This article was downloaded by: [Rutgers University] On: 04 April 2015, At: 09:38 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of Biomolecular Structure and Dynamics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tbsd20
1
H-NMR Investigation of the Interaction of the Amino Terminal Domain of the LexA Repressor with a Synthetic Half-Operator a
a
Gunnar Ottleben , Luigi Messori , Heinz Rüterjans a
b
c
, Robert Kaptein , Michèle Granger-Schnarr &
Manfred Schnarr
c
a
Institute of Biophysical Chemistry Johann Wolfgang Goethe-University of Frankfurt , Germany b
Department of Chemistry , University of Utrecht , The Netherlands c
Institute of Molecular and Cellular Biology of CNRS , Strasbourg , France Published online: 21 May 2012.
To cite this article: Gunnar Ottleben , Luigi Messori , Heinz Rüterjans , Robert 1
Kaptein , Michèle Granger-Schnarr & Manfred Schnarr (1991) H-NMR Investigation of the Interaction of the Amino Terminal Domain of the LexA Repressor with a Synthetic Half-Operator, Journal of Biomolecular Structure and Dynamics, 9:3, 447-461, DOI: 10.1080/07391102.1991.10507928 To link to this article: http://dx.doi.org/10.1080/07391102.1991.10507928
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [Rutgers University] at 09:38 04 April 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Journal of Biomolecular Structure & Dynamics, /SSN 0739-1102 Volume 9, bsue Number 3 (1991), "'Adenine Press (1991).
1
H-NMR Investigation of the Interaction of the Amino Terminal Domain of the LexA Repressor with a Synthetic Half-Operator Gunnar Ottleben\ Luigi Messori\ Heinz Riiterjans 1 *, Robert Kaptein2, Michele Granger-Schna~ and Manfred Schnarr3 Downloaded by [Rutgers University] at 09:38 04 April 2015
1
Institute of Biophysical Chemistry Johann Wolfgang Goethe-University of Frankfurt Germany 2
3
Department of Chemistry University of Utrecht The Netherlands
Institute of Molecular and Cellular Biology of CNRS Strasbourg, France
Abstract A synthetic half-operator DNA-duplex, d(GCTACTGTATGT), containing a portion of the proposed recognition sequence (CTGT) of serveral "SOS" genes, has been synthesized. The dodecamer has been characterized through 1H-NMR spectroscopy. Complete assignment of exchangeable hydrogen bonded imino protons has been acheived by applying lD NOE techniques and an analysis of the temperature dependence of the chemical shifts. In order to determine the specific role of the CTGT consensus sequence in the overall recognition process, the oligonucleotide duplex has been titrated with the amino terminal DNA binding domain of the LexA repressor. The observation of substantial changes of 1H-NMR chemical shifts in the imino proton region upon interaction with the protein strongly suggests that the protein binds specifically to the operator DNA The largest deviations of 1H-NMR chemical shifts upon protein binding have been observed for protons assigned to the CTGT segment, thus strongly suggesting a direct involvement of this sequence in the binding process. At high potassium chloride concentrations the 1H-NMR chemical shift deviations are reverted which is consistent with the known drop in the affinity constant of LexA for operator DNA at high salt concentrations.
Introduction LexA, the repressor of the "SOS" system in E. coli., regulates the transcription of about 20 SOS genes mostly involved in DNA repair, mutagenesis and cell division (1-4). These "SOS" genes are induced upon DNA damage by physical and chemical agents through a mechanism involving LexA repressor inactivation by autodigestion at the Ala 84-Gly 85 peptide bond in aRecA dependent fashion (4,5). *Author to whom correspondence should be addressed.
447
Downloaded by [Rutgers University] at 09:38 04 April 2015
448
Ott/eben et a/.
LexA is a repressor protein consisting of two functional domains: the amino terminal domain, which harbors the DNA binding site of the repressor, and the carboxy-terminal domain, which is involved in a weak dimerisation process (6,7). The isolated monomeric amino terminal domain retains most of the DNA binding properties of the entire repressor protein (8). Recent 2D NMR studies of this domain (9) have shown that it contains three a-helices- two of them separated by a tum- inferring a type of helixtum-helix motif. In contrast to canonical helix-tum-helix proteins the relative positioning of helix 2 and helix 3 differs significantly from that usually found (10,11), since the tum is two amino acids longer and contains the bulky aromatic residue Phe-37. Nevertheless the presence of a cluster ofLexA mutants within helix 3, deicient in DNA binding, suggests that this helix should be involved in DNA binding processes (12). The structural deviations observed between the DNA binding domain of LexA repressor and proteins containing a common helix-tum-helix motif raises the question how LexA recognizes specific binding sites on DNA and if it contains a new DNA binding motif or if it has to be considered as a distantly related member of the helix-tum-helix family. It has been proposed that the LexA repressor recognizes operator binding sites with the palindromic consensus sequence 5'-CTGT(at)4ACAG-3' in which CTGT and ACAG are the most highly conserved sequences ( 13). More recent studies indicate that these base pairs contain most of the information required for specific protein binding to operator DNA (11).
In order to further substantiate this point we have undertaken an extensive investigation comprising the synthesis of a number of short DNA fragments and the study of their binding properties to the amino terminal domain of the LexA repressor. In this paper we report on the interaction between the amino terminal domain of LexA and a 12 bp synthetic half-operator duplex d(GCTACTGTATGT) containing the CTGT sequence in the inner part of the oligonucleotide. The investigation is based on the assumption that this fragment contains the minimum requirements for specific recognition by the binding domain. The DNA dodecamer has been investigated by 1H-NMR analysis of the exchangeable imino protons in order to elucidate changes of the DNA conformation upon formation of a protein-DNA complex. In a first attempt to understand the mechanism of specific complex formation the present study is a prerequisite investigation for a more extensive characterization of the protein-DNA interaction in solution using 2D or 3D NMR spectroscopy.
Materials and Methods The dodecamer oligodeoxynucleotide dGCTACTGTATGT and its complementary strand were prepared by the phosphonamidate method (14) and purified by FPLC; the sequence was confirmed according to the Maxam-Gilbert procedure (15). For the 1H-NMR measurements the oligodeoxynucleotides were dissolved in 0.2 M KCl, 10 mM phosphate buffer, at pH 5.6, to a concentration of 1-4 mM. The DNA concentrations were determined from the absorbance at 260 nm, using A260 = 20 em-I for a 1 mg/ml solution. LexA protein was isolated according to published procedures (6).
LexA Repressor and A Half-Operator
449
The protein was cleaved into the carboxy- and the amino terminal domain using the autocleavage reaction at pH 9.4; the amino terminal domain was further purified as described elsewhere (8).
