Proc. Natl. Acad. Sci. USA Vol. 76, No. 6, pp. 2649-2653, June 1979
Biochemistry
lac repressor changes conformation upon binding to poly[d(A-T)] (fluorescence/stopped-flow/nonoperator DNA/cysteine residue labeling/protein-nucleic acid interaction)
DAVID E. KELSEY, THOMAS C. ROUNDS, AND SHELDON S. YORK* Department of Chemistry, University of Denver, Denver, Colorado 80208
Communicated by Norman Davidson, March 19, 1979
ABSTRACT N-Iodoacetylaminoethyl)1-naphthylamine-5sulfonate reacts with Escherichia colilac repressor to selectively label cysteine-140 with the fluorescent N(acetylaminoethyl1-naphthylamine-5-sulfonate group. The fluorescence intensity of this label decreases by 20% when labeled repressor associates with poly[d(A-T)]. Fifteen base pairs of poly[d(A-T)] per repressor tetramer are required to complete this decrease. Stopped-flow experiments have shown that the repressor undergoes at least two conformational changes as it binds to poly[d(A-T)J, with half-lives of 5.0 i 1.2 msec and 3.5 + 1.0 sec. Quite likely, these conformational changes serve to strengthen the interaction of repressor with DNA. The association of Escherichia coli lac repressor with its operator and with nonoperator DNA has been extensively studied (for recent reviews, see refs. 1-3). Both the lac operator (4) and nonoperator DNA (5, 6) undergo structural changes as a consequence of repressor binding. The large bimolecular rate constant for association (7, 8) implies that these changes occur within the repressor-DNA complex. We present evidence that the lac repressor undergoes at least two conformational changes as it binds to poly[d(A-T)]. Quite likely, one or more of these changes occurs simultaneously with changes in the DNA, serving to strengthen their interaction. The lac repressor was treated with N-(iodoacetylaminoethyl)-1-naphthylamine-5-sulfonate (I-AENS) (9) to covalently label cysteine residues with the fluorescent N-(acetylaminoethyl)-1-naphthylamine-5-sulfonate (AENS) group. Each repressor subunit has three cysteine residues (10) located in regions not thought to be directly involved in DNA binding (3). These residues display different reactivities toward the chromophoric reagents 2-chloromercuri-4-nitrophenol and 2-bromoacetamido-4-nitrophenol (11). However, the fluorescent reagent fluorescein mercuric acetate shows little selectivity among the three sulfhydryls (12). We find that I-AENS reacts selectively with one cysteine residue in each repressor subunit. Stopped-flow techniques have been used to monitor the events during the association of repressor with poly[d(A-T)] that alter the fluorescence of this AENS label.
MATERIALS AND METHODS Materials. The lac repressor was isolated from E. coli CSH 46 and assayed for operator binding activity as described (13). Other materials were obtained from the following sources: poly[d(A-T)] from Miles, I-AENS from Aldrich, iodo[14C]acetamide from New England Nuclear, and trypsin (TRL) and chymotrypsin (CDI) from Worthington. Labeling of lac Repressor. Repressor (1-3 mg/ml) in 0.2 M Tris-HCI (pH 8.6 at 40C), 0.5 M KCI, and 0.1 mM EDTA, was reacted with I-AENS (usually 8.1 mM) at 40C in the dark. At the appropriate time, the protein was separated from excess The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate
