- Email: [email protected]
Thermal denaturation of engineered Tet repressor proteins and their complexes with tet operator and tetracycline studied by temperature gradient gel electrophoresis
ANALYTICAL
BIOCHEMISTRY
175422-432
(1988)
Thermal Denaturation of Engineered Tet Repressor Proteins and Their Complexes with tet Operator and Tetracycline Studied by Temperature Gradient Gel Electrophoresis MANFREDWAGENH&ER,DIETERHANSEN,'AND
WOLFGANG
HILLEN~
Lehrstuhl fir Mikrobiologie, Institut ftir Mikrobiologie und Biochemie der Friedrich-Alexander Universitiit Erlangen-Niirnberg, Staudtstrasse 5,852O Erlangen, Federal Republic of Germany Received June lo,1988 The effects of Trp to Phe exchanges in the Tet repressor on the thermal stability of the proteins and their complexes with operator DNA and inducer have been studied by temperature gradient polyacrylamide gel electrophoresis. The denaturation temperatures obtained by this method are compared with the results from temperature-dependent fluorescence and binding activities of the proteins. It is established that exchanging the interior Trp75 to Phe reduces the thermal stability of the Tet repressor by 8’C while exchanging the exterior Trp43 to Phe has no effect on the stability of the protein. Binding of the inducer tetracycline increases the thermal stability of wild-type and Trp43 to Phe mutant Tet repressors by YC, while the ones with the Trp75 to Phe mutation are stabilized by 1o’C. The stabilizing effect of operator binding is 20°C in the Txp75 to Phe mutant and only YC in the ones with the Trp43 to Phe exchange. In addition to the denaturation temperatures, the gel mobility shifts observed in temperature gradient gel electrophoresis reveal also information about the intermediates of the denaturation reaction. The free proteins and their complexes with the inducer tetracycline exhibit monophasic transitions upon denaturation. The operator complexes of wild-type and Trp75 to Phe mutant repressors denature in more complex reactions. At low temperature they exhibit a stoichiometry of two repressor dimers per tandem tet operator DNA. Upon elevating the temperature they form complexes with only one repressor dimer per DNA fragment. When the temperature is further increased the double-stranded DNA begins to melt from one end resulting in a complex with partially single-stranded DNA which exists only in a narrow temperature range. Finally, the denatured protein and single-stranded DNA are formed at high temperature. The associated mobility shifts are analyzed by changing the ionic strength and characterizing multiphasic melting of a pure DNA fragment by temperature gradient gel electrophoresis. Q 1988 Academic Fves, Inc. KEY WORDS: temperature gradient gel electrophoresis; Tet repressor; protein engineering; DNA-protein intera&ons; gene regulation.
We have recently reported the operator and inducer binding properties of mutant Tet repressor proteins containing Trp to Phe exchanges ( 1,2). It has been shown that Trp43 is solvent exposed and contributes a sequencespecific contact to tet operator recognition (I), while Trp75 is buried in the hydrophobic core of the repressor structure and influences inducer binding weakly (2). We were inter’ Present address: Henkel KGaA, TFB Biotechnologie, Henkelstr. 67,400O Diisseldorf, FRG. * To whom correspondence should be addressed. 0003-2697/88 $3.00 CopyrisM0 1988 by Academic All rights of reproduction
FT.%, Inc. in any form -ui.
422
ested in the thermal stability of the wild-type and mutant Tet repressors because we anticipated a small effect from the surface and an increased destabilization from the core mutation (3). Furthermore, we were interested in studying the effects of inducer and operator binding on the denaturation reactions. Thermal denaturation of the X repressor, a number of mutants of it, and its N-terminal arm has been extensively studied using differential scanning calorimetry, circular dichroism, and resistance versus proteolytic cleavage (3-6). We have observed the ther-
TEMPERATURE
GRADIENT
GEL
ma1 denaturations of Tet repressor proteins and their complexes with operator and inducer using a modified temperature gradient gel electrophoresis method which has recently been described for the analysis of nucleic acid melting (7-9). We describe here temperature-dependent mobility shifts of the proteins and their complexes in the polyacrylamide gel matrix leading to conclusions regarding the denaturation reactions. In the cases of operator-repressor complexes this method reveals multiphasic mobility shift transitions which serve to characterize the intermediates of the denaturation reaction. The results are compared to denaturation studies using fluorescence spectroscopy and nitrocellulose filter binding. MATERIALS
AND
METHODS
General methods. Preparation, handling, and characterization of the wild-type and mutant Tet repressor proteins were done exactly as described previously (1,2,10). The 187- and 456-bp tet control DNAs were prepared and characterized as described (11). Fluorescence spectroscopy (2) and nitrocellulose filter binding (11) were performed exactly as described. Temperature-dependent fluorescence studies. Solutions of the repressor protein dimers (0.5 PM) either alone or with 1 PM tetracycline, and 0.2 pM solutions of repressor protein dimers with 0.1 PM 187-bp DNA in 5 mM MgC&, 5 mM Tris-HCl, pH 8.0,O. 1 mM EDTA, and 0.1 mM dithiothreitol were filled in 1 X 1 cm (repressor-operator complexes in 0.4 X 1 cm) quartz cuvettes, placed into a tempered cuvette holder, and incubated for 30 min at the temperatures indicated in the respective figures. Excitation of fluorescence was at 280 nm and the emission was recorded at the wavelength of maximal tryptophan emission (330 to 350 nm). The slit widths were 5 nm for excitation and emission for the free proteins, 8 nm for the repressor-tetracycline complexes, and 10 nm for the repressor-operator complexes. No corrections were
ELECT
ROPHORESIS
OF
PROTEINS
423
made for inner filter effects or aggregation of the proteins at elevated temperatures. The ratio of the maximal fluorescence intensity at the respective temperature to the fluorescence intensity at 25°C was plotted versus temperature to produce the denaturation profiles. Temperature gradient polyacrylamide gel electrophoresis. The temperature gradient electrophoresis was performed in an apparatus similar to the one described previously for the analysis of nucleic acids (7). A horizontal copper plate was soldered between two copper blocks containing two parallel drillings through their long sides. The blocks were heated by different circulating water baths to temperatures defining the end points of the gradient. This construction was placed between two electrophoresis buffer chambers so that the distance from the block to the buffer was minimal and that the temperature gradient was oriented perpendicularly to the electric field. During electrophoresis the buffers in the two chambers were circulated using a peristaltic pump. A 5% polyacrylamide gel containing 0.26% bis-acrylamide and 10% glycerol in 40 mM Tris-HCl, 1.25 mM EDTA, and 10 or 40 mrvr sodium acetate (compare legends to the figures) adjusted to pH 8.3 was polymerized between 20 X 20 cm glass plates. One of the plates was covered with a gelbond sheet (FMC, Rockland, ME) which was covalently linked to the gel to avoid deformation at high temperature. Gels for the analysis of repressor-operator complexes and DNA fragments were 1 mm and those for the analysis of proteins and repressor-inducer complexes were 2 mm thick. At the upper part of the gel a wide slot (138 X 6 X 1 mm) flanked by two narrow slots (6 X 6 X 1 mm) was formed. After polymerization the gels were removed from the glass plates and placed on the copper block. The linearity of the temperature gradient was checked for several times by measuring the temperatures on top and on the bottom of the gel using a thermoresistor (Pt 100, FKG 430.4, obtained from Heraeus, Hanau, FRG). The deviation
424
WAGENHGFER,
HANSEN,
from a linear gradient was always within +- I “C. To determine the transition temperatures the midpoint of the respective transition was estimated and the temperature was taken from the linear gradient. Repeat experiments yielded transition temperatures within + 1“C. The gel was loaded with 500 or 600 ~1 sample treated as described in the legends to the respective figures. It was then electrophoresed at ambient temperatures in the buffer described above at 200 V for 1.5 h to assure entry of the sample into the gel matrix. Then the voltage was reduced to 30 V, the desired temperatures were set in the water baths, and the gradient was allowed to form for 50 min. Before elevating the temperature the gel was covered with a geIbond film topped by a glass plate and a Styropor lid to achieve insulation. Then the voltage was increased to 220 V for 1.5 h to develop the gel. When repressor-tetracycline complexes were run, 3 mM MgS04 was added to the buffer described above ( 12). The gels either were silver stained as published ( 13), except that incubation of the gel was for 30 min in 10% ethanol, 0.5% acetic acid; 40 min in 0.19% AgN03; four times for 1 min in water; 12 min in a freshly prepared solution of 15 g/liter NaOH, 3.75 ml/liter 37% formaldehyde and 50 mg/liter NaBH4; and finally 10 min in 0.75% Na2C03, or were stained with Coomassie blue. RESULTS
Thermal denaturation of the wild-type and mutant Tet repressor proteins. The TnlO-encoded Tet repressor protein sequence contains two tryptophan residues at positions 43 and 75 ( 14) which are conserved in the structures of three other Tet repressors from related tetracycline resistance determinants (15). To probe and distinguish the functions of each Trp in operator and tetracycline recognition the three possible mutant Tet repressors with Trp to Phe replacements, called F43, F75, and F43F75, respectively, were constructed (1). A detailed fluorescence study revealed no indications for different conformations of these proteins (2).
