- Email: [email protected]
Non-toxic expression in Escherichia coli of a plasmid-encoded gene for phage T4 lysozyme
259
Gene, 38 (1985) 259-264 Elsevier GENE 1400
Non-toxic expression in Escherichia cofi of a plasmid-encoded gene for phage T4 lysozyme (Recombinant DNA; tat promoter; primer repair reaction; selective disadvantage)
L. Jeanne Perry”,HerbertL. Heynekerb**and Ronald WetzeP** Departments of “Biocatalysis and bMolecwlarBiology, Genentech, Inc.. 460 Point San Bruno Boulevard, South San Francisco, CA 94080 (U.S.A.) Tel. /415)952-1000. Ext. 6268 (Received February 2nd, 1985) (Revision received May 15th and July 6th, 1985) (Accepted July 8th, 1985)
SUMMARY
The phage T4 gene coding for lysozyme has been cloned into a plasmid under control of the (t~/luc) hybrid tuc promoter and expressed in Esc~e~chi~ coli with no significant toxic effect to actively growing cells. E. coli D1210 (!uc~~) transformed with this plasmid produced active T4 lysozyme at levels up to 2% of the cellular protein after induction with isopropyl-P-D-thiogalactoside. A strain producing active lysozyme was shown to be under a selective disadvantage when co-cultured with a similar strain producing inactive lysozyme. Purified strains, however, are reasonably stablle in culture and under normal storage conditions.
INTRODUCTION
T4 lysozyme is a muramidase which facilitates lysis of T4 ph~e-cited cells, thereby releasing replicated phage particles (Tsugita et al., 1968; Tsugita, 197 1). It is similar to hen egg white lysozyme in the structure of its active site and in its site * Present address: Genencor, Inc., 180 Kimball Way, So. San Francisco, CA 94080 (U.S.A.) Tel. (415)588-3475. **To whom correspondence and reprint requests should be addressed. Abbreviations: Ap, ampicillin; bp, base pair(s); IPTG, isopropylP-D-thiogalactoside; LB, Luria broth; mg/hterd,,,, see legend to Fig. 1; nt, nucleotide(s); PolIk, Klenow fragment of E. coli DNA polymerase I; R, resistance, resistant; Tc, tetracycline; [ 1, designates plasmid-carrying state. 0378-l 1~9/85/SO3.30 0 1985 Elsevier Science Publishers
of attack on the peptidoglycan cell wall (Rossmann and Argos, 1976; Remington and Matthews, 1978; Matthews et al., 198 1). In contrast to what occurs in the traditional laboratory use of the egg white enzyme, T4 lysozyme presumably lyses T4-infected cells by attack on the cell wall from the cytoplasmic side. To utilize T4 lysozyme as a model for site-directed mutagenesis studies of protein structure-function relationships (Perry and Wetzel, 1984), it was necessary to engineer the gene for T4 lysozyme for expression in E. eoli. While the gene has been cloned into a Iz vector for DNA sequence analysis (Owen et al., 1983), no attempt to express this cloned gene has been reported. Although cytoplasmic expression of this enzyme during T4 infection of E. coli plays a role in host cell lysis, we felt that non-toxic expres-
260
sion of the plasmid-encoded gene might be possible. Expression of a number of T4 genes is required for lytic release of replicated phage (Tsugita, 1971; Josslin, 1970); it thus seemed possible that expression of the lysozyme gene alone might be tolerated by E. coli. We also planned to utilize an inducible promoter, such as the tacI1 system (De Boer et al., 1982), which might allow growth of transformed strains to high density before induction of the potentially toxic protein. In this paper we report that active T4 lysozyme can be produced in E. coli at levels up to 2% of the cellular protein.
with phGH907lacII. Other strains used as plasmid hosts were E. coli294 (Bachman et al., 1976) and E. coli D1210 (Sadler et al., 1980), which carries the lucZq allele responsible for overproduction of the luc repressor. Site-directed mutagenesis to produce a gene encoding a T4 lysozyme with an Asp20+Asn mutation will be described elsewhere. (b) DNA manipulations The methods for purification and isolation of DNA, cleavage with restriction endonucleases, and elongation reactions with PolIk were as described in Maniatis et al. (1982). Standard procedures were also used for Ml3 sequencing (Messing, 1983) and primer repair reactions (Cabilly et al., 1984; Goeddel et al., 1980). Positive clones were identified by analysis of restriction digests of extracted plasmids. Positive clones were also identified by the ability of frozen/thawed cells producing lysozyme to lyse spontaneously, or by detection of lysozyme activity in extracts.
