THE JOURNAL. OF Brom~rca~ CHEM~STW Vol. 251. No. 7, Issue of April 10. pp. 2124-2132. 1976

Printed in U.S.A.

Purification and Physical Characterization of Nucleic Acid Helix-unwinding Proteins from Calf Thymus*

(Received for publication, July 21, 1975)

GLENN HERRICKS AND BRUCE ALBERTS§

From the Department of Biochemical Sciences, Princeton University, Princeton, New Jersey 08.540

We have devised a general protein fractionation procedure which selects for eukaryotic DNA-binding proteins, some of which resemble DNA-unwinding proteins from prokaryotes. Proteins were selected which (a) pass through a native DNA-cellulose column, (b) bind to a denatured DNA-cellulose column, and (c) remain bound to the latter column during a rinse with a dilute solution of the sodium salt of the polyanion dextran sulfate. When this fractionation was applied to the soluble proteins of calf thymus, three major protein species were recovered. The predominant one has an apparent molecular weight of about 24,000 in sodium dodecyl sulfate-polyacrylamide gel electrophoresis, is isoelectric near neutrality, and elutes as a monomer from denatured DNA-cellulose at moderate NaCl concentrations. This protein, designated calf-unwinding protein 1 (UPl), has been purified to homogeneity. However, isoelectric focus- ing reveals four or five subspecies (apparently separated by single-charge differences) which differ appreci- ably in their affinities for DNA.

Two other major proteins are obtained which have apparent molecular weights in sodium dodecyl sulfate of 33,000: the first, which elutes with low salt from DNA-cellulose as a homogeneous preparation, appears to be a basic protein (although it is clearly not a histone); the other, which elutes from DNA-cellulose as the major component of a “high salt eluting fraction,” is an acidic protein which co-purifies with less prominent species of higher molecular weights. Proteins similar to each of these three major calf thymus proteins have been observed by us and others in tissue culture cells of mouse, hamster, monkey, and humans, suggesting their wide occurrence among eukaryotes.

DNA-affinity chromatography provides a simple method for identifying and purifying cellular proteins which may be involved in DNA metabolism. In general, a DNA-free, soluble extract is passed through a column bed containing DNA immobilized on an inert solid support. After an extensive rinse, those proteins which have bound to the DNA on the column are eluted with high salt buffers and collected for further study. Small subclasses of such DNA-binding proteins may be

specially selected, either through the use of competing poly- mers, or by manipulation of ionic conditions and the composi- tion and strandedness of the affixed DNA (1).

DNA-cellulose columns have been used in several laborato- ries for isolation of mammalian DNA-binding proteins (l-16). In this work, we use this method to select a set of calf thymus proteins which pass through a native DNA-cellulose column, but bind to a similar column made with single-stranded DNA. To eliminate those proteins which might bind nonspecifically to any negatively charged polymer, we have chosen to study only those species which remain bound to the single-stranded DNA column during a rinse with the competing polyanion of sodium dextran sulfate.

* This research was supported by grants from the National Institutes of Health and the American Cancer Society.

$Predoctoral fellow of the National Science Foundation. Present address, Department of Molecular, Cellular and Developmental Biol- ogy, University of Colorado, Boulder, Colorado 80302

0 To whom reprint requests should be addressed.

When such a fractionation is applied to bacterial extracts,’ it selects for the known “DNA-unwinding proteins” from both T4 bacteriophage-infected (17) and uninfected Escherichia coli (18). ‘These proteins, which appear to function in DNA synthesis, share two important properties. They facilitate DNA denaturation by virtue of their strong, selective binding to single-stranded DNA, forming complexes in which the other- wise folded strands are extended and completely covered by protein. In addition, they stimulate the rate of in vitro DNA synthesis by a homologous DNA polymerase by 5- to lo-fold when tested with appropriate DNA templates (see also Refs. 19 and 20). The present work with calf thymus was undertaken to determine whether there might be DNA-binding proteins in mammalian cells which resemble these DNA-unwinding pro- teins from prokaryotes.

MATERIALS AND METHODS

Quantitation, Concentration, and Storage of Proteins-Proteins were monitored by their ultraviolet absorbance in a l-cm path length cell, and quantities are expressed as A zI)0 units, defined as the product of A,,, nm x volume in milliliters in clear solutions after correction for contaminating nucleic acids (21). Extinction coefficients were deter- mined on glycerol-free samples with the biuret reagent, comparing A,,,., with a standard curve generated with bovine serum albumin (E’” z80nm = 6.6). Samples to be saved were concentrated to at least 1 mg/ml by dialysis against solid sucrose (SchwarziMann Enzyme

’ D. Mace, N. Sigal, and G. Herrick, unpublished results.

2124

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Calf Helix-unwinding Proteins 2125

grade), dialyzed against 5 mu sodium phosphate, pH 7.8, plus 10% glycerol, clarified by high speed centrifugation (167,000 x g), and quick-frozen for storage in aliquots at -80’.

Isolation of Total Calf Thymus DNA-binding Proteins-Thymus glands from 4- to 6-week-old calves, held on ice up to 8 hours after slaughter, were trimmed free of fat and connective tissue and cubed. Pieces were frozen at -8O”, at which temperature they could be stored for at least 6 months without apparent damage to the proteins of interest. All isolation steps were performed at O-4”, with doubly distilled water used throughout.

