0022-538X/82/110586-09$02.00/0
Copyright©1982, AmericanSocietyforMicrobiology
Organization
of Chicken DNA
Sequences Homologous
to
the
Transforming
Gene of Avian
Myeloblastosis Virus
I.
Restriction Enzyme Analysis of Total DNA from Normal and
Leukemic Cells
BERNARD PERBALANDMARCEL A. BALUDA*
UCLA School ofMedicine andJonsson Comprehensive Cancer Center,LosAngeles, California90024 Received 19 April 1982/Accepted 29 June1982
Hybridization probes consisting of cloned DNA recombinants whichrepresent
different regions of the leukemogenicsequence(amv) from avian myeloblastosis viruswereusedtocarryoutamoredetailed restriction endonuclease analysis of the homologous sequences (proto-amv) presentin normal and leukemic chicken DNA.The results show that four large introns interrupt the normal cellular proto-amv sequences and that there is no majorrearrangementof these sequences in leukemicmyeloblasts.
The leukemogenic potential of avian
myelo-blastosis virus (AMV) has been associated with
asequence (amv)of approximately 1.2kilobase pairs (kb) whichhasreplacedtheviralenv gene
(18, 19). The amv insert is homologous to a
unique chicken DNA region (proto-amv) which isinterrupted by unrelated DNAsequences(12). Intronsare presentwithinthecellular homologs ofmostviraloncogenes. Forinstance, the
chick-en cellular homolog of the src gene of Rous sarcomavirus (PragueC strain) containsatleast five introns (16), and cat DNA sequences ho-mologoustotheoncogeneof theSnyder-Theilen feline sarcoma virus contain at least three
in-trons(9).
Tofurther characterize the arrangement of the
proto-amv sequences in normalchicken DNA, DNArecombinant clones representing different regions of amv were prepared and used as
hybridization probes with C/O chicken DNA
treated with various restriction endonucleases
eithersinglyorin combinations.Asimilar analy-sis was performed with DNA from leukemic chicken myeloblasts. These studies revealed
that in normal chicken DNA, the proto-amv sequences areinterrupted by atleast four large introns, and that there is no major reorganiza-tion ofthese sequences inleukemic DNA.
MATERIALS AND METHODS
Purification of total DNA. Purification of pBR322
DNAwasperformedasdescribedbyCurtissetal.(8). High-molecular-weight DNA (4.5 kb or larger) was
purified fromC/G andC/E whole chickenembryosor
tissuesasdescribed earlier(1).
Purification of DNAfragments. TheDNAfragments
obtainedby restriction enzymetreatment ofpBR322 hybrid clones were purified by electroelution as
al-ready described(12).
Preparation of Southern blots.The DNAfragments
obtained after electrophoresis in agarose gels were blotted onto nitrocellulose Milliporefilters (Millipore
Corp., Bedford,Mass.) under the conditions described earlier(12, 17).
Nick translation.Conditions used for 32plabelingof DNAfragments by nick translation (13) werereported
earlier(12).
Cloning. Escherichia coli HB101 (4) was the recipi-ent and pBR322 was the vector in transformation
experiments. Ligation conditions and transformation procedures were previously described (12). Screening of the transformant clones for the nature of their
plasmid DNAcontent was performed bya modifica-tionof the method described by Birnboim and Doly (3). Individual transformants were grown overnight at
37°Cin 5 ml of TYE medium(10 g of Tryptone [Difco Laboratories,Detroit, Mich.],5 gof yeast extract, and 5 gofNaCl per liter) containing 100 ,g of ampicillin per ml. A1.5-ml volume of the cell suspensions was thentransferredto anEppendorf tube and spun for30 s inan Eppendorf centrifuge. Afteraspirationof the
supernatant, the cell pelletwas blended inaVortex mixer for 15 s and suspended in 100 ,ul ofafreshly prepared lysozyme solution (25 mM Tris-hydrochlo-ride [pH8.0]-10 mM EDTA-50 mMglucose-5 mg of lysozyme per ml [Sigma Chemical Co., St Louis, Mo.]). After15min of incubationat4°C, 200,ulof0.2 NNaOH-1% sodiumdodecyl sulfatewasadded,and the solutionwasmixedvigorously. Afteranadditional 5minonice, 150,ul of 3 M sodiumacetate(pH 5.0) wasadded, and the mixture was left at4°C for1h. The precipitate which formed after addition of thehigh salt solution was spun down for10 min at4°C, and the
supernatant was extractedonce with400Rl ofa1:1
phenol-chloroform-isoamylmixture
(chloroform-isoam-yl is 24:1) and twice with 400RIof chloroform-isoamyl
586
on November 10, 2019 by guest
http://jvi.asm.org/
(24:1).Nucleic acids were precipitated at -70°C for 30 min after the addition of 2 volumes of absolute etha-nol,pelietedbycentrifugation,and suspended in 60 ,ul of sterile distilled water. Usually, 5 to 10p,l ofthis suspension was used to run analytical digests in the presenceof 10 ,ug of RNase A.
Colony hybridization. The method described by
Grunstein and Hogness (11) was used. The washing
and sterilization of the filters before seeding with bacteria could be omittedwithout alteringthequality
of thehybridizationobtained ifampicillin plateswere used.
Hybrkzation.The DNAblots and the
colony-bear-ingifiterswerehybridizedasdescribedpreviously (2,
12).
