‘FHE JOURNAL OF BIOLOGICAL CHEMISTRY V c t l , 258, No. 22, Issue of November 25. pp. 1:1506-1.7512.1983 I’rrnted in I J . S A.

Isolation and Comparison of Two Molecular Species of the BAL 31 Nuclease from AZteromonas espejiana with Distinct Kinetic Properties*

(Received for publication, May 19, 1983)

Chik-Fong Wei, Gary A. Alianellt, Gerard H. Bencen, and Horace B. Gray, Jr. From the Departmtnt of Biochemical and Biophysical Sciences, University of Houston, Houston, Texas 77004

The extracellular nuclease from Alteromonas espe- jiana sp. BAL 31 can be isolated as two distinct pro- teins, the “fast” (F) and “slow” (S) species, both of which have been purified to homogeneity. The F and S species of the nuclease have molecular weights, respectively, of 109 X lo3 and 85 X lo3, and both are single poly- peptide chains with an isoelectric pH near 4.2. Both species catalyze the degradation of single-stranded and linear duplex DNAs to 5’-mononucleotides. The deg- radation of linear duplex DNA occurs through a ter- minally directed hydrolysis mechanism that results in the removal of nucleotides from both the 3‘ and 5’ ends. Apparent Michaelis constants (Km) have been obtained for the exonuclease activities of both species and for the activity against single-stranded DNA of the S species. The K , for the hydrolysis of single- stranded DNA catalyzed by the F species has not been obtained because the reaction velocity was maximal even at the lowest substrate concentrations accessible in the photometric assay. The ratio of the turnover numbers for the exonuclease activities of the two spe- cies indicates that the F species will shorten linear duplex DNA at a rate 27 f 5 (S.D.) times faster than an equimolar concentration of the S species in the limit of high substrate concentration, while the correspond- ing ratio for the activities against single-stranded DNA (1.2 f 0.1) shows that the two species are similar with respect to hydrolysis of this substrate. In the limit of high substrate concentrations, the F and S species break phosphodiester bonds in single-stranded DNA at rates 1.3 f 0.3 and 33 f 2 times those for the exonu- cleolytic degradation of linear duplex DNA, respec- tively. It has not been established whether the two species are physically related.

The culture supernatant of the marine bacterium Altero- monas espejiana sp. BAL 31 (American Type Culture Collec- tion 29659) contains nuclease activities that catalyze the extensive degradation of single-stranded DNA, the terminally directed hydrolysis of linear duplex DNA so as to result in the gradual shortening of such molecules without the intro- duction of scissions away from the termini, cleavage in re-

* This work was supported by Grant GM-21839 from the National Institute of General Medical Sciences and Grant CA-11761 from the National Cancer Institute. This is Paper V in the series “Extraceilular Nucleases of Alteromonas espejiana BAL 31.” Paper IV is Ref. 4. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

4 Present address, Beckman Instruments Inc., 8920 Route 108, Columbia, MD 21045.

sponse to the presence of single strand breaks and “hairpin” regions in duplex DNA, and the cleavage of negatively super- coiled DNA (form I DNA’) to yield linear duplex DNA (1). Closed circular DNA of near zero absolute superhelix density (form Io DNA) is virtually unaffected by the nuclease activities (1). Subsequent work with highly purified nuclease samples showed that a single such preparation contained the exonu- cleolytic activity against duplex DNA (2), the endonucleolytic activity against duplex DNA containing nicks (3), and the endonucleolytic activity resulting in the cleavage of negatively supercoiled DNA (4). In addition, this preparation was shown to catalyze cleavage in form I” DNA containing covalent lesions introduced by irradiation with ultraviolet light or by reaction with certain carcinogenic or mutagenic compounds under conditions for which nontreated form I” DNA was cleaved only to a very limited extent (3).

The purified preparation used in the above studies as well as crude preparations have been shown to be remarkably stable upon extended storage in the cold (5). In addition, the purified enzyme is highly resistant to inactivation in the presence of very high salt concentrations or in the presence of high concentrations of urea ( 5 ) , while crude preparations were shown in an earlier study to be resistant to detergent- mediated inactivation (1). The optimal pH is near 8.0 and 8.8 for duplex and single-stranded substrates, respectively (5). This is a distinct advantage over other enzymes specific for single-stranded DNA and for imperfectly base-paired or dis- torted structures in nominally duplex DNA, such as the SI and mung bean nucleases, which have acidic pH optima.

The BAL 31 nuclease is a more sensitive probe for super- coiling than other single strand-specific nucleases (4) and can be used to detect alterations in helix structure caused by reaction with a variety of chemical agents (3, 5, 6) . The exonucleolytic degradation of linear duplex DNA has proven to be the most widely exploited reaction catalyzed by the nuclease, having been used as an aid to the mapping of restriction enzyme-generated fragments of small genome DNAs (2) , for the production of deletion mutations a t specific sites in cloned DNA fragments (7) , for the removal of a portion of a plasmid-cloning vector so that inserted DNA sequences can abut a signal sequence, allowing secretion of the protein products of cloned genes (8,9), for the positioning of promoter sequences at desired sites in cloning vectors (101, for the specific deletion of DNA sequences between prese- lected bases (ll), and as an aid to the sequencing of large contiguous regions of DNA cloned into the coliphage M13 vector (12); this list is by no means exhaustive.

The abbreviations used are: form I DNA, covalently intact circular duplex DNA containing superhelical turns; form Io DNA, covalently closed circular duplex DNA containing few or zero superhelical turns.

