Proliferation and Apoptosis Trafficking Signals, and Signals for Cytotoxic

CD30 Signals Integrate Expression of Cytotoxic Effector Molecules, Lymphocyte Trafﬁcking Signals, and Signals for Proliferation and Apoptosis1

Hiromi Muta, Lawrence H. Boise, Lei Fang, and Eckhard R. Podack2

Although CD30 has long been recognized as an important marker on many lymphomas of diverse origin and as activation molecule on B cells and T cells, its primary function has remained obscure. We now report that CD30 signals may serve to inhibit effector cell activity by integrating gene expression changes of several pathways important for cytotoxic NK and T cell effector function. In the large granular lymphoma line YT, CD30 signals down-regulate the expression of cytotoxic effector molecules, Fas ligand, perforin, granzyme B, and abrogate cytotoxicity. c-myc, a regulator of proliferation and an upstream regulator of Fas ligand expression, is completely suppressed by CD30. Furthermore, CD30 signals strongly induce CCR7, suggesting a role for CD30 signals in the homing of lymphocytes to lymph nodes. The up-regulation of Fas, death receptor 3, and TNF-related apoptosis- inducing ligand by CD30 indicates an increase in susceptibility to apoptotic signals whereas up-regulation of TNFR-associated factor 1 and cellular inhibitor of apoptosis 2 protect cells from certain types of apoptosis. Using gene microarrays, 750 gene products were induced and 90 gene products were suppressed >2-fold by CD30 signals. Signals emanating from CD30 use both TNFR-associated factor 2-dependent and -independent pathways. The integration of CD30 signals in a lymphoma line suggests that CD30 can down-modulate lymphocyte effector function and proliferation while directing the cells to lymph nodes and increasing their susceptibility to certain apoptotic signals. These studies may provide a molecular mechanism for the recently observed CD30-mediated suppression of CTL activity in vivo in a diabetes model. The Journal of Immunology, 2000, 165: 5105–5111.

he CD30 and CD30 ligand3 (CD30L) have been widely mediated by CD30 (17, 18). TRAF2 is regulated by TRAF-inter- studied as surface markers on a diverse array of lympho- acting protein and by susceptibility to proteolysis (19, 20). T mas and myelomas (1, 2). CD30 can induce proliferation The primary biological function of CD30 has been difficult to or apoptosis in different lymphomas (3, 4); its expression is com- establish due to the pleiotropic nature of CD30 signals and the monly associated with a TH2 phenotype (5–8). IL-4 and CD28 transience of CD30 expression in normal lymphocytes. Mice lack- up-regulate and IFN-␥ down-regulates CD30 expression on acti- ing a functional CD30 gene product have a defect in negative vated murine T cells; however, CD30 expression does not define selection of thymocytes (21), whereas CD30-transgenic mice ex- the TH2 subset (9, 10). CD30 is expressed transiently on activated pressing CD30 in the thymus have enhanced activity for thymic T cells, which is more pronounced in the CD8 subpopulation (11) negative selection (22). In accordance with this function in nega- and CD30 expressing activated memory cells produce IFN-␥ and tive selection in the thymus, CD30L is expressed in the Hassal IL-5 (12). bodies of the thymus (23). Recently, a function of CD30 in mod- CD30 belongs to the TNFR family of proteins and CD30L to the ulating cytotoxic responses was reported by using TCR-transgenic, TNF family of ligands (13, 14). The intracellular sequence of autoreactive, CD30-deficient CD8 T cells in a diabetes model. CD30 contains TNFR-associated factor (TRAF) domains, allow- CD30-deficient T cells were ϳ6000 times more active in causing ing it to interact with adaptor proteins of the TRAF family: autoimmune diabetes than their CD30-sufficient counterparts (24, TRAF1, TRAF2, and TRAF3 (15, 16). TRAF2 is thought to be 25). These data suggested that CD30 is an important negative reg- responsible for NF-␬B activation and for the antiapoptotic effect ulator for cytotoxic lymphocytes in vivo. We have reported previously that signals emanating from CD30 down-regulate the cytotoxic activity of the large granular lymphocyte (LGL) lymphoma line, YT, for B7-expressing target cells (26). CD30 signals down- Department of Microbiology and Immunology, University of Miami School of Med- regulated CD28, which is required for recognition and triggering icine, Miami, FL 33136 of cytotoxicity by YT (27). Received for publication June 12, 2000. Accepted for publication August 2, 2000. Even though the signaling pathways of CD30 and the transcrip- The costs of publication of this article were defrayed in part by the payment of page tion factors activated by CD30 signals are well characterized, little charges. This article must therefore be hereby marked advertisement in accordance is known about the genes targeted by CD30 for induction or re- with 18 U.S.C. Section 1734 solely to indicate this fact. pression. In part this is due to the difficulty of obtaining sufficient 1 This work was supported by National Institutes of Health Grants CA 39201, CA 57904, and CA80228, American Cancer Society Grant F99UM1, and Department of numbers of homogeneous lymphocyte populations expressing Defense Grant DAMDS17-8-1-8317. CD30 that would be suitable for analysis. The YT lymphoma, 2 Address correspondence and reprint requests to Dr. Eckhard R. Podack, P.O. Box expressing high levels of CD30 constitutively and endowed with 016960 (R138), Miami, FL 33101. E-mail address: [email protected] cytotoxic activity, therefore, is an ideal model to further define 3 Abbreviations used in this paper: CD30L, CD30 ligand; TRAF, TNFR-associated genes regulated by CD30 signals. As reported below, we find that factor; TRAIL, TNF-related apoptosis-inducing ligand; EST, expressed sequence tag; DR3, death receptor 3; DN, dominant negative; cIAP2, cellular inhibitor of apoptosis CD30 down-regulates the expression of major effector molecules 2; LGL, large granular lymphocyte; Fas-L, Fas ligand. for cytotoxicity and represses genes required for proliferation. In

