Mutations Increasing Asexual Plasmodium Formation in Physarum Polycephalum

Home , Heterokaryon, Plasmodium (life cycle)

MUTATIONS INCREASING ASEXUAL PLASMODIUM FORMATION IN PHYSARUM POLYCEPHALUM

PAUL N. ADLERI AND CHARLES E. HOLT

Department of Biology, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139 Manuscript received May 9, 1977 Revised copy received July 28, 1977

ABSTRACT

Rare plasmodia formed in clones of heterothallic amoebae were analyzed in a search for mutations affecting plasmodium formation. The results show that the proportion of mutants varies with both temperature (18", 26" or 30") and mating-type allele (mti, mt2, mt3, mt4). At one extreme, only me of 33 plasmoida formed by mt2 amoebae at 18" is mutant. At the other extreme, three of three plasmodia formed by mt1 amoebae at 30" are mutant. The mutant plasmodia fall into two groups, the GAD (greater asexual differentiation) mutants and the ALC (amoebaless life cycle) mutants. The spores of GAD mutants give rise to amoebae that differentiate into plasmodia asexually at much higher frequencies than normal heterothallic amoebae. Seven of eight gad mutations analyzed genetically are linked to mt and one (gad-12) is not. The gad-12 mutation is expressed in strains with different alleles of mt. The frequency of asexual plasmodium formation is heat sensitive in some (e.g., mt3 gad-li), heat-insensitive in two (mt2 gad-8 and mt2 gad-9) and cold-sensitive in one (mt1 gad-12) of twelve GAD mutants analyzed phenotypically. The spores of ALC mutants give rise to plasmodia directly, thereby circumventing the amoeba1 phase of the life cycle. Spores from five 3f die seven ALC mutants give rise to occasional amoebae, as well as plasmodia. The amoebae from one of the mutants carry a mutation (alc-1) that is unlinked to mt and is respon- sible for the ALC phenotype in this mutant. Like gad-12, ah1is expressed with different mt alleles. Preliminary observatims with amoebae from the other four ALC mutants suggest that two are similar to the one containing alc-1; one gives rise to revertant amoebae, and one gives rise to amoebae carrying an alc mutation and a suppressor of the mutation.

PHYSARUM POLYCEPHALUM is an acellular slime mold or Myxomycete. P,n the life cycle of this organism, plasmodia give rise to spores, spores germinate to yield amoebae, and amoebae form plasmodia (GRAYand ALEXOPOULOS 1968). Plasmodia are pigmented, vegetative structures in which nuclear di- vision but not cytokinesis occurs; thus, they are typically multinucleate and are of indefinite size. When starved in the light, a plasmodium gives rise to groups of spores borne on stalks. Spores, when moistened, yield amoebae. Amoe- bae are nonpigmented and uninucleate, undergo cytokinesis, and are usually

Present address: Department of Biology, University of Virginia, Charlottesville, Virginia 22901.

Genetics 87: 401-410 November, 1977. 402 P. N. ADLER AND C. E. HOLT haploid. Like plasmodia, they are vegetative cells that may be grown without change in form for an indefinite period. A multiallelic mating-type locus (mt) controls the mode and frequency of plasmodium formation. Twelve heterothallic alleles (mtl,mt2, . . . mtl2) and one “selfing” allele (mth) are known (DEE 1966; WHEALS1970; COLLINS 1975; COLLINSand TANG1977). The alleles mth, mtl, mt2, mt3 and mt4 are available in a common genetic background (COOKEand DEE 1975; ADLERand HOLT1974b). In the sexual mode of plasmodium formation, cellular and nuclear fusion of haploid amoebae carrying diff erent mt alleles produces a diploid plasmodium. Meiosis accompanies sporulation in such a plasmodium (ALDRICH 1967; LAANEand HAUGLI1976). In the asexual mode, plasmodia form within clones of amoebae without evidence of amoebal fusion (ANDERSON,COOKE and DEE 1976). Almost all the nuclei of a plasmodium derived asexually from a haploid amoeba are haploid (COOKEand DEE 1974; ADLERand HOLT1975); however, there are occasional diploid nuclei in such a plasmodium, and it is likely that only these survive sporulation (LAFFLERand DOVE1977). Sexual plasmodium formation is efficient between strains carrying any pair of mating- type alleles, with the exception that mating between mth and mt2 strains is very inefficient (ADLERand HOLT1974b; COOKEand DEE 1975; DAVIDOWand HOLT1977). Asexual plasmodium formation, or selfing, occurs at a high frequency in mth amoebae and at much lower frequencies in heterothallic amoebae. In a culture of mth amoebae, more than 10% of the cells become individual plasmodia (YOUNGMANet al. 1977). As shown earlier (ADLERand HOLT1975) and confirmed below, the most frequent selfing in heterothallic strains generates only one plasmodium in lo8 amoebae. Diploid amoebae heterozygous for mt have been isolated; these form diploid plasmodia at a high frequency (ADLER and HOLT1975). Note that although mt has a marked influence on the con- versioln from the amoebal to the plasmodial state, any genotype at mt seems to be compatible with either state. Mutational analysis has revealed additional genes that control whether or not an amoeba becomes committed to the plasmodial state. Mutations at aptA (WHEALS1973) and npfA block the asexual conversion of mth amoebae. The two loci are unlinked to mt and to one another (ANDERSONand DEE 1977). A number of other mutations that interfere with asexual plasmodium are tightly linked to mt; the mutations fall into two complementation groups (ANDERSONand DEE 1977; DAVIDOWand HOLT1977). In the present work, we sought mutations that would increase the frequency of asexual plasmodium formation in heterothallic amoebae. Our method was to pick, grow and sporulate the rare plasmodia formed by heterothallic amoebae and to examine amoebal progeny of the plasmodia for mutant characteristics. The formation of nonmutant plasmodia by the amoebae created a background that seriously interfered with the search. We found that the background could be reduced by appropriate selection of mating type and temperature, and we isolated and characterized 20 mutants. DIFFERENTIATION MUTANTS IN PHYSARUM 403

