BIOSYNTHESIS OF /3-D-GALACTOSIDASE CONTROLLED BY PHAGE-
CARRIED GENES. III: DEREPRESSION OF /3-D-GALACTOSIDASE
SYNTHESIS FOLLOWING INDUCTION OF PHAGE DEVELOPMENT
             IN LYSOGENIC BACTERIA*

       BY H. R. REVEL, S. E. LURIA, AND N. L. YOUNG

                  MASSACEUSE'ITS INSTITUTE OF TFXHNOLOQY

               Communicated October 23, 1961
When E. coli bacteria carrying the prophage Xdg, which contains the gal loci
for galactose utilization, are treated with doses of ultraviolet (UV) light sufficient
to induce phage production in X-lysogenic bacteria, the gal genes become dere-
pressed. 1  That is, synthesis of galactokinase in the absence of an external inducer
increases 20- to 30-fold over the constitutive level.  A similar result has been re-
ported for galactose-l-phosphate uridyl transferase.2 Following UV induction,
even gal+ bacteria that carry a normal X prophage show a marked constitutive syn-
thesis of these enzymes,' as though the indu&ion of prophage X, which is attached
in the immediate vicinity of the chromosomal gal genes,8 caused derepression of
these'genes.
A similar derepression by Z+ genes in prophage Pldl was independently observed
two years ago in our laboratory.  In the course of the production of high-frequency-
transducing (= HFT) lysates by UV treatment of heterogenote E. coZi la&'
(PldZ i+ z+) it was noted that the crude lysates, which were prepared in media with-
o& external inducers, contained significant amounts of fl-D-galactosidase, while the
unirradiated bacteria had much leas.  A systematic study of this phenomenon
was undertaken following the discovery of the escape phenomenon described in the
preceding paper,' and the results are reported here.
Materials and methods were as described in a previous paper.5  UV treatment
waS done on bacteria collected from growing cultures and resuspended in saline.
After irradiation the bacteria were diluted in TGA medium.
Results.-Derepression jollo@ng UV irradiatim:  Figure 1 illustrates the produc-
tion of&D-galactosidase and`its release in extracellular form by a culture of E. coli
e'+z-g+ (Pldl i- z+) following UV treatment and Pl-superinfection.  Also shown is
the production of transducing activity assayed in samples from the same culture.
There is a close parallelism between release of transdticing phage and of galactosi-
dase; both are presumably freed by bacterial lysis.  A delay occurs between syn-
thesis and release of enzyme, so that the total amounts of enzyme are at first in ex-
cess over the cell-free enzyme. Ultimately,.`however, practically all the enzyme
formed is released as free enzyme.  This is true even wiih UV treatment alpne with-
out superinfection with Pl, and indicates that all bacteria in which enzyme is pro-
duced as a response to UV are ultimately lysed.


' VOL. 47, 1961



        100

0.1

BIOCHEMISTRY: REVEL, LURIA, AND YOUNG

Total Enzyme, UV

I     I       I            I

0

60            120

Minutes Alter lrrodiolion

180

1975

FIG. I.-Production of galactosidase and of transducing phage after LJV irrsdia-
tion of the heterogenote E. coli strain3.080-1341, i+z-(Pldl i-z+).  The amount of
enzyme in a nonirradiated culture is also shown.  No inducer was present.

Figure 2 shows the effect of varying UV doses on the amounts of enzyme and of
phage produced.  For any given strain the optimal dose is the same for both effects,
indicating that they result from a common mechanism.  With the strain used in the
experiments of Figure 2, the maximum amount of galactosidase formed corresponds
to what would be produced if about 5 to 10 per cent of the bacteria had been either
genetically constitutive or grown with an inducer.

  Table 1 provides a comparison of the amounts of enzyme synthesized constitu-
tively by a variety of strains after UV treatment. The significant findings are
  as follows:  (a) UV treatment provokes constitutive p-gala&s&se synthesis
.  only when the Z+ gene is in the Pldl prophage, not when it is in the bacterial chro-
mosome, even though a normal Pl prophage may be present (at variance with the
findings with the X system1s2) ; (b) the amount of UV-initiated enzyme synthesis
is different in different strains, being generally higher for strains that carry a Pldl
i-z+ prophage; it is in all cases much less than the full TMG-induced amount would
  be.

