Xenopus Transcription Factor IIIA

Vol. 265, No. 23, Issue of August 15, pp. 13792-13799.1990

Printed in U.S. A.

Transcription

Xenopus EVIDENCE LABELING

FOR HETEROGENEITY OF CYSTEINE 287*

Factor IIIA OF Zn2+ BINDING

AFFINITIES

AND

SPECIFIC

(Received for publication, Myun

K. Han,

Francis

P. Cyran,

From the Section on Protein Chemistry, Institutes of Health, Bethesda, Maryland Baltimore, Maryland 21210

Mark

T. Fisher,

The release of Zn2+ from transcription factor IIIA (TFIIIA) was examined with the metallochromic indicator 4-(2pyridylazo)resorcinol (PAR) in the absence and presence of p-hydroxymercuriphenylsulfonate (PMPS). With 0.5 mM PAR, approximately 5 eq of Zn2+ were released from TFIIIA, but no Zn2+ release was detected from the 7 S ribonucleoprotein. The PMPSpromoted Zn2+ release from TFIIIA was 8.7 & 0.4 eq of Zn2+ of which =4 eq of Zn2+ rebound to TFIIIA upon displacement of the mercurial with excess 2-mercaptoethanol. These results suggest that at least two affinity classes of Zn2+ binding sites exist in TFIIIA, one of which is released to 0.5 XIIM PAR in the absence of PMPS. Also, 18 of the 23 cysteine residues of TFIIIA reacted with 5,5’-dithiobis-(2-nitrobenzoic acid). The kinetic data of PAR and 5,5’-dithiobis-(2-nitrobenzoic acid) reactions with TFIIIA were similar, and the spectral changes were characterized by at least three exponential terms. Both TFIIIA and the 7 S particle were reacted with the thiol-specific fluorescent probe Niodoacetyl-N’-(5-sulfo-l-naphthyl)ethylenediamine (1,5-I-AEDANS). Complete trypsin hydrolysis followed by reverse-phase high pressure liquid chromatography analysis of peptide mixtures showed only one fluorescent peak from the AEDANS-labeled 7 S particle whereas numerous fluorescent peaks were observed with AEDANS-labeled TFIIIA. This further indicates exposure of cysteine residues from Zn2+ binding domains in TFIIIA. Cys2” was identified as the site of modification by amino acid sequencing of the isolated fluorescent peptide from the derivatized 7 S particle. Limited papain cleavage of the AEDANS-labeled 7 S particle indicated that the modified cysteine is located within a 34-kDa TFIIIA fragment. Gel retardation and transcription assays showed that TFIIIA, which had been purified from the AEDANS-labeled 7 S particle, was capable of binding to the internal control region of 5 S RNA gene and retained transcription activity. Thus, Zn2+ binding domains and all but 1 cysteine residue are buried in the 7 S particle, thereby facilitating site-specific labeling of TFIIIA.

* The results

in part

were presented

at the 33rd

Sang

Laboratory of Biochemistry, 20892 and the SDepartment

Annual

Biophysical

Societv Meeting (Han. M. K.. Knutson, J. R., Kim, S. H., Fisher, M. T., Cyran, F. P.land Ginsburg, A. (1989) Biophys. J. 55,120 (ahstr.)). 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 fact.

H. Kim& National Heart, of Embryology,

and Ann

January 17, 1990)

