ABSTRACT
RNase P, an enzyme essential for tRNA biosynthesis, can be directed to cleave
any RNA when the target RNA is in a complex with a short, complementary oligonucleotide called an external guide sequence (EGS). RNase P from
Escherichia coli
can cleave phage
[lambda]
N mRNA
in vitro
or
in vivo
when the mRNA is in a complex with an EGS. The EGS can either be separate from
or covalently linked to M1 RNA, the catalytic RNA subunit of RNase P. The
requirement for Mg
2+
in the reaction
in vitro
is lower when the EGS is covalently linked to M1 RNA. Substrates made of DNA
can also be cleaved by RNase P
in vitro
in complexes with RNA EGSs. When either kind of EGS construct is used
in vivo
, burst size of phage
[lambda]
is reduced by
>=
40%. Reduction in burst size depends on efficient expression of the EGS
constructs. The product of phage
[lambda]
gene N appears to function in a stoichiometric fashion.
RNase P, an enzyme with an RNA subunit, can be employed to regulate artificially
the expression of genes
in vivo
(
1
-
3
) and
in vitro
(
4
). This regulation is achieved by creating a complex made of a target mRNA with a small oligoribonucleotide (EGS) that resembles natural
substrates for RNase P. Thus, the target mRNA is cleaved by RNase P and remains
untranslated. This technology can be used both to gain insight into the
function of genes and as a means of regulating the expression of genes that are
important in pathogenic processes. We show here for the first time that the
efficiency of bacteriophage replication in a procaryote can be diminished with
this method.
The lytic cycle of bacteriophage [lambda] is governed by a series of complex regulatory steps. The N gene, the
product of which functions as an antiterminator of transcription, is expressed
early in infection and controls the further expression of several genes
essential for the complete initiation of the lytic cycle. Since the N gene is
so critical, we have chosen N mRNA as a target for inactivation by RNase P from
E.coli
to demonstrate the potential therapeutic value of this method of gene
inactivation. We now show that N mRNA in combination with two kinds of EGS constructs can be cleaved
in vitro
by RNase P. Furthermore, when these EGS RNAs are expressed
in vivo
, the burst size of phage [lambda] is reduced by 40-60%.
IPTG, XGAL and antibiotics were purchased from Sigma Chemical Co. T4 DNA ligase,
the Klenow fragment of DNA polymerase I and restriction endonucleases were
purchased from New England BioLabs; T7 RNA polymerase and RNasin were purchased
from Promega Biotec; Rapid Hybridization buffer and radiolabeled chemicals were
from Amersham; Sequenase was from US Biochemical Corporation; T4 RNA ligase and
nucleotide triphosphates from Pharmacia; nylon membranes from Boehringer-Mannheim. C5 protein was a gift of Dr V. Gopalan of this laboratory.
RNA and DNA oligonucleotides were made by automated synthesizer (Dr J. Flory,
Yale University School of Medicine), deprotected when appropriate and purified
on 12 or 15% polyacrylamide/7M urea sequencing gels. The concentration of the
purified oligonucleotides was calculated from measurements of their absorbance
at 260 nm. The theoretical
T
m of each complex between an mRNA and its EGS was calculated as described
previously (
14
).
DNA oligonucleotides used in this work:
HHA: 5'-CGGATGTGCTTTCCGGTCTGATGAGTCCGTGAGGACGAAACCTG-3'
HHB: 5'-GATCCAGGTTTCGTCTCACGGACTCATCAGACCGGAAAGCAC-3'
NA: 5'-AATTCGGATTCTCCTGTCACCAGGTCAC-3'
NB: 5'-ATCCGGTGACCTGGTGACAGGAGAATCCG-3'
N-DNA: 5'-GCTAACTGACAGGAGAATCC-3'
MB-24 (T7 promoter, bottom strand): 5'-CTATAGTGAGTCGTATTAATTTCG-3'
MB-28 (T7 promoter, top strand): 5'-AATTCGAAATTAATACGACTCACTATAG-3'
AK20: 5'-GCTCTCTGTTGCACTGGTCG-3'
5S1: 5'-TACCAT CG GCGCTACGGCGTTTCACTTC-3'
YL-1: 5'-TGGTTTTGCGCTTACCCCAACCAACAGGGG-3'
RNA oligonucleotide: N-substrate: 5'-GCUAACUGACAGGAGAAUCC-3'
BL21(DE3), a strain of
E.coli
B that harbors a gene for phage T7 RNA polymerase (
15
) was a gift from Dr W. Studier, Brookhaven National Laboratory. T7A49, which is
temperature sensitive for RNase P function and also has the gene for T7 RNA
polymerase, has been described previously (
3
). Bacteriophage [lambda] was acquired from the American Type Culture Collection.
