Specific Interaction of Topoisomerase II and the CD3 Chain of the T Cell Receptor Complex

T cell antigen receptor (TCR)•CD3 complex is composed of six different subunits: TCRα and TCRβ and CD3, CD3, CD3, and CD3. Antigen recognition signals are transduced from TCR to the cytoplasm through the cytoplasmic domain of the CD3 chains. To understand the downstream signal transduction pathways, we cloned genes encoding proteins capable of binding to CD3 with a probe of glutathione S-transferase fused to the cytoplasmic region of CD3. One of these clones was found to encode topoisomerase IIβ (topoIIβ). The binding region of CD3 is located within the N-terminal 12 amino acids containing the motif of a basic amino acid cluster. A similar motif was found in the chain of Fc receptors (FcR) but not in the CD3 chain, and indeed, FcR but not CD3 bound to topoIIβ. The binding region of topoIIβ was determined to be the C terminus. Since this region appears to be the regulatory region of the enzymatic activity, the binding of CD3 might affect the function of topoIIβ. Although topoIIβ is localized mainly in the nucleus and CD3 is a membrane protein, we demonstrated the presence of CD3 in the nuclear fraction of thymocytes, which increased upon T cell activation. The specific interaction in cells was evidenced by co-immunoprecipitation of topoIIβ and CD3 from the nuclear fraction of T cells. The possible function of this interaction is discussed.

T cell antigen receptor (TCR)⅐CD3 complex is composed of six different subunits: TCR␣ and TCR␤ and CD3␥, CD3␦, CD3⑀, and CD3. Antigen recognition signals are transduced from TCR to the cytoplasm through the cytoplasmic domain of the CD3 chains. To understand the downstream signal transduction pathways, we cloned genes encoding proteins capable of binding to CD3⑀ with a probe of glutathione S-transferase fused to the cytoplasmic region of CD3⑀. One of these clones was found to encode topoisomerase II␤ (topoII␤). The binding region of CD3⑀ is located within the N-terminal 12 amino acids containing the motif of a basic amino acid cluster. A similar motif was found in the ␥ chain of Fc receptors (FcR␥) but not in the CD3 chain, and indeed, FcR␥ but not CD3 bound to topoII␤. The binding region of topoII␤ was determined to be the C terminus. Since this region appears to be the regulatory region of the enzymatic activity, the binding of CD3⑀ might affect the function of topoII␤. Although topoII␤ is localized mainly in the nucleus and CD3⑀ is a membrane protein, we demonstrated the presence of CD3⑀ in the nuclear fraction of thymocytes, which increased upon T cell activation. The specific interaction in cells was evidenced by co-immunoprecipitation of topoII␤ and CD3⑀ from the nuclear fraction of T cells. The possible function of this interaction is discussed.
The T cell antigen receptor (TCR) 1 ⅐CD3 complex is composed of six subunits: clonotypic ␣ and ␤ chains and invariant CD3␥, CD3␦, CD3⑀, and CD3 chains. TCR␣ and TCR␤ chains recognize antigen in association with major histocompatibility complex molecules, and the CD3 chains are responsible for trans-ducing antigen recognition signals by TCR from the membrane to the cytoplasm. CD3␥, CD3␦, CD3⑀, and CD3 contain a conserved amino acid sequence motif, ITAM (immunoreceptor tyrosine based activation motif) in their cytoplasmic domain (1), which is composed of a pair of the YXXL/I sequence with a spacer of 7-8 amino acids. ITAM is also present in the cytoplasmic tail of Ig␣ and Ig␤ of the B cell antigen receptor complex and the ␥ chain of Fc receptors (FcR␥) (2). Since the TCR⅐CD3 complex does not possess any intrinsic kinase function, tyrosine kinases associated with the complex have been shown to be important for signal transduction. Fyn and Lck associate noncovalently with the TCR⅐CD3 complex and CD4/ CD8, respectively (3,4). Upon TCR stimulation, these kinases are activated and phosphorylate several cellular substrates (5). Two tyrosine residues within ITAM are also phosphorylated and become the binding site of the SH2 domains of ZAP-70 or Syk kinase (6,7). In T cells, the recruitment of ZAP-70 to CD3⑀ or CD3 (8,9) induces the activation of this kinase and subsequently exhibits functions such as the production of lymphokines.