AND HILLEN
I
I
I
40
50
60
(“c 1
FIG. 1. Temperature gradient gel electrophoresis of Tet repressor proteins. A photograph of four temperature gradient gels containing the indicated Tet repressor proteins is shown. Each gel was loaded with 70 pg of the respective repressor protein in electrophoresis buffer (see Materials and Methods) containing 10 mM sodium acetate and 17% sucrose. The proteins were stained with Coomassie blue. The gels were adjusted with respect to the temperature as indicated at the bottom of the tigure. The temperature gradient over the sample extended from 35’C on the left to 70°C on the right.
The thermal stabilities of the wild-type and mutant Tet repressor proteins were compared by horizontal polyacrylamide gel electrophoresis with a linear temperature gradient ranging from 35 to 70°C oriented perpendicularly to the electric field. Details of the procedure are described under Materials and Methods. A photograph of the Coomassie stained gels is shown in Fig. 1. All four proteins exhibit increased mobilities at low temperature and show single transitions extending over a temperature range of 4°C to yield a nonmobile structure in the high temperature region of the gel. The denaturation of the proteins results in precipitation and immobilization or very small mobility of the repressors on the polyacrylamide gel (data not shown). The transitions are interpreted as the denaturation reaction of the proteins and indicate single cooperative denaturation processes.
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
425
OF PROTEINS
TABLE 1 THERMAL
STABILITIES
OF THE Tet REPRESWR
VARIANTS
AND THEIR COMPLEXES
WITH INDUCER
AND OPERATOR
Denaturation temperature of Tet repressor mutant (‘C) Sample Free protein Repressor-inducer
Repressor-operator
complex
complex
Method
Ionic condition“
wt
F75
F43
F43F75
tkzb eb @3b ncc fsd
10 mM Na+ 10 mM Na+, 3 mM Mg’+ lOm~Na+, 3rn~Me -= -e
Qzb k3b fsd
40mM Na+
51 49 54 59 56 64 61 58
43 41 52 59 56 63 60 n.d/
50 49 54 59 56 58 55 53
44 42 52 59 56 54 49 n.d/
lOmMNa+ -e
’ Ionic conditions are given in addition to electrophoresis buffer as described under Materials and Methods. b tgg, determined by temperature gradient gel electrophoresis with an accuracy of + 1°C. c nc, determined by nitrocellulose filter binding. d fs, determined by fluorescence spectroscopy. e -, as given under Materials and Methods. ‘n.d., not determined due to lack of signal.
The wild-type and F43 repressors denature at 5 1 and WC, respectively, while the F75 and F43F75 mutants denature at 43 and 44”C, respectively (compare Table 1). It has been shown previously that Trp43 is solvent exposed while Trp75 is located in the hydrophobic core of the protein (2). Only the latter seems to contribute to the stability of the repressor since replacing it by Phe destabilizes the respective mutants by about 8°C.
Thermal stability of thefour repressor-tetracycline complexes. The thermal denaturation profiles of the wild-type and mutant Tet repressors were compared with the ones of their complexes with tetracycline on the same temperature gradient gels. The results are shown in Fig. 2. Since these gels were run in the presence of 3 mM Mg2+ it is possible to evaluate the inhuence of Mg2+ ions on the stability of the free repressor proteins. Comparison of the respective denaturation temperatures in Table 1 indicates that 3 mM Mg2+ destabilizes the wild-type and mutant proteins slightly by about 2°C. Although this is close to the error range of the temperature determination (It l”C, see Materials and
Methods) this decreased stability is consistently observed for all four proteins. Comparison of the stabilities of the free proteins with the respective tetracycline complexes indicates that binding of the inducer stabilizes all four repressor variants. The denaturation temperatures of the wild-type and mutant complexes are nearly identical around 53°C. The wild-type and F43 repressors are stabilized by about 5°C and the F75 and F43F75 repressors by about 10°C by inducer binding. The stabilizations of proteins in the complexes are summarized in table 2. The narrow transitions indicate single cooperative denaturation reactions of the complexes. The thermal denaturations of the repressor-tetracycline complexes were also measured by nitrocellulose filter binding and tetracycline fluorescence. The results are shown in Figs. 3 and 4, respectively. All four repressor-tetracycline complexes show nearly identical thermal stabilities in the nitrocellulose filter binding assay as well as in the temperature-dependent fluorescence measurements. In both cases we observed cooperative transi-
426
WAGENHQFER,
40
50
HANSEN, AND HILLEN
60
40
50
60
I^ .
(‘C 1
FIG. 2. Temperature gradient gel electrophoresis of Tet repressor-inducer complexes. The photographs of four Coomassie stained gels are shown. Each gel contained two slots; the upper was loaded with 70 pg of the respective repressor protein in 600 ~1 of a buffer containing 3 mM MgS04 in addition to the components given in the legend to Fig. 1. The lower slot was loaded with the same solution containing a IO-fold molar excess of tetracycline over repressor preincubated for 10 min at 37°C. The gels are aligned with respect to the temperature as indicated in the middle of the figure. Each temperature gradient ranged from 35 to 7o’c.
tions of the repressor-tetracycline complexes and identical denaturation temperatures regardless of the repressor mutation. Denaturations take place at 59°C in nitrocellulose filter binding and at 56°C in the fluorescence measurements (see Table 1). The curvature of the temperature-dependent fluorescence is not as sigmoidal as that obtained from nitrocellulose filter binding. In the temperature
range from 25 to 45°C there is a gradual decrease of fluorescence which could be due to alterations of the protein structure, a process which may precede the denaturation (16). Quantification of the Trp fluorescence was complicated by aggregation of the proteins which occurs along with denaturation as can be detected by the increasing turbidity of the solution.
TABLE 2 STABILIZATION
OF Tet REPRESSOR VARIANTS COMPARED WITH
COMPLEXED WITH THE FREE PROTEINS
INDUCER
AND
OPERATOR
Increase in denaturation temperature of Tet repressor variant (“C) Ligand Tetracycline Operator
Ionic condition” 10mMNa+,3mMMg2+
lOmMNa+
F7.5
F43
F43F75
5
11
20
5 8
10
13
wt
10
11The ionic conditions are given in addition to the electrophoresis buffer described under Materials and Methods.
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
I 20
30
40
50
Temperature
60
70
Thermal stabilities of the repressor-operator complexes. The temperature-dependent denaturation of the Tet repressor variants 1.0
F /6 0.6
I
z. F43 - F75 + F43F75
35
45
Temperature
55
I 40 Temperature
("Cl
FIG. 3. Temperature-dependent nitrocellulose filter binding assaysof Tet repressor-tetracycline complexes, Solutions of represso-[3H]tetracycline complexes (1 PM) were incubated at the indicated temperatures over a period of 15 min, chilled briefly on ice, and filtered immediately over nitrocellulose. The retained radioactivity of the repressor-tetracycline complexes (given in percentage of the radioactivity bound at 4o’C) was plotted versus the incubation temperature.
0.2 -
I 30
65 (“C)
FIG. 4. Temperature-dependent fluorescence of the repressor-tetracycline complexes. Fluorescence was excited at 280 nm. Emission was recorded at 5 10 nm. The concentration of repressor-tetracycline complexes was I pM. The ratio FJF, with F = fluorescence intensity at the temperature T, and F. = fluorescence intensity at 25’C is plotted versus the temperature.
