aDYantamsnc

All articles available online at http://www.idealibrary.com on

SHORT COMMUNICATION

Genetic and Physical Mapping of the Cerebellar Deficient Folia (cdf)Locus on Mouse Chromosome 6

Chankyu Park, Chantal M. Longo, and Susan L. Ackerman1

The Jackson Laboratory, 600 Main Street, Bar Harbor, Maine 04609

Received April 24, 2000; accepted July 12, 2000

(s

c2ChtM

Dtadawr

Cerebellar deficient folia (cdf ) is a recessive mousemutation causing ataxia and cerebellar cytoarchitec-tural abnormalities, including hypoplasia, foliationdefects, and Purkinje cell ectopia. To identify the cdfgene, we have generated a high-resolution geneticmap of a 3.24 6 0.55 cM (95% CI) region encompassingthe cdf gene using 1997 F2 mice generated from a(C3H/HeSnJ-cdf/cdf 3 CAST/Ei)F1 intercross. Linkage

nalysis showed that the cdf gene cosegregates with6Mit208, D6Mit359, and D6Mit225. A contig of fiveACs, nine BACs, and three P1s was constructedcross the cdf nonrecombinant region. Based on ge-etic and physical maps, the cdf gene was localized tohe 0.28 6 0.23 cM (95% CI) interval between D6Mit209nd D6Ack1. These results will greatly facilitate theap-based cloning of the cdf gene, which in turn

hould further knowledge of the molecular mecha-isms of neuronal positioning and foliation duringerebellar development. © 2000 Academic Press

Cerebellar deficient folia (cdf ) is a recently identifiedrecessive mutation (1). Homozygous mutant mice havecerebellar cytoarchitectural abnormalities, includingfoliation defects, as well as behavioral and systemicdeficits, including unsteady gait, reduced body weight,and low fertility. In addition, ectopic Purkinje cells arefound in the white matter and inner granule cell layerof the cdf mutant mouse cerebellum, suggesting abnor-malities in Purkinje cell precursor development (1). Toidentify the cdf gene, we initiated high-resolution ge-netic mapping and contig building using YAC, BAC,and P1 clones.

The cdf gene was previously localized to an esti-mated genetic distance of 2.64 6 2.62 cM at the95% confidence interval (CI) between D6Mit16 andD6Mit70 (1). To enable positional cloning of cdf, weperformed high-resolution genetic mapping betweenD6Mit16 and D6Mit70 by typing 1997 F2 progeny of a

1 To whom correspondence should be addressed at The JacksonLaboratory, 600 Main Street, Bar Harbor, ME 04609. Telephone:

(207) 288-6494. Fax: (207) 288-6077. E-mail: sla@jax.org.

135

C3H/HeSnJ-cdf/cdf 3 Cast/Ei)F1 intercross repre-enting 3994 meioses.The proportion of genotypes among the 1869 nonre-

ombinants for the D6Mit16-D6Mit70 interval was9.1% (544), 59.1% (1105), and 11.8% (220) forAST/Ei homozygotes, heterozygotes, and C3H/HeSnJomozygotes, respectively. This suggested the exis-ence of segregation distortion from the expected 1:2:1endelian ratio (x2

2 df 5 174.5, P , 0.0001). The num-ber of heterozygotes (1105), however, was approxi-mately twice that of CAST/Ei homozygotes (544), sug-gesting that this distortion is due to the reducednumber of C3H/HeSnJ homozygotes, the cdf mutantanimals. Although we were not able to accurately fol-low differences in litter size between birth and wean-ing, the reduced number of affected animals in ourexperiment may be due to poor embryonic or postnatalsurvival.

To map the cdf gene more finely, 128 recombinantsbetween D6Mit16 and D6Mit70 were identified fromour mapping cross and further genotyped with 18 in-tervening SSLP markers (Fig. 1). Our analysis of theserecombinants yielded 12 clusters of markers; some ofthe markers were not genetically separable even after3994 meioses (Fig. 1). All animals homozygous forthe C3H/HeSnJ allele (220) between D6Mit16 and

6Mit70 were phenotypically mutant, whereas none ofhe mice heterozygous or homozygous for the CAST/Eilleles (1325) in this region was affected (Fig. 2). Re-uced numbers of cdf affected animals were also seenmong 128 recombinants; 56 were affected while 73ere wildtype (Fig. 2). There were no animals with two

ecombinant haplotypes.The cdf phenotype cosegregated with the three SSLP

markers D6Mit208, D6Mit359, and D6Mit225 with norecombination (Fig. 2). Therefore, the cdf gene maps tothe region between the proximal flanking markers ofthe cdf nonrecombinant region, D6Mit188/D6Mit209,and the distal flanking markers, D6Mit19/D6Mit246,with a recombination frequency of 11/3994 (0.28 6 0.23cM, 95% CI). Eight recombinants between the proximal

flanking and the cdf-cosegregating markers and three

Genomics 69, 135–138 (2000)doi:10.1006/geno.2000.6322

0888-7543/00 $35.00Copyright © 2000 by Academic Press

All rights of reproduction in any form reserved.

tc(

D

ccarm(ttDbrmDaal

s2htammiTGtoSwviBcc

136 SHORT COMMUNICATION

recombinants between the cosegregating markers andthe distal flanking markers were identified. Thus, as-suming uniform distribution of recombination eventswithin this region, the cdf gene is likely to be locatedcloser to the distal flanking markers.

