Functional Analysis of Mutations of Murine Chromosome 17 with the Use of Tertiary Trisomy

Functional Analysis of Mutations of Murine Chromosome 17 With the Use of Tertiary Trisomy

Anatoly Ruvinsky, Alexander Agulnik, Sergei Agulnik and Margarita Rogachova Institute of Cytology and Genetics, Academy of Sciences of the USSR, Siberian Division, Novosibirsk 630090, USSR Manuscript receivedJune 5, 1990 Accepted January 7, 1991

ABSTRACT Analysis of the functional nature of mutations can be based on comparisons of their manifestation in organisms with a deletion or duplication of a particular chromosome segment. With the use of reciprocal translocation T(16;17)43H, it is feasible to produce micewith tertiary trisomy of the proximal region of chromosome 17. The mutations on chromosome17 we tested included brachyury (T),hairpin tail (T*P),kinky (Fu'"),quaking (qk), tufted (tf),as well as tct (t complex tail interaction), and tcl (t complex lethal) that are specific to t haplotypes. The set of dominant and recessive mutations was assigned to two groups: one obligatory, manifesting itselfin the phenotype independently of the number of normal alleles in di- and trisomics, and the other facultative, phenotypically manifesting itself depending upon the dosage of mutant alleles. A model was derived from analysisof the interaction of the T and ThP mutations with t haplotypes. It seeks to explain the morphogenetic effects of the mutations observed in mice of different genotypes. The tir gene is postulated to reside on chromosome 17 within its framework. It is suggested that the gene dosage ratio at the tir and tct loci " determines tail length.

UTATIONS can be analyzed by nontraditional tfmutation), T tf/t6 +, T tf/tj2 +, ThP/+,+ qk/t12 +, homo- methods. One such method is comparison of zygotes for Robertsoniantranslocations Rb( 16.17)7Bnr M (Rb7Bnr/Rb7Bnr), Rb(8.17)lIem (RblIem/RblIem), for the manifestation of genesin the mono-, di-, and reciprocal translocationsT( 16; 17)43H (T43H/T43H). The trisomic condition. The method has been applied in mice bearing mutations in translocations T T43H/+ T43H, studies of mutations in Drosophila (MULLER 1932; Rb7Bnr t6 +/+ T tfhave been produced earlier (AGULNIK, ROBERTS1976). Its application in mammals, however, AGULNIKand RUVINSKY1985, 1986). The proximal partial has been limited because almost all the known forms haplotypes tN' and tN2,derived from tj2 and t6, respectively, have no lethal factors (AGULNIKand RUVINSKY1989). The of mono- and trisomy are lethal(DYBAN and BARANOV partial tN'" haplotype located in translocation T43H origi- 1987). The possibility of producing viable mice with nated from t6 haplotypes in T T43H/t6 + 0 X Rb7Bnr t6 +/ tertiary trisomy 17 (trisomyfor only part of the chro- + + T43H ~3cross. The tN" haplotypes retained the region mosome) by use of translocation T(16";17)43H has interacting with the T mutations of the original t6 haplotype been reported a few years ago (FOREJT,CAPKOVA and but lost its lethal factor. The mice with different sets of mutations and chromo- GREGOROVA1980). This possibility has not yet been some rearrangements involved in the crosses are listed in exploited in studies of the interaction between genes Table 1. They all have been derived from the above strains situated in the portion of the murine genome that is and stocks. "sprinkled over with different mutations" exhibiting Cytogenetic analysis: The phenotype of the offspring unusual properties(KLEIN 1986; Figure 1). The prox- resulting from the testcrosses was identified at the age of 1.5 months, thereafter bonemarrow biopsieswere per- imal region thechromosome 17 containsthe of t formed for karyotype analysis(UDALOVA 197 1). The mitotic complex with a set of dominant and recessive muta- chromosome preparations were obtained and C-banded ac- tions manifesting themselves morphologically in the cording to the standard techniques (DYBANand BARANOV most conspicuous fashion (SILVER1985). 1987). We have carried out and report here, functional analysis of a number of mutationssituated in the RESULTS proximal region of chromosome 17. The analysis in- Use of tertiarytrisomy in genedosage study: cludes comparisons of the functioning of mutant and Figure 2 is a scheme showing how tertiary trisomics normal alleles in organisms having them in different wereproduced by dintof reciprocal translocation dosage ratios and combinations. T43H. The localization ofthe breakpoints of this rearrangement is noteworthy: it is at about a third of MATERIALS AND METHODS the chromosome away from the centromere in chro- Mice: The following mouse stocksand strains were used mosome 17, and it is at the precentromeric hetero- BTBR - + Fu tf/+ + tf, TFN - tf/tf (homozygotes for the chromatin region in chromosome 16. As a result, a

