WO1999066059A1

WO1999066059A1 - Novel fragmentation vectors and uses thereof

Info

Publication number: WO1999066059A1
Application number: PCT/EP1999/004106
Authority: WO
Inventors: Jurgen Del-Favero; Christine Van Broeckhoven
Original assignee: Vlaams Interuniversitair Instituut Voor Biotechnologie Vzw
Priority date: 1998-06-12
Filing date: 1999-06-11
Publication date: 1999-12-23
Also published as: AU4513199A

Abstract

The present invention provides new vectors for the fragmentation of Yeast Artificial Chromosomes. The new vectors are characterized by the use of a short target sequence and are especially useful to isolate flanking sequences of relatively short triplet repeats that may play a role in aberant expression of human genes.

Description

Novel fragmentation vectors and uses thereof.

The present invention relates to the development of new vectors for fragmentation of yeast artificial chromosomes.

The ability of Yeast Artificial Chromosomes (YACs) to contain and propagate large DNA inserts (50-2500 kb) makes it an ideal tool for the construction of long range physical maps of the human genome. A major disadvantage of YAC based physical maps is their low resolution as compared to complementary resources like Bacterial Artificial Chromosomes (BACs) and P1 - derived Artificial Chromosomes (PACs). Nevertheless, it is an enormous endeavour to construct BACs/PACs based physical maps to the level of the existing human YAC maps, which cover nearly the entire human genome (Chumakov et al., 1995). Therefore, an attractive method to improve the resolution of YAC maps is the use of YAC fragmentation, allowing more refined mapping while using prior efforts and knowledge.

YAC fragmentation is based on homologous recombination of a sequence on the YAC with a target sequence present on the fragmentation vector. As a result of the YAC structure, fragmentation can be performed from the centric as well as the acentric YAC side. Vectors containing a selectable marker and a telomere can be used for acentromeric fragmentation, while the addition of a centromere to the vector, allows centromeric fragmentation. In either case deletion derivatives are generated with different nutrient requirements than their parent YAC. YAC fragmentation with either repetitive or unique sequences as target for homologous recombination has a number of applications for analysis, modification and selection of cloned sequences in YACs. YAC fragmentation with a repetitive target sequence, such as Alu, is useful for the construction of detailed fragmentation panels, representing deletion derivatives covering the complete parent YAC. Such a panel can be used for the construction of restriction maps of YACs (Cook and Tomilinson, 1996) and has additional application in delineating markers and genes within YACs. Furthermore, fragmentation with repetitive sequences allows removal of chimaeric parts of a YAC or can be employed in the creation of minimal YACs for the production of transgenic cell lines or animals, by removing the non-essential DNA from a YAC of interest. Targeting with specific sequences also has useful applications. Fragmentation by cDNA or specific exons of a gene allows determining the orientation and/or the estimation of the physical length of a gene (Del-Favero et al., 1999). Furthermore, it is possible to construct minigenes with intact promoter regions and spliced exons, without prior knowledge of the promoter.

Until the present invention homologous recombination between fragmentation vector and YAC was carried out with relatively long repetitive sequences such as Long Interspersed repetitive Elements (LINEs) or Short Interspersed repetitive Elements (SINEs) comprising at least 300 bp or non repetitive sequences identical to sequences known to be present in a YAC (at least 200 bp). It is thought in the art that shorter repeats or sequences would lead to unacceptably low recombination frequencies thus not giving any acceptable amount of deletion fragments.

It has now been found that it is possible to go as low as three repeats of a triplet such as CAG/CTG or CCG/GCC whereby homologous recombination still occurs to a sufficient extent and whereby thus deletion fragments are obtained which can be used for determining genome organisation and/or sequence of inserts in YACs.

Thus the present invention provides a fragmentation vector for producing deletion fragments of yeast artificial chromosomes, said vector comprising at least one telomere, at least one selectable marker and at least one repetitive element allowing for homologous recombination between said vector and said yeast artificial chromosome, whereby said repetitive element comprises at least 3 repetitions, preferably at least 7 repetitions of a certain triplet and whereby homologous recombination can occur with a frequency of at least about 1 %, preferably at least 5%. Of course the repetitive sequence does not need to be a triplet. It may as well be a doublet or longer (tetra, penta....) sequence which is repeated. In general more repetitions will be needed for doublets and less for longer sequences in order to still obtain sufficient homologous recombination. The invention is also applicable to sequences which have no or hardly any repetition at all, such as unique sequences of at least 9 bp, preferably at least 12 bp, more preferably at least 27 bp, which are much shorter target sequences than thought possible until the present invention. Thus in another embodiment the invention provides a fragmentation vector for producing deletion fragments of yeast artificial chromosomes, said vector comprising at least one telomere, at least one selectable marker and at least one sequence element identical to a sequence occurring in said yeast artificial chromosome allowing for homologous recombination between said vector and said yeast artificial chromosome, whereby said element has a length of between 9 and about 250 bp, preferentially between 9 and 150 bp and whereby homologous recombination can occur with a frequency of at least about 1%, preferably at least 5%.

