EP0973946A1 - Biological assay for testing the carcinogenic properties of a substance - Google Patents

Abstract

An assay for testing the carcinogenic properties of a test substance comprising: (i) introducing into cells a reporter gene expression vector comprising a repetitive DNA sequence which exhibits instability in cancer cells, whereby instability of the repetitive DNA sequence affects expression of the reporter gene; (ii) exposing the resulting cells to the test substance; and (iii) determining whether the test substance is carcinogenic or anticarcinogenic by comparing the frequency of reporter gene expression in the resulting cells with the frequency of reporter gene expression in cells which have not been exposed to the test substance. The invention also provides expression vectors suitable for use in the assay methods.

Description

BIOLOGICAL ASSAY FOR TESTING THE CARCINOGENIC PROPERTIES OF A SUBSTANCE

The present invention relates to a biological assay and assay reagents for testing the carcinogenic properties of a test substance. The assay is useful for screening potential anti-cancer drugs as well as for testing the carcinogenic properties of food components.

DNA repair is an essential process in all organisms from prokaryotes to eukaryotes. Defective DNA repair in higher eukaryotes such as humans is an important factor in the aetiology of both hereditary and sporadic carcinomas. According to the current model of carcinogenesis, initiation is a single-cell event which leads to the development of a precancerous lesion by clonal expansion. Progression to an invasive tumour is a prolonged process requiring the acquisition of several further mutations in genes controlling cell proliferation and differentiation. As the spontaneous mutation rate in normal cells is relatively low it suggests that an early event in the pathway of tumourgenesis is a mutation that confers a so- called "mutator" phenotype. This postulates that defects in DNA repair occur at an early stage in the sequence of events and favour accelerated tumour progression. This hypothesis is corroborated by the recent link between cancer, microsatellite instability and mutations in the genes encoding the mismatch repair machinery. Mismatch repair involves the processing of incorrectly paired nucleotides which can occur as a result of normal DNA metabolism. It also plays an important role in the recognition and correction of unpaired loop structures which form during the replication of highly repetitive regions of DNA known as microsatellite DNA. If left unrepaired these loop-structures can lead to frame-slippage and potential loss of gene function. Hereditary non-polyposis cancer (HNPCC) is a cancer of the colon characterised by microsatellite instability. HNPCC is an autosomal dominant disease in which multiple members of a family suffer early onset colon cancer in the absence of polyp formation. Of the HNPCC tumours that exhibit microsatellite instability over 50% contain mutations in the HNPCC-linked hMSH2 gene on chromosome 2 and about 20-30% contain mutations in the HNPCC-linked hMLHl gen on chromosome 3 (Umar, A. & Kunkel, T.A., 1996).

HNPCC-Lynch syndromes I and II is a common cancer predisposition syndrome that is autosomal dominant in nature. Lynch I families suffer early onset colorectal cancer, while Lynch II kindreds are also susceptible to extra colonic epithelial tumours of the endometrium, ovary, stomach, small intestine, kidney and ureter. The link between microsatellite instability and cancer is demonstrated by the fact that a subset of sporadic colon cancers and the majority of tumours occurring in HNPCC patients contain frequent mutations in the simple microsatellite sequences (A)n, (GGC)n, or (CA)n. These mutations seem to be tumour specific with each cell containing thousands of microsatellite mutations. Studies involving microsatellite instability show that it is present in a significant number of sporadic tumours including colorectal (12-28%), endometrial (17-23%), stomach (18-39%), ovarian (16%), cervical (15%), pancreatic (67%), oesophageal adenoma (22%), squamous cell skin (50%), and small-cell lung cancer (45%) (Eshleman, J.R. & Markowitz, S.D. 1995; Aaltonen et al. 1993; Merlo et al. 1994; Mironov et al. 1994; Orth et al. 1994; Modrich, M. & Lahue, R. 1996). Although not fully understood, the mismatch repair pathways of both lower and higher eukaryotes share extensive homology. For example, homologues of the human mismatch repair pathway exist in Saccharomyces cerevisiae. Henderson and Petes, Mol. and Cell. Biol. , June 1992, 12, No. 6, p.2749- 2757 have constructed reporter gene expression vectors for studying spontaneous frameshift mutations. The vectors are based on a plasmid having the LEU2 promoter and the first 12 codons of the yeast LEU2 protein fused to the eighth codon of the E. coli β-galactosidase gene (lacZ). They inserted various oligonucleotides containing simple repetitive DNAs into the Bar Αl site near the beginning of the β- galactosidase gene.

The inserts did not shift the reading frame so that β-galactosidase expression occurred unless a frameshift mutation occurred in the host (yeast) cells. Such events were visible as white colonies when the cells were grown on a medium containing Xgal.

Similar reporter gene expression vectors were constructed in which the repetitive DNA tract was inserted upstream of the URA3 gene. On a medium containing 5-fluoro-orotic acid (5-FOA) frameshift mutations were detected as URA3-(5-FOA^R) cell colonies.

Levinson and Gutman (1987), Nuc. Acids Res. , 15, p.5323-5339 also used a reporter gene expression vector to study frameshift mutations in the prokaryote E. coli K12. The vector was based on bacteriophage M13 and contained short poly-CA/TG tandem repeats linked to the lacZ gene which encodes β-galactosidase. Strand et al, Nature, 365, p.274-276, 16 September 1993, used a reporter gene expression vector comprising a yeast promoter fused to a β- galactosidase gene that contained a 29-base pair out-of-frame poly (GT) tract in the coding sequence to study repetitive DNA tract instability. They found that mutations in any three yeast genes involved in DNA mismatch repair (PMS1, MLH1 and MSH2) lead to 100- to 700-fold increases in repetitive tract instability, whereas mutations that eliminate the proof reading function of DNA polymerases have little effect.

According to a first aspect of the invention there is provided an assay for testing the carcinogenic properties of a test substance comprising: (i) introducing into cells a reporter gene expression vector comprising a repetitive DNA sequence which exhibits instability in cancer cells, whereby instability of the repetitive DNA sequence affects expression of the reporter gene; (ii) exposing the resulting cells to the test substance; and (iii) determining whether the test substance is carcinogenic or anticarcinogenic by comparing the frequency of reporter gene expression in the resulting cells with the frequency of reporter gene expression in cells which have not been exposed to the test substance.

The term "carcinogenic properties" is intended to embrace the ability of the test substance to inhibit cancer as well as to cause cancer, that is, the term embraces both carcinogenic and anti-carcinogenic properties.

By "instability" we mean a change in the size of the DNA sequence, normally by additions or deletions that are not a multiple of 3 bp. Such changes in size alter the reading frame for transcription of adjacent genes and are known as frameshift mutations.

Preferably, the repetitive DNA sequence (often referred to as microsatellite DNA) comprises a poly d(AC/TG) tract and/or a poly d(GT/CA) tract, although the tract may comprise a single nucleotide eg. poly d(G) or poly d(A).

It will be appreciated that the length of the repetitive DNA tract can be varied and is preferably selected according to the length of the repetitive sequence identified as being unstable in the cancer cell of interest. However, the length of the repetitive sequence is conveniently 8 to 60 nucleotides, more preferably 16 to 32 and especially 16.

The term "reporter gene expression vector" is intended to cover any vector into which a reporter gene has been inserted so that, on introduction into a suitable host cell, the reporter gene will be transcribed and translated to produce the protein product of the reporter gene.

A skilled person will appreciate that the expression vector can be provided in a variety of forms eg. a plasmid, a 'phage or a virus.

Preferably the reporter gene expression vector comprises a promoter region of a gene which is normally expressed in the host cell fused to a sequence encoding a reporter gene product which can be expressed in the host cell. Preferably a repetitive DNA sequence which exhibits instability in cancer cells is inserted downstream of the promoter region into the open-reading frame of the reporter gene sequence. The open-reading frame (ORF) will be understood by skilled persons to mean a DNA sequence which contains a series of triplets coding for amino acids without any termination codons.

The insertion preferably "knocks" the promoter/reporter gene fusion out- of-frame (+ 1 or -1 reading frame) so that the reporter gene is not expressed. Hence expression of the reporter gene only occurs if the inserted repetitive sequence changes size (exhibits instability) so that the correct reading frame of the reporter gene is established.

Alternatively, the insertion does not knock the promoter/reporter gene fusion out of frame so that expression of the reporter gene occurs unless the inserted repetitive sequence exhibits instability so that it knocks the promoter/reporter gene out-of-frame.

It is within the knowledge of skilled persons to select a variety of promoter sequences and reporter gene sequences which can be used in a given host cell be it yeast, human or bacterial. The reporter gene expression vectors disclosed by Henderson and Petes (1992), Levinson and Gutman (1987), and Strand et al (1993) are incorporated herein by reference.

Preferably the reporter gene expression vector is provided in the form of a low or high copy number plasmid, or an integrative plasmid, that is, a plasmid which lacks a host cell origin of replication and must therefore be integrated into the host cell genome for stable maintenance in the host cells.

Preferably the reporter gene comprises a gene whose expression product gives rise to a visible change in the host cell. For example the gene product may produce a colour change or fluorescence. A particularly preferred reporter gene system comprises the lacZ gene which encodes the enzyme β-galactosidase. β-galactosidase expression can be detected as a blue colour in colonies growth on a medium containing Xgal. Colonies which do not express β-galactosidase appear as white colonies. Hence, use of the vector comprising the lacZ reporter gene according to a preferred embodiment provides a simple blue/white colour test for screening the carcinogenic effect of a test substance.

Preferably the cells used in the assay are eukaryotic cells, preferably yeast cells or human cells, and especially eukaryotic cells which have a defect in repetitive DNA instability repair mechanisms, especially the mismatch repair pathway. As homologues of the human mismatch repair pathway exist in yeast such as Saccharomyces cerevisiae this single cell eukaryote provides an ideal model for studying the effects of test substances eg. dietary constituents on DNA repeat instability in humans.

Of course human cell lines can be used directly in the assays of the invention, the cell lines being derived from the cancer of interest eg. human colorectal cancer, especially hereditary non-polyposis cancer (HNPCC). Although not preferred, the assay may use prokaryotic cells, conveniently bacterial cells such as Escherichia coli.

In a second aspect the invention provides an assay comprising testing the carcinogenic properties of a test substance using yeast cells according to the first aspect of the invention; and further testing the test substance using human cells according to the first aspect of the invention. As yeast cells are easier to work with than human cells the first step of the assay according to the second aspect of the invention provides a primary screening step which can be used to screen out test substances exhibiting undesirable carcinogenic properties before the second step whose results have greater significance in humans.

In a third aspect the invention provides an assay for testing the carcinogenic properties of a test substance comprising the steps of: (i) screening a test substance using a reporter gene expression vector according to the first aspect of the invention; and (ii) repeating the assay using cells containing a high, medium, low or single copy number vector, the high, medium, low or single copy number vector being selected depending on the frequency of repetitive DNA instability measured in step (i).

Preferably the reporter gene expression vector in step (i) is contained in d e cells at a high copy number.

The assay according to the third aspect of the invention is particularly useful when the frequency of instability in step (i) is very high in cells which have not been exposed to the test substance. By selecting medium, low or single copy number vectors the background instability can be reduced as desired to make the assay more sensitive.

Preferred non-limiting embodiments of the invention will now be described with reference to the accompanying drawings in which:

Figure 1 shows the steps of a preferred assay according to the first aspect of the invention. The assay utilises a preferred assay plasmid of the invention;

Figure 1A shows the high copy assay plasmid pKa3-9(n) as previously described. The unique EcoRI site into which the respective tracts were cloned is underlined. Also underlined is the unique Stul site situated within the URA3 gene, used for the insertion of the KanMX4 cassette (cf. text for further details).

