• Aucun résultat trouvé

Archaebacterial histone-like protein MC1 can exhibit a sequence-specific binding to DNA

N/A
N/A
Protected

Academic year: 2021

Partager "Archaebacterial histone-like protein MC1 can exhibit a sequence-specific binding to DNA"

Copied!
8
0
0

Texte intégral

(1)

HAL Id: hal-02708673

https://hal.inrae.fr/hal-02708673

Submitted on 1 Jun 2020

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Archaebacterial histone-like protein MC1 can exhibit a sequence-specific binding to DNA

Caroline Teyssier, Bernard Lainé, Alain Gervais, Jean-Claude Maurizot, Francoise Culard

To cite this version:

Caroline Teyssier, Bernard Lainé, Alain Gervais, Jean-Claude Maurizot, Francoise Culard. Archaebac- terial histone-like protein MC1 can exhibit a sequence-specific binding to DNA. Biochemical Journal, Portland Press, 1994, 303 (2), pp.567-573. �hal-02708673�

(2)

Archaebacterial histone-like protein MC1 can exhibit a sequence-specific binding to DNA

Caroline TEYSSIER,* Bernard LAINE,t Alain GERVAIS,* Jean-Claude MAURIZOT* and Fran9oise CULARD*t

Tentre de biophysique mol6culaire, rueCharles Saolron, 45071 Orl6ans-Cedex, France and tU 325 INSERM Institut Pasteur, 1 rue Calmette, 59019 Lille Cedex, France

The binding ofMCI protein, the major chromosomal protein of the archaebacteriumMethanosarcina sp. CHTI 55, to the region preceding the strongly expressed genes encoding methyl co- enzyme reductase in aclosely related micro-organism has been investigated. By gel retardation and DNAase I footprinting assays, weidentifiedapreferential binding sequence in an open reading frame of unknown function. The large area of DNA

INTRODUCTION

MC1 is the most abundant chromosomal protein.present in various species of Methanosarcinaceae [1,2]. In the Methano- sarcina sp. CHTI 55 strain, MCI isa polypeptide of 93 amino acid residues (Mr 11000) which is mainly characterized by

ahigh number of basic and acidic residues (respectively24 and 12) distributed along the entire length of the protein [3]. With respect to the characteristics of its primary and secondary structure, andparticularly the distribution of basicresidues and the low a-helix content [4], MCI differs significantly from eukaryotic histones and from eubacterial and other archae- bacterial histone-likeproteins (forareview seeref. [5]).

MCI bindstoDNAas amonomerinanon-cooperativeway

[6]; it can protect DNA against thermal denaturation [7] and against radiolysis by fast neutrons [8]. Its DNA-interacting regionhasbeen identifiedby photochemical cross-linking [9]. As for the protein HU from Escherichia coli, MCI is able to

promote the circularization of short DNA fragments by T4

DNAligase [10,11].

Amongthe variousprokaryotichistone-like proteins, several proteins, suchas theprotein HU, arebelievedto bindto DNA only non-specifically [12,13], while others bind both non-

specifically and specifically to DNA. This is the case for three proteins, TFI from the bacteriophage SPOI [14], IHFand FIS from E. coli[15,16],thetwoformerbeingverycloselyrelatedto HU. Inaddition, it has beenshownrecentlythat theproteinH- NS,oneof themostabundantnucleoid-associatedDNA-binding proteins, exhibitshigh affinityforfragments carrying promoter

sequences in vitro; in thiscase theprotein seems torecognize a

structural feature of the DNA [17,18].

