Biochemical and Biophysical Research Communications 254, 325–329 (1999) Article ID bbrc.1998.9899, available online at http://www.idealibrary.com on
Hamster Diphtheria Toxin Receptor: A Naturally Occurring Chimera of Monkey and Mouse HB-EGF Precursors Jeong-Heon Cha, Joanna S. Brooke, and Leon Eidels 1 Department of Microbiology, University of Texas Southwestern Medical Center, 6000 Harry Hines Boulevard, Dallas, Texas 75235-9048
Received November 24, 1998
The sensitivity of mammalian cell lines to diphtheria toxin (DT) varies between species. Monkey (Mk) Vero cells are highly sensitive to DT, whereas rat and mouse (Ms) cells are resistant; hamster (Hm) cells display moderate DT sensitivity. The precursor of the Mk heparin-binding epidermal growth factor-like growth factor (proHB-EGF) functions as a DT receptor but the Ms proHB-EGF does not. In this study we have cloned, expressed, and characterized the Hm proHB-EGF/DT receptor. The expression of Hm proHB-EGF confers moderate DT sensitivity to normally DT-resistant mouse cells. The amino acid sequence of Hm preproHB-EGF shows that, overall, it more closely resembles the Ms preproHB-EGF sequence, except in the DT-binding region where it more closely resembles the Mk sequence. In the DT-binding region the Hm proHBEGF sequence differs from the Mk proHB-EGF in only four amino acid residues (124, 126, 133, and 147); one of these residues, Ile 133 in Mk proHB-EGF, has been previously reported to be important for DT binding and sensitivity. Analysis of Mk proHB-EGF mutants with residues substituted for Ile 133 suggests that Asn 133 in Hm proHB-EGF may be responsible for the moderate DT sensitivity of Hm proHB-EGF-expressing cells. © 1999 Academic Press
Diphtheria toxin (DT) is a single polypeptide of 535 amino acid residues (M r 58,342) secreted by lysogenic Corynebacterium diphtheriae (1). Limited proteolytic cleavage of the toxin yields two disulfide-linked fragments, an amino-terminal A fragment (M r 21,167) and a carboxyl-terminal B fragment (M r 37,195) (2). The toxin binds to specific cell surface receptors via the 1 To whom correspondence should be addressed. Fax: (214)6485907. E-mail:
[email protected]. Abbreviations used: proHB-EGF, precursor form of HB-EGF found at the cell surface having an extracellular, a transmembrane, and a cytoplasmic domain; preproHB-EGF, pre form of proHB-EGF still containing the signal sequence which is subsequently cleaved prior to cell-surface expression (Fig. 2).
receptor-binding domain of its B fragment, and DT: receptor complexes then are internalized via clathrindependent endocytosis into an endosomal compartment (3–5). Upon acidification of the endosome, and possibly after dissociation of the toxin from its receptor (6), the hydrophobic amino-terminal domain of the B fragment inserts into the endosomal membrane and the A fragment is translocated into the cytosol, where it catalyzes the NAD-dependent ADP ribosylation of mammalian elongation factor 2, resulting in inhibition of protein synthesis (7–9). The sensitivity of mammalian cells to DT varies among different mammalian species; monkey (Mk) cells (e.g., Vero cells) are extremely sensitive whereas mouse (Ms) (e.g., L cells) and rat cells are resistant (8, 10 –13). Hamster (Hm) cells demonstrate intermediate toxin sensitivity (12). Although all mammalian elongation factor 2s are sensitive to DT (11), only those cells expressing a DT receptor are sensitive to the toxin (12). A cell-surface receptor for DT was cloned from Mk Vero cells and identified to be a heparin-binding epidermal growth factor-like growth factor (HB-EGF) precursor (proHB-EGF) (14). HB-EGF is a member of a family of growth factors which includes epidermal growth factor, transforming growth factor a, and amphiregulin (15). In addition to Mk proHB-EGF, the proHB-EGF cDNAs of human (16), mouse (17) and rat (17, 18) have been cloned. In contrast to Mk proHBEGF, Ms proHB-EGF is unable to bind DT and, therefore, cannot act as a DT receptor. These results are consistent with the traditional belief that the DT resistance of Ms cells is due to the lack of a functional DT receptor. Analysis of the DT sensitivity and DT binding of Ms cell lines that express chimeric proHB-EGFs, composed of Ms and Mk amino acid residues, has identified a region (residues 122–148) of the proHB-EGF that is critical for DT binding (19). Using reciprocal sitedirected mutagenesis of proHB-EGF to replace Ms residues with corresponding Mk residues and vice versa, we recently demonstrated that residue Glu 141 is the
