IMPROVED DETECTION METHOD FOR PREMATURE TERMINATION MUTATIONS INTRODUCTION AND BACKGROUND TO INVENTION
This invention relates to diagnostic medicine and in particular, to the diagnosis of medical conditions characterized by the presence of genetic mutations involving premature termination codons.
Premature termination codons are well documented and known to decrease the detectable levels of most mRNA's. The decrease in detectable levels of mRNA renders the direct measurement of such a gene mutation difficult or impossible.
To date, one method has been developed to scan for premature termination mutations and is known as the Protein Truncation Test (PTT). In the PTT, either genomic DNA or cDNA generated by reverse transcription, is PCR-amplified using forward primers containing a T7-promoter sequence and translation initiation signals. Proteins are produced from the amplified products using an in vitro T7-driven coupled transcription and translation system and the mutant truncated proteins discriminated from the normal allele product by electrophoresis. Direct PCR-amplification from genes has been useful in cases where the protein is encoded by few exons, or where mutation hot-spots are known, but such an approach is impractical for scanning large multi-exon genes. By focusing solely on the coding regions of the gene by Reverse Transcriptase Polymerase Chain Reaction (RT-PCR), the complexity of analysis is reduced but success can be compromised by low levels of mutant mRNA due to destabilization caused by premature translation termination. The RNA-based PTT has been of considerable use in the definition of translation-termination mutations in many important diseases, including breast and ovarian cancer (BRCA1); familial adenomatous polyposis; hereditary nonpolyposis colorectal cancer; Duchenne muscular dystrophy, and neurofibromatosis 1. However, a significant proportion of these mutations are undetected or are difficult to detect by RT-PCR-PTT, most likely due to
nonsense-mediated mRNA decay reducing the mutant allele transcripts to levels below the sensitivity of the current RT-PCR approaches.
Furthermore, osteogenesis imperfecta (01) type I, an autosomal dominant brittle-bone disease, is commonly due to premature termination mutations in the gene for the type I collagen αl(I) chain (COLlAl), which result in mRNA instability and type I collagen haploinsufficiency.
It would be of considerable diagnostic power if a direct method of detecting for the presence of premature termination mutations could be devised which prevented or traversed the mRNA decay resulting from the presence of premature termination mutations.
It has been unexpectedly found that the use of protein synthesis inhibitors prior to RNA extraction serves to ameliorate nonsense-mediated mRNA decay. Such a finding suggested that a combination of such inhibition with an RNA based protein truncation test would provide a method allowing for the direct detection of premature termination codons and overcome many of the problems of the prior art methods discussed above including the problems of mutation induced mRNA instability thereby greatly increasing the PTT signal strength and the sensitivity and reliability of RT-PCR/PTT for the detection of such premature termination mutations. Furthermore, the invention can be applied to living cells of specific origin to the disease site for example, skin cells, or equally applied to living cells of non-specific origin including low level or illegitimate premature termination containing transcripts from blood derived cells such as lymphoblasts. The methods of the invention therefore provide very broad applicability over a wide range of diseases, particularly in the field of cancer mutation detection, as many such mutations result in truncated proteins. STATEMENT OF INVENTION
Accordingly, in one aspect the invention provides a method of detecting the presence of a gene mutation in an organism including the steps of: - providing a representative sample of cells from said organism;
incubating said sample with a protein synthesis inhibitor; extracting RNA from said sample; testing for the presence of said gene mutation in said extracted
RNA. The cells are preferably viable and capable of metabolism with the representative sample of cells derived from a specific origin of affected tissue or derived from A non-specific origin, including blood. The sample is preferably derived from lymphocytes in the form of transformed lymphoblasts.
The gene mutation is preferably due to premature termination mutations which result in destabilized mutant mRNA and may include dominant-negative mutations and excluded mutations.
The sample is preferably incubated with a protein synthesis inhibitor adapted to arrest the degradation of said destabilized mutant mRNA. The protein synthesis inhibitor can be chosen from any appropriate inhibitor; for example, puromycin, anisomycin, emetine, pactomycin or cycloheximide but is most preferably cycloheximide.