This ledustoexaminewhether,inadditiontoitsstrongnon-

specific binding [6,11], MCI interacts specifically with some

DNA regions. To study DNA sequences involved in gene

regulation, we exploited the fact that the regionpreceding the

genes encoding for the enzyme methyl coenzyme reductase (methyl CoM reductase) involved in the final step of methane productioninalargenumberofmethanogenicbacteria[19],has been characterized [20,21]. Within this sequence, a 400bp

protected against DNAaseI isinterruptedby astrongcleavage enhancementsiteoneach strand. By circularpermutationassays, weshowed that the DNA bends upon MCI binding. Furthermore weobserved that the presenceof a sequence outside thebinding site can induce an unusual electrophoretic behaviour in some complexes.

intergenic region has a 26% GC content versus the 42% GC contentof the whole chromosome [20]. This low GC content is a noticeable feature shared by sequences necessary for tran- scriptioninfront of archaebacterial genes[20,22]. Itistherefore of interest to compare the binding of themajor chromosomal protein MCI to thevarious sequences of this region.

Here, we report that MCI has a preferential binding site located inanopen readingframe (ORF)of unknown function and that this binding induces DNA bending. Furthermore we observeanunusualelectrophoretic mobility ofsomefragments bearing this site when complexed toMCI.

MATERIALS AND METHODS MC1 protein

Methanosarcina sp. CHTI 55(DSM 2906)wasgrownasindicated inChartier etal.[1].TheproteinMCl wasprepared as previously described [1,4] with the modifications indicated in Laine et al.

[11]. The protein is 99% homogeneous as ascertained by SDS/PAGE [4].Itsconcentrationwasdeterminedby absorption spectroscopywithanabsorbancecoefficientof 11 000 M-1 cm-' at280nm.

DNAfragments

The DNAfragments containingsequences upstreamofthe genes codingformethyl CoMreductase were obtainedbyrestriction digestionsofa 1.2kbfragmentcloned inpUC9[20] (Figure la).

Forconvenience thepositionswithin thisfragmentarenumbered startingfrom the EcoRI site.

For the gel-retardation assays, the 433bp fragment was obtainedby ScalandBstEIIdigestionatpositions636and1070, and the 399bp fragment by EcoRI and BglII digestion at positions1and 399.Digestionof the 399bpfragment by Bspl286 at position 190 gave two fragments of 190bp and 209bp respectively.

Fragments usedasprobesinfootprintingassayswereobtained bydouble digestionwith HpaII and AluI atpositions 256 and 422,orwithHinfl andBglIIatpositions293 and 399(Figurelb).

Abbreviations used:IHF, integration host factor; TF1, transcriptionfactor;FIS,factor forinversionstimulation;ORF,openreading frame; methylCoM reductase, methylcoenzyme reductase.

I To whomcorrespondence should be addressed.

(3)

568 C.Teyssier and others

(a)

ORF

l Il

1 1 EcoRl

Bspl286 HpallHinfl

(b)

190 256 293

Bg/llAlul 9 4 399422

ORF Methyl CoM

reductase 1200

.Hinfl 534

Hinfl 5I 534

Figure1 Schematicrepresentation and restricffonmapof the 1.2 kb DNA containing thesequence upstream to thegenes encoding for methyl CoM reductase

(a) The 1.2 kb DNA region was isolated by Allmansberger et al. [20] from the strain Methanosarcina barkeri. For conveniencetheDNA isnumbered starting fromtheEcoRI site.The transcriptionstartof thegenescoding for methylCoMreductaseandtheORFin theothersense

codingforanunknownproteinareindicated. Thepreferential MCI bindingsitededuced from Figure4is boxed.(b) RestrictionmapoftheDNAregion used for the footprinting experiments.

(c)Localization of the twofragmentsused for thepermutation assays.

CGR films. Theautoradiographswereanalysed by densitometry at580nm on aCamagTLC Scanner II densitometer.

Inthecompetition experiments, due tothenon-specific bind- ing, we always usedfragmentsofsimilar size to make the number ofnon-specificsitesequivalentonbothfragments.We also used low protein to DNA ratios to maximize specific versus non- specific binding.

DNAase I footprinting

Each reaction mixture (final volume 25#I) contained 10 mM MgCl2, 5mM CaCl2, 75 mM NaCl, 1 mM EDTA, 10 mM Tris/HCl,pH 7.5, the32P-end-labelledDNAfragment (10-9M) anddifferentconcentrations oftheprotein. After incubationat room temperature for 20min, the samples were digested with DNAase I(1 ng) for 20 s. Thereactionwasstopped with12,ul of DNAase I stopsolution(0.34 M EDTA,0.9 M sodiumacetate, 0.07,ug/mlcarrierDNA). After ethanolprecipitationthesamples were loaded on a 8% sequencing gel andrun at 40 W for2h.