TABLE 1 Strains

Strain number Relevant genotype Reference CHI mth fusA2 fusel COOKEand DEE1975 CH9 mth fusA2 fusC1 aptAI WHEALS1970 CH21 mt3 fusA2 fusC2 ADLERand HOLT1974b CH54 mt3 fusA2 fusC2 ADLERand HOLT1974b CH190 mt3 fusA2 fuse2 ADLERand HOLT1974b CH269 mtl fusA2 fusC1 COOKEand DEE 1975 CH273 mt4 fusA2 fusC2 ADLERand HOLT1974b CH274 mt2 fusAl fuel COOKEand DEE1975 CH322 mt3 actE9 emeE4 fusA2 fuse2 ADLER1975 CH326 mt4 actE8 fusA2 fusC2 ADLER1975 CH3.29 mt4 actE3 fusA2 fusC2 ADLER1975 CH343 mt3 ADLER1975 CH344. mt3 ADLER1975 CH347 mt3 fusA2 fusC2 ADLERand HOLT197413 CH348 mt3 jusA2 fusC2 ADLERand HOLT1974b CH351 mt3 fusA2 fusC2 ADLERand HOLT1974b CH393 mt2 ADLER1975 CH394 mt2 fusAl fusC1 ADLER1975 CH396 mt2 fusA2 fusC1 ADLER1975 CH50l mt4 fusA2 fmC2 ADLER1975 CH586 mt3/mt4 ADLERand HOLT1975 CH6348 mt2 ADLER1975

MATERIALS AND METHODS Media: Plasmodial rich medium agar (PRM-agar), dilutc plasmodial rich medium agar (dPRM-agar), dilute plasmodial rich medium agar pH 7 (dPRM pH 7 agar), and liver infusion agar (LIA) were prepared as described previously (ADLERand HOLT1974b; YOUNGMANer al. 1977). Strains: Strain numbers, with relevant genotype and history-, are shown in Table 1. The fus genes, which affect plasmodial fusion, were used as markers; they do not influence plasmodium formation. All strains have a Colonia (CHI) genetic background. Strain CH21 would be expected to contain about 87.5% Coloiiia genes. All other strains should contain more than 95% Colonia genes. Strain CHI has been called CL previously (ADL.ERand HOLT1974b). Strains CH269 and CH274 are LU648 and LU688, respectively, of COOKEand DEE (1 975). The GAD mutants have the following parents: mutant CH403 (parent CH396) ; CH404 (CH274) ; CH405 (CH394) ; CH4.78 (CH.269) ; CH479 (CH322) ; CH480 (CH394) ; CH484. (CH329) ; CH485 (CH322) ; CH486 (CH394); CH487 (CH394) ; CH489 (CH322); CH496 (CH21), and CH526 (CH326). Mutant CH485 (mt3 gad-7) was isolated at 26"; see Table 5 f3r the other GAD mutants. The ALC mutants have the following parents and temperatures of isolation: mutant CH5001 (parent CH21, 26"); CH5002 (CH269, 30"); CH5003 (CH269, 30"); CH5004 (CH322, 30"); CH5005 (CH326,30°) ; CH5006 (CH326, 30°), and CH5007 (CH326,26"). Culture procedures: Plasmodia were formed from amoebae on dPRM-agar, grown on PRM- agar, and induced to sporulate as described prcviously (ADLERand HOLT1974). Methods for germinating spores and growing amoebae are described in the same publication. Kinetic experiments (YOUNGMANet al. 1977) were performed as follows. Replicate amoeba1 cultures were set up by pipetting 0.05 ml of a suspension of plasmodia-free amoebae and E. coli 404 P. N. ADLER AND C. E. HOLT