Fragmentation vectors known in the art include non-centromeric as well as centromeric fragmentation vectors. The present invention is of course applicable for both kinds. In order to allow selection, the fragmented YACs need to have different nutrition requirements (preferred) or resistance than the parent YAC.

Thus the invention in a preferred embodiment provides a fragmentation vector whereby said selectable marker upon homologous recombination generates a deletion fragment having a different nutrition requirement than its parent yeast artificial chromosome.

Since most of the human genome is present in YAC libraries contained in yeast strain AB1380 (deposited at the ATCC under no 20843, obtainable from ATCC) it is preferred to be able to select and propagate the deletion fragments in said strain as well. Therefor the vector should be adapted for that purpose. A suitable marker is thus LYS2, which leads to lysine independency. In a further preferred embodiment the vectors according to the invention comprise a SP6 and T7 sequence enabling the direct sequencing of both ends of isolated fragmented YAC ends.

The vectors according to the invention may be produced in a circular form, but for homologous recombination to occur they generally need to be linear. The invention thus comprises both situations, so that in one embodiment the invention provides a fragmentation vector according to the invention which is linear and which comprises a telomere at one of its ends and recombination element at its other end. Alternatively, the invention provides a fragmentation vector according to the invention which is circular and which comprises said element and said telomere separated by a restriction site unique for said fragmentation vector. A typical vector according to the invention is presented in Figure 1.

With the event of the human sequencing project (HSP), PACs/BACs were chosen as anchor clones for constructing human "ready to sequence" contigs. PACs/BACs are advantageous over YACs since PACs/BACs are easier to manipulate and are much more stable as compared to YACs. However, since PACs/BACs contigs are still under construction it is possibly a limiting step in the HSP. This is mainly due to underrepresentation and uneven spacing of the available markers. Therefore, PAC/BAC library construction results in clusters of PAC/BAC clones separated by uncovered gaps. In a classical approach these gaps are filled by end sequencing of the PACs/BACs flanking the gaps, followed by primer design. These end probes are then used to re-screen a PAC/BAC library. This process is iterated until all gaps are covered. Although straightforward, it is a time consuming process, with an average cycle time between two weeks and 1 month.

Generating more markers in a PAC/BAC independent manner can shorten this process. Hereto, several strategies can be followed. One such strategy is to use plasmid libraries from flow sorted human chromosomes. Random sequencing of several thousand of these clones will generate enough new markers to construct complete PAC/BAC from an entire chromosome. However, not all chromosomes can be sorted as a single chromosome resulting in extra, time consuming, and verification steps to isolate these chromosome specific clones. Also, since the clones are generated ad random, no positional information is available, hampering the efficient construction of PAC/BAC contigs.

An elegant method to circumvent the above mentioned problems is the use of YAC fragmentation. In the past nearly complete YAC contigs have been constructed for each chromosome of the human genome. These can serve as a resource for the isolation of new markers. Hereto, a selection of these YACs, spanning a complete chromosome, is fragmented with acentric fragmentation vectors according to the current invention carrying a 100-bp Alu repeat in both orientations. The resulting fragmented clones have different nutrient requirements and can be easily selected for by replica plating onto growth media that reflect these differences. Next, YAC ends will be generated ad random and sequenced to obtain new markers. Also, this method provides positional information, which is of importance for an efficient PAC/BAC contig-mapping project.

This method is applicable on every chromosome or part of a chromosome or specific region of a chromosome for which a YAC contig map is available. Based on the accessible YAC information it is crucial to select these clones that assemble into the best minimal tiling path (MTP) with respect to chimerism and stability.

The proposed acentric fragmentation vectors to use in this method are based on the pDV1 vectors. As target site for homologous recombination a 100 bp Alu repeat fragment is used enabling a fragmentation efficiency of at least 80% (Del- Favero et al., 1999). A high throughput approach is preferred and is reflected in the use of 96-well microtiter plates for growth (MacMurray et al., 1991 ) and transformation (Smith et al., 1995) of the selected YAC clones. Resulting transformants will be plated on 132 mm petri-dishes containing growth media (supplemented with 5- fluoro-orotic acid) which will only the growth of fragmented YAC clones.

