MUTABLE LOCI IN MAIZE Barzgara McCriintock Study of the origin and nature of insta- bility of genic action at a number of differ- ent known loci in maize chromosomes has been continued during the past year. At any one locus of known genic action, different types of instability expression can appear. A hypothesis has been developed to account for the origin and subsequent behavior of these various types of un- stable condition. Each type is considered as reflecting the operation of a particular chromosomal system that controls the ac- tion of the genic components at the locus. These controlling systems appear to be composed of distinct chromosomal units; but, unlike the genes, these units may move from one location to another within the complement. . Two well defined classes of controlling system have been dealt with in our study. They have been termed single-unit sys- tems and two-unit systems. In the single- unit systems, only one controlling unit is recognized. When such a unit moves to a new location in the chromosomal com- plement, it imposes its specific mode of control of action of the genic components at this location. The time and place of genic action, as well as its type, will be an expression of the specificity of the par- ‘ticular controlling unit. Two-unit sys- tems operate differently. When one of these units is inserted adjacent to a par- ticular gene, it may immediately alter genic action, giving rise to a recognizable mutation. Changes in this unit may sub- sequently occur. These are reflected in altered expressions of the associated genic components. They occur only when the second unit of the two-unit system is also present in the nucleus. This second unit is independently located in the chromo- somal complement. Instability of genic ex- pression is a reflection, therefore, of the interaction of the two units. Changes 1n the gene-associated unit control the types of change in genic action, whereas the time and place of such changes are con- trolled by the second unit. The main purpose of recent studies has been to examine the same system of in- stability expression at a number of differ- ent known loci, and to compare the action of various systems that may operate at the same locus. The possibility of recognizing the oper. DEPARTMENT OF GENETICS tion of a particular system controlling in- stability of genic expression depends on its specificity. The Ac component of the two- unit Ds-Ae controlling system, described in previous reports, is highly specific in its action. Its operation is readily detected. In plants having Ds and Ac, mutability under the control of Ac has appeared at several known loci. The first recognized case was designated c™*. The origin of c™? by transposition of Ds to the locus of C has been described in previous reports. Subsequently, the Ds-Ac system of muta- tional control appeared at the Bz and Wx loci. Furthermore, several independent in- ceptions of this controlling system have occurred at each of these three loci. Other systems of mutational control have arisen, however, at the same loci. Since instability of genic action is considered to be an ex- pression of the operation of a particular controlling system, which has been incor- porated at the locus of the gene concerned, and not an expression of changes in the gene itself, it is to be expected that any one system can operate at any locus of known genic action. This expectation could be tested; and the Ds-Ac two-unit controlling system was chosen for this pur- pose. Instability at a selected locus can arise whenever the Ds component is trans- posed to the locus. By technically simple methods, it should be possible to detect the presence of this system within a few cell generations after incorporation of Ds at the locus. Two loci considered par- ticularly appropriate for the initial tests were selected: that of A: and that of Ae. The results of these tests are summarized in the following section. Oricins oF INSTABILITY AT THE 41 AND A>» Loct The genic components at the locus of A, in chromosome 3 and the locus of Az 213 in chromosome 5 are associated with the development of anthocyanin pigment in the aleurone layer of the kernel and in the plant tissues. If both Ds and Ac are pres- ent in plants homozygous for these two dominant, stable factors, instability may arise at either locus if Ds is transposed to it. (By use of the term factor instead of gene, it is hoped that misconceptions re- garding the nature of the change affecting the genic components at a locus may be avoided.) Transpositions of Ds usually occur late in the development of sporog- enous tissues. Should one such event in- sert Ds at the locus of A: or Ae in these plants, genic action at the affected locus might be altered. Such altered action could be detected in kernels on the ears of these plants if the progeny of the cell in which the transposition occurred pro- duced a female gametophyte, and also if the sperm nucleus entering this female gametophyte carried the known recessive, a1 or a2, whose action is unaffected by Ac. Color development in the aleurone layer of the kernel so produced might be modi- fied. Variegation might be exhibited if dc was also present in this kernel. The plants used as the female parents, in crosses to test this theory, had Ds located in the long arm of chromosome 5, closely linked to the factor Pr. One Ac factor was present, and