onto a dPRM agar plate. Cultures were assayed by flooding a plate with 5 ml H,O and scraping the surface of the agar with a glass rod. The resulting suspeiision was diluted with H,O and dilutions plated onto assay plates. Assay plates were either LIA or dPRM pH 7 agar (both of which retarded the differentiation of amoebae into plasmodia) and were incubated at 26" or 30" (this depended on the specific strain) for 5-9 days, and the number of amoebal and plasmodial plaques were counted. For the lowest dilution, 1/5 of the original suspension was plated; thus, the assay should detect as few as 5 plasmodia per original plate. Isolation of plasmodia from heterothallic amoebae: A suspensioii of bacteria and approximately IO4 ameobae was pipetted onto each of a series of dRPM-agar plates. 'The inoculum was allowed to remain as a spot with a diameter of about 20 mm. The plates were incubated in plastic bags (to reduce evaporation) at either IS", 26" or 30" for 4 to 6 weeks. Plates were examjned weekly for the presence of plasmodia. Once formed, the plasmodia were grown and induced to sporulate in the standard way. Plasmodia were grown ai 26" regardless of the temperature at which they formed. Determimiion of phmotype: Mating specificity, clonal plasmodium forming behavior, and plasmodial fusion behavior were determined as described previously (ADLERand HOLT1974b, 1975).

RESULTS Isolation of mutants Eflect of temperature and mating type on plasmodium formation: A number of heterothallic strains were tested for the frequency of plasmodium formation at

TABLE 2 Frequency of formation of plasmodia in cultures of heterothallic amoebae*

Number of cultures producing plasmodia Mating type Strain 18' 26" 30' mtl CH269 2 5 2 CH269 8 5 1 mt2 CH274 21 0 0 CH274 19 1 0 CH394 47 7 2 mt3 CH21 1 1 0 CHI90 5 6 0 CHI90 11 7 0 CH322 10 15 3 mt4 CH329 1 0 0 CH326 1 1 4 CH326-f 1 1 0 Summary-average frequencies mtl 0.10 0.10 0.03 mt2 0.58 0.05 0.01 mt3 0.14 0.15 0.a2 mt4 0.02 0.01 0.013

* Each raw in the upper part of the table presents the results of an experiment begun with 150 replicate amoebal cultures on dPRM agar. Fifty of the cultures were incubated at each of IS", 26" and 30". The cultures at 26" and 30" were incubated for 4 weeks and those at 18" for 6 weeks. The numbers of cultures that produced one or more plasmodia during these periods were recorded. The data were used to calculate the average frequency of cultures producing plasmodia (lower part of table). + In this experiment, amoebae were mutagenized with ethyl methane sulfonate (HAUGLI1971). DIFFERENTIATION MUTANTS IN PHYSARUM 405 18", 26" and 30" (Table 2). At 26", the bulk of the plasmodia appeared by the end of the third week of incubation and the totals at the end of four weeks are mported. At 18", plasmodia continued to appear through the fifth week of mcu. bation and the totals at the end of six weeks were reported. At 30", too few plasmodia appeared to provide useful information on the timing. Except for strain CH21, the mtl and mt3 strains behaved similarly to one another, showing about the same frequency of plasmodium production at 18" and 26" and a markedly reduced frequency at 30". The mt2 strains displayed a relatively high frequency of plasmodium production at 18" and progressively lower frequ,encies at 26" and 30". The mt4 strains gave low frequencies at all temperatures. Thus, it appears that the particular heterothallic allele of mt carried by a strain determines the frequency and temperature sensitivity of plasmodium production. It seemed possible that the effect of temperature on plasmodium formation was a trivial effect of temperature on amoeba1 growth or viability. To test this POS- sibility, we measured the growth and viability of CH322 (mt3)amoebae at three temperatures over a 4-week period. Growth determined by hemacytometer count was the same at both 26" and 30" (Figure 1) ,and showed a lag and reduced yield at 18". The viable count was generally lower than and occasionally as low as 40% of the hemacytometer count, but the reduction seemed to be due to clump- ing rather than to the presence of inviable cells. A mt2 and a mt4 strain behaved almost identically to the mt3 strain. Thus, the lower frequencies of plasmodium formation do not seem to arise from low growth rates or low viability. To test further the hypothesis that plasmodium formation by a heterothallic strain is a function of the specific mating type allele it carries, we analyzed

