cells such as Burkitt-type lymphoma cells or B cells associ- by the mAb Ki-1 against Hodgkin and Reed-Sternberg cells ated with lymphoproliferative disorders. The extracellular (1), the malignant component of Hodgkin’s disease, and has domain of CD30L shows homology to TNF, lymphotoxin, been extensively used as a clinical disease marker. CD30 was and CD40L (3). It is unclear whether CD30L also exists as a subsequently found in resting CD8 T cells, activated or vi- soluble form, as does TNF. Interaction of CD30L with cells rally transformed T and B cells, and at the surface of HIV- expressing CD30 induces signals mediated by nuclear factor infected lymphocytes. CD30 is a type I membrane protein B and TNFR-associated factor 2 that cause cell prolifera- that can be cleaved by metalloproteases producing a soluble tion or cell death. Interestingly, upon binding to CD30, form (sCD30;* reference 2). The extracellular domain of CD30L is also able to signal. One of the consequences of CD30 shows cysteine-rich domains (CRDs) characteristic this reverse signaling is cell proliferation (4). of the TNFR superfamily. CD30L is expressed as a type II The role of CD30–CD30L interaction in health and dis- membrane glycoprotein (2) in resting neutrophils and B ease is still not totally understood, in part due to the pleio- tropic nature of CD30 signals. Mice lacking a functional CD30 gene show defective negative thymocyte selection Address correspondence to Antonio Alcami, Department of Medicine, (5), whereas transgenic mice expressing CD30 in the thy- University of Cambridge, Addenbrooke’s Hospital, Level 5, Box 157, mus have enhanced thymic negative selection (6). A recent Hills Road, Cambridge CB2 2QQ, United Kingdom. Phone: 44-1223- study of genes targeted by CD30 suggests that Fas, TRAIL, 763403; Fax: 44-1223-330158; E-mail: aa258@mole.bio.cam.ac.uk CCR7, TNFR-associated factor 1, and cIAP2 are up-regu- P. Smith’s and P.G. Fallon’s present address is Department of Bio- lated whereas FasL, perforin, granzyme B, and c-myc seem chemistry, Trinity College, Dublin 2, Ireland. *Abbreviations used in this paper: AraC, cytosine arabinoside; CPV, cow- to be down-regulated (7). Finally, the reasons for increased pox virus; CRD, cysteine-rich domain; Crm, cytokine response modifier; levels of sCD30 in malignant lymphomas, viral infection EV, ectromelia virus; ORF, open reading frame; PPD, protein-purified (HIV, human T cell leukemia virus, and EBV), and several derivative; sCD30, soluble CD30; SEA, soluble egg Ags; VaV, variola vi- immunological disorders such as systemic lupus erythema- rus; vCD30, viral CD30; vTNFR, viral TNFR; VV, vaccinia virus; WR, Western Reserve. tosus or rheumatoid arthritis are not known (8). 829 J. Exp. Med.  The Rockefeller University Press • 0022-1007/2002/09/829/11 $5.00 Volume 196, Number 6, September 16, 2002 829–839 http://www.jem.org/cgi/doi/10.1084/jem.20020319 Poxviruses are a family of complex DNA viruses that en- nate mouse CD30L were purchased from Amersham Biosciences and Pierce Chemical Co., respectively. The protein A–coated code up to 200 genes and infect a wide variety of hosts (9). FlashPlates used for binding and affinity studies were purchased Smallpox was a devastating disease caused by variola virus from PerkinElmer. The mouse mAb specific for the histidine tag (VaV), one of the most virulent human pathogens. Vac- and the FITC-conjugated goat anti–mouse Igs used in flow cy- cinia virus (VV) is the best characterized poxvirus and the tometry were from CLONTECH Laboratories, Inc. and Dako, vaccine used to achieve the global eradication of smallpox respectively. Mycobacterium tuberculosis whole cell lysates (H37Rv by 1977, but its origin and natural host are unknown. strain) and protein-purified derivative (PPD) were obtained from Cowpox virus (CPV) is probably a rodent virus that spo- Mycos Research. Schistosome eggs and soluble egg Ags (SEA) radically infects other animal species. Ectromelia virus (EV) were prepared as previously described (27). is a highly virulent natural pathogen of mice that causes Cells, Viruses, and Viral DNA Preparations mousepox and has been isolated from outbreaks in labora- tory mouse colonies (10). Like VaV, EV has a restricted The growth of BSC-I, TK 143B, and K562 cells, and the sources of VV Western Reserve (WR) strain and EV isolates host range, causes severe disease with high mortality rate Hampstead and Naval have been described (15). VV and EV and skin lesions in the later stages of infection. These simi- were propagated in BSC-I cells and viral genomic DNA was pre- larities with smallpox make EV an interesting experimental pared as previously described (28). The growth of Autographa cali- model for virus–host interactions. fornica nuclear polyhedrosis virus in Spodoptera frugiperda 21 insect Poxviruses encode a unique collection of genes that cells has been described (29). Tn5 B1-4 (Hi5) insect cells were evade host immune responses. These molecules are often cultured in EX-CELL serum-free medium as suggested by the secreted and include cytokine homologues and soluble cy- supplier (European Collection of Cell Cultures). tokine receptors or binding proteins (11–13). Some of these viral genes seem to have been acquired from the host DNA Sequencing and modified during virus evolution to confer an advantage Specific oligonucleotides, CD30-1 (5 GTTCTGGATACAT- for virus replication, survival, or transmission. EV encodes GCACAAAG 3 ) and CD30-2 (5 GGAGGATAATCATTTG- receptors or binding proteins for TNF (14), IL-1 (15), CAAACG 3 ), were designed based on the sequence of CPV strain GRI90 open reading frame (ORF) D13L (30), and used to IFN- (16, 17), IFN-/ (17, 18), IL-18 (19, 20), and amplify the cognate genes from viral DNA preparations from VV chemokines (15, 21). EV also encodes antiapoptotic pro- WR and EV Hampstead and Naval by PCR using Taq DNA teins (22, 23) and an intracellular protein that confers IFN polymerase. PCR products were sequenced by the DNA Se- resistance (17). quencing Service of the Department of Biochemistry, Cambridge The TNF binding activity encoded by orthopoxviruses University, Cambridge, United Kingdom. The sequence data is particularly interesting because there are four distinct were analyzed using Genetics Computer Group computer pro- viral TNFRs (vTNFRs): cytokine response modifier grams (31). (Crm) B (24), CrmC (25), CrmD (14), and CrmE (26). These molecules show different ligand specificity and are Construction of Recombinant Baculovirus Expressing the EV Hampstead vCD30 Gene expressed at different times after infection, but their rela- tive contribution to viral pathogenesis is not well under- The EV Hampstead vCD30 gene was amplified by PCR using Pfu DNA polymerase, virus DNA as template, and oligonucle- stood. Maybe the vTNFRs bind other members of the otides CD30-3 (5 CGCAAGCTTGGATCCATGAAGATGA- growing TNF family and protect the virus from the ac- ATACTATC TTTTTATC 3 ) and CD30-4 (5 CGCGCG- tion of other ligands. GCCGCTGATGAGTATTTATGATAACAAAG 3 ), which We report the identification and characterization of a correspond to the 5 and 3 ends of the ORF and provide Hind- novel member of the TNFR superfamily, a homologue of III/BamHI and NotI sites, respectively. The resultant product CD30 encoded by EV that blocks the binding of CD30L was cloned into HindIII- and NotI-digested pBac1 (Novagen), to its receptor and induces reverse signaling in cells express- creating plasmid pMS2 (EV Hampstead vCD30). The DNA se- ing CD30L. Moreover, the viral CD30 (vCD30) abrogates quence of the insert was confirmed to not contain mutations. T cell proliferation in vitro and in vivo it blocks type 1 but The Fc fragment of the human IgG1 was cut from pIGplus not type 2 cytokine–mediated T cell responses. These stud- (R&D Systems) and subcloned into NotI/SphI sites of pMS2, ies not only describe a novel immunomodulatory strategy which created plasmid pMS18 (EV Hampstead vCD30-Fc). Re- combinant baculovirus was produced as previously described (29) of poxviruses, but also pave the way to the role of CD30– and termed AcCD30-Fc (EV Hampstead vCD30-Fc, AcMS18). CD30L in viral infections and in type I cytokine–mediated Control recombinant baculovirus expressing EV Hampstead– inflammatory diseases. truncated CrmD (AcCrmD-CRD1,2-Fc) was constructed as AcCD30-Fc (unpublished data). Purification of the Baculovirus Recombinant vCD30-Fc Protein Materials and Methods Hi5 cultures were infected with recombinant baculoviruses at Reagents 10 pfu/cell and supernatants were harvested 3–4 d later when full Recombinant mouse CD30L (ED  50  150 g/ml), hu- infection was observed. The recombinant Fc fusion proteins were man CD30 (ED  0.03  0.1 g/ml), and mouse CD30 (ED  subsequently purified using Protein A HiTrap columns (Amer- 50 50 0.03  0.1 g/ml) were purchased from R&D Systems. The io- sham Biosciences). The purified protein was then analyzed by dine-125 (103.7 mCi/ml) and iodogen reagent used to radioiodi- SDS-PAGE in 12% acrylamide gels and stained with Coomassie 830 Ectromelia Virus–encoded CD30 Homologue blue. Protein concentration was determined using the Bio-Rad Kinetics of vCD30 Production During EV Infection protein assay reagent. BSC-I cells were mock infected or infected with 10 pfu EV Hampstead per cell in the absence or presence of 40 g/ml cytosine Construction of Recombinant VV Expressing the EV Hampstead arabinoside (AraC), an inhibitor of DNA replication, and harvested vCD30 Gene at different times after infection. Supernatants were inactivated and the CD30 binding activity was tested as described above. Total The EV Hampstead vCD30 gene was amplified by PCR with RNA was extracted using the guanidine thiocyanate–based DNA/ virus DNA as template, Pfu DNA polymerase, and oligonucle- RNA Isolation Kit (Promega) according to the manufacturer’s in- otides CD30-3 and CD30-5 (5 CGCGGTACCTCATGAT- structions. Total RNA (from 7 10 cells) was then analyzed by GAGTATTTATGATAACAAAG 3 ) containing KpnI restric- RT and then by PCR. RT was performed in the presence of tion site. The DNA fragment was cloned into BamHI- and oligo(dT) (Promega), RNAsin (Amersham Biosciences), and KpnI-digested pMJ601 (provided by B. Moss, National Institutes 15 avian myeloblastosis virus RT (Boehringer). The cDNA (2.5 l of of Health, Bethesda, MD; reference 32), creating plasmid pMS12 40 l; provided by K. Shair, University of Cambridge, Cambridge, (EV Hampstead vCD30). The DNA sequence of the insert was United Kingdom) was amplified by PCR using Taq polymerase and confirmed to not contain mutations. The recombinant VV was oligonucleotides specific for vCD30, CD30-3, and CD30-4. DNA produced as previously described (29) and termed VVCD30 (EV from BSC-I cells was included as a negative control. Hampstead vCD30, vMS12). Biological Activity of the vCD30 Metabolic Labeling of VVCD30 and Electrophoretic Analysis In Vitro Studies. 1 g/ml recombinant soluble mouse CD30L BSC-I cells were infected with VV WR or VVCD30 at 10 35 preincubated with RPMI, a 25-fold excess of vCD30, or human pfu/cell. Cultures were pulse labeled with 150 Ci/ml [ S]me- IgG1 for 1.5 h at 4 C was added to 10 K562 cells and incubated thionine (1,200 Ci/mmol; Amersham Biosciences) and 150 Ci/ 35 for 2 h at 4 C. After this period, cells were incubated for 40 min ml [ S]cysteine (600 Ci/mmol; NEN Life Science Products) in at 4 C with a mouse mAb specific for the histidine tag (1 g/ml methionine- and cysteine-free medium in the absence of serum. in 0.1% BSA in PBS) that would recognize CD30L bound to the Cells or media were dissociated in sample buffer and analyzed by cell membrane. This Ab was subsequently developed with an SDS-PAGE in 12% acrylamide gels and visualized by fluorogra- FITC-labeled goat anti–mouse Ig Ab for 30 min at 4 C. Cells phy with Amplify (Amersham Biosciences). binding CD30L were then detected by FACS analysis. Un- stained cells and cells stained in the absence of CD30L were in- Preparation of VV and EV Supernatants cluded as a control. BSC-I cells were mock infected or infected with VV-WR, To test the ability of vCD30 to induce reverse signaling via VVCD30, EV Hampstead, and EV Naval at 10 pfu/cell in phe- membrane bound CD30L, 5 10 freshly isolated human neu- nol red– and serum-free medium. Supernatants were harvested at trophils were incubated in a volume of 100 l for 5 h at 37 C in 2 (for the VV infections) or 3 (for the EV and mock infections) d 96-well plates precoated with 10 g/ml of mouse, human after infection and prepared and inactivated as previously de- vCD30, human IgG1, CD30, or PBS as previously described (4). scribed (33). After this period, supernatants were harvested and the production of IL-8 was measured by ELISA (Diaclone). To address the possi- CD30L Binding Assay ble interference of vCD30 in the development of CTL responses, Recombinant mouse CD30L was radioiodinated to a specific 4 10 freshly isolated splenocytes from BALB/c mice were 6 4 activity of 10 cpm/ g using the Iodogen method (34). Approx- mixed in 96-well plates with 2 10 L929 cells in a final volume imately 150 pM of I-CD30L was incubated