The UV treatment that. causes derepression in the heterogenotes carrying Pldl z+
also leads to production of transducing particles of Pldl phage.  The production of
transducing phage is greatly increased, for most strains, by superinfection with ac-
tive Pl after irradiation.6  It is interesting to compare the production of galactosi-
dase and of lac-transducing activity after these treatments. Representative
results are given in Table 2.  Clearly, UV treatment alone is responsible for the


1976       BIOCHEMISTRY: REVEL, LURIA, AND YOUNG

PROC. N. .4. S. '

20

        40

UV Dose, Seconds

60

600

200

FIG. 2.-Production of galactosidase and of transducing phage by irradiated E.
coli strain 3.09%13-41 (see Fig. 1) aa a function of the UV dose.  The values given
are measured 200 min after irradiation.  Note in Figure 1 that by this time the
transducing phage titer may have decreased substantially.

derepression of enzyme synthesis; superinfection with Pl, which greatly increases
the production of transducing activity, does not affect enzyme production signifi-
cantly.

It was also found that the amount of enzyme formed after UV-treatment de-
pends on the medium used, being lower in TGA medium than in a tryptone-yeast
extract broth, at least for some of the heterogenote strains.

The basal levels of galactosidase present in unirradiated, uninduced cultures of
the heterogenotes are often higher than those of normal inducible E. coli strains,
especially when the endogenote is i+ and the exogenote i-.4 In order to check
whether this base-line constitutive synthesis of enzyme was due to a small minority
of i- mutant or recombinant homogenotes, a detailed study was made of strain
E. coli 3.080-13-42, which is i+ z-(Pldl i- zf) and produces constitutively 2 per
cent as much enzyme as after induction by TMG.  A culture of this strain was
plated out on tryptone agar ; isolated colonies were picked at random, grown in


VOL. 47, 1961    BIOCHEMISTRY: REVEL, LURIA, AND YOUNG

1977

                  TABLE 1
PRODUCTION OF GALA~OS~DASE VHOUT INDUCER ANTER LW IRRADIATION

  Bacterial      Chromosomal
   strain          genea

3.000           i+z+
W4032 Lx+        i+z+

W4032 lac +(Pl)
3.oSO-13-41

i+.?+
i+z-

3.OSO-13-42

i+z-

2.oSO-13-4
Sh 125-13-a
W4032-W-5

i+z-
i+z-
lade'

W4032-W-1

2.oso-w-1
Sh I-4

i+z-
i+z-

Pldz
prophage

i-z+

is+

i-z+

i-z+
i+z+

i+.z+

i+z+
i+z+

Treatmepn:,
UV

+ -
+ -
T  :
+ -
: T
++ T
+ -
+ -
++ +
: +
+ -
+ -

Ei0.D.
untreated
control
  (b)

0.99     1.82

1.03     1.03
1.78     0.97
0.98     1.21

0.61     1.24

100       1.1
115        1.1

99      19
133      20

11       2.2

17       2.2

4.9      1.2
4.5      1.1

5.5      1.4
5.7      1.6
0.76     0.29

1.1     0.98

Ratio
b)/(b)

0.55

1.0
1.8
0.8

0.5

88
101

5.3
6.7
5.0

7.7

4.1
4.3

3.9
3.5

2.6

1.1

The UV dose was 40 set; PI helper pbsga,,when present, was added to give a m.o.i. = 2 to 3.
and O.D. were measured at 30-tin intervals m treated and control culture+            E ( = enzyme)

TABLE 2

EFFECT OF PI SUPERINFECTION ON PRODUCTION OF GALMTOSIDASE hm OF TRANSDUCING
                          ACTIVITY
                        Treatment                          Transducing
   Bacterial strain          uv          PI   Maxirn;n;,e~fyme,      activity/ml
   3.oSo-1341    ++     i          11.2         5 x 10"
                                      18.5         1 x lad
   3.OSO-13-42    :     T          18            <loo
                                      26.9         1 x 106
   W4032-W-1          T          1.1         4 x 10'
                                        1.2         8 x 10s

TGA medium, and tested for constitutive galactosidase levels, comparing them with
i+.z+ and with i-z+ cells.  Out of 407 cultures tested, 401 showed low enzyme levels
like those of the parent culture; 5 had no enzyme and were probably luc-, presum-
ably having lost the Pldl prophage; only one out of 407 (or 0.25 per cent) was like
i-z+ and was probably a constitutive recombinant homogenate.' We conclude
that the constitutive enzyme present in r&radiated cultures of i+ heterogenotes is
not produced by a minority of i- cells, but is due to a certain degree of escape of the
phage-associated z+ gene from repression by a chromosomal i+ gene.

