Bmehtmwa et Bwph),swa Attn. 1088 (1991) 234-244 ' 1991 t:lsevler .Science Publishers B.V. 0167-4751/91/$03,50 A D O N I S 0167478191000812

234

It It A t~X I' 92209

Classes of autonomously replicating sequences are found among early-replicating monkey DNA Suzanne Landry and Maria Zannis-Hadjopoulos McGdl ('ancer ("('nter, Ah Gdl Um~',er~t)'. Montr~;al (Canada)

ff,ecetved 20 August 1990)

Key words: DNA replication initiation; Mammalian DNA replication: Replication origin; Autonomously replicating sequence; Early activated origin; Class of origin

Thirteen new independent chines of origin-enriched sequences (ors) that are capable of autonomous replication have been identified from a library of 100 orsi clones that had been previously isolated from early replicating monkey (CV-I) DNA. Autonomous replication was assayed by transient episomal replication in transfected HeLa cells; ors-plasmid DNA was isolated at various times after transfection and screened by the Dpnl resistance assay and the bromodeoxyuridine (BrdUrd) substitution assay to differentiate between input and newly replicated DNA. Four of the autonomously replicating clones were identified by screening the ors-library wi~h probes of ors 3, 8, 9 and 12, previously shown to be capable of autonomous replication (Frappier and Zanfiis-Hadjopouios, Proc. Natl. Acad. Sci. USA (1987) 84, 6668-6672). The other nine functional ors clones were identified among 18 randomly chosen ones, which were ~imilarly screened for autonomous replication. Nucleotide sequence analyses of 11 of the newly identified functional ors plasmids revealed, in most of them, features similar to those present in other viral or eukaryotic replication origins, notably the presence of AT-rich regions and inverted repeats. Pairwise comparisons between the newly identified ors showed no extensive sequence homologies, other than the presence of the -,-satellite repetitive sequence family in three ors and u r the r,~petitive Alu sequence family in one ors. The results suggest that there exist different classes of mammalian replication origin, highly or moderately repetitive and unique, and that their activation is most probably dependent on the presence of structural ~ctermimnt~ rather than on a particdar sequence. Introduction The notion of the replicon for mammalian cells [1] prompted the search for sequences functioning as initiation sites for DNA replication. Studies on the replication of some multigene families, such as the/8-globin gone cluster [2] and the immunoglobulin heavy chain constant region (lgCH) [3], showed temporal directionality, supporting earlier findings, and led to the identification of potential initiation sites for the/8-globin cluster. Similarly, polarity of DNA replication of the avian fl-globin [4]. histone H5 [5], and c - n ~ [6] proAbbreviations: ARS, (yeast) autonomously replicating sequence; or.J, (monkey) origin-enriched sequences; SVvA3. simian virus 40; BrdUrd, bromodeoxyuridine. The ~eq-e~ce data in this paper have been submitted to the E M B L / G e n b a n k Data Libraries under the accession numbers X56241-50. Correspondence: M. Zannis Hadjopoulos, McGill Cancer Centre, Room 710, 3655 Drumnlond StreeL Montr/zal, Qu6bec, H3G 1Yf Canada,

vided further evide~.e for tl-.e use of specific replication origins in eukarTouc cells. Moreover, a 2.4 kb fragment from the 5' region of the human c-myc locus was demonstrated to promote episomal plasmid replication in eukaryotic nuclei [7]. Studies of DNA amplificat|on have also provided evidence: for the existence of specific initiation sites of DNA replication. Mapping of the chromosomal replication origins of the D H F R amplicon by two different procedures [8,9] allowed the identification of two such origins located approx. 20 kb apart, within the previously defined 28 kb initiation locus [10]. One of these two origins had been previously localized in a 4.3 kb fragment [111. These results were confirmed by an independent mapping procedure which was also used to dem,mstrate the presence of a replication origin in the APRT domain [12]. Recently, Vaughn et al. [13] presented evidence suggesting that initiation in the amplified DHFR domain occurs at multiple sites within the 28 kb early labeled region. Finally, the isolation of amplified DNA sequences in the form of autonomously replicating episomes containing the cad gone [14] or the

