Journal of Protein Chemistry, Vol. I1, No. 5, 1992
Covalent Attachment of Ribonucleic Acids to Proteins Gernot Cremer, 1 Matthias Kalbas, 1 Hugo Fasold, ~ and Detlef Prochnow ~'2 Received January 31, 1992
As a prerequisite for the synthesis of affinity labels, we describe methods to couple histories to ribonucleic acids. For the synthesis of these covalent hybrid molecules, we used a population of histones HI, H2A, H2B, H3, and H4 from calf thymus and polyadenylic acid with an average chain length of up to 260-280 bases, representing the size of poly(A)-tails from mature mRNAs. Three methods were investigated. (a) Poly(A) containing an 8-N3-A residue was cross-linked to histones by ultraviolet irradiation. (b) The 3'-end of the polynucleotide was connected to a mononucleotide containing an aliphatic amino group, and the resulting poly(A)-derivative was coupled to histones via derivation with a bromoacetyl group. (c) The Y-end of the polynucleotide was oxidized with sodium periodate and bound covalently to an amino group of the polypeptide. To demonstrate the RNA content of the hybrid molecule, the poly(A) was removed with RNase T2. KEY WORDS: Covalent attachment; histones; ribonucleicacid; polyadenylicacid.
1. I N T R O D U C T I O N 3
Since we are working on transport signals and their receptors in nuclear envelopes, and in order to evaluate the relative predominance, we chose polyadenylic acid and histones as examples for covalent attachment at a controlled ratio. Histones contain a nuclear localization signal which directs the polypeptide into the nuclear compartment of eukaryotic cells. Polyadenylic acid constitutes one of the export signals of messenger R N A (Agutter et aL, 1977; Agutter, 1991). Three methods were selected to connect the poly(A) to the protein, and the products were: isolated in a preparative scale. The protein reactivity of nucleic acids can be achieved either by the introduction of modified desoxyribonucleotides (Evans, 1987) or ribonucleotides (Hanna and Meares, 1983) into the polymer. Photoaffinity labels (Hanna, 1989) or the appropriate chemically reactive nucleotide analogues are introduced into nucleic acids (Olomucki et aL, 1984; Rogues and Olomucki; 1987). The synthesis of protein 2'-desoxyribo-oligonucleotides can be achieved automatically (Gibson and Benkovic, 1987). Due to the more complicated synthetic methodology with ribonucleotides, the integration into oligonucleotides
The synthesis of nucleic acid-protein hybrids mostly concerns the attachment of D N A rather than R N A fragments to a protein moiety; nevertheless, the covalent attachment of nucleotides to proteins is a wellknown analytical procedure to identify D N A and R N A binding sites (Stade et al., 1989; Feavers et al., 1988). At the most, only oligonucleotide stretches were connected to proteins. Little work was done to extend the synthesis either to a preparative scale or to high molecular weight compounds as reported previously (Prochnow et al., 1990). The synthesis of analytical as well as preparative amounts of hybrid molecules still presents difficulties which are due to the chemistry of the ribose moiety of ribonucleic acids.
1Institut ffir Biochemie,Johann WolfgangGoethe Universitfit,Klinikum Haus 75A, Theodor Stern Kai 7, 6000 Frankfurt 70, Germany. 2To whom all correspondenceshould be addressed. 3Abbreviations used: AA-UMP, 5-(3-amino)allyl uridine-3'monophosphate; BRAC-UMP, 5-(3-Bromoacetamido)allyl uridine-Y-monophosphate; poly(A), polyadenylicacid; RNase TE, E.C. 3.1.27.1; PNK, polynucleotidekinase E.C. 2.7.1.78; RNA ligase, E.C. 6.5.1.3. 553
0277-8033/92/1001)-0553506.50/0 © 1992 Plenum PuNishing Corporation
554 by enzymatic methods rather than by chemical synthesis was, until now, the only convenient way. (Lemaitre et al., 1987; Beckett and Uhlenbeck, 1984). We chose one chemical and two semi-enzymatic procedures to attach the poly(A) by its 3'-end to amino- and SH-groups of histones in a strict I : 1 ratio.
2. MATERIALS AND METHODS 2.1. Preparation of Histones Histones were prepared from calf thymus as described by Johns (1964).
2.2. Synthesis of Nucleotides AA-UMP was synthesized from Y-UMP as described previously (Langer et al., 1981). 8-azidoadenosine-3'-monophosphate was synthesized with slight modifications according to the procedure described elsewhere (Ikehara et al., 1969). BRACU M P was synthesized from AA-UMP and bromoacetate-N-hydroxy succinimide ester as follows. Nucleotide (43 rag, 0.1 mmol) was dissolved in 15 ml of 5% sodium hydrogen carbonate solution, pH 8.3. The solution was kept on ice. A solution of 236 mg (1.0 mmol) of bromoacetate-N-hydroxy succinimide ester in 10-20 ml dioxane was added and the mixture was kept on ice for I hr. The solution was dried by rotaevaporation and dissolved in a 10 mM ammonium acetate solution, p H 5.0. Again, the solution was dried by rotaevaporation to remove bromoacetic acid. The procedure was repeated several times. The last residue was redissolved in 2 ml of distilled water. The product was identified on HPTLC thin-layer plates NH2 F254S from Merck (Germany); mobile phase 0.2 M NaCI in ethanol : water (30: 70). The compound was detected under UV light and by aqueous 0.1% fluoresceine solution mixed with an equal volume of hydrogen peroxide as a spray (reddish color). The Rfvalue is 0.8. The educt could not be detected.
2.3. Radiolabeling of Histones Histones were acetylated with 3H-acetic anhydride (500 mCi/mmol). For this, 5 mg ofhistones were dissolved in I ml 50 mM Na2HPO4, pH 7.4, containing 30 mM KCI, and 5 mCi of acetic anhydride were added. After 15 min, the soluton was dialyzed against 2 L of the same buffer. The treatment was repeated
Cremer et al. twice. The radioactivity was measured in a scintillation counter and the protein concentration was determined (Bradford, 1976). The specific radioactivity was 0.46 mCi/mg.
2.3.1. Synthesis and Radiolabeling of Covalent Histone-Mononucleotide Hybrid Molecules A complex of histones and BRAC-UMP was synthesized by dissolving 10 mg (~3 × 10 -7 t o o l ) of histones (containing H1, H2A, H2B, H3, and H4) in 100/11 of 50mM sodium hydrogen phosphate, pH 8.0, and incubated with the same amount of BRAC-UMP for 4 hr at 37°C. The mixture was applied on a Sephadex G-10 column. The hybrid molecule in the first peak and the low molecular weight compound (BRAC-UMP) in the second peak were collected and freeze-dried. The samples were treated with PNK (England et al., 1980) in the presence of T-32p-ATP to introduce radioactivity and applied to PAGE.