addition, CD30 up-regulates molecules critical for lymphocyte Gene microarray analysis homing and increases the expression of proapototic and antiapop- Human UniGem Microarrays (Genome Systems, St. Louis, MO) contain- totic molecules. The regulation of the expression of multiple gene ing cDNAs of 4000 named genes and an equal number of expressed se- products by CD30 can be interpreted as the synergistic suppression quence tags (ESTs) were used to analyze the induction or suppression of of cytotoxic cells by terminating cytotoxicity, diminishing prolif- mRNAs by CD30 signals. mRNA was extracted from untreated and C10- eration, redirecting lymphocytes toward lymph nodes rather than treated (17 h) YT cells according to the manufacturer’s instructions. After conversion to cDNA and differential fluorescent labeling, the fluorescent inflammatory sites, and rendering them more susceptible to certain probes of C10-treated and untreated YT cells were mixed and hybridized proapoptotic signals. to the gene microarrays. Differential binding of the two probes to the arrays is analyzed by the manufacturer and normalized according to known ref- Materials and Methods erence markers on the gene chip. A ratio of binding of one of the two probes indicates no change in gene expression due to CD30 signals. The Cells data are given as percent induction or suppression of gene products by YT cells were cultured in Iscove’s modified Dulbecco’s MEM containing CD30 signals relative to untreated YT cells. 10% heat-inactivated FBS and were treated with 5 ␮g/ml of the agonistic anti-human CD30 mAb C10 (26). Treated cells were harvested at different time points and analyzed. Results To block TRAF2 signal transmission, YT cells were stably transfected CD30 signals down-regulate Fas ligand (Fas-L) and up-regulate with an expression vector containing truncated, dominant negative, murine Fas and TNF-related apoptosis-inducing ligand (TRAIL) dominant negative (DN)-TRAF2 using the Gene Pulser (Bio-Rad, Hercules, CA). CD30 signals are transmitted through adaptor proteins of the TRAF family, resulting in the activation of transcription factors Multiprobe RNase protection assay including NF-␬B and Jun N-terminal kinase. The target genes ad- Total RNA was extracted from YT cells using the RNeasy Midi kit (Qia- dressed by CD30 signals through these and other transcription fac- gen, Chatsworth, CA). Multiprobe RNase protection assay (PharMingen, tors remain largely undefined. Because few lymphocytes express San Diego, CA) was performed with 10 ␮g of total RNA according to the CD30 under physiological conditions and CD30 expression is usu- manufacturer’s instruction. Briefly, isolated RNA was hybridized for 17 h ally transient, it is difficult to obtain information about genes reg- at 56oC with 32P-labeled multiprobe template sets and then treated with RNase. Protected RNA fragments were resolved on polyacrylamide gels ulated by CD30. Using the YT cell line (28), a LGL lymphoma and radioactive signals were analyzed by a phosphor imager (Bio-Rad). with cytotoxic activity constitutively overexpressing CD30, we Digitized signals were normalized using as reference the signals of the generated CD30 signals using an agonistic anti-CD30 Ab, C10 housekeeping genes L32 or GAPDH. The relative expression levels of the (26), and studied the regulation of several groups of genes by gene products under analysis are presented as percentage to their expression in untreated cells, which is set to 100%. CD30 signals. We have published previously that CD30 signals induced the RT-PCRs loss of YT cytotoxicity toward B7-expressing targets (26). Loss of One microgram of total RNA was used to synthesize cDNA using the cytotoxicity correlated with the down-regulation of CD28, which SuperScript Preamplification System (Life Technologies, Rockville, MD). is known to be required for YT killing of B7-expressing target For PCR, 1/200 of the volume of the first-strand cDNA sample was used cells (27). In the current study, we investigated whether cytotoxic for amplification of c-myc, CCR7, or ␤-actin. The primers used for PCR effector molecules and other gene products expressed by YT were were as follows: CCR7 (sense) 5Ј-TGGTGGCTCTCCTTGTCATT-3Ј, CCR7 (antisense) 5Ј-GCTGATGCAAAGAAGTAGGA-3Ј; c-myc (sense) regulated by CD30. YT cells contain cytolytic granules containing 5Ј-ACTGCGACGAGGAGGAGAAC-3Ј, c-myc (antisense) 5Ј-AAGC perforin and granzyme B and express functional Fas-L. To obtain CGCTCCACATACAGTC-3Ј; ␤-actin (sense), 5Ј-TCTGGCACCACAC reliable data on several groups of genes, the analyses combined the CTTCTAC-3Ј, ␤-actin (antisense) 5Ј-GAAGGAAGGCTGGAAGAGTG- results of RNA expression with protein expression and with func- Ј 3 . The PCR products were resolved on agarose gels, visualized by tional data. RNA expression data were obtained by several meth- ethidium bromide staining and quantitated using the Molecular Analyst software (Bio-Rad). ␤-actin was used as internal standard for normalization ods: RNase protection assays, RT-PCR assays, or gene microarray and relative expression levels of c-myc and CCR7 were expressed as per- assays. Protein products were assayed by Western blot, flow cy- centage of untreated cells set to 100%. tometry, or by functional assays. In all cases, CD30 signals were examined by incubating YT cells with 5 ␮g/ml of the agonistic Western blotting C10 Ab for various periods of time followed by the analysis of YT cells were lysed in 150 mM NaCl, 1.0% Nonidet P-40, 0.5% sodium RNA levels, protein levels, or function. ␮ deoxycholate, 0.1% SDS, 50 mM Tris (pH 8.0), with 1 g/ml of each To study regulation of Fas-L and a group of other gene products inhibitor, aprotinin, leupeptin, pepstatin, and 1 mM PMSF. A total of 50 ␮g of the whole-cell lysate protein was analyzed on 10% SDS-polyacrylamide involved in regulating apoptosis by CD30, RNA protection assays gels and blotted with antihuman TRAF1, TRAF2, and TRAF3 antiserum, were used (Figs. 1 and 2). The images of radiolabeled, hybridized, respectively (Santa Cruz Biotechnology, Santa Cruz, CA). Abs were de- protected RNA fragments were quantitated and normalized by ref- tected by chemiluminescence (enhanced chemiluminescence; Amersham erence to housekeeping genes (L32 and GAPDH) using the phos- Pharmacia Biotech, Piscataway, NJ). phor imager. Agonistic anti-CD30 treatment of YT completely Flow cytometry down-regulated Fas-L expression within 6 h. In contrast, TRAIL, Fas, and death receptor 3 (DR3) were up-regulated from 2- to YT cells were cultured with or without C10 for 24 h, and 1 ϫ 106 cells of each sample were stained with FITC conjugated anti-human Fas Ab or 4-fold by CD30 signals. DR3 is only weakly expressed by YT anti-human CD28 (PharMingen) and analyzed on a FACScan flow cytom- whereas TRAIL showed a somewhat higher level of expression. eter (Becton Dickinson, Franklin Lakes, NJ). The expression of other gene products such as caspase 8, Fas- associated factor, Fas-associated death domain protein, TNFR1, Chromium release (cytotoxicity) assay TNFR-associated death domain protein, and receptor-interacting P815 or P815-Fas cells were labeled with 51Cr for 2 h and washed three protein, involved in signaling for apoptosis, was not significantly times with PBS. Cytotoxicity of YT cells against P815 or P815-Fas cells affected by CD30 signals (Figs. 1 and 2). It is evident from these was examined in 6-h assays in triplicate in 96-well microtiter plates at 1:1, 3:1, 10:1, and 30:1 E:T ratios. Results are expressed as percent specific regulatory effects that CD30 signals are highly specific for selected lysis relative to HCl-lysed controls by subtracting the values of spontane- gene products rather than globally affecting transcriptional activity ous Cr release, ranging between 5 and 12%. of proapoptotic and antiapoptotic genes. The Journal of Immunology 5107