The incubation of said sample preferably occurs for a period of up to about 8 hours prior to RNA extraction with the period depending on the nature of mutation in question. The testing of extracted RNA may be as a protein truncation test.
In another aspect, the invention provides a method for detecting translation termination mutations in a patient having any one of breast cancer; ovarian cancer (BRCA1); familial adenomatous polyposis; hereditary nonpolyposis colorectal cancer; Duchenne muscular dystrophy; neurofibromatosis 1; osteogenesis imperfecta; Bethlem myopathy or Stickler syndrome. comprising the steps of: providing a representative cell sample, most preferably a lymphoblast sample, from a patient having any one of breast cancer; ovarian cancer (BRCA1); familial adenomatous polyposis; hereditary
nonpolyposis colorectal cancer; Duchenne muscular dystrophy; neurofibromatosis 1 ; osteogenesis imperfecta; Bethlem myopathy or Stickler syndrome. incubating said sample with a protein synthesis inhibitor; - extracting RNA from said sample; and testing for the presence of said gene mutation in said extracted RNA. DETAILED DESCRIPTION OF THE INVENTION
COLlAl mutations in 01 type I were used as a convenient model system to develop a RNA-based PTT which overcomes the problem of the nonsense- mediated mRNA decay by using the novel modification of pre-incubation of the fibroblasts with cycloheximide to stabilize the mutant mRNA prior to mRNA isolation and PTT.
Four 01 type I patients with characterized premature termination mutations in COLlAl were studied (Table 1). Fibroblasts from these patients had approximately half normal levels of pro l(I) mRNA indicating that the mutant mRNA was unstable and degraded. RT-PCR was performed on fibroblast total RNA using overlapping T7-primer sets spanning the regions containing the termination mutation (Table 1). In vitro translation of RT-PCR products from patient F4 and control fibroblasts are shown in Figure la. The
T7-cDNA of 1870 bp produced a protein of 70 kDa in the control (lane 1) and also from the normal allele mRNA in the patient (Figure la, lane 2). There was also a band corresponding to the predicted 53 kDa product from the mutant allele, however, the appearance of this mutant truncated protein band was variable and in most experiments, was not able to be detected (Figure lb, lane
1 and Figure 2a, lane 2). Similarly, RT-PCR followed by PTT of the other 3 patients (Figure 2a, lane 4; Figure 2b, lanes 2 and 4) also showed that the mutant allele product was vastly under-represented, or undetectable, rendering the conventional RT-PCR-PTT unreliable for screening due to the instability of the mutant mRNA.
The molecular basis of how mRNA destabilization is triggered by premature translation termination is not fully understood, and four models have been proposed; decay during co-translational nuclear export, cytoplasmic decay, nuclear scanning, and a cytonuclear feedback mechanism. Recent data lends support to the nuclear scanning model which proposes that nonsense-mediated mRNA decay is primarily a nuclear event which requires a spliceable intron downstream from the premature termination codon. EXAMPLES EXAMPLE ONE To test if these inhibitors could stabilize COLlAl mRNA with internal nonsense codons, cells were incubated with cycloheximide for up to 8 hours before RNA extraction and RT-PCR-PTT (Figure lb). Without incubation with cycloheximide (lane 1), no mutant truncated (53kDa) collagen protein was apparent, but with increasing time of pre-incubation with cycloheximide (lanes 2-5), protection of the mutant mRNA was evident, with the appearance of significant quantities of the truncated 53kDa band a 4-8 hours of preincubation (lanes 4, 5). In fact, quantification of the 35S-methionine-labelled control allele 70kDa and mutant allele 53kDa bands, and making allowance for the reduced number of methionines present in the truncated protein, indicated that the mutant and normal bands were approximately equally expressed, demonstrating the mutant mRNA was protected by cycloheximide treatment. Concentrations of cycloheximide from 10-1000 μg/ml were effective in protecting the mRNA (data not shown). EXAMPLE TWO To confirm the general applicability of the method, the effect of preincubation with cycloheximide was tested on four COLlAl cell strains in which premature termination mutations arose from point mutations or frameshift mutations (F2, F6, F7 and F8). The data shown in Figure 2 consistently demonstrate that in the absence of cycloheximide, the predicted truncated protein bands (Table 1) were not readily detected by the in vitro transcription
