THE JOURNAL OF BIOL~CICAL CHEMISTRY Vol. 251, No. 13, Issue of July 10, pp. 3868-3874, 1976

Printed in U.S.A.

The Ovalbumin Gene PARTIAL PURIFICATION OF THE CODING STRAND*

(Received for publication, December 17, 1975)

SAVIO L. C. Woo, ROY G. SMITH, ANTHONY R. MEANS,$ AND BERT W. O’MALLEY

From the Department of Cell Biology, Baylor College of Medicine, Houston, Texas 77030

Purified ovalbumin messenger RNA was employed to selectively enrich the concentration of the gene coding for ovalbumin from total chick DNA by molecular hybridization. The coding strand of the ovalbumin gene was partially purified from sheared chick DNA by affinity column chromatography using ovalbumin mRNA immobilized on phosphocellulose. The concentrations of the ovalbumin DNA sequence in various DNA fractions were quantitated by measuring their rates of hybridization with 12”I-labeled ovalbumin mRNA. When apparent Cot,+ values of these reactions were compared to the apparent Co& value obtained from the hybridization reaction between ‘261-ovalbumin mRNA and complementary DNA synthesized against ovalbumin mRNA using the enzyme reverse transcriptase, purification of the coding ovalbumin DNA strand over total chick DNA was estimated to be approximately 9,600-fold. There was no apparent degradation of the 4,000 nucleotide strands of chick DNA throughout the purification procedure. Since ovalbumin mRNA has a complexity of 1890 nucleo- tides, the resulting DNA was more than twice the length of ovalbumin mRNA and thus should contain DNA sequences adjacent to the structural portion of the ovalbumin gene.

Evidence accumulated recently has indicated that estrogen induces the synthesis of ovalbumin messenger RNA in the chick oviduct via a “transcriptional control” mechanism (l-12). To attempt definitive studies on the interaction be- tween the ovalbumin gene and oviduct DNA-dependent RNA polymerase, hormone-receptor complexes and chromosomal

proteins, the ovalbumin gene must eventually be purified. The study of these interactions and their effects on transcription of the isolated ovalbumin gene in vitro would provide further insight into the molecular mechanism of steroid hormone action as well as the general mechanism by which effector molecules regulate gene expression in eukaryotes. Such studies have been successfully carried out in several prokaryotic systems where purified operon DNAs were used to investigate the mechanism of interaction with various regulatory proteins. These studies have led to the revelation of the sequence, organization, and function of operator and promotor regions. Moreover, such investigation has greatly enhanced the under- standing of regulatory aspects of transcription of these operons by repressors and inducers at the molecular level (13-16).

Owing to the enormous complexity of eukaryotic genomes, isolation of specific genes from a variety of these organisms has been restricted to a class of reiterated genes with G + C contents that were quite distinct from that of the bulk of nuclear DNA. Examples are those genes coding for ribosomal

* This work was supported by National Institutes of Health Grant HD-08188, American-dancer Sbciety Grant BC-101, The Robert A. Welch Foundation (Q-611), and the Baylor Center for Population Research and Reproductive Biology.

*Recipient of Faculty Research Award 128 from the American Cancer Society.

RNA, transfer RNAs, 5 S RNA, and histone messenger RNAs (17-26). The gene coding for fibroin messenger RNA also has an unusually high G + C content and is the only unique sequence eukaryotic gene which has been isolated to date (27, 28). Isolation of a unique eukaryotic gene with a density which is similar to total DNA, such as that for ovalbumin, requires methods of DNA fractionation based on parameters other than size or density alone. The availability of large quantities of ovalbumin messenger RNA purified to homogeneity with respect to various physical, chemical, and biological criteria (29, 30) allowed us to partially purify its DNA complement from total chick DNA by molecular hybridization. Purified ovalbumin mRNA was immobilized onto phosphocellulose and its DNA complement was enriched 9,600-fold from total chick DNA by affinity chromatography.

EXPERIMENTAL PROCEDURE

Materials-White Leghorn laying hens were purchased from Rich- Glo Farm, La Grange, Texas; oviducts and livers were quickly frozen in liquid nitrogen after dissection. Liquified phenol and reagent grade formamide (A,,, < 0.25) were from Fisher Scientific Co., Fair Lawn, N. J.; phenol was redistilled before use. Wheat germ was a generous gift from General Mills, Inc., Minneapolis, Minn. Creatine phosphate, creatine phosphokinase, and sodium dodecyl sulfate were from Sigma Chemical Co. [“C!]Valine (260 mCi/mmol), ultrapure urea, and sucrose (ribonuclease free) were from Schwa&Mann. Oranaebura. N. Y. Pancieatic ribonuclease A and cx-amylase were from W&thin&on. Pronase was from Calbiochem. Oligo(dT)-cellulose (T3) was from Collaborative Research. Na’*LI (carrier free > 300 mCi/ml) was from Amersham/Searle. Agarose (electrophoretic grade), Chelex 100, and Cellex P powder were from Bio-Rad Laboratories, Richmond, Calif. CsCl was from EM Laboratories. Inc.. Elmsford. N. Y. l,l’-Carbonvl- diimidazole was from Aldrich Chemical Co., Milwaukee, Wise. Pyii-

3868

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Partial Purification of Ovalbumin Gene 3869

dine and N,N-dimethylformamide were from Matheson, Coleman and Bell, Norwood, Ohio. Avian myeloblastosis virus reverse transcriptase was a kind gift from Dr. J. W. Beard, Life Sciences, Inc., St. Petersburg, Fla. Deoxyribonucleoside triphosphates were from Pabst, Milwaukee, and the radioactive nucleotides were from New England Nuclear Corporation.