reagent by gel filtration through Sephadex G-25 and then dialyzed overnight against the buffer used for labeling, with 0.1 mM dithiothreitol added. The concentration of repressor and extent of AENS labeling were determined from the absorbance at 280 and 340 nm by using the following values: lac repressor subunit, E2W = 22,125 M-1 cm-1 (14), Ec40 = 105 M-1 cm-1; AENS, E280 = 1260 M-1 cm-1, E3O = 6850 M-1 cm-1 (9). Identification of Labeled Cysteine Residues. AENS-labeled repressor was reacted with iodo[14C]acetamide (1.68 Ci/mole, 1 Ci = 3.70 X 1010 becquerels) in 8 M urea at 320C. The molar concentration of iodoacetamide was 6 times that of repressor subunits. After 30 min, excess iodoacetamide was reacted with dithiothreitol, and the splution was dialyzed against 0.1 M NH4HCO3/2 M urea. This protein was then digested with trypsin (1% wt/wt) for 2.5 hr at 250C. A second addition of trypsin was followed by an additional 2.5-hr digestion. Tryptic peptides were then separated on a Sephadex G-50 Superfine column (1 X 53 cm) equilibrated with 0.1 M NH4HCO3/2% sodium dodecyl sulfate. Combined tryptic and chymotryptic digestions were done in the same manner (11). Fluorescence Measurements. Fluorescence measurements were made at 250C in a 4-mm pathlength cuvette with a Perkin Elmer MPF-3 fluorimeter as described (15). The absorbance at the excitation wavelength was kept below 0.06 to minimize inner filter effects. All buffers used for fluorescence and kinetic measurements of AENS-labeled repressor and its association with poly[d(A-T)] contained 0.1 mM EDTA and 0.1 mM dithiothreitol. Kinetic Measurements. Kinetic measurements were made with a Durrum D-110 stopped-flow spectrophotometer with a 75-W xenon arc lamp. The excitation wavelength was 340 nm, and the fluorescence emission was isolated with a Corning 3-74 filter. The data were collected on a Nicolet 1090A digital oscilloscope with a modified time base. A typical trace consisted of 1024 points, 256 points each collected at 0.2, 2, 40, and 400 msec per point. This allowed both fast and slow reactions to be analyzed from a single trace. The instrument dead time was 2.6 msec, with reagent mixing occurring 1.8 msec before the oscilloscope was triggered. For this reason, points recorded during the first millisecond were omitted from analysis. Typically, 120 points were taken for computer analysis, which was performed according to the damped nonlinear least-squares method of Laiken and Printz (16). All reactions were carried out at 250C. The concentrations after mixing are reported. RESULTS Reaction of I-AENS with lac Repressor. The number of AENS labels incorporated per repressor subunit as a function of time is shown in Fig. 1. The method of Laiken and Printz (16) Abbreviations: I-AENS, N-(iodoacetylaminoethyl)-1-naphthylamine-5-sulfonate; AENS, N-(acetylaminoethyl)-1-naphthylamine5-sulfonate group; Nbs2, 5,5'-dithiobis(2-nitrobenzoic acid). *To whom reprint requests should be addressed.
this fact.
2649
Biochemistry: Kelsey et al.
2650
Proc. Natl. Acad. Sci. USA 76 (1979) Table 1. AENS labeling of sulfhydryls
AENS/subunit 0.6 1.0 1.4 1.9 2.3
0.6 0.5 1.0 0.8 1.3 1.2 1.5 1.5 1.7 1.8 * The difference between the number of free sulfhydryls in unlabeled and AENS-labeled repressor was determined with Nbs2 (17). Unlabeled repressor gave 1.8 sulfhydryls/subunit when native and 2.7 sulfhydryls/subunit when denatured with 0.1% sodium dodecyl sulfate.
C
Sg2.0O 0
in Uj
Blocked sulfhydryls/subunit* Native Denatured
1.0 /
0
5
10
15
20
25
50
Time, hr
FIG. 1. Labeling of lac repressor with I-AENS. Repressor (10 ,M tetramers) was reacted with 8.1 mM I-AENS. The two exponential processes that adequately describe these data are shown as the solid line.