The genetic map we obtained differs from the MIT(C57BL/6J-Lepob 3 CAST)F2 intercross map (4) withrespect to four closely linked markers. Whereas therelative locus order in our map is cen–D6Mit225–D6Mit246–D6Mit282–D6Mit210, the order in the MITmap is cen–D6Mit246–D6Mit225–D6Mit210–D6Mit282.Because our genotyping data for these markers is froma much larger dataset (3994 meioses compared to 92meioses in the MIT cross), we conclude that the orderof markers shown in Fig. 1 in this region of chromo-some 6 is correct. Furthermore, the marker order inour map was consistent with that derived from ourphysical map (Fig. 3).

We also compared our genetic map with a mouseradiation hybrid (RH) map (10) for the eight commonSSLP markers. The locus order based on the com-bined data of our genetic (Fig. 1) and physical (Fig. 3)maps was D6Mit245–D6Mit126–D6Mit209–D6Mit262–D6Mit98/D6Mit127/D6Mit280–D6Mit282, while thatof the RH map was D6Mit245/D6Mit126–D6Mit209–

FIG. 1. Genetic map of the cdf region of mouse chromosome 6.This map is based on 1997 F2 progeny of the (C3H/HeSnJ-cdf/cdf 3CAST/Ei)F1 subspecies intercross. The 95% confidence interval ofthe genetic distance (cM) is indicated.

D6Mit262–D6Mit98–D6Mit127/D6Mit280–D6Mit282,

showing that the two maps are identical for the com-pared markers in terms of their order.

Eighteen overlapping YAC, BAC, and P1 clones thatspanned the cdf critical region were isolated (6, 8, 9).Nineteen new STSs were generated from the cloneends and used to build a contig between D6Mit209 andD6Mit246 (Fig. 3). Four of the rescued ends containedsequence of L1 repetitive elements. One of the P1 ends,17-1-SP6 (the assigned official locus name is D6Ack1),contained a (CA)14 repeat that was polymorphic be-ween the CAST/Ei and the C3H/HeSnJ strains andosegregated with D6Mit246 with no recombinationFig. 1). The entire cdf critical interval is spanned by a

minimum tiling path containing two YACs (Y167A7and Y434E1), one BAC (B208A9), and one P1 (P17-1)clone, although Y434E1 has a small deletion (Fig. 3).Based on the insert sizes of these clones, the size ofthe cdf-containing region between D6Mit209 and

6Mit246 is estimated to be between 2.5 and 3 Mb.Several markers that genetically cosegregated in our

ross were resolved to separate locations in the physi-al map. The proximal flanking markers, D6Mit188nd D6Mit209, of the cdf critical region were sepa-ated, and D6Mit209 was determined to be the proxi-al marker since D6Mit188 is not within our contig

data not shown). Similarly, D6Mit9 was not present inhe cdf contig, making D6Mit246 the distal marker ofhe cdf region. However, since our new genetic marker6Ack1 is proximal to D6Mit246 in our contig, D6Ack1ecame the distal flanking marker for the cdf non-ecombinant region. Among the cdf cosegregatingarkers, D6Mit225 was determined to be distal to6Mit208 and D6Mit359 by analysis of our contig. Inddition, the two nonpolymorphic markers D6Mit187nd D6Mit262 were placed to the centromeric and te-omeric sides of D6Mit208 and D6Mit359, respectively.

The cdf region of mouse chromosome 6 shares con-erved linkage homology with human chromosomep12–p13 (7). To identify cdf candidate genes from theuman radiation hybrid transcript map, we comparedhe mouse composite map compiled from both linkagend physical maps with the GB4 human transcriptap for the cdf region (2–5, 11). Analysis of the GB4ap suggested that the Siat9 gene may be located near

n the cdf critical region. Primers (Siat9L, CCACACT-CCTACACTGTCA; Siat9R, CACTTAGATGATTTG-AAGG) spanning a dinucleotide repeat in the 39 un-

ranslated region identified a SSLP. Genotype analysisf mice from our mapping cross indicated that theiat9 cosegregates with D6Mit245 (data not shown),hich would place it proximal to the cdf critical inter-al. Hexokinase 2 (Hk2) was previously localized prox-mal to Aup1 by genetic and physical mapping (11).ecause our cdf contig does not contain Hk2 and itosegregated with D6Mit129, a cdf distal marker, thedf gene must reside between Siat9 and Hk2. Investi-

gation of human ESTs in this region may producepotential candidate genes.

The Purkinje cell ectopia found in the cerebellum

tgc

BwG((Ss

137SHORT COMMUNICATION

of cdf mice may be due to migrational abnormalities ofhese neurons. In addition to the recently identifiedenes involved in neuronal migration, the cloning ofdf should provide more insight into the pathways

FIG. 2. Linkage analysis of the cdf locus using 21 SSLP markand the CAST allele are denoted by black and white boxes, respein the numerator and the number of recombinants between thdenominator.