Preparation of amv-spedclc subelones. The pBR
HAX4 clone (12) was used to prepare a subclone
containing the amv sequences located between the
HaeIIandthe EcoRIsites foundwithin the 1-kb amv insert inHAX4.The1.0-kbamvfragmentwasisolated
from pBRHAX4byBamHIdigestion, electrophoresis
in a 0.8% agarose gel, andelectroelution. It wasthen
digested withEcoRIendonuclease,andthe two
result-ingfragments (0.80 and0.20kb)wereseparated on a 2% agarose gel. The 200-base pair fragment
corre-spondingtothe 5'portionwasthen cloned inpBR322
EBpreviously electroelutedfrom a0.8%agarosegel
afterdoubledigestionwith BamHIand EcoRI endonu-cleases. Screening oftheresulting
tetracycline-sensi-tive colonies revealed that sometransformantscarried the expected sizeinsert, and hybridization with
32P-labeledHAX4showed thattheycontained thespecific
amvfragment. Oneof these transformants(pBREB3)
wasused in this work.
Clones SES3 and SX12, which contain the amv sequences located between theEcoRI-Sal sites and theSail-XbaIsites,respectively, wereprepared from
the pBR-EX11 clone(12).
Digestion of pBR-EX11 with Sail generated two DNAfragments(0.75and 4.4kb)whichwere separat-edbyelectrophoresis in a1% agarosegeland recov-ered by electroelution. The 4.4-kb fragment which contained the 3' portionbetween SalIandXbaI also contained thepBR322sequenceslocated between the BamHIandSalIsites(includingthepBR322replicon). Therefore, the electroeluted DNAfragmentwas
self-ligatedand useddirectlytotransform E. coliHB101. Thetransformantswereshowntocontain
amv-specif-ic sequences if screenedbycolonyhybridizationwith
purified 32P-labeled HAX4DNA.One of these clones
(SX12)wasused insubsequent experiments.
The other fragment (0.75 kb) generated by SalI
treatment of pBR-EX11 was subcloned in pBR322 previously digested with Sall and then was treated with alkalinephosphatase under conditions described earlier (12). Tetracycline-sensitive clones were ana-lyzedby bothminiscreeningandcolony hybridization with32P-labeledHAX4DNA.One of the cloneshaving
thecorrect insert(pBRSES3)was used as aspecific probe for the amv sequences located between the
EcoRIandSall sites.
Thederivation of these new amv subclones is sum-marized inFig. 1.
Preparationof theproto-amvE3probe. Probe E3 was derivedfrom theX-chickenrecombinant clone no,111 which contains twoEcoRIfragments (2.0 and 8.7 kb) of proto-amvsequencescorrespondingtothe3'end of
the amv sequence (12). The 8.7-kb EcoRI fragment
waspurified byelectroelution after digestion ofAlll
DNA withEcoRJendonuclease andelectrophoresis in a0.8% agarose gel. The 8.7-kbelectroelutedfragnent
(E3) was 32P labeled by nick translation to obtain
hybridizationprobe.
Physicalandbiological containment. This workwas carriedout atthe P2-EK2containmentlevels
accord-ingtotherevised guidelines of the National Institutes
ofHealth.
RESULTS
Analysis ofnormal C/O chicken DNA. (i) Hy-bridization with HAX4 probe. Todetermine the relativeposition in normal chicken DNA of the EcoRI,HindIII, and BamHI endonuclease sites which generate proto-amv DNA fragments, dou-ble enzymatic digestions were performed, and the resulting Southern blots werehybridized to the 32P-labeled HAX4 probe which contained mostof amv but no viralsequence. Single EcoRI andHindIII DNAdigests wereincluded within the same gel as internal controls in addition to standardsize markers.
The blot from a HindIII + EcoRI digest
AMV Y,T 9
I
w ,nRR02 /lAY4 _ _'
-pOnIC'In#A-t
(1.0 kb)
pBR322/EXII
IjHoeI1
I H
E o8
C w
111
oEt
ft
5.41
en x CD
0.8
kb)
ktb
ff
pBR322/EB3 5
(0.2 kb)
t
Hoel tc t
pBR322/SES3
(0.35 kb) t 1
0 E 76
pBR322/SX12 W
(0.45 kb)
3,
MBomHI linkers
FIG. 1. Derivation of the amv-specific subclones. The restriction endonuclease sites in AMV are mapped as previously reported (12, 18). Symbols: V,
HindIII; *,BamHI; V,XhoI;U,BgllI;C,EcoRI; 0, XbaI; 0, KpnI. Thesubcloning ofHAX4and EX11 from AMV has been described(12). In probes HAX4 andEB3, theHaeIIsiteoriginally present in amv has been removedduring BamHI linker ligation.
44,
1.
4
Q n
on November 10, 2019 by guest
http://jvi.asm.org/
[image:2.491.252.443.332.603.2]a Gcc
kD 293.
__ .
_4__3a
V~~~~~~~~!A
FIG. 2. Hybridization of restriction enzyme-treat-edC/O chickenDNA withHAX4. C/Ochicken DNA wasdigestedwith eitherHindIII (d), EcoRI (c),
suc-cessivelywithHindlIl andEcoRI (b),orwithHindIlI
andBamHI (a).Thedigested DNAwas
electrophor-esed in 0.8% agarose, transferred to nitrocellulose filters(Southem blots),andhybridizedto32P-labeled
HAX4 DNA. A mixture ofHindIll-digested ADNA and HaeIII-digested 4X174 RF DNA was run in parallel and used asmolecular size markers, expressed inkilobasepairs (kb).
showed four bands of 1.2, 1.9, 2.2, and 4.3 kb (Fig.2,lane b), whereas EcoRI or HindIII alone yielded the band pattern previously reported (12). Threefragments(2.1,5.4, and 8.7kb)were detected in the EcoRIdigest (Fig. 2,lanec),and twofragments (1.3 and 5.2kb)weredetected in the HindIII digest (Fig. 2, lane d). The 1.3-kb HindIlI fragmenthadpreviouslybeenestimated tobe1.4kb (12).
