ACKNOWLEDGMENTS. This work was carried out in the laboratory of Fred Sanger, .... Roth, J. R., D. N. Anton, and P. E. Hartman. 1966. Histidine regulatory ...
JOURNAL OF BACTERIOLOGY, July 1981, p. 124-134 0021-9193/81/070124-11$02.00/0
Vol. 147, No. 1
Cloning and Restriction Map of the First Part of the Histidine Operon of Salmonella typhimurium WAYNE M. BARNES MRC Laboratory of Molecular Biology, Cambridge, England CB2 2QH, and Department of Biological Chemistry, Division of Biology and Biomedical Sciences, Washington University, St. Louis, Missouri 63110*
Received 4 March 1981/Accepted 13 April 1981
The first part of the histidine operon of Salnonella typhimurium, hisGpeaGD, has been cloned onto the vector plasmid mini-ColEl(pVH51). The resulting plasmid, pWB91, has a single EcoRI site and is 11,500 base pairs in size. The HindII restriction map was determined by the method of two-dimensional crossannealing between a partial digest pattern and a complete digest pattem. The restriction fragment containing the genetic control region was identified with the aid of the small (35-base pair) internal deletion 01242 and the observation that heteroduplexed restriction fragments containing this deletion have markedly reduced mobility on polyacrylamide gels. The genetic control region was then mapped in more detail with other restriction enzymes. The genetic orientation of the restriction map was determined with the aid of several deletions of integral HindII fragments generated in vitro.
The histidine operon of Salmonella typhimurium has become a useful tool for investigations in several areas of bacterial genetics. One of these areas is the control over gene expression at the level of transcription. The major mechanism of transcriptional control of the histidine operon is an attenuator, with the most extreme sequence features of any attenuator observed to date (most control codons in a row [seven]; strongest terminator stem with 16 perfectly symmetrical base pairs; longest run of U's [nine] at the presumed end of the leader RNA [10]; and most alternative stem structures in the leader RNA [30, 30a]). In other areas of study, the histidine operon was an early example of operon organization in bacteria (22), and it continues to be a subject for study of genetic phenomena, such as frame-shift suppression (12), duplications (3, 4), and mutagenesis, when the DNA under study here is reacted with environmental mutagens (2). Hundreds of mutations in the genetic control region and structural genes of the histidine operon have been isolated and studied (17, 18, 26, 30). As a background to transcriptional studies and to DNA sequence determination of the genetic control region hisGpea (10) (this region was formerly known as hisO; new terminology from reference 6), the first two genes, hisG and hisD, and mutants of all of the above, I have constructed here a mini-ColEl plasmid carrying the first part of the operon as a substrate for determination of the Hindll restriction map of this 124
region. The mapping method used for the HinduI fragments is a novel application of the method of Hutchison (C. A. Hutchison III, manuscript in prepartion) of two-dimensional crosshybridization. I also present a more detailed restriction map of the 983-base pair HinduI fragment (R5) containing the genetic control region.
MATERALS AND METHODS Chemicals. Deoxynucleotide triphosphates were obtained from Boehringer-Mannheim Corp.; a-3P-deoxynucleotide triphosphates (at a specific activity of 100 to 150 Ci/mmol) were from New England Nuclear Corp.; ethidium bromide, mitomycin C, and histidinol were from Sigma Chemical Co.; NZ-Amine A is a pancreatic digest of casein from Humko Sheffield; tryptone and yeast extract and agar were from Difco Laboratories; acrylamide was from BDH. All other chemicals were reagent grade. Bacterial and bacteriophage strains. Strains are listed in Table 1. Media. Minimal medium was M9 salts with thiamine and glucose as described previously (36); when appropriate, separately autoclaved histidinol was added at 55 ,ug/ml. Rich medium consisted of 20 g of tryptone or NZ-Amine A, 5 g of yeast extract, and 5 g of NaCl per liter. Growth and isolation of phage. Bacteria lysogenic for A h8Odhis c1857 S7, or carrying a Hol+ (hisD+) plasnid, were inoculated from single colonies into 30 ml of selective medium (minimal + histidinol) and then subcultured into 1 liter of rich medium in a 2-liter shaking conical flask. A h8Odhis lysogens were grown at 30 to 33°C to an optical density at 550 nm of 0.5, induced at 44°C for 20 min, and then returned to
VOL. 147, 1981
HISTIDINE OPERON RESTRICTION MAP
TABLE 1. Bacterial and bacteriophage strains TA2043
F- A(his-gnd) Stre
Reference and source Smith and Tong (48)
TA1933
TA2043 (A h80dhis c1857 S7, A h80 c1857 S7)
VoU (53), Smith and Tong (48); source, S. W. Artz
TA1940
TA2043 (A
Smith and Tong (48); source, S. W. Artz
Strain
Description
h8OdhisO1242 cI857 S7, A h80dhis cI857 S7) SB3137
F- eda edd A(his-gnd) Strr (A h80dhisO2344 c1857 S7, A h80 cI857 S7)
Source, S. W. Artz
C33
Same as TA2043
TA1933 cured of phage at 42°C
EQ20
C600 (mini-ColEl = pVH51)
Hershfield et al. (24); source, S. Brenner
WB313
C33 made resistant to some unknown phage which was infecting large-scale cultures
This paper
EQ4
F- (ColEl)
Source, S. Brenner
30°C for a further 5 h of incubation. Phage were isolated as described by Barnes (Ph.D. thesis, University of Wisconsin, Madison, 1974), with a modification of the procedure described previously (54), and purified away from helper phage by banding twice in CsCl (transducing phage are the lighter [and fainter] band). A h80dhus DNA was prepared by extraction with phenol as described previously (11), followed by extensive dialysis against DNA buffer (0.01 M Tris-hydrochloride [pH 7.9]-0.01 M NaCl-0.1 mM EDTA). Isolation of plasmid DNA. Plasmid-containing clones were always checked with the toothpick assay (7) for the presence of plasmid of the right size before growth of preparative cultures. Plasmid-containing strains were inoculated as described above for A h80dhis lysogens and grown to saturation in rich medium (without any chloramphenicol amplification, which is ineffective for the mini-ColEl-His plasmids). Cleared lysates were prepared as described previously (15, 32), DNA and RNA were precipitated with 10% polyethylene glycol-0.5 M NaCl (D. Sherrat, personal communication), and supercoiled plasmid was isolated by two bandings in EtBr-CsCl (27, 38). Ethidium was removed by extraction with isopropanol, and the DNA was dialyzed against DNA buffer before ethanol precipitation. Supercoiled DNA was stored in DNA buffer at -20°C.
