Phosphorylation of DNA Polymerase A-Primase by Cyclin A-Dependent Kinases Regulates Initiation of DNA Replication in Vitro

Home , DNA polymerase, Polymerase

Phosphorylation of DNA polymerase a-primase by Cyclin A-dependent kinases regulates initiation of DNA replication in vitro

C Voitenleitner1,2, E Fanning1,2, and H-P Nasheuer1,3

1Institut fuÈr Biochemie, LMU MuÈnchen, WuÈrmtalstr. 221, 81375 MuÈnchen, Germany; 2Department of Molecular Biology, Vanderbilt University, Nashville, Tennessee 37235, USA; 3Institut fuÈr Molekulare Biotechnologie, Abteilung Biochemie, Beutenbergerstrasse 11, 07745 Jena, Germany

DNA polymerase a-primase is the only known eukar- of SV40 DNA replication, are potential substrates of yotic enzyme that can start DNA replication de novo.In Cyclin-dependent kinases (Cdks; Dutta and Stillman this study, we investigated the regulation of DNA 1992; McVey et al., 1989; Nasheuer et al., 1991, 1992). replication by phosphorylation of DNA polymerase a- Since RP-A and DNA polymerase a-primase are key primase. The p180 and the p68 subunits of DNA cellular replication proteins, and they are phosphory- polymerase a-primase were phosphorylated using Cyclin lated in a cell cycle-dependent manner, it was suggested A-, B- and E- dependent kinases. This phosphorylation that their phosphorylation might be important for the did not in¯uence its DNA polymerase activity on regulation of DNA replication (Din et al., 1990; Dutta activated DNA, but slightly stimulated primase activity and Stillman 1992; Ferrari et al., 1996; Foiani et al., using poly(dT) single-stranded DNA (ssDNA) without 1995; Nasheuer et al., 1991). changing the product length of primers. In contrast, site- To determine functional interactions of Cdks and speci®c initiation of replication on plasmid DNA replication proteins, we used the in vitro SV40 DNA containing the SV40 origin is aected: Cyclin A-Cdk2 replication system. Four Cyclin-Cdks, i.e. Cyclin A- and Cyclin A-Cdc2 inhibited initiation of SV40 DNA Cdc2, Cyclin A-Cdk2, Cyclin B-Cdc2, and Cyclin E- replication in vitro, Cyclin B-Cdc2 had no eect and Cdk2, phosphorylated DNA polymerase a-primase, but Cyclin E-Cdk2 stimulated the initation reaction. DNA only Cyclin A-Cdc2 and Cyclin A-Cdk2 prevented polymerase a-primase that was pre-phosphorylated by initiation of SV40 DNA replication in vitro.In Cyclin A-Cdk2 was completely unable to initiate the contrast, Cyclin E-Cdk2 and Cyclin B-Cdc2 showed a SV40 DNA replication in vitro; Cyclin B-Cdc2-phos- modest eect. phorylated enzyme was moderately inhibited, while Cyclin E-Cdk2-treated DNA polymerase a-primase remained fully active in the initiation reaction. Results Keywords: DNA replication; cyclins; cyclin-dependent kinases; DNA polymerase a-primase Phosphorylation of DNA polymerase a-primase by Cyclin-Cdk complexes To study the function of protein phosphorylation on Introduction DNA replication, we puri®ed highly active, recombinant DNA polymerase a-primase, Cyclin A-Cdc2, Replication of SV40 DNA in vitro serves as an Cyclin A-Cdk2, Cyclin B-Cdc2 and Cyclin E-Cdk2. excellent model system to study eukaryotic DNA The catalytic subunits Cdc2 and Cdk2 phosphorylated replication and its regulation (Challberg and Kelly, neither histone H1 nor DNA polymerase a-primase 1989; D'Urso et al., 1990; Fanning, 1994; Hurwitz et above background level (Figure 1a; data not shown). al., 1990; Li and Kelly, 1984; Stillman, 1989; Waga et The puri®ed Cyclin-Cdk complexes have reproducibly al., 1994). Recently, it was shown that ten proteins or high kinase activity on histone H1 and on DNA protein complexes are necessary and sucient for SV40 polymerase a-primase as a substrate, but they DNA replication in vitro (Hurwitz et al., 1990; Waga et phosphorylated exclusively the p180 and p68 subunits al., 1994) oering the possibility to analyse regulation (Figure 1a; data not shown). Similar activities (adjusted of DNA replication with puri®ed proteins only. with histone H1) of each kinase complex were used to Moreover, biochemical studies of the initiation of phosphorylate DNA polymerase a-primase. Normal- SV40 DNA replication allowed the development of izing kinase activity this way, Cyclin A-Cdk2 kinase an experimental model for the initiation reaction had the highest activity in phosphorylation of the p180 (Challberg and Kelly 1989; Fanning, 1994; Fanning and p68 subunits of DNA polymerase a-primase; and Knippers, 1992; Hurwitz et al., 1990; Stillman, Cyclin A-Cdc2, Cyclin B-Cdc2 and Cyclin E-Cdk2 1989, Wang, 1991). T antigen, RP-A, and DNA complexes were slightly less active in p180 and p68 polymerase a-primase, that are required for initiation phosphorylation (Figure 1a). DNA polymerase a-primase that was pre-incubated with increasing amounts of kinase complexes showed the same DNA polymerase activity on activated DNA Correspondence: H-P Nasheuer as the control enzyme, while its primase activity, that Received 15 February 1996; revised 22 November 1996; accepted 25 November 1996 was measured on poly(dT)-ssDNA, was stimulated up Regulation of DNA replication by phosphorylation C Voitenleitner et al 1612 a Cyclin A-Cdc2 Cyclin A-Cdk2 Cyclin B-Cdc2 Cyclin E-Cdk2 Cdc2 Cdk2 control

