PLASMID 9, 201-214 (1983) Gene Rearrangements Leading to the Expression of an Insertion- Inactivated Tetracycline Resistance Gene in pBR322 FFUNCOISE BRUNEL, MICHEL HEUSTERSPREUTE, MIREILLE MERCHEZ, VINH HA THI, MARIE-FRANCOISE PILAETE, AND JOHN DAVISON” Unit of Molecular Biology, Institute of CeNu[ar Pathology, 75 Avenue Hippocrate, 1200 Brussels, Belgium Received October 15, 1982; revised November 5, 1982 Cloning into the Hind111 site of plasmidpBR322 inactivates the tep promoter and usually prevents the expression of the tep gene. The corresponding clones revert to tetracycline resistance at a low frequency. Such reversion is caused by gene rearmngement within the plasmids. DNA sequence analysis reveals three classes of revertants. The first class contains plasmids with partial duplications, which result in the fusion of the promoter of the RNA I species to the tep gene. The event itself destroys the region encoding the RNA primer for replication and thus the plasmids would be replication defective if the duplication did not also include this region of the molecule. The plasmids from the second class are simple deletions which again fuse the teta region to the RNA I promoter. In one case, the junction takes place at the end of the RNA I transcript, leaving RNA I and the RNA primer virtually intact. However, it removes the promoter of the RNA primer, the latter now being read from the cloned material. The second member of this class has fused the tep gene well upstream of the RNA I region so that the RNA primer is still read from its own promoter. The low-level tetracycline resistance is probably due to partial read-through of the RNA I terminator. The third class of revertants differs from the previous two by the acquisition of foreign DNA in the form of an ISZ-type insertion element which is known to promote transcription. The plasmid pBR322 has proved a useful vector for the cloning of foreign DNA ma- terial (Bolivar et al., 1977). One of its major advantages is that unique cloning sites are sit- uated in genes which encode antibiotic resis- tance. Therefore, the initial screening of the recombinant DNA molecules can simply be based upon the phenotype of the recombinant clones. For example, the unique BumHI and SaA sites lie in the tetracycline resistance (tetR)* gene and insertion of DNA fragments results in the sensitivity of the host to the antibiotic (Bolivar et al., 1977). A similar phenomenon often occurs when foreign DNA is integrated at the unique Hind111 site. Sometimes, how- ever, the clones remain resistant to tetracy- cline, indicating that the Hind111 site is not in the structural part of the tetR gene but in its promoter, and inserts carrying a promoter transcribing in the correct direction may then ’ To whom requests for reprints should be addressed. * Abbreviations used: tep, Tetracycline resistance. restore tetracycline resistance (West et al., 1979). In this publication, we report that clones of pBR322 that have become tetracycline sen- sitive as a consequence of insertion of foreign DNA at the Hind111 site can revert to being able to grow on elevated levels of tetracycline. The molecular mechanisms of tetR gene expression and DNA replication in such re- vertant plasmids are discussed. MATERIALS AND METHODS Bacterial strains. Strain Escherichia coli K12 594 (ret+, su”) has been described pre- viously (Weigle, 1966). The C600 garlic de- rivative carrying a missense mutation in the galactokinase gene was obtained from McKenney et al. (198 1). Plasmids. pBR322 has been described pre- viously (Bolivar et al., 1977) and its entire sequence (4362 bp) determined (Sutcliffe, 1979). Its strongest promoters and their lo- 201 0147-619X/83 $3.00 Copyright 0 I983 by Academic Res. Inc. All rights of repro&don in any form resa’wd.
Transcript
PLASMID 9, 201-214 (1983)
Gene Rearrangements Leading to the Expression of an Insertion- Inactivated Tetracycline Resistance Gene in pBR322
FFUNCOISE BRUNEL, MICHEL HEUSTERSPREUTE, MIREILLE MERCHEZ, VINH HA THI, MARIE-FRANCOISE PILAETE, AND JOHN DAVISON”
Unit of Molecular Biology, Institute of CeNu[ar Pathology, 75 Avenue Hippocrate, 1200 Brussels, Belgium
Received October 15, 1982; revised November 5, 1982
Cloning into the Hind111 site of plasmid pBR322 inactivates the tep promoter and usually prevents the expression of the tep gene. The corresponding clones revert to tetracycline resistance at a low frequency. Such reversion is caused by gene rearmngement within the plasmids. DNA sequence analysis reveals three classes of revertants. The first class contains plasmids with partial duplications, which result in the fusion of the promoter of the RNA I species to the tep gene. The event itself destroys the region encoding the RNA primer for replication and thus the plasmids would be replication defective if the duplication did not also include this region of the molecule. The plasmids from the second class are simple deletions which again fuse the teta region to the RNA I promoter. In one case, the junction takes place at the end of the RNA I transcript, leaving RNA I and the RNA primer virtually intact. However, it removes the promoter of the RNA primer, the latter now being read from the cloned material. The second member of this class has fused the tep gene well upstream of the RNA I region so that the RNA primer is still read from its own promoter. The low-level tetracycline resistance is probably due to partial read-through of the RNA I terminator. The third class of revertants differs from the previous two by the acquisition of foreign DNA in the form of an ISZ-type insertion element which is known to promote transcription.