Enzymes. DNA polymerase I from Escherichia coli (grade I) and its large fragment (33, 45) were obtained from Boehringer-Mannheim Corp. One unit is defined as in reference 39. T4 ligase was supplied by Miles Research Laboratories (one unit = 1 nmol of ends joined in 20 min at 370C). Restriction enzymes
125
are reviewed by Roberts (41); EcoRI was a gift of T. Rabbits; Hha was a gift from R. Roberts; HaeII, HindII, and Hinf were a gift from N. L. Brown; HaeIII was isolated as described previously (41); and AluI was isolated as described previously (42). One unit of restriction enzyme will digest 1 jg of lambda DNA in 1 h at 370C. All enzymes were stored in 50% glycerol at -20°C. The reaction buffer for all polymerase reactions and analytical restriction digests was NaTMD (0.05 M NaCl-0.01 M Tris [pH 7.9]-0.01 M MgClr-1 mM dithiothreitol), except that EcoRI buffer was 0.09 M Tris-hydrochloride (pH 7.4)-0.009 M MgCl2 when EcoRI was used alone. Preparative restriction digests with HaeIII or HhaI were carried out in buffer with no NaCl. Restriction digests to be followed immediately by a ligase reaction (without an intervening phenol extraction) were carried out in T4 ligase buffer (0.03 M Tris-hydrochloride [pH 8]-4 mM MgCl2-1 mM EDTA-0.2 mM rATP-10 mM dithiothreitol). The restriction enzyme was inactivated by heating at 70°C for 10 min before the ligase step. Transformation and colicin selection. The procedure for transformation was that described previously (16) except that cells were grown in rich medium and washed only with 0.1 M CaCl2, and after mixing with DNA in 0.1 M CaCl2, the suspension was heated to 37°C for 30 s before the 1-h incubation at 0°C. After exposure to the DNA, it was necessary to add 1 ml of rich medium and incubate the cells for 90 min at 370C to establish the plasmids before plating on selective medium (25). Colicin El was prepared by the method of Yanofsky (personal communication) from EQ4 cells grown in Luria broth (36) and induced with mitomycin C (2 ug/ml) at an optical density at 550 nm of 0.3. After 5 h of further incubation in the dark, the cells were collected and sonicated in 1/40 the culture volume of 0.05 M potassium phosphate, pH 7.5. The sonic extract was centrifuged at 10,000 x g for 20 min, dialyzed, and stored over a drop of chloroform. This colicin preparation was titrated and used as follows. Selection was applied by adding 1 drop (50 pl) to 0.2 ml of cells and spreading the entire mixture onto a fresh rich medium plate. This procedure relies on killing on the plate. The amount of colicin used was one-third of the amount that noticeably killed plasmid-containing cells. Gels. Polyacrylamide gradient gels, 5 to 17%, were poured as described by Jeppesen (28). The gel dimensions were 20 by 40 cm by 1.5 to 3 mm, and gels were poured six or eight at a time, using the gel-pouring box of Air et al. (1). At the time of pouring, 100 ml of buffer containing 2% sucrose and 0.05% bromphenol blue was first introduced into the bottom of the gelpouring box; then the acrylamide gradient (approximately 1 liter) was let in, and finally 40% sucrose with bromphenol blue and buffer was introduced under the acrylamide gradient to lift it up even with the bottom of the gel plates. Gradient gels could be wrapped in plastic and stored for at least 4 months at 40C before use.
