direct photoaffinity labeling of an allosteric site of subunit protein

THE JOURNAL OF BIOLOGICAI. CHEMISTRY

Prrnled in I:. S A. Vol 257. No. 16, Iscue of August 25, pp. 9508-9512, 1982

Direct Photoaffinity Labeling of an Allosteric Site of Subunit Protein M1 of Mouse Ribonucleotide Reductase by dATP EVIDENCE FOR TWO INDEPENDENT BINDING INTERACTIONS WITHIN THE ALLOSTERIC SPECIFICITY SITE*

(Received for publication, February 19, 1982)

Ingrid W. Caras and David W. Martin, Jr. From the Howard Hughes Medical Institute Laboratory and Division of Medical Genetics, Department of Medicine a n d Department of Biochemistry a n d Biophysics, University of California, Sun Francisco, California 94143

The M1 subunit of ribonucleotide reductase contains two kinds of allosteric sites, the activity site and the specificity site, which regulate the overall catalytic activity and the substrate specificity of the enzyme, respectively. The effector nucleotides, dGTP and d'ITP, bind only to the specificity site; dATP and ATP bind to both sites. Partially purified protein M1 was photola- beled specifically after W irradiation in the presence of [32P]dATP. The labeling occurred exclusively at the allosteric specificity site as evidenced by 1) total inhibition of the labeling by dGTP and d"P, 2) normal photoincorporation of [32P]dATP by mutant protein M1 molecules that lack a functional activity site, and 3) co- identity of one-dimensional peptide maps of protein M1 labeled with either [32P]dATP or [32P]drrP.

A mutant protein M1 that is resistant to normal regulation by dGTP and d'ITP (indicating an alteration in the allosteric specificity site) showed normal photoincorporation of [32P]dATP (but not [32P]d'ITP). This labeling was not inhibited by dGTP or d'ITP. Our data suggest that this mutation has altered the binding of dGTP and dTTP but not dATP (or ATP) at the specificity site. Thus, by the combination of genetic and photolabeling techniques, two independent nucleotide binding interactions occurring within this one complex regulatory domain can be distinguished.

Ribonucleotide reductase catalyzes the reduction of all four ribonucleotide diphosphates to provide the deoxyribonucleo- tides required for DNA synthesis (1). The overall enzymatic activity and the substrate specificity are controlled allosteri- cally by nucleoside triphosphate effectors (1). According to the current view, the protein M1 subunit of ribonucleotide reductase contains two kinds of effector binding sites. One kind, the activity site, regulates overall catalytic efficiency by binding either ATP (stimulating) or dATP (inhibiting). The other kind, the specificity site, confers substrate specificity by binding either ATP or dATP (required for pyrimidine reduction), dTTP (required for GDP reduction), or dGTP (required for ADP reduction) (1-3). This model, based on kinetic studies with purified enzymes, has been corroborated by recent genetic evidence for the existence of two independent regulatory domains on protein M1 (4 ) .

We recently reported the use of r3'P]dTTP to photoaffinity label the specificity site of protein M1 ( 5 ) . We now show that

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this far t

protein M1 can be photoaffiity labeled by the effector nucleotide [32P]dATP. Although dATP is known to bind both to the activity and specificity sites of protein M1 (4), we provide evidence that [32P]dATP labels only the specificity site. Ex- periments with a specificity site mutant of protein M1 have allowed us to probe the fine structure of the binding interactions occurring within this complex regulatory domain.

EXPERIMENTAL PROCEDURES

Materials-[c~-~~P]dATP was obtained from Amersham and the radiopurity was verified by thin layer chromatography on polyethyl- eneimine-cellulose (Baker) developed in LiCl (6).

Cell Culture and Mutant Isolation-Mouse lymphosarcoma (S49) cells were maintained in Dulbecco's modified Eagle's medium supple- mented with 10% horse serum as described (7, 8). The isolation and characterization of the mutant line, HAT 1.5A, has been previously described (9). A detailed description of the isolation and characterization of the mutant lines, dGuo-200-1-A (200-1-A), dGuo-200-I-B (200-1-B), dGuo-L-A, and dGuo-L-B will be published elsewhere.

Preparation of Ribonucleotide Reductase Protein MI-The protein M1 subunit of ribonucleotide reductase was partially purified from mouse T-lymphosarcoma (S49) cells as described (4). The pu- rification steps included ammonium sulfate precipitation and affinity chromatography on dextran blue-Sepharose using a KC1-gradient elution (4, 9). Enzyme and protein assays were performed as previously described (4).

