Deletions and insertions in the immunity region of coliphage lambda: Revised measurement of the promoter-startpoint distance

VIROLOGY

62,

4.56451 (1%-i)

Deletions

and insertions

Lambda:

Revised

in the Immunity

Measurement

Region of Coliphage

of the Promoter-Startpoint

Distance F. R. BLATTNER’, McArdle

M. FIANDT. Laboratory,

K. K. HASS, P. A. TWOSE,

C:niuersity

of Wisconsin. Madison,

Accepted August

AND

W. SZYBALSKI

W~sconsm 53706

7, 1974

Plaque-forming XcI mutants which carry internal deletions in the immunity region, including genes c1 and rex, were isolated. Among seven independent isolates. only three demonstrably different deletions were obtained. The largest deletion (KH54 or KH115). which eliminates almost all of the DNA between promoters pI. and plc, spans the distance from 74.1 to 78.4 %A, as measured in relation to the left (0 Rh) and right (100 9%X)termini of the mature A DNA molecule. There were four representatives of a second (74.7-77.7’(x) and one of a third (‘75.2278.l”rX) deletion. In addition. a mutant carrying a 1400.nucleotide pair-long insertion in gene c1 was isolated. The cI-rex deletions served as markers for more accurate measurements of the physical positions of the endpoints of several bio insertions, a dv plasmid. and the endonucleolytic cleavage sites for the restriction nucleases by electrcm micrography ofDNA heteroduplexes. This led to a reevaluation of our earlier estimate of the interval length between the p, promoter mutation sex1 and the s,, startpoint for the major leftward operon of A. In agreement with the independent results of others. we find that the p,. to s, distance is within the limits of 24-62 nucleotide pairs 1~~:sewl) or 11 to 4,5 nucleotide pairs (s, -HinII cut in o,.), INTRODUCTION

transcription by E. coli RNA polymerase (Blattner and Dahlberg, 1972). The immunity region of’ the Escherichia The present study was undertaken to coli bacteriophage X (Fig. 1) contains three genes, which code, respectively. for the X isolate deletion mutants lacking most of repressor (cl). the “antirepressor” (cro or the DNA between these promoters. so as to ton, and the rex product, and several rec- permit setting an upper limit on the size of regions which are esognition sites, namely. promoters, opera- the nontranscribed of these promoters tors, and startpoints (Eisen and Ptashne, sential for functioning 1971; Ptashne 1971; Blattner and Dahl- and operators. Such deletions could also serve as markers for more accurate electron berg, 1972). The two divergently oriented micrographic measurement of the positions promoters, p,. and p H, are located approxiof’ the bio and dv endpoints employed in mately 2000 nucleotide pairs apart within this region. From RNA sequence analysis it our earlier estimate of’ the minimum size of was shown that neither the operators. o,, or the s,.-p,,o,. region (Blattner et al.. 1972). These deletions should also be helpful in oI{, nor the promoter p,. (and most probably physical mapping of the cleavages by the plc which overlaps with the o,{ mutations: restriction endonucleases in the lambda Ordal and Kaiser. 1973) are transcribed into RNA during the normal initiation of immunity region (Allet. 1973; Aliet et al., 1973; Allet and Solem, 1974). 1 Present address: Department of Genetics. UniVarious genetic rearrangements of’ the versity of’ Ll’isconsin. Madison. LVI ,53706. immunity region have been described, for

DELETIONS

AND INSERTIONS

IN PHAGE

X

459

Flu. 1. Genetic and physical map of the immunity-proximal region of phage h. The I and r strands of A DNA are represented by heavy lines. The positions on the X map (‘%A) are referred to the left terminus of mature X DNA taken as 0 ‘:;A and expressed as a percentage of the Apapa genome length, most rounded off to the nearest 0.1 “;X (Westmoreland et n/.. 1969). The 1 ‘(X unit corresponds to 465 nucleotide pairs (Davidson and Szybalski, 1971). The open arrows indicate the leftward and rightward RNA transcripts drawn to scale (Blattner et al.. 1972; Hayes and Szybalski, 1973). The endpoints of the KH and nin5 deletions (represented by thin lines), the KHlOO and r32 insertions (represented by loops drawn approximately to scale). and the imm434 (represented by a thin line with the internal homology segment indicated by a heavy line). imm21, hy42. and several bio substitutions are shown above the h map (Tables 2 and 3; Westmoreland et al., 1969; Fiandt and Szybalski. 1973; M. Fiandt. E. H. Szybalski, and W. Szybalski, current data). The positions of the X genes (N, rex, cl, tof or cro, CR, 0, 0 and recognition sites, including the promoters (P,. , PH, Prm ), operators (a., on), startpoint for the major leftward RNA (s,.), terminators (tH,. tRf) and the ori site for the initiation of X DNA replication, are indicated on the h map, Below the map are shown the cleavage sites for the HpaII. HinII (HindII). HinIII (Hi&III), HaeII and Hoe111 restriction endonucleases (from Hemophilus parainfluenzae, H. influenzae strain Rd and H. aegyptius, respectively), as determined by Allet and Solem (1974) and including subsequent data kindly supplied by B. Allet, and by L. Robinson and A. Landy. The length of the fragments was based on their migration in acrylamide gels and is expressed in nucleotide pairs tn.p.). The approximate placement of these cleavage sites was based on the following measurements, which included the use of the KH deletions: A. For the Hin nucleases: (a) The distance between the left dvl and KH54 endpoints in Xdvl/XKH54 heteroduplexes is 0.81 %X (Table 3; Fig. 3). This places the left dvl endpoint at 73.26 %‘oh.(b) The distance between the s.Hin cut (Allet and Solem, 1974; Maniatis and Ptashne, 1973) and the left dvl endpoint (novel joint in Xdv circle) is 0.31 ‘%‘nx,as measured in heteroduplexes between h and the Xdv fragment containing the novel joint (Fig. 3). This places the 9. cleavage at 73.57 %h and, thus the distance between the q, cleavage and the left KH54 endpoint (74.07 7X) is 0.50 “rX, i.e., 233 n.p. The distance between the o,. cleavage (73.57) and the 3h-1 endpoint (73.5) is 0.07%X (33 n.p.); (c) the Hin fusion fragment between the q, and oH cuts derived from XKH54 is 280 n,p. (Allet and Solem, 1974). This places the o,~ cut at 47 n.p. (280 minus 233). or O.l?X, to the right of the right (78.42) end of the KH,54 deletion. i.e.. at 78.n% c
460

BLATTNER

ET AL.