Genetics 127: 781-788 (April, 1991) 782 A. Ruvinsky et al. Rb (8. f ?) IIem hyperploidgametes with two doses of the proximal portion of chromosome 17 in T43H/+ female heterozygotes. Fertilization of these eggs with haploid sperm produces a zygote with tertiary trisomy. Differentmutations introduced into aparental chromosome 17 can generate a large set of trisomics ]Thp consisting of mutant gene combinations designed for our particular purposes. Crossing over in this region in females does not affect the allelic constitution of t *-to the trisomics because gametes with partial disomy carry both homologs in the proximal region of chromosome 17. tCP Males heterozygous for T43H are sterile; hetero- zygosity for Robertsonian translocation Rb( 16.17)7- 1 Bnr restores their fertility. In T43H/Rb7Bnr males, crossing over in chromosome 17 is completely sup- pressed (FOREJT,CAPKOVA and GREGOROVA1980). To ensureunambiguous identification of offspring genotype, the chromosomes of theparents were marked with Robertsonian translocations. Table 1 shows data for the crosses involving female heterozy- FIGURE 1.-Genetic c map of chromosome 17 of the house mouse, gotes for T43H and different mutations. Although showing the loci and mutations usedin the experiments: Cen - the number of offspring and that of trisomics identi- centromere; Rb(8.17)l Iem, Rb( 16.17)lBnr-Robertsonian translo- fied varied, we recovered individuals of appropriate cations; T( 16;17)43H-reciprocal translocation, arrow indicates genotypes. The total number of identified trisomics breakpoint; to- t", t haplotypes; tct, t complex tail interaction; tcP, t complex lethal-6; T, brachyury; qk, quaking; Fu'", fused kinky; Fukb, was 116.Their occurrence frequency was 12.1 & knobbly; tJ tufted; H-2, major histocompatibility complex; ThP 1.3%, on the average, in those crosses in which no (hairpin tail) and thZo,chromosome 17 deletions. The numbers single class of offspring died. indicate distance (in cM) in structurally normal chromosomes. The phenotype of the trisomics is characterized by a retardation in the development, in altered propor- large productof the translocation comprises the prox- tions of the visceral and cranial portions of the skele- imal portion of chromosome 17 and the entire chro- ton, defective formation of the vertebral column (Fig- mosome 16. During the meiotic disjunction of the ures 3 and 4) and a sharp decrease in the fertility of chromosomes, there arises acertain percentage of both sexes.

TABLE 1 Production of mice with tertiary trisomy 17

Trisornics progeny

Total No. of Genotype Po Genotype 88 progeny trisornics Percent 1. Fu" If+/++ T43H Rb7Bnr + +/Rb7Bnr + + 165 16 10+2 2. Fu'" If+/++ T43H + Fu" tflRb7Bnr + + 144 10 7+2 3. Fu" tf+/+ + T43H tf/lf 36 8 22 f 7 4. If+/+T43H tfltf 13 2 15 + 10 5. tf+/+ T43H Rb7Bnr +/Rb7Bnr + 70 6 9+3 6. + T43H/tf+ Rb7Bnr t6/+ + 38 2 5+3 7. T + T43H/+ qk + Rb7Bnr + +/Rb7Bnr + + 26 6 3+8 8. T + T43H/+ qk + Rb7Bnr + +/+ + qk 107 8 7+3 9. ThP +/+ T43H Rb7Bnr +/Rb7Bnr + 27 8 30 + 9 10. ThP +/+ T43H + Thp/Rbl Iem + 35 7 20 f 7 11. T T43H/+ + Rb7Bnr +/Rb7Bnr + 61 9 15+7 12. T T43H/+ + + T T43H/Rb7Bnr + + 58 3 5+3 13. T T43H/t6''" + Rb7Bnr +/Rb7Bnr + 80 10 13+4 14. + T43H/t" + + T T43H/Rb7Bnr + + 54 10 19 f 5 15. T T43H/t"+ + Thp/RblIem+ 1 1 100 16. tNfoT43H/t6+ + Thp/RblIem + 9 4 44?18 17. T T43H/t6 + Rb7Bnr t6/+ + 8 1 13+13 18. T T43H/t6(") tNf(N2) tf/tNI(N2) f 24 5 21+8 MutationsChromosome of Murine 17 783

N" 16 I/ 18" /6 I6 I/ i? 7-43/+ +/+

FIGURE2.-Experimental scheme for obtaining the mice with tertiary trisomy of chromosome 17 using the T(16;I 7)43H recip rocal translocation. To illustrate the portions of the translocation chromosomes derived from chromosome 16 and 17, the chromosome I7 dark G-bands are shown as solid black and the chromosome 16 bands are shown as hatched. TABLE 2 Phenotypes of mice with Fu' mutation