Figure IB shows the high copy assay plasmid pKa3-9(n)KanMX4 utilising the dominant selectable marker KanMX4. Both the EcoRI site into which the respective tracts were cloned and the additional EcoRI site derived from the KanMX4 cloning step are underlined (cf. text for further details).

Figure 2 shows a yeast/Zs. coli shuttle vector suitable for fusing yeast promoter and coding sequences to the lacZ gene of E. coli;

Figure 2A shows the centromere-based assay plasmid pKaCEN(n) as previously described. The unique EcoRI site into which the repetitive tracts were cloned is underlined. Also underlined is the unique Støl site situated within the URA3 gene, used for the insertion of the KanMX4 cassette (cf. text for further details).

Figure 2B shows the centromere-based assay plasmid pKaCEN(n)KanMX4 utilising the dominant selectable marker KanMX4. Both the EcoRI site into which the repetitive tracts were cloned and the additional EcoRI site derived from the KanMX4 cloning step are underlined (cf. text for further details).

Figure 3 shows a low copy number vector containing sequences from yeast centromere VI;

Figure 4 shows a vector which lacks a yeast origin of replication, so that it must be integrated into the yeast genome for stable maintenance; Figure 5 shows a preferred high copy number reporter gene expression vector according to the invention;

Figure 6 shows a preferred low copy number reporter gene expression vector according to the invention;

Figure 7 shows a preferred single copy integrative reporter gene expression vector according to the invention;

Figure 8 is a schematic representation which shows integration of the vector of Figure 7 at the URA3 locus of S. cerevisiae strain YN94-1;

Figures 9 and 10 show the results of experiments to confirm insertion of the integrative vector of Figure 7 into the S. cerevisiae strain as illustrated in Figure 8;

Figure 11 shows the synthetic pathway for the polyamines putrescine, spermidine and spermine in eukaryotes;

Figure 12 is a schematic representation of frame-slippage in the preferred vectors leading to β-galactosidase reporter gene expression; Figures 13a and 13b shows a Southern blot analysis of MSH2 genomic DNA (Figure 13a) and the disruption scheme (Figure 13b);

Figure 14 shows a Southern blot analysis of wild type and disrupted MLH1 gene;

Figure 15 shows a vector incorporating the SV40 promoter upstream of the luciferase gene;

Figures 15a and 15b show Southern blot analysis of YN97-150 (msh :: KanMX4);

Figure 16 shows an EBV-based vector for stable expression of DNA in human host cells; Figures 16a and 16b show Southern blot analysis of YN97-167

(rthl :: KanMX4);

Figure 17 shows the effect of a known carcinogen on the instability of the high copy number expression vector of Example 1(1). Figures 17a and 17b show Southern blot analysis of YN98-3 (pol 130-104 : : LEU2);

Figure 18 is a schematic outlining biosynthesis of purine nucleotides, from Jones and Fink, 1982.

MATERIALS

1. PREFERRED YEAST STRAINS

YN94-1: MATa, ade2-l, his3-ll, leu2-3, 112, trpl-1, ura3-l, canl-100.

2. PREFERRED E. COLI STRAINS

DH5 : F,_80/αcZΔM15 Δ(/αcZYA-argF)U169 deoR recAl endAl yR17(r_κ-, m_κ+) phoA supΕAA γ- thi-l gyrA96 relAl .

JM109: el4-(McrA-) rec Al endAl gyr A96 thi-l hdsRll (r_κ-m_κ + ) supΕAΛ rel Al A(lac-proAB) [F' trøD36 pro AB

RR1: F-, hsdS20 (r_B-, m_B-), supE44, araU, proAl, rpsL20 (str), syl-5, mlt-5, supE44, γ-.

3 PREFERRED MEDIA

SCD minimal/defined medium

20g glucose

1.7g yeast nitrogen based - without amino acids and ammonium sulphate (Difco, Detroit, USA)

5g ammonium sulphate

25ml "drop-out" mix

20g agar (if plates are required) Made up to 1 litre with glass dist. H₂0 and autoclaved at 121 °C for 15 minutes.

"Drop-out" mix (40x):

0.2g uracil 0.4g lysine O. lg adenine

0.2g arginine O. lg methionine 0.4g tyrosine

O. lg histidine 0.6g phenylalanine 0.4g tryptophan

0.6g isoleucine 0.5g threonine 0.6g leucine

"Drop-out mix" lacking uracil (SCD-U) was used for selection of assay plasmids. Food components at various concentrations were added aseptically before the plates were poured.

LB (Luria-Bertani) Medium (for growth of E. coli)

lOg Bacto-tryptone (Difco)

5g Bacto-yeast extract (Difco) lOg NaCl

15g Agar (if plates are required)

Make up to 1 litre with glass dist. H₂O, Adjust pH to 7.5 with NaOH and autoclave at 121 °C for 15 minutes.

For Amp selection, ampicillin is added to a final cone, of lOOμg/ml when the media has cooled to approximately 55 °C.

4. PREFERRED REPETITIVE OLIGONUCLEOTIDES USED IN THE CONSTRUCTION OF ASSAY PLASMIDS

(purchased from Oswel DNA service, Southampton)

Oligonucleotide 1

(47 nt, poly d(TG)₁₆, with an EcoRI restriction site overhang at its 5' end and a Smal site, indicated by underlining.) Concentration.1413mg/ml T_m 0.1 M Na+ 74.1 °C

Smal I 5 ' -AATTGCCCGGGCTGTGTGTGTGTGTGTGTGTGTGTGTG TGTGTGCCG-3'

Oligonucleotide 2

(47 nt, poly d(AC)₁₆, with an EcoRI restriction site overhang (underlined) at its 3' end). Concentration 778mg/ml

T_m 0.1 M Na+ 74.1 °C

5'-CGGGCCCGACACACACACACACACACACACACACACACAC GGCTTAA-3'

5. PREFERRED PRIMER SEQUENCES

LacZ Reverse: ₅₂₉ 5'-AAGGGGGATGTGCTGCAAGG-3' ₅₀₉ YEp Forward: ₈₂₇₁ 5'-GCAGCGAGTCAGTGAGCGAGG-3 ₈₂₉₁

The numbers in subscript indicate the positions of annealing on plasmid

YEp356R (cf. Fig. 2).

PRS3 Forward: -₅₀ 5 '-GATTCATACTCTTTTTTCTACG -3 ' _₈₁

The numbers in subscript indicate the positions of annealing relative to the ATG start codon of PRS3.

Repeat tracts of varying sequence and length were synthesized and inserted into the unique EcoRI site of pSS3-9 as previously described.

Poly (AC)₁₅A assay tract

(shifts PRS3' LacZ fusion into + 1 reading frame with respect to start ATG of RSi)

Oligonucleotide 1

(poly d(AC)₁₅A, with an EcoRI restiction site overhang at its 5 '-end and a Smal site at its 3 '-end bodi indicated by underlining, giving a total length of 46 nts).

5'-AATTCGGACACACACACACACACACACACACACACACAGCCC GGGC-3'

Oligonucleotide 2 (poly d(TG)₁₅T, with an EcoRI restiction site overhang at its 5 '-end and a Smal site at its 3 '-end botibi indicated by underlining, giving a total length of 46 nts).

5 ' - AAT GCCCGGGCTGTGTGTGTGTGTGTGTGTGTGTGTGTGTGT CCG-3 '

Poly (Cχ.₃ A) telomeric assay tract

In Saccharomyces cerevisiae the sequence poly (C^A) is found at or near the telomeres (Shampay et al. 1984). To examine the stability of these sequences a 65 bp telomeric tract was inserted into the EcoRI site of plasmid pSS3-9 as described.

(insertion shifts PRS3 ' LacZ fusion into -1 reading frame with respect to start ATG)

Telomeric oligonucleotide 1

(poly d(C_1-3A), with an EcoRI restriction site overhang at its 5 '-end and a Smal site at its 3 '-end both indicated by underlining, giving a total length of 80 nts).

5 ' -AAT CGGACCACACCACC AC ACCCACACC AC ACC AC ACACC CACACACACACCACCACCCACACACCACACAGCCCGGGC-3 '

Telomeric oligonucleotide 2 (80 nt, poly d(Gj-₃T), with an EcoRI restriction site overhang at its its 5'- end and a Smal site at its 3 '-end both indicated by underlining, giving a total length of 80 nts).

5'-AATTGCCCCiGGCTGTGTGGTGTGTGGGTGGTGGTGTGTGTG TGGGTGTGTGGTGTGGTGTGGGTGTGGTGGTGTGGTCCG-3'

Poly (A)₂₀ tract

(shifts PRS3 ' ILacZ fusion into -1 reading frame with respect to start ATG)

Oligonucleotide 1

(poly d(A)₂₀, with an EcoRI restriction site overhang at its 5 '-end and a Smal site at its 3 '-end both indicated by underlining, giving a total length of 35 nts).

' -AAΓTCGGA AAAAAAAAAAAAAAAAAAAGCCCGGGC-3 '

Oligonucleotide 2

(35 nt, poly d(T)₂₀, with an EcoRI restriction site overhang at its its 5 '-end and a Smal site at its 3' -end both indicated by underlining, giving a total length of 35 nts).

5 ' - AATTGCCCGGGCTTTTTTTTTTTTTTTTTTTTCCG-3 ' During the cloning of repetitive tracts into pSS3-9 a number of aberrant events occurred resulting in assay plasmids containing repetitive tracts as follows. Each of these tracts were subcloned into the centromeric and integrative vectors as previously described and incorporated into the assay system.

tract Reading frame

(AC)₁₆ - 1

(AC)₄₁(A)₂(C)₂(AC)₄A - 1 (AC)_n + 1

(AC)₁₂A + 1

(AC)₁₅A + 1

Construction of assay plasmids utilising the dominant selectable marker KanMX4.

Loss of growth selection may be encountered when yeast strains containing the plasmid-borne URA3 marker are grown on a non-selective medium. This can occur when whole food extract containing growth- sustainable amounts of uracil are added to the medium. This problem can be overcome using a plasmid-borne KanMX4 dominant selectable marker consisting of the E. coli transposon Tn903 fused to the transcriptional control sequences of the TEF gene of the filamentous fungus Ashbya gossypii. This hybrid molecule permits the efficient selection of yeast transformants resistant to geneticin (G418) (Wach et al. 1994).

Assay plasmids as described above containing the URA3 gene (cf Figs. 1A and 2A) were linearised by digestion at die unique Stul restriction enzyme site (Promega), (+436 relative to the URA3 start codon). 40 ng of this linearised vector was ligated to 200 ng of a gel-purified 1481 bp EcoRV/Smal fragment from pFA6-KanMX4 (Wach et al. 1994) containing the KanMX4 cassette. An aliquot of this ligation mixture was then transformed into the E. coli strain RRl and colonies growing on kanamycin were selected. Plasmid DNA was isolated from several kan- resistant colonies and subjected to restriction analysis. (The insertion of the KanMX4 cassette introduces a second EcoRI site into the assay plasmids as illustrated by comparing Figure 1A with IB and 2A widi 2B.)

When assayed for repeat instability in S. cerevisiae these G418 selectable plasmids gave similar frequencies of frame-slippage to that measured for the original constructs relying on uracil selection, (cf. Table la).

CONSTRUCTION OF MISMATCH REPAIR MUTANTS

(a) MSH2

The MSH2 gene of Saccharomyces cerevisiae (cf. Fig. 1) is one of several genes that share extensive homology with the bacterial MutS gene. Located on chromosome XV it encodes a protein of 109kda. Like the MutS protein, the MSH2p binds selectively to DNA containing mispairs and substrates containing up to 14 extra bases. Strains that contain mutated MSH2 genes have strongly elevated rates of spontaneous mutations and exhibit microsatellite instability (cf. Table 1).