For each Megabase (Mb) of YAC insert, 50 fragmented YAC clones will be picked for subsequent YAC end rescue. This corresponds to a 5 times coverage for finding a fragmented YAC clone every 100 kilobases (kb). YAC end rescue includes DNA preparation of the fragmented YAC DNA, restriction digestion of the isolated DNA and circularization of the digested DNA by ligation. All these steps are performed in a 96 well format.

Transformation of the ligation products into E. coli cells is achieved on a one by one basis and resulting colonies are plated on selective medium. 1 colony of each transformation mix is chosen ad random for sequencing.

Plasmid template DNA is prepared in 96 well plates and subsequently sequenced using high-throughput sequencing equipment. The resulting sequences are processed to remove contaminating and repetitive sequences. Also, databank searches are performed to exclude all known sequences. Next, all newly generated sequences are subjected to primer design software. The generated primers are used to PCR amplify the new chromosome specific markers.

The new markers will finally be used to screen PAC/BAC libraries by hybridization to generate PAC/BAC based ready to sequence contigs. The integrity of the contig is assessed by fingerprinting and /or fiber FISH. Detailed description of the invention

Vector construction

The basic vector pDVO (for details see figure 8), was constructed by inserting a 4.5 kb EcoRI/Sall fragment of the fragmentation vector pBCL8.1 carrying LYS2 and TEL (Lewis et al., 1992), into the plasmid vector pGEM3zf(-) digested with EcoRI/Sall. Next an End Rescue Site (ERS) was ligated in the EcoRI site. The ERS was designed to contain the recognition sequences of four restriction enzymes: EcoRI, BamHI, Kpnl and Clal. Hereto, two complementary oligonucleotides 5' TTCGGATCCGGTACCATCGAT 3' (SEQ. ID. NO.1 ) and 5' TTATCGATGGTACCGGATCCG 3' (SEQ.ID.NO.2) were synthesized, annealed and ligated into the EcoRI site partially filled in with dATP using Klenow polymerase, the endproduct is vector pDV1.

For the construction of the CAG/CTG and CCG/CGG triplet fragmentation vectors, a mixture of a (CAG)₇/(CTG)₇ and (CAG)₁₀/(CTG)₁₀ adapter or a (CCG)₁₀/(CGG)₁₀ adapter sequence was blunt end ligated in the Pstl site, after filling in this site with dNTPs and Klenow DNA polymerase, of the pDV1 basic vector. The ligation mixture was transformed in DH5α-cells and the obtained colonies were transferred to Hybond N+ membranes, treated by standard methods and hybridized with a γ- ³²P labeled (CAG)₁₀ or (CCG)₁₀ oligonucleotide as a probe. The orientation of the adapter relative to the vector telomere (TEL) was determined by sequencing different selected clones using the SP6 universal primer (Figure 1 and 2). The resulting fragmentation vector were: the pDVCAG vector with the adapter sequence in a 5' (CAG)₇3' orientation relative to TEL; the pDVCTG vector has the adapter sequences oriented in a 5' (CTG)₁₀ 3' position relative to TEL; the pDVCCG vector oriented in the 5'(CCG)₁₀3' position relative to the telomere and the pDVCGG vector containing the adapter sequence in a 5'(CGG)₁₀3' relative to the telomere (Figure 1 ). Yeast transformation and selection

YAC clones were grown in AHC medium (6.7 g/l of Yeast Nitrogen Base without amino acids, 10 g/l Casein hydrolysate acid and 55 mg/l adenine hemi- sulphate). A chemical yeast transformation protocol with alkali cations was used in all fragmentation experiments (Gietz et al., 1992). Before transformation fragmentation vectors were linearized by digestion with Sail and two μg of linearized DNA was used per transformation. Transformants were selected on minimally supplemented Synthetic Drop-out (SD) media (Sherman, 1991) lacking lysine (SDLys^"). After culturing for 3 days at 30°C transformants were replica plated. Transformants of the acentric fragmentation were replica plated to SDIys^" trp^'ura^' and SDIystrp ura^* media. Transformants only growing on SDIys^"trp ura⁺ have the correct phenotype corresponding to the replacement of URA3 by LYS2. Transformants of the centric fragmentation were replica plated to SDIyslrp ura^" and SDIys^"trp⁺ura^" media. Colonies only growing on SDIys^" trp⁺ura^" correspond to the replacement of TRP1 by LYS2.

