(Received for publication, November 27, 1996, and in revised form, March 8, 1997)
From the Institute of Molecular Biology, Academia
Sinica, Nankang, Taipei, Taiwan 115 and the § Institute of
Biochemistry, National Yang-Ming University,
Taipei, Taiwan 112, Republic of China
RNase E is encoded by the rne (also
known as ams or hmp) gene and is the principal
enzyme that controls the chemical decay of bulk mRNA in
Escherichia coli. Earlier work has shown that RNase E
degrades its own mRNA, autoregulating production of the RNase E
protein. Here we show that in cells lacking RNase E activity, the
3.6-kilobase rne gene transcript is cleaved site
specifically at two locations near its center by a novel endonuclease
whose activity is modulated by the presence or absence of amino acids in the culture medium. These cleavages produce a 2-kilobase RNase E-sensitive RNA fragment corresponding to the 3 half of the
transcript. Using primer extension and RNase protection analysis, we
mapped RNase E-independent cleavages to sites 1558 and 1576 nucleotides from the 5
end of the rne transcript (coordinates 1738 and
1747 of the rne gene). Our results indicate the existence
of a previously unknown RNase E-independent mechanism for
degradation of rne transcripts and further
demonstrate that this mechanism responds to changes in cell growth
conditions.
RNase E is encoded by the rne (1-3), also known as ams or hmp, gene (for reviews see Refs. 4 and 5) and is a key enzyme in the processing and decay of mRNA, ribosomal RNA, and regulatory RNAs in Escherichia coli (4, 5). While RNase E is essential for cell growth, conditional (temperature-sensitive) rne mutants, such as rne-3071 (6, 7) and ams-1 (8, 9) have been isolated, and the sites of mutation have been mapped to the N-terminal half of the polypeptide (10), which contains the catalytic domain of the enzyme (11, 12). Inactivation of RNase E activity in these temperature-sensitive mutants by culture at nonpermissive temperature prolongs the decay of bulk mRNA and of specific transcripts (4, 5).
Overexpression of the rne gene by high copy number plasmids
slows cell growth and can result in plasmid loss or the acquisition of
rne mutations (13). Moreover, raising the rne
gene copy number 21-fold in E. coli fails to produce a
corresponding increase in production of the RNase E protein, suggesting
that RNase E concentration is tightly controlled in vivo
(14). This appears to occur by RNase E autoregulation of the stability
of its own mRNA by cleavage of rne mRNA in the
5-untranslated region; inactivation of RNase E prolongs the half-life
of the rne transcript (14, 15). However, despite evidence
for RNase E cleavage of rne transcripts by primer extension
analysis and in vitro cleavage assay, no
rne-dependent decay intermediate was observed by
Northern blotting (14, 15), suggesting that the RNase E-generated
products are rapidly degraded.
During the course of investigation of the effect of cell growth rate on
the stability of rne mRNA, we observed a
2-kb1 RNA species corresponding to the 3
half of the rne mRNA in the E. coli
rne-3071ts mutant strain grown in minimal medium
containing casamino acids. However, accumulation of this rne
transcript fragment at a nonpermissive temperature indicated that it
was not produced by RNase E cleavage of full-length rne
mRNA. Our subsequent investigations aimed at understanding the
mechanism of production of this rne-derived transcript
species have shown that inactivation of RNase E in cells growing in
minimal medium leads to accumulation of full-length rne
mRNA, which is processed in culture medium containing casamino acids to the 2-kb fragment. Our results imply that in the absence of
RNase E activity, an rne-independent, physiologically
regulated endoribonuclease is activated, and that this previously
unidentified endonuclease cleaves full-length rne mRNA
to produce an RNase E-degraded decay intermediate.
The
E. coli strains used for studying the
rne mRNA process were N3433 and N3431
(rne-3071; Ref. 16). Plasmid pSC13 (17), which contains a
full-length rne gene, was used to isolate DNA fragments as
indicated in Fig. 1B (5 probe a (ds) and 3
probe b (ds)) for making random priming DNA probes and for
construction of other plasmids. For plasmid constructions, the
XmnI DNA fragment (Fig. 1B) was ligated to a
filled-in HindIII site of plasmid pT7/T3
19 (Life
Technologies, Inc.); we then screened recombinant plasmids for the
expression of riboprobe c from the T7 promoter (Fig. 1B). From this plasmid, the DNA fragment, which contains the 5
end portion
of the rne gene to the MluI site (Fig.