Disruption of MSH2 The MSH2 gene of Saccharomyces cerevisiae TN94-1 was disrupted with LEU2 (Fig. 13). Plasmid pRhB113 (Rhona Borts, Yeast Genetics, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford 0X3 9DU) containing die MSH2 gene disrupted with LEU2 at the SnaBI site was digested to completion with restriction enzyme Spel in buffer REactl (cf. Materials and Methods). The digestion mixture was transformed directly into S. cerevisiae YN904-1 (cf. Materials and Methods). Resulting transformants were then screened by Southern hybridization for die presence of the disrupted MSH2 gene (Figures 13a and b).

Figure 13a. Southern blot analysis of YN97-10 (msh2::LEU2)

Southern blot analysis of MSH2 genomic DNA (Fig. 13a) and corresponding disruption scheme (cf. Fig. 13b). Lane (1) Hindlll digested control YN94-1 DNA. Two bands are seen (1650 and 2510 bp corresponding to die wild type MSH2 gene on chromosome 15). Lane (2) Hwdlll digestion of genomic DNA containing disrupted MSH2. The gel was probed with a 1662 bp EcoRI fragment derived from pRhB113. A shift in the 2510 bp band is seen to 4710 bp corresponding to me insertion of LEU2. The disrupted msh2: :LEU2 strain was then assayed for the mutator phenotype (cf. Table 1).

(b) MLHl

The MLHl gene of Saccharomyces cerevisiae is one of several genes that shares extensive homology with the bacterial MutL gene. It is believed d at MLHlp forms an interaction with MSH2p during die initiation of DNA mismatch repair in yeast. Yeast strains that contain mutated MLHl genes have strongly elevated rates of spontaneous mutations and also exhibit microsatellite instability (cf. Table 1). The MLHl gene of Saccharomyces cerevisiae YN94-1 was disrupted with LEU2. Plasmid pREdl82 (Rhona Borts, Yeast Genetics, Institute of Molecular Medine, John Radcliffe Hospital, Oxford OX3 9DU) containing the MLHl gene disrupted with LEU2 was digested to completion with the restriction enzymes SacllBam l in buffer REact 3 (cf. Materials and Methods). This digestion mixture was transformed directly into S. cerevisiae YN94-1 (cf. Materials and Methods). Resulting transformants were then screened by Southern hybridization for the presence of the disrupted MLHl gene on chromosome XIII (Fig. 14).

Construction of mismatch repair pathway mutants for use in the assay system.

(c) MSH3

The MSH3 gene of Saccharomyces cerevisiae is another gene that shares extensive homolgy with the bacterial MutL gene. MSH3p forms a heterodimer with MSH2p during the initiation of insertion/deletion mismatch repair. Yeast strains mutant for MSH3 exhibit a less profound increase in microsatellite instability as compared to MSH2 and MLHl mutant strains and have slightly elevated levels of spontaneous mutations (Strand et al. 1995).

Disruption of MSH3 The MSH3 gene of Saccharomyces cerevisiae YN94-1 was disrupted widi LEU2. pREd62 (Rhona Borts, Yeast genetics, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford OX3 9DU) containing the MSH3 gene disrupted widi LEU2 was digested to completion with the restriction enzyme Aatll (cf. Materials and Methods). Resulting transformants were then screened by Southern hybridization for the presence of the disrupted MSH3 gene on chromosome III (data not shown).

(d) MSH6

The MSH6 gene of Saccharomyces cerevisiae is a further gene that shares extensive homolgy with the bacterial MutL gene. MSH6p forms a heterodimer with MSH2p during the initiation of spontaneous base-base mismatch repair (Alani et al. 1996). Yeast strains mutant for MSH6 exhibit a less profound increase in microsatellite instability as compared to MSH3 mutant strains but have elevated levels of spontaneous mutations in comparison to MSH3 mutant strains.

Disruption of MSH6

The MSH6 gene of Saccharomyces cerevisiae YN94-1 was disrupted with KanMX4. This MSH6 disruption plasmid (Rhona Borts, Yeast genetics, Institute of Molecular Medicine, John Radcliffe Hospital, Oxford OX3 9DU) was created by cloning a 4 kb PCR fragment containing the MSH6 gene into the Srβ site of pPCR script (Sratagene). A Pvull to EcoRV fragment containing the KanMX4 module (Wach et al. 1994) was men used to replace a SnaBl to Spel fragment of the MSH6 open reading frame to create plasmid pSRC9. The msh6::KanMX4 disruption cassette was released by digestion with restriction enzymes Sphl and -β-ypEI and transformed into YN94-1 (cf. Materials and Methods). Resulting yeast transformants were then screened by Southern hybridization for the presence of the disrupted MSH6 gene on chromosome IV.

Figure 15. Southern blot analysis of YN97-150 (msh6::KanMX4)

Southern blot analysis of mshό genomic DNA (Fig. 15 A) and die corresponding disrupion scheme (cf. Fig. 15B). Lane (1) BamHl digested control YN94-1 DNA. No bands are seen when probed widi the 1553 bp KanMX DNA. Lane (2) BamHl digestion of genomic DNA containing the disrupted mshό gene. A single band of 2481 bp is seen corresponding to the insertion of KanMX4.

(d) RTH1

The RTH1 (RAD27) gene of Saccharomyces cerevisiae is one of several genes encoding a 5'→3' DNA exonuclease. Subsequent mutations arising in rthl strains are duplications resulting from a novel mutagenic process, and are not due to a defect in mismatch repair (Tishkoff et al. 1997). During replication of the lagging strand, DNA polymerase extends into the downstream "Okazaki" fragment and displaces it, resulting in a 5' "flap" structure that is normally removed by RTHlp. In the absence of this exonuclease, extensive strand displacement synthesis occurs resulting in the duplication of DNA sequences. Recendy, it has been shown that expansion of CAG repeat tracts are frequent in yeast strains defective in Okazaki fragment maturation (Schweitzer et al. 1998). This result supports the hypothesis that tract expansions that occur during passage of human disease allele results from excess DNA synthesis on the lagging strand of replication.

Disruption of RTHl

The rthl::KanMX4 disruption strain was constructed using standard techniques. RTHl was amplified by PCR using primers .₃₈₀ 5'- GTTAACAGTGACTTTCGGTGACAGATA-3' -₃₅₄ (numbers in subscript indicate positions relative to the start codon of RTHl) and 5'- GTTAACAAAGCTAGGTGTCGAAGG-3' ₁₄₂₄ (extra nucleotides were added to the 5 '-end of this primer creating a Hpal site as underlined). These primers were made in-house at the 40 nmole scale.

Conditions for amplification were as follows: 50 ng YN94-1 genomic DNA template, 0.8 μl 25mM dNTPs (dATP, dCTP, dGTP, dTTP, Pharmacia) mix, lμl each primer (200 pmole), 10 μl lOx Amplitaq^® (Applied Biosystems) buffer, 2 units Amplitaq^® enzyme made up to 100 μl with dist. H₂0 and overlaid with paraffin oil. A Hybaid thermocycler was used. Programme: 92°C for 2 min for 1 cycle; then 92°C for 2 min, 50°C for 3 min, 72°C for 3 min. for 30 cycles. The reaction was then precipitated using 10 μl sodium acetate pH5.2 and 200 μl ethanol for 30 min on ice. The DNA was men centrifuged at 15000g for 20 min. The pellet was then washed in 75 % eώanol and dried in vacuo before being resuspended in 30 μl H₂O.

The resulting PCR product was cloned into S/wαl-digested pUC19 (Yannisch-Perron et al. 1985) creating plasmid pKLRTHl . A 1481 bp Smαl/EcoRV-derived fragment containing the KanMX4 cassette was removed from pFA6-KanMX4 (Wach et al. 1994) and cloned into the ΕcoRV site of pKLRTHl (+66 relative to the RTHl start codon) creating disruption plasmid pKLrthl: :KanMX4. The rthl disruption cassette was released by Hpal digestion and men transformed into YN94-1. For selection of transformants geneticin (G418, Life Technologies) was added to the media at a final concentration of 200 μg/ml. Putative yeast disruptions were checked by Southern hybridization for the disrupted gene on chromosome XI and for the rthl temperature sensitive phenotype (Sommers et al. 1995).

Figure 16. Southern blot analysis of YN97-167 (rthl::KanMX4)

Southern blot analysis of rthl genomic DNA (Fig. 16A) and the corresponding disruption scheme (cf. Fig. 16B). Lane (1) EcoRI digested control YN94-1 DNA. One band is seen when genomic DNA is hybidized with the 316 bp PCR probe. This 3794 bp ΕcoRI fragment corresponds to the wild type RTHl gene on chromosome XL Lane (2) EcoRI digestion of genomic DNA containing disrupted rthl. A shift in the 3794 bp band is seen to 2420 bp corresponding to the insertion of the KanMX4 cassette (diis cassette contains an EcoRI restriction site at its 3'- end).

(e) POL30

The POL30 gene of Saccharomyces cerevisiae encodes PCNA (proliferating cell nuclear antigen), an essential component of the DNA replication machinery. PCNA exists in vivo as a homotrimer that acts as a "sliding clamp" around the DNA helix thereby increasing the processitivity of DNA polymerases δ and ε. It has been shown that PCNA interacts strongly with the MSH2p/MSH3p heterodimer during mismatch repair (Johnson et al. , 99β) Yeast strains carrying me pol30- 104 mutation show a dramatic increase in spontaneous mutation and tract instability, (cf. Table la). Plasmid pCH1577 (Amin & Holm, 1996) containing the pol30-104 allele marked with the LEU2 gene was digested to completion with Sacl. The reaction products were transformed directly into S. cerevisiae YN94-1. Putative disruptions were patched and first checked for the cold sensitive phenotype (inability to grow at 15 C) then by Soudiern hybridization analysis and finally by sequencing which revealed the expected nucleotide exchange of a C to a T at position 752 which results in the replacement of alanine by valine in the translated product.

Figure 17. Southern blot analysis of YN98-3 (pol30-104::LEU2)

Southern blot analysis of pol30- 104 genomic DNA (Fig.17 A) and me corresponding disruption scheme (cf. Fig. 17B). Lane (1) EcoRV digested control YN94-1 genomic DNA. When hybridized widi a LEU2 probe two bands are seen (2861 and 961 bp corresponding to the wild type LEU2 gene on chromosome III. Lane (2) EcoRV digestion of genomic DNA containing pol30-104 allele. The LEU2 probe hybridizes to four fragments, firstly the two wild type LEU2 fragments as above and secondly two fragments (appearing as a doublet of 2067 and 2076 bp) corresponding to the LEU2 marker inserted upstream of me POL30 locus

(f) Δmsh2 - deletion of MSH2 in an α - type yeast strain In general, it can be said that mutations affecting repeat stability are either dependent or independent of the MSH2p repair pathway. Mutations of MSH2 together with one or more genes within this epistatic group show a mutator phenotype no greater than the single MSH2 mutant alone. Mutations that are independent of this epistatic group together with the msh.2 allele show an additive or even a multiplicative effect with regard to the mutator phenotype.

We have deleted MSH2 in an α-type yeast strain, which, when crossed with any a-type mutant yeast strain will allow die combination of the mutant and msh.2. In this way, it can be ascertained whedier the primary mutation is dependent or independent of the MSH2p pathway.

PCR-mediated disruption of MSH2 by SFH

The SFH disruption cassette (Wach et al. 1994) was amplified from pFA6- KanMX4 by PCR using Amplitaq^® (Applied Biosy stems). Two primers were designed consisting of 40 bases of flanking sequence of MSH2 and 20 bases complementary to the KanMX4 cassette.