all plants were homozygous for the stable, dominant factors 4: and A2. Pollen from plants homozygous for either a, or a2 was placed on the silks of such plants. The resulting ears were examined for kernels showing variegation of aleu- rone color. Seventy-one ears were obtained from the first cross and 120 from the sec- ond. One kernel showing variegation for aleurone color was found on an ear when a: had been introduced by the male parent, and three were found on three different ears when az had been introduced by the male parent. Plants were grown from all 214 four kernels, and tests initiated to deter- mine the nature of the instability expressed. In the plant derived from the single variegated kernel appearing in the first cross, mutability was being expressed at the locus of A:. This new mutable condi- tion, designated a:"“, proved to be Ac- controlled. It arose at the normal A: locus in the Ds-Ac-carrying female parent. ‘This could be determined from linkage studies. The female parent had been homozygous for A: and Shz, whereas the male parent had been homozygous for the recessive alleles. Sho is very closely linked with 41. Close linkage of a:"* and Shz was evi- dent in backcross tests. Three other in- dependent inceptions of instability at this Ay locus have been examined. Two of them, a and a”, are not Ac-con- trolled; but the third, a”, is Ac-con- trolled and arose in a plant having the very same constitution as that which produced av, Plants were grown from the three varie- gated kernels derived from crosses in which the pollen parent had been homo- zygous for az. In two of them, it could be shown that modifications had occurred at the locus of A2 in the Ds-Ac-carrying parent. Both modifications resulted in in- stabilities of genic action at this locus, which were designated a,* and a”. In the third plant, the chromosome 5 con- tributed by the female parent was not transmitted through either pollen or egg; therefore the nature of the alteration at 42 could not be determined. Mutability at a:™* is Accontrolled. The nature of the controlling system associated with ax™* has not yet been determined; it is not Ac- controlled, however. Initial tests were com- plicated by the presence of a defect in the chromosome 5 carrying a2", which pre- vented pollen transmission. The locus of the defect is closely linked with Pr. Trans- CARNEGIE INSTITUTION OF WASHINGTON mission of a2”* through the pollen was limited. It occurred only when the grain carried a chromosome 5 without the defect. Such a chromosome is obtained by cross- ing over between a2”° and the locus of the defect. In considering the origins of @2"* and as™*, it should be recalled that Ds was present in the same chromosome that car- ried As. In two of the three plants derived from the variegated kernels, the chromo- some 5 with a modification at A» was also defective. It is known that events at Ds produce chromosomal alterations. Al- though some of them do not produce chromosomal aberrations that result in in- viability, others do produce gross chromo- somal defects, such as deficiency. It is not surprising, therefore, to find that some dc fect in chromosome 5, affecting viability. ‘ accompanied the inception of mutability in two of these three cases. This relationship affords additional, although indirect, evi- dence that altered genic expression origi- nates through chromosome aberration. Two earlier cases in our material of mu- tability produced by changes at the /. locus are designated a2"? and a.””. Neither is Ac-controlled. INSTABILITY OF Shi Action INpucep By Ds Two independent cases of insertion ol Ds just to the left of the SA locus in chro- mosome g have been studied. With regard to the hypothesis of origin of mutation through Ds events, these two cases have been exceptionally revealing. Early studies of Ds were focused on the phenomenon of its transposition from one location to 2” other in the chromosome complement. For the sake of technical simplicity, cas¢* were selected in which Ds had been tran* posed from one position in the short a1™™ of chromosome 9 to another position with in the same arm. Over twenty such cas DEPARTMENT OF GENETICS were examined in detail. In five of them, Ds had been inserted between I and SA. These two factors lie close together in the chromosome, and approximately 4 per cent crossing over occurs between them. From linkage tests, it was determined that Ds had been inserted closer to 7 than to Sh in three of the five cases, and at approxi- mately the same position in each. Cross- ing over between I and Ds was approxi- mately one-fifth of that which occurs be- tween I and SA. In the other two cases, Ds was inserted just to the left of Sh. To determine crossover frequencies between Ds and Sh in these two cases, extensive tests were required. A number of kernels with exceptional phenotypes appeared in these tests. Studies were then initiated to determine the conditions associated with their appearance. They proved to be Ds- initiated mutations at the locus of SA. A summary of the evidence may be given. The initial tests were made in order to determine the crossover frequencies be- tween I and Ds and between Ds and Sh. The tests were conducted as follows for both case 1 (Ds 4864A) and case 2 (Ds , 5245): Plants having one or two dc fac- tors and carrying I, Ds, Sh, Bz, and Wx in one chromosome g, and C, ds, sh, bz, and wx in