Copyright 0 1994 by the Genetics Society of America Co-Segregation of Intermale Aggression With the Pseudoautosomal Region of the Y Chromosome in Mice Pierre L. Roubertoux, Michele Carlier, Herve Degrelle, Marie-Claude Haas-Dupertuis, John Phillips and Rene Moutier U.R.A. 1294 C. N. R.S. Ginitique, Neuroginitique, Comportement, Uniuersiti Paris-V ReniDescartes, 75270 Paras, Cedex 06,France Manuscript received March 23, 1993 Accepted for publication September 1 1, 1993 ABSTRACT The sexual dimorphism of aggression has led atosearch for its Y chromosomal correlates. We have previously confirmed that initiation of attack behavior against a conspecific male is Y-dependent in two strains of laboratory mice (NZB and CBA/H).We provide evidence that the non-pseudoautosomal region of the Y is not involved and that only the pseudoautosomal region of the Y is correlated with initiation of attack behavior. The autosomal correlates also contribute to this behavior in an additive or interactive manner with the pseudoautosomal correlates. parsimonious hypothesis to reconcile both arguments: HE association of differences in agonistic behava non-pseudoautosomal region of the Y chromosome ior, defined as the initiation of attack behavior (Y”’””) is transmitted from father to son, whereas all against a conspecific male, with the variants of the Y or part of the other region of the Y chromosome chromosome in mice is the subject of a long-standing et al. 1990; CARLIER, ROUBERTOUX recombines with the X chromosome and is thus debate (CARLIER termed “pseudoautosomal region” (Y‘””). This sharand PASTORET 199 1; MAXSON1992a,b). On the one ing of the YPARby both males and females is compatible hand, three independentresearch teams using inbred with a commonY correlate for attackbehavior in both strains of laboratory mice have implicated the Y chrosexes, and the so-called “autosomal” co-acting or inmosome in initiation of attack behavior against conteracting genes would, in fact, be “pseudoautosomal,” specific males, either from the comparison of strains i.e., located on the YPARas previously suggested (CARcongenic for the Y chromosome (MAXSON, GINSBURG LIER et al. 1990). Malemice fromstrain NZB/BI1979; STEWART,MANNING and and TRATTNER NJGnc (abbreviated N hereafter)display more intense BATTY 1980) or fromresults of mendelian crosses and frequent spontaneous offensive behavior against (CARLIER, ROUBERTOUX and PASTORET 199 1). Moreconspecific males than CBA/HGnc mice (abbreviated over the Y chromosome has been implicated in differH) (CARLIER and ROUBERTOUX 1986). The frequency ences between lines selected for short vs. long attack of males initiating attack behavior is higher in N than latency from a populationof feral mice (Mus musculus in H for dyadic encounters with a standard opponent. 1984;VANOORTdomesticus; VAN OORTMERSSEN Moreover, when the maternal environment is kept MERSSEN, BENUSand SLUYTER 1992).On the other constant, this difference is not modified, showing that hand, a correlated response for female aggression in the observed strain difference has genetic and/or cylines differentially selected only for male aggression toplasmic correlates. We have previously demonhas been reported (HOODand CAIRNS1988) suggeststrated that atleast one locus on the Y chromosome is ing that this behavior may not be transmitted by a associated with attack behavior in these populations locus linked to the Y chromosome. It is worth noting (CARLIER, ROUBERTOUX and PASTORET 1991), but it that in the studies using inbred strains, the Y effect was not known whether this locus is located on the always acts concomitantly with the autosomal backYNPAR or on the YPAR.In the present study we describe and TRATTNER 1979; ground (MAXSON,GINSBURG a co-segregation of the YPARwith the initiation of et al. 1990). CARLIER attack behavior. The opposing conclusions drawn between the results of HOODand CAIRNSand the otherteams could be due to differences in behavioral testing, rearing MATERIALS AND METHODS conditions, or in differences in the genetic informaMice: Identified breeders from two