EXPERIMENTAL
(a) Materials Cytosine-containing T4 DNA (O’Farrell et al., 1980; Owen et al., 1983) was a gift of Chung Liu (Genentech). The phage M13mplO (Messing, 1983) was provided by Peter Seeburg (Genentech). The tucI1 promoter (De Boer et al., 1982) was obtained from the plasmid phGH907tacII provided by Herman de Boer (Genentech). The dam -dcm - E. coli GM48 (Marinus, 1973) was used in transformations Fig. 1. Scheme digested
for cloning
the T4 lysozyme
with XhoI and a fragment
gene (e) into plasmid
of about 4000 bp isolated
with RFaI and Hind111 to yield a gel-purified the 5’ end. The plasmid HindIII.
pKCEAtet’
The gel-purified
707-bp fragment
(Cabilly
vector fragment
expression
pT4lysXHtrp
grew normally
The
plasmid
in LB or M9 media in shaker
5’-untranslated lysozyme purified
XbuI-EcoRI
(Wallace fragments
fragment + &I.
with
fragment
The lysozyme
A,,,).
reaction
Simultaneous
is provided
to
the presence
yield
a
from positives
containing
111 I-bp
generated
was used to screen
in this way were sequenced
the 5’ half of the lysozyme
gene was isolated
cells were thawed
of T4 lysozyme,
fragment.
A
at a level of about
14-nt
oligonucleotide
et al., 1984) to delete the 97 bp of
containing
the tailored
front half of the
with BamHI the plasmid transformants
pTBysXRfrpA5’.
Stringent
from the ligation
reaction.
the structure
from the pT4lysXRhpAS’
The plasmid
and the large vector fragment
of pT4lysXRtrpA5’.
plasmid
by digestion
pT4lysXHtrp
provided
the 3’ half of the gene, the TcR gene, and part of the ApR gene after cleavage
(isolated
from E. coli GM48) was cleaved with XbuI and PstI to yield a fragment
of the ApR gene. The isolated
into E. coli 294.
fragments
with
with indole acrylic acid.
purified by gel electrophoresis.
in Ml3 to confirm
with
in 50 mM
phGH907rucII
and the 3’ portion
by transformation
identified
fragments
digested
in which the closure
gene is flanked by untranslated
by induction
et al., 1980; Cabilly
and the small fragment
plasmid
codon 97 bp in from
pTBysXHtrp,
the lysozyme
level was not increased
(Goeddel
of the three
was then cleaved
by the ApR gene. E. coli 294 transformed
was then cleaved with EcoRI and the portion
ligation
T4 DNA was
fragment
gene with its initiation
to yield the plasmid
assay confirmed
The expression
XbaI + BumHI repair
of cytosine-containing
The purified
flasks. When frozen E. coli 294[pT4lysXHtrp]
et al., 1981) of dot blots of isolated
The plasmid
the tucII promoter isolated
the T4 lysozyme
(Grey et al., 1984) was cleaved with EcoRI, tilled in with PolIk, then cleaved
XbuI + EcoRI. The plasmid EcoRI
cleaved
A sample
with XbaI, tilled in with PolIk and subsequently
and selection
pBR322 was cleaved with EcoRI + BumHI
by gel electrophoresis.
hybridization 230-bp
was
was used in the primer
T4 DNA. The resulting
gene isolated.
phGH207-l*
unit at 550 nm (mgjiter
pT4lysXHrrp
(pATGAATATATlTGA)
containing
vectors.
gel electrophoresis.
an XbaI site. In plasmid pT4lysXHtrp,
Tris . HCI, pH 8,1 mM EDTA, they lysed spontaneously. 0.02 mg/liter per absorbance
expression
by agarose
was ligated to the 707-bp fragment
of the trp promoter,
is under control
Plasmid pT4lyszucII was constructed as shown and described in Fig. 1. After confirmation by (1) the presence of lysozyme activity, (2) restriction analysis
et al., 1984) was cleaved
ofthe RsaI site with the filled-in XbaI site regenerated T4 DNA,
(c) Lysozyme expression
were ligated to form the plasmid
pT4lystucI1,
The with with
containing which was
261
Bacteriophage
Xba
T4 DNA Xho
1
I
I
Isolate -4-kb fragment Rsa
Hind Ill
I, Hind III
Isolate 700-bp fragment
Xba I Pol Ik
I
Hind Ill Isolate large fragment
I
T4 DNA ligase
Xbol,BamHI Isolate Illl-bp fragment Anneal pATGAATATATTTGA Primer repair reaction EcoRI Isolate 230-bp fragment EcoRl Pol Ik BamHl Isolate large fragment
EcoRI, BamHI Isolate 37%bp
fragment
1
T4 DNA I igase
T
Xhl
EcoRl ,Pstl Isolate large fragment
Xbal,PstI Isolate l047-bp
fragment
Xbal,EcoRI Isolate 230-bp
<
1
T4 DNA ligase
frogment
262
of plasmid
DNA,
and (3) Ml3
sequencing
of the
TABLE
I
entire lysozyme gene, the plasmid was used to transform the k-repressor-overproducer strain E. coli
Relative viabilities ofE. coli 294 strains carrying
D1210. E. coli D1210 transformed
tacI1) T4 lysozyme
grew
normally
and
produced,
with this plasmid uninduced,
about
0.25 mg lysozyme per liter of one As50 cell culture. Induction with IPTG increased the yield to up to
ing expression
Day
Y0 colonies
No.
expressing
about 20 mg/liter A,,,. (d) Effect of lysozyme
expression
While lysozyme-producing dence of toxicity
or lysis during
fermentation,
Exp. 3
100
(99) 90
(90) 15 45
100
cul-
tures left standing for several days show significant lysis, and clones occasionally lose their lysozymeproducing ability. To assess the effect of expression of active lysozyme on cell viability and culture growth rate, we grew cultures of E. coli 294[ pT4lysCUCII] mixed with E. coli 294[ pT4lys20NtacII] ; this latter culture produces an inactive lysozyme with an Asp 20+Asn replacement (Wetzel, 1986) at protein levels similar to the wild type. As shown in Table I, cells expressing active lysozyme are in selective disadvantage when grown with similar cells expressing inactive lysozyme.