195 — — p180 — 130 — 87 — 66 — — p68 — 54 — — p58 — — p48 —

39 —

M PP 1 2 3 4 5 6 7

¨ dT18 ¨ dT12

1 2 3 4 5

Figure 1 Phosphorylated DNA polymerase a-primase by Cyclin-Cdk complexes. (a) DNA polymerase a-primase (lane PP) was incubated with puri®ed Cyclin-Cdk complexes, separated by SDS gel electrophoresis, and stained with Coomassie Brilliant-blue. The subunits, that were phosphorylated in the presence of Cyclin A-Cdc2 (lane 1), Cyclin A-Cdk2 (lane 2), Cyclin B-Cdc2 (lane 3), Cyclin E-Cdk2 (lane 4), Cdc2 (lane 5), Cdk2 (lane 6), and control protein (lane 7), were detected by autoradiography. (b) Equal amounts (0.4 units of primase) of DNA polymerase a-primase were pre-incubated with indicated amounts of Cyclin A-Cdc2 (~), Cyclin A-Cdk2 (&), Cyclin B-Cdc2 (*), Cyclin E-Cdk2 (~), Cdk2 (&), or control protein (6), and then primase activities were determined. (c) DNA polymerase a-primase was treated with Cyclin A-Cdk2 or control protein, and then primase products of pre-treated DNA polymerase a-primase on poly (dT) ssDNA were analysed by denaturing gel electrophoresis. Lane 1 and 2 Cyclin A-Cdk2-treated DNA polymerase a-primase (0.2 and 0.4 primase units, respectively), lane 3 and 4 DNA polymerase a-primase treated with control protein (0.2 and 0.4 primase units, respectively), and 0.4 primase units of untreated DNA polymerase a-primase (lane 5)