Downloaded by [Rutgers University] at 09:38 04 April 2015
AlllD and the 2DNOE 1H-NMRspectra wereacquiredatSOOMHzonanAM 500 spectrometer, equipped with an Aspect 3000 computer. A 20 NOE experiment of the protein-DNA complex was recorded on an AMX 600 spectrometer, interfaced with a X32 computer. The t 1 files of the 20 datasets were recorded with 2K datapoints each, and zero-filled once in t1 and ~ time prior to the Fourier transformation. A shifted cosine-bell window was applied as an apodization function in each time
B
A
13
14
12
PPM 1
Figure I: H-NMR spectra of the imino proton absorption region of the DNA dodecamer recorded with two different pulse sequences. A 1331; B. Presaturation Conditions: 4 mM DNA dodecamer concentration, pH 5.6, 10 mM phosphate buffer, 0.2 M KCI; 302 K.
450
Ott/eben eta/.
320K
315
K
Downloaded by [Rutgers University] at 09:38 04 April 2015
310K
300K
295K
292K
14
13
12
PPM Figure 2: Temperature dependence of the imino proton resonances in the range of292-320 K Conditions as in Figure I.
451
LexA Repressor and A Half-Operator
domain. Chemical shift values are reported relative to the TSP (sodium 2,2,3,3Tetradeutero-3-(-trimethylsilyl-) propionate) resonance and recorded with an accuracy of0.04 ppm. Solvent resonance suppression was achieved either by presaturation or by semi-selective excitation, namely the 1331 technique ( 16). The temperature dependence of the imino protons was investigated in the range of292-320 K. NOE enhancements are obtained from 10 difference experiments or 20 NOESY measurements. Results and Discussion
Downloaded by [Rutgers University] at 09:38 04 April 2015
Oligonucleotide Sequence
The sequence of the double stranded DNAdodecameris given below; the following numbering of the bases has been adopted:
1
2
3
4
5
6
7
8
9
10
II
12
5'-
G
c
T
A
c
T
G
T
A
T
G
T-3'
3'-
c
G
A
T
G
A
c
A
T
A
c
A-5'
24
23
22
21
20
19
18
17
16
15
14
13
The G ll-C 14 base pairs deviating from the consensus sequence given above has been introduced to facilitate resonance assignments and to improve the duplex stability. The 4 base pairs 5' to the CTGT motif have been introduced to allow the formation of all protein contacts with the DNA backbone as inferred from ethylation interference studies with the rec A operator (17). 1
H-NMR Studies
a) Sequential Assignment of the Hydrogen Bonded Imino Protons
For the assignments of the imino protons the following criteria were used: i) The AT and GC imino proton resonances appear in fairly distinct spectral regions; the former in the 14-13 ppm range and the latter in the 13-12 ppm range (18). ii) The melting of short deoxyoligonucleotides occurs sequentially and starts from the terminal base pairs. Investigations of the differential broadening or disappearance of some 1HNMR resonances with increasing temperature permits the identification of the imino protons in a sequential way, starting from those located at the ends of the sequence (19). iii) The imino protons of adjacent base pairs are about 3.5 A apart; they give rise to small but detectable nuclear Overhausereffects. Such NOE interactions can be observed in 10 and 20 experiments and represent the most straightforward method for sequential assignments (20). 1
The H-NMR spectrum of the imino proton region, recorded at 302 K. in 10 mM phosphate buffer, 0.2 M KCl at pH 5.6, is shown in Figure lA. The spectrum shows
Ott/eben eta/.
Downloaded by [Rutgers University] at 09:38 04 April 2015
452
14
13
12
PPM Figure 3: Map of the observed I D NOE connectivities of the imino protons. Some representative NOE difference spectra are also reported.
LexA Repressor and A Half-Operator
453
ten well separated signals, six of them are observed in the range of 13.4-13 ppm (signals alto a6) and four in the range ofl3-12.1 ppm (signals gl to g4) belonging to the AT and GC base pairs, respectively. With the exception of resonance g3, all signals exhibit approximately equal intensity corresponding to one proton each; only at lower temperatures signal g3 is of similar intensity, too (see below).
Downloaded by [Rutgers University] at 09:38 04 April 2015
A semiselective solvent suppression technique has been applied to detect the imino proton resonances avoiding the excitation of the water resonance (16,21); indeed, application of pulse sequences with a common presaturation step results both in a drastic loss in the intensity of signal gl and most of the AT resonances, and in the disappearance of signal g3, owing to saturation transfer processes with the solvent, as shown in Figure lB. The fact that only two out of twelve expected imino proton resonances cannot be observed may be explained by assuming that at 302 K the imino protons of the two terminal base pairs are subject of fast exchange processes associated with base fraying. By increasing the temperature we expect a progressive increase in the exchange rates of the imino protons resulting in a stepwise loss of intensities. Several 1HNMR spectra of the DNA dodecamer recorded at different temperatures are shown in Figure 2. Upon raising the temperature from 292 to 310 K we observed a selective loss of intensity of signals gl and g3; these resonances therefore have been assigned to the imino protons of the two GC pairs adjacent to the terminal ones. A further increase of the temperature up to 320 K causes a general broadening of all detected signals indicating the melting process; however, it must be recalled that a loss of intensity of the imino proton resonances does not occur at temperatures lower than 'tm. This observed line-broadening does not correspond to the collapse of the whole structure. Observation of nuclear Overhauser effects between adjacent imino protons in one dimensional experiments provides a direct way to perform sequential assignments; two NOE's are expected for base pairs in the innerpartofthe sequence whereas only one is detectable for each nuclotide located at the terminal region. Observed NO E connectivities are shown in Figure 3 together with the results of some one dimensional experiments. Additionally, a 2D NOE spectrum of the DNA dodecamer has been recorded (Figure 4). In the region of the imino proton resonances, we found most of imino proton-imino proton connectivities required for the above mentioned assignments: one NOE between signal gl and al and two between signal g2 and g4 each. This result is consistent with the experimental data of the lD spectra shown in Figure 3. With the experimental conditions used in NOESY experiment and the fact that these cross peaks have a relative low intensity, some of the expected NOE enhancements between adjacent imino protons are not observed due to saturation transfer effects (Figure 4). On the basis of the information obtained from NOE measurements and from the analysis of the temperature dependence we were able to obtain the sequential assignments often imino proton resonances. The following ordering of the resonances
454
Ott/eben et a/.
