Eur. J. Biochem. 263, 65±70 (1999) q FEBS 1999

Structural studies of the BstVI restriction±modification proteinsby fluorescence spectroscopy

Claudia Saavedra, Claudio Va squez and MarõÂa Victoria Encinas

Facultad de QuõÂmica y BiologõÂa, Universidad de Santiago de Chile

Structural studies of the proteins of the BstVI restriction±modification system of Bacillus stearothermophilus V

were carried out using intrinsic fluorescence techniques. The exposure and environments of their tryptophanyl

residues were determined using collisional quenchers. Quenching of BstVI endonuclease by iodide suggested a

heterogeneous class of tryptophan residues, while the results obtained with M.BstVI methylase were consistent

with a rather exposed tryptophan population. A comparison of the quenching efficiencies at 20 8C and 55 or

60 8C showed that their structures are more flexible and open at the temperature at which they exhibit maximal

activity. The endonuclease reached its active conformation only after 1 h of incubation at 60 8C. Fluorescence

changes were observed upon Mn2+ and Mg2+ binding, with Kd values in the range 3±5 mm. The binding of

S-adenosyl-l-methionine to the methylase produced conformational changes, which were consistent with binding

to a single site of Kd 550 and 680 mm at 20 8C and 55 8C, respectively. Quenching experiments with iodide

showed that the presence of S-adenosyl-l-methionine leads to different conformational states at 20 8C and 55 8C.

These results were interpreted in terms of differences in the structural characteristics of these restriction±

modification proteins as well as in terms of differences in the conformational states that these enzymes exhibit at

20 8C and at the temperature at which they are most active.

Keywords: BstVI; fluorescence spectroscopy; restriction±modification proteins.

For almost three decades, restriction and modification enzymeshave played an important role in the development of newmolecular and genetic techniques, which have helped in theunderstanding of gene structure and function. This is reflectedin the fact that over 2800 restriction enzymes have beendescribed to date [1]. Given that restriction endonucleases andDNA methyltransferases recognize and act on the same specificDNA sequence, they have been considered as valuablemolecular models for the study of sequence-specific protein±DNA interactions. However, the lack of sequence similarityamong endonucleases and methylases of a defined system hasbeen a puzzle to molecular biologists, and structural studies ofthese enzymes have been delayed as they have been mainlyused as molecular tools. In this context, they are usuallypurified only to an extent that ensures the exclusion ofundesired phosphatases and nonspecific nucleases. Later, thedevelopment of recombinant DNA techniques, which madepossible the molecular cloning of the corresponding structuralgenes, permitted the overexpression of these enzymes andhelped in their purification [2±4].

For some time, we have been interested in studying the BstVIrestriction±modification system, which is specified by thespore-forming thermotolerant rod Bacillus stearothermophilusV [5]. Both the BstVI endonuclease and the cognate M.BstVImethylase have been characterized [5,6] and their genes clonedinto Escherichia coli [7] and sequenced [8]. Studies on the

expression of the endonuclease and methylase in the homo-logous host or E. coli have also been performed [9,10].

In this work, we have used intrinsic tryptophan fluorescenceto obtain information on the conformation of the BstVI proteinsthemselves, as well as on the changes they undergo on inter-action with their substrates and/or cofactors.

E X P E R I M E N T A L P R O C E D U R E S

Materials

Acrylamide, NaI, S-adenosyl-l-methionine (AdoMet) andN-acetyltryptophanamide (NATA) were obtained from Sigma.All other reagents were of the purest commercially availablegrade.

Enzyme assays

BstVI restriction endonuclease and M.BstVI DNA methyltrans-ferase were purified to electrophoretic homogeneity fromappropriate E. coli recombinant clones [7] and their activitiesdetermined by electrophoresis on agarose gels, as previouslydescribed [5,6]. Protein concentrations were determined by themethod of Lowry et al. [11], using BSA as the standard.

Fluorescence measurements

All steady-state fluorescence measurements were performed ona Fluorolog Spex spectrofluorimeter. Bandwidths of 1.25 nmwere used for excitation and emission slits. Excitationwavelength was 290 nm. The buffer employed was 50 mmHepes (pH 7.5). Fluorescence quantum yields were determinedfrom the integrated area of the corrected fluorescence spectrum.NATA at 20 8C (fF = 0.15 [12]), was used as standard.

