GB2183655A - Method of increasing the half-life of an mRNA sequence - Google Patents

Abstract

The method comprises utilising recombinant DNA having a REP sequence immediately downstream (3') of the DNA encoding sequence corresponding to the said mRNA sequence. "REP sequence" means a Repetative Extragenic Palindromic sequence of approximately 30 to 50 nucleotides, which tends to form a stem loop having stem portions in palindromic relationship, and conforming to the consensus sequence shown in Figure 1 or at least 75% homologous therewith. Plasmids comprising the REP sequence from the Histidine transport operon of Salmonella typhimurium are disclosed. <IMAGE>

Description

SPECIFICATION Stabilisation of mRNA Field of the invention The present invention relates to the stabilisation of mRNA. More particularly it relates to the use of specific DNA sequences which, when transcribed, stabilise and lead to the accumulation of the resulting RNA. The ability to stabilise RNA species has a number of potential uses: (i) in expression vectors designed to maximise protein production in bacterial cells; (ii) to facilitate the in vitro production and handling of any specific RNA species; (iii) t-, stabilise 'antisense' RNA molecules used for the artificial manipulation of gene expression in both prokaryotes and eukaryotes.

Description of the prior art Most bacterial mRNA molecules have half-lives of only a few minutes although certain mRNA species, most of which encode abundant proteins, are considerably more stable. If the half-life of a mRNA molecule can be increased, the synthesis of any protein encoded by that mRNA is also likely to show a corresponding increase.

However, little is known about the factors which determine the stability of mRNA in bacterial cells.

We have now found that sequences present in multiple copies on the E. coli genome, the REP (Repetitive Extragenic Palindromic) sequences, can play an important role in the stabilisation of mRNA.

REP sequences were first identified as highly conserved nucleotide sequences present in several operons of the enteric bacteria Escherichia coli and Salmonella typhimurium. Together with several collaborators this observation was published by one of the inventors in 1982 (Higgins, C.F. etna!., Nature 298, 760-762) (Ref. 1.).

Subsequently, many additional REP sequences have been identified on the E. coli chromosome. The inventors estimate that there are between 500 and 1000 per genome, representing nearly 1% of the total cellular DNA. The characteristic features of the REP sequences have been descsribed by the inventors in a paper published jointly with Californian collaborators (Stern, M.J. etal., Cell 37, 1015-1026; 1984) (Ref. 2.).

Due to the very remarkable homology between the individual sequences, the collective term "REP sequence" was adopted by the authors of Ref. 2 and individual sequences were termed "copies". Since this terminology appears to be currently accepted, it will be employed hereafter.

The REP sequence consists of approximately 35 nucleotides, the major feature of which is an inverted repeat which can potentially form a very stable 'stem loop' structure in mRNA. The nucleotides which comprise the inverted repeat and which make up the two 'stems' of the 'stem loop', are very highly conserved and in a palindromic relationship. However, the number and nature of the intervening nucleotides, which make up the bridging portion or the 'loop' of this structure, are not conserved. REP sequences may occur singly, or in groups of two, three or four adjacent copies; adjacent copies are always in inverted orientation with respect to one another. The REP sequence is the only example of such a conserved nucleotide sequence that is present in multiple copies on the E. coli genome.The sequence conservation between the many copies of the REP sequence is greater than that between any other known class of specific functional sites (e.g.: promoters, terminators, ribosome-binding sites).

Figure I is a diagram illustrating the homology between copies of the REP sequence from sixteen independent operons which are arranged in alignment. Where more than one copy of the REP sequence occurs in an operon, these are designated A, B, C, etc. In the case of the ndhA gene, where copies of the REP sequences are found both upstream and downstream, the upstream copies are designated A' and B'. Because the REP sequence occurs in both possible orientations on the chromosome, the sequence is sometimes presented in the form of the transcribed strand and sometimes in the form of the non-transcribed strand, in order to facilitate comparison. The arrows to the left of the sequences show their orientation with respect to the direction of transcription: indicates that the sequence shown is the non-transcribed strand (i.e., equivalent to the mRNA sequence).

indicates that the sequence shown is the transcribed strand. The table at the bottom of the Figure shows the number of times each base is found in any given position in the REP sequence. The arrows below the sequences show dyad symmetries.