100
90
100
90
33
100
83
28 10
100
80
100
15 35
8
100
9
100
30
10
100
25
11
100
23
12
100
20
13
100
13
14
100
13
15
100
16
100
0
a Values shown are the “/b of isolated lysozyme
activity in culture.
carbenicillin/ml, saturated
were
294[pT4lystacII]
On day 2, after overnight
which produced
inoculated
plus
(LB, 50 fig
with
1% 294[pT4lys20NrucII];
plus 10% 294[pT4lys20NrucII].
shaking
at 37”C, cultures
were streak-
ed on LB plates (Ap, Tc) and also used to inoculate cultures
minimal medium. This showed that low expression was not toxic. In an effort to improve expression, 97 bp of untranslated T4 DNA were removed from the 5’ end of the lysozyme gene by a primer repair reaction (Goeddel et al., 1980; Cabilly et al., 1984) and the tailored gene inserted behind the tacI1 (De Boer
relatively high amounts of lysozyme uninduced, and did not respond further to IPTG (not shown). E. coli D1210 transformed with this plasmid was inducible with IPTG, showing an increase in lysozyme from 0.25 to about 10 mg/liter A,,, upon induction. These cells grew normally before and after induction, showing no evidence of lysis during fermentation. Frozen cells, however, lyse completely upon thawing. Since the conversion of pT4lysXHtrp to pT4lystacI1 involved a change in promoter as well as
al., 1982) promoter (Fig. 1). E. coli 294 transformed with the product of these manipulations produced 2-5 mg/liter A,,, T4 lysozyme; due to the low level of lac repressor in these cells, this strain produced
day 3, etc. Isolated
cultures
colonies
were
from plates
similarly
fresh 5-ml
Cloning of the T4 lysozyme gene was facilitated by the availability of the genetic/restriction map of the T4 genome (O’Farrell et al., 1980) and the DNA sequence of a restriction fragment containing the T4 lysozyme-coding gene e (Owen et al., 1983). Our initial construction, pT4lysXHtrp, directed low (0.02 mg/liter A,,,) expression of T4 lysozyme when transformed lysozyme E. coli 294 were grown in
et
(10 ~1). These
10 ~1 of
E. coli 294[pT4lysrucII];
of: Exp. 1, purified
Exp. 3, 90T0 294[pT4lysracII] DISCUSSION
colonies
On day 1, 5-ml cultures
5 pg Tc/ml)
cultures
Exp. 2, 99%
encod-
(pT4lys20N-
active lysozyme” Exp. 2
show no evi-
plasmids
or inactive
Exp. 1
on cell viability
cultures
of active (pT4lysfacII)
were grown
microtiter
dish wells, lysed by freeze/thaw,
microtiter
version
contain
on
in LB in
and assayed
in a
assay. At day 9, when lOO:h
in Exp. 3 showed no lysozyme activity, ten out of
of the colonies ten of these
of the turbidity
processed
colonies
the em mutant
were plasmid
shown
by restriction
pT4lys2ONracII
analysis
to
(see section d of
EXPERIMENTAL).
263
deletion of 5’untranslated DNA, we cannot rigorously assign the increased intracellular yield to the removal of 5’-untranslated DNA. It is unlikely, however, that the unrepressed tucI1 promoter, which contains the RNA polymerase binding elements of the trp promoter, would differ qualitatively in strength from unrepressed trp promoter. More importantly, analysis of early- and late-transcribed T4 lysozyme mRNA has implicated a hairpin loop, containing the AUG initiator, in inhibiting translation efficiency (D.S. McPheeters and L. Gold, personal communication). The sequence capable of forming this hairpin structure is present in the pT4lysXHtrp construction, but not in the pT4lystucI1.
From the work of Tsugita et al. (1968) it can be calculated that the yield of T4 lysozyme in TCinfected E. coli B is 5-10 mg/liter A,,,, based on the A,,, of the culture just prior to infection. Expression of lysozyme in pT4lystacII-transformed D1210 (an E. coli K-12 strain), after IPTG induction, reaches about the same level. That significant expression of a cloned lysozyme gene does not lead to lysis of the host cell confirms that there are other T4 genomic loci whose expression is required for lysis of infected cells (Tsugita, 197 1; Josslin, 1970). While the use of the inducible tacI1 promoter may have some effect on final yield, it is apparently unnecessary for suppression of product toxicity, since transformed E. coli 294 are not compromised by their constitutive expression of amounts of lysozyme comparable to those seen post induction in D 12 lO[pT4lystucIIl. As shown in Table I, there is a selective disadvantage to expression of active T4 lysozyme. However, the more rigorous purification of inoculant clones has eliminated the apparent instability of lysozymeproducing strains which we occasionally observed in growth and characterization of these cell lines.