Time (days) FIGURE1.-Growth of mt3 (CH322) amoebae on dPRM agar plates as determined by hemacytometer counts. 0,18"; 0,26"; 0,30". 406 P. N. ADLER AND C. E. HOLT the progeny of a cross to see whether mating specificity and the temperature- dependent frequency of plasmodium formation were linked. A cross between a mt2 and a mt4 strain was chosen since these alleles were associated with the most extreme phenotypes. Twenty progeny strains were isolated and their frequency of plasmodium production was determined at 18" and 26". One progeny strain was a mt2/mt4 heterozygote and was not analyzed further. The nine mf4 progeny and the mt4 parent formed no plasmodia at the two temperatures (Table 3).This behavior is the same as that of the mt4 strains tested previously (note that 50 cultures were examined at each temperature in the experiment of Table 2 in comparison with 10 cultures at each temperature in the present experiment). In the first tests with the 10 mt2 progeny strains and the mt2 parent strain, all except progeny number 10 produced plasmodia at 18". The strains were tested again at 18", and this time all ml2 strain produced plasomdia (Table 3). The frequencies are in adequate agreement with those seen earlier for mt2 strains. Thus, the marked difference in plasmodium forming capacity between mt2 and mt4 amoebae is due to thpemt alleles, or to genes linkec! to mt. The data in Table 2 show statistically significant differences between strains CH274 and CH394, both of which are mt2, and between CH21 and CH190 or CH322, all of which are mt3. The basis for these differences has not been investi- gated. Since strain CH21 is less fully inbred to a Colonia genetic background than the other mt3 strains (ADLERand HOLT1974b), the difference among the mt3 strains may have a genetic basis. Another possible cause of the variations within a single mating type is differences in the frequency and type of aneuploid amoebae present in cultures (ADLERand HOLT1974a). Progeny analysis: A total of 103 plasmodia formed by heterothallic amoebae

TABLE 3

Plasmodium formntion by progeny of CH396(mt2) X CH329(nt4)*

Number of cultures producing plasmodia Experiment I Experiment I1 Strain 18" 06' 18" mt4 progeny, all nine 0 0 mf4parent, CH329 0 0 mt2 progeny, serial number: 1 1 0 4 2 3 0 1 3 4 1 3 5 2 0 6 7 3 0 3 8 6 0 9 9 3 1 8 10 0 0 6 18 2 0 2 19 7 0 4 mt2 parent, CH396 5 0 4

* The experiments were conducted as in Table 2 except that only ten amoeba1 cultures were prepared for each strain at a given temperature; "-" means not measured. DIFFERENTIATION MUTANTS IN PHYSARUM 40 7

TABLE 4

Classification of plasmodia formed by heierothallic amoebae*

Temperature Number of plasmodia Parent at which (1) (2) (3) Mixed Mating Strain plasmodia Total Non- GAD ALC GADand type number formed analyzed mutant mutant mutant nonmutant mtl CH269 18" 4 4 0 0 0 26" 9 9 0 0 0 30" 3 0 1 2 0 mi2 CH274 18" 12 11 0 0 1 CH394 18" 21 21 0 0 0 26" 6 2 2 0 2 30" 1 1 0 0 0 CH396 26 1 0 1 0 0 mt3 CH2.1 26 11 9 1 1 0 CH541- 26" 4 4 0 0 0 CHI90 26" 2 2 0 0 0 CH322 18" 7 6 0 0 1 26" 13 11 2 0 0 30" 2 1 0 1 0 mt4 CH329 18" 1 0 1 0 0 CH326 18" 1 1 0 0 0 26" 1 0 0 1 0 30" 2 0 0 2 0 CH326-t 18" 1 0 1 0 0 26 O 1 1 0 0 0 Totals 103 83 9 7 4

* Plasmodia formed by individual heterothallic strains were classified according to their progeny. The classes are described in the text. + Mutagenized with ethyl methane sulfonate.