Correspondence to M. V. Encinas, Facultad de QuõÂmica y BiologõÂa,

Universidad de Santiago de Chile, Casilla 40, Correo 33, Santiago, Chile.

Abbreviations: NATA, N-acetyltryptophanamide; AdoMet, S-adenosyl-l-

methionine.

(Received 17 December 1998; revised 26 March 1999; accepted

30 March 1999)

66 C. Saavedra et al. (Eur. J. Biochem. 263) q FEBS 1999

Fluorescence quantum yields were independent of the excita-tion wavelength in the range 280±310 nm. Excitation atl , 280 nm was omitted in order to avoid excitation oftyrosine residues. The change in fluorescence intensity of theproteins was measured at both 55 and 60 8C as function of time,by adding 50 mL of protein solution (0.1 mg´mL21) to 1 mL ofthe buffer previously heated to the incubation temperature.Fluorescence intensity was continuously monitored at thewavelength of maximum emission.

Time-resolved fluorescence decay was recorded on a CD-900 Edinburgh time-correlated single-photon counting spectro-meter operating with a ns H2-filled flash lamp at 40 kHz pulsefrequency. Analysis of the fluorescence decays was carried outby a least-squares iterative convolution method based on theMarquardt algorithm, using the analysis routine provided byEdinburgh Instruments (Edinburgh, UK).

Fluorescence-quenching experiments with acrylamide andNaI were carried out by monitoring the change in intensity atthe emission maximum. Similar results were obtained when itwas calculated from the integrated emission band. Successivealiquots of freshly prepared solutions of the quencher wereadded to a cell containing the protein. Corrections for dilutionand inner-filter effect caused by acrylamide were made aspreviously described [13]. Quenching experiments at 55 and60 8C were performed after incubation of the proteins at thesetemperatures for 1 h. Fluorescence-quenching data wereanalysed according to the Stern±Volmer equation:

I8F/IF � 1� KSV�Q� �1�where I 8F is the fluorescence intensity in the absence ofquencher, IF is the fluorescence intensity at molar quencherconcentration [Q], and KSV is the Stern±Volmer quenchingconstant obtained from the slope of a plot of I 8F/IF versus [Q].KSV is related to the bimolecular rate constant (kq) by Eqn (2):

KSV � kq´t �2�where t is the fluorescence lifetime.

The effects of metal ions and AdoMet on the tryptophansinglet emission were determined from the change influorescence intensity. In experiments at high temperature, theligands were added to the respective protein solution previouslyincubated at 55 or 60 8C for 1 h. In the case of the AdoMetbinding, corrections for inner filter were made as previouslydescribed [13]. Binding data were analysed to yield the best fitby hyperbolic binding curve, and Kd values were estimated byusing the equation:

DIF/I8F � L/�L� Kd� �3�where L is the concentration of the added ligand.

R E S U L T S

Intrinsic tryptophan fluorescence of Bst VI proteins

Examination of the primary structure of M.BstVI methylase andBstVI endonuclease reveals that they contain nine and threetryptophan residues, respectively [8]. The fluorescence emis-sion spectra of these proteins after excitation at 285±295 nmare dominated by tryptophan emission. Despite these proteinscontaining a large number of tyrosine residues, no shoulder wasobserved in the 305±315 nm region. The emission spectra ofthe endonuclease at 20 8C showed a maximum at 335 nm,suggesting a hydrophobic environment for its tryptophan resi-dues. For the methylase, the maximum was shifted to 340 nm,suggesting a rather more polar average location of the

tryptophans. The decay profiles of the fluorescence in bothproteins were well fitted by two-exponential analysis. Thelifetime and the fractional contribution of each component at20 8C are shown in Table 1.