and translation (Figure 2a, lanes 2 and 4; Figure 2b, lanes 2 and 4), indicating that substantial nonsense-mediated mRNA decay occurred in all cases. Again it is important to emphasize that the level of mutant in the absence of cycloheximide was variable between experiments, but most commonly was undetectable. However, in all four cases, pre-incubation with 100 μg/ml cycloheximide for 8 hours prior to RNA extraction resulted in mutant mRNA protection and the mutant truncated protein bands were clearly evident (Figure 2a, lanes 3 and 5; Figure 2b, lanes 3 and 5). Quantification of the truncated protein bands, allowing for the differential 35S-methionine content, demonstrated a high level of protection of mutant mRNA by cycloheximide in all cases. EXAMPLE THREE
In many circumstances, it is not possible to have access to cell types that readily express the gene of interest. To test if the use of protein synthesis inhibitors would also stabilise the mutant "illegitimate" transcripts [18-21] produced by accessible cells, we examined transformed lymphoblasts from patients with a mild form of Bethlem myopathy, the premature termination mutation was in the type VI collagen COL6A1 gene (Lamande et al., unpublished data). Transformed lymphoblasts from patients were preincubated with cycloheximide for up to 4 hours before RT-PCR/PTT (Figure 3). Mutant type VI collagen mRNA produced by lymphocytes was also stabilised by cycloheximide treatment (Figure 3). With this mutation, like those of type I collagen in OI fibroblasts (Figures 1 and 2), the mRNA containing the premature termination mutation was unstable and not detectable in RT-PCR/PTT without cycloheximide treatment (Figure 3, lane 1). When preincubated with cycloheximide for 4 hours (Figure 3, lane 3), the mutant message was protected and the ratio of unncated protein to normal increased more than 20-fold. These data clearly show that even during low level illegitimate transcription, nonsense- mediated mRNA decay occurs and can be inhibited by incubation with cycloheximide.
EXAMPLE FOUR
A further example of the use of the method of the invention to stabilize mutant "illegitimate" transcripts produced by accessible cells was performed. Skin fibroblasts from patients with Stickler syndrome were used, where we suspected the premature termination mutation was in the type II collagen gene COL2A1. Fibroblasts from the patient were preincubated with cycloheximide for up to 8 hours before RT-PCR/PTT (Figure 4). Mutant type II collagen mRNA was stabilised by cycloheximide treatment, and we were able to detect the presence of a previously uncharacterised premature termination mutation in the patient which was then confirmed by cloning and sequencing the mutant RT-PCR product. Again, the mutation was not detected by RT-PCR/PTT without cycloheximide treatment. FULL DESCRIPTION - METHODS RNA isolation and RT-PCR Dermal fibroblast cultures were established from the patients and controls were maintained as described previously. Total RNA was extracted from confluent fibroblasts in 60cm2 dishes using RNeasy Total RNA kit (Qiagen) according to the manufacturer's recommendations. RNA was eluted in DEPC- treated water and stored at -70°C. In experiments to test the effect of cycloheximide (Sigma) on mutant mRNA stability, the medium was replaced with fresh medium containing varying concentrations of cycloheximide and the cells were incubated for a further 2 to 8 hours prior to the extraction of RNA. cDNA was synthesised from 200-500 ng total RNA (fibroblast derived) or 5 μg (lymphoblast derived) per 20 μl reaction, using oligo-d(T)16 priming and MuLV reverse transcriptase (GeneAmp RNA PCR kit, Perkin Elmer) according to the manufacturer's instructions for 60 minutes at 42°C. Primer sets were designed to cover the region of COLlAl mutations (Table 1). The 5' primer of each set contained, in addition to sequence complementary to COLlAl, the T7 promoter sequence and the eukaryotic initiation sequence followed by an initiating methionine. For patients F4 and F8 the 5' primer (T7-C1) was complementary