FIG. 3. Physical map of the cdf region. The markers used to buildAC, and P1 clones are not italicized, whereas SSLP markers are itaith “P”. The sizes (kb) of clones are indicated within parentheses.enBank accession numbers for the STSs developed in this study a

1-15-T7), G54327 (17-1-T7), G54328 (17-1-SP6), G54329 (K15-T7), G167L17-T7), G54333 (167L17-SP6), G54334 (295H3-T7), G54335 (29P6), G54338 (363K16-SP6), G54339 (363K16-T7), G54343 (373H

equences for the STSs are available from the GenBank dbSTS databa

involved in neuronal migration and brain develop-ment. The genetic and contig map provided here willgreatly expedite cloning of the cdf gene in the nearfuture.

. The haplotypes of affected animals are shown. The C3H alleleely. The number of affected animals for each haplotype is shownunctional markers for C3H and CAST alleles is shown in the

contig are shown above the map. STSs developed from ends of YAC,zed. YAC clones begin with “Y”, BAC clones with “B”, and P1 clonesrtical lines indicate the markers present on individual clones. Theindicated below with their STS names in the parentheses: G5432630 (K15-SP6), G54342 (147J19-SP6), G54331 (147J19-SP6), G54332-SP6), G54336 (314A13-T7), G54344 (318M2-SP6), G54337 (318M2-

T7), G54340 (373H11-SP6), and G54341 (153F6-T7). The primer

ersctive j

theliciVere5435H311-

se.

138 SHORT COMMUNICATION

ACKNOWLEDGMENTS

We thank Douglas McMinimy and Amy Lambert for assistancewith sequencing, Dr. Roderick Bronson and Susan Cook for helpfuladvice on the cdf mutant, and Drs. Gregory Cox and Juergen Nag-gert for comments on the manuscript. This work was supported byNIH NS35990 and a core grant (CA34196) from the NCI.

REFERENCES

1. Cook, S. A., Bronson, R. T., Donahue, L. R., Ben-Arie, N., andDavisson, M. T. (1997). Cerebellar deficient folia (cdf ): A newmutation on mouse chromosome 6. Mamm. Genome 8: 108–112.

2. Davies, P. O., Melanitou, E., Asano, M., Avner, P. R., andMontagutelli, X. (1995). An anchored molecular map of mousechromosome 6 with an analysis of interference. Mamm. Genome6: 738–740.

3. Deloukas, P., Schuler, G. D., Gyapay, G., Beasley, E. M., Soder-lund, C., Rodriguez-Tome, P., Hui, L., Matise, T. C., McKusick,K. B., Beckmann, J. S., et al. (1998). A physical map of 30,000human genes. Science 282: 744–746.

4. Dietrich, W. F., Miller, J., Steen, R., Merchant, M. A., Damron-Boles, D., Husain, Z., Dredge, R., Daly, M. J., Ingalls, K. A.,O’Connor, T. J., et al. (1996). A comprehensive genetic map ofthe mouse genome. Nature 380: 149–152.

5. Jones, J. M., Albin, R. L., Feldman, E. L., Simin, K., Schuster,

T. G., Dunnick, W. A., Collins, J. T., Chrisp, C. E., Taylor, B. A.,

and Meisler, M. H. (1993). mnd2: A new mouse model of inher-ited motor neuron disease. Genomics 16: 669–677.

6. Haldi, M. L., Strickland, C., Lim, P., Van Berkel, V., Chen, X.,Noya, D., Korenberg, J. R., Husain, Z., Miller, J., and Lander,E. S. (1996). A comprehensive large-insert yeast artificial chro-mosome library for physical mapping of the mouse genome.Mamm. Genome 7: 767–769.

7. Korthaus, D., Wedemeyer, N., Lengeling, A., Ronsiek, M., Jock-usch, H., and Schmitt-John, T. (1997). Integrated radiationhybrid map of human chromosome 2p13: Possible involvementof dynactin in neuromuscular diseases. Genomics 43: 242–244.

8. Pierce, J. C., Sauer, B., and Sternberg, N. (1992). A positiveselection vector for cloning high molecular weight DNA by thebacteriophage P1 system: Improved cloning efficacy. Proc. Natl.Acad. Sci. USA 89: 2056–2060.

9. Sternberg, N., Smoller, D., and Braden, T. (1994). Three newdevelopments in P1 cloning. Increased cloning efficiency, im-proved clone recovery, and a new P1 mouse library. Genet. Anal.Tech. Appl. 11: 171–180.

10. Van Etten, W. J., Steen, R. G., Nguyen, H., Castle, A. B.,Slonim, D. K., Ge, B., Nusbaum, C., Schuler, G. D., Lander,E. S., and Hudson, T. J. (1999). Radiation hybrid map of themouse genome. Nat. Genet. 22: 384–387.

11. Weber, J. S., Jang, W., Simin, K., Lu, W., Yu, J., and Meisler,M. H. (1998). High-resolution genetic, physical, and transcriptmap of the mnd2 region of mouse chromosome 6. Genomics 54:

107–115.