The position of the spectrum was independent of the tem-perature for both enzymes, but fluorescence intensity decreasedconsiderably when temperature was raised. A decrease in thefluorescence quantum yield with temperature has been reportedfor the indole chromophore as a consequence of temperature-dependence of nonradiative processes [12,14]. Absolutefluorescence quantum yields of the proteins evaluated at20 8C and 55±60 8C with respect to matched solutions ofNATA in the same buffer are shown in Table 2. The valuesobtained at temperatures at which these enzymes exhibited theirmaximal activity (60 8C for BstVI and 55 8C for M.BstVI) weretwofold lower. A similar decrease was found for the modelcompound.

Time-course of fluorescence intensity changes during theincubation of proteins at high temperature

The time-course of the fluorescence intensity changes of theendonuclease at 60 8C is shown in Fig. 1. A marked decrease inemission intensity was observed immediately after incubationof the protein at 60 8C, which was followed by a slowfluorescence decrease, reaching a constant value at approxi-mately 1 h. This result correlates well with the observed

Table 1. Fluorescence decay parameters at 208C. The buffer was 50 mm

Hepes, pH 7.5.

Protein t1 (ns) a1 t2 (ns) a2 k t la

Endonuclease 4�.64 0�.59 1�.62 0�.41 3�.4

Methylase 2�.41 0�.36 6�.87 0�.64 5�.3

NATA 3�.1

a Amplitude average lifetime.

Fig. 1. Changes in fluorescence intensity of BstVI endonuclease with

incubation time. The protein was added to the buffer previously warmed to

60 8C in a fluorimeter cell and the fluorescence intensity was monitored at

intervals of 5 s over 1 h. The excitation wavelength was 290 nm, and the

emission wavelength was 334 nm.

q FEBS 1999 BstVI restriction±modification proteins (Eur. J. Biochem. 263) 67

activation of BstVI endonuclease which had been preincubatedat 60 8C before being assayed under standard conditions(Fig. 2). This conformational change was irreversible, as theinitial conformation was not recovered when the proteinsolution was cooled to room temperature. Similar experimentswith NATA showed no time-dependence. In the case ofM.BstVI methylase, only a minor decrease in fluorescenceintensity was observed under similar conditions except thatpreincubation temperature was 55 8C. No activation of M.BstVIwas observed after preincubation at 55 8C for the same timeperiod.

Quenching of the intrinsic fluorescence of proteins

The accessibility of tryptophan residues in these thermophilicproteins to the solvent was probed by acrylamide- and NaI-quenching experiments at 20 and 55 or 60 8C. Acrylamide andiodide are diffusionally controlled quenchers of indolederivatives [15]. The fluorescence quenching of the indolechromophore varies with temperature as the result of variationsin both the diffusional rate and the excited singlet lifetime. Toexamine these variations, the effect of temperature on quench-ing experiments performed with the proteins was comparedwith the model compound NATA. Control experiments showedthat both enzymes preincubated with quenchers up to 0.1 m

were fully active, precluding the possibility of irreversiblestructural changes induced by these compounds.

Endonuclease and methylase fluorescence quenching byacrylamide, a polar uncharged water-soluble molecule that canpenetrate the matrix of a protein [16], gave linear Stern±Volmerplots (Fig. 3). The position of the fluorescence spectra wasfound to be unaltered by acrylamide in all cases. The values ofthe Stern±Volmer constants, KSV, are summarized in Table 2.These results showed that the decrease in the fluorescence ofBstVI endonuclease induced by acrylamide was higher at 60 8Cthan at 20 8C. Similar results were found for M.BstVI methyl-ase, for which KSV values were higher at 55 8C. KSV dataobtained with NATA however, showed similar values withinthis temperature range. Thus, the twofold increase in KSV

values observed when the temperature was raised is indicativeof a higher accessibility of acrylamide to the protein tryptophanpopulation.

On the other hand, methylase and endonuclease solutionswere incubated with increasing concentrations of iodide, a polaranion considered to have accessibility only to surface trypto-phans. The fluorescence intensity was markedly quenched byiodide (Fig. 4). As NaCl addition up to 0.25 m had a negligibleeffect on the emission of both proteins, we disregardedconformational changes due to variation in the ionic strengthduring ionic quenching experiments. The Stern±Volmer plots

Table 2. Fluorescence quantum yields and Stern-Volmer constants for

the quenching by acrylamide and iodide of BstVI endonuclease,

M.BstVI methylase and NATA.

