The Proximal End of Mouse Chromosome 17: New Molecular Markers Identify a Deletion Associatedwith Quaking"'"""

The Proximal End of Mouse Chromosome 17: New Molecular Markers Identify a Deletion AssociatedWith quaking"'"""

Thomas Ebersole, Okkyung Rho and KarenArtzt Department of Zoology, The University of Texas, Austin, Texas 78712-1064 Manuscript received November 4, 1991 Accepted for publication January10, 1992

ABSTRACT Five randomly identified cosmids have been mapped proximal to the Leh66D locus onmouse chromosome 17. Two of these cosmids, AulO and Aull9, map near the neurological mutationquaking. Au119 is deleted in qk-blt/qk~ab'cDNA, whereas AulO is not. Au76 maps to a gene-rich region near the Tme locus. The Au76 locus encodes a member of a low copy gene family expressed in embryos, the adult central nervous system and testis. A second member of this family has been mapped to chromosome 15 near c-sis (PDGF-B). At the centromeric end of chromosome 17, Au116 maps near the Tu1 locus, and along with Au217rs identifies a region of unusually high recombinational activity between t-haplotypes and wild-type chromosomes. Au217I and II map to the large inverted repeats found at the proximal end of the wild-type chromosome. In addition, the Au2171 and/or II loci encode testis transcripts not expressed from t-haplotypes.

HE attention given the proximal third of mouse and classical recombination analysis, many markers T chromosome 17 derives, in large part, from the have been mapped into the proximal end. The distri- analysis of t-haplotypes, a variant form found in wild bution of markers is, however, quite variable. The populations of mice. t-Haplotypes exhibit a set of well highest densities occur in the -3-Mbp interval be- defined characteristics: (1) recombinationsuppression tween Lehll9I and T (>3 markers/Mbp), and in the with wild-type chromosomes over a region of an esti- -1-Mbp Plg to Leh66D interval (>6 markers/Mbp). mated 40-Mbp, aconsequence of the existence of By comparison, the regions between T and Plg, and several large inversions (HAMMER,SCHIMENTI and proximal to Lehl191 arerelatively poorly marked (see SILVER1989; HERRMANN et al. 1986; SHIN et al. Figure 1). The combined physical distance involved 1983); (2) transmission ratiodistortion favoring t- in these underrepresented regions can only be esti- haplotype-bearing sperm in t/+ heterozygous males; mated, but a conservative minimum distance would (3) sterility of f/tY males; and (4) usually the presence be -1 0 Mbp, which would convert to less than one of one or more recessive lethal mutations (BENNETT marker per 1O6 bp. 1975). In order to enhance the genetic and physical maps The region encompassed by the inversions is called of the proximal end of chromosome 17 we report the the t-complex. The area within the two most proximal mapping of unique or low copy sequences from five inversions, In(l7)l and In(17)2, are referred toas the cosmids centromeric to theLeh66D locus in wild-type "proximal end." This region is delimited in wild-type chromosomes. These cosmids are part of a larger set chromosomes by the markers Tu1 and Leh66D. The mapping to the t-complex. They were identified by proximal end of t-haplotypes has been compared with screeninga Chinese hamster-mouse cell hybrid ge- its wild-type counterpart in terms of chromosome nomic library with mouse-specific repetitive element structure and apart from the inversions, anotable probes, followed by mapping of mouse-derived cos- structural difference involves an element of at least mids to the t-complex using t-haplotype us. wild-type 650 kbp existing in one copy in t-haplotypes which is restrictionfragment length polymorphism (RFLP) duplicated and inverted inwild-type chromosomes. analysis (EBERSOLE,LAI and ARTZT 1992). The data These elements are thought to serve as regions of presented here introduces new markers into under- matched chromatin which can yield rare recombina- represented regions of chromosome 17 proximal to tion events between t-haplotypes and wild-type chro- Leh66D. Furthermore, these new markers have al- mosomes (HERRMANN,BARLOW and LEHRACH 1987). lowed us to confirm previously described chromosome The resulting partial t-haplotypes have been valuable structure, identify aregion of unusual recombina- mapping tools for the region. tional activity between t-haplotypes and wild-type With the use of partial t-haplotypes, deletions oc- chromosomes, and show that theoriginal spontaneous curring in wild-type chromosomes or t-haplotypes, qk mutation (SIDMAN,DICKIE and APPEL 1964) is

