The Nucleic Acid Binding Activity of Nucleolar Protein B23.1 Resides ...

THEJOURNALOF BIOLOGICAL CHEMISTRY 0 1994 hy The American Society for Biochemistry and Molecular Biology, Inc

Vol. 269, No. 49, Issue of December 9, pp. 3099430996, 1994 Printed in U.S.A.

The Nucleic Acid Binding Activity of Nucleolar ProteinB23.1 Resides in Its Carboxyl-terminal End* (Received for publication, July 8, 1994, and in revised form, September 28, 1994)

Dunrui Wang, Amy Baumann, Attila Szebeni, and Mark 0. J. Olson4 From the Department of Biochemistry, the University of Mississippi Medical Center, Jackson, Mississippi 39216-4505

Protein B23 is a majornucleolar phosphoprotein proProtein B23 exists asat least two isoforms, B23.1 and B23.2. posed to be a ribosome assembly factor. Protein B23 ex- Both of these have common sequences in their NH,-terminal ists as two isoforms, B23.1 and B23.2, differing only in ends and unique sequences near their COOH termini (Chang their carboxyl-terminal sequences. The interaction of and Olson, 1989). Genomic and cDNA analyses indicate that recombinantlyproduced B23 isoforms with double- the two B23 mRNAs are alternative splicing products from a stranded DNA was studied using gel retardation and single gene (Chang and Olson, 1990). The two B23 isoforms nitrocellulose filter disk assays. Protein B23.1bound have different levels of expression at the mRNA and protein saturably to radiolabeled plasmid DNA. By competition levels with B23.1 being the predominant form in all tissues assays protein B23.1 was also capable of binding RNA examined to date (Wang et al., 1993). The two isoforms also and single-stranded DNA.On the other hand, protein appear to have different subcellular locations: B23.1 is nearly B23.2, the shorter of the two isoforms, was not capable of exclusively nucleolar whereas B23.2 is found in cytoplasmic binding double-stranded DNA. The latter result suggested that the carboxyl-terminal end of B23.1 is essen- and possibly nucleoplasmic fractions (Wang et al., 1993). The tial for DNA binding activity. This was confirmed by par-COOH-terminal end of B23.1 (the unique sequence of B23.1 tial digestion experiments using staphylococcal V8 has 37 aminoacids comparedwith 2 amino acids in B23.2) may protease which showed that a 5-kDa fragment,contain- specify the nucleolar location of the protein. The apparently higher affinity of B23.1 than B23.2 for the ing the carboxyl-terminal end of protein B23.1 retained DNA binding activity similar to that of the parent mol- nucleolus must be due to interaction with some nucleolar comecule. In contrast, a 19-kDa fragment from the amino- ponent. Since the nucleolus is composed primarily of proteins terminal halfof B23.1 didnot bind DNA. Thesequence of and nucleic acids, either of these two classes of macromolecules we have comthe carboxyl-terminal 68 amino acids comprising the is a likely target candidate. In the current study 5-kDa fragmentshowed little, if any, similarity to other pared the interactions of the two B23 isoforms with nucleic proteins, suggesting that this segment contains a previ- acids. It wasfound that proteinB23.1 but notB23.2 was capaously undiscovered nucleic acid binding motif. ble of binding DNA. Furthermore, a 5-kDa carboxyl-terminal fragment of B23.1 also binds DNA. Therefore, the COOH-terminal end of B23.1 appears to be responsible for the nucleic acid binding activity of protein B23.1.