Agarose gels were run in four-times-concentrated gradient gel buffer (7, 28) in the horizontal conformation of Shinnick et al. (47). Sometimes the melted petrolatum was only overlaid over the sample wells,
126
BARNES
the rest of the gel being covered with Saran wrap. Ethidium staining and photography of all gels was as described for the toothpick assay (7). Autoradiography of wet gels was as described previously (1, 8). DNA fragments were eluted from acrylamide gels electrophoretically as described previously (8). Two-dimensional cross-annealing restriction mapping. The two-dimensional cross-annealing restriction mapping technique (Hutchison, in preparation) overlaps two restriction patterns at a 900 angle and identifies the points of homology between them. In this application, the two patterns are an incomplete digest of pWB91 by HindII (unlabeled) and a complete digest by the same enzyme (labeled by nick translation). A 10-,ug amount of pW891 was digested with one-third of the amount of HindIl needed for a complete digest and electrophoresed in a wide slot on a 2% agarose slab gel. The appearance of this incomplete digest pattern is shown in Fig. 1A. The DNA pattern was transferred to a sheet of nitrocellulose filter (Millipore Corp.) by the Southern (49) blot method, modified for slab gels. The blotting buffer was 4x SSC (SSC is 0.15 M NaCl plus 0.015 M sodium citrate). The nitrocellulose membrane was dried in vacuo at 700C and used in the following second-dimension annealing. pWB91 DNA was labeled with 3P by nick translation, digested completely with HindII, and electrophoresed on 4% agarose, again in a wide slot. This DNA pattem was then subjected to the standard Southern blot procedure with the following variations. (i) The nitrocellulose membrane was that carrying the unlabeled incomplete digest pattern described above, turned ninety degrees to the labeled digest pattern. (ii) The blotting buffer and gel contained 4x SSC and 0.1% sodium dodecyl sulfate throughout, and (iii) the blot was carried out in an oven at 670C. Under these conditions, the labeled DNA did not stick to the nitrocellulose membrane unless it was annealing specifically to homologous sequences in the unlabeled pattern! Sequence homology between patterns was then visualized as spots on an autoradiograph (Fig. 1B). Pulse-chase restriction mapping. The method of Jeppesen et al. (29) was simplified as follows. A 0.25-pmol amount of restriction fragment was heatdenatured and annealed to 0.25 pmol of singlestranded DNA in a 20-pl volume of NaTMD buffer as described previously (8). A l-,l amount of [a-3P]dATP in 50% ethanol and 1 pl of a 1 mM solution of the other three dNTP's were added, and the solution was cooled to 1°C. A 1-U amount of DNA polymerase I large fragment was added; after 3 min, 1 t1 of 1 mM unlabeled dATP was added, and the mixture was then incubated with 1 U of restriction enzyme for 15 to 30 min at 370C. The reactions were stopped by adding 10 id of 0.1 M EDTA-50% glycerol-0.1%-bromphenol blue and then electrophoresed directly on a gradient acrylamide gel. Nick translation. Nick translation (35) was used to label DNA in vitro. A 10-,ug amount of plasmid DNA or 1 ug of fragment R5 DNA was treated in 50 pl of NaTMD buffer with 1 U of DNA polymerase I, three unlabeled dNTP's (50 uM), and one [a-3P]dNTP (2 gM). No DNase was added, since sufficient nicking activity was already present in the DNA po-
J. BACTERIOL. lymerase preparation used. After incubation at 200C for 15 to 30 min, the sample was heated to 70°C for 10 min to inactivate enzymes. After cooling, restriction digests could be carried out by adding enzyme directly to a portion of the labeled DNA. [a-3P]dNTP was not usually removed before restriction enzyme digestion or gel electrophoresis, but when it was, gel filtration over 1 ml of Sephadex G-100 was applied (14). Strand separation. pWB91 DNA was quantitatively nicked by the DNase + ethidium bromide procedure (21) or digested to linears with EcoRI before preparative separation of the complementary DNA strands by the poly(U,G) procedure (52). Pooled strands from the CsCl gradient were self-annealed and rebanded in CsCl, dialyzed, ethanol-precipitated, and resuspended in DNA buffer, with a final yield of only 10 to 20%. In vitro deletions. A 10-,ig amount of pWB91 DNA at 100 jig/ml in ligase buffer was partially digested with enzyme HindII (about one-third digest, the same as that used for the gel in Fig. 1A). The HindI enzyme was inactivated by heating at 700C for 10 min, and 1 pl of T4 ligase (0.25 Miles unit) was added to a 1-Itg portion and incubated overnight at 200C. One-fourth of this mixture was then used to transform E. coli C33, which was plated on a rich medium plate with colicin selection.
RESULTS Construction of mini-ColEl hisGpeaGD plasmid. X h8Odhis DNA and mini-ColEl (pVH51) DNA were cleaved with EcoRI and recombined in vitro by the method of Hershfield et al. (25). E. coli C33, which is deleted for the histidine operon, was transformed with the ligated DNA mixture and plated on glucose-minimal medium with added histidinol. On this medium, only those cells which received an active hisD gene (whose product converts histidinol to histidine) were able to grow. Seventeen Hol+ colonies (four independent) obtained were picked, purified, and checked for the ability to grow in the absence of histidinol (His'). (The existing literature on this operon designates bacteria which are His-, but which can utilize histidinol, as Hol+). Three clones were His', but this character was lost after purification on histidinol medium. When four independent Hol+ clones were subjected to the toothpick assay for plasmid presence and size (data not shown, but see Fig. 5 for an example of this assay), three contained plasmids over 20,000 base pairs in length, and one contained a smaller plasmid. Since the purpose of this construction was to obtain a small substrate for DNA structural studies, the smallest plasmid, designated pWB91, was chosen for further study. pWB91 is His- Hol+ and contained only one EcoRI restriction site. Evidently, the other RI site expected from the manner of its construction was deleted somehow. Comparative double di-
VOL. 147, 1981
.
8
HISTIDINE OPERON RESTRICTION MAP 6
5 43 2
6
5
43 2
127
1
_1/+4i-6 _
No
r
5+34-6 L
No*7
::
3+-66
4
I bl
b
2 + 7e
2+8+
2-|7
i3 M
_..
4 5
7
5
!1* 57
A B
_ __
FIG. 1. Two-dimensional cross-annealing. The procedure ofHutchison (in preparation) is described in the text. (A) The appearance of the unlabeled partial HindII digest of pWB91 DNA, electrophoresed on 2% agarose. This gel is not the actual one used for the experiment in (B), but it was prepared identically. (B) Final autoradiograph. The unlabeled bands were oriented horizontally, as in (A). The labeled bands of the complete digest are oriented vertically (having been electrophoresed from right to left), and their positions are indicated at the top of the autoradiograph. Spots represent homology between the completely digested, labeled pattern and the partially digested, unlabeled pattern. The identifications deduced for the bands in the partial digest pattern are indicated on the right of the autoradiograph. Identifications marked with an asterisk comprise a data set sufficient to determine the map. The other identifications represent redundant data. (C) A line tracing of the spots used in the mapping analysis. The vertical dotted lines approximately indicate the position of the radiolabeled complete Hindu digest. The horizontal solid lines indicate the positions of the complete HindII digest products of the cold dimension, which was actualy a partial digest. The spots indicated with dotted lines were observed once in a replicate experiment and were not used in the analysis, although they are consistent. All other resolved spots were seen in two separate experiments.