the homologue, were crossed to plants homozygous for C, ds, sh, bz, and wx and having no Ac factor. From one such test with case 1, only a single kernel was found among a total of 4291 that could be considered to have a chromosome derived from crossing over between Ds and Sh. In phenotype, it was I sh bx wx; and Ds was present in the /-carrying chromosome. In 26 kernels, the phenotype was I sh Bz; and in 17 of them Ds was certainly present, located between I and Bz. (The presence of Ds may be detected only in those kernels that also have Ac.) In normal stocks, crossing over between 215 Sh and Bz is approximately half that be- tween J and Sh, or close to 2 per cent. If the kernels with the J sh Bz phenotype car- ried a chromosome derived from a double crossover in the two adjacent short seg- ments, I to Sh and S& to Bz, then kernels carrying the reciprocal crossover chromo- some should also have been present. They should have had the C Sh bz phenotype. No kernels showing this phenotype were present. It would be even more difficult to explain in terms of crossing over the origin of the 17 kernels with Ds located between 7 and Bz. Double crossing over, involving the Ds-to-Sh and Sh-to-Bz seg- ments, would have been required. This seems even less probable when it is recalled that only one kernel had a phenotype that could have appeared as a consequence of a single crossover between Ds and Sh. Obvi- ously, the I sh Bz phenotype does not rep- resent a product of crossing over. Another mechanism is responsible. The results in similar tests of case 2 were much the same as those described for case 1. In one such test, 6683 kernels were obtained. Seven of them were I sh bz wx in phenotype, and Ds was present in the J chromosome. They could be interpreted as having originated through crossing over between Ds and SA. Twenty- nine kernels were I sh Bz; and in 17 of them Ds was certainly present, located be- tween I and Bz. No kernels having a C Sh bz phenotype appeared. The argu- ments given above for case 1, which ex- cluded crossing over as a mechanism re- sponsible for the appearance of sh in kernels having an I sh Bz phenotype, ap- plied with equal force for case 2. On the other hand, this phenotype could be at- tributed to mutation at the locus of SA. Since mutation to sh had not been observed to occur with such a high frequency in other genetic stocks of maize, or in those 216 cases in our material where Ds was inserted at other locations in the short arm of chromosome g, it could be suspected that these mutations were produced by some event undergone by Ds. Tests were initi- ated to determine whether this was true. Events at Ds will occur only if Ac is present in the nucleus. If the sh pheno- type arises from a particular event at Ds, then no mutations to sh should occur if Ae is absent. Thus, had the experiment de- scribed above been conducted with plants having the same constitution with respect to markers in their chromosomes 9, but having no Ac factor, kernels with an I sh Bz phenotype would not have ap- peared on the ears, for no events at Ds would have occurred in any of the cells of these plants. Such a comparative test | was conducted with case 2. All the plants in one culture had the same constitution as described above with respect to markers in chromosome 9, but differed with respect to Ac, which was present in some and ab- sent from others. When those having Ac were crossed to plants homozygous for C, sh, bz, and wx, and carrying no Ac factor, some kernels with an I sh Bz phenotype appeared on the resulting ears, as in the experiment described earlier. When four of the sister plants that had no Ac were similarly crossed, no I sh Bz ker- nels appeared, among a total of 2553, on the resulting ears, and none of the kernels was variegated; for no Ac was present in any of them and therefore no breaks at Ds could occur. Pollen from these same four plants was placed on silks of plants homo- zygous for C, sh, bz, and wx and also carrying an Ac factor. Ds events could now occur, but only during the develop- ment of the kernels that had received Ds from the male parent and Ac from the female parent. Mutation to sh might origi- nate in some of these kernels, which would CARNEGIE INSTITUTION OF WASHINGTON then show segments of the sh phenotype but would not be totally sh. Among 1657 kernels produced from this cross, none having 7 and Bz was totally sh in pheno- type. In the reciprocal cross, 1213 kernels were obtained. In such a cross, the B- phenotype can be detected only in those kernels that also have Ds and Ac. No I s/ Bz phenotypes were found among them. It should be emphasized that in these re- ciprocal crosses no evidence of crossing over between Ds and Sh appeared; that is. in none of the kernels with an J sh bz wx phenotype was Ds detected. They ap- parently carried a chromosome derived from crossing over between J and Ds. In both case 1 and case 2, the frequency of crossing over between I and Ds seemed to be the same as that between J and SA in other stocks. Similarly, no change in cross- over frequency between SA and Bz was noted in either case. The comparative tests outlined above support the hypothesis that the sh phenc- type in the J sh Bz kernels originates through events at Ds, when the latter is lo cated just to the left of SA; for this phen type appears only when Ac is also present. A second test of this