Proc. Natl. Acad. Sci. USA Vol. 77, No. 11, pp. 6521-6525, November 1980 Biochemistry

Identification of two factor Y effector sites near the origins of replication of the ColE1 and pBR322 (site specificity/ATP hydrolysis/initiation of DNA synthesis) S. L. ZIPURSKY AND K. J. MARIANS Department of Developmental Biology and Cancer, Division of Biological Sciences, Albert Einstein College of Medicine, Bronx, New York 10461 Communicated by Jerard Hurwitz, August 11, 1980

ABSTRACT The Escherichia coli replication factor Y has MATERIALS AND METHODS been characterized as a OX174 (+) strand specific DNA-de- pendent phosphohydrolase. In conjunction with other E. coil Enzymes and Assay Conditions. Factor Y was purified replication proteins, factor Y is involved in the formation of according to the procedure of Wickner and Hurwitz (10), using heterogeneous primers that are elongated by the E. coli DNA as an assay the differential ATPase activity of factor Y with III elongation machinery. We report here that the cX174 and bacteriophage fd DNAs as effectors. Fraction VII heat-denatured DNAs of plasmids pBR322 and CoIEl serve as (glycerol gradient) was used for all experiments described in effectors for the hydrolysis of ATP by factor Y. The DNA se- this report. One unit of ATPase activity corresponds to the quences of pBB322 responsible for factor Y effector activity have cleavage of 1 ,umol of ATP to ADP and in 30 min at 30'C been localized. Two separate regions of the pBR322 chromo- Pi some support Y ATPase activity. These sequences are near the with OX174 DNA as effector. The specific activity of fraction replication origin and are located on opposite DNA strands. VII was 320 units/mg and represented a 2000-fold purification from the ammonium sulfate fraction. Assay for factor Y ATPase In Escherichia colh three enzymatic mechanisms have been activity was as described (10), with the addition of E. coli characterized that could account for de novo priming on the ssDNA-binding protein (DBP) (a gift of Jack Chase of this in- lagging strand. A primer could be formed by RNA polymerase stitution) at a ratio of 1 ,ug of DBP per 180 pmol of nucleotide (1-3), by the dnaG product acting alone at a unique residues (11). The concentration of DNA in assays was main- chromosomal site (4-6), or by the dnaG gene product directed tained at 0.36 nM with respect to molecules. Standard assays to the lagging strand template by factor Y (7, 8). were 25 ,ul and were incubated at 30'C for 30 or 60 min. The In dtro DNA replication of recombinant DNA plasmids specific activity of the [y-32P]ATP used varied from 10 to 50 containing the leading strand origin of bacteriophage bX174 cpm/pmol. Pi was determined by the method of Conway and RFI (replicative form I) DNA suggested the presence of factor Lipmann (13). Y DNA Restriction enzymes were purchased from New England effector sites on pBR322 (9). Biolabs and were assayed in the buffers suggested by the Factor Y described by Wickner and Hurwitz (10) and by manufacturer. Bacterial alkaline phosphatase (BAPF) was from Shlomai and Kornberg (11) (n' in their nomenclature) is a Worthington and bacteriophage T4-polynucleotide kinase was OX174 specific DNA-dependent ATPase. The latter group from P-L Biochemicals. identified a 54-nucleotide region of the OX174 viral strand that Preparations of DNAs. kX174 and fd (+) strand DNAs were functioned as an effector for factor Y ATPase activity. Pre- purified as described (14). kX174 RFI [3H]DNA was prepared sumably factor Y directs the priming proteins involved in from OX174 am3-infected E. coli HF4704 in the presence of kX174 single-stranded DNA (ssDNA) RF replication to a chloramphenicol at 30 ,ug/ml (9). pBR322 and ColEl unique chromosomal site (11) from which a "mobile " [3H]DNAs were prepared from the E. coil strains CR34 and apparatus is assembled that is capable of synthesizing hetero- JC411, respectively, after amplification overnight in the pres- geneous primers complementary to contiguous DNA sequences. ence of chloramphenicol at 180,ug/ml (9). Bacteriophage fi (12). RFI DNA was the gift of P. Model of the Rockefeller University. We report here that DNAs of both pBR322 and the related DNA fragments were purified preparatively from polyacryl- plasmid CoIEl contain two discrete chromosomal segments that amide gels by electroelution followed by dialysis overnight when denatured support factor Y ATPase activity. Our analysis against 10 mM Tris-HCI, pH 8 at 4°C/10 mM NaCl/1 mM indicates that these segments are located near the plasmid origin EDTA (TEN buffer). Residual acrylamide was removed by of DNA replication. Both segments are also downstream from centrifugation in an Eppendorf microcentrifuge. the replication origin; one is located on the L strand and the other further downstream on the H strand.