Pulsed field gel electrophoresis (PFGE) and hybridization

High molecular weight (HMW) yeast DNA, embedded in low melting point agarose plugs was prepared as described by Southern et al. (1987). PFGE analysis was carried out using the CHEF Mapper XA apparatus (BioRad). Conditions for optimal separations were as determined by the embedded algorithm. YACs were visualized on an UV transilluminator after ethidium bromide staining. Their size was estimated using as length markers the Lambda concatemere ladder (Boehringer Mannheim) and the chromosomes of yeast strain AB1380. YACs were also visualized after Southern blotting and hybridization with radiolabeled LYS2 or Alu probes. Hybridizations were incubated overnight in 7% SDS, 0.5 M Na₂HP0₄/NaH₂P0₄ pH7.2, 1 mM EDTA at 65°C and subsequently washed at 65°C in 2χSSC (1χSSC is 0.15 M sodiumchloride and 0.015 M sodium citrate), 1 % SDS for 15 min. 1χSSC, 1 % SDS for 15 min. and O.δxSSC, 1 % SDS for 30 min., followed by an overnight exposure to a Kodak X-ray film.

End rescue of fragmented YACs and sequencing

HMW yeast DNA of fragmented YACs was isolated in agarose plugs and YAC end sequences were obtained by end rescuing. Hereto, 20 μl of a agarose plug was equilibrated with 1 χTE (10 mM Tris-HCI pH8, 1 mM EDTA) and agarose was removed by treatment with gelase as described by the manufacturer (Epicentre Technologies). Next, the DNA was digested with one of the 4 restriction enzymes of the ERS in the presence of 2χOPA buffer (1 χOPA is10 mM Tris- acetate pH7.5, 10 mM magnesium acetate, 50 mM potassium acetate) in a total volume of 50 μl for 3 hours. Following heat inactivation, the digested DNA was circularized by ligation in a total volume of 100 μl in the presence of 1 ^χOPA buffer, 1 mM ATP and 1 unit DNA ligase (Boehringer Mannheim) at roomtemperature for 16 hours. The ligation mixture was dialyzed against 0.5^χTE for at least 2 h and 2 μl were electroporated into electrocompetent XL1-blue cells and plated on LB medium containing 100 μg/ml ampicilline. Three individual transformants from each transformation were grown overnight in liquid medium (supplemented with 100 μg/ml ampicilline) and DNA was prepared with the Wizard DNA purification system (Promega). Sequences were analyzed on an automated DNA sequencer model ABI 373A (Applied Biosystems) with the fluorescent T7 dye-termination system (Applied Biosystems) and the universal SP6 primer (Promega).

Examples

CAG/CTG fragmentation

A general outline of the triplet based YAC fragmentation is depicted in Figure 2. As a proof of principal experiment YAC clone 965a3, which was shown to contain the SCA7 gene was fragmented. SCA7 is the causative gene for autosomal dominant cerebellar ataxia with retinal degeneration if the normal (CAG)₁₀ repeat present in the first exon is expanded above 38 repeats (Del- Favero et al., 1998).

YAC clone 965a3 was separately fragmented with pDVCAG and pDVCTG and transformation efficiencies were determined after LYS2 selection. A 2-fold higher transformation efficiency was obtained with pDVCTG (360 colonies/μg vector DNA). Replica plating on selective medium identified transformants that had integrated LYS2 into the YAC insert by recombination. A recombination efficiency of 10% was obtained with pDVCTG versus 2% with pDVCAG. To examine whether the observed differences were the result of specific recombination events between the SCA7 (CAG)₁₀ and the target sequence on pDVCTG or pDVCAG, we prepared for each vector yeast DNA of 12 independent fragmented YACs, separated them by PFGE, blotted and hybridized the fragmented YACs with LYS2 (Fig. 3). None of the 12 clones obtained with pDVCAG contained fragmented YACs of the same size, whereas 6 out of 12 YACs obtained with pDVCTG contained the same fragment of 1400 kb (Fig. 3), suggesting a specific recombination event. To confirm whether the latter are true recombination products, the fragmented YAC ends were isolated by YAC end rescue followed by sequence analysis. This showed 6 sequences containing a CAG/CTG repeat with a flanking sequence identical to the sequence downstream of the SCA7 (CAG)_n repeat (Genbank Ace N° AF032102), confirming that the YAC fragmentation using short CAG/CTG repeats as target sequence was highly specific. These data clearly show that it is possible to use short (repeat) sequences to fragment YACs in a sequence specific manner.