427
OF PROTEINS
50
60
70
(“C)
FIG. 5. Temperature-dependent fluorescence of the wt and F43 repressor-operator complexes. Fluorescence was excited at 280 nm and emission was recorded at 330 nm. The concentrations were 0.2 pM repressor dimer and 0.1 pM 187-bp DNA containing the tandem ret operator arrangement (11). The ratio F/F,, with F = maximal fluorescence intensity at 25°C was plotted versus temperature.
complexed with the 187-bp tet operator DNA (11) has been studied employing fluorescence spectroscopy and temperature gradient gel electrophoresis. Since the fluorescence of Trp43 is totally quenched in the complex with tet operator and is not recovered by denaturation of the complex (data not shown) only the Trp75 containing Tet repressors could be studied with this method (1). Figure 5 displays the respective denaturation curves. The fluorescence properties of Trp75 are not affected by complex formation with the tet operator (I). Thus, the temperature-dependent fluorescence intensities of the complexes in Fig. 5 reflect only the denaturation of the proteins. Between 25 and 45°C the same gradual decrease of fluorescence intensity is observed as described above for the complexes with tetracycline. Above 45°C the denaturation curve is sigmoidal. The denaturation temperatures are 53°C for the F43-tet operator complex and 58°C for the wildtype-tet operator complex (see Table I). The reduced stability of the mutant complex corresponds to its reduced affinity for the tet operator (1). The denaturation of the F43 mu-
428
WAGENHijFER,
HANSEN, AND HILLEN
tant in this complex is clearly monophasic while the wild-type denaturation appears to be more complex. The resolution of the fluorescence measurements does not allow a quantitative analysis of this denaturation curve; however, it can be compared to the results obtained by temperature gradient gel electrophoresis (see below). Temperature gradient gels of the repressor-operator complexes were run at 10 and 40 mM sodium acetate. This was done because the T, of the DNA fragment is strongly ionic strength dependent (17) while the thermal stability of the complex is expected to be less sensitive to it. The results of the temperature gradient gel electrophoresis experiments are displayed in Fig. 6. The respective gels were silver stained under conditions visualizing predominantly the DNA. When the gels were run with a lo-fold increased amount of complex and stained with Coomassie blue the denatured protein is located at the high temperature region of the gel (data not shown). The denaturation profiles are complex in all cases consisting of at least two transitions. At lower temperatures the repressoroperator complexes show reduced electrophoretic mobilities in comparison to the uncomplexed 187-bp DNA. At higher temperatures the repressor-operator complexes denature leading to the free operator DNA as indicated by the increased mobility. At the high temperature end of the gel the doublestranded DNA melts and the single strands exhibit a reduced mobility compared to the free double-stranded DNA (compare also below). In some cases at low ionic strength these denaturation events are not well separated. The temperature gradient gels of these complexes yield two kinds of information with respect to denaturation. The first contains the denaturation temperatures. The wild-type and F75 repressor-operator complexes denature at 64 and 63°C at 10 mM Na+ and at 61 and 60°C at 40 mM Na+, respectively. The F43 and F43F75 repressor-operator complexes denature at 58 and 54°C in 10 mM Na+ and at 55 and 49°C at 40 mM Na+,
. I. @. 1
FIG. 6. Temperature gradient gel electrophoresis of Tet repressor-tet operator complexes. Photographs of silver stained gels are shown. The four gels displayed on the left side contained 400 ng 187-bp tet operator DNA and an eight-fold molar excess of the wild-type or F75 repressor or a 20-fold molar excessof the F43 or F43F75 repressor as indicated. The mixtures were preincubated in 500 ~1 of electrophoresis buffer (see Materials and Methods) containing 10 mM sodium acetate and 5% glycerol for 15 min at 37°C. Electrophoresis was carried out in the same buffer. The four gels displayed on the right side contained the same samples except that the concentration of sodium acetate was 40 mM. The two vertical rows of gels are adjusted with respect to the temperature which is indicated at the bottom. The filled triangles on the left side denote the position of the 187-bp DNA complexed with two Tet repressor dimers while the open triangles denote the positions of the 187-bp DNA complexed with one Tet repressor dimer. The lanes M contain DNA hagments resulting from a HaeIII/EcoRI digest of pWH802 ( 15) and the 187-bp DNA; the lanes S contain only the free 187-bp DNA fragment. The temperature gradients extended from 35 to 70°C.
respectively (compare also Table 1). Table 2 shows the thermal stabilizations of the Tet repressor proteins in the complexes over the free proteins. It may be derived that the wildtype repressor is stabilized by 13°C and the F75 mutant by 2o’C to reach roughly identical denaturation temperatures. The destabilizing effect of the Phe75 mutation in the free protein is completely neutralized by the stabilizing effect of tet operator binding in the complex. The Tet repressor mutants with the
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
Trp43 to Phe substitution, on the other hand, are stabilized only by 8°C (F43) and 10°C (F43F75). This corresponds well to their reduced binding constant for operator DNA (1). Comparing the denaturation temperatures at 10 and 40 mM Na+ reveals that the complexes are slightly destabilized by this small increase in ionic strength. This also corresponds to the destabilizing effect of ions on the association constant of Tet repressor&t operator binding ( 18). The second kind of information drawn from the temperature gradient gel electrophoresis concerns the intermediates of the denaturation reactions. Upon increasing the temperature in the gels shown in Fig. 6 the 187-bp DNA complexed with two dimers of wild-type or F75 Tet repressors dissociate in two steps at both ionic strengths. The F43and F43F7Soperator complexes, on the other hand, denature at both salt concentrations in a single transition. In the complexes with F43 and F43F75 repressors at the high and low salt concentrations the melting of the double-stranded 187-bp DNA is clearly resolved from the denaturation of the protein-DNA complexes (compare also below). In the low salt experiments with the wild-type and F75 repressors the DNA melting and denaturation of the protein-DNA complexes occur in a narrow temperature range. In these experiments an intermediate species shows up in the denaturation reaction which exhibits a greatly reduced mobility in the gel. It exists during 1 or 2°C and denatures then to the nonmobile protein (not shown) and the single-stranded DNA (see Fig. 6). In order to facilitate the interpretation of these results we have examined the denaturation of pure DNA fragments on temperature gradient gels. Denaturation of the 187- and 456bp tet operator DNAs on temperature gradient gels. Figure 7 displays the temperature gradient gels of the 187-bp DNA fragment at 10 and 40 mM Na+. Under both conditions the DNA fragment denatures in a single sharp transition. The melting temperatures are 62 and
429
OF PROTEINS
M
8
40 50 60
(“0
FIG. 7. Temperature gradient gel electrophoreses of the 187-bp DNA fragment. Photographs of the silver stained gels run with 400 ng of free I87-bp DNA in 500 ~1 of the buffer given in the legend to Fig. 6 with 10 mM sodium acetate (upper) and 40 mM sodium acetate (lower) are shown. The gels were aligned with respect to temperature as indicated on the bottom of the figure. The lanes M and S are as in Fig. 6. The temperature gradient extended from 35 to 70°C.