highly inbred strains tion borne either on theautosomes or on the Y chroof mice were used: H and N. They were supplied by the mosome. Obligatory crossing over between the X and CSEAL (CNRS, Orlians la Source, France) at 127 and 1 14 the Y chromosomes (SIMLERet al. 1985; ELLISand brother X sister generations of mating, respectively, and GOODFELLOW 1989) at the male meiosis leads toa had been maintained using the same mating system, in our T Genetics 135 225-230 (January, 1994) 226 P. L. Roubertoux et al. laboratory, for a further 8 generations at the beginning of the experiment. Rearing conditions:The following conditions have been kept constant over the years: temperature 23.5" & 0.5"; photoperiod, 12:12 withlights on at 8:30 AM; food (IM UAR) and tap water ad libitum; dust-free sawdust bedding; weaning 30 & 1 days after birth. Each pregnant female was isolated from the mating cages. The litters having fewer than four pups were discarded and the others culled to six pups. At weaning, each male was housed with one female (generally a littermate) in an opaque cage (42 X 27 X 17 cm) until testing (66 If: 4 days after birth). Behavioral testing: The conditions of observation and measurement (described by CARLIERand ROUBERTOUX 1986) were maintained unchanged for all groups over the years. Briefly, the test was performed in a transparent cage (42 X 26 X 18 cm) with a transparent lid. The floor of the cage was covered with sawdust from several cages housing males of the tested groups. This procedure accelerates the appearance of the first attack (CARLIER and ROUBERTOUX 1986). Several observations were carried out on a given day and the groups were matched for order of observation. Each test was a dyadic encounter with an A/JOrl (A) male as the opponent. This strain had been chosen for its low scores of aggression and, in fact, the A males rarely initiated attack. The males tested in such a situation were discarded. The A opponent came from a group male cage. The behavioral records started when the tested male sniffedthe Aopponent for the first time, and the test lasted 6 min in cases where no attack occurred. Several variables were measured but only the frequency of attacking males will be considered here. Crosses: Only crosses informative for the demonstration of a co-segregation betweenthe YPARand initiation of attack behavior are reported here. First, congenic strains for YNPARwere developed in 1983 in by CARLIER. The H-YNPAR was substituted by the N-YNPAR the H strain to obtain its congenic H.N-YNPAR for this region of the Y chromosome and secondly by substituting the HYNPAR in the place of the N-YNPARin the N strain to obtain its congenic N.H-YNPAR. The congenic N.H-YNpAR was developed with N as recipient strain andH as donor, the N females being sired by NHF, males (in each cross the mother's genotype will be given first, separated by x from the father's genotype; within each genotype the first letter+) indicate the strain of the mother) and the backcross males subsequently sired again with N. The symmetrical design was used to obtain the second congenic H.N-YNPAR. The N.H-YNpAR and the H.N-YNPAR reached 29 and 30 backcross generations, respectively, and were observed between the 2nd and 21st repeated backcross. At every generation it is assumed that the congenic progeny lose 50% of the alleles from the congenic donor. Thus more than 99.99% of the allelic forms of the parental recipient strain, located throughout the genotype, including the YPAR, are expected in the congenic, at the 21st backcross. The isogenicity of the genetic background was directly confirmed by skin graft acceptance and by the similarity of mandible shape. First, tail skinreciprocal graftings were performed in females from H and H.N-YNPAR on one hand, and from N and N.H-YNPAR on the other, at the 9th backcross generation. One rejection out of 11 was observed in the first group and none in the second, 347 days afterthe graft. Second, 11 mandible measurements were performed on the same four groups. There is a high similarity for pairwise squared distances for a pair