13506

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Two Molecular Species of BAL 31 Nuclease 13507

The highly purified preparation used in some of the above work (2-5) displayed only two bands, with approximately 70% of the mass in one band, upon electrophoresis in polyacryl- amide gels in the presence of sodium dodecyl sulfate. The preparations used in other work (6-12) were either cruder preparations supplied from this laboratory or commercially available preparations, which also contain more than one protein species. Hence, the possibility remained that the various nuclease activities might not all be ascribable to a single protein species. More recently, it has become apparent that the kinetics of the exonucleolytic hydrolysis of linear duplex DNA a t a given number of units (1, 5, 13) of activity against single-stranded DNA can vary markedly from prepa- ration to preparation ( 5 ) . The present study describes the purification and some properties of two distinct molecular species of the BAL 31 nuclease from the culture supernatant obtained starting from a single clone of A. espejiana BAL 31.

MATERIALS AND METHODS

DNAs-Phage PM2 form I DNA was obtained using a modification (3) of Richardson’s (14) method. Lysates of $X174 am3 phage were obtained using the Escherichia coli C amber suppressor strain HF4714 (15) grown in Luria broth (16). The nonsuppressor strain HF4704 (thy-) (17) was infected using these lysates, and the phage was obtained as described from infected cells (18) and the cell supernatant (19) and pooled before banding in CsCl density gradients. Phage DNA was extracted as per Sinsheimer (20). Calf thymus DNA was obtained from Sigma.

Chromatographic Media-Agarose-hexane-adenosine 5’-mono- phosphate, in which 5’-AMP is coupled to agarose through a diami- nohexane linker attached to the C-8 position of the adenine moiety, was purchased from P-L Biochemicals. Sephacryl S-200, Sephadex G-100, and Sephadex G-100 superfine were products of Pharmacia Fine Chemicals.

Assay and Kinetics of BAL 31 Nuclease-All assays and kinetic determinations were carried out in 600 mM NaC1, 12.5 mM CaCl,, 12.5 mM MgCl,, 20 mM Tris. HCl, 1 mM EDTA (pH 8) at 30 “C. The photometric assay using denatured calf thymus DNA (5) and the unit of enzyme activity (1, 5, 13) have been described. The assay for exonucleolytic activity against linear duplex DNA of the fractions from chromatography on Sephadex G-100 superfine resin (see “Re- sults”) was carried out on 10-pl aliquots of the fractions at a fixed concentration (22 pg/ml) of PM2 form I DNA in total reaction volumes of 0.2 ml. The kinetics of exonucleolytic degradation of linear duplex DNA was determined using various concentrations of PM2 form I DNA and 98.0 and 76.5 units/ml of the “slow” (S) and “fast” (F) species of the nuclease, respectively, in 0.2-ml reaction mixtures. The form I DNA is quantitatively converted to the linear duplex form in a matter of seconds under these conditions, so that the kinetics are unaffected by starting with a supercoiled substrate. The hyper- chromicity associated with the reaction was monitored at 260 nm in a thermostatted cuvette (1-cm path length) in a Cary 118C spectro- photometer. The average hyperchromicity associated with complete hydrolysis (to constant AZm) was 60.0%, and the reciprocal extinction for the starting duplex DNA was taken to he 50 pg/ml-A::gm, which constants were used to calculate the reaction velocities in moles of nucleotide/liter-min from the initially linear recordings of AZM) versus time. Samples extensively but not completely degraded were subjected to analytical band sedimentation in alkaline solution as described (l), and the shortened molecules were observed to sediment as discrete bands, indicating as expected that internal scissions were not intro- duced by the nuclease.

A Beckman DU-7 spectrophotometer equipped with a kinetics unit was used for the photometric determination of reaction velocities of viral 4x174 am3 and denatured calf thymus DNAs using 0.4-ml reaction mixtures and 0.713 rt 0.005 and 0.60 0.03 units/ml of S and F species of the enzyme, respectively. A reciprocal extinction of 36 pg/ml-A:GP,“” has been reported for 4x174 viral DNA (20) and was used for the calf thymus DNA also; these DNAs displayed average hmerchromicities of 43.3 and 43.1%, respectively, upon quantitative hydrolysis.

Protein Determinations-The method of Bradford (21) was used with bovine serum albumin as the standard, employing the reagent kit for this assay supplied by Bio-Rad. This assay was used because

it is performed at acidic pH and is therefore not subject to interference from M$+, which is present in the nuclease preparations and may precipitate in the basic solutions used in other assays.

Gel Electrophoresis-Electrophoresis was done in tube gels (0.5 X 15 cm) containing 7.5% (w/v) polyacrylamide and in linear gradient slab gels (0.3 x 8 X 8 cm, 4-30% (w/v) polyacrylamide) (Pharmacia Fine Chemicals). The procedure for protein denaturation and the electrophoresis buffer were as described (22). Gradient gels were subjected to pre-electrophoresis to introduce the detergent-containing buffer and were otherwise used as suggested by the supplier. Gels were stained with Coomassie blue R-250 (23) and destained in a diffusion destainer (Bio-Rad).

Molecular Weight Determinations-A t,otal of nine standard pro- teins (Pharmacia Fine Chemicals) were subjected to electrophoresis in 4-30% polyacrylamide slab gels containing sodium dodecyl sulfate with some lanes of the gels containing both S and F nuclease species. The molecular weights used for the standards were as given (24). The molecular weights of the two species of the nuclease lay within the range comprised by the standard proteins. Plots of log(M,) uersus distance migrated showed excellent linearity (average correlation coefficient -0.998). Duplicate determinations were made for each species of the nuclease and agreed within 0.7%.

Isoelectric p H Determination-The isoelectric pH of each species of the enzyme was determined using a Servalyt Precote thin layer polyacrylamide gel (Serva) according to instructions from the sup- plier. The pH gradient was estimated using the protein standards of known isoelectric pH supplied by the manufacturer.