Photoaffinity Labeling-Protein M1 in 50 m~ Tris-HCI, pH 7.5, 5 m~ MgC12,2 mM dithiothreitol was mixed with 1 p~ [m3'P]dATP (10 pCi; 1 Ci = 3.7 X IO'" Bq) in 50-100 pl. The mixture was placed as a drop on parafilm on ice and irradiated for 10 min using a General Electric germicidal lamp (G8 shortwave UV lamp) at a distance of 8 cm. The UV dose was 1.8 milliwatts/cm* as determined by an UV meter (black-ray shortwave UV meter 5255, Ultraviolet Products, San Gabriel, CA). After irradiation, 50 nmol of dATP were added and the proteins were precipitated with cold 5% trichloroacetic acid. The precipitates were washed with 5% trichloroacetic acid, dissolved in sample buffer, and analyzed on 7.5% sodium dodecyl sulfate slab gels (5). The stained and dried gels were autoradiographed a t -70 "C using Kodak X-R5 f i i and intensifying screens. The protein M1 (89 kilodaltons) band was identified by co-electrophoresis with pure protein M1 (5). Band intensities were determined by densitometric scanning and integration of the area under the peaks.

One-dimensional peptide mapping and preparation of [35S]methionine-labeled protein M1 were as previously described (5).

RESULTS

[32PJdATP Labels the Protein MI Subunit of Ribonucle- otide Reductase-Ultraviolet irradiation of mixtures contain- ing [w3'P]dATP and partially purified protein M1 from wild type S49 cells resulted in photolabeling of an 89-kilodalton protein (Fig. 1, lane I ) with the kinetics shown in Fig. 2. Evidence that the labeled protein is protein M1 is provided by competition experiments and peptide mapping described below.

9508

Photoaffinity Labeling of Ribonucleotide Reductase Protein M I 9509

1 2 3 4 5 6 7 8

FIG. 1 . Photoaffinity labeling of partially purified protein M1 with [:"P]dATP. Protein M I from wild type S49 cells was nlixeti with 1 ILM ["l']dA'l'I' (10 pCi/(iO 111) and 1JV-irradiated as descrihcd under "Experimental I'rocedures." The laheled proteins were analyzed hv sodium dodecyl sulfate-gel electrophoresis and autoradiog- raphy. 1. no addition; 2, d(;?'l' (,%) pM): : I . (i'rl' ( j 0 pM): 4. d'yr1' (50 p ~ ) : 5 , thvmidine (50 p ~ ) : 6'. dCTI' (50 p ~ ) : 7. deoxyadenosine (:%X0 p ~ ) ; 8, ATP (380 p ~ ) . The 89-kilodalton protein M1 hand is indicated by an urrow.

(m id

with [''"PldATP. Mixtures o f protein M 1 and [~"I'](tATI' ( 1 ILM, 1 0 FIG. 2. Time course of photoaffinity labeling of protein M1

pCi/60 pl) were CJV-irradiated and then analvzed as described under "Experimental I'rocedures." Hand densities were quantitated hy densitometric scanning of the autoradiogram.

Competition by Nucleotides a n d Nucleosides-Fig. 1 shows the effects of added nucleotides and nucleosides on the labeling of protein M1 by [,'"P]dATP. A quantitative interpretation of these data is given in Table I. Low concentrations of dATP (less than 50 PM) decreased the labeling as expected, indicating that the radiolabel indeed labels a dATP-binding site and that the binding is saturable. The labeling of protein M1 by ['"PI dATP was inhibited by the allosteric effector nucleotides, dGTP and dTTP, which are known from kinetic (2, 3 ) and genetic studies (4) to compete for binding a t a single regulatory site, the substrate specificity site. The labeling of protein M1 was not significantly affected by ATP (less than 350 PM) or by nucleotides and nucleosides which play no role in the regulation of ribonucleotide reductase (see Table I).