Immediately above the table at the bottom there is shown the consensus sequence which was modified from that originally indicated in Ref. 1 following a computer search of the GenBank database (November 1983). It is possible in theory that further minor modification may be made to this consensus sequence.

The finding that many REP sequences are in intergenic regions of multicistronic operons originally led to the suggestion in Ref. 1 that the function of these sequences might be to regulate the relative expression of genes within these operons. However, the experiments published in Ref. 2 showed that in some, if not all cases this is incorrect. Despite much speculation, the function of the REP sequences has, until recently, remained obscure.

Definitions For the put'poses of the specification and claims: "REP sequence" means a stem loop sequence of approximately 30 to 50 nucleotides having stem portions in palindromic relationship and conforming to the consensus sequence shown in Figure 1 or at least 75% homologous therewith.

'Upstream' denotes a direction towards the promoter [distal (5')] and downstream the opposite direction [distal (3')].

Summary of the invention The inventors have now discovered a role for the REP sequence in determining the stability, i.e., the half-life of mRNA. In particular, when transcribed into RNA, the REP sequence results in the accumulation of upstream RNA, extending from the promoter to the REP sequence. This discovery permits the REP sequence to be usefully manipulated by recombinant DNA techniques e.g. to stabilise specific mRNA species. This facility is useful in many areas. For example: 1. In the construction of plasmid vectors designed to maximise the overproduction of proteins in E. coli or other bacterial cells.

2. In the production of more stable RNA hybridisation probes (e.g.: from the SP6 or T7 promoter systems).

3. As a tool for stabilising RNA in vivo and in vitro to facilitate studies of RNA processing and degradation.

4. To stabilise, and therefore make more effective, 'antisense' mRNA molecules which are used to manipulate gene expression in both prokaryotes and eukaryotes (see Weintraub, H. etna!., Trends in Genetics 1, 22-25; 1985).

In accordance with one aspect of the invention there is provided a method of increasing the half-life of an mRNA sequence transcribed from DNA which comprises utilising recombinant DNA having a REP sequence immediately downstream (3') of the DNA sequence which corresponds to the said mRNA sequence.

The invention is inclusive of a method of stimulating the production of a protein or sequence thereof in which the encoding mRNA is caused to accumulate utilising a method as described above.

The invention further includes a plasmid containing the DNA sequence: [S]x.J.REP.K.[S]y where S signifies a unique endonuclease restriction site, x and y are each at least 1, x + y is greater than 2, and REP indicates a REP sequence.

The invention further includes use of a plasmid containing a REP sequence bounded by endonuclease restriction sites as a tool in the selective accumulation of mRNA.

Brief description of the drawings Figure 1 is a diagram illustrating a number of examples of the REP sequence which have been identified on the E. coli or Salmonella typhimurium chromosome, as already described; Figure 2 is a diagram (not to scale) of the histidine transport operon of Salmonella typhimurium showing the location of the REP sequences and DNA fragments used as hybridisation probes; Figure 3 is a 'Northern' blot autoradiograph showing the accumulation of an approximately 950 nucleotide transcript from the histidine transport operon amongst mRNA isolated from the parental strain (TA271), or from its derivative (TA3808) from which the REP sequence has been deleted; Figure 4 is a diagram (not to scale) showing the construction of plamids pWJ01 and pWJ1 51;; Figure 5 is a 'Northern' blot autoradiograph of mRNA isolated from an E. collstrain carrying plasmids pWJ01 (containing REP) and pKG1800 (lacking REP); Figure 6A and 6B are graphs showing the amount of downstream (A) and upstream (B) mRNa synthesised from plasmids pKG1800 (0-0) or pWJ01 (o-O) as a function of growth state (0.D.600); and Figure 7 is a diagram (not to scale) illustrating the construction of plasmids pWJ61 and pWJ62 and the location of hybridisation probes used.