325
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Vol. 254, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
most critical residue, that residue Ile 133 is the next most critical residue, and that His 135 may also contribute significantly to DT binding and sensitivity (20). The DT sensitivity and DT binding of Mk proHBEGF expressed in Ms L-M(TK 2) cells is increased by the co-expression of a Mk CD9 molecule on the cell surface (21). Ms L-M(TK 2) cells expressing Mk CD9 (L-M(TK 2)/CD9) have been used in our laboratory as host cells to examine the DT binding and sensitivity properties of wild-type and mutant proHB-EGFs (19, 20). Therefore, in order to compare the DT sensitivity and DT binding of the Hm proHB-EGF with that of the Mk proHB-EGF, we used the L-M(TK 2)/CD9 cell line as the host cell in this study. Since Hm cells are moderately DT sensitive (11), we hypothesized that the expression of Hm proHB-EGF in DT-resistant Ms cell lines would convey moderate DT sensitivity to these cell lines. In this study, we tested this hypothesis by cloning the Hm preproHB-EGF 2 cDNA and characterizing the DT binding and sensitivity of Hm proHB-EGF cells. The amino acid sequence predicted from this cDNA revealed it to be a natural chimera of Mk and Ms proHB-EGFs, sharing a high level of identity with the DT-binding region of Mk proHB-EGF in an otherwise Ms proHB-EGF-like background. We analyzed the DT binding and sensitivity of L-M(TK 2)/CD9 cells expressing the Hm DT receptor on their cell surface and observed that the expression of Hm proHB-EGF conveys moderate DT sensitivity to DT-resistant mouse cells. MATERIALS AND METHODS Materials. The vector pCR2.1, and the bacterial host strains of E. coli INVaF9 and DH5a were purchased from Invitrogen. [ 125I]NaI (IMS 30; 13–17 mCi/mg), and L-[4,5- 3H]leucine (60 Ci/mmol) were obtained from Amersham. The Chinese hamster ovary (CHO) cell cDNA library was generously provided by Rockford K. Draper, University of Texas at Dallas. Partially purified DT was purchased from Connaught Laboratories (Ontario, Canada), purified further by anion-exchange chromatography (22) with modifications (23), and was radioiodinated as described previously (23). Restriction enzymes were from Boehringer-Mannheim. Cloning cylinders were from BELLCO Glass, Inc. Disposable electroporation cuvettes (4 mm gap) were from BTX. All other reagents were as previously described (24). Cell culture. We employed the L-M(TK 2)/CD9 cell line which stably expresses Mk CD9 cell-surface molecules and is DT resistant (21). As a DT-sensitive cell line control, a wild-type Mk proHB-EGFexpressing cell line was constructed by the transfection of pMkHBEGF into L-M(TK 2)/CD9 cells by electroporation (20). This cell line will be referred to as Mk proHB-EGF. Cloning of Hm DT receptor. The Hm DNA sequence of the DT receptor was amplified from CHO cell cDNA library using two oligonucleotides 59-GGGCCCAAGCTTCGGGACCATGAAGCTGCTGCCGTCGG-39 and 59-CCTATGGTACCTAAACATGAGAAGCCCCACGATGAC-39 in a PCR. The former primer was engineered to contain a HindIII site and the latter primer sequence originally contained a KpnI site (restriction sites are indicated by the underlined sequences). The PCR product (583 bp) was extracted from a 0.8% agarose gel using a Prep-A-Gene DNA purification Kit (Bio-Rad) and then was cloned into the TA cloning vector pCR2.1, resulting in