TheBamHI + HindIII digest revealed three fragments (1.3, 4.3, and 5.2 kb) with the 32P_
labeled HAX4probe (Fig. 2a). Previously (12), we had reported that BamHI digestion of C/O DNA generates two proto-amv fragments (5.4
and 19.6kb).
A similar hybridization pattern was obtained with DNApurified fromembryonic fibroblasts, spleencells,oryolksaccells obtained fromC/E chickens(datanot shown).
Theproportionofproto-amv sequences
pres-ent within the EcoRI and HindIll fragments
cannot be accurately determined, and
conse-quently the relativeintensityof the bandscannot
beused forquantitative estimations.
Neverthe-less, the strong intensity of the 2.1-kb EcoRI
band suggested thepossibility that itrepresents hybridization of HAX4 with two proto-amv DNA fragments of similar size. Previously, we
had shown thatthe EcoRI 2.1-kb fragment de-tectedwithHAX4included sequences located in
amv beyond the SalI site towards the 3' end. Therefore, an EcoRI + Sail digest was per-formedontotal C/O DNAtodeterminewhether this 2.1-kb band correspondedto adoublet. The resultsshowed that in additiontothe EcoRI 8.7-and5.4-kb bands, twootherbands, of 1.45 and 2.0kb,werepresent(Fig. 3). The lowintensity of the 1.45-kb bandrelative to that of the 2.0-kb bandpresumably resulted from inefficient cleav-age of the 2.0-kb fragment with Sall. In three independent Sall treatments, it was notpossible
to obtain complete digestion if total chicken DNA was used as substrate. The 1.45-kb band also hybridized with the SES3 probe, which consisted of the amv sequences locatedbetween the EcoRI andSail sites(datanotshown).This positioned the Sall site at 1.45 kb downstream from the EcoRI site within the corresponding proto-amv region, thereby revealing an intron between these two enzyme sites. The position of the Sail site was confirmed by analysis of A proto-amv recombinantclones presented in our second paper on this subject (B. Perbal, J. M. Cline, R. L. Hillyard, and M. A. Baluda, submit-tedfor publication). The presence of the intact 2.0-kbfragment and the 1.45-kbfragment is in agreement with the possibility that the 2.1-kb bandgeneratedbyEcoRI alone is adoublet.
To determine the location of the additional EcoRI 2.0-kb proto-amv fragment, in normal chicken DNA, several double enzymatic diges-tions were performed on C/O DNA, and the resulting blots were hybridized with DNA probes representing different portions of the amvsequence. Thepreparationandlocation of
2O -.3.7
4.3
FIG. 3. Doubledigestionof C/O DNA withSall + EcoRI. C/O DNAwasdigestedwithSail andEcoRI endonucleases in the EcoRI running buffer. After
electrophoresisina0.8% agarosegeland transferto nitrocellulose,theresultingblotwashybridizedto
32P-labeledHAX4DNA.HindIll-digestedA DNAwasrun assize markers.
on November 10, 2019 by guest
http://jvi.asm.org/
[image:3.491.112.192.75.248.2] [image:3.491.335.373.454.611.2]proto-amv ORGANIZATION IN CHICKEN DNA 589
TABLE 1. HybridizationofrestrictionenzymefragmentsfromC/OchickenDNAwithdifferentamVprobes Fragnents(kb) detected withamv probe:
Enzymes HAX4 EB3 SES3 SX12
HindII + EcoRl 1.2,1.9, 2.2, 4.3 2.2 1.9 1.2, 1.9,4.3
BamHI + EcoRI 2.0,5.4 5.4 2.0 2.0, 5.4
SmaI+EcoRI 2.0, 4.9, 8.7 4.9 2.0 2.0,8.7
Hindlll+BamHI 1.3, 4.3, 5.2 5.2 5.2 1.3,4.3
theseprobes withinamv aredescribed in Materi-als andMethods and shown in Fig. 1.
(ii) Hybridization of double enzymatic digests
with subclones of
HAX4.
Four differentdouble digests were set up. In addition to HindIlI, EcoRI, and BamHI endonucleases, SmaI was used because of the location of its recognition site150 base pairs towards the 3' terminus from the EcoRI site in amv (12, 14), The double digests were electrophoresed in 0.8% agarose gels, transferred to nitrocellulose, and hybrid-izedwith one of the probesHAX4,
EB3,SES3, orSX12 (forasummary, see Table 1).(i) Among thefourproto-amv fragments gen-erated in the HindlIl + EcoRI digest (Fig. 4, lanesd), the2.2-kbfragment hybridized with the amv sequences locatedbetween the HaeII and EcoRI sites (probe EB3), whereas the 1.9-kb fragment hybridized with both the SES3 and SX12 probes which include amv sequences
lo-cated between the EcoRI and XbaI sites. The
faintness
ofthe 1.9-kb band detected with the SX12probe indicatesthatonlyashort stretchof amv sequences which layon the 3' side of the Sall site ofamv was present in this fragment.Thetwoother
fragments
(1.2 and 4.3kb) gener-ated in the EcoRI + HindlIl digesthybridized only with the SX12 probe. These findings indi-cated that the 3'portion ofamvfoundbetween the Sail and XbaI siteswasdistributed in three separate proto-amv fragments generated by EcoRI and HindIll double digestion, i.e., 1.9, 1.2, and 4.3 kb. Therefore, this portiQn of amv mustbelocated inatleast two domains of exonin proto-amv.