to base pairs 2088-2107 and the 3' primer (C2) to base pairs 3939-3958. For F6 and F7 the 5' primer (T7-D1) was complementary to base pair 3057-3076 and the 3' primer (D2) to base pairs 4616-4645. Each PCR reaction was undertaken in a 50 μl reaction volume with 100 ng of the appropriate forward and reverse primers, 2mM MgCl2, 2.5 Units Tag polymerase (Amplitaq, Perkin Elmer Cetus). Each cycle consisted of an initial denaturation at 94°C for 2 minutes, followed by 35 cycles of denaturation at 94°C for 30 seconds, annealing at 60°C for 1 minute and extension at 72°C for 1 minute. Following PCR, cDNA products were analysed on 1.5% agarose gels. For illegitimate transcript amplification of type VI collagen al(VI) cDNA bases 508-1235 were amplified using primers 5'-T7-ATTGTGGTGACCGACGGGCA-3' and 5'- GGCCCCTCGTCTCCAGATGG-3'. Each 50 μl PCR was performed with 300 ng of primers, 3mM MgCl2, 2.5 Units Taq polymerase, for 40 cycles of denaturation at 95 °C for 1 minute, annealing at 64°C for 1 minute and extension at 72°C for 2 minutes. For illegitimate transcript amplification of type II collagen, αl(II) cDNA bases 937-1974 were amplified using primers 5'-T7- GTGAAAGAGGACGGACTGGC-3' and 5'-TCACCTGGTTTTCCACCTTC-3'. Each 50μl PCR was performed with lOOng of primers, 2mM MgCl2, 2.5 Units Taq polymerase, for 40 cycles of denaturation at 95°C for 1 minute, annealing at 62°C for 1 minute and extension at 72°C for 2 minutes. Protein truncation test
4 μl of the T7-modified RT-PCR products were added directly, without purification to T7 polymerase-coupled transcription and translation system (TNT™ Coupled Reticulocyte Lysate, Promega), as previously described in a reaction volume of 12.5 μl. Transcription and translation was performed at 30°C for 90 minutes, and the translated protein products labelled with 10 μCi of translation grade L-[35S] methionine (1000 Ci/mmol, DuPont NEN). At the completion of the translation 3 μl aliquots of each reaction were mixed with gel sample buffer (2% SDS, lOmM DTT), denatured at 65°C for 10 minutes and analysed on 14% (w/v) SDS/polyacrylamide gels. Radioactively labelled protein
bands were detected by fluorography or imaged and quantified using a phosphorimager (Molecular Dynamics, STORM™).
The method described here overcomes the problem of mutation-induced mRNA instability and has universal applicability, greatly increasing the PTT signal strength and thus the sensitivity and reliability of RT-PCR-PTT for the detection of premature termination mutations. The stabilisation of mutant unstable mRNA is not specific to mRNAs expressed at high level, such as collagen mRNA transcripts in fibroblasts, but very low level mutant mRNAs such as illegitimate transcripts produced by lymphoblasts are also stabilised by cycloheximide .
The rapid and reliable detection of premature termination mutations in type I collagen in patients with osteogenesis imperfecta type I is indicative of the importance of the invention. Collagen mutations in osteogenesis imperfecta, and indeed all other collagenopathies, fall into two general categories, included dominant-negative mutations or excluded mutations. Dominant-negative structural mutations, commonly including missense mutations which compromise the structure of the collagen triple helix, have a major effect on collagen molecular assembly and disturb matrix integrity leading to a clinically severe phenotype. On the other hand excluded mutations such as nonsense mutations which result in a functionally-null mutant allele, cause haploinsufficiency. This reduced amount of collagen, which nonetheless is structurally-normal, results in a much milder osteogenesis imperfecta phenotype. It is thus important to be able to discriminate between these two patient groups since those patients with nonsense mutations are likely to respond better to therapeutic strategies which upregulate collagen synthesis, while those with dominant negative mutations would not benefit from such strategies, which may even exacerbate their matrix structural defects.