* We suggest that Abbreviations: Factor Y, E. coli replication factor Y (10); RF, circular the site located on the L strand functions as an origin for lag- duplex DNA molecule (replicative form); RFH, negatively supercoiled ging-strand DNA synthesis in pBR322 replication and that the circular duplex DNA molecules; ssDNA, single-stranded DNA mole- site located on the H strand is involved in converting leading- cule; DBP, E. coil ssDNA-binding protein; (+) strand, viral strand; (-) strand DNA synthesis from a continuous to a discontinuous strand, complementary strand; R-loop, structure in which a "bubble" is formed in duplex DNA by hybridization of an RNA segment to one mode. strand of the duplex. * We refer to the pBR322 DNA leading and lagging strand templates The publication costs of this article were defrayed in part by page as H and L, respectively. In ColEl replication the leading strand charge payment. This article must therefore be hereby marked "ad- template is the H strand and the lagging strand the L strand. The H vertisement" in accordance with 18 U. S. C. §1734 solely to indicate and L designations of the two strands of ColE1 reflect heavy and light this fact. density, respectively. 6521 Downloaded by guest on September 30, 2021 6522 Biochemistry: Zipursky and Marians Proc. Natl. Acad. Sci. USA 77 (1980)

RESULTS Table 1. Factor Y ATPase activity with denatured DNA Factor Y ATPase Activity Dependent on Denatured 4X174 fragments RFI and pBR322 DNAs. As shown in Fig. 1, the ATPase ac- 32pi formed, tivity of factor Y using effector DNA saturated with DBP was DNA effector added nmol highly specific for q5X174 (+) strand DNA compared to fd (+) strand DNA. An identical specificity was evident when heat- OX174 RFI 9.72 denatured RF1 DNA was compared to denatured fI RFI fl RFI 0.10 OX174 pBR322 5.95 DNA. Equimolar amounts of 4X174 (+) strand DNA and de- HindII + Pvu II fragments of pBR322 natured OX174 RFI DNA gave identical rates of ATP hydrolysis A (2065-3906) 3.03 by factor Y (data not shown). Denatured pBR322 DNA sup- B (650-2065) 0.12 ported significant ATPase activity in the presence of factor Y. C (3906-650) 0.12 The major difference between the activity observed with Pst I + Pvu II fragments of pBR322 OX174 RFI and pBR322 DNAs was that the shape of the curve A (3608-2065) 0.09 with the latter DNA was sigmoidal. The reason for this is not B (2065-3608) 2.77 clear, but the shape may reflect differences in the affinity of Hae II fragments of pBR322 factor Y for different effector sites. GolEl DNA gave results A (2718-232) 0.07 essentially identical to those with pBR322 DNA (data not B (1726-2348) 4.29 shown). The ATPase activities with denatured 4X174 RFH DNA E (2348-2718 2.87 and pBR322 DNA were identically sensitive to N-ethylmalei- Hae II fragments of ColEl mide and heat (data not shown). A 0.08 Two Regions of pBR322 and ColEl DNA Support Factor B 0.15 Y ATPase Activity. To identify the region of the pBR322 and C 2.14 ColEl that supported factor Y ATPase activity, we D 0.10 purified a series of DNA fragments of plasmids generated with E 4.53 restriction endonucleases. Denatured fragments were assayed F ND for their ability to support factor Y ATPase activity. The results The indicated DNA preparations and DNA fragments were heat of these experiments are shown in Table 1. denatured. Reaction mixtures (25 ,ul) contained 18 fmol of single- Digestion of pBR322 DNA with a combination of the HindIl stranded molecules with 0.08 unit of factor Y and were incubated for and Pvu II endonucleases yields three fragments, only one of 60 min. DNA fragments were prepared as follows: various amounts (10-30 ug) of pBR322 [3H]DNA (5.7 cpm/pmol) and ColEl [3H]DNA which, the A fragment, had effector activity when denatured. (0.57 cpm/pmol) were digested overnight with the indicated or combinations of enzymes. The digests were heated at 650C for 10 min, extracted with phenol, and precipitated with ethanol. The regions of the gel containing the fragments were excised and the DNA was recovered by electroelution. DNA concentrations were calculated A from the specific activity of the starting DNA. The numbers in pa- rentheses after each pBR322 DNA fragment are the positions of the 15 5' ends of the recognition sequence that bracket the fragment. See ref. 15 for the Hae II cleavage map of CoIEl DNA. ND, not determined.