Ginsburg

Lung, and Blood Institute, National Carnegie Institution of Washington,

Transcription factor IIIA (TFIIIA)’ interacts with 5 S RNA to form a 7 S cytoplasmic ribonucleoprotein particle (1, 2), and it also binds specifically to 50 bp of the ICR of 5 S RNA genes (3). TFIIIA is required for the initiation of transcription (4, 5). TFIIIA from Xenopus laeuis consists of a polypeptide chain of 344 amino acids (M, = 38,500) (6). This protein includes nine imperfect repeats of a 27-amino acid motif (7, 8). Each domain, known as a “zinc finger,” contains 2 cysteine and 2 histidine residues that are proposed to bind one Zn2+ (7, 9). Limited proteolytic cleavage of the 7 S particle with papain showed that an amino-terminal 30-kDa TFIIIA fragment gave a DNA footprint pattern that was almost identical to that obtained with the intact 38.5kDa factor but supported transcription weakly (10). A TFIIIA mutant, in which the first 50 amino acids from the carboxyl terminus were deleted, also showed 90% reduction of transcription activity (11). It was concluded that the carboxyl-terminal lo-kDa fragment is responsible for most of the transcription-enhancing activity of the intact protein but that it does not bind directly to DNA. Since the transcription of 5 S RNA genes requires at least two other factors, TFIIIB and TFIIIC, in addition to RNA polymerase III (12, 13), the carboxyl-terminal region of TFIIIA is proposed to interact with the other factors that are essential to form a transcription complex or with RNA polymerase III. The observation that TFIIIC enhances the stabilization of TFIIIA binding to the 5 S RNA gene by interacting with the carboxyl-terminal domain of TFIIIA supports this view (14). Fluorescence spectroscopic techniques can readily be used to examine TFIIIA-TFIIIC and TFIIIA-DNA interactions. Fluorescence studies can be conducted utilizing intrinsic fluorophores or extrinsic covalent probes. The advantage of covalently attaching fluorophores to the surface of a protein is that probes having the desirable spectroscopic properties can be used. Recently, Hanas et al. (15) reported selective modification of TFIIIA with 1,5-I-AEDANS. In this paper, we demonstrate site-specific labeling of TFIIIA with 1,5-IAEDANS by identification of the modified site and characterization of the steady-state fluorescence of AEDANS-labeled TFIIIA. In addition, we have studied PMPS-promoted ’ The abbreviations used are: TFIIIA, transcription factor IIIA from X. la&s immature oocytes; TFIIIB, transcription factor IIIB; TFIIIC, transcription factor IIIC; ICR, internal control region of the X. laeuis oocyte 5 S RNA gene; DTT, dithiothreitol; SDS, sodium dodecyl sulfate; I-AEDANS, N-iodoacetyl-N’-(5-sulfo-1-naphthyl) ethylenediamine; HPLC, high pressure liquid chromatography; PMPS,p-hydroxymercuriphenylsulfonate; PAR, 4-(2-pyridylazo)resorcinol; HEPES, 4-(2-hydroxyethyl)-I-piperazineethanesulfonic acid; bp, base pair(s); DTNB, $5’.dithiobis-(2-nitrobenzoic acid); TPCK, l-tosylamido-2-phenylethyl chloromethyl ketone.

13792

TFIIIA:

Zn2+ Binding

Zn*+ release from TFIIIA as well as the kinetics of Zn2+ release monitored with a metallochromic indicator, PAR. Accessibility of sulfhydryl groups of TFIIIA also was examined with a thiol-specific reagent DTNB. EXPERIMENTAL