The plasmid pBR322 has proved a useful vector for the cloning of foreign DNA ma- terial (Bolivar et al., 1977). One of its major advantages is that unique cloning sites are sit- uated in genes which encode antibiotic resis- tance. Therefore, the initial screening of the recombinant DNA molecules can simply be based upon the phenotype of the recombinant clones. For example, the unique BumHI and SaA sites lie in the tetracycline resistance (tetR)* gene and insertion of DNA fragments results in the sensitivity of the host to the antibiotic (Bolivar et al., 1977). A similar phenomenon often occurs when foreign DNA is integrated at the unique Hind111 site. Sometimes, how- ever, the clones remain resistant to tetracy- cline, indicating that the Hind111 site is not in the structural part of the tetR gene but in its promoter, and inserts carrying a promoter transcribing in the correct direction may then
’ To whom requests for reprints should be addressed. * Abbreviations used: tep, Tetracycline resistance.
restore tetracycline resistance (West et al., 1979).
In this publication, we report that clones of pBR322 that have become tetracycline sen- sitive as a consequence of insertion of foreign DNA at the Hind111 site can revert to being able to grow on elevated levels of tetracycline. The molecular mechanisms of tetR gene expression and DNA replication in such re- vertant plasmids are discussed.
MATERIALS AND METHODS
Bacterial strains. Strain Escherichia coli K12 594 (ret+, su”) has been described pre- viously (Weigle, 1966). The C600 garlic de- rivative carrying a missense mutation in the galactokinase gene was obtained from McKenney et al. (198 1).
Plasmids. pBR322 has been described pre- viously (Bolivar et al., 1977) and its entire sequence (4362 bp) determined (Sutcliffe, 1979). Its strongest promoters and their lo-
201 0147-619X/83 $3.00 Copyright 0 I983 by Academic Res. Inc. All rights of repro&don in any form resa’wd.
202 BRUNEL ET AL.
cations on the circular molecule have been identified (Backman et al., 1978; Morita and Oka, 1979; West and Rodriguez, 1980; Stiiber and Bujard, 198 1). The elements involved in its DNA replication have also been inten- sively studied (Itoh and Tomizawa, 1980; Tomizawa et al., 198 1).
Plasmid pBR322: :T5 Hind111 L carrying the 3.9-kb Hind111 fragment of bacteriophage T5 DNA has been described (Brunel et al., 1979, Brunei et al., 1981). Recombinants L8 and L51 contain this fragment in opposite orientations. Both plasmids contain a func- tional T5 D21 gene (Brunel et al., 1979).
pKO-1 was a gift from Dr. K. McKenney. It is 3900 bp long and contains the complete galactokinase gene, though this is not ex- pressed due to absence of a suitable promoter. The insertion of an active promoter at the Hind111 or nearby SmaI site can be detected by the change in color of the C600 galK- recipient when plated on McConkey galactose medium (McKenney et al., 1981).
Enzymes and chemicals. Restriction en- zymes were purchased from New England Biolabs. Bacterial alkaline phosphatase was obtained from Worthington and polynucle- otide kinase (from TCinfected E. coli B) from P-L Biochemicals. The reagents for DNA se- quencing were obtained from Eastman Ko- dak (hydrazine), Aldrich Chemical Company (dimethylsulfate), and BDH (pyridine and pi- peridine). [32P]ATP (sp act 3000 Ci/mmol) and [14C]galactose (sp act 40-60 mCi/mmol) were purchased from Amersham.
Analysis ofrevertant DNAs. The methods used for preparing plasmid DNAs and for their restriction analyses have been described pre- viously (Brunel et al., 1979, 1981). DNA se- quencing was by the method of Maxam and Gilbert (1977).
RESULTS
Isolation of Tetracycline Resistant Clones
E. coli 594 carrying the plasmids L8 and L5 1 was used to isolate tep revertants. These plasmids carry the 3.9-kb T5 Hind111 L fiag- ment in opposite orientations (Brunel et al.,
198 1) and are shown diagramatically in Fig. 2. L8 is sensitive to the presence of 3 &ml of tetracycline in the growth of medium, whereas L5 1 grows poorly on 12 pg/ml. Rare revertants were selected from both strains on 6 and 20 j&ml, respectively. Eight were re- tained for further study: L5 1- 11 and L5 l- 19 were resistant to 50 pg/ml of the antibiotic; L8-5, L8-16, LX-136, L8-106, and L8-66 grew on 36 pg/ml; while L8-1, on the contrary, was sensitive to a concentration superior to 6 p&ml.
Reversion Involving Acquisition of 1.3 kb of Foreign DNA
The restriction enzyme analysis of tetR re- vertants L5 1-19, L5 l-l 1, and L8-5 is pre- sented in Fig. 1. In all three revertant plas- mids, the size of one of the two Bg/II frag- ments is larger than that of the parental fragment by 1.3 kb (Fig. 1 A, slots b and c, Fig. 1 B, slot b). The 1.3-kb insertions all con- tain a Hind111 site (Fig. 1 A, slots e and f, Fig. lB, slot d) and can be found either in the small Bg/II-Bg/II fragment (L5 1-19, Fig. 1 A, slot i) or in the BglII-Hind111 part of T5 Hind111 L which is near the tetR gene (L51- 11, Fig. 1 A, slot h; L8-5, Fig. lB, slot f). These and other data are summarized in Fig. 2.