64’C, respectively. They are in fair agreement with results obtained from optical melting studies of the same DNA (17). The same sharp transitions are also found in most denaturation experiments of the protein-DNA complexes. In order to interpret the intermediate species with reduced mobility described above, the melting of a 456-bp tet control DNA fragment was examined by the temperature gradient gel method. In optical melting studies it showed two clearly separated transitions (17). The temperature gradient gel is shown in Fig. 8. In the first transition the DNA fragment denatures to a partially single-stranded structure at one end ( 17). Interestingly, the mobility of this DNA is greatly reduced in the gel matrix. Its mobility is gradually further reduced upon increasing the temperature. In a second transition this partially singlestranded DNA melts completely resulting in the two single strands. These exhibit an increased mobility compared with the partially denatured structure, which increases further upon raising the temperature. DISCUSSION
We have used temperature gradient gels to analyze the denaturation of wild-type and
430
WAGENHijFER,
HANSEN,
AND HILLEN
are higher with this method (see Table 1). We conclude that temperature gradient gels may lead to slightly different denaturation temperatures as compared to free solutions. However, the comparison of thermal stabilities of mutants and their complexes leads to consistent results. This conclusion is also derived from the melting of the DNA fragments. When studied in solution under slightly lower I I I ionic strength (17) the T, values are roughly 50 60 70 ("Cl 5°C lower. This could be attributed to the FIG. 8. Temperature gradient gel electrophoresis of the cage effect assumed to stabilize complexes in 456-bp tet control DNA. A photograph of the silver gels (2 1). On the other hand, stained gel is shown. It was loaded with 400 ng of the polyacrylamide the T, difference between the two transitions 456-bp DNA in 500 pl electrophoresis buffer (see Materials and Methods) containing 10 mM sodium acetate and in the 456-bp fragment (formerly called 501 5% glycerol. The temperature scale is given on the botbp) is about 3°C as was observed in solution tom of the figure. The lane M contains the standard as in ( 17). Since the first transition is monomolecFig. 6 with the 456-bp DNA and the lane S contains only ular and the second is bimolecular, caging the 456-bp DNA. The temperature gradient extended would be expected to increase this T,,, differfrom 40 to 80°C. ence. This argues against a cage effect and is in agreement with observations made with dissociation reactions (22). mutant Tet repressor proteins alone and in protein-DNA their complexes with inducer and operator. The small differences in the denaturation The temperature gradient gels offer an advan- temperatures would then be attributed to tage compared to other methods because, in changes in ionic strength. Comparison of the gel mobility shift transiaddition to the determination of denaturation temperatures, they allow the analysis of tions resulting from the denaturation of progel mobilities of the native and denatured teins and DNA fragments indicates that the forms detecting possible intermediates in the latter is a highly cooperative process occurdenaturation reaction. The limitation is, of ring in a narrow temperature range. Denaturcourse, that only those transitions which alter ation of the proteins and their complexes, on the other hand, occurs within a broader temthe electrophoretic mobility can be analyzed. Therefore, we compared the results obtained perature range of approximately 4°C. Thus, from the temperature gradient gels with the this property limits the accuracy of T, deterones obtained from temperature-dependent mination rather than the reproducibility of the temperature gradient. fluorescence changes and binding activities The Tet repressor denatures in a single determined by nitrocellulose filter binding assays. These results are interpreted under the transition in the temperature gradient gel (see assumption that the decrease in fluorescence Fig. 1). This implies that, unlike, e.g., the X results from denaturation of the protein as repressor (4), this protein may consist of only supported by the cooperativity of the respec- a single domain. Alternatively, the denaturation of potential other domains may not intive transition curves. Since these measurements were done in different buffers and at fluence the gel mobility. Scanning calorimedifferent heating rates one cannot expect try of these proteins reveals also single sharp identical denaturation temperatures. For the transitions supporting the first assumption repressor-inducer complexes they tend to be (Hinz et al., to be published). The Phe75 mulower in the temperature gradient gels, while tation, which was shown to be located in the the ones of the repressor-operator complexes hydrophobic core of the repressor (2), has a
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
large destabilizing effect while the Phe43 mutation on the surface of the protein (2) does not reduce the stability. This resembles results obtained with X repressor mutants (5) and confirms the locations of those residues in the repressor structure. Binding of the inducer tetracycline stabilizes the wild-type repressor by 5°C and the ones containing the Phe75 mutation by 10°C. Thus, inducer binding compensates fully for the destabilizing Phe75 mutation. All protein-tetracycline complexes denature at the same temperature in denaturation studies using temperature gradient gels (see Fig. 2), nitrocellulose filter binding (see Fig. 3), and fluorescence spectroscopy (see Fig. 4). The denaturation temperatures obtained with the three methods vary from about 53°C in the temperature gradient gels to 56°C in the fluorescence measurements and 59°C in the nitrocellulose filter binding experiments. This may be due to different buffer conditions or differences in handling of the samples in these methods. The results from all three methods indicate that the repressor-inducer complexes denature cooperatively in a single transition. It is noted that the free proteins are slightly destabilized in the presence of 3 mM Mg*+ (compare Figs. 1 and 2 and Table 1). The reason for this is not clear. The denaturations of the repressor-operator complexes clearly reveal the additional information obtained from temperature gradient gel electrophoresis in comparison to other methods of studying these complex reactions. In the simple cases the temperature-induced decompositions of the repressor-operator complexes yield in the first step the precipitated protein and double-stranded DNA followed by melting of the DNA. These mechanisms are found for the F43 and F43F75 complexes at both salt concentrations. The repressors binding tet operator DNA with high affinity (wild-type and F75) (1) form complexes which denature via a different pathway. The 187-bp DNA containing the tandem tet operator arrangement ( 11,19) can bind one or two Tet repressor dimers result-
OF PROTEINS
431
ing in different mobility shifts (20). First the complex of the 187-bp DNA with two repressor dimers denatures to the complex of the 187-bp DNA with one repressor dimer which then denatures at high salt to the doublestranded DNA followed by DNA melting. At low salt the thermal stability of the DNA is decreased to an extent where part of it denatures to single-stranded ends at temperatures below protein denaturation in the proteinDNA complex. This is indicated by the intermediate with greatly reduced gel mobility in Fig. 6. The same denaturation mechanism had been proposed before from melting studies at low salt ( 17,23). The denaturation temperatures of the complexes clearly correlate with the affinities of the Tet repressor variants for tet operator (1) and not with the stability of the free proteins. This is indicated by the data obtained with the F43 and F75 mutants (see Table 2). The free 187- and 456-bp DNA fragments show quite different behavior on the temperature gradient gels. By hyperchromicity studies it had been shown previously that the 187bp DNA denatures in a single cooperative transition while the 456-bp DNA denatures in two cooperative transitions separated by about 3.5”C (17,23). The same behavior is found on the temperature gradient gel (see Figs. 7 and 8). The first transition of the 456-bp DNA leads to a partially singlestranded structure (17). This exhibits a lower mobility in the gel as the single strands. The same observation has been made for the denaturation of double-stranded RNA species previously (7,8). The partially single-stranded structure shows decreased mobility in the polyacrylamide gel as the temperature is raised. This could indicate increased fraying of the double stranded ends. However, this is unlikely because the ends of the completely double-stranded DNA should also increase their fraying on the low temperature side of the gel and yet its mobility increases. Thus, it is concluded that the flexibility of the singlestranded ends causes the temperature-dependent increase in gel retardation.
WAGENHGFER,
432
HANSEN,
ACKNOWLEDGMENTS We thank Dr. D. Riesner and V. Rosenbaum for their help with the temperature gradient gels and Mrs. R. Kisse1for typing the manuscript. This work was supported by grants from the Deutsche Forschungsgemeinsch and the Fonds der chemischen Industrie.
REFERENCES 1. Hansen, D., and Hillen, W. (1987) J. Biol. Chem.
262,12,269-12,274. 2. Hansen, D., Altschmied, L., and Hillen, W. ( 1987) J. Biol.Chem.262,14,030-14,036.
AND HILLEN
10. Gehmichen, R., Klock, G., Altschmied, L., and Hillen, W. (1984)EMBO J. 3,539-543. 11. Hillen, W., Klock, G., Kaffenberger, I., Wray, L. V., and Reznikoff, W. S.(1982) J. Biol. Chem.257, 6605-6613. 12. Takahashi, M., Altschmied, L., and Hillen, W. (1986)J. Mol. Biol.187,341-348. 13. Sammons, D. W., Adams, C. D., and Nishizawa, E. E. (1982) Electrophoresis 2,135-141. 14. Postle, K., Nguyen, T. T., and Bertrand, K. P. (1984)
NucleicAcidsRes.12,4849-4863. 15. Unger, B., Klock, G., and Hillen, W. (1984) Nucleic
Acids.Res.12,7693-7703.
16. privalov,P. L. (1979) Adv. ProteinChem.33, 1673. Hecht, M. H., Nelson H. C. M., and Sauer, R. T. 241. (1983)Proc.Natl.Acad.Sci. USA80,2676-2680. 17. Hillen, W., Unger, B., and Klock, G. (1982) Nucleic 4. Pabo, C. O., Sauer, R. T., Sturdevant, J. M., and AcidsRes.10,6085-6097. Ptashne, M. ( 1979) Proc.Natl.Acad.Sci.USA76, 18. Kleinschmidt, C., Tovar, K., Hillen, W., and Piir1608-1612. schke, D. (1988) Biochemistry27,1094-1104. 5. Hecht, M. H., Sturdevant, J. M., and Sauer, R. T. 19. Bertrand, K. P., Postle, K., Wray, L. V., Jr., and Rez(1984)Proc. Nati. Acad.Sci. USA81,5685-5689. nikoff, W. S. (1983) Gene23, 149-156. 6. Hecht, M. H., Sturdevant, J. M., and Sauer, R. T. 20. Hillen, W., Gatz, C., Altschmied, L., Schollmeier, (1986)Proteins:Struct.Funct.Genet.1,43-46. K., and Meier, I. (1983) J. Mol. Biol. 169,7077. Rosenbaum, V., and Riesner, D. (1987) Biophys. 722. Chem.26,235-246. 21. Fried, M., and Crothers, D. M. (1981) NucleicAcids 8. PO, T., Steger, G., Rosenbaum, V., Kaper, J., and Res.9,6505-6516. Riesner, D. (1987) NucleicAcidsRes.15, 5069- 22. Revzin, A., Ceglarek, J. A., and Garner, M. M. ( 1986)Anal. Biochem.153,172-l77. 5084. 9. Steger, G., PO, T., Kaper, J., and Riesner, D. ( 1987) 23. Hillen, W., and Unger, B. (1982) Nature(London) 297,700-702. NucleicAcids Res.15,5085-5104.
