organization of highly repeated sequences in mouse main-band dna · 2009-05-17 · j. mol. biol....

J. Mol. Biol. (1976) 100, 227-256

Organization of Highly Repeated Sequences in Mouse Main-band DNA

THOMAS R. C~CH t AND JOHN E. HEAEST~:

Department of Chemistry University of California

Berkeley, Calif. 94720, U.S.A.

(Received 19 May 1975, and in revised form 7 October 1975)

Ronaturat ion analysis shows that the main CsC1 density band of mouse DNA (at 400 bases mol. wt) contains a min imum of four second-order kinetic components : non-repetitive DNA; two major classes of repeated DNA, called "moderately repeated" and "highly repeated", each comprising 8 to 16~/o of the genome; and a minor sub-class of the highly repeated DNA, about 0.5% of the genome with a satellite-like complexity. The highly repeated main-band sequences are isolated from high molecular weight DNA (12,000 bases) by renaturat ion to low Cot w and hydroxyapati te chromatography. This DNA is spread for electron microscopy by the formamide isodenaturing technique. Most of the bimoleeular renaturat ion products show interspersion of renatured and single-stranded regions. The renatured regions average about 700 base-pairs in length, most being in the range of 200 to 1000. On molecules where two or more renaturod regions can be identified, they are separated by a wide range of distances averaging 1800 bases. A minori ty fraction (about 12% by weight) of the bimolecular rena tura t ion products shows no evidence of interspersion, and may consist of tandemly repeated sequences.

Tile renaturat ion rate of the higllly repeated DNA fraction (at initial mol. wts of 6000 and 12,000 bases) is studied before and after shearing the DNA to low molecular weight. The kinetic complexity of the interspersed, single-stranded DNA that is at tached to renatured, highly repeated DNA is thereby determined. Most of the unrenatured DNA contains highly repeated sequences similar in complexity to the sequences that renatured at low Cot. Only about 20% of the interspersed DNA has a low repetition frequency, approximately non-repetitive. We conclude that about 15% of the mouse main-band DNA contains highly repeated sequences, most of which are organized as permuted tandem repeats. About 5O/o of the non-repetitive DNA of mouse is interspersed with this fraction of the genome. This model of sequence organization differs from those that have been demonstrated for more moderately repeated sequences of higher organisms, where much more non-repetitive DNA is interspersed with the repeated sequences.

Present address: Department of Biology, Building 16, Room 717, Massachusetts Institute of Technology, Cambridge, Mass. 02139, U.S.A.

:~ To whom reprint requests should be sent. w Abbreviations used: Cot, product of initial DNA tenon (mol. DbIA PO4/l) and time (s); equiv.

(equivalent) Cot, Cot corrected to the standard renaturation condition of 0.12 M-phosphate buffer. This definition of equiv. Co t differs from that used in Cech et al. (1973), where Cot w a s corrected for both salt concentration and DNA molecular weight to give equiv. Cot; ds, double-stranded; ss, single-stranded.

1~ 227

228 T . R . CECH AND J. E. HEARST

1. Introduction

Largely through renaturation kinetics studies of the DNA of a wide variety of eukaryotic organisms, two general patterns of organization of repeated DNA sequences are emerging. Satellite DNAs are composed of simple (oligomeric) sequences tandemly repeated, with variable accuracy, hundreds or thousands of times without interruption (Waring & Britten, 1966 ; Bond et al., 1967 ; Southern, 1970; Gall & Atherton, 1974; Goldring et al., 1975; Walker, 1971 (review)). In the main CsCI density bands of all eukaryotic DNAs that have been studied, middle repetitive sequences are interspersed with non-repetitive sequences (see reviews by Davidson & Britten, 1973; Lewin, 1975).

Main-band DNA usually contains highly repeated as well as moderately repeated sequences. Some human and Drosophila main-band DNA components are simple sequence DNAs that occur in heterochromatie regions of chromosomes; these differ from satellites by base composition only (Corneo et al., 1971 ; Jones & Corneo, 1971 ; Marx, Allen & Hearst, 1976; Peacock et al., 1974). Other highly repeated main-band sequences in mouse and human DNA (Cech et al., 1973; Corneo et al., 1970; Marx et al., 1976) appear to be interspersed with other sequences rather than tandemly repeated. I n situ hybridization shows that these highly repeated sequences are not restricted to centromeres. The average repetition frequency is intermediate between that of simple sequence DNAs and the major middle repetitive component of, for instance, Xenopus; to minimize confusion, they will be called "highly repeated" main-band sequences. The same sequences, when isolated in renatured form, have been called "h.a.r .r .DNA" for hydroxyapatite-isolated rapidly renaturing DNA.

In our earlier study (Cech et al., 1973), we analyzed the hydroxyapat i te yield, CsC1 buoyant density, Sl-nuclease resistance, and hyperchromicity of renatured, highly repeated mouse main-band DNA as a function of DNA fragment size. We concluded that repeated sequence DNA of average length 1-5-4-0.5 kb~ is interspersed with other DNA of average length 2.24-1.1 kb. In the present study, electron microscopic observation of the renatured molecules gives direct confirmation of this interspersion model for most of the highly repeated DNA. Electron microscopy gives the length distribution of the renatured segments as well as the average length. Hydroxyapat i te renaturation studies result in a better estimate of the repetition frequency of the highly repeated DNA and show tha t it differs from more moderately repeated DNA. Most significant is the unexpected finding that the highly repeated DNA is interspersed with other highly repeated sequences, as well as with some non-repetitive DNA. Different highly repeated sequences seem to be organized in a permuted tandem repeat arrangement.

2. Materials and Methods

(a) Mouse D N A

SVT2 mouse tissue culture cells were originally obtained from Theodore Gurney, Jr (University of Utah). The cells are transformed by simian virus 40, and they have a di- ploid number of chromosomes. The labeling, extraction and purification of DNA from these cells has been described previously (Ceeh et al., 1973). Unlabeled Balb/c mouse DNA

t Abbreviations used: kb (kilobase), 1000 bases of ss or 1000 ba~e-pairs of ds DNA; HAP, hydroxyapatito,

H I G H L Y R E P E A T E D M O U S E M A I N - B A N D D N A 229

was e x t r a c t e d f r o m c o m b i n e d l ivers , spleens , t e s t e s a n d b ra ins . T h e p r o c e d u r e fol lowed t h a t desc r ibed b y F l a m m etal. (1966), w i t h t h e a d d i t i o n of a P r o n a s e t r e a t m e n t before t h e f inal d e p r o t e i n i z a t i o n . P o l y s a c c h a r i d e s were r e m o v e d b y a 30-ra in c e n t r i f u g a t i o n a t 45,000 r e v s / m i n in t h e B e c k m a n SW50.1 ro tor .

(b) E s c h e r i c h i a coli D N A

Frozen , a l l - l abe l ed E. coli ( s t r a in H F 4 7 3 3 F - thy- gal- E n d o I - ) , o r ig ina l ly a gif t of A. J o h n Clark ( U n i v e r s i t y of Cal i fornia , Berke ley) , were o b t a i n e d f r o m D a v i d A p p l e b y (Berkeley) . T h e cells were t h a w e d a n d t r e a t e d w i t h lysozyme, a f t e r w h i c h t h e D N A was e x t r a c t e d a n d pur i f i ed b y a m e t h o d s imi la r to t h e SVT2 m o u s e D N A m e t h o d . T h e meas - u r e d specific activity( was 1.3 • 10 s e t s / m i n pe r ~g w i t h D N A c o n c e n t r a t i o n d e t e r m i n e d b y a n a l y t i c a l CsC1 d e n s i t y g r a d i e n t e e n t r i f u g a t i o n of t h e u n s h e a r e d E. eoli D N A w i t h a k n o w n a m o u n t of m o u s e sa te l l i t e D N A . Th i s m e t h o d o b v i a t e d t h e poss ib i l i ty t h a t t h e m e a s u r e d op t i ca l d e n s i t y o f t h e D N A w o u l d inc lude a c o n t r i b u t i o n f r o m s ~ a t t e r l h g or c o n t a m i n a t i n g R N A , r e s u l t i n g in a n u n d e r e s t i m a t e d specific a c t i v i t y .

(c) Shearing of DNA and determination of molecular weights

Our p rocedure s for shea r ing n a t i v e D N A h a v e b e e n desc r ibed p rev ious ly (Cech et al., 1973). Brief ly , need le s h e a r i n g was u sed to p r o d u c e D N A of 5 to 12 k b s ing l e - s t r and molecu la r w e i g h t a n d s o n i c a t i o n for mo lecu l a r we igh t s of 0"3 to 0.4 kb. Molecu la r we igh t s were d e t e r m i n e d b y a n a l y t i c a l b o u n d a r y s e d i m e n t a t i o n in a lka l ine so lu t ion , us ing t he e q u a t i o n s of S tud ie r (1965). T h e s e e q u a t i o n s a re b a s e d on d a t a a t m o l e c u l a r we igh t s a b o v e 8 kb; so t h e i r use for t h e s o n i c a t e d s a m p l e s is a n e x t r a p o l a t i o n .

I n s h e a r i n g r e n a t u r e d D N A , i t is i m p o r t a n t to k n o w w h e t h e r d o u b l e - s t r a n d e d a n d single- s t r a n d e d reg ions are b e i n g r e d u c e d to s imi la r f r a g m e n t sizes. There fore , b o t h n a t i v e a n d d e n a t u r e d h igh m o l e c u l a r w e i g h t m o u s e D N A samples were s h e a r e d b y two m e t h o d s , a n d s i n g l e - s t r a n d e d m o l e c u l a r we igh t s d e t e r m i n e d . As s h o w n in T a b l e 1, r a p i d d e c o m p r e s s i o n

TABLE 1

Shearing of D N A

Single-stranded mol. wt (kb) Nat ive DNA Dena tured DNA

Bomb 400 lb / in : 1.2 1"I 1200 lb/in 2 1-1 0.7

Sonicate 0-30 r~-PO4 0-33 0-54 0.05 ~ - P 0 4

Driver 1 0.32 Driver 2 0.40

All molecular weights were determined on single-stranded DNA; "na t i ve" and " d e n a t u r e d " refer to the s ta te of the DNA before shearing. I n shearing DNA in the decompression bomb, each sample was equi l ibrated with argon for 10 rain a t the indicated pressure and then rapidly decom- pressed as i t exited from the bomb th rough a needle valve. Sonication of n i t rogen-sa tura ted solutions was done in a n ice b a t h in 5 30-s blasts. All samples wore sheared in 0"05 M-phosphate buffer except where indicated. They were then dialyzed into 1 M-I~aC1 and made 0.1 M in NaOH for the sedimenta t ion analysis. The unsheared single-stranded mol. wt was' a t least 40 t imes the sheared value in each case. "Dr iver 1" is the Balb/c DNA used in the dr iven rena tu ra t ion experi- men t of Table 4; "dr iver 2" is t h a t used in all t he exper iments summarized in Fig. 6. Dr iver 2 was sonicatod in a much larger volume t h a n was used in the previous 2 experiments, p robably account ing for its higher mol. wt. The relat ive resistance of dena tured DNA to shear by sonication may be due to the higher ionic s t rength used in t h a t exper iment ; denatured DNA in 0.30 M- phospha te buffer was found to be quite res is tant to shear by the bomb method (data no t shown).

230 T . R . C E C H A N D J . E . H E A R S T

in a b o m b and sonicat ion sheared single- and double-s t randed D N A to somewha t different extents . B o t h techniques , however , were adequa te for our purposes.