to two- to threefold of that of control enzyme (Figure Cyclin A-Cdc2 and Cyclin A-Cdk2 inhibit initiation of 1b, data not shown). The length of the primase SV40 DNA replication in vitro products was not altered by phosphorylation, since phosphorylated and untreated DNA polymerase a- To study whether the initiation step of replication on primase as well as DNA polymerase a-primase treated SV40-origin containing double-stranded DNA with control protein synthesized equivalent products (dsDNA) is modulated by phosphorylation, Cyclin- (Figure 1c). These data suggest that basic enzymatic Cdk complex or control protein were added directly to functions of DNA polymerase a-primase were not or the initiation reaction. In this assay, Cyclin A-Cdc2 or only moderately altered by phosphorylation with Cdks. Cyclin A-Cdk2 signi®cantly reduced the initiation Regulation of DNA replication by phosphorylation C Voitenleitner et al 1613 2 2 2 2 2 Cyclin A-Cdc Cyclin A-Cdk Cyclin B-Cdc2 Cyclin E-Cdk2 control Cyclin A-Cdk Cyclin B-Cdc Cyclin E-Cdk control

active Cdk – + – + – + – + – – – – units 0.4 0.2 0.4 0.2 0.4 0.2 0.4 0.2 0.4

¨ dT18

¨ dT12 ¨ dT12

1 2 3 4 5 6 7 8 9 10 11 12 Arb units 41 5.3 29 4.9 33 23 25 27 22 22 23 Rel Inc.(%) 18 100 15 100 91100124100 95100 1234 5678910M Figure 2 Initiation of SV40 DNA replication in the presence of Arb units 69 0.7 4 18 41 28 58 30 66 Cyclin-dependent kinases. The initiation reaction of dsDNA in Rel Inc.(%)104 2.3 6 60 63 92 88 100 100 the puri®ed system was performed in the presense of immunoanity puri®ed Cyclin A-Cdc2 (lane 2 and 3), Cyclin Figure 3 Initiation of SV40 DNA replication in vitro by A-Cdk2 (lane 4 and 5), Cyclin B-Cdc2 (lane 6 and 7), Cyclin E- phosphorylated DNA polymerase a-primase. For comparing the Cdk2 (lane 8 and 9), or proteins from mock infected cells that initiation activity of the DNA polymerase a-primase complexes, bind to 12CA5 antibodies (control protein, lane 10 and 11). In primase activity of each DNA polymerase a-primase was parallel, the initiation reactions were performed with either active measured shortly before the initiation assay. The initiation (lanes 2, 4, 6 and 8), heat inactivated (10 min at 708C) kinase products of 0.4 units of untreated immunoanity-puri®ed DNA complexes (lanes 3, 5, 7 and 9), or with control protein polymerase a-primase were analysed (lane 1). In parallel, 0.2 and (untreated, lane 10, and heat treated, lane 11). Reaction products 0.4 primase units of Cyclin A-Cdk2- (lane 2 and 3), Cyclin B- were analysed by denaturing gel electrophoresis and autoradio- Cdc2- (lane 4 and 5), or Cyclin E-Cdk2- (lane 6 and 7) graphy. The initiation products of 0.4 units of untreated DNA phosphorylated DNA polymerase a-primase, or DNA polymer- polymerase a-primase are shown in lane 1. Radioactive material ase a-primase treated with control protein (lane 8 and 9) were that is detectable in the absence of DNA polymerase a-primase is incubated under identical conditions and the initiation products were analysed. The radioactive material that is detectable in the shown in lane 12. The length of 5'-end labeled oligo(dT12 ± 18)is indicated at the right by arrows. The amount of initiation absence of DNA polymerase a-primase is shown in lane 10. Lane products was determined by microdensity