(,"'P]dATP Labeling of Mutant Protein M1 Molecules That Lack a Functional Acticity Site-The above data (competition by dGTP and dTTP) suggested that [,"'P]dATP labels primarily the specificity site of protein M1. To assess the

?'ABLE I Effect of nucleotides and nucleosides on ["'P]dATI' luheling of

protein M I Partially purified protein M 1 from wild tyDe S49 cells was irradiated

at 2% nm in the presence of 0.4 p~ ["'I']dATI' and the indicated additions for 10 min at 0 "C and analyzed by sodium dodecyl sulfate- gel electrophoresis. The level of labeling was determined hy densiro- metric scanning of autoradiograms as described under '%xperimental Procedures."

- ~~ ~ ~

HelativCc'IJl

prowin .M 1 Addition Concentration dA7'I' latwling of

_ ~ _ _ _ _ _ _ _ ~ _ ~ ~~ ~"

P .n None 100 dATI' 14 60

50 4 100 0

ATP 100 30 1'5

112 Ileoxyadenosine 30 loo

50 90 100 91

dGTP 12 3 2 5 1

GTP 150 86 Ileoxyguanosine 25 9r) dTTP 25 3

50 6 Thymidine 50 70 dCT1' 50 83

___-______- " .. " ~~

contribution of the activity site to the photoincorporation of [""PIdATP, we carried out a series of experiments using mutants of protein M1 that lack a functional activity site. Mutant protein M1 was prepared from two independently selected mutant cell lines, 200-1-A and dGuo-L-A-B, which have the following properties: 1) they are resistant to deoxyadenosine toxicity, 2) they contain altered ribonucleotide reductase (protein M1) molecules which do not respond normally to regulation by dATP; 3 ) they express only one protein M1 allele, producing exclusively mutant protein M1. In addition, wild type protein M1 was prepared from two analogous cell lines, 200-1-R and dGuo-L-A, which also express only one protein M1 allele, in these cases, the wild type allele. (A detailed description of the isolation and characterization of these cell lines will be published elsewhere.) The effect of dATP on the CDP-reductase activity reconstituted from protein M1 preparations of these cell lines is shown in Fig. 3. Mutant protein M1 from 200-1-A and dGuo-L-A-B cells was resistant to inhibition by dATP, while wild type protein M1 from 200-1-B and dGuo-L-A cells was normally sensitive. However, ['"PI dATP labeled the mutant and wild type protein M 1 preparations with equal efficiency. Fig. 4 shows the ['"PIdATI' labeling of the dGuo-LA (wild type) and dGuo-L-A-€3 (mutant) protein M1 preparations. A similar result was o5tained with protein M1 from 200-1-A and 200-1-B cells (data not shown). In addition, none of the preparations showed significant residual labeling after blocking of the specificity site with dGTP (also see below). Since genetic loss of a functional allosteric activity site does not appear to affect the labeling of protein M1 by ['"PIdATP, we conclude that ['"P]dATP does not label the activity site.

Inhibition of /'"PpIdATP Labeling by dGTP-Although both nucleotide competition experiments and experiments with activity site mutants of protein MI suggested that ['"PI dATP labels the allosteric specificity site rather than the activity site, we considered the possibility that there might exist low level labeling at the activity site. Such labeling would presumably be resistant to inhibition by dGTP. Fig. 5B shows the effect of dGTP on the labeling of protein M1 by 1 PM ['l"P]

95 10 Photoaffinity Labeling of Ribonucleotide

FIG. 3 . Effect of dATP on the CDP reductase activity reconstituted from protein M 1 from dGuo-200-I.-A, dGuo-LA-B, 200-1-A, and 200-1-B cells. Oextran blue-Sepharose-fractionated protein M2 (0.25 mg) from wild type cells and protein MI (0.04-0.06 mg) prepared from mutant cells were added to each assay. CI)I'- reductase activity was assaved as previously described (9). I'rotein MI was from the cell lines dGuo-L-A (0). dGuo-L-A-R (O), 200-I-A (A), and 200-1-B (A).