Protein B23 (also called nucleophosmin, numatrin, or N038) is a major nucleolar phosphoproteinthat is proposed to particiEXPERIMENTALPROCEDURES pate in ribosome biogenesis at several different levels (Olson, RecombinantProteinB23Isoforms-RecombinantproteinsB23.1 1990). First, proteinB23 is capable of binding nuclearlocalization signals (Goldfarb, 1988)and has been shown t o shuttle be- and B23.2 used in these studieswere produced in Escherichia coli and al., 1993) tween nuclei and cytoplasm of hybrid cells (Borer et al., 1989). purifiedessentiallyasdescribedpreviously(Umekawaet l l c vector except that the respective cDNAs were inserted into the PET Furthermore, protein B23 is able form to specific complexes with for expression (Novagen). other proteins such as HIV-1 the Rev protein (Fankhauser et al., Gel RetardationAssay-Plasmid pGEM-4Z DNA was labeledby nick 1991). Thus, one of the proposed functions is to serve asa trans- translation (NEN kit) using [CY-~~PI~CTP (3000 mCi/mM). The labeled port vehicle or receptor for ribosomal proteins and other pro- DNA was incubated with recombinant protein B23.1 or B23.2 in TBE teins entering thenucleolus. Second, protein B23 is capable of buffer (90 m~ Tris, 90 mM boric acid, and 2 mM EDTA, pH 8.3) at room binding nucleic acids (Fankhauseret al., 1990) and exhibitshe- temperature for 15 min followedby 0.5% agarose gel electrophoresis in TBE buffer. In some experiments competing nucleic acids such a synlix destabilizing activity (Dumbar et al., 1989). The latter ob- thetic oligonucleotide (29 mer: GGGGAATTCGTTCTCTTCCCAAAGTservation plus theprotein’s association with ribosomal RNA in GGAA) or a mixture of 16 and 23 S E. coli ribosomal RNA (Boehringer the fibrillar and granularregions of the nucleolus (Spectoret al., Mannheim) were added to the binding mixture before electrophoresis. and 1984; Biggiogera et al., 1989) suggests itsinvolvement with as- The DNAinthe gels was fured with 0.7% trichloroacetic acid1h,for sembly of preribosomal RNP’ particles. Finally, recent work the gels were dried using paper towels. The dried gels were subjected to showing that proteinB23 has ribonuclease activity’ implicates autoradiography on x-ray film. Filter Binding Assay-DNA binding by protein B23 was quantified this protein inpreribosomal RNA processing. by the nitrocellulose filter assay (Riggs et al., 1970; Olson et al., 1983). Prior t o use the nitrocellulose filters (0.45pm, Costar) were soaked in * This work was supported in part by National Institutes of Health distilled water for at least 1 h and with TBE buffer for 15 min. RadioGrant 5 R 0 1 GM28349. The costs of publication of this article were labeled DNAwas incubated with recombinant proteins in TBE bufferas defrayed in part by the payment of page charges. This article must described above. The reactions were stopped by filtration through nitherefore be herebymarked“aduertisement”inaccordancewith18 trocellulose filters in a dot blot apparatus (Bio-Rad). The filters were U.S.C. Section 1734 solely toindicate this fact. $ Towhom correspondence should be addressed: Dept.of Biochemis- then washedwith 1 mlofTBEbuffer.Proteinconcentrationswere or by absorbance at 280 try, University of Mississippi Medical Center, 2500 N.State St., Jack- determined by the Pierce Coomassie Blue assay nm (Gill and von Hippel, 1989). The radioactivity in the filters was son, MS 39216-4505. Tel.: 601-984-1500; Fax: 601-984-1501. measured by an Ambis radioanalytic imaging system. The abbreviations used are: RNP, ribonucleoprotein; HPLC, high Western Blots-Following separation of proteins on Laemmli (1970) performance liquid chromatography; base pair, kb, kilobase pair. type gels, the proteins were blotted onto nitrocellulose paper. filters The J. E. Herrera and M. 0. J. Olson, manuscript in preparation.

30994

Carboxyl-terminal End of Protein B23.1

30995

1 2 3 4 5 6 7 -23.1 -9.4 -6.6 .4,4

0

50

100

150

200

823 ( w )

23 -2.0

FIG.1. Gel retardation assay for protein B23-DNA complexes. Recombinant B23isoforms were mixed with 32P-labeledDNA (1.4 ng of pGEM-4Z plasmid) in the presenceor absence of unlabeled competing nucleic acids. TheDNA and DNA-protein complexes were separated by 0.5% agarose gel electrophoresis and visualized by autoradiography. Lane I , labeled DNA with no added proteins; lane 2, labeled DNA incubated with1.5 pg of B23.1; lune 3, B23.1 and labeledDNA plus 250 ng unlabeled pGEM-4Z; lune 4, B23.1 and labeled DNA plus 250 ng of 29-mer synthetic oligonucleotide; lune 5,B23.1 and labeled DNA plus 250 ng of E. coli 16 and 23 S ribosomal RNA; lune 6, labeled DNA plus 1.5 pg of B23.2; lune 7,labeled DNA plus 1.5 pg of bovine serum albumin.