for 12 h with 5 ng of 200 l RPMI in the presence or absence of 10 g/ml vCD30- purified vCD30-Fc or recombinant mouse CD30 in a protein Fc or IgG1 supplemented with 10% FCS, sodium pyruvate, and A–coated FlashPlate. The binding medium was phenol red–free nonessential amino acids for 5 d at 37 C, 5% CO . After this pe- MEM, 0.1% BSA, 20mM Hepes, pH 7.5. The amount of riod, IL-2 (Roche) was added to a final concentration of 50 U/ml CD30L bound to the viral receptor was measured in a Packard and the incubation was held for an additional 2 d. Finally, cells Topcount microplate counter. Nonspecific binding was deter- were harvested and the viable cells counted by trypan blue exclu- 125 5 4 mined by incubating I-CD30L with binding medium only. sion. 10 activated splenocytes were mixed again with 10 L929 in For the competition studies, a 500-fold molar excess of cold the presence or absence of 10 g/ml vCD30-Fc or IgG1 and the mouse CD30L was added to the recombinant mouse or viral re- number of cells producing IFN- were measured using an ceptors before the addition of I-CD30L. To test the CD30 ELISPOT assay (R&D Systems). binding activity in supernatants of EV Hampstead and Naval (VV In Vivo Studies. Type 1 and 2 cytokine–dominated pulmo- WR or VVCD30), 50 l supernatant equivalent to 1.5 10 nary granulomas were induced by mycobacterial or SEA, respec- cells were preincubated with I-CD30L before being added to tively, as previously described (36). Female BALB/c mice were the recombinant mouse or viral receptors. For the determination obtained from Harlan. vCD30-Fc or controls CrmD-CRD1,2-Fc of the affinity constant of both mouse and vCD30 to the and IgG1 (Sigma-Aldrich) were injected intraperitoneally (10 g CD30L, binding assays with increasing amounts of I-CD30L per injection) throughout sensitization and elicitation of bead against a fixed amount of recombinant CD30 (2 and 0.5 ng granulomas, i.e., on days 0, 7, 14, and 16, or were injected only mouse or viral protein, respectively) were performed. The results during elicitation of granulomas on days 14 and 16. On day 0, were analyzed with the LIGAND software (35). For the deter- mice were sensitized by intraperitoneal injection of 20 g M. tu- mination of membrane-bound activity of vCD30, BSC-I cells berculosis whole cell lysates in complete Freund’s adjuvant (type 1 were mock infected or infected with VV WR, VVCD30, and cytokine sensitization) or 5,000 Schistosoma mansoni eggs (type 2 EV Hampstead at 10 pfu/cell. 24 h later, human I-CD30L was cytokine sensitization). 14 d later mice were injected intravenously added and bound I-CD30L was determined by phthalate oil with 5,000 Sepharose 4B beads covalently coupled with PPD or centrifugation (29). SEA. Mice were killed 4 d after bead injection on day 18. The left 831 Saraiva et al. lung lobe was snap frozen and used for cytokine analysis. The re- under accession numbers AJ507059 and AJ507060, respec- mainder of the lung was fixed for histological studies. The diame- tively. ters of the granuloma surrounding at least 50 individual beads per mouse were measured. Group mean granuloma volumes from four or five mice per group are presented. The statistical differ- Results ences between groups were determined using Student’s t test. Identification of a Novel Member of the TNFR Superfamily Lung tissue cytokines were determined as previously described Encoded by EV. Analysis of the CPV strain GRI90 se- using ELISA protocols (37). Control naive mouse lungs were pro- quence (30) revealed the presence of an ORF (D13L) with cessed to determine basal lung cytokine levels. Data were ex- sequence similarity to host CD30, a TNFR superfamily pressed as nanogram cytokine per milligram lung protein. The member, and distinct from the previously identified poxvi- spleen and draining mediastinal LN were removed on the day the rus TNFRs (CrmB, CrmC, CrmD, and CrmE). PCR and mice were killed and used for cell culture and intracellular cyto- sequence analysis of the cognate gene in EV isolates Hamp- kine staining (27). In brief, spleen cell suspensions were cultured for 6 h in media alone or in the presence of 6 g/ml Con A (Sigma- stead and Naval showed the existence of an intact vCD30 Aldrich) and 10 g/ml Brefeldin A (Sigma-Aldrich) added during gene (Fig. 1 a). The predicted viral molecule lacked N-gly- the last 4 h. All intracellular detection reagents were from Caltag. cosylation sites, was considerably smaller (12 kD) than the Cells were surface stained with tri-color–conjugated anti-CD4 or mouse or human counterparts (52 and 120 kD, respec- CD8 mAbs. After cell permeabilization, cells were incubated with tively), and aligned with the second and third CRDs found an FITC-conjugated anti–IFN-–mAb, PE-conjugated anti–IL-4 in the extracellular domain of the host CD30, with some mAb, or FITC- or PE-conjugated isotype control mAbs. For motifs highly conserved (Fig. 1 b). Interestingly, vCD30 FACS analysis CD4 or CD8 lymphocytes were gated and showed a higher similarity to mouse (63.7%) than to hu- quadrants were set using isotype control mAbs. The frequencies of man (56.7%) CD30. The existence of a signal peptide and IFN-– and IL-4–stained cells are expressed as percentages. the lack of transmembrane domain suggested that vCD30 Nucleotide Sequence Accession Number may act as a soluble decoy receptor for host CD30L. Characterization of the vCD30 Protein. A recombinant The sequence data for the EV Hampstead– and Naval-encoded CD30 homologue are available from GenBank/EMBL/DDBJ VV WR expressing the EV vCD30 under a strong pro- Figure 1. Sequence of vCD30. Pairwise align- ment of the predicted amino acid sequence of vCD30 in (a) three distinct orthopoxviruses and (b) of human, mouse, and vCD30. Dark shadows rep- resent (a) differences or (b) similarities. Gray boxes, regions of high similarity; , deletions; *, stop codons. The predicted signal peptide (SP), CRDs, and transmembrane domain (TM) are indicated. These sequence data are available from GenBank/ EMBL/DDBJ under accession numbers: Y11842 (CPV-GRI90 vCD30), M83554 (human CD30), U25416 (mouse CD30), AJ507059 (EV Hampstead vCD30), and AJ507060 (EV Naval vCD30). 832 Ectromelia Virus–encoded CD30 Homologue moter was constructed. PCR analysis indicated the absence human IgG1. The recombinant product (vCD30-Fc) was of the vCD30 gene in VV WR (not depicted). Pulse label- purified by affinity chromatography on protein A Sepharose ing experiments with [ S]methionine and cysteine showed and its molecular size was consistent with vCD30 encoding that vCD30 was efficiently secreted from infected cells as a a 12-kD protein (Fig. 2 b). In addition, matrix-assisted laser 12-kD protein (Fig. 2 a). The EV vCD30 gene was also ex- desorption ionization time-of-flight mass spectrometry pressed in the baculovirus system fused to the Fc region of analysis of the purified vCD30-Fc (performed by S. Mas and B. Aguado, Medical Research Council Human Ge- nome Mapping Project, Cambridge, United Kingdom) and the minor contaminant band only identified peptides corre- sponding to vCD30 or the Fc region of human IgG1, dem- onstrating the high purity of the vCD30-Fc preparation and indicating that the minor contaminant band results from proteolytic degradation of vCD30-Fc (not depicted). CD30L Binding Activity, Specificity, and Affinity of vCD30. CD30L binding activity was determined using a scintil- lation proximity assay. Purified recombinant vCD30-Fc or mouse CD30–Fc were incubated for 12 h with radio- labeled mouse CD30L in protein A–coated FlashPlates containing a thin layer of scintillant in the interior of each well. In this assay, recombinant CD30 proteins are immo- bilized by the Fc portion and bound I-CD30L induces a signal detectable in a scintillation counter. Free I-CD30L is not detected and there is no need to remove it. As shown in Fig. 3 a, both vCD30-Fc and mouse CD30–Fc specifi- Figure 2. Expression of vCD30 in the VV expression system and as an Fc fusion protein. (a) BSC-I cells were infected with VV-WR or Figure 3. (a) CD30L binding activity of vCD30-Fc and (b) recombi- 35 35 VVCD30 and pulse labeled with [ S]cysteine and [ S]methionine from 4 nant vCD30 expressed from VVCD30. (a) 5 ng vCD30-Fc or mouse to 8 h after infection. Proteins present in cells and media were analyzed CD30–Fc were mixed with 150 pM mouse I-CD30L in the presence or by SDS-PAGE in the presence of 2-ME and visualized by fluorography. absence of unlabeled CD30L and added without preincubation to protein The position of the expressed protein in supernatants and cell extracts is A–coated FlashPlates. (b) 75 l supernatants, equivalent to 1.5 10 cells, indicated. (b) Hi5 insect cells were infected with the recombinant bacu- from cultures mock infected or infected with VVCD30 or WR were pre- lovirus expressing vCD30-Fc, harvested 3 d after infection, and the re- incubated with 200 pM mouse I-CD30L, mixed with 5 ng vCD30-Fc, combinant protein was purified by affinity chromatography in a protein A and then added to the protein A–coated FlashPlates. Bound I-CD30L column. The fractions containing the purified vCD30-Fc were then con- was determined in a Packard Topcount microplate counter. The back- centrated and analyzed by SDS-PAGE. The position of vCD30-Fc is in- ground radioactivity in the absence of recombinant protein has been sub- dicated. In both a and b, molecular masses in kD are shown. tracted. I-CD30L binding of duplicate samples (mean SD) is shown. 833 Saraiva et al. 125 cally bound I-CD30L. Interestingly, the viral protein The EV-encoded CD30 Is Expressed at Late Times after In- seemed to have a better binding capacity than the mouse fection and Binds CD30L. To investigate the binding ac- counterpart and this was observed over a range of protein tivity of natural vCD30, supernatants from cells unin- concentrations (not depicted). This difference may reflect a fected or infected with EV isolates Hampstead or Naval slightly higher affinity of vCD30 or the loss of some bio- were tested in binding assays. The naturally produced EV logical activity of the mouse protein during the purification protein efficiently blocked the binding of mouse I- procedure. Control IgG1 did not show binding activity in CD30L to vCD30-Fc (Fig. 5 a). Moreover, vCD30 was this assay (not depicted). vCD30 did not bind iodinated expressed at late times after infection because supernatants human TNF in similar binding assays (not depicted). prepared in the presence of AraC, an inhibitor of DNA The activity of the recombinant vCD30 expressed from replication that allows early protein expression but pre- VV (VVCD30) was determined by measuring its ability to vents synthesis of late viral proteins, did not show binding block the binding of I-CD30L to vCD30-Fc. Radiola- activity (Fig. 5 b). Failure to detect by RT-PCR vCD30- beled CD30L was incubated with supernatants from cells specific transcripts in cell extracts at early times after infec- uninfected or infected with VVCD30 or VV WR before tion in the presence of AraC confirmed this result (not its addition to vCD30-Fc in the FlashPlate binding assay. depicted). As a positive control for the RT-PCR, we de- Only the VVCD30 supernatant blocked the binding of tected transcripts specific for the early gene encoding the I-CD30L to the vCD30-Fc. This confirmed the absence viral epidermal growth factor (not depicted). There is no of CD30L binding activity in VV WR and demonstrated consensus poxvirus late promoter sequence upstream of that the ORF cloned into the recombinant virus VVCD30 the vCD30 gene (9). was expressed as a secreted protein that binds CD30L (Fig. 3 b). TNF binding activity at the surface of cells infected with VV strain Lister has been reported (29). However, binding assays of I-CD30L to cells infected with VVCD30 failed to detect CD30L binding activity at the cell surface (not depicted). The binding affinity of mouse and vCD30 for mouse CD30L was determined in binding assays using the protein A–coated FlashPlates with increased doses of labeled ligand. Scatchard analyses showed an affinity of 0.66 0.30 nM (mean of three independent experiments, one of which is represented in Fig. 4) for vCD30. The affinity determined for mouse CD30 was 0.93 0.11 nM (mean of three inde- pendent experiments, not depicted), comparable to that determined by other methods for the interaction of human CD30 with mouse or human CD30L (2.5 0.3 nM; ref- erence 3). Figure 5. CD30L binding activity and kinetics of production of the natural EV vCD30. (a) EV sCD30L binding activity. 