A Locus Determining P-Galactosidase Activity in the Mouse*

SUMMARY A group of inbred mouse strains, typified by C3H/HeJ, have twice as much /3-galactosidase activity in their tissues as mice of a second group of strains represented by DBA/Z J. In all organs tested, the higher activity in C3H/HeJ is expressed throughout development. The difference in activity is determined by a locus designated P-galactosidase (Bgs) on chromosome 9. The alleles present in C3H/HeJ (Bgsh) and in DBA/LiHa (Bgsd) show additive inheritance. Studies of pH optima, heat lability, intracellular location, and molecular weight of the fl-galactosidase activity in crude homogenates suggest that only a single enzyme component is present. Partially purified enzyme preparations from C3H/HeJ and DBA/2J were indistinguishable with respect to heat lability, K, for fi-nitrophenylgalactoside, and molecular weight. The presence of a single enzyme component, the absence of detectable structural differences, and the additive inheritance of Bgs alleles suggest that this locus may be either a structural gene duplication or a regulatory site con-trolling the transcription or translation of a closely linked fl-galactosidase structural gene.

A group of inbred mouse strains, typified by C3H/HeJ, have twice as much /3-galactosidase activity in their tissues as mice of a second group of strains represented by DBA/Z J. In all organs tested, the higher activity in C3H/HeJ is expressed throughout development.
The difference in activity is determined by a locus designated P-galactosidase (Bgs) on chromosome 9. The alleles present in C3H/HeJ (Bgsh) and in DBA/LiHa (Bgsd) show additive inheritance. Studies of pH optima, heat lability, intracellular location, and molecular weight of the fl-galactosidase activity in crude homogenates suggest that only a single enzyme component is present.
Partially purified enzyme preparations from C3H/HeJ and DBA/2J were indistinguishable with respect to heat lability, K, for fi-nitrophenylgalactoside, and molecular weight.
The presence of a single enzyme component, the absence of detectable structural differences, and the additive inheritance of Bgs alleles suggest that this locus may be either a structural gene duplication or a regulatory site controlling the transcription or translation of a closely linked fl-galactosidase structural gene.
The biochemical genetics of P-glucuronidase has been extensively studied in the mouse, and genetic variants affecting the structure of the enzyme (l),' its intracellular location (2), its control by steroid hormones (3), and its developmental pattern (4) have been described.
However, we do not know to what estent the genetic controls affecting fi-glucuronidase are unique to this enzyme and to what extent they apply to other acid hydrolases.
To approach this question, we have begun to define the genetic control of fl-galactosidasc activity. These studies have already shown that the activities of P-galactosidase and fl-glucuronidase are coordinately determined during developmcnt (5 of human congenital defects and we hope that information gained from animal studies of this enzyme will be of assistance in understanding the human disorders.
In this report, we show that the common inbred strains of mice are readily divisible into two groups on the basis of their brain P-galactosidase levels. Mice of strain C3H, chosen to represent one of these groups, have twice the activity of C57BL/Ks and DBA/2 mice, chosen to represent the other group, in all organs tested at all stages of development. This difference is determining by a single locus on chromosome 9, with two alleles exhibiting additive inheritance.
In a later report we shall describe a second variant in which only the liver activity is altered. Enzyme Assay In our standard assay, reaction mixtures contained: 0.9 ml of 0.1 M citrate buffer, pH 3.5; 0.05 ml of 0.1 M p-nitrophenyl-fl-ngalactoside (Pierce); and 0.05 ml of tissue homogenates.
Reactions were started by addition of substrate, incubated for 30 to GO min at 37", and stopped by cooling in an ice bath followed by rapid addition of 0.2 ml of 30% trichloroacetic acid. After centrifugation for 10 min to remove precipitated protein, the supernatant solutions were transferred to tubes containing 0.2 ml of 5.0 M 2-amino%methyl-1:3-propanediol (Aldrich). The absorbance of released p-nitrophenol was measured in a Zeiss spectrophotometer at 415 nm (E,, = 14,000). The reaction rate was proportional to enzyme concentration and to time of incubation for more than 3 hours. Enzyme activity is expressed as micromoles of p-nitrophenol produced per hour.
Automated Assay of fl-Galactosidase For analysis of the large numbers of animals generated in genetic crosses, an auto-analyzer was designed which measures hydrolysis of the fluorogenic substrate 4-methylumbelliferyl-&n- soluble and particulate activity were assayed in liver. To detect regulatory and developmental variants, the quantitative levels of enzyme activity in a set of adult tissues were compared. No mutants in structure or subcellular location were detected among the set of 45 strains.
However, we did find considerable variation in total enzyme activity among adult organs.
The strains could readily be divided into two groups on the basis of their brain enzyme activity.