DNAsequencingwascarried outby theMaxam-Gilbert method.

For the study of, the ejectrophoretic behaviour. pf, jhe.

complexes,weusedfragments of 133, 150, 203 and 236 bp, which have no specific binding sequence for MC 1. These fragments

werefromE. coli and havebeen previously described [6,11].

For thecircular-permutationassays,fragments of 142 bp and 241 bp, derivedfrom digestion of the 1.2 kb DNA fragment with XhoIIand Hinfl digestions respectively (Figure lc),werelabelled andcircularized.The DNAligationswereperformedataDNA concentration of 0. 1,ug/ml by T4 DNA ligase (BRL) according to the manufacturer's instructions. To enhance the yield of the 142bp DNA minicircle, DNA ligation was performed in the

presence of MCl (at a protein-to-DNA molar ratio of 150), taking advantageof the fact that MCI binding favours the ring formation of short DNA fragments [11]. After incubation, samples were deproteinized, loaded on preparative poly- acrylamide gels [4% acrylamide, 0.2% bisacrylamide in TBE buffer(89mMTris/boric acid, 2 mM EDTA) in thepresenceof 0.6,ug/ml of ethidium bromide). Each DNA minicircle was

extracted from the gel, and purified by phenol-chloroform treatmentsfollowed by Prepac chromatography columns(BRL).

The 142bp DNA minicircle was linearized by digestion with Hinfl, HpaII or RsaI to obtain a set of circularly permutated fragmentsasshown inFigure7(b), The 241 bp DNA minicircle

was linearizedbydigestion with BglIIorDdeI.

DNA fragments were 5'-end labelled by T4 polynucleotide kinase with [y32P]ATP. The DNA concentrations were deter- minedby absorptionspectroscopywithanabsorbance coefficient of13000M-1 cm-' per bpat260nm.

Gel-retardation assays

The binding of MCI to defined restriction fragments was

performedin10 mMTris/HCl,pH7.5, containing 75 mM NaCl and 1 mM EDTA. After equilibration for 20 min at 20°C, samples were mixed with 0.2 vol. of loading buffer (0.01% Bromophenol Blue, 50% glycerolin 10mM Tris/HCl, pH 7.5, 1mMEDTA), and loadedonpolyacrylamide slab gels in 89mM Tris/boric acid, 2 mM EDTA (TBE). Electrophoresis was

performed at 1OV cm-'. The gels were dried and exposed to

Analysis of thebinding data

Theanalysis ofthebinding datawasperformed usingthebinding polynomial procedureas developedby6'Clore etal.'[23]. As the quantitative analysis was performed at a low binding ratio (experimentswerelimitedto proteintoDNA ratiossuch that a maximumofthree to fourMCI proteinswereboundper DNA fragment), onecantruncate thebindingpolynomialto adegree 5. Wechecked,byincreasing thisdegree, that this was sufficient foraccurateanalysis ofourbinding data. Undertheseconditions thebinding polynomialwasforafragment without a specific site

j-5

5 HNKi Li

Z= l+ZJ-l

i-1

andfora fragmentwith a specific site

J 5 Ks

5 [NN +KN)K -Li Z=I+ZJ-1

i-l

where KS is the binding constant to the specific site, KN the bindingconstant toothersites, N thenumberof base pairs of the fragmentand L is the free concentration ofMCI protein.

For agiven free concentration of protein, L, the value of Z can becalculated forvalues of KN andKS. Fromthe valueof Z, the amountofbound MCI iscalculated as indicatedbyClore et al.