200 pM mouse 125 4 I-CD30L was preincubated with supernatant, equivalent to 1.5 10 cells, from BSC-I cells mock infected or infected with EV strains Hamp- stead or Naval, and then incubated with 5ng vCD30-Fc in a protein A–coated FlashPlate. The binding of I-CD30L was determined in a Pack- ard Topcount microplate counter. (b) BSC-I cells were mock infected or infected with EV strain Hampstead in the absence or presence of AraC. 6 Figure 4. Affinity of vCD30 for mouse CD30L. Half a nanogram of (Early) or 24 h after infection (Late), supernatants were harvested and an aliquot, equivalent to 5 10 cells, was tested for its ability to block the vCD30 was incubated in the protein A–coated FlashPlates with different 125 125 125 amounts of mouse I-CD30L, and the radioactivity bound was deter- binding of 200 pM I-CD30L to 5 ng vCD30-Fc. Bound I-CD30L mined in a Packard Topcount microplate counter. Data were converted to was determined as described above. The background radioactivity corre- the Scatchard coordinate system and analyzed with the LIGAND software. sponding to the binding medium has been subtracted. Specific I-CD30L binding of duplicate samples (mean SD) is shown. Specific I-CD30L binding of duplicate samples (mean SD) is shown. 834 Ectromelia Virus–encoded CD30 Homologue Biological Activity of vCD30. The ability of vCD30 to block CD30L binding to cell surface receptors was investi- gated. First, we screened by flow cytometry human and mouse cell lines for CD30 expression using soluble recom- binant mouse CD30L that cross reacts with human recep- tors (3). Staining profiles indicated that human monocyte K562 cells expressed high levels of CD30 (Fig. 6 a). The addition of a 25-fold excess of vCD30-Fc, but not of IgG1, efficiently blocked CD30L binding to K562 cells (Fig. 6 a). As little as a 10-fold excess of vCD30-Fc was sufficient to interfere with the binding of CD30L to its cellular recep- tors (not depicted). In addition to blocking the CD30–CD30L interaction, vCD30 had the potential to bind CD30L expressed at the cell surface and trigger intracellular signals. Neutrophils constitutively express CD30L, but not CD30, and rap- idly produce IL-8 when stimulated by CD30 (4). There- fore, we analyzed the production of IL-8 by freshly iso- lated human neutrophils in response to immobilized human, mouse, or vCD30-Fc proteins or IgG1. As shown in Fig. 6 b, the viral homologue induced a response com- parable to that of the mammalian receptors, whereas hu- man IgG1 had no effect on IL-8 production. The latter indicated that in this experimental system, CD30– CD30L interactions leading to IL-8 production do not occur in the absence of vCD30. Therefore, the effect ob- served is due to signaling through CD30L after interac- tion with vCD30 and not to the inhibition of CD30– CD30L interactions by vCD30 acting as a soluble decoy receptor. Finally, we investigated a possible role of vCD30, and indirectly of the CD30–CD30L pair, in the development Figure 6. Biological activities of vCD30. (a) Blockade of the binding of T cell responses in vitro. We determined the influence of soluble CD30L to CD30 expressed at the cell membrane. K562 cells of vCD30 on the activation of IFN-–producing cells in an were incubated with a 6 histidine-tagged CD30L and binding was de- MLR. As shown in Fig. 6 c, vCD30 almost completely ab- tected with the mouse Ab specific for the histidine tag on the CD30L rogated the production of IFN- by splenocytes from molecule followed by an FITC-goat anti–mouse Ab (filled histogram). This binding was competed by a 25-fold excess of vCD30-Fc (right, solid BALB/c mice exposed to L929 cells of different haplotype. line), but not by the same excess of IgG1 (right, dashed line). The profile The presence of vCD30 in the priming phase was sufficient of unstained cells (left, dashed line) and of cells stained in the absence of to cause this effect. This result suggested an important role CD30L (left, solid line) is also shown. (b) IL-8 production by neutrophils of the CD30–CD30L interaction for the establishment of T induced by human, mouse, or vCD30-Fc. Freshly isolated neutrophils were incubated for 5 h in the presence of the indicated immobilized Fc cell responses, particularly at early stages of activation. fusion proteins or IgG1. Supernatants were harvested and assayed for IL-8 Moreover, the viral protein might be targeting this interac- production by ELISA. IL-8 secretion of duplicate samples (mean SD) is tion to protect the virus against host T cell responses. The shown. The figure shows one representative experiment of three done presence of IgG1 had no effect on the development of the with neutrophils from different donors. (c) Role of vCD30 in in vitro T T cell response (not depicted). cell responses. Freshly isolated splenocytes from BALB/c mice were mixed with irradiated L929 cells in the absence or presence of vCD30. 5 d Biological Role of vCD30 In Vivo. To address the po- later, IL-2 was added and the incubation was held for an additional 2 d. tential immunomodulatory activity of vCD30 in vivo, the After this priming phase, the viable cells were harvested and incubated effects of vCD30-Fc treatment on inflammation in a with irradiated L929 for 20 h in the presence of vCD30. Cell activation pulmonary granuloma model were investigated. In this was measured by their ability to produce IFN-. The number of cells se- creting IFN was assayed by ELISPOT. The result of duplicate samples model, type 1 and 2 cytokine–mediated granulomas are in- (mean SD) is shown. vCD30/, vCD30 added only in the priming duced in mouse lungs by bead-immobilized mycobacterial phase; vCD30/vCD30, vCD30 added in both phases. PPD or SEA, respectively. Mice were treated with vCD30-Fc or control IgG1 during Ag sensitization and elicitation (four treatments over 18 d), or only during elic- treated mice (Fig. 7, a and b). This effect was observed itation (two treatments over 4 d). The administration of when vCD30 was administered throughout both the sensi- vCD30-Fc to mice caused a significant impaired type 1 cy- tization and elicitation response or only during elicitation tokine–mediated inflammatory response with pulmonary (Fig. 7 b). In contrast, type 2 cytokine–mediated pulmo- granuloma size reduced 80% compared with IgG1- nary inflammation was not modified by vCD30 adminis- 835 Saraiva et al. Figure 7. Role of vCD30 during in vivo type 1 or 2 cytokine–mediated pulmonary inflamma- tion. BALB/c mice were type 1 or 2 sensitized and pulmonary granulomas were elicited as de- scribed in Materials and Methods. Mice were treated with vCD30-Fc or control IgG1 through- out the sensitization (S) and elicitation (E; four treatments) or only during elicitation (two treat- ments). (a) Photomicrographs of hematoxylin and eosin–stained sections of representative type 1 (lymphocytes-, monocytes-, and neutrophils- rich) or 2 (eosinophil-dominated) pulmonary granulomas surrounding Ag-coated beads in mice treated with vCD30-Fc or IgG1. (b) Quantifica- tion of the granuloma volume (mean SD) from four or five mice per group. Significantly smaller type 1 granulomas (*, P  0.01; Student’s t test) were found in mice treated two or four times with vCD30 compared with IgG1-treated mice. (c) vCD30-Fc treatment reduced lung tissue IFN- and IL-12 in type 1–sensitized mice. The levels of IFN-, IL-12, IL-4, and IL-5 in lung tissue homogenates were tested in ELISA and are expressed as nanogram cytokine per milligram lung protein. (d) Intracellular detection of alter- ations in the frequencies of IFN-– and IL-4– producing CD4 and CD8 T cells in type 1– or 2–sensitized mice. Data in quadrants represent the percentage of positively stained cells. All data are representative from two separate experiments (n four or five mice per group). Cytokine and FACS data presented are from mice treated four times with vCD30-Fc. Comparable data was ob- tained in two separate experiments when mice were only treated two times. tration, as vCD30-Fc–treated mice had comparable or 1 cytokine levels occurred when vCD30 was administered marginally larger granulomas than obtained in control ani- throughout the experiment (Fig. 7 c) or only during elici- mals (Fig. 7, a and b). The composition of the cellular in- tation (not depicted). These results demonstrate that the filtrate surrounding the type 1 or 2 granulomas was not al- formation of type 1, but not type 2, cytokine–mediated tered by vCD30 treatment (not depicted). pulmonary inflammation was impaired by vCD30 in vivo. The effect of vCD30 on pulmonary cytokine responses To address if the reduced IFN- production in vCD30- was measured in lung tissue homogenates. Levels of type 1 treated mice was associated with alterations in the produc- or 2 cytokines in the lungs of control IgG1–treated mice tion of type 1 (IFN-) or type 2 (IL-4) cytokines by T were biased in the respective type 1– (elevated IFN- and cells, we performed intracellular cytokine staining on IL-12) or 2– (IL-4 and IL-5) mediated pulmonary inflam- CD4 or CD8 T cells from spleens. Cells from the spleens matory responses (Fig. 7 c). vCD30 treatment of mice with of type 1 granuloma–sensitized mice treated with vCD30 type 1 granulomas caused a substantial reduction in lung had two- to threefold lower frequencies of both Th1 and IFN- and albeit to a lesser degree, IL-12 levels (Fig. 7 c) Tc1 cells compared with control IgG1–treated mice (Fig. 7 with no alterations in IL-4 and IL-5 levels. Cytokine levels d). Similarly, vCD30 treatment also reduced the numbers in mice with type 2 granulomas were not altered by of IFN-–secreting T cells in type 2 granuloma–sensitized vCD30, although IFN- and IL-12 levels were lower in mice (Fig. 7 d). These results support a preferential effect of vCD30-treated mice (Fig. 7 c). Changes in pulmonary type vCD30 on type 1 T cells. 836 Ectromelia Virus–encoded CD30 Homologue In addition, the role of vCD30 in the development of transmembrane or cytoplasmic domains. The extracellular Th1-like responses was controlled with a truncated version domain of human CD30 consists of a duplicate structure of of another EV vTNFR. A recombinant inactive form of three CRDs, whereas mouse CD30 only has the first clus- CrmD consisting of the first two CRDs (CrmD-CRD1,2- ter and the viral counterpart have the second and third Fc) and unable to protect cells from TNF–induced necrosis CRDs. This demonstrates direct involvement of the sec- was produced in the baculovirus system and purified ac- ond and third CRDs in ligand binding. Moreover, the cording to the same protocols used for vCD30-Fc (unpub- conservation of some motifs in the three molecules may lished data). Mice exposed to mycobacterium or schisto- point at specific residues involved in CD30L binding. soma antigens as described above were treated with Both mouse and human CD30 can be proteolytically vCD30-Fc, CrmD-CRD1,2-Fc, or PBS, and the levels of cleaved by a zinc metalloprotease to produce a soluble cytokines in the lung were determined. Table I shows form (sCD30) that is larger than vCD30. The viral protein that vCD30-Fc, but not CrmD-CRD1,2-Fc, specifically appears to retain the minimal structure necessary for effi- down-regulated Th1 cytokine production, confirming the cient CD30L binding activity. The higher similarity of previous result and demonstrating that the role attributed vCD30 to mouse CD30 is consistent with mice being the to vCD30-Fc was specific. natural host for EV. vCD30 is the fifth member of the TNFR superfamily identified in poxviruses and the only one that does not bind Discussion TNF, which suggests a distinct role during viral infection. The sequence similarity between vCD30 and the other We report the identification and characterization of a homologue of mammalian CD30, designated vCD30 and vTNFRs (CrmB, CrmC, CrmD, and CrmE) is confined to the CRDs. CrmB and CrmD show an extended COOH- encoded by EV. vCD30 is a new member of the TNFR superfamily that binds CD30L. Comparative studies in 10 terminal region with no sequence similarities in databases, which is not required for TNF binding (unpublished data). other EV isolates showed that all of them encode an active vCD30 (unpublished data). The complete genome se- Interestingly, the ORF downstream vCD30 in the EV strain Naval genome has similarity to the COOH-terminal quence of EV has been recently determined (available at http://www.sanger.ac.uk) and shows that the vCD30 gene extension found in some vTNFRs and no intergenic re- gion is found (unpublished data). We propose that the an- described here is the only CD30 homologue encoded by the virus (unpublished data). Thus it is most likely that the cestral vCD30 gene may have contained the COOH-ter- minal extension that has been lost in the gene presently gene we have characterized is responsible for the activity expressed in EV infections. The presence of an active found in EV. vCD30 is a 12-kD protein secreted at late times during vCD30 seems restricted to a few poxviruses. An intact vCD30 gene has been identified in EV isolates and CPV EV infection that binds CD30L with high affinity and in- hibits its binding to cell surface CD30. This function of GRI-90 (30), whereas it is not found in VV strain WR (this paper) or in 18 other poxviruses including 3 strains of vCD30 is similar to that of the poxvirus soluble cytokine decoy receptors, including vTNFRs. However, vCD30 has VaV, whose complete genome has been sequenced and is available in public databases. Maybe different immuno- an additional unique property. It induces reverse signaling in cells expressing CD30L. Therefore, vCD30 may not be modulatory genes compensate for the lack of vCD30 in other poxviruses. classified as a genuine decoy receptor and we propose that the mechanism of action of vCD30 is totally different from Amino acid sequence similarity of vCD30 with the hu- man and mouse counterparts is confined to a region of the that adopted by vTNFRs. Because secreted CD30L has not been identified, vCD30 is acting at the cell surface, both extracellular domain because the viral protein lacks the Table I. Modulation of Cytokine Production by vCD30 in a Mouse Model of Inflammation Th1 response Th2 response vCD30-Fc CrmD-CRD1,2-Fc PBS vCD30-Fc CrmD-CRD1,2-Fc PBS IFN- 0.33 0.2 3.31 0.57 3.64 0.47 0.29 0.16 0.94 0.19 0.81 0.26 IL-12 0.82 0.28 3.35 0.6 3.20 0.73 0.17 0.08 0.39 0.15 0.43 0.16 IL-4 0.34 0.09 0.05 0.04 0.09 0.05 3.12 0.49 3.81 0.44 2.93 0.63 IL-5 0.88 0.2 0.38 0.12 0.32 0.21 5.93 0.89 5.72 0.67 5.75 0.98 Cytokine responses in the lungs of mice exposed to mycobacterium (Th1 response) or schistosoma antigens (Th2 response) and treated on days 14 and 16 with 10 g vCD30-Fc, CrmD-CRD1,2-Fc, or PBS. Levels of IFN-, IL-12, IL-4, and IL-5 in lung tissue homogenates as tested by ELISA and expressed as nanogram cytokine per milligram lung protein (mean SD) from five mice per group. 