at 37”. the infective centers were diluted into 50 ml of tryptone broth and shaken for 14-%O hr at 37”. The lysates were treated with 0.5 ml of chloroform and clarified by low-speed centrifugation. Bacteriophage crosses. We employed a rapid method to score recombinants between the X&Pputative immunity-deletion mutants and several hbio phages listed in Table 1. Crosses were performed on plates overlaid with E. coli TC600 in soft top agar. Using platinum wire, Xbio phages (log/ml) were deposited at right angles to similar st,reaks of the XcImPm deletion phages. After overnight incubation at 37” the points of streak intersection were impaled with a Pasteur pipet and the agar plug resuspended in 0.5 ml 480 buffer supplemented with 0.1 ml MgCa solution. adsorbed to E. co/i NlOO (recA su’) indicator, which permits plaque formation only for XFec+P+ recombinants (Zissler et al., 1971), and plated in soft top agar. The number and the ratio of turbid to clear plaques are listed in Table 2. The samples MATERIALS AND METHODS collected from the intersection between Strains and media. The strains of bacte- two streaks of the same phage servred as controls. ria, lysogens, and bacteriophages employed Construction of QR48(XindmP80)/h lysoin this study are listed in Table 1. They were propagated in tryptone broth (0.8’;: gen. Phage hind-P80 (Table 1) was plated Tryptone (Difco). O.iiCr NaCl, 0.01 M on E. coli QR48 and the putative lysogens from the center of well-isolated turbid MgCl,). Plating media contained tryptone plaques were cross-streaked with phages X broth supplemented with 1.15% (bottom) or 0.655 (top) agar, respectively. The $80 and Ximm434. Bacteria immune to X were and plated on tryptone agar buffer solution contained 0.1 M NaCl and purified 0.01 M Tris, pH 7.9. The MgCa solution seeded with lo9 Xb2imm434c. X-resistant survivor colonies were purified and crosscontained 0.01 M CaCl, and 0.01 M MgCl,. Plating and propagation of bacteria and streaked with A, Ximm434, 480immX and phages. For plaque assay of phages. 0.1 ml 480Ximm434. Bacteria resistant to all but of stationary-phase bacteria was mixed @OXimm434 were further purified and tested for the multiplicity of prophage with 0.1 ml of the appropriate phage dilution in d80 buffer and 0.1 ml of MgCa copies by the 480 “Ter” test described by solution, incubated for 10 min at 37”, Nijkamp et al. (1971). A lysogen which mixed with 3 ml of top agar. and poured carried a single prophage copy was seover solidified bottom agar (40 ml of bot- lected. “Clear” test for single and multiple protom agar per loo-mm plate). Phage stocks were propagated from a phage copies. A test far simpler than the single plaque. by impaling it with a Pas- 480 “Ter” test of Dove (Nijkamp et al., teur pipet, resuspending the agar plug in 1 1971) gave reliable results in the present Isolated colonies of ml 480 buffer. and preadsorbing lo6 PFU system. the (about 0.1 ml) to an appropriate bacterial QR48(XindmP80)lX lysogen were inocuhost, as in the plaque assay. After 10 min lated into 200 ml of tryptone broth and example, the defective Xbio phages. which lack one or both (p, and p,{) promoters (Kayajanian. 1968), their nondefective derivatives (Court and Sate. 1969). and phages with heterologous immunity regions and new promoters, including the hybrids Ximm434. AimmA ,434 and Ximm21 (Kaiser and Jacob, 1957; Wilgus et al., 1973; Liedke-Kulke and Kaiser, 1967). To isolate plaque-forming mutants with internal deletions in the immunity region we selected for mutations which would inactivate the repressor gene, causing release of phage from the lysogen, and at the same time would render these phage particles resistant to treatment with EDTA. The EDTA technique has yielded many deletion mutants in the central part of the lambda genome, but never in the immunity region (Parkinson and Huskey. 1971). Preliminary results of these studies were already summarized by Blattner et al. (1974b) and cited by Blattner (1973).

DELETIONS

AND INSERTIONS

LYSOGENS,

AND BACTERIOPHAGES

TABLE BACTERIA,

Strains

Relevant

Bacteria (E. coli) TC600

Phages Xh2c XhZimm434c @80inmX (hy5) $~BOhinn434 (hy5imm434 hy80i434) Xind-P80 hbioN2-lnin5 hbio3h-lnin5 hbio24-5nin5 XbiolG-3nin5 Xbiot75nin5 Xbio30-7nin5 Xbiotl24nin (ht124-10) XA2

recA

EMPLOYED

Source (reference)

C. R. Fuerst; derived from C600, Appleyard (1954) Campbell (1965) Signer and Weil (1968) Strain 152 of M. Meselson; Gottesman and Yarmolinsky (1968)

suI1

This stud)

or

Aint-N Aint-s,, Aint-rex Aint-rex Aint-rex Aint-c1 Aint-c1 IS2 insertion y region

461

IS THIS STUD)

genotype

WC+ su” SM’ recA suI1 recA- suO

Lysogens QR48 (Xind-P80)iX

X

1

ret+ SUII

594 QR48 NlOO

IN PHAGE

in the

grown to saturation. After spinning down the bacteria, the supernatant was plated on the TC600 indicator bacteria. The titer was generally lo3 phage per ml. With lysogens which carried only single prophage copies, about half of the plaques were clear, whereas lysogens with multiple prophage copies yielded hardly any clear plaques. Presumably, this is because any c1 clear mutant arising in the multiple lysogen would be repressed by the ~1’ product of the other nonmutant prophage(s). It was ascertained in several independent experiments that the results of the 480 “Ter” test (Nijkamp et al., 1971) are in complete agreement with the “clear” test for single and multiple prophage copies described here. Buoyant densities of phages. The phages were banded in a CsCl density gradient at 25” in the Spinco model E analytical

Kaiser and Jacob (1957); Westmoreland et al.. (1969) Signer (1964); see Fiandt et al. (1971) H. Echols; Wu et al. (1971) see Hayes and Szyhalski (197.3) This study F. R. B.; see Blattner et al. (1972) G. Kagajanian; see Blattner et al. (1972) F. R. B.; see Blattner et al. (1972) F. R. B.; see Hayes and Szyhalski (1973) W. S.; see Hayes and Szyhalski (1979) D. Court; Court and Sato (1969) C. R. Fuerst; Guhaet al. (1971); see Hayes and Szybalski (1973) Brachet et al. (1970); Fiandt et al. (1972)

ultracentrifuge. Phages Xpapa (1.508 g/ cm3) and Xb2 (1.491 g/cm”) served as density markers. and the densities were calculated as described by Szybalski and Szybalski (1971). Electron micrographic heteroduplex mapping. Heteroduplexes were prepared by mixing the 1 and r DNA strands, separated by CsCl-poly(U,G) isopycnic centrifugation, in bicarbonate-buffered 50% formamide, as originally described by West moreland et al. (1969), and annealing for 2 hr at 25”. Alternatively, the DNA was released and denatured by heating in 1005 formamide (42’ , 2 hr) the CsCl gradientpurified phages, and then annealed as described above after adding an equal volume of bicarbonate buffer. Spreading in a cytochrome c matrix, grid preparations. uranium oxide shadow casting, electron microscopy, length measurements and

cl-rex

1.5040 1.5024

KH112 KH115

,..