pCHA1028 (Fig. 1). The PCR Hm HB-EGF insert of two independent clones was sequenced in both forward and reverse directions using the ABI PRISM dye terminator cycle sequencing ready reaction kit (Perkin–Elmer) with the ABI Prism 377 DNA Sequencer at the sequencing facilities of our Institution. The DNA sequence for the Hm HB-EGF insert of both clones was identical. The DNA sequence for the Hm DT receptor was submitted to GenBank/EMBL (accession number AF069753). The PCR insert was digested with HindIII and KpnI and a 570-bp fragment containing the Hm DT receptor was used to replace the equivalent fragment of the Mk DT receptor in plasmid pMkHB-EGF (Fig. 1). This new construct, pHmHB-EGF (Fig. 1), was used to characterize the Hm DT receptor. This Hm DT receptor contained the amino acid residues 1-186 (a signal sequence, an extracellular domain, a transmembrane domain and a part of the cytoplasmic domain) of Hm preproHB-EGF and residues 187 to 208 (a part of the cytoplasmic domain) of the Mk proHB-EGF. Mutagenesis of Mk proHB-EGF. Site-directed mutagenesis was performed using the Quickchange kit purchased from Stratagene. Oligonucleotides were synthesized by GIBCO/BRL and used as primers to generate and to sequence the preproHB-EGFs used in this study. The mutated sequences of the preproHB-EGF cDNAs were confirmed by automated sequencing as described above. Transfection. Electroporation (25) was used to transfect plasmid DNAs into the L-M(TK 2)/CD9 cells, using the BTX ECM 600 System with modifications as previously described (20). The L-M(TK 2)/CD9 cells were electroporated with plasmid DNAs and individual colonies were selected for in the presence of geneticin (1 mg/ml); the transfected cells were tested for DT sensitivity using a single concentration of 1 mg/ml DT (20). Six DT-sensitive isolated colonies were then further purified by a cloning cylinder method (20). The DT sensitivity of the six purified cell lines was analyzed further by the cytotoxicity assay described below. Also, the cell-surface expression of the proHB-EGF and the CD9 molecules in these cell lines was monitored by ELISAs as described below. Cytotoxicity assay. The DT sensitivity of purified DT receptorexpressing cell lines was assayed using cytotoxicity assays that have been described previously (20). Each assay included the Mk proHBEGF cell line and the L-M(TK 2)/CD9 cell line as positive and negative controls, respectively. DT sensitivity was measured by the degree of protein synthesis inhibition. All pure DT-sensitive cell lines employed showed at least 80% protein synthesis inhibition at 1 mg/ml of DT. All cytotoxicity assays were performed in duplicate and at least two separate cytotoxicity assays were performed for each cell line. Screening for Mk CD9 and proHB-EGF molecules on the cell surface by ELISA. Cell-surface expression of Mk CD9 and proHB-EGF molecules was monitored using whole-cell ELISAs as described previously (20, 21). Each Mk CD9 ELISA assay included the Mk proHBEGF/CD9 cell line as a positive control and the L-M(TK 2) cell line as a negative control. Each proHB-EGF ELISA assay included the Mk proHB-EGF cell line as a positive control and the L-M(TK 2)/CD9 cell line as a negative control. Radiolabeled DT binding assay. The apparent DT binding affinity of DT receptor-expressing cells was measured by procedures previously described (14, 21). Each assay included the Mk proHBEGF cell line and the L-M(TK 2)/CD9 cell line as positive and negative controls, respectively. For each cell line, a representative clone expressing cell-surface MkCD9 and proHB-EGF was selected for the binding assay. All DT binding assays were performed in duplicate and at least two separate DT binding assays were performed for each cell line. Sequencing of proHB-EGF cDNA integrated into the chromosome of cell lines. Chromosomal DNA was isolated from the representative cell line selected for the DT binding assay and sequenced to confirm the fidelity of the stably integrated proHB-EGF cDNA as previously described (20).
326
Vol. 254, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
of the cytoplasmic domain, with the remainder of the cytoplasmic domain provided by Mk proHB-EGF (Figs. 1 and 2).
FIG. 1. Construction of pHmHB-EGF. The fragment (Hm HBEGF9) of HindIII and KpnI from pCHA1028 replaces the equivalent fragment (Mk HB-EGF9) of pMkHB-EGF (see Materials and Methods).