of congenic strains under thehypothesis of isogenicity, since BAILEY(1985) estimated the number of genes responsible for morphological changes on murine mandible to be over 100. No significant difference was found within the same sex between H vs. H.N-YNPAR and between N vs. N.HYNpAR (MOUTIERand CARLIER 1991). Second, the parental H and N females were sired with the parental and congenic males, thus providing two sets of recirocal Fls: HNF,,NHF, on one hand and HN.HYNp RF1,NH.N-YNPARFI on the other. Theselast two groups were constituted afterthe8th generation ofbackcross. Results on HNF, and NHF, have already been published ROUBERTOUX (CARLIER and ROUBERTOUX 1986; CARLIER, and PASTORET 1991). New groups have since been studied and because these gave very similar results, the data of all the males from the same F, were grouped. Third, two sets of backcrosses were developed with H and N parental males on reciprocal HNF, and NHF, mothers: HN X H, HN X N, NH X H, and NH X N. Fourth, reciprocal backcrosses on FI mothers whose ovaries had been removed and replaced in situ by ovaries from parental strains were obtained. As NHF, females are histocompatible with H andN tissues, their ovaries wereremoved and replaced by ovaries from N or H females according to a technique described previously (ROUBERTOUX and CARLIER 1988; ROUBERTOUX, NOSTEN-BERTRAND and CARLIER 1990; CARLIER,NOSTEN-BERTRAND and MICHARD1992). The females were labeled NoFl (for those providing N ova) or HoFl (for those providing H ova). Two controls were performed: first, to test for a transplantation effect per se, intrastrain ovary transplantations wereused: males born from NoNfemales (N femaleswhoseovaries had been removed and replaced in situ by ovaries from the N strain) were compared to those with identical genotype born from ungrafted N females, a symmetrical procedure being employed for HoH and H females. No effect due to grafting has been shown for initiation of attack behavior (ROUBERTOUX and CARLIER 1988). Second, controls for thegenotype of the pups were required because part of the ovary of the host female canremain after ovarectomy. Four independent genetic markers (coat color and threeelectrophoretic markers) were used to detect discrepancies between the expected and observed genotypes. Individuals were first observed in the behavioral test, theywere then typedemploying the four genetic polymorphic markers. Only those for which the observed genotype fitted the expected were considered for the subsequent statistical analysis. One subject was discarded for this reason in the NoFl X HN backcross (leading to an observed error rate of 1/65). Genetic polymorphism for the YPAR;measurement of liver steroid sulfatase(STS)activity: The hypothesis of cosegregation of the YpAR and initiation of offense behavior, can be tested using known differences between N-YPARand H-YPAR. The liver STS (EC 3.1.6.2) activity is controlled by the Sts locus (mapped on the YPAR; KEITCES et al. 1985) and was therefore used as a marker. The technique of measurement of the initial production rate of ['Hlestrone from ['HI estrone sulfate (expressed as picomoles of ['Hlestrone produced per min and permg protein; PROSTand ADFSSI 1983) was modified from BURSTEINand DORFMAN(1 963). Briefly, (1) the substrate concentration must not be rate-limiting since the catalysis is inhibited by an excessof substrate. Thus, to obtain the V,,, and K , from the Lineweaver-Burk reciprocal plot, we performed three substrate concentrations and a zero time to detect possible free steroid (each in duplicate). (2) Chromatographic purification of the radioactive substrate was needed since it can be partially hydrolyzed and/or radiolyzed. (3) The reaction was stopped with NanCOs (and not NaOH) to prevent ionization of free estrone, which was extracted using petroleum benzine. The liver STS activity was measured in males from the parental and their congenic strains, and alsoin NH.NyNPARFI and HN.H-YNpARFI males. R Intermale Aggression Statistics and experimental design: The differences between proportions of attacking