Utilizing chimeric molecules such as CD8 or Tac-⑀ (the extracellular domain of CD8 or the ␣ chain of the IL-2 receptor, fused to the cytoplasmic domain of or ⑀, respectively) (10,11), it has been demonstrated that each of the cytoplasmic domains of CD3 chains induces similar activation events to those through the TCR⅐CD3 complex, including tyrosine phosphorylation, Ca 2ϩ mobilization, and IL-2 production. On the other hand, there is evidence to indicate that the TCR complex is composed of two activation modules, CD3␥␦⑀ and CD3 (12,13). Molecules important for signaling pathways associated with each of the activation modules have to be determined. Signals mediated through the chain are required for some activation such as Thy-1 (12, 13)-and CD2-mediated stimulation (14). The relationship between signals through these two modules has not yet been clarified. In addition, it has been shown that stimulation through CD3⑀ of immature thymocytes without undergoing rearrangement of TCR genes induces differentiation of thymocytes, indicating that signals through CD3⑀ play important roles in T cell development (15,16). Furthermore, although phosphorylation-dependent signals in T cells have been extensively analyzed, the molecules associated with unphosphorylated forms of CD3⑀ or CD3 other than Fyn have not yet been identified (3,17).
In order to understand the downstream signaling events through CD3⑀, one of the TCR activation modules, we cloned genes encoding CD3⑀-binding proteins. One of these clones was found to encode topoisomerase II␤ (topoII␤) (18). We demonstrated the presence of CD3⑀ in the nuclear fraction upon T cell activation and showed the evidence of the specific association in vivo between CD3⑀ and topoII␤ in this fraction of T cells. * This work was supported by grants-in-aid for Scientific Research from the Ministry of Education, Science, and Culture, and from the Agency for Science and Technology, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ Present address: Dept. of Immunology, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113, Japan.
Detailed mapping of the binding regions of both CD3⑀ and topoII␤ indicates that the association depends on a novel motif in CD3⑀ and suggests that the binding might modulate the in vivo function of topoII␤.
Preparation of 32 P-Labeled GST Fusion Proteins-32 P-Labeled GST fusion proteins were prepared as described previously (21). Briefly, purified GST fusion proteins were adsorbed onto glutathione-Sepharose beads. The beads were washed once with a kinase buffer (20 mM Tris (pH 7.5), 100 mM NaCl, 12 mM MgCl 2 ) and then resuspended in 2-3 bead volumes of the buffer containing the catalytic subunit of cAMPdependent protein kinase (Sigma), [␥-32 P]ATP, and 1 mM dithiothreitol. The kinase reaction proceeded for 30 min and was terminated by the addition of 1 ml of a stop buffer (10 mM sodium phosphate (pH 8.0), 10 mM sodium pyrophosphate, 10 mM EDTA, 1 mg/ml bovine serum albumin). After the supernatant was removed, the beads were washed with a TNEN buffer (20 mM Tris (pH 8.0), 100 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40). Labeled GST fusion proteins were eluted by agitating the beads for 15 min in 10 -50 bead volumes of 20 mM reduced glutathione, 100 mM Tris (pH 8.0), 120 mM NaCl. The probes were labeled to high specific activity (approximately 5-10 ϫ 10 5 cpm/g of protein).
Screening of cDNA Library-A gt11 library, constructed from mRNA from an HTLV-1 transformed human T cell line, HAT109, was obtained from Dr. M. Yoshida (The Research Institute of Medical Science, University of Tokyo). A total of 1.6 ϫ 10 6 plaques were screened with a 32 P-labeled GST-⑀ fusion protein. After incubation of the plates at 42°C for 4 h, the plates were overlaid with nitrocellulose filters that had been impregnated with IPTG and incubated at 37°C overnight. The filters were then removed, washed with a TBST buffer (10 mM Tris (pH 8.0), 150 mM NaCl, 0.05% Triton X-100) at room temperature, denatured with 6 M guanidine hydrochloride in a Hepes balanced buffer (HBB) (20 mM Hepes (pH 7.5), 5 mM MgCl 2 , 1 mM KCl) and renatured as described previously (21). Thereafter, the membranes were blocked in the HBB buffer containing 5% dry milk (Yukijirushi Ltd., Hokkaido, Japan) at 4°C for 1 h and then with 1% dry milk HBB. Labeled probes were added at approximately 5 ϫ 10 4 cpm/ml to HBB and incubated overnight. The filters were then washed three times with phosphatebuffered saline (PBS) containing 0.2% Triton X-100 at 4°C, dried, and exposed at Ϫ80°C. Positive phages were subsequently isolated and the cDNA inserts were sequenced after subcloning into pBluescript.