Smith and Genentech scientists Chung Liu, Fred Young, Harvey Miller, Bill Holmes, Michael Rey, and Lisa Coussens for helpful discussions.
REFERENCES Bachman,
K., Ptashne,
plasmids Acad. Cabilly,
carrying
M. and Gilbert,
W.: Construction
the CI gene of bacteriophage
of
1. Proc. Natl.
Sci. USA 73 (1976) 4174-4178. S., Riggs, A.D., Pande,
H., Shively, J.E., Holmes,
Rey, M., Perry, L.J., Wetzel, R. and Heyneker, ation of antibody
activity
chains
in Escherikhia coli. Proc.
produced
W.E.,
H.L.: Gener-
from immunoglobulin
polypeptide
Natl. Acad.
Sci.
USA 81 (1984) 3273-3277. De Boer, H., Heyneker,
H., Comstock,
M. and Horn, T.: Construction
in Escherichia coli, in Ahmad,
their properties
Smith, E.E. and Whelan, Translation
J.,
Winter
Symposia,
E., Leung, D., Crea,
S.: Synthesis
of human
fibroblast
by E. coli. Nucl. Acids Res. 8 (1980) 4057-4074.
interferon
G.L., McKeown,
Heyneker, rectly
and
F., Schultz,
1982, pp. 309-327.
H.M., Yelverton,
R., Sloma, A. and Pestka, Grey,
Miami
Press, New York,
D.V., Shepard,
A., Vasser,
W.J. (Eds.), From Gene to Protein:
into Biotechnology;
Vol. 19. Academic Goeddel,
L., Wieland,
of three hybrid promoters
Seeburg,
P.S. and
H.L.: Pseudomonas aeruginosa secretes
and cor-
processes
K.A., Jones,
human
growth
A.J.S., hormone.
Biotechnology
2
(1984) 161-165. Josslin,
R.: The lysis mechanism
lysis. Virology Maniatis,
T., Fritsch,
A Laboratory Spring
E.F. and Sambrook,
Manual.
Harbor,
Marinus,
of phage T4: mutants
affecting
40 (1970) 719-726.
M.G.:
J.: Molecular
Cold Spring Harbor
Cloning.
Laboratory,
Cold
NY, 1982. Location
of DNA
Escherichia coli K-12 genetic
methylation
map.
genes
Mol. Gen.
on the
Genet.
127
(1973) 47-55. Matthews,B.W.,
Remington,
M.G. and Anderson,
S.J., Gruetter,
W.F.: Relation between hen egg white lysozyme
and bacterio-
phage T4 lysozyme:
J. Mol. Biol.
evolutionary
implications.
147 (1981) 545-558. Messing,
J.: New Ml3 vectors
for cloning.
Methods
Enzymol.
101(1983)20-78. O’Farrell,
P.H., Kutter,
ofthe bacteriophage
E. and Nakanishi,
M.: A restriction
map
T4 genome. Mol. Gen. Genet. 179 (1980)
421-435. Owen, J.E., Schultz, tide sequence ACKNOWLEDGEMENTS
analysis
D.W., Taylor,
A. and Smith, G.R.: Nucleo-
of the lysozyme
of mutations
involving
gene of bacteriophage repeated
sequences.
T4: J. Mol.
Biol. 16 (1983) 229-248.
We are indebted to Gerald Smith and Larry Gold for communication of their results prior to publication and to Chung Liu (Genentech) for the gift of cytosine-containing T4 DNA. We thank Genentech’s organic chemistry group for the oligonucleotide used in this work. We acknowledge Gerald
Perry,
L.J. and Wetzel,
lysozyme: vation. Remington,
R.: Disulfide
stabilization
Science
toward
thermal
into T4 inacti-
226 (1984) 555-557.
S.J. and Matthews,
assess similarity bacteriophage
bond engineered
of the protein
ofprotein lysozyme.
(1978) 2180-2184.
B.W.: A general
structures, Proc.
Natl.
method
with applications Acad.
to
to T4
Sci. USA
75
264 Rossmann, M.G. and Argos, P.: Exploring structural homology of proteins. J. Mol. Biol. 105 (1976) 75-95. Sadler, J.R., Tecklenburg, M. and Betz, J.L.: Plasmids containing many tandem copies of a synthetic lactose operator. Gene 8 (1980) 279-300. Tsugita, A.: Phage lysozyme and other lytic enzymes, in Boyer, P.D. (Ed.), The Enzymes. Academic Press, New York, 1971, pp. 344-411. Tsugita, A., Inouye, M., Terzaghi, E. and Streisinger, G.: Purification of bacteriophage T4 Iysozyme. J. Biol. Chem. 243 (1968) 391-397.
Wattace, R.B., Johnson, M.J., Hirose, T., Miyake, T., Kawashima, E.H. and Itakura, K.: The use of synthetic oligonucieotides as hybridization probes, II. Hybridization of ohgonucleotides of mixed sequences to rabbit /I-globin DNA. Nucl. Acids Res. 9 (1981) 879-894. Wetzel, R.: Investigation of the structural roles of disuhides by protein engineering; a study with T4 lysozyme, in Inouye, M. and Sarma, R. (Eds.), Protein Engineering. Academic Press, New York, 1986, in press. Communicated by A.D. Riggs.