FIG.2. Nitrocellulose filter assay for binding of doublestranded DNA to protein B23 isoforms. Increasing amounts of recombinant protein B23 isoforms were added to 1.4 ng of "P-labeled plasmid DNA pGEM-4Z in binding buffer. 'El protein B23.1; , protein B23.2.

+

0

E0 E

were incubated with anti-B23polyclonal rabbit serum (1:lOO dilution) FOLD COMPETITOR or anti-B23monoclonal antibody (1:10,000 dilution) in buffer A (50 mM FIG. 3. Competition by RNA and single-stranded DNA with Tris-HCI, pH 7.5,200 m~ NaCl and0.10% Triton X-100)containing 5% dry milk. Following three washes with buffer A, the filters were washed binding of double strandedDNA by B23.1. Fixed amounts of B23.1 and incubated with alkaline phosphatase conjugated goat anti-rabbit or (200 ng) and radiolabeled pGEM-4Z (1.4 ng) were incubated with inanti-mouse antiserum IgG fraction (1:2, 500 dilution in buffer A con- creasing amountsof unlabeled competitors. B23.1-DNAcomplexeswere taining 5% dry milk). After successive washes with buffer A, distilled detected by the filter binding assay. CPMICPMo represents ratio of water and 1 M Tris-HC1 (pH 8.6), the filters were incubated with ni- labeled DNA bound in the presenceof unlabeled competitor to labeled DNA bound in the absenceof competitors. The competitors were: Q 1 6 troblue tetrazolium and 5-bromo-4-chloroindolylphosphate solution for + 23 S E.coli ribosomal RNA; +,denatured 16+ 23 S RNA, B,5 S RNA; developing. 0, m13 single-strandedDNA. Fold competitor indicates the ratioof the Proteolytic Digestion of Protein B23.1-Recombinant protein B23.1 competing nucleic acid to pGEM-4Z DNA in nanograms. was incubated withstaphylococcal V8 protease (Sigma) in 50pl of 50 mM NH,HCO, (pH 8.0) at room temperature or a t 37 "C for different times as indicated in the text. tected when B23.2 (Fig. 1, lane 6) was incubated with the Reverse-phase HPLC-The V8 digests of protein B23.1 were applied to a 2.2 x 25-cm C-4 reverse-phase HPLCcolumn (Vydac) equilibrated labeled plasmid DNA. This suggested that protein B23.2 was with 0.1% trifluoroacetic acid. The acetonitrile gradient forthe peptide not capable of binding DNA. To confirm the above findings and to quantify the binding of separation was 10% for 10 minutes,followed by 40% for 120 min, then 60% for 20 min, and 90% for 10 min. The flow rate was 0.5 mumin. B23.1 to DNA, we used the nitrocellulose filter binding assays.

The assay employed plasmid DNA which was 32P-labeled as above. The 32P-labeledDNAs typically had specific activities of Protein B23.1, but Not B23.2, Binds Nucleic Acids-Protein 1x lo7cpdpg. The effect of ionic strength on binding efficiency B23 has previously been shown to bind DNA as well as RNA was tested over a range of salt concentrations (not shown). The (Dumbar et al., 1989; Feuerstein et al., 1990). However, for binding was optimal in 1 x TBE, and this buffer was used in most of the current work, double stranded DNA was used to subsequent experiments. Fig. 2 shows that at low protein constudy the nucleic acid binding activities of B23 isoforms and centrations theradioactivity retained on the nitrocellulose filderived fragments. Plasmid pGEM-4Z was chosen as a source ter was nearly proportional to the amount of protein added, of DNA for the binding assays because its sequence and length reaching saturation at a ratio of approximately 70 pg of are defined and it is relatively easy tolabel with radioactivity. proteidpg of DNA. At saturation, protein B23.1 was capable of DNA gel shift andnitrocellulose filter bindingassays were used binding about 90% of the radiolabeled plasmid DNA contained to detect the interaction betweenprotein B23 and double- in the binding mixture. Using the protein concentration at stranded DNA. half-saturation the apparent dissociation constant was calcuFig. 1shows that protein B23.1 forms a complex with plas- latedto be approximately M. However, when recombinant mid DNA that migrates more slowly than the plasmid DNA protein B23.2 was assayed for DNA binding activityunder the alone. The complex did not appear when excess unlabeled plas- same conditions, very little radioactivity was retained on the mid was added to the binding mixture prior to addition of filters. This confirmed that protein B23.2 was essentiallyincaB23.1. Similarly, excess single-stranded synthetic oligonucleo- pable of binding DNA. tide or E. coli ribosomal RNA prevented the formation of the Previous work in this laboratory (Dumbar et al., 1989) and labeled DNA-protein complex indicating that protein B23.1 the initial gel retardation experiments (Fig. 1) of this study also binds single-stranded DNA and RNA. Fig. 1 also showed suggested that protein B23.1 also interacts with RNA and that when bovine serum albumin was incubated with radiola- single-stranded DNA. To confirm this, competition experiments beled plasmid, there was no shift in mobility, indicating that were performed using nitrocellulose filter binding assays. In the retardation was due to interaction with B23.1 and not any the competition assays, various amounts of unlabeled competiprotein in general. On the other hand, no such shift was de- tors were incubated with a fixed amount of protein B23.1 and RESULTS