gests (with HindII and EcoRI) of pWB91 and X h8Odhis DNA (data not shown) indicated that the histidine operon boundary of the inserted DNA was intact, the fortuitous deletion having removed some phage DNA and some miniColEl DNA corresponding to the small HindIIEcoRI fragment (140 base pairs) of mini-ColEl. Comparative digests of pWB91 and X h80dhis DNA (Fig. 2) indicated that all of the HindII sites on pWB91 were in the cloned DNA and that the histidine HindII fragments were cloned intact, since fragments 2 through 8 had electrophoretic mobilities identical to HindII fragments of the transducing phage DNA, even when the two digests were mixed together (Fig. 2, last channel). (This conclusion neglects the unlikely possibility that accidental alterations to the
DNA created misrepresentative HindII fragments of fortuitously identical mobilities.) The EcoRI-digested linear form of pWB91 was sized by comparison to phage lambda EcoRI fragments on 1% agarose gels (23). It is 11,500 base pairs long. Restriction site mapping. Since restriction enzyme HindH was found to have no site on ColEl DNA and only one site on mini-ColEl DNA, the problem of mapping an insert into these vectors was not complicated by vector DNA fragments (the vector DNA is part of the largest fragment). The eight HindII fragments of pWB91, designated Rl to R8, were ordered by the two-dimensional cross-hybridization technique of Hutchison (in preparation). The original technique was used to identify homolo-
128
BARNES
J. BACTERIOL.
tN ,>, --) IC)I sn 3 +
ca. CL
-+*
0 O
RI
gies between restriction digests of two different enzymes. In the present application, an unlabeled, incomplete digest (Fig. 1A) was analyzed against a labeled, complete digest of the same enzyme. In this favorable case, in which the HindI fragments are evenly spaced by gel electrophoresis, the entire map could be determined by inspection of the autoradiograph in Fig. 1B. This figure presents the data with bands in the
unlabeled, incomplete HindlI digest oriented horizontally, and the bands in the labeled, complete digest oriented vertically (Fig. 1C). This experiment essentially identifies the incomplete R2 digest products in the unlabeled pattern. The fragments thus identified are marked in Fig. 1B. Information from the labeled fragments 7 and 8 could not be used directly for these identifications, since these small fragments were run R 3R3^ -off _ the gels. Fragments 7 and 8 were assigned to their incomplete digest products where compelR4 n.g i logical on the basis of mobility shifts con2lWy sistent with their size. R5 One pair of fragments in the incomplete digest had nearly identical mobilities (2 + 7 + 8 and 3 + 5), but this complication could be resolved by application of the following rule. This rule, which can be observed to hold for the partial digestion product 2 + 7 + 5 and 8 + 2 + 7 + 5, R6 is that the larger nick-translated fragments, being more radioactive in proportion to their size, gave rise to darker spots than did the smaller nick-translated fragments which annealed to the same partial fragment. This rule resolved the data for the comigrating particles 2 + 7 + 8 and 3 + 5. The spot from nick-translated fragment two was weaker than the spots from fragments 3 and 5; therefore, the spot from 2 could not have been annealing to the same partial digestion product. The identifications shown are consistent with the sizes of the fragments and indicate a unique map order in the two large segments, 1 + 4 + 6 and 3 + 5 + 7 + 2 + 8. The orientation of these two large segments vis-a-vis each other remains to be established. The definite absence of a 3 + 6 or a 5 + 3 + 6 fragment R7 is strong, negative evidence that 1 must be next to 3, and 6 must be next to 8. Consistent with this conclusion, there is no indication of a 1 + 8 fragment but there is an apparent 1 + 3 fragment, although this upper region of the gel has less resolution than the rest. There is no eviR8 R8 dence as to whether the lowest spot in Fig. 1B is 6 or 6 + 8, although a 6 + 8 fragment must exist FIG. 2. Identification of fragment containing deletion 01242. This 5 to 17% acrylamide gel presents digests of DNA from (left to right) pVH51, pWB91, A h80dhis+, A h8OdhisO1242, and a mixture ofpWB91
and A h80dhis+. The HindII fragments ofpWB9l are labeled on the left. It can be seen that fragment R5 is the one affected by 01242.
VOL. 147, 1981
HISTIDINE OPERON RESTRICTION MAP
procedure.) Thus, the only clear data obtained by this experiment were: 6 -- 8 (6 is adjacent to 8), 2 -. 7, and 4 -- 6. These data confirm the weakest parts of the HindII map data deduced above (Fig. 3). The restriction map of the hisGpea region HindIl fragment R5, which was found to contain deletion 01242 (see next section), was determined in more detail. Nine Hha fragments obtained by digestion of purified HindII fragment R5 were designated RH51 to RH59 and sized by comparison with a Hha digest of 4X174 (sizes from reference 44). Fragments RH51 to RH58 were isolated from a 5 to 17% acrylamide gradient gel and ordered by the pulse-chase mapping method, using poly(U,G)-isolated Hstrands of pWB91 (see above). Very clear results were obtained (data not shown), indicating the map order shown in Fig. 3. The three HaeIII sites on HindII fragment R5 were mapped in the same way. A double digest with Hha and Hinf determined the location of the single Hinf site in HindII fragment R5 as shown. The two HaeII sites were located on the basis of the sizes of the
for the map order shown. The data set identified with an asterisk in Fig. 1B is sufficient to determine the map, and the rest of the data comprise an independent set of data that allow the same conclusion. Thus, this one compact experiment determined the Hindl fragment order, if all of the data were given equal weight. However, those data from the top part of the gel, where the partial fragments containing fragment Rl are poorly resolved, and the negative data consisting of the absence of certain spots could use some confirmation. To confirn the map order, the pulse-chase mapping technique of Jeppesen et al. (29) was applied, using the H strand from a poly(U,G) strand separation (52) of pWB91 DNA. This technique identifies nearest neighbors of restriction fragments, but it was found that the large (>800 base pairs) nearest neighbors could not be identified, due to an apparent failure of DNA polymerase I to continue unlabeled synthesis over long enough stretches. (This problem was encountered and solved by Summers [51] but I have not repeated the experiment with the improved F' or Sohuonalla DNA
Phage DNA
pWB32
Histidine DNA pa G
a
-
pWB 3 pWB 10
DNA
Phenotype:
i i~~~~~~~4
I
R2 2080
.