The eflect of X prophuge on UV-provoked derepressiolz:  When an i+ heterogenote
strain carrying Pldl z+ also carries a prophage X, which is readily induced to vegeta-
tive replication and maturation by small doses of UV light, the UV-induced syn-
thesis of galactosidase is almost fully suppressed. Data for strains of the W-5
type are given in Table 3; other strains behave similarly.  Note that the produc-
tion of Zac+ transducing particles is also reduced.  Development of phage x after
irradiation is fairly rapid, lysis beginning at. 60 min. and being complete by 100
mm8  The supression of both galactosidase synthesis and production of transducing


1978

BIOCHEMISTRY: REVEL, LURIA, AND YOUNG    PROC. pu'. A. S.

                        TABLE 3
EFFECT OF k PROPHAGE ON UV DEREPRESSION OF GALALTOSIDASE IN HETEROGENOTES
                       Treatment
  Bacterial strain          uv         Pl        Enzyme
                                      units/ml        Transducing
                                                 sotivity/ml
W4032-W-5( A)    :     T         2.9          4.4 x 103
                                       2.6         9.6 X lo3
                     -        +          1.0         9.8 x 10'
                     -          -          1.3          1.4 x 103
  W4032-W-5X*         +     T        38.8          2.1 x 105
   (X-cured)         +                  29.7         1.6 x 10'
                     -        +         0.77        3.6 X IO5
                     -                  0.85         1.8 x lo3
W4032-W-5(h)        t-   J         o .83        6.5 x 103
  (X-reinfected)        +                  0.58        2.2 x 10'
                     -                  0.87         4.0 x 103
                     -          -         0.75         1.2 x 103

   Before and after treatment the bacterial cells were grown in tryptone-yea& extract broth.
  and transducing activity (on strain 2.050 aa recipient) were measured 3 hr after trratment. Enzyme

Pldl phage may be due to a general interference with a variety of biosynthetic
processes by the developing phage X.

Derepression following injection with a virulent mutant of phage Pl:  Confirma-
tory evidence that the partial derepression of the z+ gene following UV treatment
of heterogenotes reflects the removal of immunity, rather than some unknown action
of UV light, was provided by the observation that derepression also follows another
kind of immunity release.  A virulent mutant for phage Pl, called Plkc tir, was
available, which upon infecting PI-lysogenic bacteria causes a large proportion of
them to lyse and to release phage.  An experiment in which cells of a heterogenote
strain were superinfected with phage Plkc vir is presented to Figure 3.  Production
of galactosidase, in amounts corresponding to the complement for 2 X lo6 fully-
induced cells, was produced by infection of about 5 X 10' cells.  Similar results
were obtained with all heterogenote strains.  With E. coli heterogenotes, most of the
enzyme was released from the cells by 90 min. after superinfection, in parallel with
release of phage; with Shigella heterogenotes, the release was slower.  Note in
Figure 3 that the lysate also contained transducing activity, the amount being about,
1 transduction per 1,000 active phages.  This indicates that the Pldl phage, whose
Z+ gene controls the derepressed enzyme synthesis, participated in the phage yield.
Control experiments showed no derepression of a chromosomally located Z+ gene
in Pl-sensitive or PI-lysogenic bacteria superinfected by Plkc vir.  The presence
of a h prophage did not alter the results of superinfection with Plkc vir.

Discussion.-Phage PI in the prophage state is partially inducible to vegetative
replication, maturation and lysis by UV irradiation; 5 to 20 per cent of Pl-lysogenic
cells release phage after suitable UV-treatment.  The UV provoked derepression
of galactosidase synthesis under the control of a Pldl z+ prophage is probably anal-
ogous to the derepression of phage genes that would lead to phage production from
a normal Pl prophage.  Likewise, superinfection with the virulent mutant Plkc vir
overcomes prophage immunity and derepresses the function of genes in the prophage.

Most or all of the enzyme made constitutively after derepression by UV light is
released as free enzyme into the medium.  This indicates that most and possibly
all the cells that produce enzyme also undergo lysis.  The derepression of the z+
gene is accompanied by derepression of the phage genes of Pldl that control the
lytic process. There is apparently no "abortive" derepression of the lac genes


VOL. 47,1961

100

IO

BIOCHEMISTRY: REVEL, LURIA, AND YOUNG

Free Phoge-
Free Phoge-

Transductions
Transductions

0

60            120

Minutes Afler Superinfecfion

160

FIQ. 3.-Production of galactosidaae by E. coti strain 3.OSO-13-41 after multiple in-
            fection with phage Plkc vir.

analogous to the abortive transduction, in which the z+ genes in Pldl phage func-
tion but lysis and enzyme release are not observed.6