Penetrance Genotype" Phenotype Genotype" (W) ~~ +ID? Normal 100 Fu"/+ tail Kinky 90-95 Fu"/+/+ tail(25) Kinky a4 Fu'/Fu"/+ tail(5)' Kinky 100 Fuk'/Fub Embryonic lethal 100 Fu'"/Dfd Embryonic lethal 100 Number of trisomics scored is indicated in parentheses. Df- t"" deletion. ' Obtained in cross 2, Table 1. There may be a few Fu" tf +/+ + T43H/+ + + mice among thesetrisomics as a result of crossin FIGURE3.-Mouse with Fu" tf/Fu" tj/+ + genotype on the left, over in Rb7Bnr + +/+ Fu tfmales between centromere and Fu E Fu" tf/++ sib on the right. (less than 8%). The breeding data, considered together with the developmental stage (GLUECKSOHN-SHOENHEIMER cytogenetic and phenotypic descriptions, allowed us 1949; LYON and BECHTOL1977; GREENSPANand to carry out Mullerian gene dosage studies of the O'BRIEN 1986). Taking all this into consideration, it various mutations. becomes apparent that the phenotypic manifestation Functional analysis of the mutations kinky,tufted of Fukiis unrelated to a hyperfunction of the gene in and quaking:Kinky (Fuki)is a mutation that maps 10 question. If this were the case, mice of Fuki/Fu'/+ cM distal to the centromere of chromosome 17. It genotype would exhibit more severe morphological produces kinky-tailed phenotype and otherabnormal- disorganization than Fu'/Fuki homozygotes. ities in heterozygotes. Homozygotes die on days 8-9 From this comparative analysis of mice ofdifferent of embryonic development (GREEN1989). genotypes, it follows that themanifestation ofthe Fuki Table 2 shows the data for themanifestation ofFuki mutation is due to some altered product competing in mice witheuploid and aneuploid chromosome sets. with the product of the wild type allele. The effect of There aretwo noteworthy observations. First, trisom- the Fu' mutation on the viability and phenotype of ics of Fu'/+/+ genotype show a mutant phenotype. mice of different genotypes can be ordered as: Fu'/ This was interpreted as indicating a highlevel of Fuki= Fuki/Df> Fuki/Fuki/+2 Fu ki/+ 2 Fu ki/+/+ > dominance of the Fu' gene. The expressivity of the +/+ = +/Df(Df - th2' deletion). Fukigene in trisomics is almost the same as in Fuki/+ The results for therecessive mutation tufted, which heterozygotes. It may be inferred that the Fukimuta- is closely linked to Fukiand produces specific waves of tion does not produce a hypofunction of the normal hair loss (LYON 1956), are given in Table 3. In addi- allele. The second observation concerns the viability tion to the earlier described genotypes, we have pro- of individuals of Fuki/Fuki/+genotype; these, how- duced trisomics carrying the mutant gene in a single ever, were much more severely disorganized morpho- (tf,/+/+)dose. These mice were phenotypically nor- logically (Figure 3). It is known that homozygotes mal. (Fu'/Fu') and hemizygotes (Fu'/-) die at an early Some trisomics with two doses of tufted (tfltf,/+) 784 A. Ruvinsky et al.

TABLE 3 TABLE 4 Phenotypes of mice with tfmutation Phenotypes of mice with qk mutation

Penetrance Penetrance Genotype" PhenotypeGenotype" (%) Genotype" Phenotype (%) +/of" Normal 100 +/of" Normal 100 tfl+ Normal 100 qk/+ Normal 100 tf/+/+ (27) Normal 100 qk/+/+ (1 1) Normal 100 tf/tf/+>33 (9)" Tufted qk/qk/+ (3)" Normal 100 tfM Wf Tufted 100 qklqk, qk@f Normal 100

Number of trisomics scored is indicated in parentheses. a Number of trisomics scored is indicated in parentheses. 'Of - thZ0deletion. Df - Thp deletion. Five trisomics obtained in cross 2, Table 1. There may be a few Obtained in cross 6, Table 1. There may be a few T + T43H/ Fu*' tf +/+ + T43H/+ + + mice among these trisomics as a result + qk +/+ + + mice among these trisomics as a result of crossing of crossing over in Rb7Bnr + +/+ Fu" tfmales between centromere over in Rb7B nr +/+qk males between centromere and qk (less than and tf(less than 9%). 4%). were phenotypically mutant, otherswere normal. The TABLE 5 penetrance of the tf mutation in these trisomics was estimated as 33%. The pattern of hair loss was differ- Phenotypes of mice with T, Thp and tct mutations ent in mice with overtly manifest mutation and in tf/ Phenotype" tf homozygotes (Figure 3). As in the case of the Fuki mutation, tfdoes not produce a loss of function (nor- Short Normal tail Tailless mal phenotype in +/Of mice), nor does it cause a Genotype hyperfunction (normal phenotype in tf/+/+mice and Tlt +/t TI+ decrease in the penetrance of the tf/tf/+ mice). Con- t/t ThP/+ Thp/t sequently, formation of the altered product, probably tl+/+ (2) TIT/+ (3) TITb (2) t/t/+ (2) Thp/Thp/+(2) TAP/ThP/t(1) competing with a certain concentration of the product v+/+ (23) Tltlt (6) of the wild-type allele, is to reiterate, a featureof the Thp/+/+(1 3) Thp/t/t(2) mutation Fukiand that of tf, too. T/t/+ ( 18)