Filter Binding Experiments-The lysogens from the positive phages were prepared by standard procedure (22) for filter binding experiments. The lysogens of phages, GST-topoII␤1 and GST-topoII␤2, were induced with IPTG. The induced and uninduced proteins were fractionated by SDS-PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane (Millipore, Bedford, MA) with a transfer buffer without methanol as described previously (21). The membranes were denatured with 6 M guanidine hydrochloride, renatured, and blocked with 5% dry milk-PBS for 1 h at room temperature. The membranes were probed with 0.1-1 g/ml of GST-⑀ or GST alone, followed by incubation with anti-GST mAb or anti-␤ galactosidase antibody in PBS-0.2% Tween 20 (PBS-T) at 4°C overnight. After washing with PBS-T, the membranes were incubated with peroxidase conjugated anti-mouse Ig antibody at room temperature for 1 h, washed three times with PBS-T, and developed with an ECL detection system (Amersham).
In Vitro Solution Binding Experiments and Western Blots-After 5 g of each GST fusion protein was preabsorbed on the glutathione-Sepharose by incubating at 4°C for 1 h, the nuclear extracts of cells were added to the mixtures and further incubated at 4°C for 2 h. The mixtures were washed three times with PBS-0.1% Nonidet P-40 (PBS-N). The lysates were subjected to SDS-PAGE, and the proteins were transferred to PVDF membranes. The membrane were blocked with Block Ace (Yukijirushi Ltd., Hokkaido, Japan) at room temperature for 1 h, incubated with anti-topoII␤ mAb (␤5A7) for 1 h at room temperature, washed three times with a TBS-T buffer (20 mM Tris (pH 7.6), 137 mM NaCl, 0.2% Tween 20), and incubated with peroxidase-conjugated anti-mouse Ig antibody at room temperature for 1 h. After three washes, the membranes were developed with an ECL detection system. Preparation of Nuclear Extracts-Cells were lysed in a lysis buffer (1% digitonin (Aldrich) or 1% Nonidet P-40, 50 mM Tris (pH 7.6), 150 mM NaCl, 1 mM EDTA, 0.15 unit/ml aprotinin, 1 mM PMSF, 1 mM sodium orthovanadate, and 1 mM sodium fluoride) at 4°C for 30 min. The lysate was centrifuged at 1500 rpm for 10 min, and the pellet was washed twice with the lysis buffer. The supernatants were used as the cytosolic fractions. The nuclear extracts were prepared as described by Drake et al. (23) with some minor modifications. Briefly, the nuclear pellets were resuspended in a nuclear extraction buffer (5 mM potassium phosphate (pH 7.0), 2 mM MgCl 2 , 0.1 mM EDTA, 1 mM PMSF, 10 g/ml leupeptin, 10% glycerol, 0.5 mM iodoacetoamide). 5 M NaCl solution was added slowly to make a final concentration of 0.35 M, and nuclear fractions were extracted at 4°C for 60 min with constant stirring. Extracts were centrifuged at 15,000 rpm for 20 min, and the 2 N. Nozaki and A. Kikuchi, manuscript in preparation.
supernatants were then dialyzed overnight at 4°C against PBS containing 0.5 mM PMSF. After centrifugation at 15,000 rpm for 20 min, the supernatants were collected and used as the nuclear extracts.
Immunoprecipitation-The cell lysates or the nuclear extracts from 2B4 or DO11.10 T cell hybridoma cells were immunoprecipitated with 5 g of anti-CD3⑀ mAb 2C11 or anti-human CD3 mAb OKT3 as a control and protein A-Sepharose beads. The precipitates were washed with the lysis buffer containing 0.1-0.5% Nonidet P-40 for the digitonin lysate or with 1% Brij 96-containing buffer for the Brij lysate. Immunoprecipitates were analyzed on 8% SDS-PAGE.
