Archlvee of
Arch. Toxicol. 38, 45-51 (1977)
TOXICOLOGY 9 by Springer-Verlag1977
A Biochemical Specific Locus Mutation System in Mice* H. V. Mailing and L. R. Valcovic Environmental Mutagenesis Branch, National Institute of Environmental Health Sciences, P. O. Box 12233, Research Triangle Park, North Carolina 27709, USA
Abstract. Two mouse strains DBA/2J and C57BL/6J are heteromorphic with respect to the electrophoretic mobility of at least 8 enzymes and the/5-hemoglobin chain. The genotype of DBA/2J for these markers is: Es-1, Es-3, Gpd-1, Gpi-1, Id-1, Mod-1, Pmg-1, Dip-1 and Hbb; and of C57B1/6J: Es-1, Es-3, Gpd1, Gpi-1, Id-1, Mod-1, Pgm-1, Dip-1 and Hbb. Electrophoresis on tissues of interstrain hybrids will show the two parental bands and additional hybrid bands if the enzyme is a polymeric structure. In specific locus mutations which result in loss of activity (deletions, nonsense mutations), the hybrid resembles the nonmutated parent. If the mutation results in a change in electrophoretic mobility, some of the bands on the gel will either run faster or slower compared to the hybrid bands in a normal F 1 animal. Fifty DBA/2J males were irradiated with ~,-rays from a Co 6~ source with two doses of 500 R at 24-h intervals a t a dose rate of 95 R/min. After this irradiation the males were sterile for approximately 3 months. After the males regained their fertility, they were continuously mated to C57BL/6 females. Thus far somewhat over 2600 animals have been tested and four new mutations detected, giving a frequency of approximately 1.7 x 10-4 mutations per locus per generation. The four mutations are: two independent mutations at Hbb, one at Mod-1 and one at Id-1.
Key words: Biochemical specific locus mutation system C57BL/6 mice -- X-ray -- Electrophoresis.
DBA/2J and
Zusammenfassung. Die beiden M/iusest/imme DBA/2J und C57BL/6J sind heteromorph hinsichtlich der elektrophoretischen Mobilit/it von mindestens 8 Enzymen und der/3-H/imoglobinkette. Der Genotyp fiir DBA/2J-M/iuse f'dr diese Marker ist: Es-1, Es-3, GpD-1, Gpi-1, Id-1, Mod-1, Pgm-1, Dip-1 und Hbb; der f/Jr C57BL/6J-M/iuse ist: Es-1, Es-3, Gpd-1, Gpi-1, Id-1, Mod-1, Pgm-1, Dip-1 und Hbb. * Presented at the 3rd Meeting of the Gesellschaftf'tir Umwelt-Mutationsforschunge. V., Neuherberg, July 1-2, 1976
46
H.V. Mailing and L. R. Valcovic Bei elektrophoretischen Untersuchungen des Gewebes von Hybriden der beiden St/imme findet man die beiden Banden der Eltern und eine zus/itzliche Hybridbande, wenn das Enzym von polymerer Struktur ist. Bei Mutationen vom ,,specific locus type", die aufgrund von Deletionen oder ,,nonsense mutations" einen Aktivit/itsverlust zur Folge haben, gleicht der Hybrid dem mutierten Elternteil. Wenn die Mutation eine ,~nderung der elektrophoretischen Mobilit/it zur Folge hat, werden einige der Banden auf dem Gel entweder schneller oder langsamer sein als die entsprechenden Hybridbanden bei normalen F1-Tieren. 50DBA/2J-M/innchen wurden im Abstand von 24 h mit 2 Dosen von 500 R F-Strahlen mittels Co60 bestrahlt und zwar mit einer Dosierung von 95 R/min. Nach dieser Bestrahlung waren die m/innlichen Tiere ffir ca. 3 Monate steril. Nach R/ickerlangung der Fertilit~it wurden die M/innchen kontinuierlich mit 50 C57BL/6-Weibchen gepaart. Bisher wurden etwas mehr als 2600 Tiere getestet, wobei 4 neue Mutationen festgestellt wurden. Dies entspricht einer Mutationsh/iufigkeit von ca. 1,7 x 10-4 per locus/Generation. Bei den 4 Mutationen handelt sich um zwei unabh/ingige Mutationen aus Hbb, eine aus Mod-1 und eine aus Id-1.