Carboxyl-terminal End of Protein B23.1

30996

A

0

2

4

6

8

1 0 1 2 1 4

623.1 OR 5K AMOUNT (pmOl)

FIG.5. Nitrocellulose filter assay for binding of double stranded DNA to 5 kDa peptide and recombinant proteinB23.1. Increasing amounts in pmol of HPLC-purified 5-kDa peptide (0)or protein B23.1 (a)wereaddedto 1.4 ng of 32P-labeledplasmidDNA pGEM in binding buffer.

d 40

5000

50

60

70

00

FRACTION NUMBER

5

4000

3000 2000

lo00 0

NONE

B 32 27.5

-

~

.:

FIG.4. Purification of 5-kDapeptide from V8 protease digest. T w o hundred fifty pg of recombinant protein B23.1 was incubated with V8 protease at room temperature for 8 h. The 5-kDa COOH-terminal fragment of protein B23 was isolated and purified by chromatography on a reverse-phase C-4 HPLC column.A, HPLC chromatographof protein B23.1V8 protease products. Fractions 66-68 contained 5-M)a peptide (5 K).B , Coomassie Brilliant Blue stain of the 5-kDa peptide in fraction 67 after SDS-polyacrylamidegel electrophoresis.

radiolabeled double-stranded DNA in binding bufTer.Fig. 3 shows that protein B23.1 also binds single stranded DNA and ribosomal RNA. DNA from m13 phage as well as 16sand 23s RNA were able to compete with the binding of protein B23.1 to double-stranded DNA. Ribosomal RNA from both prokaryotic and eukaryotic cells had similar competitive abilities against radiolabeled DNA (data not shown). There was little, if any, effect of denaturation on the competition by ribosomal RNA. Furthermore, the four competition curves in Fig. 3 were not significantly different from each other. Thus, there was no detectable preference of B23.1 for a specific type of nucleic acid. A5-kDa COOH-terminal Fragment from B23.1 Contains DNA Binding Activity-The above studies which showed that protein B23.1 but not B23.2 binds DNAsuggested that theDNA binding activity of B23.1 resides in theCOOH-terminal end of the molecule. To test thishypothesis, an attemptwas made to isolate a DNA binding fragment from this partof the molecule. Protein B23.1 was digested for various times with staphylococcal V8 protease (Chan et al., 1986). The DNA binding activity was monitored by filter binding assays as was the presence of the COOH-terminal end of protein B23.1 which is recognized by a monoclonal antibody flung and Chan, 1987; Wang et al.,

dsONA ssONA RNA

FIG.6. Competition of single-stranded DNAand RNAwith double-stranded DNA bindingby 5-kDa peptidefrom recombinant protein B23.1. The 5-kDa COOH-terminal peptideof recombinant protein B23.1 wasisolated as inFig. 5. Six pg of 5-kDa peptide and radiolabeled pGEM-4Z (1.4 ng) was incubated with different competitors (0.5 pg each). B23.1-DNA complexes were detected by the filter binding assay. The competitors are indicated on the bottom.