1B), was further deleted and self-ligated to generate a
second plasmid in which the expression of the riboprobe d was under the
control of the T3 promoter as shown (Fig. 1B).
Luria-Bertani broth (18) was used to grow E. coli for plasmid DNA preparation. The medium used to grow cultures for RNA studies was MOPS minimal medium (19) supplemented with 0.4% glucose, 1.32 mM K2HPO4, 0.02 mM (equal to 6.75 mg/ml) thiamine, and 0.04 mg/ml of the appropriate amino acids. 1% casamino acids (Difco) or a mixture of 20 individual amino acids (final concentration of 50 µg/ml, Sigma) was added in certain experiments as indicated.
RNA Isolation, Northern Blot, RNase Protection, and Primer Extension AnalysesFresh overnight cultures grown at 33 °C in
MOPS/glucose were diluted 1:100 with fresh medium as described in each
experiment and cultured at 33 °C, or alternatively shifted to
43 °C, for different lengths of time prior to RNA isolation. Cell
growth was monitored by taking the optical density reading at 460 nm using a Beckman DU-62 spectrophotometer. Total cellular RNA was isolated as described (20) at various times after rifampicin (250 µg/ml culture) treatment or after temperature upshift of logarithmic
phase cultures (A460 = 0.4), except that after
RNA precipitation at 20 °C, cellular chromosomal DNA that forms
strands in the solution was removed by spooling on a micropipette tip, and the precipitated RNA was then centrifuged at 4 °C for 15 min. For RNA sample preparation for gel electrophoresis analysis, the RNA
pellet was washed with 70% ethanol and was then vacuum-dried and
resuspended in 15 µl of 1 × formaldehyde gel-loading dye (21) containing 40 µg/ml ethidium bromide. RNA samples were denatured at
65 °C for 15 min and separated on 1% agarose, 6% formaldehyde gels
(20 x 24 x 0.5 cm) in 1 × formaldehyde-MOPS buffer at 75 V
for 10 h (bromphenol blue migrated approximately 16 cm from the
well). A photograph of the individual gel showing ribosomal RNAs in
each RNA sample was taken using Type 665 Polaroid film (Polaroid Corp.,
Cambridge, MA) and was used to control the amount of RNA loaded in each
well. The RNA gel was then soaked in 10 × SSC for 30 min at room
temperature and transferred to a Zeta-Probe blotting membrane (Bio-Rad)
using 10 × SSC (1 × SSC = 0.15 M NaCl, 0.015 M trisodium citrate) for 12 h. RNA on membranes was
then cross-linked to the membrane using a UV-Stratalinker 1800 (0.12 J,
2 min, Stratagene). The membrane was baked in an 80 °C oven for
1 h and kept in a sealed bag at room temperature until used.
Hybridization procedures using either 32P-labeled random-priming double-stranded DNA probe (22) or riboprobe were performed according to the vendor's instructions (Bio-Rad). The radioactively hybridized bands were analyzed using either a Molecular Dynamics PhosphorImager or an x-ray autograph. Films showing ribosomal RNAs in each RNA sample were scanned and quantitated by the 300S MD computing densitometer.
The riboprobes used for the RNase protection assays are indicated in
Fig. 4A, and RNase protection was performed as described by
the vendor's instruction (Ambion). Primer extension analysis was done
as described previously (23); the primers used are primer A1912
(5-CAGAACGTTTTTCATTCAGCAGGTAAG-3
, rne gene coordinates 1912 to 1886) and primer AMS (5
-CCGCTAACTGCCTGAAAGATC-3
,
rne gene coordinates 294 to 274).
Restriction endonucleases, T7 and T3 RNA polymerases, placental
ribonuclease inhibitor, and RNase-free DNase I were purchased from
Amersham Corp., Life Technologies, Inc., Promega, Boehringer Mannheim,
or New England Biolabs and were used according to the vendor's
instructions. Plasmid DNA was isolated using an alkaline-lysis procedure (21). [-32P]dCTP was used for random priming
from the double-stranded DNA probe. [
-32P]UTP was used
for riboprobe synthesis. [
-32P]ATP was used for end
labeling of oligoprimers. Agarose and polyacrylamide gel
electrophoresis and DNA fragment isolations were performed as described
by Sambrook et al. (21).