1.5040

KH103

3.0 4.3

2.8 3.0

3.0 2.9

4.3

(-;“^I

74.27

74.;:< 74.68

7.5.23

74.07 74.73

Left

78.45

77.62 77.74

78.05

78.4'b 77.660

Right

Termini

1.18

2.89 3.06

2.82

4.:15 2.9:1

(&

Electron micrographic mappinga

1

150 4?J

9 19

73

58

tu

250 321

125 118

337

10

312

cl

N2-1

TABLE

2

.600 .104

,072 ,161

.217

.lOO

.I86

tu/c1

DELETIONS

3 1x

I" 9

2’1 -

I

:34

tu

120 96 53 238

30

216 242

cl

:3hm1

.057 ,076

.lOO .094

.o:n

.157 .091

tu/cl

1 0

6 3

4 2

0

tu

75 > 200

1253 265

120

521 76

cl

24-5

.oo.i ,011 ,013 0

0’

,008 ,026

tu/c1

0 0

0

0

0 0 21

tu

0 0 0

>200 >200

0

0' 0 ,067

tuicl

>2OO

350

> 50 >5O 367

cl

16-3

Scoring of turbid and clear plaques in crosses with Xbm phages

OF THE cI-rex

0 0

0

0

0 0 0

tu

', 200 > 200

>200

>800 > 800

2 50

> <50

t 12-l

-

.-

._

.I__

_-

_-

__

-

--

._--

-

--

.I-

.--

-

-

-I

0

0 0

0

0

0

0'

--~ ~___ cl tu/cl

t75: 30-Y:

~_ 0 The left termini of the KH deletions are the average of measurements versus the bio endpoints of X2-l (7X.0 “;X). 3h-1 (X,.5 ?hI. and 24-5 (74.5 ‘YXI. as specified in Tables 3 and 4. The right termini are the average of measurements versus the nin.i deletion loop (83. -( “;X), as specified in Table 4 and Fig. 1. The standard deviations and standard errors of the mean are specified in Table 4. Symbol A indicates a deletion. ’ Measured versus the r32 insertion loop (at 79.28 ‘7 XI. ah specified in Table 4 (see Fig. 1). Figure 0 represents a tu/cl ratio less than .OOl or .Ol.

1.5044

KH86

~g/cnP)

Density

1.50’24 1.5041 1.5042

-

C&l gradient centrifugation

KH54 KH67 KH70

deletion phage

MAPPING

&

2

i

5

F

DELETIONS

AND

INSERTIONS

their evaluation were according to Westmoreland et al. (1969) with modifications by Fiandt et al. (1971).

IN PHAGE

X

463

phages. whose endpoints were positioned by electron microscopy (Fiandt and Szybalski, 1973: Blattner et a/., 1972 and unpublished data), was emploved to map RESULTS the left endpoints of the putative c1 deleIsolation of the cl deletions. A tions. The crosses were performed by crossrecA -( Xind -P )/X lysogen (see Materials streaking on lawns of E. co/i TC600, as and Methods), which contained a single described in Materials and Methods, and copy of the prophage, served as the source the turbid (tu) and clear (cl) recombinants of deletion mutants. The roles of the vari(see Fig. 2) were scored on E. co/i NIOO ous mutations were as follows. Mutations (recA-su”), which plates only the Fec+P’ recA- and ind- were to depress the fre- phage (Zissler et al. (1971)). In crosses with quency of spontaneous induction, resist- Xbiot75 seven of the 116 putative c1 deleance to h (/A) was to prevent readsorption tion mutants did not produce any turbid of the released phages, and the P mutarecombinants. Since Xbiot75 contains the tion served as a genetic marker in the entire c1 gene and a part of the rex gene to subsequent mapping of the deletion. From the left of it (Gussin et a/., 1973). the the culture fluid of this lysogen we isolated absence of turbid recombinants indicated many independent h clones with the that these seven mutants had deletions “clear” phenotype which survived a lowhich cut into genes c1 and rex and overlap min treatment with 0.01 M EDTA at 37”. with the biot’75 endpoint. In the present For each isolation, a separate colony of study only these seven deletion mutants QR48(Xind-P80)/X was inoculated into 1 were examined further, since well-characliter of tryptone broth and grown in a terized deletions removing most of the left shaker overnight at 37”. The volume of 1 part of the immunity region were of greatliter was found to be necessary. because est interest for precise mapping of the p,,o,, region. 250-ml cultures frequently failed to yield The results of the crosses between Xbio any clear candidates for the desired delephages and the rex-c1 deletion mutants are tion mutants. apparently for statistical reasons. A 30-ml sample was clarified by shown in Table 2. Failure to observe any low-speed centrifugation and the supernaturbid recombinants indicates that the c1 tant spun at 30.000 rpm for 1.5 hr in the deletion (Arex-c1) overlaps the bio endSpinco rotor 30. The pellet containing the point (Fig. 2, cross B). The left endpoints spontaneously released phage was resus- of the cI-rex deletions appear to fall into pended in 0.5 ml of 0.01 M EDTA (pH 8). one of three intervals: (1) between the bioincubated at 37” for 10 min to enrich for 3h-1 and 24-5 endpoints (KH54, KH115). deletion mutants, and then plated on a (2) between the bio24-5 and 16-3 endpoints QR48 lawn, as described in Materials and (KH67. KH86, KH103, KH112). or (3) between the bio16-3 and t75 endpoints Methods. The resulting plaques were virtually all clear and numbered from 1 to 200 (KH70). The tu/cl ratio is a measure of the per 30 ml of culture. One plaque was distance between the bio and left Arex-cl endpoints in relation to the variable interisolated. purified on the TC600 indicator bacteria, and stored for further study. The vals between the right endpoint of’ the entire procedure was repeated 120 times, in deletion and the P80 mutation. We shall defer interpretation of the tu/cl ratios to order to obtain 116 independently derived the Discussion section. clear mutants. In four cases no c1 deletion Huo>,ant densities of the phages. The candidate was found. This suggests that such prophage mutations occur at a fre- buoyant densities of several clear mutants quency of about lOV”~lOV”‘, assuming a are listed in Table 2. The observed deburst size of 50. to 100-phage particles per creases in density are consistent with the genetic evidence that these are deletion lysogenic bacterium. mutants. The approximate extent of deleGenetic mapping of the left deletion tion (A %X, Table 2) was calculated by endpoint. A series of six bio transducing