The ordering into seriesa ofdecreasing phenotypic a Number of trisomics scored is indicated in parentheses. severity of the tfallele is as follows: tf/tf= $,/Of> tf/ tf/+ 2 tf/+/+=tf/+ = +/Of. tailed, this is in contrast to T/+ individuals whose tails Quake (qk) is located 4 cM from the centromere. are short. The same is truefor the Thpdeletion- Heterozygotes have a normal phenotype, hemi- and bearers. Miceof T/T/+, Thp/Thp/+ genotypeshave homozygotes have marked rapid tremor of the body, short tails and are viable, whereas TIT and Thp/Thp that appears at about 10 days after birth (SIDMAN, homozygotes are characterized by a recessive lethal DICKIEand APPEL1964). Trisomics having a single effect (JOHNSON 1974; BENNETT1975). Thus,the (qk/+/+), or two (qk/qk/+) doses of the mutant allele addition of the wild-type allele to trisomics abolished are phenotypically normal (Table 4). The qk mutation, the phenotypic manifestation of the mutation T and therefore, does not cause a hyperfunction of the rel- ThP in heterozygotes and its lethal effect in homozy- evant gene. The more plausible consequences of the gotes. Comparisons of the mutations T and Thpdis- mutation seem to be a complete or partial loss of the close full similarity of the phenotypic severity of all gene function, or, perhaps, a formation of an altered the examined genotypes (Table 5). Since ThPis a product not capable of competing with the normal deletion, the conclusion that the mutation T and ThP allele in the given concentration range. The data produces a complete loss of function appears the most obtained allowus to arrive at the ordering of the justified. phenotype severity of the mutant allele qk into the The tct mutation seems to be due to a decrease in series: qk/qk = qk/Df > qk/qk/+ = qk/+/+ = qk/+ = functional activity relative to the normal allele. This +/Of -(Of ThP deletion). is apparent from the following comparisons.T/+ mice Functional analysis of the mutations T, Thp,tct, have short tails; the addition of the tct allele produces tcI6: The dominant mutations T, Thp,when heterozy- tails whose length is normal in T/tct/+ trisomics (Fig- gous, produce short tails, and, when in a compound ure 4). T/tct/tct and Thp/tct/tct mice (obtained in with the tct mutation (specific to the t haplotypes), it crosses 17, 18 Table 1) have short tails, i.e., they are results in taillessness(SILVER 1985). ThPis an extensive phenotypic equivalents of T/+, Thp/+(Figure 5). For deletion stretching a distance of about 3-4 cM in the these reasons the tct mutation is thought to be acting precentromeric region of chromosome I7 (BENNETT in the same direction as the wild-type allele, although 1975).We have produced trisomics carrying the mu- not as severely. tations set out in Table 5. T/+/+trisomics are normal- The results of analysis of +/t6/t6 and t6/t6/t6 em- MutationsChromosome of Murine I7 785

TABLE 6 Phenotypes of mice with tcZ6 mutation

Genotype" Phenotype Genotype" tcl"+ Normal tcl 'I+/+( 16) Normal tCP/tC16/+ (1) Normal tcl"1cl~ Lethality

a Number of trisomics scored is indicated in parentheses. was further support of the conclusion that the tcL6 is in close vicinity to tf, proximal to the T43H breakpoint.