The size of the insertions as well as their ability to promote transcription are properties reminiscent of those of the IS2 insertion ele- ment previously discovered by Hirsch et al. (1972). We therefore set out to explore the potential similarities. First, the HaeIII, HpaI, Hinff, and TaqI restriction patterns of the re- vertant plasmid DNAs were compared to those of the parents L5 1 and L8. They were found to contain restriction fragments of the size predicted from the known DNA sequence of the classical IS2 (data not shown). The 1300- bp insertion sequences also contain Hind111 and SmaI sites, as does IS2. Their HhaI di- gests, on the other hand, lacked the 2 12-bp fragment normally obtained by HhaI restric- tion of IS2 (position 627 to 839). Subsequent DNA sequencing of the 72-bp HindIII-HaeIII region which should have contained the HhaI
GENE REARRANGEMENTS IN pBR322 203
L51
Ps
Hc
Hd
tic
+lL5
04+1
220+
,125
0,
+ pP
_(
L51-
19
151-
11
. Ps
“c
Hd
H
c Bg
-
Hd
Hc
Bg
PS
Hd
HC
Ps
Ps
Hc
Hd
L-e
---I
petq
pP
+
m
2350
+%
? 15
00
%?
2700
-
717
-3O
SOV2
25O
d A3
LOO
-1
1800
--*
L8
L8-5
PS
“c
Hd
PS
89
Bg
Hc
Hd
“c
PS
-m-J
~125
0 ~1
22O
+clL
50--*
+ ,te
tq
PS
nc
“d
PS
89
8.g
MC
“d
HC
Md
HC
PS
I%.
2. P
hysi
cal m
ap o
f td
reve
rtant
s ca
rryin
g an
IS-2
type
ins
ertio
n. A
ll of
the
plas
mid
s ar
e re
pres
ente
d as
thou
gh li
near
ized
thr
ough
the
uni
que
Pst
I re
stric
tion
site
of
pBR
322.
The
par
enta
l L8
and
L5 1
pla
smid
s di
ffer
by t
he o
rient
atio
n of
the
TS H
ind1
11 L
fragm
ent w
ith r
espe
ct to
pBR
322.
Sol
id li
nes
(-)
repr
esen
t pBR
322
DN
A an
d br
oken
line
s (-
--) r
epre
sent
the
part
of th
e m
ap w
hich
is
not
draw
n to
sca
le. T
he o
pen
box
(0)
indi
cate
s th
e T5
Hin
d111
L fra
gmen
t an
d th
e cl
osed
box
(I)
indi
cate
s th
e 1.
3-kb
inse
rtion
seq
uenc
e. Th
e si
zes o
f the
fra
gmen
ts a
re g
iven
in b
ase
pairs
. Res
trict
ion
site
abb
revi
atio
ns:
Ps,
PM
; H
c, H
incI
I; H
d, H
indI
II;
Bg, B
g/II.
GENE REARRANGEMENTS IN pBR322 205
site in position 839 showed that this was iden- tical to the published DNA sequence of IS2 (Ghosal et al., 1979) except that the terminal C of the GCGC HhaI recognition site (posi- tion 839) was deleted. Such microvariations in IS sequences have previously been reported (Musso and Rosenberg, 1977; Brosius and walz, 1982).
Second, the orientations of the IM-like se- quences were also determined with respect to the tep region. This was achieved by sizing the fragments obtained by the double HincII- HindIII digestion of the L5 l-l 1, L5 1-19, and LS-5 plasmid DNAs (Fig. lC, slots c, d, f). The results obtained are only compatible with the orientation drawn in Fig. 2. This orien- tation is also that in which IS2 turns on the expression of the genes located furthest away from Hind111 (Saedler et al., 1974). Therefore, apart from minor variations in the sequences, it is likely that the insertion elements which restore the tep gene function are similar to IS2.