464

HLATTNER

FIG. 2. Genetic mapping of the KH deletions. In cross I, both turbid (tu) and clear (cl) recombinants are formed since the bio substitution and KH deletion do not overlap. If they do overlap (cross II) only clear (cl) recombinants are produced. Since the XbioIl is cut in gene ret and thus contains the entire c1’ gene. the latter result strongly indicates that the clear phenotype of’ the xKH mutant is due to a deletion (see text). The Fee phenotype (no plaque formation on recA host) of Ahlo mutants is caused by the removal of x genes red and gum (Zissler et nl., 1971). The frequencies and ratios of tu to cl recombinant5 are listed in Table 2.

linear interpolation, taking the DNA lengths in Xpapa (X) and Xb2 as 100 %:IX and 87 %A, respectively (Westmoreland et al., 1969). Electron micrographic mapping of the KH deletions. Heteroduplexes were constructed between DNA strands of the rexc1 deletion strains and several Xbio-nin5 phages or Xr32, as described in Materials and Methods. The positions of the left and right termini of the rex-cI deletion were measured as duplex DNA in relation to the bio endpoints and the nin5 or r32 loops. respectively. From our previous measurements (Fiandt et al., 1971; Blattner et a/., 1972; Fiandt and Szybalski. 1973) and present data the locations of these markers are known (see Fig. 1). The positions of the endpoints of seven rex-c1 deletions are shown in Table 2. Judging from the positions of their endpoints, KH54 and KH115 are very similar if not identical. Since in heteroduplexes between XKH54nin5 and XKH115, where the nin5 loop served as an indicator of heteroduplex formation, only perfect duplex DNA was observed in the immunity region, we conclude that the KH54 and KH115 deletions are identical or differ by less than 30 nucleotide pairs at either end. Likewise, it appears that KH67. KH86, KH103, and K112 may form a second group of identical deletions (Table 2) although no heteroduplexes between

E7’ AL

them vvere examined. The deletion KH70 is unique in the present collection. Three representat i1.e KH deletions are shown in Fig. 1. ReeL~nluntion of the distance betu,cen the promoter (p,.) and startpoint (s,.) for the leftward mRi’VA. The distance from s,, to the promoter mutation sex1 constitutes a minimum estimate of the size of the pro moter-operator region for the major leftward h operon (Blattner et al., 1972). The original determination of this distance combined sequencing, electron microscopy and recombination techniques. Each of the components of that measurement has been the subject of continued investigation and refinement. The sequence of the major leftward RNA has now been completed, and the distance from s,. to the dvl endpoint is exactly 116 nucleotide pairs (Dahlberg and Blattner 1974). The linearity of the relationship between recombination frequency and physical map distance in the h immunity region has been verified using markers with established positions (Blattner et al., 1974a). In this section we discuss the refined electron micrographic measurements made possible by the availability of the KH deletions, and reevaluate the s,. to sex1 distance using the original method and other approaches. The earlier electron micrographic measurements of the positions of the bioN2-1 and bio3h-1 endpoints employed the r32 insertion loop as marker. whereas the dvl endpoint was measured relative to the imm434 substitution, and the two sets of measurements were tied together by the measurement of the bioN2-1 to imm434 distance (Blattner et a/., 1972). The present measurements employ the KH deletions, which are distinctly more suitable markers since they are located up to 10 times closer to the bio3h-1 endpoint than was the r32 insertion. Moreover, all three of the key positions. namely, N2-1, dvl. and 3h-1 were measured versus the same KH markers (Table 3. Fig. 3). To position the scxl mutation, a graphical approach shown in Fig. 3 was adopted. In this figure the KH54 deletion serves as the fixed reference point for the bioN2.1, dvl and bio3h-1 endpoints. Mathemati-

DELETIONS

AND

INSERTIONS TABLE

POSITIONS OF THE bio AND dvl

3

Physical distance (‘;A)” NZ-1 to KH

KH67

465

X

ENDPOINTS IN RELATION TO THE LEFT END OF THE KH DELETIONS

cI-rex deletion KH54

IN PHAGE

1.07

(16)

.14 SD .04 SEM

1.73 (13)

.I1 SD .03 SEM

dvl to KH 231 (14)

.05 SD .Ol SEM

3h-1 to KH .57 (12)

.09 SD .03 SEM

1.23 (14)

.I3 SD .03 SEM

” The figures represent the distances measured as %A length (see Fig. I). the standard deviation (SD), and the standard error of the mean (SEM), the latter also shown by error flags in Fig. 3. The figures in parentheses represent the number of heteroduplex molecules measured. The circular DNA of phage PM2 (relaxed form) served as the standard with its length corresponding to 20.44’;X.

tally, the graphical extrapolation is equivalent to the algebraic method employed previously (Blattner et al., 1972). This analysis is based on the sex+/sexI (tu/cl) ratios obtained in crosses between XsexlsusP3 and XbioN2-1 or hbio3h-1 (data from Fig. 2 of Blattner et al., 1972). Linear extrapolation of these data indicates that the position of sex1 relative to KH54 is 218 =t 18 (SEM) nucleotide pairs, i.e., 0.47 + 0.04 %X (the composite standard error of the mean, SEM, was estimated graphically as shown by the dotted lines in Fig. 3 drawn through the error flags for the N2-1 and 3h-1 measurements). To position the sl, startpoint in relation to the KH54 endpoint, one must subtract the distance of exactly 116 nucleotides (i.e., distance from the 5’ end ofp,, RNA ( = s,,) to the dvl endpoint, as based on the sequencing data of Dahlberg and Blattner, 1974; see Fig. 3) from the distance of 377 * 6 (SEM) nucleotide pairs (i.e., distance from dvl to KH54 equal to 0.81 %X, SD = 0.05, SEM = 0.01; see Fig. 3). Thus, by this subtraction, the startpoint is 261 * 6 (SEM) nucleotide pairs from KH54 endpoint. Since both sex1 and s,. have been located relative to the same coordinate frame defined by KH54, the distance between them can be obtained by subtraction, and the error flag computed by rms (root mean square) addition of the component SEM (standard error of the mean) values. Thus, we conclude that the s,. to sex1 distance is 43 * 19 (SEM) nucleotide pairs, i.e., 24-62 nucleotide pairs.