2.3.2. Synthesis and Radiolabeling of HistonePoly(A ) Hybrid Molecules The complexes of histone and poly(A) were synthesized by the following three procedures: 1. 8-N3-3'-AMP was phosphorylated with the aid of polynucleotide kinase and z-32P-ATP to the 3',5'-diphosphate following the procedure of England et al. (1980) with slight modifications, and was ligated to poly(A) at the 3'-end (Wower et aL, 1988). The polyribonucleotide was washed by repeated precipitations with ethanol:water (2: 1) and mixed with histories (H1, H2A, H2B, H3, and H4) in a ratio of 1 : 1. After irradiation with UV-light at 366 nm, the hybrid molecule was precipitated with ethanol: water (2: 1). The mixture was kept at -20°C for 60 min and centrifuged. The washing/precipitation procedure was repeated several times. The last pellet was applied to a SDS-PAGE. The gel was dried and exposed to a Kodak X-ray film for autoradiography. 2. AA-UMP was phosphorylated, radiolabeled, and ligated to poly(A) (England et al., 1980; Richardson, 1965; Wower et al., 1988). The polymer was washed as described under step 1 and dissolved in 500/tl of 5% sodium hydrogen carbonate in water. The solution was kept on ice and mixed with a 20-fold excess of bromoacetylN-hydroxy succinimide ester in dioxane. After 1 hr, twice the volume of ethanol was added to
Covalent Attachment of Ribonucleic Acids to Proteins the solution to precipitate the polymer. The mixture was cooled at -20°C for 1 hr and centrifuged. The washing step was repeated at least twice. The pellet was dissolved in 50 mM sodium phosphate, pH 7.5, containing 100 mM sodium chloride and adjusted to 1/lg polynucleotide per /~I buffer. This solution was mixed with histones dissolved in the same buffer. The RNA and protein were adjusted to a 1 : 1 molar ratio. The solution was incubated at 37°C for 24 hr and the hybrid molecule was precipitated and washed as described above. The product was used for SDSPAGE. 3. Poly(A) was oxidized with sodium periodate (Belitsina and Spirin, 1979) and linked to an amino group of the histone. The resulting Schiff base was reduced with the aid of sodium cyanoborohydride (Borch and Durst, 1971). For this approach we used a mixture of histones H1, H2A, and H2B. (These fractions were strongly enriched in these appropriate proteins, but they did still contain all the other histones in small amounts as well.) To determine the histone-poly(A) ratio, 3H-labeled histones were linked to 32p-labeled poly(A) (4.5/~Ci/mg) (Richardson, 1965). The resulting complexes were applied to a SDS-PAGE (Laemmli, 1970) and the appropriate bands were excised. The gel pieces were treated with 300/ll of hydrogen peroxide and 30/tl of ammonia at 35°C for 30 min and with 1 ml Protosol (Dupont) at the same temperature for at least 8 hr. The amount of 3H and 32p was determined with the aid of a Beckman LS 3801 Scintilation counter. For separation from unreacted histones and poly(A), the hybrid molecules were run on a preparative SDS-PAGE, stained with Coomassie R-250, and excised from the gel. The hybrid molecules were reeluted from the gel pieces using a Bio-Rad electroeluter model 422 over 5 hr at 60 mA in a solution containing 25 mM Tris base, 192 mM glycine, and 0.1% SDS. The protein-containing solutions were desalted with a Sephadex G-10 column using water as mobile phase. The protein-containing fractions were freeze-dried and redissolved in 50mM Tris-HC1, pH7.4, containing 25mM KC1, 0.5raM CaCI2, 2.5 mM MgC12, 5 mM NaCI, 2.5 mM Na2HPO4, and 5 mM spermidine for the transport measurement. Alternatively, the reeluted histone-poly(A) hybrid molecules were dialyzed directly against the transport buffer.
555 2.4. Digestion of Histone-Poly(A) Hybrid Molecules Two hundred micrograms of hybrid molecule in 10/11 of 10 mM Na-acetate buffer, pH 5.5, were mixed with 20 U of RNase T2 in the same buffer. After incubation at 22°C for 14 hr, the sample was prepared for PAGE (Laemmli, 1970).
3. RESULTS 3.1. Attachment of Poly(A) to Histones Method A: Using the PNK kinase reaction (Wower et al., 1988), mononucleotide-3'-monophosphates were radiolabeled in the Y-position. We chose 8-N3-adenosine monophosphate. The resulting nucleoside Y,5'-diphosphates were attached to poly(A) with the help of RNA ligase (Lemaitre et al., 1987). The procedure is indicated in Scheme I. Azidoadenosine Y-monophosphate is a known substrate for polynucleotide kinase and (after conversion into the appropriate nucleoside-3'-,5'-disphosphate) for polyribonucleotide ligase, too (Wower et al., 1988). The nucleoside 5',3'-disphosphate was attached to the Y-end of poly(A) with an average molecular weight of 100,000 D. This furnishes a poly(A) containing an azido-adenosine residue at the Y-end of the polyribonucleotide. For the formation of polyribonucleotide protein hybrid molecules, this poly(A,8-N3A) was incubated with histones and irradiated with UV-light. Method B: In a comparable procedure, 5-(3-amino)allyl uridine-3'-monophosphate was phosphorylated at the Y-hydroxyl position with the help of r-32PATP and PNK. Again, the attachment of the resulting uridine-3',5'-diphosphate analogue to the 3'-end of poly(A) was achieved with the aid of RNA ligase. This leads to a poly(A,AA-U) derivative containing a single aliphatic amino group in the polymer. This amino group is suitable to attach protein-reactive components. Therefore, poly(A,AA-U) was treated with bromoacetyl-N-hydroxy succinimide ester. The ester reacts predominantly with this mentioned primary aliphatic amino group of the polyribonucleotide (Scheme I). This bromoacetyl group acts as a proteinreactive component and the polyribonucleotide was attached to histones under the appropriate conditions (see Materials and Methods). Figure 1 shows hybrid molecules in lane B with an apparent molecular weight of 33 kD (for explanations see legend of Fig. 1).
C r e m e r et aL
556
IHz
R1 =
(9.0
•
0d~
R2== ~NHI
I C--N~I HC~,I~C'-.N"
I
®.o - pA{'
=
pUp/~f~NHz
RNA-Ligase
/Nj I) n
Ap(AP)nA =
+
pAp
) •
,,Nj Ap (Ap) nApAp
260
RNA-Ligase
,•/"•" NHz 2)
Ap(AP) nA
+
pup
> •
Ap(Ap) n A p U ~
NH'
Broloacetate-NHydroxysucclnimidaat .r
Ap(ap).Apup "v'~'
0
Scheme I. Protein-reactive 3'-mononucleotides (8-N3-adenosine-Y-monophosphate) or the precursors of protein-reactive nucleotides (AA-UMP) were phosphorylated in the Y-position with PNK in the presence of radioactive ATP. The diphosphates were attached at the 3'-end of poly(A) by a ligation step with the aid of RNA ligase. The resulting photoreactive poly(A,8-N3-A) was used for the coupling procedure with histories, while the poly(A,AA-U) was activated with a bromoacetate group prior to the coupling with histones.