R7? R5 A
.
,A
R3 L
{I')
RI 4600 base pairs
,;" I000'"' 1500 '(600) 1-'
hol+ holv.weak
!.:
I
R4 I R6 I.IIp Hind II .IL-_ 1siteS 5 on pWB91 1250 ' 730 ISO .-
mini colEl
D -
pWB II
129
250
hol*
I1 I hol+
le
".
EcoRi
A01242 is located in this Hha fragment ,'
Detail of Hind
fragrnent RS
350
I
1
RH51
I
_
Hho I Has
I
220 RH52
I
153
RH53
.
I
.1 00
a
N
I RH541
J
has III
HinfI G Hph ~
~~goen
FIG. 3. Physical map ofpWB91 DNA. The plasmid is a circle 11,500 base pairs in length. The circle has been cut arbitrarily at one of the HindII sites. The apparent source of the DNA in pWB91 is given at the top of the figure as deduced from comparative HindII and EcoRI digests of A h80dhis, A h80, and mini-ColEI DNA. The fragments upstream from the hisGp region could possibly be F' DNA, from the manner of construction of the source A h80dhis, which was excised from F' his (53). p, a, G, and D indicate the order and approximate locations of histidine operon genes. Fragments generated by restriction enzyme HindII are designated RI, R2, etc., according to size. Fragments generated by digesting fragment R5 with Hha are designated RH51, RH52, etc. Fragment 1' is the product of a HindII + EcoRI double digest. Approximate sizes in base pairs are indicated for some of the fragments. The DNA retained in various deleted plasmids is indicated by solid lines, with deletion boundaries located exactly at HindII sites. The phenotype indicated for each plasmid is that of WB313 cells carrying the plasmid. The location of the G gene boundary was determined from DNA sequence data (10) compared with the N-terminal amino acid sequence of G protein (37).
130
BARNES
products formed, identification of the Haell fragment containing 01242 (see below), and the knowledge that HaeII sites (RGCGCY) must coincide with HhaI sites (GCGC) (41). Identification of the HisO fragment. The fragment containing the histidine operon control sequence was identified in various digests by experiments- such as that shown in Fig. 2. Advantage was taken of the existence of the small internal deletion 01242, which deletes part of the attenuator of the control region (5, 30, 31, 43; Barnes, manuscript in preparation). Comparison ducing phage DNA with digests of histidine-t and without deletion 01242 identified the target fragment as the only one with decreased mobility as a result of the deletion. The effect of the deletion could be seen in digests with AluI, HaeIII, EcoRI*, Hinf, HaeII, and Hindll. Figure 2 presents the data for this type of experiment with enzyme HindI. The affected fragment can be seen to be the one called R5 on pWB91. From the magnitude of the mobility shift, the size of the deletion can be estimated to be 30 to 50 base pairs. No effect of 01242 could be seen in digests with HhaI, since apparently the affected Hha fragment was a small one in a crowded, poorly resolved region of the analyzing gels. A more complicated procedure was therefore utilized to identify the target Hha fragment. In this experiment labeled (by nick translation), purified HindH fragment R5 was digested with H/ia and then heteroduplexed with a 10-fold excess of Hha digests of pVVB91 DNA (hisGa+), A h8OdJis01242 DNA, or A h80dhis02355 (point mutant) DNA. (Plasmids carrying any histidine mutations are not yet available.) The heteroduplexed Hha fragments thus formed were electrophoresed on a 5 to 17% acrylamide gradient gel and autoradiographed (Fig. 4). The heteroduplexes with wild-type or point mutant histidine operon DNA gave rise to the same pattern as the undenatured, labeled Hha digest of fragment R5, except for the boundary fragments, which were larger since they overlapped the terminal HindII sites of fragment R5. The heteroduplexes with 01242 DNA were identicaL except for the 220-base pair fragment known as RH52, whose mobility was hypersensitively decreased. Figure 4, column D compared with columns C and E shows that the heteroduplex form (wild type/ 01242) of RH52 is apparently greatly retarded in electrophoretic mobility, and can be seen to migrate at or slightly ahead of the position of RH51 in column D. The magnitude of this effect is unexpected from the small size of the deletion (35 base pairs; 30, and unpublished data). It was concluded that a small, single-stranded loop has
J. BACTERIOL.
a large effect on the electrophoretic mobility of a restriction fragment, and that the affected fragment contained the deletion 01242. Since I made this observation, this effect has been independently observed and expanded upon to form the basis of a new separation method, denaturation gradient electrophoresis (19, 20). (When this experiment was carried out with labeled pWB91 and HinduI digests only, fragment R5 appeared as two slower-moving bands when it was heteroduplexed to A h80dhisO1242 DNA; data not shown.) Genetic map orientation. A series of deletions of integral numbers of Hindl firagments was created for pWB91 by a procedure involving incomplete digestion with restriction enzyme HindIu, followed by recircularization with the
52
5 3_ 54
MR,
55 56 57
58
FIG. 4. Identification of Hha firagment carrying deletion 01242 by gel analysis of heteroduplexed restriction fragments. Pure HindII restriction fragment R5 was labeled by nick translation and then digested with HhaI (column A). This DNA was then heat denatured and reannealed in the presence of a fivefold molar excess of a Hha digest of the following
cold DNAs: none (column B); A h80dhis (column C); A h80dhis01242 (column D); pWB91 (colwnn E). The numbers on the left identify the Hha fragments of HindIIfragment R5. The arrow on the right indicates the fragment (RH52) affected by deletion 01242.