In the following Examples the inventors have directly examined the effect of the REP sequence on RNA molecules in which it is present, by 'Northern' blotting, 'dot-blotting' and filter hybridisation. These Examples demonstrate that the presence of the REP sequence in a transcription unit leads to the accumulation of upstream (5') RNA, and also that this RNA species is an intermediate which accumulates during RNA degradation as a result of stabilisation.

EXAMPLE 1 Demonstration that the REP sequence causes accumulation of upstream chromosomally-encoded mRNA In this example, mRNA from two similar bacterial strains was examined, the strains differing only by the presence or absence of a REP sequence at one intragenic region. The experiment successfully demonstrated the accumulation of mRNA corresponding to the region between the promoter and the REP sequence, only in the strain in which the REP sequence was present.

The histidine transport operon of Salmonella typhimurium is shown in Figure 2 and consists of four genes hi & , hisO, hisM and hisP. The entire operon has been sequenced and details appear in Ref. 1. Two copies of the REP sequence are found in the hisJ-hisO intragenic region in inverted orientation with respect to each other.

The inventors previously described (in Ref. 2.) the construction of a strain of S. typhimurium (TA3808) in which the REP sequences are deleted from this intergenic region.

mRNA was isolated from cells grown exponentially in LB of the parental strain (TA271) and also from the derivative with the REP sequences deleted (TA3808), by the SDS-phenol exraction method described by Peck and Wang (Cell 40, 129-137; 1985). Total RNA was separated on a 1.3% agaroseformaldehyde gel and transferred to a nitrocellulose filter as described by Maniatis (Maniatis et a/., "Molecular cloning. A laboratory manual", Cold Spring Harbor, New York. 1982). The filter was probed for mRNA sequences encoded by the histidine transport operon using an appropriate DNA probe labelled by nick-translation (Rigby, P.W.J. et al, J. Mol.Biol. 113 237-245; 1977).The 225 bp Hindlll-Kpnl fragment from the histidine transport operon, which includes only DNA from the hi & gene (Figure 2), was used as a probe. The location of this fragment is shown at "225" in Figure 2.

Figure 3 shows the results of such a 'Northern' blot. Two important hybridisation bands are seen on this blot.

Both lanes show full-length mRNA of about 3310 nucleotides (This RNA is somewhat shorter from strain TA3808 due to the 141 bp deletion of the hisJ-hisQ intergenic region in this strain). However, TA271, in which the REP sequence is present also shows considerable accumulation of an RNA species of about 950 nucleotides.

This corresponds closely in size to RNA extending from the his promoter to the REP sequence and was shown to encode the entire hisJ gene.

To confirm that this RNA species extends from the promoter up to, but not beyond, the REP sequence, the blot was also probed with a 955 base pair Kpnl-Hindlll fragment, the 5' end of which is located close to the beginning of hisO and immediately downstream (3') to the REP sequence (the location of the fragment is shown as "955" in Figure 2). As expected, this probe still detected full-length RNA but failed to detect the 950 nucleotide RNA species, demonstrating that this 950 nucleotide mRNA species does not extend downstream from the REP sequence to any significant extent (data not shown).

EXAMPLE 2 Demonstration that the REP sequence causes accumulation of upstream RNA encoded by plasmids In order to ascertain that the REP-dependent accumulation of upstream mRNA noted above is not specific to the histidine transport operon, the REP sequence was cloned in various plasmids and the RNA examined by 'Northern' blot analysis as described above.

As an example, plasmids pWJO1 (Figure 4d) and pWJ1 51 (Figure 4c) were constructed by cloning the 218 bp Rsal fragment from the histidine transport operon (containing the REP sequence and shown in Figure 4a) into the Smal site of pKG1800 Figure 4(b)).