al a2
a4
gl
aS a6
g2
g4
0
Downloaded by [Rutgers University] at 09:38 04 April 2015
12.5
0 13.0
D
~
D
0
~q pprn
13.0
13.5
·13.5
@
pprn
12.5
Figure 4: Plot between the imino imino proton connectivities of the DNA dodecamer. The 500 MHz NOE experiment was acquired using a mixing time, 'm• of 130 msec and 450 t 1 files, 2k data points, each. The resonance assignment of the imino protons is shown on top of the spectrum; sequential pathway of some imino proton resonances are indicated by a line. Conditions as in Figure lA
is obtained that matches the DNA dodecamer sequence:
assigned signals
:00
sequence : G I
gl
al
a4
g2
a2
g4
a5
a6
a3
C2
T3
A4
C5
T6
G7
T8
A9
TIO Gll Tl2
g3
00
C24 G23 A22 T21 G20 Al9 CIS Al7 Tl6 Al5 Cl4 A13
455
LexA Repressor and A Half-Operator
H
G
Downloaded by [Rutgers University] at 09:38 04 April 2015
F
E
D
c 8
A
14
13
12
PPM Figure 5: Chemical shifts of the imino proton resonances in the presence of various concentrations of the amino terminal domain ofLexA at 302 K. The DNA concentration was 1.5 mM. Protein: DNA ratios for each spectrum are: A) 0.0: B) 0.16: C) 0.33: D) 0.50; E) 0.67: F) 0.84: G) 1.00: H) 1.15.
b) Interaction of DNA Dodecamer with the Amino Terminal Fragment the LexA Protein The chemical shifts of the imino protons within the double stranded DNA are highly sensitive to the base pair ring currents and hence, are significantly influenced by
456
Ott/eben et a/.
protein/DNA ratios
ppm 13,9
13,9
T3 13,7
13,7
Downloaded by [Rutgers University] at 09:38 04 April 2015
.
T 10
T6
13,15
T 21
13,3
..- -r-T 16
T8
13,1
---
13,15
13,3
13,1
12,9
12,8 0,15
ppm 12,9
12,9
G 23 G 20
12,7
12,7
G 11 12,15
12,15
• 12,3
12,3
G7
12,1
.-
12,1
11,9
11,9 0.5
Figure 6: Chemical shift variation of imino proton resonances in dependence of serve raJ protein-DNA ratios.
LexA Repressor and A Half-Operator
457
Downloaded by [Rutgers University] at 09:38 04 April 2015
changes oflocal geometry. Upon addition of the DNA-binding domain ofLexA the imino proton resonances of the DNA dodecamer exhibit substantial variations in their chemical shift values indicating small changes in the DNA helix conformation (see Figure 5). As previously reported for similar systems, the position of the imino proton resonances can be followed at various degrees of saturation ofDNA with the amino terminal domain despite the increase of resonance line-widths (22,23). The chemical shift values of the imino protons for each step of the titration are shown in Figure 6. Complex formation affects mainly four signals out of the ten detected resonances in the imino proton resonance region. For the 1:1 ratio of operator DNA to protein the resonance frequencies g4, g2, a6 and a2 are shifted by more than 0.15 ppm (e.g. g4=212 Hz, a2= 143 Hz), whereas the variations of the chemical shift values of all other imino proton resonances are smaller than 0.1 ppm. It is remarkable to note that all signals with a large change in the chemical shift value correspond to base pairs which are located in the inner part of the dodecamer operator sequence; all of them except one (a6) belong to the proposed recognition motif CTGT. Signals of protons belonging to nucleotides located near the ends of the sequence (namely gl and g3) are almost unaffected upon complex formation. It is evident that the central part of the DNA fragment, and in particular the CTGT segment, is responsible for the recognition process between protein and DNA Moreover, this result qualitatively suggests that the terminal regions are not involved in this process. However, if the CTGT recognition motif is not preceded by 4 base pairs at the 5' site, but only by 2 base pairs as in the palindromic 20-mer CA-CTGT-ATGTI AGAT-ACAG-TG, no substantial chemical shift variations are observed suggesting that in this case no stable specific complex has been formed. This is not unexpected, since at least one contact between LexA and the DNA backbone in each halfoperator is lost in this sequence; in this case the DNA duplex is too short (17).
Upon progressive formation of the complex, we observe significant line-broadening of all resonances that show chemical shift variations larger than 0.15 ppm. Signal g4 for example, shows broadening and the decrease in intensity up to disappearance of the line at 12.4 ppm and the simultaneous appearance of a new signal (g4') at higher field (11.9 ppm) corresponding to the same resonance in the bound state. Careful inspection of the spectra at different degrees of saturation suggests the presence of something like "intermediate exchange" condition. According to a small chemical shift difference between the free and bound state a coalescence signal is observed at titration point C for signal g2 whereas for signal g4 the chemical shift variation is too large to obtain such a signal. To support this interpretation we have recorded spectra at increasing temperatures for a sample with a protein-DNA stoichiometry of0.5, where both signals g4 and g4' are observed. Upon progressively raising the temperature from 302 to 312 K we note that signals g4 and g4' broaden and coalesce into one signal, which subsequently narrows; the signal is located at a position between g4 and g4' as it is expected for a coalescence signal. However, in spite of the fact that the exchange rate for the four exchange processes is believed to be similar, the appearance of the involved signals is different; the remaining signals a2 and a6 undergo chemical shift changes upon complex formation in a more crowded region excluding the observation of their exchange behavior.
Downloaded by [Rutgers University] at 09:38 04 April 2015
Figure 7: Part of a NOESY spectrum of the protein-DNA complex, recorded at 600 MHz. The mixing time tm, was 150 msec and the data were accumulated in 512 t 1 files at an F 1 sweep width of 12000 Hz. The resonance assignment of the imino protons is shown on top of the spectrum; sequential pathway of some imino proton resonances are indicated by a line. Conditions as in Figure lA.
At the end of the titration, the respective signals are relatively narrow. Overall, despite these rather complex exchange processes, it is possible to follow most of the imino proton signals through the titration. In order to consolidate the assignments of the endpoint of the titration, we recorded a NOESY spectrum of the protein-DNA complex (Figure 7). Most of the imino proton imino proton connectivities are absent due to saturation transfer, but the cross peaks between signal g4 and a5 as well as between a6 and a3 show clearly the indicated pathway in Figure 5. It should be mentioned that most of the resonances of the bound species exhibit
line-widths moderately larger than the respective line-widths of the free DNA
459
LexA Repressor and A Half-Operator
Downloaded by [Rutgers University] at 09:38 04 April 2015
c
8
A
14
12
13
11
PPM Figure 8: Dissociation of the complex after the addition ofKCI. (A) Spectrum of the protein-DNA complex. (B) The same absorption region of A for the complex in the presence of I M KCI. (C) Spectrum of the free DNA dodecamer.
dodecamer. On the average the line-widths of the bound species are by a factor 1.3l.Slarger than those of the free species. This increase in line-width can be explained
460
Ott/eben et a/.