Gene, 38 (1985) 259-264 Elsevier GENE 1400
Non-toxic expression in Escherichia cofi of a plasmid-encoded gene for phage T4 lysozyme (Recombinant DNA; tat promoter; primer repair reaction; selective disadvantage)
L. Jeanne Perry”,HerbertL. Heynekerb**and Ronald WetzeP** Departments of “Biocatalysis and bMolecwlarBiology, Genentech, Inc.. 460 Point San Bruno Boulevard, South San Francisco, CA 94080 (U.S.A.) Tel. /415)952-1000. Ext. 6268 (Received February 2nd, 1985) (Revision received May 15th and July 6th, 1985) (Accepted July 8th, 1985)
SUMMARY
The phage T4 gene coding for lysozyme has been cloned into a plasmid under control of the (t~/luc) hybrid tuc promoter and expressed in Esc~e~chi~ coli with no significant toxic effect to actively growing cells. E. coli D1210 (!uc~~) transformed with this plasmid produced active T4 lysozyme at levels up to 2% of the cellular protein after induction with isopropyl-P-D-thiogalactoside. A strain producing active lysozyme was shown to be under a selective disadvantage when co-cultured with a similar strain producing inactive lysozyme. Purified strains, however, are reasonably stablle in culture and under normal storage conditions.
INTRODUCTION
T4 lysozyme is a muramidase which facilitates lysis of T4 ph~e-cited cells, thereby releasing replicated phage particles (Tsugita et al., 1968; Tsugita, 197 1). It is similar to hen egg white lysozyme in the structure of its active site and in its site * Present address: Genencor, Inc., 180 Kimball Way, So. San Francisco, CA 94080 (U.S.A.) Tel. (415)588-3475. **To whom correspondence and reprint requests should be addressed. Abbreviations: Ap, ampicillin; bp, base pair(s); IPTG, isopropylP-D-thiogalactoside; LB, Luria broth; mg/hterd,,,, see legend to Fig. 1; nt, nucleotide(s); PolIk, Klenow fragment of E. coli DNA polymerase I; R, resistance, resistant; Tc, tetracycline; [ 1, designates plasmid-carrying state. 0378-l 1~9/85/SO3.30 0 1985 Elsevier Science Publishers
of attack on the peptidoglycan cell wall (Rossmann and Argos, 1976; Remington and Matthews, 1978; Matthews et al., 198 1). In contrast to what occurs in the traditional laboratory use of the egg white enzyme, T4 lysozyme presumably lyses T4-infected cells by attack on the cell wall from the cytoplasmic side. To utilize T4 lysozyme as a model for site-directed mutagenesis studies of protein structure-function relationships (Perry and Wetzel, 1984), it was necessary to engineer the gene for T4 lysozyme for expression in E. eoli. While the gene has been cloned into a Iz vector for DNA sequence analysis (Owen et al., 1983), no attempt to express this cloned gene has been reported. Although cytoplasmic expression of this enzyme during T4 infection of E. coli plays a role in host cell lysis, we felt that non-toxic expres-
260
sion of the plasmid-encoded gene might be possible. Expression of a number of T4 genes is required for lytic release of replicated phage (Tsugita, 1971; Josslin, 1970); it thus seemed possible that expression of the lysozyme gene alone might be tolerated by E. coli. We also planned to utilize an inducible promoter, such as the tacI1 system (De Boer et al., 1982), which might allow growth of transformed strains to high density before induction of the potentially toxic protein. In this paper we report that active T4 lysozyme can be produced in E. coli at levels up to 2% of the cellular protein.
with phGH907lacII. Other strains used as plasmid hosts were E. coli294 (Bachman et al., 1976) and E. coli D1210 (Sadler et al., 1980), which carries the lucZq allele responsible for overproduction of the luc repressor. Site-directed mutagenesis to produce a gene encoding a T4 lysozyme with an Asp20+Asn mutation will be described elsewhere. (b) DNA manipulations The methods for purification and isolation of DNA, cleavage with restriction endonucleases, and elongation reactions with PolIk were as described in Maniatis et al. (1982). Standard procedures were also used for Ml3 sequencing (Messing, 1983) and primer repair reactions (Cabilly et al., 1984; Goeddel et al., 1980). Positive clones were identified by analysis of restriction digests of extracted plasmids. Positive clones were also identified by the ability of frozen/thawed cells producing lysozyme to lyse spontaneously, or by detection of lysozyme activity in extracts.
EXPERIMENTAL
(a) Materials Cytosine-containing T4 DNA (O’Farrell et al., 1980; Owen et al., 1983) was a gift of Chung Liu (Genentech). The phage M13mplO (Messing, 1983) was provided by Peter Seeburg (Genentech). The tucI1 promoter (De Boer et al., 1982) was obtained from the plasmid phGH907tacII provided by Herman de Boer (Genentech). The dam -dcm - E. coli GM48 (Marinus, 1973) was used in transformations Fig. 1. Scheme digested
for cloning
the T4 lysozyme
with XhoI and a fragment
gene (e) into plasmid
of about 4000 bp isolated
with RFaI and Hind111 to yield a gel-purified the 5’ end. The plasmid HindIII.
pKCEAtet’
The gel-purified
707-bp fragment
(Cabilly
vector fragment
expression
pT4lysXHtrp
grew normally
The
plasmid
in LB or M9 media in shaker
5’-untranslated lysozyme purified
XbuI-EcoRI
(Wallace fragments
fragment + &I.
with
fragment
The lysozyme
A,,,).
reaction
Simultaneous
is provided
to
the presence
yield
a
from positives
containing
111 I-bp
generated
was used to screen
in this way were sequenced
the 5’ half of the lysozyme
gene was isolated
cells were thawed
of T4 lysozyme,
fragment.