(d) Isolation of renatured D N A

I ) N A was alkali dena tu red a t r o o m t empera tu re , w a r m e d to 60~ and neut ra l ized wi th NaH2PO4 such t h a t the final buffer was 0-12 M-phosphate buffer ( K r a m et al., 1972). Phospha t e buffer will be used to indicate an equ imola r m ix tu r e of N a 2 H P 0 4 and NaH2PO4 wi th a p H of 6"8. Af te r incuba t ion at 60~ to the desired Cot, r ena tu red D N A was isolated by ba tch h y d r o x y a p a t i t e c h r o m a t o g r a p h y (Cech et al., 1973). Unless s t a ted otherwise, the cr i ter ia for base-pair ing were of "h igh s t r i ngency" - -0 .12 , 0.16 and 0.45 M-phosphate buffer elut ions were pe r fo rmed a t 70~ "Less s t r i ngen t " cr i ter ia (omi t t ing the 0.16 ~- phospha te e lut ion and e lut ing a t 60~ were somet imes used, especial ly in isolat ing foldback DNA. The f rac t ion of the D N A rena tu red (fract ion bound to H A P ) was ca lcu la ted as the to ta l r ad ioac t iv i ty or opt ical dens i ty uni ts e lu ted a t 0.45 M-phosphate d iv ided by the to ta l recovered a t all 3 phospha t e concentra t ions . Recover ies averaged 100 ~: 5% of the init ial c ts /min.

(e) Hydroxyapatite assay of renaturation kinetics

The use of an excess of sonicated, unlabeled D N A to dr ive the r ena tu ra t ion of a t race a m o u n t of [aHJDNA, moni to r ing r ena tu ra t ion by H A P binding, has been ex tens ive ly descr ibed by Br i t t en and his co-workers (Br i t ten & Smith , 1970; Dav idson et al., 1973),

l cm uni ts of soni- and our m e t h o d was t aken f rom theirs. E a c h sample conta ined 40.D.~60 ,,,n cared to ta l Ba lb /c mouse D N A (0.32 to 0.40 kb) to dr ive the rena tura t ion . W h e n this D N A was ch roma tog raphed on H A P wi thou t dena tu ra t ion , 99"0~/o e lu ted in the double -s t randed fract ion. D N A concent ra t ions neve r exceeded 50 O.D. un i t s /ml (2"5 mg/ml) , t he reby avoid ing high viscosities. E a c h sample also con ta ined 2000 to 12,000 c t s /min of 3H- labeled SVT2 t racer D N A at 4.0 • 104 or 2.65 • l0 s c t s /min per ~g. The ra t io of dr iver to t racer was 1500 in the Tab le 4 exper iment , and ranged f rom 2500 to 22,000 for the rest. Fo r r ena tu ra t ion t imes of less t han 1 h the s~mples were dena tu red and r ena tu red as descr ibed in section (d). F o r longer t ime points, samples were dena tu red wi th N a O H a t r o o m tempera tu re , neut ra l ized ~dth ice cold N a H 2 P 0 4 , sealed into glass ampoules or capillaries, and hea ted quick ly to 6O~ by submers ion in a wa te r bath. Fo r ve ry low Cot values ( ~ 1 0 - 6 ) , sonica ted Micrococcus lysodeikticus D N A was subs t i tu ted for the Ba lb /c mouse D N A , wi th Cot based on the concen t ra t ion of 3H-labeled mouse D N A only. Some ve ry high Cot samples were r ena tu red in 0.90 ~1-phosphate buffer a t 70~ (to m a i n t a i n the tm -- 25~ criterion), in which case the Cot was mul t ip l i ed by 8"06 (Br i t ten et al., 1974) to give equiv. Cot. A t the end of the r ena tu ra t ion t ime, small vo lume samples were d i lu ted to give 4 ml wi th a final concen t ra t ion of 0.12 M-phosphate. Samples were neve r frozen, bu t were e i ther ch roma tog raphed immed ia t e ly or d i lu ted and s tored on ice. H A P frac- t iona t ion was done by the ba tch m e t h o d as descr ibed in sect ion (d), using the less s t r ingen t e lut ion condit ions. (Al though 70~ and 0" 16 M elut ions can be useful me thods of select ing for good hybrids , the i r use in a hybr id iza t ion assay would effect ively e l iminate the possi- b i l i ty of some molecules ever being scored as r ena tu red , because the assay cr i ter ion would be more s t r ingent t h a n the hybr id iza t ion cri terion.) The vo lumes of the 8 e luants were measured and the fract ions were cent r i fuged a t 2000 g for i0 rain to pel le t any finely suspended H A P . Opt ical densit ies were measured a t 260 n m and 340 n m a t room t empera - ture in t he Cary 15 spec t ropho tomete r , using h y d r o x y a p a t i t e - w a s h e d phospha te buffers (which absorb less t h a n unwashed buffers) as reference solutions. R e c o v e r y of O.D. uni ts ranged f rom 111 -{- 4~/o af ter 8-min incubat ions to 97~ af ter 100 h incubat ions . The

100~o recoveries are due to t he hype reh romic i ty of bo th s ingle-s t randed and r e n a t u r e d D N A . Af te r opt ical dens i ty measurements , 3.75 ml of each 0.12 M-eluant, or 1 In] of a 0.45 M-eluant d i lu t ed wi th 2-75 ml of water , was dissolved in 15 ml of cocktai l (1 pa r t T r i t on X l 0 0 : 2 par t s P P O - P O P O P in toluene) and coun ted for 5 to 10 min in a B e c k m a n 200 l iquid sc int i l la t ion counter . R e c o v e r y of r ad ioac t iv i t y ranged f rom 1004-4~ af ter shor t incubat ions to 88-4-4~ af ter long t imes. Subsequen t to the expe r imen t of Tab le 4, all r ena tu ra t ion samples con ta ined 2 • 10-7 mol E D T A . This gave an E D T A concen t ra t ion

H I G H L Y R E P E A T E D MOUSE MAIN-BAND DNA 231

of 1 mM in 0.2 ml samples, which were incubated for the longest times, and a proportionally lower concentration in larger volume samples. Because the 0.2 ml samples were diluted 20-fold before HAP fractionation, it is not surprising that no effect of the EDTA on the fractionation or the O.D. readings was noticed.

(f) Electron microscopy

Formamide spreading of DNA was done using the isodenaturing technique of Davis et al. (1971 ). The formamide concentration in the spreading solution was 45~/o or 50~/o. At room temperature, the denaturing ability of these spreading solutions is equivalent to 62 and 66~ respectively, in 0-12 M-phosphate buffer. Circular phage DNA was added to some preparations before spreading to give internal length standards. Double-stranded phage PM2 DNA was taken to have a molecular weight of 9-4 kb, and single-stranded fd DNA a molecular weight of 6-4 kb. Further details of our electron microscopic methods are given in Cech & Hearst (1975).

3. Results

(a) Isolation of highly repealed sequences from mouse main-band DNA

Our procedure for isolating highly repeated sequences from the main CsCl buoyant density class of mouse DNA is described in Cech et al. (1973). First, the simple sequence light satellite DNA is removed by Ag § density gradient centrifugation. The main-band DNA is then sheared, to a molecular weight of 6 or 12 kb in the experiments described here. Next, fotdback DNA (Walker & McLaren, 1969; Brit ten & Smith, 1970) is removed by denaturing the main-band DNA, renaturing it for one and a half to two minutes at low concentration (Cot ~ 10 -~ to 10-3), and removing strands containing base-paired regions by hydroxyapat i te (HAP) chromatography. The foldback DNA is largely the result of intramolecular renaturat ion of inverted repeat sequences (Wilson & Thomas, 1974; Cech & Hearst , 1975). Our method of isolating foldback DNA removes 90% of the HAP-bindable inverted repeat sequences tha t can be identified in the electron microscope. Very little bimolecularly renaturing DNA is contained in this fraction.

The DNA not bound in the first H A P chromatography (i.e. the single-stranded fraction) is further incubated at 60~ in 0-12 M-phosphate buffer to a Cot ranging from 4 • 10- 3 to 8 • 10- 2, when bimolecular renaturat ion products are isolated by a second H A P chromatography. Depending on the DNA molecular weight and the Cot, 4 to 10% of the main-band DNA is collected in the second H A P chromatography. (For a more systematic s tudy of the H A P yield of this DNA, see Cech eta/., 1973.) Because the fraction of the DNA bound to H A P increases continuously with Cot in this range, the procedure does not result in the isolation of a whole class of repeated sequences, l~ather, it selects for the most highly repeated sequences. Da ta presented in sections (d) and (e) will demonstrate tha t the isolated sequences are in fact highly repeated, not simply the first moderately repeated sequences tha t happen to renature. This data will also show tha t the kinetic class or classes represented by the isolated highly repeated sequences make up 8 to 16% of the mouse genome.

The highly repeated DNA bound in the second H A P chromatography usually undergoes a further kinetic purification. The DNA is diluted to a phosphate concen- t ra t ion of 0"05 M, alkah denatured, neutralized and renatured in 0.12 M-phosphate in the same volume and for the same t ime as the previous renaturation. Double-stranded DNA is then collected on HAP. (This is the third H A P chromatography.) This pro-

232

(o) Isolated interspersed

T. R . O E C H A N D J . E . H E A R S T

i

ii ~ t l i l l . . . . .

(b) Alternating interspersed i + ii + iii + iv

I l l [ I I ] ] ~ i %

/

vi ;, ,- . . . . . . . ' c:z=:~.

(c) Tandem repeated RI RI RI RI R I vii

viii ~ ~ ! = l l , i

(d) Permuted tandem repealed RI Rz Rs R2 R4

Highly repealed DNA

More complex DNA

ix Rt ~R2 1 1 1 1 1 ]

R'2 R~, +

Rz R5 xi R ~

a'4 R'~

+ others like those in (a), (b), (c)

FIG. 1. Models of highly repeated DNA sequence organization and their predicted rena tura t ion products. The left half of the Figure shows 4 ways in which repeated sequences could be organized in the nat ive DNA. Each arrow represents random shearing of the D)~TA, denaturat ion, and rena tura t ion to a Co t sufficient for the reassociation of highly repeated sequences. The r ight half of the Figure shows some of the resulting types of rena tm'a t ion products, wi th rena tu red regions indicated (I I I I I I I). I n most c~os , more complex s t ructures would form wi th subsequent nucleat ion events. (a) The repeated sequence ( ) is interspersed wi th more complex DNA ( -) and is isolated from other repeated sequences, such t h a t a shear f ragment will contain.only one repeated sequence region, The expected rena tura t ion products are (i) l inear ds/ss molecules (partly double-stranded and par t ly single-stranded), (ii) single-forked molecules, and (iii) double-forked molecules. (b) Regions of repeated and more complex DNA al ternate, such t h a t a t moderately high molecular weights m a n y shear f ragments will conta in 2 or more


cedure is called "recycling". The highly repeated DNA isolated after the recycling step amounts to only 0.6 to 1"5~0 of the main band DNA (6.7 to 9 .9~ isolated as double-stranded at Cot 6• 10 -2, and 9 to 15% of this reisolated after recycling at Cot 6 x 10-3)t. Although the recycling provides a rigorous purification of the DNA according to its renaturation kinetics, it seemed dangerous to s tudy such a small fraction of the genome and attribute its properties to a larger class. Therefore, a portion of the unrecycled highly repeated DNA (isolated on the second HAP chromatography) was often saved and characterized along with the recycled fraction.

(b) Models of sequence organization

When single-stranded DNA containing highly repeated sequences is renatured and isolated on hydroxyapatite, the types of structures that are formed should be in- dicative of the way the sequences are organized in the native DNA. Figure 1 diagrams the primary (bimoleeular) and secondary renaturation products expected for four possible types of sequence arrangement. In many cases the structures contain regions of highly repeated sequence that are not base-paired, so more complex structures should form upon prolonged renaturation.

Models (b) and (d) in Figure 1, in which stretches of highly repeated DNA alternate with stretches containing different sequences (either highly repeated or more complex), are consistent with our previous data for mouse main-band sequences (Cech eta l . , 1973). We preferred model (b) because, even when incubated to an equivalent Co t of 50, the renatured DNA retained a considerable amount of single- strandedness. In our earlier work, however, we made no direct determination of the complexity of the interspersed, single-stranded sequences.