In a standard preparation, frozen tissue (500 g) was extracted with 1100 ml of 20 mM Tris-HCl, pH 8.8 at 20”/50 mM NaCl/l mM Na,EDTA/O.l rnM dithiothreitolm% glycerol (v/v; Eastman Spectra grade) (Buffer A), plus an additional 0.9 mM dithiothreitol. Before the tissue thawed, it was reduced to an ice slurry in a Waring Blendor run at slow speed; the slurry was then homogenized at highest speed in three or four I-min bursts spaced by cooling periods (all nuclei appear to be broken by this treatment). The homogenate was centrifuged 30 min at 13,000 x g, and the large sediment was discarded. The supernatant (about 50 mg/ml of protein) was filtered through buffer- soaked cheesecloth and brought to 2 M NaCl by adding solid salt; the blender was then run at low speed to dissolve the salt. In a similar manner, the extract was made 10% (w/w) in PEG2 (Union Carbide Carbowax 6000). After standing 30 min, the PEG/2 M NaCl extract was centrifuged for 30 min at 13,000 x g to pellet any remaining DNA and particulate matter (1, 22). The clear red supernatant was subjected to a second filtration through cheesecloth and dialyzed in taut dialysis bags (1.1.inch diameter) against three lo-hour changes of 14 liters of Buffer A without glycerol. Mixing during dialysis was achieved by gentle bubbling with nitrogen; in addition, the bags were manually inverted at 2-hour intervals, except for the final 5 hours, when they were left undisturbed to permit lipids to coalesce. The impermeate is removed through a pinhole in the bottom of each bag, with the final lipid-containing volume being discarded. Glycerol (1/9 volume) was added and the extract was centrifuged a third time to remove precipitates formed during dialysis.

This extract (2.5 liters) was loaded at 200 to 300 ml/hour sequen- tially onto Buffer A-equilibrated columns of native DNA-cellulose (200 ml packed volume in a column bed (5 x 10 cm) containing 200 mg of DNA), and then denatured DNA-cellulose (400 ml packed volume in a column bed (5 x 20 cm) containing 300 mg of DNA). These columns are linked in series, so that only those proteins which fail to bind to native DNA-cellulose are exposed to the denatured DNA-cellulose column. Each column was automatically washed with Buffer A after the last of the extract passed through it, since a 1500-ml portion of Buffer A (containing 5% glycerol) was carefully layered above the denser extract (10% in glycerol) in a graduated cylinder which served as the intake reservoir (via a peristaltic pump). The DNA-celluloses were prepared as previously described (1); for convenience, commercial calf thymus DNA (Worthington Biochemical) and dry unwashed Whatman CFll cellulose were used.

After 1 liter of wash had been pumped through, the columns were uncoupled and separately rinsed with 1 column volume of Buffer A containing 0.50 mg/ml of sodium dextran sulfate 600 (Pharmacia). The proteins which resisted removal by sodium dextran sulfate were eluted from their respective columns with a subsequent rinse with Buffer A containing 2 M NaCl. After each use, columns were stored in this high salt buffer to retard growth of microorganisms, being rinsed with sev- eral column-volumes of low salt Buffer A just prior to reuse. The bind- ing capacities of these columns had been checked by several control experiments, which demonstrated that the yields we report are at least close to the maximum attainable by our methods, and, more impor- tantly, that no major species of DNA-binding protein has been missed because of the low DNA-cellulose to extract ratio used (23).

Gel Permeation Chromatography-Concentrated protein samples (6 ml or less in Buffer A) were clarified by centrifugation and applied to a 29 cm high, llO-ml bed-volume Bio-Gel A-0.5m agarose column (packed in Buffer A), run at 11 ml/hour. The column was calibrated for Stoke’s radius by fractionating a mixture of dextran blue, [“Clthymi- dine, beef-liver catalase, hemoglobin, and myoglobin.

Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis-Sam- ples were reduced by sucrose dialysis to 0.2 to 0.3 ml in 2 mM sodium phosphate, pH 6.8, plus 10% glycerol, and then adjusted to 1% sodium dodecyl sulfate (Pierce Chemical Co., Sequanal grade), 1% 2-mercap- toethanol, and 0.005% bromphenol blue. The mixture was then boiled

*The abbreviations used are: PEG, polyethylene glycol; UPl, unwinding protein 1.

for 1 min. Gels (10% acrylamide, 0.5 x 8 cm) were prepared and run as described by Shapiro et al. (24), with 0.27% ethylene diacrylate used as the cross-linking agent. Top and bottom electrode buffers were exchanged by continuous pumping to maintain them at a constant composition. Before staining with Coomassie brilliant blue R-250 (ICI America, Inc.), gels were marked by stabbing them with a needle dipped in India ink. Protein standards used for molecular weight calibration were: bovine serum albumin (M, = 67,000) (25), porcine pepsin (A4? = 32,700) (26), bovine trypsin (M, = 23,300) (27), sperm whale myoglobin (M, = 17,200) (28), and egg white lysozyme (M, = 14,300) (27).

Isoelectric Focusing in Urea-containing Polyacrylamide Gels- Protein samples, containing less than 5 rmol of salts, were photopo- lymerized into gels containing pH 3 to 10 ampholines, as described by Wrigley (29), but with 6 M urea added (Schwarz/Mann Ultrapure grade, freshly dissolved). During the initial 20 min of electrofc:using, ihe voltage was continuously raised from 200 V to 350 V, where it was maintained until the colored standards run in a parallel gel (cyto- chrome c, hemoglobin, methyl red or bromphenol blue, and xylene cyanole FF) were completely focused (1.5 to 2 hours at 20”). Gel tubes were held vertically with the cathode at the top. At the end of each run, pH gradients were estimated by briefly exposing the focused gels to a few drops of a wide-range, universal pH indicator solution; proteins were then stained with bromphenol blue (30).