1 2 3 4

FIG. 5

FIG. 4 . Effect of an altered activity site on the photolabeling of protein M 1 by ["'PPjdATP. I'rotein MI was labeled with [ . " P I dATI' as described under "Experimental Procedures." 1 and 2, wild type protein M I from dGuo-L-A cells: 3 and 4, mutant protein M I from dGuo-L-A-B cells. Additions: 2 and 4, 20 CLM dGT1'; I and .3, 20 p~ deoxvguanosine (added to control for nonspecific IJV absorption effects of purine nucleotide). An arrow indicates the position of the protein M 1 band.

dATP. As controls, we measured the effects of dGTP on [:,'PI dTTP labeling of protein M1 (occurring exclusively at the specificity site ( 5 ) ) (Fig. SA) and on ['"PIdATP labeling of the mutant protein from 200-1-A cells (lacking a functional activity site) (Fig. 5C). A quantitative interpretation of the data, obtained by scanning the autoradiograms, is shown below each panel. The inhibition of labeling by dGTP was identical in all three cases. The level of labeling at the highest concentration of dGTP was less than 0.5% of the control value. We were unable to demonstrate significant dGTP-resistant labeling of protein M1 by using concentrations of ['"PIdATP as high as 40 PM, for periods as long as 30 min and at tempera- tures as low as -80 "C (data not shown).

One-dimensional Peptide Mapping of [:"'P]dATP-labeled and [:'"P]dTTP-labeled Protein MI-We previously dem- onstrated that proteolytic digestion of ['"PIdTTP-labeled protein M1 produced a single labeled peptide (5). We wished to

Reductase Protein MI

A B C 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

0 50 150 150 150 dGTP (pM)

Inhibition of Dhotolabeline hv dGTP. I'hotolabelinp. analysis. and quantitation of the tleta were as descrihed under "Ex- perimental I'rocedures." A. protein MI from wild type cells was labeled with ['"I']dTTl' ( I p ~ . 1.4 pCi/60 pl); B, protein MI from wild type cells was labeled with [~"I']dATI' ( 1 p ~ . IO pCi/(iO PI): C, protein M I from 200-I-A cells was labeled with ["l'IdATI' ( I PM. I O pCi/60 pl). DeoxyGTP was added as follows: 1, no addition: 2. 10 p ~ : 3 , 25 p ~ ; 4, 50 p ~ , and 5 , 150 PM, The protein MI hand is indicated by an arrow.

_, -

determine whether digestion of ['"PIdATP-labeled protein M1 produced a similar labeled peptide.

Protein M1, labeled with either ['"PIdATP or ['"PIdTTP, was cut from an acrylamide gel and digested with increasing concentrations of Staphylococcus aureus (V8) protease. The digestion patterns, analyzed by acrylamide gel electrophoresis, were identical (Fig. 6). Exhaustive digestion of protein M1 with V8 protease (30 pg) produced a single ["'PIdATP-labeled peptide which appeared to be identical with the ['"PIdTTP- labeled peptide. By contrast, similarly digested ["'SJmethio- nine-labeled protein M1 gave rise to a number of labeled peptides. This result suggests that ["'PIdATP labels a single site on protein M1 that is identical with, or closely related to, the site of ['"PIdTTP labeling. This also provides direct evidence that the ['"PIdATP-labeled 89 kilodalton protein is protein M1.

[:'"P]dATP labeling of a Mutant Protein MI With an Altered Specificity Site-Having shown that ["'PIdATP labels exclusively the specificity site of protein M1, we anticipated that specificity site mutants of protein M1 would show reduced photoincorporation of ['''PIdATP. We previously described the isolation and properties of a mutant cell line, HAT 1.5-A, resistant to the toxic effects of thymidine and deoxyguanosine (9). Mutant protein M1 molecules from these cells do not respond to normal allosteric regulation by dGTP or

Photoaffinity Labeling of Ribonucleotide Reductase Protein MI 9511

30 K-t

20 K*

14 K-t

5 .7K-c

FIG. 6. One-dimensional peptide mapping of ["PIdATP-labeled, [3ZP]dTTP-labeled, and ['%]methionine-labeled protein M1. Labeled protein M1 bands were cut from a dried acrylamide gel, equilibrated with 0.125 M Tris-HCI, pH 6.8, 107 glycerol, 1 mM 2- mercaptoethanol, 1 mM EDTA, 0.15"; sodium dodecvl sulfate, and digested with Staphylococcus aureus V8 protease at 22 "C overnight. The eluted peptides were separated on a 7.5% acrylamide. 7 M urea gel having an acrylamide/bis-acrylamide ratio of 101. 2-3, ["]dATI'- labeled protein M1 digested with 0, 10, and 30 pg of V8 protease; 4-7, ['"PIdTTP-labeled protein M 1 digested with 0, 2, 10, and 30 pg of V8 protease; 8 and 9, [""S]methionine-labeled protein M1 was digested with 10 and 30 pg of V8 protease.