The primers SFH1 :

5 ' -TTATCTGCTGACCTAAC ATC AA AATCCTC AGATTAA AAGTAG CTGAAGCTTCGTACGCTG -3' SFH2:

5 ' -TATCTATCGATTCTCACTTAAGATGTCGTTGTA ATATTAAATA GGCCACTAGTGGATCTG -3' Conditions for the amplification were as follows: 50 ng pFA6-KanMX4, 0.8 μl 25 mM dNTPs mix, 1 μl each primer (200 pmole), 10 μl lOx Amplitaq^® (Applied biosy stems) buffer, 2 units Amplitaq^® made up to 100 μl with dist. H.-.0 and overlaid with paraffin oil. For PCR programme see "disruption of RTHl ". The PCR product was then precipitated using 10 μl 3 M sodium acetate pH 5.2 and 200 μl edianol for 30 min on ice. The DNA was the centrifuged at 15,000 g for 20 min. The pellet was then washed in 75% edianol and vacuum dried before being resuspended in 50 μl of dist. H₂O. Transformation of yeast was carried out according to the method of Gietz and Woods (1994). For selection, geneticin (G418)-supplemented medium was used. Putative yeast disruptions were checked by Direct Colony PCR (Pearson and McKee, 1991) using the following primers.

Checking primer 1 :

-₂₆₉ 5'- AGCACTCCGTATAAACAAAG -3' -₂₅₀ Checking primer 2: sin 5'- TAGTGACAGTGGAATAAAGG -3' ₃₁₃₀ (Numbers in superscript indicate positions relative to the MSH2 start codon).

On checking by PCR a wild type MSH2 strain gives a product of approximately 3399 bp whereas strains in which MSH2 has been disrupted by die KanMX4 cassette yield a product of approximately 2016 bp.

Figure 14. Southern blot analysis of YN97-11 (mlhl::LEU2)

Lane (1) EcoRV digested control YN94-1 DNA. Two bands are seen (approximately 1000 and 3000 bp) corresponding to the wild type LEU2 gene on chromosome III. Lane (2) EcoRV digests of yeast genomic DNA containing disrupted MLHl. The gel was probed with a 2200 bp LEU2 fragment. Three bands are seen in the disruptant, two corresponding to the wild type LEU2 on chromosome III and die odier larger band corresponding to me LEU2 gene used to disrupt MLHl on chromosome XIII (approximately 5800 bp).

METHODS

1. E. COLI TRANSFORMATION USING THE "CALCIUM CHLORIDE" METHOD

(Mandel & Higa 1970)

Solutions:

0.1 M MgCl₂ 0.1 M CaCl₂

Protocol:

An overnight culture was diluted (1: 100) in 50ml of LB medium and grown at 37°C with shaking until OD₆₅₀ = 0.3-0.35 (2-2.5 hr approx.). The culture was divided into 2x25ml tubes and incubated on ice for 20 minutes. Tubes were spun at 4°C for 3 minutes at 3500rpm in a centrifuge (Sigma 4kl0). The supernatant was discarded and 20ml 0.1 M MgCl₂ was added, vortexed and dien immediately spun. 10ml of sterile ice-cold CaCl₂ was added and left on ice for 10 minutes. Tubes were spun and pellet was resuspended in 2ml 0.1 M CaCl₂. Cells were then frozen at -70 °C or used directly for transformation. For transformation, lOOμl of cell suspension was added to a pre-cooled reaction tube together with 4μl (1-lOng) of transforming DNA. This was then left on ice for 30 minutes. Cells were then heat-shocked for 45 seconds at 42 °C and dien incubated on ice for 4 minutes. 1ml of LB medium was then added and the cells were incubated at 37 °C for 1 hour. This was then plated directly onto selective media LB plates and incubated overnight at 37 °C.

2 YEAST TRANSFORMATION USING THE "PLATE" METHOD

(Elble, 1992)

Solutions:

PLATE: 90ml 45 % PEG 4000 10ml 1M Li-acetate lml 1 M Tris-HCI, pH 7.5 0.2ml 0.5 M EDTA

Protocol:

The yeast strain was streaked out for single colonies onto YEPD agar and incubated overnight at 30°C. A single colony was picked and grown overnight in 10ml YEPD at 30°C with shaking. 0.5ml of ie culmre was spun down in a microcentrifuge (Sigma 112) and me supernatant was decanted by inversion. lOμl of carrier DNA (lOOμg) and lμg transforming DNA was men added (no transforming DNA was added for control) and vortexed. 0.5ml PLATE was then added, vortexed and incubated overnight at room temperature. Cells were then pelleted, washed and resuspended in 125μl glass dist. H₂O. This was then plated directly onto selective media and incubated at 30 °C for 3 days.

3. RAPID ALKALINE EXTRACTION PROCEDURE FOR SCREENING RECOMBINANT PLASMID DNA

(MINISCREENS)

(Birnboim & Doly. 1979)

Solutions:

Resuspension buffer: 50 mM Tris-HCI

10 mM EDTA, pH 8.0 Lysis buffer: 200 mM NaOH

1 % SDS

Neutralization Buffer: 3 M Na acetate, pH 5.5 Isopropanol Phenol/Chloroform/isoamyl alcohol (25 : 24 : 1 ) Chloroform/isoamyl alcohol (24: 1) Na-acetate 3 M Ethanol 100% and 70%

Protocol:

1.5ml of an overnight culture was spun in a bench centrifuge (Sigma 112) for 3 minutes and the supernatant decanted. 200μl of resuspension buffer was then added and vortexed. 200μl of lysis buffer was then added and the tubes inverted six times 200μl of neutralization buffer was added immediately and the tubes inverted six times. This was then spun for 15 minutes in a bench centrifuge and the supernatant transferred to a new mbe. DNA was then precipitated with 0.6 vol. isopropanol, spun and the pellet washed with 70% ethanol. This was then dried in vacuo and resuspended in 50μl glass dist. H₂0.

4. PREPARATIONS OF YEAST GENOMIC DNA (Sherman et al. 1982)

Solutions:

Sorbitol/Tris/EDTA: 1.2 M sorbitol 10 mM EDTA

0.1 M Tris-HCI, pH 8.0 Dithiothreitol (DTT) 1 M

Zymolyase 20T, 5mg/ml (ICN Biomedicals Inc. , Oxford, UK)

RNase A, lOmg/ml (Sigma, Dorset, UK) - boiled for 10 minutes, then snap cooled.

Pronase E, lOmg/ml (Sigma, Dorset, UK)

SDS 10%

Phenol/chloroform/isoamyl alcohol (25:24: 1)

Chloroform/isoamyl alcohol (24: 1) Na-acetate 3 M

Isopropanol

Ethanol 75 % Protocol:

An individual yeast colony was used to inoculate 10ml SCD-U and grown up overnight at 30 °C with shaking. The culmre was then spun at 3000rpm for 3 minutes and resuspended in 5ml sorbitol/Tris/EDTA, lOμl DTT was added and the culture incubated at 30 °C for 30 minutes. The cells were pelleted as before and resuspended in 0.5ml sorbitol/Tris/EDTA. lOμ Zymolyase 20T and lOμl RNase A were added and incubated at 30°C. lOμl pronase E and 50μl SDS were added and incubated at 37 °C for 2 hours. One phenol/chloroform/isoamyl alcohol and three chloroform/isoamyl alcohol extractions were then carried out. 0.1 vol. 3 M Na-acetate and 0.6 vol. isopropanol were then used to precipitate the DNA which was pelleted and washed with 70% ethanol. This was then air dried and dissolved in lOOμl glass dist. H₂O.

5. RESTRICTION DIGESTS

(Maniatis et al. 1989)

Solutions:

Restriction Buffers (working concentrations) (Life Technologies Inc. , Paisley, UK):

REact™l 50 mM Tris-HCI (pH 8.0) 10 mM MgCl₂

REact™2 50 mM Tris-HCI (pH 8.0) 10 mM MgCl₂ 50 mM NaCl

REact™3 50 mM Tris-HCI (pH 8.0) 10 mM MgCl₂ 100 mM NaCl

REact™6 50 mM Tris-HCI (pH 7.4) 6 mM MgCl₂ 50 mM KCI 50 mM NaCl

Protocol:

For purification of DNA fragments, restriction digests were performed in a total volume of 30μl. 20μl of DNA (0.2-lμg) was digested with 2μl of enzyme, 3μl of lOx restriction buffer and 5μl glass dist. H₂O. The reaction was incubated according to the manufacturer's instructions for a minimum of 3 hours. For restriction analysis, digests were performed in a total volume of lOμl. lμl of DNA was digested with lμl of enzyme, lμl of the appropriate buffer and 7μl glass dist. H₂0. This was then incubated according to the manufacturer's instructions for 1 hour. Enzymes were then inactivated by heating at 60 °C for 10 minutes.

6. RECOVERY OF DNA FROM LOW-MELTING- TEMPERATURE AGAROSE

(Weislander, 1979)

Solutions: Tris-borate (lxTBE) buffer: 0.089 M Tris-base

0.089 M boric acid 0.002 M EDTA pH 8

Efhidium bormide (working cone. 0.5 mg/ml) Tris-HCI 20 mM pH 8 EDTA 1 mM pH 8

Phenol/Chloroform/isoamyl alcohol (25:24: 1) Chloroform/isoamyl alcohol (24: 1) Isopropanol Ethanol 70%

Low-melting-temperature agarose (Life Technologies Inc. , Paisley, UK).

Protocol:

0.25g of low-melting agarose was dissolved in 25ml lxTBE and heated to 70 °C. The gel was dien cast and allowed to set at 4 °C. DNA samples were then loaded into the gel and electrophoresis was carried out at 60v. The gel was men stained for 30 minutes in ethidium bromide and me band of interest was localised using a long-wave-lengdi (300-360 mn) uV lamp to minimize damage to the DNA. The band was then cut out of the gel with a flamed scalpel and added to 5 vol. 20 mM Tris-HCI and 1 mM EDTA. This was then heated at 65 °C to melt the gel. DNA was then extracted using standard procedures and redissolved in 50μl glass dist. H₂0. 7 DNA LIGATIONS

(Maniatis et al. 1983, Pheiffer & Zimmerman, 1983)

Solutions:

T4 DNA ligase (Promega-Biotec, Madison, USA) lx ligase buffer (Promega-Biotec, USA): 50 mM Tris-HCI, pH 7.8

10 mM MgCl₂ 10 mM ditiiiothreitol 1 mM ATP

25 mg/ml bovine serum albumin

Protocol:

Ligations were carried out in a total volume of lOμl. Approximately 0.1 μg of vector DNA and the appropriate amount of insert DNA was added to a sterde microfuge mbe. Reaction was made up to 7.5μl with glass dist. H₂0. For cohesive end ligation, the reaction was warmed to 65 °C for 5 minutes, 37 °C for five minutes, room temperamre for five minutes, then 4°C for five minutes. lμl ligase buffer and lμl T4 ligase was added and the reaction was men incubated overnight at 20 °C for cohesive ends, or at 10 °C for the ligation of blunt-ended DNA. l-4μl of the ligation mixture could men be used for transformation of E. coli.

SOUTHERN HYBRIDIZATION

(Southern, 1975, ECL™ Technical Manual, Amersham, UK, 1951) This direct nucleic acid labelling and detection system is based on enhanced chemiluminescence. The system involves directly labelling probe DNA with the enzyme horseradish peroxidase, achieved by denaturing the probe so that it is in single stranded form. Subsequent addition of glutaraldehyde causes the formation of chemical cross-links so that the probe is covalently labelled with enzyme. Reduction of hydrogen peroxide by the enzyme is then coupled to a light producing reaction involving luminol, which on oxidation produces blue light which can be detected on light sensitive film.