FIGURE 1. Regulation of proapoptotic and antiapoptotic genes by CD30. YT cells were incubated with 5 ␮g/ml agonistic anti-CD30 (C10) for the indicated time periods and analyzed by multiprobe RNase protection assays. A, Image of protected bands after electrophoresis. B, Quanti- tative analysis of the phosphor imager analysis of individual mRNAs after normalization expressed relative to uninduced cells. The complete set of gene products and position of the probe is shown in Fig. 2. FIGURE 2. CD30 signals are suppressed by overexpression of DN- TRAF2. A, Protein expression of DN-TRAF2 and wild-type (WT) TRAF2 in wild-type YT and bulk-transfected, uncloned (DN) YT as measured in TRAF2 transmits signals for Fas and Fas-L regulation by CD30 Western blots. B, Multiprobe RNase protection assays of wild-type and To determine whether the TRAF2-dependent pathway was used by DN-TRAF2 containing YT cells following treatment with agonistic anti- CD30 signals for the regulation of death receptor and ligand genes, CD30 (C10) for 24 h (ϩ) or without Ab (Ϫ). Left lane, The position of the YT cells were transfected with the DN mutant of TRAF2 (Fig. 2). labeled probes, connected by lines to the corresponding protected frag- DN-TRAF2 protein, after selection in G418, was expressed at a ments. L32 and GAPDH were used for normalization. C, Quantitation of high level in bulk-transfected, uncloned YT, as seen in Western normalized Fas-L and Fas expression in the presence and absence of DN- TRAF2 following CD30 signaling. blots (Fig. 2A). The presence of DN-TRAF2 blocked CD30-mediated down-regulation of Fas-L by Ͼ50%. Similarly, DN-TRAF2 inhibited the up-regulation of Fas to the same degree (Figs. 2 and 3C). CD30 signals regulating the expression of death receptors and ligands, therefore, use primarily the TRAF2 pathway. These analyses were repeated in three independent experiments with essen- tially identical results. The effects of CD30 signals were also de- ilar to Fas-L. YT cells do not express granzyme A. The analysis tected at the protein level or by functional assays. was conducted by RNase protection assay and by hybridization to Fas-L expression on untreated YT cells predicted that YT gene microarrays of fluorescently labeled cDNAs obtained from should induce apoptosis of Fas-expressing target cells. Fas-trans- untreated YT and from cells treated for 17 h with C10, respec- fected P815 (P815-Fas), but not wild-type P815 (Fig. 3A) are in- tively. The microarray analysis gives information about the abso- deed lysed by YT in 6-h assays, proving the use of Fas-L for the lute level of gene expression (Table I, right column) and relative lytic activity. Treatment of YT with anti-CD30 Ab (C10) for 16 h changes following CD30 signaling. For validation, the results of completely abolished lysis of P815-Fas by YT, indicating that the microarray analysis were confirmed for the relevant genes by Fas-L protein in addition to its mRNA is down-regulated by CD30 independent assays for RNA expression (Table I). Quantitative signals. In DN-TRAF2-transfected YT in contrast, P815-Fas lysis analysis of normalized expression of perforin and granzyme B was not inhibited by C10 treatment of YT cells (Fig. 3B). mRNA after CD30 signaling revealed a 2-fold suppression from In accord with the RNase protection assay, CD30 signals up- 100 to about 50% of the RNA levels for both gene products in both regulated Fas protein expression by YT about 3-fold as measured microchip assays and by RNase protection assay (Table I and Fig. by the increase in fluorescence intensity in the flow cytometer, 4). Cathepsin C mRNA, coding for a lysosomal protease related to whereas CD28 expression, as reported previously (26), decreased granzyme B, was also suppressed. Signals from CD30 thus down- 3-fold. Both effects were blocked by the presence of DN-TRAF2 regulate the two major cytotoxic pathways of lymphocytes; the in YT (Fig. 3, C–F). apoptotic pathway important for lymphocyte homeostasis by down-regulating Fas-L and the cytolytic pathway important for CD30 signals down-regulate perforin and granzyme B immune defense by suppressing perforin and granzyme B expres- expression sion. Since granzymes and perforin are stored in cytoplasmic gran- We next determined whether the effector molecules perforin and ules, the effect of decreased mRNA levels will have only a delayed granzyme B, expressed by YT cells, were regulated by CD30 sim- effect on the corresponding protein levels. 5108 LYMPHOCYTE EFFECTOR DOWN-REGULATION BY CD30