TemperatureKSV (m21)

Protein (8C) fF Acrylamide NaI

Endonuclease 20 0�.041 5�.6 ±

60 0�.022 11�.3 ±

Methylase 20 0�.096 6�.0 6�.0

55 0�.047 10�.0 6�.0

NATA 20 0�.15a 24�.6 10�.3

55 0�.075 25�.4 7�.3

60 0�.063 25�.7 6�.2

a From [12].

Fig. 2. Activation of BstVI restriction endonuclease at 60 8C. Aliquots

containing 0.2, 0.1 or 0.05 units of purified BstVI enzyme were

preincubated at 20 8C (lanes a±c) or 60 8C (lanes f±h) for 1 h, chilled on

ice and then assayed for endonocleolytic activity for 10 min at 60 8C under

the standard conditions described earlier [5]. Lanes d and e contain negative

and positive controls of digestion, respectively. The substrate was the hybrid

plasmid pBR322-pTF62 [5].

Fig. 3. Stern±Volmer plots for the intrinsic fluorescence quenching of

BstVI proteins by acrylamide. (A) Effect of increasing concentrations of

acrylamide to a 0.8 mm BstVI endonuclease solution at 20 8C (W) and 60 8C

(X) and to NATA at 20 8C (K) and 60 8C (O). (B) The effect of acrylamide

on M.BstVI methylase (0.8 mm) at 20 8C (A) and 55 8C (B). All

experiments were carried out in 50 mm Hepes buffer, pH 7.5. The

excitation wavelength was 290 nm, and the emission wavelength was

335 nm.

68 C. Saavedra et al. (Eur. J. Biochem. 263) q FEBS 1999

shown in Fig. 4 showed the markedly different behaviour ofthese proteins. Plots at 20 8C and 60 8C for the endonucleaseare notably curved downwards and the emission spectra wereslightly blue-shifted. This behaviour is indicative of tryptophanheterogeneity when considering their accessibility to the anion.For the methylase, however, deviations from linearity werealmost negligible and the position of the spectrum showed noappreciable changes. These facts may result from mixedtryptophan populations, which do not differ significantly intheir individual susceptibility to the anionic quencher. KSV

values for methylase fluorescence quenching are included inTable 2. The different behaviour of these thermophilic proteins

towards iodide quenching at a given temperature suggests adifferent average location of the tryptophan population, theseresidues being more deeply buried in the protein matrix of theendonuclease than in the case of M.BstVI methylase.

The changes in quenching efficiency of the proteins withtemperature were opposite to that found for the modelcompound NATA. In this case, KSV values showed that iodidewas more effective as a quencher at lower temperatures,whereas KSV values for the quenching of methylase fluores-cence remained almost unchanged when the temperature wasincreased and only a slight increase was observed for theendonuclease. This behaviour towards iodide quenching isindicative of greater exposure of tryptophan residues to thesolvent at higher temperatures.

Fig. 4. Stern±Volmer plots for the fluorescence quenching of BstVI

proteins by iodide. (A) Effect of increasing concentration of iodide to

BstVI endonuclease (0.8 mm) at 20 8C (W) and 60 8C (X) and to NATA at

20 8C (K) and 60 8C (O). (B) The effect of iodide on M.BstVI methylase

(0.8 mm) at 20 8C (A) and 55 8C (B). All experiments were carried out in

50 mm Hepes buffer, pH 7.5. Excitation wavelength was 290 nm, and the

emission was measured at 335 nm or 340 nm.

Table 3. Binding constants for Mg2+ and Mn2+ to the BstVI endo-

nuclease and for AdoMet to M.BstVI methylase.

Protein Ligand Temperature (8C) Kd (mm)

Endonuclease Mg2+ 20 4.3 �^ 0.4

Mg2+ 60 4.7 �^ 0.7

Mn2+ 20 3.0 �^ 0.6

Mn2+ 60 4.1 �^ 0.6

Methylase AdoMet 20 640 �^ 60

AdoMet 55 510 �^ 45

Fig. 5. Changes in the intrinsic fluorescence of BstVI endonuclease as a

function of metal ion concentration. Effect of Mg2+ (B) and Mn2+ (X) at

60 8C. Enzyme concentration was 1 mm. The solid lines represent fitting of

the data to a hyperbolic function (Eqn 3).