Solutions:

Tris-borate TBE buffer: (cf. earlier) Agarose (Life technologies Inc. , Paisley, UK) 0.25 M HCl Denaturation solution: 1.5 M NaCL

0.5 NaOH

Neutralization solution: 1.5 M NaCl

0.5 M Tris-HCI pH 7.5 20xSSC 3 M NaCl

0.3 M Tri-sodium citrate pH 7.0

Hybridization buffer 0.2 M: (Amersham Life Sciences, ECL Gold hybridization buffer, with 5 % (w/v) blocking agent and 2.2 M NaCl) SDS 1 %

Hybond-N (Life Science, Amersham, UK) ECL Nucleic Acid Labelling and Detection Kit (Life Science, Amersham, UK).

Protocol:

Pre-digested genomic DNA was run on a 1 % agarose gel in lxTBE. The gel was then immersed in 0.25 M HCl and agitated for 7 minutes. This was then rinsed twice with dist. H₂O and immersed in denaturation solution for 30 minutes. After rinsing with deionised H₂O die gel was immersed in neutralization solution for 30 minutes. The DNA was then transferred to Hybond-N with 20xSSC using a vacuum blotter (Appligen, Durham, UK) at 60 mbar for 1 hour. DNA was cross-linked onto the membrane in a Stratalinker 2400 (Stratagene, Cambridge, UK), using the autocrosslink mode (120,000 μjoules/30 sec). The blot was then placed in a glass hybridization mbe (80 x 200 mm, Techne, Cambridge, UK), and 25ml hybridization buffer was added. The blot was then pre-hybridized for 15 minutes at 42°C.

9 PREPARATION OF PROBE

The DNA fragment to be labelled was diluted to a concentration of lOng/μl in glass dist. H₂0. lOOng (lOμl) was then denatured by boiling for 5 minutes. The DNA was immediately cooled on ice for 5 minutes and briefly spun in a microcentrifuge. lOμl of ECL DNA labelling reagent was added to the cooled DNA and mixed. lOμl of ECL glutaraldehyde was added and mixed. This was dien incubated at 37 °C for 10 minutes. The probe was then added to me hybridization solution at a final concentration of lOng/ml. The blot was hybridized overnight in a Techne-oven (Hybridizer HB-1D, Techne, Cambridge, UK) at 42°C and then washed at 65 °C with 2x SSC and 1 % SDS. A second wash was carried out widi 0.5x SSC and 0.1 % SDS. The blot was men exposed to luminescence detection film (Hyperfilm-ECL, Amersham, UK) for various lengths of time.

10. DIRECT COLONY PCR

(Pearson & McKee, 1992)

Solutions:

Glass dist. H₂0

Taq polymerase buffer lOx (Perkin Elmer, Cheshire, UK) 500 mM KCI

100 mM Tris-HCI pH 9 1 % Triton X- 100

Taq polymerase (Perkin Elmer, Cheshire, UK)

Primer: LacZ Reverse 1 :50 dilution Primer: YEp Forward 1:50 dilution dNTPs, 25 mM

Sterile Oil

PCR purification resin (Promega Biotech, Madison, USA)

Protocol:

Single yeast colonies were picked widi a sterile tip and mixed widi lOμl glass dist. H₂0 in a 0.5ml reaction tube and placed on ice. A mix containing lOOμl polymerase buffer, 780μl H₂O, 8μl dNPTs, 4μl of each primer and 5μl Taq polymerase was then made up and vortexed briefly. 90μl of this mix was added to each reaction mbe and sterile oil was added to prevent evaporation.

PCR Programme (Thermal reactor, Hybaid, Middlesex, UK): 92°C - 2 min 1 cycle

92°C - 2 min 45°C - 3 min 72 °C - 2 min 30 cycles

The majority of the oil was pipetted off. The PCR product was purified using PCR purification resin according to the manufacturer's instructions and redissolved in lOOμl glass dist. H₂0.

The reaction products were identified by running 15μl of the assay on a 2% agarose gel in lxTBE.

11. YEAST β-GALACTOSIDASE ASSAY

(Miller, 1972, Guarente, 1983)

Solutions:

Phosphate buffer (KPP) 0.1M: 1 M KH₂P04

1 M K₂HPO4 Mix both until pH 6.5, then dilute 1 : 10

Z buffer: 60 mM Na₂HPO₄

40 mM NaH₂PO₄.H₂0 10 mM KCI

1 mM MgS0₄

Na₂C0₃ 1 M β-mercaptoethanol (Millipore, Bedford, UK) 0-Nitrophenyl β-D-Galacto-pyranoside (ONPG) (Sigma, Dorset, UK): 4mg/ml 0.10-0.11 mm acid- washed glass beads (B. Braun Melsungen, Germany)

Protocol:

10 ml overnight yeast culmre in SCD-U was spun for 3 minutes at 5000rpm (Sigma 5kl0). The supernatant was decanted and die pellet washed twice with KPP buffer. The pellet was then frozen at -20 °C for 3 hours to "crack" the cells. The pellet was placed on ice and the same volume of glass beads was added, this was then vortexed at maximum speed for 1 minute. 1.2ml of Z buffer was then added and vortexed. This was men spun in a pre-cooled centrifuge (4°C) for 5 minutes at 3000rpm and me extract decanted into fresh chilled tubes. For the assay 25 μl of extract was then added to 1ml Z buffer + β-mercaptoethanol (175μl β- mercaptoethanol in 50 ml Z buffer) and warmed to 30 °C (control; 1 ml Z buffer / β-mercaptoethanol + 25μl Z buffer). 200μl ONPG soln was added and die time measured for a yellow colour to appear. The reaction was then stopped witii 500μl M Na₂C0₃ soln and die OD₄₂₀ measured (Ultraspec 2000, Pharmacia Biotech, St. Albans, UK).

12. DETERMINATION OF VOLUME ACTIVITY (U/MI IN CELL EXTRACTS

Volume activity was then calculated for each sample using die following equation:

VOLUME ACTIVITY = 1000 x V_τ x OD 420

4.5 x V_F x dt

Where V_E = extract volume (25 μl)

V_τ = total reaction volume (1725μl) dt = time (min)

Unit Definition (U) = one unit (U) of β-galactosidase hydrolyses 1 nMol ONPG per minute under me above conditions.

13 DETERMINATION OF TOTAL PROTEIN IN CELL EXTRACTS

To determine specific β-glactosidase activity, total cell extract protein must be quantified.

Solutions:

Bradford solution (Bio-Rad, Munich) Protocol:

25μl cell extract was made up to 800μl widi glass dist. H₂0. 200μl of Bradford solution was then added, vortexed and incubated at room temperamre for 10 minutes. OD₅₉₅ was then measured (Ultraspec 2000, Pharmacia Biotech, St. Albans, UK) and protein concentration (mg/ml) was determined using a bovine serum albumin calibration curve.

Determination of specific activity (U/mg) in cell extracts

Specific activity of the cell extracts could then be calculated using the following equation:

SPECIFIC ACTIVITY = 1000 x V_τ x OD 420

4.5 x V_E x dt x Cp

Where V_E = extract volume (25 μl)

V_τ = total Reaction volume (1725μl) Cp = protein Concentration (mg/ml) dt = time (min) Unit Definition (U) = one unit (U) of β-galactosidase hydrolyses 1 nMol of ONPG per minute under the above conditions.

14. GEL PURIFICATION OF OLIGONUCLEOTIDES

Solutions: % polyacrylamide gel: 2.5ml glycerol

2.5ml acrylamide/N, N ' -Methylenebisacrylamide 40% (w/v) (Millipore, Bedford, UK) 5ml 5xTBE

15ml H₂O

200μl 10% (w/v) ammonium persulphate (APS) 12μl N,N,N' ,N'-Tetra-memyl-ethylenediamine (TEMED), (Bio Rad, Herts UK)

Protocol:

50μg of each oligo was run on a 4% acrylamide gel in lxTBE at 100 volts. Single stranded DNA bands were shadowed by UV using a fluorescent TLC plate and cut out. DNA was eluted from gel fragments into 200μl glass dist. H₂0 by shaking overnight and precipitated using standard procedures.

15 DNA SEQUENCING (Sanger et al. , 1977)

Protocol:

Nucleotide sequences were determined by automated DNA sequencing based on me chain-termination method using die ABI 373A sequencer (Applied Biosy stems, Foster City, California, USA). Double stranded DNA was sequenced using the 'Taq DNA polymerase dideoxy terminator cycle sequencing kit' (Applied Biosystems), with primers; lacZ Reverse and YEp Forward (cf. Materials and Methods).

EXAMPLE 1:

Construction of preferred yeast assay plasmids pKa3-9(32-l), pKaCEN(32-l) & pKaINT(32-l)

The preferred system utilises a unique group of yeast vectors containing die bacterial lacZ gene (minus die promoter and first 7 codons) fused to die first 29 codons and promoter region of the yeast gene PRS3 (5- phospho-ribosyl-l(α)-pyrophosphate syndietase) (Carter et al. 1994). This functional gene fusion was preferably knocked out-of-frame (-1 reading- frame) by the insertion of a poly d(AC)ι₆ tract at die EcoRI site downstream of the PRS3 promoter and within the coding region initiated from the ATG of PRS3. The resulting PRS3'/lacZ gene fusion containing this out-of-frame insertion was inserted into each of three yeast vectors: a high copy vector, a low copy vector and a single copy integrated into the yeast genome.

1 (i) pKa3-9(32-l) (high copy construct)

The initial PRS3'/lacZ fusion was constructed by die insertion into the multiple cloning region of YΕp356R (Fig. 2, Materials and Methods) (Myers, A.M. , et al. 1986) of a 371 bp DNA fragment comprising the promoter region and regulatory elements of the yeast gene PRS3. This 371 bp HpallClal fragment (-284 - +85 relative to the PRS3 start codon) was rendered blunt and die resulting fragment ligated into the unique Smal site of YEp356R, creating plasmid pSS3-9. Oligonucleotides 1 and 2 (cf. Materials and Methods) containing 16 repetitive AC or TG units were annealed and cloned into the EcoRI site situated within the multiple cloning region of pSS3-9 (50μg of each oligonucleotide was purified on a 4% acrylamide gel and redissolved in glass dist. H₂0. 200ng of each olinucleotide pair were then annealed in glass dist. H₂0 by heating to 90 °C and allowing to cool slowly to room temp. pSS3-9 was cut to completion wi i EcoRI in buffer RΕact™3 and subsequent ligation was carried out widi 100-fold excess of insert to vector at 12°C (cf. Materials and Mediods)). Ligations were transformed directly into S. cerevisiae YN94-1 using the "plate" method (cf. Materials and Methods) and resulting colonies were screened for the presence of insert by "direct colony PCR" (cf. Materials and Methods). Insertion of this repetitive tract creating assay plasmid pKa3-9(32-l) resulted in shifting the coding region of the PRS3 '/lacZ fusion to die -1 reading frame. Also contained witiiin plasmid pKa3-9(32-l) are sequences from the yeast 2 micron origin of replication which allows this construct to exist episomally at a copy number of 50-150 per cell. pKa3-9(32-l) also contains the ColEl origin permitting replication in Escherichia coli. Selection for the assay plasmid is allowed by the incorporation of the ampicdlin resistance gene of E. coli and the URA3 gene of yeast respectively (see. Fig. 5).

l(ii) pKaCEN(32-l) (low copy construct)

Construction of this low copy centromere-based plasmid involved removal of the PRS3'1lacZ fusion containing the poly d(AC/TG)₁₆ tract from pKa3- 9(32-1) on a 3680 bp Bam iNsH fragment (cf. Fig. 5). Initially pKa3- 9(32-1) was digested to completion with -BαmHI and Nsϊ in restriction buffer REact™3 (Materials and Methods). KCI was added to adjust die salt concentration to that of REact™6 and digestion widi Seal was ien carried out to eliminate incorporation of the smaller BamHl/ Nsil fragment.