scanning of the M shows 5' end labeled oligo(dT12 ± 18) markers as indicated at the autoradiograph and displayed at arbitary units (arb units). The right by arrows. The amount of initiation products was amount of unspeci®c radioactive material (lane 12; 0.36 arb units) determined by microdensity scanning of the autoradiograph and was subtracted. In addition, the amount of initiation products in displayed as arbitary units (arb units). The amount of unspeci®c the presence of active kinase was normalized to that in the radioactive material (lane 10; 0.41 arb units) was subtracted. In presence of inactive kinases addition, the amount of initation products synthesized in the presence of 0.2 primase units (lanes 2, 4, 6 and 8) was normalized to those in lane 8, and the amount of products synthesized by enzyme equivalent to 0.4 primase units (lanes 1, 3, 5, 7 and 9) was normalized to those in lane 9 activity of DNA polymerase a-primase to about 15% and 18% of the activity of the inactivated kinase which served as positive control (Figure 2, lanes 2 to 5). In the presence of Cyclin B-Cdc2 the initiation reaction control protein did not in¯uence the initiation of was slightly reduced to about 90% of that in the DNA replication (Figure 3, compare lanes 1 and 9), presence of heat-treated kinase (Figure 2, lanes 6 and since the amount of initiation products that was 7), while with the control protein the amounts of synthesized by 0.4 primase units of untreated or initiation product stayed nearly unchanged (Figure 2, pretreated DNA polymerase a-primase showed about lanes 10 and 11). In the presence of Cyclin E-Cdk2 the 4% dierence between both reactions. However, the initiation activity reproducibly increased to about amounts of initiation products synthesized by Cyclin 120% of the level that was determined in the presence A-Cdk2-phosphorylated DNA polymerase a-primase of heat-treated kinase (Figure 2, lanes 8 and 9). The decreased to about 2 to 6% of those synthesized by the assay that did not contain kinase or control protein control-treated DNA polymerase a-primase (Figure 3, showed reproducibly higher initiation activity than lanes 2, 3, 8 and 9), and the amount of radioactive those that contained kinase or control protein. Since products was only slightly above that of the negative this decrease of initiation activity varied with protein control in Figure 3 lane 10. In contrast, the Cyclin E- preparations, the initiation activity in the presence of Cdk2-phosphorylated DNA polymerase a-primase active kinase was compared to that in the presence of initiated DNA replication nearly as eciently as the heat-inactivated kinase. controls in this experiment, but showed slightly higher To study whether phosphorylation of DNA poly- initiation activity in other experiments (Figure 3, lanes merase a-primase aects initiation of DNA replication, 6 to 9; data not shown). In the experiments presented we pre-phosphorylated DNA polymerase a-primase here, the Cyclin E-Cdk2-treated DNA polymerase a- with Cyclin-Cdks, and then both enzyme complexes primase synthesized a high amount of products with were completely separated from each other. Pre- short length, while the lane with control treated DNA incubation of DNA polymerase a-primase with polymerase a-primase contained