pET3040, a plasmid containing the phage T7 promoter and terminator sequences (
16
), was obtained from Dr W. Studier. pGFIB, a plasmid containing a lipoprotein
promoter, was from Masson and Miller (
17
). pLT19xW, a plasmid containing gene N of phage [lambda], was a gift from Dr T-C Lin (Yale University). pGEM3Z was obtained from Promega Biotech.
pACYC184 was obtained from New England BioLabs. pJA1 and pJA2 are derivatives
of pUC19 that contain the gene for M1RNA under the control of T7 promoter (
18
); pJA2' is a subclone of pJA2 (
19
); p[Delta][94-204] and p[Delta]65 are derivatives of pJA2' that contain mutated M1 genes-a deletion of M1 between nucleotides 94 and
204, and a deletion of nucleotide 65 (
11
). pNT7DS1EGS and pNT7APEGS, which target [beta]-galactosidase and alkaline phosphatase respectively, were
constructed as described (
3
). Other plasmids used in this work are described elsewhere (
5
).
pNEGS, a plasmid that contains NEGS under T7 promoter control, followed by a
hammer-head and T7 terminator, was constructed by ligating pGEM3Z vector
(digested by
Eco
RI and
Hin
dIII restriction enzymes), hybrids of oligonucleotides NA/NB, HHA/HHB and the T7 terminator fragment (
Bam
HI-
Hin
dIII fragment of pET3040). Before ligation, the DNA oligos were annealed and
their 3' ends phosphorylated. pNT7NEGSs were subcloned from pNEGS to delete the
lac
repressor binding site on the plasmids (
Pvu
II digestion).
Stem NEGS was transcribed by T7 RNA polymerase using the linearized template
pNEGS (or pNT7NEGS) completely digested by
Bst
NI.
pM1NEGS is a pUC plasmid that contains a M1NEGS gene (M1 gene followed by a 22
nt spacer and NEGS) under T7 promoter control, followed by a hammerhead
structure and T7 terminator (
3
). p[Delta]65NEGS and p[Delta][94-204]NEGS contain the same cloning fragments of pM1NEGS
except that the M1 gene bears a deletion at nucleotide 65 or a deletion from
nucleotides 94 to 204 of M1, respectively, pM1NEGS, p[Delta]65NEGS and p[Delta][94-204]NEGS were constructed by ligating the M1 gene vector
(pJA2', p[Delta]65 or p[Delta][94-204] digested by
Pst
I and
Hin
dIII, with
Pst
I end made blunt by T4 DNA polymerase) and an NEGS insert generated by digestion
of pNEGS DNA with
Eco
RI and
Hin
dIII. The
Eco
RI end was made blunt with Klenow polymerase. The M1NEGS hammerhead construct
was transcribed with T7 RNA polymerase on a template made by digestion of
plasmid DNA with
Bam
HI. After self-cleavage by the hammerhead ribozyme, the M1NEGS transcript has four
nucleotides downstream from the 3' CCA sequence as indicated in the legend to Figure
1
.
pGEM3Z-N is a plasmid that contains phage [lambda] gene N under T7 promoter control. It was constructed by ligating
pGEM3Z vector (cut by
Xma
I and
Hin
cII) and the insert (fragment between
Bfa
I and
Bsp
EI from pLT19xW). The new plasmid contains ~2700 bp and the gene for Amp resistance.
N mRNA was transcribed by T7 RNA polymerase using the linearized template pGEM3Z-N (cut by
Hin
dIII or
Hpa
I).
pACYC184, which contains a different origin of replication from pBR322 or pUC19
derivatives (such as pM1NEGS), was chosen as the vector for gene N. To put gene
N under the constitutive expression promoter, the lipoprotein promoter, control
p1ppN was constructed by ligating the gene N fragment (insert of pGEM3Z-N plasmid from
Eco
RI to
Hin
dIII) to vector pGFIB1 (cut with
Pst
I and
Eco
RI). Thus, plppN is a pBR322 plasmid that carries a gene N under the control of
the lipoprotein promoter. The insert in plppN was cut out with
Nar
I and
Hin
dIII and ligated to pACYC184 vector (cut by
Nru
I and
Hin
dIII). The latter plasmid, named pAlppN, can co-exist with pUC plasmid in the same cell, and should express N mRNA
constitutively.