The resulting 3680 bp fragment containing the -PRS5- 'lacZ fusion was gel purified (cf. Materials and Mediods). pRS416 was also digested to completion wi i NsiL and BamHl in buffer REact™3 with the aim of removing the existing lacZ region on a 923 bp fragment. The resulting 3975 bp vector band containing die ARS and centromere sequences was gel purified and ligated to die 3680 bp fragment of pKa3-9(32-l) containing the PRS31LacZ fusion. In this way the 7655 bp plasmid - pKaCEN(32-l) (Fig. 6) was generated. The ligation mixture was then transformed into E. coli DH5α (cf. Materials and Mediods) and placed onto LB + ampicillin plates. DNA from resulting colonies was purified using the rapid alkaline extraction procedure (cf. Materials and Mediods) and screened by restriction analysis. The low copy centromere-based plasmid exists episomally at levels of 1-5 copies per cell (see Fig. 6).

l(iii) pKaINT(32-l) (single copy integrative construct)

Construction of pKaINT(32-l) involved removal of the PRS3 'llacZ fusion containing the poly d(AC)₁₆ tract from pKa3-9(32-l) on a BamHl/ Nsil fragment. This fragment was subsequently cloned into BamHl/Nsil digested integrative plasmid YIp352 (cf. Fig. 4, Materials and Methods (Myers et al. 1986)). Before pKaINT(32-l) could be integrated into the yeast genome it was linearised by restriction at its unique Ncol site located witiiin the URA3 gene. Integration then occurred by homologous recombination between the URA3 gene of pKaINT(32-l) and the genomic URA3 locus of S. cerevisiae YN94-1 (Fig. 8). The resulting URA3 strain was checked by Soumern hybridization to confirm the insertion of the integrative plasmid (cf. Materials and Methods). (See Figs. 7, 8, 9 and 10).

2. Southern blot analysis of integrated vector

Southern hybridization was used to confirm the integration of pKaINT(32- 1) into the yeast genome (cf. Materials and Methods). A 423 bp EcoRI derived restriction fragment from pKaINT(32-l) containing die yeast PRS3 promoter region was used to probe YN94-1 genomic digests (Fig. 7). In 5 clones transformed with pKaINT(32-l), probing BamHl/ Nsil- restricted yeast genomic DNA revealed two bands; firstly, a 2719 bp fragment arising from the wild type PRS3 promoter region on chromosome VIII, and secondly a 3680 bp fragment corresponding to the engineering PRS3 promoter of pKaINT(32-l), integrated at the URA3 locus on chromosome V. Therefore the plasmid has been stably integrated into the genome (Figs. 8, 9 and 10).

From the Soutiiern blot (Fig. 10) we see that lane 6 gives the same banding pattern as the control YN94-1 in lane 1. This pattern may be due to recombination between the repeated URA3 genes on chromosome V. In this way me integrated vector is lost and the URA3 mutation reverts to wild type, hence growth on selective media and loss of the 3680 bp band on the blot.

EXAMPLE 2: Sensitive screening system in S. cerevisiae utilising the luciferase reporter gene

Here we describe die use of an alternative reporter gene in a screening in S. cerevisiae, utilising the light emitting properties of the luciferase gene product. The development of iis system involves simple sub-cloning steps with a fragment containing the luciferase gene derived from "pGL3- Promter vector" (Promega-Biotec) (cf. Fig. 15). This fragment is fused to our PRS37(AC)₁₆ tract using die three assay plasmids described above which have had the lacZ reporter gene previously removed. DNA stability is then monitored by alterations in levels of measured light emittance using a Lumat luminometer LB9501 (Berthold, UK) with the luciferin substrate obtained from Promega Biotec.

2(i) Further example of a suitable reporter gene:

The URA3 (orotidine-5 ' -phosphate decarboxylase) gene from Saccharomyces cerevisiae

Isolated and sequenced by Mark Rose, Paula Grisafi and David Botstein. (1984); Structure and function of the yeast URA3 gene: expression in Escherichia coli. Gene, 29, 113-124.

SGD Name: URA3/YEL021W (http: //genome- www. stanford.edu/)

Cells in which the URA3 gene is expressed can be selected against by growth on plates containing 5-fluoro-orotate (5-F0A).

EXAMPLE 3:

Preferred human assay vectors

Screening system utilising die β-galatosidase reporter gene for use in human cell lines

This screening system involves the construction of a similar plasmid to that used in die yeast system. In mis case a fragment containing the luciferase gene is removed from pGL3-Promter vector (Fig. 15) and replaced with a fragment derived from one of the three assay plasmids described in Example 1. This fragment derived from the assay plasmid contains me poly(AC)₁₆ tract fused to die β-galactosidase reporter gene minus the yeast promoter and start codon. Cloning of mis fragment into pGL3-Promter vector results in an SV40/lacZ fusion containing a poly(AC)₁₆ tract in the open reading frame. Replication in human cell lines is allowed through the incorporation of the EBV (Epstein-Barr virus) origin of replication derived from plasmid pDR2 (cf. Fig. 16). Selection in human cultured cells is by incorporation of a fragment containing the hygromycin resistance gene also derived from plasmid pDR2 (cf. Fig. 16).

Rather than yeast cells as described in die previous Examples, in anotiier embodiment human cell lines are utilized.

Preferably the cells are derived from mmours of colorectal cancer patients or other mmours which exhibit instability in repetitive (microsatellite) DNA sequences. Alternatively, human cell lines can be engineered to contain mutations in genes implicated in mismatch repair pathways, such as hMSH2 and/or hMLHl, using standard mutagenesis techniques wimin the knowledge of a skilled person.

The following vectors are used to transform human cell lines in me same way mat me vectors of Examples 1 and 2 are used to transform yeast cells according to preferred embodiments of the assay according to the first aspect of the invention.

EXAMPLE 4: Assay for testing the carcinogenic properties of a test substance

Using me test system of the invention me background frequency of microsatellite instability associated widi wdd type (mismatch repair- competent) and strains defective in DNA mismatch repair has been assayed. As shown in Table 1 die yeast strains defective in mismatch repair show extensive DNA instability. This situation can be compared to the DNA instability observed in cell lines derived from colorectal carcinomas.

Table 1: Frequency of alternation in lengths of poly (AC) tracts in wild type yeast strains and DNA mismatch repair mutants.

Basal frequency of microsatellite instability of the various genetic backgrounds within the yeast assay system

Using each type of assay construct frequencies of microsatellite instability were measured in relation to wild type yeast and disruptions as described in Table la. Slippage events responsible for the white to blue colony transition were determined by sequencing.

(I) Wild type

As shown in Table la the wild-type yeast strain for mismatch repair - YN94-1 - shows a low background white to blue frame-shifting frequency when transformed with constructs described above in the context of high copy 2 micron-based yeast origin of replication. Frequency of instability increases relative to the lengdi of the repetitive tract (obviously a longer tract is more likely to undergo mutational change than a shorter one). Frequency of instabilities are approximately 10-fold lower in constructs in the context of the centromere-based vectors compared to the high copy constructs and approximately 100-fold lower in the single copy integrative assay constructs. For the high copy construct containing the (C^A)-, telomeric tract the frequency of alteration was approximately 3xl0^"5 suggesting mat telomeric repeats are much more stable than poly (AC) tracts. In this way, the overall sensitivity of the assay system can be altered dirough a choice of plasmid copy number and repetitive tract leng i. Little difference in slippage-frequency was seen between die constructs in the -1 reading-frame compared to those in the + 1 reading frame when measured in me wild type strain, indicating no bias towards insertions or deletions. The sequence of the d(AC) tract in plasmids rescued from blue colonies was determined as described. As an example, for assay constructs containing the poly d(AC)₁₆ tract (-1 reading-frame) the only alteration detected was die loss of one (AC) pair which resulted in a poly d(AC)₁₅ tract. This seemed to be irrespective of plasmid copy number or tract location. Using mis strain (cf. Table la) 14 of the 20 high copy plasmids in the -1 reading-frame analysed had altered tract lengths. The 6 blue colonies with unaltered tracts presumably had a mutation elsewhere in the LacZ gene. Alternatively as diis is a high copy construct the subpopulation of altered tracts responsible for the white to blue transition may not have been rescued. This latter theory was strengtiiened by the finding mat 10 out of the 10 low copy centromere- based (-1) plasmids rescued (cf. Table la) contained altered tract lengms. For blue yeast colonies containing chromosomally integrated poly(AC)_n tracts a PCR was carried out as described. Four PCR products corresponding to (AC)_n tracts integrated at me URA3 locus on chromosome V were analysed. Sequencing showed tiiat each repetitive tract had lost a single (AC) pair. The loss of this (AC) pair resulted in a shift of the coding region from the -1 reading frame back into the correct reading frame thereby giving rise to a functional gene product. For assay constructs containing the poly d(AC)_!2A tract (+ 1 reading-frame) the most frequent alteration detected was the gain of one (AC) pair which resulted in a poly d(AC)_i3A tract. Again, this was irrespective of plasmid copy number or tract location. In the wild type strain YN94-1 (cf. Table 1) 9 of the 12 high copy plasmids analysed contained altered tract lengms. 8 of these contained one extra (AC) pair, the 9th contained an extra 4 (AC) pairs giving a poly d(AC)₁₆A tract. The gain of a single (AC) pair or in one case 4 (AC) pairs resulted in die shifting of the coding region of the PRS37lacZ fusion from the + 1 reading frame back into the correct reading frame giving rise to a functional gene product.

(II) msh2

A 100-600 fold increase in me frequency of microsatellite instability was measured in yeast strain YN97-10 carrying a disrupted msh2 allele (Table la). Difficulty was encountered in determining frequencies of slippage in this strain carrying high copy plasmids due to the number of mutational events observed. The number of slippage events indicated by ie white to blue colour transition was found to be higher in studies utilising the assay constructs in the -1 reading-frame compared to studies utilising assay constructs in the + 1 reading-frame, indicating a bias for deletions over insertions. This finding is consistent widi die previous results of Strand et al. (1993) who also found a bias towards deletions for this strain. The values clearly illustrate MSH2p as being a key component in the mismatch repair pathway with a strategic and unique role in me identification and correction of insertion/deletion loop structures in the genetic material. (III) mlhl

Mutation in the yeast mismatch repair gene MLHl (Table la) also has a profound effect on the rate of microsatellite instability within all the assay constructs. The degree of instability is in the same order of magnitude to that of mshl mutants (approx. 200-800 fold, cf. Table 2). Again frequencies of frame-slippage are higher in studies utilising the assay constructs in the -1 reading-frame. Although die exact function of MLHlp is unknown these results again illustrate the importance of this gene product in the correction of mispaired loop structures.