hypothesis was con- ducted with both case 1 and case 2. In plants that are homozygous for I Ds Si and carry an Ac factor, mutations to :/ should occur in a few cells late in the d- velopment of the sporogenous tissue, oF 1” cells of the gametophyte. Kernels that are I sh in phenotype should appear on th: ears from reciprocal crosses with plant homozygous for C ds sh. The frequen: of appearance of such kernels in these ' ciprocal crosses is recorded in table : Many of the sé kernels certainly carne: Ds in the I chromosome. Its location b= tween I and Bz could be affirmed in t majority of cases because of the pattern ' variegation produced by breaks at Ds. DEPARTMENT OF GENETICS If the described mutation results from inhibition of SA action by Ds itself, then reversion to SA might result from a subse- quent event at Ds that removed this in- hibitory action. In order to obtain more evidence about the nature of the events that are responsible for mutations at this locus, plants were grown from some of the 1 sh Bz kernels in which both Ds and Ac were present. Tests were conducted to de- termine the viability of the mutant sh when homozygous, the position of Ds, and whether or not reversion to SA would occur. TABLE 3 FREQUENCY OF MUTATION TO sh wn RECIPROCAL CROSSES CrOSS. oo cee eee eee eee eee eens IDs Shf CdsshQ IDs Sh; by I Ds Sh/ 1Ac ¢ I Ds Sh; by Cdssh J 1AC PH Kernel type.........-.-+-00e. ISh Ish ISh Ish Case 1 (Ds 4864A)....... 644 7 6349 128 Case 2 (Ds 5245)........ 9082 18 5004 34 For case 1, ten kernels were selected that had the J sh Bz phenotype and carried both Ds and Ac. Only six of them germinated. A mutant sf factor was present in four of the plants. It was not present in one plant, whose constitution was like that of the heterozygous parent: I Ds Sh Bz Wx/ C ds sh bz wx; t Ac. The sixth plant was triploid, with the following chromosome-9 constitutions: 1 Ds Sh Bz Wx/I Ds sh Bz Wx/C ds sh bz wx. Two Ac factors were present. A mutant sf factor was present in one of these /-carrying chromosomes. An ear derived by self-pollination was obtained from each of the four diploid plants having a mutant sf factor. The ratio of I to C kernels indicated no reduc- tion in viability of the mutant when homo- zygous. In crosses of these four plants to those homozygous for C, ds, and bz, a 217 few kernels with an Sh phenotype ap- peared, but only on one of the test ears from crosses involving one of the four plants. It is impossible to be certain that they were not the result of contamination, but the associated phenotypes make it un- likely that they were. In three of the four plants, crossing over between J and Ds and Ds and Bz was the same as that which occurred in the heterozygous parent plants, indicating no decided shift in the position of Ds in the origin of the sh mutation. In the fourth plant, the frequency of crossing over between I and Ds was unchanged, but that between Ds and Bz appeared to be reduced. In case 2, 21 kernels having an I sh Bz phenotype and also carrying Ds and Ac were selected for testing. Only 13 of them germinated. A mutant sf factor proved to be present in twelve plants. The constitu- tion of the thirteenth plant was similar to that of the heterozygous parent: I Ds Sh Bz Wx/C ds sh bz wx; 1 Ac. The I-to-C ratio on ears derived from self-pollination of the twelve plants carrying a mutant sh factor indicated no reduction in viability of the kernels homozygous for this factor. Crossing over between J and Ds was un- modified in all twelve plants. That be- tween Ds and Bz was unmodified in eight plants, but in four plants it appeared to be reduced. On the ears derived from crosses to plants homozygous for C, sh, bz, and wx, kernels having an I Sh Bz phenotype appeared in tests of three of the eight plants in which crossing over between Ds and Bz was unmodified. The rates of reversion were low in two plants, but markedly higher in the third plant (5933-1). This applied to rates of germinal mutation, that is, mutation occurring in the sporogenous or gametophytic cells. If mutation occurs during the development of the kernel, a sector with the Sh phenotype will be 218 formed. To be detected, such sectors must be large; in other words, the mutation must occur early in the development of the kernel. All lJate-occurring mutations will be overlooked. Some sectorial kernels developed on the ears derived from crosses of all three plants mentioned above, the frequency of such kernels being highest on the ears derived from crosses of plant 5933-1. It appeared that this plant carried a mutable sh locus; and tests conducted with the progeny of the plant during the winter of 31951-1952 fully confirmed its presence. Although Ac control of mutation is probable in this case, it cannot yet be demonstrated with certainty. Inability to detect all kernels showing mutation to SA necessitates extended tests for this deter- mination. It is known from studies of Ds-initiated mutable loci involving factors associated with color development, which may be de- tected even in individual cells, that the fre- quency of reversion depends on the state of Ds. For some states, the rate is high; for others it is very low. Only three of the twenty examined Ds-initiated mutations to sh have provided certain evidence of sub- sequent reversion, although a fourth in- stance may possibly have occurred. The original J sk Bz kernels selected