[23],andconsequentlythe total amountofprotein isknown. One cantherefore get the fraction offreeDNAfragment,(1/Z),as a function of the total protein concentration. K. and KN were deduced from comparison between this theoretical decrease of the fraction of free DNA fragment and that experimentally determined afterscanningof theautoradiographs.Alternatively, we tried to use the intensities of each complex to follow the binding process but less precise results were obtained. The bindingpolynomialwasalsocalculatedaccordingtotheEpstein theory [24] assuming a site size of 11 base pairsas determined previously [6]. Similar results wereobtained. This is due to the fact that thefragmentlength (about200bpin thecase wehave analysed)islargeincomparisonwith the site size and because the binding ratio is lowaspreviously mentioned.

., IT

Xholl Xholl

(c) -ME---A

228 Hinfl 370

RRRO

293 -55&&a

(4)

RESULTS AND DISCUSSION

Localization of a preferential binding site of protein MC1

Wepreviously studied the non-specific binding ofMCI protein to DNA by the gel-retardation method. We showed that upon

MC1 binding, DNA fragments exhibit a gel electrophoretic

pattern consisting of a ladder made of several discrete sharp bands with retardedmobilities when compared with the onesof the free DNA fragments. Quantitative analysis of the intensities of these bands allowedustodemonstrate that theycorrespond to

complexes made with theDNA fragments bearing1,2, 3,...MCI molecules [6].

Tolook for aputative MCI preferential bindingweused the

same method. The MCI-DNA binding was investigated with various fragments derived from a 1.2 kb DNA isolated from Methanosarcina barkeri that contains sequences preceding the methyl CoM reductasegenes [20]. In ordertodetecteven a low specific binding, and due to the strong non-specific binding of MC1 protein,weused thegel-retardationassaysunder particular conditions. First, in competition conditions between fragments of similar size to get an equivalent number of competing non-

specific sites on both fragments. Secondly, with low protein to

DNAratiostoreduce thepresenceof non-specificcomplexes. We first compared the MCI binding to two DNA fragments, a 433bp DNA fragment (position 636-1070) anda 399bpDNA fragment (position* 1-399). The first fragment contains the intergenic region between the methyl CoM reductasegeneandan

ORF of unknown function, whereas the second fragment is located in this latter ORF. Figure 2 shows the titration of the mixture of thetwo DNAfragmentswithincreasingMC1 protein concentrations. To differentiate between the two species, com- plexes with the two separate fragments are also shown in the

- + 399bp + + + + + +

+ - 433bp + + + + + +

399 433

bp bp

2

0

11 12

-

0

A B C D E F

Figure 2 Competition between two DNA fragments derived from the 1.2 kb DNA upstream of the genes coding for methyl CoM reductase for MC1 binding

Autoradiograph of gel-retardation assay under competition conditions between a 433 bp (position636-1070) anda399bp(position 1-399)DNAfragment.Theprotein concentrations

werein lanes A-F: 0, 0.5 nM,1 nM, 1.6 nM, 2.1 nM and 3.2nM respectively. The DNA fragment concentrationwas0.21 nMfor the 399bp and 0.19nMfor the 433 bpfragments.

Thearrowsindicatethenumberof bound protein moleculesperDNAfragment.ThelanesTl and T2displaythepositions of complexesconstituted of MCi and either the 399bporthe 433bpDNAfragments.Thesampleswereseparatedona7%acrylamide/0.1 % bisacrylamide gel,in TBEbuffer.

figure. For all the ratios of protein to DNA tested, slightly more complexes were formed with the 399 bp DNA fragment than with the 433 bp one. Upon increasing further the concentration of MCI, complexes with more than one bound protein per fragment are formed. This is apparent in Figure 2, lane F,where a band appearswhich can beidentified as that corresponding to two MCI proteins bound on the 399 bp fragment.