Since Fas-L mRNA was completely down-regulated by CD30 signals, we investigated whether c-myc expression was also suppressed (Table I and Fig. 5). Using gene microarray analysis and RT-PCR, c-myc was found to be significantly suppressed by CD30 signals, despite the finding that proliferation was not significantly affected at that time (data not shown). In timed RT-PCR analysis, CD30 signals completely down-regulated c-myc expression in YT within 24 h. Suppression of c-myc by CD30 therefore appears to be responsible for the suppression of down stream Fas-L. ␤-actin used as control in the RT-PCR analysis was not affected by CD30 signals. Gene microarray analysis gives simultaneous information about the regulation of many gene products. We found a good correlation between gene microarray assays and assays for individual gene expression by other techniques (Table I). The microarray allows one to obtain a global overview of gene regulation in addition to the assessment at the individual gene level. In a global analysis of the microarray assay, CD30 signals up-regulated many more gene products than were down-regulated by CD30. We used as thresh- old for significant regulation a 2-fold or greater change of gene expression induced by CD30 (Table II). Of the ϳ8000 gene products analyzed by microarray analysis, 750 were up-regulated Ͼ2- fold by CD30, including ESTs, whereas only 90 gene products, including ESTs, were down-regulated to a similar extent (Table II). The remaining gene products were either not expressed by YT or not significantly regulated by CD30 signals. Twenty-four of the most highly induced or suppressed gene products by CD30 signals are given in Tables III and IV. One of the up-regulated gene products that was discovered by FIGURE 3. CD30 signals change surface protein expression of Fas and microarray analysis is the chemokine receptor CCR7. Together CD28 and suppress the apoptotic function of Fas-L. A, Six-hour Cr release with other receptors, CCR7 controls lymphocyte homing (30–32). assay of P815 and Fas-transfected P815 (P815-Fas) killed by YT at the CD30 signals increased the expression of CCR7 mRNA 5.8-fold indicated E:T ratio. B, Wild-type (WT)-TRAF2 containing mock-trans- by gene chip analysis. The microarray analysis was confirmed by fected YT and DN-TRAF2-transfected YT were treated without or with 5 RT-PCR of the same samples that were also used for RT-PCR of ␮ ϩ g/ml anti-CD30 ( C10) for 24 h and analyzed for cytotoxicity on P815- c-myc and quantitated relative to actin (Fig. 5). CCR7 mRNA was Fas. Inhibition of killing of P815-Fas by C10-treated YT indicates Fas-L induced by CD30 signaling almost 3-fold within 24 h above the down-regulation which is blocked by DN-TRAF2. E:T ratio was 30:1, Cr level of unstimulated YT as measured by PCR. Expression of release assays as in A. C and D, Up-regulation of Fas by C10 measured by flow cytometry on wild-type YT and blockade by DN-TRAF2 following CCR7 during the next 72 h slowly decreased (Fig. 5). According stimulation of CD30. E and F, Down-regulation of CD28 by C10 and its to microarray analysis, CCR5 and CCR6 are also expressed by YT blockade in the presence of DN-TRAF2. cells but not affected by CD30 signals, whereas CCR1 and CCR2 and CXCR1 are not expressed.

CD30 signals suppress expression of c-myc and CD30 signals induce TRAF1 and cellular inhibitor of apoptosis up-regulate CCR7 2 (cIAP2) in a TRAF2-independent pathway myc Expression correlates with cell proliferation and is known to TRAF proteins are important signal-transducing adaptors for regulate Fas-L expression in activated T cells via TGF-␤ (29). members of the TNFR family. TRAF1 is up-regulated by TNFR

Table I. mRNA expression levels in YT in the presence of CD30 signalsa

CD30-Induced (suppressed) mRNA Levels Relative to Untreated YT Control

RNAse protection RT-PCR Microarray Microarray, untreated control Gene Product (% of control) (% of control) (% of control) (arbitrary units)