Preparation of Purified Ovalbumin n&VA-Ovalbumin mRNA was purified from total nucleic acid extracts of hen oviducts as described previously (29, 30) with the following modifications: (a) nitrocellulose filtration was replaced by oligo(dT)-cellulose column chromatography, and (b) the RNA was heated at 70° for 1 min and quick cooled before chromatography on the Sepharose 4B column and the second oli- go(dT)-cellulose column. Better resolution and recovery of ovalbumin mRNA were obtained with these modifications. Throughout the purification procedure, ovalbumin mRNA and total mRNA activities were quantitated by the cell-free wheat germ translation assay (29), and the purity of the mRNA was monitored by acid-urea agarose gel electrophoresis (29). The purified ovalbumin mRNA was labeled with Na? to a specific radioactivity of 10’ cpm/pg as previously described (30). The labeled product co-migrated with untreated ovalbumin mRNA in acid-urea agarose gels.

Synthesis of DNA Complementary to Ovalbumin mRNA-DNA complementary to purified ovalbumin mRNA was synthesized to various specific radioactivities using reverse transcriptase isolated from avian myeloblastosis virus as previously described (31). To obtain a specific radioactivity of lOa cpm/pg, the cDNA,,’ was synthesized as follows: 0.2 mCi each of [SH]dGTP, [SH]dATP, and [SH]dCTP were lyophilized to dryness and redissolved in 1 ml of a solution containing 50 rnM Tris-HCl, pH 8.3, containing 20 mM dithiothreitol, 6 mM MgCl,, 200 PM dTTP, 2.5 pg/ml of oligo(dT) 18.10, 36 &ml of actinomycin D, and 25 @g/ml of purified mRNA,,. The final concentrations of the [SH]dGTP, [SH]dATP, and [8H]dCTP were adjusted to 35 /.LM each. The mixture was chilled in ice for 5 min before reverse transcriptase was added to a final concentration of 60 units/ml. This solution was mixed and immediately incubated at 46” for 5 min. The reaction mixture was adjusted to contain 0.5% of sodium Sarkosyl, 10 mM

Na,EDTA, and 100 pg/ml of sheared Escherichia coli DNA. The sample was passed through a Sephadex G-50 column at room tempera- ture in 0.01 M Tris-HCl, pH 7.6, containing 0.1 M NaCl. Nucleic acid in the excluded fractions was precipitated with ethanol and redissolved in 0.5 ml of 0.1 M NaOH, containing 10 mM Na,EDTA, and heated at 60” for 30 min. The solution was chilled, adjusted to pH 5.0 with 2 M

sodium acetate, pH 4.5, and the [SH]cDNA,, was again precipitated with alcohol. Approximately 15 to 20% of the ISH]cDNA,, represented complete complements of ovalbumin mRNA and greater than 50% revealed a complexity of 1000 nucleotides or more (31). These characteristics were determined by several methods including analysis by alkaline sucrose gradient centrifugation, polyacrylamide gel electro- phoresis in the presence of 98% formamide and electron microscopy using the formamide-urea procedure previously described (31). To synthesize cDNA,, with very low specific radioactivities all ingredients in the synthetic reaction were identical to those described above, except only [S2P]dGTP was used as label. Specific radioactivity of the [?‘]dGTP in the reaction mixture was adjusted to 4 &i/pmol, and the specific radioactivity of the [‘?P]cDNA,, product was 5380 cpm/pg.

Oualbumin [S2P]cDNA Excess Hybridization with ‘Z51-Oualbumin mRNA-Hybridization between ‘*‘I-ovalbumin mRNA and [s*P]- cDNA,, in excess was carried out as previously described (30). ‘*9-ovalbumin mRNA, 10s cpm ( lo7 cpm/Mg), was mixed with 0.5 ml of [31P]cDNA,, (5380 cpm/pg) in hybridization buffer (0.01 M N-tris(hy- droxymethyl)methyl-2-amino ethanesulfonic acid, 0.75 M NaCl, 0.005 M Na,EDTA, and 50% formamide, pH 7.0). Aliquots of 50 ~1 were sealed in capillary tubes, heated for 10 min at 70”, and then incubated for various times at 41°. The reaction was terminated by a 50-fold dilution in 2.5 ml of cold 0.24 M sodium phosphate, pH 7.0. The diluted samples were divided into two l-ml portions. One portion was treated with 50 &ml of pancreatic ribonuclease A for 1 hour at 25”, chilled in ice, and adjusted to 10% with respect to trichloroacetic acid. The other portion was similarly treated with trichloroacetic acid but ribonuclease was not added. Sheared chick DNA, 100 pg, was added to each tube for co-precipitation. The trichloroacetic acid-insoluble radioactivity was collected on glass fiber filters, washed with 5% trichloroacetic acid, and solubilized in 0.5 ml of Nuclear Chicago solubilizer at 60” for 1 hour.

‘The abbreviations used are: cDNA,,, enzymatically synthesized DNA complementary to ovalbumin mRNA; mRNA,,, ovalbumin mRNA.

The radioactivity content was determined by liquid scintillation spectrometry in a spectrofluor-toluene scintillation fluid (32). Per cent hybridization is the fraction of total trichloroacetic acid-precipitable radioactivity that has been rendered ribonuclease-resistant during the course of hybridization.

Preparation of Sheared Chick DNA-Chick liver DNA, free of protein and RNA, was prepared by a modification of the procedure of Marmur (33) as described previously (34). The DNA was adjusted to 1 mg/ml in 0.1 x SSC (0.015 M sodium chloride/O.0015 M sodium citrate, pH 7.0), containing 0.2 M NaCl and 0.005 M Na,EDTA and sheared in a French Press to a mean length of 4000 nucleotides. The length of the sheared DNA was estimated from sedimentation in alkaline sucrose gradients. Sheared DNA solutions were further treated with 50 pg/ml of pronase for 1 hour at 37” and extracted twice with the same buffer (phenol/sodium dodecyl sulfate buffer, pH 8.0) used for total nucleic acid extraction. The DNA was finally precipitated from solution with alcohol and redissolved in 0.001 M Na,EDTA, pH 7.0. Two volumes of a buffer (0.015 M N-tris(hydroxymethyl)methyl-2.amino ethanesulfonic acid, 0.0075 M Na,EDTA, 1.125 M NaCl, and 75% formamide, pH 7.0) was added to the DNA solution. This DNA solution was passed through a Chelex 100 column before use for hybridization reactions.