A second method to estimate this distance is provided by the electron micrographic analysis of the fragment obtaihed by digesting Xdv DNA with the HinIl nuclease. As shown by Allet and Solem (1974) as well as by Maurer et al. (1974) this nuclease cuts within six nucleotide pairs of the sex1 mutation site. In a heteroduplex between the novel joint-containing hdvl fragment and X DNA the distance from the dvl endpoint (novel joint) to the HinII cleavage site is 0.31 %h (SEM-0.03) (see Fig. 3), i.e., 144 * 14 nucleotide pairs. Subtracting the 116 nucleotide pair s,, to dvl distance (Fig. 3) we arrive at the figure of 28 + 14 nucleotide pairs for the distance from s,, to the HinII cut in o,,. The measure of the s,. to sex1 distance includes an additional uncertainty of six nucleotide pairs (28 + 17). Thus, we conclude that the nontranscribed region between the promoter mutation sex1 and the mRNA startpoint s,, is shorter than originally estimated by Blattner et al. (1972). This difference is principally due to the earlier overestimation of the distance between the bioN2-1 and bio3h-1 endpoints, when employing the more distant r32 marker. Order of sites in the p,,o,, region. As shown in Fig. 3, as far as the electron micrographic measurements relative to the KH54 deletion are concerned, the positions of the left end of imm434 and of the bio3h-1 endpoint are indistinguishable from each other and from the position of the sI. startpoint (Fig. 3). However, the sequence data derived from the X and Ximm434 templates show that s,. must be to the left

BLATTNER

ET AL

of’ the imm-434 region. Moreover. there are two lines of’ evidence that the bio3h- 1 endpoint must be inside the imm43-l region or at least extremely close to its left end: ti) no recombinants were found in Ximm434 Y Xbio3h-1 crosses (Kayajanian, 1966; Blattner et al.. 19i’:! and unpublished data). and (ii) no duplex DNA was observed vvithin the bio-imm nonhomology (“bubble”) in 70 heteroduplexes between DNA strands of the Ximm434 and Xbio3h-limmhnin5 phages (electron microscopic data not shown). Thus. the 3h-1 endpoint is vvithin (or not more than 10 nucleotide pairs to the left of) the imm434 region. Concerning the positions of promoter and operator mutants;. Mr. Shiu-Lok Hu and Mrs. J. K. Boettiger in our laborator! FIG. 3. Mapping of the p,. promoter mutant sexl. conducted pairnise four-factor Two Xblon1n.5 phages. N2-I and 3h-I. were each have crosses which indicate that the mutants ~2, crossed with the CZ’:~lO2~XsexlcIH;i~su.sP:l) Ivsogen. sexl. and sex3 map in that order from left The ratios of the turbid (sex’) to clear (sex11 recombinants ttu/cl 409/2.669 0. l,i:< for IL’?- 1. and 610/ to right. and all of these are to the right of 23,427 = 0.026 for 3h-11. as determined by Blattner bio3h-1. This is consistent with the posiet al. (1972) in their Fig. 2, were plotted as a function tion (14 nucleotide pairs upstream from the of the physical positions of the bio endpoints. The s,. start point) assigned for the c2 operator measured distances from KH54 to the bio3h-1, dvl. mutant by Dahlberg and Blattner (1974) in and bioN2-1 endpoints are given in Table 3. Other the sequence of Maniatis et a/. (1974) and measurements, their standard deviations (SD) and with the observation that both sex1 and standard errors of the mean (SEM) are as follows: sex3 mutations eliminate the HinII cleavdvl to imm434: 0.25 TX 0.05 SD, 0.02 SEM (Xdvll age in the p,.-o, region (Allet and Solem. Ximm434 heteroduplex. seven measurements), bio1974). It will be interesting to see whether N2-1 to imm434: 0.50 GA 0.06 SD, 0.01 SEM (XbioN2the genetically assigned order of’ sex mutalnin5/Ximm434 heteroduplex. 28 measurements). dvl to Hid1 q cut: 0.31 RX 0.08 SD, 0.03 SEM (heterotions agrees with future detailed sequencduplex between X and a Xdvl fragment generated by ing data. the Hin restriction enzyme and containing the o,. In summary, the order of’ the six ele(a,) site and the novel Xdv joint, six measurements). ments appears to be s,, imm333, bio3h-1, These distances are drawn below the plot. and the ~2. and sexl. sex3. from left to right on the error flags represent the standard error of the mean, X map. which takes into consideration the number of heteroInsertion in the immunity region. During duplexes measured. According to these measurescreening for clear phage phenotypes (see ments the positions of the bio3h-1 and imm434 end2) a phage mutant KHlOO was points are not distinguishable; they are placed at 73.5 Table %X units from the left terminus of h DNA. Similarly, isolated. the density of which was higher the distance of 116 nucleotides (NUCL.) from the dvl (1.5112 g/cm”) than that of’ Xpapa. correendpoint to the a,. startpoint, obtained from sequence sponding to about a 2.4 %X insertion. This data (Blattner et al., 1972; Dahlberg and Blattner. was very surprising. since the EDTA tech-

1974). place the s, RNA startpoint in the same position. The intercept of the plot at 73.6 RX determines the position of the sex1 mutation. The intercepts of the dotted lines indicate the range for this intercept defined by the standard errors of the mean for the bio measurements. The solid error flag above represents the 90’7 confidence limits on the intercept considering the plaque counts only. The dotted error flag was calculated to include the intercepts of the dotted lines

and the solid error flag. The inverted triangle identifies the position of the HinII cleavage site: the sex1 mutation which abolishes this cleavage presumably is within three nucleotide pairs from this site (Allet and Solem. 1974; Maniatis et al., 1974). The map of the p,-proximal region of X and of the strong tsi) and weak ts2-sgl repressor binding sites (Maniatis and Ptashne. 1973) is drawn at the top of the figure.