VOL. 147, 1981
HISTIDINE OPERON RESTRICTION MAP
blunt-end ligase activity of T4 ligase (46; see above). After these enzyme treatments and transformation with 0.25 ,ug of DNA, 50 colicinresistant colonies were obtained, of which 70% were Hol+ and the rest were Hol- (probably the desired deletions). When 14 of these HolV strains were checked for plasmid size by the toothpick assay (Fig. 5), all but one were indeed deletions of various sizes. In fact, 7 out of 32 Hol+ clones were also deletions (two colonies had no visible plasmid in the toothpick assay), for a total deletion frequency of 43%. To determine how exact the blunt-end ligation was, and to determine which HindII fragments were deleted, several plasmids were prepared in quantity and digested with HindH (Fig. 6). All but one of the deleted planids tested have restriction patterns consistent with the restriction map and recircularization of partial digest products containing fragment Rl, which contains the selected replicon and colicin resistance gene. The inconsistent plasmid, pWB2, contains fragments 1, 2, 4, and 6 only. Evidently this plasmid resulted from a bimolecular event involving fragment 1 + 4 + 6 and fragment 2. This event was possible because of the high DNA concentration necessary to achieve efficient blunt-end ligation (50). pWB4 contains a new fragment that, judging from the restriction map and the size of the fragment, is an inexact fusion of fragments 3 and 4 in which a HindII site was destroyed. Plasmids pWB1, 3, 10, and 32 were found to have deleted the exact HindUI fragments indicated in the map in Fig. 2. The orientation of the histidine operon with respect to the restriction map was determined by careful comparison of the phenotypes of plasmids pWBl and pWB3. pWB1 is definitely Hol(hisD+), but it was noticed that, in fact, pWB3
1 ,m=ww hL
2 3 4 5 .-Ilmww R.-
6 7 8 Apiollbft
-
. -
131
is weakly Hol+ (hisD+), with colonies of various small sizes appearing after 3 days of incubation on minimal medium plus histidinol Presumably, this plasmid retains an intact hisD gene but has lost the his promoter. (A few faster-growing colonies which appear with time apparently result from spontaneous events such as genetic duplications [3, 4] which put the hisD gene under control of a new promoter. The toothpick assay of 25 of these indicated several positive changes in plasmid size, and some with no change, but no deletions were apparent. These effects were not investigated further.) Thus, the hisD gene must be at least partly contained on HindUI ragment R3, and since hisGa is in fragment R5, the genetic map must have the relative orientation shown in the map in Fig. 2. Similar logic concerning the location of the promoter indicates that, based on the phenotypes of plasmids pWB91, 32 and 3 (Fig. 2), the promoter must be on HindUI fragments R5 or R7, which is consistent with the previously determined location of the promoter-linked mutation 01242 on fragment R5 (see above). From the approximate lengths ofDNA needed to code for the G protein (900 base pairs) (37; R. N. Husson and W. M. Barnes, manuscript in preparation) and the D protein (1,300 base pairs) (W. Gray and T. Kohno, personal communication), the EcoRI site located some 2,400 base pairs downstream from hisGa and the start of the G gene must be in the first part of the third gene of the operon, hisC. DISCUSSION The advantages of a small molecule (l10 kilobases) for detailed DNA structural analysis include economy of DNA preparation and simplicity of restriction fragment analysis. Accord-
9 10 11 12 91 91 91 -~~ ~¢-_t
-..
_.
fikl.
.-.or igin I
rRNA
1
---
7~~~~~.
FIG. 5. Toothpick assay (7) of HindII-generated deletions of pWB91. Colonies containing the original pWB91 are analyzed in the right-most three channels.
132
J. BACTERIOL.
BARNES -.
e
I u i _ 'D L. d,w.
.,
R2 R3 R4 R5
R6
ment approximately 220 base pairs in lergth. Several deletions were created on the plasmid in vitro by partial HindII digestion followed by ligation. The actual object of this exercise was to create a small Hol- deletion that could be used to cross regulatory mutants onto the plasmid by in vivo recombination. This purpose has not yet been realized, but the deletions did serve to establish the genetic orientation of the cloned histidine operon DNA. With the knowledge of the EcoRI and HindII sites determined here, the same hisGpeaGD DNA has been cloned onto the single-stranded phage M13 (9). The HhaI and HaellI map determined here was then used as a basis for DNA sequencing of part of the histidine operon genetic control region (8, 10) and the rest of the histidine DNA between the fragment R7 and the EcoRI site in hisC (Barnes and Husson, in
preparation). ACKNOWLEDGMENTS This work was carried out in the laboratory of Fred Sanger,
whom I thank for his hospitality. I thank Clyde Hutchison for sharing his method during its development.
LITERATURE CITED
R7
1. Air, G. M., F. Sanger, and A. R. Coulson. 1976. Nucleotide and amino acid sequence of gene G of phiX174. J. Mol. Biol. 108:519-533. 2. Ames, B. N., J. McCann, and E. L. Yamasaki. 1975. Methods for detecting carcinogens and mutagens with 3. 4.