837 Saraiva et al. mimicking signal transduction mediated by CD30 and in- other example of proteins from virulent viruses that have hibiting the effect of CD30L in cells expressing CD30. We promise as therapeutic reagents. also demonstrate that vCD30 is a soluble molecule with no We thank Sebastian Mas and Begonia Aguado for the mass spec- membrane-associated binding activity, as it has been de- trometry analysis of vCD30-Fc. scribed for TNF binding activity in VV Lister (29). This work was funded by the Wellcome Trust (grants 051087/Z/ The expression of CD30 is associated with the activated 97/Z and 057381). M. Saraiva is funded by Fundacao para a Ciencia status of T cells. In vitro, it has been mainly associated with e Tecnologia-Praxis XXI (grant BD-18081/98). P.G. Fallon was a Th2/Th0 phenotype (38), although in vivo studies sug- supported by a Wellcome Trust Career Development Fellowship gest that the relationship between CD30 T cells and Th1 and A. Alcami is a Wellcome Trust Senior Research Fellow. or Th2 profiles is very complex. Recent studies support a Submitted: 26 February 2002 novel regulatory mechanism for CD30 in Th1 polarized Revised: 21 June 2002 responses such as rheumatoid arthritis (39). We show that Accepted: 8 August 2002 the blockade of the binding of CD30L to CD30 by the vi- ral protein and/or the activation of CD30L by vCD30 is responsible for the inhibition of IFN- production by acti- References vated splenocytes in MLR. The potent in vivo inhibition 1. Schwab, U., H. Stein, J. Gerdes, H. Lemke, H. Kirchner, M. of pulmonary granuloma formation by vCD30 in type 1, Schaadt, and V. Diehl. 1982. Production of a monoclonal but not 2, cytokine–sensitized mice supports a preferential antibody specific for Hodgkin and Sternberg Reed cells of role for CD30–CD30L in type 1 cytokine–mediated re- Hodgkin’s disease and a subset of normal lymphoid cells. Na- sponses. Consistent with our in vitro data, a potential di- ture. 299:65–67. rect affect of vCD30 on type 1 T cells was shown by the 2. Horie, R., and T. Watanabe. 1998. CD30: expression and diminished frequencies of IFN-–producing T cells de- function in health and disease. Semin. Immunol. 10:457–470. tected in vCD30-treated mice. These data support increas- 3. Smith, C.A., H.J. Gruss, T. Davis, D. Anderson, T. Farrah, ing evidence for a key role of CD30 in type 1 responses in E. Baker, G.R. Sutherland, C.I. Brannan, N.G. Copeland, vivo. Interestingly, vCD30 treatment was effective in sup- N.A. Jenkins, et al. 1993. CD30 antigen, a marker for pressing type 1 cytokine–mediated inflammation when ad- Hodgkin’s lymphoma, is a receptor whose ligand defines an ministered either throughout the Ag sensitization and elic- emerging family of cytokines with homology to TNF. Cell. 73:1349–1360. itation phase or only during elicitation. In the context of 4. Wiley, S.R., R.G. Goodwin, and C.A. Smith. 1996. Reverse viral infection, this is probably compromising a protective signaling via CD30 ligand. J. Immunol. 157:3635–3639. antiviral immune response, which is known to be Th1-like 5. Amakawa, R., A. Hakem, T.M. Kundig, T. Matsuyama, J.J. T cell mediated. Simard, E. Timms, A. Wakeham, H.W. Mittruecker, H. CD30 is also known to be expressed in activated B cells Griesser, H. Takimoto, et al. 1996. Impaired negative selec- (2). CD30L was found to be a potent mediator of mouse B tion of T cells in Hodgkin’s disease antigen CD30-deficient cell growth and differentiation in vitro, although different mice. Cell. 84:551–562. results were found with human B cells (2). Therefore, 6. Chiarle, R., A. Podda, G. Prolla, E.R. Podack, G.J. Thor- vCD30 may also interfere with B cell responses to the virus. beecke, and G. Inggirami. 1999. CD30 overexpression en- Finally, vCD30 may also modulate signaling between B and hances negative selection in the thymus and mediates pro- grammed cell death via Bcl-2-sensitive pathway. J. Immunol. T cells, a process in which CD30 has been implicated. 163:194–205. Viral mechanisms of immune modulation may point at 7. Muta, H., L.H. Boise, L. Fang, and E.R. Podack. 2000. those host molecules that contribute to antiviral responses. CD30 signals integrate expression of cytotoxic effector mole- The remarkable finding of a virus-encoded CD30 homo- cules, lymphocyte trafficking signals, and signals for prolifera- logue uncovers a potential role of the CD30–CD30L sys- tion and apoptosis. J. Immunol. 165:5105–5111. tem in antiviral immune responses. Increased levels of 8. Gruss, H.J., A. Pinto, J. Duyster, S. Poppema, and F. Herr- sCD30 have been reported in pathological conditions and mann. 1997. Hodgkin’s disease: a tumor with disturbed im- after infection with HIV, EBV, hepatitis virus B and C, munological pathways. Immunol. Today. 18:156–163. measles virus, and varicella-zoster virus, but the biological 9. Moss, B. 1996. Poxviridae: the virus and their replication. In significance is not known. Viral proteins have been opti- Virology. B.N. Fields, R.M. Chanock, D.M. Knipe, and P.M. Howley, editors. Lippincott-Raven Publishers, Phila- mized during virus–host coevolution to become potent in- delphia. 2637–2671. hibitors of host immune responses. Maybe the biological 10. Fenner, F., and R.M. Buller. 1997. Mousepox. In Viral properties of vCD30 differ from those of the host sCD30. Pathogenesis. N. Nathanson, F. Gonzalez-Scarano, R. Characterization of viral immunomodulatory proteins Ahmed, H.L. Robinson, D.E. Griffin, K.V. Holmes, and may also shed light into the function of the host counter- F.A. Murphy, editors. Lippincott-Raven Publishers, Phila- parts. By using vCD30, we demonstrate a role of CD30– delphia. 535–553. CD30L interactions in the generation of Th1 inflammatory 11. Alcami, A., and U.H. Koszinowski. 2000. Viral mechanisms responses. vCD30 may also provide alternative strategies to of immune evasion. Immunol. Today. 21:447–455. effectively block the activity of CD30 in vivo, which might 12. McFadden, G., and P.M. Murphy. 2000. Host-related im- be applied to modulate an overreactive immune response munomodulators encoded by poxviruses and herpesviruses. in a number of human disease conditions (8). vCD30 is an- Curr. Opin. Microbiol. 3:371–378. 838 Ectromelia Virus–encoded CD30 Homologue 13. Tortorella, D., B.E. Gewurz, M.H. Furman, D.J. Schust, and type 2 cytokine-producing mouse CD4 and CD8 T cells in H.L. Ploegh. 2000. Viral subversion of the immune system. acute Schistosoma mansoni infection. Eur. J. Immunol. 28: Annu. Rev. Immunol. 18:861–926. 1408–1416. 14. Loparev, V.N., J.M. Parsons, J.C. Knight, J.F. Panus, C.A. 28. Esposito, J., R. Condit, and J. Obijeski. 1981. The prepara- Ray, R.M. Buller, D.J. Pickup, and J.J. Esposito. 1998. A tion of orthopoxvirus DNA. J. Virol. Methods. 2:175–179. third distinct tumor necrosis factor receptor of orthopoxvi- 29. Alcami, A., A. Khanna, N.L. Paul, and G.L. Smith. 1999. ruses. Proc. Natl. Acad. Sci. USA. 95:3786–3791. Vaccinia virus strains Lister, USSR and Evans express soluble 15. Smith, V.P., and A. Alcami. 2000. Expression of secreted cy- and cell-surface tumour necrosis factor receptors. J. Gen. Vi- tokine and chemokine inhibitors by ectromelia virus. J. Virol. rol. 80:949–959. 74:8460–8471. 30. Shchelkunov, S.N., P.F. Safronov, A.V. Totmenin, N.A. 16. Mossman, K., C. Upton, R.M. Buller, and G. McFadden. Petrov, O.I. Ryazankina, V.V. Gutorov, and G.J. Kotwal. 1995. Species specificity of ectromelia virus and vaccinia vi- 1998. The genomic sequence analysis of the left and right rus interferon- binding proteins. Virology. 208:762–769. species-specific terminal region of a cowpox virus strain re- 17. Smith, V.P., and A. Alcami. 2002. Inhibition of interferons veals unique sequences and a cluster of intact ORFs for im- by ectromelia virus. J. Virol. 76:1124–1134. munomodulatory and host range proteins. Virology. 243:432– 18. Colamonici, O.R., P. Domanski, S.M. Sweitzer, A. Larner, 460. and R.M. Buller. 1995. Vaccinia virus B18R gene encodes a 31. Genetics Computer Group. 1994. Program Manual for the type I interferon-binding protein that blocks interferon  Wisconsin Package. transmembrane signaling. J. Biol. Chem. 270:15974–15978. 32. Davison, A.J., and B. Moss. 1990. New vaccinia virus recom- 19. Smith, V.P., N.A. Bryant, and A. Alcami. 2000. Ectromelia, bination plasmids incorporating a synthetic late promoter for vaccinia and cowpox viruses encode secreted interleukin-18- high level expression of foreign proteins. Nucleic Acids Res. binding proteins. J. Gen. Virol. 81:1223–1230. 18:4285–4286. 20. Born, T.L., L.A. Morrison, D.J. Esteban, T. VandenBos, 33. Alcami, A., J.A. Symons, P.D. Collins, T.J. Williams, and L.G. Thebeau, N. Chen, M.K. Spriggs, J.E. Sims, and R.M. G.L. Smith. 1998. Blockade of chemokine activity by a solu- Buller. 2000. A poxvirus protein that binds to and inactivates ble chemokine binding protein from vaccinia virus. J. Immu- IL-18, and inhibits NK cell response. J. Immunol. 164:3246– nol. 160:624–633. 3254. 34. Markwell, M.A., and C.F. Fox. 1978. Surface-specific iodi- 21. Graham, K.A., A.S. Lalani, J.L. Macen, T.L. Ness, M. Barry, nation of membrane proteins of viruses and eucaryotic cells L.Y. Liu, A. Lucas, I. Clark-Lewis, R.W. Moyer, and G. using 1,3,4,5-tetrachloro-3,6-diphenylglycouril. Biochem- McFadden. 1997. The T1/35kDa family of poxvirus-secreted istry. 17:4807–4817. proteins bind chemokines and modulate leukocyte influx into 35. Munson, P.J., and D. Rodbard. 1980. LIGAND: a versatile virus-infected tissues. Virology. 229:12–24. computerized approach for characterization of ligand-binding 22. Turner, S.J., J. Silke, B. Kenshole, and J. Ruby. 2000. Char- systems. Anal. Biochem. 107:220–239. acterization of the ectromelia virus serpin, SPI-2. J. Gen. Vi- 36. Chensue, S.W., K. Warmington, J.H. Ruth, N. Lukacs, and rol. 81:2425–2430. S.L. Kunkel. 1997. Mycobacterial and schistosomal antigen- 23. Brick, D.J., R.D. Burke, A.A. Minkley, and C. Upton. 2000. elicited granuloma formation in IFN- and IL-4 knockout Ectromelia virus virulence factor p28 acts upstream caspase-3 mice: analysis of local and regional cytokine and chemokine in response to UV light-induced apoptosis. J. Gen. Virol. 81: networks. J. Immunol. 159:3565–3573. (See published erra- 1087–1097. tum 162:3106.) 24. Hu, F.Q., C.A. Smith, and D.J. Pickup. 1994. Cowpox virus 37. Fallon, P.G., and D.W. Dunne. 1999. Tolerization of mice contains two copies of an early gene encoding a soluble se- to Schistosoma mansoni egg antigens causes elevated type 1 creted form of the type II TNF receptor. Virology. 204:343– and diminished type 2 cytokine responses and increased mor- 356. tality in acute infection. J. Immunol. 162:4122–4132. 25. Smith, C.A., F.Q. Hu, T.D. Smith, C.L. Richards, P. Smo- 38. Del Prete, G., M. De Carli, F. Almerigogna, C.K. Daniel, lak, R.G. Goodwin, and D.J. Pickup. 1996. Cowpox virus M.M. D’Elios, G. Zancuoghi, F. Vinante, G. Pizzolo, and S. genome encodes a second soluble homologue of cellular Romagnani. 1995. Preferential expression of CD30 by hu- TNF receptors, distinct from CrmB, that binds TNF but not man CD4 T cells producing Th2-type cytokines. FASEB J. LT . Virology. 223:132–147. 9:81–86. 26. Saraiva, M., and A. Alcami. 2001. CrmE, a novel soluble tu- 39. Gerli, R., C. Lunardi, F. Vinante, O. Bistoni, G. Pizzolo, and mor necrosis factor receptor encoded by poxviruses. J. Virol. C. Pitzalis. 2001. Role of CD30 T cells in rheumatoid ar- 75:226–233. thritis: a counter-regulatory paradigm for Th1-driven dis- 27. Fallon, P.G., P. Smith, and D.W. Dunne. 1998. Type 1 and eases. Trends Immunol. 22:72–77. 839 Saraiva et al. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journal of Experimental Medicine Pubmed Central

Inhibition of Type 1 Cytokine–mediated Inflammation by a Soluble CD30 Homologue Encoded by Ectromelia (Mousepox) Virus

The Journal of Experimental Medicine , Volume 196 (6) – Sep 16, 2002

Loading next page...