1993).The optimal time of digestion at room temperature while retaining DNA binding activity was 8 hours (data not shown). This digest was separated by reverse-phase HPLC (Fig. 4A) and thefractions were monitored forDNA binding activity. Of the fractions analyzed, only one (no. 67) had a high DNA binding activity: this fraction contained a fragment with a molecular weight of approximately 5,000 by SDS-gel electrophoresis (Fig. 4B). Amino acid sequencing indicated that this peptide begins at residue 225 and probably contains the 68amino acids at theCOOH-terminal end (data not shown). The HPLC-isolated5-kDa fragment was tested for DNA binding activity using the nitrocellulose filter disc assay in parallel with the parent protein. Fig. 5 shows that on a molar basis both protein B23.1 and the 5-kDa peptide have similar DNA binding properties. In other experiments, fractions 50-52 from the HPLC separation in Fig. 4A which contained predominantly the 19-kDa fragment were tested for DNA binding activity. These had no detectable ability to bind DNA (data not shown). Protein sequencing indicated that the 19-kDa fragment originated from the amino-terminal half of the molecule. This further supports the idea that theCOOH-terminal end of protein B23.1, but not the NH,-terminal region, is responsible for its DNA binding activity. To show that the 5-kDa fragment bound other types of nucleic acids competition experiments were done using unlabeled RNA or single-stranded DNA as competitors in the filter binding assay. Both of the latter unlabeled nucleic acids were able to compete for labeled plasmid DNA (Fig. 6). Thus, the 5-kDa fragment is responsible for the RNA- and singlestranded DNA binding activity of the molecule as well as the double-stranded DNA binding activity. Sequences are now available for B23.1 from five species.To determine the degree of conservation of the COOH-terminal end, the 5 sets of 68 residues comprising the 5-kDa fragment were aligned (Fig. 7). This region of the molecule was highly conserved with 100% identity among the rat, mouse and hu-

30997

Carboxyl-terminal End of Protein B23.1 225

292

R

SFKK--QE-K--TPKTPKGP-SSVEDIKAKNQASIEKGGSLPKVEAKFINYVKNCFRMTDQEAIQDLWQWRKSL

M

SFM--QE-K--TPKTPKGP-SSVEDIKAKMQASIEKGGSLPKVEAKFINYVKNCFRMTDQEAIQDLUQWRKSL

n C

SFKK--QE-K--TPKTPKGP-SSVEDIKAKHQASIEKGGSLPKVEAKFINYVKNCFRM~DQE~IQDLUQWRKSL pdsKk--D-KslTPKTPKvPlS-LEEIKAKMQASVDKGcSLPKLEpKFaNYVMCFRteDQkVIQaLUQWRqTL

X

peqKgkQDtKpqTPKTPKtPlSS-EEIKAKMQtyLDKGnvLPKLEVKFaNYVKNCFReenQkVIeDLWkWRQSLkdgk

FIG.7. Comparison of carboxyl-terminalsequences of protein €323.1 from various species. All sequences were compared torat 5-kDa staphylococcal V8 protease digestion fragment. Capital letters designate identical or conservative substitutions. Lower case letters designate nonconservative substitutions. R = rat B23.1 (Chang andOlson 1989);M = mouse B23.1 (Schmidt-Zachmann and Franke, 1988);H = human B23.1 (Chan etal., 1988);C = chicken B23.1 (Maridor and Nigg, 1990);X = Xenopus B23.1 (Schmidt-Zachmann et al., 1987).