1 2 3 4 a b 1 2 3 4 5 6

- " FIG. 7. Photoaffinity labeling by ["2P]dTTP and ["'PIdATP of

protein M1 from wild type and HAT 1.5A cells. The photolabeling and analysis were as described in under "Experimental I'rocedures." a, photolabeling by ['"P]dTTP (2 p ~ , 1.4 pCi/60 pl); I and 2, protein M1 from HAT 1.5A cells; 3 and 4, protein M1 from wild type cells. I and 3 contained 50 p~ deoxyguanosine. 2 and 4 contained 50 PM dCTP. b, photolabeling by ['"PIdATP (1 p ~ , 10 pCi). 2-3, protein M I from HAT 1.5A cells; 4-6, wild type protein M1. Additions: I and 4, no addition; 2 and 5,50 p~ dCTP; 3 and 6.50 p~ dTTP. An arrour indicates the position of the 89-kilodalton protein M1 band.

dTTP (indicating an alteration in the specificity site) and photoincorporate considerably less ["'P]dTTP than do wild type protein M1 preparations (5). We measured the relative labeling of protein M1 from HAT 1.5-A and wild type cells by ["'PIdATP and by ['"'PIdTTP (Fig. 7 ) . As previously reported (5), wild type protein M1 photoincorporated approximately 40-fold more [,'"P]dTTP than did protein M1 from HAT 1.5- A cells. By contrast, both preparations of M1 photoincorporated similar levels of [""PIdATP. Whereas the labeling of wild type protein M1 by ["'PIdATP (and ["'PIdTTP) was inhibited by both dGTP and dTTP, the labeling of HAT 1.5-A protein M1 was resistant to inhibition by dGTP and dTTP (Fig. 7 , a and 6). These data suggest that the mutation affecting the specificity site of protein M1 in HAT 1.5 cells has altered the binding affinity for dGTP and dTTP at that site,' while dATP (and presumably ATP) binds normally.

DISCUSSION

We recently reported the use of [:"P]dTTP to photolabel specifically the allosteric specificity site of ribonucleotide reductase protein M1 (5). We have now shown that [:'"P]dATP also photolabels partially purified protein M1. However, the relative photoincorporation of ["'PIdATP is about 100-fold lower than that of [:"P]dTTP. The ['"PIdATP labeling is specific as shown by the following observations: 1) low concentrations (less than 50 PM) of dATP inhibited the labeling of protein M1, indicating that ['"PIdATP labels a dATP- binding site and that the binding is saturable; 2) extensive digestion of ["PIdATP-labeled protein M1 with V8 protease yielded a single labeled polypeptide; and 3) known allosteric effector nucleotides of ribonucleotide reductase inhibited the labeling, while other nucleotides and nucleosides had no effect.

Unlike dTTP, dATP can affect both allosteric sites of protein MI, the activity site and the specificity site (4). We therefore anticipated using ["'PIdATP to probe the activity site. However, the evidence presented above shows that ['"'PI dATP labels only the specificity site. This conclusion was verified in three ways. Firstly, we compared the effects of dGTP on the labeling of protein M1 by ['"PIdTTP and ['"PI dATP. We reasoned that whereas ['"'PldTTP labeling would be totally inhibited by dGTP (which competes with dTTP for binding exclusively at the specificity site), the appearance of a low level of dGTP-resistant, [""PIdATP labeling would indicate labeling of the activity site. We observed, with both photolabels, identical inhibition by dGTP, suggesting that this was not the case. Secondly, we compared the dATP labeling of wild type protein M1 and of a mutant protein M1 that is unresponsive to dATP regulation at the activity site (presumably lacking the ability to bind dATP at that site). We reasoned that if both allosteric sites are indeed labeled by dATP, then the mutant protein M1 should show reduced photoincorporation of [:"P]dATP and any residual labeling of wild type protein M1 in the presence of dGTP (which blocks the specificity site) should be absent in the mutant protein M1 preparation. The photoincorporation of ['"PIdATP by mutant and wild type protein M1 preparations was indistin- guishable, both in the absence and presence of dGTP, suggesting that the activity site is not involved in [:'2P]dATP labeling. Finally, we compared the proteolytic digestion patterns of [32P]dATP-labeled and [3'P]dTTP-labeled wild type

' dCTP binding by the mutant protein MI appears to have been totally abolished, while dTTP binding has been reduced to approximately 3% of the wild type level. We deduce this from the observations that: ( i ) dTTP partially reduced the incorporation of [:"I']dATI' by the mutant protein M1 and (ii) the incorporation of ["I']TTP by the mutant protein M1 was 3% relative to wild type protein MI.