was used to fit these data to two exponential processes introducing 0.9 and 1.7 AENS labels per subunit with pseudo-firstorder rate constants of 1.3 hr-t and 0.083 hr-', respectively. These correspond to bimolecular rate constants of 4.3 X 10-2 M-I sec- and 2.8 X 10-3 M-' sec-t, respectively. In order to determine if I-AENS reacted exclusively with the three sulfhydryl groups per repressor subunit, we measured the number of blocked sulfhydryl groups at different extents of AENS la-
beling by using 5,5'-dithiobis(2-nitrobenzoic acid) (Nbs2). Nbs2 reacts with only two sulfhydryl groups per subunit under native conditions (6, 1 1), but will react with all three sulfhydryls under denaturing conditions (6). The results shown in Table 1, together with the data from Fig. 1, indicate that one sulfhydryl group per subunit reacts rapidly and a second sulfhydryl, together with an as yet unidentified residue, react slowly. The two sulfhydryls reactive toward I-AENS appear to be the same two that react with Nbs2 in the native state. Cysteine-140 Is Selectively Modified. Repressor, labeled to varying extents, was denatured; then unreacted sulfhydryls were alkylated with iodol '4Clacetamide. The protein was then digested with trypsin and the peptides were separated. Tryptic digestion produces three cysteine-containing peptides (10, 18) with the following molecular weights: Cys-140, 5361; Cys-281,
2860; and Cys-107, 693. Fig. 2A shows the results for repressor labeled with 0.8 AENS groups per subunit. Most of the fluorescence (72%) is associated with the peptide containing Cys140. An additional 14% of the fluorescence is associated with the peptide containing Cys-107. The position of this peptide
C 0)
250 Ci
o5
I0
~ ~
~
~
5
iI
A
Fraction FIG. 2. Separation of tryptic peptides from AENS-labeled repressor. The position otfthe sulfhydryl-containing peptides is marked with 14C (0 --- 0). The fluorescence intensity of AENS-containing peptides (O--*) was monitored with excitation anld emission wavelengths of 1340 and 490 nm, respectively. Both cpm and fluorescence intensities are plotted on relative scales. (A ) Procedure begun with 1.7/ mg of repressor with 0.8 AENS labels per subunit. Peak cpm (fraction 31) was 2058. (B) Procedure begun with 1.3< mg of repressor with 2.0 AENS labels per subunit. Peak cpm (fraction 31) was 1560. The relative scales for B have been normalized with respect, to A so that the total fluorescence intensity eluted in H is 2.5 times that eluted in A.
Proc. Natl. Acad. Sci. USA 76 (1979)
Biochemistry: Kelsey et al. in the elation profile is shifted by the AENS label, as shown by a comparison of Fig. 2 A and B. These results demonstrate that Cvs-140 reacts rapidly while Cys-107 reacts slowly. The same conclusion was reached by separating peptides produced by combined tryptic and chymotryptic digestion. Fig. 2B shows the results obtained starting with 2.0 AENS labels per subunit. Essentially all of Cys-140 and more than half of Cys- 107 have reacted, while Cys-281 remains unlabeled by AENS. The fluorescence under the Cys- 140 peak corresponds to 1.23 AENS labels per cysteine residue, indicating that another residue has reacted in addition to cysteine. Because the peptide containing Cys-140 is significantly larger than the next largest tryptic peptide (Mr = 42.33), this unidentified residue appears to be located in the same peptide as Cys-140, which spans residues 119-168. Most of our experiments have used repressor labeled with approximately 0.5 AENS groups per subunit to achieve selective labeling of Cys-140. This level of labeling does not alter the operator binding activity of repressor. Neither does it alter the ability of repressor to bind to polyld(A-T)J, as assayed by sedimenting mixtures of the two through glycerol gradients (13). Both unlabeled and AENS-labeled repressor are >95% active in complexing with polyfd(A-T)]. Repressor-Polyid(A-T)l Association Alters AENS