A comparison of the sizes of the DNA
frag-ments detected after a single digestion with either EcoRI or Hindlll (12) with the results obtained above allows us to order the
fragments
obtained in the HindlIl + EcoRI digest as fol-lows from the 5' to the 3' end: 2.2, 1.9, 1.2, and 4.3 kb. Therelative order of the 1.2- and 4.3-kb fragments was deduced from the order of the HindlIl fragments: 5.4, 1.3, and 5.4 kb,
assum-ing that the1.2-kbfragmentin thedouble EcoRI
+HindIII digest correspondstotheHindlIl 1.3-kbfragment. Thiswasconfirmed by the results obtained with the HindlIl + BamHI double digestion (see below). The appearance of the 2.2-kb
fragment
fixedthe position ofaHindlIla b c d
1-HAX4
FIG. 4. Hybridization ofC/O DNA double digests with different amv probes. C/O DNA was digested successivelyby eitherHindlllandEcoRI(d), BamHI and EcoRI (c),SmaIand EcoRI(b),orHindllland BamHI (a). Afterelectrophoresis in 0.8% agarose gels and transfertonitrocellulose, the blotswerehybridizedto the indicated probes 32Plabeledby nick translation. The size markers were eitherHindIII-digested A DNAor a
mixtureofHindIII-digestedA DNAandHaeIII-digested+X174RF DNA.Thegelused for SX12hybridization waselectrophoresed for a longer period than were the other gels.
VOL.44, 1982
on November 10, 2019 by guest
http://jvi.asm.org/
[image:4.491.98.385.452.615.2]site within the 5'-proximal 5.4-kb EcoRI frag-ment, and the presence of the 4.3-kb fragment fixed theposition ofaHindIII site within the 3'-proximal8.7-kb EcoRI fragment.
Since only two bands (5.2 and 1.3 kb)were
detected if a HindIII digest ofC/O DNA was hybridized to HAX4 (see above), the present resultssuggest that there isasecond 3'-proximal HindIIIfragmentof 5.2 kbwhich
overlaps
with the EcoRI 8.7-kb fragment and includes the HindIII-EcoRI 4.3-kb fragment. This would place mostof the 1.3-kb HindlIl fragmentwithin oneof the two EcoRIfragmentsof 2.0 and 2.1 kb detected with HAX4 and the 1.9-kb EcoRI-HindIIIfragment withinthe other one.(ii) When a double BamHI + EcoRI digest (Fig. 4, lanes c) ofC/ODNAwas hybridizedto HAX4
32P-labeled
DNA, two very dark bands (5.4and 2.0kb) were detected. The 5.4-kb band was probably a doublet consisting of the 5'-proximal 5.4-kb fragment detected after diges-tion with EcoRI alone and the3'-proximal5.4-kb fragment detected after digestion with BamHI alone (12). The 2.0-kb band was also a doublet consisting of the 2.0-kb fragmentgenerated by EcoRI alone and a 2.0-kb fragment resulting from the presence ofaBamHI sitenear the 3' end of the 2.1-kb fragment also generated by EcoRIalone.Hybridizationof the BamHI +EcoRI double digest with the probe EB3 revealed the 5'-proximal 5.4-kb EcoRI
fragment,
whereas hy-bridizationwith theprobeSES3 revealedonlya2.0-kbfragment. TheprobeSX12hybridized as
expectedwith twofragments of 2.0 and 5.4kb, the latter presumably corresponding to the 3'-proximalBamHIfragment
previously
described (12).(iii) SmaI + EcoRI double digestion (Fig. 4, lanesb) generated three bands(8.7, 4.9, and 2.0 kb) which hybridized with HAX4. The 4.9-kb band hybridized to the EB3 probeand allowed us to localize a SmaI site within the 5.4-kb EcoRIfragment.Thesizes(8.7and 2.0kb)of the two other bands corresponded to the sizes of fragmentsgeneratedbyEcoRIalone.
IfC/O DNA digested with SmaI alone was
hybridized toHAX4, two bands (16 and6.0kb) were detected (data notshown). The existence of the 6.0-kb SmaI fragment and the results given above establish that there isoneSmaIsite
at0.5 kb from the 5' end within the 5.4-kb EcoRI fragment whichhybridizes with the EB3 probe. Also, there must be another SmaI site 1.1 kb outside the 3' endof the same
EcoRI fragment
since a 4.9-kb fragment which hybridizes with the EB3 probe was generatedbyadoubleSmaI
+ EcoRI digestion. Because the 2.0-kb SmaI-EcoRI fragment hybridized with the SES3 probe, it contains amv sequences located
be-tweenthe
EcoRI
andSallsites andmustcorre-spond
to the 1.9-kbEcoRI-HindIII
fragment
which also
hybridizes
to the SES3probe.
Wecanassumethatasin amv,theSmaI site in this proto-amv
region
islocated150basepairs
tothe 3' side of the EcoRI site.Therefore,
thisEcoRI sitewould belocatedat0.95kbonthe 3'sideof theEcoRIsite which is at the 3' end of the5'-proximal
5.4-kb EcoRI fragment. This was de-terminedasfollows: SmaI-SmaI (6.0kb)minusSmaI-EcoRI
(4.9
kb)
minus EcoRI-Smal(0.15
kb)
equals
EcoRI-EcoRI(0.95kb).