Perforin 44 ND 50 2460 Granzyme B 58 ND 50 4237 Cathepsin C ND ND 43 1692 c-myc ND 0 52 2043 CCR7 ND 300 580 316

a Analysis of expression of selected gene products by microarray analysis, RT-PCR, and RNase protection assay. YT cells were untreated or treated for 17 h with the agonistic anti-CD30 Ab C10 and used for RNA isolation. RNA was analyzed directly by RNase protection assay or converted to cDNA for analysis by RT-PCR or microarray analysis. RT-PCR data are normalized with reference to ␤-actin and expressed as percentage of the control (untreated YT). Microarray analysis was done by differ- entially labeling the cDNAs from untreated and C10-treated YT and hybridization to 8000 cDNAs printed on gene chips. Results are expressed as percentage of untreated control YT cells. The arbitrary units in the untreated control column correlate with absolute mRNA expression levels of untreated YT; the values represent the relative levels of untreated YT cDNA. The Journal of Immunology 5109

Table II. Effect of CD30 signals on the expression of multiple genes microarrayed on gene chips for analysis

Effect of CD30 Signals No. of Gene Products

Up-regulated Ͼ2-folda 750 Up-regulated 1.4–2-fold 1600 Not signiﬁcantly regulated 2560 Down-regulated 1.4–2-fold 510 Down-regulated Ͼ2-fold 90 Genes not expressed by YT 3000

a Fold regulation refers to the difference to the uninduced state (without anti- CD30).

categories: cytotoxic and proapoptotic activity, proliferation, lymphocyte traffic, and susceptibility to apoptosis. The concerted mod- ulation of gene expression by CD30 signals as described here has FIGURE 4. Suppression by C10 of perforin and granzyme B mRNA the potential to down-regulate cytolytic lymphocyte effector func- expression in YT and in DN-TRAF2-transfected YT upon CD30 signaling tion by several synergistic pathways. Although the data were ob- (5 ␮g C10, 24 h). Upper panel, Analysis by RNase protection assays. tained with a cytotoxic lymphoma line expressing CD30, the path- Lower panels, Quantitative analysis of perforin and granzyme B expression ways and gene products regulated by CD30 may also be affected by phosphor imager analysis and of the RNA protection array and normal- in CTL and may provide a molecular mechanism for the previ- ization with GAPDH. ously reported effect of CD30 suppressing the cytotoxic activity of autoreactive, TCR-transgenic CD8 cells in a diabetes model in signals. CD30 signals likewise strongly induced TRAF1 mRNA mice (24). Indeed, ongoing studies in our laboratory using CD30L- (Fig. 6, A and B) and protein (Fig. 6C) within6htoabout 12-fold deficient mice confirm the suppressive effect of CD30L-CD30 in- above the level of unstimulated cells as measured by RNase pro- teraction on CTL in vivo (H. Muta, G. Caceres, R. Levy, and E. R. tection assay. TRAF1 levels subsequently slowly decreased over Podack, unpublished observations). 72htoϳ6-fold above that in control cells. TRAF2 and TRAF4 The most direct effect of CD30 on cytotoxic activity is mediated levels did not change upon CD30 signaling; TRAF3 was not ex- via the complete down-regulation of Fas-L, eliminating the ability pressed by YT. The apoptosis inhibitor cIAP2, which can block of YT to lyse P815-Fas. Regulation of Fas-L expression by CD30 activation of caspase 3, 7, and 8, was induced to a similar extent is a novel finding and has important implications in the homeo- and with similar kinetics as TRAF1. In Western blots, TRAF1 static regulation of lymphocytes. CD30 also down-regulates per- protein was also induced in accordance with the results of RNase forin and granzyme B. Since perforin and granzyme B are storage protection assay, whereas TRAF2 protein remained unchanged proteins, the effect of transcriptional regulation on protein levels during the entire time course of CD30 signaling (Fig. 6C). TRAF3 depends on the half-life of the cytotoxic proteins and may take was not detected in Western blots (data not shown). The presence of DN-TRAF2 in YT had little effect on TRAF1 and cIAP2 induction by CD30 in YT, in contrast to its effect on Fas-L, sug- Table III. Genes and ESTs up-regulated by CD30a gesting that these genes are induced independent of the participa- tion of TRAF2 (Fig. 6D). Name GenBank ID Fold induction Homo sapiens BCE-1 AI248988 8 Discussion Chemokine receptor 7 L08176 5.8 The genes targeted by CD30 for induction or for down-regulation Liver carboxylesterase-2 U60553 5.2 as revealed in this study can be grouped into distinct functional Apoptosis inhibitor 2 AI581499 4 EST AI081356 3.6 Human mRNA for KIAA0033 gene D26067 3.2 Cadherin 14 D83542 3.2 EST AI341451 3.1 rab 8-interacting protein AF000145 3.1 Brain-heart-protocadherin AB006757 3 EST AI692207 3 EST Incyte 2598990 3 Pre-B cell leukemia transcription factor 3 X59841 2.9 EST N51730 2.9 KIAA0432 AA219672 2.9 Vinculin M33308 2.9 Aldehyde dehydrogenase M63967 2.9 Caspase 1 X65019 2.9 EST AI123482 2.9 EST AA830746 2.9 FIGURE 5. Suppression of c-myc and up-regulation of CCR7 by CD30 Phosphofructokinase M59741 2.8 signals. RT-PCR analysis of c-myc, CCR7, and ␤-actin in YT following EST AI609112 2.8 CD30 signaling after culture for the indicated time with 5 ␮g/ml C10 EST AI400515 2.8 (lower panel). The ethidium bromide-stained bands are normalized by the EST AI298083 2.8 ␤ use of -actin and quantitated relative to untreated YT cells which are set a Twenty-four of the most highly up-regulated genes, as detected by gene mi- to 100%. croarray analysis, are listed. 5110 LYMPHOCYTE EFFECTOR DOWN-REGULATION BY CD30