males were tested here with a logit model analysis using the SAS CATMOD procedure. All the main effects and interactions are included in this model and tested according to ananalysis ofvariance design as indicated in the SAS/STAT Guidefor Personal Computers (1 987). The contribution of the YNPAR to initiation of attack behavior was tested by comparing males from each parental and its congenic strain. The behavioral observations were not performed at each generation of the congenics after backcross 4. The statistical analysis was performed by pooling the generations 2-6, 7-1 3 and 14-21 within each congenic. A group of each corresponding parental strain was observed each year allowing a comparison between the congenic andtheir parental strains. The proportions of attacking males were thus compared with a 3 X 2 X 2 desi n; three groups of generations and parental strains, two YN A R (H and N) and two backgrounds (H and N). The two sets ofreciprocal Fls from parental females sired with parental and congenic males constitute a segregating population for theseparation of the possible contribution of YN AR and YPAR. The HNFl and HN.H-YNPARFI share the same YpAR from N and have different non-pseudoautosomal re ions. Similarly, NHFl and NH.N-YNPARF1 share the same YpB from H but have different non-pseudoautosomal regions. A logit model analysisof the attack proportions following a 2 x 2 design was thus performed to test for the effect of each part of the Y (pseudoautosomal and nonpseudoautosomal). However, other candidate factors (the maternal environment, differential genomic imprinting, mitochondrial DNA and the X chromosome) also covary with yNPAR and YPAR,in these crosses, and thus might be responsible for an apparent co-segregation of one part of the Y with initiation of attack behavior in these reciprocal Fls. It was thus necessary to eliminate the possible effects of these factors. The contribution of differences in maternal environment can be eliminated if the co-segregation of one part of the Y and initiation of attack behavior stillpersistswhenthis environment is kept constant. Moreover, possible differential genomic imprinting (for example, the X from H inhibiting the expression of N-YNPARin the subjects having received H-YPAR) is not compatible with a co-segregation which is alwaysfound in the backcrosses, regardless of whether the genotypic contribution of the mother is H or N. The effect of these potential factors were tested employing reciprocal backcrosses on F1 mothers bearing in situ grafted ovaries from N and H. The mitochondrial DNA (mtDNA) is almost entirely maternally transmitted (GYLLENSTEIN et ul. 1991) and H and N mothers have mtDNA from different origins, domesticus and brwirostris, respective1 (YONEKAWA et al. 1982). Consequently, the HN.H-YNPA$,males that bear the N-YPAR have alsoinherited the mtDNA from H, whereas the NH.NYNPARFl, bearing the H-yPAR,have received the mtDNA from N. Backcrosses from the reciprocal F1 females provide the opportunity to test for an implication of mtDNA on initiation of attack behavior. Offspring of the first set of backcrosses with reciprocal F1 females HN X H and HN X N compared to those of the second NH X H and NH X N, have, in probability, the same number of autosomal or linked allelic forms from N and H, an identical Y chromosome and are exposed to identical maternal effects. They only differ by the origin, H us. N, of mtDNA and were used, here, to test for its implication in initiation ofattack behavior. us. the NHFl and NH.NThe HNFI and HN.H-YNPARFI YNPARF 1 also differ by the X chromosome from H us. N. The 9 x- 227 sources of variation between the backcrosses from F1 females with parental males and the backcrosses from FI females bearing N or H ovaries with HNF, and NHFl males are presented in Table 1. In each pair, 1 us. 1’ or 2 us. 2’, the two backcrossesshare identical autosomal information and maternal environment (uterine and postnatal). They differ by the origin of the yNPAR and a parental N or H or a recombinant X chromosome. A possible involvement of this chromosome in