Thus,there isanextraEcoRIsite within the proto-amv
region
underdiscussion, establishing
the existence of another intronconsisting
of at least 0.95 kb betweentwoEcoRI
sites. A similar conclusioncanbedrawn from the results obtainedwiththe Hindlll + BamHI double
digestion
of C/O DNA. Theexistenceof the 0.95-kb EcoRIfrag-ment wasconfirmed
by
analysis ofA proto-amv recombinantclones (Perbal etal. submittedforpublication).
(iv)
Doubledigestion
with HindIII + BamHI(Fig. 4,
lanesa)andhybridizationtoHAX4gave risetothreebands of5.2,
4.3, and 1.3 kb. The 5.2-kbband,
which alsoappearedafterdigestion
with Hindlll
alone,
wasdetected afterhybrid-ization with the EB3 andSES3probes,
confirm-ing
the presence ofthis HindIllfragmentonthe 5' side of the proto-amv sequences. This5.2-kb HindIIIfragment
waspreviouslyshowntocarry proto-amv sequences common to both the 5.4-andthe 2.0-kb EcoRIfragments(12).This obser-vation alsoconfirmsthatthere isnoBamHIsite inthe 5.2-kbHindlIl fragment (12).The 4.3- and 1.3-kb fragments generated
by
HindIlI + BamHI digestion were detected
by
the probe SX12. This suggests that the 4.3-kb fragmentwasthe result ofaHindlllcutin the 3'-proximal5.4-kb BamHI fragmentwhich ispart of the
EcoRI
8.7-kbfragment. The1.3-kbfrag-ment was generated by HindlIl alone and is
equivalent
to the 1.4-kb HindIll fragment de-scribedpreviously(12).
These results show that the proto-amv
se-quences corresponding to the portion of amv
delineatedbythe HaeII andSail sitesare locat-edon asingle5.2-kb HindIlI fragment,whereas the proto-amv sequences corresponding to the
amv sequences located between the Sall and XbaI sitesarelocatedontwoadditionalHindlll
fragments.
One of them (1.3 kb) hasalready
been characterized (12). The other contains a
BamHIsite such thataHindIII+BamHI
diges-tion generates a4.3-kb fragment which hybrid-izes with the SX12 probe. Since only two
HindIII bands (5.2 and 1.3 kb) were detectedby hybridizationwithHAX4, the 5.2-kb bandmust
beadoublet. Thiswas confirmed by hybridiza-tion of HindIII-digested C/O DNA with the
J.VIROL.
on November 10, 2019 by guest
http://jvi.asm.org/
CHICKEN 591
a b
-23.7
a, -- 9.5
- 6.6
_W 4.3
- 2.1
1.9
1.35 1.0
-0.87
FIG. 5. Hybridization ofEcoRI-or
HindIII-digest-edC/ODNAwith the proto-amv E3 probe. C/O DNA
digested with either EcoRI (a) or HindIII (b) was
electrophoresedin agarose gel,Southernblotted, and
hybridizedto 32P-labeled E3 DNA,which represents the 3' end of the proto-amv sequences. A mixture of
HindIll-digested A DNA andHaeIII-digested
+X174
RFDNA was used as molecular size markers.
proto-amv E3 probe which corresponds to the 3'-terminalamvsequences(seeabove).
(iii)
Hybridization
ofC/ODNA withproto-wnv E3probe. C/ODNA wasdigested with HindIIIor EcoRI endonuclease, electrophoresed in a
0.8% agarosegel, blot transferred, and hybrid-ized to 32P-labeled E3 DNA. As expected, the 8.7-kbEcoRlfragmentwasdetectable(Fig. 5a). However, the E3probehybridized totwoDNA
fragments
(5.2 and 4.2kb) inC/ODNAdigested with HindlIl (Fig. 5b). This confirms that there are two5.2-kbHindlIl fragments,oneofwhich contains the proto-amv sequences present in the 8.7-kbEcoRI fragment. The 4.2-kb band mustrepresenthybridizationwith non-proto-amv cel-lular DNA sequences presentwithin the EcoRI 8.7-kbfragment. These data allowus toposition the EcoRI, HindIII, SmaI, BamHI, and Sail sites within the region of normal C/O DNA which contains the proto-amv sequences (see Fig. 7).
Analysis of leukemic DNA. To determine whether there was a similar arrangement of proto-amv sequences in leukemic cells, DNA purified from peripheral blood myeloblasts of leukemic C/E chicken no. 21710 was treated withvariousrestriction endonucleases and ana-lyzed with amv probes. This DNA was the
sourceof the Arecombinant library from which the AMV provirus recombinant clone AllA1-1 wasisolated (18). Knowledge of the location of many endonucleasesites within the AMV
provi-rus (12, 14, 18)permits ustoidentifytheAMV
fragmentswhich shouldhybridize to the HAX4 probe.
Hybridization of HindIll-digested leukemic DNA(Fig. 6a) to the HAX4 probegave rise to three bands (5.2, 4.3, and 1.3 kb). The 4.3-kb fragmentwas not generated fromnormal DNA and corresponded to the 3'-proximal HindIII fragment ofthe AMV genome.
AfterEcoRIdigestionofleukemic DNA, two AMV fragments hybridized with the HAX4 probe (4.4 and 3.6 kb) in addition to the three proto-amv fragments (8.7, 5.4, and 2.0kb)(Fig. 6b). One band (3.6 kb) is the internal AMV EcoRI fragment which hybridizes lightly to
HAX4
because it contains only 200 base pairs homologous tothis probe. The otherband (4.4 kb), whichis notdetectedbyhybridization with EB3 (data not shown), probably represents anAMV provirusjuncture fragment with chicken DNAadjacenttoits 3'end.Thepresenceof this presumptivejuncturefragmentwasunexpected since the DNA wasfromtotalleukemic myelo-blasts. Also, itis different fromthe 3'juncture fragment(2.0 kb) present in the AMV-A clone
hAll-1.