TCR Stimulation-Stimulation was performed as described previously (24). Briefly, 2B4 cells (4 ϫ 10 7 cells) or murine thymocytes (7 ϫ 10 7 cells) were incubated with 2C11 (10 g/ml) for 30 min on ice, washed twice with RPMI 1640, and stimulated by adding prewarmed GAH (100 g/ml) at 37°C for 2 min. The reaction was stopped by adding ice-cold PBS, and cells were lysed for 30 min at 4°C in a lysis buffer. Cytosolic extracts and nuclear extracts were prepared according to the same protocol as described above.
Biotinylation on Sorbent-Biotinylation on sorbent was performed as described previously (25). Briefly, cytosolic or nuclear extracts of 2B4 hybridomas (4 ϫ 10 7 cells) and thymocytes (7 ϫ 10 7 cells) were immunoprecipitated with protein A-Sepharose beads coupled with anti-CD3 mAb (HMT3-1) at 4°C for 2 h. The beads were washed three times with the lysis buffer and once with PBS, and proteins on the beads were biotinylated by incubation with 100 g/ml biotin (Pierce) in 1 ml of the labeling buffer (0.01 M Hepes (pH 8.0), 150 mM NaCl) at 4°C for 1 h. The beads were washed three times with the lysis buffer, and the proteins were eluted by boiling in the SDS-PAGE sample buffer and were then analyzed by nonreducing-reducing two-dimensional SDS-PAGE (12% for the first dimensional gel under nonreducing condition and 14% for the second dimensional gel under reducing condition).

RESULTS
Cloning of cDNAs Encoding the CD3⑀-binding Proteins-A gt11 expression library prepared from a human T cell line was screened with a probe of 32 P-labeled GST fusion protein containing the cytoplasmic domain of the CD3⑀ chain (GST-⑀). Eight positive clones (clones 3, 6, 10, 16, 25, 36, 43, and 46) encoding ␤-galactosidase fusion proteins were picked up and subjected to further analysis. Subsequent analysis demonstrated that three (10, 43, 46) and two clones (6, 36) contained overlapping cDNA sequences, respectively. Among the resulting five cDNAs encoding CD3⑀-binding proteins, we described the characterization of clone 3 (C3) in the present study.
Topoisomerase II␤ Specifically Binds to CD3⑀-To confirm the binding specificity of C3 to GST-⑀, we performed a filter binding assay with lysogenic phage lysates. As shown in Fig. 1, the GST-⑀ probe bound to a protein of approximately 180 kDa from the IPTG-induced lysate of C3, but not from the uninduced lysate of this clone nor from the induced lysate of an irrelevant clone C1. GST alone as a control did not show any specific interaction. The 180-kDa protein from C3 also interacted with anti-␤-galactosidase antibody, confirming that this protein was a ␤-galactosidase fusion protein (Fig. 1). Sequence analysis showed that C3 contained the C-terminal region of topoII␤. TopoII␤ has just been cloned recently (18), whereas its isoform, topoII␣, had been characterized previously. Both isoforms are considered to be involved in DNA replication and transcriptional regulation (26,27). However, the function of topoII␤ and the functional difference from its isoform are not yet understood. The region of topoII␤ corresponding to C3 was abundant in acidic charged amino acids and potential phosphorylation sites and is considered to be the regulatory region of the enzyme (18).
N Terminus of the Cytoplasmic Region of CD3⑀ Binds to TopoII␤-To define the precise binding region of CD3⑀ to to-poII␤, serial deletion mutants of CD3⑀ were constructed. The cytoplasmic region was tentatively divided into three regions; the N-terminal region: aa 135-161; the central region, aa 162-169; and the C-terminal region, aa 170 -189. The N-terminal region, the central portion, and the C-terminal region contain a basic aa cluster, a proline-rich sequence, and ITAM, respectively ( Fig. 2A). We prepared serial deletion constructs by PCR and site-directed mutagenesis in the tyrosine residues of ITAM as shown in Fig. 2A. Each GST fusion protein was induced with IPTG in E. coli and purified on glutathione-Sepharose. Purified proteins were resolved in SDS-PAGE. Coomassie Blue staining showed that all constructs made the expected size of the proteins (Fig. 2B).