Generics 151: 183-190 (May, 1992) 184 T. Ebersole, 0. Rho and K. Artzt associated witha deletion. Also included in this report been added to 7%,and hybridization was continued for 12- are the map positions of some members of two low 24 hr. A 100-200-n BPsam le of probe DNA was random copy gene families detected with these cosmids. One primed to at least 10 cpm/pg. Three to four million cpm/ ml were added to 200 pg/ml sonicated salmon sperm DNA, family, from a locus duplicated on wild-type chromo- denatured and added to the bag. For probes containing some 17 (loci: Dl 7Au217I and Dl 7Au217II), encodes repetitive sequences, 100-200 pg/ml sonicated cold mouse two testis transcripts not expressed in t-haplotypes. DNA were also added,the mixture was denatured and The other family, a member of which is found at the incubated for 1-2 hrat 42" in 8-10 ml hybridization Dl 7Au76 locus, encodes transcripts found in testicular solution, and then added to the bag. All blots were washed twice in 2 X SSC, 0.1 % SDS at 65",followed by a final wash germ cells, early embryos, and, in later embryos and in 0.1 X SSC, 0.1% SDS at 65". adults, in restricted parts of the central nervous system. RESULTS

MATERIALS AND METHODS AuZO and Ad29 are two loci near qR: The probes P10 and P119 both detect polymorphisms between DNA and probe sources: Characterization of the break- points of the partial t-haplotpes t6 and P,and of the wild type and t-haplotypes with TaqI. P10 detects a deletion chromosomes Thp, gr,TzzH, TP" and TPb2are 5.0-kbp or 3.8-kbp wild-type band, depending on the found in Committee on the Mouse Chromosome 17 (199 1). strain. In t-haplotypes a 2.6-kbp band is present and The partial t-haplotypes r5=and t""'tf are described in also a faint doublet at about 1.2 kbp (Figure 2). The HOWARDet al. (1990). Additional references for some chro- wild-type bands are missing in PP and Ttorl,but not mosomes follow: Thp, JOHNSON (1974); p,SELBY (1973); also see BENNETTet al. (1975); BABIARZ(1 983); Tto", SARV- p or FZH(refer to Figure 1 for a depiction of these ETNICK et d. (1986); fd, VOJTISKOVA et al. (1976); pb2, chromosomes). Furthermore, the locus is duplicated WINKINGand SILVER(1984); t6, HERRMANN,BARLOW and in kub2, that is, Tto"/kub2 DNA hasboth wild-type and LEHRACH(1987). DNA was prepared from liver usingstand- t-type bands, whereas Ttorlcarries neither. An identi- ard methods or from spleen in 0.5% low-melting-point cal situation existswith P119: wild-typebands are agarose plugs. Wild type refers either to the C3H/DiSn or BTBRTF/Nev mouse strains. Identification of the five cos- deleted in pfi and Ttorl,but not 70' or TZ", and the mids is reported in a companion study (EBERSOLE, LAIand locus is duplicated in Pb2(data not shown). The ARTZT,1992). Probes for thenew loci described here were D17RP17 locus has been previously mapped to this derived from genomic fragments from cosmid clones or, in same region using the samechromosomes (HERR- the case of P217c, cDNA isolated using the cosmid clone MANN et al. 1986; MANN,SILVER and ELLIOTT1986). 217. The 1.4-kbp cDNA (DI5Au76rse) was detected in a testicular cell cDNA library using pooled BglII fragments The qk locus is known to be deleted in Ttorland pp, from the chromosome I7 cosmid 76. The cDNA, when but not p (BENNETTet al. 1975; ERICKSON, LEWIS hybridized back to cosmid 76, will cross-hybridize to it and and SLUSSER1978), and it has been mapped within detect a 1.7-kbp BamHI fragment (P76). All blots were 0.3 cMof D17RP17 (KING and DOVE1991). We, washed under stringent conditions: 0.1 X SSC, 0.1% sodium therefore, expected AulO and Au119 to map near qk dodecyl sulfate (SDS), 65", 10-15 min final wash. Nomenclature: Because allthe probes derive from mouse and be useful in the cloning of the gene affected in chromosome 17, for simplicity, each locus will be referred this interesting neurological mutation. We tested to by an abbreviation of the formal name, hence Dl 7AuIO thesetwo loci in quaking miceby hybridization of will be called AulO, etc. Loci on chromosomes other than precompeted whole cosmid to restriction enzyme di- I7 will be referred to with the full name. The original qk gests of qk"/qk" liver DNA. As shown in Figure 3, the mutation has been renamed qkvinbl' to distinguish it from the Aull9 cosmid detects several bands in control lanes recessive lethal ethylnitrosurea (ENU)-induced alleles (KING and DOVE1991). but not any in qk"/qk" DNA. This indicates that most The probes for each locus are named with a P followed if not all of the -40-kbp Au119 cosmid is deleted in by the locus name, P10 is a 4.2-kbp BamHI genomic frag- qk"/qk" mice. This result was verified using two other ment from locus AulO; P116, a 2.8-kbp Hind111 genomic enzymes and two other DNA preparations, one from fragment; P119, a 8.0-kbp EcoRI genomic fragment; P76, a spleen and the other from liver. The AulO cosmid 1.7-kbp BamHI genomic fragment; P217a, a 5.0-kbp EcoRI genomic fragment; P217b, a 3.8-kbp EcoRI genomic frag- detected bands in all lanes of the same blot and is ment; and P2 17c, a 920-bp EcoRI cDNA fragment isolated outside the deleted region (data not shown). from a mouse 10.5-day embryo cDNA library. Au76 maps near a gene cluster: With MsPI, P76 Electrophoresis, blotting and hybridizations: Genomic detects an RFLP between C3H (1.7-kbp) and t-hap- digestions of 8-10 pg of DNA were electrophoresed in TPE lotypes (3.8-kbp and fainter 4.0-kbp bands) (Figure buffer through a 0.7% agarose gel for 15-17 hr at-0.7 V/ cm. Transfer to Hybond-N membranes (Amersham) was 4). The wild-type band is missing in pp,but present accomplished by standard capillary blotting techniques. Hy- in p and F". Ttorris duplicated for the locus, having bridizations were done as follows. Prehybridization of the both band types, while kubZ is deleted for the t-haplo- blot was done in 50% formamide, 50 mM NaP04, pH 6.5, type band. This suggests that Au76 maps to the same 2.5 X Denhardt's solution, 5 X SSC,0.1% SDS, 1 mM gene-rich region as Plg, Sod-2, tw73,Tcp-1, Tme, I@r, EDTA, with 200 pg/ml sonicated salmon sperm DNA for 2 hr at 42". Hybridization of probe was done with the same and near the closely linked genes TcplOb and C, Tcte- solution as in prehybridization except dextran sulfate had 2 and Hst-1 (Committee on the Mouse Chromosome Deletion Associated withqk" 185