PROCEDURES

Materials X. laeuis were purchased from Nasco. TPCK-treated trypsin and RNase were obtained from Worthington Enzymes. Molecular weight markers, Sephadex G-25, and Sephacryl300 HR were obtained from Pharmacia LKB Biotechnology Inc. 1,5-I-AEDANS was obtained from Molecular Probes. PAR-was obtained from Eastman Organic Chemicals. PMPS. DTNB. and urotein A-Senharose were nurchased from Sigma. Chelex 100 (160-260 mesh) was obtained from Bio-Rad. Water was distilled and then deionized and filtered through a Millipore Q2 reagent grade system prior to use. Methods Purification of 7 S Particles and TFIIIA from X. lo&s-Immature ovaries were hand isolated from 150-250 6-7-month-old X. luevis and were homogenized in 50 mM HEPES, pH 7.5, 25 mM KCl, 5 mM MgCl*, 1 mM DTT, 10 pM ZnSOI, and 0.5 mM phenylmethylsulfonyl fluoride (buffer A). The crude extract was then centrifuged for 30 min at 18,000 rpm in an SS-34 rotor. The supernatant was chromatographed on a 440-ml Sephacryl S-300 column (2.5 x 90 cm) equilibrated with buffer A. The pooled solution was then chromatographed on a 14-ml DEAE-52 column (2.5 X 3 cm) equilibrated with-26 mM HEPES. DH 7.5.1.5 mM MsCl,. 1 mM DTT. and 10 uM ZnSOa (buffer B). The’column’ was washed v&h buffer B ‘plus 0.25 M KCl, and the 7 S particle was eluted with buffer B plus 0.4 M KCl. The pooled 7 S particle was mixed with an equal volume of buffer C containing 50 mM HEPES, pH 7.5, 5 mM MgC12, 1 mM DTT, 10 pM ZnSO1, and 20% glycerol and chromatographed on a 5-ml CM-Sephadex column (1.25 x 3 cm) equilibrated with buffer C plus 0.1 M KCl. The eluant was treated with RNase for 30 min at room temperature and then mixed with an equal volume of 10 M urea in buffer C. The sample was once again chromatographed on a 5-ml CM-Sephadex column equilibrated with buffer C plus 0.1 M KCl. The column was washed with buffer C plus 0.25 M KCl, and TFIIIA was eluted with buffer C plus 0.4 M KC1 and again with buffer C plus 1.0 M KCl. Purified 7 S particle and TFIIIA were stored at -80 “C. Preparation of AEDANS-labeled TFZZZA-The 7 S particle eluted with 0.4 M KC1 from the DEAE-52 column was dialyzed overnight against 6 liters of buffer B without DTT. A 100 molar excess of IAEDANS solution (dissolved in buffer B without DTT) was added to the dialyzed sample. After a 2-h incubation at 25 “C, DTT was added to stop the reaction (the final concentration of DTT was 5 mM), and the sample was dialyzed extensively against buffer B. The modified TFIIIA was obtained by chromatography on a CM-Sephadex column following RNase and 5 M urea treatment. Excess dye was removed by filtration of the protein through a Sephadex G-25 column followed bv extensive dialvsis. AEDANS-labeled TFIIIA was stored in buffer Cat -80 “C. The”degree of labeling was determined from the absorption spectrum using c336 “,,, of 6,600 M-’ cm-‘, which is the molar extinction coefficient for the product of the reaction of 1,5-I-AEDANS with N-acetylcysteine (is). Protein Determination-The TFIIIA protein concentration was determined by AzW nm = 6.7 for 10 mg/ml (17). The 7 S particle or AEDANS-labeled TFIIIA protein concentration was determined by the method of Bradford (18) or with the Pierce bicinchoninic acid protein assay using bovine serum albumin as the standard. Spectroscopic Measurements-Absorption spectra were measured with a Hewlett-Packard 8450A diode array spectrophotometer. Steady-state fluorescence spectra and emission anisotropy were recorded with an SLM 8,000 photon-counting spectrophotofluorometer with lo-mm Glan-Thompson polarizers. Fluorescence emission spectra were collected under “magic angle” emission conditions (19). The absorbance of all fluorescence samples was less than 0.1 at the wavelength of excitation to avoid inner filter effects. Reaction of TFIIIA with PAR and DTNB-TFIIIA stored in buffer C plus 1.0 M KC1 was chromatographed on a lo-ml G-25 column equilibrated with Chelex-treated 50 mM HEPES, pH 7.0, and 50 mM KCl. DTNB was dissolved in Chelex-treated buffers of 50 mM HEPES, pH 7.0 or pH 7.5, and 50 mM KCl. PMPS, PAR, and Chelextreated buffer solutions were prepared and tested by the procedures

and Labeling

of CYS~‘~

13793

of Hunt et al. (20). The concentration of Zn2+ was determined by using a molar extinction coefficient for (PAR)*. Zn complex formation of 66,000 cm-’ at 500 nm (20). A molar extinction coefficient of 13,600 cm-’ at 412 nm (21) was employed to estimate the concentration of TNB anion produced. The kinetic data were analyzed by a nonlinear least squares procedure (22) using a model with three concurrent pseudo-first order reactions leading to the same products. Since the change in absorbance provided a direct measure of the overall extent of the reaction, the following equation was used At = crle-klt

+ ageWktt + cu3e-k3’3’+ A,

where 01,. (Ye, and 01~ represent the concentrations of the species; A, represents the final absorbance; k,, kp, and ka represent the rate constants for the reactions with DTNB and PAR: and t is time in seconds. The best fit was determined by reduced xi value and residuals. Tryptic Peptide Analysis-TPCK-treated trypsin (l:lOO, w/w) was added to 1 ml of AEDANS-modified TFIIIA (0.234 mg/ml) and incubated at 37 “C overnight. An analytical Dynamax-150A (4.6 mm x 30 cm) 12-pm C-18 reverse phase column was used on an HP 1090M liquid chromatography system at room temperature. Ultraviolet absorbance was monitored with a diode array spectrophotomer, and fluorescence was monitored with a model 121 Gilson fluorometer. A linear gradient of solvent A (5% acetonitrile in H,O and 0.1% trifluoroacetic acid) and solvent B (75% acetonitrile in Hz0 and 0.1% trifluoroacetic acid over 100 min was used at a flow rate of 1.0 ml/ min. The isolated peptide was sequenced on an Applied Biosystems pulse liauid seauenator (model 477A) eouipped _ __ with an on-line phenylthiohydantoin analyzer by the Macromolecular Structure Facility at Michigan State University in East Lansing. Transcription Assay-X. lnevis nuclear extract was prepared from hand-isolated oocyte nuclei, and TFIIIA in the extract was depleted by using purified IgG against TFIIIA and protein A-Sepharose. Transcription reactions using these extracts and cloned class III gene were performed as described previously (23). Gel Retardation Assay-Protein-DNA complexes were resolved on 4% acrylamide nondenaturing gels using conditions described previously (10). A lOO-bp oligonucleotide that has the identical sequence (l-100) of ICR of the 5 S RNA gene was prepared as follows. Two primers of 30-mer (started from both ends) were initially synthesized on an Applied Biosystems 380B DNA synthesizer. The oligonucleotides were then purified by using a Du Pont Zorbax Bio Series Oligo column on an HP 1090M liquid chromatography system. A lOO-bp double-stranded oligonucleotide was prepared by using the Tao polymerase chain reaction with the Perkin-Elmer Cetus DNA thermal cycler, two 30-mer primers, and plasmid DNA, pXP-14, containing the ICR of 5 S RNA gene, which was used as a template. RESULTS