Coupling Purified Ovalbumin mRNA to Phosphocellulose-Purified ovalbumin mRNA was chemically linked to Cellex P powder using a method based on that described by Saxinger et al. (35) and by Shih and Martin (36). Cellex P powder, 10 g, was repeatedly suspended in water and decanted until no fine particles remained. The powder was collected on a Buchner funnel and washed successively with large volumes of water, 0.25 M NaCl in 0.25 M HCl, water, 0.25 M NaOH, and finally with water to neutrality. The powder was resuspended in 400 ml of 0.1 M tributylamine hydrochloride, pH 6.0, and filtered. This procedure was repeated four times and the powder was finally dried by suction. It was then dried in oacuo over P,O,/KOH for 7 to 10 days, and activated by gently shaking for 16 hours at room temperature with 100 ml of N,N’-carbonyldiimidazole dissolved in anhydrous dimethyl- formamide (100 mg/ml). The activated powder was collected in a Buchner funnel, washed extensively with anhydrous dimethylformam- ide followed by acetone, and finally dried in U~CUO over Drierite overnight. Purified ovalbumin mRNA, 0.7 mg, and 1 pg of 12sI-oval- bumin mRNA (10’ cpm/pg) in 1 ml of water were added to 2 g of the activated phosphocellulose powder using a Pasteur pipette. The mixture was stirred with a glass rod for 5 min and again dried in oacuo over Drierite overnight. The coupling reaction was carried out by shaking the mixture gently in 100 ml of anhydrous pyridine at 50” for 24 hours. The resin was collected by centrifugation at 3,000 x g for 15 min and the supernatant fluid was removed. The excess activated phospho groups were inactivated by gently mixing with 100 ml of 15% NH,OH in an icewater bath for 20 min. The mixture was again centrifuged and the supernatant fluid was removed. The pellet was subsequently washed three times with 0.1 M sodium phosphate, pH 5.0, once with 50% formamide and five times with water in a similar manner. About 30% of the radioactivity was solubilized from the resin during the ammonium hydroxide treatment, but all of the other washing combined constituted no more than 10% of the total input radioactive ovalbumin mRNA. Very little radioactivity could be eluted from the resin by hybridization buffer at 70”, and the binding capacity of the resin did not diminish appreciably after 3 months of continuous use.

Affinity Chromatography-The procedure was carried out by a modification of the procedure described by Shih and Martin (37). Two grams of phosphocellulose-ovalbumin mRNA matrix was packed in a jacketed column (2 x 15 cm). The column was connected in series with a peristaltic pump, a reservoir, and a second jacketed column (2 x 15 cm) which was back-connected to the column containing the affinity resin in a closed circuit. The second column was maintained at 70” (8” above the Z’, of chick DNA) so that all DNA molecules would be denatured before coming into contact with the ovalbumin mRNA affinity resin in the first column which was regulated at 46“. Six thousand A,., units of sheared chick DNA in 500 ml of hybridization buffer was recycled through this system for 3 days at a flow rate of approximately 2 ml/min. At the end of the hybridization period, the two columns were disconnected. The resin was washed with fresh hybridization buffer until no further A,,,-adsorbing material could be eluted. The column temperature was then raised to 55’ for 15 min and the resin was again washed with fresh buffer. The bound DNA was finally eluted with hybridization buffer at 70”. Aliquots of 1.5 ml were collected, and the fractions containing A,,,-adsorbing material were pooled. This step was repeated four times and the 70” eluates were

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

3870 Partial Purification of Oualbumin Gene

pooled and diluted to 500 ml with hybridization buffer. The diluted DNA solution was recycled through the mRNA affinity column for 70 hours in a similar manner for a second time. At the end of the hybridization period, the affinity resin was washed with hybridization buffer successively at 46, 55, and finally at 70”. The 70” eluate from the second affinity chromatography step was diluted to 300 ml with hybridization buffer and recycled through the affinity resin for a third time. After 70 hours of hybridization, the bound nucleic acid was again eluted at 70”.

Chick DNA/‘251-Ovalbumin mRNA Hybridization-Ovalbumin DNA sequence present in various chick DNA fractions was detected by hybridization with 1251-ovalbumin mRNA as described above.2 The DNA fractions were treated with 0.1 N NaOH at 62” for 1 hour, chilled, neutralized with 2 M sodium acetate, pH 4.5, and precipitated with alcohol. The precipitate was redissolved in hybridization buffer. ‘2SI-labeled ovalbumin mRNA (lo7 cpm/hg), lo6 cpm, was added to 0.5 ml of the DNA solution and 50.~1 aliquots were sealed in capillary tubes. The tubes were heated for 10 min at 70” and hybridization was carried out at 41” for various lengths of time. The reaction was terminated by a 50.fold dilution into 2.5 ml of cold 0.24 M sodium phosphate buffered at pH 7.0. Aliquots of 1 ml were treated with pancreatic ribonuclease A while the other l-ml aliquots were directly treated with 10% trichloroacetic acid. Radioactivity content was determined by liquid scintillation spectrometry and per cent hybridi- zation was calculated as described above.