FIG. 6. HindII digests of HindII-generated deletions. The original pWB91 is analyzed in the leftmost channel. This is a 5 to 17% acrylamide gradient gel.
ingly, to obtain a suitable substrate for DNA sequence analysis of the genetic control region of the histidine operon from S. typhunurium, the first part of this operon was cloned onto the plasmid vector mini-ColEl (pVH51) (24, 25). The HindII restriction map of the plasmid was determined, and the 983-base pair fragment containing the target genetic control region was mapped in detail with several other restriction enzymes, notably HhaI, which maps the fragment in great detail all by itself (Fig. 3). The fragment containing the genetic control DNA was identified in several restriction digests by comparing the restriction patterns of A h80dhis+ and A h8OdhisO1242 DNA. 01242 was thus verified to be a short deletion by its effect on the mobility of the restriction fragment containing it. 01242 was localized to a HhaI frag-
5.
6.
the Salmonella/manmala-microsome mutagenicity test. Mutat. Res. 31:347-364. Anderson, R. P., and J. R. Roth. 1978. Tandem genetic duplications in Salmonelld typhimurium: amplification of the histidine operon. J. Mol. Biol. 126:53-71. Anderson, R. P., and J. R. Roth. 1978. Tandem chromosomal duplications in Salmonella typhimurium: fusion of histidine genes to novel promoters. J. Mol. Biol. 119:147-166. Artz, S., and J. Broach. 1975. Histidine regulation in Salnonella typhimurium: an activator-attenuator model of gene regulation. Proc. Natl. Acad. Sci. U.S.A. 72:3453-3457. Bachmann, B. J., and K. B. Low. 1980. Linkage map of Escherichia coli K-12, edition 6. Microbiol. Rev. 44:1-
56. 7. Barnes, W. M. 1977. Plasmid detection and sizing in single colony lysates. Science 195:393-394. 8. Barnes, W. IL 1978. DNA sequencing by partial ribosubstitution. J. Mol. Biol. 119:83-89. 9. Barnes, W. IL 1979. Construction of an M13-histidine trnsducing phage: a single-stranded cloning vehicle with one EcoRI site. Gene 5:127-139. 10. Bares, W. IL 1978. DNA sequence from the histidine operon control region: seven histidine codons in a row. Proc. Natl. Acad. Sci. U.S.A. 75:4281-4285. 11. Barnes, W. M., W. S. Reznikoff, F. R. Blattner, R. C. Dickson, and J. Abelson. 1975. The isolation of RNA homologous to the genetic control elements of the lactose operon. J. Biol. Chem. 250:8184-8192. 12. Bosal, L., and J. R. Roth. 1980. The influence of codon context on genetic code translation. Nature (London)
288:123-127. 13. Brenner, M., and B. N. Ames. 1971. The histidine operon and its regulation, p. 349-387. In H. J. Vogel
VOL. 147, 1981
HISTIDINE OPERON RESTRICTION MAP
(ed.), Metabolic regulation. Academic Press, Inc., New York. 14. Brownlee, G. G., and E. Cartwright. 1977. Rapid gel sequencing of RNA by primed synthesis with reverse transcriptase. J. Mol. Biol. 114:93-117. 15. Clewell, D., and D. R. Helinski. 1970. Properties of a supercoiled deoxyribonucleic acid-protein relaxation complex and strand specificity of the relaxation event. Biochemistry 9:4428-4440. 16. Cohen, S. N., A. C. Y. Chang, and C. L. Hsu. 1972. Nonchromosomal antibiotic resistance in bacteria: genetic transformation of Escherichia coli by R-factor DNA. Proc. Natl. Acad. Sci. U.S.A. 69:2110-2114. 17. Ely, B. 1974. Physiological studies of Salmonella histidine operator-promoter mutants. Genetics 78:593-606. 18. Ely, B., D. B. Fankhauser, and P. E. Hartman. 1974. A fine structure map of the Salmonella histidine operator-promoter. Genetics 78:607-631. 19. Fischer, S. G., and L S. Lerman. 1979. Length-independent separation of DNA restriction fragments in two-dimensional gel electrophoresis. Cell 16:191-200. 20. Fischer, S. G., and L S. Lerman. 1980. Separation of random fragments of DNA according to properties of their sequences. Proc. Natl. Acad. Sci. U.S.A. 77:44204424. 21. Greenfield, L., L. Simpson, and D. Kaplan. 1975. Conversion of closed circular DNA molecules to singlenicked molecules by digestion with DNase I in the presence of ethidium bromide. Biochim. Biophys. Acta 407:365-375. 22. Hartman, P. E., J. C. Loper, and D. Serman. 1960. Fine structure mapping by complete transduction between histidine-requiring Salmonella mutants. J. Gen. Microbiol. 22:323-353. 23. Helling, R. B., H. M. Goodman, and H. W. Boyer. 1974. Analysis of endonuclease R.EcoRI fragments of DNA from lambdoid bacteriophages and other viruses by agarose-gel electrophoresis. J. Virol. 14:1235-1244. 24. Hershfield, V., H. W. Boyer, L Chow, and D. R. Helinski. 1976. Characterization of a mini-ColEl plasmid. J. Bacteriol. 126:447-453. 25. Hershfield, V., H. W. Boyer, C. Yanofsky, M. Lovett, and D. R. Helinski. 1974. Plasmid ColEl as a molecular vehicle for cloning and amplification of DNA. Proc. Natl. Acad. Sci. U.S.A. 71:3455-3459. 26. Hoppe, I., M. Johnston, D. Biek, and J. Roth. 1979. A refined map of the hisG gene of Salmonella typhimurium. Genetics 92:17-26. 27. Hudson, B., J. H. Dawson, R. Desiderio, and C. W. Mosher. 1977. Ethidium analogues with improved resolution in the dye-bouyant density procedure. Nucleic Acids Res. 4:1349-1359. 28. Jeppesen, P. G. N. 1980. Separation and isolation of DNA fragments using linear polyacrylamide gradient gel electrophoresis. Methods Enzymol. 65:305-319. 29. Jeppesen, P. G. N., L. Sanders, and P. M. Slocombe. 1976. A restriction cleavage map of PhiX174 DNA by pulse-chase labelling using E. coli DNA polymerase. Nucleic Acids Res. 3:1323-1339. 30. Johnston, H. M., W. M. Barnes, F. G. Chumley, L. Bossi, and J. R. Roth. 1980. Model for regulation of the histidine operon of Sabnonella. Proc. Natl. Acad. Sci. U.S.A. 77:508-512. 30a.Johnston, H. AL, and J. R. Roth. 1981. DNA sequence changes of mutations altering attenuation control of the histidine operon of Salnonella typhimurium. J. Mol. Biol. 145:735-756. 31. Kasai, T. 1974. Regulation of the expression of the histidine operon in Sabnonella typhinurium. Nature (London) 249:523-527. 32. Katz, L, D. T. Kingsbury, and D. R. Helinski. 1973. Stimulation by cyclic adenosine monophosphate of
33.