--
97.4 66.2
---
42.7
---
31.0
--
21.5
--
14.4
Method C: In a third a p p r o a c h , the Y-end o f p o l y a d e n y l i c acid was oxidized with s o d i u m p e r i o d a t e p r i o r to c o u p l i n g with histones. The p r o d u c t , a dialdehyde, f o r m s Schiff bases with a m i n o g r o u p s o f
Fig. 1. SDS-PAGE, Coomassie stain (Laemmli, 1970). Histone-poly(A) hybrid molecules were synthesized as described under Materials and Methods and applied on the gel. The left lane shows hybrid molecules of histones (H2A, H2B, and H4) and the BRAC-mononucleotide as a control. As demonstrated in Fig. 3A, histones and histone-mononucleotide hybrids are indistinguishable from each other on SDS-PAGE. The center lane shows hybrid molecules of histones (H2A, H2B, and H4) and poly(A). The additional band at 33,000 D presents the histonepoly(A) hybrid molecule. This band contains equimolar amounts of histones and polyadenylic acid. In the right lane, the positions and the molecular weights of marker proteins are indicated. From the top: phosphorylase b, bovine serum albumin, ovalbumin, carbonic anhydrase, soy bean trypsin inhibitor, lysozyme.
histones (see Materials and Methods). These linkages were stabilized by t r e a t m e n t with N a C N B H 3 ( B o r c h and D u r s t , 1971). Results are shown in Fig. 2.
Covalent Attachment of Ribonucleic Acids to Proteins
Fig. 2. SDS-PAGE, Coomassie stain (Laemmli, 1970).Lanes M1 and M2: marker proteins, lane MI, from the top: myosin (200 kD), fl-galactosidase(116.3 kD), phosphorylase b (97.4 kD), bovine serum albumin (66.2kD), ovalbumin (42.7kD). Lane A : Histone H2A- and histone H2B-poly(A) hybrid molecules, 10/zg protein. Lane B: as lane A but in the presence of 20 U RNase T2. Lane C: 20 U RNase T2 as a control. Lane Mz: from the top: phosphorylase b (97.4 kD), bovineserum albumin (66.2 kD), ovalbumin (42.7 kD), carbonic anhydrase (31 kD), soybean trypsin inhibitor (21.5 kD), lysozyme (14.4 kD). The disappearance of a distinct band (arrow) at about 120,000D and the faintness of the band (arrow) at about 33,000D in lane B indicates the cleavage of the poly(A) moiety of the hybrid molecules.
557
qllltlD
M1
3.2. Attachment of 5-(3-Bromoacetamido)Allyl Uridine-3'-Monophosphate (BRAC-UMP) to Histones Figure 3A (PAGE, Coomassie stain) shows the effects of the treatment of a mixture of histones with 3'-BRAC-UMP (for details see Materials and Methods). This nucleotide contains a bromoacetyl group and reacts with SH-groups in histones. The resulting hybrid molecule carries a mononucleotide 3'-monophosphate (3'-BRAC-UMP) residue. Therefore, the hydroxyl group at the Y-position of the B R A C - U M P is accessible for polynucleotide kinase and serves to label the hybrid molecule with r-32p-ATP. The reaction mixture was applied to a Sephadex G-10 column. The protein-containing fraction (Fig. 3, lane A) in the void volume and the mononucleotide-containing fraction (Fig. 3, lane B) were treated with r-32p-ATP and P N K and applied to PAGE. The histones carrying a BRAC-UMP-residue (Fig. 3, lane A) appear as a pattern comparable to histones without any mononucteotide residue (Fig. 3, lane C). The appropriate autoradiograph (Fig. 3B) reveals that the radioactivity is incorporated almost completely into the histone-mononucleotide hybrid (Fig. 3B, lane A). As shown in lane B of Fig. 3B, a negligible amount of the applied radioactivity is left in the low molecular weight fraction. This proves the complete attachment of B R A C - U M P to histones and,
A
B
C
¸
H2
therefore, a 1 : t ratio of B R A C - U M P and histone in the hybrid molecule.
3.3. Identification of Hybrid Molecules To identify histon-poly(A) hybrid molecules, all the appropriate samples were applied to SDS-PAGE. The results are shown on the appropriate Coomassie stained SDS-PAGE (Fig. 1). Histones carrying a single mononucleotide are shown in lane A as a control, while the histone-poly(A) derivative appears in lane B as an additional band with an apparent molecular weight of approximately 33,000 D (Fig. 1). Comparable results are shown in Fig. 2, where additional bands of 33,000 D and 120,000 D indicate the formation of the hybrid molecules (for explanations see legend of Fig. 2).
3.4. Removal of the RNA Moiety of the Hybrid Molecules by RNase Treatment As a further control, the complex was digested with RNase T2 (Fig. 2, lane B). The treatment removes the additional Coomassie stained bands and dembnstrates that the additional protein bands contain RNase T2 labile material.
558
Cremer et aL
M
A
B
C Fig. 3. (A) 12.5% SDS-polyacrylamide gel (Laemmli, 1970). Coomassie stain. Lane M: marker proteins; 5 ~tg each; from the top: phosphorylase b (97.4kD), BSA (66.2kD), ovalbumin (42.7kD), bovine carbonic anhydrase (31 kD), soybean trypsin inhibitor (21.5kD), lysozyme (14.4kD). Histones after reaction with protein reactive BRAC-3'-UMP and treatment with polynucleotide kinase in the presence of r-32p-ATP. The reaction mixture was separated on a Sephadex G10 column and the histonecontaining fraction and the low material-containing fraction were treated as described under Materials and Methods. Lane A: protein-containing fraction (15/lg protein). Lane B: low material compounds. Lane C: untreated histones as a control (30/~g protein per lane). (B) Autoradiograph of Fig. 3A. Lanes A and B correspond to the appropriate Coomassie stain. The radioactivity in the gel was visualizedafter exposing on a Kodak X-ray film. Lane A shows that radioactivity is incorporated almost completely into the histone bands. Lane B shows only traces of radioactiyity in the low material fraction. Therefore, the incorporation of radioactivity into the histones (and the attachment of the mononucleotide) is almost 100%. This corresponds to a 1:1 ratio between mononucleotide and histone.