RESULTS

Purification of Two Species of the BAL 31 Nuclease-A. espejiana BAL 31 was subcultured starting from a single colony obtained from an AMs-agar plate ( 2 5 ) , and 1 liter of an overnight culture was used to inoculate 15 liters of modified (1) AMs-4 medium (26) which was aerated near 27 “C for 96 h using the apparatus described (5). The bacteria were filtered from the culture, and the culture supernatant was concen- trated to 600 ml using a Millipore Pellicon cassette ultrafil- tration apparatus as described in detail elsewhere ( 5 ) . The concentrated culture fluid (Fraction I) was further concen- trated by the addition of 1 volume of acetone which had been precooled to -20 “C ( 5 ) . After storage at -20 “C for 90 min, the precipitate was collected by centrifugation at 5,000-10,000 x g for approximately 10 min and resuspended in 85 ml of CAM buffer (1) (100 mM NaCl, 5 mM CaCl,, 5 mM MgC12, 20 mM Tris.HC1, 1 mM EDTA (pH 8)). The suspension was clarified by centrifugation at 100,000 X g for 90 min and the supernatant (Fraction 11, 80 ml) retained.

Affinity chromatography on agarose-AMP ( 5 ) was done on Fraction I1 using a column (1 X 13 cm) of this resin which had been washed with CAM buffer. After sample loading, the column was washed with 150 ml of CAM buffer modified by the increase of the NaCl concentration to 1 M. The nuclease activities were eluted from the column with the same buffer containing 20 mM 5’-AMP. The first 40 ml of the elution buffer contained all the nuclease activity eluted and were concentrated by dialysis against dry sucrose followed by di- alysis against CAM buffer (Fraction 111, 3 ml). Fraction I11 was adjusted to 5% (w/v) in sucrose and applied to a column (2.5 X 70 cm) of Sephadex G-100 superfine which had been equilibrated with CAM buffer. Elution of 2-ml fractions with CAM buffer was carried out at a flow rate of 8 ml/h.

TWO peaks of nuclease activity against both single-stranded and linear duplex substrates were eluted (Fig. 1) with that eluted first, characterized by a partition coefficient for the peak fraction near 0.21, corresponding to the F species. The peak corresponding to the S species had a partition coefficient of 0.34. Homogeneous F and S nucleases were obtained by pooling the fractions indicated in Fig. 1 (Fractions IV and V, respectively). The ratios of exonuclease reaction velocities against linear duplex DNA (moles of nucleotide/liter-min) to

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from

13508 Two Molecular Species of BAL 31 Nuclease

units of activity/liter on single-stranded DNA were 1.5 & 0.2 x lo-’‘’ and 1.28 f 0.08 x 10”’ mol/unit-min for the individ- ual fractions comprising Fractions IV and V, respectively. This demonstrates that the fractions pooled to yield homo- geneous nucleases are constant with regard to this ratio for each of the two species as must be the case if Fractions IV and V each contain only a single nuclease.

The peak corresponding to the S species had a shoulder on the low molecular weight side which appears to represent yet a third species of the nuclease, as the nuclease activity profile (at least against single-stranded DNA) displayed a shoulder paralleling the absorbance profile (Fig. 1). When fractions from the trailing portions of this peak were examined in denaturing gels, the band corresponding to the S species was trailed by a minor component, indicating that the shoulder contains an additional protein species.

A summary of the purification scheme is presented in Table I. The recovery, based on the activity against single-stranded DNA, after the acetone precipitation and agarose-AMP col- umn steps is quite good. The pooled fractions representing pure F and S species contained 44% of the units of nuclease activity of Fraction I; the total recovery is approximately 90% if the total activity represented in Fig. 1 is considered. Fig. 2 shows the results of gel electrophoresis of aliquots of Fractions IV and V under nondenaturing and denaturing conditions. It may be concluded that the nondenatured F and S nucleases are single polypeptide chains, since the single bands observed after electrophoresis under denaturing conditions rule out nonidentical subunits and any structures involving identical

subunits for either species would elute in the void volume of a Sephadex G-100 column.

Until the chromatography on Sephadex G-100 superfine resin, nuclease activities had always eluted as a single peak from all the types of columns used, including the ion exchange and gel filtration columns of the earlier work (1). The first evidence that distinct molecular species were responsible for the variance in kinetic properties noted from preparation to preparation was obtained when the exonuclease activities/ unit of activity against single-stranded DNA were assayed across the single peak of nuclease activity obtained from chromatography on Sephacryl S-200 and were observed to vary widely (data not shown). It should be noted that chro- matography on “regular” Sephadex G-100 (particle size 40- 120 pm) gave only partial resolution; the G-100 resin denoted superfine (particle size 20-50 pm) is needed to obtain homo- geneous F and S nucleases. Attempts to resolve the two species by eluting agarose-AMP columns with step gradients of 5’- AMP (0-24 mM in 4 mM increments) were not successful.

Physical Properties-The molecular weights, isoelectric pH, and extinction coefficients of both species of BAL 31 nuclease are presented in Table 11. The error estimates for the extinc- tion coefficients are based on the standard deviations obtained upon repeated protein concentration determinations. The ac- tual uncertainties may be significantly larger as the absorb- ances of these samples were rather small. These determina-

A B -

90 110 130 - v

0 W > -

Fraction number

FIG. 1. Elution profile of BAL 31 nuclease chromato- graphed on Sephadex G-100 superfine gel filtration resin. 0, lo2 X A:?:”; 0, lo-” X nuclease activity against denatured calf thymus DNA (units/ml); W, activity against linear duplex PM2 DNA (mi- cromoles of nucleotide released per liter-min). Bars represent frac- tions pooled to obtain Fractions IV and V of Table I. Conditions for chromatography are described under “Materials and Methods.”