 
/lp/pubmed-central/inhibition-of-type-1-cytokine-mediated-inflammation-by-a-soluble-cd30-XVPPrkEGBa

References (79)

Publisher
Pubmed Central
Copyright
Copyright © 2002, The Rockefeller University Press
ISSN
0022-1007
eISSN
1540-9538
DOI
10.1084/jem.20020319
Publisher site
See Article on Publisher Site

Abstract

CD30 is up-regulated in several human diseases and viral infections but its role in immune reg- ulation is poorly understood. Here, we report the expression of a functional soluble CD30 ho- mologue, viral CD30 (vCD30), encoded by ectromelia (mousepox) virus, a poxvirus that causes a severe disease related to human smallpox. We show that vCD30 is a 12-kD secreted protein that not only binds CD30L with high affinity and prevents its interaction with CD30, but it also induces reverse signaling in cells expressing CD30L. vCD30 blocked the generation of interferon –producing cells in vitro and was a potent inhibitor of T helper cell (Th)1- but not Th2-mediated inflammation in vivo. The finding of a CD30 homologue encoded by ec- tromelia virus suggests a role for CD30 in antiviral defense. Characterization of the immuno- logical properties of vCD30 has uncovered a role of CD30–CD30L interactions in the genera- tion of inflammatory responses. Key words: poxviruses • immunomodulation • cytokine • TNFR superfamily • Th-1 Introduction CD30 and CD30L (CD153) are members of the TNFR cells, activated T cells and macrophages, and in neoplasic and TNF superfamilies, respectively. CD30 was identified cells such as Burkitt-type lymphoma cells or B cells associ- by the mAb Ki-1 against Hodgkin and Reed-Sternberg cells ated with lymphoproliferative disorders. The extracellular (1), the malignant component of Hodgkin’s disease, and has domain of CD30L shows homology to TNF, lymphotoxin, been extensively used as a clinical disease marker. CD30 was and CD40L (3). It is unclear whether CD30L also exists as a subsequently found in resting CD8 T cells, activated or vi- soluble form, as does TNF. Interaction of CD30L with cells rally transformed T and B cells, and at the surface of HIV- expressing CD30 induces signals mediated by nuclear factor infected lymphocytes. CD30 is a type I membrane protein B and TNFR-associated factor 2 that cause cell prolifera- that can be cleaved by metalloproteases producing a soluble tion or cell death. Interestingly, upon binding to CD30, form (sCD30;* reference 2). The extracellular domain of CD30L is also able to signal. One of the consequences of CD30 shows cysteine-rich domains (CRDs) characteristic this reverse signaling is cell proliferation (4). of the TNFR superfamily. CD30L is expressed as a type II The role of CD30–CD30L interaction in health and dis- membrane glycoprotein (2) in resting neutrophils and B ease is still not totally understood, in part due to the pleio- tropic nature of CD30 signals. Mice lacking a functional CD30 gene show defective negative thymocyte selection Address correspondence to Antonio Alcami, Department of Medicine, (5), whereas transgenic mice expressing CD30 in the thy- University of Cambridge, Addenbrooke’s Hospital, Level 5, Box 157, mus have enhanced thymic negative selection (6). A recent Hills Road, Cambridge CB2 2QQ, United Kingdom. Phone: 44-1223- study of genes targeted by CD30 suggests that Fas, TRAIL, 763403; Fax: 44-1223-330158; E-mail: aa258@mole.bio.cam.ac.uk CCR7, TNFR-associated factor 1, and cIAP2 are up-regu- P. Smith’s and P.G. Fallon’s present address is Department of Bio- lated whereas FasL, perforin, granzyme B, and c-myc seem chemistry, Trinity College, Dublin 2, Ireland. *Abbreviations used in this paper: AraC, cytosine arabinoside; CPV, cow- to be down-regulated (7). Finally, the reasons for increased pox virus; CRD, cysteine-rich domain; Crm, cytokine response modifier; levels of sCD30 in malignant lymphomas, viral infection EV, ectromelia virus; ORF, open reading frame; PPD, protein-purified (HIV, human T cell leukemia virus, and EBV), and several derivative; sCD30, soluble CD30; SEA, soluble egg Ags; VaV, variola vi- immunological disorders such as systemic lupus erythema- rus; vCD30, viral CD30; vTNFR, viral TNFR; VV, vaccinia virus; WR, Western Reserve. tosus or rheumatoid arthritis are not known (8). 829 J. Exp. Med.  The Rockefeller University Press • 0022-1007/2002/09/829/11 $5.00 Volume 196, Number 6, September 16, 2002 829–839 http://www.jem.org/cgi/doi/10.1084/jem.20020319 Poxviruses are a family of complex DNA viruses that en- nate mouse CD30L were purchased from Amersham Biosciences and Pierce Chemical Co., respectively. The protein A–coated code up to 200 genes and infect a wide variety of hosts (9). FlashPlates used for binding and affinity studies were purchased Smallpox was a devastating disease caused by variola virus from PerkinElmer. The mouse mAb specific for the histidine tag (VaV), one of the most virulent human pathogens. Vac- and the FITC-conjugated goat anti–mouse Igs used in flow cy- cinia virus (VV) is the best characterized poxvirus and the tometry were from CLONTECH Laboratories, Inc. and Dako, vaccine used to achieve the global eradication of smallpox respectively. Mycobacterium tuberculosis whole cell lysates (H37Rv by 1977, but its origin and natural host are unknown. strain) and protein-purified derivative (PPD) were obtained from Cowpox virus (CPV) is probably a rodent virus that spo- Mycos Research. Schistosome eggs and soluble egg Ags (SEA) radically infects other animal species. Ectromelia virus (EV) were prepared as previously described (27). is a highly virulent natural pathogen of mice that causes Cells, Viruses, and Viral DNA Preparations mousepox and has been isolated from outbreaks in labora- tory mouse colonies (10). Like VaV, EV has a restricted The growth of BSC-I, TK 143B, and K562 cells, and the sources of VV Western Reserve (WR) strain and EV isolates host range, causes severe disease with high mortality rate Hampstead and Naval have been described (15). VV and EV and skin lesions in the later stages of infection. These simi- were propagated in BSC-I cells and viral genomic DNA was pre- larities with smallpox make EV an interesting experimental pared as previously described (28). The growth of Autographa cali- model for virus–host interactions. fornica nuclear polyhedrosis virus in Spodoptera frugiperda 21 insect Poxviruses encode a unique collection of genes that cells has been described (29). Tn5 B1-4 (Hi5) insect cells were evade host immune responses. These molecules are often cultured in EX-CELL serum-free medium as suggested by the secreted and include cytokine homologues and soluble cy- supplier (European Collection of Cell Cultures). tokine receptors or binding proteins (11–13). Some of these viral genes seem to have been acquired from the host DNA Sequencing and modified during virus evolution to confer an advantage Specific oligonucleotides, CD30-1 (5 GTTCTGGATACAT- for virus replication, survival, or transmission. EV encodes GCACAAAG 3 ) and CD30-2 (5 GGAGGATAATCATTTG- receptors or binding proteins for TNF (14), IL-1 (15), CAAACG 3 ), were designed based on the sequence of CPV strain GRI90 open reading frame (ORF) D13L (30), and used to IFN- (16, 17), IFN-/ (17, 18), IL-18 (19, 20), and amplify the cognate genes from viral DNA preparations from VV chemokines (15, 21). EV also encodes antiapoptotic pro- WR and EV Hampstead and Naval by PCR using Taq DNA teins (22, 23) and an intracellular protein that confers IFN polymerase. PCR products were sequenced by the DNA Se- resistance (17). quencing Service of the Department of Biochemistry, Cambridge The TNF binding activity encoded by orthopoxviruses University, Cambridge, United Kingdom. The sequence data is particularly interesting because there are four distinct were analyzed using Genetics Computer Group computer pro- viral TNFRs (vTNFRs): cytokine response modifier grams (31). (Crm) B (24), CrmC (25), CrmD (14), and CrmE (26). These molecules show different ligand specificity and are Construction of Recombinant Baculovirus Expressing the EV Hampstead vCD30 Gene expressed at different times after infection, but their rela- tive contribution to viral pathogenesis is not well under- The EV Hampstead vCD30 gene was amplified by PCR using Pfu DNA polymerase, virus DNA as template, and oligonucle- stood. Maybe the vTNFRs bind other members of the otides CD30-3 (5 CGCAAGCTTGGATCCATGAAGATGA- growing TNF family and protect the virus from the ac- ATACTATC TTTTTATC 3 ) and CD30-4 (5 CGCGCG- tion of other ligands. GCCGCTGATGAGTATTTATGATAACAAAG 3 ), which We report the identification and characterization of a correspond to the 5 and 3 ends of the ORF and provide Hind- novel member of the TNFR superfamily, a homologue of III/BamHI and NotI sites, respectively. The resultant product CD30 encoded by EV that blocks the binding of CD30L was cloned into HindIII- and NotI-digested pBac1 (Novagen), to its receptor and induces reverse signaling in cells express- creating plasmid pMS2 (EV Hampstead vCD30). The DNA se- ing CD30L. Moreover, the viral CD30 (vCD30) abrogates quence of the insert was confirmed to not contain mutations. T cell proliferation in vitro and in vivo it blocks type 1 but The Fc fragment of the human IgG1 was cut from pIGplus not type 2 cytokine–mediated T cell responses. These stud- (R&D Systems) and subcloned into NotI/SphI sites of pMS2, ies not only describe a novel immunomodulatory strategy which created plasmid pMS18 (EV Hampstead vCD30-Fc). Re- combinant baculovirus was produced as previously described (29) of poxviruses, but also pave the way to the role of CD30– and termed AcCD30-Fc (EV Hampstead vCD30-Fc, AcMS18). CD30L in viral infections and in type I cytokine–mediated Control recombinant baculovirus expressing EV Hampstead– inflammatory diseases. truncated CrmD (AcCrmD-CRD1,2-Fc) was constructed as AcCD30-Fc (unpublished data). Purification of the Baculovirus Recombinant vCD30-Fc Protein Materials and Methods Hi5 cultures were infected with recombinant baculoviruses at Reagents 10 pfu/cell and supernatants were harvested 3–4 d later when full Recombinant mouse CD30L (ED  50  150 g/ml), hu- infection was observed. The recombinant Fc fusion proteins were man CD30 (ED  0.03  0.1 g/ml), and mouse CD30 (ED  subsequently purified using Protein A HiTrap columns (Amer- 50 50 0.03  0.1 g/ml) were purchased from R&D Systems. The io- sham Biosciences). The purified protein was then analyzed by dine-125 (103.7 mCi/ml) and iodogen reagent used to radioiodi- SDS-PAGE in 12% acrylamide gels and stained with Coomassie 830 Ectromelia Virus–encoded CD30 Homologue blue. Protein concentration was determined using the Bio-Rad Kinetics of vCD30 Production During EV Infection protein assay reagent. BSC-I cells were mock infected or infected with 10 pfu EV Hampstead per cell in the absence or presence of 40 g/ml cytosine Construction of Recombinant VV Expressing the EV Hampstead arabinoside (AraC), an inhibitor of DNA replication, and harvested vCD30 Gene at different times after infection. Supernatants were inactivated and the CD30 binding activity was tested as described above. Total The EV Hampstead vCD30 gene was amplified by PCR with RNA was extracted using the guanidine thiocyanate–based DNA/ virus DNA as template, Pfu DNA polymerase, and oligonucle- RNA Isolation Kit (Promega) according to the manufacturer’s in- otides CD30-3 and CD30-5 (5 CGCGGTACCTCATGAT- structions. Total RNA (from 7 10 cells) was then analyzed by GAGTATTTATGATAACAAAG 3 ) containing KpnI restric- RT and then by PCR. RT was performed in the presence of tion site. The DNA fragment was cloned into BamHI- and oligo(dT) (Promega), RNAsin (Amersham Biosciences), and KpnI-digested pMJ601 (provided by B. Moss, National Institutes 15 avian myeloblastosis virus RT (Boehringer). The cDNA (2.5 l of of Health, Bethesda, MD; reference 32), creating plasmid pMS12 40 l; provided by K. Shair, University of Cambridge, Cambridge, (EV Hampstead vCD30). The DNA sequence of the insert was United Kingdom) was amplified by PCR using Taq polymerase and confirmed to not contain mutations. The recombinant VV was oligonucleotides specific for vCD30, CD30-3, and CD30-4. DNA produced as previously described (29) and termed VVCD30 (EV from BSC-I cells was included as a negative control. Hampstead vCD30, vMS12). Biological Activity of the vCD30 Metabolic Labeling of VVCD30 and Electrophoretic Analysis In Vitro Studies. 