Northern blot hybridization to a 32P-labeled double-stranded DNA probe that contains the rne coding region (325-3589 nucleotides; see Refs. 17 and 24) was used to detect the rne message and its decay intermediates. As previously observed (14, 15), we found that a transcript approximately the size of full-length rne mRNA accumulates in N3431 at a temperature nonpermissive for cell growth (43 °C), but not at a permissive temperature (33 °C) (Fig. 1A). However, in N3431 cultured at 43 °C in the same medium supplemented with casamino acids (CAA), but not in media lacking CAA, in addition to the full-length rne mRNA, an rne transcript species about 2 kb in length was observed (Fig. 1).
Using Northern blot analyses and double-stranded DNA probes (indicated
as ds in Fig. 1, B and C)
corresponding to either the 5-end region (5
probe a (ds)) or 3
-end
region (3
probe b (ds)) to further identify the origin of the 2-kb
transcript species, we found that the 3
-end probe but not the 5
-end
probe hybridized to the 2-kb species (Fig. 1C). Additional
studies (Fig. 1C) using single strand-specific riboprobes
confirmed that this species is transcribed in the same direction as the
full-length rne transcript rather than in the orientation of
a transcript known to extend in an antisense direction into the 3
end
of the rne gene coding region from a downstream promoter
(17) and thus implies that the 2-kb transcript species is a decay
intermediate produced from full-length rne transcripts. As
RNase E produced in strain N3431 is inactivated at 43 °C (16), the
results show that the production of the 2-kb transcript species occurs
by an RNase E-independent mechanism and furthermore that this mechanism
is dependent on the presence of CAA in the culture medium. Analysis of
the kinetics of decay of full-length rne mRNA and
formation of the 2-kb transcript species at various times following
rifampicin treatment of a culture that was temperature-upshifted from
33 to 43 °C for 15 (T15) or 30 (T30) min before the rifampicin was
added, which inhibits new RNA synthesis, showed that the 2-kb
transcript species was formed continuously (Fig. 2,
T15 and T30, respectively) and that the intracellular concentration of full-length mRNA declined. Thus, production of the 2-kb species occurs despite inhibition of new RNA
synthesis and therefore is in fact an rne decay
intermediate. The 2-kb RNA intermediate observed at T30 is more
abundant (Fig. 2) because the rifampicin was added 15 min later than
T15 culture thus allowing the production of the very stable (see Figs.
2 and 3) decay intermediate to continue for another 15 min. Although E. coli N3433 whose RNase E is active under both 33 and
43 °C growing temperatures was studied, we found no detectable
rne transcript (either full-length or decay intermediate) in
cells cultured in minimal medium supplemented with or without CAA (data
not shown). These results are consistent with a previous report (14)
showing that in the presence of RNase E activity the rne
message is extremely unstable.
Production of the 2-kb Decay Intermediate Requires Amino Acids in the Culture Medium Prior to Temperature Upshift to 43 °C
To investigate the apparent dependence of production of the 2-kb decay intermediate on the presence of CAA in the culture medium, we examined accumulation of this RNA species during several intervals following RNase E inactivation in medium either lacking or containing CAA. We found that a decrease in the level of full-length rne mRNA and significant accumulation of the decay intermediate started about 20 min after temperature upshift to 43 °C. (Fig. 3A, lanes 1-6, and B, bottom panel). However, accumulation of the 2-kb decay intermediate did not occur when CAA was added to minimal medium 10 min after the culture was shifted to 43 °C (Fig. 3A, lanes 7-11) or just prior to the temperature upshift (data not shown). When CAA was added only one-generation time (i.e. about 40 min) before the temperature upshift, accumulation of the decay intermediate RNA occurred about 30 min after the upshift and was delayed for roughly 10 min compared to the culture grown continuously in CAA (Fig. 3C, lanes 1-6 versus lanes 7-12). With further shortening of the interval between addition of CAA and temperature upshift, the delay in production of the 2-kb species was further prolonged (data not shown). However, in all cases accumulation of the intermediate RNA was associated with a decrease in the level of the full-length rne transcript (Figs. 2 and 3). Similar kinetics of production of the decay intermediate were observed in minimal medium containing a mixture of 20 individual amino acids (Fig. 4). Collectively, these results indicate that formation of the decay intermediate in the temperature-sensitive rne3071 mutant requires the presence of amino acids prior to RNase E inactivation.
The 5RNase protection and primer extension analyses were used
to map the 5 terminus of the decay intermediate. Using two
32P-labeled RNA riboprobes, which are indicated in Fig.