Downloaded by [Rutgers University] at 09:38 04 April 2015
with the higher molecular weight of the complex as compared to the free dodecamer. However, signal g4 reveals a more extensive line-broadening (an increase in line-width by a factor 5) which must be due to a different mechanism. In general, two possible explanations for such selective line-broadening can be proposed (24): i) A selective increase in the T2 -I relaxation rate owing to a decrease in mobility of a particular region of the molecule as a result of complex formation with the protein. This mechanism is not unlikely in this case, since (g4) that undergoes broadening belongs to the portion of the molecule which is beleived to be directly involved in binding the protein. ii) The line-broadening can be due to an exchange process between two or more conformations of the bound oligonucleotide which can be intermediate on the NMR time scale. We can rule out that the broadening is related to an increase in the exchange rate with the solvent since increasing the temperature causes a marked narrowing ofline g4' and no loss in intensity. In order to distinguish between the above mentioned mechanism, we performed several line-widths experiments at various field strengths. Increasing line-broadening with the square of the field is expected in case of mechanism "ii", whereas litle change should be observed for mechanism "i". Thorough comparison of the imino proton resonances recorded at 270 MHz, 500 MHz and 600 MHz showed no significant change in line-broadening. As a result, the decrease in mobility of the particular part of the recognition motif seems to be a more reasonable explanation for this line-broadening upon complex formation. All obtained results provide evidence for the formation of a specific protein-DNA complex between the DNA binding domain ofLexA repressor and the half-operator, making this system suitable for a more detailed structural investigation by multidimensional NMR spectroscopy.
c) Dissociation of the Specific Complex with Increasing Concentrations of KCI By adding increasing amounts ofKCl to the solution of the protein-DNA complex, the 1 H-NMR spectrum of the free DNA dodecameris almost recovered (Figure 8). High salt concentrations apparently lead to the disruption of the protein-DNA complex; the double stranded structure of the DNA dodecamer is kept intact as concluded from the observation of the imino proton resonances. This result is consistent with the reported decrease of the association constant of LexA protein for the uvrA operator and for non-operator DNA upon increasing the salt concentrations (25,26).
Conclusions The binding of the amino terminal domain ofLexA to a dodecamer oligonucleotide, containing the proposed recognition sequence CTGT, has been investigated by 1HNMR experiments. The chemical shift changes of the imino proton resonances describe the specific binding of the domain to the operator sequence. The signals most affected belong to the proposed recognition region CTGT, which suggests that these base pairs play a key role in the binding process. The complex pattern of the line-widths
LexA Repressor and A Half-Operator
461
observed during the titration with the protein has been explained on the basis of exchange processes and a restriction of dynamics. It has been explained on the basis of exchange processes and a restriction of dynamics. It has been also shown that high concentrations ofKCl causes the breakdown of the protein-DNA complex.
Acknowledgements
Downloaded by [Rutgers University] at 09:38 04 April 2015
The authors thank Dr. Brian Fera for the preparation of the DNA fragment. This work was supported by grants from the European Communities (ST2J-0291), from INSERM (871 007) from the" Association de Recherche contre le Cancer", from the "Ligue Nationale de la Luthe contre le Cancer". References and Footnotes
I. 2. 3. 4. 5. 6. 7. 8. 9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
Little, J.W. and Mount, D.W., Cell 29, 11-22 (1982). Walker, G.C., Microbiol. Rev. 48, 60-95 (1984). Howard-Flanders, P., West, S.C. and Stasiak, A., Nature 309,215-220 (1984). Friedberg, E.C., "DNA Repair", Freeman, San Francisco (1985). Little, J.W. and Hill, SA, Proc. Nat/. Acad. Sci. USA 28, 4199-4203 (1985). Schnarr, M., Pouyet, J., Granger-Schnarr, M. and Daune, M., Biochemistry 24,2812-2818 (1985). Schnarr, M. Granger-Schnarr, M., Hurstel, S. and Pouyet, J., FEBS Lett. 234, 56-60 ( 1988). Hurstel, S., Granger-Schnarr, M., Daune, M. and Schnarr, M., EMBOJ 5, 793-798 (1986). Lamerichs, R.M.J.N., Padilla, A., Boelens, R., Kaptein, R., Ottleben, G., Raterjans, H., GrangerSchnarr, M., Oertel, P. and Schnarr, M., Proc. Nat/. Acad. Sci. USA 86, 6863-6867 (1989). Brennan, R.G. and Matthews, B.W.,J Bioi. Chern. 264, 1903-1906 (1989). Strub!, K., Trends Biochem. Sci. 14, 137-140 (1989). Oertel-Buchheit, P. Schnarr, M. and Granger-Schnarr, M., Mol. Gen. Genet. 223,40-48 (1990). Wertmann, K.F. and Mount, D.W.,J Bacteriol. 163, 376-384 (1985). Matteucci, M.P. and Caruthers, M.H.,J Am. Chern. Soc. 103,3185-3191 (1981). Maxam, AM. and Gilbert, W., Methods Enzymol. 65, 499-560 ( 1980). Hore, P.J.,J Magn. Reson. 55,283-300 (1983). Hurstel, S., Granger-Schnarr, M. and Schnarr, M., EMBO J 7, 269-275 (1988). Zuiderweg, E.R.P., Scheek, R.M., Veeneman, G., van Boom, J.H., Kaptein, R., Raterjans, H. and Beyreuther, K., Nuc/. Acid Res. 9, 6553-6569 (1981). Weiss, M.A., Patel, J.D., Sauer, R.T. and Karplus, M., Nuc/. Acid. Res. 12, 4035 ( 1984). Buck, F., Hahn. K.D .• Brill, W.. Raterjans, H .. Chernov. B.K.. Skryabin, K.G .. Kirpichnikov, M.P. and Bayev, A.A,J Biomol. Struct. Dynam. 3, 899-911 (1986). Rajagopal, P .. Gilbert, D.E., van der Maarel, G.A., van Boom, J.H. and Feigon, J.,J Magn. Reson. 78, 526-537 (1988). Hahn, K.D .. Buck, F .• Rtlterjans, H .. Chernov, B.K.. Skryabin, K.G. and Kirpichnikov. M.P.,Eur.J Biochem. 12, 87-95 (1985). Scheek, R.M .• Zuiderweg. E.R.P., Klappe, K.J.M., van Boom, J.H .• Kaptein, R., Rtlterjans H.and Beyreuther, K., Biochemistry 22, 228-235 (1983). Schumacher, R.. Buck, F. and Rtlterjans, H .. Nuc/. AcidY Res. 13,5097-5105 (1989). Betrand Burggraf, E., Hurstel, S.. Daune, M. and Schnarr. M.,J Mol. Bioi. 193,293-302 (1987). Hurstel, S., Granger-Schnarr. M. and Schnarr. M., Biochemistry 29, 1961-1970 (1990).
