Genetic Regulation of Coumarin Hydroxylase Activity in Mice EVIDENCE FOR SINGLE LOCUS CONTROL ON CHROMOSOME

phenobarbital-induced activity of coumarin is in 16 as is a of coumarin hydroxylase in hepatic strains of activity, while the

Linkage of the Locus Controlling Coumarin Hydroxylase Activity-Examination of the offspring of the backcross AKD2F1/J to AKR/J revealed, as expected, the equal numbers of mice with intermediate (F,) and low (AKR) hydroxylase activity and with pigmented (+/c) and albino (c/c) coats (Table III). However, only 11 of the 52 mice, or 21%, were recombinants with respect to these two traits, i.e. having low hydroxylase activity and a pigmented coat (six mice) or intermediate activity and an albino coat (five mice). This nonrandom assortment of the two traits indicates that the locus controlling coumarin hydroxylase activity is on the same chromosome as the albino locus (c), with an estimated map distance of 21 + 6 centimorgans (CM). The albino locus is located on chromosome 7 (Linkage Group I) of the mouse (14). Coh is proposed as the locus symbol for coumarin hydroxylase, with superscripts h and 1 designating the high and low activity alleles, respectively.
Inheritance of Coumarin Hydroxylase Activity in DBA/2J and C57BL/6J Mice- Table  IV summarizes coumarin hydroxylase activity in whole liver homogenates of phenobarbital-pretreated DBA/BJ, C57BL/6J, and B6D2F1/J mice, as well as 24 RI strains derived by inbreeding the Fz generation of DBA/W X C57BL/6J crosses. As previously reported (4), hydroxylase activity in the B6D2F1/J hybrid is the arithmetic mean of the parental activities. The 24 recombinant inbred strains can be divided into two principal classes with respect to coumarin hydroxylase. Four strains (BXD-2,6, 12, and 14) have high activity (mean of 264 units) resembling the DBA/ 25 progenitor, while the remaining 20 strains have low activity (mean of 57 units) resembling the C57BL/6J progenitor. These results are also consistent with single gene inheritance, although the preponderance of low activity strains is unusual. The genetic loci which control the electrophoretic variations of glucose phosphate isomerase (Gpi-1) and the hemoglobin p chain (Hbb) are separated by a distance of 33 CM and are located on the centromeric and noncentromeric side, respectively, of the albino locus on chromosome 7 (10). The alleles at both loci differ in DBA/BJ and C57BL/6J mice, and the 24 RI strains have been typed for the parental allele which they   contain. The Hbb typing of 20 BXD RI strains has been reported previously (15). A comparison of the distribution of the high and low activity alleles for coumarin hydroxylase with the alleles for Gpi-1 and Hbb shows that recombination between Coh and Gpi-1 occurs in only 4 of the 24 strains (BXD-15, 27, 28, and 30), while recombination between Coh and Hbb exists in 10 of 24 strains (BXD-2, 5, 6,9,13,18,24,25,27,30). Thus, the Coh locus is evidently considerably closer to the Gpi-1 locus than to the Hbb locus. The data are compatible with either the gene order Coh-Gpi-I-Hbb or Gpi-I-Coh-Hbb.
Also shown in Table IV is the lack of linkage between any of the three chromosome 7 loci and the locus controlling the induction of aryl hydrocarbon hydroxylase (Ah), another cytochrome P-450-dependent microsomal mixed function oxidase. Lack of linkage between the Ah and Hbb loci had been previously reported (11).
Inheritance of Coumarin Hydroxylase Activity in SWR/J and C57L/J Mice-Examination of coumarin hydroxylase activity in whole liver homogenates of C57L/J (high activity) and SWR/J (low activity) mice, their FI hybrid, and nine SWXL RI strains indicates additive inheritance in the F, hybrid, five high activity RI strains (SWXL-4, 6, 7, 14, and 17), and four low activity RI strains (SWXL-2, 12, 15, and 16) ( Table V). SWR/J and C57L/J mice are allelic at both the Gpi-1 and the c locus (9,16). The complete correspondence of parental phenotypes between Coh and Gpi-1 loci in the RI strains provides further evidence of close linkage between these two loci, but does not help in assigning Coh to the The Genetics of Coumarin Hydroxylation in Mice centromeric or noncentromeric side of Gpi-1.  Gpi-1 (16, 17). Two crosses were carried out to establish the location of Coh relative to Gpi-1 and p. The fist cross consisted of successive backcrosses of C57L/J to the C57BL/6J-p"" stock, which carries the unstable allele of p that arose in C57BL/6J (18). The cross can be represented as (C57L/J x C57Bl/GJ-pun/pun), X C57BL/6J-p""/p"", where n represents the number of crosses of the heterozygous female parent to the C57BL/6J-p"" stock. In each generation, one or more females bearing the normal allele of p and the Gpi-1" allele of C57L/J were selected to continue the incipient congenie line. Siblings were tested for their Gpi-1 allele and coumarin hydroxylase activity in order to accumulate threepoint linkage data and to assure that the Cob* allele from C57L/J had not been lost by crossing over in the previous generation. A total of 69 progeny were tested and the results are summarized in Table VI. There were 15 recombinants between Gpi-1 and p, and 4 recombinants between Coh and Gpi-1. No double crossovers were observed and since the p locus is distal to the Gpi-1 locus the data indicate that the correct gene order is Coh-Gpi-l-p.
In the second cross, the progeny of the mating of (C57BL/6J x 129/J)F1 x C57BL/ 6Jyp""/p"" were analyzed (Table VII). Among the 18 progeny tested, two crossovers were observed between Gpi-1 andp and one was observed between Coh and Gpi-1. No double cross-Progeny of (C57BL/&l x 129/J) FI x C57BL/6J -pun/p"" classified for alleles at Coh, Gpi- overs were observed and again the gene order Coh-Gpi-l-p is indicated. Utilizing the data from both experiments, the estimate of the recombination frequency between Coh and Gpi-1 is 5.8 & 2.5%. The estimated recombination frequency between Gpi-1 andp is 19.5 f 4.3% which is in agreement with previous estimates (16, 17).