3.5. Determine of the RNA-Protein Ratio in the Hybrid Molecules T o determine the the histones, we used labeled p o l y ( A ) for molecules were mixed
a m o u n t of p o l y ( A ) b o u n d to 3H-labeled histones a n d 32p_ the synthesis. Both m a c r o in a 1 : 1 ratio ( m o l / m o l ) a n d
the resulting hybrid molecules were purified o n a SDSP A G E . The b a n d s were excised f r o m the S D S - P A G E a n d the a m o u n t of 3H a n d 32p in the h i s t o n e - p o l y ( A ) h y b r i d molecules was determined. Prior to the synthesis, the radiolabeled p o l y ( A ) was separated o n H P L C a n d c o n c e n t r a t e d using a n A m i c o n m i c r o c o n c e n t r a t o r (cutoff M W 10,000 D) ( P r o c h n o w et al., 1990). This
Covalent Attachment of Ribonucleic Acids to Proteins
559
enables us to remove the excess of r-32p-ATP and other low molecular weight compounds completely. We calculated a ratio of 1.1 molecules histones by 1 molecule poly(A) in the 33 kD band (Fig. 1). This applies to a molecular weight of 100,000 D for poly(A) and 15,000D for the histones (average molecular weight).
The other way to activate the polynucleic acid is to attach a protein-reactive 3'-5'-nucleotide diphosphate with the help of RNA ligase. The mononucleotide has to be accepted by the enzyme~ and an excellent candidate is the 8-azido-adenosine analogue (Wower et al., 1988). The nucleotide is activated by UV-irradiation and binds to electron-rich components. The obvious disadvantage is a cross-link yield which is considered to be about 10%. As demonstrated in previous experiments, photoreactive polyadenylic acid shows a high affinity for poly(A) binding sites, but only low specifity for other proteins (Prochnow et al., 1990). This was the reason why we chose the polynucleotide with the bromoacetyl group in our experiments. Protein reactivity was achieved by coupling with bromoacetate-N-hydroxy succinimide ester (Scheme I). The ester reacts with primary aliphatic amino groups, and therefore exclusively with the amino allyl residue of the AA-UMP in the polymer. Figure 1 shows that the complex does appear as an additional higher molecular weight band (i.e., the band with an apparent molecular weight of 33,000 D). This is the resulting major band after the coupling procedure and corresponds to hybrid molecules where single poly(A) molecules with a chain length of about 260-280 nucleotides are attached to single histone molecules. This is supported by the measurement with 3H labeled histones and 32p labeled poly(A), where both the moieties are attached in a 1:1 ratio (see
4. DISCUSSION Our aim was to synthesize covalent histonepoly(A) hybrid molecules where the protein moiety is connected to a single polyadenylic acid molecule. A combination of chemical and enzymatic synthesis was used to generate protein-reactive polyribonucleotides; the activated polynucleic acids were attached to histones. The connection of mononucleotides like BRAC-UMP to histones in comparable experiments was possible, too (Fig. 3A and B). To synthesize covalent hybrid molecules, one of the components has to be activated prior to the coupling step. We activated polyadenylic acid rather than the histones because the primary structure is uniform and side reactions can be avoided more easily. Any activation has to be localized at a definite region of the molecule. We chose the T-end to ensure that the protein, a globular particle, can be linked to the poly(A), which has a rodqike structure (Porschke and Jung, 1985). The T-terminus of the polymer contains vicinal hydroxyl groups, which form a dialdehyde after treatment with sodium periodate (Belitsina and Spirin, 1979). This treatment is gentle enough to avoid any degradation of other parts in the polymer. The activated poly(A) is able to react with primary amino groups or SH-groups, which are present in histones. As an intermediate product, a Schiff base was expected (Borch and Durst, 1971). Treatment with sodium borohydride or sodium cyanoborohydride leads to a tertiary amine; particularly, the cyanoborohydride reacts under mild conditions, avoiding any degradation of the molecule. The resulting histone-poly(A) hybrids are synthesized in high yields, which is demonstrated by the appearance of a 120 kD band in a Coommassie stained gel (Fig. 2). However, as shown in the following paragraphs, the hybrid molecules (1:1 ratio) appear as a 33kD band. The 120 kD band might present a hybrid molecule where several poly(A) tails are attached to a single histone. Therefore, the desired I:1 ratio between poly(A) and histone moiety is doubtful, and those hybrids (120 kD molecular weight) may be not quite suitable for our purposes.
Results).
The bromomethyl group of the poly(A,UU-A) reacts primarily with -SH groups, and this avoids the masking of the nuclear location signal of histones, which is cysteine-free. This procedure seems to be the best way to synthesize histone-poly(A) hybrid molecules. These complexes will be useful tools for studying the relative effects of import and export signals across nuclear pore complexes.
ACKNOW~LEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft; Schwerpunktprogramm: KernZytoplasma-Transport.
REFERENCES
Agutter, P. S. (1991).In Between Nucleus and Cytoplasm, Chapman and Hall, London. Agutter, P. S., Harris, J. R., and Stevenson,I. (1977). Bioehem. J. 162, 671-679.
560 Beckett, D., and Uhlenbeck, O. C. (1984). In Oligonucleotide Synthesis: A Practical Approaeh (Gait, M. J., ed.), IRL Press, Oxford, U.K., pp. 185-205. Belitsina, N. V., and Spirin, A. S. (1979). Methods Enzymol. 60, 745-760. Botch, R. F., and Durst, H. D. (197l). J. Amer. Chem. Soe. 93, 2897. Bradford, W. G. (1976). AnaL Biochem. 72, 353. England, T. E., Bruce, A. G., and Uhlenbeck, O. C. (1980). Methods Enzymol. 65, 65-74. Evans, R. K. (1987).. Bioehemisoy 26, 269-276. Feavers, I. A., McEwan, I. J., Haimin, L., and Jost, J.-P. (1989). J. Biol. Chem. 264, 9114-9117. Gibson, K. J., and Benkovic, S. J. (1987). Nue. Acid. Res. 15, 64556467. Hanna, M. M. (1989). Methods EnzymoL 180, 383-409. Hanna, M. M., and Meares, C. F. (1983). Biochemistry 22, 35463551. Ikehara, M., and Uesugi, S. (1969). Chem. Pharm. Buff. 17, 348.
Cremer et aL Johns, E. W. (1964). Biochem. J. 92, 55-59. Laemmli, U.K. (1970). Nature 227, 680-685. Langer, P. R., Waldrop, A. A., and Ward, D. C. (1981). Proc. Natl. Acad. Sci. USA 78, 6633-6637. Lemaitre, M., Bayard, B., and Lebleu, B. (1987). Proc. Natl. Acad. Sci. USA 84, 648-652. Olomucki, M., Le Gall, J. Y., and Colinart, S. (1984). Tetrahedron Lett. 25, 371-374. Porschke, D., and Jung, M. (1985). J. Biomol. Struet. Dyn. 6, 11731183. Prochnow, D., Riedel, N., Agntter, P. S., and Fasold, D. (1990). J. Biol. Chem. 265, 6536-6539. Roques, P., and Olomucki, M. (1987). Eur. J. Biochem. 167, 103-109. Richardson, C. C. (1965). Proe. Natl. Acad. Sci. USA 54, 158-165. Stade, K., Rinke-Appel, J., and Brimacombe, R. (1989). Nuc. Acid Res. 17, 9889-9908. Wower, J., Hixson, S. S., and Zimmermann, R. A. (1988). Biochem. 27, 8114-8121.