more long products, Regulation of DNA replication by phosphorylation C Voitenleitner et al 1614 which are also initiation products under the assay kinases. In vitro the viral initiation reaction was under conditions (Figure 3, lanes 6 to 9; Stadlbauer et al., negative control; viral properties to eliminate initiation 1996). DNA polymerase a-primase that was pre- control by phosphorylation might be absent or inactive phosphorylated by Cyclin B-Cdc2 showed a moder- in this assay composed of puri®ed proteins only. To ately reduced initiation activity (reduction to about overcome some regulatory functions at the initiation two-thirds of that of the control reactions) compared step, the viruses have evolved an additional strategy: to that of DNA polymerase a-primase incubated with The virus supplies one essential initiation protein, the control protein (Figure 3, lanes 4, 5, 8 and 9). viral T antigen, and therefore becomes partially independent from the host initiation machinery and its regulation. Studies to investigate implications of the Discussion presented ®ndings on viral and cellular DNA replication in vivo are currently underway. In this report, we demonstrated that four highly puri®ed Cyclin-Cdk complexes phosphorylated in vitro the p180 and p68 subunits of DNA polymerase a- primase, but not the primase subunits (Figure 1a; Materials and methods Nasheuer et al., 1991). The phosphorylation of DNA polymerase a-primase barely in¯uenced its basic Protein puri®cation enzymatic functions (Figure 1b and 1c; data not shown). These ®ndings are in agreement with previous The S. pombe protein suc1 was puri®ed from lysates of published results that phosphorylation does not or only E. coli BL21 containing pRK172-suc1 (Moreno et al., 1989). The cells were disrupted by sonication, the super- modestly change enzymatic properties of DNA natant of the centrifugated lysates (15 min 20 000 g)was polymerase a-primase (Nasheuer et al., 1991; Podust adjusted to 30% ammonium sulphate, and was applied to et al., 1990; Prussak and Tseng, 1989). These results 50 ml of phenyl Sepharose (Pharmacia, Freiburg (Ger- suggest that the primase and DNA polymerase activity many)). After washing 300 ml of 0.85 M ammonium that are required for Okazaki fragment synthesis sulphate (pH 7.8) the bound proteins were eluted by a during lagging strand synthesis are probably not gradient from 0.85 M to 0 M ammonium sulphate. Proteins modulated by Cdk-dependent phosphorylation. in pooled fractions were precipitated with ammonium A dierent picture emerged, when the initiation sulphate (65%), and then applied to gel ®ltration activity of DNA polymerase a-primase was examined (Sephacryl S100, Pharmacia). 8 on SV40-origin containing dsDNA: Cyclin A-Cdc2 For the puri®cation of Cyclin-Cdk complexes 2.5610 to 36108 High Five (ITC Biotechnology GmbH, Heidelberg), and Cyclin A-Cdk2 inhibited the initiation reaction SF9, or SF9X cells were infected with 10 p.f.u./cell of each (Figure 2, lanes 2 to 5). In contrast, the initiation recombinant baculovirus and incubated for 44 to 48 h at 278C activity of DNA polymerase a-primase was only (Rosenblatt et al., 1992; Desai et al., 1992). Cells were slightly diminished by Cyclin B-Cdc2, and was even homogenized in lysis buer (50 mM HEPES-KOH pH 7.5,