Fig. 6. AdoMet binding as monitored by quenching of the intrinsic

fluorescence of M.BstVI methylase. Protein was incubated in 50 mm

Hepes at 20 8C (W) and 55 8C (X) and the fluorescence was recorded to the

maximum emission (340 nm). The curves have been fitted to Eqn (3).

q FEBS 1999 BstVI restriction±modification proteins (Eur. J. Biochem. 263) 69

Effect of bivalent metal ions on endonuclease fluorescence

The intrinsic fluorescence of BstVI endonuclease was modifiedon the addition of bivalent metal ions. At saturating concentra-tions of the activating metal ion Mg2+, fluorescence intensitydecreased 23% at both 20 8C and 60 8C, without any significantshift in the emission spectrum. The addition of Mn2+ to theprotein also resulted in quenching of the emission intensity. Amaximum decrease of 20% was reached on addition of 0.1 mmMn2+ (as MnCl2), indicating that this metal was also able tointeract with the enzyme. Support for this comes from the factthat Mn2+ can effectively replace Mg2+ in DNA cleavage(C. VaÂsquez, unpublished work).

The binding of these metal ions to the endonuclease wasanalysed by measuring the fluorescence changes that occur onbinding. Saturation curves were hyperbolic in all cases (Fig. 5).The dissociation constants (Kd) calculated from the hyperbolicfittings are shown in Table 3. These data can be interpreted as asimilar affinity of the endonuclease for these ions.

Binding of AdoMet to M.Bst VI methylase

Addition of AdoMet to the methylase at 20 8C and 55 8Cproduced a decrease in its intrinsic fluorescence, whichremained centred at 340 nm. The quenching of methylasetryptophan residues observed in the presence of AdoMet maybe a consequence of both quenching of tryptophan singletexcited state by the cofactor and/or a conformational change inthe protein caused by AdoMet binding. This would bring sometryptophan residues into a more highly quenching environment.To test the first possibility, we carried out quenching experi-ments with NATA. The addition of AdoMet up to 1.5 mm didnot have any significant effect on NATA emission. Therefore,the protein quenching by the methyl donor under the conditionsemployed here indicate a protein conformational change as aconsequence of AdoMet binding. The fluorescence decreaseversus AdoMet concentration could be fitted to a hyperbolicequation for a single binding site (Fig. 6). The calculated Kd

values are included in Table 3.We also assessed changes in the tryptophan microenviron-

ments as a result of AdoMet interaction, by the quenchingemission of the protein with NaI in the presence of AdoMet.Stern±Volmer plots (Fig. 7) showed very different behaviour at20 8C and 55 8C. At the higher temperature, the presence ofAdoMet had almost no effect on the quenching efficiency by

iodide, whereas at 20 8C there was a notable decrease in thequenching efficiency, indicating that AdoMet binding results insome tryptophan residues being shielded from the quencher.

D I S C U S S I O N

In this paper, we demonstrate that measurements of the intrinsicfluorescence of proteins from restriction±modification systemscan give useful information about their conformation and/ortheir interaction with specific ligands.

Most of the tryptophan residues of BstVI restriction endo-nuclease appear to be buried inside the native protein, asdeduced from the rather low emission maximum (335 nm).This behaviour holds even at 60 8C, the optimal temperature forBstVI activity [5]. In the case of M.BstVI, the average popu-lation of tryptophan residues appears to be located in morepolar environments.