DISCUSSION

Genetic basis of T-tct interaction: The t complex that occupies most of the proximal part of chromosome 27 has stirred theinterest of geneticistsfor more than 50 years (SILVER1985). The extent of the complex is about 17 cM. More than 10 dominant mutations within the t complex have been described (BEN- NET 1975). Their effects are variable ranging from defects of the vertebral column, complete taillessness to stumps, and early embryonic death of homozygotes (GREEN1989). Chromosomes carrying extensive in- FIGURE4.-Mouse with + T T43/+ t" +/Rb7 + + genotype on verted regions, t haplotypes, have been isolated from the left, + t"/Rb7 + sib on the right. natural populations of mice (KLEIN 1986). The presence of at least four inversions suppresses crossing over, thereby providing the persistence of the allelic constitution of the t haplotypes (ARTZT,SHIN and b BENNETT1982; HERMANNet al. 1986; JUSTICE and BODE1988a; HAMMERet al. 1989). The results of our analysis demonstrate that Tis a complete loss-of-function mutation like the deletion Thp.The lines of evidence are (1) loss of dominance in trisomics: mice of T/+/+ genotype are normal-tailed; (2) overcoming of FIGURE5.-Mouse with T + T43H/t" + +/t"'? tf + genotype. the lethal effect in TIT/+ tertiary trisomics and the This mouse producedfrom T T43H/tIZ + 0 X 1"' tf/t"2 tfd crosses. same tail length as in T/+; (3) identity of phenotypic bryonic development suggest that thetcZ6 lethal effect severity of the T mutation and ThP.Our previous data may be caused by a hypofunction of the mutant allele concerning the influence of T on the segregation of (MCGRATHand HILLMAN1982). Our data concerning homologs in females heterozygous for Robertsonian the manifestation of the recessive lethal factor of the translocations involving chromosome 27 also suggest t6 haplotype (tcL6)are presented in Table 6. Trisomics that the T mutation might have resulted from a chro- with two doses of the lethal t6 haplotype (+ T T43H/ mosomal rearrangement (RUVINSKYet aZ. 1987, + t6 +/Rb7Bnr t6 +) are viable. The wild-type allele 1988). The data of HERMANNet al. (1990) really resides on the chromosome + T T43H (17'6) only. demonstrated that T is a result of deletion. This large product of translocation T43H (17'6) en- The tct (t complex tail interaction) mutation is an compasses the proximal part of chromosome 27 and obligatory element of the t haplotypes. It does not chromosome 26 in its entirety. The breakpoint in manifest itself when in heterozygous or homozygous chromosome 27 is located about1 cM distal to tf condition. T/tct compounds are tailless. Absence of (FOREJT,~APKOVA and GRECOROVA1980; AGULNIK, recombination between T and tct argues for their AGULNIKand RUVINSKY1986). The tcZ6 is close to tf allelism (JUSTICE and BODE 1988b). There arecontra- (ARTZT1984). The normal allele corresponding to dictory data not exluding the possibility that they may the lethal tc16 has been incorporated into the large be closely linked loci (AGULNIK,RUVINSKY and Po- product during theformation of translocation T43H, LYAKOV 1990). To our knowledge, experimental data as may be judged by the viability ofthe trisomics. This explaining why T-tct interact are scant, if any. 786 A. Ruvinsky et al.

TABLE 7

Dosage of tir and tct loci and their ratio in mice with different genotypes

~ ____~ Allelic dosage Coeffkient for Postulated dosage ratio of Phenotype Genotype genotype ti?+ tct tct fir and tcf Normal tail +I+ tir+tct+/tir+tct+ 2 2 0 1 +It tir+tct+/tir+tct 2 1 1 1.2 < k < 1.33 t/t tir+tct/tir+tct 2 0 2 1.5