RESULTS

Enrichment of Ovalbumin DNA Sequence by Affinity Chromatography-The coding ovalbumin DNA strand was enriched from total chick DNA after hybridization with ovalbumin mRNA that was coupled to the inert phosphocel- lulose matrix. The matrix was washed free of unhybridized DNA and the ovalbumin DNA could then be eluted from the matrix by raising the temperature sufficiently to allow the hybrids to dissociate. To characterize the affinity column, 130,000 cpm of [“P]cDNA,, (10’ cpm/Fg) in the presence of excess chick DNA was allowed to recycle through the resin at 46” in a total volume of 60 ml. Aliquots of 1 ml were taken at various times of hybridization for radioactivity determination. Radioactivity content in the hybridization medium diminished gradually to a level below 35% of the initial radioactivity after 70 hours of hybridization (Table IA). After extensive washing at 46”, bound nucleic acid was eluted at 70” and the radioactiv- ity recovered in the eluate was 80,000 cpm, corresponding to 61.5% of the starting radioactivity (Table IB). During the course of the reaction, 6 A,,, units of DNA were recovered from a total of 750 A,,, units, representing a mass yield of 0.8%. The purification of [SZP]cDNA,, by this procedure was therefore 76.Cfold.

Six thousand A,,, units of total chick DNA, sheared to a mean length of 4,000 nucleotides, was thus allowed to recycle through the ovalbumin mRNA affinity column. An additional washing of the resin at 55” was employed before elution at 70”. Twenty-six and 16.5 A,,, units of DNA were eluted at 55” and 70°, respectively. Aliquots of these fractions were assayed for ovalbumin DNA sequences by hybridization with 1251-oval- bumin mRNA. Total chick DNA reacted very slowly, and the phosphocellulose-unbound DNA reacted at a further reduced rate (Fig. 1). On the other hand, DNA eluted at 70’ hybridized with the mRNA to a level of 60% and at a much higher rate with an apparent Cot,+ value of 40. The extent of hybridization

2“DNA excess” conditions were not established in these experi- ments. These conditions were not necessary because rate of hybridiza- tion is not a function of the driver/tracer ratio, but a function of the concentration of the reactants (38). Thus, relative concentrations of the ovalbumin coding DNA strand in various DNA fractions can be obtained by comparison of the Cot,+ values of the hybridization reactions.

TABLE I

Affinity chromatogrdphy of [“P]cDNA,,,

A. Recychng time Radioactivty

content DNAcontent

hours CPm AZ.0

0 130,000 750 20 100,000 750 40 70,000 750 70 50,000 750

B DNA samples Radioactivity DNA

Yield Purifi-

content content cation

CPm A%0 % -fold

Initial hybridization 130,000 750 100 1 medium

70” eluate 80,000 6 61.5 76.8

1ou 1OL 10' lb'

c cd

FIG. 1. Hybridization of ‘2sI-ovalbumin mRNA to various DNA fractions obtained by affinity column chromatography using oval- bumin mRNA coupled to phosphocellulose. Each reaction point contained 10,000 cpm of ‘*4-ovalbumin mRNA (10’ cpm/pg). Condi- tions of hybridization were as described under “Experimental Proce- dure.” Duration of hybridization ranged from 30 s to 100 hours. Hybrid formation was assayed using ribonuclease as described under “Experi- mental Procedure,” and the data points were corrected by a 5% background ribonuclease resistance as determined by treating an aliquot of the reaction mixture in an identical manner except that the mixture was not incubated. A-A, total chick DNA; O-O, DNA fraction not bound to the ovalbumin mRNA affinity resin; O-O, DNA fraction eluted from the affinity resin at 55’; O-0, DNA fraction eluted from the affinity resin at 70”.

did not reach completion because the ovalbumin DNA to Y-ovalbumin mRNA ratio in the reaction might not have been sufficient (39), since only a small aliquot of the DNA solution was used in the experiment.3 Although the 55’ eluate also reacted with ‘ZsI-ovalbumin mRNA faster than total chick DNA, its apparent Cot, value was approximately 9 times greater than that of the 70’ eluate (Fig. 1).

Sixty-five A,,, units of the DNA recovered from the affinity column was rechromatographed through the same resin. Ap- proximately 2.4 A,,, units of DNA were recovered in the subsequent 70” eluate. This time the DNA content in the 55” wash was insignificant. The unbound DNA and the 70’ eluate

SDNA was originally sheared randomly and the DNA/Y-mRNA hybrids might contain single strand ‘Z”I-mRNA tails. Since we were assaying hybrid formation by ribonuclease digestion, the tails would be degraded and escape scoring. This effect would contribute to the observation that the extent of hybridization was less than complete. Furthermore, these tails were capable of hybridizing with additional DNA molecules. But the rate of this reaction was slower than the original hybridization reaction and would cause the curves to deviate from ideal C,t, curves (40).

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Partial Purification of Ovalbumin Gene 3871

were again treated with alkali and assayed for ovalbumin DNA sequence by hybridization with Y-labeled ovalbumin mRNA. The unbound DNA was inactive whereas the 70” eluate reacted at a further accelerated rate with an apparent CotH value of 2.3 (Fig. 2). The extent of hybridization was 70% and the apparent reduction of the CotH value was 17.4-fold when compared to that of the 70” eluate from the first affinity column. When chromatographed through the affinity column for a third time, the 70” eluate hybridized with 12SI-ovalbumin mRNA with an apparent CotH value of 1.25 (Fig. 3). Apparent purification was therefore 1.84-fold over the second affinity column step and the DNA mass recovered was 0.9 A,,, units.

Estimation of Extent of Enrichment of Ovalbumin DNA -Although partially purified DNA fractions hybridized with 12SI-ovalbumin mRNA much faster than did chick DNA, the precise extent of enrichment and the recovery of ovalbumin DNA sequence from total chick DNA could not be readily determined. This is because the Cot,+ value of the DNA-excess hybridization reaction between total chick DNA and ‘Y-oval- bumin mRNA could not be accurately assessed (Fig. 1). These parameters, however, could be determined by comparison of

cot FIG. 2. Hybridization of ‘ZSI-ovalbumin mRNA to various DNA

fractions from the second affinity column chromatography step. Conditions of hybridization and assay of hybrid formation were identical to that described in the legend of Fig. 1. O-Cl, DNA fraction not bound to the mRNA affinity resin; O----O, DNA fraction bound to the affinity resin that was eluted at 70”; A-A, DNA fraction eluted at 70” from the first affinity chromatography step taken from Fig. 1.