Fluorescence. The fluorescence emission spectrum of repressor labeled with 0.46 AENS groups per subunit is shown in Fig. 3. The addition of saturating amounts of polyld(A-T)I decreased the fluorescence intensity at the emission maximum by 20% and red-shifted the emission maximum by 4 nm. Essentially identical results were obtained in 0.1 M Tris-HCI (pH 8.0) with repressor labeled with 0.25, 0.50, or 0.76 AENS groups per subunit. With 1.19 AENS groups per subunit the fluorescence decrease was only 14%, even though the protein was still fully active in binding to poly[d(A-T)I. At extents of labeling greater than 1 AENS group per subunit, the fraction of labels attached to residues other than Cys- 140 sharply increases. These results
2651
imply that the AENS labels attached to Cys-140 are primarily responsible for the fluorescence decrease. Changing the pH from 7.2 to 8.5 had little effect on the emission spectrum of repressor labeled with 0.47 AENS groups per subunit. The decrease in fluorescence intensity upon addition of polyId(A-T)I was independent of pH. AENS-labeled repressor was titrated with poly[d(A-T)J in order to determine the size of the repressor binding site on nonoperator DNA. With 0.50 AENS labels per subunit, 15 base pairs of poly[d(A-T)] per tetramer were required to complete the decrease in fluorescence (Fig. 4). The same result was obtained with repressor labeled with 0.25, 0.76, or 1.19 AENS groups per subunit. This again emphasizes that the ability of the repressor to bind to poly[d(A-T)] is not altered by these extents of AENS labeling. This titration was repeated in 10 mM Tris-HCI (pH 8.0) with the ionic strength adjusted to 30 mM with NaCl. Again, 15 base pairs per tetramer were required to reach the end point when repressor labeled with 0.46 AENS groups per subunit was used. Stopped-Flow Experiments Detect Two Conformational Changes. Stopped-flow techniques were used to monitor this decrease in fluorescence as AENS-labeled repressor bound to polyfd(A-T)I. Fig. 5 shows the results of one experiment in which repressor labeled with 0.66 AENS groups per subunit was reacted with 30 base pairs of poly[d(A-T)] per tetramer. In this reaction, the fluorescence decreased in two exponential processes, with half-lives of 4.4 msec and 4.2 sec. After extrapolation to the time of mixing, these two reactions accounted for 71% of the total decrease in fluorescence (11.3%) observed in the stopped-flow instrument by monitoring the emission from AENS-labeled repressor before and after the addition of poly[d(A-T)J. The ratio of the magnitudes of the fluorescence decreases in these fast and slow reactions was 1.4. The small differences between the calculated and observed points may be attributed to one or more of the following causes: instrumental noise and drift, selection of only 120 out of 1024 points for analysis, or additional reactions with magnitudes too small to be adequately characterized. These same two processes were observed under a wide variety of experimental conditions, as shown in Table 2. Varying the amount of polyfd(A-T)j from 5 to 60 base pairs per tetramer, while holding the concentration of repressor essentially con-
c
aL) C
a) a)
CD
c
V) 0
C
a)
U-
c
a1) el a) 0
U-
420
460 540 500 Wavelength, nm
580
Fi(,. 3. Fluorescence emission spectra of AENS-labeled repressor in the absence and presence of poly[d(A-T)l. Repressor (1.9 ,uM tetramers), labeled with 0.46 AENS labels per subunit, in 10 mM TrisHCIL pH 8.0/50 mM NaCi. (-) Without DNA; (---) with 48 base pairs of polyld(A-T)l per tetramer. Excitation wavelength, 340 nm.
10 50 40 20 30 Poly [d (A-T)I base pairs/repressor tetramer
FIG. 4. Titration of AENS-labeled repressor with polyld(A-T)l. Excitation and emission wavelengths, 340 and 475 nm, respectively. Starting solution: 250 ,tl, containing repressor (1.6 pM tetramers) with 0.50 AENS labels per subunit, in 0.10 M Tris-HCI (pH 8.0).
Biochemistry: Kelsey et al.