This functional analysisdiscloses a situation that normal-short-tailless. What is the variation range of appears to be contrary to expectation. The presence the activity of tct compared to its wild type? To deter- of an additional chromosome with mutations T and mine the lower range, normal tailed mice with the ThPin T/+/+, Thp/+/+,and T/+/t does not affect the highest tirltct are to be compared with short-tailed phenotype of mice when compared to +/+, +/t, nor with the smallestvalues, ie., miceof t/t and Thp/t does it affect that of T/T/+, Thp/Thp/+when com- genotypes. The comparisons yield 2 tir+/2tct < 2 tir+/ pared to T/+ and Thp/+ disomics. From the data in 1 tct, the sign < indicates that short-tailed mice have Table 5, it follows that tct is a partial loss of function a higher value for tir/tct ratio, and, hence, the activity mutation, not significantenough to affect the normal- of tct > 1/2 tct+. Similarly, the upper range of the tailed phenotype of tctltct. T/tct/tctand Thp/tct/tctmice variation can be derived from comparison of short- are short-tailed, however. The question, then is: why tail ThP/ThP/+with tailless Thp/tmice. In this case the the completeloss-of-function alleles T and ThP do inequality for the tirltct ratio can be expressed as 3 affect the phenotype of the T/tct/tct and Thp/tct/tct tir+/l tct+ < 2 tir+/l tct; hence, tct < 2/3 tct'. Clearly, trisomics. the coefficient relating the activity of tct and tct+ lies A hypothesis is offered that would reconcile our within the 1/2 < m < 2/3 range. observations. There may exist a putative genetic fac- The conditional coefficient for the dosage ratios of tor nonallelic to T and tct residing on chromosome the tzr and tct genes is calculated by means of the 17; it may be involved in the determination of the formula K = nl/(n2+ mns), where n1,n2,ns, are tir+,tct, development of the tail phenotype. All the versions of and tct-allele number, m denotes the activityof tct chromosome 17 carry the tir (tail interaction) gene. compared to that of kt+. When the coefficient values Since Thp/t/ttrisomics are short-tailed as opposed to range from 1 to 2, mice with normal tail length fall t/t disomics, a chromosome with the deletion Thp into the same group, when they range from 2 to 3, affects tail length. Therefore, the tzr locus has to be short-tailed individuals fall into the same group. The outside the ThP deletion. The products of the tir and values range from 3 to 6 in the case of taillessness tct genes interact, thereby providing the normal phe- (Table 7). Thus, there is fairly good fit between tail notype at certain ratios of product activities that are length and gene dosage ratio. specific to +/+, +/t, etc. genotypes (Table 5). When In further treatmentof our hypothesis, we took into a deletion of the tct gene causes a dosage disequili- consideration that there are exceptional t haplotypes brium, tail development becomes defective. suppressing the effect of T and, thereby giving rise to Table 7 lists the genotypes and the attendant phe- normal tail phenotype. A case in point is twLub2which notypes, allele dosagesof the tir and tct genes. As the produces normal tailwhen in compound with T model postulates,the allelic combinations carried are: (WINKINGand SILVER1984). twLubZhas beendescribed tir+ tct+, wild-type chromosome; tir+tct, chromosome as a result of an unequal crossing over involving the with the t haplotype; tir+Df(tct), chromosome with wild and t haplotype-bearingchromosomes; it contains the mutation T or ThP.The model postulates that the duplications and deletions (SARVETNICKet d. 1986). activity of the tct allele (or its products) is lower than With this in mind, we thought it possible, by way of that of the wild allele tct+. As Table 7 shows, the tir/ analogy, thatthe constitution of turLubZmay be tct ratios increase with decreasing tail lengths: from tir+tct+tct,ie., the tct gene is duplicated, one allele tct+ Mutations of Murine Chromosome 17 787 being wild, the other tct is like the one present in the These exceptional allelic versions accommodate well original twlublhaplotype. In T/twLUb2compound gene into our model. Thus, T"/+ heterozygotes may be dosage would be 2 tzr+/(l kt++ 1 tct), ie., such that then described as tir+tct'/tir+tct+; and if the product would provide the formation of the normal pheno- of the tct" is inactive and competitive with the product type. of the tctf allele, the conditional gene dosage coeffi- The analogy may be extended further to explain cient would be 3-6 and mice would be tailless. the formation of the normal phenotype in T/tTu3com- The genetic behavior of the exceptional alleles sup- pound (STYRNAand KLEIN 1981). tTU3is presumably ports our hypothesis. A molecular characterization of a recombinant