In order to isolate unknown CAG/CTG repeats, 4 YACs from the SPG4 candidate region, located on chromosome 2 (2p21-24), were used (De Jonghe et al., 1996). SPG4 is one of the loci for dominant spastic paraplegia's (SPG), a clinical and genetically heterogeneous group of neurodegenerative disorders, mainly characterized by spasticity of the lower limbs. Clinical and genetical analysis of Belgian pedigrees confirmed previous data localizing the SPG4 gene between the flanking markers D2S400 and D2S376, separated by a genetic distance of 4 cM (De Jonghe et al., 1996). Furthermore, within certain families, anticipation was observed leading to the possible involvement of triplet repeats in the etiology of the disease.

For this purpose, 4 YAC clones 937d3 (1800 kb); 931 e1 (1700 kb); 802a5 (300 kb) and 895c12 (1500 kb) spanning the SPG4 region (De Jonghe et al., 1996), were fragmented separately with pDVCAG and pDVCTG . YAC clones 802a5 and 895c12 yielded transformation efficiencies of 100-200 colonies per μg vector DNA and upon replica plating yielded recombination efficiencies of 2% independently of the vector used suggesting the absence of CAG/CTG sequences in these YACs. Repetition of the experiment yielded the same results and therefore both YACs were not further analyzed. A significant difference in transformation efficiencies was observed between fragmentation vectors for 931 e1 and 937d3, with transformation efficiencies of 420 versus 160 colonies per μg vector for 931 e1 and 180 versus 400 colonies for 937d3 with respectively pDVCAG and pDVCTG. Replica plating also showed a significant difference in recombination efficiencies between 931 e1 and 937d3 of respectively 8% and 2.5% for pDVCAG and 2% and 17% for pDVCTG. Subsequently, yeast DNA was prepared from 24 individual fragmented YACs of 931 e1 obtained with pDVCAG and 22 individual fragmented YACs of 937d3 obtained with pDVCTG, which were analyzed by PFGE, southern blotting and hybridization with LYS2. YAC 931 e1 resulted in 2 sets of equally sized fragmented YACs containing respectively 18 clones of 1600 kb (B36) and 2 of 1500 kb (B35), while YAC 937d3 contained 4 sets of respectively 4 clones of 1600 kb (C43); 10 of 900 kb (C44); 2 of 350 kb (C42) and 2 of 50 kb (C45) (Table 1 and Fig. 4). To test whether these different sets of fragmented YACs each represented a unique recombination event, we isolated for all fragmented YACs end sequences by plasmid YAC end rescue. The resulting plasmids were analyzed by restriction digestion with EcoRI and Sphl and showed an identical digestion pattern for each fragmented YAC clone from each of the 6 fragmentation sets. Subsequent sequence analysis of most clones showed the presence of identical sequences in each set, confirming the restriction digestion data and consequently proving that each set represents products of a specific recombination event.

Characterization of the CAG/CTG repeats

A BLASTN similarity search (Altschul et al., 1990) with the sequence of each fragmentation set against GenBank division's dbEST and dbSTS, showed only BLASTN hits for the sequences of C44, C43, B36 and B35 (Table 1 ). Sets C44 and B36 were respectively identical to the 3' and 5' sequences flanking the (CAG)₉ repeat contained within the human marker UT2172 (GenBank Ace. N° L18017). C43 matched perfectly with a human macronuclear mRNA sequence (GenBank Ace. N°: L37700)(30). Set B35 showed identity to marker WI-9543 and to expressed sequence tags (ESTs) located in UniGene cluster Hs.27287.

Ultimately, the CAG/CTG repeat content and the opposite flanking sequence were determined for each fragmentation set using the Genome walker kit (Clontech). The number of repeat units was determined in both human and YAC DNA by PCR amplification with flanking primer sets (Table 1) followed by sequence analysis. No differences were observed in the number of repeat units between human and YAC DNA. This analysis showed that B35 (GenBank Ace N° AF154411) contained 1 CAG unit; C42 (GenBank Ace N° AF154410) 5 CAG units, C45 (GenBank Ace N° AF154408) 6 CAG units while C43 (GenBank Ace N° AF154409) contained an imperfect CAG repeat sequence: cacCAGcacCAGCAG (Table 1).