Since no common nucleolar localization sequence has been man sequences. The chicken andxenopus sequences showed 82 and 75% identity with therat sequence, respectively. Thus, this identified so far, it seems likely that each nucleolar proteinhas its own mechanism for nucleolar localization. This localization segment of protein B23.1 is highly conserved. may depend on the existence of receptors or target molecules in DISCUSSION the nucleolus. The present study suggests that the COOHThe studies reported here indicate that thecarboxyl-termi- terminal end of protein B23.1 directs the protein t o a target nal end of protein B23.1 is responsible for the nucleic acid nucleic acid. However, protein B23.1 binds DNAas well as RNA binding activity of the molecule. In fact, a 5-kDa COOH-termi- and the target nucleic acid has not yet been identified. In addition t o being a mechanism for nucleolar localization nal fragment retains nearly allof the DNA binding activity of the parent protein. Isolation of the 5-kDa fragment was pos- the nucleic acid binding activity of B23.1 would be expected to sible because of presence of a protease susceptible segment of have more substantial biological roles. The subnucleolar localthe molecule located near residue 225 of protein B23.1. Diges- ization of B23.1 in both dense fibrillar and granular compotion with V8 protease produces a relatively stable 19-kDa NH,- nents (Biggiogera et al., 1989; Raska and Dundr, 1993) sugof B23.1 is preribosomal RNA and that it terminal fragment and the 5-kDa COOH-terminal fragment. gests that the target Thus, protein B23.1 appears to havea modular structure with may be involved in thetwo stages of ribosome assembly reprethe NH,-terminal end possibly responsible for interactions with sented by these components. The release of protein B23 from the nucleolus by inhibition of preribosomal RNA synthesis proteins which contain nuclear localization signals and the COOH-terminal end retaining the nucleic acid binding activity. (Yung et al., 1985) further supports itsassociation withRNA. It Searches of protein sequence data bases have not revealed is possible to envision that the binding of the COOH-terminal appreciable similarity of the 5-kDa fragment to segments of end of B23.1 to RNA may facilitate other functions of the reother nucleic acid-bindingproteins. Only a short sequence maining portion of the polypeptide chain. One of these funcof NLS-containing proteinsas suggested by (LWQW, residues 285-288) was found in a few nucleic acid- tions is the binding binding proteins including the negative regulator of mitosis Goldfarb (1988); e.g. some ribosomal proteins may be directed (Engle et al., 1990), transposon TN7 transposition protein T to their siteson preribosomal RNA by protein B23.1. A second (Flores et al., 1990) and DNA primase (Marks andWood, 1993). function is the newly discovered ribonuclease activity of the On the other hand, the 5-kDa fragment is relatively rich in protein which appears to be contained in the NH,-terminal aromatic amino acids interspersed among basic residues. Al- portion of the molecule.' This activity may also be directed to its though this combination of classes of amino acids is a charac- target by the COOH-terminal nucleic acid-binding tail of proteristic of segments of certain RNA binding proteins (Dreyfuss tein B23.1. In other words, the COOH-terminal end of B23.1 et al., 19931, there is little or no sequence similarity between may be a means of directing the otherfunctions of the molecule the 5 kDa fragment and these proteins. The net charge of the to their sites of action. 5-kDa fragment is moderately positive with 14 basic and 8 Recent studies suggest that B23.1 also interacts with DNA, acidic residues in thesequence. This feature together with the possibly as a factor in ribosomal DNA replication. Takemura et presence of 6 aromatic residuesmay account for its affinity for al. (1994) found that protein B23.1 stimulated the activity of nucleic acids. It isalso noteworthy that from residue 228 to the DNA polymerase a,but not DNA polymerase p or y. InterestCOOH terminus there is complete conservation of aromatic and ingly, protein B23.2 had no effect on DNA polymerase a,again basic residues among all species for which the sequence of suggesting afunctional role for the COOH-terminal end of B23.1 is known. In any event, thesequence of the 5-kDa frag- B23.1. In earlier studies, Feuerstein et al. (1990) found that ment may representa new nucleic acid binding motif. protein B23 copurified with DNA polymerase (Y primase and Studies on the localization of protein B23.1 and its equiva- suggested that the protein is part of a complex of enzymes lent amphibian oocyte nucleolar protein NO38 showed the im- required for DNA replication. Thus, it is possible that interacportance of the COOH-terminal end of protein B23 in nucleolar tion between the COOH-terminal end of B23.1 and template localization (Peculis and Gall, 1992; Wang et al., 1993). Peculis DNA molecules is important to enhanceor sustain the process and Gall (1992)prepared a series of deletions in theportion of of the DNA replication. A systematic investigation into the the cDNA for protein NO38 coding for its carboxyl-terminal possible preferential binding of B23.1 to specific DNA or RNA end. The oocytes injected with RNA encoding truncated forms sequences will be required to more precisely define its nucleic of NO38 were examined for altered patterns of protein accu- acid binding function. mulation. It wasfound that a domain of about 24 amino acids Acknowledgments-We thank Will Edmonson for preparing recombinear thecarboxyl terminus of NO38 was essentialfor nucleolar nant protein B23 isoforms; Michael 0. Wallace for protein sequencing; localization. It is striking that this domain begins 35 amino Hayato Umekawa, Julio Herrera, andMiroslav Dundr for helpful disacids upstream from the COOH-terminal end of the protein and cussion; and Romie Brown for typing the manuscript, to theuniquesegment of protein isessentiallyequivalent REFERENCES B23.1. Wang et al. (1993) found that protein B23.1 was predominantly located in the nucleoli of Novikoff hepatoma cells, Biggiogera, M., Fakan, S., Kaufmann, S. H., Black,A., Shaper,J. H., and Busch,H. (1989) J . Histochem. Cytochem. 37, 1371-1374 whereas protein B23.2 was found in cytoplasmic fractions after R. A., Lehner, C. F., Eppenberger, H. M., and Nigg, E. A. (1989) Cell 56, cell fractionation. Thus,the COOH-terminal end of B23.1 Borer, 379-390 seems t o be essential for nucleolar localization. Chan, P. K., Chan, W. Y., Yung, B. Y.-M., Cook, R. G., Aldrich, M. B., Ku, D.,