BIOCHEMISTRY
175422-432
(1988)
Thermal Denaturation of Engineered Tet Repressor Proteins and Their Complexes with tet Operator and Tetracycline Studied by Temperature Gradient Gel Electrophoresis MANFREDWAGENH&ER,DIETERHANSEN,'AND
WOLFGANG
HILLEN~
Lehrstuhl fir Mikrobiologie, Institut ftir Mikrobiologie und Biochemie der Friedrich-Alexander Universitiit Erlangen-Niirnberg, Staudtstrasse 5,852O Erlangen, Federal Republic of Germany Received June lo,1988 The effects of Trp to Phe exchanges in the Tet repressor on the thermal stability of the proteins and their complexes with operator DNA and inducer have been studied by temperature gradient polyacrylamide gel electrophoresis. The denaturation temperatures obtained by this method are compared with the results from temperature-dependent fluorescence and binding activities of the proteins. It is established that exchanging the interior Trp75 to Phe reduces the thermal stability of the Tet repressor by 8’C while exchanging the exterior Trp43 to Phe has no effect on the stability of the protein. Binding of the inducer tetracycline increases the thermal stability of wild-type and Trp43 to Phe mutant Tet repressors by YC, while the ones with the Trp75 to Phe mutation are stabilized by 1o’C. The stabilizing effect of operator binding is 20°C in the Txp75 to Phe mutant and only YC in the ones with the Trp43 to Phe exchange. In addition to the denaturation temperatures, the gel mobility shifts observed in temperature gradient gel electrophoresis reveal also information about the intermediates of the denaturation reaction. The free proteins and their complexes with the inducer tetracycline exhibit monophasic transitions upon denaturation. The operator complexes of wild-type and Trp75 to Phe mutant repressors denature in more complex reactions. At low temperature they exhibit a stoichiometry of two repressor dimers per tandem tet operator DNA. Upon elevating the temperature they form complexes with only one repressor dimer per DNA fragment. When the temperature is further increased the double-stranded DNA begins to melt from one end resulting in a complex with partially single-stranded DNA which exists only in a narrow temperature range. Finally, the denatured protein and single-stranded DNA are formed at high temperature. The associated mobility shifts are analyzed by changing the ionic strength and characterizing multiphasic melting of a pure DNA fragment by temperature gradient gel electrophoresis. Q 1988 Academic Fves, Inc. KEY WORDS: temperature gradient gel electrophoresis; Tet repressor; protein engineering; DNA-protein intera&ons; gene regulation.
We have recently reported the operator and inducer binding properties of mutant Tet repressor proteins containing Trp to Phe exchanges ( 1,2). It has been shown that Trp43 is solvent exposed and contributes a sequencespecific contact to tet operator recognition (I), while Trp75 is buried in the hydrophobic core of the repressor structure and influences inducer binding weakly (2). We were inter’ Present address: Henkel KGaA, TFB Biotechnologie, Henkelstr. 67,400O Diisseldorf, FRG. * To whom correspondence should be addressed. 0003-2697/88 $3.00 CopyrisM0 1988 by Academic All rights of reproduction
FT.%, Inc. in any form -ui.
422
ested in the thermal stability of the wild-type and mutant Tet repressors because we anticipated a small effect from the surface and an increased destabilization from the core mutation (3). Furthermore, we were interested in studying the effects of inducer and operator binding on the denaturation reactions. Thermal denaturation of the X repressor, a number of mutants of it, and its N-terminal arm has been extensively studied using differential scanning calorimetry, circular dichroism, and resistance versus proteolytic cleavage (3-6). We have observed the ther-
TEMPERATURE
GRADIENT
GEL
ma1 denaturations of Tet repressor proteins and their complexes with operator and inducer using a modified temperature gradient gel electrophoresis method which has recently been described for the analysis of nucleic acid melting (7-9). We describe here temperature-dependent mobility shifts of the proteins and their complexes in the polyacrylamide gel matrix leading to conclusions regarding the denaturation reactions. In the cases of operator-repressor complexes this method reveals multiphasic mobility shift transitions which serve to characterize the intermediates of the denaturation reaction. The results are compared to denaturation studies using fluorescence spectroscopy and nitrocellulose filter binding. MATERIALS
AND
METHODS
General methods. Preparation, handling, and characterization of the wild-type and mutant Tet repressor proteins were done exactly as described previously (1,2,10). The 187- and 456-bp tet control DNAs were prepared and characterized as described (11). Fluorescence spectroscopy (2) and nitrocellulose filter binding (11) were performed exactly as described. Temperature-dependent fluorescence studies. Solutions of the repressor protein dimers (0.5 PM) either alone or with 1 PM tetracycline, and 0.2 pM solutions of repressor protein dimers with 0.1 PM 187-bp DNA in 5 mM MgC&, 5 mM Tris-HCl, pH 8.0,O. 1 mM EDTA, and 0.1 mM dithiothreitol were filled in 1 X 1 cm (repressor-operator complexes in 0.4 X 1 cm) quartz cuvettes, placed into a tempered cuvette holder, and incubated for 30 min at the temperatures indicated in the respective figures. Excitation of fluorescence was at 280 nm and the emission was recorded at the wavelength of maximal tryptophan emission (330 to 350 nm). The slit widths were 5 nm for excitation and emission for the free proteins, 8 nm for the repressor-tetracycline complexes, and 10 nm for the repressor-operator complexes. No corrections were
ELECT
ROPHORESIS
OF
PROTEINS
423
made for inner filter effects or aggregation of the proteins at elevated temperatures. The ratio of the maximal fluorescence intensity at the respective temperature to the fluorescence intensity at 25°C was plotted versus temperature to produce the denaturation profiles. Temperature gradient polyacrylamide gel electrophoresis. The temperature gradient electrophoresis was performed in an apparatus similar to the one described previously for the analysis of nucleic acids (7). A horizontal copper plate was soldered between two copper blocks containing two parallel drillings through their long sides. The blocks were heated by different circulating water baths to temperatures defining the end points of the gradient. This construction was placed between two electrophoresis buffer chambers so that the distance from the block to the buffer was minimal and that the temperature gradient was oriented perpendicularly to the electric field. During electrophoresis the buffers in the two chambers were circulated using a peristaltic pump. A 5% polyacrylamide gel containing 0.26% bis-acrylamide and 10% glycerol in 40 mM Tris-HCl, 1.25 mM EDTA, and 10 or 40 mrvr sodium acetate (compare legends to the figures) adjusted to pH 8.3 was polymerized between 20 X 20 cm glass plates. One of the plates was covered with a gelbond sheet (FMC, Rockland, ME) which was covalently linked to the gel to avoid deformation at high temperature. Gels for the analysis of repressor-operator complexes and DNA fragments were 1 mm and those for the analysis of proteins and repressor-inducer complexes were 2 mm thick. At the upper part of the gel a wide slot (138 X 6 X 1 mm) flanked by two narrow slots (6 X 6 X 1 mm) was formed. After polymerization the gels were removed from the glass plates and placed on the copper block. The linearity of the temperature gradient was checked for several times by measuring the temperatures on top and on the bottom of the gel using a thermoresistor (Pt 100, FKG 430.4, obtained from Heraeus, Hanau, FRG). The deviation
424
WAGENHGFER,
HANSEN,
from a linear gradient was always within +- I “C. To determine the transition temperatures the midpoint of the respective transition was estimated and the temperature was taken from the linear gradient. Repeat experiments yielded transition temperatures within + 1“C. The gel was loaded with 500 or 600 ~1 sample treated as described in the legends to the respective figures. It was then electrophoresed at ambient temperatures in the buffer described above at 200 V for 1.5 h to assure entry of the sample into the gel matrix. Then the voltage was reduced to 30 V, the desired temperatures were set in the water baths, and the gradient was allowed to form for 50 min. Before elevating the temperature the gel was covered with a geIbond film topped by a glass plate and a Styropor lid to achieve insulation. Then the voltage was increased to 220 V for 1.5 h to develop the gel. When repressor-tetracycline complexes were run, 3 mM MgS04 was added to the buffer described above ( 12). The gels either were silver stained as published ( 13), except that incubation of the gel was for 30 min in 10% ethanol, 0.5% acetic acid; 40 min in 0.19% AgN03; four times for 1 min in water; 12 min in a freshly prepared solution of 15 g/liter NaOH, 3.75 ml/liter 37% formaldehyde and 50 mg/liter NaBH4; and finally 10 min in 0.75% Na2C03, or were stained with Coomassie blue. RESULTS
Thermal denaturation of the wild-type and mutant Tet repressor proteins. The TnlO-encoded Tet repressor protein sequence contains two tryptophan residues at positions 43 and 75 ( 14) which are conserved in the structures of three other Tet repressors from related tetracycline resistance determinants (15). To probe and distinguish the functions of each Trp in operator and tetracycline recognition the three possible mutant Tet repressors with Trp to Phe replacements, called F43, F75, and F43F75, respectively, were constructed (1). A detailed fluorescence study revealed no indications for different conformations of these proteins (2).