5A (top panel) as probe a and b,
respectively, and known RNA size markers in an RNase protection assay
(see "Materials and Methods"), we detected two RNA fragments whose
sizes are about 270 and 280 nucleotides in length (Fig. 5A
(bottom panel), lanes 5 and 10). Using
RNA size markers on the same gel, the 5
termini of these species were
mapped to the vicinity of rne coordinate 1750 (indicated with arrows in Fig. 5A). Based on these results,
we designed an oligonucleotide complementary to a sequence downstream
from the 5
termini locations mapped by RNase protection
(i.e. primer A1912, coordinates 1912 to 1886) and used
primer extension to map the precise locations of the 5
termini; DNA
sequence reactions using the same primer (A1912) were used as the size
standard. As shown in Fig. 5B, the 5
termini of the
intermediate RNA were detected in N3431 grown at 43 °C but not at
33 °C and were mapped at rne coordinates 1738 and 1747. The intensity of the terminus at position 1738 is stronger than that at
1747, consistent with our RNase protection data showing that the larger
species is more abundant (Fig. 5A, bottom panel,
indicated by arrows).
Additional primer extension analysis demonstrated that the 2-kb decay
intermediate is generated from intact rne mRNA rather than from the previously reported (14) RNA species produced by RNase E
cleavage 48 nucleotides from the 5 end of the full-length rne transcript. A 5
terminus at rne gene
coordinate 180 that is identical to the rne initiation site
mapped by Jain and Belasco (14) was detected at both temperatures (Fig.
5C) as was a second site at coordinate 146 in N3431 grown at
43 °C (Fig. 5C, 43 °C). The primer extension signal at
the nucleotide position 180 was much greater at 43 °C than at
33 °C, and the second 5
end at nucleotide position 146 was detected
only at 43 °C. These results, which are consistent with our results
showing that the CAA-dependent cleavage that produces the
2-kb decay intermediate becomes evident when RNase E is inactivated,
indicate that the 2-kb decay intermediate is the cleavage product of
the intact full-length rne transcript.
Previous work has shown that RNase E autoregulates its production by degrading its own transcript (14, 15). We have now found that in the absence of RNase E activity a default mechanism involving cleavage by an endoribonuclease whose activity is regulated by the presence of amino acids in minimal medium initiates degradation of rne transcripts. In addition, our finding that amino acids in minimal medium can regulate the endoribonuclease activity in vivo has not been reported previously. This kind of default mechanism of regulating RNA decay pathway may also exist for other transcripts to compensate for deficiency in RNase E activity or to respond to different cell growth conditions, including oxygen concentration, which have been shown to modulate RNase E activity (25).2
The cleavages we observed at coordinates 1738 and 1747 in the
rne message do not result from RNase III, which also has
been shown to cleave its own message (i.e. rnc transcript),
since in an rnc-105/rne-3071 double mutant strain (SW001),
the 2-kb decay intermediate was formed as in N3431 (data not shown).
Another major endoribonuclease RNase P, which is essential for tRNA
processing, has also recently been reported (26) to be involved in
mRNA degradation. Whether the observed endoribonuclease whose
activity regulated by amino acids in minimal medium is RNase P is
unknown. However, there is no reported evidence that RNase P activity
is affected by the presence or absence of amino acids. Therefore, production of the 2-kb decay intermediate appears to result from an
unidentified endonuclease whose activity requires specific amino acids
in the culture medium. Whether there was an initial cleavage at an
upstream site coupled with subsequent processing to these termini
cannot be ruled out. However, no 5-3
exoribonucleases have been
identified in E. coli, and other decay intermediates resulting from endonuclease digestion would have been detected by the
experiment shown in Fig. 1. The 2-kb decay intermediate observed here
was not observed in previous studies (14, 15) where LB medium rather
than CAA minimal medium was used.
It is worth noting that an essential glycolytic enzyme, enolase, that catalyzes the conversion of 2-phosphoglycerate to phosphoenolpyruvate has been found in the RNase E complex or degradosome independently by us and others (27, 28). Bacteria mutated in the eno gene cannot be cultured in glycerol minimal medium, but it can grow if either CAA, succinate, or malate is provided (29). As tight control of RNase E expression is essential for cell growth (13, 14), it is possible that this CAA or medium composition-rescued phenotype is related to the activation of the observed endoribonuclease activity that hydrolyzes rne mRNA.
We thank S.-W. Wu for construction of SW001. We also thank S. N. Cohen for helpful discussion of the results and useful comments on the manuscript.