Reversion Involving Duplication of Plasmid DNA
The three tetracycline-resistant revertants LS- 136, L8-66, and L8- 106 also contain plas- mid DNA of higher molecular weight than the parental L8 plasmid. One of them, LS- 136 was studied in greater detail. A restriction analysis of the same type as that performed on the L5 1 derivatives demonstrated that the reversion comes from the acquisition of DNA material in the largest L8 PstI restriction frag- ment and therefore is most probably due to a DNA insertion. Restriction with HindIII, however, showed that the plasmid retains the original pBR322 and HindIII L fragments, but carries in. addition a third HindIII fragment of 3 110 bp. Since it is difficult to imagine how any insertion could leave intact both original Hi&III fragments, it is necessary to discard the insertion hypothesis and test the possi- bility of gene duplication. In fact, the L8- 136 HincII, SalI, and BumHI restriction digests all produce an additional fragment of 3 110 bp (data not shown). This result can only eas-
ily be obtained if a segment of pBR322.T5 Hind111 L is duplicated and maintained in the same orientation, as hypothesized diagramat- ically in Fig. 3. Since the Hi&II site farthest to the right, representing the junction between pBR322 and T5 Hind111 L, is repeated, it is necessary to suppose that a portion of both the pBR322 and the T5 insert are duplicated. In support of this suggestion is the observa- tion that the nearby Awl site of the T5 Hind111 L fragment is also duplicated. However, the T5 duplication does not extend as far as the HincII restriction site located 700 bp inside the insert (data not shown). It is not imme- diately obvious how this proposed duplica- tion of DNA could result in expression of tet- racycline resistance. The T5 DNA does not change in relative position to the inactive tetR genes and the fact of now having two inactive tep genes seems unlikely to confer the high level of resistance seen in the L8- 136 revertant as compared to the parent. Thus, the dupli- cation has probably resulted in the placement of an appropriately oriented promoter up- stream of one of the duplicated tetR genes. To define which promoter was involved, it is nec- essary to ascertain the exact fusion point of the duplication. The hypothesis depicted in Fig. 3 predicts that after cleavage with SalI, the large SalI fragment would contain all of the genetic material present in L8 whereas the smaller SulI fragment would contain the du- plicated region. The position of the fusion point can therefore be determined by electro- phoretic separation of the large and small SalI fragments followed by digestion with a second restriction enzyme (Hue111 or HhaI), the re- striction coordinates of which are exactly known (Sutcliffe, 1978). The results of such an analysis are shown in Fig. 4. The large SulI fragment of L8- 136 (slot c) contains all of the Hue111 fragments present in pBR322 (slot b), except fragments 7 and 9 which contain, re- spectively, the WI site used for cleavage and the Hind111 site used for cloning. In contrast, the smallest S&I fragment of LS- 136 (slot d) lacks, in addition, fragments 1,4, and 6. These form a block of fragments located between map coordinates 2952 and 4344 (Fig. 3, Sut-
0 1
2 3
4 Kb
I
1
b)
EC
Hd
Bg
Bg
He
Av
Hd
HC
AV
HC
EC
9 I
78
3 25
46
Ha
/ /II
II I
1 ,,/
I r
I
11
/I/
11
7 6
8 2
j’ ‘9
” 1
“3’4
‘7
Hh
, 15
I L
te
tr-
ori
i am
p5
- R
NA
prim
er
RN
A IN
EC
Hd
Bg
Sa I
Bg
Hc
Av
Hd
HC
AV
Sa I
Av
Hd
+k
AV
Hc
EC
T5
liLte
t’-
ori
iT5,
itetr
1 or
i La
mp5
- R
NA
prim
er
fusi
on
RN
A __
____
t R
NA
It
FIG
. 3.
Sch
emat
ic
repr
esen
tatio
n of
L8
and
L8-1
36.
The
plas
mid
s ar
e dr
awn
as th
ough
lin
eariz
ed
by c
uttin
g at
the
pBR
322
EcoR
l si
te (
EC).
The
cloc
kwis
e di
rect
ion
on
pBR
322
DN
A (-)
is
fro
m
left
to r
ight
. Th
e nu
mbe
ring
of t
he H
haI
(Hh)
an
d H
ue11
1 (H
a)
rest
rictio
n si
tes
is t
aken
fro
m
Sutc
liffe
(1
978)
, bu
t on
ly
the
larg
est
fragm
ents
ar
e sh
own.
R
estri
ctio
n si
tes
(Hd,
H
indU
I; B
g, B
gflI;
H
c,
Hi&
I; Av
Ava
l) on
the
T5
DN
A (I)
w
ere
dete
rmin
ed
by D
avis
on
(unp
ublis
hed
data
). Th
e or
igin
of
rep
licat
ion
(ori)
is
indi
cate
d as
wel
l as
the
resp
ectiv
e lo
catio
ns
of t
he R
NA
prim
er
(-)
RN
A I
( ti)
, an
d th
e fu
sion
R
NA
in L
8-13
6 (i)
. Th
e di
agra
m
is d
raw
n to
sca
le.
GENE REARRANGEMENTS IN pBR322 207
31 5
8.- 1
l-
6- 7- -. O--9-
abc de fghi j
1
- 5 -
7 -
.*3,4
6
!9
FIG. 4. Restriction analysis of the &/I fragments of L8- 136. LS- 136 was cleaved with &/I and the large and small fragments separated by electrophoresis in 0.7% agarose. Restriction digests of the purified fiag- ments were then compared to similar digests of the parent L8 and pBR322. Slots a-e contain HaeIII digests of DNA of a, L8; b, pBR322; c, L8-136 large Nazi fragment; d, L8-136 small Sun fragment; and e, L8- 136 intact DNA. Slots f-j contain HhaI digests of DNA off, L8; g, pBR322; h, LS- 136 large SafI fragment; i, LS-136 small &A fragment; and j, L8-136 intact DNA. The numbers on the right and left indicate the largest pBR322 fragments produced by cleavage with HaeIII and HhaI, respectively.
cliffe, 1978). Thus, it is clear that the fusion coordinates 2995, again within the HueIII-4 point lies beyond coordinate 2952 within fragment (it is not possible from Fig. 4 to de- fragment Hue111 4. Studies of the HhaI digests termine the presence or absence of the HhuI- confirm this result. The largest MI fragment 13 fragment which lies just to the right of (Fig. 4, slot h) contains all of the HhaI frag- Hhu-9). When revertant L8-136 is cut by ments of plasmid pBR322 (slot g), except frag- Hinfl, a fragment of about 260 bp is generated ment 7 which contains the Hind111 site used which comes from the small Sun fragment for cloning (the fragment cleaved by SulI is and is not present in the parent plasmid (data not seen to be missing since it comigrates with not shown). That it should be obtained from other fragments of the same molecular weight the revertant only indicates that it contains (Sutcliffe, 1978). However, the smaller SulI the fusion of the duplication. It can therefore fragment (slot i) lacks, in addition, the HhuI be predicted that this fragment starts in 1, 3, and 4 fragments. These form a block of pBR322 at the Hinff site 2845, includes the fragments between coordinates 2995 and Hue111 2952 site, and finishes at a Hinff site 4259. Thus, the transition point lies beyond in the T5 DNA. This was verified by DNA
3085 . . . . , I CTCTTGATCCGGCAAACAAACCACCGCTGGTAGCGGTGGTTTTTTTGTTT
. TATAGCACAGCAAC.. .