235 mdr-1 gene [15] is consistent with the presence of replication origins in these episomes. Several different methods have been employed for the isolation of DNA replication origins. Shot-gun cloning experiments performed in some laboratories have led to the isolation of human [16] and mouse 117,18] sequences with ARS activity in yeast, and, more recently, of human genomic sequences that enabled a defective Epstein-Barr virus vector to replicate in human cells [19]. Another approach, based on the binding properties of some genomic DNA fragments to the c-myc protein [201 or the p53 protein [211, led to the isolation of additional autonomously replicating sequences. Our laboratory developed a novel method whereby short nascent DNA strands containing newly activated replication origins can be extruded from small replication but~bles by branch migration [22]. We used this method for the isolation and cloning of origin-enriched sequences (ors) activated at the onset of the S phase 122,231. Initially, 12 ors clones obtained by this procedure were characterized in detail and their properties [241, as well as their primary nucleotide sequence [25] have been reported. Two of those ors are unique, two are members of the middle-repetitive O-family sequence, three contain the highly reiterated Alu family sequence and one contains the highly reiterated a-satellite sequence. At least four of them (ors 3, 8, 9 and 12) were found capable of autonomous semi-conservative replication upon transfection in HeLa cells [26] and have been termed "functional ors'; by electron microscopy all four ~ere found to contain replication bubbles, which for ors 12 were mapped within the ors sequence [26]. The same four ors sequences have also been found capable of initiating semi-conservative replication in an in vitro system that uses HeLa cell extracts (Pearson et al., unpublished data). h~ this paper we describe the replication patterns and sequence characteristics of 13 new functio .ai ors clones that were identified among our library of early replicating ors, either on the basis o~ ".'~mology to the previously characterized four functional ors (ors 3, 8, 9 and 12) or I y random selection. Materials and Methods

Cells and plasmids HeLa cells were propagated in Dulbecco's minimal essential medium (DMEM) supplememed with 596 fetal calf serum (FCS). pBR322 and pBR/ors plasmid~ were propagated in E. coli HB101 ant' plasmid DNAs were prepared as previously described [221. Screening of the ors library by dot blot hybridization Plasmid DNA from each clone was obtained from rapid plasmid preparations [27] dot blotted on a Gene

Screen Plus membrane (New England Nuclear, Boston, MA) and hybridized with nick-translated purified fragments of ors 3, 8, 9, or 12 at 65°C, as described previously [241. 12 positive clones were selected and similarly submitted t~ a second screening, using purified plasmid DNA. Dpni-resistance assay Logarithmically growing tlcLa cells were transfected with 3 /~g of pBR322 or 3 /~g or pBR/ors plasmid, consisting primarily of supercoiled monomers, by the calcium phosphate coprecipitation method described by Graham and Van der Eb [281 as modified by Chen and Okayama [29]. Following an overnight incubation at 37°C the cells were washed twice with PBS, and given fresh medium containing 5% FCS. 72 h later, lowmolecular-weight DNA was isolated by the method of Hirt [30]. extracted twice with phenol equilibrated in TE (10 mM Tris, 1 mM EDTA (pH 8)), and diethyl ether and finally precipitated in 7096 ethanol. The DNA was resuspended and digested with 3 U of Dpnl for 1 h to cleave unreplicated (fully methylated) input DNA 1311, then run on a 196 agarose gel, Southern blotted [321 on a Gene Screen Plus membrane (New England Nuclear, Boston, MA) and treated as described previously [26]. Alternatively, the Dpnl-digested DNA was used to transform E. coil for assaying Dpnl-resistant (autonomously replicating) plasmid in Hirt ~upcr,~ataat DNA. Semiconservative replication assay Cells were transfected by calcium phosphate coprecipitation [28] or by electroporation using the Gene Pulser (Bio-Rad) at 0.4 kV, 25 p F in a phosphatebuffered sucrose (0.095396 KH2PO 4, 996 sucrose, 1 mM MgCI, PBS (pH 7.2)). After washing with PBS. the cells were grown for 24 h in DMEM (10% FCS) containing bromodeoxyuridine (BrdUrd) at 12.5 # g / m l [261. Hirt [301 supernatants were prepared and centrifuged in cesium chloride (CsCI) gradients (initial refractive index 1.4150) in a Beckman vertical Vti 80 rotor at 65 000 rpm for 16-18 h, or at 70000 rpm for 4-5 h. Fractions of 250 ~tl were collected from the bottom of the gradient, and the refractive index (RI) of every second fraction was read using a refractometer (Fisher Scientific). Aliquots of 25 p.! (or 50 pl for dot blots) were taken from each fraction, diluted 1:1 with double-distilled water and were either electrophoresed on 196 agarose gel for Southern blot analysis or they were dot blotted. The blot transfer and hybridization were performed as described above. Sequence analysis Sequencing was performed using the commercial Sequenase kit (United States Biochemical Corporation (USB), CI. eland, OH). 2 p.g of denatured plasmid DNA were annealed to 30 ng of primer. Two oligo-