Covalent Attachment of Ribonucleic Acids to Proteins Gernot Cremer, 1 Matthias Kalbas, 1 Hugo Fasold, ~ and Detlef Prochnow ~'2 Received January 31, 1992
As a prerequisite for the synthesis of affinity labels, we describe methods to couple histories to ribonucleic acids. For the synthesis of these covalent hybrid molecules, we used a population of histones HI, H2A, H2B, H3, and H4 from calf thymus and polyadenylic acid with an average chain length of up to 260-280 bases, representing the size of poly(A)-tails from mature mRNAs. Three methods were investigated. (a) Poly(A) containing an 8-N3-A residue was cross-linked to histones by ultraviolet irradiation. (b) The 3'-end of the polynucleotide was connected to a mononucleotide containing an aliphatic amino group, and the resulting poly(A)-derivative was coupled to histones via derivation with a bromoacetyl group. (c) The Y-end of the polynucleotide was oxidized with sodium periodate and bound covalently to an amino group of the polypeptide. To demonstrate the RNA content of the hybrid molecule, the poly(A) was removed with RNase T2. KEY WORDS: Covalent attachment; histones; ribonucleicacid; polyadenylicacid.
1. I N T R O D U C T I O N 3
Since we are working on transport signals and their receptors in nuclear envelopes, and in order to evaluate the relative predominance, we chose polyadenylic acid and histones as examples for covalent attachment at a controlled ratio. Histones contain a nuclear localization signal which directs the polypeptide into the nuclear compartment of eukaryotic cells. Polyadenylic acid constitutes one of the export signals of messenger R N A (Agutter et aL, 1977; Agutter, 1991). Three methods were selected to connect the poly(A) to the protein, and the products were: isolated in a preparative scale. The protein reactivity of nucleic acids can be achieved either by the introduction of modified desoxyribonucleotides (Evans, 1987) or ribonucleotides (Hanna and Meares, 1983) into the polymer. Photoaffinity labels (Hanna, 1989) or the appropriate chemically reactive nucleotide analogues are introduced into nucleic acids (Olomucki et aL, 1984; Rogues and Olomucki; 1987). The synthesis of protein 2'-desoxyribo-oligonucleotides can be achieved automatically (Gibson and Benkovic, 1987). Due to the more complicated synthetic methodology with ribonucleotides, the integration into oligonucleotides
The synthesis of nucleic acid-protein hybrids mostly concerns the attachment of D N A rather than R N A fragments to a protein moiety; nevertheless, the covalent attachment of nucleotides to proteins is a wellknown analytical procedure to identify D N A and R N A binding sites (Stade et al., 1989; Feavers et al., 1988). At the most, only oligonucleotide stretches were connected to proteins. Little work was done to extend the synthesis either to a preparative scale or to high molecular weight compounds as reported previously (Prochnow et al., 1990). The synthesis of analytical as well as preparative amounts of hybrid molecules still presents difficulties which are due to the chemistry of the ribose moiety of ribonucleic acids.
1Institut ffir Biochemie,Johann WolfgangGoethe Universitfit,Klinikum Haus 75A, Theodor Stern Kai 7, 6000 Frankfurt 70, Germany. 2To whom all correspondenceshould be addressed. 3Abbreviations used: AA-UMP, 5-(3-amino)allyl uridine-3'monophosphate; BRAC-UMP, 5-(3-Bromoacetamido)allyl uridine-Y-monophosphate; poly(A), polyadenylicacid; RNase TE, E.C. 3.1.27.1; PNK, polynucleotidekinase E.C. 2.7.1.78; RNA ligase, E.C. 6.5.1.3. 553
0277-8033/92/1001)-0553506.50/0 © 1992 Plenum PuNishing Corporation
554 by enzymatic methods rather than by chemical synthesis was, until now, the only convenient way. (Lemaitre et al., 1987; Beckett and Uhlenbeck, 1984). We chose one chemical and two semi-enzymatic procedures to attach the poly(A) by its 3'-end to amino- and SH-groups of histones in a strict I : 1 ratio.
2. MATERIALS AND METHODS 2.1. Preparation of Histones Histones were prepared from calf thymus as described by Johns (1964).
2.2. Synthesis of Nucleotides AA-UMP was synthesized from Y-UMP as described previously (Langer et al., 1981). 8-azidoadenosine-3'-monophosphate was synthesized with slight modifications according to the procedure described elsewhere (Ikehara et al., 1969). BRACU M P was synthesized from AA-UMP and bromoacetate-N-hydroxy succinimide ester as follows. Nucleotide (43 rag, 0.1 mmol) was dissolved in 15 ml of 5% sodium hydrogen carbonate solution, pH 8.3. The solution was kept on ice. A solution of 236 mg (1.0 mmol) of bromoacetate-N-hydroxy succinimide ester in 10-20 ml dioxane was added and the mixture was kept on ice for I hr. The solution was dried by rotaevaporation and dissolved in a 10 mM ammonium acetate solution, p H 5.0. Again, the solution was dried by rotaevaporation to remove bromoacetic acid. The procedure was repeated several times. The last residue was redissolved in 2 ml of distilled water. The product was identified on HPTLC thin-layer plates NH2 F254S from Merck (Germany); mobile phase 0.2 M NaCI in ethanol : water (30: 70). The compound was detected under UV light and by aqueous 0.1% fluoresceine solution mixed with an equal volume of hydrogen peroxide as a spray (reddish color). The Rfvalue is 0.8. The educt could not be detected.
2.3. Radiolabeling of Histones Histones were acetylated with 3H-acetic anhydride (500 mCi/mmol). For this, 5 mg ofhistones were dissolved in I ml 50 mM Na2HPO4, pH 7.4, containing 30 mM KCI, and 5 mCi of acetic anhydride were added. After 15 min, the soluton was dialyzed against 2 L of the same buffer. The treatment was repeated
Cremer et al. twice. The radioactivity was measured in a scintillation counter and the protein concentration was determined (Bradford, 1976). The specific radioactivity was 0.46 mCi/mg.
2.3.1. Synthesis and Radiolabeling of Covalent Histone-Mononucleotide Hybrid Molecules A complex of histones and BRAC-UMP was synthesized by dissolving 10 mg (~3 × 10 -7 t o o l ) of histones (containing H1, H2A, H2B, H3, and H4) in 100/11 of 50mM sodium hydrogen phosphate, pH 8.0, and incubated with the same amount of BRAC-UMP for 4 hr at 37°C. The mixture was applied on a Sephadex G-10 column. The hybrid molecule in the first peak and the low molecular weight compound (BRAC-UMP) in the second peak were collected and freeze-dried. The samples were treated with PNK (England et al., 1980) in the presence of T-32p-ATP to introduce radioactivity and applied to PAGE.