RESULTS

In searching for possible mammalian analogues for the DNA-unwinding proteins found in bacteria, we take advantage of the fact that these DNA-binding proteins greatly destabilize regions of DNA double helix by virtue of their selective affinity for single-stranded over double-stranded DNA. Thus, under appropriate conditions, such proteins should pass through a native DNA-cellulose column while adhering strongly to a column containing single-stranded DNA. To analyze a hypo- thetical situation in quantitative terms, one estimates (from the relationship, l/R, = 1 + K [free DNA sites]) that a protein with an association constant (K) for native DNA I 10’ M- ’ and an association constant for denatured DNA 2 10’ Mm’ would “pass through” a standard native DNA-cellulose column during the first 8 liquid column volumes of rinse, but remain “bound” to a denatured DNA-cellulose column after 70 liquid column volumes of rinse.’ Thus, the protocol described under “Materials and Methods” should select DNA-unwinding pro- teins which have as little as a IO-fold differential affinity for single-stranded DNA. We estimate that such a protein would depress the melting temperature of DNA by less than lo.’ This procedure should of course also select DNA-unwinding proteins of much greater differential affinities, which are thus capable of denaturing DNA more strongly. Note that at a low enough ionic strength, a particular DNA-unwinding protein might

bind to double-stranded DNA with a K > lo8 Mm’, and thus be retained by such a native DNA column. However, the same

SUsing the notation of Martin and Synge (31), we set l/R, = A/AL.lIR and Kerr = (As/AL) a, where l/R, is the number of liquid column volumes of rinse necessary to bring the protein peak to the bottom of the column and K.,, = [bound protein]/[free protein]. Martin and Synge derive l/R = AL/A (1 + (As/A,) a) or l/R, = 1 + K.,, As the association constant K = K.,,/[free DNA sites], we obtain the equation in the text: l/R, = 1 + K[free DNA sites]. For the calculations we assume DNA-cellulose with 1 mg of DNA per packed ml and a relative liquid column volume AL/A = 0.5. We further as- sume that the DNA remains in large excess during the chromatogra- phy, and that a binding site can start at any DNA nucleotide (thus, [free DNA sites] = [total DNA nucleotides]).

’ For this calculation we arbitrarily choose a molar ratio of protein to DNA sites of 1 to 1, a DNA concentration of 6 pg/ml, a DNA site 10 nucleotides long, and an ionic strength such that the melting tempera- ture of DNA in the absence of protein is 50”. The equations used are likely to overestimate melting point depressions: for details, see Ref. 32.

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

2126 Calf Helix-unwinding Proteins

protein is likely to display exclusive affinity for a denatured DNA column at some higher ionic strength, where its general affinity for DNA is weakened.

We have chosen to screen for selective binding by mamma- lian proteins in a buffer containing 0.05 M NaCl, because the previously characterized Escherichia coli and T4 DNA- unwinding proteins bind selectively to denatured DNA col- umns at this salt concentration. We chose to fractionate the single-strand-specific proteins further on the basis of resistance to elution by a sodium dextran sulfate rinse, hoping to elimi- nate proteins that bind DNA nonspecifically. Hence, we began our studies by isolating the total single-strand-specific, sodium dextran sulfate-resistant DNA-binding proteins from a readily available mammalian source, calf thymus glands.

Elution profiles for a typical experiment are presented in Fig. 1. It can be seen that the denatured DNA-cellulose column bound about twice as much total protein as the native DNA-cellulose column which preceded it. Moreover, whereas sodium dextran sulfate removed essentially all of the proteins that bound to the native DNA-cellulose column (Fig. lA), about 15% of the proteins on the denatured DNA column were not removed until a subsequent 2 M NaCl elution (Fig. 1B). Thus, an apparently single-strand-specific, sodium dextran sulfate-resistant protein fraction constitutes about 10% of the total calf DNA-binding proteins recovered (0.67 x 15%), totaling 0.60 AgaO units/g of homogenate protein.

The different protein fractions eluting from both the native and denatured DNA columns in Fig. 1 have been analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and photographs of stained gels are presented in Fig. 2, A and

II f c $3 2 UN

IA

3.0 t

2 - 2.0M

i=r""-

20 1 j .3

1.0

IL

c i L

L -- 4

0 200

ELUTION VCi. (ml)

FIG. 1. Fractionation of the soluble calf thymus proteins that bind to native and denatured DNA-cellulose columns linked in series. Elution profiles after uncoupling the columns are shown; for details see under “Materials and Methods.” Elution volumes shown are those from the beginning of the sodium dextran sulfate (NaDS) rinse; the 2.0 M NaCl rinse was begun where indicated. A, elution of proteins bound to the native DNA column. The total amount of protein eluted was 1.9 A,,, units/g of homogenate protein (estimated by biuret assay at 15% of the net weight of tissue). B, elution from the denatured DNA column of proteins that had passed through the native DNA column. The total amount of protein eluted was 4.1 A,,, units/g of homogenate protein. The failure of sodium dextran sulfate to elute all the proteins from this column cannot be due simply to slow equilibration, since the same fractionation can be performed with sodium dextran sulfate present in the extract, as well as in the rinse, without a major effect on the composition or yield of sodium dextran sulfate-resistant proteins (the yield of UP1 is considerably lower, however, as expected from its partial sodium dextran sulfate sensitivity; see under “Discussion”).

FIG. 2. Sodium dedecyl sulfate-polyacrylamide gel electrophoretic analysis of fractions of calf thymus DNA-binding proteins from Fig. 1. Electrophoresis was performed as described under “Materials and Methods;” the volume of each aliquot was chosen to give an amount of protein sufficient for readily visible bands after Coomassie blue staining, and, therefore, a comparison of band intensities between gels is not meaningful. Migration was to the right; apparent molecular weights are designated below the major bands; Z’D indicates the tracking dye position. A and A, samples from native (A) and denatured (I?) DNA-cellulose elutions; gel numbers correspond to those fractions so marked in Fig. 1. C, band prolife of single-strand-specific, sodium dextran sulfate-resistant proteins. A gel similar to that shown below (D) was scanned with a Joyce-Loebl densitometer. D, comparison of 1 he single-strand-specific, sodium dextran sulfate-resistant proteins with calf thymus histones (Worthington Biochemical); histone bands are designated according to Ref. 33.