To verify the chromosomal localization of the CAG/CTG repeats, we PCR amplified DNA from a monochromosomal mapping panel (Athwal et al., 1985). This analysis showed that all CAG/CTG repeats mapped to chromosome 2 except C43, which is located on chromosome 18. Subsequently, the polymorphic character of the CAG repeats was determined in 30 control individuals by a fluorescence-based PCR test allowing analysis of CAG repeat fragments on an ABI automated sequencer equipped with GENESCAN software for size determination of the PCR fragments. The size differences between the polymorphic fragments were interpreted in terms of number of triplet repeats. This analysis showed that UT2172 (C44 and B36) was bi-allelic with either 8 or 9 CAG repeats, while the others were monomorphic (Table 1 ). Also, STR/STS content mapping of the parental YACs together with CAG/CTG fragmented YAC clones was performed by PCR amplification for the CAG/CTG primer sets together with STS and STR markers from the SPG4 locus (Fig. 6).

CCG/CGG fragmentation

To further explore the triplet based fragmentation technique we constructed the pDVCGG and pDVCCG acentric fragmentation vectors, containing a CGG/CCG repeat instead of a CAG/CTG repeat as target site for homologous recombination. Two YAC clones 937d3 (1800 kb) and 931 e1 (1700 kb) from the SPG4 region (De Jonghe et al., 1996), were fragmented separately with pDVCGG and pDVCCG. Transformation efficiencies between 2000 and 3000 colonies per μg vector were obtained for each YAC/vector combination. Also, subsequent replica plating showed no significant difference in recombination efficiencies between 931 e1 and 937d3 for either pDVCGG or pDVCCG. A recombination efficiency of 4.4% was obtained for the combination 931 e1 /pDVCGG; 5.1% for 931e1/pDVCCG; 6.3% for 937d3/pDVCGG and 7.4% for 937d3/pDVCCG.

Subsequently, yeast DNA was prepared for 24 individual fragmented YACs from each of the 4 experiments and was analyzed by PFGE, southern blotting and hybridization with LYS2. Fragmentation products from YAC 931 e1 with pDVCGG showed the presence of 7 equally sized fragmented YACs, while fragmentation products of pDVCCG contained 3 sets of respectively 10, 4 and 2 similar fragmentation products. For YAC 937d3 3 sets containing 13, 7 and 2 equally sized fragmented YACs were obtained with pDVCGG, while 7, 4 and 3 equal sized fragmented YACs were found with pDVCCG (Table 2).

To test whether these different sets of fragmented YACs each represented a unique recombination event, we isolated all fragmented YACs ends by plasmid YAC end rescue. The resulting plasmids were analyzed by restriction digestion with Rsal and showed an identical digestion pattern for each plasmid derived from the different fragmentation sets. Subsequent sequence analysis of most clones showed the presence of identical sequences in each set, confirming the restriction digestion data and consequently proving that each set represents products of a specific recombination event.

Characterization of the CGG/CCG repeats

A BLASTN similarity search (Altschul et al., 1990) with the sequence of each fragmentation set against GenBank showed hits for some of the obtained sequences (Table 2). Overall, CGG/CCG repeats could be isolated from the equally sized fragmented YACs (Fig. 7).

These data show that it is possible to isolate CGG/CCG repeats in a region specific manner using the triplet fragmentation technique. Moreover, data show that some region of recombination contain only 1 or a few CGG/CCG repeats but are located in a C-G rich region (Fig. 7; AC007156).

These data also explain the differences observed between CAG/CTG and CGG/CCG fragmentation with respect to the transformation and recombination efficiencies. Higher transformation efficiencies and more even recombination efficiencies are observed when CGG/CCG vectors are used since more recombination events are possible due to recombination events in C-G rich regions.

FIGURES and TABLES

Table 1 : Summary of the CAG/CTG fragmentation data obtained with YAC clones 931 e1 and 937d3 from within the SPG4 locus. Table 2: Summary of the preliminary CGG/CCG fragmentation data obtained with YAC cones 931 e1 and 937d3 from within the SPG4 locus.

Figure 1 : Schematic representation of the new fragmentation vectors, linearized with Sal I, between target sequence (TS) and telomere (TEL) sequences. A) acentromeric fragmentation vector. B) centromeric fragmentation vector; CEN4 centromere for yeast chromosome IV. Filled rectangle represents the target sequence (T.S.). Relevant restriction sites are indicated. Arrows indicate the orientation of the target sequence (5'-> 3'). Arrowheads within LYS2 indicate the transcription direction.

Figure 2: General outline of the triplet fragmentation method.

Figure 3: Pulsed Field Gel analysis of the fragmentation products of the SCA7 YAC.