The nuclear extracts prepared from 2B4 hybridoma cells were precipitated with each GST-protein prebound on glutathione beads, separated on SDS-PAGE, and transferred to a PVDF membrane. The membrane was blotted with anti-to-poII␤ mAb ␤5A7. GST-⑀, -EM3, -EM4, -ED3, -ED5, -ED1, but not GST-ED4 bound to topoII␤ (Fig. 2C). Since GST-ED1 contains only the N-terminal 12 aa, it is likely that the binding region is located in the basic aa cluster in the N-terminal region of CD3⑀ (Fig. 3A).
We next wanted to determine the specificity of the binding, whether the binding of topoII␤ is specific for CD3⑀ or whether it also binds to similar signaling molecules such as CD3 and FcR␥. As shown in Fig. 2C, none of GST-, GST-fyn, or GST alone bound to topoII␤. In contrast, GST-FcR␥ was found to bind to topoII␤ (Fig. 2C). Comparing the aa sequences of CD3⑀ and FcR␥, we found that the homologous sequence in the 12 aa binding region of CD3⑀ was present in FcR␥ (Fig. 3A). This novel motif contains a basic aa cluster. The CD3 chain contains clusters of basic aa such as KKRAR and RRR, but it did not bind to topoII␤, demonstrating that the binding is not due to nonspecific interactions with any clusters of basic aa.
Collectively, these data demonstrate that the binding of CD3⑀ to topoII␤ is sequence-specific, and only 12 aa of the N terminus of the cytoplasmic region of CD3⑀ are sufficient for the interaction.
The Most C-terminal Region of TopoII␤ Binds to CD3⑀-Since C3 contained a 1.6-kb fragment corresponding to the C-terminal region of topoII␤, we employed a filter binding assay to specify the precise binding region to CD3⑀. The cDNA fragment was divided into two regions, and the GST fusion constructs were prepared for each region. GST-topoII␤1 and GST-topoII␤2 contained the N-and C-terminal halves of C3, respectively (Fig. 4A). Each GST protein was examined for the binding ability to CD3⑀. IPTG-induced or uninduced lysate from the bacteria containing each construct was blotted with GST-⑀ or GST alone. As shown in Fig. 4B, only the induced lysate from GST-topoII␤2, but not from GST-topoII␤1, bound specifically to GST-⑀. This region of topoII␤ contains a cluster of acidic aa, suggesting that this cluster may be responsible for binding to CD3⑀.
CD3⑀ and CD3 Are Present in the Nucleus-Unlike topoII␣, the expression of topoII␤ is rather restricted, being especially high in the thymus (data not shown) (29), suggesting that topoII␤ might be involved in T cell function. Although CD3⑀ binds specifically to topoII␤, the obvious question raised was concerning the localization of these two proteins. Whereas CD3⑀ is a membrane protein, topoII␤ is localized mainly in the nucleus. Therefore, we examined the possibility of whether CD3⑀ also exists in the nucleus. One piece of evidence which may support this possibility is that both CD3⑀ and CD3 have the consensus sequence corresponding to nuclear localizing signal (NLS). As depicted in Fig. 3B, CD3⑀ possesses the motif homologous to the NLS of SV40 T antigen (28), and CD3 contains the nucleoplasmin-like NLS (bipartite) (29), respectively. Since we were unable to detect CD3⑀ or CD3 in the nuclear fraction by Western blotting (data not shown), a highly sensitive biotinylation method was employed (30) to detect even small amounts of CD3⑀ or CD3 in the nuclear fraction.