These plasmids were constructed as follows: The REP sequences from the S. typhimurium histidine transport operon were isolated as a 218 bp/Rsal fragment of DNA from plasmid pFA7 (described in Higgins, C.F. & Ames, G.F-L. Proc.Natl.Acad.Sci.USA 78, 6038-6042, 1981). This DNA fragment was cloned into the vector pKG1800 in two alternative ways. pKG1800 has been described elsewhere (McKenney, K. et al, In: Gene Amplification and Analysis, Vol II, J.C. Chirikjian & T.S. Papa, eds, pp383-415, New York/North Holland Elsevier, 1981). Briefly this vector contains the promoter from the galactose operon (Pgal) which promotes transcription of the galK (galactokinase) gene. Between the promoter (Pgal) and the initiation codon of the galK gene is a unique Smal site into which any DNA fragment can be inserted.In construction I, the 218 bp Rsal fragment was ligated directly into the Smal site of pKG1800, using T4 DNA ligase, to construct plasmid pWJ01. In construction II, octanucleotide BamH1 linkers were ligated to the 218 bp Rsal fragment and to pKG1800 digested with Smal.

Excess linkers were removed by digestion with BamH1 and subsequent purification by gel exclusion chromatography (with Sephadex G-50 resin). The two were mixed and ligated with T4 DNA ligase to produce plasmid pWJ1 51. pWJO1 and pWJ1 51 differ only in the presence of BamH1 linkers (ie, two BamH1 enzyme sites) flanking the 218 bp REP fragment. The structure of both pWJO1 and pWJ1 51 and the orientation of the REP DNA fragment was confirmed by digestion with appropriate restriction endonucleases and separation by agarose gel electrophoresis. The usefulness of pWJ1 51 was in that the additional BamH1 sites facilitated the isolation of appropriate hybridization probes for a number of experiments.In the diagram (Figure 4) of pWJ1 51 there are shown a BstEIl-BamH1 fragment and a Pvul-Accl fragment present on the operon fragment, and therefore present also in pKG1800 and pWJ01.

RNA was isolated from cells containing these plasmids (grown in M63 medium supplemented with 2% fructose and 5 x 1 0-4M fucose) and analysed on a 'Northern' blot as described above and shown in Figure 5.

Tracks A and B show the hybridisation patterns obtained with RNA isolated from strains carrying pKG1800 and its derivative pWJ1 01, respectively.

The 406 bp BstEll-BamH1 fragment of pWJ1 51 was used as a probe. (It could alternatively have been obtained from pKG1800 using the BstEIl and Smal sites). This probe will only detect RNA containing sequences upstream (5') from the REP sequence. Full-length mRNA (2.8 kb) extending from the promoter (Pgal) to the terminator, is present in both tracks. However, quite clearly the plasmid pWJO1 which contains the REP sequence, but not the parental plasmid pKG1 800, shows accumulation of an additional RNA species of about 600 nucleotides. This species corresponds in size to RNA extending from the promoter to the REP sequence.

To show that this 600 nucleotide RNA species does not extend downstream from the REP sequence, the same RNA samples were probed using the Accl-Pvul fragment of pWJ1 51 (Tracks C and D). This probe (Figure 4), which will detect only RNA molecules containing sequences downstream from the REP, does not detect the 600 nucleotide RNA species, confirming that it only includes sequences upstream of the REP sequence.

Several other different plasmids, constructed in different ways, also show accumulation of mRNA upstream of the REP sequence (data not shown).

EXAMPLE 3 Demonstration that the REP-dependent accumulation of upstream RNA is not the result of a termination event It is important to establish whether the RNA species accumulated upstream of the REP sequence is an intermediate in synthesis (i.e. the result of premature transcription termination) or in degradation.