haplotype and its constitution may be the proximal portion of chromosome 17 would vali- tir+tct tct+, differing from twLubZin the order of the tct date it in the ultimate sense. alleles. An approach to classification of dominant and re- If so, the conditional coefficient for the gene dosage cessive mutation: The approach we presently apply ratio for T/tTu3genotype would be in the 1.2 < k < has been first used in genetic studies of Drosophila 1.3 range, and it would fall withinthe variation range melanogaster (MULLER 1932). The approach has been of the normal-tailed mice. Another exceptional hap subsequently widely applied in genetic investigation lotype th7 has been thought to contain a duplication of plants (GERSTELand MANN1966). The bewildering in the proximal part (LYONand MEREDITH 1964). complexity of the organization and function of the This gives reasonfor assuming a constitution of tir+tct genome prevented the application of thisapproach to tct for th7, and further extending the analogy to in- mammals. The presence of an additional chromo- clude th7. The behaviorof the th7 derivatives, the some, a portion of it, or a deletion, inmost cases partial haplotypes th3,th14,th15,thJ7(LYON and MERE- causes death at early stages ofembryonic development DITH 1964) when in compound with T has been also (DYBANand BARANOV1987). The finding of the explained by the presence of one or two tct alleles. T(16; 17)43H reciprocal translocation and the feasi- An alternative explanation is offered within the bility of using it to produce viable mice tertiary tri- framework of our hypothesis: a deletion of the tir somy for a partof chromosome 17, paved the way for locus would explain the properties of twLub2,tTu', and functional analysis of mutations located in this region. th7. Upon acceptance of that, the conditional coeffi- By varying the dosage of genes and the combination cient for the tirltct dosagewould decrease and T/ of alleles one can analyze the nature of the various twLub29Tu3,h7individuals would be phenotypically nor- mutations. mal. Tests of the properties of the trisomics would The results of this functional analysis of genes lo- probably prove which hypothesis, the deletion or du- cated on chromosome 17 allowed us to carry out a plication, is correct. classification of perhaps a more general significance. Thus, the situation envisaged above seems to be The above suggestiondoes not ignore the classical H. best explicable upon the assumption that there may Muller approach of classification of mutations. The be two alleles at the tct locus, tct+ and tct. May there mutations we considered are assigned to four groups. exist other allelic versions? As a matter of fact there (I) Mutations manifesting themselves in the pheno- are lines of evidenceindicating that tAE5,tW7'Jr1produce type independently of the number of normal alleles, short-tail phenotype in homozygotes (VOJTI~KOVAet termed obligate dominant (Fd'). (2) Mutations losing dominance in trisomics with two doses of the wild al. 1976; NADEAU,VARNUM and BURKART1989). allele, termed facultative dominant (T, Thp).(3) Mu- Upon the assumption that tctAE5and are func- tations like tfmanifesting themselves inthe phenotype tionally less active relative to tct allele, the conditional in trisornics, termed facultative recessive, i.e., gene dosage coefficient for tir and tct in tAE5/tAE5and tf,tf/+ tW7lJ~l/tw7lJ~Jhomozygotes would be 2-3. Thus, with prone to lose recessiveness depending on the ratio of the mutant to normal allele dosage. (4) Mutations not this assumption, our model's expectation of short-tail manifesting in the phenotype of trisomicsin the pres- phenotype is fulfilled. ence of at least one dose of a normal allele, termed Altered activity oftct gene may be accounted for by obligate recessive (qk, tcP). change in the number of its copies,each copy differing This functional classification ofmutations with mor- in activity (NADEAU,VARNUM and BURKART1989). phogenetic effects seems to broaden our understand- Exceptional allelic versions ofthe tct gene presum- ing of the nature of dominance and recessiveness. ably occur not only in the t haplotypes, but also in Dominance and recessiveness are then viewed not chromosomes carrying T dominant mutations. lllus- only as properties of characters, or results of gene trative in this respect appears to be the T" mutation interaction, but also as the manifestations of the in- resulting in a tailless phenotype in heterozygous mice. trinsic essence of the mutations. This mutation accomplishes itseffect by antagonizing the function of a wild-type allele, and it hasbeen The authors are grateful to A. FADEEVAfor translating this paper termed antimorph (MACMURRAYand SHIN 1988). from Russian into English. 788 A. Ruvinsky et al.