There are 3 reasons for this low recovery: (1) much of the DNA isolated at Co t 6• 10 -= consists of sequences with Cot t > 6 • 10 - 2. Each recycling step will therefore lead to the renaturation of less than one-half of such sequences. (2) The second HAP chromatography is done under conditions of low stringency (see Materials and Methods), but the third is done at high stringency, resulting in the loss of marginally-stable duplexes. (3) Single-strand breaks occur during the alkali denaturat ion steps, 60~ incubations, and/or HAP ehromatographies. These breaks release more- complex sequences from a t tachment to highly repeated sequences, preventing the former from binding in the double-stranded fraction upon recycling. Reduction of the size of the highly repeated sequences also slows their renaturation, as evidenced by data presented in section (f) of Results.

repeated sequence regions. The renaturation products include those of model (a), but in addition several types of multiple-forked structures (iv, v, vi) can form. I f neighboring repeated sequence elements are similar, or if they are different but often occur in the same order, then s t ructures (iv] and (vi) will be common. Structures like (v) give no information about the order of repeateit sequence elements. They could result from a second nucleation on a single- or double-forked molecule, or from a single-strand break in structure (iv) or (vi). (e) Identical or closely related sequences are tandemly repeated. Many satellite DNAs are organized in this manner. The renaturation products do not contain structures in which a double-stranded region forks into 2 single- s t randed tails, which distinguishes this model from the others. Subsequent nucleation events produce branched structures (viii) and hyperpolymers (Britten e$ al., 1974). (d) Repeated sequences are randomly arranged, but the various sequence elements do not always follow each other in the same order. I t is not necessary tha t adjacent sequence elements be unrelated, but only tha t they be different enough to prevent base-pairing under the hybridization criteria used. Most of the pr imary and secondary renaturation products predicted for this model would look identical in the electron microscope to those of model (b): (ix) looks like (i), (x) looks like (iii), (xi) looks like (iv), and so on. Although one can conceive of structures tha t would be unique to the permuted tandem repeated model, these involve many forks and would therefore probably be identified only as "messes"- -s t ruc tures tha t are not unambiguously traceable.

234 T. R. CECH AND J. E. H E A R S T

(c) Electron microscopy of the highly repeated DNA

(i) Types of molecular structures observed

H i g h l y r e p e a t e d mouse m a i n - b a n d D N A was i so la ted as descr ibed in sect ion (a), above , a n d sp read for e lec t ron mic roscopy in 45 or 50% formamide . A t room tem- p e r a t u r e in these sp read ing solutions, base -pa i red regions of DI~A were more s t ab le t h a n in t he 70~ 0.12 ~ - p h o s p h a t e b u f f e r - H A P e lu t ion (see Mate r ia l s and Methods , sec t ion (e)). F u r t h e r m o r e , in the op t ica l mel t ing profile of th is class of DNA, mos t of t he h y p e r c h r o m i c i t y occurred above 70~ (Cech et al., 1973). Therefore, t he spread ing was no t expec ted to dena tu re the r e n a t u r e d regions t h a t were responsible for t he D N A elut ing f rom H A P in t he doub l e - s t r a nde d f ract ion.

The h igh ly r e p e a t e d D N A frac t ions con ta ined the t y p e s of molecules shown in F igures 2 a n d 3. Molecules were classified on the basis of the n u m b e r of forks and free ends in each s t ruc ture , as descr ibed in t he legend to F igu re 2. The double- s t r a n d e d regions of r e n a t u r e d molecules could usua l ly be ident i f ied b y the fac t t h a t t h e y were more dense ly s t a ined and less k i n k y t h a n s ing le - s t randed regions. The mos t diff icul ty arose in specifying the s t r andedness of shor t regions, because the re was less d i s tance over which the s t r a n d dens i ty could be averaged . Conversely , s ing le - s t randed phage fd D N A and doub l e - s t r a nde d PM2 DNA, a d d e d as molecu la r weight s t anda rds , con ta ined long regions of the same s t r andedness ; these D N A s could a lways be pos i t ive ly d i f fe ren t ia ted b y the cr i ter ia of s t r a n d de ns i t y and kinkiness .

A s u m m a r y of t he t y p e s of molecules seen is g iven in Tab le 2. A t in i t ia l D N A molecular weights of b o t h 6 and 12 kb , and in t he un recyc led as well as recyc led p repa ra t ions , the larges t class of s t ruc tu res is the forked class, mak ing up to 32 to 410/o of t he samples . I n addi t ion , m a n y of t he messes p r o b a b l y con ta in h idden b imolecu la r forks. The low p ropor t i on of l inear (ds/ss) s t ruc tu res suggests t h a t mos t of the mouse m a i n - b a n d h igh ly r e p e a t e d sequences are no t a r r anged as s imple t a n d e m repea t s (Fig. 1, model (c)). A n y mouse sa te l l i te D N A c o n t a m i n a t i n g the m a i n - b a n d D N A would no t con t r ibu te s igni f icant ly to e i ther t he forked or l inear categories , because sa te l l i te forms d i s t inc t ive mess s t ruc tures even when r e n a t u r e d to low Cot values (Cech, 1975).

Tab le 3 subdiv ides t he forked class in to the t y p e s of s t ruc tu res d i a g r a m m e d in F igu re 2. Compar i son of the observed molecules and those expec ted for t he var ious models in F igu re 1 suggests t h a t a l t e rna t ing in te r spers ion and /o r p e r m u t e d t a n d e m

stranded (ss) tails at both ends. Arrows indicate approximate junctions of ss and ds regions. Manning et al. (1975) estimate that, on an unforked molecule, the accuracy with which the junctions between single- and double-stranded regions can be specified is "perhaps ~0"5 kb"; we agree with this value. Molecules i, vii, and ix of Fig. 1 would be in this category. (b) Single-forked molecule, one renatured (ds) region plus three ss tails. See also molecule ii of Fig. 1. (c) Double- forked molecule, 1 ds region plus 4 ss tails. See also molecules iii and x of Fig. 1. (d) Multiple-forked molecule, 2 (or more) ds regions, each pair joined by 1 single-strand. See also molecule v, Fig. 1. (e) Multiple-forked molecule, 2 (or more) ds regions, each pair joined by 2 single strands. See also molecules iv and xi, Fig. 1. (f) Complex forked structure, contains at least 1 clear bimolecular fork, but the whole structure is not unambiguously decipherable. The bottom region of this example could be much more tangled and the molecule would still be scored in this category. (g) Complex forked structure that includes a "branch", defined as an ss tail in the middle of a ds region. Branches may result from the type of renaturation diagrammed in Fig. 1 (viii). I f they occur on a structure with many tails, as in this example, identification of the renaturatien events often becomes ambiguous, so such structures are scored as "complex". (h) Hairpin, an intramolecular renaturation product with two ss tails and one or more ds "stem" regions. This example shows a looped hairpin; about 40% of the hairpins were unlooped.

Linear ds/ss

Single fork


(a)

Double fork

Multiple f e)

C

\ Complex

L

Hairpin

FIe. 2. Examples of the types of rena tured molecules seen in the highly repeated ma in -band DNA fraction. These molecules were on the same grids used for the measurements of Fig. 5(a), (b) and (d). The bar has a length corresponding to 1.0 kb of double-s tranded or 1"2 kb of single- s t randed DNA. In the in terpre ta t ion to the left of each photo, presumed double-s t randed regions are indicated by heavy lines, s ingle-stranded regions by th in lines, and regions of uncer ta in ty by dashed lines. (a) Linear ds/ss molecule, an unforked bimolecular r ena tu ra t ion p roduc t wi th single-

236

(a)

T. R. C E C H A N D J . E . H E A R S T

.~,~,. ~.., :!~.~.'~-~:.':',~:-,.~'~.~ "" , .-'... ~:~-',~4~ .,-~....~':;,. ~-~2";.';,." :-'.~ .--~';.? - .-7',~-' ~k ."',~."~- ~-""-i..." -~ ~.'-.-; ~: . -~ ' :'...'

(c) (b)

FIG. 3. Highly repeated main-band D N A isolated on hydroxyapat i te and spread for electron microscopy in 45% formamido, as in Fig. 2. The bar again represents a length of 1"0 kb (ds). In the interpretation, double- and single-stranded regions are indicated as in Fig. 2.

(a) A molecule of the "multiple fork" category plus 3 single-stranded fd D N A circles. Although the presumed ronatured region on the right is too shor~ to appear double-stranded, the low

H I G H L Y R E P E A T E D MOUSE MAIN-BAND DNA 237

repetition (models (b) and/or (d)) may be the predominant t~Tes of sequence organization. The high proportion of multiple-forked and complex structures (Fig. 4) would not be expected if the repeated sequence elements were isolated in the genome (model (a)). As expected for molecules containing several segments of repeated DNA, more forked molecules are in the form of complex structures at the higher molecular weight. An increased number of forked molecules are also probably hidden in messes, which would explain the apparent decrease in the frequency of forked structures (from 40 to 32%) with higher molecular weight.

The single-stranded (ss) molecules contain no discernable base-paired regions. Because it is difficult to identify short double-stranded regions, some of these molecules could be ds/ss bimolecular renaturation products. Most ss molecules are short, however, and therefore probably result from breaks in the single-stranded tails of forked molecules or hairpins, during or after HAP fractionation.

( concentration of mouse DNA on this grid makes it unlikely that it is simply a cross-over of 2 strands (see text). (b) A molecule of the "complex" category plus 2 fd DNA circles. (c) Single- stranded fd DNA (above) and double-stranded PM2 DNA (below), added to the mouse DNA as molecular weight standards. As discussed in the text, the strandedness of these DNAs was always clearly differentiable.

238 T. R . C E C H A N D J . E. H E A R S T

TABLE 2

Electron microscopic categorization of molecules in hydroxyapatite-fractionated DNA

HAP fraction % of main- % of molecules band DNA n Forked ds]ss ss Hairpin Mess

6 kb DNA Highly repeated DNA

(7x10 - 4 < C o t < 6 x 10 -2, 1. string.) 9"9 274 404- 1 84-2 254-2 74-0 19/= 1

Same, recycled at Got 6x 10 -a (h. string.) 1.5 388 414-2 384-0* 104-1 114-3

Unbound 77.8 283 5 0 83 8 3 12 kb DNA

Highly repeated DNA (3x10 -4 < Cot < 4 x 10 -3, h. string.) 3.9 550 324-2 74-1 294-3 154-1 174-1

Unbound N 80 583 5 0 78 10 7

The mol. weights are initial values determined by alkaline sedimentation. After the 1st HAP (foldback DNA) fractionation, number average mol. weights (/~n) were determined by electron microscopy. -~n was 5.3 kb for the 6 kb DNA and 10 kb for the 12 kb DNA. -~n for forked molecules in the 12 kb highly repeated DNA fraction (which has now been through 2 renaturat ion and HAP fraetionation steps) was 7.2 kb per strand. A < Co t < B means tha t the fraction was isolated on HAP at Co t B after being stripped of foldback DNA by a prior HAP fractionation at Oct A. The stringency of the HAP fraetionation criteria, defined in section (d) of Materials and Methods, is designated as "h. string." for high stringency and "1. str ing." for less stringent. The "unbound" fractions contain the DNA tha t eluted as single-stranded after the 2nd HAP fraetionation.

HAP fractions were dialyzed and spread for electron microscopy in 50~o formamide. To avoid possible confusion, no circular phage DNAs were added. "n" is the number of random molecules scored. Small fragments ( < 1 kb) are excluded from the scoring. The "forked" category includes all of the molecules shown in Fig. 2 except the linear and hairpin types. "Hairpins" are defined in Fig. 2; see also Wilson & Thomas, 1974, and Ceeh & Hearst , 1975. I f a hairpin is seen on the tail of another structure (forked or mess), it is scored in one of the lat ter categories only. "Messes" cannot be unambiguously traced, usually because of multiple renaturat ion events and crossovers. I f such a structure includes an obvious bimoleeular fork (Fig. 2(f)), it is scored in the forked category. Circles and lariats (Thomas et al., 1970) make up 1 to 4% of the molecules in the renatured DNA fractions, and are not listed in the Table. The " 4 - " numbers measure reproducibility of scoring, and do not include all sources of error. They give the maximum difference found when two grids of the same DNA preparation were scored, or when one grid was scored on 2 different days. On one grid (*) the discrimination between double- and single-stranded regions was not good enough to positively differentiate ds/ss from ss molecules, so the total number of linear structures is listed.