The level of downstream mRNA can be quantitated by a 'dot-blot' procedure. Cells containing pWJ101 (+ REP) or pKG1800 (-REP; Figure 4), were grown in M63 + 0.2% fructose + 5 x 1 0.4M fucose. 1 ml samples of these cells were withdrawn at appropriate time intervals when growth had reached the Or6"0 indicated at the x axis of Figure 6. RNA was isolated from each of these samples as follows.The cells from 1 ml of culture were resuspended in 100 us of ice-cold sucrose (20%): Tris-HC1 (10 mM, pH 8.0): EDTA (10mM): lysozyme (3 mg ml~' ) . Cells were lysed by three rounds of rapid freeze-thawing and then 100 u1 of formaldehyde, 200 ul of 20 x SSC (Maniatis et al., op.cit) and 30F1 of 10% SDS were added. The samples were heated at 60"C for 15 min and the debris removed by centrifugation in an Eppendorf centrifuge for 2 min.The sample was diluted 1:4 with 4 x SSC and 100,t11 dotted onto nitrocellulose filters using a Bio-Rad 'dot-blouer'. The RNA was baked onto the filter at 80"C for 2h and then hybridised with an appropriate 32P-labelled probe at 42"C in 50% formamide as described by Maniatis et al., (op.cit.). The amount of 32P-labelled probe which hybridises to each spot is directly proportional to the amount of complementary mRNA and can be quantitated by scintillation counting. The results from an experiment using the Accl-Pvul fragment of pWJ151 as a probe are shown in Figure 6A. This probe will only hybridise with mRNA downstream of the REP sequence.Clearly, the amount of RNA downstream of the REP sequence synthesised from pKG1800 (lacking the REP sequence) is identical to that synthesised from pWJO1 (containing the REP sequence). Thus, the REP sequence does not significantly reduce the amount of promoter-distal (3') mRNA and cannot therefore function as a terminator.

EXAMPLE 4 Ouantitation of mRNA accumulated upstream from the REP sequence using dot-blots RNA was prepared from cells harbouring either pKG1800 or pWJO1 grown to various optical densities as described above. RNA was quantitated by hybridisation using the BstEII-BamH1 fragment from pWJ1 51 (Figure 4) as a probe: the results are shown in Figure 6B. At early stages of growth there is little difference between the two plasmids in the amount of upstream mRNA synthesised. However, at later stages of growth there is clearly at least 5-fold more upstream mRNA synthesised from the plasmid containing the REP sequence (pWJO1).As described in Example 3 above, the same mRNA samples from these two plasmids show no difference in the amount of downstream (promoter-proximal) RNA, providing a good internal control and confirming that the REP sequence specifically causes the accumulation of upstream RNA.

EXAMPLES Quantitation of accumulated mRNA by filter hybridisation As a further confirmation that the steady-state level of upstream RNA is increased by the REP sequence, a very different procedure for measuring upstream mRNA was adopted. Plasmids pWJ61 and pWJ62 were constructed as shown in Figure 7.

PWJ53 contains the galactokinase gene (galK) transcribed from the trp operon promoter Ptrp. Between this promoter and the galK gene is the REP sequences from the S. typhimurium histidine transport operon in a 218 bp fragment (Figure 4(a)) inserted using BamH1 linkers. The construction of pWJ53 has been described in Ref.

2. In order to place a gene upstream (3') to the REP sequence the chloramphenicol acetyl transferase (cat) gene was excised from pCM1 (Close, T.J. and Rodriguez, R: Gene 20, 305, 1982) as a Sall DNA fragment and ligated into the Sall site of pWJ53 with T4 DNA ligase. This produced plasmid pWJ62 which contains the cat and galK genes as part of a single transcription unit with the REP sequence located between the two. In order to produce a derivative of pWJ62 from which the REP sequence was deleted, pWJ62 was digested with BamH1 and religated using T4 DNA ligase. Derivatives were screened for the absence of the 218 bp BamH1 fragment which contained the REP sequences. pWJ1 61 is one such derivative.The galK gene sequence includes a 189 bp fragment between BC71 sites indicated in the diagram of pWJ62 and the cat gene sequence includes a 535 bp EcoR1 -Sall fragment indicated in the diagram of pWJ61.