FIGURE1 .-Chromosomesused for mapping proximal to Leh66B. Clear boxes indicate wild-type chromatin, crosshatched boxes indicate t- chromatin. The stippled boxes in Pb2 are regions known to be deleted in this chromosome. The markers described in this report are noted below the two middle chromosomes. The relative order of markers in the first tier above the chromosomes are known. The position of markers in the upper tier are not defined with respect to neighboring loci. The breakpoint of the f""" and t""' chromosomes at the proximal end have not been distinguished. The line of dots next to indicates that this Q J+ ty"' chromosome is wild type distal to 2 Leh66A. a* c I Tb b I I 1 T&

1 I T 22"

17 1991; BARLow et ul. 1991; FOREJT et Ut!. 1991). mouse chromosome 17 and 18 (data not shown). The The Au76 locus also potentially encodes a member 1.4-kbp cDNA was mapped using a TaqI RFLP be- of a small gene family. A 1.4-kbptesticular cell cDNA tween C57BL/6 and DBA/2J mice and the BXD RI identified with genomic fragments from cosmid 76 strains. The parental strains share several TuqI frag- detects a 2.8-kbp message in adult testis, and a 3.0- ments, however DBA/2J contains 1.9- and 2. l-kbp kbp messagein adultbrain and in an embryonic bands not present in C57BL/6 (data not shown). Of carcinoma cell line (Figure 5). In situ hybridization of 22 informative RI lines 20 were concordant with c-sis 6.5to 18.5 days post coitumembryos shows that (PDGF-B)on chromosome 15 (see Figure 6). Our data message is present in all stages, but becomes confined give a distanceof 2.6 f 2.05 cM between c-sis and this mainly to the developing central nervous system in new locus, named Dl5Au76rse (related sequence ex- late stages (data not shown). pressed). In an independent analysis using an inter- Although the 1.4-kbp cDNA cross-hybridizes specific backcross, no recombination was detected be- strongly with cosmid 76, it proved notto map to tween these loci in 158 animals, which maps this chromosome 17. This cDNA detects fourBumHI frag- cDNA no more than 1.2 cM from c-sis (95% confi- ments in C3H DNA, but only one of these is found in dence level) (M. J. JUSTICE,personal communication). the somatic cell hybrid, R4 4.1, which contains only Comparison of the sequence of the 1.4-kbp cDNA 186 T. Ebersole, 0.Rho and K. Artzt

h YI

-3.8- - 3.8 -

"2.6 \

-1.2

of AuIV locus. FIGURE2.-Mapping the TaqI-digested DNA 8 hybridized with P10. ppand Tf"' delete this locus. It is duplicated in PbZ. FIGURE4.-Mapping of the Au76 locus. Hybridization of MspI- digested DNA with probe P76. Fpdeletes this locus. TIf"' has both wild-type and f-type bands. Pb2deletes the f-type band at least, and probably the wild-type one as well. Since Mspl (recognition site = CCGG) is sensitive to methylation at the first cytosine, numerous faint hands can be seen.