Table IV. Genes and ESTs down-regulated by CD30a genes like TRAF1 and cIAP2 on the other hand, which is associated with resistance to TNF-induced apoptosis and to caspase 3, 7, Name GenBank ID Fold Suppression and 8 activation, may render the cells more resistant to specific apoptotic signals (33, 34). Among CD4 cells, CCR7 is expressed EST AA679059 10 EST AI015241 4.5 exclusively on TH1 cells (29). Reentry into lymph nodes of CCR7- Guanylate-binding protein 2 M55543 3.2 positive memory cells triggered by CD30 and expression of anti- Heterogeneous nuclear apoptotic molecules, therefore, may serve to provide TH1 helper ribonuleoprotein F L28010 2.6 function for the generation of additional effector cells. Further Poly (A)-binding protein-like 1 Y00345 2.6 studies will be needed to understand the role of CD30 regulating Glutamate-cysteine ligase L35546 2.5 Human pim-2 protooncogene lymphocyte traffic and apoptosis in vivo. homolog AI634033 2.4 The importance of TRAF2 in mediating CD30 signals is well Lymphocyte cytosolic protein 2 U20158 2.4 documented (18, 20). As shown here, TRAF2 is used by CD30 for Guanylate-binding protein 1 M55542 2.4 up- or down-regulating many genes on YT including Fas-L, CD28, Ribosomal protein L7a AI929387 2.4 NK cell group 7 sequence S69115 2.4 and Fas. Interestingly, however, the up-regulation by CD30 of IFN-␥-inducible protein 16 S75433 2.4 TRAF1 and cIAP2 is not dependent on TRAF2 since the DN mu- Ornithine decarboxylase 1 M16650 2.4 tant was unable to block the induction of these genes. It has been Protein tyrosine phosphatase D11327 2.4 reported previously, and is confirmed here, that TRAF-indepen- EST W03282 2.3 dent signaling can emanate from CD30 (35). However, the mo- Human NK cell receptor (KIR) AF034773 2.3 Pyrroline-5-carboxylate lecular nature of that signaling pathway remains to be defined. synthetase X94453 2.3 The analysis of gene products modulated by CD30 signals on an Ribosomal protein, large, PO AI929696 2.3 LGL lymphoma allows the formulation of the hypothesis that Cathepsin C U79415 2.3 CD30 terminates cytotoxic lymphocyte responses by several syn- DNA segment on chromosome X expressed protein AA532479 2.2 ergistic pathways. The molecular mechanisms appear to comprise AHNAK (meaning giant in a synergistic pattern for suppression of cytotoxicity by decreasing Hebrew) nucleoprotein M80899 2.2 the expression of cytotoxic effector molecules, down-regulating Glycogenin U31525 2.2 costimulatory CD28, slowing down proliferation, potentially redi- Killer cell lectin-like receptor recting lymphocyte traffic, and increasing lymphocyte susceptibil- subfamily C, member 2 X54869 2.2 ity to apoptosis. It remains to be established that these molecular a Twenty-four of the most strongly suppressed genes, as detected by gene microarray analysis, are listed. several days to become detectable. CD30 thus suppresses the granule-dependent and granule-independent cytotoxic pathways of cytolytic lymphocytes. A similar effect of CD30 signals on cytotoxic lymphocytes infiltrating tissues in vivo, that may be mediated by local CD30L expression on monocyte/macrophages or activated T cells, is likely to substantially dampen cytotoxic activity in situ. Down-regulation of Fas-L by TGF-␤ is mediated by c-myc; the DN c-myc mutant has been shown to block TGF-␤-mediated Fas-L down-regulation (32). CD30 signals, as shown here, completely down-regulate c-myc in addition to Fas-L supporting the link between c-myc and Fas-L expression. The down-regulatory signal for c-myc and Fas-L is traveling via TRAF2, because DN-TRAF2, the DN mutant, blocks the effect of CD30 signaling on Fas-L down- regulation. The down-regulation of c-myc by CD30 may also be responsible for the diminished proliferative capacity of CD30-expressing Ag-specific CTL in tissues (24). The strong up-regulation of CCR7 by CD30 was an unexpected finding resulting from gene microarray analysis. CCR7 expression allows cells to enter lymph nodes via secondary lymphoid tissue chemokine expressed on high endothelial venules (29–31). CCR7- expressing, central memory T cells have reduced effector function, including sharply diminished cytokine production (31). Our studies show that CD30 signals can strongly up-regulate CCR7, coin- FIGURE 6. Regulation of TRAF1 and cIAP2 by CD30 signals is cident with the loss of cytotoxic function. CD30L signals delivered TRAF2 independent. A, Image of the multiprobe RNase protection assays to CD30 expressed on tissue-infiltrating cytotoxic cells, therefore, of YT treated with agonistic anti-CD30 (5 ␮g/ml) for the indicated periods have the potential to shut off cytotoxicity within the tissue and of time. B, Quantitation of the relative mRNA expression, after normal- redirect CTL to the draining lymph node. ization with GAPDH, of TRAF1 and cIAP2 (uninduced level of expression is set to 100%). C, Western blot analysis of TRAF1 and TRAF2 expression CD30 signals can up-regulate Fas and DR3, making cells po- in YT after different periods of CD30 signaling with 5 ␮g/ml C10. D, tentially more susceptible to apoptotic signals by Fas-L or the TRAF1 and cIAP2 up-regulation is not blocked by DN-TRAF2. Wild-type DR3-L. It is possible that cells that are redirected to lymph nodes (WT) and DN-TRAF2-transfected YT were treated or not treated with C10 via CCR7 expression, reencounter APCs and undergo activation- for 24 h before RNA isolation and RNase protection assays. Image of induced cell death due to the increased level of Fas and DR3 ex- protected bands, left panel; quantitative analysis of normalized signals, pression. The concurrent up-regulation by CD30 of antiapoptotic right panels; uninduced levels are set to 100%. The Journal of Immunology 5111