FIG. 5. The DNA sequences of the new pBR322-T5 junction in L8-I 36 and L8- 16. The T5 DNA material is represented in italics whereas pBR322 sequences are in roman typescript. The prominent features of RNA I promoter and terminator have been indicated (0) as well as the start point of the message f.y). In L8-136, the junction takes place afte.r base 3052 and in LS-16, after base 3085. DNA sequencing data show that 85 bp of T5 DNA remain in L8-136 and 330 bp in L8-16 (data available on request). The 5’ + 3’ strand is shown.
sequencing following the procedure of Maxam and Gilbert (1977). Figure 5 shows that the transition from the pBR322 to the T5 DNA material actually occurs after position 3052. This pBR322 region is known to be tran- scribed in a clockwise direction from a pro- moter located upstream of position 2977 (Backman, 1978; Morita and Oka, 1979) to give the RNA I transcript of about 110 bp. Therefore the L8-136 resistance to tetracy- cline is most probably due to the transciption of the tep gene from this promoter; the re- sulting transcript containing 76 bases of RNA I, 87 bases of T5 DNA, and the tep region.
Restriction analysis of revertants L8-66 and L8-106 shows that, like L8-136, these are du- plications resulting in the fusion of the RNA I and the tep regions, though the exact fusion point within the transcript differs in all three revertants (data not shown).
It is known that the RNA I transcript and the RNA primer for initiation of plasmid DNA replication are transcribed in opposite direc- tions from the same segment of DNA (Itoh
and Tomizawa, 1980; Tomizawa et al., 198 1). Thus, a deletion of part of the RNA I tran- script would result in a replication-defective molecule, which probably accounts for the observation that fusions of this type are all duplications with one intact primer region. This was further investigated by cleavage of the L8-136 revertant with AvaI (Fig. 3) fol- lowed by religation and transformation. Of 308 clones selected for amp, only 18 were also tetR. In contrast, 100% of clones selected for tetn were also amp. This result reinforces the view that in L8- 136 the functional primer is closely linked to the amp region, while the tetR part of the molecule is replication defec- tive. Finally, we note that the LS- 136 plasmid has two origins of replication but only one RNA primer. It is not known whether both of the origins are active or only that adjacent to the primer transcript.
Reversion Due to Deletion of DNA Material
The tep revertants L8-16 and L8-1 have lost two of the genetic properties of the L8
GENE REARRANGEMENTS IN pBR322 209
parent since they have become sensitive to ampicillin and incapable of complementing a T5 D21 amber mutant for growth (Brunel et al., 1979). The most likely explanation for this phenomenon is that a deletion covering both functions has occured. This was con- firmed by restriction enzyme analysis; L8-16 and L8-1 are not cleaved by EcoRI and L8- 1 is not sensitive to HindIII, whereas L8- 16 is cut only once by this enzyme (data not shown). When HpuII is used to define more precisely which region of the pBR322 plasmid is not present in L8- 16, it can be seen that fragments 3 and 5 are missing. The &a11 bands coming from Hind111 L are also mostly lost which indicates that T5 DNA material has also been deleted. However, a new 5 15- bp fragment appears in the revertant, which is not present in either pBR322 or in the L8 parent. This fragment contains the new junc- tion between pBR322 and the T5 DNA. The DNA sequence of this fragment shows that the deletion has removed all DNA between pBR322 coordinate 3085 and the last 330 bp of the T5 Hind111 L fragment (Figs. 5,6). The deletion, therefore, takes away both the &lac- tamase and the T5 D21 genes. As with re- vertant L8-136, it is probable that the tetR region of L8-16 is expressed from the RNA I promoter. The RNA I transcript terminates variably at pBR322 coordinates 3084, 3085, and 3086 (Morita and Oka, 1979) and the fusion of L8-16 removes the last base of the longest transcript.