30998

Carboxyl-terminal Endof Protein B23.1

Goldknopf, I. L., and Busch, H. (1986) J . B i d . Chem. 261, 14335-14341 Chan, W.Y., Liu, Q. R., Bojigin. J., Busch, H., Rennert, 0, M., Tease, L. A,, and Chan P. K. (1989) Biochemistry 28, 1033-1039 Chang, J. H., and Olson, M. 0.J. (1989) J. B i d . Chem. 264, 11732-11737 Chang, J. H., and Olson, M. 0.J. (1990) J. B i d . Chem. 266, 18227-18233 Dreyfuss, G., Matunis, M. J., Pinol-Roma, S., and Burd, C. G. (1993)Annu. Rev. Biochem. 62, 289-321 Dumbar, T. S., Gentry, G. A., and Olson, M. 0.J. (1989) Biochemistry 28, 94959501 Engle, D. B., Osmani, S. A,, Osmani, A. H., Rosborough, S., Xiang, X., and Morris, N. R. (1990)J. Biol. Chem. 266, 16132-16137 Fankhauser, C., Izaurralde, E., Adachi, Y., Wingfield, P., and Laemmli, U. K. (1991) Mol. Cell. B i d . 11, 2567-2575 Feuerstein, N., Mond, J. J., Kinchington, P. R., Hickey, R., Kajalainen Lindsberg, M. K., Hay, I., and Ruyechan W. T. (199O)Biochim. Biophys. Acta 1087,127-136 Flores, C., Qadri, M. I., and LichtensteinC. (1990) Nucleic Acids Res.18,901-911 Gill, S. G., and von Hippel, P. H. (1989)AnaL Biochem. 182,319-326 Goldfarb, D. S. (1988) Cell B i d . Int. Rep. 12, 809-832 Laemmli, U. K. (1970) Nature 227,680485 Maridor, G., and Nigg, E. A. (1990) Nucleic Acids Res. 18, 1286 Marks, G. L., and Wood, D. 0.(1993) Gene (Amst.)123,121-125

Olson, M. 0.J. (1990) The EukaryoticNucleus: Molecular Biochemistryand Macromolecular Assemblies (Strauss, P. R., and Wilson, S. H., eds) Vol. 2, pp. 541-546, Telford Press, West Caldwell, N J Olson, M. 0.J., Rivers, 2. M., Thompson, B. A., Kao, W.Y., and Case, S. T. (1983) Biochemistry 22, 3345-3351 Peculis, B. A,, and Gall, J. G . (1992) J. Cell B i d . 116, 1-14 Raska, I., and Dundr, M. (1993) Chromosomes W a y 11, 101-119 Riggs, A. D., Suzuki, H., and Bourgeois, S. (1970) J. Mol. B i d . 48, 67-83 Schmidt-Zachmann, M. S., and Franke, W. W. (1988) Chromosoma (Berl.) 96, 417426 Schmidt-Zachmann, M. S., Hugle-Dorr, B., and Franke, W.W. (1987) EMBO J. 6, 1881-1890 Spector D. L., Ochs, R. L., and Busch, H. (1984) Chromosoma (Berl.)90,139-148 Takemura, M., Ohta, N., Furuichi.Y., Takahashi, T., Yoshida, S., Olson, M. 0. J., and Umekawa, H. (1994)Biochem. Biophys. Res. Commun. 1 9 9 , 4 6 5 1 Umekawa, H., Chan, J. H., Correia, J.,Wang, D., Wingfield, P. T., and Olson, M. 0. J. (1993) Cell. Mol. Biol. Res. 39, 635445 Wang, D., Umekawa, H., and Olson, M. 0. J.(1993)Cell. &Mol. Biol. Res.3 9 , 3 3 4 2 Yung, B. Y.-M., and Chan, P. K. (1987) Biochim. Biophys. Acta 926, 74-82 Yung, B. Y.-M., Busch, H., and Chan, P. K. (1985) Biochim. Biophys. Acta 826, 167-173