TABLE 3

Types of forked structures in renatured highly repeated DNA

Highly repeated % of forked structures DNA fraction n Single Double Multiple Complex Branch

6 kb unrecycled 108 284-2 24=h9 1 3 • 254-3 94-1 12 kb unrecycled 175 204-4 254-3 94-3 394-1 74-1

This Table subdivides the forked category of Table 2 for 2 of the DNA fractions. "n" is the number of forked structures scored. The types of forked structures are defined in Fig. 2. "Branch" molecules look like the diagram in Fig. l(viii}. A sample of 35 multiple-forked structures in the 12 kb DNA contained an average of 2.3 regions of renaturat ion per structure; 84~o of these were double forks, 16% were single forks.


kb 0 2 4 6 8 .~ - ' " - : I t ~ i k i , [ |

0 2 4 d a m x ~,0 " s : ~ .' _ : : . ,

ss Mx aO-i 6 ,

I0 !

(a)

:i (d)

(k)

(n)

, T

(e) ~

(f)

(h) ]I

(k) ~ L__

(p) -

m ds

(q) _ ~

Fzo. 4. Multiple-forked and complex structures seen in the 12 kb highly repeated DNA fraction {Tables 2 and 3). Traceable molecules containing more than one renatured region were photographed and measured. The first 17 such molecules tha t were encountered are interpreted here, with crossovers eliminated and the lengths of single-stranded tails averaged. These structures are the source of some of the data in Fig. 5(b) and (d). All would be scored aa "multiple forks" except {k) and (o}, which are "complex" structures. Structures (c), (e), (f), (h), (j), (1} and (n) contain inverted repeat sequences in the form of looped and unlooped hairpins. Structures (o}, (p) and (q) contain inverted repeats tha t may have been distant from each other in the native DNA. Although these 17 molecules are a small sample (about 10 -x~ of DNA), they illus- t ra te the large variety of renaturod structures encountered in many hours of surveying the grids.

240 T. R . C E C H A N D J . E . H E A R S T

The finding of intramolecular renaturation products (hairpins) in the highly repeated DNA fraction was initially surprising because foldback DNA had already been removed in the first HAP chromatography. As shown in Cech & Hearst (1975), the hairpins in the highly repeated fraction contain a lower average number of base-pairs than those isolated in the foldback fraction (0.55 kb compared to 1.0 kb). Furthermore, the frequency of hairpins observed in the highly repeated fraction can be explained by HAP being 90% rather than 100% efficient in fractionating this class of molecules.

(ii) Length measurements

In order to determine the length distribution of the highly repeated elements of the sequence interspersion pattern, randomly encountered double-forked and multiple-forked molecules were photographed, traced, and measured. The double- stranded regions of these structures are accurately defined by the branch points, making them easy to measure. Furthermore, these structures are the most informative because the ends of the base-paired regions are determined by an end of sequence homology (i.e. insufficient sequence homology for stability under the spreading conditions used)t. This is not the case for linear ds/ss and single-forked molecules, in which double-stranded regions end at the point where one of the strands has been broken.

The lengths of base-paired regions in double-forked and multiple-forked molecules, given in the histograms of Figure 5(a) and (b), have similar distributions. The forma- tion of one type of structure or the other may depend only on random processes-- where the sequences happened to be sheared, and whether renaturation continued long enough to form multimolecular products.

The same sequences that formed double-forked renaturation products could have formed single-forked or linear ds/ss products if the DNA had been sheared at a different random place (see Fig. l(a)). The expected proportion of these three products is a function of the length of the repeated sequence and the length of the whole strand (Manning et al., 1975). For the 0-8 kb renatured DNA lengths seen here (at ~ n ---- 7.2 kb), the theory of Manning et al. (1975) predicts a ratio of 64 double-forked:34 single-forked:2 linear molecules. The observed ratio (from the data of Table 3, including the forked regions of multiple-forked molecules) is 65 double-forked:35 single-forked molecules. Therefore, the single-forked molecules probably result from the same size class of repeated DNA sequences that produced the double-forked

t The ending of sequence homology at a fork is only an indication tha t 1 of the 2 strands changes sequence at tha t point. For instance, a sequence element A of length 0.1 kb could be repeated within a single renatured region of length 0-4 kb. Let A' be the complementary sequence to A, and X and Y any other sequences. Measurement of the renatured portions of the renaturation products

XA X X XA X X A X A X

A A A A and A A A A A'A 'A 'A ' A'A'A'A"

Y Y Y A' YY YY YY A'Y

would suggest a native DNA length of 0.4 kb for the repeated stretches, while the actual average lengths would be longer. Therefore, the lengths of the double-stranded regions provide a min imum estimate for the lengths of the repeated sequences in the native DNA. Conversely, lengths of single-stranded regions joining two double-forked regions provide a ~naxim~m estimate of thA length of the non-homologous DNA.


I0-

0- 15-

I0-

,~ 5-

d O- z

I0-

5-

O-

5-

O' 0

T T r r

�9 - '" (a) -I ~ , . _ _ [ ~ _ |n )~rmnn,,, _

(b)

& ".L.,~,d~; ' "i~(-'"

. . . . , , ,

(c)

".. . ,S_ ~_ ."

m

0-5 i'9 ~ 5 20 2"5 5 4 5 ,5 7 kb

Fza. 5. Elect ron microscopic measurements of hydroxyapat i te - f rac t ionated DNA, spread in 45% formamide with circular fd and PM2 DNAs. (a) Length " d " of base-paired regions in double- forked molecules of the 12 kb highly repeated DNA fraction scored in Tables 2 and 3. The number- average molecular weight of the base-paired regions, ~,, is 0.80 kb. /1~,, the number-average single-strand molecular weight, is 6.9 kb (per s t rand) for these structures. This is less t h a n the initial 12 kb molecular weight (determined by alkaline sedimentat ion) par t ly because it is an number-average value, and par t ly because of single-strand breaks t h a t occur during the 2 rena tu ra t ion and H A P chromabography steps. (b) Length of base-paired regions in mult iple-forked molecules on the same grid as (a). Only regions te rmina ted on each end by 2 single s t rands are included. The dashed line refers to measurements of rena tured regions in the type of multiple- forked s t ructure shown in Fig. 2(e), and the solid line to those of all multiple-forked s t ructures (Fig. 2(d) and (e)). ~n = 0"66 kb; -~r n = 8"0 kb per s t rand. (c) Length of base-paired regions in presumed double-forked and mult iple-forked s t ructures seen in the 12 kb unbound fract ion (Table 2), the DNA t h a t chromatographed as single-strand on HAP, an = 0.33 kb; -~rn ~ 7"0 kb. (d) Length " s " of s ingle-stranded DNA joining 2 base-paired regions in the mult iple-forked molecules of pa r t (b). The dashed line again refers to the type of s t ruc ture shown in Fig. 2(e). an ~ 1"8 k b ; / ~ r given in (b).

242 T. R. CECH AND J. E. HEARST

molecules. About 18% [7~/o](32 + 7~/0)] of the bimolecnlar structures are linear ds/ss molecules, a significantly higher fraction than the predicted 20/o. Therefore, about 16% of the renaturation products probably result from highly repeated sequences occurring in longer stretches ( ~ 8 kb).

As given in Table 2, the frequency of forked molecules in the DNA unbound by HAP was about seven times less than in the highly repeated DNA fractions. The presumed base-paired regions in these molecules were measured (Fig. 5(e)). They are mostly small, an observation consistent with their being either random crossovers (not base-paired) or short base-paired regions that might be inefficiently isolated on HAP. Even if all of the double-forked molecules in the unbound fraction were really random crossovers, and there were an equal number of such crossovers in the highly repeated DNA distributions (Fig. 5(a) and (b)), the average base-paired lengths would be only slightly underestimated.

The distance between adjacent renatured regions of multiple-forked molecules was measured. As shown in Figure 5(d), these single-stranded regions are extremely heterogeneous in length. The number average length is 1.8 kb, but the distribution ranges from small gaps of ~0 .1 kb to separations of 6 to 8 kb. The distribution's maximum is limited by the fragment length, which averages 8 kb per strand.

In order to convert the number fractions of Table 2 to approximate weight fractions, measurements were made of all single- and double-stranded regions of about 200 double-forked, multiple-forked, hairpin, and ss molecules in the 12 kb highly repeated DNA. From these measurements, and excluding the mess category, the 12 kb highly repeated DNA fraction was estimated to contain 69% by weight of bimolecular products (59% forked plus 9.5% ds/ss molecules), 19% single-strands, and 12% hairpins. An average of 20% by weight of the highly repeated DNA fraction was base-paired and 80% was single-stranded. Considering the small number of molecules measured and the exclusion of the mess category from the analysis, this value is in good agreement with the 30=t=10% double-strandedness determined for the highly repeated DNA fraction by Sl-nuelease resistance, CsC1 buoyant density, and hyperchromicity (Cech et al., 1973).

(d) Renaturation kinetics of mouse D N A

Kinetic complexities were determined by the technique of Bri t ten & Davidson (e.g., Davidson et al., 1973): a tracer quanti ty of 3H-labeled DNA is renatured with a large excess of unlabeled, sonicated total DNA (see Materials and Methods, section (e)). Cot is determined by the concentration of the unlabeled "driver" DNA, because it is in excess. Monitoring the renaturation of the driver DNA (by optical density) serves as an internal standard for renaturation of the tracer DNA (monitored by scintillation counting).

In the experiments described here, the unlabeled driver DNA was isolated from Balb/c mice and the [3H]DNA from SVT2 tissue culture cells. As shown in Table 4, the renaturation of these DNAs usually differed by one per cent or less over a range of more than six orders of magnitude of Cot. Therefore, the 3H-labeled SVT2 DNA is interchangeable with Balb/e mouse DNA in a renaturation kinetics experiment. Brit ten & Rake (1969) came to the same conclusion for L-cell and mouse DNAs.

Figure 6 shows the renaturation profile of the Balb/c mouse DNA (0-4 kb), and includes all data points from the experiments done subsequently to the one described above. This Co t curve is in substantial agreement with other mouse DNA Cot curves

H I G H L Y R E P E A T E D M O U S E M A I N - B A N D D N A

TABLE 4

Renaturation of 3H-labeled S V T2 D N A driven by excess Balb/c mouse D N A

243

F r a c t i o n b o u n d Cot h:~ Balb]c w SVT2 H A�82

7.3 • 10 -3 0-1 0.100 0-095 - -0 .005

6-7 • lO -2 1.2 0.143 0-136 - -0 .007

7 .7 • 10 -2 0.1 0.184 0.179 - -0 .005

2.3 • 10 -1 0-4 0.233 - - - -

1.2 2.1 0.265 0.273 3 0 - 0 0 8

3.7 6.5 0.268 0.275 3 0 - 0 0 7

1.1 • 10 1.0 0.321 0-332 3 0 ' 0 1 1

1.5 • 10 27 0.319 0-319 0

4.3 • 10 76 0-315 0.299 - -0 .016

1.1 • l02 10 0-366 0.376 3 0 . 0 1 0

3-5 • 102 31 0-441 0.451 3 0 . 0 1 0

9.9 x 102t 10 0.543 0.550 + 0.007

1.1 • 10 a 103 0.518 0.536 3 0 " 0 1 8

3.0 • 10at 31 0.715 0.712 - -0 .003

1.0 • 104t 103 0.801 0"824 + 0.023

Balb]c D N A a n d a l l - l abe led SVT2 D N A were m i x e d in a ra t io o f 1500:1 a n d son i ca t ed to a s i ng l e - s t r and size of 0.32 kb. P o r t i ons o f t h e m i x t u r e were a lka l i - dena tu red , r e n a t u r e d (0-12 M- p h o s p h a t e buffer , 60~ a t d i f ferent t i m e s a n d concen t r a t i ons , a n d c h r o m a t o g r a p h e d on H A P . E a c h s a m p l e c o n t a i n e d 0.2 m g un labe l ed Ba lb /c D N A a n d 0.13 pg (5300 c t s ]min) a l l - l abe led SVT2 D N A . U n l a b e l e d D N A c o n c e n t r a t i o n s va r i ed f rom 0.005 to 1'0 m g / m l .

t R e n a t u r e d in 0.90 M -phospha t e a t 70~ the re fo re Cot was mu l t i p l i ed b y 8-06 to g ive t h e i nd i ca t ed equ iv . Cot.