For tests of stem EGS, strains of BL21(DE3) containing pUC19, pNT7NEGS,
pNT7QEGS, or pNT7APEGS were used as host. Cells were grown at 37oC in LBC (LB + 100 [mu]g/ml carbenicillin) overnight, diluted 100-fold and grown to 2 * 10
8
cells/ml, then chilled on ice. Cells were collected by centrifugation at 4000
g
for 10 min at 4oC, washed once with SM buffer (100 mM NaCl, 8 mM MgSO
4
, 50 mM Tris-HCl, pH 7.5, 0.01% gelatin), and resuspended in SM buffer. These exponential starter culture cells were kept on ice
and could be used for up to 4 days.
Exponential starter cultures were diluted 10-fold in LBC + 0.2% maltose and grown at 37oC to a cell density of 1 * 10
8
cell/ml. (OD
600
= 0.4). Each culture was then separated into two equal portions, and a final
concentration of 2 mM IPTG was added to one of the portions. An additional 15
min shaking was then carried out to allow induction of T7 RNA polymerase and
EGS expression. An aliquot of 1 ml was harvested at this point, and total RNA
extracted. An aliquot of 0.5 ml was then infected with phage [lambda] (MOI = 3) simultaneous with addition of MgSO
4
(final concentration 10 mM). The mixture was gently vortexed and incubated at 37oC for 15 min. Phage adsorption was stopped by dilution of the mixture 10
4
-fold in the same medium. Diluted cells (1 ml) were transferred into 25 ml flasks and shaken at 280 r.p.m. for 45 min at 37oC. The cultures were further diluted to 10
2
-fold, and a final concentration of 0.5 mg/ml lysozyme was added to 1 ml of the dilution culture. After 30 min incubation on ice, the cell
lysates were centrifuged and the supernatants were diluted one more time (10
2
-fold) and titrated. Burst sizes were determined by dividing the number of
the progeny phage by the cell density of the infected cultures.
Total RNA extracted from cells was subjected to electrophoresis on
polyacrylamide gels that contained 7 M urea. Bands of RNA were stained with ethidium bromide and the gel was soaked in transfer buffer (10
mM Tris-acetate, pH 7.8, 5 mM NaOAc, 1 mM EDTA) for ~20 min. The bands of RNAs were then electrotransferred to a nylon membrane over the course of 12-15 h at 250 mA. After transfer, the RNA was cross-linked to the membrane with a Stratalinker (Stratagene). Hybridization was performed
in Rapid Hybridization buffer (Amersham) according to the protocol from
Amersham.
Northern analysis used for detecting the N mRNA in phage [lambda] infected cells was performed using a 1% agarose gel following the protocol of Sambrook
et al
. (
20
). Total RNA (10 [mu]g) obtained from phage-infected cells were used in each sample. DNA oligo YL-1 was used as a probe for N mRNA; AK20 was used for M1 RNA.
Standard reactions catalyzed by M1EGS RNA or mutant M1EGS RNA were incubated at
37oC for 15-60 min with 10 [mu]M M1EGS and an equal amount of substrate (2000-5000 c.p.m.) in PEG buffer (50 mM Tris-HCl, pH 7.5, 100 mM MgCl
2
, 100 mM NH
4
Cl) 4% polyethylene glycol 6000-7500; or in 3 M NH
4
OAc buffer (50 mM Tris-acetate, pH 7.5, 50 mM Mg(OAc)
2
, 3 M NH
4
OAc and 0.01% NP-40). Standard reactions carried out in the presence of C5 protein were incubated at 37oC for 15-60 min in 10 [mu]l total volume, in a reaction mixture that contained PA
buffer (50 mM Tris-HCl, pH 7.5, 10 mM MgCl
2
, 100 mM NH
4
Cl), 0.01 [mu]M M1EGS, 0.1 [mu]M C5 protein and 0.1 [mu]M substrate (2000-5000 c.p.m.). The reactions were stopped by the addition of
an equal volume of 9 M urea dye (0.02% of XC and BPB). In experiments to
determine optimum Mg
2+
concentration, reactions were stopped by precipitating the RNAs with EtOH to
remove salt. The dry samples were dissolved in 9 M urea dye. The samples were
then analyzed on a 12% polyacrylamide/7M urea sequencing gel.