F S F S FIG. 2. Gel electrophoretic patterns of the two species of

BAL 31 nuclease under nondenaturing (A) and denaturing (B) conditions. Direction of migration was from top to bottom. Conditions of electrophoresis are described under “Materials and Methods.”

TABLE I Purification of F and S species of BAL 31 nuelease

Step Protein Protein

mdml Crude concentrated supernatant (Fraction I ) 0.13 Acetone precipitation (Fraction 11) 0.66 Agarose-AMP chromatography (Fraction 111) 0.50 Sephadex G-100 chromatography (Fraction IV)b 0.036 SeDhadex G-100 chromatograDhv (Fraction V)‘j 0.094

X nu- clease activ-

itf

total mg unitslml 79.2 0.081 53.1 0.620

1.49 14.0 0.291 0.92 0.470 2.80

X nu- clease activ-

itf

total units 48.6 49.6 42.0

14.0 7.36

s p ~ ~ ~ ~ ~ - Yield Purification

unitslpg % -fold 0.61 100 1 0.93 102 1.5

28.1 86 46 25.3 15‘ 41 29.8 29‘ 49

Against denatured calf thymus DNA. *Fractions 106-112 (Fig. 1) pooled to yield pure F nuclease.

Based on pooled fractions only. Fractions 119-123 (Fig. 1) pooled to yield pure S nuclease.

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Two Molecular Species of BAL 31 Nuclease 13509

TABLE I1 Some Dhysical properties of F and S species of BAL 31 nuclease

x mo- lo6 x ab- Weight &sow. lecular z:z sorption coeffi- tion coefficient weight cient

liter/mol-cm dl/g-cm F species 109 4.2 1.10 * 0.09 10.2 * 0.8 S species 85.0 4.2 1.00 * 0.03 11.7 * 0.3

tions could also be subject to systematic error depending on the sensitivity of the Bradford (21) assay to differences in amino acid composition and the difference in the (as yet unknown) amino acid compositions of the BAL 31 nucleases and the bovine serum albumin standard. A highly purified sample of S nuclease did not show detectable staining in a carbohydrate assay under conditions for which pancreatic DNase I (4.6% carbohydrate by mass) was readily stained (5), indicating that carbohydrate moieties, if any, are not present in sufficient amounts to introduce errors into the molecular weight and protein concentration determinations for the S species. No nuclease activity was observed in crude prepara- tions to bind to columns of concanavalin A covalently coupled to agarose, which is consistent with an absence of carbohy- drate moieties in either species.

The isoelectric pH for both nucleases is quite low and strongly suggests a marked preponderance of acidic over basic amino acid side chains. The absorption spectra of the two species are indistinguishable and display no unusual features, showing a maximum a t 277.5 nm and a shoulder at 283.5 nm. The weight extinction coefficients are near the value expected for a protein of average aromatic amino acid content.

The question arises as to whether the two species of the nuclease are related polypeptides, perhaps bearing such a relationship that one can be derived from proteolytic cleavage of the other. Although this point requires further study, preliminary evidence is strongly in the affirmative. Digestion of the F nuclease with pronase (CAM buffer, 1 mg/ml, 37 'C, 60 min) resulted in a drastic loss of exonuclease activity on duplex DNA with a much lower percentage loss of the activity against single-stranded DNA and the concomitant migration of all the observable protein at a position indistinguishable from that of the S species in both nondenaturing and dena- turing polyacrylamide gels.' When corrected for the 45% loss of activity against single-stranded DNA, the calculated ratio of exonuclease reaction velocities using the kinetic parameters given below is 0.043 if it is assumed that all the nuclease activity remaining after proteolysis corresponds to S nuclease; the observed ratio is 0.050. Subtilisin, used in CAM buffer at the very low concentration used by Klenow et al. (27) to cleave Escherichia coli DNA polymerase I into a polymerase frag- ment and an exonuclease fragment, catalyzes the cleavage of the F species into several discrete fragments,* the smallest of which migrates in gels indistinguishably from S enzyme.

Kinetic Properties-Plots of reaction velocity versus sub- strate concentration (expressed in terms of the molar concen- tration of DNA nucleotide residues) for the hydrolysis of single-stranded DNAs by the two nucleases are presented in Fig. 3. For the F nuclease, the velocity was independent of substrat,e concentration within the scatter of the data even at the lowest DNA concentrations (approximately 2.3 pg/ml of viral 4x174 DNA) accessible in the photometric assay. Thus, the velocities from Fig. 3 (excluding the two determinations using calf thymus DNA) were averaged to obtain the maxi- mum velocity ( VmaX) which was used in subsequent calcula-

C.-F. Wei and H. B. Gray, Jr., unpublished work.

tions. For the S nuclease, the velocity did increase with substrate concentration but apparently also had reached its maximum value over much of the range of substrate concen- trations studied (Fig. 3). Values of V,,, were calculated from the average of the eight points from Fig. 3 corresponding to substrate concentrations above 40 pM (not including the data for calf thymus DNA) and also from a double reciprocal (Lineweaver-Burk) plot of the first seven points from Fig. 3. The values differed less than 3%, but the standard deviation of the value from the Lineweaver-Burk analysis was nearly six times that of the value from averaging the data where V,,, has apparently been achieved. The value of K, was also estimated for the S nuclease from the Lineweaver-Burk treat- ment, but has a rather large standard deviation (+23%) (Table 111) as expected both from the paucity of the data and from the fact that the substrate concentrations range from approx- imately K, = 3.5 to 8, well above the optimal range (28) for such determinations. No estimate of K, for the F enzyme can be made from the present data except to note that it would be below 5 pmol of nucleotides/liter. When denatured calf thymus DNA, highly diluted from the same stocks used to assay the nuclease, was used in these experiments, values of the reaction velocities were indistinguishable from those ob- served for 4x174 DNA (Fig. 3).