T h e r e n a t u r a t i o n t i m e in h o u r s = t/3600. w T h e f r ac t ion o f t he recovered opt ica l d e n s i t y u n i t s t h a t b o u n d to H A P , i nd i ca t i ng t h e frac-

t i on o f t h e un l abe l ed Ba lb / c m o u s e D N A r e n a t u r e d a t e ach Cot. I] T h e f r ac t ion of t h e recovered r a d i o a c t i v i t y t h a t b o u n d to H A P , i nd i ca t i ng t h e f r ac t ion of t h e

a l l - l abe led SVT2 m o u s e D N A r e n a t u r e d a t each Cot. �82 A = [ f rac t ion b o u n d (SVT2)] -- [ f rac t ion b o u n d (Balb]c)]. Th i s ini t ia l Cot cu rve shows ev idence o f D N A d e g r a d a t i o n a t long t i m e po in t s , a n d t h e ave rage

mo lecu la r we igh t of t h e s amp l e i n c u b a t e d in 0.12 M-phospha t e for 103 h was r e d u c e d to 0.24 kb . T h e a d d i t i o n o f E D T A to s a m p l e s in s u b s e q u e n t r e n a t u r a t i o n e x p e r i m e n t s p r e v e n t e d t h e degra- da t ion .

at similar fragment size and renaturation conditions (Britten & Kohne, 1966; Laird, 1971; Straus & Birnboim, 1974). The renaturation of Escherichia cell DNA, used as the standard for calculating kinetic complexities, is shown also in Figure 6; the Cot~ of 4.0 is in close agreement with previous values (Britten & Kohne, 1966; Davidson et al., 1973; Bonner et al., 1973).

The renaturation of mouse DNA is fit with five kinetic components, summarized in Table 5. The magnitude of the most rapid transition (the foldback DNA) is known from previous work (Cech et al., 1973; Wilson & Thomas, 1974). The quantity of light satellite DNA was determined by CsC1 buoyant density centrifugation, and its renaturation rate is known (Southern, 1975; Waring & Britten, 1966; Hutton & Wetmur, 1973).

17


0|,._ SAT I I I , I , ,

I - ' ~ ~ HR ,o t - '

20L M o u s e F4"4~'~..~ MR

- \ 50 OVA

70 40

80 60 n

90 80 " ~ ~ " < ~

I O N I - - I I I I I I

0- 3 10 - 2 I0-i i00 i0 j i02 i03 104 105 Equiv. Cot

Fro. 6. Renaturat ion profiles of total Balb/c mouse DNA and E. coli DNA, both 0-4 kb. Samples were renatured in 0.12 M-phosphate buffer at 60~ or, for most Cot values over 102, in 0.90 M- phosphate buffer at 70~ with equiv. Cot calculated as described in Materials and Methods, section (e). ([~) Renaturat ion of unlabeled Balb]c mouse DNA, combined data from 7 experimenbs. The numbers (4 and 5) designate overlapping data points. (Z~) Renatura t ion of unlabeled Balb/e mouse DNA in the E. coli experiment. The line through the points is calculated for 4 second-order components plus foldback DNA, as described in Table 5. The 4 second-order components are designated as SAT, mouse satellite DNA; HR, highly repeated main-band DNA; MR, moderately repeated DI~A; NR, non-repetit ive DNA. The arrows designate the range of Cotj values tha t fit the data well; in the case of the SAT component, the Cotj was held constant. ( �9 ) Renaturat ion of 3H-labeIed E. coli DNA in the presence of excess Balb/e DNA. The 2 DNAs were co-sonicated, insuring equal molecular weights. Each point was determined with 1-7 • 105 ets/min. Because the Balb]c DNA was in 1000-fold excess, each Balb/c point is displaced by 3 decades of Cot from the corresponding E. coli point. The line through the points is calculated for one second-order component (86% of the DNA) with a Cot~ = 4-0, plus a 6% component renatured at Cot < 10 -2, plus 8% of the DNA not renaturing in the time course of the experiment.

The region of the Cot curve between mouse satellite and single-copy DNA can be adequately fit only if a m/nimum of two components are used. The faster of these two components is called "highly repeated main-band DNA" because, as will be shown, it includes the sequences we have been characterizing. At the s tandard renaturat ion criteria, it comprises 8 to 16% of the mouse genome and has an un- corrected repetition frequency of (1.3 to 2.7) • 10 4 copies. As described in footnote (e) of Table 5, we have sufficient information about the base composition and sequence divergence of these sequences to make approximate corrections to the complexity and repetition frequency. The corrected values are (3-1 to 6.5) • 104 copies of a sequence with a 3000 to 12,000 base-pair complexity. (This region of the Cot curve could of course be fit with more than one component, each of which would then have a lower complexity.) The "moderate ly repeated" class ( H I 1% of the mouse genome with an average complexity of (6 to 12) x 105 base-pairs) has a complexity similar to tha t of the major repetition class in the Xenopus (Davidson et al., 1973) and ra t (Rice, 1971 ; Holmes & Bonner, 1974) genomes, which are about the size of the mouse genome.

H I G H L Y R E P E A T E D M O U S E M A I N - B A N D D N A

T AB L E 5

Kinetic classes of the mouse genome

245

Class

Corn - % of g e n o m e a Cott (obs.) b Cotj p lex i ty d Repe t i -

t i on Fig. 6 (Range ) Fig. 6 (Range ) (pure) c (base- pai rs) f r equencyr

F o l d b a e k 2.5 . . . . . . Sate l l i te 8.5 - - 4 . 5 • -3 - - 3 . 6 • -4 4 . 1 • 2 5 . 6 • 10 6 H i g h l y

r e p e a t e d 14 ( 8 - 1 6 ) 0.2 (0.1-0-2) 2 " 8 • -2 2 . 9 • 4~ 1 . 3 • 4g Mode ra t e ly

r e p e a t e d 11 ( 8 - 1 6 ) 10 (5-10) 1.1 1 . 2 x 106 2 .5 • 102 N o n - r e p e a t e d 58 (56-62) 1500 (1400- 8-7 x 102 9.1 x 108 1.7

1700)

T h e 5 k ine t ic c lasses descr ibed in t he t e x t were u sed to fit t h e m o u s e D N A r e n a t u r a t i o n d a t a o f Fig. 6. T h e d a t a was p l o t t e d a n d fit w i t h a Ca lcomp 565 p lo t t e r a n d Dig i ta l P D P 8 e c o m p u t e r . F i t t i n g was done b y t r ia l a n d error, a f t e r wh i ch t h e r o o t - m e a n - s q u a r e y dev i a t i ons (S~) f r om t h e f i t t ed c u r v e s were ca lcu la ted .

a T h e p e r c e n t a g e s a re va l id on ly for t h e f r a g m e n t size (0-4 kb) a n d r e n a t u r a t i o n cond i t i ons (0-12 M-phospha t e buffer , 60~ or equ i va l en t ) u sed in t he e x p e r i m e n t s o f Fig. 6. Because 6 % of t h e D N A fai led to r e n a t u r e , t he p e r c e n t a g e s a d d to 94 i n s t e a d of 100%. "F ig . 6" va lue s g ive t h e bes t fit (Sy = 1.8 in u n i t s o f %), a n d a " r a n g e " o f accep tab le va lue s ~s l i s ted (Sy _<= 2.1).

b T h e Cotj o b s e r v e d for each class o f D N A w h e n d i lu t ed b y t he r e m a i n d e r o f t he genome . Aga in , b o t h t h e bes t fit to t h e d a t a a n d a r ange of accep tab l e va lue s a re g iven .

c T h e Co t j e s t i m a t e d for t h e pu re k inet ic c o m p o n e n t , ca l cu la t ed a t Cot t (obs.) • (% of g e n o m e / 100). T h e "F ig . 6" va l ue s in t h e p reced ing c o l u m n s are u s e d in t h e ca lcu la t ion .

a T h e kinet ic c o m p l e x i t y is re la t ive to E. cell D N A , a n d is ca l cu la t ed as

c o m p l e x i t y (E. coli) Cot t (E, coli) • Cotj (pure).

Cotj (E. cell) = 4.0 f rom Fig. 6 a n d c o m p l e x i t y (E. coli) ~ 4.2 • 106 base -pa i r s (Cairns, 1963). e Th i s is a m a x i m u m va lue . A s s u m i n g a n 8 % h i gh ly r e p e a t e d D N A class w i th C6tj (obs.) ~ 0.1,

n u m b e r s wh ich fit t h e d a t a a l m o s t as well as t hose used , t h e ca l cu l a t ed c o m p l e x i t y is 8 • 10 a. Also, r e n a t u r a t i o n r a t e a n d the re fo re k ine t ic c o m p l e x i t y a re a f u n c t i o n o f base c o m p o s i t i o n (Wet- t ou r & D a v i d s o n , 1968) a n d of s e q u e n c e d ive rgence ( B o n n e t et al., 1973). B o t h fac tors cause t h e c o m p l e x i t y to be o v e r e s t i m a t e d . G i ven t h e base c o m p o s i t i o n (41% G + C) a n d At,, (9~ of h i g h l y r e p e a t e d m o u s e m a i n - b a n d D N A (Cech et al., 1973), i t s co r rec ted c o m p l e x i t y can be est i- m a t e d as [ (0-8--2 .9) x 104] • 0"83 • 0 . 5 ~ (0"3 - -1 .2 ) • 104. F u r t h e r m o r e , t h i s k ine t ic c lass cou ld be t h e ave r age o f m a n y subc lasses , e ach of wh i ch wou ld t h e n h a v e a lower complex i t y .

f T a k i n g t h e m o u s e hap l o i d gonomo size as 2.7 • 109 base -pa i r s (Laird , 1971), r epe t i t i on fre- q u e n c y is ca l cu l a t ed as 2.7 • 1 0 9 x (% of g e n o m e / 1 0 0 ) / e o m p l e x i t y .

* I n t h e ease o f a n 8 % h i g h l y r e p e a t e d D N A class w i th a c o m p l e x i t y o f 8000 base -pa i r s (foot- no t e e), t h e c a l cu l a t ed r epe t i t i on f r e q u e n c y is 2.7 x 104, or 6 . 5 • 104 w h e n cor rec ted for base c o m p o s i t i o n a n d s e q u e n c e d ivergence .

(e) Kinetic complexity of the highly repeated DNA

To determine the spectrum of repetition frequencies in the 6 to 12 kb highly repeated mouse DNA isolated on HAP, a tracer quanti ty of this DNA was denatured and renatured with excess sonicated Balb/c DNA. The tracer DNA was in one case the same sample as used for the electron microscopy, and in two other cases was an analogous preparation. Therefore, the 12 kb tracer was expected to include 12~ by weight of hah'pin-containing strands (not all of which is expected to renature after a second denaturation), 19~/o single-stranded molecules of unknown repetition frequency, and the remaining 69% molecules with one or more regions of highly repeated DNA.


T h e 6 k b t r a c e r was e x p e c t e d to c o n t a i n 5 % h a i r p i n s , 1 7 % ss m o l e c u l e s , a n d 7 9 %

r a p i d l y r e n a t u r i n g m o l e c u l e s b y w e i g h t .