Date Received: March 1, 1991
Communicated by the Editor Wolfram Saenger
Journal of Biomolecular Structure and Dynamics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tbsd20
1
H-NMR Investigation of the Interaction of the Amino Terminal Domain of the LexA Repressor with a Synthetic Half-Operator a
a
Gunnar Ottleben , Luigi Messori , Heinz Rüterjans a
b
c
, Robert Kaptein , Michèle Granger-Schnarr &
Manfred Schnarr
c
a
Institute of Biophysical Chemistry Johann Wolfgang Goethe-University of Frankfurt , Germany b
Department of Chemistry , University of Utrecht , The Netherlands c
Institute of Molecular and Cellular Biology of CNRS , Strasbourg , France Published online: 21 May 2012.
To cite this article: Gunnar Ottleben , Luigi Messori , Heinz Rüterjans , Robert 1
Kaptein , Michèle Granger-Schnarr & Manfred Schnarr (1991) H-NMR Investigation of the Interaction of the Amino Terminal Domain of the LexA Repressor with a Synthetic Half-Operator, Journal of Biomolecular Structure and Dynamics, 9:3, 447-461, DOI: 10.1080/07391102.1991.10507928 To link to this article: http://dx.doi.org/10.1080/07391102.1991.10507928
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.
Downloaded by [Rutgers University] at 09:38 04 April 2015
This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions
Journal of Biomolecular Structure & Dynamics, /SSN 0739-1102 Volume 9, bsue Number 3 (1991), "'Adenine Press (1991).
1
H-NMR Investigation of the Interaction of the Amino Terminal Domain of the LexA Repressor with a Synthetic Half-Operator Gunnar Ottleben\ Luigi Messori\ Heinz Riiterjans 1 *, Robert Kaptein2, Michele Granger-Schna~ and Manfred Schnarr3 Downloaded by [Rutgers University] at 09:38 04 April 2015
1
Institute of Biophysical Chemistry Johann Wolfgang Goethe-University of Frankfurt Germany 2
3
Department of Chemistry University of Utrecht The Netherlands
Institute of Molecular and Cellular Biology of CNRS Strasbourg, France
Abstract A synthetic half-operator DNA-duplex, d(GCTACTGTATGT), containing a portion of the proposed recognition sequence (CTGT) of serveral "SOS" genes, has been synthesized. The dodecamer has been characterized through 1H-NMR spectroscopy. Complete assignment of exchangeable hydrogen bonded imino protons has been acheived by applying lD NOE techniques and an analysis of the temperature dependence of the chemical shifts. In order to determine the specific role of the CTGT consensus sequence in the overall recognition process, the oligonucleotide duplex has been titrated with the amino terminal DNA binding domain of the LexA repressor. The observation of substantial changes of 1H-NMR chemical shifts in the imino proton region upon interaction with the protein strongly suggests that the protein binds specifically to the operator DNA The largest deviations of 1H-NMR chemical shifts upon protein binding have been observed for protons assigned to the CTGT segment, thus strongly suggesting a direct involvement of this sequence in the binding process. At high potassium chloride concentrations the 1H-NMR chemical shift deviations are reverted which is consistent with the known drop in the affinity constant of LexA for operator DNA at high salt concentrations.
Introduction LexA, the repressor of the "SOS" system in E. coli., regulates the transcription of about 20 SOS genes mostly involved in DNA repair, mutagenesis and cell division (1-4). These "SOS" genes are induced upon DNA damage by physical and chemical agents through a mechanism involving LexA repressor inactivation by autodigestion at the Ala 84-Gly 85 peptide bond in aRecA dependent fashion (4,5). *Author to whom correspondence should be addressed.
447
Downloaded by [Rutgers University] at 09:38 04 April 2015
448
Ott/eben et a/.
LexA is a repressor protein consisting of two functional domains: the amino terminal domain, which harbors the DNA binding site of the repressor, and the carboxy-terminal domain, which is involved in a weak dimerisation process (6,7). The isolated monomeric amino terminal domain retains most of the DNA binding properties of the entire repressor protein (8). Recent 2D NMR studies of this domain (9) have shown that it contains three a-helices- two of them separated by a tum- inferring a type of helixtum-helix motif. In contrast to canonical helix-tum-helix proteins the relative positioning of helix 2 and helix 3 differs significantly from that usually found (10,11), since the tum is two amino acids longer and contains the bulky aromatic residue Phe-37. Nevertheless the presence of a cluster ofLexA mutants within helix 3, deicient in DNA binding, suggests that this helix should be involved in DNA binding processes (12). The structural deviations observed between the DNA binding domain of LexA repressor and proteins containing a common helix-tum-helix motif raises the question how LexA recognizes specific binding sites on DNA and if it contains a new DNA binding motif or if it has to be considered as a distantly related member of the helix-tum-helix family. It has been proposed that the LexA repressor recognizes operator binding sites with the palindromic consensus sequence 5'-CTGT(at)4ACAG-3' in which CTGT and ACAG are the most highly conserved sequences ( 13). More recent studies indicate that these base pairs contain most of the information required for specific protein binding to operator DNA (11).
In order to further substantiate this point we have undertaken an extensive investigation comprising the synthesis of a number of short DNA fragments and the study of their binding properties to the amino terminal domain of the LexA repressor. In this paper we report on the interaction between the amino terminal domain of LexA and a 12 bp synthetic half-operator duplex d(GCTACTGTATGT) containing the CTGT sequence in the inner part of the oligonucleotide. The investigation is based on the assumption that this fragment contains the minimum requirements for specific recognition by the binding domain. The DNA dodecamer has been investigated by 1H-NMR analysis of the exchangeable imino protons in order to elucidate changes of the DNA conformation upon formation of a protein-DNA complex. In a first attempt to understand the mechanism of specific complex formation the present study is a prerequisite investigation for a more extensive characterization of the protein-DNA interaction in solution using 2D or 3D NMR spectroscopy.
Materials and Methods The dodecamer oligodeoxynucleotide dGCTACTGTATGT and its complementary strand were prepared by the phosphonamidate method (14) and purified by FPLC; the sequence was confirmed according to the Maxam-Gilbert procedure (15). For the 1H-NMR measurements the oligodeoxynucleotides were dissolved in 0.2 M KCl, 10 mM phosphate buffer, at pH 5.6, to a concentration of 1-4 mM. The DNA concentrations were determined from the absorbance at 260 nm, using A260 = 20 em-I for a 1 mg/ml solution. LexA protein was isolated according to published procedures (6).
LexA Repressor and A Half-Operator
449
The protein was cleaved into the carboxy- and the amino terminal domain using the autocleavage reaction at pH 9.4; the amino terminal domain was further purified as described elsewhere (8).