A
at a level of about
14-nt
oligonucleotide
et al., 1984) to delete the 97 bp of
containing
the tailored
front half of the
with BamHI the plasmid transformants
pTBysXRfrpA5’.
Stringent
from the ligation
reaction.
the structure
from the pT4lysXRhpAS’
The plasmid
and the large vector fragment
of pT4lysXRtrpA5’.
plasmid
by digestion
pT4lysXHtrp
provided
the 3’ half of the gene, the TcR gene, and part of the ApR gene after cleavage
(isolated
from E. coli GM48) was cleaved with XbuI and PstI to yield a fragment
of the ApR gene. The isolated
into E. coli 294.
fragments
with
with indole acrylic acid.
purified by gel electrophoresis.
in Ml3 to confirm
with
in 50 mM
phGH907rucII
and the 3’ portion
by transformation
identified
fragments
digested
in which the closure
gene is flanked by untranslated
by induction
et al., 1980; Cabilly
and the small fragment
plasmid
codon 97 bp in from
pTBysXHtrp,
the lysozyme
level was not increased
(Goeddel
of the three
was then cleaved
by the ApR gene. E. coli 294 transformed
was then cleaved with EcoRI and the portion
ligation
T4 DNA was
fragment
gene with its initiation
to yield the plasmid
assay confirmed
The expression
XbaI + BumHI repair
of cytosine-containing
The purified
flasks. When frozen E. coli 294[pT4lysXHtrp]
et al., 1981) of dot blots of isolated
The plasmid
the tucII promoter isolated
the T4 lysozyme
(Grey et al., 1984) was cleaved with EcoRI, tilled in with PolIk, then cleaved
XbuI + EcoRI. The plasmid EcoRI
cleaved
A sample
with XbaI, tilled in with PolIk and subsequently
and selection
pBR322 was cleaved with EcoRI + BumHI
by gel electrophoresis.
hybridization 230-bp
was
was used in the primer
T4 DNA. The resulting
gene isolated.
phGH207-l*
unit at 550 nm (mgjiter
pT4lysXHrrp
(pATGAATATATlTGA)
containing
vectors.
gel electrophoresis.
an XbaI site. In plasmid pT4lysXHtrp,
Tris . HCI, pH 8,1 mM EDTA, they lysed spontaneously. 0.02 mg/liter per absorbance
expression
by agarose
was ligated to the 707-bp fragment
of the trp promoter,
is under control
Plasmid pT4lyszucII was constructed as shown and described in Fig. 1. After confirmation by (1) the presence of lysozyme activity, (2) restriction analysis
et al., 1984) was cleaved
ofthe RsaI site with the filled-in XbaI site regenerated T4 DNA,
(c) Lysozyme expression
were ligated to form the plasmid
pT4lystucI1,
The with with
containing which was
261
Bacteriophage
Xba
T4 DNA Xho
1
I
I
Isolate -4-kb fragment Rsa
Hind Ill
I, Hind III
Isolate 700-bp fragment
Xba I Pol Ik
I
Hind Ill Isolate large fragment
I
T4 DNA ligase
Xbol,BamHI Isolate Illl-bp fragment Anneal pATGAATATATTTGA Primer repair reaction EcoRI Isolate 230-bp fragment EcoRl Pol Ik BamHl Isolate large fragment
EcoRI, BamHI Isolate 37%bp
fragment
1
T4 DNA I igase
T
Xhl
EcoRl ,Pstl Isolate large fragment
Xbal,PstI Isolate l047-bp
fragment
Xbal,EcoRI Isolate 230-bp
<
1
T4 DNA ligase
frogment
262
of plasmid
DNA,
and (3) Ml3
sequencing
of the
TABLE
I
entire lysozyme gene, the plasmid was used to transform the k-repressor-overproducer strain E. coli
Relative viabilities ofE. coli 294 strains carrying
D1210. E. coli D1210 transformed
tacI1) T4 lysozyme
grew
normally
and
produced,
with this plasmid uninduced,
about
0.25 mg lysozyme per liter of one As50 cell culture. Induction with IPTG increased the yield to up to
ing expression
Day
Y0 colonies
No.
expressing
about 20 mg/liter A,,,. (d) Effect of lysozyme
expression
While lysozyme-producing dence of toxicity
or lysis during
fermentation,
Exp. 3
100
(99) 90
(90) 15 45
100
cul-
tures left standing for several days show significant lysis, and clones occasionally lose their lysozymeproducing ability. To assess the effect of expression of active lysozyme on cell viability and culture growth rate, we grew cultures of E. coli 294[ pT4lysCUCII] mixed with E. coli 294[ pT4lys20NtacII] ; this latter culture produces an inactive lysozyme with an Asp 20+Asn replacement (Wetzel, 1986) at protein levels similar to the wild type. As shown in Table I, cells expressing active lysozyme are in selective disadvantage when grown with similar cells expressing inactive lysozyme.