B. Note the very heterogeneous mixtures of proteins in the main peaks from both the native (Fig. 2A, Gels 1 to 4) and the denatured (Fig. 28, Gels 5 to 8) columns. In striking contrast, the sodium dextran sulfate-resistant fraction from the dena- tured column contains only a few prominent protein species (Fig. 2B, Gels 10 to 12). In general, this fraction contains protein bands corresponding to apparent molecular weights of about 24,000, 33,000, 40,000, 11,000, 17,000, and a triplet at

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Calf Helix-unwinding Proteins 2127

about 70,000, listed in approximate order of decreasing relative abundance. A densitometer tracing of such a stained gel pattern is presented in Fig. 2C. In Fig. 20, the migration of these proteins on sodium dodecyl sulfate-polyacrylamide gels is compared to that of calf thymus histones: as explained below, none of these proteins appear to be histones. When reapplied to native DNA-cellulose in 0.05 M NaCl, all seven of these proteins pass through the column without detectable affinity, whereas they rebind if reapplied to a denatured DNA column (see below).

As described under “Materials and Methods,” one of the fit steps in the preparation of the extract is a 13,000 x g centrifugation under low salt conditions, which is sufficient to remove both mitochondria (34) and chromatin (35). About 95% of the total DNA and 50% of the total protein are removed by this step. Yet this step does not lower the yield or change the composition of the single-strand-specific, sodium dextran sul- fate-resistant proteins obtained, as compared with alternative procedures in which the chromatin is thoroughly disrupted (either by high salt or by pancreatic DNase I digestion, see Ref. 1). This finding suggests that these proteins are not primarily chromosomal or mitochondrial in origin. From further experi- ments which utilized centrifugation conditions sufficient to sediment even small particulates (2 hours at 160,000 x g), we conclude that a substantial fraction of these sodium dextran sulfate-resistant proteins are present in the soluble cell fraction (23).

Separation and Properties of Three Major Single-strand- specific, Sodium Dertran Sulfate-resistant Fractions-After rebinding the single-strand-specific, sodium dextran sulfate- resistant fraction to a second denatured DNA column, three major proteins can be distinguished by elution with increasing concentrations of NaCl. Either a series of step elutions (Fig. 3) or a salt gradient (Fig. 5) may be used. In either case, a fairly clean separation of three major protein species may be ob- tained. As seen from the sodium dodecyl sulfate-polyacryla- mide gel patterns in Fig. 3B, a protein with an apparent molecular weight of 33,000 is eluted first, followed by the 24,ooO dalton species, which elutes over a wide range of salt, a small portion being eluted only with 2 M NaCl. A second, less pure 33,000 dalton species elutes only at the high NaCl concentration, along with several less prominent proteins of higher molecular weight. These three protein fractions will henceforth be designated as the “33,000 dalton; low salt elut- ing protein,” the “calf-unwinding protein 1” (UPl), and the “high salt eluting proteins,” respectively. In both 5% and 10% polyacrylamide sodium dodecyl sulfate gels, the two proteins with apparent molecular weights of 33,000 migrate as single bands near the position of porcine pepsin (M, = 32,700), while UP1 migrates in both gels near bovine trypsin (M, = 23,300).5

High Salt Eking Proteins-These p;oteins can be sepa- rated from UP1 either by NaCl-gradient elution from dena- tured DNA-cellulose (see Fig. 5, below) or by gel permeation chromatography; the latter is shown in Fig. 4. In Fractions 4 to 8, the 33,000 dalton, high salt eluting protein appears in association with smaller amounts of the 40,000 dalton protein and other higher molecular mass species, but some also appears to chromatograph separately (e.g. in Fraction 9). There is a suggestion that the 33,000 and 40,000 dalton species

61t should be recognized that molecular weights so obtained are estimates only. In fact, we have consistently observed UP1 runs at a position corresponding to a molecular weight of 18,000 to 19,000 in a Tris-buffered sodium dodecyl sulfate gel electrophoresis system (36).

A .mo A “i%l”

2.OM NaCl

-T

FRACTION NUMBER

B

FIG. 3. Further fractionation of single-strand-specific, sodium dex- tran sulfate-resistant proteins. A, step elution with NaCl from a denatured DNA-cellulose column. Proteins were isolated essentially as described under “Materials and Methods” and shown in Fig. 1. An aliquot containing 1.3 A,,, units of protein was dialyzed against Buffer A and applied to a l-ml packed volume, denatured DNA-cellulose column containing 1.1 mg of bound DNA; 0.4-ml fractions were collected every 15 min. The sodium chloride concentration was increased in the steps noted. B, sodium dodecyl sulfate-polyacrylamide gel analysis of fractions. Samples were prepared and gels displayed as described for Fig. 2. Because the salt concentration required to elute a given species is observed to decrease as the total protein/DNA ratio on the column increases, the position at which each protein elutes is very sensitive to the exact chromatographic conditions used. (However, the order of elution is not affected by the protein/DNA ratio or by pH changes between pH 7 and pH 9.) For this reason, gradient elutions (Fig. 5) were normally preferred. In columns run parallel to this one, it was found that either 1% Triton X-100 or 1 M urea may be added to all column buffers without affecting the elution patterns obtained. With some hatches of DNA-cellulose, loading of these proteins leads to removal of a small portion of DNA from the column, and sometimes trailing of 260 nm absorbance in the rinse is severe. Proteins begin to break through in the rinse when a level of about 1 A,., unit of protein/mg of column DNA is exceeded (Gel I).

may form a loose complex, since the two co-focus to pH 5.2 to 5.6 in a nondenaturing isoelectric focusing column (23) and co-elute from denatured DNA-cellulose in a salt gradient (see Fig. 5, below). After various attempts to purify the 33,000 dalton species away from the other high salt eluting proteins,

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

2128 Calf Helix-unwinding Proteins

FIG. 4. Fractionation of a single-strand-specific, sodium dextran sulfate-resistant protein fraction on the basis of molecular weight. A, gel permeation chromatography (Bio-Gel A-0.5m agarose gel). The starting sample (7.5 A,,, units in 6 ml of Buffer A) was similar in protein composition to that used in Fig. 3, and was prepared and fractionated as described under “Materials and Methods.” B, sodium dodecyl sulfate-polyacrylamide gel analysis of fractions. Samples were prepared and gels were displayed as in Fig. 2, but with apparent native molecular weights also listed (calculated from the elution volume of the respective column fraction, assuming spherical molecules, as described under “Materials and Methods”). The bottom gel shows the composition of the sample applied to the column.

our purest fractions contained 60 to 70% 33,000 dalton protein and had a relatively low extinction coefficient (Ej& “,,, = 4). In further work with these proteins (32, 39), samples have been used with a composition comparable to that of Gel 16, Fig. 6A below.