236 nucleot~dc primers (15-mer each) were used that are located 14 bp upstream (3') and 26 bp downstream (5') from lhe Nrul cloning site in pBR322. Replication proceeded in presence of the sequenase enzyme, I~S] dATP, and unlabeled dNTPs; the reaction was stopped by the addition of ddATP, ddCTP, ddGTP or dd'Vl'P, and the samples were run on a 8.3 M urea-65~, acrylamide gel. Sequence analyses were performed using the LKB 2020 DNAsis software or the BESTFIT-Nucleic Acids Sequence Analysis System (VAX - Universit~ de Montreal).

-II II -~

II1"--

Results

.

~'111 °1

I"Screening of the ors libra,, b.v dot blot hybridization

In order to identify ors clones with possible sequence homology to the four (ors 3, 8, 9 and 12) previously shown to be capable of autonomous replication [26], we screened by dot blot hybridtzation a library of 100 ors clones that we had generated from early-replicating monkey (CV-I) cell DNA [23]. The hybridization results are shown in Table I. Three clones, 2Aa2 (ors 13), 2Acl (ors 14) and 2Aa6 hybridized with ors 3 and ors 9; one clone 1Ba6 (ors 19) hybridized with ors 9 only; and two clones, 2Acl (ors 14) and lAc2 (ors 15) hybridized ~'ith ors 12. Results of nucleotide sequence analyses (see below) suggested that ors 13, 14 and 19 cross-react with different short stretches of the non-Alu portions of ors 3 and ors 9, whereas ors 14 and 15 cross-react with the a-satellite portion of ors 12. These five positive clones were then subjected to functional analyses, similar to those we had used previously for or:' 3, 8, 9 and 12 in order to assess wh~ther the apparent sequence homologies translated to similar function (i.e., ability to sustain autonomous replication). Dpn] resistance assay DNAs from ors 13, 14, 15, 19 :nd 2Aa6 plasmids

were Iransfected in HeLa cells and 72 h later Hirt supernatants weJe prepared as described in Materials and Methods. Plasmid DNA from Hirt supernatants was digested with Dpnl to distinguish input plasmids from plasmids replicated in HeLa cells. The plasmids

TABLE l S c r e e m n g o f the ors library by dot blot hyhridtzatton

+ + and + + . refer to the relative intensities of the hybridization signal; - , denotes absence of hybridization sigral. Probe

2Aa2

2Act

(orsl3)

(orsl4)

l At2 (orslS)

(orslg)