FIG. 3. Hybridization of ‘211-ovalbumin mRNA to DNA eluted at 70” from the third affinity column chromatography step (04) and to 0.22 pg/ml (A-A), 0.55 pg/ml (O-Cl), and 1.1 pg/ml (O-O) of ovalbumin complementary [sT]DNA (5380 cpm/Fg). Conditions of hybridization and the assay of per cent 1251-ovalbumin mRNA hybridized were identical with that described in the legend of Fig. 1.

the apparent Cots values obtained from the hybridization reactions using the partially purified DNA fractions and that using enzymatically synthesized cDNA,, which should hy- bridize with 1261-ovalbumin mRNA at the highest rate. Under identical hybridization conditions, the reaction was essen- tially complete at a C,t value of 1.0 with an apparent Cots value of approximately 0.011 (Fig. 3). Moreover, identical C,t curves were obtained at three different cDNA,, to mRNA ratios. Using this number as a reference point, the Cot,+ value in the total chick DNA/‘251-ovalbumin mRNA hybridization reaction was calculated to be approximately 12,000. Thus the enrichment of ovalbumin DNA from total chick DNA was approximately 9,600-fold by affinity chromatography (Table II).

Characterization of Ovalbumin DNAl’2SI-Ovalbumin mRNA Hybrid-Although the DNA purified by affinity chro- matography hybridized with 12SI-ovalbumin mRNA at a rate about 9,600-fold faster than that of total chick DNA, the fidelity of the hybrid must be verified. Fig. 4 shows the melting profile of the hybrid obtained using hydroxylapatite column chromatography. The melting profile of a control hybrid containing enzymatically synthesized cDNA,, and 12SI-oval- bumin mRNA was also obtained at the same time. Under the conditions employed, the T,,, values of the two hybrids were 85 and 86”, respectively. The lack of substantial melting of the hybrids at lower temperatures strongly indicates that the 12SI-ovalbumin mRNA was indeed in a true hybrid form since the mRNA by itself does not bind to hydroxylapatite at 60” in the presence of 0.14 M sodium phosphate, pH 7.0.

The hybrid was also characterized with respect to density by equilibrium sedimentation in CsCl gradients (Fig. 5). The hybrid banded at a density which ranged from 1.7050 to 1.7650 and was lower than that of the hybrid formed between [3*P]cDNA,, and ovalbumin mRNA (p = 1.7720). Native chick DNA and [S2P]cDNA,, banded at densities of 1.6960 and 1.7000, respectively, while ‘251-ovalbumin mRNA banded at

1.8100. The relative broadness of the radioactivity peak sug- gests that the size of the single-stranded DNA region relative to the size of the DNA/RNA double-stranded region in the hybrid was polydisperse. These observations would be expected since

TABLE II

Purification of oualbumingene from total chick DNA

Material Dh’A content

Apparent cat,,*

A,., units mollliter -fold %

Total DNA 24,000 12,000” 1 100

DNA bound to 1st mRNA,, 65 400 300 81

affinity column

DNA bound to 2nd mRNA,, 2.4 2.3’ 5200 52

affinity column

DNA bound to 3rd mRNA,, 0.9 1.25” 9600 36

affinity column

“This is an approximate value calculated from the apparent CJ,,~

value of the cDNA,V/‘2SI-mRNA,, hybridization reaction: Apparent

C,t,,,(total chick DNA) = ((2 x 108)/(1890P62)) x 0.011 where 2 x log

is the number of DNA nucleotides present in haploid chick genome,

1890 is the complexity of ovalbumin mRNA, 62 is the number average

chain length of poly(A) within the mRNA,, molecule, and 0.011 is the

apparent Cot,,Z value of the cDNA,JZSI-mRNA,, reaction (Fig. 3).

*Taken from Fig. 1.

c Taken from Fig. 2.

d Taken from Fig. 3.

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

3872 Partial Purification of Ovalbumin Gene

90-

80-

70 -

60 -

50 -

40 -

30-

20 -

IO-

60 65 70 75 80 85 90

TEMPERATURE (%)

FIG. 4. Hydroxylapatite column chromatography of the hybrids formed between ‘*sI-ovalbumin mRNA and cDNA,, (04) or ovalbumin DNA eluted from the third affinity column (0-O). The hybrids were applied to l-ml hydroxylapatite columns at 60” in 0.14 M

sodium phosphate, pH 7.0. The columns were washed with the same buffer until no radioactivity was eluted. The temperature was then raised at 2.5” increments and the columns were eluted with 3 x 2 ml of the same buffer at each increment. lZ?-ovalbumin mRNA eluted at each temperature increment was plotted commulatively as per cent of total 12sI-ovalbumin mRNA eluted.

the DNA was originally sheared to more than twice the length of ovalbumin mRNA.

The purified DNA, although 9,600-fold enriched in the DNA sequence that is complementary to ovalbumin mRNA, is useful for future studies only if the DNA has not been degraded during the purification process. The size of the partially purified DNA preparation was thus analyzed on an alkaline sucrose gradient and compared to total chick DNA sheared to a mean length of 4,000 nucleotides. Since the migration of the two DNA peaks in gradients were identical (Fig. 6), the mean length of the partially purified DNA was still more than twice that of the ovalbumin mRNA. Thus, no significant degradation of DNA occurred during the isolation procedure.