The pulsed field gel was run under the following conditions: 6V/cm; 14°C for 28 h with a pulse time from 30s to 2m30s. Subsequently the gel was blotted and hybridized with the Lys2 probe. The 840 kb band which is present in every lane represents the endogenous Lys2 gene, located on yeast chromosome II of the yeast strain. The left panel [(CAG)₇ fragmentation vector] shows only fragmentation products in lanes 3, 4, 9 and 12, indicating the absence of a major CAG repeat. The right panel [(CTG)₁₀ fragmentation vector] shows more fragmentation products and more important six YACs of the same length, indicating an identical recombination site (Lanes 2, 3, 4, 6, 7 and 11). λ represents the λ concatemere size marker (Boehringer Mannheim).

Figure 4: Pulsed Field Gel analysis of the fragmentation products of YACs 937d3 and 931 e1.

The pulsed field gel was run under the following conditions: 6V/cm; 14°C for 28 h with a pulse time from 30s to 2m30s. Subsequently the gel was blotted and hybridized with the Alu probe. Figure 4A represents the analyzed fragmentation products of YAC clone 937d5 with the (CTG)₁₀ fragmentation vector. Figure 4B shows the analyzed fragmentation products of YAC clone 931 e1 with the (CAG)₇ fragmentation vector.

Figure 5: Nucleotide sequences of the obtained CAG repeats. (SEQ. ID. NO's: 3-7) For each triplet repeat the complete sequence is shown. The underlined sequences represent the obtained sequence after end rescue of the fragmented YACs. The recombinatory CAG repeat is indicated in bold. The double underlined sequence in sequence C represents the part of the sequence that was identical to the macronuclear mRNA sequence present in the database.

Figure 6: YAC contig map spanning the SPG4 candidate region. Horizontal lines represent YAC clones: thick lines represent parental YACs, thin lines represent fragmented YACs. Black circles represent positive STS/STR hits. Open squares represent the absence of a positive STS/STR hit. The SPG4 candidate regions based on informative recombinants are indicated: SPG4 candidate region between markers D2S400 and D2S367 (De Jonghe et al. 1996).

Figure 7: Nucleotide sequences of the obtained CGG/CCG repeats.(SEQ.ID.NO's: 8-12)

For each triplet repeat the complete known sequence is shown. The underlined sequences represent the obtained sequence after end rescue of the fragmented

YACs. The recombinatory CGG/CCG repeat is indicated in bold.

In the case of AC007156, the YAC was fragmented twice in a 15 kb interval indicating the sensitivity of the CGG/CCG method to recombine with C-G rich regions.

Figure 8: Basic fragmentation vector: pDVO, length 8126 bp.(SEQ.ID.N0.13) Bases 1-14: End rescue site containing EcoRI-BamHI-Kpnl and Clal Bases 14-4819: Lysine-2 gene of yeast (Selectable marker)

Bases 4820-4961 : Yeast telomere sequence

Bases 4962-8126: pGem3zf(-) vector sequence.

Bases 4961-4966: Sail restriction site. Function: linearization of the fragmentation vector.

Bases 4967-4972: Pstl restriction site. Function: cloning site for Target sequence (e.g. triplet repeat).

Bases 4973-4978: Sphl restriction site. Function: isolation of cloned fragmented end in combination with the restriction site used to generate the plasmid.

&) σ

Φ

CO

rπ

CO m

=1 c; r—

rn r

Table 2

REFERENCES

Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, D.J. (1990) Basic local alignment search tool. J Mol Biol 215: 403-410.

Athwal, R.S., Smarsh, M., Searle, B.M., and Deo, S. (1985) Integration of a dominant selectable marker into human chromosomes and transfer of marked chromosomes to mouse cells by microcell fusion. Somat Cell Mol Genet 11 : 177-187.

Chumakov, I.M., Rigault, P., Le Gall, I., Bellanne-Chantelot, C, Billaut, A., Guillou, S., Soularue, P., Guasconi, G., Poullier, E., Gros, I. (1995). A YAC contig map of the human genome. Nature, 377 (6547 supplement): 175-297.

Cook, G-P., and Tomilinson, l-M. (1996). High resolution restriction mapping of YACs using chromosome fragmentation. Nucl. Acids Res. 24: 1585-1586.

De Jonghe, P., Krols, L., Michalik, A., Hazan, J., Smeyers, G., Lόfgren, A., Weissenbach, J., Martin, J-J., Van Broeckhoven, C. (1996). Pure familial spastic paraplegia: clinical and genetic analysis of nine Belgian pedigrees. Eur J Hum Genet 4: 260-266.

Del-Favero, J., Goossens, D., Van den Boscche, D., and Van Broeckhoven, C. (1999) YAC fragmentation with repetitive and single copy sequences: detailed physical mapping of the presenilin 1 gene on chromosome 14. Gene 229: 193-201.