DELETIONS

AND

INSERTIONS

nique should have discriminated against any insertions. We have examined this phage by electron micrographic mapping (Table 4) and found that it contains an insertion corresponding to about 2.9 %h (1350 nucleotide pairs) located at 77.2 %;1x from the left X DNA terminus (Fig. 1). We are presently determining whether this insertion might be identical to the IS2 insertosome represented by the r32 and bl insertions in X. or some other insertion sequences, e.g., IS3 or IS4 (see Fiandt et al., 1972). No homology was found in l/r heteroduplexes between the insertions XKHlOO and Xr32. indicating that KHlOO is not an IS2 insertion with the same orientation as in Xr32 (Fiandt et al., 1972). Position of the KH deletions and insertion in relation to restriction sites. It was shown that restriction enzymes produced by Hemophilus species and by RTF factors of E. co/i cleave mature X DNA at specific sites (see Allet et al., 1973; Allet. 1973; TABLE

IN PHAGE

Maniatis and Ptashne, 1973). Using our preparation ofphage DNA, Dr. B. Allet has determined which of these sites are located within the KH54 and KH67 deletions. As shown in Fig. 1, the cleavage site at about 75.2 %A, produced by the Hemophilus parainfluenzae (HpaII) enzyme, and the cuts at about 76.3, 77.4, 77.5, and 77.7 (ind~)W, produced bv the Hemophilus influenzae strain Rd (H&II + III) enzymes. must be within the KH54 deletion, since the HpuII fragments 570 and 2150 n.p. and the HinII + III fragments 1200, 520, <50, 100 and 375 n.p., are not present in the XKH54 DNA digests and are replaced by the 650 and 280 n.p. fusion fragments, respectively (Fig. 1). Deletion KH67 does not remove the ind- cut at 77.7%. Similarly, several cleavages produced by the HaeII and III enzymes are located within the KH deletions (Fig. 1). The detailed data on the positions of the restriction enzyme cleavage sites were presented else4

DISTANCES BE.I.WEEN THE TERMINI OF THE KH DELETIONS AND THE bio ENDPOINTS INSEHTION AND nin.5 DELETION (TO THE RIGHT)”

cl-rex deletion

467

X

Physical distance (SD: SEMI”

(TO THE LEFT) AND THE r32

(c(x)

N2-1 to KH-left

3h-1 to KH-left

24-5 to KH-left

KH-right

to r32

KH-right

to rzin5

0.86 (0.07: ,021 KH54

5.60 C.44; .12) 5.33 (.40; .lO)

1.07 f.14; .04) .57 t.09: .03) 1.62 C.15: .04)

KH67 1.73 C.11; .03)

5.85 (24; .09) 5.97 (57: .I61

1.23 c.13: .03) .27 C.06: .03) KH70

KH86

1.73 (.lO: ,041

.74 (.lO; .03)

6.67 (33; ,141 5.63 (30; ,081

.32 C.08; .02)

6.04 (.“O; .05) 6.12 (34; .1X)

1.6.5 t.14; .04)

.03)

5.96 (.1,5; .05)

KH115

.77 c.09: .O”)

5.25 t.23: ,061

KHlOO

3.78 (.24: .06)

KHlO4

1.18 (.ll:

2.05 (23: .06)

6.56 (36: .lO)

~For experimental details see Materials and Methods, Fig. 1, and the legend to Table 3. No homology region between the bio endpoints and the KH deletion was found in heteroduplex XKH70/Xbiot75nin5. The immunity region appeared as a homoduplex in the XKHl15/XKH.ilnin.5 heteroduplex. h Standard deviation (SD) and standard error of the mean (SEM) are specified in parentheses. in that order.

468

RLATTNER

where (Allet and Solem. 197-1). but the present results include several revised positions for these sites in and near the immunity region, as based on our electron micrographic measurements and on the unpublished data of L. Robinson and of B. Allet. The Hin cleavage at 77.7%X, described by Allet and Solem (1974) to occur in xc1857 and attributed to the Hi&II nuclease by L. Robinson (personal communication), is probably caused by the indsince this extra cut (dotted mutation, arrow on Fig. 1) is observed in “ind-” XKH67in&, which does not carry the ~I857 mutation, and in Xind-cI857 DNA (B. Allet. personal communication). In addition, a cleavage at a very close or the same site was found in XcIts2ind’ DNA after digestion with the HinII+III mixture (T. Maniatis, personal communication). We feel that the most likely explanation is that the ind and cIts2 mutations each create an extra Hin-specific sequence and that the two mutations map very close to each other. This appears to be consistent with the cIts2 mapping data of Blattner et al. (1974a) and F. R. Blattner, T. M. Shinnick and J. D. Bore1 (manuscript in preparation) but not with that of Lieb (1966). Obviously, the cleavages cannot be both HinIII specific and at the same position, if they are caused by the ind- and cIts2 mutations, since the mutants have different phenotypes. Indeed XcIts2 is not only devoid of the Ind- phenotype but in fact is superinducible (Lieb, 1966). which could reflect different alterations affecting the inducer-binding portion of the repressor molecule. It is also possible that this “ind-” cleavage is a result of nonisogenicity or point mutation unrelated to the Ind or ts2 phenotype of the c1 mutations. We examined the digestion pattern of XKHlOO with two of the restriction nucleases and found that this insertion contains near its right terminus a cleavage site for E. coli EcoRl restriction endonuclease. Digestion with the HinIII nuclease showed no cleavage within the KHlOO insertion and confirmed the location of KHlOO within the .i’tO nucleotide-pair fragment (Fig. 1).

ET AL

The selective method employed in this study led to the isolation of more than 100 phage X mutants, many of which carry an internal deletion in the immunity region, including part of gene ~1. A unique feature of the method is that the deletion event is lethal for the cell in which it occurred, so that all mutants isolated from separate single colonies must be completely independent. Seven of these mutants. all of which delete at least a part of gene rex and do not recombine with Xbiot75, were analvzed in more detail (Table 2). Only three distinguishable deletions were found (Fig. 1). It appears significant that among as few as seven independent genetic events. one kind of deletion occurred twice and another four times. This may suggest some kind of sequence specificity in the illegitimate recombination mechanism leading to these deletions. There is a good agreement between the physical and genetic data on the mapping of the KH deletions. However, when calculating the positions of the left end of the KH deletions from the tu/cl ratio obtained in crosses with Xbio phages (Table 2), one must take into consideration that the deletion reduces the opportunity for recombination in the bio to Psus80 interval. Such a correction was unnecessary in the case of analogous mapping of the c1 point mutations (Blattner et al., 1972a). Many other c1 deletions that are to the right of the biot75 endpoint are in the process of being analyzed. Moreover. internal deletions which remove the p,, promoter are being isolated, starting with the N-independent Xnin5chi lysogen and an appropriate excision helper. It is surprising that the XKHlOO mutant, which has an insertion in the c1 gene, was isolated by this method. We have no explanation why this mutant was able to pass through the EDTA screening procedure. unless such c1 insertions are rather frequent. This insertion might be very useful in studies on the immunity transcription, and its cleavage site for the E. coli EcoRl restrict ion endonuclease will render it helpful in constructing x phages incor-