Downloaded by [Rutgers University] at 09:38 04 April 2015
AlllD and the 2DNOE 1H-NMRspectra wereacquiredatSOOMHzonanAM 500 spectrometer, equipped with an Aspect 3000 computer. A 20 NOE experiment of the protein-DNA complex was recorded on an AMX 600 spectrometer, interfaced with a X32 computer. The t 1 files of the 20 datasets were recorded with 2K datapoints each, and zero-filled once in t1 and ~ time prior to the Fourier transformation. A shifted cosine-bell window was applied as an apodization function in each time
B
A
13
14
12
PPM 1
Figure I: H-NMR spectra of the imino proton absorption region of the DNA dodecamer recorded with two different pulse sequences. A 1331; B. Presaturation Conditions: 4 mM DNA dodecamer concentration, pH 5.6, 10 mM phosphate buffer, 0.2 M KCI; 302 K.
450
Ott/eben eta/.
320K
315
K
Downloaded by [Rutgers University] at 09:38 04 April 2015
310K
300K
295K
292K
14
13
12
PPM Figure 2: Temperature dependence of the imino proton resonances in the range of292-320 K Conditions as in Figure I.
451
LexA Repressor and A Half-Operator
domain. Chemical shift values are reported relative to the TSP (sodium 2,2,3,3Tetradeutero-3-(-trimethylsilyl-) propionate) resonance and recorded with an accuracy of0.04 ppm. Solvent resonance suppression was achieved either by presaturation or by semi-selective excitation, namely the 1331 technique ( 16). The temperature dependence of the imino protons was investigated in the range of292-320 K. NOE enhancements are obtained from 10 difference experiments or 20 NOESY measurements. Results and Discussion
Downloaded by [Rutgers University] at 09:38 04 April 2015
Oligonucleotide Sequence
The sequence of the double stranded DNAdodecameris given below; the following numbering of the bases has been adopted:
1
2
3
4
5
6
7
8
9
10
II
12
5'-
G
c
T
A
c
T
G
T
A
T
G
T-3'
3'-
c
G
A
T
G
A
c
A
T
A
c
A-5'
24
23
22
21
20
19
18
17
16
15
14
13
The G ll-C 14 base pairs deviating from the consensus sequence given above has been introduced to facilitate resonance assignments and to improve the duplex stability. The 4 base pairs 5' to the CTGT motif have been introduced to allow the formation of all protein contacts with the DNA backbone as inferred from ethylation interference studies with the rec A operator (17). 1
H-NMR Studies
a) Sequential Assignment of the Hydrogen Bonded Imino Protons
For the assignments of the imino protons the following criteria were used: i) The AT and GC imino proton resonances appear in fairly distinct spectral regions; the former in the 14-13 ppm range and the latter in the 13-12 ppm range (18). ii) The melting of short deoxyoligonucleotides occurs sequentially and starts from the terminal base pairs. Investigations of the differential broadening or disappearance of some 1HNMR resonances with increasing temperature permits the identification of the imino protons in a sequential way, starting from those located at the ends of the sequence (19). iii) The imino protons of adjacent base pairs are about 3.5 A apart; they give rise to small but detectable nuclear Overhausereffects. Such NOE interactions can be observed in 10 and 20 experiments and represent the most straightforward method for sequential assignments (20). 1
The H-NMR spectrum of the imino proton region, recorded at 302 K. in 10 mM phosphate buffer, 0.2 M KCl at pH 5.6, is shown in Figure lA. The spectrum shows
Ott/eben eta/.
Downloaded by [Rutgers University] at 09:38 04 April 2015
452
14
13
12
PPM Figure 3: Map of the observed I D NOE connectivities of the imino protons. Some representative NOE difference spectra are also reported.
LexA Repressor and A Half-Operator
453
ten well separated signals, six of them are observed in the range of 13.4-13 ppm (signals alto a6) and four in the range ofl3-12.1 ppm (signals gl to g4) belonging to the AT and GC base pairs, respectively. With the exception of resonance g3, all signals exhibit approximately equal intensity corresponding to one proton each; only at lower temperatures signal g3 is of similar intensity, too (see below).
Downloaded by [Rutgers University] at 09:38 04 April 2015
A semiselective solvent suppression technique has been applied to detect the imino proton resonances avoiding the excitation of the water resonance (16,21); indeed, application of pulse sequences with a common presaturation step results both in a drastic loss in the intensity of signal gl and most of the AT resonances, and in the disappearance of signal g3, owing to saturation transfer processes with the solvent, as shown in Figure lB. The fact that only two out of twelve expected imino proton resonances cannot be observed may be explained by assuming that at 302 K the imino protons of the two terminal base pairs are subject of fast exchange processes associated with base fraying. By increasing the temperature we expect a progressive increase in the exchange rates of the imino protons resulting in a stepwise loss of intensities. Several 1HNMR spectra of the DNA dodecamer recorded at different temperatures are shown in Figure 2. Upon raising the temperature from 292 to 310 K we observed a selective loss of intensity of signals gl and g3; these resonances therefore have been assigned to the imino protons of the two GC pairs adjacent to the terminal ones. A further increase of the temperature up to 320 K causes a general broadening of all detected signals indicating the melting process; however, it must be recalled that a loss of intensity of the imino proton resonances does not occur at temperatures lower than 'tm. This observed line-broadening does not correspond to the collapse of the whole structure. Observation of nuclear Overhauser effects between adjacent imino protons in one dimensional experiments provides a direct way to perform sequential assignments; two NOE's are expected for base pairs in the innerpartofthe sequence whereas only one is detectable for each nuclotide located at the terminal region. Observed NO E connectivities are shown in Figure 3 together with the results of some one dimensional experiments. Additionally, a 2D NOE spectrum of the DNA dodecamer has been recorded (Figure 4). In the region of the imino proton resonances, we found most of imino proton-imino proton connectivities required for the above mentioned assignments: one NOE between signal gl and al and two between signal g2 and g4 each. This result is consistent with the experimental data of the lD spectra shown in Figure 3. With the experimental conditions used in NOESY experiment and the fact that these cross peaks have a relative low intensity, some of the expected NOE enhancements between adjacent imino protons are not observed due to saturation transfer effects (Figure 4). On the basis of the information obtained from NOE measurements and from the analysis of the temperature dependence we were able to obtain the sequential assignments often imino proton resonances. The following ordering of the resonances
454
Ott/eben et a/.