LITERATURE CITED KLEIN, J.,1986 Natural History ofMajor Histocompatibility Complex. John Wiley & Sons, New York. ACULNIK,A. I., S. I. ACULNIKand A. 0. RUVINSKY,1985 Mouse LYON,M. F. 1956 Hereditary hair loss in the tufted mutant of the News Lett. 73: 28-3 1. house mouse. J. Hered. 47: 101-103. AGULNIK,S. l., A. 1. AGULNIK and A.0. RUVINSKY,1986 Mouse LYON,M. F., and K.B. BECHTOL,1977 Derivationof mutant t- News Lett. 75: 39-42. haplotypes of the mouseby presumed duplication or deletion. AGULNIK, A.I., and A. 0. RUVINSKY,1989 New partial t haplo- Genet. Res. 30: 63-76. types of the mouse: characterization of thebasic properties (in LYON,M. F. and R. MEREDITH,1964 Investigations of the nature Russian) Genetika 25: 894-901. of t-alleles in the mouse. 11. Genetic analysis of an unusual AGULNIK,A. I., A. 0. RUVINSKY and A. V. POLYAKOV,1990 tctN mutant allele and its derivatives. Heredity19 313-325. mutation causing taillessness in mice, heterozygous for T gene MCGRATH,J., and N. HILLMAN,1982 The phenotypic expression (in Russian) Genetika (in press). ofa t6/t6/t6 genotype. J. Embryol. Exp. Morphol.69: 107-1 13. ARTZT, K., 1984 Genemapping within the T/t complexof the MACMURREY, A., andH.3. SHIN, 1988 The antimorphic nature mouse. 111. t-Lethal genes are arranged in three clusters on of the T' allele at the mouse T locus. Genetics 120 545-550. chromosome 17. Cell 39: 565-572. MULLER,H. J. 1932 Further studies on the nature and causes of ARTZT, K., H-S. SHINand D. BENNETT, 1982 Genemapping gene mutations, pp. 213-255 in Proceedings of the Sixth Inter- within the T/t complex of the mouse. 11. Anomalous position national Congress of Genetics. Brooklyn Carden, New York. of H-2 complex in t-haplotypes. Cell 28: 417-476. NADEAU,J. H., D. VARNUMand D. BURKART,1989 Genetic BENNETT, D.,1975 The T-Locus of the mouse. Cell 6: 441-454. evidence for two t complex tail interaction (tct) loci in t haplo- DYBAN,A. P., and V. S. BARANOV,1987 CytogeneticsofMammalian types. Genetics 122: 895-903. Embryonic Pevelopment. Clarendon Press, London. ROBERTS,P. A. 1976 The genetics of chromosome aberrations, FOREJT,J., J. CAPKOVAand S. GREGOROVA,1980 T(16;17)43H pp. 68-173 in The Genetics and Biology of Drosophila, Vol. la, translocationas a tool in analysis ofthe proximal part of edited byM. ASHBURNERand E. NOVITSKY.Academic Press, chromosome 17 (including T-t gene complex) of the mouse. London. Genet Res. 35: 165-177. RUVINSKY, A.O., S. I. AGULNIK,A. I. AGULNIKand D. K. BELYAEV, GERSTEL, D.U., and T. J. MANN,1966 Aspects of chromosome 1987 The influenceof mutations on chromosome 17 upon manipulation: a resume, pp. 1 13-1 20 in Chromosome Manipu- the segregation of the homologies in female mice heterozygous lations and Plant Genetics, edited by RILEYand K. R. LEWIS. for Robertsonian translocations. Genet. Res. 50: 235-237. Plenum Press, New York. RUVINSKY,A. 0..S. I. AGULNIK, A.I. ACULNIK~~~D. K. BELYAEV, GLUECKSOHN-SHOENHEIMER,S., 1949 The effect of a lethal mu- 1988 Structural changes of the homologuesa as possible cause tation responsible for duplications and twinning in mouse em- of abnormal disjunction in female mice heterozygous for Rob- bryos. J. Exp. Zool. 100 47-76. ertsonian translocations. Genet. Sel.Evol. 20 299-306. GREEN,M. C., 1989 Catalogof mutant genes and polymorphic SARVETNIK, N.,H. S. Fox, E. MANN,P. E. MAINS,R. W. ELLIOTT loci, pp. 12-403 in Genetic Variants and Strainsof the Laboratory and L. M. SILVER,1986 Nonhomologouspairing in mice Mouse, edited by M. LYONand A. SEARLE,Oxford F. G. heterozygous for a t-haplotype can produce recombinant chro- University Press, Oxford. mosomes with duplications and deletions. Genetics 113: 723- GREENSPAN, R.J., and M. C. O'BRIEN, 1986 Genetic analysis of 734. mutations at the locus in the mouse. Proc. Natl. Acad. fused SIDMAN, R. L.,M. M. DICKIEand S. H. APPEL, 1964 Mutant mice Sci. USA 83: 441 3-441 7. (quaking and jimpy) with deficient myelination in the central HAMMER,M. H.,J. C. SCHIMENTI~~~L. M. SILVER, 1989 On the nervous system. Science 144: 309-31 1. origins oft complex inversions. Proc. Natl. Acad. Sci. USA 86 SILVER, L. M., 1985 Mouse t haplotypes. Annu. Rev. Genet. 19: 3261-3265. 179-208. HERRMANN,B., M. BUCAN, P. E. MAINS,A,". FRISCHAUF, L.M. STYRNA,J., and J. KLEIN,1981 Evidence for two regions in the SILVER andH. LEHRACH,1986 Genetic analysis of the prox- mouse t complex controlling transmission ratios. Genet. Res. imal portion of the mouse t complex: evidence for a second 38: 315-325. inversion within t haplotypes. Cell 44: 469-476. UDALOVA,L. D., 1971 Themethod of bone marrow biopsy in HERRMANN,B. G., S. LABEIT,A. POUSTKA,T. R.KING and H. karyotypestudies of laboratoryanimals (in Russian). Arch. LEHRACH,1990 Cloning of theT gene required in mesoderm Anat. Gistol. Embryol. 9: 87-88. formation in the mouse. Nature 343: 617-622. VOJTISKOVA,M., V. VIKLICKY,B. VORACOVA,S. E. LEWIS,and S. JOHNSON, D. R, 1974 Hairpin-tail: a case of post-reductional gene GLUECKSOHN-WAELSCH,1976 The effects of a t-allele (tA"') in action in the mouse egg? Genetics76 795-805. the mouse on the lymphoid system and reproduction. J. Em- JUSTICE, M. L., and V. C. BODE, 1988a Genetic analysis of mouse bryol. Exp. Morphol. 36 443-451. t haplotypes using mutations induced by ethylnitrosourea mu- WINKING,H., and L. M. SILVER,1984 Characterizationof a tagenesis: theorder of T and q' is invertedin t mutants. recombinant mouse t-haplotype that expresses a dominant le- Genetics 120: 533-543. thal maternal effect. Genetics 108 1013-1020. JUSTICE,M. J. and V. C. BODE,1988b Mouse News Lett. 80 168- 170. Communicating editor: K. ARTZT