A Similarly, only one of the two fragments (fragment B) from a double digestion of pBR322 DNA with the endonculeases Pst 0~~ I and Pvu II supported factor Y ATPase activity when dena- 10 tured. These data and the restriction endonuclease cleavage map of pBR322 DNA (Fig. 2) allowed us to localize the effector region to the sequences between nucleotide residues 2065 and 3608 of pBR322 DNA. The lack of equivalent activity when '00 the denatured HindII + Pvu II-A fragment of pBR322 DNA was compared to intact denatured pBR322 DNA on an equi- molar basis was clarified by the next series of experiments. Three of the fragments produced when pBR322 DNA was digested with Hae II (the A, E, and B fragments; Fig. 2) overlap the HindII + Pvu II-A fragment. Of these, the E fragment includes the of pBR322 DNA (17). When these fragments were purified and assayed, two (the E and B 2 fragments) possessed effector activity (Table 1). The B fragment contained a Pvu II recognition site. This, coupled with the data presented above, suggested that the Pvu II endonuclease cleaved within the effector region of the Hae II-B fragment, 0 0.1 0.2 0.3 thereby inactivating it. Factor Y, units To see if a similar situation was present in the ColEl , FIG. 1. ATPase activity of factor Y using different DNAs as ef- the DNA was digested with the Hae TI endonuclease and the fectors. Assays (50 ul) contained 9 fmol of OX174 (+) strand DNA (A), fd (+) strand DNA (0), kX174 RFI DNA (A), pBR322 DNA (-), and five largest fragments (15) were purified. As with pBR322 fl RFI DNA (-). The double-stranded DNAs were denatured by DNA, two Hae II fragments of ColEl DNA were active (Table boiling (10 min) and quick cooling. Factor Y was added as indicated. 1). The Hae II-C fragment of ColEl DNA corresponds to the Assay mixtures were incubated for 60 min. position of the Hae IT-B fragment of pBR322 DNA. the Hae Downloaded by guest on September 30, 2021 Biochemistry: Zipursky and Marians Proc. Nati. Acad. Sd.USA 77(1980) 6523

Table 2. Factor Y effector activity of the denatured pBR322 Hae II-B DNA fragment Restriction enzyme 32Pi formed, added nmol None 2.10 Pvu II 0.25 Hae III 1.50 Hae III + Hpa II 0.10 Hga I 1.56 The indicated digests of the Hae II-B fragment of pBR322 DNA were heat denatured and assayed as described in the legend to Table 1. The Hae II-B fragment of pBR322 DNA (0.12-0.25 ,ug) isolated as described in the legend to Table 1 was treated overnight with the in- dicated restriction enzyme or combinations thereof. Controls were carried through the same procedures except that restriction enzymes were omitted. Half of the digest was used for the assays described above, while the other half was labeled with 32P as described in the legend to Fig. 4, displayed on a polyacrylamide gel, and examined for extent of digestion (data not shown).