2652
Proc. Natl. Acad. Sci. USA 76 (1979)
Finally, varying the number of AENS labels per subunit from 0.41 to 0.91 had no effect on these two reactions. When all 29 trials at pH 8.0 and ionic strength 56 mM were included, the average half-lives ± SD of the fast and slow reactions were 5.0 1.2 msec and 3.5 + 1.0 sec, respectively. When the 16 trials at a different pH or ionic strength were averaged, the half-lives of the fast and slow reactions were 5.1 ± 1.1 msec and 3.5 + 1.0 sec, C
a)
5
92
Biochemistry
lac repressor changes conformation upon binding to poly[d(A-T)] (fluorescence/stopped-flow/nonoperator DNA/cysteine residue labeling/protein-nucleic acid interaction)
DAVID E. KELSEY, THOMAS C. ROUNDS, AND SHELDON S. YORK* Department of Chemistry, University of Denver, Denver, Colorado 80208
Communicated by Norman Davidson, March 19, 1979
ABSTRACT N-Iodoacetylaminoethyl)1-naphthylamine-5sulfonate reacts with Escherichia colilac repressor to selectively label cysteine-140 with the fluorescent N(acetylaminoethyl1-naphthylamine-5-sulfonate group. The fluorescence intensity of this label decreases by 20% when labeled repressor associates with poly[d(A-T)]. Fifteen base pairs of poly[d(A-T)] per repressor tetramer are required to complete this decrease. Stopped-flow experiments have shown that the repressor undergoes at least two conformational changes as it binds to poly[d(A-T)J, with half-lives of 5.0 i 1.2 msec and 3.5 + 1.0 sec. Quite likely, these conformational changes serve to strengthen the interaction of repressor with DNA. The association of Escherichia coli lac repressor with its operator and with nonoperator DNA has been extensively studied (for recent reviews, see refs. 1-3). Both the lac operator (4) and nonoperator DNA (5, 6) undergo structural changes as a consequence of repressor binding. The large bimolecular rate constant for association (7, 8) implies that these changes occur within the repressor-DNA complex. We present evidence that the lac repressor undergoes at least two conformational changes as it binds to poly[d(A-T)]. Quite likely, one or more of these changes occurs simultaneously with changes in the DNA, serving to strengthen their interaction. The lac repressor was treated with N-(iodoacetylaminoethyl)-1-naphthylamine-5-sulfonate (I-AENS) (9) to covalently label cysteine residues with the fluorescent N-(acetylaminoethyl)-1-naphthylamine-5-sulfonate (AENS) group. Each repressor subunit has three cysteine residues (10) located in regions not thought to be directly involved in DNA binding (3). These residues display different reactivities toward the chromophoric reagents 2-chloromercuri-4-nitrophenol and 2-bromoacetamido-4-nitrophenol (11). However, the fluorescent reagent fluorescein mercuric acetate shows little selectivity among the three sulfhydryls (12). We find that I-AENS reacts selectively with one cysteine residue in each repressor subunit. Stopped-flow techniques have been used to monitor the events during the association of repressor with poly[d(A-T)] that alter the fluorescence of this AENS label.
MATERIALS AND METHODS Materials. The lac repressor was isolated from E. coli CSH 46 and assayed for operator binding activity as described (13). Other materials were obtained from the following sources: poly[d(A-T)] from Miles, I-AENS from Aldrich, iodo[14C]acetamide from New England Nuclear, and trypsin (TRL) and chymotrypsin (CDI) from Worthington. Labeling of lac Repressor. Repressor (1-3 mg/ml) in 0.2 M Tris-HCI (pH 8.6 at 40C), 0.5 M KCI, and 0.1 mM EDTA, was reacted with I-AENS (usually 8.1 mM) at 40C in the dark. At the appropriate time, the protein was separated from excess The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate
reagent by gel filtration through Sephadex G-25 and then dialyzed overnight against the buffer used for labeling, with 0.1 mM dithiothreitol added. The concentration of repressor and extent of AENS labeling were determined from the absorbance at 280 and 340 nm by using the following values: lac repressor subunit, E2W = 22,125 M-1 cm-1 (14), Ec40 = 105 M-1 cm-1; AENS, E280 = 1260 M-1 cm-1, E3O = 6850 M-1 cm-1 (9). Identification of Labeled Cysteine Residues. AENS-labeled repressor was reacted with iodo[14C]acetamide (1.68 Ci/mole, 1 Ci = 3.70 X 1010 becquerels) in 8 M urea at 320C. The molar concentration of iodoacetamide was 6 times that of repressor subunits. After 30 min, excess iodoacetamide was reacted with dithiothreitol, and the splution was dialyzed against 0.1 M NH4HCO3/2 M urea. This protein was then digested with trypsin (1% wt/wt) for 2.5 hr at 250C. A second addition of trypsin was followed by an additional 2.5-hr digestion. Tryptic peptides were then separated on a Sephadex G-50 Superfine column (1 X 53 cm) equilibrated with 0.1 M NH4HCO3/2% sodium dodecyl sulfate. Combined tryptic and chymotryptic digestions were done in the same manner (11). Fluorescence Measurements. Fluorescence measurements were made at 250C in a 4-mm pathlength cuvette with a Perkin Elmer MPF-3 fluorimeter as described (15). The absorbance at the excitation wavelength was kept below 0.06 to minimize inner filter effects. All buffers used for fluorescence and kinetic measurements of AENS-labeled repressor and its association with poly[d(A-T)] contained 0.1 mM EDTA and 0.1 mM dithiothreitol. Kinetic Measurements. Kinetic measurements were made with a Durrum D-110 stopped-flow spectrophotometer with a 75-W xenon arc lamp. The excitation wavelength was 340 nm, and the fluorescence emission was isolated with a Corning 3-74 filter. The data were collected on a Nicolet 1090A digital oscilloscope with a modified time base. A typical trace consisted of 1024 points, 256 points each collected at 0.2, 2, 40, and 400 msec per point. This allowed both fast and slow reactions to be analyzed from a single trace. The instrument dead time was 2.6 msec, with reagent mixing occurring 1.8 msec before the oscilloscope was triggered. For this reason, points recorded during the first millisecond were omitted from analysis. Typically, 120 points were taken for computer analysis, which was performed according to the damped nonlinear least-squares method of Laiken and Printz (16). All reactions were carried out at 250C. The concentrations after mixing are reported. RESULTS Reaction of I-AENS with lac Repressor. The number of AENS labels incorporated per repressor subunit as a function of time is shown in Fig. 1. The method of Laiken and Printz (16) Abbreviations: I-AENS, N-(iodoacetylaminoethyl)-1-naphthylamine-5-sulfonate; AENS, N-(acetylaminoethyl)-1-naphthylamine5-sulfonate group; Nbs2, 5,5'-dithiobis(2-nitrobenzoic acid). *To whom reprint requests should be addressed.
this fact.
2649
Biochemistry: Kelsey et al.
2650
Proc. Natl. Acad. Sci. USA 76 (1979) Table 1. AENS labeling of sulfhydryls
AENS/subunit 0.6 1.0 1.4 1.9 2.3
0.6 0.5 1.0 0.8 1.3 1.2 1.5 1.5 1.7 1.8 * The difference between the number of free sulfhydryls in unlabeled and AENS-labeled repressor was determined with Nbs2 (17). Unlabeled repressor gave 1.8 sulfhydryls/subunit when native and 2.7 sulfhydryls/subunit when denatured with 0.1% sodium dodecyl sulfate.
C
Sg2.0O 0
in Uj
Blocked sulfhydryls/subunit* Native Denatured
1.0 /
0
5
10
15
20
25
50
Time, hr
FIG. 1. Labeling of lac repressor with I-AENS. Repressor (10 ,M tetramers) was reacted with 8.1 mM I-AENS. The two exponential processes that adequately describe these data are shown as the solid line.