mechanisms occur in T cells in vivo due to local CD30L expres- 19. Duckett, C. S., and C. B. Thompson. 1997. CD30-dependent degradation of sion. The reported finding that CD30 dampens cytotoxic activity of TRAF2: implications for negative regulation of TRAF signaling and the control of cell survival. Genes Dev. 11:2810. CTL in a diabetes model in mice (24) provides strong support for 20. Lee, S. Y., and Y. Choi. 1997. TRAF-interacting protein (TRIP): a novel com- this hypothesis. Moreover, in our own initial studies, CD30L-de- ponent of the tumor necrosis factor receptor (TNFR)- and CD30-TRAF signaling ficient mice show increased in vivo CTL activity supporting the complexes that inhibits TRAF2-mediated NF-␬B activation. J. Exp. Med. 185: 1275. role of CD30 as a negative regulator of cytotoxic lymphocytes. 21. Amakawa, R., A. Hakem, T. M. Kundig, T. Matsuyama, J. J. Simard, E. Timms, A. Wakeham, H. W. Mittruecker, H. Griesser, H. Takimoto, et al. 1996. Impaired References negative selection of T cells in Hodgkin’s disease antigen CD30-deficient mice. Cell 84:551. 1. Younes, A., and A. Carbone. 1999. CD30/CD30 ligand and CD40/CD40 ligand in malignant lymphoid disorders. Int. J. Biol. Markers 14:135. 22. Chiarle, R., A. Podda, G. Prolla, E. R. Podack, G. J. Thorbecke, and G. Inghirami. 2. Gattei, V., M. Degan, F. M. Rossi, A. de Iuliis, F. T. Mazzocco, D. Serraino, 1999. CD30 overexpression enhances negative selection in the thymus and me- V. Zagonel, D. Aldinucci, and A. Pinto. 1999. CD30 ligand (CD30L)-expressing diates programmed cell death via a Bcl-2-sensitive pathway. J. Immunol. 163: acute myeloid leukemias: a new model of paracrine interactions for the regulation 194. of blast cells proliferation. Leuk. Lymphoma 35:21. 23. Romagnani, P., F. Annunziato, R. Manetti, C. Mavilia, L. Lasagni, C. Manuelli, 3. Gruss, H. J., N. Boiani, D. E. Williams, R. J. Armitage, C. A. Smith, and G. B. Vannelli, V. Vanini, E. Maggi, C. Pupilli, and S. Romagnani. 1998. High R. G. Goodwin. 1994. Pleiotropic effects of the CD30 ligand on CD30-expressing CD30 ligand expression by epithelial cells and Hassal’s corpuscles in the medulla cells and lymphoma cell lines. Blood 83:2045. of human thymus. Blood 91:3323. 4. Lee, S. Y., C. G. Park, and Y. Choi. 1996. T cell receptor-dependent cell death 24. Kurts, C., F. R. Carbone, M. F. Krummel, K. M. Koch, J. F. Miller, and of T cell hybridomas mediated by the CD30 cytoplasmic domain in association W. R. Heath. 1999. Signalling through CD30 protects against autoimmune dia- with tumor necrosis factor receptor-associated factors. J. Exp. Med. 183:669. betes mediated by CD8 T cells. Nature 398:341. 5. Chiarle, R., A. Podda, G. Prolla, J. Gong, G. J. Thorbecke, and G. Inghirami. 1999. CD30 in normal and neoplastic cells. Clin. Immunol. 90:157. 25. Heath, W. R., C. Kurts, I. Caminschi, F. R. Carbone, and J. F. Miller. 1999. CD30 6. Brugnoni, D., P. Airo, F. Facchetti, L. Blanzuoli, A. G. Ugazio, R. Cattaneo, and prevents T-cell responses to non-lymphoid tissues. Immunol. Rev. 169:23. L. D. Notarangelo. 1997. In vitro cell death of activated lymphocytes in Omenn’s 26. Bowen, M. A., K. J. Olsen, L. Cheng, D. Avila, and E. R. Podack. 1993. Func- syndrome. Eur. J. Immunol. 27:2765. tional effects of CD30 on a large granular lymphoma cell line, YT: inhibition of 7. Horie, R., and T. Watanabe. 1998. CD30: expression and function in health and cytotoxicity, regulation of CD28 and IL-2R, and induction of homotypic aggre- disease. Semin. Immunol. 10:457. gation. J. Immunol. 151:5896. 8. Mavalia, C., C. Scaletti, P. Romagnani, A. M. Carossino, A. Pignone, L. Emmi, 27. Azuma, M., M. Cayabyab, D. Buck, J. H. Phillips, and L. L. Lanier. 1992. In- C. Pupilli, G. Pizzolo, E. Maggi, and S. Romagnani. 1997. Type 2 helper T-cell volvement of CD28 in MHC-unrestricted cytotoxicity mediated by a human nat- predominance and high CD30 expression in systemic sclerosis. Am. J. Pathol. ural killer leukemia cell line. J. Immunol. 149:1115. 151:1751. 9. Gilfillan, M. C., P. J. Noel, E. R. Podack, S. L. Reiner, and C. B. Thompson. 28. Wano, Y., T. Uchiyama, K. Fukui, M. Maeda, H. Uchino, and J. Yodoi. 1984. 