3.0 - - 3.0 - 2.4 2.8 -

FIGURE3.-Deletion of the At1119 locus in qk"/qP DNA. Above, the entire cosmid clone 1 19 was used as the probe onEcoRldigested FIGURE5.-l'ranscripts encoded at D15Au76rse or Au2171/11. DNA after precompeting away repetitive elements. Below, a control hybridization to the sameblot with probe P76 detects a high Left, the 1.4-kbp cDNA from D15Au76rse hybridized to brain, testis and embryonal carcinoma (EC) cell total RNA. Right, lanes molecular mass fragment (1 1 kbp) in all lanes. qk"' is an ENU- from a single blot showing the P217c cDNA probe hybridized to induced allele on a [-haplotype background. relevant testis RNA. The blot was treated with RNase A. f/f', Tzz"/ f"',and T"/t-' RNA lack a detectable transcript. Inspection of the (locus: D15Au76rse) and thestrongly cross-hybridizing stained gel and a pre-RNase A autoradiogram showed that intact 1 .7-kbp BamHI genomic fragment (probe P76) from RNA had transferred to the blot from all lanes (data not shown). chromosome 17 identified a region of several hundred basepairs with 90% sequence similarity. P76 will de- ship among the family members is unexplained be- tectthe same set of transcripts on aNorthern as yond the small sequence homology we identified. described above for the 1.4-kbp cDNA. When probe Ad16 is at the centromeric end of the t-complex:. P76 is hybridized to BamHI-digested mouse DNA and The P116 probe is polymorphic with EcoRI between washed understringent conditions, fourdifferent- C3H (9.6 kbp), and the haplotypes tw5 and tW1' (4.3 sized bands are evident (data not shown). It is, there- kbp), and t" and tnc5 (1 3.0-kbp) (Figure 7). The locus fore, possible that there are fourclosely related mem- is deleted in T"", but not PP or Y'' which retain the bers in the gene family: one on chromosome 17 in the 9.6-kbp wild-type band. This suggests a map position Plg to Leh66D interval, one on chromosome 15 near near Tul, Au9, and Leh48. Surprisingly, the partial t- c-sis, and two of unknown map position. The relation- haplotype, t""' shows no t"-type 13.0-kbp band, al- Deletion Associated with qk" 187 B X D RI Strain

12 5 6 8 911 12 13 14 15 16 18 1920 21 22 23 24 25 27 28 29 30 31 32

D15Au76rse -DDBDD B B D D D B D D - D D - D B D B DD BD

c-sis BDDBDDBBDD-BDDDDDBDBDBDDDB

FIGURE6.-Strain distribution pattern detected by the 1.4-kbp cDNA from D15Au76rse and by c-six in 26 strains of the R X D RI set. Only strains 3 1 and 32 are discordant.