AND HILLEN
I
I
I
40
50
60
(“c 1
FIG. 1. Temperature gradient gel electrophoresis of Tet repressor proteins. A photograph of four temperature gradient gels containing the indicated Tet repressor proteins is shown. Each gel was loaded with 70 pg of the respective repressor protein in electrophoresis buffer (see Materials and Methods) containing 10 mM sodium acetate and 17% sucrose. The proteins were stained with Coomassie blue. The gels were adjusted with respect to the temperature as indicated at the bottom of the tigure. The temperature gradient over the sample extended from 35’C on the left to 70°C on the right.
The thermal stabilities of the wild-type and mutant Tet repressor proteins were compared by horizontal polyacrylamide gel electrophoresis with a linear temperature gradient ranging from 35 to 70°C oriented perpendicularly to the electric field. Details of the procedure are described under Materials and Methods. A photograph of the Coomassie stained gels is shown in Fig. 1. All four proteins exhibit increased mobilities at low temperature and show single transitions extending over a temperature range of 4°C to yield a nonmobile structure in the high temperature region of the gel. The denaturation of the proteins results in precipitation and immobilization or very small mobility of the repressors on the polyacrylamide gel (data not shown). The transitions are interpreted as the denaturation reaction of the proteins and indicate single cooperative denaturation processes.
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
425
OF PROTEINS
TABLE 1 THERMAL
STABILITIES
OF THE Tet REPRESWR
VARIANTS
AND THEIR COMPLEXES
WITH INDUCER
AND OPERATOR
Denaturation temperature of Tet repressor mutant (‘C) Sample Free protein Repressor-inducer
Repressor-operator
complex
complex
Method
Ionic condition“
wt
F75
F43
F43F75
tkzb eb @3b ncc fsd
10 mM Na+ 10 mM Na+, 3 mM Mg’+ lOm~Na+, 3rn~Me -= -e
Qzb k3b fsd
40mM Na+
51 49 54 59 56 64 61 58
43 41 52 59 56 63 60 n.d/
50 49 54 59 56 58 55 53
44 42 52 59 56 54 49 n.d/
lOmMNa+ -e
’ Ionic conditions are given in addition to electrophoresis buffer as described under Materials and Methods. b tgg, determined by temperature gradient gel electrophoresis with an accuracy of + 1°C. c nc, determined by nitrocellulose filter binding. d fs, determined by fluorescence spectroscopy. e -, as given under Materials and Methods. ‘n.d., not determined due to lack of signal.
The wild-type and F43 repressors denature at 5 1 and WC, respectively, while the F75 and F43F75 mutants denature at 43 and 44”C, respectively (compare Table 1). It has been shown previously that Trp43 is solvent exposed while Trp75 is located in the hydrophobic core of the protein (2). Only the latter seems to contribute to the stability of the repressor since replacing it by Phe destabilizes the respective mutants by about 8°C.
Thermal stability of thefour repressor-tetracycline complexes. The thermal denaturation profiles of the wild-type and mutant Tet repressors were compared with the ones of their complexes with tetracycline on the same temperature gradient gels. The results are shown in Fig. 2. Since these gels were run in the presence of 3 mM Mg2+ it is possible to evaluate the inhuence of Mg2+ ions on the stability of the free repressor proteins. Comparison of the respective denaturation temperatures in Table 1 indicates that 3 mM Mg2+ destabilizes the wild-type and mutant proteins slightly by about 2°C. Although this is close to the error range of the temperature determination (It l”C, see Materials and
Methods) this decreased stability is consistently observed for all four proteins. Comparison of the stabilities of the free proteins with the respective tetracycline complexes indicates that binding of the inducer stabilizes all four repressor variants. The denaturation temperatures of the wild-type and mutant complexes are nearly identical around 53°C. The wild-type and F43 repressors are stabilized by about 5°C and the F75 and F43F75 repressors by about 10°C by inducer binding. The stabilizations of proteins in the complexes are summarized in table 2. The narrow transitions indicate single cooperative denaturation reactions of the complexes. The thermal denaturations of the repressor-tetracycline complexes were also measured by nitrocellulose filter binding and tetracycline fluorescence. The results are shown in Figs. 3 and 4, respectively. All four repressor-tetracycline complexes show nearly identical thermal stabilities in the nitrocellulose filter binding assay as well as in the temperature-dependent fluorescence measurements. In both cases we observed cooperative transi-
426
WAGENHQFER,
40
50
HANSEN, AND HILLEN
60
40
50
60
I^ .
(‘C 1
FIG. 2. Temperature gradient gel electrophoresis of Tet repressor-inducer complexes. The photographs of four Coomassie stained gels are shown. Each gel contained two slots; the upper was loaded with 70 pg of the respective repressor protein in 600 ~1 of a buffer containing 3 mM MgS04 in addition to the components given in the legend to Fig. 1. The lower slot was loaded with the same solution containing a IO-fold molar excess of tetracycline over repressor preincubated for 10 min at 37°C. The gels are aligned with respect to the temperature as indicated in the middle of the figure. Each temperature gradient ranged from 35 to 7o’c.
tions of the repressor-tetracycline complexes and identical denaturation temperatures regardless of the repressor mutation. Denaturations take place at 59°C in nitrocellulose filter binding and at 56°C in the fluorescence measurements (see Table 1). The curvature of the temperature-dependent fluorescence is not as sigmoidal as that obtained from nitrocellulose filter binding. In the temperature
range from 25 to 45°C there is a gradual decrease of fluorescence which could be due to alterations of the protein structure, a process which may precede the denaturation (16). Quantification of the Trp fluorescence was complicated by aggregation of the proteins which occurs along with denaturation as can be detected by the increasing turbidity of the solution.
TABLE 2 STABILIZATION
OF Tet REPRESSOR VARIANTS COMPARED WITH
COMPLEXED WITH THE FREE PROTEINS
INDUCER
AND
OPERATOR
Increase in denaturation temperature of Tet repressor variant (“C) Ligand Tetracycline Operator
Ionic condition” 10mMNa+,3mMMg2+
lOmMNa+
F7.5
F43
F43F75
5
11
20
5 8
10
13
wt
10
11The ionic conditions are given in addition to the electrophoresis buffer described under Materials and Methods.
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
I 20
30
40
50
Temperature
60
70
Thermal stabilities of the repressor-operator complexes. The temperature-dependent denaturation of the Tet repressor variants 1.0
F /6 0.6
I
z. F43 - F75 + F43F75
35
45
Temperature
55
I 40 Temperature
("Cl
FIG. 3. Temperature-dependent nitrocellulose filter binding assaysof Tet repressor-tetracycline complexes, Solutions of represso-[3H]tetracycline complexes (1 PM) were incubated at the indicated temperatures over a period of 15 min, chilled briefly on ice, and filtered immediately over nitrocellulose. The retained radioactivity of the repressor-tetracycline complexes (given in percentage of the radioactivity bound at 4o’C) was plotted versus the incubation temperature.
0.2 -
I 30
65 (“C)
FIG. 4. Temperature-dependent fluorescence of the repressor-tetracycline complexes. Fluorescence was excited at 280 nm. Emission was recorded at 5 10 nm. The concentration of repressor-tetracycline complexes was I pM. The ratio FJF, with F = fluorescence intensity at the temperature T, and F. = fluorescence intensity at 25’C is plotted versus the temperature.