1 g/ml recombinant soluble mouse CD30L BSC-I cells were infected with VV WR or VVCD30 at 10 35 preincubated with RPMI, a 25-fold excess of vCD30, or human pfu/cell. Cultures were pulse labeled with 150 Ci/ml [ S]me- IgG1 for 1.5 h at 4 C was added to 10 K562 cells and incubated thionine (1,200 Ci/mmol; Amersham Biosciences) and 150 Ci/ 35 for 2 h at 4 C. After this period, cells were incubated for 40 min ml [ S]cysteine (600 Ci/mmol; NEN Life Science Products) in at 4 C with a mouse mAb specific for the histidine tag (1 g/ml methionine- and cysteine-free medium in the absence of serum. in 0.1% BSA in PBS) that would recognize CD30L bound to the Cells or media were dissociated in sample buffer and analyzed by cell membrane. This Ab was subsequently developed with an SDS-PAGE in 12% acrylamide gels and visualized by fluorogra- FITC-labeled goat anti–mouse Ig Ab for 30 min at 4 C. Cells phy with Amplify (Amersham Biosciences). binding CD30L were then detected by FACS analysis. Un- stained cells and cells stained in the absence of CD30L were in- Preparation of VV and EV Supernatants cluded as a control. BSC-I cells were mock infected or infected with VV-WR, To test the ability of vCD30 to induce reverse signaling via VVCD30, EV Hampstead, and EV Naval at 10 pfu/cell in phe- membrane bound CD30L, 5 10 freshly isolated human neu- nol red– and serum-free medium. Supernatants were harvested at trophils were incubated in a volume of 100 l for 5 h at 37 C in 2 (for the VV infections) or 3 (for the EV and mock infections) d 96-well plates precoated with 10 g/ml of mouse, human after infection and prepared and inactivated as previously de- vCD30, human IgG1, CD30, or PBS as previously described (4). scribed (33). After this period, supernatants were harvested and the production of IL-8 was measured by ELISA (Diaclone). To address the possi- CD30L Binding Assay ble interference of vCD30 in the development of CTL responses, Recombinant mouse CD30L was radioiodinated to a specific 4 10 freshly isolated splenocytes from BALB/c mice were 6 4 activity of 10 cpm/ g using the Iodogen method (34). Approx- mixed in 96-well plates with 2 10 L929 cells in a final volume imately 150 pM of I-CD30L was incubated for 12 h with 5 ng of 200 l RPMI in the presence or absence of 10 g/ml vCD30- purified vCD30-Fc or recombinant mouse CD30 in a protein Fc or IgG1 supplemented with 10% FCS, sodium pyruvate, and A–coated FlashPlate. The binding medium was phenol red–free nonessential amino acids for 5 d at 37 C, 5% CO . After this pe- MEM, 0.1% BSA, 20mM Hepes, pH 7.5. The amount of riod, IL-2 (Roche) was added to a final concentration of 50 U/ml CD30L bound to the viral receptor was measured in a Packard and the incubation was held for an additional 2 d. Finally, cells Topcount microplate counter. Nonspecific binding was deter- were harvested and the viable cells counted by trypan blue exclu- 125 5 4 mined by incubating I-CD30L with binding medium only. sion. 10 activated splenocytes were mixed again with 10 L929 in For the competition studies, a 500-fold molar excess of cold the presence or absence of 10 g/ml vCD30-Fc or IgG1 and the mouse CD30L was added to the recombinant mouse or viral re- number of cells producing IFN- were measured using an ceptors before the addition of I-CD30L. To test the CD30 ELISPOT assay (R&D Systems). binding activity in supernatants of EV Hampstead and Naval (VV In Vivo Studies. Type 1 and 2 cytokine–dominated pulmo- WR or VVCD30), 50 l supernatant equivalent to 1.5 10 nary granulomas were induced by mycobacterial or SEA, respec- cells were preincubated with I-CD30L before being added to tively, as previously described (36). Female BALB/c mice were the recombinant mouse or viral receptors. For the determination obtained from Harlan. vCD30-Fc or controls CrmD-CRD1,2-Fc of the affinity constant of both mouse and vCD30 to the and IgG1 (Sigma-Aldrich) were injected intraperitoneally (10 g CD30L, binding assays with increasing amounts of I-CD30L per injection) throughout sensitization and elicitation of bead against a fixed amount of recombinant CD30 (2 and 0.5 ng granulomas, i.e., on days 0, 7, 14, and 16, or were injected only mouse or viral protein, respectively) were performed. The results during elicitation of granulomas on days 14 and 16. On day 0, were analyzed with the LIGAND software (35). For the deter- mice were sensitized by intraperitoneal injection of 20 g M. tu- mination of membrane-bound activity of vCD30, BSC-I cells berculosis whole cell lysates in complete Freund’s adjuvant (type 1 were mock infected or infected with VV WR, VVCD30, and cytokine sensitization) or 5,000 Schistosoma mansoni eggs (type 2 EV Hampstead at 10 pfu/cell. 24 h later, human I-CD30L was cytokine sensitization). 14 d later mice were injected intravenously added and bound I-CD30L was determined by phthalate oil with 5,000 Sepharose 4B beads covalently coupled with PPD or centrifugation (29). SEA. Mice were killed 4 d after bead injection on day 18. The left 831 Saraiva et al. lung lobe was snap frozen and used for cytokine analysis. The re- under accession numbers AJ507059 and AJ507060, respec- mainder of the lung was fixed for histological studies. The diame- tively. ters of the granuloma surrounding at least 50 individual beads per mouse were measured. Group mean granuloma volumes from four or five mice per group are presented. The statistical differ- Results ences between groups were determined using Student’s t test. Identification of a Novel Member of the TNFR Superfamily Lung tissue cytokines were determined as previously described Encoded by EV. Analysis of the CPV strain GRI90 se- using ELISA protocols (37). Control naive mouse lungs were pro- quence (30) revealed the presence of an ORF (D13L) with cessed to determine basal lung cytokine levels. Data were ex- sequence similarity to host CD30, a TNFR superfamily pressed as nanogram cytokine per milligram lung protein. The member, and distinct from the previously identified poxvi- spleen and draining mediastinal LN were removed on the day the rus TNFRs (CrmB, CrmC, CrmD, and CrmE). PCR and mice were killed and used for cell culture and intracellular cyto- sequence analysis of the cognate gene in EV isolates Hamp- kine staining (27). In brief, spleen cell suspensions were cultured for 6 h in media alone or in the presence of 6 g/ml Con A (Sigma- stead and Naval showed the existence of an intact vCD30 Aldrich) and 10 g/ml Brefeldin A (Sigma-Aldrich) added during gene (Fig. 1 a). The predicted viral molecule lacked N-gly- the last 4 h. All intracellular detection reagents were from Caltag. cosylation sites, was considerably smaller (12 kD) than the Cells were surface stained with tri-color–conjugated anti-CD4 or mouse or human counterparts (52 and 120 kD, respec- CD8 mAbs. After cell permeabilization, cells were incubated with tively), and aligned with the second and third CRDs found an FITC-conjugated anti–IFN-–mAb, PE-conjugated anti–IL-4 in the extracellular domain of the host CD30, with some mAb, or FITC- or PE-conjugated isotype control mAbs. For motifs highly conserved (Fig. 1 b). Interestingly, vCD30 FACS analysis CD4 or CD8 lymphocytes were gated and showed a higher similarity to mouse (63.7%) than to hu- quadrants were set using isotype control mAbs. The frequencies of man (56.7%) CD30. The existence of a signal peptide and IFN-– and IL-4–stained cells are expressed as percentages. the lack of transmembrane domain suggested that vCD30 Nucleotide Sequence Accession Number may act as a soluble decoy receptor for host CD30L. Characterization of the vCD30 Protein. A recombinant The sequence data for the EV Hampstead– and Naval-encoded CD30 homologue are available from GenBank/EMBL/DDBJ VV WR expressing the EV vCD30 under a strong pro- Figure 1. Sequence of vCD30. Pairwise align- ment of the predicted amino acid sequence of vCD30 in (a) three distinct orthopoxviruses and (b) of human, mouse, and vCD30. Dark shadows rep- resent (a) differences or (b) similarities. Gray boxes, regions of high similarity; , deletions; *, stop codons. The predicted signal peptide (SP), CRDs, and transmembrane domain (TM) are indicated. These sequence data are available from GenBank/ EMBL/DDBJ under accession numbers: Y11842 (CPV-GRI90 vCD30), M83554 (human CD30), U25416 (mouse CD30), AJ507059 (EV Hampstead vCD30), and AJ507060 (EV Naval vCD30). 832 Ectromelia Virus–encoded CD30 Homologue moter was constructed. PCR analysis indicated the absence human IgG1. The recombinant product (vCD30-Fc) was of the vCD30 gene in VV WR (not depicted). Pulse label- purified by affinity chromatography on protein A Sepharose ing experiments with [ S]methionine and cysteine showed and its molecular size was consistent with vCD30 encoding that vCD30 was efficiently secreted from infected cells as a a 12-kD protein (Fig. 2 b). In addition, matrix-assisted laser 12-kD protein (Fig. 2 a). The EV vCD30 gene was also ex- desorption ionization time-of-flight mass spectrometry pressed in the baculovirus system fused to the Fc region of analysis of the purified vCD30-Fc (performed by S. Mas and B. Aguado, Medical Research Council Human Ge- nome Mapping Project, Cambridge, United Kingdom) and the minor contaminant band only identified peptides corre- sponding to vCD30 or the Fc region of human IgG1, dem- onstrating the high purity of the vCD30-Fc preparation and indicating that the minor contaminant band results from proteolytic degradation of vCD30-Fc (not depicted). CD30L Binding Activity, Specificity, and Affinity of vCD30. CD30L binding activity was determined using a scintil- lation proximity assay. Purified recombinant vCD30-Fc or mouse CD30–Fc were incubated for 12 h with radio- labeled mouse CD30L in protein A–coated FlashPlates containing a thin layer of scintillant in the interior of each well. In this assay, recombinant CD30 proteins are immo- bilized by the Fc portion and bound I-CD30L induces a signal detectable in a scintillation counter. Free I-CD30L is not detected and there is no need to remove it. As shown in Fig. 3 a, both vCD30-Fc and mouse CD30–Fc specifi- Figure 2. Expression of vCD30 in the VV expression system and as an Fc fusion protein. (a) BSC-I cells were infected with VV-WR or Figure 3. (a) CD30L binding activity of vCD30-Fc and (b) recombi- 35 35 VVCD30 and pulse labeled with [ S]cysteine and [ S]methionine from 4 nant vCD30 expressed from VVCD30. (a) 5 ng vCD30-Fc or mouse to 8 h after infection. Proteins present in cells and media were analyzed CD30–Fc were mixed with 150 pM mouse I-CD30L in the presence or by SDS-PAGE in the presence of 2-ME and visualized by fluorography. absence of unlabeled CD30L and added without preincubation to protein The position of the expressed protein in supernatants and cell extracts is A–coated FlashPlates. (b) 75 l supernatants, equivalent to 1.5 10 cells, indicated. (b) Hi5 insect cells were infected with the recombinant bacu- from cultures mock infected or infected with VVCD30 or WR were pre- lovirus expressing vCD30-Fc, harvested 3 d after infection, and the re- incubated with 200 pM mouse I-CD30L, mixed with 5 ng vCD30-Fc, combinant protein was purified by affinity chromatography in a protein A and then added to the protein A–coated FlashPlates. Bound I-CD30L column. The fractions containing the purified vCD30-Fc were then con- was determined in a Packard Topcount microplate counter. The back- centrated and analyzed by SDS-PAGE. The position of vCD30-Fc is in- ground radioactivity in the absence of recombinant protein has been sub- dicated. In both a and b, molecular masses in kD are shown. tracted. I-CD30L binding of duplicate samples (mean SD) is shown. 833 Saraiva et al. 125 cally bound I-CD30L. Interestingly, the viral protein The EV-encoded CD30 Is Expressed at Late Times after In- seemed to have a better binding capacity than the mouse fection and Binds CD30L. To investigate the binding ac- counterpart and this was observed over a range of protein tivity of natural vCD30, supernatants from cells unin- concentrations (not depicted). This difference may reflect a fected or infected with EV isolates Hampstead or Naval slightly higher affinity of vCD30 or the loss of some bio- were tested in binding assays. The naturally produced EV logical activity of the mouse protein during the purification protein efficiently blocked the binding of mouse I- procedure. Control IgG1 did not show binding activity in CD30L to vCD30-Fc (Fig. 5 a). Moreover, vCD30 was this assay (not depicted). vCD30 did not bind iodinated expressed at late times after infection because supernatants human TNF in similar binding assays (not depicted). prepared in the presence of AraC, an inhibitor of DNA The activity of the recombinant vCD30 expressed from replication that allows early protein expression but pre- VV (VVCD30) was determined by measuring its ability to vents synthesis of late viral proteins, did not show binding block the binding of I-CD30L to vCD30-Fc. Radiola- activity (Fig. 5 b). Failure to detect by RT-PCR vCD30- beled CD30L was incubated with supernatants from cells specific transcripts in cell extracts at early times after infec- uninfected or infected with VVCD30 or VV WR before tion in the presence of AraC confirmed this result (not its addition to vCD30-Fc in the FlashPlate binding assay. depicted). As a positive control for the RT-PCR, we de- Only the VVCD30 supernatant blocked the binding of tected transcripts specific for the early gene encoding the I-CD30L to the vCD30-Fc. This confirmed the absence viral epidermal growth factor (not depicted). There is no of CD30L binding activity in VV WR and demonstrated consensus poxvirus late promoter sequence upstream of that the ORF cloned into the recombinant virus VVCD30 the vCD30 gene (9). was expressed as a secreted protein that binds CD30L (Fig. 3 b). TNF binding activity at the surface of cells infected with VV strain Lister has been reported (29). However, binding assays of I-CD30L to cells infected with VVCD30 failed to detect CD30L binding activity at the cell surface (not depicted). The binding affinity of mouse and vCD30 for mouse CD30L was determined in binding assays using the protein A–coated FlashPlates with increased doses of labeled ligand. Scatchard analyses showed an affinity of 0.66 0.30 nM (mean of three independent experiments, one of which is represented in Fig. 4) for vCD30. The affinity determined for mouse CD30 was 0.93 0.11 nM (mean of three inde- pendent experiments, not depicted), comparable to that determined by other methods for the interaction of human CD30 with mouse or human CD30L (2.5 0.3 nM; ref- erence 3). Figure 5. CD30L binding activity and kinetics of production of the natural EV vCD30. (a) EV sCD30L binding activity. 