the initiation of attack behavior could be deduced for thewithinpair com arisons pending the demonstration of an absence of the YZAR effect, deduced from the congenic strains. The liver STS activity was compared by analysis of variance. RESULTS The parental and their congenic strains: The percentages of attacking males are presented in Figure 1 for the three groups of congenic and parental strains. The three groups of pooled backcrosses and their corresponding parental groups differ (x& = 8.85, P = 0.01). Inspection of Figure 1 shows that males tend to attack less in the groups of the more recent generdeations.However,theinteractionsbetweenthis crease andthe background on one hand and the YNPAR on the other are not significant (x& = 0.36, P = 0.83, and x& = 0.98, P = 0.61). The H and N background effect is significant (x(?)= 52,35, P < 0.0001): the percentage of attacking males is higher in N than in a n d ROUBH , as we have previously shown (CARLIER ERTOUX 1986). The YNPAR effect is not significant (x(?,= 0.14, P = 0.70) nor is the interaction between the H and the N background and the YNPAR (x(:, = 0.30, P = 0.58). The reciprocal Fls: The percentage of attacking of reciprocal Flsare presented in males in each group Figure 2. The effect ofthe non-pseudoautosomal part of the Y and the interaction between the two parts of t h e Y a r e n o significiant t (x$,= 0.000). However, the effect of the pseudoautosomal region plus maternal effects, including the mtDNA and theX contribution is significant (x(?) = 10.43, P = 0.001): the two groups having the pseudoautosomal part of the Y from the N strain (HNF1 and HN.H-YNPARFl) attack more frequently, suggesting that if the difference between the reciprocalFls is borne on the Y chromosome it is independant of the genetic information borne on the p P A R The backcrosses with reciprocal F1 females: The sample sizes and the percentages of males that initiated attack behavior in the offspring of the backcrosses were: H N X H ( n = 21; 38.09%), HN X N (n = 20; SO%), NH X H ( n = 24; 29.17%) and NH X N (n = 26; 69.23%). Only the origin of the father reached the level of significance (x(?)= 14.80; P < 0.0002). T h e males born from HNFI or NHFl mothers did not differ (x(?)= 1.06; P = 0.30). Consequently, as n o effect of the mother was detected, the present data suggest that attack behavior differences are independant of mtDNA origin in these strains. 228 P. L. R o u b e r t o u x et al. TABLE 1 Sources of variation between reciprocal backcrosses on F1mothers bearing H and N ovaries. NoFI:FIfemales bearing N ovaries; HoFI:FIfemales bearing H ovaries; NPAR, non-pseudoautosomalregion; PAR, pseudoautosomal region PAR Backcross generations Maternal environment mtDNA FI FI FI FI N N H H NOFIX N H ( 1 ) NHFl X N (1 ') HoFl X H N (2) HNF1 X H (2') EN.H-Y ;I x, 51) 128 Autosomes 314 N 314 N 114 N 114 N oo 1; 11, 02 2 to6 7 to 13 14to21 Generations of backcrosses FIGURE1 .-Percentage (+sE) of attacking males for thecongenic and parental strains at three stages of backcrosses. N , number of subjects [-n. ? z: 5: 38 N114 H 114 H H314 H 314 H H N N H FI H FI On Y XNPAR FI N FI H N FI FI 1z1 H.N-Y NPAR NPAR -li I O I n 51) c On X yNPAR z 151 FIGURE2.-Percentage (+sE) of attacking males for four groups of reciprocal Fls between parental strains H and N and between parental and congenic strains H.N-ppARand N.H-ppAR.Number of subjects are indicated under the x axis. Reciprocal backcrosses from HNFl, NHFI, NoFl and HoFl females: There was no effect of the YNPAR on the phenotype under study and thus the equations 1 , 1 ', 2, 2' could be used to test for the effect of the parental (H or N) us. X chromosome recombining H and N information. The comparisons for proportions of attacking males differ neitherbetween ( 1 ) offspring of NoFl X NH (n = 15; 93.33%)and ( 1 ') offspring of NHFl X N (n = 26; 69.23%)(x:(1,2 = 1.97; P = 0. lo), nor between (2) offspring of HoFl X HN (n = 22; 13 I FIGURES."Percentage (+SE) of attacking males in offspring of reciprocal backcrosses usingFI mothers bearing H or N ovaries and FI males. Number of subjects are indicated under the x axis (from CARLIER, ROUBERTOUX and PASTORET199 1). 