427
OF PROTEINS
50
60
70
(“C)
FIG. 5. Temperature-dependent fluorescence of the wt and F43 repressor-operator complexes. Fluorescence was excited at 280 nm and emission was recorded at 330 nm. The concentrations were 0.2 pM repressor dimer and 0.1 pM 187-bp DNA containing the tandem ret operator arrangement (11). The ratio F/F,, with F = maximal fluorescence intensity at 25°C was plotted versus temperature.
complexed with the 187-bp tet operator DNA (11) has been studied employing fluorescence spectroscopy and temperature gradient gel electrophoresis. Since the fluorescence of Trp43 is totally quenched in the complex with tet operator and is not recovered by denaturation of the complex (data not shown) only the Trp75 containing Tet repressors could be studied with this method (1). Figure 5 displays the respective denaturation curves. The fluorescence properties of Trp75 are not affected by complex formation with the tet operator (I). Thus, the temperature-dependent fluorescence intensities of the complexes in Fig. 5 reflect only the denaturation of the proteins. Between 25 and 45°C the same gradual decrease of fluorescence intensity is observed as described above for the complexes with tetracycline. Above 45°C the denaturation curve is sigmoidal. The denaturation temperatures are 53°C for the F43-tet operator complex and 58°C for the wildtype-tet operator complex (see Table I). The reduced stability of the mutant complex corresponds to its reduced affinity for the tet operator (1). The denaturation of the F43 mu-
428
WAGENHijFER,
HANSEN, AND HILLEN
tant in this complex is clearly monophasic while the wild-type denaturation appears to be more complex. The resolution of the fluorescence measurements does not allow a quantitative analysis of this denaturation curve; however, it can be compared to the results obtained by temperature gradient gel electrophoresis (see below). Temperature gradient gels of the repressor-operator complexes were run at 10 and 40 mM sodium acetate. This was done because the T, of the DNA fragment is strongly ionic strength dependent (17) while the thermal stability of the complex is expected to be less sensitive to it. The results of the temperature gradient gel electrophoresis experiments are displayed in Fig. 6. The respective gels were silver stained under conditions visualizing predominantly the DNA. When the gels were run with a lo-fold increased amount of complex and stained with Coomassie blue the denatured protein is located at the high temperature region of the gel (data not shown). The denaturation profiles are complex in all cases consisting of at least two transitions. At lower temperatures the repressoroperator complexes show reduced electrophoretic mobilities in comparison to the uncomplexed 187-bp DNA. At higher temperatures the repressor-operator complexes denature leading to the free operator DNA as indicated by the increased mobility. At the high temperature end of the gel the doublestranded DNA melts and the single strands exhibit a reduced mobility compared to the free double-stranded DNA (compare also below). In some cases at low ionic strength these denaturation events are not well separated. The temperature gradient gels of these complexes yield two kinds of information with respect to denaturation. The first contains the denaturation temperatures. The wild-type and F75 repressor-operator complexes denature at 64 and 63°C at 10 mM Na+ and at 61 and 60°C at 40 mM Na+, respectively. The F43 and F43F75 repressor-operator complexes denature at 58 and 54°C in 10 mM Na+ and at 55 and 49°C at 40 mM Na+,
. I. @. 1
FIG. 6. Temperature gradient gel electrophoresis of Tet repressor-tet operator complexes. Photographs of silver stained gels are shown. The four gels displayed on the left side contained 400 ng 187-bp tet operator DNA and an eight-fold molar excess of the wild-type or F75 repressor or a 20-fold molar excessof the F43 or F43F75 repressor as indicated. The mixtures were preincubated in 500 ~1 of electrophoresis buffer (see Materials and Methods) containing 10 mM sodium acetate and 5% glycerol for 15 min at 37°C. Electrophoresis was carried out in the same buffer. The four gels displayed on the right side contained the same samples except that the concentration of sodium acetate was 40 mM. The two vertical rows of gels are adjusted with respect to the temperature which is indicated at the bottom. The filled triangles on the left side denote the position of the 187-bp DNA complexed with two Tet repressor dimers while the open triangles denote the positions of the 187-bp DNA complexed with one Tet repressor dimer. The lanes M contain DNA hagments resulting from a HaeIII/EcoRI digest of pWH802 ( 15) and the 187-bp DNA; the lanes S contain only the free 187-bp DNA fragment. The temperature gradients extended from 35 to 70°C.
respectively (compare also Table 1). Table 2 shows the thermal stabilizations of the Tet repressor proteins in the complexes over the free proteins. It may be derived that the wildtype repressor is stabilized by 13°C and the F75 mutant by 2o’C to reach roughly identical denaturation temperatures. The destabilizing effect of the Phe75 mutation in the free protein is completely neutralized by the stabilizing effect of tet operator binding in the complex. The Tet repressor mutants with the
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
Trp43 to Phe substitution, on the other hand, are stabilized only by 8°C (F43) and 10°C (F43F75). This corresponds well to their reduced binding constant for operator DNA (1). Comparing the denaturation temperatures at 10 and 40 mM Na+ reveals that the complexes are slightly destabilized by this small increase in ionic strength. This also corresponds to the destabilizing effect of ions on the association constant of Tet repressor&t operator binding ( 18). The second kind of information drawn from the temperature gradient gel electrophoresis concerns the intermediates of the denaturation reactions. Upon increasing the temperature in the gels shown in Fig. 6 the 187-bp DNA complexed with two dimers of wild-type or F75 Tet repressors dissociate in two steps at both ionic strengths. The F43and F43F7Soperator complexes, on the other hand, denature at both salt concentrations in a single transition. In the complexes with F43 and F43F75 repressors at the high and low salt concentrations the melting of the double-stranded 187-bp DNA is clearly resolved from the denaturation of the protein-DNA complexes (compare also below). In the low salt experiments with the wild-type and F75 repressors the DNA melting and denaturation of the protein-DNA complexes occur in a narrow temperature range. In these experiments an intermediate species shows up in the denaturation reaction which exhibits a greatly reduced mobility in the gel. It exists during 1 or 2°C and denatures then to the nonmobile protein (not shown) and the single-stranded DNA (see Fig. 6). In order to facilitate the interpretation of these results we have examined the denaturation of pure DNA fragments on temperature gradient gels. Denaturation of the 187- and 456bp tet operator DNAs on temperature gradient gels. Figure 7 displays the temperature gradient gels of the 187-bp DNA fragment at 10 and 40 mM Na+. Under both conditions the DNA fragment denatures in a single sharp transition. The melting temperatures are 62 and
429
OF PROTEINS
M
8
40 50 60
(“0
FIG. 7. Temperature gradient gel electrophoreses of the 187-bp DNA fragment. Photographs of the silver stained gels run with 400 ng of free I87-bp DNA in 500 ~1 of the buffer given in the legend to Fig. 6 with 10 mM sodium acetate (upper) and 40 mM sodium acetate (lower) are shown. The gels were aligned with respect to temperature as indicated on the bottom of the figure. The lanes M and S are as in Fig. 6. The temperature gradient extended from 35 to 70°C.