2.0 1

I I I I I I I /

0 40 80 120

Substrate concentration, pLM

FIG. 3. Reaction velocity as a function of substrate concen- trations for the hydrolysis of viral 41x174 DNA or denatured calf thymus DNA catalyzed by the two species of BAL 31 nuclease. 0, fast species, $X174 DNA substrate; ., fast species, calf thymus DNA substrate; 0, slow species, $X174 DNA substrate; X, slow species, calf thymus DNA substrate. Substrate concentration and velocity are expressed in terms of total molar nucleotide concen- tration and moles of nucleotide released per liter/min, respectively. Reaction conditions and details of the photometric assay are given under "Materials and Methods."

TABLE I11 Kinetic properties of F and S species of BAL 31 nuclease with respect . .

. to single-stranded DNA Maximum velocity/

activity* constant' unit,liter

X x Michaelis numberb number'

nmollunit- min p~ rnin" min" units/pmol

F species 7.1 f 0.3 7.7 * 0.2 2.54 -t 0.08 3.0 f 0.7 2.80 k 0.08 S species 8.3 f 0.9 8.3 * 0.7 2.7 0.21 3.0 k 0.2

"Determined from Lineweaver-Burk analysis of the first seven

Calculated from units/pg (Table I) and molecular weight (Table

Calculated from molar activity and maximum velocity from Fig.

points of Fig. 3.

11).

3.

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from

13510 Two Molecular Species of BAL 31 Nuclease

Kinetic properties of the two species of the nuclease with respect to hydrolysis of single-stranded DNA are listed in Table 111. The turnover number kp is defined by V,,, = k p [ E ] , where [ E ] , is the total nuclease concentration. It is noted that kp is approximately the same for each species of the enzyme whether it is calculated from the value of Vmax and the corresponding enzyme concentration from Fig. 3 or from the specific activity of the F and S enzymes from Table I. This indicates that the values of V,,, from the data of Fig. 3, where the substrate concentration does not exceed 40 pg/ ml, are compatible with those corresponding to the standard enzyme assay, where the substrate concentration is near 650 wg/ml. The data of Table I11 indicate that individual mole- cules of the two species are not well, if at all, distinguished by their ability to catalyze the hydrolysis of single-stranded DNA. The ratios of the turnover number of the F nuclease to that of the S enzyme range from 1.1 to 1.2 (kO.1) (depending on which pairs of values from Table 111 are used) so that the h, values could even be identical within experimental error.

The catalytic efficiency (kp/K, ) can be estimated only for the S enzyme activity on single-stranded DNA because K, for the F nuclease-catalyzed hydrolysis could not be deter- mined with the photometric assay (Fig. 3) and kp and K, for the exonuclease activity on duplex DNA are in different units (moles of nucleotide/liter-min and moles of termini/liter for V,,, and K,,,, respectively). The catalytic efficiency for the action of the S nuclease on single-stranded DNA is 4 k 1 x IO7 liters/mol-s which is comparable with some of the higher values reported for other enzymes in the literature (29).

Lineweaver-Burk plots representing the kinetics of the terminally directed hydrolysis of linear duplex PM2 DNA catalyzed by the F and S nucleases are presented in Fig. 4.

60 8orA

1 /Substrate concentration, FM”

FIG. 4. Lineweaver-Burk plots for the exonuclease activity of the two species of BAL 31 nuclease against linear duplex PM2 phage DNA. A, slow species; E , fast species. Reciprocal of substrate concentration is expressed in terms of (molar concentration of duplex termini)”, and reciprocal of velocity is expressed in terms of (moles of nucleotide released per liter/min)”. Reaction conditions and details of the photometric assay are given under “Materials and Methods.” Least squares analysis of the data to obtain values for K, and V,,, was made assuming a constant percentage error in substrate eoncentrations and weighting the reciprocals accordingly.

Here, velocities are expressed in moles of nucleotides/liter- min, while substrate concentrations are in moles of DNA termini/liter. Although the scatter in the two sets of data is similar, the standard deviations in K, and V,,, were percent- agewise significantly greater in the case of the F nuclease (Table IV). This is a reflection of the fact that the range of ratios of substrate concentrations to K,,, was somewhat below the optimal range (28) in the case of the F enzyme and closer to this range for the S nuclease.

The kinetic parameters for the exonucleolytic hydrolysis of linear duplex PM2 DNA are presented in Table IV. The unit of enzyme activity refers to that determined using single- stranded DNA as substrate. While 1 unit of BAL 31 exonu- clease activity on duplex DNA could be defined (and has been by one commercial supplier), it seems more practical to deter- mine nuclease activity using a single-stranded substrate, where it is clear that the reaction velocity is maximal down to much lower DNA concentrations than would be used in such assays. By contrast, rather high substrate concentrations (on a weight basis of DNA) are required to approach maximal velocities for the exonuclease activity; for DNA fragments containing 1000 base pairs, it can be calculated from the Michaelis-Menten equation and the values of K,,, (Table IV) that to reach 90% of the maximal reaction velocity would require concentrations of 280 and 120 pg/ml for the F and S nucleases, respectively, with larger fragments requiring pro- portionally larger weight concentrations. Where lower con- centrations are used, the resulting dependence of velocity upon the concentration of duplex ends introduces an addi- tional variable into the determinations.