Metabolism
of coumarin to 7-hydroxycoumarin is catalyzed by coumarin hydroxylase, a cytochrome P-450-dependent mixed function oxidase found principally in the endoplasmic reticulum of liver. Previous in viuo and in vitro studies (4,5,19) have shown that coumarin hydroxylase activities are markedly strain-dependent in the mouse. The results of the present study indicate that a single gene locus on mouse chromosome 7 is primarily responsible for the expression of coumarin hydroxylase activity.   Table IV. h "S" and "L" are used as generic symbols for alleles inherited from the SWR/J and C57L/J strains. resnectivelv. The SWR/J eenotvne is Cob' Gpi-1" c Ahd/Coh' Gpi-1" c Ah". The C57L/J genotype is Cob Gpi-1" + Ahh/Cohh Gpi-1" + Ahb. Regions where crossovers have resulted in recombination of the parental alleles in the different SWXL lines are denoted by a X.
' SWXL-2 and SWXL-6 are now extinct. A survey of 16 inbred strains of mice (Table I) indicated that low coumarin hydroxylase activity is more prevalent than high coumarin hydroxylase activity by a ratio of about 2:l. The five strains having high coumarin hydroxylase activity lack a common origin and strains such as C57L/J and C57BL/ 6J, which had a common origin (16), have markedly different levels of coumarin hydroxylase activity. Despite the apparent lack of common ancestry, the coumarin hydroxylase from all of the strains with high activity could be preferentially inhibited by aniline and the coumarin hydroxylase from strains with low activity were preferentially inhibited by metyrapone (Table II). These data suggest a common structural feature within, but not between, the two classes of coumarin hydroxylase. Fig. 2 shows a partial linkage map of mouse chromosome 7 (16, 20). Inheritance and linkage data using a total of five loci from six strains of mice have been presented ( Fig. 1 and Tables III to VII) and all the data are consistent with the conclusion that a single locus, designated Coh, on chromosome 7 is the principal determinant of coumarin hydroxylase activity. The recombination frequency of 21 f 6% between the Coh and c loci in the backcross of AKR/J with AKDBFl/J mice and the coincidence of parental phenotypes at the Coh and Gpi-1 loci in the BXD and SWXL lines indicate the close proximity of the Coh and Gpi-1 loci. The data obtained with the three-point crosses between p, Gpi-1, and Coh loci clearly indicate the gene order as Coh-Gpi-l-p rather than Gpi-l-Coh-p since the former sequence of genes requires only a single chromosome break and crossover between gene pairs. The frequency of crossing over between Gpi-1 and Coh (5 of 87 mice, or 5.8%) in the three-point crosses agrees well with the frequency estimated from the data with the recombinant inbred mice, 3.7 f 2.1% (see Ref. 21 for the basis for this estimate and its standard error). In the BXD recombinant inbred strains (Table IV), there is an excess of lines fiied for the Cob" allele. Were it not for the fact that close linkage is observed between Coh and Gpi-1, one might question whether a single locus model adequately accounts for the data. However, since Gpi-1 is certainly controlled by a single locus (9), there is no inconsistency in ascribing single locus control to coumarin hydroxylase activity. The predominance of the Coh' and Gpi-lb alleles in the BXD strains may be attributable to a closely linked gene in the DBA/PJ strain that is detrimental to survival or reproduction (or both).
In addition to the data just described, results of studies with two additional lines of mice support our assignment of the Coh locus proximal to the Gpi-1 locus on chromosome 7. As shown in Tables I and II, the 129/J strain has high coumarin hydroxylase activity, while the C57BL/6J strain has low activity. The PRO/Re mouse strain was derived from the cross of C57BL/6J with 129/J (22) with selection for the 129/J alleles at the albino (c) and pinkeye dilution (p) loci. The strain carries predominantly C57BL/6J alleles at loci unlinked top and c, but is homozygous for the 129/J alleles at both the Hbb and Gpi-1 loci. Presumably, the PRO/Re strain inherited most, if not all, of the segment of chromosome 7 from somewhere proximal to Gpi-1 to somewhere distal to Hbb from its 129/J progenitor.
When the PRO/Re strain was tested for coumarin hydroxylase activity it was found to have low activity like its C57BL/6J progenitor strain. This finding suggests that Coh is proximal to Gpi-1 since assignment of the locus to the distal side of Gpi-1 would require a double crossover. The BlO.l29(21M) congenic strain was developed through insertion of the histocompatability locus 4 allele (H-4") and the closely linked pinkeye dilution mutation (p) from the 129/J strain into the C57BL/lOSn genetic background by successive crossing and intercrossing (23). The linked segment of chromosome 7 from 129/J is bounded by Gpi-1 and c (24). The BlO.l29(21M) was tested for coumarin hydroxylase activity and found to be Cob' like the inbred partner, again suggesting the gene order Coh-Gpi-l-p.
A congenic strain has been started in which the Cohh allele of the C57L/J strain is being transferred to the genetic backaround of the C57BL/6J strain. When sufficient backcross with which to extend those studies and better define the regulation of the cytochrome P-450-dependent mixed function oxidase systems.