Predicted amino acid sequence of Hm preproHBEGF. The overall structure of Hm preproHB-EGF is similar to that of all other preproHB-EGF proteins; it contains a signal sequence (residues 1–23), a pro region (residues 24 – 62), a mature growth factor (residues 63–148), a juxtamembrane domain (residues 149 –159), a transmembrane domain (residues 160 –184), and a carboxyl-terminal cytoplasmic domain (residues 185– 208) (Fig. 2). The juxtamembrane and transmembrane domains, as well as the heparin-binding domain (residues 92–123), show regions of high sequence conservation. Sequence differences occur in the carboxylterminal end of the pro regions and in the amino- and carboxyl-terminal ends of the mature growth factors. The EGF-like domain (residues 108 –143) of the Hm HB-EGF shows the six conserved cysteine residues that are spaced in the characteristic pattern of the other EGF-like domains (15). A comparison of the amino acid sequences among the Hm, Mk, and Ms HB-EGF precursors revealed that in the DT binding region (amino acids 122-148) there are 10 amino acids that are not identical (Fig. 2). Six of these residues are identical in the Hm and Mk sequences (60%), two are identical in the Hm and Ms
RESULTS AND DISCUSSION Cloning of Hm preproHB-EGF. The two primers that were used to clone the Hm preproHB-EGF cDNA were selected because they correspond to a highly conserved region at the DNA level in the human, monkey, rat, and mouse preproHB-EGFs, with 100% identity for the 59 primer and 92% identity for the 39 primer among these species. Since the amino acid residues encoded by these primers share 100% identity among four mammalian species, we predicted that it is unlikely that the amplified Hm preproHB-EGF sequence would contain incorrect amino acids due to the use of these primers. These primers enabled us to amplify the signal sequence, the extracellular domain, the transmembrane domain, and two amino acid residues of the cytoplasmic domain of Hm preproHB-EGF. This PCR product was cloned into the TA cloning vector pCR2.1, resulting in pCHA1028 as described under Materials and Methods (Fig. 1). This Hm preproHB-EGF nucleotide sequence was excised from pCHA1028 and inserted into pMkHB-EGF to replace the equivalent Mk preproHB-EGF sequence in pMkHB-EGF (Fig. 1). This final construct, pHmHB-EGF, contains the Hm HBEGF signal sequence, the extracellular domain, the transmembrane domain and two amino acid residues
FIG. 2. Comparison of the monkey (upper line) (14), hamster (middle line), and mouse (lower line) (17) preproHB-EGFs. Amino acid residues identical to the monkey sequence are denoted by the character, 2, and different residues are denoted by their single-letter amino acid code. The signal sequence, mature growth factor domain, and the transmembrane domain are indicated by black boxes and are based on the designations of Abraham et al. (17). The DT binding region (residues 122–148) is denoted by asterisks. Residues 133, 135, and 141, which are important for DT binding (20), are indicated by arrows. The amino acids corresponding to the primer sequences used to amplify the hamster preproHB-EGF are indicated by lowercase letters.
327
Vol. 254, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
DT Sensitivity and DT-Binding Ability of Cell Lines Cell line a
IC 50 (ng/ml) b
Kd (nM) b
n (3 10 25) b
Hm proHB-EGF Mk proHB-EGF Mk I133A Mk I133N
7.6 6 3.7 0.5 6 0.1 0.4 6 0.1 4.2 6 1.2
15.1 6 1.9 2.7 6 0.7 3.8 6 0.8 13.2 6 9.7
10 6 2.8 4.7 6 2.3 1.4 6 0.5 1.4 6 0.5
a The constructs Mk I133A and Mk I133N have the monkey Ile 133 residue replaced by Ala and Asn, respectively. b All cytotoxicity assays were done in duplicate and were performed at least twice for the selected representative cell line with the average IC 50 value 6 variation from the mean being reported. All binding assays were done in duplicate and were performed at least twice for each cell line with the average calculated K d and number of binding sites per cell (n) values 6 variation from the mean being reported.