1998. Expression of the costimulatory receptor CD30 is regulated by both CD28 Characterization of human interleukin 2 receptor (Tac antigen) in normal and and cytokines. J. Immunol. 160:2180. leukemic T cells: co-expression of normal and aberrant receptors on Hut-102 10. Nakamura, T., R. K. Lee, S. Y. Nam, B. K. Al-Ramadi, P. A. Koni, K. Bottomly, cells. J. Immunol. 132:3005. E. R. Podack, and R. A. Flavell. 1997. Reciprocal regulation of CD30 expression 29. Genestier, L., S. Kasibhatla, T. Brunner, and D. R. Green. 1999. Transforming on CD4ϩ T cells by IL-4 and IFN-␥. J. Immunol. 158:2090. growth factor ␤1 inhibits Fas-L expression and subsequent activation-induced 11. Bowen, M. A., R. K. Lee, G. Miragliotta, S. Y. Nam, and E. R. Podack. 1996. cell death in T cells via downregulation of c-myc. J. Exp. Med. 189:231. Structure and expression of murine CD30 and its role in cytokine production. 30. Stein, J. V., A. Rot, Y. Luo, M. Narasimhaswamy, H. Nakano, M. D. Gunn, J. Immunol. 156:442. A. Matsuzawa, E. J. Quackenbush, M. E. Dorf, and U. H. von Andrian. 2000. The 12. Alzona, M., H. M. Jack, R. I. Fisher, and T. M. Ellis. 1994. CD30 defines a subset CC chemokine thymus-derived chemotactic agent 4 (TCA-4, secondary lym- ␥ of activated human T cells that produce IFN- and IL-5 and exhibit enhanced B phoid tissue chemokine, 6Ckine, exodus-2) triggers lymphocyte function-associ- cell helper activity. J. Immunol. 153:2861. ated antigen 1-mediated arrest of rolling T lymphocytes in peripheral lymph node 13. Durkop, H., U. Latza, M. Hummel, F. Eitelbach, B. Seed, and H. Stein. 1992. high endothelial venules. J. Exp. Med. 191:61. Molecular cloning and expression of a new member of the nerve growth factor receptor family that is characteristic for Hodgkin’s disease. Cell 68:421. 31. Yoshida, R., T. Imai, K. Hieshima, J. Kusuda, M. Baba, M. Kitaura, 14. Smith, C. A., H. J. Gruss, T. Davis, D. Anderson, T. Farrah, E. Baker, M. Nishimura, M. Kakizaki, H. Nomiyama, and O. Yoshie. 1997. Molecular G. R. Sutherland, C. I. Brannan, N. G. Copeland, N. A. Jenkins, et al. 1993. CD30 cloning of a novel human CC chemokine EBI1-ligand chemokine that is a spe- antigen, a marker for Hodgkin’s lymphoma, is a receptor whose ligand defines an cific functional ligand for EBI1, CCR7. J. Biol. Chem. 272:13803. emerging family of cytokines with homology to TNF. Cell 73:1349. 32. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two 15. Gedrich, R. W., M. C. Gilfillan, C. S. Duckett, J. L. Van Dongen, and subsets of memory T lymphocytes with distinct homing potentials and effector C. B. Thompson. 1996. CD30 contains two binding sites with different specific- functions. Nature 401:708. ities for members of the tumor necrosis factor receptor-associated factor family 33. Schwenzer, R., K. Siemienski, S. Liptay, G. Schubert, N. Peters, P. Scheurich, of signal transducing proteins. J. Biol. Chem. 271:12852. R. M. Schmid, and H. Wajant. 1999. The human tumor necrosis factor (TNF) 16. Mizushima, S., M. Fujita, T. Ishida, S. Azuma, K. Kato, M. Hirai, M. Otsuka, receptor-associated factor 1 gene (TRAF1) is up-regulated by cytokines of the T. Yamamoto, and J. Inoue. 1998. Cloning and characterization of a cDNA en- TNF ligand family and modulates TNF-induced activation of NF-␬B and c-Jun coding the human homolog of tumor necrosis factor receptor-associated factor 5 N-terminal kinase. J. Biol. Chem. 274:19368. (TRAF5). Gene 207:135. 17. Lee, S. Y., G. Kandala, M. L. Liou, H. C. Liou, and Y. Choi. 1996. CD30/TNF 34. Wang, C. Y., M. W. Mayo, R. G. Korneluk, D. V. Goeddel, and A. S. Baldwin, ␬ receptor-associated factor interaction: NF-␬B activation and binding specificity. Jr. 1998. NF- B antiapoptosis: induction of TRAF1 and TRAF2 and c-IAP1 and Proc. Natl. Acad. Sci. USA 93:9699. c- IAP2 to suppress caspase-8 activation. Science 281:1680. 18. Duckett, C. S., R. W. Gedrich, M. C. Gilfillan, and C. B. Thompson. 1997. 35. Horie, R., S. Aizawa, M. Nagai, K. Ito, M. Higashihara, T. Ishida, J. Inoue, and Induction of nuclear factor ␬B by the CD30 receptor is mediated by TRAF1 and T. Watanabe. 1998. A novel domain in the CD30 cytoplasmic tail mediates NF␬B TRAF2. Mol. Cell. Biol. 17:1535. activation. Int. Immunol. 10:203.