al a2
a4
gl
aS a6
g2
g4
0
Downloaded by [Rutgers University] at 09:38 04 April 2015
12.5
0 13.0
D
~
D
0
~q pprn
13.0
13.5
·13.5
@
pprn
12.5
Figure 4: Plot between the imino imino proton connectivities of the DNA dodecamer. The 500 MHz NOE experiment was acquired using a mixing time, 'm• of 130 msec and 450 t 1 files, 2k data points, each. The resonance assignment of the imino protons is shown on top of the spectrum; sequential pathway of some imino proton resonances are indicated by a line. Conditions as in Figure lA
is obtained that matches the DNA dodecamer sequence:
assigned signals
:00
sequence : G I
gl
al
a4
g2
a2
g4
a5
a6
a3
C2
T3
A4
C5
T6
G7
T8
A9
TIO Gll Tl2
g3
00
C24 G23 A22 T21 G20 Al9 CIS Al7 Tl6 Al5 Cl4 A13
455
LexA Repressor and A Half-Operator
H
G
Downloaded by [Rutgers University] at 09:38 04 April 2015
F
E
D
c 8
A
14
13
12
PPM Figure 5: Chemical shifts of the imino proton resonances in the presence of various concentrations of the amino terminal domain ofLexA at 302 K. The DNA concentration was 1.5 mM. Protein: DNA ratios for each spectrum are: A) 0.0: B) 0.16: C) 0.33: D) 0.50; E) 0.67: F) 0.84: G) 1.00: H) 1.15.
b) Interaction of DNA Dodecamer with the Amino Terminal Fragment the LexA Protein The chemical shifts of the imino protons within the double stranded DNA are highly sensitive to the base pair ring currents and hence, are significantly influenced by
456
Ott/eben et a/.
protein/DNA ratios
ppm 13,9
13,9
T3 13,7
13,7
Downloaded by [Rutgers University] at 09:38 04 April 2015
.
T 10
T6
13,15
T 21
13,3
..- -r-T 16
T8
13,1
---
13,15
13,3
13,1
12,9
12,8 0,15
ppm 12,9
12,9
G 23 G 20
12,7
12,7
G 11 12,15
12,15
• 12,3
12,3
G7
12,1
.-
12,1
11,9
11,9 0.5
Figure 6: Chemical shift variation of imino proton resonances in dependence of serve raJ protein-DNA ratios.
LexA Repressor and A Half-Operator
457
Downloaded by [Rutgers University] at 09:38 04 April 2015
changes oflocal geometry. Upon addition of the DNA-binding domain ofLexA the imino proton resonances of the DNA dodecamer exhibit substantial variations in their chemical shift values indicating small changes in the DNA helix conformation (see Figure 5). As previously reported for similar systems, the position of the imino proton resonances can be followed at various degrees of saturation ofDNA with the amino terminal domain despite the increase of resonance line-widths (22,23). The chemical shift values of the imino protons for each step of the titration are shown in Figure 6. Complex formation affects mainly four signals out of the ten detected resonances in the imino proton resonance region. For the 1:1 ratio of operator DNA to protein the resonance frequencies g4, g2, a6 and a2 are shifted by more than 0.15 ppm (e.g. g4=212 Hz, a2= 143 Hz), whereas the variations of the chemical shift values of all other imino proton resonances are smaller than 0.1 ppm. It is remarkable to note that all signals with a large change in the chemical shift value correspond to base pairs which are located in the inner part of the dodecamer operator sequence; all of them except one (a6) belong to the proposed recognition motif CTGT. Signals of protons belonging to nucleotides located near the ends of the sequence (namely gl and g3) are almost unaffected upon complex formation. It is evident that the central part of the DNA fragment, and in particular the CTGT segment, is responsible for the recognition process between protein and DNA Moreover, this result qualitatively suggests that the terminal regions are not involved in this process. However, if the CTGT recognition motif is not preceded by 4 base pairs at the 5' site, but only by 2 base pairs as in the palindromic 20-mer CA-CTGT-ATGTI AGAT-ACAG-TG, no substantial chemical shift variations are observed suggesting that in this case no stable specific complex has been formed. This is not unexpected, since at least one contact between LexA and the DNA backbone in each halfoperator is lost in this sequence; in this case the DNA duplex is too short (17).
Upon progressive formation of the complex, we observe significant line-broadening of all resonances that show chemical shift variations larger than 0.15 ppm. Signal g4 for example, shows broadening and the decrease in intensity up to disappearance of the line at 12.4 ppm and the simultaneous appearance of a new signal (g4') at higher field (11.9 ppm) corresponding to the same resonance in the bound state. Careful inspection of the spectra at different degrees of saturation suggests the presence of something like "intermediate exchange" condition. According to a small chemical shift difference between the free and bound state a coalescence signal is observed at titration point C for signal g2 whereas for signal g4 the chemical shift variation is too large to obtain such a signal. To support this interpretation we have recorded spectra at increasing temperatures for a sample with a protein-DNA stoichiometry of0.5, where both signals g4 and g4' are observed. Upon progressively raising the temperature from 302 to 312 K we note that signals g4 and g4' broaden and coalesce into one signal, which subsequently narrows; the signal is located at a position between g4 and g4' as it is expected for a coalescence signal. However, in spite of the fact that the exchange rate for the four exchange processes is believed to be similar, the appearance of the involved signals is different; the remaining signals a2 and a6 undergo chemical shift changes upon complex formation in a more crowded region excluding the observation of their exchange behavior.
Downloaded by [Rutgers University] at 09:38 04 April 2015
Figure 7: Part of a NOESY spectrum of the protein-DNA complex, recorded at 600 MHz. The mixing time tm, was 150 msec and the data were accumulated in 512 t 1 files at an F 1 sweep width of 12000 Hz. The resonance assignment of the imino protons is shown on top of the spectrum; sequential pathway of some imino proton resonances are indicated by a line. Conditions as in Figure lA.
At the end of the titration, the respective signals are relatively narrow. Overall, despite these rather complex exchange processes, it is possible to follow most of the imino proton signals through the titration. In order to consolidate the assignments of the endpoint of the titration, we recorded a NOESY spectrum of the protein-DNA complex (Figure 7). Most of the imino proton imino proton connectivities are absent due to saturation transfer, but the cross peaks between signal g4 and a5 as well as between a6 and a3 show clearly the indicated pathway in Figure 5. It should be mentioned that most of the resonances of the bound species exhibit
line-widths moderately larger than the respective line-widths of the free DNA
459
LexA Repressor and A Half-Operator
Downloaded by [Rutgers University] at 09:38 04 April 2015
c
8
A
14
12
13
11
PPM Figure 8: Dissociation of the complex after the addition ofKCI. (A) Spectrum of the protein-DNA complex. (B) The same absorption region of A for the complex in the presence of I M KCI. (C) Spectrum of the free DNA dodecamer.
dodecamer. On the average the line-widths of the bound species are by a factor 1.3l.Slarger than those of the free species. This increase in line-width can be explained
460
Ott/eben et a/.