Reactions of TFIIIA with PMPS and PAR-The metallochromic indicator PAR forms a 2:l PAR.Zn” complex with Pz’ = 1O1’ M-’ (20, 24, 25), and the rate of (PAR)P.Zn2’ formation (>>107 M-b-‘) is non-rate limiting in studies of Zn2+ release from proteins (20). Moreover, relatively high concentrations of mercurial and thiol reagents (such as DTT and 2-mercaptoethanol) do not interfere with the absorbance changes accompanying (PAR)2. Zn2+ formation, which makes it possible to use these reagents in the presence of PAR for studies of Zn2+ release and uptake (20, 26). To estimate the total Zn*+ concentration in TFIIIA, free Zn2+ (10 pM ZnSOJ was removed from TFIIIA by passage through a Sephadex G25 gel filtration column that had been equilibrated with Chelex-treated buffer containing 50 mM HEPES/KOH, pH 7.0, and 50 mM KCl. Aliquots of TFIIIA were added to the same buffer solution containing 0.54 mM PAR and 0.11 mM PMPS. The concentration of Zn2+ was determined by using a molar extinction coefficient of 66,000 cm-’ at 500 nm (20). As shown in Fig. 1, PMPS-promoted Zn2+ release was linear, and the slope obtained from the linear regression analysis indicates that 8.7 f 0.4 mol of Zn2+ was released per mol of TFIIIA. At the end of the titration, an excess of 2-mercaptoethanol was added, and a decrease of absorbance corresponding to 3.6 Zn2+ occurred (Fig. 1, inset). Thus, approximately 4

TFIIIA:

Zn2+ Binding

and Labeling of CYS~‘~

k4 +’ 01 r3 N 2 1 0

ON’ 0

0.1 j-

I

0.2

0.4

0.6

0.8

1

TFMA (&I) FIG. 1. PMPS-promoted Zn2+ release from TFIIIA in the presence of PAR. Aliquots of TFIIIA (0.37 mg/ml) in Chelextreated buffer containing 50 mM HEPES, pH 7.0, and 50 mM KC1 were added to the same buffer containing 0.54 mM PAR and 0.11 mM PMPS. Concentration of Zn2+ was determined spectroscopically by using a molar extinction coefficient of 66,000 M-’ cm-’ for (PAR),. Zn complex formation (20). The slope (dashed line) obtained by linear regression indicates that 8.7 f 0.4 mol of Zn’+ were released per mol of TFIIIA. Inset, reversible binding of Zn*’ to TFIIIA in the presence of 0.5 mM PAR. After PMPS-promoted Zn2+ release was complete with 0.8 PM TFIIIA, 14.3 mM 2-mercaptoethanol (2-ME) was added to displace the protein-bound PMPS. The decrease in the absorbance at 500 nm shows that -4 of 9 eq of Zn*+ from the (PAR)* +Zn complex are rebound to TFIIIA.