The 399 bp DNAfragment was cleaved with the endonuclease Bsp1286to obtain two fragments ofnearly the same length, one corresponding to the region 1-190 (190 bp) and the other to the region 191-299 (209 bp) Their relative affinities forMCl protein were compared by a competitive gel-retardation assay using conditions mentioned above (low protein to DNA ratios). Upon increasing amounts of protein, we observed that the free 209 bp fragment disappears more rapidly than the 190 bp one (Figure 3).This clearly demonstrates that the binding is stronger for the 209bp DNAfragment than for the 190 bp one. Simultaneously with the disappearing of the free DNA fragments, higher bands regularly appear. As expected, they are more abundant with the 209 bpfragment. So, for the higher protein to DNA ratio (lane D) essentially only a complex bearing one MCI molecule is visible with the 190 bp fragment, whereas complexes containing one, two, and three MC 1 molecules (complexes 1, 2 and 3 respectively) are visible with the 209 bp fragment. As shown later, with the 209 bp fragment, the complex 1 results from the specific bindingof one MCI- molecule, and the complexes 2 and 3 are due to the non-specific binding of one and two MC1 molecules in addition to one specific binding.

Furthermore, we have to point out that complex 1 with the 209 bp DNA migrates slightly faster than the one formed with the 190bp DNA fragment, although the two free DNAs and the two complexes containing two MC1 molecules per fragment migrate as expected according to the length of the DNA fragments used.

To gofurther into the location of the preferential binding of protein MCI, we subjected the DNA-protein complexes to DNAase Ifootprinting. Figure 4(a) shows the results for the top strand. A region of partial protection is clearly observed between positions 318 and 338. The amounts of protein necessary to detect thefootprint and the specific complex by gel retardation under the same binding conditions are similar. The footprint is interrupted at two positions: a lack of protection adjacent to a highly enhanced cleavage site. Protection of the bottom strand is extended over about the same DNA region (Figure 4b). As for the top strand an enhanced cleavage site is observed. Relative to that of the top strand, itsposition is located with an offset of 3 bp on the 3' side of the sequence (Figure 4c).

Theprecise boundaries of the protection are difficult to locate, because on one hand the protection is not total, and on the other hand several positions are not cleaved by DNAase I; this is particularly evident for the stretch of six adenines at the 5' end onthe top strand(Figure 4a). Nevertheless, we estimate that the extent of protection by MCI ranges from 20 to 30 bp. This sequenceexhibits no sequence similarity with other sequences of the DNA banks. Its basecomposition (about 60%/ AT) is close to that of the whole 1.2kb DNA fragment.

When reported on an 'unwrapped' cylindrical projection, the protected area draws a largesurface. It spreads on one 'side' of the DNA overabout two double-helix turns and surrounds a minor groove region which is DNAase-hypersensitive (Figure 4c).

Experiments with larger amounts of protein did not reveal otherareasofprotection beforethe saturation of thefragment.

This stepwas accompanied by severalenhanced cleavages that probably reflect some important DNA conformational changes

... ...

(5)

570 C.Teyssier and others

190 + + + + -

209 + + + + +

A B C D E

190 209

4- 5 4- 4 4-3 4-2 4- 2 4-i

4- 0 4-1 4-0

Figure3 Competition betweentwoDNAfraigmentsIsolated from anORF sequence for MCI binding

The 209bp (position191-399)and the190bp (position1-190)DNAfragmentsderived from adigestionofthe399bpDNAfragmentwith BspI1286.Equimolaramounts(0.45 nM)ofthe twofragmentsweretitrated with thefollowingtotal concentrations of MCi inlanes A-D:0, 4nM, 8nMand 20 Lane E shows thepositionof thecomplexesformed withthe 209bp DNAfragment.Thesampleswere separatedona6%acrylamide/0.3% bisacrylamidegel,in TBEbuffer.