In most investigations, the average number of base pairs removed per duplex terminus for a particular set of reaction conditions is the quantity of interest. This is obtained simply by dividing both sides of the Michaelis-Menten equation for a unisubstrate reaction by the substrate concentration

where u(, is the initial reaction velocity, [SI is the substrate concentration, [ I V l represents the concentration of released nucleotides, and t is time. Since u ~ , and Vmax are expressed in moles of nucleotide/unit volume/unit time if [SI and K, are in moles of termini/unit volume, Equation 1 gives the number of nucleotide residues released per DNA terminus/unit time. V,,, is obtained from the first column of Table IV multiplied by the number of units/liter. Because [SI remains constant until some of the molecules are completely degraded, Equation 1 is integrable with respect to t so that the total number of nucleotides removed per DNA terminus may be calculated for a given time of reaction. This is the basis on which Equation 1 of Legerski et al. (2) was obtained.

The velocities do decrease with time of reaction, however, presumably due to product inhibition by the released nucleo- tides, so that the use of Equation 1 will lead to low estimates of the extent of degradation when appreciable fractions of a linear duplex DNA are hydrolyzed. At the highest substrate concentration, the velocity had decreased to approximately

TABLE IV Kinetic properties of F and S species of BAL 31 nuclease with respect

to linear duplex DNA Maximum velocity/ Michaelis X turnover

unitfliter constant number

nmolfunit-min nM min-’ F species 2.3 & 0.3 96 & 15 6.2 * 1.0 S sDecies 0.093 & 0.005 41 * 3 0.235 & 0.015

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from

Two Molecular Species of BAL 31 Nuclease 13511

45% of u~, for the F nuclease by the time that one-half of the original DNA had been degraded. The inhibition correspond- ing to 50% degradation decreased with decreasing DNA con- centration. A similar magnitude of inhibition was observed for the S enzyme. In the concentration range of these studies (6 X 10-'-3 X M duplex termini) or below, it would appear that Equation 1 will provide an accurate estimate of the reaction velocity if the extent of degradation is limited to e.g. 20% or less of the duplex substrate. The constants in Equation 1 will also depend upon the mole fraction of guanine + cytosine of the DNA, leading to low estimates of uo for DNAs significantly higher in G + C content than PM2 DNA (42%).

The ratio of the turnover numbers for the F to the S nuclease from Table IV is 27 & 5, indicating that the molecules of the two species are very well distinguished by their relative capacities to shorten linear duplex DNA. This ratio corre- sponds to saturating substrate concentration; in the limit of low substrate concentration, where [SI is negligible with respect to K,, the corresponding ratio becomes k , F K , , S /

k,,SKm,F and has the value 11 f 3. The F species is thus a much more efficient exonuclease than the S enzyme at a given substrate concentration. The relative rates of internucleotide bond cleavage mediated by each species for single-stranded linear and duplex substrates may be calculated, in the limit of saturating substrate concentrations, from the turnover numbers in Tables I11 and IV. The F and S enzymes will cleave, respectively, 1.3 f 0.3 and 33 k 2 bonds in single- stranded DNA/bond cleaved in duplex DNA.

DISCUSSION

The present study demonstrates that the enzymes respon- sible for the BAL 31 nuclease activity in crude Alteromonas culture supernatants can comprise a t least two distinct pro- teins, each of which is capable in pure form of catalyzing a number of nucleolytic reactions. In addition to the activities against single-stranded, supercoiled, nicked circular duplex and linear duplex DNAs, aliquots of the F and S nucleases of this study mediate the cleavage of PM2 form I" DNA contain- ing a low average number (three or less) of apurinic sites/ molecule' as well as the breakage of form I' plasmid DNAs a t sites corresponding to the junctions between B-form and Z- form (left-handed) regions (30). The latter reaction uses as substrate plasmid DNAs produced by recombinant DNA tech- niques which contain tracts ofpoly(dG-dC) .poly(dG-dC) (31); these sequences, but not the rest of the molecule, adopt the Z-configuration in the presence of 4.5 M NaCl (32), the salt concentration a t which BAL 31 nuclease-catalyzed cleavage occurs. These DNAs are affected very little by the nuclease at salt concentrations insufficient to induce the B- to Z-form transition. Each pure species of the nuclease also has a ribonuclease activity.''

The endonuclease activity against single-stranded DNA, supercoiled DNA, B- to Z-DNA junctions in form Io DNA, and form I" DNAs containing a variety of covalent alterations can be regarded as a very general property of the BAL 31 nucleases, since this activity is elicited by a wide variety of imperfectly base-paired or covalently distorted structures in nominally duplex molecules. Such structures, in addition to the above, include those examined by Legerski et al. (3) using a highly purified sample of S nuclease (originally thought to be homogeneous as only a single band was observed upon electrophoresis in nondenaturing polyacrylamide gels) unless the rather unlikely assumption is made that the catalytic activity in that preparation, which contained only two protein

C:F. Wei, G. H. Bencen, and H. B. Gray, Jr., unpublished work.

species at detectable levels, was predominantly due to a spe- cies other than the BAL 31 S nuclease. Cruder preparations of the F nuclease, purified using only affinity chromatography on agarose-AMP, have been shown to cleave PM2 form Io DNA which had been reacted with a variety of methylating agents, ethylating agents, Ag' or Hg'+ (6), and it is likely, in the light of the above, that these represent still more types of altered duplex structures which can elicit cleavage by the BAL 31 nucleases.

The exonuclease activity against linear duplex DNA could be placed in the above category if it is regarded as arising from the relative thermodynamic instability at the ends of the duplex. Some evidence for a terminally directed activity mediated by such instability has been obtained in the case of the mung bean nuclease (33). However, the evidence in the case of the BAL 31 nuclease, at least for the s species, favors an exonucleolytic attack on one strand (e.g. the 3"terminated strand) at each end of the duplex, followed by removal of the protruding strand by either an endonucleolytic or exonucleo- lytic mechanism (5, 34). There is apparently a fraction of partially degraded duplexes which has fully base-paired ter- mini, as coliphage T4 DNA ligase-catalyzed joining of BAL 31 nuclease-shortened duplexes to linkers and other fully base-paired termini does occur (8, 9, 12). It has been shown, however, that the fraction of termini ligated by T4 DNA ligase in linear duplex DNA populations partially degraded by com- mercial samples of the nuclease increases dramatically when such samples are subjected to the action of the Klenow fragment of E. coli DNA polymerase I and/or T4 DNA polym- erase (35). This indicates the presence of protruding single strands in a significant fraction of the molecules which is consistent with the mode of degradation suggested above.