This 4.4-kb fragment was notpreviously detectable with viral probes because it comi-grateswith severalEcoRIfragments
ofthesamesize which areinternal tothehelper myeloblas-tosis-associated virus and endogenous
provi-ruses.
Five amv-specific bands (1.2, 1.9, 2.2, 2.9,
b d e
-23.7
_ 9.5
- ~ -6.6~
- 4.3
2.1 -' 9
_ _ -- 1.35
-.0
0.87
0.60
FIG. 6. Hybridizationof restriction enzyme-digest-ed leukemic DNA with HAX4 probe. DNA from leukemic myeloblasts of chicken no. 21710 (18) was digested with either HindIll (a), EcoRI (b), HindIII and EcoRI (c),BamHIand
EcoRI
(d), orHindIII andBamHI(e).The DNA digests wereelectrophoresed in a0.8% agarose gel, Southernblotted, andhybridized to 32P-labeled HAX4 DNA. A mixture of
HindIll-digested)DNAandHaeIII-digested+X174RF DNA wasused as molecular size markers.
44,
on November 10, 2019 by guest
http://jvi.asm.org/
[image:6.491.111.170.78.248.2] [image:6.491.291.399.409.582.2]and 4.3 kb)weredetectedafterhybridization of
HAX4 to C/E leukemic DNA digested with EcoRJ + HindIlI. Two fragments (1.2 and 2.9 kb)containingamvsequences wereexpectedto
be generated by EcoRI + HindIII digestion of the AMV provirus. Only the 2.9-kb band was
detected (Fig. 6c) in addition to the pattern
obtained with normal DNA, becausethe 1.2-kb bandcomigrated with the 1.2-kb band character-istic forproto-amv sequences (see above). The
autoradiographic intensity of the 1.2-kb band generated from EcoRI +HindIII-digested DNA
wasinagreementwith thepresenceoftwo
amv-containing fragments.
Ifleukemic DNA digested withboth BamHI
and EcoRi was hybridized tothe HAX4 probe (Fig. 6d), three bands (5.4, 2.0, and 1.8 kb)were
detected. The 2.0-kb band isbroad andintense
inthe autoradiogram because it corresponds to twoproto-amvfragments (see above). The
1.8-kb band corresponds to an EcoRI fragment
generated from the AMV provirus, whereasthe
5.4-kbband isacellular fragment carrying proto-amvsequences.
The BamHI + HindIII digestion of leukemic DNA(Fig. 6e)gaverisetofour bands (5.2, 4.3, 2.8, and 1.3 kb) when hybridized with HAX4 DNA.Onlyoneband(2.8 kb) resulted from the
double digestion ofthe integrated AMV
provi-rus.The other bandscorrespondedtothe
proto-amv-containing fragments. As mentioned
earli-er, the5.2-kbbandconsisted oftwoproto-amv
fragments.
Thesedataare summarized in Table2.
DISCUSSION
A previous preliminary analysis had shown
thatatleastoneinterveningsequenceinterrupts
theproto-amv sequencesin chicken DNA (12).
Theuseofdoubleenzymatic digestions together
withprobes of subcloned definedregions of the
viral amv sequence now establishes that there
are atleast four largeintrons in theproto-amv
sequences.
The double digestion of C/O DNA with
HindIII+EcoRIrevealed that the 2.2-kb EcoRI
bandwhichhybridizestotheHAX4probe (12)is in fact a doublet consisting of two adjacent fragments whosemoreaccuratesizesare2.0 and
2.1 kb. Each of them contains a HindIll site
such that HindIIl digestion generates a single
1.3-kbfragment (12). Hybridizationwith
specif-ic subclonesrepresenting differentamvregions
showed that the 2.0-kb EcoRI fragment
con-tainedproto-amv sequences homologousto the amvsequenceslocated from the EcoRI siteupto 150to200 basepairs beyondtheSallsite.These
EcoRI andSallsitesareseparated by1.45kb in proto-amv, whereasthey areseparated by only
350 basepairsinamv.This revealsanewintron
in proto-amv. The remaining portion of the 3'-proximal amv sequences are carried by the 2.1-and 8.7-kb EcoRIfragments.
The proto-amv sequences corresponding to the amv sequences located between the HaeII and EcoRI sites are present in a 2.2-kb HindIII-EcoRI fragment generated by HindIII + EcoRI doubledigestion. This fragment makes up the 3' endof the 5.4-kb EcoRI fragment. The 5'-proxi-mal5.4-kbEcoRIfragment and theinternal 2.0-kb EcoRI fragment which contain proto-amv sequences are notcontiguous but areseparated by a 0.95-kb EcoRI fragment which therefore defines the minimal length ofan additional in-tronwithin the proto-amv sequences.
The locationof the3'-proximal 5.2-kb HindIII fragment which contains proto-amv sequences homologous to 0.2 kb of 3'-proximal amv se-quences wasestablished (i) by hybridization of single HindIII digests of C/O DNA with the A111-derived E3 probe which contains the 3'-proximal proto-amv sequences (12) and (ii) by hybridization of double BamHI + HindIII and HindlIl + EcoRIdigests with the amv subclone SX12.