(resultsnotshown).Thepresenceofaunique preferentialbinding sitealongthe studied DNAfragmentallowedustoquantitatively analysethe binding competition shown in Figure3. This isnot

straightforward since the very large number ofpotential non-

specificsitesonthefragments hastobe taken intoaccount. On

a random DNA fragment each base is thebeginning ofanon-

specific site except then terminal bases, where n represents the numberof bases covered bytheprotein. Forthe mixture of the twofragmentsof about 200bp, onecan calculatethat there are

about.700 potential non-specific sites in competition with the preferential site for the MCI binding. The binding features of MCI werequantitatively analysed asindicated in the Materials and methods section. The decrease of the quantity of the free DNA has been calculated for several values of ratio of the specifictonon-specificbindingconstantsandcomparedwiththe experimentally observed decreases of the intensity of the free DNAbands. Wefind that this ratio is about 2000. This indicates

adifferenceofbindingenergyof about4kcal which is far smaller than that calculatedbetween thespecificandnon-specific binding ofprokaryotic repressors. For example, in the Lac operator- repressor system Ks/KN- 106-l07 [25]. However, the concen-

tration of MCI in the cell is--markedly larger than that of a

prokaryotic repressor: it is estimated to be one MCI1 molecule

for every 150bp, corresponding to a concentration of

10-4-10-5M[26].With suchaconcentrationwe canassumethat thispreferentialsite ispermanentlymoreoccupied byMCI than itsneighbouring sequences in vivo.

Behaviour of the complexesin polyacryiamide gels

Themobilityofcomplexesmade withvarious DNAfragmentsof different lengths containing the preferential binding sequence werecomparedwith theonesofcomplexesmade withfragments lacking this sequence. An example of such an experiment is shown in Figure5. TheDNA-protein complexes are progress- ively andregularly retarded withincreasing amounts of bound protein, except with the fragment containing the preferential bindingsequence(lane E).The relativemobilities of the different

(a)

AG CT 0 0

A B C C D E F G H

...

AM

.4"W!

...

...

:MW

X. W.-4w

..... qAW-ow *W

(b)

AG +

A B C

...

Ai"

-:%::-:.

Aw

...

ANN

...

Ow

...

AW:

AW:

(c)

Figure4 DNAase I protection analysisofMCl-DNAcompiexes

(a)Topstrand labelled(position 257-422).LanesA andBcorrespondtoMaxam andGilbert sequencingreactions.LaneCis theDNAase digestintheabsence ofprotein. LanesD to H arethe DNAase digestin the presenceof increasing MCi concentrations(26nm,52nM, 87nM,140nM 170nM).The DNAfragmentconcentrationwas9 nMineach lane.(b)Bottom strand labelled(position 191-399).Lane Ais the A+Gsequencing reaction. LaneB isthe DNAase digestintheabsence ofprotein.LaneCis the DNAase digestin the presenceof protein (50nM).The DNAfragmentconcentrationwas5 nM.(C) Planarrepresentationof the cylindrical projectionof the DNA molecule(10.5bpperturn)intheregionof thepreferential binding site of MC1. The internucleotidic bonds protected from DNAase cleavage are

symbolized bythe full circles.The twoarrows indicatethe bonds with enhancedcleavage in the presence ofMCi protein.

DNAfragments(thequotientof themigrat~iondistancesof free DNA and complexed DNA) werecalculated, and plotted as a functionof the number of boundproteinmolecules perfragment.

Forthenon-specific fragments, alinearrelationship isobserved

(6)

A B C D E F G

... . ... ....;.=... . .. .... (a)

Figure5 Comparison ofmobilities ofcomplexes formed betweenMC1 and various DNAfragments

Autoradiographof gel-retardation assay with DNA fragments of, respectively,fromlanes A to F, 133, 150,203, 236, 399 and 433 bp.Onlythe 399 bpfragment (position 1-399) contains the preferential binding site of MCI protein.The 399 and 433bpfragmentswerederivedfrom the 1.2kbDNAfromMethanosarciniabarkeriwhereas the otherfragmentswerefromE.coli. Lane G:size markersobtained fromapBR322-MSP digest.Thesamples wereseparatedon a7%

acrylamide/0.1% bisacrylamidegel, in TBEbuffer.

(Figure 6a). Inthisfigure only the data obtained with protein to DNA ratios up to three are shown, but the graphs were still linear with higher ratios when observed. This is in accordance with the behaviour observed by Bading [27] on the basis of experimentsperformedwith the lac repressorbindingto a203bp DNA fragment. On the contrary, the complexes formed with three DNA fragments containing the preferential binding site havemobilities which donotfollow thesamerule. The curves are notlinear and thespecific complexes migratefaster thanexpected (Figure 6b).