34.
35. 36.
37.
38.
39. 40.
41. 42.
43.
44.
45. 46.
47. 48. 49. 50.
133
plasmid deoxyribonucleic acid replication and catabolite repression of the plasmid deoxyribonucleic acidprotein relaxation complex. J. Bacteriol. 114:577-591. Klenow, H., K. Overgaard-Hansen, and S. A. Patkar. 1971. Proteolytic cleavage of native DNA polymerase into two different catalytic fragments. Eur. J. Biochem. 22:371-381. Lewis, J. A., and B. N. Ames. 1972. Histidine regulation in Salmonella typhimirum. XI. The percentage of transfer RNAhi, charged in vivo and its relation to the repression of histidine operon. J. Mol. Biol. 66:131-142. Maniatis, T., A. Jeffrey, and D. G. Kelid. 1975. Nucleotide sequence of the rightward operator of phage lambda. Proc. Natl. Acad. Sci. U.S.A. 72:1184-1188. Miller, J. H. 1972. Experiments in molecular genetics, p. 431-433. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Piszkiewicz, D., B. E. Tilley, T. Rand-Meir, and S. M. Parsons. 1979. Amino acid sequence of ATP phosphoribosyltransferase of Salmonella typhimurium. Proc. Natl. Acad. Sci. U.S.A. 76:1589-1592. Radloff, R., W. Bauer, and J. Vinograd. 1967. A dyebouyant-density method or the detection and isolation of closed-circular duplex DNA: the closed circular DNA in HeLa cells. Proc. Natl. Acad. Sci. U.S.A. 57:15141512. Reznikoff, W. S., R. B. Winter, and C. K. Hurley. 1974. The location of the repressor binding sites in the lac operon. Proc. Natl. Acad. Sci. U.S.A. 71:2314-2318. Richardson, C. C., C. L. Schildkraut, H. V. Aposhian, and A. Kornberg. 1964. Enzymatic synthesis of deoxyribonucleic acid. XIV. Further purification and properties of DNA polymerase from E. coli. J. Biol. Chem. 239:222-232. Roberts, R. J. 1976. Restriction endonucleases. Crit. Rev. Biochem. 6:123-163. Roberts, R. J., J. B. Breitmeyer, N. F. Tabachnik, and P. A. Meyers. 1975. A second specific endonuclease from Haemophilus aegyptius. J. Mol. Biol. 91:121123. Roth, J. R., D. N. Anton, and P. E. Hartman. 1966. Histidine regulatory mutants in Salmonella typhimurium. I. Isolation and general properties. J. Mol. Biol. 22:305-323. Sanger, F., G. M. Air, B. G. Barrell, N. L. Brown, A. R. Coulson, J. C. Fiddes, C. A. Hutchison, P. M. Slocombe, and M. Smith. 1977. Nucleotide sequence of bacteriophage PhiX174 DNA. Nature (London) 265: 687-695. Setlow, P., D. Brutlag, and A. Kornberg. 1972. Deoxyribonucleic acid polymerase: two distinct enzymes in one polypeptide. J. Biol. Chem. 247:224-231. Sgaramella, U. 1972. Enzymatic oligomerization of bacteriophage P22 DNA and of linear simian virus 40 DNA. Proc. Natl. Acad. Sci. U.S.A. 69:3389-3393. Shinnick, T. M., E. Lund, 0. Smithies, and F. R. Blattner. 1975. Hybridization of labeled RNA to DNA in agarose gels. Nucleic Acids Res. 2:1911-1921. Smith, G. R., and B. Tong. 1974. Construction of p80 dhis carrying Salmonella typhimurium histidine operon mutations. J. Bacteriol. 120:1223-1226. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-517. Sugino, A., H. M. Goodman, H. L. Heyneker, J. Shine, H. W. Boyer, and N. R. Cozzarelli. 1977. Interaction of bacteriophage T4 RNA and DNA ligases in joining of duplex DNA at base-paired ends. J. Biol. Chem. 252:
3987-3994. 51. Summers, J. 1975. Physical map of polyoma viral DNA fragments produced by cleavage with a restriction enzyme from Haemophilus aegyptius, endonuclease R-
134
BARNES
HaeI. J. Virol. 15:946-953. 52. Szybalski, W., H. Kubin , Z. Hradechna, and W. C. Summers. 1971. Analytical and preparative separation ofthe complementary DNA strands. Methods Enzymol. 210:383-413.
53. Voll, ML J. 1972. Derivation of an F-merogenote and a
J. BACTERIOL. 4,80 high-frequency transducing phage carrying the histidine operon of Salnwnella. J. Bacteriol. 109:741-750. 54. Wilcox, G., J. Singer, and L Hefferman. 1971. Deoxyribonucleic acid-ribonucleic acid hybridization studies on the L-arabinose operon of Escherichia coli B/r. J. Bacteriol. 108:1-4.