Downloaded by [Rutgers University] at 09:38 04 April 2015
with the higher molecular weight of the complex as compared to the free dodecamer. However, signal g4 reveals a more extensive line-broadening (an increase in line-width by a factor 5) which must be due to a different mechanism. In general, two possible explanations for such selective line-broadening can be proposed (24): i) A selective increase in the T2 -I relaxation rate owing to a decrease in mobility of a particular region of the molecule as a result of complex formation with the protein. This mechanism is not unlikely in this case, since (g4) that undergoes broadening belongs to the portion of the molecule which is beleived to be directly involved in binding the protein. ii) The line-broadening can be due to an exchange process between two or more conformations of the bound oligonucleotide which can be intermediate on the NMR time scale. We can rule out that the broadening is related to an increase in the exchange rate with the solvent since increasing the temperature causes a marked narrowing ofline g4' and no loss in intensity. In order to distinguish between the above mentioned mechanism, we performed several line-widths experiments at various field strengths. Increasing line-broadening with the square of the field is expected in case of mechanism "ii", whereas litle change should be observed for mechanism "i". Thorough comparison of the imino proton resonances recorded at 270 MHz, 500 MHz and 600 MHz showed no significant change in line-broadening. As a result, the decrease in mobility of the particular part of the recognition motif seems to be a more reasonable explanation for this line-broadening upon complex formation. All obtained results provide evidence for the formation of a specific protein-DNA complex between the DNA binding domain ofLexA repressor and the half-operator, making this system suitable for a more detailed structural investigation by multidimensional NMR spectroscopy.
c) Dissociation of the Specific Complex with Increasing Concentrations of KCI By adding increasing amounts ofKCl to the solution of the protein-DNA complex, the 1 H-NMR spectrum of the free DNA dodecameris almost recovered (Figure 8). High salt concentrations apparently lead to the disruption of the protein-DNA complex; the double stranded structure of the DNA dodecamer is kept intact as concluded from the observation of the imino proton resonances. This result is consistent with the reported decrease of the association constant of LexA protein for the uvrA operator and for non-operator DNA upon increasing the salt concentrations (25,26).
Conclusions The binding of the amino terminal domain ofLexA to a dodecamer oligonucleotide, containing the proposed recognition sequence CTGT, has been investigated by 1HNMR experiments. The chemical shift changes of the imino proton resonances describe the specific binding of the domain to the operator sequence. The signals most affected belong to the proposed recognition region CTGT, which suggests that these base pairs play a key role in the binding process. The complex pattern of the line-widths
LexA Repressor and A Half-Operator
461
observed during the titration with the protein has been explained on the basis of exchange processes and a restriction of dynamics. It has been explained on the basis of exchange processes and a restriction of dynamics. It has been also shown that high concentrations ofKCl causes the breakdown of the protein-DNA complex.
Acknowledgements
Downloaded by [Rutgers University] at 09:38 04 April 2015
The authors thank Dr. Brian Fera for the preparation of the DNA fragment. This work was supported by grants from the European Communities (ST2J-0291), from INSERM (871 007) from the" Association de Recherche contre le Cancer", from the "Ligue Nationale de la Luthe contre le Cancer". References and Footnotes
I. 2. 3. 4. 5. 6. 7. 8. 9.
10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
Little, J.W. and Mount, D.W., Cell 29, 11-22 (1982). Walker, G.C., Microbiol. Rev. 48, 60-95 (1984). Howard-Flanders, P., West, S.C. and Stasiak, A., Nature 309,215-220 (1984). Friedberg, E.C., "DNA Repair", Freeman, San Francisco (1985). Little, J.W. and Hill, SA, Proc. Nat/. Acad. Sci. USA 28, 4199-4203 (1985). Schnarr, M., Pouyet, J., Granger-Schnarr, M. and Daune, M., Biochemistry 24,2812-2818 (1985). Schnarr, M. Granger-Schnarr, M., Hurstel, S. and Pouyet, J., FEBS Lett. 234, 56-60 ( 1988). Hurstel, S., Granger-Schnarr, M., Daune, M. and Schnarr, M., EMBOJ 5, 793-798 (1986). Lamerichs, R.M.J.N., Padilla, A., Boelens, R., Kaptein, R., Ottleben, G., Raterjans, H., GrangerSchnarr, M., Oertel, P. and Schnarr, M., Proc. Nat/. Acad. Sci. USA 86, 6863-6867 (1989). Brennan, R.G. and Matthews, B.W.,J Bioi. Chern. 264, 1903-1906 (1989). Strub!, K., Trends Biochem. Sci. 14, 137-140 (1989). Oertel-Buchheit, P. Schnarr, M. and Granger-Schnarr, M., Mol. Gen. Genet. 223,40-48 (1990). Wertmann, K.F. and Mount, D.W.,J Bacteriol. 163, 376-384 (1985). Matteucci, M.P. and Caruthers, M.H.,J Am. Chern. Soc. 103,3185-3191 (1981). Maxam, AM. and Gilbert, W., Methods Enzymol. 65, 499-560 ( 1980). Hore, P.J.,J Magn. Reson. 55,283-300 (1983). Hurstel, S., Granger-Schnarr, M. and Schnarr, M., EMBO J 7, 269-275 (1988). Zuiderweg, E.R.P., Scheek, R.M., Veeneman, G., van Boom, J.H., Kaptein, R., Raterjans, H. and Beyreuther, K., Nuc/. Acid Res. 9, 6553-6569 (1981). Weiss, M.A., Patel, J.D., Sauer, R.T. and Karplus, M., Nuc/. Acid. Res. 12, 4035 ( 1984). Buck, F., Hahn. K.D .• Brill, W.. Raterjans, H .. Chernov. B.K.. Skryabin, K.G .. Kirpichnikov, M.P. and Bayev, A.A,J Biomol. Struct. Dynam. 3, 899-911 (1986). Rajagopal, P .. Gilbert, D.E., van der Maarel, G.A., van Boom, J.H. and Feigon, J.,J Magn. Reson. 78, 526-537 (1988). Hahn, K.D .. Buck, F .• Rtlterjans, H .. Chernov, B.K.. Skryabin, K.G. and Kirpichnikov. M.P.,Eur.J Biochem. 12, 87-95 (1985). Scheek, R.M .• Zuiderweg. E.R.P., Klappe, K.J.M., van Boom, J.H .• Kaptein, R., Rtlterjans H.and Beyreuther, K., Biochemistry 22, 228-235 (1983). Schumacher, R.. Buck, F. and Rtlterjans, H .. Nuc/. AcidY Res. 13,5097-5105 (1989). Betrand Burggraf, E., Hurstel, S.. Daune, M. and Schnarr. M.,J Mol. Bioi. 193,293-302 (1987). Hurstel, S., Granger-Schnarr. M. and Schnarr. M., Biochemistry 29, 1961-1970 (1990).
Date Received: March 1, 1991
Communicated by the Editor Wolfram Saenger