36.36%) and (2') offspring of HNFl X H (n = 21; 38.10%)(x(?,= 0.14; P = 0.9 1). Reciprocal backcrosses on F1 mothers bearing H or N ovaries: The four groups were: NoF, X NH (n = 15), HoFl X NH (n = 14), NoFl X HN (n = 13) and HoFl X HN (n = 22). Here the maternal environment effects are kept constant. Underthese conditions (Figure 3), there is an effect of the genotype of the mother H us. N (x(?)= 5.03; P < 0.03): i.e., the male progeny attack lesswhen born from H rather than from N females. An effect of the genotype of the father is also shown (x:l) = 5.94; P < 0.02). The partial comparisons demonstratethatthe male offspring from NHF, fathersthat have received the H-YNPAR and the N-YPARattack more frequently than themales from which have received the N-YNPARand theH-YPAR HNFl fathers, whatever the genotype or the maternal (CARLIER, environmentprovided by themother ROUBERTOUXand PASTORET199 1). The liver STS activity :The activity ofthis enzyme (Table 2) had a significantly lower value in the H and H.N-YNPAR which share the same H-YPAR compared to share the same N-YPAR, the N and N.H-YNPAR which the other sources of variation, including the interaction between the YNPARand the autosomes, were not significant. This result indicates that two different allelic forms of the Sts gene correspondingto different levels of enzymatic activity are borne on the N-YPAR Intermale Aggression 229 TABLE 2 Steroid sulfatase activity (pmol/min/mg protein) in parental and congenic strains and one of their reciprocal hybrids(n, number of subjects; see text for other abbreviations) Genotypes 9 n R f SE N 5 167.2 f 29.1 141.4 (4 N.H-PpAR * 14.5 77.8 (a) H.N-PpAR H 8 5 f 16.7 56.8 (b) 8 f 5.6 (b) H N .NHH- Y . NN- ~Y~N~~F~, ~ F , 6 32.0 f65.43 9.3 (4 f 5.1 (4 The strains with the same letter in the last row did not differ. The difference between (a) and (b) is significant: F(2,26) = 23.48 (P< 0.001), as is the difference between (c) and (d): t(9) = 2.97 (P< 0.05). and H-YPAR.Moreover, the significantly lower STS activity found in the NH.N-YNPARFI compared to the HN.H-YNPARF1, suggests an under-expression of Sts when carried on the XPAR inthese strains. The identical activities for the Y- and X-borne loci reported by JONES et al. (1 989) are most probably limited to the Sts silent allele employed by these authors. DISCUSSION The percentage of males initiating attack behavior is higher in strain N than in H for dyadic encounters with an A male as standard opponent as shown in Figure 1. We know that this difference has at least onegeneticcorrelate(CARLIER,ROUBERTOUX and PASTORET1991). The present results indicate that one locus is located on the YPAR. Co-segregation of initiation of attack behavior with YNPARwas first examined using the permutation of the yNPAR between strains N and H (Figure 1). Independent segregation between initiation of attack behavior and YNPARappears as soon as the first backcross and still persists at the2 1st. This clearly demonstrates that the locus is not on the YNPAR. On the contrary, the absence of dissociation between the initiation of attack behavior and the YPAR in different segregating generations, strongly suggests that the genetic correlates for attack behavior are located on the YPAR chromosomal region. The groups HN.H-YNPARF1, and HNFl (sharing the same YPAR from N and having different Y non-pseudoautosomal regions) have ahighattackbehavior, whereas NH.N-YNPARFl and NHFl (sharing the same Y non-pseudoautosomal yPAR fromHbutdifferent regions) have a low attack behavior (Figure 2). Other factors,presented in the experimentaldesign, also covary with YPARbut the results presented here show that they have no significant effect. (1) Initiation of attack behavior is independent of the origin of mtDNA, as shown here by the comparison of backcrosses with reciprocal F1 females. (2) For several reasons, the X chromosome cannot contribute to the difference in frequency of initiation of attack behavior. First, there is a poor likelihood of its