200 pM mouse 125 4 I-CD30L was preincubated with supernatant, equivalent to 1.5 10 cells, from BSC-I cells mock infected or infected with EV strains Hamp- stead or Naval, and then incubated with 5ng vCD30-Fc in a protein A–coated FlashPlate. The binding of I-CD30L was determined in a Pack- ard Topcount microplate counter. (b) BSC-I cells were mock infected or infected with EV strain Hampstead in the absence or presence of AraC. 6 Figure 4. Affinity of vCD30 for mouse CD30L. Half a nanogram of (Early) or 24 h after infection (Late), supernatants were harvested and an aliquot, equivalent to 5 10 cells, was tested for its ability to block the vCD30 was incubated in the protein A–coated FlashPlates with different 125 125 125 amounts of mouse I-CD30L, and the radioactivity bound was deter- binding of 200 pM I-CD30L to 5 ng vCD30-Fc. Bound I-CD30L mined in a Packard Topcount microplate counter. Data were converted to was determined as described above. The background radioactivity corre- the Scatchard coordinate system and analyzed with the LIGAND software. sponding to the binding medium has been subtracted. Specific I-CD30L binding of duplicate samples (mean SD) is shown. Specific I-CD30L binding of duplicate samples (mean SD) is shown. 834 Ectromelia Virus–encoded CD30 Homologue Biological Activity of vCD30. The ability of vCD30 to block CD30L binding to cell surface receptors was investi- gated. First, we screened by flow cytometry human and mouse cell lines for CD30 expression using soluble recom- binant mouse CD30L that cross reacts with human recep- tors (3). Staining profiles indicated that human monocyte K562 cells expressed high levels of CD30 (Fig. 6 a). The addition of a 25-fold excess of vCD30-Fc, but not of IgG1, efficiently blocked CD30L binding to K562 cells (Fig. 6 a). As little as a 10-fold excess of vCD30-Fc was sufficient to interfere with the binding of CD30L to its cellular recep- tors (not depicted). In addition to blocking the CD30–CD30L interaction, vCD30 had the potential to bind CD30L expressed at the cell surface and trigger intracellular signals. Neutrophils constitutively express CD30L, but not CD30, and rap- idly produce IL-8 when stimulated by CD30 (4). There- fore, we analyzed the production of IL-8 by freshly iso- lated human neutrophils in response to immobilized human, mouse, or vCD30-Fc proteins or IgG1. As shown in Fig. 6 b, the viral homologue induced a response com- parable to that of the mammalian receptors, whereas hu- man IgG1 had no effect on IL-8 production. The latter indicated that in this experimental system, CD30– CD30L interactions leading to IL-8 production do not occur in the absence of vCD30. Therefore, the effect ob- served is due to signaling through CD30L after interac- tion with vCD30 and not to the inhibition of CD30– CD30L interactions by vCD30 acting as a soluble decoy receptor. Finally, we investigated a possible role of vCD30, and indirectly of the CD30–CD30L pair, in the development Figure 6. Biological activities of vCD30. (a) Blockade of the binding of T cell responses in vitro. We determined the influence of soluble CD30L to CD30 expressed at the cell membrane. K562 cells of vCD30 on the activation of IFN-–producing cells in an were incubated with a 6 histidine-tagged CD30L and binding was de- MLR. As shown in Fig. 6 c, vCD30 almost completely ab- tected with the mouse Ab specific for the histidine tag on the CD30L rogated the production of IFN- by splenocytes from molecule followed by an FITC-goat anti–mouse Ab (filled histogram). This binding was competed by a 25-fold excess of vCD30-Fc (right, solid BALB/c mice exposed to L929 cells of different haplotype. line), but not by the same excess of IgG1 (right, dashed line). The profile The presence of vCD30 in the priming phase was sufficient of unstained cells (left, dashed line) and of cells stained in the absence of to cause this effect. This result suggested an important role CD30L (left, solid line) is also shown. (b) IL-8 production by neutrophils of the CD30–CD30L interaction for the establishment of T induced by human, mouse, or vCD30-Fc. Freshly isolated neutrophils were incubated for 5 h in the presence of the indicated immobilized Fc cell responses, particularly at early stages of activation. fusion proteins or IgG1. Supernatants were harvested and assayed for IL-8 Moreover, the viral protein might be targeting this interac- production by ELISA. IL-8 secretion of duplicate samples (mean SD) is tion to protect the virus against host T cell responses. The shown. The figure shows one representative experiment of three done presence of IgG1 had no effect on the development of the with neutrophils from different donors. (c) Role of vCD30 in in vitro T T cell response (not depicted). cell responses. Freshly isolated splenocytes from BALB/c mice were mixed with irradiated L929 cells in the absence or presence of vCD30. 5 d Biological Role of vCD30 In Vivo. To address the po- later, IL-2 was added and the incubation was held for an additional 2 d. tential immunomodulatory activity of vCD30 in vivo, the After this priming phase, the viable cells were harvested and incubated effects of vCD30-Fc treatment on inflammation in a with irradiated L929 for 20 h in the presence of vCD30. Cell activation pulmonary granuloma model were investigated. In this was measured by their ability to produce IFN-. The number of cells se- creting IFN was assayed by ELISPOT. The result of duplicate samples model, type 1 and 2 cytokine–mediated granulomas are in- (mean SD) is shown. vCD30/, vCD30 added only in the priming duced in mouse lungs by bead-immobilized mycobacterial phase; vCD30/vCD30, vCD30 added in both phases. PPD or SEA, respectively. Mice were treated with vCD30-Fc or control IgG1 during Ag sensitization and elicitation (four treatments over 18 d), or only during elic- treated mice (Fig. 7, a and b). This effect was observed itation (two treatments over 4 d). The administration of when vCD30 was administered throughout both the sensi- vCD30-Fc to mice caused a significant impaired type 1 cy- tization and elicitation response or only during elicitation tokine–mediated inflammatory response with pulmonary (Fig. 7 b). In contrast, type 2 cytokine–mediated pulmo- granuloma size reduced 80% compared with IgG1- nary inflammation was not modified by vCD30 adminis- 835 Saraiva et al. Figure 7. Role of vCD30 during in vivo type 1 or 2 cytokine–mediated pulmonary inflamma- tion. BALB/c mice were type 1 or 2 sensitized and pulmonary granulomas were elicited as de- scribed in Materials and Methods. Mice were treated with vCD30-Fc or control IgG1 through- out the sensitization (S) and elicitation (E; four treatments) or only during elicitation (two treat- ments). (a) Photomicrographs of hematoxylin and eosin–stained sections of representative type 1 (lymphocytes-, monocytes-, and neutrophils- rich) or 2 (eosinophil-dominated) pulmonary granulomas surrounding Ag-coated beads in mice treated with vCD30-Fc or IgG1. (b) Quantifica- tion of the granuloma volume (mean SD) from four or five mice per group. Significantly smaller type 1 granulomas (*, P  0.01; Student’s t test) were found in mice treated two or four times with vCD30 compared with IgG1-treated mice. (c) vCD30-Fc treatment reduced lung tissue IFN- and IL-12 in type 1–sensitized mice. The levels of IFN-, IL-12, IL-4, and IL-5 in lung tissue homogenates were tested in ELISA and are expressed as nanogram cytokine per milligram lung protein. (d) Intracellular detection of alter- ations in the frequencies of IFN-– and IL-4– producing CD4 and CD8 T cells in type 1– or 2–sensitized mice. Data in quadrants represent the percentage of positively stained cells. All data are representative from two separate experiments (n four or five mice per group). Cytokine and FACS data presented are from mice treated four times with vCD30-Fc. Comparable data was ob- tained in two separate experiments when mice were only treated two times. tration, as vCD30-Fc–treated mice had comparable or 1 cytokine levels occurred when vCD30 was administered marginally larger granulomas than obtained in control ani- throughout the experiment (Fig. 7 c) or only during elici- mals (Fig. 7, a and b). The composition of the cellular in- tation (not depicted). These results demonstrate that the filtrate surrounding the type 1 or 2 granulomas was not al- formation of type 1, but not type 2, cytokine–mediated tered by vCD30 treatment (not depicted). pulmonary inflammation was impaired by vCD30 in vivo. The effect of vCD30 on pulmonary cytokine responses To address if the reduced IFN- production in vCD30- was measured in lung tissue homogenates. Levels of type 1 treated mice was associated with alterations in the produc- or 2 cytokines in the lungs of control IgG1–treated mice tion of type 1 (IFN-) or type 2 (IL-4) cytokines by T were biased in the respective type 1– (elevated IFN- and cells, we performed intracellular cytokine staining on IL-12) or 2– (IL-4 and IL-5) mediated pulmonary inflam- CD4 or CD8 T cells from spleens. Cells from the spleens matory responses (Fig. 7 c). vCD30 treatment of mice with of type 1 granuloma–sensitized mice treated with vCD30 type 1 granulomas caused a substantial reduction in lung had two- to threefold lower frequencies of both Th1 and IFN- and albeit to a lesser degree, IL-12 levels (Fig. 7 c) Tc1 cells compared with control IgG1–treated mice (Fig. 7 with no alterations in IL-4 and IL-5 levels. Cytokine levels d). Similarly, vCD30 treatment also reduced the numbers in mice with type 2 granulomas were not altered by of IFN-–secreting T cells in type 2 granuloma–sensitized vCD30, although IFN- and IL-12 levels were lower in mice (Fig. 7 d). These results support a preferential effect of vCD30-treated mice (Fig. 7 c). Changes in pulmonary type vCD30 on type 1 T cells. 836 Ectromelia Virus–encoded CD30 Homologue In addition, the role of vCD30 in the development of transmembrane or cytoplasmic domains. The extracellular Th1-like responses was controlled with a truncated version domain of human CD30 consists of a duplicate structure of of another EV vTNFR. A recombinant inactive form of three CRDs, whereas mouse CD30 only has the first clus- CrmD consisting of the first two CRDs (CrmD-CRD1,2- ter and the viral counterpart have the second and third Fc) and unable to protect cells from TNF–induced necrosis CRDs. This demonstrates direct involvement of the sec- was produced in the baculovirus system and purified ac- ond and third CRDs in ligand binding. Moreover, the cording to the same protocols used for vCD30-Fc (unpub- conservation of some motifs in the three molecules may lished data). Mice exposed to mycobacterium or schisto- point at specific residues involved in CD30L binding. soma antigens as described above were treated with Both mouse and human CD30 can be proteolytically vCD30-Fc, CrmD-CRD1,2-Fc, or PBS, and the levels of cleaved by a zinc metalloprotease to produce a soluble cytokines in the lung were determined. Table I shows form (sCD30) that is larger than vCD30. The viral protein that vCD30-Fc, but not CrmD-CRD1,2-Fc, specifically appears to retain the minimal structure necessary for effi- down-regulated Th1 cytokine production, confirming the cient CD30L binding activity. The higher similarity of previous result and demonstrating that the role attributed vCD30 to mouse CD30 is consistent with mice being the to vCD30-Fc was specific. natural host for EV. vCD30 is the fifth member of the TNFR superfamily identified in poxviruses and the only one that does not bind Discussion TNF, which suggests a distinct role during viral infection. The sequence similarity between vCD30 and the other We report the identification and characterization of a homologue of mammalian CD30, designated vCD30 and vTNFRs (CrmB, CrmC, CrmD, and CrmE) is confined to the CRDs. CrmB and CrmD show an extended COOH- encoded by EV. vCD30 is a new member of the TNFR superfamily that binds CD30L. Comparative studies in 10 terminal region with no sequence similarities in databases, which is not required for TNF binding (unpublished data). other EV isolates showed that all of them encode an active vCD30 (unpublished data). The complete genome se- Interestingly, the ORF downstream vCD30 in the EV strain Naval genome has similarity to the COOH-terminal quence of EV has been recently determined (available at http://www.sanger.ac.uk) and shows that the vCD30 gene extension found in some vTNFRs and no intergenic re- gion is found (unpublished data). We propose that the an- described here is the only CD30 homologue encoded by the virus (unpublished data). Thus it is most likely that the cestral vCD30 gene may have contained the COOH-ter- minal extension that has been lost in the gene presently gene we have characterized is responsible for the activity expressed in EV infections. The presence of an active found in EV. vCD30 is a 12-kD protein secreted at late times during vCD30 seems restricted to a few poxviruses. An intact vCD30 gene has been identified in EV isolates and CPV EV infection that binds CD30L with high affinity and in- hibits its binding to cell surface CD30. This function of GRI-90 (30), whereas it is not found in VV strain WR (this paper) or in 18 other poxviruses including 3 strains of vCD30 is similar to that of the poxvirus soluble cytokine decoy receptors, including vTNFRs. However, vCD30 has VaV, whose complete genome has been sequenced and is available in public databases. Maybe different immuno- an additional unique property. It induces reverse signaling in cells expressing CD30L. Therefore, vCD30 may not be modulatory genes compensate for the lack of vCD30 in other poxviruses. classified as a genuine decoy receptor and we propose that the mechanism of action of vCD30 is totally different from Amino acid sequence similarity of vCD30 with the hu- man and mouse counterparts is confined to a region of the that adopted by vTNFRs. Because secreted CD30L has not been identified, vCD30 is acting at the cell surface, both extracellular domain because the viral protein lacks the Table I. Modulation of Cytokine Production by vCD30 in a Mouse Model of Inflammation Th1 response Th2 response vCD30-Fc CrmD-CRD1,2-Fc PBS vCD30-Fc CrmD-CRD1,2-Fc PBS IFN- 0.33 0.2 3.31 0.57 3.64 0.47 0.29 0.16 0.94 0.19 0.81 0.26 IL-12 0.82 0.28 3.35 0.6 3.20 0.73 0.17 0.08 0.39 0.15 0.43 0.16 IL-4 0.34 0.09 0.05 0.04 0.09 0.05 3.12 0.49 3.81 0.44 2.93 0.63 IL-5 0.88 0.2 0.38 0.12 0.32 0.21 5.93 0.89 5.72 0.67 5.75 0.98 Cytokine responses in the lungs of mice exposed to mycobacterium (Th1 response) or schistosoma antigens (Th2 response) and treated on days 14 and 16 with 10 g vCD30-Fc, CrmD-CRD1,2-Fc, or PBS. Levels of IFN-, IL-12, IL-4, and IL-5 in lung tissue homogenates as tested by ELISA and expressed as nanogram cytokine per milligram lung protein (mean SD) from five mice per group. 