DNAconformationcangreatlyaffect DNAmigrationthrough polyacrylamide gel matrix. So, intrinsically curved DNA and protein-induced bent DNA are strongly retarded [28,29]. The degree of theretard induced bythe protein greatly dependson the positionof the binding siteon the DNA fragment [29]. To determine whether DNA bending accounts for the abnormal mobilityobserved, we used thecircular-permutation assay. We therefore circularized a 142bp XhoII fragment containing the preferential binding site, then we cut it by different single site restriction enzymes yielding a set of four permuted fragments.

The result of this circularpermutation isshown inFigure7.The mobilities of the free DNA fragments are nearly identical, indicating that there is little or no sequence-directed DNA bending in the free DNA fragment investigated. On the other hand the mobilities of the specific complex bearing one MCI protein vary considerably, depending on the location of the preferentialbindingsitewithin thefragment.The lowestmobility is observed with the fragment with a centrally located binding site, as deduced from the footprint experiments. Clearly this

0

E

a)

._4cc

0 1 2 3

BoundMC1 per DNA fragment (b)

2.6

- 106bp *- 209bp -_-399bp

2.4-

2.21

` 2.0

.0

E 1.8

0

1.8 a) 1.6

cr

1.4 1.2 1.0

0 1 2 3

BoundMC1 per DNAfragment

Figure 6 Analysis of mobilities of complexesformed between MC1 and fragments containing and notcontaining thepreferential bindingsite The relativemobilitiesof thecomplexes(free/complexed DNA)areplottedas afunction of the amountof boundMC1 molecules per DNAfragment. (a) Non-specific bindingDNAfragments.

The 150bp and the 203bpareE coli DNAfragments [11],and the 433bp DNA encompasses sequence 636-1070 within the 1.2kb DNA from Methanosarci/na barkeri. (b) Fragments containingthepreferentialMCi bindingsite.Thefragmentsof106,209and399bpencompass sequences294-399,191-399 and1-399respectively.

permutation assay strongly indicates that MCI binding to the preferential binding site induces DNAbending. Itislikelythat, duetothe moderatespecificityof thebinding,the observedeffect is underestimated. Indeed assuggested insimilarcaseswith the histone-like proteinsIHFandTFI [30-32], the DNA mobilities arecertainlyaveraged bythedifferentpositionsoccupied bythe protein on the DNA fragments during the time of the electrophoresis.Wetherefore think that thepermutationcurveis flattened and that MC1 binding to thepreferential bindingsite induces quitea largeDNA bending.

DNA bending requiresmultipleprotein-DNAcontacts. This isconsistent with the resultshowing that the site sizeprotected against DNAase I is quite large (20-30bp). It islikely that the DNAiswrapped around theprotein,ashasbeen shown for other complexes between DNA and histone-like proteins [30,33].

Furthermore, as enhanced sensitivity to DNAase I can be generated byexpanding the minor groove[341,thetwostrongly enhanced cleavage sites are probably-ldocated in an outward

(7)

572 C.Teyssier and others (a)

Free DNA

A B C D

....

:.:.::.:

..~~~~~~~~~~~~~~~~~...

......

...

~ ~ ~ ~~~~..:

:.: ....

...... .Free

.... ....

:. ::... :D

A B C D

Figure8 Unusual mobilityofthecomplexesformed with BgIllfragments

A 241 bpH/nflfragmentwascircularized andcutby BglllorDdeltoobtain threecircularly permuted fragmentsas explained in the Materials and methods section. Lanes B-D are

respectivelytheH/nfl, the Bglll,and theDdelDNAfragments in thepresence ofMCi. Lane Ais the free H/nfl DNAfragment. Thesampleswere separated ona5%acrylamide/0.25%

bisacrylamide gel,inTBE.Thepositionsofthesites relativetotheleft end of thefragmentare:

H/nfl (0.12), Bg/ll (0.68)and Ddel(0.78).

sequence outside the binding site and lacking the permutated

fragmentcould benecessaryto observe thisparticular mobility

D- , - --. _ . of the specific complexes. In fact the sequence XhoII-BglII

1 30 60 90 120 142bp (position370-399)is absent in the permutated fragment, whereas

I a. I, I thethreefragmentsused in theexperimentshown inFigure 6(b) Hpall Hinfl Rsal Xholl Hpall span up to the position 399 (BglII cut). A second circular- permutationexperimentwithafragment encompassingtheBglIl 7 Bending of DNAby MC1 restriction site (Hinfl fragment, position 293-534) was then performed.The result ofthis permutation is shown in Figure8.

bpXhodlfragment was circularized and cutbysingle-siterestrictionenzymestoobtain Whenthe DNAislinearized withBglIl,andonlyinthatcase,the urcircularly permutedfragmentscontainingthepreferentialbindingsite asindicated abnormal behaviour of the specific complex is again observed terialsandmethodssection. DNAfragmentswere:laneA,Hinfl;laneB, Hpall;lane (laneC): it migrates abnormally fast in regard to the complex lane D, Rsal. Thesampleswereseparatedona6%acrylamide/O.3%bisacrylamide containingtwoMCIproteinsonthesamefragment,furthernore

IE buffer.(b) The relative mobilities of thecomplexeswereplottedas afunction of ithasthehighestmobility althoughthe preferential bindingsite

)n ofthe site relative to thefragmentends. The closed boxdenotes theMCl-binding ishate nearest to theprefrentialfromte

is located nearest to the centre of the fragment (68 00 from the leftend). Thetwo otherspecificcomplexesmigrateasexpected (Figure 8, lanes BandD).

Toourknowledgethat is the first time that such behaviour has been reported. Only the mobility of thecomplexes is affected, position on the bent DNA, with the minor groove and neither their order of appearance nor their respective

led. proportionismodified.Wedidnotfind any clear-cutexplanation

lis permutationassay,thereisnoabnormally highmobility of this phenomenon, but we can formulate some of its specific complexes, and particularly the behaviour de- characteristics. (1)A sequence with amaximal lengthof 30bp

I inFigure 6(b)isnolongerobserved whatever theposition located between 30 and 60bpfromthepreferentialbindingsite specific siteon the fragments. We thensuspected thata isrequired(XhoII-BglII).(2)Thepositionof this sequence in the

(b) 1.40

1.351 Hpall HpaIll

Xholl

Iinfl Rsal

>- 1.30- 0

E 1.25-

.2>

Ix 1.20-

1.15 -

1.1(

Figure 7 (a)A142 asetoffo in theMal C, Xholl;

gel,in TB thepositic site.

facing expand

Inth of the scribed of the

Hinfl*.'

Références

Documents relatifs

Thereafter, Krivine also proves the converse, which is that every arithmetical for- mula which is true in the ground model is realized by the term which implements the trivial

Keywords: dependent types, sequent calculus, classical logic, control opera- tors, call-by-value, delimited continuations, continuation-passing style transla- tion, value

We will focus on the in fluence of the aspect ratio of NN NW’s and NN P’s piezoelectric particles as well as the dielec- tric permittivity of the matrix on the piezoelectric

In order to establish the time necessary to reach equilibrium between prion protein and nucleic acid interaction by the fluorescence properties of YOYO (20 nM, used this

‘unzipping’ of the originally bound protein; (2) that a model in which solution-phase proteins bind non-specifically to DNA can describe experiments on exchanges between the

Ils se modifient sans trop de mal pour des entiers de Z mais parfois en s’alourdissant un peu ; ainsi dans Z on ne peut plus affirmer l’existence d’un entier d unique tel que n

The idea of shuttling a small amount of liquid between different temperature zones by means of a pneumatic pump offers decisive advantages over methods published so far: (a) a

Les familles professionnelles ayant la plus grande proportion de salariés exposés. Proportion de salariés exposés Effectifs