Iodide was especially useful for differentiating the locationof tryptophan residues in these proteins. The downwardscurvature of the Stern±Volmer plots for endonuclease at both20 8C and 60 8C indicates that not all the tryptophan residuesare accessible to the aqueous phase in the endonuclease. Thisassumption is also supported by the blue shift of the emissionspectrum after iodide addition. However, the situation forM.BstVI methylase was quite different. Stern±Volmer plots foriodide quenching were linear up to high NaI concentrations andno blue shift of the emission spectrum was observed (Fig. 4).Thus, the position of the fluorescence spectra as well as thebehaviour towards the ionic quencher indicate notable differ-ences in the molecular environments of the tryptophan residuesin these proteins, those of the methylase being much moreexposed to the protein surface. This high exposure of trypto-phan residues to the solvent is somewhat surprising as thisprotein contains nine tryptophan residues, thus most of themshould be accessible to the solvent.

Although acrylamide is a diffusional quencher of indolechromophores in both polar and nonpolar solvents [15], it isless suitable for differentiating tryptophan residues topo-graphically. This uncharged molecule is able to penetrate theprotein and to quench residues located on the surface or insidethe apolar interior of a protein [16]. Bimolecular quenching rateconstants (at 20 8C) calculated using Eqn (2) and the amplitudeaverage lifetime are 1.7 � 109 and 1.1 � 109 m21´s21 for theendonuclease and methylase, respectively. These values repre-sent a high average accessibility of tryptophan residues toacrylamide [17]. If it is assumed that the quenching processoccurs by penetration of the probe into the protein matrix, theseresults indicate that the matrix of the endonuclease, even atroom temperature, is flexible enough to be penetrated byacrylamide, while a slightly lower flexibility can be deduced forthe methylase.

Stern±Volmer constant values at 20 8C and 55 8C (or 60 8C)for the quenching of NATA by acrylamide are similar. In thecase of BstVI proteins, however, the quenching efficiencyincreased with temperature. These results indicate a highdynamic penetration of the quencher into the protein matrix onan increase in temperature. This implies high fluidity and openstructure at the temperature at which these proteins are mostactive. As DNA-binding proteins, one would expect bothendonuclease and methylase to have a structure of sufficientflexibility to allow them to interact with the nucleic acid.

On the other hand, the thermophilic enzymes used in thisstudy were purified from recombinant E. coli clones [7]. Boththe endonuclease and DNA methyltransferase retained theirintrinsic thermostabilities when assayed in vitro. In vivo,

Fig. 7. Effect of AdoMet on the fluorescence quenching of M.BstVI

methylase by iodide. Stern±Volmer plots in the presence of 1 mm AdoMet

at 20 8C (A) and 55 8C (B).

70 C. Saavedra et al. (Eur. J. Biochem. 263) q FEBS 1999

however, the endonuclease, at least, exhibited poor restrictingactivity against incoming phage DNA in the heterologous hostE. coli [10]. This may be because a higher temperature isneeded to achieve its native conformation for optimal activity,as suggested by the results shown in Figs 1 and 2.

Changes in the intrinsic fluorescence of the endonucleaseinduced by the binding of bivalent ions indicate that metalbinding induced conformational changes that were sensed bythe tryptophan residues. It is interesting to note that bivalentmetal ions are essential for the activity of restriction endo-nucleases and nucleases in general. BstVI restriction enzymebound Mn2+ and Mg2+ with high affinity (Fig. 5), and this isprobably true for other bivalent metal ions as some activity isalso seen in the presence of Co2+ or Ni2+ (C. VaÂsquez, unpub-lished work). Kd values obtained at 20 8C and 60 8C showed asimilar affinity for these ions. These results, together with thesimilar quenching induced on protein fluorescence, suggest aunique binding site for both bivalent metal ions.

The decrease in fluorescence emission upon AdoMet bindingto M.BstVI methylase can be interpreted as the consequence ofligand-induced local protein structural changes. It is of interestto note that very different conformations are produced at 20 8Cand 55 8C, as evidenced by iodide quenching. Binding ofAdoMet at 55 8C produced a structural state in which thetryptophan populations did not differ with respect to theiraccessibility to iodide when compared with that exhibited in theabsence of the cofactor. At 20 8C, however, the presence ofAdoMet resulted in a significant decrease in the effectivequenching ability of iodide. This result suggests that thebinding of AdoMet induces relocation of tryptophan residues toa region of lower accessibility to the ionic quencher. Again, itshould be noted that this is not the conformation required foroptimal enzyme activity, as shown by the negligible activityobserved at 20 8C. It is tempting to speculate that theseAdoMet-induced conformational changes are responsible, atleast in part, for the low methylase activity observed at lowertemperatures.