DELETIONSAND INSERTIONSIN PHAGEh porating segments of eukaryotic and prokaryotic DNA. The properties of XKH54 indicate that no essential p,.ol. or pKo H elements are located within this deletion, since the phage grows normally and is not virulent. Thus, the notion of specific large antennae upstream from the promoter and operator is rather implausible (Blattner et al., 1972). Also, any genes (~1, rex) or other elements (lit RNA; Hayes and Szybalski. 1973) located within the KH54 deletion do not appear to be required for lytic phage growth. The KH54 deletion can be used to place upper limits on the sizes of the left and right promoter-operator regions. Since the sex1 to KH54 distance is 218 * 18 nucleotide pairs, the upper limit on the left is rather large, certainly sufficient to include all of the multiple repressor binding sites described by Maniatis and Ptashne (1973). The experiment showing that the o,~operator region is cut by the HinII enzyme (Maniatis and Ptashne, 1973) and that RNA polymerase protects against this cut (Allet and Solem, 1974) permits the placement also of the rightward P,~o,~ site in relation to the KH54 deletion. This HinII cut is 78.5 %;Ix, i.e., at about 0.1 %X from the right terminus of KH54 (Fig. 1). Thus, the right terminus of the KH54 deletion comes within about 50 nucleotide pairs of the HinII cut in ~~~0,~region, without perceptibly affecting its functions, although it might well delete a part of the repressor binding sites which according to Maniatis and Ptashne (1973) extend for about 75 nucleotide pairs to the left of this cut in oH . It would be interesting to assay the repressor binding capability of XKH54 DNA and the behavior of the u2 and u3 mutants of XKH54. The KH deletions were also useful as markers to permit more precise physical measurement of the minimal size of the promoter-operator regions. Using the KH54 and KH67 “loops” as markers, we found that the distance between the p,. promoter mutation sex1 and the s,, startpoint for the major leftward RNA is between 24 and 62 nucleotide pairs long. This distance was overestimated in our earlier

469

measurements. The present result is in agreement with the estimate of Maniatis and Ptashne (1973). Moreover, the current DNA and RNA sequencing data of Maniatis et al. (1974), which place the 5’.pppA s,. startpoint (Dahlberg and Blattner, 1974) 30-36 base pairs from the site of the sex1 mutation, and currently published results of Maniatis et al. (1973) and Maurer et,al. (1974) are within the limits of the present measurements (see Fig. 3). The basic findings of Blattner and Dahlberg (1972). (1) that the promoter region of X comprises two physically separated sites, p,. and s,., (2) that p,, could be mutated or substituted by a heterologous p,, from phage 434 without affecting s,,, and (3) that p,, is not transcribed into RNA, are not affected by our more accurate measurements. However, the reduced minimum distance from sex1 to s,, necessitates reconsideration of our earlier proposal that RNA polymerase drifts from DNA entry site to start site. The elimination of the drift concept would require that the RNA polymerase recognize an entire sequence at least as large as the distance between sex1 and s,,, or possibly two short sequences separated by such a distance, without unwinding more than five’ base pairs of DNA (Saucier and W’ang, 1972). Recent unpublished data on the protection of DNA by tightly bound RNA polymerase, presumably at the start site, indicate that the entry site defined by promoter mutations is outside of the region protected by the RNA polymerase, both in the case of the promoter-operator region of the lac operon (W. Gilbert, D. Pribnow, and J. Gralla, personal communication: W. Barnes, W. S. Reznikoff’, R. Dickson, and J. Abelson, unpublished) and for the p,.o,, region of x (R. Maurer, personal communication). Moreover, the promoter DNA fragment from phage fd protected by RNA polymerase retains the startpoint and can act as template for transcription, but cannot rebind the RNA polymerase (B. Heyden and H. Schaller, personal communip More recent data of J. C. Wang (personal communication) indicate that up to ten base pairs might be unwound per RNA polymerase molecule hound.

470

BLATTNER

cation), as if’ it had lost its entry site. Furthermore. there appears to be a region of’ drift between the oop and lit transcripts in h DNA (Hayes and Szybalski. 1973). We thus believe that the notion of’ polymerase drift must still be considered as a reasonable possibility. ACKNOWLEDGMENTS We acknowledge the fruitful collaboration with Dr. Bernard Allet which led to the correlation between his restriction site map and our deletion maps. Dr. Linda Robinson and Dr. Arthur Landy were very kind to communicate to us their unpuhlished results (manuscript in preparation) on the HmIII cut in the XcIind,857 mutant and the HinII cut at 77.4 %#A, which they found to generate the “<50 n.p.” fragment. Dr. W. Gilbert, Dr. W. S. Reznikoff. Dr. W. Barnes, Dr. M. Ptashne, Dr. T. Maniatis, Dr. R. Maurer, and Dr. H. Schaller have kindly permitted us to cite their unpublished data on the RNA polymerase binding. We also thank Dr. Barry Egan for reminding us of F. Jacob’s trick of looking for clear mutants in the culture fluid of lysogenic bacteria, and Dr. Elizabeth Szyhalski for her unpublished data on several electron micrographic measurements and for editing the manuscript These studies were supported by a grant from the National Cancer Institute (CA-07175). REFERENCES ALLET, B. (1973). Fragments produced by cleavage of X deoxyribonucleic acid with the Hemophilus parainfluenzae restriction enzyme HpaII. BiochemistT 12, 3972-3977. ALLET, B., JEPPESEN, P. G. N.. KATAGIRI, K. ,J.. and DELILX, H. (1974). Mapping the DNA fragments produced by cleavage of X DNA with endonuclease Rl. Nature (London) 241, 120-123. ALLET, B. and QOLEM, R. (1974). Separation and analysis of promoter sites in bacteriophage lambda DNA by specific endonucleases. J. Mol. Biol. 85, 475-484. APPLEYARD, R. (1954). Segregation of new lysogenic types during growth of a douhly lysogenic strain derived from Escherichia coli K12. Genetics 39, 440-452. BLATTNF,H, F. R. (1973). Some considerations regarding the structure and function of promoter-operator regions. In “Molecular Cytogenetics” (B. A. Hamkalo and d. Papaconstantinou, eds.). pp. 29%300. Plenum Press, New York. BLATTNEH, F. R.. BOKEI., .J. D., SHINNICK, T. M.. and SZYBAIZKI, W. (1974a). Mapping of point mutations on the physical map of coliphage lambda: absence of clustering for odd-numbered exchanges. In “Mechanisms in Recombination” (R. F. Grell. ed.1 Plenum Press. New York.