Del-Favero, J., Krols, L., Michalik, A., Theuns, J., Lόfgren, A., Goossens, D., Wehnert, A., Van den Boscche, D., Van Zand, K., Backhovens, H., van Regenmorter, N., Martin, J-J., and Van Broeckhoven, C. (1998) Molecular genetic analysis of autosomal dominant cerebellar ataxia with retinal degeneration (ADCA type II) caused by CAG triplet repeat expansion. Hum Mol Genet 7: 177-186.

Gietz, D., St Jean, A., Woods, R., and Schiestl, R. (1992) Improved method for high efficiency transformation of intact yeast cells. Nucl Acids Res 20: 1425-1428.

Lewis, B.C., Shah, N.P., Braun, B.S. and Denny, C.T. (1992). Creation of a Yeast Artificial Chromosome fragmentation vector based on lysine-2. GATA 9: 86-90.

MacMurray, A.J., Weaver, A., Shin, H.S., and Lander, E.S. (1991 ). An automated method for DNA preparation from thousands of YAC clones.

Nucl Acids Res 25: 385-390.

Sherman, F. (1991) Getting started with yeast. In: Guthrie C and Fink G (Eds) Guide to yeast genetics and molecular biolog. Methods of enzymolog, Academic press, San Diego, pp 3-21.

Smith, H.D., Ragoussis, J., and Markie, D. (1995). Integrative modification of YAC clones in a 96-well microplate format.

Nucl Acids Res 25:1830-1831

Southern, E., Anand, R., Brown, W., and Fletcher, D. (1987). A model for the separation of large DNA molecules by crossed field gel electrophoresis. Nucl. Acids Res. 15:5925-5943.

Claims

1. A fragmentation vector for producing deletion fragments of yeast artificial chromosomes, said vector comprising at least one telomere, at least one selectable marker and at least one repetitive element allowing for homologous recombination between said vector and said yeast artificial chromosome comprising at least 3 repetitions, preferably at least 4 repetitions, more preferably at least 9 repetitions.

2. A fragmentation vector according to claim 1 , whereby said repetitive element comprises between about 5 and about 35 repetitions of a sequence comprising at least a doublet.

3. A fragmentation vector according to claim 1 , whereby said repetitive element comprises between about 5 and about 35 repetitions of a sequence comprising at least a triplet.

4. A fragmentation vector for producing deletion fragments of yeast artificial chromosomes, said vector comprising at least one telomere, at least one selectable marker and at least one sequence element identical to a sequence occuring in said yeast artificial chromosome allowing for homologous recombination between said vector and said yeast artificial chromosome, whereby said element has a length of between 9 and about 150 base pairs.

5. A fragmentation vector according to anyone of the aforegoing claims further comprising a centromere.

6. A fragmentation vector according to anyone of the aforegoing claims whereby said selectable marker upon homologous recombination generates a deletion fragment having a different nutrition requirement than its parent yeast artificial chromosome.

7. A fragmentation vector according to anyone of the aforegoing claims wherein said selectable marker allows for selection and propagation in yeast strain AB1380.

8. A fragmentation vector according to anyone of the aforegoing claims, whereby said selectable marker is LYS2.

9. A fragmentation vector according to anyone of the aforegoing claims further comprising an SP6 primer.

10. A fragmentation vector according to anyone of the aforegoing claims further comprising a fluorescent T7 dye-termination system.

11. A fragmentation vector according to anyone of the aforegoing claims which is linear and which comprises said telomere at one of its ends and said element at its other end.

12. A fragmentation vector according to any of the claims 1-10 which is circular and which comprises said element and said telomere separated by a restriction site unique for said fragmentation vector.

13. A fragmentation vector according to anyone of the aforegoing claims as represented in Figure 1.

14. A method for producing deletion fragments of Yeast Artificial Chromosomes, comprising allowing for homologous recombination between said yeast artificial chromosome and a fragmentation vector according to anyone of the aforegoing claims and selecting the resulting deletion fragments.

15. A method according to claim 14, which is carried out in yeast strain AB1380.

16. A method according to claim 14 or 15, whereby said yeast artificial chromosomes comprise sequences from the human genome.

17. Use of a vector according to any one of claims 1-13 or a method according to claim 14, 15 or 16 in determining organization or sequence of at least part of the human genomic sequences present in said yeast artificial chromosomes.

18. DNA sequences obtainable or obtained by a method or a use according to anyone of claims 14-16.