To this end, cytosolic and nuclear extracts from thymocytes and 2B4 hybridomas either unstimulated or stimulated by cross-linking with anti-CD3⑀ mAb were prepared as described under "Materials and Methods." These extracts were immunoprecipitated with anti-CD3⑀ mAb HMT3-1 and protein A-Sepharose. Precipitated proteins on the beads were labeled with biotin and analyzed on two-dimensional SDS-PAGE. As shown in Fig. 5, A and B, precipitation of the cytosolic fraction showed TCR␣␤ and CD3 dimers as off-diagonal spots and CD3⑀ as the spot slightly above the diagonal. As expected, both CD3⑀ and CD3 were detected in the nuclear fractions of both plasmic domains of ⑀, various mutants of ⑀, , c, and FcR␥. Schematic structures of these fusion proteins were shown. B, purified GST fusion proteins described in A. GST fusion proteins were purified on glutathione-Sepharose beads, separated on SDS-PAGE, and stained with Coomassie Blue. The molecular size markers are indicated at the left margin. C, minimum binding region of CD3⑀ to topoII␤ and the specificity of the binding among signaling molecules. The nuclear extracts of T cell hybridomas were precipitated with each GST fusion protein and subjected to SDS-PAGE. Proteins were transferred to membranes, and the membranes were blotted with anti-topoII␤ mAb. The arrow indicates topoII␤. The molecular size markers are indicated at the left margin. thymocytes (Fig. 5, C and D) and 2B4 hybridomas (not shown). The chain in the nuclear fraction was observed only upon TCR stimulation, and the amount of CD3⑀ also increased upon TCR stimulation in thymocytes. CD3⑀ from the cytosol fraction showed two spots (⑀ and ⑀Ј in Fig. 5B). Since ⑀Ј was slightly larger than ⑀ and reacted with anti-phosphotyrosine mAb 4G10 by Western blotting (data not shown), ⑀Ј appeared to be phosphorylated ⑀. On the other hand, most of CD3⑀ in the nuclear fraction appeared to be the same size as ⑀Ј, suggesting that most of CD3⑀ in the nuclear fraction is probably phosphorylated. TCR ␣ and ␤ dimers were barely detected in this fraction, suggesting that the detected CD3⑀ and CD3 in the nuclear fraction were not present as components of the whole TCR⅐CD3 complex and that the presence of CD3⑀ and CD3 in this fraction did not merely reflect the contamination of the cytosolic fraction.
To confirm the latter issue, the same cytosolic and nuclear fractions used to detect CD3⑀ were immunoprecipitated and blotted with anti-tubulin mAb, since tubulin is present exclusively in the cytosol (31). As shown in Fig. 6, tubulin was detected only in cytosolic fraction but not at all in the nuclear fraction of our preparations. This result clearly demonstrated that the contamination of the nuclear fraction by cytosol was negligible and that CD3⑀ is likely to be present in the nucleus.
These results suggest that CD3⑀ exists in the nucleus especially after TCR activation and that it binds to topoII␤ in the nucleus in vivo.
Specific Interaction between TopoII␤ and CD3⑀ in T Cells-To demonstrate the direct interaction between CD3⑀ and topoII␤ in vivo, the nuclear fraction was prepared from DO11.10 hybridoma cells by solubilization with either digitonin or Brij 96 and immunoprecipitated by anti-mouse CD3⑀ (2C11) mAb and anti-human CD3⑀ (OKT3) mAb as a control and then analyzed by Western blot with anti-topoII␤ Ab. As shown in Fig. 7, topoII␤ was co-precipitated from the nuclear fractions of both digitonin (Fig. 7, lanes 1 and 2)-and Brij 96 (Fig. 7, lanes 3 and 4)-lysed cells. The fact that the association was only observed in digitonin or Brij lysates and that this association was not observed when the immunoprecipitate was washed with the buffer containing a higher concentration of FIG. 6. Negligible carryover of the cytosol fraction into the nuclear fraction of the thymocyte preparation. The total lysates in the same cytosol (C) and nuclear fractions (N) described in the legend to Fig. 5 were subjected to 14% SDS-PAGE, and the proteins were transferred to membranes. The membrane was stained with Coomassie Blue (A) or blotted with anti-tubulin mAb followed by development with an ECL detection system (B). The arrow indicates tubulin. The molecular size markers are indicated at the left margin.
Nonidet P-40 than 0.3% suggested that the interaction between topoII␤ and CD3⑀ was not strong in vivo (data not shown). DISCUSSION We have cloned topoII␤ as a CD3⑀-binding molecule by west-Western screening procedure. We determined the binding regions of both CD3⑀ and topoII␤ as well as the specificity of the interaction. The binding region of CD3⑀ to topoII␤ was localized within 12 aa in the N-terminal region containing a novel motif of basic aa cluster. A similar motif was also found in FcR␥. Since ITAM has been thought to be the only functional domain within CD3⑀, this is the first report that a specific binding site other than ITAM exists in the cytoplasmic region of CD3⑀. The binding region was composed of a novel motif of a basic aa cluster. Although CD3 has other basic aa clusters, did not bind to topoII␤, confirming the specific binding of topoII␤ to CD3⑀. It has been suggested that the TCR-CD3 complex is composed of two distinct activation modules (12,13). Distinct signals are transduced through CD3⑀ and CD3. The binding of ⑀, but not , to topoII␤ may represent one of such differences in signal transduction.