stimulated by Cyclin E-Cdk2 (Figure 2, lanes 6 to 9). 100 mM NaCl, 5 mM KCl, 1 mM MgCl2,5mM NaF, 5 mM Inhibition and stimulation of the initiation reaction by EGTA, 2 mM EDTA, 1 mM DTT, 0.2% Nonidet-P40, 0.1 mM Cdks in vitro were most likely not due to phosphor- Leupeptin, 1% Trasylol1; a generous gift of Bayer Leverkusen ylation of RP-A, since mutation of Cdk phosphoryla- (Germany)) and the complexes were puri®ed by immunoaffi- tion sites RP-A did not change the function of RP-A nity chromatography with the monoclonal antibody 12CA5 in DNA replication (Brush et al., 1994; Henricksen Sepharose 4B or by anity chromatography with suc1- Sepharose 4B using 4 mg/ml of HA-peptide or suc1 for and Wold, 1994). elution. The complexes were dialyzed against 50 mM KPi pH To determine whether the DNA polymerase a- 7.5, 1 mM EDTA, 1 mM 2-mercaptoethanol, 10% glycerol and primase is regulated by Cdks, the enzyme complex stored at 7808C. For control of background activities, was preparatively phosphorylated by these kinases. proteins were puri®ed by suc1-anity or immunoanity Although primase of phosphorylated DNA polymerase chromatography from mock infected cells (control protein). a-primase was fully active on ssDNA (Figures 1c, data DNA polymerase a-primase was prepared as described not shown), Cyclin A-Cdk2-phosphorylated DNA before (Stadlbauer et al., 1994) with slight modi®cations. polymerase a-primase was no longer active in the After binding to monoclonal antibody SJK 237-71 Sephar- initiation reaction (Figure 3, lanes 2 and 3). These ose (Tanaka et al., 1982), the enzyme complex was results suggest that DNA polymerase a-primase might extensively washed with 50 mM Tris ± HCl, pH 8.6, 400 mM NaCl, 150 mM KCl, and 1 mM EDTA. For pre- be controlled by phosphorylation and might catalyse phosphorylation this resin was equilibrated to histone-kinase the initiation reaction on dsDNA and the discontin- buer (20 mM HEPES/KOH pH 7.5, 1 mM DTT, 10 mM uous synthesis on ssDNA in dierent modes. The later MgCl2,4mMEGTA, pH 7.8, 5 mM NaF, 1 mM EDTA, 0.1 interpretation is supported by studies of species-speci®c mg/ml bovine serum albumine (BSA), 0.1 mM ATP) and viral DNA replication (BruÈ ckner et al., 1995; Schneider incubated for 30 min at 378C with kinase complexes that et al., 1994; Stadlbauer et al., 1996). incorporate 50 mmol32P on histone H1 per h. The resin was Our ®ndings suggest that SV40 DNA replication is washed twice with tenfold column volume of buer under phosphorylation control. In vivo, however, SV40 containing 50 mM Tris HCl, pH 8.6, 150 mM KCl, 1 mM DNA replication occurs in an ampli®cation mode EDTA, to remove kinase activity. The enzyme complex was making an inhibitory control obviously unnecessary. eluted by pH shift. Bovine RP-A, Topoisomerase I, and SV40 T antigen were This apparent discrepancy can eventually be explained puri®ed according to Nasheuer et al. (1992) and Moare® et by assuming that papovaviruses have developed al., (1993). The DNA polymerase and primase activity were functions to escape phosphorylation control of cellular measured using activated DNA and poly(dT) ssDNA as proteins, either by inhibiting negatively acting kinases, previously described (Nasheuer and Grosse, 1987, 1988). or by activating phosphatase(s) that counteract these Analysis of primase products was performed as described Regulation of DNA replication by phosphorylation C Voitenleitner et al 1615 (Nasheuer and Grosse, 1988). Initiation reactions were Acknowledgements performed according to BruÈ ckner et al., (1995), and WethankFGrosse,FMuÈller, and S Dehde for critical Stadlbauer et al. (1996), additionally all assays contained reading the manuscript; K Weiûhart for sharing proteins, 5mM NaF. and for useful comments on the manuscript; A Brunahl, VFoÈrster, and M Hauser for excellent technical assistance; D Morgan for generous gift of baculoviruses expressing Kinase assays human Cdc2, Cdk2, Cyclin A, Cyclin B or Cyclin E; J For kinase assays 1 mgDNApolymerasea-primase was Schneider-Mergener for providing HA-peptide; for provid- incubated for 15 min at 378C in histone-kinase buer that ing baculoviruses expressing human p180, p68, p58 and included 1 mCi g-[32P]-ATP per assay and kinase complex as p48, we thank T Wang, C Rehfuess, and A BruÈ ckner, indicated. The proteins were separated by SDS ± PAGE respectively. The ®nancial support of the Deutsche and the incorporation of phosphate was visualized by Forschungsgemeinschaft (Fa 138/5-2, 6-1, Na 190/6-3, autoradiography. For quantitative results the protein and WI 319/11-2 to E Fanning and H-P Nasheuer), NIH bands were excised and the radioactivity was measured (1RO1 GM52948-01) to EF and European Community by liquid scintillation counting. (CHRX-CT93-0248 DG 12) is gratefully acknowledged.