sequences (20%), and two are unique to the Hm sequence (20%). Since Hm cells are moderately DT sensitive, it is not surprising that the DT binding region of Hm proHB-EGF shares a higher degree of homology with the equivalent region in the Mk proHB-EGF than with this region in Ms proHB-EGF. Outside the DT binding region the amino acid sequence of Hm preproHB-EGF (excluding the cytoplasmic domain, as this was provided by Mk proHB-EGF in our cloning strategy) shows 32 different amino acid residues when compared to Mk and Ms preproHB-EGFs. Sixteen of these residues (50%) are identical to the Ms preproHBEGF, eight residues (25%) are identical to the Mk preproHB-EGF, and eight (25%) are unique to Hm preproHB-EGF (Fig. 2). These data, therefore, show that the amino acid sequence, excluding the DT binding region, of Hm preproHB-EGF resembles more closely the Ms preproHB-EGF than the Mk preproHBEGF. These results suggest that the Hm DT receptor is a naturally occurring chimera containing a Mk-like DT binding region in a Ms proHB-EGF-like background. Analysis of the DT sensitivity and DT binding of Hm proHB-EGF cells. L-M(TK 2)/CD9 cells were transfected with pHmHB-EGF, resulting in Hm proHBEGF-expressing cells. Analysis of the DT binding ability and DT sensitivity of this new cell line (Hm proHBEGF) allowed us to compare the Hm DT receptor with the Mk DT receptor in the same cell line L-M(TK 2)/ CD9 background. A comparison of Hm proHB-EGF cells with Mk proHB-EGF cells showed that the Hm proHB-EGF cells are less sensitive to DT (;15-fold) and have a lower affinity for DT (;6-fold) (Table 1). This result is interesting as it has been reported previously that hamster cells (BHK-21) display moderate DT sensitivity (;15-fold less sensitive) in comparison with highly DT sensitive Mk Vero cells (12). Together, these observations suggest that the difference in DT
sensitivity of hamster and monkey cells may be due to inherent differences in their DT receptors. The DT binding region of Hm proHB-EGF (residues 122-148) contains four amino acids that differ from those in the Mk proHB-EGF (residues 124, 126, 133, and 147) (Fig. 2). Three of these different amino acids represent conservative substitutions (residues 124, 126, and 147). Previous results from our laboratory suggest that in the Mk DT receptor Glu 141 plays the most critical role in toxin binding and sensitivity, that Ile 133 is the next most important residue, and that His 135 also contributes significantly to DT binding and sensitivity (20). Hm proHB-EGF contains Glu 141 and His 135 but, surprisingly, contains Asn 133 instead of Ile 133. Thus, we chose to examine the contribution of the Asn residue at position 133. For this purpose we prepared cells expressing mutant Mk proHB-EGFs, containing Ala or Asn residues instead of Ile 133, and compared them with the Mk wild-type proHB-EGF cells. Substitution of Ile 133 with Ala in Mk proHB-EGF resulted in similar toxin sensitivity (IC 50 5 0.4 ng/ml) and binding (K d 5 3.8 nM) to that of wild-type Mk proHB-EGF (IC 50 5 0.5 ng/ml; K d 5 2.7 nM) (Table 1). In contrast, substitution of Ile 133 with Asn—the amino acid found in position 133 in Hm proHB-EGF— resulted in a decrease in DT sensitivity (;8-fold; IC 50 5 4.2 ng/ml) and in a decrease in DT affinity (;5-fold; K d 5 13.2 nM) (Table 1). A similar decrease in sensitivity and affinity was previously observed when Ile 133 was substituted with Lys, the amino acid found at position 133 in Ms proHB-EGF (20). Recently, the crystal structure of a complex of DT with a fragment of human HB-EGF (residues 73–147) has been elucidated (26). Human HB-EGF is identical to Mk HB-EGF except for a single residue at position 87, which is Asn in human and Ser in Mk HB-EGF (14, 17). From this crystal structure it is apparent that residue Ile 133 is involved in an hydrophobic interaction with DT as it is situated at the predominantly nonpolar interior of the DT:HB-EGF interface. Substitution of this Ile residue with the similar non-polar residue Ala in Mk proHB-EGF resulted in almost no change in toxin binding and sensitivity. In contrast, substitution with the more polar Asn residue (27) disrupts this hydrophobic interaction; this disruption is reflected in the lower DT affinity and the decreased DT sensitivity. ACKNOWLEDGMENTS We thank Robert S. Munford for critical review of the manuscript. We also thank Katherine N. Ivey and Jason G. Ilgen for technical assistance. The editorial assistance of Eleanor R. Eidels is greatly appreciated. This research was supported by U.S. Public Health Service Grant AI-16805.