cosmid DNA as the probe. The two genomic probes indicate that there are two copies of the Au217 locus in wild-type chromosomes (loci: Au217I and Au217Z1) and only one copy in t-haplotypes (locus: Au217A). However, mapping of the P2 17c cDNA probe in t- haplotypes indicates that an additional copy hybridizing to the cDNA sequence exists uniquely in t-haplotypes at a very proximal position (locus: Au217rs, (TS, related sequence)). The evidence is as follows: P2 17a detects 5.0-kbp and 2.6-kbp wild-type bands and 2.9- , 2.6- and 2.2-kbp t-haplotype bands with HindIII (Figure SA). P217b detects 1.8- and 1 .6-kbp wild-type - 4.3 bands and a 3.8-kbp t-haplotype band with HindIII (Figure 8B). In 7"'/tw' DNA only the P2 17a 5.0-kbp wild-type band is missing while withP2 17b both wild- FIGURE7.-Mappingofthe Aul 16 locus. fhR1digest ofrelevant type bands are absent. This indicates that the two DNA hybridized with probe PI 16. Neither f""" nor 1"'" have the wild-type loci are deleted in 70' (the P217a 2.6-kbp 4.3-kbp band their "complete" f-haplotype counterparts carry. r5 band is from t"'). These two wild-type loci are prob- carries a 13.0-kbp band (asterisk) also found in t" and related 1- ably not extremely close to one anothersince they are haplotypes, of which I""' is one, but I"' haslost proximal f- chromatin. The 9.6-kbp wild-type band is deleted in TZ2''. separated by pp, which deletesthe P217b 1.6-kbp band, but not the.8-kbp 1 band. Although P2 17a does though tf2was the parental t-haplotype in the cross- not show any missing bands in pp/tw5DNA, we pre- over event that generated it. The presence of only a sume that the 2.6-kbp band, which is shared by t- wild-type band is evidence for asecond recombination haplotypes and wild type, is from t"' only, the wild- event very proximal in t-chromatin. P116 detects a type fragment being deleted in pp. The distal wild- similar event in the twrx chromosome and in another type copy ofAu217 should reside in the deletedregion shared by and pp. The proximal wild-typecopy chromosome (now extinct)segregating in our tWI2 stock which was represented in our DNA collection. should map to DNA present in ppand deleted in p. We have renamed this chromosome tWf2' by analogy t-haplotypes carry one copy of the genomic probes with twrx. In both casesonly the wild-type band is P2 1 7a and P2 17b. We beleivethat theAu217 genomic present. locus maps to the two inverted repeats found in wild- The tw5' chromosome has been examined previously type chromosomes and to thesingle copy ofthe region at the Tu1 locus, which maps proximal to Zn(17)I in t-haplotypes. This structural difference between t- (HOWARDet al. 1990). twrx is known to have lost t- haplotypes and wild-type chromosomes has been pre- chromatin at Tul, thus, we analyzed the Tu1 locus in viously characterized using the Tu1 19, Tu66 andAu3 twf2xand tWlffand found that they too have lost the t- probes (HERRMANN,BARLOW and LEHRACH1987; specific bands at Tu1 (data not shown). Both tW"' and HOWARDet al. 1990). However, suprising results were tWfzxretain t-chromatin around the Tctex-1 locus (data obtained with the P2 17c cDNA probe. Similar to the not shown) and at Au217rs (see below). This data results with genomic probes P217a andb, P217c suggests that Au116 maps proximal to Tctex-1 and detects related sequences at Au217Z and ZI inwild- Au217rs in t-haplotypes, in the vicinity of Tul. type chromosomes and at Au217A in t-haplotypes. In Au217 maps to the large inverted repeats: The contrast, P2 17c detects additional copies of itself that mapping of Au217 is based on thecumulative data for map very proximal in t-haplotypes, because the 2.9 two genomic probes, P217a andP217b, subcloned TaqI t-haplotype band detected by P217c is missing from cosmid 2 17, and cDNAa probe, P2 17c isolated in thand tW5' (Figure 8C). This suggests a transposition- from a 10.5-day embryo cDNA library using whole like event carried the transcribed sequence, but not 188 T. Ebersole, 0. Rho and K. Artzt A B C