F i g u r e 7 s h o w s t h e r e n a t u r a t i o n o f 6 a n d 12 k b h i g h l y r e p e a t e d m o u s e m a i n - b a n d

D N A f r a c t i o n s , d r i v e n b y exces s 0-4 k b t o t a l D N A . T a b l e 6(A) l i s t s t h e k i n e t i c com-

p o n e n t s t h a t b e s t f i t t h e d a t a o f F i g u r e 7. B e c a u s e o f t h e s m a l l n u m b e r o f d a t a p o i n t s ,

t h e n u m b e r s a r e n o t as e x a c t as t h o s e o f T a b l e 5. N e v e r t h e l e s s , t h e d a t a is a c c u r a t e

e n o u g h t o s h o w t h a t t h e b u l k o f t h e t r a c e r D N A s r e n a t u r e d i n o n e o r t w o v e r y r a p i d

s e c o n d - o r d e r t r a n s i t i o n s , ~ d t h Coti v a l u e s o f ~ 6 • 10 -3 a n d ~ 0 . 1 0 . T h i s r e n a t u r a t i o n

r a t e c o r r e s p o n d s t o t h a t o f t h e h i g h l y r e p e a t e d m a i n - b a n d c o m p o n e n t s (Coti = 0 .10

t o 0.20) s e e n in t h e s o n i c a t e d d r i v e r D N A . B e c a u s e r e n a t u r a t i o n r a t e is a f u n c t i o n

0 ~ . ~ .1 I I I I I

\

o 40 - 0 " ' . . 0

x '~ 6eCo "'" "'" k.m..x

I I I I 1 I I I IOC _ - - 6 ~ I -4 10-5 10-2 i0-1 i0 0 i01 10 2 10 5 10 4

Equiv. Cot 105

FIe. 7. The rena tura t ion of a t racer quan t i ty of 3H-labeled highly repeated mouse main-band DNA, driven by an excess of unlabeled, sonioated to ta l mouse DNA.

( . . . . . ) Rena tu ra t ion curve for the driver DNA, from Fig. 6. ( • Da t a points for driver DNA rena tu ra t ion in the 12 kb unrecycled tracer exper iment

described below. The individual points for the other 2 sets of da ta are shown in Fig. 9. * ) , Range of the Cot t of the highly repeated component of 0.4 kb mouse DNA, from

Fig. 6. Rena tu ra t ion of the 3H-labeled highly repeated DNA, each set of points fit wi th 2 or 3 second-

order components as given in Table 6(A) : - - O - - O - - , 12 kb unrecycled tracer (foldback DNA removed a t Cot 1 • 10-3; 2nd HAP iso-

lation a t Cot 8 x 10 -2, low str ingency elutions, yield = 6.7% of main-band). - - m - - E - - , 12 kb recycled t racer (above sample recycled at Cot 5 • 10 -3, high stringency,

yield after recycling = 0.6% of main-band) . - - A - - A - - , 6 kb recycled tracer (foldback DNA removed a t Cot 7 x 10-4; 2nd HAP isolation

at Co t 6 x 10-2, low stringency, 9-9% yield; recycled at Co t 6 x 10-3, high stringency, yield after recycling = 1.5 % of main-band). This sample was examined in the electron microscope (Table 2).

Each point was determined with 1.I • 104, 2-2 x 103 or 2'0 • 103 c ts /min of t racer DNA in the 3 experiments, respectively. The DNA specific act ivi ty was 2-6 x 105 ets /min per ~g for the 12 kb DNA, 4.0 • 104 for the 6 kb DNA. The unlabeled driver DNA was present in a 3.7 x 103, 2.2 • 104, or 2.5 • 103 t imes excess in the 3 experiments, respectively.

( 0 ) 12 kb unrecycled tracer, same DNA as (O), bu t incubated wi th carrier M. lyaodeikticus DNA instead of wi th unlabeled mouse DNA driver. Cot for these points is calculated as if the mouse DNA driver had been present. As expected for such a rapidly rena tur ing DNA, t racer DNA self-reaction does occur a t high Cot values. Because the self-reaction points are displaced by several decades of Cot from the driven rena tura t ion curve, this DNA would renature wi th driver DNA long before significant self-reaction could occur.

H 1 G H L Y R E P E A T E D MOUSE M A I N - B A N D DNA

TABLE 6

Kinetic components of highly repeated mguse main-band DNA, before and after shearing

247

Region of Co t curve Highly repeated FB SAT HR MR MR

tracer DNA % % C0t~ % C0tj % Cot~ % C0t~

A. Unsheared (Fig. 7) 6 kb recycled ( � 9 5.4 25 0.004 62 0-08 0 - - 0 - - 12 kb recycled (m) (4.1) 36 0.008 43 0.11 0 - - 11 400 12 kb unrecy. ( 0 ) 4.1 0 - - 62 0.13 15 20 14 1000

B. Sheared (Fig. 9) 6 kb recy., shear to 0.9 kb (A) 4.9 26 0.008 54 0.20 0 - - 9 300 12 kb recy., shear to 0.5 kb (E]) (0-7) 39 0.021 35 0.15 0 - - 21 300

These kinetic components describe the graphs in Figs 7 and 9. They do not fit the data uniquely, but do give a good indication of the repetition frequency components that are involved. SAT, HR, MR and MR refer to regions of the Co t curve defined in Fig. 6. Here the SAT components are mostly main-band DNA with a satellite-like Cot j, rather than actual satellite sequences (see Fig. 8). The foldback (FB) values were determined by renaturation of the tracer DNA to Cot

10 -6 in the absence of driver DNA. The 12 kb recycled tracer values in parentheses were not experimentally determined, but were assigned the values measured for the same tracer DNA before it was recycled.

of molecu la r weight , t he d a t a for sheared t r ace r D N A (section (f)) is necessa ry to es tab l i sh th is correspondence.

I n t he expe r imen t wi th t he recyc led 6 kb h igh ly r epea t ed t racer , t he re was no de t ec t ab l e c o m p o n e n t wi th a Cott > 0.10. I n t he two expe r imen t s wi th 12 kb DNA, however , 11 to 29% of t he D N A r e n a t u r e d as s lower components . The non- repea ted D N A in these p r e p a r a t i o n s m u s t have been f reed f rom a t t a c h m e n t to fo ldback D N A or r e p e a t e d D N A b y s ing le -s t rand breaks . The e lec t ron microscopic obse rva t ion of 19% single s t r ands in an unrecyc led r ap id ly r e na tu r i ng f rac t ion a t th is molecu la r weight (Table 2, 12 kb DNA) is cons is ten t w i th t he a m o u n t of non - r e pe a t e d D N A seen in t he cor responding Cot curve. The on ly Cot curve t h a t showed de t ec t ab l e amoun t s of m o d e r a t e l y r e p e a t e d D N A was the p r e p a r a t i o n t h a t had n o t been recycled . Thus, the r igorous k ine t ic recycl ing s tep does seem to serve the func t ion of pur i fy ing h igh ly r e p e a t e d f rom m o d e r a t e l y r e p e a t e d sequences.

Because a s ignif icant f rac t ion (about one- th i rd) of t he recycled h igh ly r e p e a t e d D N A r e n a t u r e d wi th a Cot�89 of (4 to 8) x 10 -3, s imi lar to t h a t of mouse sa te l l i t e DNA, i t was necessa ry to see i f th is c o m p o n e n t could be due to sa te l l i te D N A c o n t a m i n a t i n g the or iginal m a i n - b a n d D N A p repa ra t i on . A 0-2 to 0 .5% c o n t a m i n a t i o n would exp la in t he r e n a t u r a t i o n kinet ics , and such a smal l a m o u n t of sa te l l i te would no t have been d e t e c t e d in t he ana ly t i ca l CsC1 cen t r i fuga t ion of the un f rac t iona ted , n a t i v e ma in - b a n d DNA. To tes t for sa te l l i te D N A in the t racer , por t ions of the low-C0 t r e n a t u r e d f rac t ions f rom two of t he F igu re 7 Cot curves were b a n d e d in p r e p a r a t i v e CsC1 gradien ts . The un labe l ed d r ive r D N A se rved as an in te rna l m a r k e r for t he pos i t ions of r e n a t u r e d sa te l l i te a n d h igh ly r e p e a t e d m a i n - b a n d DNAs. As shown in F igu re 8,

t he t r ace r D N A s b a n d m o s t l y wi th t he m a i n - b a n d h igh ly r e p e a t e d DNA. (The


(a)

0"8-

0.6-

o

0.4-

0 - 2 -

c

1.0-

o

I'O-

1,j

/ op.O "~

0.8-

0-6-

0-4-

oIL

cb) i

41

d

I0 20 ~0

Fraction no.

R

-iO00

-800

- 6 0 0 ~ m

"T

�9 400 z

-200

,0

-200

150

.c_

m .3

,noo~

5O

i -

40

1~Io. 8. A tes t for satellite DNA in the 12 kb highly repeated mouse ma in -band DNA tracer. (a) Unrecycled 12 kb tracer, ne t 3H c ts /min ( - - O - - O - - ) ; sonicated driver DIqA, A2eonm ( - - O - - O - - ) . Port ions of 3 rena tured fractions f rom the unrecycled 12 kb tracer Got curve of Fig. 7 were pooled. EDTA, Sarkosyl, and solid CsC1 were added, and the DNA was allowed to aggregate a t 60~ to an equiv. Cot of 50. After ad ju s tme n t to a vol. of 6 ml and a refractive index of 1.3992, the solution contained 1 mM-EDTA/0-04% Sarkosyl/0.30 ~a-phosphate buffer. Centri- fugat ion was a t 35,000 revs /min in the Spinco fixed-angle 65 rotor a t room tempera tu re for 36 h. The polyallomer tube was t hen punc tured and fractions collected th rough a 23 G needle. The optical densi ty of each 0.12 ml fraction was determined in a microcuvet te , after which the fraction was dried on a W h a t m a n GF/A glass filter. The filters were counted on Omnlfiuor (New England


experiment was repeated after shearing the tracer DNA to 0.5 kb, with the same result.) The small amount of tracer DNA that bands with the satellite is insufficient to account for the fastest-renaturing component of the tracer DNA, so it must be primarily of main-band orion. This component makes up only about 0 . 5 ~ of the main-baud DNA (1/3 of the 1.5~ recycled highly repeated DNA), and therefore is not detected in the unrecycled Cot curve of Figure 7 or the total DNA Co t curve of Figure 6.

(f) Kinetic complexity of the unrenatured sequences attached to highly repeated DNA

In the experiments of Figure 7, where long DNA strands containing highly repeated sequences are renatured with total mouse DNA, the renaturation kinetics are determined by the most highly repeated sequences on each strand. In order to determine the complexity of the DNA tha t is adjacent to the highly repeated stretches (the single-stranded "tails" seen by electron microscopy) the Cot analysis was repeated on the long tracer strands after shearing. The 6 kb recycled tracer of Figure 7 was sheared by rapid decompression in a bomb (850 lb/in 2) to a molecular weight of 0.9 kb, and the 12 kb recycled tracer was sonicated to a molecular weight of 0.5 kb. (The sheared molecular weights were not determined for the tracer DNA itself, but in parallel experiments described in Table 1.) Random shearing is expected t to free 7 6 • of the DNA in the forked structures of the 12 kb tracer DNA from attachment to renatured regions. For the 6 kb tracer DNA, 6 4 • of the DNA in forked structures should be sheared from renatured regions. Therefore, if the unrenatured tails are largely non-repetitive DNA, the Cot curve should change dramatically after the tracer is sheared.