2.3.2. Synthesis and Radiolabeling of HistonePoly(A ) Hybrid Molecules The complexes of histone and poly(A) were synthesized by the following three procedures: 1. 8-N3-3'-AMP was phosphorylated with the aid of polynucleotide kinase and z-32P-ATP to the 3',5'-diphosphate following the procedure of England et al. (1980) with slight modifications, and was ligated to poly(A) at the 3'-end (Wower et aL, 1988). The polyribonucleotide was washed by repeated precipitations with ethanol:water (2: 1) and mixed with histories (H1, H2A, H2B, H3, and H4) in a ratio of 1 : 1. After irradiation with UV-light at 366 nm, the hybrid molecule was precipitated with ethanol: water (2: 1). The mixture was kept at -20°C for 60 min and centrifuged. The washing/precipitation procedure was repeated several times. The last pellet was applied to a SDS-PAGE. The gel was dried and exposed to a Kodak X-ray film for autoradiography. 2. AA-UMP was phosphorylated, radiolabeled, and ligated to poly(A) (England et al., 1980; Richardson, 1965; Wower et al., 1988). The polymer was washed as described under step 1 and dissolved in 500/tl of 5% sodium hydrogen carbonate in water. The solution was kept on ice and mixed with a 20-fold excess of bromoacetylN-hydroxy succinimide ester in dioxane. After 1 hr, twice the volume of ethanol was added to
Covalent Attachment of Ribonucleic Acids to Proteins the solution to precipitate the polymer. The mixture was cooled at -20°C for 1 hr and centrifuged. The washing step was repeated at least twice. The pellet was dissolved in 50 mM sodium phosphate, pH 7.5, containing 100 mM sodium chloride and adjusted to 1/lg polynucleotide per /~I buffer. This solution was mixed with histones dissolved in the same buffer. The RNA and protein were adjusted to a 1 : 1 molar ratio. The solution was incubated at 37°C for 24 hr and the hybrid molecule was precipitated and washed as described above. The product was used for SDSPAGE. 3. Poly(A) was oxidized with sodium periodate (Belitsina and Spirin, 1979) and linked to an amino group of the histone. The resulting Schiff base was reduced with the aid of sodium cyanoborohydride (Borch and Durst, 1971). For this approach we used a mixture of histones H1, H2A, and H2B. (These fractions were strongly enriched in these appropriate proteins, but they did still contain all the other histones in small amounts as well.) To determine the histone-poly(A) ratio, 3H-labeled histones were linked to 32p-labeled poly(A) (4.5/~Ci/mg) (Richardson, 1965). The resulting complexes were applied to a SDS-PAGE (Laemmli, 1970) and the appropriate bands were excised. The gel pieces were treated with 300/ll of hydrogen peroxide and 30/tl of ammonia at 35°C for 30 min and with 1 ml Protosol (Dupont) at the same temperature for at least 8 hr. The amount of 3H and 32p was determined with the aid of a Beckman LS 3801 Scintilation counter. For separation from unreacted histones and poly(A), the hybrid molecules were run on a preparative SDS-PAGE, stained with Coomassie R-250, and excised from the gel. The hybrid molecules were reeluted from the gel pieces using a Bio-Rad electroeluter model 422 over 5 hr at 60 mA in a solution containing 25 mM Tris base, 192 mM glycine, and 0.1% SDS. The protein-containing solutions were desalted with a Sephadex G-10 column using water as mobile phase. The protein-containing fractions were freeze-dried and redissolved in 50mM Tris-HC1, pH7.4, containing 25mM KC1, 0.5raM CaCI2, 2.5 mM MgC12, 5 mM NaCI, 2.5 mM Na2HPO4, and 5 mM spermidine for the transport measurement. Alternatively, the reeluted histone-poly(A) hybrid molecules were dialyzed directly against the transport buffer.
555 2.4. Digestion of Histone-Poly(A) Hybrid Molecules Two hundred micrograms of hybrid molecule in 10/11 of 10 mM Na-acetate buffer, pH 5.5, were mixed with 20 U of RNase T2 in the same buffer. After incubation at 22°C for 14 hr, the sample was prepared for PAGE (Laemmli, 1970).
3. RESULTS 3.1. Attachment of Poly(A) to Histones Method A: Using the PNK kinase reaction (Wower et al., 1988), mononucleotide-3'-monophosphates were radiolabeled in the Y-position. We chose 8-N3-adenosine monophosphate. The resulting nucleoside Y,5'-diphosphates were attached to poly(A) with the help of RNA ligase (Lemaitre et al., 1987). The procedure is indicated in Scheme I. Azidoadenosine Y-monophosphate is a known substrate for polynucleotide kinase and (after conversion into the appropriate nucleoside-3'-,5'-disphosphate) for polyribonucleotide ligase, too (Wower et al., 1988). The nucleoside 5',3'-disphosphate was attached to the Y-end of poly(A) with an average molecular weight of 100,000 D. This furnishes a poly(A) containing an azido-adenosine residue at the Y-end of the polyribonucleotide. For the formation of polyribonucleotide protein hybrid molecules, this poly(A,8-N3A) was incubated with histones and irradiated with UV-light. Method B: In a comparable procedure, 5-(3-amino)allyl uridine-3'-monophosphate was phosphorylated at the Y-hydroxyl position with the help of r-32PATP and PNK. Again, the attachment of the resulting uridine-3',5'-diphosphate analogue to the 3'-end of poly(A) was achieved with the aid of RNA ligase. This leads to a poly(A,AA-U) derivative containing a single aliphatic amino group in the polymer. This amino group is suitable to attach protein-reactive components. Therefore, poly(A,AA-U) was treated with bromoacetyl-N-hydroxy succinimide ester. The ester reacts predominantly with this mentioned primary aliphatic amino group of the polyribonucleotide (Scheme I). This bromoacetyl group acts as a proteinreactive component and the polyribonucleotide was attached to histones under the appropriate conditions (see Materials and Methods). Figure 1 shows hybrid molecules in lane B with an apparent molecular weight of 33 kD (for explanations see legend of Fig. 1).
C r e m e r et aL
556
IHz
R1 =
(9.0
•
0d~
R2== ~NHI
I C--N~I HC~,I~C'-.N"
I
®.o - pA{'
=
pUp/~f~NHz
RNA-Ligase
/Nj I) n
Ap(AP)nA =
+
pAp
) •
,,Nj Ap (Ap) nApAp
260
RNA-Ligase
,•/"•" NHz 2)
Ap(AP) nA
+
pup
> •
Ap(Ap) n A p U ~
NH'
Broloacetate-NHydroxysucclnimidaat .r
Ap(ap).Apup "v'~'
0
Scheme I. Protein-reactive 3'-mononucleotides (8-N3-adenosine-Y-monophosphate) or the precursors of protein-reactive nucleotides (AA-UMP) were phosphorylated in the Y-position with PNK in the presence of radioactive ATP. The diphosphates were attached at the 3'-end of poly(A) by a ligation step with the aid of RNA ligase. The resulting photoreactive poly(A,8-N3-A) was used for the coupling procedure with histories, while the poly(A,AA-U) was activated with a bromoacetate group prior to the coupling with histones.