Ors3

+++

+++

-

-

Ors8

++

-

-

_

Ors 9

~ +

+ +

-

+ +

+ +

+ +

-

Ors 12

l Ba6

2Aa6 ++ + +

Fig. l. D p n l resistance assay of ors 13, 14, 15, 19 and pBR322. HeLa cells were transfected with 3 p g of plasmid DNA. The recovered pl; ~mids were digested with D p n l , electrophoresed on 1% agaros¢. blot-transferred, and probed with pBR322. The position of superc~,iled (I). nicked circular (It), and linear ( l i d plasmid D N A s is indicated on the left for ors 13, 14 and 15, and on the right for ors 19.

were not linearized prior to digestion with Dpnl, so that the formation of supercoiled (form 1) DNA product could be detected in addition to nicked circular (form II) and linear (form 111) DNAs: Dpnl-resistaat supercoiled molecules are thought to be a more reliable index of replication in animal cells [33l. The Dpnl-digested DNA was subjected to agarose gel electrophoresis, blotted and probed with pBR322 (Fig. 1). Four of the five ors plasmids, namely ors 13, 14, !5 and 19, tested positive in this assay, resulting in supercoiled (I), nicked circular !.l) and linear OlD Dpnl-resistant bands of 5to 50-fold higher intensity than those observed for the vector pBR322 ( D p n l resistance background). The recovered Dpnl-resistant plasmid DNAs were of the correct size expected for ors 13 (4451 bp), 14 (4501 bp), 15 (4637 bp) and 19 ( > 6863 bp), arguir,g against the possibility that replication of these plasnuds occurred as a result of their integration into, and then re-excision from, the cellular chromosomal DNA. Furthermore, previous electron microscopic analyses of replicating ors 3, 8, 9 and 12 [26], as well as other ors (our urpublished data) have indicated the presence of closed circular plasmids of the correct size. In addition to these five non-randomly selected ors clones, another 18 clones were picked at random from the library and similarly tested for function by the Dpnl resistance assay. Nine of the 18 (ors 16-18 and ors

237 (D I~. CO 0

00000

t " e4 ~

q ' It) M)

00000

D p n l treatment (i.e.. Amp 'Tot / A m p 'Tot ~ = approx. I), whereas after D p n l treatment invariably no

Amp~Tet ÷ transformants were recovered (data not shown). A time course experiment performed with ors 13, 15 and 20 using this method, over a period of 96 h, showed a doubling of D p n l resistant plasmids approximately every 24 h (data not shown), as was previously found with other ors [26].

"

o

-.-II "111

S e m i c o n s e r v a t i v e replication ( B r d U substitution)

"1

In order to obtain independent confirmation of the results obtained by the D p n l assay, we tested the replication of these ors plasmids by another assay for autonomous replication, namely the BrdUrd substitution assay. BrdUrd, a heavy analogue of thymidine, is incorporated into DNA during replication and increases the density of the replicated molecules causing them to band at t t~nsity higher than that of unsubstituted DNA, in a :esium chloride gradient. One cycle of replication in the presence of BrdUrd leads to a hybrid density (HL) DNA with only one strand substituted, whereas two or more cycles of replication lead to the synthesis of fully substituted (HH) DNA. First, the replication of ors 13, 14, 15 and 19 plasmids, as well as of the vector pBR322, was ~ssayed by density labeling with BrdUrd. Between 19 and 22 fractions were collected for each gradient; the refractive index of every other fraction was measured to verify the linearity of the density gradients and the samples were subjected to Southern blot analysis (Fig. 3). For all four ors plasraids a peak of unreplicated li3ht-light (LL) plasmid DNA is seen near the top of each gradient (centered around fraction 18 for ors 13, !~ and 19; and around fraction 16 for ors 14), as well as two additional peaks of HL (fractions 6-17 for ors 13 and 15, and fractions 4-14 for ors 14 and 19) and HH (fractions 1-5 for ors 13 and 15, and fractions 1-3 for ors 14 and 19) plasmid DNAs. For pBR322 all of the plasmid DNA is present in the form of LL DNA (fractions ~:~,~-21), indicating that the vector by itself did not undei~o a~ly replication under the same conditions. Although the mass of the unreplicated pBR322 detected by this method is less than the mass of the unreplicated portion of the ors plasmids, we have verified both by transformation of E. coli and Southern blot analyses that the amounts of pBR322 and ors DNAs taken up by the transfected HeLa cells are comparable, as are those recovered in the Hirt supernatants (data not shown). Similar results were obtained when ors 16, 17, 18, 20, 21, 23, 24 and 25 were tested for semi-conservative rt:plication, except that these ors were analyzed by dot blot hybridization. The results, plotted as DNA content in each dot, as measured by densitometer scanning, are shown in Figs. 4 and 5. For ors 17, 18, 20, 21, 23 and 25 two peaks of higher density (HL and HH) DNA are observed, indicating at least two rounds of replicatiot~