A C K N O W L E D G E M E N T S

This work was supported by FONDECYT 1950440 and by DICYT. We are

grateful to Dr E. Cardemil for useful discussions and Dr S. Bertolotti for

helping us with the lifetime measurements.

R E F E R E N C E S

1. Roberts, R. & Macelis, D. (1998) Rebase-restriction enzymes and

methylases. Nucleic Acids Res. 26, 338±350.

2. Wilson, G. (1991) Organization of restriction-modification systems.

Nucleic Acids Res. 19, 2539±2566.

3. Wilson, G. & Murray, N. (1991) Restriction and modification systems.

Annu. Rev. Genet. 25, 585±627.

4. Wilson, G. (1993) Restriction enzymes: a brief overview. NEB

Transcript, New England Biolabs. 5, 1±5.

5. VaÂsquez, V. (1985) Isolation and partial characterization of BstVI, a

thermostable isoschizomer of XhoI. Biochem. Int. 10, 655±662.

6. Barra, R., Chiong, M., GonzaÂlez, E. & VaÂsquez, C. (1988) A DNA-

modification methylase from Bacillus stearothermophilus V. Bio-

chem. J. 255, 699±703.

7. VaÂsquez, C., Saavedra, C. & GonzaÂlez, E. (1991) Cloning the

BstVI restriction-modification system in Escherichia coli. Gene 102,

83±85.

8. GonzaÂlez, E. & VaÂsquez, C. (1993) Characterization of the BstVIRM

genes encoding the Bacillus stearothermophilus V restriction-

modification system. Gene 131, 103±106.

9. GonzaÂlez, E., Padilla, C., Saavedra, C. & VaÂsquez, C. (1994)

The expression of the BstVIM gene from Bacillus stearothermo-

philus V is restricted to vegetative cell growth. Microbiology 140,

1337±1340.

10. Saavedra, C., GonzaÂlez, E. & VaÂsquez, C. (1998) Studies on the

heterologous expression of BstVI restriction endonuclease in Escheri-

chia coli. Biochem. Mol. Biol. Int. 44, 391±397.

11. Lowry, O., Rosebrough, A., Farr, A. & Randall, J. (1951) Protein

measurement with the Folin phenol reagent. J. Biol. Chem. 193,

265±275.

12. Robbins, R.J., Fleming, G.R., Beddard, G.S., Robinson, G.W.,

Thistlethwaite, P.J. & Woolfe, G.J. (1980) Photophysics of aqueous

tryptophan: pH and temperature effects. J. Am. Chem. Soc. 102,

6271±6279.

13. Encinas, M.V., Rojas, M.C., Goldie, H. & Cardemil, E. (1993)

Comparative steady-state fluorescence studies of cytosolic rat liver

(GTP), Saccharomyces cerevisiae (ATP) and Escherichia coli (ATP)

phosphoenolpyruvate carboxykinases. Biochim. Biophys. Acta 1162,

195±202.

14. Smith, G.J. & Melhuish, W.H. (1991). Effect of the temperature and

viscosity on the excited singlet state of indoles in polar solvents.

J. Phys. Chem. 95, 4288±4291.

15. Lissi, E.A., Encinas, M.V., Bertolotti, S., Cosa, J.J. & Previtali, C.M.

(1990) Fluorescence quenching of indolic compounds in reverse

micelles of AOT. Photochem. Photobiol. 51, 53±58.

16. Krautwurst, H., Berti, M., Encinas, M.V., Marcus, F., Latshaw, S.P.,

Kemp, R.G., Frey, P.A. & Cardemil, E. (1995) Saccharomyces

cerevisiae phosphoenolpyruvate carboxykinase: revised amino

acid sequence, site-directed mutagenesis, and microenvironment

characteristics of cysteines 365 and 458. Biochemistry 34,

6382±6388.

17. Eftink, M.R. & Ghiron, C.A. (1981) Fluorescence quenching studies

with proteins. Anal. Biochem. 114, 199±227.