The 33,000 Dalton, Low Salt Eluting Protein-This protein elutes from DNA-cellulose >99% pure (as judged by sodium dodecyl sulfate-polyacrylamide gels). It constitutes up to 0.04% of the total cell protein; however, in some experiments (e.g. see Fig. 5 below) we have failed to obtain significant amounts of this protein. An amino acid analysis of this 33,009

TABLE I Amino acid compositions of two single-strand-specific, sodium dextran

sulfate-resistant proteins of calf thymus

The calf thymus protein samples and a control sample of egg white lysozyme (about 0.80 mg of each protein per analysis) were dialyzed exhaustively against 5 mM potassium phosphate, pH 7.4, hydrolyzed (105’ for 24 hours in 6 N HCI), and chromatographed by standard techniques (40). Peak areas were normalized for ninhydrin color yield measured for a standard amino acid mix and for equal recovery of internal standards, and results for the protein hydrolysates were corrected by subtracting corresponding values (primarily for ammonia) from a no-protein control. The molecular weight of UP1 was assumed to be 24,000. Tryptophan was determined as the product of the measured tyrosine value times the trypt.ophan/tyrosine ratio calcu- lated from the protein spectrum in alkali (41). Results obtained for a single hydrolysate are shown; the values obtained for the lyzozyme control were within 15% of established values (27), with the exception of cysteine (low by 29%) and tryptophan (high by 42%).

Residues per assumed monomer molecular weight

33,ooO daltm, low salt eluting protein UP1

Lysine Histidine Arginine Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Cysteine Valine Methionine Isoleucine Leucine Tyrosine Phenylalanine Tryptophan Ammonia, approximate A = Asp + Glu Net A = A - NHs, approxl-

mate B = Lys t Arg

25 18 8.3 8.4

13 16 30 20 13 13 20 16 27 30 11 8.0 25 18 20 12 6.0 1.6

28 16 11 5.1 19 8.0 31 9.5 7.4 4.7 8.2 11

11 1.5 26 15 57 50 31 35

38 34

dalton, low salt eluting protein is presented in Table I. Its aromatic amino acid composition is consistent with its extinc- tion coefficient, E$O,,, = 11. The content of charged amino acids suggests that it is a basic protein. From Fig. 20 it can be seen that calf thymus histone I migrates at the same position as this protein on 10% sodium dodecyl sulfate-polyacrylamide gels (much slower than predicted from its molecular weight, see Ref. 42). However, histone I has a clearly different amino acid content: it has a ratio of basic to acidic amino acids of 4.6 (as compared with 0.67 found for this protein) and a much lower extinction coefficient (43).

WV-This protein also elutes from DNA-cellulose as a well separated peak. It is obtained reproducibly, constituting at least 0.1% of the total cell protein. In the nondenaturing conditions of Fig. 4, it elutes from the gel permeation column with an apparent native molecular weight of 19,000 as judged by its elution relative to standard proteins (see under “Mate- rials and Methods”). Comparing this value to the sodium dodecyl sulfate gel-derived molecular weight of 24,000, we conclude that the presumed native protein is a roughly

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Calf Helix-unwinding Proteins 2129

symmetric monomer. Amino acid analysis (Table I) of a pure UP1 sample (see Fig. 7, below) shows a paucity of tyrosine, consistent with the protein’s relatively low extinction coeffi- cient, Es.,,,,,, = 4.45. From its composition, one would predict that it should have a near neutral isoelectric point. This is the case, both in sucrose gradient-stabilized isoelectric focusing in nondenaturing conditions (p1 = 7.8, Ref. 23) and in polyacryl- amide gel isoelectric focusing in urea (see below).

Identification of UP1 Heterogeneity and Purification of Subfractions-Various initial efforts to purify UP1 by standard chromatographic techniques were frustrated by its heterogene- ous behavior. For example, in the experiment shown in Fig. 5, single-strand-specific, sodium dextran sulfateresistant pro- teins were fractionated by rebinding them to a small denatured DNA-cellulose column and eluting them with a linear gradient of NaCl. From sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the fractions (Fig. 6A, also see Fig. 3), it is clear that UP1 elutes from DNA over a broad range of salt. Isoelectric focusing in urea-containing, polyacrylamide gels (Fig. 6B) reveals that this broadness reflects an isoelectric point heterogeneity: UP1 appears to be composed of four or five subspecies with isoelectric points near neutrality, spaced roughly equally over a range of 0.5 to 1.0 pH unit. These subspecies elute from the column in the order of their relative basicities, the three most acidic eluting close to one another at a low NaCl concentration, and the most basic eluting consider- ably later.

In order to remove contaminants of different molecular weights from UPl, selected fractions from the salt gradient elution in Fig. 5 were pooled into three samples (called

“Acidics,” “Mix,” and “Basics”) and subjected to gel permea- tion chromatography on agarose as described in Fig. 7. The peak fractions from this column were dialyzed and concen- trated for storage at -80”. As judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the resulting sam-

FRACTION NUMBER

FIG. 5. Preparative fractionation of the single-strand-specific, dex- tran sulfate-resistant proteins by NaCl gradient elution from dena- tured DNA-cellulose. Such proteins, isolated from 509 g of calf thymus and dialyzed against Buffer A, were clarified by a 30-min centrifuga- tion at 100,000 x g; 97% of the original protein was recovered in the supernatant and applied to a small denatured DNA-cellulose column at 60 ml/hour. The column was rinsed with Buffer A until absorbance reached background; gradient elution (0.05 to 1.0 M NaCl) was then initiated (Fraction 12), and at the end a 2 M NaCl step was employed (Fraction 45). The column had a packed volume of 60 ml and contained 100 mg of denatured DNA; 17-ml fractions were collected every 16 min. Sodium chloride concentrations were determined by conductivity. About 75% of the applied protein was recovered by the elution; of this, about 66% was estimated to be UPl.