i

_ _

A B Fig. 2. Dpnl resistanceassay of: (A) ors 16. 17, 20-26 and pBR322; and (B) ors 18, 21 and pBR322. Transfectionsand assays were done as in Fig. I.

20-26) tested positive by this assay, giving rise to D p n l - r e s i s t a n t progeny DNA bands (Fig. 2). The production of D p n l - r e s i s t a n t supercoiled (form I) DNA can be clearly seen for ors 20, 22, 24, 25 and 26 (Fig. 2A), is faintly visible for ors 16 and 17 and not detectable for ors 23. Ors 18 and 21 revealed Drml-resistant

bands in separate experiments (Fig. 2B~ but, in general, required longer exposure tim,.s; lower detectable levels of D p n l - r e s i s t a n t DNA generated by the latter ors plasmids might be the result of either lower overall replicating efficiencies by comparison to the other ors, as also seen previously [26], or failure by this assay to completely discriminate between replicated and unreplicated DNA, as has also been reported for polyoma virus [34]. Ors 22 did not give consistently positive results in repeat experiments of D p n l resistance nor in subsequent functional assays and was presumed to be overall replication negative. As an average of many independent experiments, we have estimated the amount of the recovered D p n l - r e sistant material to be approx. 300 pg to 3 ng (6 • 104 to 6 . 1 0 ~ copies) or between 10 -4 ~o 10 -~ of the input (3 /~g) plasmid. For further verification of the D p n i - r e s i s t a n t ors clones and quantitative comparisons of D p n l - r e s i s t a n t ors plasmid versus the vector alone, we cotransfected HeLa cells with the ors (Amp+Tet - ) and pBR3222 (Amp ÷Tet + ) and used the Hirt supernatants before and after D p n l treatment to transform E. coil, as previously described by Vassilev and Johnson [35]. The results of multiple experiments showed that at 72 h post-traasfection the number of E. coil transformants generated by the two plasmids was approximately the same before

238 of the lrtptJt p l a s m i d D N A s . In the e x p e r t m e n t d e p i c t e d in Fig, 4, o r s 16 h a s s e e m i n g l y u n d e r g o n e o n l y o n e r o u n d of replication, b u t a peak c o r r e s p o n d i n g to H H D N A h a s been s o m e t i m e s o b s e r v e d in s e p a r a t e experim e n t s ( d a t a n o t shown). Similarly, the m a j o r i t y of the replicated D N A for o r s 24 is recovered as H H material (Fig. 5), s u g g e s t i n g that only a s m a l l fraction of t h e i n p u t D N A h a s been used as t e m p h t t e for at least two r o u n d s of replication, Typically, however, b o t h H L a n d H H p l a s m i d D N A s were recovered in the m a j o r i t y of

the cases. These data are characteristic of a semiconservative mode of DNA replication and argue against repair-type synthesis. Incorporation of BrdUrd by nick-repair results in broad DNA distribution front LL to HH positions instead of the characteristic well defined HL and HH peaks that arise from semiconservatire DNA replication [36]. Moreovt~r, density shifts, such as those obtained with the o r s plasmids, are not observed when the vector alone, without the inserted o r s fragment, is analyzed.

I 4ZO ¸

L,~o.