The degradation of single-stranded substrates, at least in the case of the S nuclease, apparently proceeds predominantly via an exonucleolytic mechanism, since the products of partial digests of single-stranded DNA are 5"mononucleotides and remaining high molecular weight DNA (5). Nonetheless, en- donucleolytic activity is clearly present against single- stranded substrates, as seen from the degradation of the single-stranded circular DNA of phage 6x174 by both pure F and S enzymes. This indicates a strong, but not absolute, preference for the exonucleolytic hydrolysis mode. The ques- tion as to whether the activity may be processive arises with respect to the degradation of both duplex and single-stranded substrates. A fully processive mechanism, in which a molecule of enzyme remains bound to a substrate molecule terminus until degradation is complete, is ruled out for the exonuclease activities on linear duplex DNA of both species and for the activity against single-stranded DNA of the S species by the fact that the velocity depends upon substrate concentration under conditions for which the molar concentration of sub- strate molecules is in large excess over that of enzyme. In the case of the activity of the F nuclease against single-stranded DNA, the velocity was independent of substrate concentration over the experimentally accessible concentration range (Fig. 31, leaving open the possibility of a fully processive mecha- nism. This was ruled out by gel electrophoretic examination of the products of very limited hydrolysis of 6x174 phage DNA under conditions where the ratio of the number of DNA molecules to the number of enzyme molecules was approxi- mately 10. No material migrated at the position of intact single-stranded circles under conditions for which there was material migrating as partially degraded linear single- stranded DNA. It should be noted, however, that a quasi- processive mechanism for the activities against linear duplex and single-stranded DNAs, in which more than one nucleotide

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from

13512 Two Molecular Species of BAL 31 Nuclease

may be removed per enzyme binding event, has not been eliminated as a possibility by any of these experiments.

The average values of V,,,/unit-liter and possibly also K,,, for the exonucleolytic activity against duplex DNA (Table IV) will be affected by the G + C content of the DNA. In very limited digests, the extent of the reaction will accordingly depend upon local sequence features (e.g. regions of consecu- tive dG.dC base pairs). This effect is significant; Legerski et al. (2) noted that degradation of Xb2irnrnZ1 DNA from the end of higher G + C content proceeded a t approximately 2/3 the rate for the other terminus with a highly purified sample of S nuclease. The two halves of this DNA differ in G + C content by approximately 10% (36). Direct sequencing of BAL 31 nuclease-shortened pBFC322 DNA has reveaIed a tendency for the activity to terminate in regions of several consecutive dG. dC base pairs (8,9). Poly(dG-dC) .poly(dG-dC) sequences are very slowly digested compared to natural DNA sequences. In the work demonstrating nuclease-catalyzed cleavage in B- to Z-DNA junction regions, only sequences whict terminated at both ends with (dG-dC),, where n has a maximum value of 16, remained essentially intact after sufficient incubation with the F enzyme to cleave an appreciable fraction of the mole- cules in the junction regions; those not so terminated were degraded extensively (30). The (dG-dC), segments were in the left-handed (Z-form) conformation in these experiments. Comparison of the rates of degradation of poly(dG-dC). poly(dG-dC) under moderate and high (Z-form-inducing) con- ditions of salt concentration showed, when the effects of very high salt concentration are taken into account, that the alter- nating copolymer is degraded more rapidly in the B-form but that this would still be slow compared to the rate of terminally directed hydrolysis of a naturally occurring DNA (30). Hence, the errors associated with the use of the kinetic constants of Table IV in Equation 1 or its integrated form could be substantial for experiments designed to remove a small num- ber of nucleotides, where local sequence features may predom- inate, or for more extensive degradations of DNAs differing in G + C content from that of PM2 phage DNA (42%) by more than e.g. 5-10%.

It was thought that the different kinetic species of the nuclease arose depending on the purification procedure used (5); several preparations starting with chromatography of concentrated culture supernatant on columns of DEAE-cel- lulose reproducibly gave rise to S enzyme, while the F nuclease character was observed only (but not always) when agarose- AMP chromatography was used. However, it is now clear that the affinity chromatography step can yield a mixture of the two species. The results may largely be due to varying relative amounts of the two species from one concentrated culture supernatant to the next, which could in turn be mediated by the action of proteases in crude preparations to convert F nuclease to S nuclease in accord with the preliminary findings noted above.

The F nuclease appears to be more efficacious as a probe for covalent alterations or unstacked or weakly hydrogen- bonded base pairs in duplex DNAs (5). In more recent work, this has been found to be the case for cleavage at apurinic sited and for hydrolysis in junctions between B-form and Z- form region^.^ The S species might be preferable in exonu- clease applications requiring the removal of a very limited number of nucleotide residuesfduplex terminus. I t is noted that commercial preparations are not fractionated so as to separate the two species and may consist of mixtures which will vary, especially with respect to the kinetics of the exo-

M. W. Kilpatrick, C.-F. Wei, H. B. Gray, Jr., and R. D. Wells, unpublished work.

nuclease activity against linear duplex DNA, from preparation to preparation. It thus appears prudent to characterize sam- ples from each commercial batch with respect to its activity on linear duplex substrates if these are used without further purification.