was used to fit these data to two exponential processes introducing 0.9 and 1.7 AENS labels per subunit with pseudo-firstorder rate constants of 1.3 hr-t and 0.083 hr-', respectively. These correspond to bimolecular rate constants of 4.3 X 10-2 M-I sec- and 2.8 X 10-3 M-' sec-t, respectively. In order to determine if I-AENS reacted exclusively with the three sulfhydryl groups per repressor subunit, we measured the number of blocked sulfhydryl groups at different extents of AENS la-
beling by using 5,5'-dithiobis(2-nitrobenzoic acid) (Nbs2). Nbs2 reacts with only two sulfhydryl groups per subunit under native conditions (6, 1 1), but will react with all three sulfhydryls under denaturing conditions (6). The results shown in Table 1, together with the data from Fig. 1, indicate that one sulfhydryl group per subunit reacts rapidly and a second sulfhydryl, together with an as yet unidentified residue, react slowly. The two sulfhydryls reactive toward I-AENS appear to be the same two that react with Nbs2 in the native state. Cysteine-140 Is Selectively Modified. Repressor, labeled to varying extents, was denatured; then unreacted sulfhydryls were alkylated with iodol '4Clacetamide. The protein was then digested with trypsin and the peptides were separated. Tryptic digestion produces three cysteine-containing peptides (10, 18) with the following molecular weights: Cys-140, 5361; Cys-281,
2860; and Cys-107, 693. Fig. 2A shows the results for repressor labeled with 0.8 AENS groups per subunit. Most of the fluorescence (72%) is associated with the peptide containing Cys140. An additional 14% of the fluorescence is associated with the peptide containing Cys-107. The position of this peptide
C 0)
250 Ci
o5
I0
~ ~
~
~
5
iI
A
Fraction FIG. 2. Separation of tryptic peptides from AENS-labeled repressor. The position otfthe sulfhydryl-containing peptides is marked with 14C (0 --- 0). The fluorescence intensity of AENS-containing peptides (O--*) was monitored with excitation anld emission wavelengths of 1340 and 490 nm, respectively. Both cpm and fluorescence intensities are plotted on relative scales. (A ) Procedure begun with 1.7/ mg of repressor with 0.8 AENS labels per subunit. Peak cpm (fraction 31) was 2058. (B) Procedure begun with 1.3< mg of repressor with 2.0 AENS labels per subunit. Peak cpm (fraction 31) was 1560. The relative scales for B have been normalized with respect, to A so that the total fluorescence intensity eluted in H is 2.5 times that eluted in A.
Proc. Natl. Acad. Sci. USA 76 (1979)
Biochemistry: Kelsey et al. in the elation profile is shifted by the AENS label, as shown by a comparison of Fig. 2 A and B. These results demonstrate that Cvs-140 reacts rapidly while Cys-107 reacts slowly. The same conclusion was reached by separating peptides produced by combined tryptic and chymotryptic digestion. Fig. 2B shows the results obtained starting with 2.0 AENS labels per subunit. Essentially all of Cys-140 and more than half of Cys- 107 have reacted, while Cys-281 remains unlabeled by AENS. The fluorescence under the Cys- 140 peak corresponds to 1.23 AENS labels per cysteine residue, indicating that another residue has reacted in addition to cysteine. Because the peptide containing Cys-140 is significantly larger than the next largest tryptic peptide (Mr = 42.33), this unidentified residue appears to be located in the same peptide as Cys-140, which spans residues 119-168. Most of our experiments have used repressor labeled with approximately 0.5 AENS groups per subunit to achieve selective labeling of Cys-140. This level of labeling does not alter the operator binding activity of repressor. Neither does it alter the ability of repressor to bind to polyld(A-T)J, as assayed by sedimenting mixtures of the two through glycerol gradients (13). Both unlabeled and AENS-labeled repressor are >95% active in complexing with polyfd(A-T)]. Repressor-Polyid(A-T)l Association Alters AENS Fluorescence. The fluorescence emission spectrum of repressor labeled with 0.46 AENS groups per subunit is shown in Fig. 3. The addition of saturating amounts of polyld(A-T)I decreased the fluorescence intensity at the emission maximum by 20% and red-shifted the emission maximum by 4 nm. Essentially identical results were obtained in 0.1 M Tris-HCI (pH 8.0) with repressor labeled with 0.25, 0.50, or 0.76 AENS groups per subunit. With 1.19 AENS groups per subunit the fluorescence decrease was only 14%, even though the protein was still fully active in binding to poly[d(A-T)I. At extents of labeling greater than 1 AENS group per subunit, the fraction of labels attached to residues other than Cys- 140 sharply increases. These results