? t t

33,000 24,000 TD eimc

FIG. 6. Molecular weight and charge characterization of the protein for the sodium dodecyl sulfate-polyacrylamide gel electrophoresis and species eluted from the denatured DNA-cellulose column shown in Fig. two-thirds for electrofocusing gels. A, sodium dodecyl sulfate-polya- 5. The gel numbers listed correspond to the fractions designated in Fig. crylamide gel electrophoresis; apparent molecular weights are indi- 5, with the gels at the bottom representing the original material cated. B, Isoelectric focusing of proteins in polyacrylamide gels applied to the column. For each pair of gels, a sample volume containing urea, pH 3 to 10 gradient. Gels are displayed with the corresponding to 0.045 A,,, units of protein was dialyzed against 5 rnH cathode (basic end) to the left; similar bands are aligned and pH 7 is Tris-HCl, pH 8.8/10% glycerol; it was then split, with one-third used marked with an arrow.

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

2130 Calf Helix-unwinding Proteins

ELUTION VOL. (ml)

4 4 . 24,000 TD

C ACIDICS MIX

i BASICS

BASIC 4 PH?

ACIDIC

FIG. 7. Final purification of UP1 samples with different isoelectric compositions. A, gel permeation chromatography (Bio-Gel A-0.5m agarose gel). Fractions from the gradient-eluted denatured DNA column shown in Fig. 5 were pooled and named as follows: Acidics, Fractions 3 and 4; Mix, Fractions 6, 7, and 8; Basics, Fractions 9, 10, and 11. The pooled materials were concentrated and dialyzed with about 60% recovery and were applied to the same column at about 125 ml apart (see under “Materials and Methods”). Elution volumes are given from the beginning of the sample applications. Peak fractions eluted from the gel permeation column were pooled (more specifically, the top 3, 4, and 2 such fractions from the Acidics, Mix, and Basics, respectively). The resulting three samples were each concentrated, dialyzed, and stored at -60’ as described under “Materials and

ples appeared >99% pure (Fig. 7B), and electrofocusing showed that they had retained their original isoelectric point heterogeneity (Fig. 7C).

Nature of UP1 Heterogeneity-Due to the neutral isoelectric point of UP1 and its content of groups which buffer in this pH range (histidines and the a amino group, Table I), a unit charge change caused by the appearance or disappearance of a strongly basic or acidic group should cause only a minor change in the isoelectric point. Calculations of the expected incre- ments in fact suggest that each isoelectric focusing subspecies differs from its neighbor by a single unit charge (23). Various experiments, including chemical analyses, argue that the charge differences between subspecies are not caused by sulfur damage, deamination, nucleic acid contamination, or by heterogeneity of carbohydrate or phosphate content; i.e. none of the normally considered in uiuo or in vitro modifications of aminoacyl side groups seems to be responsible for UP1 hetero- geneity. These results, in conjunction with tryptic peptide mapping (23), point to two possible explanations for the heterogeneity observed: (a) the subspecies are coded for by different but similar genes (controls show that one lobe of a single thymus gland contains the usual mixture of subspecies (23)), or (b) the subspecies arise as the result of limited exopeptidase action on a single gene product. In the latter context, it is interesting to note that a small difference in apparent molecular weight between UP1 subspecies is ob- served, both in sodium dodecyl sulfate-polyacrylamide electro- phoresis (23) and in gel permeation chromatography (see legend to Fig. 7); in both cases, the most basic subspecies behaves as much as 3,999 daltons larger than the most acidic. While such a molecular weight heterogeneity is specifically predicted by the second model suggested, it is of course also fully compatible with the alternative view that these differ- ences originate genetically.

DISCUSSION

In this communication we have described three calf thymus single-strand-specific, sodium dextran sulfate-resistant DNA- binding protein fractions, two of which are single proteins readily purified free of other species. The first, the 33,999 dalton, low salt eluting protein, is a basic protein present in an estimated 209,099 copies per cel1.O The second and most prominent, UPl, is isolated as a monomer of 19,099 to 24,999 daltons, has a neutral isoelectric point and is present in at least 809,999 copies per cell. The third fraction consists of several polypeptides, the most prominent weighing 33,999; the latter has an acidic isoelectric point (Fig. 6) and is present in about 50,999 copies per cell. It has not been obtained pure and may participate in a loose complex with other polypeptides in the high salt eluting fraction.

The DNA-binding of these three fractions is Nadl-sensitive, presumably reflecting the interaction of these proteins with the polyphosphate DNA. That the binding is resistant to sodium dextran sulfate suggests that sodium dextran sulfate cannot

OBased on 4.6 x lo* cells/g of thymus (44) and 0.15 g of total cell protein/g of thymus (see legend to Fig. 1).

Methods.” Final yields of UP1 ranged from 70 to 85% of the total materials applied to the column. B, sodium dotlecyl sulfate-polyacryla- mide gels. Aliquots containing 0.002,0.010, and 0.025 A,.. units of each set of subspecies were applied, and gels are displayed as in Fig. 2A. The Acidics, Mix, and Basics fractions show apparent molecular weights of 23,700, 24,200, and 24,800, respectively. C, Urea polyacrylamide isoelectric focusing gels, pH 3 to 10. Aliquots representing 0.020 A,,, units were applied, and gels are displayed as in Fig. 6B.

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Calf Helix-unwinding Proteins 2131

mimic the particular polyphosphate conformations of single- stranded DNA to which these proteins bind. However, we have found (23) that UPl, but not the other two fractions, can be completely eluted from single-stranded DNA with more con- centrated sodium dextran sulfate (8.75 mg/ml of sodium dextran sulfate, total [Na+ ] = 0.05 M). Thus, while UP1 has been isolated on the basis of its sodium dextran sulfate-resist- ant binding, it is only partially resistant. It is interesting to note that, like the calf UPl, both T4 bacteriophage gene 32 protein and the E. coli DNA-unwinding protein are only partially sodium dextran sulfate-resistant.’