Nonetheless, the structure of L8- 16 leaves open the question of which regulatory ele- ment controls the transcription of the RNA primer since its promoter Pp is deleted from the plasmid molecule. Two nonexclusive hy- potheses can be formulated as to the identity of the new controlling element. First, tran- scription could be initiated from the PI pro- moter described by Sttiber and Bujard ( 198 1) as overlapping the tet promoter but un- touched by insertions at the Hind111 site and reading in the anticlockwise direction. Sec- ond, the remaining 330 bp of T5 DNA could also contain a promoter reading in the anti- clockwise direction (Fig. 6). To test these hy-
potheses, the region was examined for the presence of promoters. The 388-bp Hue111 fragment containing 243 bp of the T5 DNA and 145 bp of the neighboring pBR32? HindIII-BumHI region (Fig. 6) was sub- cloned in the SmuI site of the promoter-probe vector pKO-1 (McKenney et al., 198 1, Fig. 6). In this vector, the galactokinase gene can only be expressed efficiently if the inserted fragment contains a promoter sequence. A galK- derivative of E. coli C600 was trans- formed by the ligation mixture of pKO- 15’muI and LS-16 Hue111 fragments, and 32 of the red colonies which grew on MacConkey ga- lactose ampicillin plates (i.e., expressed the galactokinase gene) were retained for further study. Their plasmid DNAs were digested with Hind111 and four of them contained a Hind111 fragment of 162 bp and thus had inserted the 388-bp Hue111 fragment in the orientation compatible with expression of the galactoki- nase gene from Pl (Fig. 6, LKl). The 16Zbp fragment was then eliminated from LKl by Hind111 restriction followed by ligation (thereby removing the Pl promoter). The cor- responding plasmid (LK2) was still able to confer the red phenotype to the g&C strain. Therefore, the 243-bp fragment of T5 must contain a promoter which is likely to take part (possibly as well as Pl) in the transcription of L8- 16 RNA primer.
A puzzling result obtained in this same ex- periment is that 3 out of 32 red colonies were found to contain plasmids which carry the Hue111 fragment in the opposite orientation (LK3). These colonies could not have been expressing the galactokinase gene from any of the two promoters described above, since, in this orientation, these initiate transcription in the opposite direction. In addition, guK bacteria which were transformed by the LK3 Hind111 deletion derivative (LK4) lacking the T5 DNA material remained white on McConkey galactose medium. The transcrip- tion of the galactokinase gene, therefore, takes place from T5 in the clockwise direction. This is seemingly in contradiction with the fact that the parental plasmid L8 was incapable of growth even on low concentrations of tetra-
BRUNEL ET AL.
ild
lid
FIG. 6. Schematic representation of U-16 and its derivatives. Plasmid U-16 is derived from L8 by deletion of the DNA indicated by the dotted lines, resulting in the fusion of promoter P4 to the te?’ region. The presence of promoters on the remaining TS DNA was investigated by subcloning the 388&p HaeIII Fragment of L8- 16 onto plasmid pKO-I, in which the presence of a promoter is indicated by the expression of the galactokinase gene. Two kinds of recombinants were obtained, which contained the 388&p H&II fragment in one orientation (LKI) or the other (LK3) differentiated by the relative distances between the Hind111 sites on the plasmid and in the insert. The respective contribution of pBR322 and T5 material to the galactokinase expression was then asserted by removing the HindI&HindIII lkagments from each plasmid (LKl, - LK2, LK3 - LK4) and determining the galK phenotype of the corresponding trans- formants.
GENE REARRANGEMENTS IN pBR322 211
cycline and could be a consequence of the differential sensitivities of the two methods used to assess promoter activity.
As indicated above, the final revertant plas- mid, L&l, is a deletion removing the &lac- tamase gene and all T5 DNA material. When L8-1 DNA is cut with BstNI, the pBR322 fragment 1 (position 2634- 129) is replaced by a fragment of 733 bp which contains the ends of the deletion (data not shown). The DNA sequencing of the deleted molecule was there- fore achieved from the BstNI site in position 129 (Fig. 7). It shows that the new junction is located between two regions situated be- tween bases 75 and 8 1, on one hand, and 33 13 and 33 19 on the other. More resolution could not be achieved since the seven base pairs contained in these two regions are identical. It should be noted that the deletion of LS-1 does not remove any of the genetic material postulated to be concerned with tep since the first open reading frame in this region begins at coordinate 86 (Sutcliffe, 1979). These re- vertants exhibit rather low tetracycline resis- tance. It is not clear from which promoter L8- 1 is expressing tep. It is possible that this is due to low level read-through of the RNA I transcript through the stop signal located at 3085. Alternatively, there may be another clockwise promoter located between 3085 and 3320.
DISCUSSION
Three kinds of DNA rearrangements were found to restore the expression of tetracycline resistance from a pBR322 plasmid in which the tep promoter has been destroyed by the insertion of a foreign DNA fragment (i.e., T5 Hind111 L). All three, insertions, deletions, or duplications, result in the positioning of a new promoter in front of the tetR gene.
The three tep revertants containing foreign DNA (e.g., L5 1-19) have acquired an inser- tion element nearly identical to the IS2 (Ghosal et al., 1979). Minor variations were found between their DNA sequences and that of the originally sequenced IS2. Others have observed similar polymorphisms (Musso and Rosenberg, 1977; Brosius and Walz, 1982).
IS2 has been shown to promote transcription in one orientation. In all three cases, the IS2 was in the correct orientation compared to the tep gene, so that transcription of the tep gene took place from the IS2 promoter. It is nonetheless noteworthy that the insertions oc- cured at various locations in the T5 Hind111 L fragments, one of them being further away than 1500 bp from the beginning of the tetR gene. This indicates the absence of any tran- scription terminator sequences in this portion of T5 DNA in that orientation.