I 416 ¸

=

~

1.401

:QO=' ÷ F,~ctlon

1

4

I

I

i¢

,'

Fr4¢lion

,;

=~

,:

,

I|

, o , II

ors

13

ors

t*o*

8: i.loi

, *t = - ~

÷

t*~¢

F~=CIto~

14

i.*o* 1.4¢0

4. z

,

Fr=¢l~on

ors

15

Fr=¢tlon

pBR322

*

,~

ors

19

Fig 3, BrdUrd incorporation in o r s 13. 14, 15, 19 and pBR322. HeLa cells were transfected with 5 lag of plasmid DNA and incubate! in medium containing BrdUrd at 12.5 #g/ml. At 24 h post-transfection, the cells were lyzed, and the recow:red plasmid DNAs were c,mtrifuged to equilibrium on CsCI gradients. Fractions were collected and an aliquot from each fraction was subjected to blot-hybridization analysis as in Fig. 1. The refractive index (RI) of every second fraction was taken for each gradient, and the resulting profiles are plotted above each gel. The position of :,;upercoded (I), nicked circalar (IlL and linear (!I!) plasmid DNAs are indicated by the arrows.

239 The density scanning of all dot blots indicates that between 12% (ors 24) and 85% ( o r s 23) of the ors plasmid DNA recovered in the Hirt supernatants, consist of newly replicated (HL and HH) DNA; this is not always apparent by Southern blot analyses of D p n l resistant bands (compare for example ors 18 and 23 in Figs. 2, 4 and 5), suggesting once more a lesser sensitiv-

Ors

ity of this technique for detecting ors replication, as previously discussed [261. Sequence analysis

The primary nuclec,tide sequences of 1! of the above newly identified functional ors (excepting ors 19 and ors 21) were obtained using the Sequenase dideoxy

Ors 17

16

4206

-

-

25

1 4~00

- ~5

' ~0

4

)0-

'.

i 4C50 -__

4r, q

................

I

2

_:=. -7-.-~:

""

3

4

"1 ,

,

.........................................

~1 I 0

~

II

I~ i]

il

-%~

2_

16 17 10 13 d )

l~

|'r~ct

Frac(|orl

•

Ors 1 4~00

OOOOoOOOOOO,,

t~n •

* • oO0~JI~Oe,

Ors 20

18 I 4200

.

.~\

aO

;\.\

•"-...

_

I 4150 --

I (IO0

--

14650

~

t 4100

~7-il

--

k 2

I 1050

-n

h

3

4

5

5

", 8

I 4!,'0t~ . . . . . . . [ 2 3

9 lUtll~lS1415161zlS19202122~3

Fractmn

•

o • OOO0

•

•

•

,'

, •

;

~ "t

~

6

7

i24 . . . . . '' B 9 loll

' I~'LI4

) 1516~7l~192021

Frtlct io#i

• •

e ~O00

e

o~oo, adBil~BO

O0 • • o

~

pBR322 ----T _

\',,

:

I Io¢,o

,

2

~

3 4

i

i

i

=1

~-~

,

i 1

l~

"'°

o

5 (l 7 G fb loll I g 1 3 1 4 1 5 ~ 6 I T l S l [ a g 0 2 1

rractlon Q

~

O

• •

Fig. 4. BrdUrd incorporation in o r s 16, 17, 18, 20 and p B R 3 2 2 . Transfections and assays were d o n e as in Fig 3, except that fractions were analyzed by dot-blot hy*~ridization. The RI of each fraction as well as the histograms generated by densitometry scanning o f the dots, are plotted above each blot.

240 chain termination method (Fig. 6). The sizes of these or.~ range bstweer. 88 bp ( o r s 13) and 274 hp (ors 15). Ors 19 and ors 21, which have not been sequenced yet, are approx. 2.5 kb and 800 bp, respectively, as estimated by tl-eir rate of migration on agarose or polyacrylamide gels aloogside size markers. The primary sequence data were subjected to computer-assisted anal~¢sis using the DNAsis software or the BESTF1T sequence analysis program, in a ~imilar manner to that we used previously for the analysis of eight o r s clones that are members of the same early-tel licating monkey ors library [25]. Briefly, our current analysis centered around the same salient features that had emerged from our previous study, namely the search for A/T-rich regions, inverted repeats, repetitive family sequencer and known consensus sequences [25]. As before, we found that the newly identified functional o r s contain extensive ATrich regions (i.e., regions in which st;etches of between 10 bp and 67 bp are >i 70% AF); however, only o r ~ 13

Ors KtO;"

.