out of 9 eq of Zn’+ are rebound to TFIIIA after displacing the protein-bound mercurial by the addition of excess 2-mercaptoethanol in the presence of 0.5 mM PAR. When TFIIIA was incubated with 0.5 mM PAR at pH 7.0 and 25 “C (in the absence of a mercurial reagent), a slow Zn2+ release was observed, as evidenced by the slow absorbance increase at 500 nm as the (PAR)?.Zn*+ complex was formed (Fig. 2, top, curue A). Analysis of the kinetic data by a nonlinear least squares method indicated that the kinetics of Zn2+ release to 0.5 mM PAR was best fit by at least three exponential terms (Table I). Biexponential fits gave substantially higher reduced x2 values, and as illustrated in the bottom panels of Fig. 2, the residual plot for a three-exponential fit for the data of curue A in the top was significantly better than that of the biexponential analysis. The rate constants are 5.4 f 0.5 X 10-*/s, 1.0 + 0.2 X 10-*/s, and 1.61 f 0.3 X 10m3/s for the release of approximately 1, 2, and 2, e.g. of Zn’+, respectively. In contrast, when TFIIIA was added to 0.5 mM PAR with 5 M urea present at pH 7.0, a total of 5 eq of Zn2+ was released instantaneously (Fig. 2, top, curue B). Such Zn2+ release to PAR was not observed with the 7 S particle (Fig. 2, top, curue C) or from a partial RNase digest of the 7 S particle (Fig. 2, top, curue D). Furthermore, horse liver alcohol dehydrogenase, which contains 2 catalytic and 2 structural Zn’+/ dimer (27), showed no such Zn*+ release to PAR, but released 4 eq of Zn*+ per dimer upon addition of PMPS (28). When 7 S particle was digested with RNase and chromatographed on a CM-Sephadex column, two protein samples containing TFIIIA were eluted with 0.4 and 1.0 M KCl. The protein eluted with 0.4 M KC1 was less than 10% of the TFIIIA eluted with 1.0 M KC1 (as judged by SDS-polyacrylamide gel electrophoresis analysis). Absorption spectra indi-

A 1 _ uj 0.05 o] :-: - -y---_- _ _ < -0.05 _=--- -- - __-- - _1

-0.1

I

3

--Lt. -- ----- - 0

-

- - _-

500

2 _- -

1000 TIME (SEC)

FIG. 2. Top, kinetics of Zn2’ release from TFIIIA (curue A), TFIIIA in 5 M urea (curue B), 7 S particle (curue C), and partial RNase digest of 5 S RNA bound to TFIIIA (curue D) in the presence of 0.5 mM PAR at 25 “C. TFIIIA was eluted with buffer C plus 1.0 M KCl, and partial 5 S RNA-bound TFIIIA was eluted with buffer C plus 0.4 M KC1 from the CM-Sephadex column (see “Methods” for the sample preparations). All three samples were chromatographed on a Sephadex G-25 column, equilibrated with Chelex-treated buffer containing 50 mM HEPES, pH 7.0, and 50 mM KCl. Protein concentrations of the samples were: 1 pM for TFIIIA and for TFIIIA in 5 M urea, 2 fiM for 7 S particle, and 1.6 pM for a partial digest of 7 S particle in which RNA is still bound to TFIIIA (see “Results”). Bottom, the residuals for two- and three-exponential fits of the data of curue A (aboue) are plotted in panels 1 and 2, respectively. Reduced x2 values for two- and three-exponential fits were 4.1 x 10m6 and 1.3 X 10e7, respectively.

TABLE

I and -SH exposure from isolated TFZZZA All experiments were performed with TFIIIA concentrations of l2 PM in Chelex-treated 50 mM HEPES, pH 7.0, and 50 mM KC1 at 25 “C. The final concentration of PAR was 0.5 mM and the DTNB concentration was 2 mM in the same buffer. The standard deviation was calculated from three separate experiments (see “Methods” for data analysis). Zn2+ release from TFIIIA to PAR was monitored at 500 nm (20). and sullhvdrvl _- group _ reactivity of TFIIIA to DTNB was measured at 412 nm (21). Pseudo-first order Reaction measured Amplitudes rate constants Kinetic

Zn*+

for Zn2+ release

parameters

release

-SH reactivity DTNB

to PAR

to

5.4 -c 0.5 x 10-*/s 1.0 -c 0.2 x 10-2/s 1.6 f 0.3 x W3/s

IZn*+1/1TFIIIAl . 0.8’; 0.4 1.7 +- 0.2 1.7 f 0.2

5.5 f 0.9 x 10-2/s 9.4 f 0.2 x 10-3/s 1.2 + 0.1 x 10-3/s

[-SH]/[TFIIIA] 7.9 * 0.7 5.6 + 0.2 4.2 + 0.2

-

TFIIIA:

Zn2+ Binding

and Labeling

cated that the TFIIIA fraction eluted with 0.4 M KC1 still contained some residually bound RNA whereas the protein

eluted with 1.0 M KC1 did not (Fig. 3). Since the protein eluted with 0.4 M KC1 and the 7 S particle did not release Zn’+ to PAR (Fig. 2, top), Zn2+ release from TFIIIA is prevented by the binding of 5 S RNA. From the above results, we concluded that there are at least two affinity classes of

Zn2+ binding sites in TFIIIA; lower affinity sites release Zn*+ to PAR in the absence of a mercurial reagent. Reaction of TFIZIA with DZ’NB-The DTNB reaction with TFIIIA produced TNB anions slowly, which indicated a slow exposure

of reacting

sulfhydryl

groups.