involvement since the individuals which have received the X from H, the less attacking strain, attack more frequently than those which have received the X from N, the more attacking strain. Second, the association of NYPARwith a high incidence of attack behavior is always present in the reciprocal backcrosses on F1 mothers whether they bear H or N ovaries (and consequently an H or an N X chromosome). Third, the inheritance of an X chromosome either from H or N compared to an X recombining H and N information, did not produce any difference in the incidence of attack behavior. (3) The covariation of the frequency of attack behavior with YPAR still persists when the maternal environment is kept constant, in the reciprocal backcrosses on F1 mothers bearing H or N ovaries: in the experiment illustrated in Figure 3 only the pseudoautosomalregion was correlated with the attack behavior in the expected direction. (4) Possible differential genomic imprinting (X from H inhibiting the expression of N-YNPAR)in the subjects having received H-YPAR, is not compatible with the frequency of attack behavior differences always found to be in the same direction in the backcrosses, regardless of whether the genotypic contribution of the mother is H or N. The agreement of the data presented herewith the hypothesis of a genetic correlate of attack behavior located on the pseudoautosomal part of the Y chromosome is reinforced by the demonstration of an enzymatic polymorphism for H-YPARand N-YPAR. However, high attack behavior is associated with the N genetic information of the pseudoautosomal region only when it is borne on the Y chromosome and not on the X. On the other hand, the low frequency of attack behavior is associated with the genetic information from H, only when it is borne on the YPARand not on the XPAR.The description of a polymorphism for the Sts gene, previously mapped on this region, strengthens this hypothesis. We have observed (Table 2) thattheSTS enzymatic activity is higher when measured in individuals bearing the N-YNPAR and the H-XNPAR compared to those bearing the H-YNPAR and the N-XNPAR, suggesting that the contribution to the phenotype could vary, according to theregion (on the X or on the Y) on which it is borne. However, the differences in attack behavior do not seem to be the consequence of a pleiotropic effect of allelic substitutions at the Sts locus, as suggested by the positive correlation between attack behavior and STS enzymatic activity, in N and H strains. This correlation 230 Roubertoux P. L. becomes negative with other strains: the C57BL/10 males that attack less than DBA/1 (SELMANOFF, MAXSON and GINSBURG 1976) have a significantly higher STS enzymatic activity (67.9 & 15.2 vs. 41.9 & 6.7; unpublished). The YPARalone cannot be considered responsible for the difference in this behavior since the offspring of the backcrosses from a mother belonging to the more attacking strain N attack more frequently than those of the backcrosses from an H mother (less attacking strain), when the strain of the father and the maternalenvironment are held constant (NoFI X HNFl > HoFl X HNFl and NoFl X NHFl > HoFl X NHF]); see Figure 3. Thus autosomal homozygotic alleles clearly also contribute to the difference either in an additive or interactive manner with the YPAR. We thank W. E. CRUSIO,B. DRESP, P.MANDELand B. MARTIN for useful advice and J. HIRSCHand S. C. MAXSON for discussion and suggestions during thepreparation of the manuscript. We also thank our anonymous referees for their helpful comments. This workwas supported by CNRS (URA 1294), DRED (Universitk Paris V Renk Descartes), Fondation pour la Recherche Mkdicale, Naturalia et Biologia, NATO Grant 0235/89, and Universiti de Reims Champagne Ardenne. L I T E R A T U R E CITED BAILEY, D. W., 1985 Genes that effect the shape of the murine mandible. Congenic strain analysis.J. Hered. 