Some of the important present tasks within genetic toxicology are development, verification and implementation of (a) reliable test systems for prescreening of numerous compounds for mutagenic activity, (b) monitoring systems of the human population to detect exposure to mutagenic environment and (c) risk assessment systems closely related to the human situation. The latter type of system should be used (a) when a chemical possesses certain benefit factor to the society such that it may be undesirable to remove it from man's environment or (b) pollutants which production is unavoidable and the society has to set limits on the degree of exposure of the population. Laboratory model systems to provide such data must approximate, as closely as possible, the pharmacological and genetical situation as it exists in man. In mutagenesis over the past number of years, the morphological specific locus test has provided a data base from which exposure limits of ionizing radiation have been set for the human population. When the genetic risk of a certain compound to the human population has to be evaluated, it is important that a battery of test systems be used which covers most types of transmissible mutagenic events. This includes a whole array of chromosome aberrations such as transmissible translocations, addition or loss of chromosomes or parts thereof, and the various types of point mutations ranging from physical removal of the gene to a single base-pair change. Comparison of the mutagenic specificity of chemicals in Neurospora has clearly shown that there is no consistent correlation, if any at all, between chromosomal and point mutational events. In our laboratory, Dr. Valcovic and I have developed a system to measure point mutations (Valcovic and Malling, 1973), which we call the biochemical specific locus mutation system (BSL). In this system we utilize electrophoresis for the visualization of specific proteins and enzymes for the detection of the mutant phenotype. One of the main features of this procedure is that it is possible to detect mutations in which the enzyme is not functionally destroyed, but merely has an altered charge
47
A Biochemical Specific Locus Mutation System in Mice Fig. 1. Electrophoretic pattern of parents normal F1 and possible mutant types in the biochemical specific locus mutation system. P~ = parents with fast moving band; P2 = parent with slow moving band; F 1 = normal hybrid between P1 and P2; I A E 4 = inactive polypeptide of P~; IAE-2 = inactive polypeptide found of P1 but able to form complex with P2 polypeptide; E P M = electrophoretic mobility mutants
'
i
i
m
=
m
m
o
~m
t~ Pi
P2
FI
IAEI
IAE-2 F'PM
m
EPM
and hence a shift in the electrophoretic mobility. However, due to technological problems, it is difficult to detect the loss of enzyme activity in an FI; this requires seeing a band with half the intensity of the normal F r We have overcome this problem by using two inbred strains of mice in which there are allelic differences at the loci under test. For a given enzyme, one strain (P1) will show a band of fast mobility and the other strain (P2) a slow band (Fig. 1). If we assume that the enzyme is a dimer, then the interstrain hybrid will show three bands; a fast band and a slow band corresponding to each parental strain and a third band in the middle representing the hybrid molecule. Now let us mutagenize the strain (P1) with the fast band and again look at the interstrain hybrid. If the mutation resulted in no enzyme produced or an inactive form, then the F I shows only the slow band; it looks like the untreated parent (IAE1), or if the produced polypeptide can interact with the active polypeptide, two bands may be formed (IAE-2). On the other hand, if the mutation was a base-pair substitution that merely altered the charge of the protein, the hybrid will still show three bands but the fast and hybrid bands will be in an altered position (EPM). It is, therefore, very easy to differentiate two types of mutations, only one of which is detected in the morphological specific locus test. In this example we have considered only the types of mutations one sees if one parent is mutated. If the other parent were also treated, then the mutant types in F~ could look like the mirrow image of any one of the mutant types in Figure 1 and is easily differentiated from mutations induced in the other parent. So in the F~ it is possible to identify whether the gene from the male or female has mutated. The F~ can, therefore, be the progeny of a cross where both parents are treated.