Curve

A of Fig. 4

shows kinetics of the DTNB reaction with TFIIIA at pH 7.5. Approximately 18 sulfhydryl groups of TFIIIA were accessible 0.30

1

. . . . . . . . . . . . . . . . . . . . . . . . . .._...._._....... 0.0 260

260

300

320

340

WAVELENGTH (nm) FIG. 3. Peak-normalized absorption spectra of TFIIIA (solid Iine) and partial RNase digest of 5 S RNA bound to TFIIIA (&shed line). Both protein samples were in buffer containing 50 mM HEPES, pH 7.0, and 50 mM KCl. The inset shows a photograph of Coomassie Blue-stained 10% SDS-polyacrylamide gel electrophoresis analysis of TFIIIA fractions (20 ~1 of 1.5-ml fraction) eluted from the CM-Sephadex column with 1.0 M KC1 plus buffer C (see “Methods”); the arrow marks the 38.5-kDa position.

24 ,

2

16-

E t

12-

A

’

1

DTNB

FIG. 4. Kinetics of the DTNB reaction with TFIIIA at 25 “C. The reaction was initiated by adding 0.5 mM DTNB (curve A) and 0.5 mM DTNB in the presence of 2 mM EDTA (curue B) to 1 pM TFIIIA in Chelex-treated buffer containing 50 mM HEPES, pH 7.5, and 50 mM KCl. Curve C, 1 pM TFIIIA was preincubated with 2 mM EDTA for the indicated time, and 0.5 mM DTNB was added at the arrow mark.

13795

to DTNB. When TFIIIA was treated with EDTA prior to DTNB addition, slightly more than 19 -SH groups reacted instantaneously with DTNB (Fig. 4, curve C). The same number of -SH groups reacted with DTNB when EDTA and DTNB were added together (Fig. 4, curve B). The DTNB reactions with TFIIIA were repeated at pH 7.0, and the kinetic parameters, obtained by nonlinear least squares analysis, are summarized in Table I. As in the case of Zn” release to PAR, the kinetics of the reaction of TFIIIA with DTNB was best fit by at least three exponentials. The pseudo-first order rate constants obtained by the kinetic analysis (see “Methods”) are 5.5 f 0.9 x 10-‘/s, 9.4 f 0.2 X 10-“/s, and 1.2 & 0.1 X 10-“/s for the exposure of eight, six, and four -SH groups, respectively. Of the 23 cysteine residues in TFIIIA, 22 are located within nine zinc finger domains (6), and 18 are postulated to be involved in binding nine Zn*+ ions (7,9). Since 18 sulfhydryl groups reacted with DTNB, a substantial number of the reacting sulfhydryls must be located in the zinc finger domains. Moreover, the three rate constants for -SH group exposure correspond to the rate constants for Zn2’ release to PAR within experimental error. The most direct interpretation is that -SH groups become available for attack by DTNB as Zn2+ is released from TFIIIA. In fact, we measured 7-8 eq of Zn2+ release per TFIIIA at the end of the DTNB reaction from the absorbance increase at 500 nm upon addition of 0.5 mM PAR. Once a cysteine residue is reacted with DTNB, Zn2+ cannot rebind to the protein. It is therefore not surprising that more Zn*+ ions are released in the DTNB reaction with TFIIIA than in the reaction of TFIIIA with 0.5 mM PAR, which involves an equilibrium between (PAR)2. Zn*+ and TFIIIA. One may interpret the existence of the three rate constants given in Table I for Zn2+ release as an indication of three different Zn2+ affinities. However, one cannot rule out the possibility that the rate of Zn2+ release changes as a consequence of conformational changes associated with initial loss of Zn2+. Modification

of TFIIIA

in the 7 S Particle

with

1,5-I-

AEDANS-Modification of 7 S particle with 1,5-I-AEDANS was performed with a 100-fold molar excess of the dye at pH 7.5 and 25 “C. After a 2-h incubation, modified TFIIIA was purified from the labeled 7 S particle according to the procedure described under “Methods.” Excess dye was removed by gel filtration and extensive dialysis. The degree of labeling in AEDANS . TFIIIA was determined spectroscopically using t336 “In = 6.0 X lo3 M-’ cm-’ (16). The labeling stoichiometry was 1.45 mol of AEDANS per mol of TFIIIA. When the 7 S particle was reacted with 1,5-I-AEDANS for 30 min at 25 “C, the labeling stoichiometry was 1.05 mol of AEDANS per mol of TFIIIA. Identification DANS-Cysteine