837 Saraiva et al. mimicking signal transduction mediated by CD30 and in- other example of proteins from virulent viruses that have hibiting the effect of CD30L in cells expressing CD30. We promise as therapeutic reagents. also demonstrate that vCD30 is a soluble molecule with no We thank Sebastian Mas and Begonia Aguado for the mass spec- membrane-associated binding activity, as it has been de- trometry analysis of vCD30-Fc. scribed for TNF binding activity in VV Lister (29). This work was funded by the Wellcome Trust (grants 051087/Z/ The expression of CD30 is associated with the activated 97/Z and 057381). M. Saraiva is funded by Fundacao para a Ciencia status of T cells. In vitro, it has been mainly associated with e Tecnologia-Praxis XXI (grant BD-18081/98). P.G. Fallon was a Th2/Th0 phenotype (38), although in vivo studies sug- supported by a Wellcome Trust Career Development Fellowship gest that the relationship between CD30 T cells and Th1 and A. Alcami is a Wellcome Trust Senior Research Fellow. or Th2 profiles is very complex. Recent studies support a Submitted: 26 February 2002 novel regulatory mechanism for CD30 in Th1 polarized Revised: 21 June 2002 responses such as rheumatoid arthritis (39). We show that Accepted: 8 August 2002 the blockade of the binding of CD30L to CD30 by the vi- ral protein and/or the activation of CD30L by vCD30 is responsible for the inhibition of IFN- production by acti- References vated splenocytes in MLR. The potent in vivo inhibition 1. Schwab, U., H. Stein, J. Gerdes, H. Lemke, H. Kirchner, M. of pulmonary granuloma formation by vCD30 in type 1, Schaadt, and V. Diehl. 1982. Production of a monoclonal but not 2, cytokine–sensitized mice supports a preferential antibody specific for Hodgkin and Sternberg Reed cells of role for CD30–CD30L in type 1 cytokine–mediated re- Hodgkin’s disease and a subset of normal lymphoid cells. Na- sponses. Consistent with our in vitro data, a potential di- ture. 299:65–67. rect affect of vCD30 on type 1 T cells was shown by the 2. Horie, R., and T. Watanabe. 1998. CD30: expression and diminished frequencies of IFN-–producing T cells de- function in health and disease. Semin. Immunol. 10:457–470. tected in vCD30-treated mice. These data support increas- 3. Smith, C.A., H.J. Gruss, T. Davis, D. Anderson, T. Farrah, ing evidence for a key role of CD30 in type 1 responses in E. Baker, G.R. Sutherland, C.I. Brannan, N.G. Copeland, vivo. Interestingly, vCD30 treatment was effective in sup- N.A. Jenkins, et al. 1993. CD30 antigen, a marker for pressing type 1 cytokine–mediated inflammation when ad- Hodgkin’s lymphoma, is a receptor whose ligand defines an ministered either throughout the Ag sensitization and elic- emerging family of cytokines with homology to TNF. Cell. 73:1349–1360. itation phase or only during elicitation. In the context of 4. Wiley, S.R., R.G. Goodwin, and C.A. Smith. 1996. Reverse viral infection, this is probably compromising a protective signaling via CD30 ligand. J. Immunol. 157:3635–3639. antiviral immune response, which is known to be Th1-like 5. Amakawa, R., A. Hakem, T.M. Kundig, T. Matsuyama, J.J. T cell mediated. Simard, E. Timms, A. Wakeham, H.W. Mittruecker, H. CD30 is also known to be expressed in activated B cells Griesser, H. Takimoto, et al. 1996. Impaired negative selec- (2). CD30L was found to be a potent mediator of mouse B tion of T cells in Hodgkin’s disease antigen CD30-deficient cell growth and differentiation in vitro, although different mice. Cell. 84:551–562. results were found with human B cells (2). Therefore, 6. Chiarle, R., A. Podda, G. Prolla, E.R. Podack, G.J. Thor- vCD30 may also interfere with B cell responses to the virus. beecke, and G. Inggirami. 1999. CD30 overexpression en- Finally, vCD30 may also modulate signaling between B and hances negative selection in the thymus and mediates pro- grammed cell death via Bcl-2-sensitive pathway. J. Immunol. T cells, a process in which CD30 has been implicated. 163:194–205. Viral mechanisms of immune modulation may point at 7. Muta, H., L.H. Boise, L. Fang, and E.R. Podack. 2000. those host molecules that contribute to antiviral responses. CD30 signals integrate expression of cytotoxic effector mole- The remarkable finding of a virus-encoded CD30 homo- cules, lymphocyte trafficking signals, and signals for prolifera- logue uncovers a potential role of the CD30–CD30L sys- tion and apoptosis. J. Immunol. 165:5105–5111. tem in antiviral immune responses. Increased levels of 8. Gruss, H.J., A. Pinto, J. Duyster, S. Poppema, and F. Herr- sCD30 have been reported in pathological conditions and mann. 1997. Hodgkin’s disease: a tumor with disturbed im- after infection with HIV, EBV, hepatitis virus B and C, munological pathways. Immunol. Today. 18:156–163. measles virus, and varicella-zoster virus, but the biological 9. Moss, B. 1996. Poxviridae: the virus and their replication. In significance is not known. Viral proteins have been opti- Virology. B.N. Fields, R.M. Chanock, D.M. Knipe, and P.M. Howley, editors. Lippincott-Raven Publishers, Phila- mized during virus–host coevolution to become potent in- delphia. 2637–2671. hibitors of host immune responses. Maybe the biological 10. Fenner, F., and R.M. Buller. 1997. Mousepox. In Viral properties of vCD30 differ from those of the host sCD30. Pathogenesis. N. Nathanson, F. Gonzalez-Scarano, R. Characterization of viral immunomodulatory proteins Ahmed, H.L. Robinson, D.E. Griffin, K.V. Holmes, and may also shed light into the function of the host counter- F.A. Murphy, editors. Lippincott-Raven Publishers, Phila- parts. By using vCD30, we demonstrate a role of CD30– delphia. 535–553. CD30L interactions in the generation of Th1 inflammatory 11. Alcami, A., and U.H. Koszinowski. 2000. Viral mechanisms responses. vCD30 may also provide alternative strategies to of immune evasion. Immunol. Today. 21:447–455. effectively block the activity of CD30 in vivo, which might 12. McFadden, G., and P.M. Murphy. 2000. Host-related im- be applied to modulate an overreactive immune response munomodulators encoded by poxviruses and herpesviruses. in a number of human disease conditions (8). vCD30 is an- Curr. Opin. Microbiol. 3:371–378. 838 Ectromelia Virus–encoded CD30 Homologue 13. Tortorella, D., B.E. Gewurz, M.H. Furman, D.J. Schust, and type 2 cytokine-producing mouse CD4 and CD8 T cells in H.L. Ploegh. 2000. Viral subversion of the immune system. acute Schistosoma mansoni infection. Eur. J. Immunol. 28: Annu. Rev. Immunol. 18:861–926. 1408–1416. 14. Loparev, V.N., J.M. Parsons, J.C. Knight, J.F. Panus, C.A. 28. Esposito, J., R. Condit, and J. Obijeski. 1981. The prepara- Ray, R.M. Buller, D.J. Pickup, and J.J. Esposito. 1998. A tion of orthopoxvirus DNA. J. Virol. Methods. 2:175–179. third distinct tumor necrosis factor receptor of orthopoxvi- 29. Alcami, A., A. Khanna, N.L. Paul, and G.L. Smith. 1999. ruses. Proc. Natl. Acad. Sci. USA. 95:3786–3791. Vaccinia virus strains Lister, USSR and Evans express soluble 15. Smith, V.P., and A. Alcami. 2000. Expression of secreted cy- and cell-surface tumour necrosis factor receptors. J. Gen. Vi- tokine and chemokine inhibitors by ectromelia virus. J. Virol. rol. 80:949–959. 74:8460–8471. 30. Shchelkunov, S.N., P.F. Safronov, A.V. Totmenin, N.A. 16. Mossman, K., C. Upton, R.M. Buller, and G. McFadden. Petrov, O.I. Ryazankina, V.V. Gutorov, and G.J. Kotwal. 1995. Species specificity of ectromelia virus and vaccinia vi- 1998. The genomic sequence analysis of the left and right rus interferon- binding proteins. Virology. 208:762–769. species-specific terminal region of a cowpox virus strain re- 17. Smith, V.P., and A. Alcami. 2002. Inhibition of interferons veals unique sequences and a cluster of intact ORFs for im- by ectromelia virus. J. Virol. 76:1124–1134. munomodulatory and host range proteins. Virology. 243:432– 18. Colamonici, O.R., P. Domanski, S.M. Sweitzer, A. Larner, 460. and R.M. Buller. 1995. Vaccinia virus B18R gene encodes a 31. Genetics Computer Group. 1994. Program Manual for the type I interferon-binding protein that blocks interferon  Wisconsin Package. transmembrane signaling. J. Biol. Chem. 270:15974–15978. 32. Davison, A.J., and B. Moss. 1990. New vaccinia virus recom- 19. Smith, V.P., N.A. Bryant, and A. Alcami. 2000. Ectromelia, bination plasmids incorporating a synthetic late promoter for vaccinia and cowpox viruses encode secreted interleukin-18- high level expression of foreign proteins. Nucleic Acids Res. binding proteins. J. Gen. Virol. 81:1223–1230. 18:4285–4286. 20. Born, T.L., L.A. Morrison, D.J. Esteban, T. VandenBos, 33. Alcami, A., J.A. Symons, P.D. Collins, T.J. Williams, and L.G. Thebeau, N. Chen, M.K. Spriggs, J.E. Sims, and R.M. G.L. Smith. 1998. Blockade of chemokine activity by a solu- Buller. 2000. A poxvirus protein that binds to and inactivates ble chemokine binding protein from vaccinia virus. J. Immu- IL-18, and inhibits NK cell response. J. Immunol. 164:3246– nol. 160:624–633. 3254. 34. Markwell, M.A., and C.F. Fox. 1978. Surface-specific iodi- 21. Graham, K.A., A.S. Lalani, J.L. Macen, T.L. Ness, M. Barry, nation of membrane proteins of viruses and eucaryotic cells L.Y. Liu, A. Lucas, I. Clark-Lewis, R.W. Moyer, and G. using 1,3,4,5-tetrachloro-3,6-diphenylglycouril. Biochem- McFadden. 1997. The T1/35kDa family of poxvirus-secreted istry. 17:4807–4817. proteins bind chemokines and modulate leukocyte influx into 35. Munson, P.J., and D. Rodbard. 1980. LIGAND: a versatile virus-infected tissues. Virology. 229:12–24. computerized approach for characterization of ligand-binding 22. Turner, S.J., J. Silke, B. Kenshole, and J. Ruby. 2000. Char- systems. Anal. Biochem. 107:220–239. acterization of the ectromelia virus serpin, SPI-2. J. Gen. Vi- 36. Chensue, S.W., K. Warmington, J.H. Ruth, N. Lukacs, and rol. 81:2425–2430. S.L. Kunkel. 1997. Mycobacterial and schistosomal antigen- 23. Brick, D.J., R.D. Burke, A.A. Minkley, and C. Upton. 2000. elicited granuloma formation in IFN- and IL-4 knockout Ectromelia virus virulence factor p28 acts upstream caspase-3 mice: analysis of local and regional cytokine and chemokine in response to UV light-induced apoptosis. J. Gen. Virol. 81: networks. J. Immunol. 159:3565–3573. (See published erra- 1087–1097. tum 162:3106.) 24. Hu, F.Q., C.A. Smith, and D.J. Pickup. 1994. Cowpox virus 37. Fallon, P.G., and D.W. Dunne. 1999. Tolerization of mice contains two copies of an early gene encoding a soluble se- to Schistosoma mansoni egg antigens causes elevated type 1 creted form of the type II TNF receptor. Virology. 204:343– and diminished type 2 cytokine responses and increased mor- 356. tality in acute infection. J. Immunol. 162:4122–4132. 25. Smith, C.A., F.Q. Hu, T.D. Smith, C.L. Richards, P. Smo- 38. Del Prete, G., M. De Carli, F. Almerigogna, C.K. Daniel, lak, R.G. Goodwin, and D.J. Pickup. 1996. Cowpox virus M.M. D’Elios, G. Zancuoghi, F. Vinante, G. Pizzolo, and S. genome encodes a second soluble homologue of cellular Romagnani. 1995. Preferential expression of CD30 by hu- TNF receptors, distinct from CrmB, that binds TNF but not man CD4 T cells producing Th2-type cytokines. FASEB J. LT . Virology. 223:132–147. 9:81–86. 26. Saraiva, M., and A. Alcami. 2001. CrmE, a novel soluble tu- 39. Gerli, R., C. Lunardi, F. Vinante, O. Bistoni, G. Pizzolo, and mor necrosis factor receptor encoded by poxviruses. J. Virol. C. Pitzalis. 2001. Role of CD30 T cells in rheumatoid ar- 75:226–233. thritis: a counter-regulatory paradigm for Th1-driven dis- 27. Fallon, P.G., P. Smith, and D.W. Dunne. 1998. Type 1 and eases. Trends Immunol. 22:72–77. 839 Saraiva et al.

Journal

The Journal of Experimental MedicinePubmed Central

Published: Sep 16, 2002

There are no references for this article.