(65.95), o r s 16 (63.6q~,), o r s 20 (63.2%), o r s 24 (69.2%) and o r s 25 (626%) have higher overall AT content than normal monkey (CV-1) D N A (56%j while o r s 17 (52.7%) and o r s 26 (51.3%) have a lower overall AT content than the normal. Nearly all of the eleven sequenced o r s have one (ors 14, 15, 17, 18 and 25), two(ors 16 and 20) or three (ors 24) stretches of five or six consecutive As or 'I',~ the exceptions being o r s 13, 23 and 26, which are also the smallest in size. A few small direct repeats were found, as before [25] N o significant open reading frames were observed in any of the ors clones. The computerassisted analysis revealed significant homology of o r s 14, 15 and 23 (97%, 95% and 84%, respectively) to the African green monkey a-satellite repetitive sequence family [37]; and 70% homology of o r s 17 to the Alu repetitive family consensus sequence [38]. Ors 26 showed 90% homology to human and bovine mitochondrial D N A s 139-411 and was not studi,-d further. N o other o r s s~:)wed significant homology to any sequence listed

~1

O r s ~:3

-

I 42'))0 ~

............................................

1~

" •

÷

I~

--

1,)

2')

4:5~7 • ':

"O

e'~ I ~ 9 5 , ,

. . . . . .

r

::

1

~

3

4

5

6

7

O

I')li

1~13141516171BI~0~I

Frttctton

Fract~orl

~B,O O 0

~'-t

O odnh'at'dL~ O O • • • • o O ' ~ b

()t'~ "51

O r s 25

i [ 4150

--

e~

L~

] ' i ,L' [ t ' ~ I i

• . o•

•

• Q O O ~ S

• • oootoe

Fig. 5. BrdUrd incorporation in ors 21.23, 24 and 25. As in Fig. 4.

• • •

e O0~)000

241 in G E N B A N K (GenBank R55.0; March, 1988); nor to the scaffold attachment regions (SAR) A and B ~42]. as we have found previously with other o r s 125]; the core origin of SV40 [43,441; or the sequences p H L M Y C I 120] and pRJ531 [21], which have been reported as autonomously replicating. Comparison w0h Ihe yeast ARS conse~sus sequence W r I ' T A T R T T I ' W [45] revealed a 10/1 ! match for o r s 20, and 9/11 matches for o r s 16, 24. and 25. A total of !1 perfect inverted repeats with free energies favorable to cruciform formation (i.e., ~< - 1 0 kcal) and with stems varying in size from 5 to 10 bases, were found in 7 of the 11 o r s , as shown in Fig. 7. As we have previously shown, the likelihood of finding 11 energetically favorable inverted repeats in a single sequence of 1383 bp (i.e., if the seven sequences were linked to give a single sequence of 1383 bp) is significant [25]. We have reported previously that earlyactivated monkey replication origins are ennched for inverted repeats [24,461 and have presented evidence suggesting that cruciforms form at or near origins of D N A replication [471. Finally, the nucleotide sequence of these new functional o r s were compared among themselves as well as with o r s 3, 8, 9 and 12, and except for the clones containing a-satellite ( o r s 12, 14, 15 and 23) or Alu sequences ( o r s 3, 9 and 17), no extensive sequence homologies were found, confirming and reinforcing our previous findings that no extensive sequence homologies exist among the early-replicating o r s [25]. Curiously, o r s 17 was not identified by the dot blot hybridization (la

011513

bp}

a l v z r ; A r : ~ A A G T I I t ( A r t ; Tr': ;, rl