The members of one group show brain enzyme activities of 6 to 9 pmoles per hour per g and those of the other group 13 to 16 pmolcs per hour per g. Brain activity was chosen as the distinguishing characteristic because there is very little individuai variation in this activity, and because the distinction between the two groups is unambiguous (Fig. 1). The values of brain fl-galactosidase for these inbred strains is presented in Table I. We have chosen one strain from the high activity group, C3II/HeJ, and two strains from the low activity group, DISA/2J and C57BL/Ks, for further comparison. The value of P-galactosidase specific activity in adult tissues of C3H/HeJ mice is twice as high as the corresponding DEA/2J tissues (Table II) I   TABLE   II Variation in specific activity of brain p-galactosidase among inbred strains of mice Enzyme activity and protein concentration were determined on homogenates prepared from pooled tissues of three adult males from each strain.-@-Galactosidase activity in CSHjHeJ and DBA/$J mice -Organs from 12 mice of each strain were individually homogenized and assayed. Activity is expressed as the mean micromoles per hour per g f 1 S.E. Mixtures of samples from the 2 strains have the predicted additive activity, indicating that low molecular weight effecters are not responsible for the activity difference. DeveZopment-The higher activity in C3H/HeJ relative to that in DBA/2J and C57BL/Ks does not appear to arise from a difference in the developmental pattern of the enzyme in these strains.
Enzyme activities were measured in brain, liver, and heart of mice of each strain between gestational Day 14 and 60 days of age ( Fig. 2A). The higher activity of C3H was maintained throughout this period, the activity in C3H being almost exactly twice that of the other strains in each tissue at all stages of development (Fig. 2B).
Enzyme Properties-The characteristics of the fi-galactosidase activity present both in crude homogenates and in partially purified preparations from C3H/HeJ and DBA/SJ livers were compared to determine whether the difference in activity re-3269 Brain. . . . . 7 2. &Galactosidase activity in developing mouse organs Mice were obtained from the Production Deoartment of the Jackson Laboratory and tissues were prepared~as described (5) Birth was on Day 0. Liver, brain, and heart were studied in on high activity strain (C3H/HeJ (O-0)) and in two low activit: strains (DBA/BJ (O-O) and C57BL/Ks (O---O)). A activities per g of tissue. B, relative activities.
The data arl expressed relative to the average value at each time point calcu lated from the DBA/B, C57BL/Ks, and one-half of the C3I activities.
fleeted any difference in the properties of the enzyme. Severa lines of evidence suggest that a single enzyme component i responsible for the P-galactosidase activity present in the tissue we have studied.
About 85% of the total &galactosidase activity of liver wa estimated to be in lysosomes from measurement of the relativ amounts of soluble and particulate activity and the observation Residual enzyme activity was then determined under standard assay conditions. Rates of inactivation were calculated with the assumption that denaturation followed first order kinetics; this was confirmed in control experiments. that nearly all of the particulate activity was released when the particulate fraction was exposed to hypotonic buffer (9). There was no difference between strains in this regard.
The enzyme from the two strains had the same elution volume after gel filtration (see "Materials and Methods"). In every case the kinetics of heat denaturation of the activity present in crude homogenates was monophasic.
There was no indication of more than one enzyme component.
The sensitivity to thermal denaturation was measured over the range 37-49" for partially purified enzyme (Fig. 3). From these data, a value of 39,000 cal per mole for the energy of activation of the denaturation reaction was calculated for both strains.
The kinetics of denaturation was tested at 43" and was first order with a rate constant of -0.07 min+ for both strains.
We have had considerable difficulty in carrying out acrylamide gel electrophoresis.
Single bands with similar mobilities in all strains were observed at pH 8.1; however, the staining reaction requires high protein concentrations and the resolution was poor.
The enzyme present in both C3H/HeJ and DBA/SJ liver has optimal activity at pH 3.5, and there was no difference in the pH activity curves of the two strains.
Enzyme from both strains is completely inhibited by lop5 M p-chloromercuribenzoate. The affinity of the enzyme for the substrate p-nitrophenyl-fl-n-galactoside was determined (Fig. 4). A value of 5.2 x 10m4 M was obtained for the K, of enzyme preparation from both strains.
Taken collectively, these results suggest that a single enzyme component is present in both strains, and that the difference in &galactosidase activity between C3H and DBA/S does not result from a change in either the intracellular location or structure of this enzyme.
Genetics-The difference in activity between C3H and DBA mice appears to be controlled by a single locus with two alleles showing additive inheritance. The brain P-galactosidase ac- livers for the substrate p-nitrophenyl-&ngalactosidase.
The apparent K, for enzyme from both strains is 5.2 X 10-* M. The data are plotted using the Eadie-Hofstee (10) linear transform of the Michaelis-Menten equation. Eighty offspring of the reciprocal backcrosses between (C3H/ HeHa X DBA/LiHa)FI mice and the DBA parent were scored for coat color and for brain fi-galactosidase.