Figure 9(a) shows the renaturation profile of the 6 kb tracer before and after shearing, and Figure 9(b) shows a similar experiment with the 12 kb tracer. The

t Let Fss be the fraction of the renatured DNA tha t would chromatograph as single-stranded on HAP after shearing, i.e. would no longer be a t tached to renatured regions. Then, for double- forked molecules, 2' , , = 1 -- [(d + L -- 2 S)]LI], where d is the length of the renatured segment, L is the fragznent size after shearing, S is the minimum stable duplex length in HAP chromatography, and LI is the initial s t rand length before shearing. (This equation comes from the analysis done in the Appendix of Cech et al., 1973. I t is true for L > d as well as L < d.) For double- forked molecules in the 12 kb tracer DNA, using the electron microscopic number average values of d and L,, using S = 0-06 kb (Wilson & Thomas, 1973), and using L = 0-5 kb from Table 1 (although this is a weight-average value), F , , ~ 1 -- [(0.7+0.5--0.12)/7] = 0.85. Taking into account the est imated number of forked structures with more than one renatured region will increase d and thereby decrease F , . For the whole forked category of the 12 kb tracer, / ~ , is calculated to be 0.76 4-0.09. The uncertainty arises from different estimates of the average number of renatured regions in more complex structures like those shown in Fig. 2(f) and (g).

Nuclear). The unlabeled driver DNA contains those Balb/c mouse DNA sequences which renature a t low O0t values: foldback DNA, which bands as a broad peak in the gradients; mouse satellite DNA, which forms the sharp, light density peak (fractions 35 to 40); and main-band highly repeated DNA, which forms the sharp, heavy density peak (fractions 25 to 33). The identi ty of these peaks is known from our previous work (Cech eC al., 1973). 1.3% of the tracer DNA bands with the satellite peak. Because only the fastest-renaturing 60~o of the tracer is included in the gradient, the 1"3~o satellite represents 0"8~o in the whole tracer. (b) Recycled 12 kb tracer DNA from the C0t curve of Fig. 7 ( - - 0 - - 0 - - 7 ; sonicated driver DNA ( - - 0 - - 0 - - ) . The procedure was the same as described in (a). 11 ~o of the recycled tracer is in the satellite peak. The radio- activity represents the fastest-renaturing 76~/o of the tracer, giving a corrected value of 8~ satellite DNA in the recycled tracer. Whether or not this last correction is made, it is obvious tha t the fastest component of the highly repeated DNA is primarily of main-band origin.


0

Z0

40

60

80

:= I 00 2

1 N 0

I l , i s e~ . . oe , , . I K I J I I

I0 I I I I I -6 I;e I - 4 iO-a 10-2 io-i i 0 0 i01 102

o,

o

~ 'L.

~

I I I0 3 10 4 105

20

40

60

80

- I00 , ' - ' - -~ -~ I . I0 ''~ I0-"

( �9 P, o,.--%, "," '~- .x . .

~ �9 " . ; x . . . . , x ~ . ~ ~, " -o ,+

% ~ \~0 " X .

%\ ~P. % ~

,, 'H- o ~ - - o x,. . : "

" 0 'Xo

i I I i i I I I 10-3 10-2 i0-1 I0 0 I01 10 2 I0 ~ 10 4 I0 @

Equiv. COt

Fio. 9. Rena tura t ion of highly repeated mouse main-band DNA, before and after shearing. (a) (A) 6 kb recycled tracer DNA rena tu ra t ion from Fig. 7. (/~) Rena tu ra t ion of the same tracer after shearing to 0.9 kb. ( • ) Da ta points for driver DNA renatura t ion, unsheared expt. ( + ) Da t a points for driver DNA renatura t ion , sheared expt.

( . . . . . ) Rena tu ra t ion curve for the driver DNA, from Fig. 6. <- >, range for the Cot ~ of the highly repeated component of 0.4 kb mouse DNA from Fig. 6. (b) (m) 12 kb recycled t racer DNA rena tura t ion from Fig. 7. (E]) Rena tu ra t ion of same t racer af ter shearing to 0-5 kb. ( • ), (W) driver DNA rena tura t ion as in (a). Each set of t racer DNA points is fi t ted wi th 2 or 3 second-order components as given in Table 6.

The amount of 3H-tracer DNA and unlabeled driver DNA used to determine each point was the same for the sheared as for the corresponding unsheared point , and therefore is given in the legend to Fig. 7.

k i n e t i c c o m p o n e n t s t h a t b e s t f i t t h e s e c u r v e s a r e g i v e n i n T a b l e 6 B . B e c a u s e o f t h e

s m a l l n u m b e r o f d a t a p o i n t s , o t h e r f i ts t o t h e d a t a a r e poss ib l e , a n d t h e l i s t e d Cot t

v a l u e s m u s t b e c o n s i d e r e d o n l y a p p r o x i m a t e . T h e r e a r e t h r e e i m p o r t a n t o b s e r v a t i o n s

t h a t c a n b e m a d e f r o m t h e d a t a o f F i g u r e 9.

(1) M o s t o f t h e i n t e r s p e r s e d , u n r e n a t u r e d s e q u e n c e s i n t h e h i g h l y r e p e a t e d D N A

f r a c t i o n a r e a lso h i g h l y r e p e a t e d s e q u e n c e s , a l t h o u g h t h e r e a r c s o m e i n t e r s p e r s e d

H I G H L Y R E P E A T E D M O U S E MAIN-BAND DNA 251

complex sequences. The most striking fact about the Figure 9 Co t curves is how little they change upon shearing the tracer DNA. In both experiments, only about 10% non-repetitive DNA is detached from rapidly renaturing sequences upon shearing. (The complex sequences have 100 < Cot t < 1000 in Fig. 9(a) and Coti ~ 300 in (b); this component will be designated as "non-repetitive" DNA even though it seems to be slightly repetitious.) This 10% is more than can be accounted for by non-repetitive sequences sheared from foldback DNA. (The sequences near foldback DNA "hairpins" in Xenopus, Drosophila, and human DNA are largely non-repetitive sequences (Davidson et al., 1973; Schmid et al., 1975; Schmid & Deininger, unpublished observations).) Thus, there appears to be some interspersion of highly repeated and single- copy DNA sequences as in model (b), Figure 1. Most of the tracer DNA, however, continues to renature rapidly after shearing. This is the result expected for a permuted tandem repeat model of sequence organization, diagrammed in Figure l(d). The relative amount of the non-repetitive DNA sequence interspersion is calculated as

( A N R - - A F B ) I n r

A N R is the increase in the non-repetitive DNA component after shearing, A F B is the estimated amount of non-repetitive DNA detached from foldback DNA by shearing, and f is the weight fraction of forked molecules in the tracer. For the 6 kb tracer,

[(0.09-[-0.02) -- (0.01=J=0-01)] ---- 0.18•

[(0.64=l=0.06) • (0.68• I n r

For the 12 kb tracer,

n r

[ (0-10• (0"02i0"01)] ---- 0"18=t=0"07.

[(0-76i0"09) • (0-59-4-0-10)]

At 12 kb there is an additional 11 ~/o non-repetitive DNA that was presumably broken from attachment to rapidly renaturing regions prior to the deliberate shearing; this gives a total of 0.29-k 0-07 non-repetitive interspersion at the higher starting molecular weight.

Thus, analysis of both experiments leads to the same conclusion: 184-7% of the interspersed sequences within a few kilobases of highly repeated sequences are non- repetitive (or of very low repetition frequency), while 82 • contain sequences with about the same repetition frequency as the sequences initially responsible for the whole strand binding to HAP. This 82 ~/o does not have to be pure highly repeated sequences--it could still contain more complex sequences, as long as each 0.5 kb fragment contained enough highly repeated DNA to give the observed rapid renaturation.

(2) The repeated sequence elements in the high molecular weight tracer DNA belong to the highly repeated class (~14% of the total DNA, Cot~ = 0.1 to 0.2). This conclusion could only be made tentatively in section (e), because of the unknown effect of molecular weight on the tracer's renaturation rate. Now (especially in the Fig. 9(b) experiment), the sheared tracer has a molecular weight about equal to that of the driver DNA, and the Cott values can be directly compared. The Cott of the major repeated sequence class of the sheared tracer (Table 6B) is the same as the highly repeated class in total DNA (Table 5), within experimental error.


(3) The increase in G0t~ of the highly repeated DNA component upon shearing gives additional evidence for the presence of more than one stretch of repeated DNA per high molecular weight tracer strand. In the HAP-assayed renaturation of long tracer strands and short driver strands, the rate of the tracer-driver renaturation is faster than the driver-driver renaturation in proportion to the relative number of nucleation sites on the two sizes of strands (Wetmur, 1971 ; Davidson et al., 1973). In our case, assuming that 70 to 80% of the sequences on each tracer strand contain highly repeated DNA, the tracer should bind to HAP with a rate about [0.75 • L (tracer)/L(driver)] faster than that of the driver. The tracer renaturation curve should therefore be displaced to higher Cot values after shearing, and the amount ef the shift should be about [0.75 • L(unsheared traeer)/L(sheared tracer)]. For the experiment of Figure 7(a) this ratio is about (0.75 • 5/0.9) ~ 4; for Figure 9(b), about (0-75 • 7/0.5) ---- 10. In both cases, the observed ratio is about 3 - - the sheared curve is displaced by half a decade of Cot. Qualitatively, the decreased C0t at high molecular weight means that there are more nucleation sites on the long strands, which is consistent with the permuted tandem repeat model. Quantitatively, there is a dis- crepancy between the expected and observed shifts in one of the two experiments (Fig. 9(b)). This is probably due to the unsheared tracer having a lower than expected molecular weight at the time of the renaturation experiment. Fortunately, even a large decrease in molecular weight does not affect the previous conclusions about interspersion. When a starting molecular weight of 3.7 kb is substituted in the equation for Fss, the resulting value of Inr (the fraction of non-repetitive sequence interspersion) is 0-24 instead of 0.184-0.07.

4. Discussion

(a) A model for the organization of highly repeated mouse ~nain-band sequences

The electron microscopy of renatured DNA shows that, in fragments with a number- average length of 7 to 8 kb, approximately 84% by number (~88~o by weight) of highly repeated mouse main-band sequences occur in an interspersion pattern. That is, renatured regions alternate with non-homologous regions. Most interspersed highly repeated DNA regions are 0.2 to 1.0 kb in length, with a number-average length of 0-7 kb. On those molecules where two or more renatured regions can be identified, the average spacing between regions varies widely, averaging about 1.8 kb. The remaining 16~ (12% by weight) of the repeated DNA sequences may be tandemly repeated rather than interspersed, or may be interspersed sequences with a longer average repeated sequence length ( > 8 kb). These percentages refer only to the obvious bimolecular renaturation produets--intramoleeular renaturation products, single-stranded DNA, and circles were also seen in the highly repeated DNA fraction. The percentages assume that the 17% messes seen in this fraction arise largely from further renaturation of molecules with the same type of sequence arrangement as the traceable molecules.

In an earlier s tudy (Ceeh et al., 1973) we determined the HAP yield and per cent double-strandedness of highly repeated mouse main-band DNA at three molecular weights. We concluded that segments of renatured DNA with an average length of 1.54-0.5 kb were interspersed with stretches of uurenatured DNA averaging 2"24-1.1

HIGHLY REPEATED MOUSE MAIN-BAND DNA 253

kb. These numbers are in excellent agreement with the electron microscopic averages of 1.1 kb [(0.88 • kb for forked molecules) + (0.12 x ~ 4 kb for ds/ss)] and ~1.8 kb, for double- and single-stranded regions, respectively.

Our earlier study suggested that the interspersed, unrenatured sequences were of higher complexity than the renatured sequences. The arguments were indirect, based on the high degree of single-strandedness of the renatured molecules even after they were aggregated to a high Cot (~50). Apparently the single-strandedness was largely due to excluded volume effects that inhibited further renaturation of the large aggregates. The driven renaturation experiments described here lead us to conclude that only ]8• of the unrenatured DNA in the 6 kb repeated DNA fraction is made up of non-repetitive sequences. At a starting molecular weight of 12 kb there is an additional 11% non-repetitive DNA, presumably broken from attachment to rapidly renaturing regions prior to the deliberate shearing. In both cases, the bulk of the single-stranded tails seem to have a repetition frequency similar to that of the renatured regions.