DISCUSSION

Although ovalbumin comprises more than 50% of the pro- teins synthesized in the estrogen-stimulated chick oviduct, the protein is coded by a unique sequence gene which is not amplified during estrogen stimulation (5, 6, 30, 31). Since there are 2 x lo9 nucleotides in the DNA of a haploid chick genome and the complexity of the ovalbumin mRNA is approximately 1800 nucleotides, only (1800/2 x log) of the chick DNA is

22 -

a-

18 -

16 -

14 -

"b 12-

; lo-

8 a-

6-

4-

2-

0 2 4 6 8 10 12 14 16 18

FRACTION NUMBER

FIG. 5. CsCl equilibrium centrifugation of the hybrid formed be- tween l*sI-ovalbumin mRNA and ovalbumin DNA eluted from the third affinity column. Ovalbumin DNA eluted from the third affinity column was incubated for 38 hours with 50,000 cpm of ‘2”I-ovalbumin mRNA. The mixture was precipitated with alcohol and redissolved in 2.5 ml of 0.01 M Tris-HCl, pH 7.6, containing 5 mM Na,EDTA. CsCl was added to yield a refractive index of 1.4035 and the gradient was developed by centrifugation for 70 hours at 53,000 rpm in a Ti 75 angle rotor at 25”. Aliquots of 0.15 ml were collected and their refractive indices were measured (IX-Cl). O.l-ml aliquots were then diluted ith 2.5 ml of 0.24 M sodium phosphate, pH 7.0, and subsequently divided into two l-ml portions. One portion was treated with 50 rg/ml of rihonuclease A (O--O) and the other portion was directly precipitated with trichloroacetic acid (A-A). The precipitates were collected on glass fiber filter papers and radioactivity contents determined as described under “Experimental Procedure.” The arrows a, b, c, and d represent densities of native chick DNA, [T]cDNA,,, a hybrid formed between mRNA,, and [3zP]cDNA,,, and mRNA,, alone in parallel gradients, respectively.

represented by the coding strand of the structural ovalbumin gene. Thus, the isolation of this DNA strand to homogeneity requires a purification of greater than 1 million-fold. The ovalbumin gene, as indicated by the base composition studies of purified ovalbumin mRNA (30), does not have a G + C con- tent substantially different from that of total DNA and there- fore cannot be isolated by equilibrium centrifugation alone. Molecular hybridization was thus employed as a tool to effect a partial purification of this gene sequence.

Initially, we allowed total chick DNA to hybridize with excess ovalbumin mRNA in liquid medium. The majority of the complementary DNA strand was hybridized to the mRNA instead of its DNA complement since the mRNA was in large excess (39). The DNA/mRNA hybrid was next removed from the bulk of DNA by an oligo(dT)-cellulose column. The fractionation was effective due to the presence of a polyade- nylic acid tract on the 3’ terminus of the mRNA. Although a 35-fold purification of the coding ovalbumin DNA strand was achieved in this manner, the yield was relatively poor because fractionation of the hybrid from total DNA depended greatly on the integrity of the mRNA at the end of the hybridization reaction. Should there be a slight degradation of the mRNA during hybridization, only those DNA strands that were hybridized to the 3’ terminus of the mRNA would be co-puri- fied with the mRNA in the oligo(dT)-cellulose column chroma- tography step. To overcome this difficulty, we next covalently immobilized ovalbumin mRNA onto an inert matrix and this

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

Partial Purification of Ovalbumin Gene 3873

by limiting the C,t values attained during the reassociation. To determine the rate of hyperpolymer formation between total chick DNA sheared to 4,000 nucleotides in length, the dena- tured DNA was allowed to reassociate to various C,t values, followed by centrifugation and analysis of the DNA content of the pellets. Indeed it was observed that the formation of hyperpolymers was minimal if, under the hybridization condi- tions employed, the C,t value did not reach 100. It was thus possible to isolate the ovalbumin gene from sheared chick DNA using purified ovalbumin mRNA as long as the DNA reassocia- tioion C!! values during hybridization could be limited to less

0.21 B

0- TOP BOTTOM

FIG. 6. Alkaline sucrose gradient centrifugation of chick DNA sheared to a mean length of 4.000 nucleotides (Panel A). and DNA eluted from the third affinity column chomatography step (Panel B). The migration of the open (16.1 S) and closed (18.4 S) forms of @X 174 DNA in parallel gradients are indicated by the arrows a and b, respectively.

matrix was incubated with a solution of denatured chick DNA. The matrix of choice was phosphocellulose because of its higher coupling efficiency with RNA as compared to neutral cellulose (36) and the stability of this affinity resin over a long period of time. Indeed, SV40 DNA was purified from large quantities of non-specific DNAs using SV40 RNA coupled to phosphocel- lulose (36). For the present study, this approach offers the advantage of providing a high yield of ovalbumin gene se- quences because we did not have much difficulty in fractionat- ing the hybrid from the starting material at the end of hybridization. Nevertheless, a long incubation time was re- quired since the mRNA was immobilized.

A major difficulty encountered in our attempts to isolate the

The closed recycling system described by Shih and Martin (37) was thus employed to enrich for the ovalbumin DNA sequences. Chick DNA was denatured in a second column immediately prior to re-exposure to the immobilized mRNA. Since the denatured DNA solution only remained in the hybridization column for a short time, the extent of hyper- polymer formation was limited and prolonged hybridization times were possible. After 3 days of recycling, greater than 60% of the ovalbumin DNA sequence was bound to the resin (Tables IB and II). When this procedure was repeated (three times), the bound DNA reacted with lZ?-ovalbumin mRNA with an apparent Cot,+ of 1.25 (Fig. 3). The extent of enrich- ment of ovalbumin gene over total chick DNA, however, was difficult to estimate because the Cot,+ value of the reaction using total chick DNA cannot be accurately measured in the same assay. In order to achieve a C,t of 10,000 the DNA concentration must be greater than 10 mg/ml in order to prevent the hybridization time from exceeding 100 hours. At 10 mg/ml, the 4,000-nucleotide chick DNA fragments are quite viscous even in the presence of 50% formamide. When hybridi- zation was carried out to a C,t of greater than 10,000, the viscosity of the DNA hyperpolymer formed was such that hybrid formation could not be accurately measured. This technical difficulty could be overcome if the Cot,+ value of the reaction using pure ovalbumin gene and l*?-ovalbumin mRNA could be determined under the same hybridization conditions.