The segregation of enzyme levels into low and intermediate classes is evident in Fig.  5. The substrate used was 4-methylumbelliferyl-P-n-galactoside. The activity against this substrate is lower than with p-nitrophenyl-fl-n-galactoside.  (Fig. 5). Similarly, progeny of the backcross to the DBA parent segregated into the intermediate and low classes in approximately equal numbers (Fig. 5). The Fz generation segregated into the expected three classes. The locus has been designated Bgs (fl-galactosidase) ; the allele present in C3H/HeJ is Bgsh and that present in DBA/ 25 is Bgsd.
Linkage tests to known loci (Table III) 5. Inheritance of p-galactosidase activity. Brain enzyme activity was determined for the parental strains C3H/HeHa and DBA/LiHa, and for the I?,, Fz, and both backcross generations. In all cases, reciprocal crosses were made, with no significant difference in the results. p-Galactosidase was determined by automated assay with the substrate 4.methyluInbelliferyl-/3-ogalactoside.
All animals were between 9 and 12 weeks of age. Each filled circle represents one animal 9 marker dilute. This agrees with the observations of Hakansson and Lundin.
Levels of P-galactosidase in liver and brain appear to be determined by the same locus. Animals segregating into the three brain activity classes, low, intermediate, and high, simultaneously segregate into the corresponding classes of liver activity (Fig. 6). DISCUSSION Inbred mouse strains show significant differences in levels of fi-galactosidasc; all of the strains examined could be classified as high fl-galactosidase (specific activities between 13 and 16 ~rnoles per hour per g wet. weight of brain tissue) or low P-galactosidasc strains (6 to 9 pmoles per hour per g). In a representative high strain (CSH/HeJ), the P-galactosidase activity in all tissues at all stages of development is twice that of the low strains DlU/BJ and C57BL/Ils. In genetic crosses between these strains, this quantitative difference segregated as a single Mendelian factor. Linkage tests demonstrated that the responsible locus, designated Bgs, is located on chromosome 9 in the mouse, approximately 21 ccntimorgans from the dilute locus. The enzyme level in heterozygotes was equal to the mean of the parental activities, i.e. the locus exhibits additive inheritance. Two alternative models for the effect of the Bgs locus may be considered.
First, the Bgsh allele (in C3H mice) may encode a structurally altered enzyme whose catalytic activity per molecule is twice as great as the enzyme encoded by the Bgs" allele (in DBA/2 mice). Alternatively, the effect of the Bgsh allele may be to increase the concentration of enzyme molecules in the tissues of C3H mice. The first model depends upon a difference in primary structure between the fl-galactosidases present in C3H and DBA/2 mice, while the second model predicts no difference in structure.
To test these predictions we compared the heat lability, electrophoretic mobility, and kinetic properties of fl-galactosidase from the two strains; no differences in these parameters were observed. The probability that a structural change would be reflected in one of these parameters can be estimated.
Empirically, 70% (39/55) of amino acid substitutions were found to alter the heat lability of bacterial /% galactosidase (II), and 67 y0 (14/21) of variants in activity of glucose g-phosphate dehydrogenase exhibited altered K, for substrate (12). Approximately 30 y0 of base substitutions theoretically should produce a protein with altered net charge; however, we are not certain that a single charge difference would have been detected in our electrophoretic system. We feel that the absence of detectable differences in these three properties make it unlikely that the @-galactosidases of C3H/HeJ and DBA/2J mice differ in their primary structures and hence in their catalytic activity per molecule.
However, definitive evidence on this point will require the purification of the enzyme and more detailed structural comparisons.
It appears more likely that the Bgs locus controls the number of enzyme molecules present in mouse tissues. Two models which account for the failure to detect any structural change and for the additive inheritance of the Bgs alleles are that Bgsh represents a duplication of the P-galactosidase structural gene, or that it represents an alternate form of a regulatory site controlling the efficiency of transcription or translation of a closely linked fi-galactosidase structural gene.
In anticipation of further studies of the Bgs locus we wish to 4 E. Hakansson and L. Lundin, personal communication, 1973. point out that all presently reasonable models for its action in-dicate that Bgs is either identical with or closely linked to the structural gene for @-galactosidase. In considering possible models of genetic regulation in mammals, it is of considerable interest that while fl-galactosidase and fl-glucuronidase are both present in lysosomes, and share a coordinated pattern of development (5), the genetic factors controlling their expression are not associated physically on the same chromosome; the Bgs locus is 011 chromosome 9, and the P-glucuronidase structural gene with its attendent regulatory sites is 011 chromosome 5 (l-4, 13).' Acknow2edgments-We are grateful to Dr. Douglas Coleman of the Jackson Laboratory, Bar Harbor, Maine, for sponsoring our studies at the Jackson Laboratory during the summer months of 1970 to 1972. The st,rain survey and developmental studies were great,ly facilitated by the cooperation of the Laboratory staff.