In terms of the physiological interaction of these two proteins, the question of their localization then arises. TopoII␤ is considered to be located in the nucleus whereas CD3⑀ is a membrane protein, and this question was therefore answered by demonstrating the translocation of CD3⑀ into the nucleus. Our finding that CD3⑀ and CD3 have NLS-like sequences in their cytoplasmic region supports the idea of translocation of a part of CD3⑀ and CD3 into the nucleus. There have been several reports about transmembrane-type receptors such as epidermal growth factor receptor and platelet-derived growth factor receptor, similar to CD3⑀, translocating to the nucleus (32). Luton et al. (33) reported that the TCR⅐CD3 complex translocated to the cytoskeleton-associated insoluble fraction upon TCR stimulation. Considering that the nuclear fraction was involved in the insoluble fraction and nuclear translocation was linked with cytoskeleton-associated proteins such as actin filament, their observation may partly reflect the translocation of the CD3 chains to the nucleus. Although we failed to detect CD3⑀ or CD3 in the nuclear fraction by Western blotting, we eventually succeeded in detecting both CD3⑀ and CD3 in the nuclear fractions of normal thymocytes and T hybridoma cells by labeling with a sensitive biotinylation method. Importantly, the amounts of CD3⑀ and CD3 were increased upon TCR activation while those in the cytosolic fraction did not seem to be changed. Taken together with the result that our preparation of the nuclear fraction did not contain any detectable contamination by the cytosolic fraction, we demonstrated for the first time that CD3⑀ and exist in the nucleus in normal T cells and increased upon T cell activation. By immunoprecipitation with anti-CD3⑀ mAb, we clearly demonstrated the direct in vivo association between CD3⑀ and topoII␤ in the nuclear fraction of T cells. These data indicate that CD3⑀ is translocated into the nucleus upon T cell activation and interacts with topoII␤.
In addition, there is a possibility that CD3⑀ may also bind to topoII␤ present in cytosol. Indeed, it has recently been observed that topoII␤ is transiently distributed to the cytoplasm during the mitotic stage, whereas topoII␣ is associated tightly with chromosomes constantly throughout the cell cycle. 3 Previous analyses of T cells expressing extensive deletion constructs of CD3⑀ showed that ITAM is both necessary and sufficient for IL-2 production (11). These results indicate that the topoII␤-binding motif of CD3⑀ is not prerequisite for IL-2 production. However, it has recently been reported that inhibitors of topoII, quinolon derivatives, up-regulated IL-2 production upon TCR stimulation (34). Since the CD3⑀-binding site is the regulatory region of topoII␤ in vitro, which was also suggested by the fact that an anti-topoII␤ mAb specific for the C terminus region of topoII␤ inhibited the enzymatic activity, 3 the binding of CD3⑀ may also block the enzymatic function and result in the super induction of IL-2 production, similar to the treatment of T cells with the topoII inhibitors. Moreover, topoII inhibitors are known to induce apoptosis (35,36). Growth arrest and subsequent apoptosis is induced in T cells upon TCR stimulation. The binding of CD3⑀ to the regulatory region of topoII might modulate the function similarly to the inhibitors in vivo. Although we tested this hypothesis by performing a decatenation assay with nuclear extracts, we failed to detect any significant effect of the GST-⑀ binding on the in vitro function of topoII␤ (data not shown). We assumed that the failure of modulation in the overall decatenation assay may be due to the dominant function of topoII␣ even under the condition that CD3⑀ binding may modify the function of topoII␤ protein. Analysis by the use of recombinant topoII␤, which is not available yet, will be required.
Collectively, CD3⑀ appears to possess two distinct functional domains; whereas ITAM is phosphorylated and stimulates ZAP70 and the following activation pathway in the cytoplasm, the N-terminal motif of CD3⑀ found in this study plays a functional role in the nucleus after translocation.