100
90
100
90
33
100
83
28 10
100
80
100
15 35
8
100
9
100
30
10
100
25
11
100
23
12
100
20
13
100
13
14
100
13
15
100
16
100
0
a Values shown are the “/b of isolated lysozyme
activity in culture.
carbenicillin/ml, saturated
were
294[pT4lystacII]
On day 2, after overnight
which produced
inoculated
plus
(LB, 50 fig
with
1% 294[pT4lys20NrucII];
plus 10% 294[pT4lys20NrucII].
shaking
at 37”C, cultures
were streak-
ed on LB plates (Ap, Tc) and also used to inoculate cultures
minimal medium. This showed that low expression was not toxic. In an effort to improve expression, 97 bp of untranslated T4 DNA were removed from the 5’ end of the lysozyme gene by a primer repair reaction (Goeddel et al., 1980; Cabilly et al., 1984) and the tailored gene inserted behind the tacI1 (De Boer
relatively high amounts of lysozyme uninduced, and did not respond further to IPTG (not shown). E. coli D1210 transformed with this plasmid was inducible with IPTG, showing an increase in lysozyme from 0.25 to about 10 mg/liter A,,, upon induction. These cells grew normally before and after induction, showing no evidence of lysis during fermentation. Frozen cells, however, lyse completely upon thawing. Since the conversion of pT4lysXHtrp to pT4lystacI1 involved a change in promoter as well as
al., 1982) promoter (Fig. 1). E. coli 294 transformed with the product of these manipulations produced 2-5 mg/liter A,,, T4 lysozyme; due to the low level of lac repressor in these cells, this strain produced
day 3, etc. Isolated
cultures
colonies
were
from plates
similarly
fresh 5-ml
Cloning of the T4 lysozyme gene was facilitated by the availability of the genetic/restriction map of the T4 genome (O’Farrell et al., 1980) and the DNA sequence of a restriction fragment containing the T4 lysozyme-coding gene e (Owen et al., 1983). Our initial construction, pT4lysXHtrp, directed low (0.02 mg/liter A,,,) expression of T4 lysozyme when transformed lysozyme E. coli 294 were grown in
et
(10 ~1). These
10 ~1 of
E. coli 294[pT4lysrucII];
of: Exp. 1, purified
Exp. 3, 90T0 294[pT4lysracII] DISCUSSION
colonies
On day 1, 5-ml cultures
5 pg Tc/ml)
cultures
Exp. 2, 99%
encod-
(pT4lys20N-
active lysozyme” Exp. 2
show no evi-
plasmids
or inactive
Exp. 1
on cell viability
cultures
of active (pT4lysfacII)
were grown
microtiter
dish wells, lysed by freeze/thaw,
microtiter
version
contain
on
in LB in
and assayed
in a
assay. At day 9, when lOO:h
in Exp. 3 showed no lysozyme activity, ten out of
of the colonies ten of these
of the turbidity
processed
colonies
the em mutant
were plasmid
shown
by restriction
pT4lys2ONracII
analysis
to
(see section d of
EXPERIMENTAL).
263
deletion of 5’untranslated DNA, we cannot rigorously assign the increased intracellular yield to the removal of 5’-untranslated DNA. It is unlikely, however, that the unrepressed tucI1 promoter, which contains the RNA polymerase binding elements of the trp promoter, would differ qualitatively in strength from unrepressed trp promoter. More importantly, analysis of early- and late-transcribed T4 lysozyme mRNA has implicated a hairpin loop, containing the AUG initiator, in inhibiting translation efficiency (D.S. McPheeters and L. Gold, personal communication). The sequence capable of forming this hairpin structure is present in the pT4lysXHtrp construction, but not in the pT4lystucI1.
From the work of Tsugita et al. (1968) it can be calculated that the yield of T4 lysozyme in TCinfected E. coli B is 5-10 mg/liter A,,,, based on the A,,, of the culture just prior to infection. Expression of lysozyme in pT4lystacII-transformed D1210 (an E. coli K-12 strain), after IPTG induction, reaches about the same level. That significant expression of a cloned lysozyme gene does not lead to lysis of the host cell confirms that there are other T4 genomic loci whose expression is required for lysis of infected cells (Tsugita, 197 1; Josslin, 1970). While the use of the inducible tacI1 promoter may have some effect on final yield, it is apparently unnecessary for suppression of product toxicity, since transformed E. coli 294 are not compromised by their constitutive expression of amounts of lysozyme comparable to those seen post induction in D 12 lO[pT4lystucIIl. As shown in Table I, there is a selective disadvantage to expression of active T4 lysozyme. However, the more rigorous purification of inoculant clones has eliminated the apparent instability of lysozymeproducing strains which we occasionally observed in growth and characterization of these cell lines.
Smith and Genentech scientists Chung Liu, Fred Young, Harvey Miller, Bill Holmes, Michael Rey, and Lisa Coussens for helpful discussions.