This suggests a permuted tandem repeat model of sequence organization, where a highly repeated sequence R 1 is sometimes adjacent to R2, but sometimes adjacent to R3. If the interspersed non-repetitive DNA is also adjacent to the permuted tandem repeats, the simplest model of sequence organization consistent with our data is : R1R2S R4R~. . . R1R3S R6R2 . . . . S indicates a single-copy or non-repetitive sequence. Each repeated sequence element (Rn) has an average length of roughly 0.7 kb. An alternate possibly--R1R2R3R1R 4 . . . S S R1 S S . . . , would give a larger increase in HAP yield with molecular weight than we have observed (Cech et al., 1973). Our data does not preclude smaller stretches (~0.5 kb) of non-repetitive DNA in addition to the 18% predicted by the shearing experiments of section (f). Britten (1972), however, has degraded intermediate repetition classes of mouse DNA (including both the highly repeated and moderately repeated sequences as defined here) to lower molecular weights than we have used. He found a fine-scale interspersion of repeated and non-repeated sequences only in a low thermal stability (tin ~ 68~ fraction, a fraction excluded from our highly repeated DNA by 70~ HAP elutions. Although the sequences responsible for the HAP isolation of our highly repeated DNA fraction must be of high thermal stability, it is possible that some of the interspersed, unrenatured repeated sequences in the permuted tandem repeat arrangement are of low thermal stability (Rice, 1971).

We have several ways of estimating the amount of the mouse genome that contains highly repeated sequences (interspersed to some extent with non-repeated sequences). From the yield of highly repeated DNA at 12 kb molecular weight, a minimum of 4% of the mouse main-band contains such sequences. This is clearly an under- estimate, because the Cot of isolation was less than the Cot~ of the highly repeated DNA. Approximately 12• of the sonicated driver DNA renatures in the same transition as the sheared highly repeated tracer, giving a maximum estimate for the amount of highly repeated DNA. Adding the amount of interspersed non-repetitive DNA that is no longer attached to highly repeated DNA at 0.5 kb molecular weight increases this estimate to 15• The ~15 % of the DNA containing highly repeated main-band sequences is probably distributed among most or all of the mouse chromosomes (Cech et al., 1973). Since only about 18% of the 6 kb highly repeated DNA fraction consists of interspersed non-repetitive DNA, the amount of the non-repetitive DNA of mouse that is so arranged is estimated as (18% • 15%)/58% = 5%. The


more moderately repeated fraction of mouse DNA seems to be more widely interspersed with non-repetitive DNA (Britten, 1972), in an arrangement similar to that found in the frog, sea urchin, and human genomes (Davidson et al., 1973; Graham et al., 1974; Schmid & Deininger, unpublished observations).

We do not yet know the generality of the sequence arrangement proposed here for highly repeated mouse main-band sequences. Human homogeneous main-band DNA (Corneo et al., 1970) probably contains highly repeated sequences interspersed with other repeated sequences (Marx, Allen & Hearst, 1976). About 6% of Xenopus DNA is made up of highly repeated sequences (Davidson et al., 1973), and permuted tandem repetition has not been ruled out for their sequence organization.

(b) Biological function of the highly repeated sequences

Reasonable estimates for the reiteration frequency and complexity of the mouse main-band highly repeated DNA range as high as 64,000 repeats of a 3000 base-pair sequence (from the footnotes to Table 5). This repetition frequency, so much higher than that of the more moderately repeated class of mouse DNA (or the major repetition frequency class of Xenopus DNA), suggests that the two may have quite different origins and/or functions. Of course, the intermediate region of the mouse Cot curve could be fit by more than two components, and may well represent a continuum of repetition frequency classes. In this case, it is the large range of repetition frequencies that is difficult to reconcile ~dth a single model of sequence function.

Besides the low kinetic complexity, there are three preliminary indications that mouse main-band highly repeated sequences may have some characteristics of simple sequence (satellite) DNAs. (i) The permuted tandem repeat sequence organization of the highly repeated main-band DNA may be qualitatively similar to that of the mouse satellite. High molecular weight mouse satellite DNA shows anomalous renaturation kinetics (Hutton & Wetmur, 1973), and the renatured product has an increased buoyant density (Bond et al., 1967; Corneo et al., 1968; discussed by Flamm et al., 1969) and shows a series of single-stranded loops when viewed in the electron microscope (Cech, 1975). All of these properties might be explained by a permuted tandem repeat sequence organization. (ii) Both mouse satellite and highly repeated main-band DNA form renaturation products of high thermal stability. Rice (1971) has found that repeated mouse sequences of high thermal stability do not hybridize with the corresponding sequences of rat or hamster DNA. These sequences, unlike other repeated sequences which occur in all three rodents, have presumably arisen since the divergence of the species. (iii) Short runs of pyrimidine nucleotides are characteristic of rodent satellite DNAs (Southern, 1970; Walker, 1971). Approxi- mately 1.3% of mouse L-cell DNA contains long pyrimidine tracts, 50 to 150 nucleotides (Straus & Birnboim, 1974). These are located in DNA with the same repetition as the highly repeated mouse main-band DNA characterized here.

Preliminary hybridization experiments have detected sequences complementary to the highly repeated main-band DNA in the whole cell RNA of SVT2 cells (Cech, 1975). In an RNA excess hybridization at an RNA Cot of 820, 21% of the highly repeated DNA formed RNA:DNA hybrids, detected by HAP chromatography after RNase treatment. This value is corrected for the DNA self-reaction. The 21% is a minimum estimate, since saturation was not achieved. The Cs2SO, buoyant density of the hybrids after RNase treatment corresponded to an RNA/DNA ratio of 0.2/1.0. Such a ratio would be expected for the hybridization of permuted repeat RNA

H I G H L Y R E P E A T E D MOUSE M A I N - B A N D DNA 255

t r ansc r ip t s to the p e r m u t e d r epea t DNA. More de ta i l ed s tudies of t he t r ansc r ip t i on of these sequences are necessary to d iscover the i r biological funct ion.

One of us (T. R. C.) is grateful to Harumi Kasam~tsu for many hours spent teaching him electron microscopy techniques. We have apprecia ted a continual exchange of da ta and ideas with Carl Schmid. We thank Alan Levin for his help in computer programming. James Wang and Leroy Liu generously provided circular fd and PM2 DNAs. This work was suppor ted in pa r t by Uni ted States Public Heal th Service grant no. GM11180 and Nat ional Science Founda t ion grant no. GB36799 to one of us (J. E. H.). T. R. C. was suppor ted by a Nat ional Science Founda t ion Graduate Fellowship.

R E F E R E N C E S

Bond, H. E., F lamm, W. G., Burr, H. E. & Bond, S. B. (1967). J. Mol. Biol. 27, 289-302. Bonner, T. I. , Brenner, D. J . , Neufeld, B. R. & Bri t ten, R. J . (1973). J. Mol. Biol. 81,

123-135. Bri t ten, R. J. (1972). In Evolution of Genetic Systems, Brookhaven Symp. no. 23, pp. 80-

94, Gordon & Breach, New York. Bri t ten, R. J . & Kohne, D. E. (1966). Carnegie Inst. Wash. Yearb. 65, 78-106. Bri t ten, R. J . & Rake, A. V. (1969). Carnegie Inst. Wash. Yearb. 67, 325-327. Bri t tcn, R. J . & Smith, J . (1970). Carnegie Inst. Wash. Yearb. 68, 378-386. Bri t ten, R. J. , Graham, D. E. & Neufeld, B. R. (1974). In Methods in Enzymology (Colo-

wick, S. P. & Kaplan , N. O., eds), vol. 29E, pp. 363-418, Academic Press, New York. Cairns, J. (1963). Cold Spring Harbor Symp. Quant. Biol. 28, 43-46. Ceeh, T. R. (1975). Ph.D. thesis, Univers i ty of California, Berkeley. Cech, T. R. & Hears t , J. E. (1975). Cell, 5, 429-446. Cech, T. R., Rosenfeld, A. & Hearst , J . E. (1973). J. Mol. Biol. 81, 299-325. Corneo, G., Ginelli, E., Soave, C. & Bernardi , G. (1968). Biochemistry, 12, 4373-4379. Corneo, G., Ginelli, E. & Polli, E. (1970). J. Mol. Biol. 48, 319-327. Corneo, G., Ginelli, E. & Polli, E. (1971). Biochim. Biophys. Acta, 247, 528-534. Davidson, E. H. & Brit ten, R. J . (1973). Quart. Rev. Biol. 48, 565-613. Davidson, E. H., Hough, B. R., Amenson, C. S. & Bri t ten, R. J . (1973). J. l~lol. Biol. 77,

1-23. Davis, R. W., Simon, M. & Davidson, N. (1971). In Methods in Enzymology (Grossman, L.

& Moldave, K., eds), vol. 21D, pp. 413-428, Academic Press, New York. F lamm, W. G., Bond, H. E. & Burr, H. E. (1966). Bioehim. Biophys. Aeta, 129, 310-317. F lamm, W. G., Walker , P. M. B. & McCallum, M. (1969). J. Mol. Biol. 42, 441-455. Gall, J . G. & Atherton, D. D. (1974). J. Mol. Biol. 85, 633-664. Goldring, E. S., Brutlag, D. L. & Peacock, W. J. (1975). In The Eukaryote Chromosome

(Peacock, W. J. & Brock, R. D., eds), pp. 47-59, Austral ian Nat ional Univers i ty Press, Canberra, in the press.

Graham, D. E., Neufeld, B. R., Davidson, E. H. & Bri t ten, R. J. (1974). Cell, 1, 127-137. Holmes, D. S. & Bonner, J . (1974). Biochemistry, 13, 841-848. Hut ton , J . R. & Wetmur , J . G. (1973). Bioehem. Biophys. Res. Commun. 52, 1148-1155. Jones, K. W. & Cornco, G. (1971). _Nature New Biol. 233, 268-271. Kram, R., Botchan, M. & Hears t , J . E. (1972). J. Mol. Biol. 64, 103-117. Lai ld , C. D. (1971). Chromosoma, 32, 378-406. Lewin, B. (1975). Cell, 4, 77-93. Manning, J. E., Schmid, C. W. & Davidson, N. (1975). Cell, 4, 141-155. Mar:~ K . A . , Allen, J . R. & Hears t , J. E. (1976). Bioehim. Biophys. Acta, in the press. Peacock, W. J. , Bmltlag, D., Goldring, E., Appels, R., Hinton, C. W. & Lindsley, D. L.

(1974). Cold Spring Harbor Symp. Quant. Biol. 38, 405-416. Rice, N. (1971). Carnegie Inst. Wash. Yearb. 69, 472-485. Sehmid, C. W., Manning, J . E. & Davidson, N. (1975). Cell, 5, 159-172. Southern, E. M. (1970). _Nature (London), 227, 794-798. Southern, E. M. (1975). J. Mol. Biol. 94, 51-69. Straus, N. A. & Birnboim, H. C. (1975). Proe. Nat. Acad. Sei., U.S.A. 71, 2992-2995. Studicr, F. W. (1965). J . Mol. Biol. 11, 373-390.

256 T . R . CECH AND J. E. H E A R S T

Thomas, C. A., Jr , Hamkalo, B. A., Misra, I). M. & Lee, C. S. (1970). J. Mol. Biol. 51, 621-632.

Walker, P. M. B. (1971 ). In Progress in Biophysics and Molecular Biology (Butler, J. A. V. & Noble, D., eds}, vol. 23, pp. 145-190, Pergamon Press, Oxford.

Walker, P. M. B. & McLaren, A. (1969). Molec. Gen. Genet. 104, 104-I06. Waring, M. & Brit ten, R. J . (1966}. Science, 154, 791-794. Wetmur , J . G. (1971). Biopolyrners, 10, 601-613. Wetmur , J. G. & Daviclson, N. (1968). J. Mol. Biol. 31,349-370. Wilson, D. A. & Thomas, C. A., J r (1973). Biochim. Biophys. Acta, 331, 333-340. Wilson, D. A. & Thomas, C. A., J r (1974). J. ~ o l . Biol. 84, I15-138.

organization of highly repeated sequences in mouse main-band dna · 2009-05-17 · j. mol. biol....

Documents