The second group of tep revertants (e.g., L8-136) are partial duplications in which the tep gene is fused to the RNA I transcription unit thought to control plasmid copy number and plasmid incompatibility within the bac- terial cell (Tomizawa and Itoh, 1981). Tet- racycline resistance here is easily explained by transcription from the RNA I promoter (P4) though the tetR gene. An interesting conse- quence of these RNA I-tetR fusions is that they remove the promoter and part of the coding region of the RNA primer concerned with initiation of DNA synthesis at the origin of pBR322. Thus, the RNA I-teta fusion part of the molecule is replication defective (al- though it has a functional origin of replica- tion). The partial duplication is therefore nec- essary because one part of the molecule is amp” and replication proficient, while the other part is tep but replication deficient.
The third group of tep revertants com- prises large deletions of plasmid DNA and has two members, L8- 16 and L8-1, which differ in their mechanisms of expression of tetra- cycline resistance and in their mechanisms of DNA replication primer synthesis. Like the partial duplications, L8-16 expresses the tep gene from the RNA I promoter. It is known that the RNA I transcript terminates variably at coordinates 3084, 3085, and 3086 (Morita and Oka, 1979) to give RNA I transcripts of 108, 109, or 110 bp. Thus the L8-16 deletion, which has its fusion point at 3085 removes only 1 bp from the terminus of the longest RNA I transcript. At the same time however, it removes 2 bp from the beginning of the primer RNA transcript (which begins at 3087
212 BRUNEL ET AL.
w h v . u
I . ” ”
4 A
h h w W 4 ,"-. 4
m P
. CJ m-u W h u 4
4 h h h v V W W h h h .4
. a: 4 4 u 4 v h a: h h W h 4 W v 4 4 u
," .h
GENE REARRANGEMENTS IN pBR322 213
and proceeds in the opposite direction) and also deletes the promoter for this transcript located around 3097 (Tomizawa and Itoh, 1981). The fact that the plasmid replicates shows that, despite deletion of its promoter, the RNA primer is nonetheless synthesized. The subcloning into a promoter-probe vector of the T5 243-bp HaeIII-Hind111 fragment, which is located immediately upstream of the RNA primer region, demonstrates that this fragment contains at least one T5 promoter, active in the anticlockwise direction, which may be responsible for synthesis of the primer RNA. In addition, the P 1 promoter of pBR322 (Stiiber and Bujard, 1981) is also correctly oriented to fulfill this function. Presumably the substitution of two bases of the 5’ end of the RNA primer and its fusion to another transcript does not affect its priming activity. In fact, yields of L8-16 DNA are very high indicating that the new promoter may en- hance transcription of the primer and thereby replication of the DNA (unpublished results). There is great variability in promoter struc- ture (Rosenberg and Court, 1979) so that it may be difficult to recognise a promoter in a sequence of DNA. Inspection of the 330-bp T5 DNA sequence remaining in L8-16 shows numerous potential TATAAT boxes al- though none has a convincing -35 sequence. Several other weak promoters are known to have poor -35 homology (Rosenberg and court, 1979).
The revertant plasmid L8-1 is a deletion between pBR322 regions [3313-33201 and [75-811, removing all phage T5 material in the process. An exact determination of the deletion endpoint cannot be made by DNA sequencing since the relevant sequences in these two regions are identical over seven base pairs and have 24 of 36 base pairs in common over the adjacent region (Fig. 7). Despite this close homology, the two regions have quite different ancestries (reviewed by Sutcliffe, 1979). The region around 33 12 is derived from the Tn3 transposon and contains the terminal end of the /3-lactamase gene (but not the ter- minal repeat of Tn3). In contrast, the 75-81 region is derived from pSC 10 1, being located
very near the beginning of the tep region. The evolutionary origin of the observed homology is thus obscure. It should be noted that short homologous sequences (such as those in- volved in the L8-1 deletion) have been im- plicated in deletion formation in both recA+ and recA- E. coli (Albertini et al., 1982). However this may not be the exclusive mech- anism since the other deletions and partial duplications (which may be regarded as a deletion in a dimer) show no apparent ho- mology at the point of fusion.
Revertant L8-1 differs from the other dele- tion and duplication revertants in that it does not involve an obvious fusion of the tep gene to the RNA I transcript since the deletion oc- curs about 228 bp beyond the terminator sig- nal for the RNA I transcript (3085). The plas- mid confers only low-level resistance and probably expresses tep by read-through of the RNA I transcript through the terminator in the 3080 region. Alternatively, it is conceiv- able that it does so from an unidentified pro- moter located between 3085 and 3320. A sim- ilar situation has previously been encountered by Chang et al., 1977, who detected mouse dehydrofolate reductase activity expressed from a cDNA inserted into the PstI site in the opposite orientation to that of the /I-lacta- mase gene.
ACKNOWLEDGMENTS
F.B. is particularly grateful to Professor W. Fiem for his hospitality, to Mr. F. Molemans who taught her the Maxam and Gilbert technique with much patience and thoroughness, and to Dr. K. McKenney who provided the promoter-probe plasmid vector prior to publication.
REFERENCES
ALBERTINI, A. M., HOFER, M., CALOS, M. P., AND MILLER, J. H. (1982). Gn the formation of spontaneous deletions: The importance of short sequence homolo- gies in the generation of large deletions. Cell 29,3 1% 328.