FIG. 2. Some restriction enzyme cleavage sites in pBR322 DNA. The numbering system is that ofSutcliff (16). The thick arrow indi- fragment contains all the information required to support factor cates the origin and direction of replication. Y ATPase activity. Localization of the Region Needed to Support Factor Y II-E fragments from both genomes are essentially identical ATPase Activity in the Hae II-E Fragment of pBR322 DNA. upstream but diverge by 15% downstream from the origin of Cleavage of the Hae II-E fragment of pBR322 DNA with the replication (18). The origin of replication is defined as the restriction endonuclease Taq I did not alter factor Y ATPase junction between DNA and RNA in the newly synthesized effector activity. Subsequent purification and evaluation of leading strand (15). For the remainder of this report only effector activity of the two fragments produced by digestion pBR322 DNA will be considered. Results obtained with ColEl of the Hae II-E fragment with the Taq I endonuclease revealed DNA will be reported elsewhere. that only the larger, fragment A (nucleotides 2354-2574), was Localization of the Region Needed to Support Factor Y active (data not shown). ATPase Activity in the Hae II-B Fragment of pBR322 DNA. The region was defined further by digesting this active Taq Because the nucleotide sequence of pBR322 DNA has been I fragment separately and in combination with the Alu I and completely determined (16) there are a large number of re- FnuDII endonucleases. The maximal limits of the effector site striction endonuclease recognition sites catalogued. A detailed in the Hae II-E fragment of pBR 322 DNA were between nu- restriction endonuclease map of the Hae II-E-B region of cleotides 2418 (the Alu I site) and 2574 (the Taq I site) (Table pBR322 DNA is shown in Fig. 3. This knowledge was exploited 3). to localize the sites within this region responsible for effector The Two Regions of Factor Y ATPase Effector Activity activity. Reside in Different Strands of pBR322 DNA. Because factor Table 2 shows the effect of treating the purified Hae II-B Y requires ssDNA for ATPase activity, we determined which fragment of pBR322 DNA with various restriction nucleases. strands of the Hae TI-B and E fragment of pBR322 DNA were Digestion with the Hae III and the Hga I enzymes had essen- active. The two strands of 5'-32P-end-labeled Hae TI-B and E tially no effect, whereas digestion with the Pvu II enzyme and fragments of pBR322 DNA were separated on polyacrylamide a combination of the Hae III and Hpa II enzymes inactivated gels, electroeluted, and assayed for effector function. As shown the ATPase effector activity of the Hae II-B fragment. in Fig. 4, only one strand of each fragment was active. Sequence The Hga I endonuclease-generated fragment that spans analysis of the separated strands (data not shown) followed by roughly the same region as that between the Hae III and Hpa comparison to the sequence of Sutcliff (16) indicated that the II recognition sites within the Hae TI-B fragment of pBR322 two sites were on opposite strands of pBR322 DNA; the site in DNA was purified from a Hga I endonuclease digest of pBR322 the Hae TI-B fragment was on the H strand and the site in the DNA and assayed for effector activity. It was found to be fully Hae II-E fragment was on the L strand. active (data not shown). Therefore, this 170-base-pair-long Table 3. Factor Y effector activity of the denatured Taq I-A Hae III Pvu II Hae III Hae III fragment of the pBR322 DNA fragment Hae II-E Hae II Hga I Hga I HaeII f Hga I HaeII .1 . I . I I I I Restriction enzyme 32p, formed, 1700 1900 2300 2800 TaqI I added nmol Alu I Alu I None 0.37 FnudII Alu I 0.22 FnudII <0.03 HpaII I Hpa II FnudIl + Alu I <0.03 The indicated digests of the Taq I-A fragment of the Hae II-E FIG. 3. Detailed restriction enzyme map in the Hae ]l-B-E region fragment of pBR322 DNA were heat denatured and assayed as de- of pBR322 DNA. Numbering and the positions of the 5' end of the scribed in the legend to Table 1. The digestion procedures and assay restriction enzyme recognition sites are taken from the data ofSutcliff for extent of digestion were as described in Table 2 except that the (16). digestions were performed with 65 ng of fragment. Downloaded by guest on September 30, 2021 6524 Biochemistry: Zipursky and Marians Proc. Natl. Acad. Sci. USA 77 (1980) DNA are well understood, the mechanism of chain elongation subsequent to DNA polymerase I action is unclear. It has been reported that leading-strand DNA synthesis requires the dnaB gene product but not the dnaG gene product (24). These data 0) are difficult to reconcile with in vitro experiments utilizing extracts from temperature-sensitive DNA ligase mutants which indicate that leading-strand synthesis proceeds discontinuously (25). The presence of a factor Y ATPase effector site on the H strand located in close proximity to the DNA polymerase I-DNA polymerase III transition site suggests that this site may 44...L-.....I..