2651
imply that the AENS labels attached to Cys-140 are primarily responsible for the fluorescence decrease. Changing the pH from 7.2 to 8.5 had little effect on the emission spectrum of repressor labeled with 0.47 AENS groups per subunit. The decrease in fluorescence intensity upon addition of polyId(A-T)I was independent of pH. AENS-labeled repressor was titrated with poly[d(A-T)J in order to determine the size of the repressor binding site on nonoperator DNA. With 0.50 AENS labels per subunit, 15 base pairs of poly[d(A-T)] per tetramer were required to complete the decrease in fluorescence (Fig. 4). The same result was obtained with repressor labeled with 0.25, 0.76, or 1.19 AENS groups per subunit. This again emphasizes that the ability of the repressor to bind to poly[d(A-T)] is not altered by these extents of AENS labeling. This titration was repeated in 10 mM Tris-HCI (pH 8.0) with the ionic strength adjusted to 30 mM with NaCl. Again, 15 base pairs per tetramer were required to reach the end point when repressor labeled with 0.46 AENS groups per subunit was used. Stopped-Flow Experiments Detect Two Conformational Changes. Stopped-flow techniques were used to monitor this decrease in fluorescence as AENS-labeled repressor bound to polyfd(A-T)I. Fig. 5 shows the results of one experiment in which repressor labeled with 0.66 AENS groups per subunit was reacted with 30 base pairs of poly[d(A-T)] per tetramer. In this reaction, the fluorescence decreased in two exponential processes, with half-lives of 4.4 msec and 4.2 sec. After extrapolation to the time of mixing, these two reactions accounted for 71% of the total decrease in fluorescence (11.3%) observed in the stopped-flow instrument by monitoring the emission from AENS-labeled repressor before and after the addition of poly[d(A-T)J. The ratio of the magnitudes of the fluorescence decreases in these fast and slow reactions was 1.4. The small differences between the calculated and observed points may be attributed to one or more of the following causes: instrumental noise and drift, selection of only 120 out of 1024 points for analysis, or additional reactions with magnitudes too small to be adequately characterized. These same two processes were observed under a wide variety of experimental conditions, as shown in Table 2. Varying the amount of polyfd(A-T)j from 5 to 60 base pairs per tetramer, while holding the concentration of repressor essentially con-
c
aL) C
a) a)
CD
c
V) 0
C
a)
U-
c
a1) el a) 0
U-
420
460 540 500 Wavelength, nm
580
Fi(,. 3. Fluorescence emission spectra of AENS-labeled repressor in the absence and presence of poly[d(A-T)l. Repressor (1.9 ,uM tetramers), labeled with 0.46 AENS labels per subunit, in 10 mM TrisHCIL pH 8.0/50 mM NaCi. (-) Without DNA; (---) with 48 base pairs of polyld(A-T)l per tetramer. Excitation wavelength, 340 nm.
10 50 40 20 30 Poly [d (A-T)I base pairs/repressor tetramer
FIG. 4. Titration of AENS-labeled repressor with polyld(A-T)l. Excitation and emission wavelengths, 340 and 475 nm, respectively. Starting solution: 250 ,tl, containing repressor (1.6 pM tetramers) with 0.50 AENS labels per subunit, in 0.10 M Tris-HCI (pH 8.0).
Biochemistry: Kelsey et al.
2652
Proc. Natl. Acad. Sci. USA 76 (1979)
Finally, varying the number of AENS labels per subunit from 0.41 to 0.91 had no effect on these two reactions. When all 29 trials at pH 8.0 and ionic strength 56 mM were included, the average half-lives ± SD of the fast and slow reactions were 5.0 1.2 msec and 3.5 + 1.0 sec, respectively. When the 16 trials at a different pH or ionic strength were averaged, the half-lives of the fast and slow reactions were 5.1 ± 1.1 msec and 3.5 + 1.0 sec, C
a)
5
92