ovalbumin gene from sheared chick DNA by affinity chroma- tography using immobilized ovalbumin mRNA was the tend- ency of the denatured chick DNA to form DNA hyperpolymers during the prolonged incubation time. A hyperpolymer is a multistranded structure of reassociated DNA (41, 42). The formation of these hyperpolymers probably occurs through the reassociation of the middle repetitive sequences which consti- tute about 30% of the chick genome (41-43). When reassociated to high C,t values, the hyperpolymers formed were so large that they formed DNA gels and could be readily removed from solution by low speed centrifugation. Thus, subsequent frac- tionation of a DNA/RNA hybrid fro,m the bulk of chick DNA would be less difficult if the formation of these hyperpolymers could be prevented. Since a hyperpolymer is formed by the reassociation of many DNA strands, its rate of formation should be slower than the rate of a simple repeated sequence reassociation where only two DNA strands are involved. The formation of DNA hyperpolymers could therefore be controlled

Complementary DNA generated against ovalbumin mRNA using reverse transcriptase is part of the ovalbumin gene. Our capability to generate a complete complementary DNA tran- script of ovalbumin mRNA in large quantities (31) makes this approach feasible. The complementary DNA preparation was thus allowed to hybridize in large excess with ‘Y-ovalbumin mRNA and the Cot,+ value obtained was 0.011 using three concentrations of cDNA,, in the hybridization medium (Fig. 3). Although comparison of this value with that obtained from the hybridization reaction between the final DNA preparation and 1*4-ovalbumin mRNA (C,t, = 1.25) is not quantitative, an approximate estimation of enrichment of the ovalbumin coding strand with respect to total chick DNA by repeated affinity chromatography should be (O.OlU1.25) x ((2 x log)/ (1890 - 62)) = 9600-fold, where 2 x loo is the number of DNA nucleotides present in haploid chick genome, 1890 is the com- plexity of ovalbumin mRNA and 62 is the number average chain length of polyadenylic acid at the 3’ terminus of the mRNA (30). A greater degree of precision was difficult to obtain because the DNA/Y-ovalbumin mRNA reactions did not reach completion.

The fidelity of the partially purified ovalbumin DNAsse- quence was indicated by its ability to form stable hybrids with 1*9-ovalbumin mRNA. The density and melting temperature of the hybrid composed of this DNA fraction and ‘Y-oval-

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

3874 Partial Purification of Ovalbumin Gene

bumin mRNA were quite similar to those of the hybrid formed between cDNA,, and ovalbumin mRNA (Figs. 4 and 5). Furthermore, the size distribution of the partially purified DNA was identical to that of the starting DNA (Fig. 6), indicating that no DNA degradation had occurred during the purification process. Since chick DNA was originally sheared to a mean length of 4,000 nucleotides, the partially purified DNA preparation described in this communication should contain some DNA sequences that are adjacent to the struc- tural ovalbumin gene, the complexity of which should be identical to that of the ovalbumin mRNA (1890 - 62 nucleo- tides). Whether these “extra” DNA sequences contain the regulatory elements of the structural gene coding for oval- bumin is not known at the present time. However, DNA fragments much longer than 4,000 nucleotides containing the structure gene coding for ovalbumin can be partially purified in the same manner if necessary.

Additional purification of the ovalbumin gene (both coding and anticoding strands) followed by amplification in bacterial plasmids (44-46) are logical extensions of the present study. Only in this manner can we hope to obtain sufficient quantities of specific DNA sequences to study interactions of steroid hormone-receptor complexes and other regulatory proteins at a specific transcriptional locus.

Acknowledgments-We wish to thank Drs. John J. Mona- han, Stephen E. Harris, and Jeffrey M. Rosen for their valuable comments during the course of this work. The excellent technical assistance from Mr. Timothy Adams, Mrs. Serlina Robinson, Mrs. Marty Mills, Mr. Scott Supowit, and Mrs. Celine Yu is gratefully acknowledged.

REFERENCES

1. O’Malley, B. W., and Means, A. R. (1974) Science 183, 610-620 2. Woo, S. L. C., and O’Malley, B. W. (1975) Life Sci. 7,1039-1048 3. Means, A. R., Comstock, J. P., Rosenfeld, G. C., and O’Malley, B.

W. (1972) Proc. Natl. Acad. Sci. U. S. A. 69, 1146-1150 4. Chan, L., Means, A. R., and O’Malley, B. W..(1973) Proc. N&l.

Acad. Sci. U. S. A. 70, 1870-1874 5. Harris, S. E., Means, A. R., Mitchell, W. M., and O’Malley, B. W.

(1973) Z’roc. N&l. Acad. Sci. U. S. A. 70. 3776-3780 6. Sullivan, D., Palacios, R., Stavnezer, J., Taylor, J. M., Faras, A.

J., Kiely, M. L., Summers, N. M., Bishop, J. M., and Schimke, R. T. (1973) J. Biol. Chem. 248,7530-7539

7. Palmiter, R. D. (1973) J. Biol. Chem. 248, 8260-8270 8. Harris, S. E., Schwartz, R. J., Tsai, M.-J., Roy, A. K., and

O’Malley, B. W. (1976) J. Biol. Chem. 251,524-529 9. Tsai, M.-J., Schwartz, R. J., Tsai, S. Y., and O’Malley, B. W.

(1975) J. Biol. Chem. 250, 5165-5174 10. Schwartz, R. J., Tsai, M.-J., Tsai, S. Y., and O’Malley, B. W.