Figure Legends:
Fig. 1 SDS-PAGE analysis of COLlAl fibroblast mRNA PTT products.
RT-PCR was performed using primer set T7-C1/C2 (see Methods), spanning proαl(I) mRNA base pair 2088-3958 (Table 1). a, control (lane 1) and patient F4 (lane 2). b, effect of pre-incubation of patient F4 fibroblasts with 100 μg/ml cycloheximide prior to mRNA isolation. No cycloheximide (lane 1), preincubation with cycloheximide for 2 hours (lane 2), 4 hours (lane 3), 6 hours (lane 4) and 8 hours (lane 5). The full-length protein product produced by in vitro transcription/translation of the RT-PCR product (70 kD) and the predicted mutant truncated protein (53 kD) are indicated.
Fig. 2 Effect of cycloheximide on PTT analysis of four COLlAl premature termination mutations, a, RT-PCR was performed using primer set T7-C1/C2 (see Methods), spanning proαl(I) mRNA base pair 2088-3958 (Table 1). Control (lane 1), patient F4 without (lane 2) and with preincubation with cycloheximide (lane 3), patient F8 without (lane 4) and with preincubation with cycloheximide (lane 5). The full-length protein product produced by in vitro transcription/translation of the RT-PCR product (70 kD) and the predicted mutant truncated proteins (F4, 53 kD; F8, 48 kD) are indicated, b, RT-PCR was performed using primer set T7-D1/D2 (see Methods), spanning proαl(I) mRNA base pair 3057-4645 (Table 1). Control (lane 1), patient F6 without (lane 2) and with preincubation with cycloheximide (lane 3), patient F7 without (lane 4) and with preincubation with cycloheximide (lane 5). The full-length protein product produced by in vitro transcription/translation of the RT-PCR product (53 kD) and the predicted mutant truncated proteins (F6, 26 kD; F7, 28 kD) are indicated.
Fig 3. Effect of cycloheximide on RT-PCR/PTT of lymphoblastoid cells from a patient with a premature termination mutation causing Bethlem myopathy. RT-PCR was performed on Bethlem myopathy lymphoblast total RNA using a T7-primer set spanning the mutation in the type VI collagen gene coding region (see Methods). Lane 1, RT-PCR/PTT of RNA from cells without preincubation with cycloheximide (CHX); Lane 2-3, cells preincubated for 1 hour (lane 2) or 4 hours (lane 3) with cycloheximide. The full-length type VI collagen protein product produced by in vitro transcription/translation of the RT- PCR product (27 kDa) and the predicted mutant truncated protein (17 kDa) are indicated.
Fig 4. Effect of cycloheximide on RT-PCR/PTT of fibroblasts from a patient with a premature termination mutation causing Stickler syndrome. RT- PCR was performed on Stickler syndrome fibroblast total RNA using a T7- primer set spanning the mutation in the type II collagen coding region (see Methods) . Lane 1 , RT-PCR/PTT of RNA from cells without preincubation with cycloheximide (CHX); Lane 2-3, cells preincubated for 4 hours (lane 2) or 8 hours (lane 3) with cycloheximide. The full-length type II collagen protein product produced by in vitro transcription/translation of the RT-PCR product (47 kDa) and the mutant truncated protein (35 kDa) are indicated.
TABLE 1
COLAIA premature termination mutation and predicted truncated protein product
Footnotes: a designates the base pair position in the proαl(I) coding sequence, numbered from the transcription start site [28].
b indicates site of the premature termination codon relative to site of the mutation.