--
97.4 66.2
---
42.7
---
31.0
--
21.5
--
14.4
Method C: In a third a p p r o a c h , the Y-end o f p o l y a d e n y l i c acid was oxidized with s o d i u m p e r i o d a t e p r i o r to c o u p l i n g with histones. The p r o d u c t , a dialdehyde, f o r m s Schiff bases with a m i n o g r o u p s o f
Fig. 1. SDS-PAGE, Coomassie stain (Laemmli, 1970). Histone-poly(A) hybrid molecules were synthesized as described under Materials and Methods and applied on the gel. The left lane shows hybrid molecules of histones (H2A, H2B, and H4) and the BRAC-mononucleotide as a control. As demonstrated in Fig. 3A, histones and histone-mononucleotide hybrids are indistinguishable from each other on SDS-PAGE. The center lane shows hybrid molecules of histones (H2A, H2B, and H4) and poly(A). The additional band at 33,000 D presents the histonepoly(A) hybrid molecule. This band contains equimolar amounts of histones and polyadenylic acid. In the right lane, the positions and the molecular weights of marker proteins are indicated. From the top: phosphorylase b, bovine serum albumin, ovalbumin, carbonic anhydrase, soy bean trypsin inhibitor, lysozyme.
histones (see Materials and Methods). These linkages were stabilized by t r e a t m e n t with N a C N B H 3 ( B o r c h and D u r s t , 1971). Results are shown in Fig. 2.
Covalent Attachment of Ribonucleic Acids to Proteins
Fig. 2. SDS-PAGE, Coomassie stain (Laemmli, 1970).Lanes M1 and M2: marker proteins, lane MI, from the top: myosin (200 kD), fl-galactosidase(116.3 kD), phosphorylase b (97.4 kD), bovine serum albumin (66.2kD), ovalbumin (42.7kD). Lane A : Histone H2A- and histone H2B-poly(A) hybrid molecules, 10/zg protein. Lane B: as lane A but in the presence of 20 U RNase T2. Lane C: 20 U RNase T2 as a control. Lane Mz: from the top: phosphorylase b (97.4 kD), bovineserum albumin (66.2 kD), ovalbumin (42.7 kD), carbonic anhydrase (31 kD), soybean trypsin inhibitor (21.5 kD), lysozyme (14.4 kD). The disappearance of a distinct band (arrow) at about 120,000D and the faintness of the band (arrow) at about 33,000D in lane B indicates the cleavage of the poly(A) moiety of the hybrid molecules.
557
qllltlD
M1
3.2. Attachment of 5-(3-Bromoacetamido)Allyl Uridine-3'-Monophosphate (BRAC-UMP) to Histones Figure 3A (PAGE, Coomassie stain) shows the effects of the treatment of a mixture of histones with 3'-BRAC-UMP (for details see Materials and Methods). This nucleotide contains a bromoacetyl group and reacts with SH-groups in histones. The resulting hybrid molecule carries a mononucleotide 3'-monophosphate (3'-BRAC-UMP) residue. Therefore, the hydroxyl group at the Y-position of the B R A C - U M P is accessible for polynucleotide kinase and serves to label the hybrid molecule with r-32p-ATP. The reaction mixture was applied to a Sephadex G-10 column. The protein-containing fraction (Fig. 3, lane A) in the void volume and the mononucleotide-containing fraction (Fig. 3, lane B) were treated with r-32p-ATP and P N K and applied to PAGE. The histones carrying a BRAC-UMP-residue (Fig. 3, lane A) appear as a pattern comparable to histones without any mononucteotide residue (Fig. 3, lane C). The appropriate autoradiograph (Fig. 3B) reveals that the radioactivity is incorporated almost completely into the histone-mononucleotide hybrid (Fig. 3B, lane A). As shown in lane B of Fig. 3B, a negligible amount of the applied radioactivity is left in the low molecular weight fraction. This proves the complete attachment of B R A C - U M P to histones and,
A
B
C
¸
H2
therefore, a 1 : t ratio of B R A C - U M P and histone in the hybrid molecule.
3.3. Identification of Hybrid Molecules To identify histon-poly(A) hybrid molecules, all the appropriate samples were applied to SDS-PAGE. The results are shown on the appropriate Coomassie stained SDS-PAGE (Fig. 1). Histones carrying a single mononucleotide are shown in lane A as a control, while the histone-poly(A) derivative appears in lane B as an additional band with an apparent molecular weight of approximately 33,000 D (Fig. 1). Comparable results are shown in Fig. 2, where additional bands of 33,000 D and 120,000 D indicate the formation of the hybrid molecules (for explanations see legend of Fig. 2).
3.4. Removal of the RNA Moiety of the Hybrid Molecules by RNase Treatment As a further control, the complex was digested with RNase T2 (Fig. 2, lane B). The treatment removes the additional Coomassie stained bands and dembnstrates that the additional protein bands contain RNase T2 labile material.
558
Cremer et aL
M
A
B
C Fig. 3. (A) 12.5% SDS-polyacrylamide gel (Laemmli, 1970). Coomassie stain. Lane M: marker proteins; 5 ~tg each; from the top: phosphorylase b (97.4kD), BSA (66.2kD), ovalbumin (42.7kD), bovine carbonic anhydrase (31 kD), soybean trypsin inhibitor (21.5kD), lysozyme (14.4kD). Histones after reaction with protein reactive BRAC-3'-UMP and treatment with polynucleotide kinase in the presence of r-32p-ATP. The reaction mixture was separated on a Sephadex G10 column and the histonecontaining fraction and the low material-containing fraction were treated as described under Materials and Methods. Lane A: protein-containing fraction (15/lg protein). Lane B: low material compounds. Lane C: untreated histones as a control (30/~g protein per lane). (B) Autoradiograph of Fig. 3A. Lanes A and B correspond to the appropriate Coomassie stain. The radioactivity in the gel was visualizedafter exposing on a Kodak X-ray film. Lane A shows that radioactivity is incorporated almost completely into the histone bands. Lane B shows only traces of radioactiyity in the low material fraction. Therefore, the incorporation of radioactivity into the histones (and the attachment of the mononucleotide) is almost 100%. This corresponds to a 1:1 ratio between mononucleotide and histone.
3.5. Determine of the RNA-Protein Ratio in the Hybrid Molecules T o determine the the histones, we used labeled p o l y ( A ) for molecules were mixed
a m o u n t of p o l y ( A ) b o u n d to 3H-labeled histones a n d 32p_ the synthesis. Both m a c r o in a 1 : 1 ratio ( m o l / m o l ) a n d
the resulting hybrid molecules were purified o n a SDSP A G E . The b a n d s were excised f r o m the S D S - P A G E a n d the a m o u n t of 3H a n d 32p in the h i s t o n e - p o l y ( A ) h y b r i d molecules was determined. Prior to the synthesis, the radiolabeled p o l y ( A ) was separated o n H P L C a n d c o n c e n t r a t e d using a n A m i c o n m i c r o c o n c e n t r a t o r (cutoff M W 10,000 D) ( P r o c h n o w et al., 1990). This
Covalent Attachment of Ribonucleic Acids to Proteins
559
enables us to remove the excess of r-32p-ATP and other low molecular weight compounds completely. We calculated a ratio of 1.1 molecules histones by 1 molecule poly(A) in the 33 kD band (Fig. 1). This applies to a molecular weight of 100,000 D for poly(A) and 15,000D for the histones (average molecular weight).