(IV) msh3 and mshό

Mutation in the yeast genes MSH3 and MSHό have less profound effects on the rate of microsatellite instability within this yeast system as compared to mutations in MSH2 and MLHl (Table la). Mutation in me MSH3 gene causes an approx. 30-fold increase in frame-slippage compared to wild type. An approximate 2-fold increase in white to blue transition was seen in smdies using die high copy (-1) construct compared to the (+ 1) construct and a 4-fold increase was seen in the levels of slippage using the centromere-based (-1) construct compared to the centromere-based ( + 1) version, again indicating a bias towards deletions ratiier dian insertions in this strain. This finding is also consistent widi die previous findings of Strand et al. (1995) who found a bias towards deletions for the strain deleted in MSH3. Mutation in me MSHό gene causes an 8-12 fold increase in insertion/deletion events indicating die secondary role diis mutS homologue plays in me pathway. (V) rthl

We have shown tiiat strains mutated for the gene encoding diis 5'→3' exonuclease RTHlp, cause a 20-260 fold increase in the frequency of frame-slippage compared to wild type. We have found that this is also dependent on die reading-frame of the assay construct used. Previous data (Johnson et al. 1995) have shown that mutations in RTHl cause up to a 280-fold increase in instability of simple repeats compared to wild-type strains. Secondly, tract changes involving single repeats are exclusively insertions in die rthl strain consistent with the theory of displacement synthesis. Similar to these results we have found a 260-fold increase in instability for centromere-based assay constructs in the + 1 reading-frame and a 20-fold increase for assay constructs in the -1 reading-frame (Table 1). On sequencing of these tracts we found iat in assay constructs containing tracts in the + 1 reading-frame all alterations were due to the insertion of one single repeat unit whereas all the alterations in tracts originating in the -1 reading-frame showed die loss of one (AC) pair or in one case the loss of four (AC) pairs. No insertions were found. These results suggest that as well as the prevention of displacement synmesis leading to the duplication of DNA sequences, RTHlp may also play a role in the prevention of loops occurring in the template strand.

(VI) pol30

The pol30-104 mutation in PCNA causes an approximately 20- 100-fold increase in me frequency of tract alteration as compared to wild type. The effect of pol30-l 04 on tract alterations, however, is not quite as severe as that of the null mutations in the mismatch repair genes (cf Table la). Epistatic analyses of pol30-104 with null mutations in mismatch repair genes MSH2, MLHl and PMS1 have shown that rates of tract instability were the same in double mutants of pol30-104 with null mutations in mismatch repair genes and in single mismatch repair mutants (Johnson et al. 1996). Therefore, it can be said that hypermutability in this mutant results from a defect in mismatch repair.

4(i) Determination and verification of frame-slippage events

When transformed into yeast me preferred three assay plasmids produce white colonies on medium containing a chromogenic substrate for β- galactosidase and no β-galactosidase activity can be measured (cf. Materials and Methods). Frame-slippage within the repetitive poly d(AC/TG)₁₆ region due to faulty DNA mismatch repair results in in-frame variants which are detected as blue colonies (Fig. 12). The slippage events can be verified after plasmids rescued from S. cerevisiae (cf. preparation of yeast genomic DNA, Materials and Methods) are transformed into commercially available strains of E. coli. (Materials and Mediods). The acmal slippage event was identified by DNA sequencing using primer "PRS3 Forward" (cf. Materials and Mediods). The procedure used was based on the chain-termination method (Sanger et al. 1977) using the ABI 373 A sequencer (Applied Biosy stems, Foster City, California, USA) (cf. Materials and Mediods).

Sequencing of the repetitive region within the three assay plasmids revealed mat the blue yeast colony on a plate corresponded to a slippage event at the DNA level. The most frequent slippage event was the loss of one (AC) pair, the result of which shifted the coding region of the PRS3 'lacZ fusion back into the correct reading frame giving rise to a functional gene product (Fig. 12). Once sequenced, resulting in-frame variants of the high copy, low copy integrative assay plasmids were measured for specific β-galactosidase activity in yeast (Fig. 12).

4(ii) Influence of certain food components on DNA instability and repair

4(iii) Screening for the potential negative influences of compounds on DNA stability and mismatch repair

As well as being able to screen for the protective (anticarcinogenic) effects of human dietary components on the stability of microsatellite DNA, it is also possible to screen for the negative (carcinogenic) influences of compounds using the preferred simple colour selection system in yeast according to the first aspect of the invention. When screening for carcinogens, a wild type yeast strain for mismatch repair is transformed with the high copy assay construct for increased sensitivity. This transformed strain is then grown in the presence or absence of carcinogens and die frequency of white to blue colonies arising from the different growth conditions are subsequently compared. Carcinogens increase the rate of white to blue transition as compared to "normal" growth conditions (cf Fig. 1).

The intercalating agent ethidium bromide and its effect on microsatellite instability Ethidium bromide - a known potent carcinogen causes frame-slippage through its action of intercalation. This and other intercalating agents preferentially target monotonic runs or alternating nucleotide sequences. As shown in Figure 17, at higher concentrations (6-7μg/ml) this agent increases the frequency of frame-slippage 300 fold compared to an untreated strain (cf. Figure 1).

Method

Yeast strain YN94-1 was transformed with pKa3-9(32 1) and plated onto selective medium (cf. Materials and Methods). A single colony was inoculated into 100ml SCD-UracU (cf. Materials and Methods) and grown to mid-log phase. Culmre was divided into 10 x 10ml sterile bottles. Cells were harvested, washed in 0.1 M potassium phosphate (KPP) buffer (pH 6.5) then resuspended in KPP buffer at a concentration of approximately 10⁶ cells/ml. Ethidium bromide was then added to concentrations between 0 and 9μg/ml, and die cells were incubated at 30 °C with agitation for approximately 7 hours. Following incubation the cells were washed with 0.1 M KPP buffer, diluted in 10ml H₂0 and plated on SCD Uracil for single washed with 0.1 M KPP buffer, diluted in 10ml H₂0 and plated on SCD Uracil for single colonies (4 plates for each ethidium bromide concentration). This experiment was repeated three times.

Polyamines and DNA stability

Polyamines - a group of flexible polycations are normal constituents of the cell and are essential for many cellular processes. They are found in high concentrations in red meat, fish and vegetables. Under physiological conditions putrescine, spermidine and spermine are protonated and possess two, three and four positive charges respectively. Spermine, with its four positive charges binds two phosphate groups in each strand of the DNA helix. This spanning of the major and minor groove by spermine stabilises the DNA helix (Heby, O. & Persson, L. , 1990).

The frequency of frame-slippage for the preferred centromere-based assay plasmid pKaCEN(32-l) of Example l(i) has been determined in yeast strains disrupted in steps of polyamine synthesis (Fig. 11 and Table 2). Results so far indicate that polyamines are required for the stabilisation of DNA during replication and could therefore aid in the prevention of frame-slippage.

Table 2: Frequency of alteration in lengths of poly(AC) tracts in wild type yeast strains and polyamine mutants.

A simple colour assay for monitoring spontaneous base-base mismatches in S. cerevisiae.

As described, the mismatch repair pathway of eukaryotes repairs both insertion/deletion mismatches and spontaneous base-base mismatches. Therefore, using the above yeast strains we can use our assay to monitor the genetic and dietary influences on spontaneous base/base mismatches in yeast.

All of the above mentioned yeast strains are derived from the wild type yeast YN94-1 available from Dr Michael Stark, Department of Biochemistry, University of Dundee, DD1 4HN, Dundee, Scotland, UK and Professor Michael Schweizer, Genetics & Microbiology Department, Institute of Food Research, Norwich Laboratory, Norwich Research Park, Norwich, NR4 7UA, England, UK and as such contain die ade2-l point mutation which affects the biosynmesis of purine nucleotides (Figure. 18). The red pigment that accumulates in adel and ade2 mutants derives from an intermediate formed in reaction 5. This step involves the closure of an imidazole ring by phosphoribosylaminoimidazole syndietase to yield phosphoribosylaminoimidazole (AIR). Obviously, wild type ADE2 cells do not accumulate this pigment and therefore remain white. Therefore, subsequent reversion of red ade - strains to wild type enables us to monitor the frequency of spontaneous mutations by a red to white colour transition. We have subsequently measured me basal frequency of red to white reversion in the various genetic backgrounds of this system (Table 3) Table 3 Frequency of ADE⁺ reversion in various yeast genetic backgrounds

Strain Frequency of ADE Rate relative to reversion wild type

wdd type 2.51 x 10^"4 1 msh2 1.60 x 10^~2 66 msh3 4.80 x 10^"4 2 mshό 3.57 x 10^"3 14 mlhl 1.87 x lO^"2 74.5 rthl 3.42 x 10^"3 13.6 pol30 6.30 x 10^"3 25

4(iii) Industrial Applicability

The identification of human dietary components mat protect against DNA instability and therefore some types of cancer by use of the present invention will contribute to the scientific basis for a healthy diet. The simple blue/white colour test according to a preferred embodiment can be provided in kit-form or scaled up for use in the food or pharmaceutical industries.

Once a test substance is identified as containing "protective" factors, programmes can be undertaken to characterize and elucidate the mode of actin of the protective factor within me foodstuff. The results from the assay of the invention should be of enormous value to plant and crop breeders who wish to produce foodstuffs of greater nutritional value. It has been observed diat drug resistant human ovarian carcinoma cell lines aquire a mutator phenotype and a deficiency in hMLHl repair activity, with loss of expression of the hMLHl subunit occurring in 9/10 independently derived cisplatin resistant sublines (Hirst et al. 1997). This observation shows that loss of hMLHl expression is a frequent event in the development of drug resistance and support the involvement of mismatch repair in mediating die cytotoxic action of chemofherapeutically important drugs. The assays of the invention include yeast strains carrying this mlhl mutation and so could be used in an in vivo s dy looking at the effects of cytotoxic agents and subsequent resistance.

Table la Frequency of alteration in lengths of poly(AC)n tracts in wild type yeast strains and mismatch repair mutants

Strain Relevant Tract Reading- Frequency of Frequency relative genotype frame tract alteration to wild type

High copy 2 μ based assay constructs

YN94-1 wild type (AC)_t5 1.35xl0^"4 YN94-1 wild type (AC)_4I(A)₂(C)₂(AC)₄A l . lβxlO^"2 YN94-1 wild type (AC),, + 1 1.37xl0^"4 YN94-1 wild type (AC),₂A + 1 1.40xl0^"4 en YN94-1 wild type (AC)_I5A + 1 3.27xl0^"4 YN94-1 wild type (AC)₂₄A + 1 5.40x10^"4

•Vi YN94-1 wild type (AC,.,)_n 2.95xl0⁵ σ YN97-10 mshl (AC) -1 6.80xl0^"2 504 YN97-10 msh.2 (AC) + 1 7.41xl0^"2 529 m m YN97-11 mlhl (AC) -1 1.14x10^"' 843

30 YN97-11 mlh 1 (AC) + 1 8.$0xl0^"2 635 cr

-1 6.06x10^"3 45

1N3 YN97-147 msh3 (AC) YN97-147 msh3 (AC) A + 1 2.90xl0^"3 21

YN97-150 mshό (AC) - 1 1.20x 10-' 9 YN97-150 mshό (AC) ,A + 1 1.40x10°

YN97-167 rthl (AC) 5.43xl0^'3 40 YN97-167 rthl (AC) ,A + 1 4.76xl0^"2 340

YN98-3 pol30-104 (AC) 9.47x ; o- 70 YN98-3 pol30-104 (AC) ,A + 1 3.03x : o- 21

Table la, continued overleaf

Table la, continued low copy centromere-based constructs

YN94-1 wild type (AC),₆ -1 1.40x10^"' YN94-1 wild type (AC)₄,(A)₂(C)₂(AC)₄A -1 1.72xl0^"4 YN94-1 wild type (AC)_I2A + 1 1.65x10^"' YN94-1 wild type (AC)_!5A + 1 2.27x10^"' YN94-1 wild type (AC)₂₄A + 1 4.40x10^"'

YN97-167 rthl (AC),₆ -1 2.72x10^"4 20 YN97-167 rthl (AC)₁₂A + 1 4.30xl0^"3 260

YN98-3 pol30-104 (AC)₁₆ -1 1.72xl0^"3 123 YN98-3 pol30-104 (AC)_I2A + 1 3.29x10^"4 20

Table la, continued overleaf

Table la, continued single copy integrated constructs

YN94-1 wild type (AC)₁₆ -1 <5.0xlO^"δ 1 YN94-1 wild type (AC),₂A fl <5.0xl0^"6 1

YN97-10 mshl (AC)₁₆ -1 1.59X10-¹ 318 YN97-10 mshl (AC)_I2A fl 1.80xl0^"4 180

YN97-11 mlhl (AC)_I6 -1 1.82xl0^"3 363

CO

CZ YN97-11 mlhl (AC)₁₂A fl 2.19xl0^"3 438 CO

<3

— 1 YN97-147 msh3 (AC),₆ -1 1.45xl0^"4 29 CZ — 1 YN97-147 msh3 (AC)₁₂A fl l.OOxlO^"4 20 m

C Os

-O--- σ m

m ro 3

REFERENCES

Aaltonen, L.A. , Peltomaki, P. , Leach, F.S. , Sistonen, P. , Pylkkanen, L.