Selection o f the Strains
Our first goal before starting this system then, once we had set these rules, was to examine the available data on all the inbred mouse strains to find the combination of two strains in which there was a maximum number of loci showing these allelic differences. Not surprisingly, the two strains which we discovered to be most widely different were the DBA/2 and C57BL/6 strains. These are known to be different not only by electrophoresis but by a great many other parameters. There are indeed nine different loci in which there are allelic differences. It is interesting to note on this
48
H.V. Mailing and L. R. Valcovic
that when we look at these enzymes that another feature of the system is apparent, all of these enzymes have homologous enzymes and, therefore, homologous genes in man. So that in our risk evaluation system, while recognizing that we do not have the capability of using man and the human metabolism, we can at least look at genes in our test species which, with the exceptions of probably only minor sequence variation, are similar genes to those which exist in the human population. Additionally, since we are studying mutations in enzymes, that is the immediate gene product, we have then the capability of doing biochemical studies to evaluate and identify the molecular event leading to the mutation; that is to be able to define whether the induced mutation is indeed a base-pair substitution, a frameshift mutation or, in fact, a short intragenic deletion. Now I would like to go into a description of our first experiment with this system which we started approximately two years ago. We felt that since we were developing a new specific locus system and were going to compare it to the morphological specific locus system in which there was a great deal of data, especially with ionizing radiation, that we should obtain a data point in our system to compare with the morphological specific locus system, and thus we selected radiation for our first mutagen. A Co 6~ gamma source was employed and the dose rate was 95 R/min. DBA/2J and C57BL/6J males were exposed to a total dose of 1000 R which was delivered as two 500 R doses with a 24-h interfraction period. These males were then mated to females of the opposite strain. Males from the two strains differed drastically in their radiosensitivity. The DBA males were sterile for approximately 3 months and of the total of 50 males that had been irradiated, all 50 returned to fertility after the three-month sterile period. However, in the C57BL male group, an equal number being radiated, only two males ever recovered any fertility and each of those gave not more than two litters. So, C57BL males are highly sensitive to radiation and the DBA males are much more resistant. The data on mutation induction will be primarily that from the DBA males. We have tested until now somewhat over 23,000 loci and have discovered four new mutations which give us an induced frequency of approximately 17 x 10 -5 mutants per locus per generation. Using the same irradiation protocol and dose, Russell found 49.9 x 10-5 mutants per locus per generation (Russel, 1963); under similar irradiation conditions but with a different set of markers in the specific locus test, Lyon got 13.5 x 10-5 mutants per locus per generation (Lyon and Morris, 1969). It is interesting that all four mutations which we have so far detected are mutations of the null activity or inactive enzyme type. This is not surprising since the mutants are induced by ionizing radiation. Two of these mutations which have occurred at different times in totally different animals were at the hemoglobin beta locus, one was at the malic dehydrogenase locus and the fourth was at the isocitrate dehydrogenase locus. The starch block that we use can be cut into 5 slices. Unfortunately, the 9 enzymes polymorphic in the two strains cannot be run in the same buffer system and several slices are discarded. In order to utilize this material we now stain for enzymes which are not electrophoretically different between the two strains. Through this added screening a variant was found at the LDH-loci, but this variant is not yet confirmed through breeding. It is not surprising that we should find a variant in the LDH-system; this enzyme is a tetramer, made up of two different polypeptides, each most likely coded by a different gene. In certain tissues all permutations of the two
A Biochemical Specific Locus Mutation System in Mice
49
polypeptides are found resulting in five bands on the gel. L D H is, therefore, analogous to a heterozygote within one strain. Of 50 C57BL/6J irradiated males only 2 regained fertility after the sterile period. A mating of these two males to DBA/2J females resulted in 14 F 1 which were screened electrophoretically (in total 126 loci). In one of these F 1 animals both the phosphoglucomutase locus allele and the isocitrate dehydrogenase allele from the C57 were inactive. These two loci are not linked and with the low induced mutation rate, it is highly unlikely that this unusual animal is the product of two independent mutational events in the two structural genes. The simplest hypothesis that comes to mind is that of a mutation in a regulator gene which is somehow responsible for the transcription of these two enzymes, one of which is in the glycolytic pathway and the other is in the Krebs cycle pathway. Certainly there are other explanations, for example, a possible position effect that could explain the phenomenon but we have not as yet done the chromosome analysis or any other tests to be able to ascertain exactly the mutational event which has led to this strange, but very interesting, phenotype. The utility of this system in relation to its providing relevant data for evaluating the mutagenic risk of chemicals depends on the number of loci which can be screened simultaneously. It is clear that in terms if sensitivity this new system with its nine loci is not significantly more efficient than the seven loci systems of