Cells containing pWJ61 or pWJ62 were pulse-labelled for 3 minutes with 3H-uridine (1 mCi ml-') and RNA isolated according to the procedures of Peck and Wang (Celi 40, 129-137; 1985). RNA from 30ml cells was resuspended in 900 ul of hybridisation buffer (50% formamide; 5 x SSC; 50mM sodium phosphate, pH 6.5; 0.1% SDS; 0.02% Ficoll; 0.2% polyvinyl pyrollidone; 0.02% bovine serum albumin) and 150 ul of this used per hybridisation. Hybridisations were carried out by a modification of methods described elsewhere (Hansen and Hatfield, Proc. Natl. Acad. Sci.USA. 81,76-79; 1984). 6 pg of single-stranded DNA of an appropriate probe cloned into an M13 vector was spotted onto a nitrocellulose filter in 10 x SSC and fixed to the filter by baking for 2 hours. Each filter was prehybridised for 42"C for 6 hours in hybridisation buffer containing 40 ug mI-1 sonicated salmon-sperm DNA. The tritiated RNA was added and hybridisation carried out at 42"C for 16 hours.

The filters were then washed in 0.3M NaCI: 0.03M sodium citrate: 1 OmM magnesium chloride and incubated for 30 min at room-temperature in the same solution containing 10 ug mI-1 RNaseA. The filters were washed five times in 2 x SSC: 0.1% SDS, dried, and the amount of tritiated RNA retained by the filter determined by scintillation counting. Two DNA fragments cloned into M13 were used as probes for these experiments. The upstream probe was a 535bp EcoR1-Sall fragment from the cat gene; the downstream probe was a 289bp BC1l fragment from galK (Figure 7).

In addition to measuring mRNA levels, the protein product of the cat gene was also assayed by standard methods as described by Shaw W.V. (Methods in Enzymol. 43, 736; 1975).

Briefly, cells were grown to an OD600 of about 0.5, the cells pelleted by centrifugation and resuspended in 0.5 ml of extraction buffer (50 mM Tris-CI pH 7.8, 30 I1M dithiothreitol). The cells were broken by sonication and the debris removed by centrifugation. CAT activity in the supernatant was determined by incubating with acetyl-CoA (0.1 mM) and 5,5'-dithiobis-Z-nitrobenzoic acid (DTNB; 0.4 mg my~'). To this mix chloramphenicol (0.1 mM) was added and the rate of increase in absorption at 412 nm determined as a measure of CAT activity.

The generation of product absorbing at 412 nm is due to the reduction of DTNB by reduced CoA which is generated during the acetylation of chloramphenicol.

The results of these experiments are detailed in Table 1 below. Clearly, the REP sequence causes about a ten-fold increase in the amount of upstream mRNA but has little effect on the amount of downstream RNA. This is reflected in an increased amount of upstream (CAT) gene product synthesised from the plasmid containing the REP sequence.

TABLE 1 REP Upstream Downstream CAT Protein Plasmid Sequence RNA RNA Activity pCH61 absent 793 185 100 pCH62 present 10188 192 490 Notes: The amount of RNA is expressed in arbitrary units determined from the 3H retained by the filters after adjusting for background retention, the probe size and plasmid copy number.

CAT activity is expressed as a percentage of the amount of enzyme synthesised from the parental plasmid pWJ 161.

Example 5 serves to illustrate the increased production of protein which can be obtained by introducing a REP sequence in an appropriate position downstream of DNA encoding the desired protein. Clearly this technique can be utilised to great advantage in the production of protein on a commercial basis.

Plasmids containing REP sequences bounded on each side by an endonuclease restriction site have been utilised in the study of the REP sequences themselves. For use as a commercial tool e.g. in the production of proteins, a plasmid containing a REP sequence would normally require sites for a plurality of endonucleases at least on one side and preferably on both sides of the REP sequence. Such plasmids are considered novel and within the scope of the invention.

Claims (4)

1. A method of increasing the half-life of an mRNA sequence transcribed from DNA which comprises utilising recombinant DNA having a REP sequence immediately downstream (3') of the DNA encoding sequence corresponding to the said mRNA sequence.

2. A method of stimulating the production of a protein or sequence thereof in which the encoding mRNA is caused to accumulate utilising a method according to claim 1.

3. A plasmid containing the DNA sequence: tS]x.J.REP.K. [S]y where S signifies a unique endonuclease restriction site, x andy are each at least 1, x + y is greater than 2, and REP indicates a REP sequence.

4. Use of a plasmid containing a REP sequence bounded by endonuclease restriction sites as a tool in the selective accumulation of mRNA.