4.3 - 3.8 - 3.3 - 2.9 - 2.9 - 2.6 - 1.8 - 2.2 - 1.6 -

FIGURE8.-Mapping of the Au2J7 loci. (A) Hybridization of~gcnotnicprobe P217a to HindIIIdigested DNA. T" is deleted for both wild- type loci; since the 2.6-kbp wild-type is shared by t-haplotypes. only the i.0-kbp band is missing. Similarly in 7"@,the 2.6-kbp fragment should be deleted. (B) Hybridization of genomic probe P217b to Hindlll-digested DNA. Both Wild-type fragments are deleted in T",but only the smaller 1.Bkbp band is missing in Fp.(C) Hybridization of P2 17c, a cDNA, to Tagl-digested DNA. Note that tu'" is missing the amplified 2.9-kbp band while retains it. Wild-type and t-haplotypes carry a 4.3-kbp fragment. the flanking DNA from Au217A to a proximal position and is collectively known as the proximal end of the in t-haplotypes (Au217rs). In addition, judging from t-complex. Au116 maps to the centromeric end of the the band intensity seen on several different blots, the t-complex near Tul. Au217I and II are found in the transcription unit seems to have been amplified sev- invertedrepeats in wild-type chromosomes and eralfold. We have no evidence that the genomic DNA Au217A maps to the single copy of this region found at the Au2J 7A locus flanking the transcribed sequence in t-haplotypes. t-Haplotypes carry additional copies also maps in this proximal position in t-haplotypes. of the Au217A encoded transcription unit proximal The P2 17c cDNA probe detects messages of -2.4- to Tctex-I. Au76 is located in the region deleted in and 3.0-kbp primarily in testis. (Trace amounts are pb2.Both AulO and Aul19 are near thequaking locus. seen in a few other adulttissues.) Au217I and Au217II Recombination near the start of Zn(27)Z: The ser- may separately encode the two messages, but this has endipitous detection of three recombination events not been determined. The apparent highercopy num- occurring in or near In(l7)l has allowed the ordering ber in t-haplotypes does not amplify message abun- in t-haplotypes of markers just distal to Tu1 as dance. In fact, the opposite is true: Neither f/tY nor cen . . . (Tul/Aull6).. . Au217rs.. . Tctex-1. There F"/t testis RNA from C3H congenic animals contain may be a higher frequencyof crossover in this region any detectable message on northern blots (see Figure than normally seen between wild-type chromosomes 5). It may be that the failure of t-haplotypes to tran- and t-haplotypes. This raises two possibilities: Either scribe this sequence is a result of genomic alterations the region around Tul/Aull6 is outside of In(l7)I that took place during the transposition and amplifi- and not under recombinationsuppression like the loci cation events. within the inversions or elements exist within In(l7)l The proximal position of Au217rs in t-haplotypes which can act as sites of exchange between t-haplo- allowed us to reexamine the tw"' and t'""'haplotypes types and wild-type chromatin. The latter case might at this locus. In contrast to Aul16, which exists in its be roughly analogous to theinverted repeatsto which wild-type form in these two exceptional recombinants, Au217I and II map which can generate rarecrossovers the P2 17c probe in both cases detects the 2.9-kbp t- between wild-type and t-haplotypes in In(l7)2 (HERR- type TuqI fragment (Figure7C). This places the break- MANN,BARLOW and LEHRACH1987). points of these two chromosomes between Aul16/TuI The P217c probe shows that the twsx breakpoint in and Au217rs. Furthermore, since tWrxhas lost the 2.9- t-haplotypes occurred more distal to those of tu'"' and kbp Tug1 fragment detected by P2 17c, we can order tw/2x It is possible that all three exchanges occurred proximal endmarkers in t-haplotypes as: cen . . . . (Aull6/Tul) . . . Au217rs . . . Tctex-I . . . (Au9/ proximal to In(l7)l in t-haplotypes since the position Leh48). of Au217rs with respect to the proximal breakpoint of In(l7)I is not known. Onthe other hand, if t- chromatin (which is defined here as DNA under re- DISCUSSION combination supression whether it is within an inver- The five chromosome 17 markers described here sion or adjacent to one)is being lost by recombination, are located in the interval between Tu1 and Leh66D. it may be informative to assess the impact of this loss This region encompasses In(l7)l and 2 of t-haplotypes on the transmission ratio of t-haplotypes, as others Deletion Associated with Associated Deletion qk" 189 have proposed a locus, Tcd-I, at the proximal end of proposed, based on analysis of the t6 partial distal t- t-haplotypes which influences this phenotype (LYON haplotype and derivative chromosomes when hetero- 1984; SILVERand REMIS1987). zygous with a complete t-haplotype. The t6 haplotype If the proximal position of Au217rs in t-haplotypes is a partial t-haplotype which carries wild-type chro- is a consequence of a transposition-like event as we matin proximal to the breakpoint between Lehll9I propose, we should not expect an allele at the corre- and Leh66EI (HERRMANN,BARLOW and LEHRACH sponding site in wild-type chromosomes.However, 1987). Since t6 retains the 3.3-kbp wild-type band with DNA flanking the insertion site in t-haplotypes could the P217c probe (datanot shown), Au217I should yield a markerat this locus in wild-type chromosomes. map proximal to this breakpoint. Evidence that t6 has