64’C, respectively. They are in fair agreement with results obtained from optical melting studies of the same DNA (17). The same sharp transitions are also found in most denaturation experiments of the protein-DNA complexes. In order to interpret the intermediate species with reduced mobility described above, the melting of a 456-bp tet control DNA fragment was examined by the temperature gradient gel method. In optical melting studies it showed two clearly separated transitions (17). The temperature gradient gel is shown in Fig. 8. In the first transition the DNA fragment denatures to a partially single-stranded structure at one end ( 17). Interestingly, the mobility of this DNA is greatly reduced in the gel matrix. Its mobility is gradually further reduced upon increasing the temperature. In a second transition this partially singlestranded DNA melts completely resulting in the two single strands. These exhibit an increased mobility compared with the partially denatured structure, which increases further upon raising the temperature. DISCUSSION
We have used temperature gradient gels to analyze the denaturation of wild-type and
430
WAGENHijFER,
HANSEN,
AND HILLEN
are higher with this method (see Table 1). We conclude that temperature gradient gels may lead to slightly different denaturation temperatures as compared to free solutions. However, the comparison of thermal stabilities of mutants and their complexes leads to consistent results. This conclusion is also derived from the melting of the DNA fragments. When studied in solution under slightly lower I I I ionic strength (17) the T, values are roughly 50 60 70 ("Cl 5°C lower. This could be attributed to the FIG. 8. Temperature gradient gel electrophoresis of the cage effect assumed to stabilize complexes in 456-bp tet control DNA. A photograph of the silver gels (2 1). On the other hand, stained gel is shown. It was loaded with 400 ng of the polyacrylamide the T, difference between the two transitions 456-bp DNA in 500 pl electrophoresis buffer (see Materials and Methods) containing 10 mM sodium acetate and in the 456-bp fragment (formerly called 501 5% glycerol. The temperature scale is given on the botbp) is about 3°C as was observed in solution tom of the figure. The lane M contains the standard as in ( 17). Since the first transition is monomolecFig. 6 with the 456-bp DNA and the lane S contains only ular and the second is bimolecular, caging the 456-bp DNA. The temperature gradient extended would be expected to increase this T,,, differfrom 40 to 80°C. ence. This argues against a cage effect and is in agreement with observations made with dissociation reactions (22). mutant Tet repressor proteins alone and in protein-DNA their complexes with inducer and operator. The small differences in the denaturation The temperature gradient gels offer an advan- temperatures would then be attributed to tage compared to other methods because, in changes in ionic strength. Comparison of the gel mobility shift transiaddition to the determination of denaturation temperatures, they allow the analysis of tions resulting from the denaturation of progel mobilities of the native and denatured teins and DNA fragments indicates that the forms detecting possible intermediates in the latter is a highly cooperative process occurdenaturation reaction. The limitation is, of ring in a narrow temperature range. Denaturcourse, that only those transitions which alter ation of the proteins and their complexes, on the other hand, occurs within a broader temthe electrophoretic mobility can be analyzed. Therefore, we compared the results obtained perature range of approximately 4°C. Thus, from the temperature gradient gels with the this property limits the accuracy of T, deterones obtained from temperature-dependent mination rather than the reproducibility of the temperature gradient. fluorescence changes and binding activities The Tet repressor denatures in a single determined by nitrocellulose filter binding assays. These results are interpreted under the transition in the temperature gradient gel (see assumption that the decrease in fluorescence Fig. 1). This implies that, unlike, e.g., the X results from denaturation of the protein as repressor (4), this protein may consist of only supported by the cooperativity of the respec- a single domain. Alternatively, the denaturation of potential other domains may not intive transition curves. Since these measurements were done in different buffers and at fluence the gel mobility. Scanning calorimedifferent heating rates one cannot expect try of these proteins reveals also single sharp identical denaturation temperatures. For the transitions supporting the first assumption repressor-inducer complexes they tend to be (Hinz et al., to be published). The Phe75 mulower in the temperature gradient gels, while tation, which was shown to be located in the the ones of the repressor-operator complexes hydrophobic core of the repressor (2), has a
TEMPERATURE
GRADIENT
GEL ELECTROPHORESIS
large destabilizing effect while the Phe43 mutation on the surface of the protein (2) does not reduce the stability. This resembles results obtained with X repressor mutants (5) and confirms the locations of those residues in the repressor structure. Binding of the inducer tetracycline stabilizes the wild-type repressor by 5°C and the ones containing the Phe75 mutation by 10°C. Thus, inducer binding compensates fully for the destabilizing Phe75 mutation. All protein-tetracycline complexes denature at the same temperature in denaturation studies using temperature gradient gels (see Fig. 2), nitrocellulose filter binding (see Fig. 3), and fluorescence spectroscopy (see Fig. 4). The denaturation temperatures obtained with the three methods vary from about 53°C in the temperature gradient gels to 56°C in the fluorescence measurements and 59°C in the nitrocellulose filter binding experiments. This may be due to different buffer conditions or differences in handling of the samples in these methods. The results from all three methods indicate that the repressor-inducer complexes denature cooperatively in a single transition. It is noted that the free proteins are slightly destabilized in the presence of 3 mM Mg*+ (compare Figs. 1 and 2 and Table 1). The reason for this is not clear. The denaturations of the repressor-operator complexes clearly reveal the additional information obtained from temperature gradient gel electrophoresis in comparison to other methods of studying these complex reactions. In the simple cases the temperature-induced decompositions of the repressor-operator complexes yield in the first step the precipitated protein and double-stranded DNA followed by melting of the DNA. These mechanisms are found for the F43 and F43F75 complexes at both salt concentrations. The repressors binding tet operator DNA with high affinity (wild-type and F75) (1) form complexes which denature via a different pathway. The 187-bp DNA containing the tandem tet operator arrangement ( 11,19) can bind one or two Tet repressor dimers result-
OF PROTEINS
431
ing in different mobility shifts (20). First the complex of the 187-bp DNA with two repressor dimers denatures to the complex of the 187-bp DNA with one repressor dimer which then denatures at high salt to the doublestranded DNA followed by DNA melting. At low salt the thermal stability of the DNA is decreased to an extent where part of it denatures to single-stranded ends at temperatures below protein denaturation in the proteinDNA complex. This is indicated by the intermediate with greatly reduced gel mobility in Fig. 6. The same denaturation mechanism had been proposed before from melting studies at low salt ( 17,23). The denaturation temperatures of the complexes clearly correlate with the affinities of the Tet repressor variants for tet operator (1) and not with the stability of the free proteins. This is indicated by the data obtained with the F43 and F75 mutants (see Table 2). The free 187- and 456-bp DNA fragments show quite different behavior on the temperature gradient gels. By hyperchromicity studies it had been shown previously that the 187bp DNA denatures in a single cooperative transition while the 456-bp DNA denatures in two cooperative transitions separated by about 3.5”C (17,23). The same behavior is found on the temperature gradient gel (see Figs. 7 and 8). The first transition of the 456-bp DNA leads to a partially singlestranded structure (17). This exhibits a lower mobility in the gel as the single strands. The same observation has been made for the denaturation of double-stranded RNA species previously (7,8). The partially single-stranded structure shows decreased mobility in the polyacrylamide gel as the temperature is raised. This could indicate increased fraying of the double stranded ends. However, this is unlikely because the ends of the completely double-stranded DNA should also increase their fraying on the low temperature side of the gel and yet its mobility increases. Thus, it is concluded that the flexibility of the singlestranded ends causes the temperature-dependent increase in gel retardation.
WAGENHGFER,
432
HANSEN,
ACKNOWLEDGMENTS We thank Dr. D. Riesner and V. Rosenbaum for their help with the temperature gradient gels and Mrs. R. Kisse1for typing the manuscript. This work was supported by grants from the Deutsche Forschungsgemeinsch and the Fonds der chemischen Industrie.
REFERENCES 1. Hansen, D., and Hillen, W. (1987) J. Biol. Chem.
262,12,269-12,274. 2. Hansen, D., Altschmied, L., and Hillen, W. ( 1987) J. Biol.Chem.262,14,030-14,036.
AND HILLEN
10. Gehmichen, R., Klock, G., Altschmied, L., and Hillen, W. (1984)EMBO J. 3,539-543. 11. Hillen, W., Klock, G., Kaffenberger, I., Wray, L. V., and Reznikoff, W. S.(1982) J. Biol. Chem.257, 6605-6613. 12. Takahashi, M., Altschmied, L., and Hillen, W. (1986)J. Mol. Biol.187,341-348. 13. Sammons, D. W., Adams, C. D., and Nishizawa, E. E. (1982) Electrophoresis 2,135-141. 14. Postle, K., Nguyen, T. T., and Bertrand, K. P. (1984)
NucleicAcidsRes.12,4849-4863. 15. Unger, B., Klock, G., and Hillen, W. (1984) Nucleic
Acids.Res.12,7693-7703.
16. privalov,P. L. (1979) Adv. ProteinChem.33, 1673. Hecht, M. H., Nelson H. C. M., and Sauer, R. T. 241. (1983)Proc.Natl.Acad.Sci. USA80,2676-2680. 17. Hillen, W., Unger, B., and Klock, G. (1982) Nucleic 4. Pabo, C. O., Sauer, R. T., Sturdevant, J. M., and AcidsRes.10,6085-6097. Ptashne, M. ( 1979) Proc.Natl.Acad.Sci.USA76, 18. Kleinschmidt, C., Tovar, K., Hillen, W., and Piir1608-1612. schke, D. (1988) Biochemistry27,1094-1104. 5. Hecht, M. H., Sturdevant, J. M., and Sauer, R. T. 19. Bertrand, K. P., Postle, K., Wray, L. V., Jr., and Rez(1984)Proc. Nati. Acad.Sci. USA81,5685-5689. nikoff, W. S. (1983) Gene23, 149-156. 6. Hecht, M. H., Sturdevant, J. M., and Sauer, R. T. 20. Hillen, W., Gatz, C., Altschmied, L., Schollmeier, (1986)Proteins:Struct.Funct.Genet.1,43-46. K., and Meier, I. (1983) J. Mol. Biol. 169,7077. Rosenbaum, V., and Riesner, D. (1987) Biophys. 722. Chem.26,235-246. 21. Fried, M., and Crothers, D. M. (1981) NucleicAcids 8. PO, T., Steger, G., Rosenbaum, V., Kaper, J., and Res.9,6505-6516. Riesner, D. (1987) NucleicAcidsRes.15, 5069- 22. Revzin, A., Ceglarek, J. A., and Garner, M. M. ( 1986)Anal. Biochem.153,172-l77. 5084. 9. Steger, G., PO, T., Kaper, J., and Riesner, D. ( 1987) 23. Hillen, W., and Unger, B. (1982) Nature(London) 297,700-702. NucleicAcids Res.15,5085-5104.