328
Vol. 254, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
REFERENCES 1. Greenfield, L., Bjorn, M. J., Horn, G., Fong, D., Buck, G. A., Collier, R. J., and Kaplan, D. A. (1983) Proc. Natl. Acad. Sci. USA 80, 6853– 6857. 2. Pappenheimer, A. M., Jr. (1977) Annu. Rev. Biochem. 46, 69 –94. 3. Keen, J. H., Maxfield, F. R., Hardegree, M. C., and Habig, W. H. (1982) Proc. Natl. Acad. Sci. USA 79, 2912–2916. 4. Morris, R. E., Gerstein, A. S., Bonventre, P. F., and Saelinger, C. B. (1985) Infect. Immun. 50, 721–727. 5. Moya, M., Dautry-Varsat, A., Goud, B., Louvard, D., and Boquet, P. (1985) J. Cell Biol. 101, 548 –559. 6. Brooke, J. S., Cha, J. H., and Eidels, L. (1998) Biochem. Biophys. Res. Commun. 248, 297–302. 7. Collier, R. J. (1975) Bacteriol. Rev. 39, 54 – 85. 8. Eidels, L., Proia, R. L., and Hart, D. A. (1983) Microbiol. Rev. 47, 596 – 620. 9. Middlebrook, J. L., and Dorland, R. B. (1984) Microbiol. Rev. 48, 199 –221. 10. Boquet, P., and Pappenheimer, A. M., Jr. (1976) J. Biol. Chem. 251, 5770 –5778. 11. Middlebrook, J. L., and Dorland, R. B. (1977) Can. J. Microbiol. 23, 183–189. 12. Middlebrook, J. L., Dorland, R. B., and Leppla, S. H. (1978) J. Biol. Chem. 253, 7325–7330. 13. Stenmark, H., Olsnes, S., and Sandvig, K. (1988) J. Biol. Chem. 263, 13449 –13455. 14. Naglich, J. G., Metherall, J. E., Russell, D. W., and Eidels, L. (1992) Cell 69, 1051–1061.
15. Carpenter, G., and Wahl, M. I. (1990) in Handbook of Experimental Pharmacology (Sporn, M. B., and Roberts, A. B., Eds.), p. 69, Springer-Verlag, New York. 16. Higashiyama, S., Abraham, J. A., Miller, J., Fiddes, J. C., and Klagsbrun, M. (1991) Science 251, 936 –939. 17. Abraham, J. A., Damm, D., Bajardi, A., Miller, J., Klagsbrun, M., and Ezekowitz, R. A. (1993) Biochem. Biophys. Res. Commun. 190, 125–133. 18. Temizer, D. H., Yoshizumi, M., Perrella, M. A., Susanni, E. E., Quertermous, T., and Lee, M. E. (1992) J. Biol. Chem. 267, 24892–24896. 19. Hooper, K. P., and Eidels, L. (1995) Biochem. Biophys. Res. Commun. 206, 710 –717. 20. Cha, J. H., Brooke, J. S., and Eidels, L. (1998) Mol. Microbiol. 29, 1275–1284. 21. Brown, J. G., Almond, B. D., Naglich, J. G., and Eidels, L. (1993) Proc. Natl. Acad. Sci. USA 90, 8184 – 8188. 22. Pappenheimer, A. M., Jr., Uchida, T., and Harper, A. A. (1972) Immunochemistry 9, 891–906. 23. Cieplak, W., Gaudin, H. M., and Eidels, L. (1987) J. Biol. Chem. 262, 13246 –13253. 24. Almond, B. D., and Eidels, L. (1994) J. Biol. Chem. 269, 26635– 26641. 25. Baum, C., Forster, P., Hegewisch-Becker, S., and Harbers, K. (1994) Biotechniques 17, 1058 –1062. 26. Louie, G. V., Yang, W., Bowman, M. E., and Choe, S. (1997) Mol. Cell 1, 67–78. 27. Kyte, J., and Doolittle, R. F. (1982) J. Mol. Biol. 157, 105– 132.
329