The remarkable observation is that the effectiveness of repression mechanisms on
the Z+ gene in phage depends on the state of the phage element.  ((x)  In a newly-
entered Pldl phage, the a+ gene is subject to full repression by an i+ gene in cis,
to partial repression by i+ in trans, and to variable degrees of repression by Pl-
immunity.  (b)  In an established Pldl prophage, the z+ gene is subject to variable
but almost complete repression by i+, whether in cis or trans, and is insensitive to
Pl-immunity. (c) In a Pldl element that has been activated, by W or by
Plkc vir, the z+ gene is again partially relieved from i+ repression.  A significant
conclusion that seems justified by these tindings is that the various repression
mechanisms must act at the gene level, as has already been postulated on the basis
of other evidence.g  It is difficult to see how repression mechanism acting at the
level of the protein-forming system-messenger-ribosome level-could discriminate


1980        BIOCHEMISTRY: REVEL, LURIA, AND YOUNG    PROC. N. A. S.

between messages coming from genes in the chromosome, or in a loose phage, or in
a prophage.

The working of the repressive controls can hardly be speculated upon on the basis
of present data.  Since Pl-immunity, which suppresses multiplication of a newly
arrived phage and also depresses the function of the z+ gene in such a phage, does
not affect, it in an established phage, it may be that Pl-immunity per se is not the
primary mechanism that supresses the function of most phage genes in the prophage
state.  A similar conclusion has recently been proposed on different grounds. lo
We have no evidence as to whether the derepression of z+ after UV-treatment on
Pldl heterogenotes simply reflects a change in state of the prophage or is a result of
multiplication of PM.  In the case of Xdg-carrying heterogenotes, there is evidence
that Xdg does not multiply unless a normal phage X is also present ;ri yet, derepression
of the gal genes occurs in the absence of normal X.  Hence, multiplication of the
gal-containing element is apparently not needed.
Clearly, topographical considerations play a role in regulatory mechanisms, either
because of variable local concentrations of repressors or because of neighborhood
relations between genes.  An illustration of the importance of such neighborhood
relations is the derepression of chromsomal gal+ genes upon activation of an adjacent
X prophage by UV t.reatment.  In analogous experiments with Pl-lysogenic bac-
t,eria there is no derepression of the chromsomal z+ gene.  All available evidence
indicates that a Pl or Pldl prophage, if it is attached to any chromosomal site at
all,** is not near the lac 10cus.`~  The chromosomal gal genes and the X prophage
may be considered as part of a superoperon, some of whose component genes become
repressed or derepressed in a coordinate fashion; lac genes and phage genes in Pldl
may constitute another such superoperon.

Summary.-UV-irradiation or infection by virulent phage mutant Pllcc vir cause
heterogenote strains carrying Pldl z+ phage to produce some /3-D-galactosidase in
t,he absence of an external inducer.  The enzyme produced in this derepressed
synthesis is released free into the medium by lysis.  The derepressed synthesis is
attributed to the induction of the prophage that carries the z+ gene.  No derepres-
sion of chromosomal z+ gene is observed when similar treatments, are used on
bacteria carrying a normal Pl prophage.

* Aided by grants from the National Science Foundation, G-8808, and the National Institute of
Allergy and Infectious Diseases, National Institutes of Health, E-3038 (C-l 1.
1 Buttin, G., F. Jacob, and J. Monod, C R. Ad. Sci. (Paris), 250,247l (1960).
* Yarmolinsky, M. B., and H. Wiesmeyer, these PROCEEDINGS, 46, 1626 (1960).
* Lsderberg, E. M.: and J. Lederberg, Generics, 38, 51 (1953).
4 Revel, H. R., and S. E. Luria, these PROCEEDINGS, 47,196s (1961).
6 Revel, H. R., S. E. Luria, and B. Rotman, these PROCEEDINGS, 47,1956 (1961).
6 Luria, S. E., J. N. Adams, and R. C. Ting, .Virology. 12, 348 (1960).
7 Morse, M. L., E. M. Lederberg, and J. Lederberg, Genetics, 41,758 (1956).
8 Weigle, J. J., and M. Delbriick, J. Bucleriol., 62,301 ( 1951).
9 Jacob, F., and J. Monod, J. Mol. Bid., 3, 318 (1961).
10 Melechen, N. E., and P. D. Skaar, Virology, in press (1961).
11 Arber, W., G. Kellenberger, and J. Weigle, S&weir. 2. allg. Path. Bakt., 20, 659 (1957).
I* Jacob, F., and E. L. Wollman, in Recinl Progress in Microbiology, VIIth Int. tong. Micro-
biol., Stockholm (1959), p. 15.

13 Boice. L. B., and S. E. Luria, Bacterial. Proc., p. 197 (1961).