(1975) J. Biol. Chem. 250, 51755182 11. Tsai, S. Y., Tsai, M.-J., Schwartz, R., Kalimi, M., Clark, J. H.,

and O’Mallev, B. W. (1975) Proc. Natl. Acad. Sci. U. S. A. 72, 4228-4232 -

12. Harris, S. E., Rosen, J. M., Means, A. R., and O’Malley, B. W. (1975) Biochemistry 14, 2072-2081

13

14 15.

16.

17.

18. 19.

20. 21.

22. 23.

24.

25. 26.

27.

28.

29.

30.

31.

32.

33. 34.

35.

36. 37.

38.

39.

40.

41.

42.

43.

44.

45.

46.

Gilbert, W., and Maxim, A. (1973) Proc. Natl. Acad. Sci. CJ. S. A. 70. 3581-3584

Maizels, N. (1973) Proc. Natl. Acad. Sci. U. S. A. 70, 3585-3589 Maniatis, T., and Ptashne, M. (1973) Proc. Natl. Acad. Sci.

U. S. A. 70,1531-1535 Nakanishi, S., Adhya, S., Gottesman, M., and Pastan, I. (1975)

J. Biol. Chem. 250,8202-8208 Birnstiel, M. L., Wallace, H., Sirlin, J., and Fischberg, M. (1966)

Natl. Cancer Inst. Monogr. 23, 431-437 Brown, D. D., and Surgimoto, K. (1973) J. Mol. Biol. 78,397-415 Brown, D. D., Wensink, P. C., and Jordan, E. (1972) J. Mol. Biol.

63, 57-73 Quagliarotti, G., and Ritossh, F. M. (1968) J. Mol. Biol. 36,57-67 Kersten, W., Kersten, H., and Szybalski, W. (1966) Biochenistry

5, 236-244 Stern, R. (1973) Carnegie Inst. Wash. Year. B. 72, 15-18 Kedes, L. H., and Birnstiel, M. L. (1971) Nature New Biol. 230,

165-169 Clarkson, S. G., Birnstiel, M. L., and Purdom, I. F. (1973) J. Mol.

Biol. 79, 411-429 Smith, I., Smith, H., and Pi&o, S. (1972) Anal. Biochem. 48,27-32 Birnstiel, M., Telford, J., Weinberg, E., and Stafford, D. (1974)

Proc. Natl. Acad. Sci. U. S. A. 71, 2900-2904 Suzuki, Y., Gage, L. P., and Brown, D. D. (1972) J. Mol. Biol. 70,

637-649 Lizardi, P., and Brown, D. D. (1974) Cold Spring Harbor Symp.

&ant. Biol. 38, 701-706 Rosen. J. M.. Woo. S. L. C.. Holder. J. W.. Means, A. R.. and O’Mailey, B. ‘W. (1975) Biochemistry’ll, 69-78 Woo. S. L. C.. Rosen. J. M.. Liarakos. C. D.. Robberson, D. L..

Choi, Y. C.; Busch; H., Means, A. ‘R., anh O’Malley,’ B. WI (1975) J. Biol. Chem. 250.7027-7039

Monahan, J. J., Harris, S. E., Woo, S. L. C., Robberson, D. L., and O’Malley, B. W. (1976) Biochemistry 15, 223-234

Means, A. R., Abrass, I. B., and O’Malley, B. W. (1971) Biochem- istrv 10. 1561-1570

Mar&, j. (1961) J. Mol. Biol. 3,208-218 Rosen. J. M.. Harris. S. E.. Rosenfeld. G. C.. Liarakos. C. D.. and

O’calley, B. W. (1974) Cell Diff. 3; 103%i16 Saxinger, W. C., Ponnamperuma, C., and Gillespie, D. (1972)

Proc. N&l. Acad. Sci. U. S. A. 69, 2975-2979 Shih, T. Y., and Martin, M. A. (1974) Biochemistry 13,3411-3418 Shih, T. Y., and Martin, M. A. (1973) Proc. Natl. Acad. Sci.

U. S. A. 70,1697-1700 Young, B. D., Harrison, P. R., Gilmour, R. S., Birnil, G. D., Hill,

A., Humphries, S., and Paul, J. (1974) J. Mol. Biol. 84,555-568 Ross, J., Gielen, J., Packman, S., Ikawa, Y., and Leder, P. (1974)

J. Mol. Biol. 87, 697-714 Smith, M. J., Britton, R. J., and Davidson, E. H. (1975) Proc.

Natl. Acad. Sci. U. S. A. 72, 4805-4809 Britten, R. J., Graham, D. E., and Neufeld, B. R. (1974) Methods

Enzymol. 29, 363-418 Davidson, E. H., Hough, B. R., Klein, W. H., and Britten, R. J.

(1975) Cell 4, 217-238 Rosen, J. M., Liarakos, C. D., and O’Malley, B. W. (1973)

Biochemistry 12, 2803-2809 Kedes, L. H., Chang, A. C. Y., Houseman, D., and Cohen, S. N.

(1975) Nature 255.533-538 Glover, H. M., White, R. L., Finnegan, D. J., and Hogness, D. S.

(1975) Cell 5, 149-157 Marrow, J. F., Cohen, S. N., Chang, A. C. Y., Boyer, H. W.,

Goodman, H. M., and Helling, R. B. (1974) Proc. Natl. Acad. Sci. U. S. A. 71,1743-1747

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from

S L Woo, R G Smith, A R Means and B W O'MalleyThe ovalbumin gene. Partial purification of the coding strand.

1976, 251:3868-3874.J. Biol. Chem. 

  http://www.jbc.org/content/251/13/3868Access 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/251/13/3868.full.html#ref-list-1

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

by guest on July 18, 2020http://w

ww

.jbc.org/D

ownloaded from