The major portion of the mammalian single-strand-specific, sodium dextran sulfate-resistant proteins appears not to be chromatin-associated, as judged by our cell fractionation studies. The possibility that the binding of these proteins to DNA is an artifact of in vitro conditions seems unlikely in view of the strong specificity of their binding (i.e. to single-stranded DNA, but not to double-stranded DNA or sodium dextran sulfate). However, while they may slightly prefer DNA, both UP1 and the high salt eluting proteins also have a substantial affinity for single-stranded RNA (32). In view of the possible excess of single-stranded RNA over single-stranded DNA in viuo, much of this protein could be bound to RNA at any instant.

If these proteins have DNA, rather than RNA, as their primary site of action, their apparent non-chromatin location might be an artifact caused by cell disruption, or the cell might contain a pool of proteins that bind to DNA only for a short time during the cell cycle, e.g. during the S period. The high molecular weight eukaryotic DNA polymerase, which presum- ably binds to nuclear DNA at some point and may in fact be in the nucleus in vivo,8 is likewise found in the non-chromatin fraction (45); in an accompanying paper we show that high

molecular weight DNA polymerase is stimulated in vitro both by UP1 and by the high salt eluting proteins (39).

Calf UP1 shows a reproducible isoelectric point heterogene- ity, which, from available data, could arise either genetically or from limited exopeptidase action. With regard to its possible physiological significance, it is important to note that this small charge difference between subspecies has a marked effect on DNA-affinity (Figs. 5 and 6). Thus, different subspecies could have related but different functions; alternatively, only the most basic subspecies might be functional, the less basic subspecies being inactive “turnover” products. However, the heterogeneity might also reflect heterogeneity in the thymus cell population, the state of the protein being a differentiated characteristic. None of these possibilities can be eliminated a priori, and further study of this protein, particularly in synchronized tissue culture cells, will be necessary to evaluate them adequately.

As judged by a variety of criteria (electrophoretic mobility in sodium dodecyl sulfate-polyacrylamide gels, salt elution from DNA-cellulose, insensitivity to sodium dextran sulfate elution, and specificity for single-stranded DNA), we have isolated proteins corresponding to the calf UP1 and to the calf 33,000 dalton, low salt eluting protein from extracts of cultured mouse cells (3T3 and I-10 cells; Ref. 1 and Footnote 9) and cultured monkey cells (primary African green monkey kidney and CV-1 cells). lo Moreover, a large scale extraction of 0.5 g of mouse I-10

’ G. Herrick and B. Alberts, unpublished results. BHerrick, G., Spear, B. B., and Veomett, G. (1976) Proc. N&l. Acad.

Sci. U. S. A., in press. 9 G. Herrick, unpublished results.

‘” G. Herrick and A. J. Levine, unpublished results,

cells (Leydig cell tumor; Ref. 46) yielded a single-strand-spe- cific, sodium dextran sulfate-resistant protein fraction compa- rable in quantity and composition to the typical calf thymus preparation. Thus, these proteins which we have characterized from calf thymus may be present in all mammalian cells.

Tsai and Green (47) have identified in both mouse and

human cultured fibroblasts a cytoplasmic protein, designated P8, which they purified by taking advantage of its selective affinity for single-stranded DNA. P8 is present in very large amounts, consisting of 3% of the soluble cell protein in the human cells used. This protein migrates slightly slower than our 33,000 dalton proteins on a sodium dodecyl sulfate-polya- crylamide gel. ‘I Stein (14) finds a single-strand-specific DNA- binding protein, which she calls P8, in elevated levels in senescent lines of WI38 human cells; it migrates together with the P8 of Tsai and Green on sodium dodecyl sulfate electrophoresis.‘* She also sees a 25,000 dalton, single-strand- specific DNA binding protein which presumably corresponds to the calf UPl. Fox and Pardee (4) have demonstrated two prominent single-strand-specific, DNA-binding proteins in hamster cell cultures; the one that elutes from DNA-cellulose at 0.15 M NaCl has subsequently been found to co-migrate on sodium dodecyl sulfate-polyacrylamide gels with the calf thymus 33,000 dalton proteins; the other, which eluted at 0.60 M NaCl, probably corresponds in molecular weight to the calf UPl.” Both hamster proteins appear to be synthesized throughout the cell cycle (4). A low salt eluting, 33,000 dalton, single-stranded DNA-binding protein has been demonstrated in the cytoplasm of human cells (15). These various 33,000

dalton DNA-binding proteins may correspond to a rat liver protein (30,000 daltons, pI = 9.6) which stimulates RNA polymerase B (48).

Hotta and Stern (2, 49) have described a presumably different single-strand-specific DNA-binding protein which appears in greatly elevated amounts during meiotic prophase in germ-line tissues of both mammals and plants. Banks and Spanos (50) have isolated a single-strand-specific DNA-bind- ing protein from the basidiomycete fungus Ustilugo maydis; it behaves in many respects like the prokaryotic DNA-unwinding proteins.

In the following two articles (32, 39) we further characterize the calf single-strand-specific, sodium dextran sulfate-resist- ant DNA-binding proteins, with the special view of comparing them with the prokaryotic DNA-unwinding proteins.

Acknowledgments-We would like to thank Dr. Bruce Cun- ningham (Rockefeller University) for instructing one of us (G. H.) in the techniques of peptide mapping, Barbara Bam- man and the Whitehall Foundation of Princeton University for help with the amino acid analyses, and Drs. R. Karpel and J. Fresco for helpful comments on this manuscript.