ET AL. HIATTSEK, F. R.. and DAIU.HEH(:. .J. E. I 197”1. RN.1 synthesis startpoints in bacteriophage X: Are the promoter and operator transcribed? .Yalurc’ :VfvI Hioi. 237, ‘27-“:3”. BLAT.IXEK, F. K.. DAIII.HF.HG, -1. E.. BOXI ~IGEK, -1. K.. FIANIV~. Xl.. and SZI.HALSKI, FV’. I 19721. Distance from a promoter mutation to an RNA synthesis startpoint on bacteriophage X DNA. Nature Neu BLATTNEK, F. R.. FIANI~I.. M.. HASS. K. K.. Twost;. P. A.. and SZ~RALSKI. W. (1974b). Deletions and insertions in the immunity region of coliphage lamhda: measurement of the promoter-startpoint distance. Abntr. Annu. Meet. Amer. Sot. Microbioi. p. 228. BHACHE.I, P.. EISE~. H., and RAMHACH. A. (1970). Mutat ions of coliphage X affecting the expression of replicative functions 0 and P. MI,/. (ien. Genet. 108, 266 ~276. CAMPRELI.. A. (1965). The steric effect in Iysogenization by bacteriophage lamhda I. Lysogenization of a partially diploid strain of Escherichia co/i K12. Virology 27, 329-X19. COURT, D.. and SATO, K. (1969). Studies of novel transducing variants of lambda: dispensability of gene N and Q. Vfrolog?, 39, 348-35’1. DAHI.BEHG, SKY, M. (1968). Integration-negative mutants of bacteriophage lambda. J. Mol. Bd. 31, 487~50.5. GVHA, A.. SATLMEX. Y.. and SZ~HALSKI. LV. 119711.

DELETIONS

AND

INSERTIONS

Divergent orientation ot’ transcription from the biotin locus. J. Mol. tliol. 56, 53-6?. GUSSI?;, G. N., PETERSON, V., and LOEH, N. t1973). Deletion mapping of the h ren gene. Genetics 74, 385-39”. HAYES. S.. and S~~EIALSKI.W. (19X). Control of short leftward transcripts from the immunity and ori regions in induced coliphage lamhda. Mol. Gen. Genet. 126, 255m’QO. I KAISEH. A. D.. and JACOB. F. (1957). Recombination hetween related temperate bacteriophages and the genetic control of immunity and prophage localizat ion. Viro/og>’ 3, 509-521. KAYAJANIAN. G. (1968). Studies on the genetics of hiotin-transducing, defective variants of hacteriophage X. Virology, 36, 30-41. LIEB, M. (1966). Studies on heat-inducible lambda bacteriophage I. Order of genetic sites and properties of mutant prophages. J. Mol. Hiol. 16, 149-163. LIEDKE-KULKE. M.. and KAISER, A. D. (1967). The c-region of coliphage 21. Virology 32, 475-481. MANIATIS, T., and PTASHNF., M. (1973). Structure of the X operators. Nature (London) 246, 133-136. MANIATIS. ‘I’., P,TASHNE, M., BAHHELL. B. G.. and DONELSON, J. (1974). Sequence of a repressor hinding site in the DNA of bacteriophage X. Nature (London) (in press). MANIATIS, I’., PTASHNF., M.. and MAUHEH. R. (1973). Control elements in the DNA of bacteriophage X. Co/d Spring Harbor Symp. Quant. Rio/. 38, 857-868. MAIXEH, H.. MASIATIS, T.. and PTASHNE, M. (1974). The promoters are in the operators in phage lamhda. Nature (London) 249, 221~223. NIJKAMP, H. J. .J.. SZYBAISKI, W.. OHASHI. M.. and DOVE, W. F. (1971). Gene expression by constitutive mutants of coliphage lambda. Mol. Gen. Genet. 114, 80-88. ORUAI.. G. W.. and KAISER. .4. D. (197%). Mutations in the right operator of bacteriophage lambda: evidence of operator-promoter interpenetration. J. Mol. Biol. 79, 709-722.

IN PHAGE

X

471

PARKINSON, J. S., and HUSKES, R. J. (1971). Deletion mutants of bacteriophage lambda. I. Isolation and initial characterization. J. Mol. Hiol. 56, 369-384. PTASHNE, M. (1971). Repressor and its action. In “The Bacteriophage Lamhda” (A. D. Hershey. ed.). pp. 221-237. Cold Spring Harbor Laboratory. Cold Spring Harbor, NY. SAI C‘IEH. J-M.. and WAX(.. J. (‘. (1972). Angular alteration of the DNA helix hy E. coli RNA polymerase. Nature New Bio/. 239, 1677170. SIGNER E. R. (1964). Recombination between coliphages X and $80. Virology 22, 650-651. SIGNER, E. R.. and WEIL J. (1968). Recombination in bacteriophage X I. Mutants deficient in general recombination. J. MU/. Hiol. 34, 261~271. SZYBALSKI, W.. and SZYBAUKA E. H. (1971). Equilihrium density gradient centrifugation. In “Procedures in Nucleic Acid Research”, Vol. 2 (G. L. Cantoni and D. R. Davies. eds.). pp. 311-354, Harper and Row, New York. WESTMOREI,AND, B. C., SZYBAISKI, W., and RIS, H. (1969). Mapping of deletions and substitutions in heteroduplex DNA molecules of bacteriophage lamhda hy electron microscopy. Science 163, 13431348. WILGU, G. S., MURAL, R. J.. FRIEDMAN, D. I., FIANDT, M., and SZYBALSKI, W. (1973). XimmX.434: a phage with hybrid immunity region. Virology 56, 46-58. WV, M. A., GHOSH, S., WILI.ARD, M., DAVISON, J., and ECHOLS, H. (1971). Negative regulation by lamhda: repression of lambda RNA synthesis in vitro and host enzyme synthesis in uiLw. In “The Bacteriophage Lambda” (A. D. Hershey. ed.), pp. 589-598. Cold Spring Harbor Laboratory, Cold Spring Harhor, NY. ZISSLEQ J., SICNE~ E.. and SCHAEFER F. (1971). The role of recombination in growth of bacteriophage lamhda. I. The gamma gene. In “The Bacteriophage Lamhda” (A. D. Hershey, ed.), pp. 4555468. Cold Spring Harbor Lahoratorg. Cold Spring Harbor, NY.