*...Lig.olo Io I IO have two functions: (i) it prevents the continuous extension of 0 4 8 12 16 20 0 4 8 12 16 20 the 6sL fragment and (ii) it serves as an initiation point for DNA, fmol subsequent discontinuous DNA synthesis of the L strand. FIG. 4. (A) Factor Y ATPase activity dependent on the separated Fig. 5 diagramatically represents the role we envision for strands ofthe Hae IH-B fragment ofpBR322 DNA. Hae 11-B fragment both factor Y effector sites in ColEl-like DNA replication. By (0.5 ,g) isolated as described in the legend to Table 1 was treated with the combined action of RNA polymerase, RNase H, and DNA alkaline phosphatase, extracted with phenol, and precipitated with is formed that extends past the factor ethanol. T4 polynucleotide kinase was then used to add 32p to the 5' polymerase I, an R-loop end. The fragment was precipitated again with ethanol and redis- Y effector site on the L strand, allowing it to assume an active solved in 0.33 M NaOH. Strands were separated by electrophoresing configuration (Fig. SB). We propose that a priming complex the denatured fragment through a 20 X 20 X 0.15 cm 6.5% polyac- for discontinuous lagging-strand DNA synthesis is assembled rylamide gel (acrylamide to bisacrylamide, 30:1) in 0.1 M Tris-borate on the L-strand of the Hae II-E fragment at the factor Y ef- electrophoresis buffer at 300 V for 3 hr. The indicated amount ofDNA fector site. This complex migrates along the L-strand template was incubated with 0.08 unit of factor Y in a 25-,Ml reaction volume in an overall 5' - 3' direction (Fig. 5B). Unwinding of the for 60 min. 0, H strand; A, L strand. (B) Factor Y ATPase activity dependent on the separated strands of the Hae II-E fragment of double helix in advance of the newly formed replication fork pBR322 DNA. The fragment (0.2 ,g) was treated as above except that will unravel the factor Y effector site on the H strand. We sus- an 8% polyacrylamide gel run at 300 V for 7.5 hr was used to separate pect that the active configuration of this effector site prevents the strands. 0, L strand; O, H strand. continuous extension of the 6sL fragment and serves as a focus for the initiation of discontinuous leading-strand synthesis (Fig. DISCUSSION 5C). The priming complex assembled at this site proceeds in We have demonstrated that in addition to 4X174 (+) strand a 3' -k 5' direction. In this model the 6sL fragment acts only as DNA, denatured ColEl DNA and the denatured DNA of the an origin activator. This type of model can be extended to ac- ColEl-like plasmid pBR322 can serve as effectors for factor Y count for the initiation of replication of the E. coli chromo- ATPase activity. Studies of kX174 ssDNA -o RF replication some. have established that factor Y is essential for priming of the The isolation of factor Y effector sequences near the repli- 4X174 (-) strand DNA synthesis. Factor Y is classified as a cation origin of pBR322 DNA suggests a note of caution on the pre-priming enzyme and appears to be involved in providing of replication origins. The assignment of DNA frag- the priming apparatus an entrance site to the OX174 chromo- ments as replication origins by their ability to direct DNA some (7, 8). The priming apparatus, once assembled [containing synthesis of DNA sequences incapable of self-replication may at least the dnaB and dnaG (primase) gene products], is capable be misleading. These fragments may represent only leading- of mobilizing itself along the single-stranded template, poly- strand origins or incomplete but functional origins. merizing heterogeneous primers (12). Our findings that two We do not know the functional configuration of the factor factor Y effector sites exist in ColEl and pBR322 DNAs may Y effector sites in pBR322 DNA. Shlomai and Kornberg (11) have important implications concerning a possible role(s) for have demonstrated that the site in 4X174 (+) strand DNA is factor Y in duplex DNA synthesis. probably a rather large hairpin. Computer examination of the Initiation of ColEl DNA synthesis [evidence suggests a DNA sequence in the regions of the factor Y effector sites using similar mechanism is likely for ColEl-like plasmids (19)] re- quires the E. coil enzymes RNA polymerase, RNase H, DNA 1600 2000 2400 2800 3200 DNA I In