for testing did not show any detectable somatic rever- son to Sh. They may represent, therefore, a selected class in which the state of Ds is one producing relatively few reversions. Tests of an unselected sample are now under way. They should give a better in- dication of the frequency of production of highly mutable sh loci through the me- dium of Ds events. Nevertheless, from the several tests outlined above, there appears to be little question that events at Ds, fol- lowing its insertion just to the left of SA, are capable of inducing mutation to sh. It is also certain that some of the mutants are CARNEGIE INSTITUTION OF WASHINGTON capable of subsequent reversion to SA. Since one of the selected s# mutants proved to be highly mutable, additional examples of high mutability may appear when ap- propriate methods for their selection are employed. SUMMARY Instability of genic action, under the con- trol of the two-unit system of which Ac is one of the components, has been examined at six different loci. (See table 4.) Four of these six loci are associated with antho- cyanin pigment formation (C, Bz, Ai, and Az), one with composition of the starch in pollen and endosperm (Wx), and the sixth with a morphological structure of the endosperm (SA). It is clear that this two- unit system may operate at loci concerned in quite different types of phenotypic ex- pression. It cannot be stated that the sys- tem could operate at any known locus; but it should be capable of operating at many. Mutational behavior under the control of Ac always follows a distinct pattern, re- gardless of the locus involved. No muta- tions occur when dc is absent. They ap- pear only when Ac is present, and the mc and place of their occurrence are an expres- sion of the particular state and dose of Ac. At five of the above-mentioned loci (col- umn 3 of table 4), other systems control- ling genic expression are known to operate. Some aspects of the behavior of ai”~° arc unique in our studies. Two distinct classes of mutation occur. One class produces phenotypes indistinguishable from that given by the normal Ai factor: deep pis: mentation in the aleurone layer and intensc pigmentation in all parts of the plant. The other class produces a complex series o! changes affecting anthocyanin develop- ment in plant and aleurone. Mutations i9 the latter class form a graded series with respect to the intensity of pigmentation produced. In the plant, however, pigm¢™ DEPARTMENT OF GENETICS tation is always restricted to the root, stalk, sheath, auricle, and glume; none appears to be present in the leaf, except in the mid- rib and along the edge. The effects of a TABLE 4 KNown Loci (coLUMN I) AT WHICH MUTABILITY, UNDER THE CONTROL OF 4c, HAS ARISEN (cOL- UMN 2), AND MUTABILITY AT THE SAME LOCI CONTROLLED BY OTHER SYSTEMS (COLUMN 3) (The figures following the symbols represent the sequence of appearance in the Cold Spring Harbor cultures.) Symbol for normal, Instability instability dominant factor controlled seston other ystem at locus by Ac ¥ than Ac Coc ccc cence cm} ons gm-2 coma—4 Shyo ccc cee eee ee See text for cases Bo. ee eee bem} b2m—3 bam? 1 wan? wan? wens wens was wam4 A Lecce eee eee aes ay ay"! a yn ayn? A Deore e ener eeceees aym—4 a,m! am? ag particular mutation of this class on inten- sity of pigmentation may not correspond in plant and aleurone. For example, a mutation producing very slight pigmenta- tion in the aleurone may give very intense pigmentation in the affected plant tissues. Also, some of the mutations observed pro- duced totally colorless kernels and in- tensely pigmented plants. Another distinc- tive feature of a:""* is related to the effect 219 of crossing over in the immediate vicinity of this locus, that is, between ai”? and Sh2. The majority of crossovers within this short segment result in mutation that ap- pears to belong to the second class men- tioned above. In contrast, tests of crossing over between Sz and a:""', ai", or ai™* have given no such clear indication of mu- tation accompanying crossing over. Still other aspects of behavior peculiar to a." have been observed, but description will be postponed until further evidence is available. It is clear from our studies that different systems of control of genic expression can arise at any one locus in the chromosome complement, and that the same system may operate at different loci. The organization of the chromosomes with respect to functional units, and the types of functional units that may be pres- ent, are not clearly understood. Our studies indicate that at least two classes of func- tional genetic unit are carried by the chro- mosomes: one of them potentially capable of determining a particular course of cellu- lar reactions, the other associated with the control of this potential action. It has been possible to distinguish between these unit components at a locus and to describe, in terms of phenotypic expression, their modes of operation. Our studies also sug- gest that many mutations may be expres- sions of changes in controlling systems, the potential capacities of the gene units re- maining unchanged. At present, there is no way to distinguish, on the basis of muta- tion alone, between alterations in potentiali- ties of genic action and alterations in the controlling systems that leave the potential unchanged, except in those cases where the latter is clearly evident.