Russell. However, presently 17 other electrophoretic allelic differences have been identified in various inbred strains. Taylor (1972) has described the biochemical relationship between the various mouse strains in a two-dimensional plate. Most of the common inbred strains fall in a cluster to which DBA belongs. C57BL/6 stands alone and quite far away from any of the other strains. C57BL/6J was, therefore, chosen to be the tester strain and in collaboration with Womack and Roderick at Jackson Laboratory (Bar Harbor, Maine, USA) these other markers are systematically backcrossed into the C57BL strain, the results will eventually be a new inbred strain with as many as 28 allelic differences from DBA, including the two new differences uncovered recently between DBA and C57BL/6. DBA is closely related to a series of other strains which nevertheless differ from DBA and each other at many physiological traits such as resistance to radiation. The new C57BL/6 will have a high number of electrophoretic differences form the strains closely related to DBA. By using this new C57BL/6 strain as a tester strain which each treated strain is mated to, it is possible to compare the effect of the physiological variations on the induced mutation rates. This increases our sensitivity almost four-fold over the seven loci morphological specific locus test. We are working with normal occurring variants of enzymes which are fully functional. There should, therefore, not be other than practical limit for the number Of genes we can accommodate in one strain. Additionally, we are working to develop new electrophoresis methods for more enzymes which have not yet been adequately screened in the mouse to detect polymorphisms and it is hoped that we will uncover more polymorphisms either within these strains or with other inbreds and further increase the number of loci that can be tested. Thus, if, as we suppose, we can increase the number of polymorphic loci in the system to approximately 40 and, remembering that we can treat both parents in a single experiment, then in a single F 1 animal we can screen for mutations in 80 different loci. Thus, you can see we can
50
H. V. Malling and L. R. Valcovic
very soon decrease the total number of animals that are required to provide reliable data on which to base a risk estimation for the human population. With the new alleles we are presently adding to the test, many of them are going to be located on the same chromosome. There are at least 6 chromosomes in which that is the case. In the morphological specific locus mutation experiment, there have been found some mutants in which two linked loci have appeared mutated simultaneously; the probability for this being two independent events is so low that this possibility can be excluded. These events have been explained as double non-disjunction. In a couple of years when we have combined these markers we will be able to detect non-disjunction in 6 - 7 different chromosomes. Furthermore, God-1 is on chromosome 19, which is a small chromosome. It will be interesting to see whether it is possible to obtain a mouse which is monosomic for one of these chromosomes, but there is probably little hope for that. This is the system that we have developed and are now evaluating in our laboratory. However, we are not yet totally satisfied with this. We feel that there may be other methods of detecting biochemical mutations by methods other than the use of electrophoresis; and one of these which we are currently looking into is that of measuring enzyme activity after incubating the enzymes at different temperatures, thereby, in essence, doing heat stability analysis of the enzyme. We have looked at the heat stability of several enzymes in these two strains, DBA and C57BL, and thus far have an indication that we have as many differences in heat stability as we have of electrophoretic differences but that these two characters are not correlated. There are several enzymes for which we cannot detect a difference in electrophoresis but can detect a difference in the heat stability of these enzymes. So while this is in very early stages of development, it is possible that heat stability, by way of measuring the enzyme activity, may eventually turn out to be the method of choice in detecting mutations. I would say one point in favor of this method; and that is that to measure enzyme activity, one can employ standard spectrophotometric techniques; and, of course, this means the possible application of automatic enzyme analyzing equipment. This could thereby reduce the time and probably the cost for mutational analysis. Electrophoresis, while it is a very fine technique, is a technique which is very difficult to modify to be totally automated; and so it will continually require a significant amount of manpower to perform such analysis. This allows me to stress an important factor that we must consider in designing methods for mutational screening and risk evaluation and even for human population monitoring. We must recognize not only accuracy with which we are measuring the biological event or the mutational event; in other words, the sensitivity of the biological system; but also the technological ease with which such tests can be performed and the reliability associated with these tests so that they need not always be done highly specialized and highly trained scientists.
A Biochemical Specific Locus Mutation System in Mice
51
References Lyon, M. F., Morris, T.: Gene and chromosome mutation after large fractionated or unfractionated radiation doses to mouse spermatogonia. Mutation Res. 8, 191 (1969) Russell, W. L.: The effect of radiation dose and fractionation on mutation in mice. In: Repair from genetic radiation damage (F. H. Sobels, Eds.), p. 205. Oxford: Pergamon 1963 Taylor, B. A.: Genetic relationships between inbred strains of mice. J. Heredity 63, 83 (1972) Valcovic, L. R., Mailing, H. V.: An approach to measuring germinal mutations in the mouse. Env. Hlth Perspect. 6, 201 (1973)
Received December 21, 1976