REFERENCES

1. Alberts, B. & Herrick. G. (1971) Methods Enzymol. 21D, 198-217 2. Hotta, Y. & Stem, H. (1971) Nature New Biol. 234, 83-86 3. Salas, J. & Green, H. (1971) Nature New Biol. 229, 165-169 4. Fox, T. 0. & Pardee, A. B. (1971) J. Biol. Chem. 246, 6159-6165 5. van der Vliet, P. C. & Levine, A. J. (1973) Nature New Eiol. 246,

170-174 6. van den Broek, H. W. J., Nooden, L. D., Sevall, J. S. & Bonner, J.

(1973) Biochemistry 12, 229-236

‘IG. Herrick and B. Alberts, unpublished results and H. Green, personal communication, 1974.

‘* G. Stein and H. Green, personal communication. Is G. Herrick and T. Fox, unpublished results.

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

2132 Calf Helix-unwinding Proteins

7. Portugal, F. H., Simonds, J., Twardzik, D. R., Mulroy, P. F. & Oskarsson, M. (1973) Biochem. Biophys. Res. Commun. 51, 789-796

8. Poliskv. B. & Kates. J. (1972) Virology 49. 168179 9. Kleinsmith, L. J. (i973) J. Biol. Che”m. 248, 56485653

10. Choe. B. K. & Rose, N. R. (1974) EXP. Cell Res. 83, 271-280 11. Carter, D. B. & Chae, C. (1974) Biochem. Biophys. Res. Commun.

58, 100-1007 12. Rubio, V. & Long, C. (1974) Int. J. Cancer 14, 348358 13. Carroll, R. B., Hager, L. & Dulbecco, R. (1974) Proc. Natl. Acad.

Sci. U. S. A. 71, 3754-3757 14. Stein, G. H. (1975) Exp. Cell Res. 90, 237-248 15. Shanmugam, G., Bhaduri, S., Arens, M. & Green, M. (1975)

Biochemistry 14, 332-337 16. Huang, A. T., Riddle, M. M. & Koons, L. S. (1975) Cancer Res. 35,

981-986 17. Alberts. B. M. & Frey, L. (1970) Nature 227, 1313-1318 18. Sigal, N., Delius, H.: Komberg, T., Gefter, M. L. & Alberts, B.

(1972) Proc. Natl. Acad. Sci. U. S. A. 69, 3537-3541 19. Reuben, R. C. & Gefter, M. L. (1973) Proc. Natl. Acad. Sci.

U. S. A. 70, 1846-1850 41.

20. Scherzinger, E., Litfin, F. & Jost, E. (1973) Mol. Gen. Genet. 123, 42. 247-262 43.

21. Layne, E. (1957) Methods Enzymol. 3, 447-454 22. Yamamoto, K. R., Alberts, B. M., Benzinger, R., Lawhorne, L. &

Treiber, G. (1970) Virology 40,734-744 - 23. Herrick, G. A. (1973) Ph.D. thesis, Princeton University, Prince-

ton, N.J. 24. Shapiro, A. L., Viriuela, E. & Maizel, J. V., Jr. (1967) Rio&em.

Biophys. Res. Commun. 28, 815-820 25. Phelps, R. A. & Putnam, F. W. (1960) in The Plasma Proteins

(Putnam, F. W., ed) Academic Press, N.Y. 26. Williams, R. C., Jr. & Rajagopalan, T. G. (1966) J. Biol. Chem.

241, 4951-4954

27.

28. Edmundson, A. B. (1965) Nature 205,883~887 29. Wrigley, C. (1968) Sci. Tools 15, 17-23 30. Awdeh, 2. L. (1969) Sci. Tools 16, 42-43 31. Martin, A. J. P. & Synge, R. L. M. (1949) Biochem. J. 35,

1358-1368 32. 33. 34. 35. 36. 37. 38. 39.

Herrick, G. & Alberts, B. (1976) J. Biol. Chem. 251,2133-2141 Elgin, S. C. R. & Banner. J. (1970) Biochemistry 9.4440-4447 Tyler, D. D. & Gonse, J. (1967) Methods Enzymol. 10, 101-103 Zubay, G. & Doty, P. (1959) J. Mol. Biol. 1, l-20 Laemmli, U. K. (1970) Nature 227, 680-685 Deleted in proof. Deleted in proof. Herrick, G., Delius, H. & Alberts, B. (1976) J. Biol. Chem. 251,

2142-2146 40. Spackman, D. H., Stein, W. H. & Moore, S. (1958) Anal. Chem.

30, 1190-1206 Beaven, G. H. & Holiday, E. R. (1952) Adu. Protein Chem. 7,

319-386

44. 45. 46.

Panyim, S. & Chalkley, R. (1971) J. Viol. Chem. 246, 7557-7560 Johns, E. W. (1971) in Histone and Nucleohistones (Phillips, D.

M. P., ed) pp. l-45, Plenum Press, New York Bollum, F. J. (1975) Prog. Nucleic Acid Res. 15, 109-144 Yoneda, M. & Bollum, FIJ. (1965) J. Biol. Chem. 240.3385-3391 Yasumura, Y., Tashjian, A. H. & Sato, G. H. (1966) Science 145,

1186-1189 47. Tsai, R. L. & Green, H. (1973) J. Mol. Biol. 73, 307-316 48. Seifart, K. H., Juhasz, P. P. & Benecke, B. J. (1973) Eur. J.

49. 50.

Dayoff, M. 0. (1972) Atlas ofProtein Sequence and Structure, Vol. 5, National Biomedical Research Foundation, Washington, D. C.

Biochem. 33, 181-191 Hotta, Y. & Stem, H. (1971) Deu. Biol. 26, 87-99 Banks, G. R. & Spanos, A. (1975) J. Mol. Biol. 93,63-77

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from

G Herrick and B Albertsfrom calf thymus.

Purification and physical characterization of nucleic acid helix-unwinding proteins

1976, 251:2124-2132.J. Biol. Chem. 

  http://www.jbc.org/content/251/7/2124Access 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/251/7/2124.full.html#ref-list-1

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

by guest on January 6, 2020http://w

ww

.jbc.org/D

ownloaded from