Hybridizationof singleSmaIanddouble SmaI
+ EcoRIdigestswith amvsubcloneslocates the proto-amv internal SmaI siteat 1.1 kb outside the 3' end of the 5.4-kb EcoRI fragment. This confirms that a 0.95-kb EcoRI fragment sepa-ratesthe 2.0- and5.4-kbEcoRI fragments which contain proto-amv sequences. The location of the SmaI site within the2.0-kbEcoRIfragment (12, 14) also suggests that the5'-proximalEcoRI site ofthis fragment corresponds to the single EcoRIsite present in amv (12, 14). It isunlikely that in thegenerationof theamvinsert,splicing occurredbetweentwoEcoRI sitestoregenerate
anEcoRIsite (5,6,15).Therefore,it ispossible that avery small stretch (less than 30
nucleo-TABLE 2. Restriction enzymefragmentsfrom leuke-mic and normal DNAs whichhybridizedwith HAX4
probe
Fragments(kb) from thefollowing
Enzymes type of DNA:
Leukemic Normal
HindIIl 1.3, 4.3,a 5.2 1.3, 5.2
EcoRI 2.0,3.6,0a 4.4,0 2.0,5.4, 8.7
5.4, 8.7
HindIlI + EcoRI 1.2,b 1.9, 2.2, 1.2,1.9, 2.2, 4.3
2.9,°4.3
BamHI+EcoRI 1.8,a 2.0,5.4 2.0,5.4
HindIlI + BamHI 1.3,2.8,a4.3, 1.3, 4.3, 5.2 5.2
aFragments originating from AMV provirus
inte-gratedinleukemic DNA.
bThis bandconsists of at least one proto-amv and one amvfragment.
on November 10, 2019 by guest
http://jvi.asm.org/
[image:7.491.257.454.497.661.2]proto-amv ORGANIZATION IN CHICKEN DNA 593
I
AMV _
Chicken I / I
K,I
r o o E
n
L4: _;-..
;i
I
I
I I
r I % I % /
w
I
It
ii I / //I/
//
II / i / /I I/ /,
BomHI Eco RSrnI H[ndd B )K EcoRI
HindM EcoRIK EcoRI
SmiaI
kb
FIG. 7. Restriction endonucleasemapofthe chicken cellularDNAregion containingproto-amvsequences.
(Top) The amv insert of the AMV provirus has been subdivided into four regions corresponding to the hybridization probes used in this study. (Bottom) The recognition sites for various restrictionenzymeshavebeen mapped in the chicken DNA region containing theproto-amvsequences. Theexactlocations of theproto-amv
sequenceswithin the various DNAsegmentsdelineated by restrictionenzymesitesarenotknown,exceptfor the amvsequenceslocatedbetween the EcoRI andSalI sites (see the text). The question markatthe 5' end ofamv indicates thatwedonotknow whether this5'segmentis contiguouswith theHaeII site inchicken DNA. The size scales (in kilobases) for the viral(top)and cellular (bottom) sequences aredifferent.
tides long) ofproto-amv sequences, not long
enoughtobe detectedby hybridization, is
locat-ed upstream of the corresponding proto-amv EcoRIsite. These fewnucleotides would permit splicingtogeneratetheamvregion within which the internal EcoRI site is located. The present
resultsaresummarized inFig. 7, which showsa
partial restriction map of the chicken DNA
region containing proto-amv sequences. From our previous study (12) with the AMV
KpnI-XbaI probe, we know that the amv sequences
located between the 5' end of theamvinsert and the HaeII site are present within the 2.2-kb HindIII-EcoRI fragment, but we do not know whether their 3' end is contiguous with the HaeIIsite. Theamountofproto-amvsequences
carriedby the 2.1- and 8.7-kb EcoRI fragments
cannotbeprecisely determined. However, from therelativeintensity of the 1.2- and 4.3-kb bands obtained in the doubleHindIII + EcoRIdigest hybridizedtoHAX4, itappearsthat the number ofproto-amvsequencespresentonthe
3'-proxi-mal 4.3-kb fragment (i.e., within the 8.7-kb EcoRI fragment) is slightly greater than the
numberofproto-amv sequencespresenton the
1.2-kb fragment (i.e., within the 2.1-kb EcoRI
fragment). A similar arrangement ofproto-amv sequencesoccurred inallC/O and C/E chicken tissues examined, indicating their stability in development anddifferentiation.
Under theexperimental approach usedinthis
study, we detected the presence of four large
introns which interrupt the proto-amv se-quences, butwedonotruleoutthe existence of
additional smallerintervening sequences. They would notbe detectable if they did notcontain recognition sites for the restriction
endonucle-ases weused. Therefore, the four introns
detect-ed in thisstudyrepresentthe minimumnumber. Also,aswillbe shown inournextpaper onthis
subject (Perbaletal., submitted for publication),
an additional intron was discovered with the
samerestrictionenzymes asthose used here if
proto-amv DNArecombinant clones were
ana-lyzed instead of total chicken DNA.
The restriction endonuclease analysis
per-formed on DNAisolated fromleukemic myelo-blasts didnotrevealanymajorrearrangement of
the cellularproto-amv sequences as aresult of
leukemogenesis. Characterization of chicken-X
DNA recombinants carrying proto-amv se-quences isolated from leukemic cells confirms
that the proto-amv arrangement in leukemic chicken DNA is similartothat innormal DNA.
This implies that the leukemogenic effect of AMVmaybeentirely dependentupon
transcrip-tion of theintegrated AMV provirus. In support
of this hypothesis,the amvtranscript, which is shorter(2.5 versus4.5 kb) than the RNA which
is transcribed from theproto-amv sequencesin
hemopoietic chicken tissues (7, 10, 20; unpub-lished data), is very abundant inleukemic
my-eloblasts, whereas the proto-amv transcript is
absent.