The band intensity of the transposed Au217rs se- lost some of the t-genes involved in sterility has been quence indicates it is amplified in copy number, per- presented by LYON (1986). It may be that the wild- hapsthree- to fivefold. A similar event has been type copy of the gene at Au217I (assuming that it described by DISTECHE,GANDY andADLER (1 987). In expresses at least one of the two transcripts) is one of this case, a 20-copy 9-kbp nonretroviral element in the loci involved. inbred mice was shown to have been translocated from Detection of a gene family expressed during de- the X chromosome centromeric region to an autoso- velopment: The Au76 locus contains one member of mal centromeric region where it is amplified about a small gene family that maps to the region deleted in fivefold. These events occured independently in five the partial t-haplotype, Pb2.We have identifieda seperate mouse strains, often on a different autosome. sequence of at least 300-bp shared among the family Also, UEHARAet al. (1990)identified anelement members which is responsible for the observed cross- (TSE, t-specific element) amplifiedto different extents hybridization. In situ hybridization with the 1.4-kb in different t-haplotypes. Some t-haplotypes carried cDNA indicates that message is present as early as 6.5 only a few copies while others carried over 100. In days P.c., and throughout embryogenesis. However, neither case, however, was a transcription unit iden- the signal becomes confined mainly to the CNS in late tified. stages (0. RHO, unpublished data). This expression The Au217 locus is not expressedin t-haplotypes: data does not distinguish among messages from dif- The P2 17a and P2 17b genomic probes show that the ferent loci. Au2Z7 genomic DNA is duplicated in wild-type mice The quaking locus: Analysis of the region between (loci: Au217I and Au217II), but is single copy in t- T and Dl 7RP17 has suffered from a lack of markers. haplotypes (locus: Au217A). The mappingdata are In particular, the qk locus, -0.3 cM proximal to RP17 consistent with a position onthe inverted repeats is an interesting target for cloning. The genetics of found in wild-type chromosomes (HERRMANN,BAR- the qk alleles available suggest the existence of one or LOW and LEHRACH 1987).We have verified this data more genes at this locus functioning in myelination of by linking Au217I/II with Leh66EI/EII and Lehll9Il the central nervous system, spermiogenesis and em- IZ on pulsed-field gel fragments in wild-type mice (data bryogenesis. When homozygous, the original sponta- not shown). neousmutation, qk", causes dysmyelination of the The Au217I and II loci encode two testicular cell CNS,resulting in the quakingphenotype (SIDMAN, transcripts in wild-type mice. (We have notdeter- DICKIEand APPEL 1964),and completearrest of mined if they are expressed from both loci or only spermiogenesis resulting in male sterility (BENNETTet one of the loci). These transcripts are not expressed al. 197 1). There are five independent ENU-induced from the Au217A locus in t-haplotypes or from the alleles of qk which are presumed to be point mutations. transposed and amplified copies in t-haplotypes. When these alleles are heterozygous with qk", quaking Hence, no transcript could be detected in testis RNA but fertile males are obtained. Homozygosity of the from f/tymales or from p/tand TZZH/tmales since ENU-induced alleles results in embryonic lethality PIrand TZZHdelete both wild-type copies. Could these (JUSTICE and BODE1988; KING and DOVE199 1). genes play a role in the sterility or transmission ratio Preliminary pulsed-field gel data indicate that the distortion associated with t-haplotypes? These pheno- markers AulO and AuZ19 can be physically linked to types are very likely the result of alterations in multi- the RPZ 7 locus. Both are probably proximal to RPI 7 ple factors acting in a concerted fashion (LYON 1984, in wild type. We have shown that at least 40-kbp at 1986; SILVERand REMIS1987). Because ?O'/t males the Au119 locus is deleted in qk" homozygotes, while are fertile (BENNETTand ARTZT 1990),loss of expres- the neighboring locus, AulO, is not. Furthermore, the sion from Au2Z7A alone is not sufficient to induce RPI7 locus is not deleted in qk"/qk" mice (data not sterility; it may be that the loss of these transcripts in shown). The size of the deletion has not yet been complete t-haplotypes acts in combination with muta- ascertained, however it is not large enough to signifi- tions at otherproximal or distal loci to produce steril- cantly distort recombination frequencybetween T and ity in f/tymales. These sorts of interactions have been RPI 7(KING and DOVE199 1).No transcripts in brain, 190 T. Ebersole, 0.Rho and K. Artzt testis or embryo have been found encoded at Au119 MICKOVA,199 1 Genetic mapping of the t-complex region on or AulO at the level of sensitivity of northern blots. It mouse chromosome 17 including the Hybrid sterility-1 gene. Mammal. Genome 1: 84-91. seems plausible that the male sterility associated with HAMMER,M. F., J. SCHIMENTI andL. M. SILVER,1989 Evolution gk"/qk", but not found with the ENU-induced muta- of mouse chromosome 17 and the origin of inversions associ- tions, could be due to a deleted gene necessary for ated with t-haplotypes. Proc. Natl. Acad. Sci. USA 