REFERENCES Bachman,
K., Ptashne,
plasmids Acad. Cabilly,
carrying
M. and Gilbert,
W.: Construction
the CI gene of bacteriophage
of
1. Proc. Natl.
Sci. USA 73 (1976) 4174-4178. S., Riggs, A.D., Pande,
H., Shively, J.E., Holmes,
Rey, M., Perry, L.J., Wetzel, R. and Heyneker, ation of antibody
activity
chains
in Escherikhia coli. Proc.
produced
W.E.,
H.L.: Gener-
from immunoglobulin
polypeptide
Natl. Acad.
Sci.
USA 81 (1984) 3273-3277. De Boer, H., Heyneker,
H., Comstock,
M. and Horn, T.: Construction
in Escherichia coli, in Ahmad,
their properties
Smith, E.E. and Whelan, Translation
J.,
Winter
Symposia,
E., Leung, D., Crea,
S.: Synthesis
of human
fibroblast
by E. coli. Nucl. Acids Res. 8 (1980) 4057-4074.
interferon
G.L., McKeown,
Heyneker, rectly
and
F., Schultz,
1982, pp. 309-327.
H.M., Yelverton,
R., Sloma, A. and Pestka, Grey,
Miami
Press, New York,
D.V., Shepard,
A., Vasser,
W.J. (Eds.), From Gene to Protein:
into Biotechnology;
Vol. 19. Academic Goeddel,
L., Wieland,
of three hybrid promoters
Seeburg,
P.S. and
H.L.: Pseudomonas aeruginosa secretes
and cor-
processes
K.A., Jones,
human
growth
A.J.S., hormone.
Biotechnology
2
(1984) 161-165. Josslin,
R.: The lysis mechanism
lysis. Virology Maniatis,
T., Fritsch,
A Laboratory Spring
E.F. and Sambrook,
Manual.
Harbor,
Marinus,
of phage T4: mutants
affecting
40 (1970) 719-726.
M.G.:
J.: Molecular
Cold Spring Harbor
Cloning.
Laboratory,
Cold
NY, 1982. Location
of DNA
Escherichia coli K-12 genetic
methylation
map.
genes
Mol. Gen.
on the
Genet.
127
(1973) 47-55. Matthews,B.W.,
Remington,
M.G. and Anderson,
S.J., Gruetter,
W.F.: Relation between hen egg white lysozyme
and bacterio-
phage T4 lysozyme:
J. Mol. Biol.
evolutionary
implications.
147 (1981) 545-558. Messing,
J.: New Ml3 vectors
for cloning.
Methods
Enzymol.
101(1983)20-78. O’Farrell,
P.H., Kutter,
ofthe bacteriophage
E. and Nakanishi,
M.: A restriction
map
T4 genome. Mol. Gen. Genet. 179 (1980)
421-435. Owen, J.E., Schultz, tide sequence ACKNOWLEDGEMENTS
analysis
D.W., Taylor,
A. and Smith, G.R.: Nucleo-
of the lysozyme
of mutations
involving
gene of bacteriophage repeated
sequences.
T4: J. Mol.
Biol. 16 (1983) 229-248.
We are indebted to Gerald Smith and Larry Gold for communication of their results prior to publication and to Chung Liu (Genentech) for the gift of cytosine-containing T4 DNA. We thank Genentech’s organic chemistry group for the oligonucleotide used in this work. We acknowledge Gerald
Perry,
L.J. and Wetzel,
lysozyme: vation. Remington,
R.: Disulfide
stabilization
Science
toward
thermal
into T4 inacti-
226 (1984) 555-557.
S.J. and Matthews,
assess similarity bacteriophage
bond engineered
of the protein
ofprotein lysozyme.
(1978) 2180-2184.
B.W.: A general
structures, Proc.
Natl.
method
with applications Acad.
to
to T4
Sci. USA
75
264 Rossmann, M.G. and Argos, P.: Exploring structural homology of proteins. J. Mol. Biol. 105 (1976) 75-95. Sadler, J.R., Tecklenburg, M. and Betz, J.L.: Plasmids containing many tandem copies of a synthetic lactose operator. Gene 8 (1980) 279-300. Tsugita, A.: Phage lysozyme and other lytic enzymes, in Boyer, P.D. (Ed.), The Enzymes. Academic Press, New York, 1971, pp. 344-411. Tsugita, A., Inouye, M., Terzaghi, E. and Streisinger, G.: Purification of bacteriophage T4 Iysozyme. J. Biol. Chem. 243 (1968) 391-397.
Wattace, R.B., Johnson, M.J., Hirose, T., Miyake, T., Kawashima, E.H. and Itakura, K.: The use of synthetic oligonucieotides as hybridization probes, II. Hybridization of ohgonucleotides of mixed sequences to rabbit /I-globin DNA. Nucl. Acids Res. 9 (1981) 879-894. Wetzel, R.: Investigation of the structural roles of disuhides by protein engineering; a study with T4 lysozyme, in Inouye, M. and Sarma, R. (Eds.), Protein Engineering. Academic Press, New York, 1986, in press. Communicated by A.D. Riggs.