Table 1
To determine kinetic parameters of various reactions, incubation was in PEG
buffer and aliquots were taken during the linear portion of the kinetics of the
cleavage reaction. Parameters were calculated from Lineweaver-Burk plots (
21
) for multiple turnover or as described by Fedor and Uhlenbeck (
22
) for single turnover.
RNAs over 30 nt long were transcribed by T7 RNA polymerase as described by
Vioque
et al
. (
17
) after digestion of plasmid DNA with the appropriate restriction enzyme. RNAs
smaller than 30 nt were transcribed by T7 RNA polymerase as described by Milligan and Uhlenbeck
et al
. (
23
). RNAs were purified on polyacrylamide/7M urea sequencing gels. UV shadowing,
or X-ray film exposure was used to locate the RNAs on the gel. The identified
RNAs were cut out of the gels, eluted and then precipitated. The dry RNA
pellets were dissolved in distilled H
2
O.
RNAs that were internally labeled were prepared by using [[alpha]-
32
P]GTP in the transcription reaction and purified on acrylamide/7M urea gel.
These RNAs were directly used after purification. The methods of Sambrook
et al
. (
20
) were used to label RNAs at their 5' termini.
RNAs used for enzymatic assays were heated to 65oC in PA buffer (see above) for 5 min, then slowly cooled to 22oC. This treatment was not performed with the M1NEGS used for the metal
ion dependence experiments.
RNAs from cells grown at 43oC: after growth in LBC overnight at 30oC, cells that contained two plasmids (pAlppN and pM1NEGS, or other
control plasmids) were re-inoculated in the same medium at OD
600
= 0.05 and grown at 30oC to OD
600
= 0.3. The cells were then shifted to 43oC for 20 min and then split into two portions. IPTG (final concentration 2
mM) was added to one of them. Both sets of the cells (with or without IPTG)
were incubated at 43oC for another 25 min. Total RNAs were extracted from the cells and used for
Northern analysis. To confirm that RNase P was inactivated in cells at 43oC, samples were collected at the time points before the temperature shift,
after 20 min at 43oC, and after 25 min IPTG induction, and used for preparation of the cell
extracts.
RNA from cells grown at 30oC: cells were treated as above except for the following difference. Instead
of shifting the temperature, cells were kept at 30oC: the time for IPTG induction at 30oC was twice as long as that at 43oC.
A target site in N mRNA for cleavage by RNase P was chosen according to its
susceptibility to RNase T
1
in vitro
(
4
,
5
). Two different EGSs were used to direct RNase P to this site which is located 50-60 nt downstream from the AUG at the 5' end of the message: one, a simple oligonucleotide called a stem
EGS (
4
; Fig.
1
) and the second, an EGS covalently linked to the 3' end of M1 RNA (M1NEGS), the catalytic RNA subunit of RNase P (
2
; Fig.
1
), which was developed from studies of substrates linked to M1 RNA (
5
-
9
). Complexes of N mRNA with both such EGSs can be cleaved readily
in vitro
as shown in Figure
2
. Panels A and B show that the target mRNA in a complex with stem NEGS can be
cleaved by both M1 RNA alone and the holoenzyme complex (M1 RNA plus C5
protein), the latter at much lower concentrations of enzyme and lower ionic
strength than the former. Panel C shows cleavage by M1NEGS: unlike the reaction
shown in panel B, on addition of C5 protein to the reaction with M1NEGS, no
significant increase in cleavage rate is seen (data not shown). All these
reactions can be driven to completion by the addition of more enzyme.
The reaction of M1 RNA, the catalytic RNA subunit of
E.coli
RNase P, and the holoenzyme complex (M1 RNA with C5 protein) with stem EGSs has been described previously (
4
,
10
) but the properties of M1EGS, particularly of that with M1NEGS, have not been
fully characterized. We showed, as have others for different covalent M1-substrate constructs (
2
,
7
,
8
,
9
) that the latter reaction is intramolecular in buffers that contain low concentration of Mg
2+
but acquires significant intermolecular properties at high Mg
2+
(
5
). The reaction of M1NEGS reaches an optimum at ~50 mM Mg
2+
(see Fig.