BACKMAN, K., BETLACH, M., BOYER, H. W., AND YAN- OFSKY, S. (1979). Genetical and physical studies on the replication of ColEI-type plasmids. Cold Spring Harbor Symp. Quant. Biol. 43, 69-76.
BOLIVAR, F., RODRIGUEZ, R. L., GREENE, P. J., BET- LACH, M. C., HEYNEKER, H. L., BOYER, H. W., CROSA, J. H., AND FALKOW, S. (1977). Construction and char-
214 BRUNEL ET AL.
acterization of new cloning vehicles II. A multipurpose cloning system. Gene 2, 95-l 13.
BROSIUS, J., AND WALZ, A. (1982). DNA sequences flanking an E. coli element IS2 in a cloned yeast TRPS gene. Gene 17,223-228.
BRUNEL, F., DAVISON, J., AND MERCHEZ, M. (1979). Cloning of bacteriophage T5 DNA fragments in plas- mid pBR322 and bacteriophage XgtWES. Gene 8, 53-68.
BRUNEL, F., DAVIWN J., HA THI, V., AND REEVE, J. (1981). Cloning of bacteriophage T5 DNA fragments III. Expression in E. coli mini-cells. Gene 16, 107-l 18.
CHANG, A. C. Y., NUNBERG, J. H., KAUFMANN, R. J., EHRLICH, H. A., SCHIMKE, R. T. AND COHEN, S. N. (1978). Phenotypic expression in E. coli of a DNA se- quence coding for mouse dihydrofolate reductase. Na- ture (London) 275,6 17-624.
GHOSAL, D., SOMMER, H.,. AND SAEDLER, H. (1979). Nucleotide sequence of the transposable DNA element IS2. Nucl. Acids Res. 6, 111 l-l 122.
HIRSCH, H. J., STARLINGER, P., AND BRACHET, P. (1972). Two kinds of insertion in bacterial genes. Mol. Gen. Genet. 119, 191-198.
ITOH, T., AND TOMIZAWA, J. (1980). Formation of an RNA primer for initiation of replication of ColEI DNA by ribonuclease H. Proc. Nat. Acad. Sci. USA 77,2450- 2454.
MAXAM, A. M., AND GILBERT, W. (1977). A new method for sequencing DNA. Proc. Nat. Acad. Sci. USA 74, 560-564.
MCKENNEY, K., SHIMATAKE, H., COURT, D., SCHMEIS- SNER, U., BRADY, C., AND ROSENBERG, M. (1981). A system to study promoter and terminator signals rec- ognized by E. coli RNA polymerase. In “Gene Am- plification and Analysis” (J. G. Chirikjian and T. Pa- pas, Eds.), Vol. II, pp. 383-415. Elsevier/North-Hol- land, New York.
MORITA, M., AND OKA, A. (1979). The structure of a transcriptional unit on colicin EI plasmid. Eur. J. B&hem. 97,435-443.
Mosso, R. E., AND ROSENBERG, M. (1977). Nucleotide sequences of two sites for IS2 DNA insertion. In “DNA Insertion Element, Plasmids and Episomes” (A. I. Buk- hari, J. A. Shapiro, and S. I. Adhya, Eds.), pp. 597- 60 1. Cold Spring Harbor Laboratory, New York.
ROSENBERG, M., AND COURT, D. (1979). Regulatory se- quences involved in the promotion and termination of RNA transcription. Annu. Rev. Genet. 13, 319-353.
SAEDLER, H., J~RG, H. J., Hu, S., AND DAVIDSON, N. (1974). IS2; A genetic element for turn-off and tum- on of gene activity in E. coli. Mol. Gen. Genet. 132, 265-271.
STOBER, D., AND BUJARD, H. (1981). Organization of transcriptional signals in plasmids pBR322 and pACYCl84. Proc. Nat. Acad. Sci. USA 78, 167-17 1.
SUTCLIFFE, J. G. (1978). pBR322 restriction map derived from the DNA sequence: Accurate DNA size markers up to 4361 nucleotide pairs long. Nucl. Acids Res. 5, 272 l-2728.
SUTCLIFFE, J. G. (1979). Complete nucleotide sequence of the E. coli plasmid pBR322. Cold Spring Harbor Symp. Quant. Biol. 43, 77-90.
TOMIZAWA, J., ITOH, T., SELZER, G., AND SOM, T. (198 1). Inhibition of ColEI RNA primer foundation by a plas- mid specified small RNA. Proc. Nat. Acad. Sri. USA 78, 1421-1425.
TOMIZAWA, J., AND ITOH, T. (1981). Plasmid ColEI in- compatibility determined by interaction of RNA I with primer transcript. Proc. Nat. Acad. Sci. USA 78,6096- 6100.
WEIGLE, J. (1966). Assembly of phage X in vitro. Proc. Nat. Acad. Sci. USA 55, 1462-1466.
WEST, R. W. JR, NEVE, R. L., AND RODRIGUEZ, R. L. (1979). Construction and characterization of E. coli promoter probe plasmid vectors. I. Cloning of pro- moter containing DNA fragments. Gene 7, 27 l-288.
WEST, R. W. JR, AND RODRIGUEZ, R. L. (1980). Con- struction and characterization of E. coli promoter probe plasmid vectors; II. RNA polymerase binding studies on antibiotic resistance promoters. Gene 9, 175- 193.