86: 3261- spermiogenesis. However, the association of the quak- 3265. ing phenotype with a deletion presents a dilemma; HERRMANN,B. G., D. P. BARLOWand H. LEHRACH,1987 A large invertedduplication allows homologousrecombination be- since a presumed point mutation in the qk structural tween chromosomes heterozygous for the proximal t-complex gene is an embryonic lethal, it is difficult to imagine inversion. Cell 48: 813-825. that a null mutation resulting from a deletion would HERRMANN,B., M. BU~AN,P. E. MAINS,A,". FRISCHAUF,L. M. be homozygous viable. More likely, the deletion inter- SILVER andH. LEHRACH,1986 Genetic analysis of the prox- feres with regulatory elements necessary in the CNS, imal portion of the mouse t-complex: evidence for a second inversion within t-haplotypes. Cell 44 469-476. butnot with those neededduring embryogenesis. HOWARD,C. A,, G. R. GUMMERE,D. BENNETTand K. ARTZT, Based on these assumptions, one could predict that 1990 Genetic and molecular analysis of the proximal region the structural locus for qk should be found on one of the mouse t-complex usingnew molecular probes and partial side or the other of the deletion breakpoint close to t-haplotypes. Genetics 126: 1103-1 114. JOHNSON,D. R., 1974 Hairpin-tail: a case of post-reductional gene the junctionsite in qk"/qV DNA. action in the mouse egg? Genetics76: 795-805 JUSTICE, M. J., and V. C. BODE,1988 Three ENU-induced alleles The authorswould like to thank Y. C. HSU for use of the 10.5- of the murine quaking locus are recessive embryonic lethal day mouse embryo cDNA library, M. JUSTICE for mapping of the mutations. Genet. Res. 51: 95-102. Au76-derived cDNA,J. CAMPSfor helpin preparing the manuscript KING, T. R.,and W. F. DOVE, 1991 Pleiotropicaction of the and P. CENTILLI for technical assistance. This workwas supported murinequaking locus: structureof the qk" allele.Mammal. by grants CA21651 and HD10668 from the National Institutes of Genome 1: 47-52. Health. LYON,M. F., 1984 Transmissionratio distortion in mouse t- haplotypes is due to multiple distorter genes acting on a re- LITERATURECITED sponder locus. Cell 37: 621-628. LYON,M. F., 1986 Male sterility of the mouse t-complexis due to BARIARZ,B. S., 1983 Deletion mapping of the T/t complex: evi- homozygosity of the distorter genes.Cell 44: 357-363. dence for a second region of critical embryonic genes. Dev. MANN,E. A,, L. M. SILVERand R. W. ELLIOT, 1986 Genetic Biol. 95: 342-351. analysis of a mouse T complex locus that is homologous to a BARLOW,D. P., R. ST~GER,B. G. HERRMANN,K. SAITOand N. kidney cDNA clone. Genetics 114 993-1006. SCHWEIFER,1991 The mouse insulin-like growth factor type- SARVETNICK, N., S. H. Fox, E. MANN,P. E. MAINS,R. W. ELLIOTT 2 receptor is imprinted and closely linked to the Tme locus. and L. M. SILVER,1986 Non-homologouspairing in mice Nature 349: 84-87. heterozygous for a t-haplotype can produce recombinant chro- BENNETT,D., 1975 The T-locus of the mouse. Cell 6 441-454. mosomes with duplications and deletions. Genetics 113: 723- BENNETT,D., and K. ARTZT, 1990 Deletionanalysis of male 734. sterility effects of t-haplotypes in the mouse. Genet. Res. 56: SELBY,P. B., 1973 X-ray-induced specific-locus mutation rate in 179-183. newborn male mice. Mutat. Res. 18: 63-75. BENNETT,D., L. C. DUNN,K. SPIEGELMAN,K. ARTZT,J. COOK- SHIN, H.-S.,L. FLAHERTY,K. ARTZT,D. BENNETTand J. RAVETCH, INGHAM and E. SCHERMERHORN,1975 Observations on a set 1983 Inversion in the H-2 complex of t-haplotypes in mice. of radiation induced dominant T-like mutations in the mouse. Nature 306: 380-383. Genet. Res. 26 95-108. SIDMAN,R. L., M. M. DICKIE andS. H. APPEL,1964 Mutant mice BENNETT, W.I., A. M. GALL,J. L. SOUTHARD and R.L. SIDMAN, (quaking and jimpy) with deficient myelination in the central I97 1 Abnormal spermiogenesis in quaking, a myelin-deficient nervous system. Science 144: 309-3 1 1. mutant mouse. Biol. Reprod. 5: 30-58. SILVER,L. M.,and D. REMIS,1987 Five ofthe nine genetically Committee on the Mouse Chromosome 17, 1991 Maps of mouse defined regions of mouse t-haplotypes are involved in trans- chromosome 17First report. Mammal. Genome 1: 5-29. mission ratio distortion. Genet. Res.49: 51-56. DISTECHE,C. M., S. L. CANDYand D. M. ADLER, UEHARA, H., T. EBERSOLE, D. BENNETTand K. ARTZT, 1990 Submegabaseclusters ofunstable tandem repeats 1987 Translocationand amplification of an X-chromosome DNA repeat in inbred strains of mice. Nucleic Acids Res. 15: unique to the Tla region of mouse t haplotypes. Genetics 126: 4393-4401. 1093-1 102. VOJTISKOVA,M., V. VIKLICKY,B. VORACOVA,S. E. LEWISand S. EBERSOLE,T.,F. L~~andK. ARTZT, 1992 New molecular markers GLUECKSOHN-WAELSCH,1976 The effects of a t-allele (Y') in for the distal end of the t-complex and their relationships to the mouse on the lymphoid system and reproduction. Em- mutations affecting mouse development. Genetics 131: 175- J. bryol. Exp. Morphol. 36: 443-45 1. 182. WINKING,H., andL. M. SILVER,1984 Characterization of a ERICKSON,R., S. LEWISand K. SLUSSER,1978 Deletion mapping recombinant mouse t-haplotype that expresses a dominant le- of the t-complex of chromosome17 of the mouse. Nature274: thal maternal effect. Genetics 108: 1013-1020. 163-164. IOREJT, J.. V. VINCEK,J. KLEIN, H. LEHRACH andM. LOUDOVA- Communicating editor: N. A. JENKINS