7 6 107-1 14. BURSTEIN S., and R. I. DORFMAN, 1963 Determination ofmammalian steroid sulfatase with7-aSH 36-hydroxyandrost-5-en17-one sulfate. J. Biol. Chem. 238 1656-1660. CARLIER, M., M. NOSTEN-BERTRAND and C. MICHARD, 1992 Separating genetic effects from maternal environmental effects, pp. 1 11-1 26 in Techniquesf o r the Genetic Analysisof the Brain and Behavior: Focus on the Mouse, edited by D. GOLDOWITZ,D. WAHLSTEN and R. WIMER.Elsevier, Amsterdam. CARLIER, M., AND P. L. ROUBERTOUX, 1986 Differences between CBA/H and NZB mice on intermaleaggression. I. Comparison between parental strains and reciprocal Fls, pp. 47-57in Genetic Approaches to BehavioralPhenotypes, edited by J. MEDIONI and G. VAYSSE. Privat, Toulouse. CARLIER, M., P. L. ROUBERTOUX and C. PASTORET, 1991 The Y chromosome effect on intermale aggression in mice depends on the maternal environment. Genetics 129 231-236. CARLIER,M., P. L. ROUBERTOUX, M. L. KOTTLER and H. DEGRELLE,1990 Y chromosome and aggression in strains of laboratory mice. Behav. Genet. 20: 137-156. ELLIS,E., and P. N. GOODFELLOW, 1989 The mammalian pseudoautosomal region. Trends Genet. 5: 406-4 10. GYLLENSTEN, U.,D. WHARTON, A. JOSEFSSON and A. C. WILSON, 1991 Paternal inheritance of mitochondrial DNAinmice. Nature 352: 255-257. HOOD,K. E., and R.B. CAIRNS,1988 A developmental-genetic et al. analysisof aggressivebehavior in mice.11. cross-sexinheritance. Behav. Genet. 1 8 605-619 JONES,J., J. P. PETERS,C.RASBERRY and B. M. CATTANACH, 1989 X-inactivation of the Sts locus in mouse: anomaly of the dosage compensation mechanism. Genet. Res. 53: 193-199. KEITGES, M. RIVEST,M. SINISCALCoandS. M. GARTLER, 1985 Xlinkage of steroid sulphatase in the mouse is evidence for a functional Y-linked allele. Nature 315: 226. MAXSON,S. C., 1992a Methodological issuesin genetic analyses of an agonistic behavior (offense) in male mice, pp. 349-374 in Techniquesfor the Genetic Analysis of the Brain and BehavMr: D. WAHLSTEN Focus on the Mouse, edited byD. GQLDOWITZ, and R. WIMER.Elsevier, Amsterdam. MAXSON, S. C., 1992b Potential genetic models of aggression and violences in males, pp. 174-188 in Genetically Dejned Animals Models of Neurobehavioral Dysfunctions, edited by P. DRISCOLL. Birkhauser, Boston. MAXSON, S. C., B. GINSBURG and A. TRATTNER, 1979 Interaction ofY-chromosomal gene@) in the development of intermale aggression in mice. Behav. Genet. 9 219-226. MOUTIER, R., and M. CARLIER,1991 Mandible shape analysisin Y-congenic strains of mice. Lab. Anim. 2 5 303-307. PROST,O., and G. L. ADESSI,1983 Estrone and dehydroepiandrosterone sulfatase activities in normal and pathological human endometrium biopsies. J. Clin. Endocrinol. Metab. 5 6 653-661. ROUBERTOUX, P. L., and M. CARLIER, 1988 Differences between CBA/H and NZB mice on intermale aggression. I1 Maternal effects. Behav. Genet. IS: 175-184. ROUBERTOUX, P. L., M. NOSTEN-BERTRAND and M. CARLIER, 1990 Additive and interactive effects between genotype and maternal environments, concepts and facts. Adv. Study Behav. 1 9 205-247. SAS/STAT, 1987 Guide for Personal Computers, Version 6 Edition, SAS Institute Inc., Cary, N.C. B. SELMANOFF,M. K., S. C. MAXSON and GINSBURG, 1976 Chromosomal determinants of intermale aggressive behavior in inbred mice. Behav. Genet. 6: 53-69. SIMMLER M. C., F. ROUYER, G. VERGNAUD, M.,NYSTROM-LAHTI, K. Y. NGO, A. DE LA CHAPELLEand J. WEISSENBACH, 1985 Pseudoautosomal DNA sequences in the pairing region ofthe human sex chromosomes. Nature 317: 692-697. STEWART, A. D., A. MANNINGand J. BATTY, 1980 Effects on y chromosome variants on themale behaviour of themouse MUS musculus. Genet. Res. 35: 261-268. VANOORTMERSSEN, G. A., 1984 A behavioral-genetic approach to variation in attack latency, serum testosterone level, and other agonistic parameters. Aggressive Behav. 1 0 177-1 78. VAN OORTMERSSEN, G. A,, R. F. BENUSand F. SLUYTER, 1992 Studies on wild house mice. IV. On the heredity of testosterone and readiness to attack. Aggressivee Behav. 1s: 143-148. YONEKAWA, H., K. MORIWAKI,0. COTOH,N. MIYASHITA,s. MIGITA, F. BONHOMME, J. P. HJORTH,M. L. PETRASand y. TAGASHIRA,1982 Origins of laboratory mice deduced from restriction patterns of mitochondrial DNA. Differentiation 22: 222-226. Communicating editor: R. E. GANSCHOW
© Copyright 2026 Paperzz