“j/

of Cys287

of the Tryptic

Peptide

Containing

an AE-

Residue-The modified 7 S particles and AEDANS. TFIIIA (purified from the AEDANS-7 S particle) were completely digested with trypsin (l:lOO, w/w) overnight, and the resulting tryptic peptides were applied to an HPLC C-18 reverse phase column. The peptide elution patterns obtained with both samples monitored by fluorescence are shown in Fig. 5. Only one major fluorescence peak was observed from AEDANS-7 S particle (prepared as indicated above). Similar results were obtained from AEDANS . TFIIIA (purified from AEDANS-7 S particles), but traces of minor peaks were observed. The integration of these fluorescence peaks indicated that the area of the major peak represents 90% in Fig. 5A and 70% in Fig. 5B of the total fluorescence peak integral for AEDANS-7 S and AEDANS.TFIIIA, respectively. The results are in good agreement with the labeling

TFIIIA:

13796

Zn2+ Binding

and Labeling of CY.S*‘~

abcdefgh 94K 67K 43K

38.5K 34K

30K

10

20 ilme

30 Crnlrl.

40 >

FIG. 5. Analysis of tryptic peptides by HPLC reverse phase C-18 column chromatography. Elution profiles of tryptic peptides from A, 100 ~1 of AEDANS-7 S particle (0.32 mg of protein/ml); B, 100 ~1 of AEDANS.TFIIIA (0.23 mg/ml) purified from AEDANSlabeled 7 S particle; and C, 100 ~1 of isolated TFIIIA modified with I-AEDANS (0.05 mg/ml). All samples were digested with TPCKtreated trypsin (1:50, w/w) overnight at 37 “C, and the peptides were chromatographed on an analytical reverse phase C-18 column with a flow rate of 1.0 ml/min. A linear gradient of solvent A (5% acetonitrile. 0.1% trifluoroacetic acid) and solvent B (75% acetonitrile, 0.1% trifluoroacetic acid) was used. Elution profiles were monitored with a fluorescence detector (excitation = 340 nm; emission = 490-575 nm). stoichiometry of AEDANS . TFIIIA and AEDANS-7 S particles determined by spectroscopic methods. When I-AEDANS was reacted with purified TFIIIA (Fig. 5C), more than nine fluorescent peaks were resolved by the HPLC C-18 reverse phase column. These data suggest that 1,5-I-AEDANS reacts with numerous free sulfhydryl groups in TFIIIA. To identify the sequence of the peptide containing AEDANS, the major fluorescent peak from AEDANS-7 S particle (Fig. 5A) was isolated and sequenced. The amino acid sequence of the peptide was found to be X-Pro-Arg-Pro-Lys. Note that the arginine residue was not cleaved by trypsin due to the presence of an adjacent proline residue (29). Examination of the amino acid sequence of TFIIIA, deduced from the nucleotide sequence (6), revealed that the derivatized peptide corresponds to Cys2R7-Pro2R8-Arg28g-Pro2g0-Lys2g1 in the TFIIIA sequence. Thus, the modified amino acid is CYS~‘~. Smith et al. (10) reported that limited proteolysis of the 7 S particle with papain yielded a 30-kDa breakdown product that was later reported to be 34 kDa (14). To establish the position of the AEDANS group in TFIIIA, we digested the AEDANS-labeled 7 S particle with papain and analyzed the products by 10% SDS-polyacrylamide gel electrophoresis. As shown in Fig. 6, fluorescence was observed in a 38.5-kDa band as well as in a 34-kDa band. Hence, the fluorescent peptide is located within the 34-kDa fragment from the amino-terminal region of TFIIIA. Properties of AEDANS in TFIIIA and 7 S Particle-Fig. 7 shows fluorescence spectra of AEDANS-modified TFIIIA (isolated from the AEDANS-7 S particle) and 7 S particle. The wavelength of maximum emission is reported to be 500 nm and 498 nm for 1,5-acetylcysteine-AEDANS and mercaptoethanol-AEDANS, respectively (16). The maximum emission wavelength for AEDANS.TFIIIA and AEDANS-7 S particles is 495 nm, indicating that the AEDANS group is

2O.lK DF FIG. 6. SDS-polyacrylamide gel electrophoresis analysis of AEDANS-labeled 7 S particle and a partial papain digest. Lanes u-d show the 10% Coomassie-Stained gel. Lanes e-h show the same samples photographed under ultraviolet illumination. All lanes contained 5 rg of the labeled 7 S particle. Lanes b and f, c and g, d and h show the 7 S particle after digestion with papain for 5 min at 25 “C where 0.2, 1, and 5 pg of papain were added, respectively.