References

BruÈ ckner A, Stadlbauer F, Guarino LA, Brunahl A, Moreno S, Hayles J and Nurse P. (1989). Cell, 58, 361 ± 372. Schneider C, Rehfuess C, Prives C, Fanning E and Nasheuer H-P and Grosse F. (1987). Biochemistry, 26, Nasheuer H-P. (1995). Mol. Cell. Biol., 15, 1716 ± 1724. 8458 ± 8466. Brush GS, Anderson CW and Kelly TJ. (1994). Proc. Nat. Nasheuer H-P and Grosse F. (1988). J. Biol. Chem., 263, Acad. Sci. USA, 91, 12520 ± 12524. 8981 ± 8988. Challberg M and Kelly TJ. (1989). Annu. Rev. Biochem., 58, Nasheuer H-P, Moore A, Wahl AF and Wang TS-F. (1991). 671 ± 717. J. Biol. Chem., 266, 7893 ± 7903. Desai D, Gu Y and Morgan DO. (1992). Mol. Biol. Cell., 3, Nasheuer H-P, Winkler DV, Schneider C, Dornreiter I, 1021. Gilbert I and Fanning E. (1992). Chromosoma, 102, S52 ± Din S-U, Brill SJ, Fairman MP and Stillman B. (1990). Genes S59. Dev., 4, 968 ± 977. Podust V, Bialek G, Sternbach H and Grosse F. (1990). Eur. D'Urso G, Marraccino RL, Marshak DR and Roberts JM. J. Biochem., 193, 189 ± 193. (1990). Science, 250, 786 ± 791. Prussak CE and Tseng BY. (1989). Biochem. Biophys. Res. Dutta A and Stillman B. (1992). EMBO J., 11, 2189 ± 2199. Commun., 159, 1397 ± 1403. Fanning E. (1994). Trends Cell. Biol. 4, 250 ± 255. Rosenblatt J, Gu Y and Morgan DO. (1992). Proc. Natl. Fanning E and Knippers R. (1992). Annu. Rev. Biochem., 61, Acad. Sci. USA, 89, 2824 ± 2828. 55 ± 85. Schneider C, Weiûhart K, Guarino LA, Dornreiter I and Ferrari M, Lucchini G, Plevani P and Foiani M. (1996). J. Fanning E. (1994). Mol. Cell. Biol., 14, 3176 ± 3185. Biol. Chem., 271, 8661 ± 8666. Stadlbauer F, Brueckner A, Rehfuess C, Eckerskorn C, Foiani M, Libergi G, Lucchini G and Plevani P. (1995). Mol. Lottspeich F, FoÈ rster V, Tseng BY and Nasheuer H-P. Cell. Biol., 15, 883 ± 891. (1994). Eur. J. Biochem., 222, 781 ± 793. Henricksen LA and Wold MS. (1994). J. Biol. Chem., 269, Stadlbauer F, Voitenleitner C, BruÈ ckner A, Fanning E and 24203 ± 24208. Nasheuer H-P. (1996). Mol. Cell. Biol., 16, 94 ± 104. Hurwitz J, Dean FB, Kwong AD and Lee S-H. (1990). J. Stillman B. (1989). Ann. Rev. Cell Biol., 5, 197 ± 245. Biol. Chem., 265, 18043 ± 18046. Tanaka S, Hu S-Z, Wang TS-F and Korn D. (1982). J. Biol. Li JJ and Kelly TJ. (1984). Proc. Natl. Acad. Sci. USA, 81, Chem. 257, 8386 ± 8390. 6973 ± 6977. Waga S, Bauer G and Stillman B. (1994). J. Biol. Chem., 269, McVey D, Brizuela L, Mohr I, Marshak DR and Gluzman 10923 ± 10934. Y. (1989). Nature (London), 341, 503 ± 507. Wang TS-F. (1991). Ann. Rev. Biochem., 60, 513 ± 552. Moare® I, Small D, Gilbert I, HoÈ pfner M, Randall S, Schneider C, Russo AAR, Ramsperger U, Arthur AK, Stahl H, Kelly TJ and Fanning E. (1993). J. Virol., 67, 4992 ± 5002.