Acknowledgment-We thank Beckman Instruments for the use of the DU-7 spectrophotometer and kinetics accessory.

REFERENCES 1. Gray, H. B., Jr., Ostrander, D. A., Hodnett, J . L., Legerski, R. J.,

2. Legerski, R. J., Hodnett, J . L., and Gray, H. B., Jr. (1978) Nucleic

3. Legerski, R. J., Gray, H. B., Jr., and Robberson, D. L. (1977) J.

and Robberson, D. L. (1975) Nucleic Acids Res. 2, 1459-1492

Acids Res. 5, 1445-1464

Biol. Chem. 252,8740-8746 4.

5.

6.

7.

8.

9. 10.

11.

12.

13. 14. 15.

16. 17.

18.

19. 20. 21 22 23. 24. 25. 26.

27.

28.

29.

30.

31.

32.

33.

34. 35.

36.

Lau, P. P., and Gray, H. B., Jr. (1979) Nucleic Acids Res. 6, 331- 357

Gray, H. B., Jr., Winston, T. P., Hodnett, J. L., Legerski, R. J., Nees, D. W., Wei, C.-F., and Robberson, D. L. (1981) in Gene Amplification and Analysis (Chirikjian, J. G., and Papas, T. S., eds) Vol. 2, pp. 169-203, Elsevier/North-Holland, New York

Wei, C.-F., Legerski, R. J., Robberson, D. L., Alianell, G. A,, and Gray, H. B., Jr . (1983) in DNA Repair: A Laboratory Manual of Research Procedures (Friedberg, E. C., and Hanawalt, P. C., eds) Vol. 2, pp. 13-40, Marcel Dekker, Inc., New York

Myers, R. M., and Tjian, R. (1980) Proc. Natl. Acad. Sci. U. S. A. 77,6491-6495

Talmadge, K., Stahl, S., and Gilbert, W. (1980) Proc. Natl. Acud. Sci. U. S. A. 77, 3369-3372

Talmadge, K., and Gilbert, W. (1980) Gene (Amst.) 12, 235-241 Taniguchi, T., Guarente, L., Roberts, T. M., Kimelman, D.,

Douhan, J., 111, and Ptashne, M. (1980) Proc. Nutl. Acad. Sci.

Panayotatos, N., and Truong, K. (1981) Nucleic Acids Res. 9,

Poncz, M., Solowiejczyk, D., Ballantine, M., Schwartz, E., and Surrey, S. (1982) Proc. Natl. Acad. Sci. U. S. A. 79,4298-4302

Vogt, V. M. (1973) Eur. J . Biochem. 33, 192-200 Richardson, J. P. (1973) J. Mol. Biol. 78, 703-714 Benbow, R. M., Hutchinson, C. A,, 111, Fabricant, J. D., and

Sinsheimer, R. L. (1971) J. Virol. 7, 549-558 Luria, S. E., and Barrous, J. W. (1957) J. Bacteriol. 74, 461-476 Linqvist, B. H., and Sinsheimer, R. L. (1967) J. Mol. Biol. 28,

Dumas, L. B., Darby, G., and Sinsheimer, R. L. (1971) Biochim.

Truffaut, N., and Sinsheimer, R. L. (1974) J. Virol. 13,818-827 Sinsheimer, R. L. (1966) Proc. Nucleic Acids Res. 1, 569-576 Bradford, M. M. (1976) Anal. Biochem. 72, 248-254 Laemmli, U. K. (1970) Nature (Lond.) 227,680-685 Weber, K., and Osborn, M. (1969) J. Bioi. Chem. 244,4406-4412 Lambin, P. (1978) Anal. Biochem. 85, 114-125 Espejo, R. T., and Canelo, E. S. (1969) Virology 34, 738-747 Espejo, R. T., Canelo, E. S., and Sinsheimer, R. L. (1971) J. Mol.

Klenow, H., Overgaard-Hansen, K., and Patkar, S. A. (1971) Eur.

Segel, I. H. (1975) Enzyme Kinetics, pp. 46-48, Wiley-Intersci-

Fersht, A. (1977) Enzyme Structure and Mechanism, p. 132, W .

Kilpatrick, M. W., Wei, C.-F., Gray, H. B., Jr., and Wells, R. D.

Klysik, J., Stirdivant, S. M., and Wells, R. D. (1982) J . Biol.

Klysik, J., Stirdivant, S. M., Larson, J . E., Hart, P. A., and Wells,

Kroeker, W . D., Kowalski, D., and Laskowski, M., Sr. (1976)

Legerski, R. J. (1977) Ph.D. dissertation, University of Houston Shishido, K., and Ando, T. (1982) in Nucleases (Linn, S. M., and

Roberts, R. J., eds) pp. 155-185, Cold Spring Harbor Labora- tory, Cold Spring Harbor, NY

Davidson, N., and Szybalski, W. (1971) in The Bacteriophage Lambda (Hershey, A. D., ed) pp. 45-82, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

U. S. A . 77,5230-5233

5679-5688

87-94

Biophys. Acta 228,407-422

Biol. 56,597-621

J , Biochem. 22, 371-381

ence, New York

H. Freeman, San Francisco

(1983) Nucleic Acids Res. 11, 3811-3822

Chem. 257, 10152-10158

R. D. (1981) Nature (Lond.) 290,672-677

Biochemistry 15,4463-4467

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from

C F Wei, G A Alianell, G H Bencen and H B Gray, JrAlteromonas espejiana with distinct kinetic properties.

Isolation and comparison of two molecular species of the BAL 31 nuclease from

1983, 258:13506-13512.J. Biol. Chem. 

  http://www.jbc.org/content/258/22/13506Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/258/22/13506.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on April 9, 2018

http://ww

w.jbc.org/

Dow

nloaded from