RNase gyrase, and polymerase (20). the presence of A H, RNA polymerase catalyzes the formation of a unique tran- 5' I 3 H script complementary to the plasmid H strand (21). This tran- L script is located at the origin of DNA synthesis and is elongated at least 100 nucleotides by DNA polymerase 1 (20). This early leading-strand replication intermediate is referred to as the 6sL B fragment. Subsequent elongation, however, requires DNA polymerase III (22). Because L and H strand DNA syntheses occur nearly simultaneously (23) it seemed reasonable to us that the displaced parental L-strand in the origin region would C contain a lagging-strand origin. We have demonstrated that this region of the L-strand serves as a factor Y effector. Stauden- bauer et al. (24) have shown that ColEl lagging-strand DNA synthesis requires the dnaB, dnaG, and dnaC gene products. FIG. 5. Model for the initiation of replication of pBR322 and These observations lead us to suggest that lagging-strand DNA ColEl DNAs. The line at the top ofthe diagram indicates the positions ofthe nucleotide residues in pBR322 DNA according to Sutcliff (16). synthesis of CoEl and pBR322 DNAs is initiated within the The arrow in A indicates the position of RNA polymerase initiation plasmid Hae II-E fragments in a fashion similar to the initiation for the 6sL fragment (20). Broken lines indicate RNA; solid lines in- of 4X174 (-) strand DNA synthesis (7, 8). dicate DNA. The ovals on the template strands indicate the positions Although early events in leading-strand synthesis of ColEl of the factor Y effector sites. Downloaded by guest on September 30, 2021 Biochemistry: Zipursky and Marians Proc. Natl. Acad. Sct. USA 77 (1980) 6525

the Sumex-Aim programs revealed no regions of extensive 9. Zipursky, S. L., Reinberg, D. & Hurwitz, J. (1980) Proc. Natl. hairpin formation. Most secondary structure implicit in the Acad. Sci. USA 77,5182-5186. sequence of the effector sites would exist as stems with rather 10. Wickner, S. & Hurwitz, J. (1975) Proc. Nati. Acad. Sci. USA 72, large loops. The computer analysis also revealed no significant 3342-3346. 11. Shlomai, J. & Kornberg, A. (1980) Proc. Natl. Acad. Sci. USA 77, regions of homology among the three effector sites. 799-803. We thank Dr. J. Hurwitz for helpful discussion and a critical reading 12. McMacken, R., Ueda, K. & Kornberg, A. (1977) Proc. Nati. Acad. of the manuscript. Computer analysis was carried out by the use of the Sci. USA 74, 4190-4194. Stanford Molgen Project and the National Institutes of Health 13. Conway, T. & Lipmann, F. (1964) Proc. Natl. Aced. Sci. USA Sumex-Aim facility. This work was supported by Grant IR01 GM26410 52, 1462-1469. from the National Institutes of Health and Grant JFRA-14 from the 14. Franke, B. & Ray, D. (1970) Virology 44, 168-187. American Cancer Society to K.J.M. and by Grant 5RO1 GM13344 from 15. Oka, A. & Takanami, M. (1976) Nature (London) 264, 193- the National Institutes of Health to J. Hurwitz. S.L.Z. was supported 196. by National Institutes of Health Training Grant GM07491-04. 16. Sutcliff, J. G. (1978) Cold Spring Harbor Symp. Quant. Biol. 43, 77-90. 1. Wickner, W., Brutlag, D., Schekman, R. & Kornberg, A. (1972) 17. Tomizawa, J. I., Ohmori, H. & Bird, R. (1977) Proc. Nati. Acad. Proc. Natl. Acad. Sct. USA 69,965-969. Sci. USA 74, 1865-1869. 2. Vicuna, R., Hurwitz, J., Wallace, S. & Girard, M. (1977) J. Biol. 18. Bolivar, F., Betlach, M. C., Heyneker, H., Shine, J., Rodriguez, Chem. 252,2524-2533. R. & Boyer, H. (1977) Proc. Natl. Acad. Sci. USA 74, 5265- 3. Vicuna, R., Ikeda, J.-E. & Hurwitz, J. (1977) J. Biol. Chem. 252, 5269. 2534-2544. 19. Backman, K., Betlach, M., Boyer, H. & Yanofsky, S. (1978) Cold 4. Bouche, J.-P., Zechel, K. & Kornberg, A. (1975) J. Biol. Chem. Spring Harbor Symp. Quant. Biol. 43, 69-76. 250,5995-6001. 20. Itoh, T. & Tomizawa, J.-I. (1980) Proc. Natl. Acad. Sci. USA 77, 5. Bouche, J.-P., Rowen, L. & Kornberg, A. (1978) J. Biol. Chem. 2450-2454. 253,765-769. 21. Tomizawa, J. (1975) Nature (London) 257,253-254. 6. Benz, E., Jr., Reinberg, D., Vicuna, R. & Hurwitz, J. (1980) J. Biol. 22. Staudenbauer, W. (1976) Mol. Gen. Genet. 149, 151-158. Chem. 255, 1096-1106. 23. Tomizawa, J., Sakakibara, Y. & Kakefuda, T. (1974) Proc. Natl. 7. Wickner, S. & Hurwitz, J. (1974) Proc. Natl. Acad. Sci. USA 71, Acad. Sci. USA 71, 2260-2264. 4120-4124. 24. Staudenbauer, W., Scherzinger, E. & Lanka, E. (1979) Mol. Gen. 8. Schekman, R., Weiner, J., Weiner, A. & Kornberg, A. (1975) J. Genet. 177, 113-120. Biol. Chem. 250,5859-5865. 25. Sakakibara, Y. (1978) J. Mol. Biol. 124, 373-389. Downloaded by guest on September 30, 2021