ACKNOWLEDGMENTS
This researchwas supported by researchgrantCA-10197 from the National Cancer Institute, National Institutes of
0.1 kb
i I I
I I
I II
II II
HindM BamHI EcoRI VOL. 44,1982
7z1
_i
on November 10, 2019 by guest
http://jvi.asm.org/
[image:8.491.102.391.68.230.2]Health. B.P. waspartlysupported by the Centre National de la Recherche Scientifique and was therecipient of a travel fellowship from the North Atlantic Treaty Organization.
LITERATURE CITED
1. Baluda, M. A., and W. N. Drohan. 1972. Distribution of
deoxyribonucleic acid complementarytothe ribonucleic acid of avian myeloblastosisvirus in tissues ofnormaland
tumor-bearing chickens. J. Virol. 10:1002-1009. 2. Bermnn,D. G., L. M. Souza, and M. A.Baluda.1981.
VertebrateDNAs containnucleotidesequencesrelatedto
the transforminggene ofavian myeloblastosis virus. J. Virol. 40:450-455.
3. BIrboim, H. C., and J. Doly. 1979. A rapid alkaline
extractionprocedurefor screening recombinant plasmid
DNA.NucleicAcids Res. 7:1513-1523.
4. Boyer, H. W., and D.RoulandDussoix. 1969.A
comple-mentationanalysis oftherestriction andmodification of
DNA inEscherichia coli.J.Mol. Biol.41:459-472.
5. Breathmach, R., C. Benoist, K. O'Hare, F. Gannon, and P. Chambon. 1978. Ovalbumingene: evidence foraleader
sequence in mRNA and DNA sequences at the exon-intron boundaries. Proc. Natl. Acad.Sci. U.S.A. 75:4853-4857.
6. Catteral, J. F., B. W. O'Malley, M. A. Robertson, R.
Staden,Y.Tanaka,and G.G.Brownlee.1978.Nucleotide
sequence homology at 12 intron-exonjunctions in the
chick ovalbumingene.Nature(London)275:510-513. 7. Chen, J.H.1980. Expressionofendogenous avian
mye-loblastosis virus information in different chicken cells. J. Virol.36:162-170.
8. Curtss,R., M. Inoue, B. Pereira, J. C.Hsu, L. Alexander, and L. Rock.1977.Constructionand use ofsafer bacterial host strainsfor recombinantDNAresearch, p.99-111.In
W. A. Scott and L. Rock (ed.), Molecular cloning of
recombinantDNA.AcademicPress, Inc., NewYork.
9. Franchini,G., J. Evan, C. J.Sherr,and F. Wong-Staal.
1981. onc sequences (v. fes) of Snyder-Theilen feline
sarcomavirusarederived fromnoncontiguous regionsof a catcellular gene(c-fes).Nature(London)290:154-157.
10. Gonda, T. J., D. K. Sheiness, L. Fanshier, J. M.Bishop,C.
Moscovci,and M. G.Moscovid.1981. The genome and theintracellularRNAsof avian myeloblastosisvirus.Cell
23:279-290.
11. Grunstein, M., and D. S. Hogness. 1975. Colony hybrid-ization: a methodfortheisolationof cloned DNAs that contain a specificgene. Proc. Natl. Acad. Sci. U.S.A. 72:3691-3695.
12. Perbal, B., and M. A. Baluda. 1982. Avianmyeloblastosis
virus transforming gene is related to uniquechicken DNA regions separated by at least oneinterveningsequence. J. Virol. 41:250-257.
13. Rlgby,P. W., M. Dieckmann, C. Rhodes, and P. Berg. 1977. Labeling deoxyribonucleic acid to high specific
activity in vitro bynick translation withDNApolymerase
I. J.Mol. Biol. 113:237-251.
14. Rushlow, K. E., J. A.Lautenberger, T. S.Papas,M. A.
Baluda, B. Perbal, J. G. ChirlkJlan, and E. P. Reddy. 1982. Nucleotide sequenceofthetransforming geneof
avianmyeloblastosis virus.Science216:1421-1423. 15. Self, I., G. Khoury, and R. Dhar. 1979. BKV splice
sequences based on analysis of preferred donor and
acceptorsites.NucleicAcidsRes.6:3387-3398.
16. Shalloway, D., A. D. Zelnetz,andG. M. Cooper. 1981. Molecular cloning and characterization ofthe chicken
genehomologous to thetransforminggeneofRous sarco-mavirus. Cell 24:531-541.
17. Southern, E. M. 1975. Detection ofspecific sequences among DNAfragments separatedbygelelectrophoresis.
J.Mol. Biol.98:503-517.
18. Souza, L. M., M. J.Brlskin,R.L. Hillyard, and M. A. Baluda.1980. Identification of theavianmyeloblastosis
virus genome. II. Restriction endonuclease analysis of
DNAfromAproviral recombinantsandleukemic myelo-blast clones. J.Virol. 36:325-336.
19. Souza, L. M., J. N. Stroumer, R. L. Hillyard, M.C. Komaromy, and M. A.Baluda.1980.Cellularsequences arepresent in thepresumptive avianmyeloblastosis virus
genome. Proc. Natl. Acad.Sci. U.S.A. 77:5177-5181.
20. Westin, E. H., R. C.Gaflo, S. K. Arya, A. Eva, L. M.
Souza,M. A.Baluda, S. A. Aaronson, and F.Wong-Staal.
1982.Differential expression of the amv gene in human
hematopoietic cells. Proc. Natl. Acad. Sci. U.S.A. 79:2194-2198.
on November 10, 2019 by guest
http://jvi.asm.org/