WAVELENGTH (nm) FIG. I. Fluorescence excitation and emission spectra of AEDANS-labeled TFIIIA isolated from the labeled 7 S particle (solid curues) and AEDANS-labeled 7 S particle (dashed curves). The spectra have not been corrected for the wavelength dependence of the lamp intensity or the detection system sensitivity. Excitation spectra were recorded with an emission wavelength set at 500 nm, and emission spectra were recorded with an excitation wavelength at 340 nm.

located in a relatively aqueous environment. An excitation peak from 275 to 290 nm in AEDANS.TFIIIA was observed and is indicative of Forster energy transfer between intrinsic tryptophan fluorescence and AEDANS. Steady-state emission anisotropy of AEDNAS. TFIIIA was measured to be 0.043 f 0.011 at 25 “C. This low anisotropy clearly does not represent a typical value for a protein of 38.5 kDa. Possibly this low anisotropy value arises from either segmental motion of the carboxyl-terminal domain or rotation of an elongated molecule. Accurate hydrodynamic properties of AEDANS. TFIIIA are being examined with nanosecond time-resolved fluorescence techniques. The solvent accessibility of AEDANS was studied by KI

TFIIIA:

Zn2+ Binding

fluorescence quenching. Quenching experiments were performed by recording emission spectra (exciting at 340 nm) as a function of KI concentration. Wavelength-dependent SternVolmer plots were examined. Fig. 8 shows a linear SternVolmer relationship for both AEDANS-labeled TFIIIA and 7 S particle obtained from emission intensities at 500 nm; this indicates that there is a single class of fluorophores. The Stern-Volmer constant obtained from the slope was 1.52 M-’ and was wavelength independent. In addition, no spectral shift was observed in the quenched spectra. The fluorescence quenching results indicate that the fluorophore in AEDANS. TFIIIA and AEDANS-7 S particle is fully accessible to solvent. Furthermore, these data indicate that the environments of AEDANS moieties in TFIIIA and 7 S particles are very similar. This, in turn, implies that there are no significant structural changes in the Cys”’ region of TFIIIA upon dissociation of 5 S RNA from the TFIIIA.

and Labeling of C~S“‘~

A 1

2

B 3

1234

9!_.**CL -

DNA Binding Assay and Transcriptional Assay of AEDANS. TFUZA-Gel mobility shift analysis shows that AEDANS-labeled TFIIIA binds to the lOO-bp oligonucleotide that has an ICR sequence identical to the 5 S RNA gene (Fig. 9A). As shown earlier, the modification site was located just outside of zinc finger domains and was not involved in 5 S RNA binding. Therefore, binding of AEDANS-labeled TFIIIA to DNA was expected. As shown in Fig. 9B, the AEDANS-labeled TFIIIA also retained full activity in the transcription activity assay. Thus, labeling CyszR5 of TFIIIA with 1,5-I-AEDANS does not interfere with the specific protein-protein interactions and nucleotide sequence recognition involved in promoting 5 S RNA synthesis. Since the labeling stoichiometry for AEDANS . TFIIIA used in the transcription assay was 1.45 mol of AEDANS per mol of TFIIIA, we can state that most transcription activity came from the modified TFIIIA and not by residual unmodified TFIIIA. In fact, this was the main reason that we prepared AEDANS . TFIIIA, which had a higher degree of labeling ’ order to ensure that all TFIIIA was modified.

I,

llrlr

--5SRNA

- *

1tRNA

m mobilitv shift assav of AEDANS-labeled TFIIIA. Lane I, 0.1 pg of DNA containing theICR of the 5 S RNA gene (see “Methods” for DNA preparation); lane 2, 0.1 Gg of DNA plus 0.1 pg of AEDANS-labeled TFIIIA; lane 3, 0.1 pg of DNA plus 0.2 pg of AEDANS-1abeledTFIIIA. The gel was stained with ethidium bromide and photographed under ultraviolet illumination. Panel B, transcription assav of AEDANS-labeled TFIIIA. Transcrintion assays contained 2.5 ~1 of oocyte nuclear extract (lane I) and 4 gl of TFIIIAdepleted oocyte nuclear extract (lanes 2-4). TFIIIA-depleted extract was supplemented with 20 ng of native TFIIIA (lane 3) and AEDANSlabeled TFIIIA (lane 4). The nuclear extract and complemented extract were incubated with a mixture of plasmids, pXIsl1 (25 ng), containing X. laeuis somatic 5 S RNA gene and pXItmet (50 ng), containing tRNA”” gene.

/'I

l.l-

14' 1