The other way to activate the polynucleic acid is to attach a protein-reactive 3'-5'-nucleotide diphosphate with the help of RNA ligase. The mononucleotide has to be accepted by the enzyme~ and an excellent candidate is the 8-azido-adenosine analogue (Wower et al., 1988). The nucleotide is activated by UV-irradiation and binds to electron-rich components. The obvious disadvantage is a cross-link yield which is considered to be about 10%. As demonstrated in previous experiments, photoreactive polyadenylic acid shows a high affinity for poly(A) binding sites, but only low specifity for other proteins (Prochnow et al., 1990). This was the reason why we chose the polynucleotide with the bromoacetyl group in our experiments. Protein reactivity was achieved by coupling with bromoacetate-N-hydroxy succinimide ester (Scheme I). The ester reacts with primary aliphatic amino groups, and therefore exclusively with the amino allyl residue of the AA-UMP in the polymer. Figure 1 shows that the complex does appear as an additional higher molecular weight band (i.e., the band with an apparent molecular weight of 33,000 D). This is the resulting major band after the coupling procedure and corresponds to hybrid molecules where single poly(A) molecules with a chain length of about 260-280 nucleotides are attached to single histone molecules. This is supported by the measurement with 3H labeled histones and 32p labeled poly(A), where both the moieties are attached in a 1:1 ratio (see
4. DISCUSSION Our aim was to synthesize covalent histonepoly(A) hybrid molecules where the protein moiety is connected to a single polyadenylic acid molecule. A combination of chemical and enzymatic synthesis was used to generate protein-reactive polyribonucleotides; the activated polynucleic acids were attached to histones. The connection of mononucleotides like BRAC-UMP to histones in comparable experiments was possible, too (Fig. 3A and B). To synthesize covalent hybrid molecules, one of the components has to be activated prior to the coupling step. We activated polyadenylic acid rather than the histones because the primary structure is uniform and side reactions can be avoided more easily. Any activation has to be localized at a definite region of the molecule. We chose the T-end to ensure that the protein, a globular particle, can be linked to the poly(A), which has a rodqike structure (Porschke and Jung, 1985). The T-terminus of the polymer contains vicinal hydroxyl groups, which form a dialdehyde after treatment with sodium periodate (Belitsina and Spirin, 1979). This treatment is gentle enough to avoid any degradation of other parts in the polymer. The activated poly(A) is able to react with primary amino groups or SH-groups, which are present in histones. As an intermediate product, a Schiff base was expected (Borch and Durst, 1971). Treatment with sodium borohydride or sodium cyanoborohydride leads to a tertiary amine; particularly, the cyanoborohydride reacts under mild conditions, avoiding any degradation of the molecule. The resulting histone-poly(A) hybrids are synthesized in high yields, which is demonstrated by the appearance of a 120 kD band in a Coommassie stained gel (Fig. 2). However, as shown in the following paragraphs, the hybrid molecules (1:1 ratio) appear as a 33kD band. The 120 kD band might present a hybrid molecule where several poly(A) tails are attached to a single histone. Therefore, the desired I:1 ratio between poly(A) and histone moiety is doubtful, and those hybrids (120 kD molecular weight) may be not quite suitable for our purposes.
Results).
The bromomethyl group of the poly(A,UU-A) reacts primarily with -SH groups, and this avoids the masking of the nuclear location signal of histones, which is cysteine-free. This procedure seems to be the best way to synthesize histone-poly(A) hybrid molecules. These complexes will be useful tools for studying the relative effects of import and export signals across nuclear pore complexes.
ACKNOW~LEDGMENTS This work was supported by the Deutsche Forschungsgemeinschaft; Schwerpunktprogramm: KernZytoplasma-Transport.
REFERENCES
Agutter, P. S. (1991).In Between Nucleus and Cytoplasm, Chapman and Hall, London. Agutter, P. S., Harris, J. R., and Stevenson,I. (1977). Bioehem. J. 162, 671-679.
560 Beckett, D., and Uhlenbeck, O. C. (1984). In Oligonucleotide Synthesis: A Practical Approaeh (Gait, M. J., ed.), IRL Press, Oxford, U.K., pp. 185-205. Belitsina, N. V., and Spirin, A. S. (1979). Methods Enzymol. 60, 745-760. Botch, R. F., and Durst, H. D. (197l). J. Amer. Chem. Soe. 93, 2897. Bradford, W. G. (1976). AnaL Biochem. 72, 353. England, T. E., Bruce, A. G., and Uhlenbeck, O. C. (1980). Methods Enzymol. 65, 65-74. Evans, R. K. (1987).. Bioehemisoy 26, 269-276. Feavers, I. A., McEwan, I. J., Haimin, L., and Jost, J.-P. (1989). J. Biol. Chem. 264, 9114-9117. Gibson, K. J., and Benkovic, S. J. (1987). Nue. Acid. Res. 15, 64556467. Hanna, M. M. (1989). Methods EnzymoL 180, 383-409. Hanna, M. M., and Meares, C. F. (1983). Biochemistry 22, 35463551. Ikehara, M., and Uesugi, S. (1969). Chem. Pharm. Buff. 17, 348.
Cremer et aL Johns, E. W. (1964). Biochem. J. 92, 55-59. Laemmli, U.K. (1970). Nature 227, 680-685. Langer, P. R., Waldrop, A. A., and Ward, D. C. (1981). Proc. Natl. Acad. Sci. USA 78, 6633-6637. Lemaitre, M., Bayard, B., and Lebleu, B. (1987). Proc. Natl. Acad. Sci. USA 84, 648-652. Olomucki, M., Le Gall, J. Y., and Colinart, S. (1984). Tetrahedron Lett. 25, 371-374. Porschke, D., and Jung, M. (1985). J. Biomol. Struet. Dyn. 6, 11731183. Prochnow, D., Riedel, N., Agntter, P. S., and Fasold, D. (1990). J. Biol. Chem. 265, 6536-6539. Roques, P., and Olomucki, M. (1987). Eur. J. Biochem. 167, 103-109. Richardson, C. C. (1965). Proe. Natl. Acad. Sci. USA 54, 158-165. Stade, K., Rinke-Appel, J., and Brimacombe, R. (1989). Nuc. Acid Res. 17, 9889-9908. Wower, J., Hixson, S. S., and Zimmermann, R. A. (1988). Biochem. 27, 8114-8121.