(1993), Clues to the pathogenesis of familiar colorectal cancer, Science, 260, 812-816.

Birnboim, H.C. , and Doly, J. (1979) A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids

Research, 7, 1513. Carter, A.T. , Narbad, A. , Pearson, B.M. , Beck, K.F. , Logghe, M. , Contreras, R. , Schweizer, M. (1994).

Phosphoribosylpyrophosphate synthetase (PRS): a new gene family in Saccharomyces cerevisiae. Yeast, 10, 1031-1044. Elble, R. (1992). A simple and efficient procedure for transformation of yeast. Biotechniques, 13, 18-20. Eshelman, J.R. and Markowitz, S.D. (1995). Curr. Opin. Oncol. 7, 83-

89. Heby, O. and Persson, L. (1990). Molecular genetics of polyamine synthesis in eukaryotic cells. TIBS. 15, 153-158. Henderson, S.T. , Petes, T.D. (1992). Instability of simple sequence DNA in Saccharomyces cerevisiae. Mol. Cell. Biol., 12, 2749-2757 ^'.

Mandel, M. and Higa, A. (1970). Calcium dependent bacteriophage DNA infection. J. Mol. Biol , 53, 154. Maniatis, T. , Fritsch, E.F. , Sambrook, J. (1983). Molecular cloning a laboratory manual, 7th Ed. Cold Spring Harbour, N.Y. Merlo, A. , Mabry, E. , Gabrielson, R. , Vollmer, R. , Baylin, S.B. (1994),

Frequent instability in primary small cell lung cancer. Cancer Res.

54, 2098-2100. Miller, J.H. (1972). Experiments in molecular genetics. Cold Spring Harbour Laboratory, Cold Spring Harbour, N.Y. Mironov, N.M. , Augelon, A.M. , Potapova, G.I. , Omori, Y. , Gurbunov, O.V. (1994), Alterations of (CA)n DNA repeats and tumour supressor genes in human gastric cancer. Cancer Res. 54, 41-43.

Modrich, P. and Lahue, R. (1996), Mismatch repair in replication fidelity, genetic recombination and cancer biology. Annu. Rev. Biochem., 65, 101-133. Myers, A.M. , Tzagoloff, A. , Kinney, D.M., and Lusty, C. (1986). Yeast shuttle and integrative vectors with multiple cloning sites suitable for construction of lacZ fusions. Gene, 45, 299-310. Orth. K., Hung, A. , Gezdar, A. , Bowcock, J.M. , Mathis. (1994), Genetic instability in human ovarian cancer cell lines. Proc. Natl. Acad. Sci. USA, 91, 9495-9499. Pearson, B.M. and McKee, R.A. (1992). Rapid identification of Saccharomyces cerevisiae, Zygosaccharomyces bailii and Zygosaccharomyces rouxii. Int. J. Food Micro, 16, 63-67. Sanger, F., Nicklen, S. and Coulson, A.R. (1977). DNA sequencing with chain-termination inhibitors. Proc. Natl. Sci. USA, 74, 5463-5467. Sikorski, R., and Hieter, P. (1989). Genetics, 122, 17-19.

Strand, M. , Prolla, T.A. , Liskay, R.M. , Petes, T.D. (1993). Destabilization of tracts of simple repetitive DNA in yeast by mutations affecting DNA mismatch repair. Nature, 365, 274-278. Strand, M. , Earley, M.C. , Crouse, G.F., Petes, T.D. (1995). Mutations in the MSH3 gene preferentially lead to deletions within tracts of simple repetitive DNA in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA, 92, 10418-10421. Umar, A. and Kunkel, T.A. (1996). DNA-replication fidelity, mismatch repair and genome instability in cancer cells. Eur. J. Biochem., 238, 297-307. Weislander, L. (1979). A simple method to recover intact high molecular weight RNA and DNA after electrophoretic separation in low gell temperature agarose gels. Anal. Biochem, 98, 305.

Amin, S. N. and Holm, C. (1996) In vivo analysis reveals that the interdomain region of the yeast proliferating cell nuclear antigen is important for DNA replication and DNA repair. Genetics, 144, 479-493. Alani, E. (1996). The Saccharomyces cerevisiae Msh2 and Mshό proteins form a complex that specifically binds to duplex oligonucleotides containing mismatched DNA base pairs Mol. Cell. Biol. 16 p5604- 5607. Gietz, R. D. and Woods, R. A. (1994). High efficiency transformation with lithium acetate. In Johnson, J. R. (Ed.)

Molecular Genetics of Yeast. A Practical Approach. IRL Press, Oxford, pp. 121-131. Henderson, S. T. , Petes, T. D. (1992). Instability of simple sequence DNA in Saccharomyces cerevisiae. Mol. Cell. Biol. 12: 2749- 2757.

Hirst, G. L. Mcllwrath, A. J. Anthoney, D. A. Van Der Zee, A. and Brown, R. (1997). Loss of hMLHl expression and drug resistance in ovarian tumour cells. Poster presentation, Ninth Tenovus - Scotland Symposium. Johnson, R. E. Kovvali, G. K. Prakash, L & Prakash, S. (1995) Requirement of the yeast RTHl 5' to 3' exonuclease for the stability of simple repetitive DNA. Science 269: 238-239. Johnson, R. E. Kovvali, G. K. Guzder, S. N. Amin, N. S. Holm, C.

Habraken, Y. Sung, P. Prakash, L. Prakash, S. (1996). Evidence for involvement of yeast proliferating cell nuclear antigen in DNA mismatch repair. J. Biol. Chem. 271, 27987-27990. Jones, E. W. and Fink, G. R. (1982) Amino acid and nucleotide biosynthesis. In Strathern, Jones and Broach (Ed.) The molecular biology of the yeast

Saccharomyces - metabolism and gene expression), p.271-278.

CHS Press. Schweitzer, J. K. and Livingston, D. M. (1998). Expansions of CAG repeat tracts are frequent in a yeast mutant defective in Okazaki fragment maturation. Hum. Mol. Genet. 7,69-74 Shampay, J. S. Szostak, J. W. Blackburn, E. H. (1984) DNA sequences of telomeres maintained in yeast. Nature 310, 154-157. Sia, E. A. Jinks-Robertson, S. & Petes, T. D. (1997). Genetic control of microsatellite stability. Mutation Research 383: 61-70. Sia, E. A. Kokoska, R. J. Dominska, M. Greenwell, P. & Petes, T. D.

(1997).

Microsatellite instability in yeast: Dependence on repeat unit size and DNA mismatch repair genes. Mol. Cell. Biol. 17: 2851-2858 Sommers, C. H. Miller, E. J. Dujon, Prakash, S. & Prakash, L. (1995)

Conditional lethality of null mutations in RTHl that encodes the yeast counterpart of a mammalian 5'- to 3 '-exonuclease required for lagging strand DNA synthesis in reconstituted systems. J. Biol.

Chem 270: 4193-4196. Strand, M. , Prolla, T. A. , Liskay, R. M. , Petes, T. D. (1993).

Destabilization of tracts of simple repetitive DNA in yeast by mutations affecting DNA mismatch repair.

Nature 365: 274-278. Strand, M. , Earley, M. C , Crouse, G. F. , Petes, T. D. (1995).

Mutations in the MSH3 gene preferentially lead to deletions within tracts of simple repetitive DNA in Saccharomyces cerevisiae. Proc.

Natl. Acad. Sci. USA 92: 10418-

10421. Tishkoff, D. X. Filosi, N. Gaida, G. M. Kolodner, R. D. (1997). A novel mutation avoidance mechanism dependent on S. cerevisiae

RAD27 is distinct from DNA mismatch repair. Cell 88: 253-263. Wach, A. Brachat, A. Pohlmann, R. and Philippsen (1994) New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae. Yeast 10: 1793-1808.

Wierdl, M. Dominska and Petes, T. D. (1997) Microsatellite instability in yeast: Dependence on the length of the microsatellite. Genetics 146:

769-779. Yanisch-Perron, C. Vieira, J. and Messing, J. (1985). Improved M13 phage cloning vectors and host strains: nucleotide sequence of the

M13mpl8 and pUC19 vectors. Gene 33: 103-119.

Claims

1. An assay for testing the carcinogenic properties of a test substance comprising: (i) introducing into cells a reporter gene expression vector comprising a repetitive DNA sequence which exhibits instability in cancer cells, whereby instability of the repetitive DNA sequence affects expression of the reporter gene;

(ii) exposing the resulting cells to the test substance; and (iii) determining whether the test substance is carcinogenic or anticarcinogenic by comparing d e frequency of reporter gene expression in the resulting cells with the frequency of reporter gene expression in cells which have not been exposed to the test substance.

2. An assay as claimed in Claim 1 wherein the cells are eukaryotic cells.

3. An assay as claimed in Claim 1 or 2 wherein the cells are yeast or human cells.

4. An assay as claimed in Claim 1 , 2 or 3 wherein the cells are defective in one or more repair processes, which defects lead to repetitive DNA sequence instability.

5. An assay as claimed in Claim 3 wherein the cells are human cells containing one or more mutations in the hMSH2 gene on chromosome 2 and/or the hMLHl gene on chromosome 3.

6. An assay as claimed in Claim 3 wherein the cells are yeast cells which contain one or more mutations in the MLHl gene and/or MSH2 gene and/or PMS1 gene.

7. An assay as claimed in any one of Claims 1 to 6 wherein the unstable repetitive DNA sequence comprises poly d(AC) and/or poly d(GT).

8. As assay as claimed in Claim 7 wherein the unstable repetitive DNA sequence comprises poly d(AC)_n and/or poly d(GT)_n, wherein n is from 8 to 60, preferably from 16 to 32.

9. An assay as claimed in any one of Claims 1 to 8 wherein the repetitive DNA sequence exhibits instability in human colorectal cancer cells.

10. An assay as claimed in any one of Claims 1 to 9 wherein the reporter gene is selected from lacZ, URA3 or luciferase.

11. An assay as claimed in Claim 10 wherein the reporter gene comprises the lacZ gene.

12. An assay comprising :

(i) testing the carcinogenic properties of a test substance in yeast cells in an assay as claimed in Claim 3 or Claim 4 or any of their dependent claims; and

(ii) further testing the carcinogenic properties of the test substance in an assay in human cells, as claimed in Claim 3 and any of its dependent claims.

13. An assay for testing the carcinogenic properties of a test substance comprising the steps of: (i) screening a test substance using a reporter gene expression vector as claimed in any one of Claims 1 to 11; and

(ii) repeating the assay using cells containing a high, medium, low or single copy number reporter gene expression vector, the high, medium, low or single copy number vector being selected depending on the frequency of repetitive DNA instability measured in step (i).