9512 Photoaffinity Labeling of Ribonucleotide Reductase Protein M1

protein M1. Exhaustive digestion with VS protease produced a single labeled peptide in each case. These peptides were identical on the basis of both size and charge (the latter determined by isoelectric focusing, data not shown). This result confirms that both reagents label the same (or closely related) site($ on protein M1.

The inability of [32P]dATP to photolabel the allosteric activity site of protein M1 cannot be attributed to the lower binding affinity for dATP at that site (K , specificity site, 1-2 ,UM uersus K, activity site, 20-30 ,uM (4)) since we were unable to detect activity site labeling using concentrations of [32P] dATP up to 40 ,UM. It is likely that the activity site does not contain a suitable acceptor side chain (e.g. a thiol group (10)) for photoactivated dATP.

While genetic techniques have clearly established the exist- ance of two regulatory domains on protein M1 (the activity site and the specificity site) (4), they have not resolved the nature of the complex binding interactions occurring at the specificity site. Whereas a single mutation at the specificity site can abolish (or reduce) the binding of both dGTP and dTTP (4), the effect on dATP (or ATP) binding is usually obscured in kinetic studies by dATP binding at the activity site. We have used the photolabeling of protein M1 to probe this question. Protein M1 from HAT 1.5A cells (4, 9) is resistant to normal regulation by dGTP and dTTP (indicating an alteration in the specificity site) and shows a markedly reduced ability to photoincorporate [32P]dTTP (5). In view of the above conclusion that dATP and dTTP photolabel the same protein M1 site, we speculated that protein M1 from HAT 1.5A cells would similarly be resistant to [32P]dATP labeling. However, the HAT 1.5.4 protein M1 preparation photoincorporated normal levels of [32P]dATP. In addition, labeling of the mutant protein M1 (unlike the wild type) was resistant to inhibition by dTTP and dGTP. These data suggest

site of protein M1 such that the binding of dGTP and dTTP is severely reduced or abolished, while dATP (and presumably ATP) binding remains unaffected. By the combination of genetic and photolabeling techniques, two independent binding interactions occurring within the regulatory domain termed the specificity site can therefore be distinguished, one for dATP and another for dGTP and dTTP.

The above observations are consistent with kinetic evidence that ribonucleotide reductase reconstituted with HAT 1.5A protein M1 is competent to catalyze pyrimidine-reduction (requiring the binding of ATP or dATP at the specificity site), but not ADP reduction (requiring the binding of dGTP) or GDP reduction (requiring the binding of dTTP) (4, 9).

These results demonstrate the usefulness of photolabeling to extend genetic techniques in probing the complex regulation of ribonucleotide reductase.

REFERENCES

1. Thelander, L., and Reichard, P. (1979) Annu. Rev. Biochem. 48,

2. Eriksson, S., Thelander, L., and Akerman, M. (1979) Biochemistry

3. Thelander, L., Eriksson, S., and Akerman, M. (1980) J. Biol. Chem. 255, 7426-7432

4. Eriksson, S., Gudas, L. J., Clift, S. M., Caras, I. W., Ullman, B., and Martin, D. W., Jr. (1981) J. Biol. Chem. 256,10193-10197

5. Eriksson, S., Caras, I. W., and Martin, D. W., Jr. (1982) Proc. Natl. Acad. Sei. U. S. A. 79,81-85

6. Stahl, E. (1969) Thin-Layer Chromatography p. 799, Springer, Berlin

7. Friedrich, U., and Coffino, P. (1977) Proc. Natl. Acad. Sci. U. S A. 74,679-683

8. Sibley, C. H., and T o m b s , G. M. (1974) Cell 2, 213-220 9. Ullman, B., Gudas, L. J., Caras, I. W., Eriksson, S., Weinberg, G.

L., Wormsted, M. A,, and Martin, D. W., Jr. (1981) J. Biol. Chem. 256,10189-10192

133-158

18,2948-2952

that the mutation in HAT 1.5A cells has altered the specificity 10. Shetlar, M. D. (1980) Photobiol. Rev. 5, 105-197

direct photoaffinity labeling of an allosteric site of subunit protein

Documents