2
C), whereas that with a separate EGS reaches an optimum at ~200 mM Mg
2+
(data not shown). We also confirmed that there is an optimum length of the
spacer linking the EGS to M1 in M1EGS constructs (~25 nt in this case), and that the pH dependence of the reaction slopes gently upward between pH 6 and 9 in
single-turnover reactions, and that Ca
2+
and Mn
2+
can substitute for Mg
2+
in the reaction buffer although at lower rates at equivalent concentrations of
divalent cations (
5
).
An independent measure of the ability of M1NEGS RNAs to target N mRNA was
carried out in cells that contained a plasmid coding for over-expression of N mRNA and another compatible plasmid coding for M1NEGS. In
these cells (T7A49; see Materials and Methods), which were also temperature
sensitive for function of C5 protein, reduction of N mRNA levels from an
initial value of 600-800 copies/cell was 37% for [Delta]65NEGS and 66% for M1NEGS (Table
4
). At the restrictive temperatures, the inhibition was somewhat less effective.
Levels of 5S RNA were used as an internal standard in these experiments.
RNA enzymes can be used to inhibit gene expression in mammalian cells in tissue
culture (
2
,
12
) as well as in
E.coli
(
3
,
13
). The inhibition of procaryote genes attained in
E.coli
is less than the corresponding levels attained in eucaryotes, possibly because
of the tighter coupling of transcription and translation in
E.coli
. We have found that RNase P, in combination with EGSs targeted to N mRNA can be
used to reduce the burst size of bacteriophage [lambda] by 40-60%. This result is statistically significant and can be
correlated with corresponding reductions in the levels of N mRNA
in vivo
as measured by Northern blots. The latter results, however, measure intact N
mRNA as well as fragments that still hybridize to our probe. The methodology we
used must be developed further (e.g. through the use of multiple EGS genes
targeting different sites in [lambda] mRNAs) to be an effective agent in controlling virus infection of
bacteria, but it may have further utility now for the study of gene function,
especially for genes the products of which act in a stoichiometric fashion. Our
data suggest that the N gene product acts in this way. As indicated above, a
stem EGS targeted to the Q gene of phage [lambda], a late gene that controls transcription of other genes, inhibited
burst size by only 10%. It is possible that our choice of target sites in both
the N and Q genes, while advantageous
in vitro
, were not optimal in guiding RNase P
in vivo
. More effective sites could be identified by chemical footprinting
in vivo
of target mRNAs rather than by nuclease susceptibility
in vitro
.
We have also characterized
in vitro
, in part, the properties of the reaction of covalently linked M1 RNA and an EGS
RNA (M1NEGS) with its target, [lambda] N mRNA. This catalytic construct binds and cleaves substrate, DNA or
RNA, more efficiently than does M1 RNA alone in combination with a separate
EGS. A construct containing a large deletion [Delta][94-204] NEGS, in M1 RNA responds only poorly to the presence of C5
protein
in vitro
(
11
) and served as a control
in vivo
: it has much less activity
in vivo
than wild type M1NEGS, thereby showing that the inhibitory effect of M1NEGS
alone is not due solely to an anti-sense effect as has also been demonstrated by Liu and Altman (
2
) and Guerrier-Takada
et al
. (
3
).
We thank our colleagues, especially Dr C. Guerrier-Takada, for advice, encouragement and exchange of data. This research was
supported by grant GM19422 from the National Institutes of Health of the USA.
+
Present address: Boehringer-Ingelheim Corporation, Danbury, CT 06877, USA
Multiple turnover
Single turnover
E/S
K
m
(nM)
k
cat
(min
-1
)
k
cat
/
K
m
(min
-1
/[mu]M)
k
cat
/
K
m
(min
-1
/[mu]M)
M1N/RNA
148
4.75
32.2
292
M1N/DNA
99.4
0.01
0.10
0.70
(M1+N)/RNA
327
0.67
2.04
60.5
The amount of N mRNA was calculated using 5S RNA as
an internal control. The relative amount of N mRNA was calculated by dividing
the amount of N mRNA from cells involved with IPTG by the amount from cells
with no IPTG induction
REFERENCES
Return