in the above and other experiments were screened for the presence of mutations (Table 4). Spores were obtained from each of the plasmodia, the spores were germinated, and the resulting vegetative cells examined. Most (83) of the plasmodia gave rise only to heterothallic amoebae with the same mating specificity as the parent amoebae and were therefore classified as nonmutant. The plasmodia that were the result of mutation fell into three classes. Spores of the first class released CPF* amoebae, that is, amoebae that form plasmodia in clones at a much higher frequency than heterothallic amoebae. These mutants are referred to as GAD* (greater asexual differentiation) mutants (Table 4). Spores of the second class released mainly plasmodia instead of amoebae. Since it was found that the sequence plasmodium-spore-plasmodium could be obtained repeatedly, these mutants are referred to as ALC* (amoebaless life cycle) mutants (ADLER,

* Abbreviations. As in previous publications, amoebae that readily form plasmodia in clones are described as CPF (clonal plasmodium formmg) amoebae. Note that mth, mtx gad, mtx/mty, and mtx alc amoebae are all CPF, where mtx and mty represent two different heterothallic alleles of mt. The use of ALC and GAD to designate mutant class is described above ALC is also used as follows an ALC spore is one that gives rise to a plasmodium dlrectly and an ALC plasmodium is one formed by an ALC spore. 408 P. N. ADLER AND C. E. HOLT DAVIDOWand HOLT1975). Spores of the third class of mutant plasmodia released CPF ammbae and heterothallic amoebae of the parental mating type in approximately equal numbers. Such plasmodia could have arisen by several mechanisms, among them: (a) two piasmodia, one due to a mutation and the other not, fused to make a heterokaryon; (b) the amoeba that differentiated into the plasmodium contained the DNA sequences for both the wild-type and mutant gena (the amoeba could have been a heterozygote in G2 or could have had a region of DNA heteroduplex), or (c) the plasmodium was the result of a cross bctween a mutant amoeba and a wild-type heterothallic amoeba. It is shown below :hat the CPF amoebae from these mutants do not cross with heterothallic amoebae of the mating type from which they were derived. Thus, the class (3) mutants seem to have arisen either by mechanism (a) or mechanism (b) , and on this basis thvzy have been treated as GAD mutants.

Characterization of GAD mutants Kinetics of plasmodium formalion: Initial characterization of GAD mutants was carried out by setting up a series of dPRM agar cultures of the GAD mutant amoebae, incubating them at 26" or 30", and recording the day when plasmodia became evident to the naked eye. Typical data are shown in Figure 2. From data plott.?d in this fashion, the mean time (T(50)) for plasmodium formation was estimated. The values for T(50) €or twelve of the mutants, a mth strain and a mt3/mt4 strain are given in Table 5. Differences between a strain at 26" and 30" were reproducible. Thus one can conclude that CH489 (mt3 gad-10) forms plasmodia more slowly at 30" than at 26". Some quantitative variation (Iyz day) was seen between experiments; therefore: one cannot conclude that CH489 pro-

Time (days) FIGURE2.-Time of plasmodium formation by a GAD mutant. Approximately IO4 CH405 (mt2gad-I) amoebae were inoculated onto each of 20 dPRM agar plates. Ten plates were incubated at 26" and ten at 30". Each plate was examined daily for the presence of macroscopic plasmodia. The value of T(50), which is the time for 50% of the cultures to form plasmodia, was determined by interpolation as shown. 0,26"; A, 30". DIFFERENTIATION MUTANTS IN PHYSARUM 409

TABLE 5 Time of plasmodium formation by GAD mutants*

T(50) in days Mutant isolation Strain Genotype+ 2G0 30" Difference temperature CH405 mt2 gad-l 4.5 8.1 3.6 96" CH40.3 mt2 gad2 3.5 6.5 3.0 26" CH404 mt2 gad-3 5.4 8.4 3.0 18" CH480 mt2 gad-4 4.5 9 4.5 26 CH479 mt3 gad-5 2.8 4.2 1.4 26" CH484 mt4 gad-& 6.4 7.8 1.4 18" CH486 mt2 gad4 3.4 3.5 0.1 26" CH487 mt2 gad-9 2.6 2.6 0 26" CH489 mt3 gad-IO 2.7 3.3 Q.6 18" CH496 mt3 gad-ll 4.5 >28 >23.5 26" CH478 mtl gcrd-12 13.7 8.5 -5.2 30" CH526 mt4 gad-13 3.5 7.2 3.7 18" CHI mth 4.4 7.5 3.1 - CH586 mD/mt4 3.6 3.7 0.1

* The twelve GAD mutant strains shown and two control strains were analyzed as described in Figure 2. Strain CH485 was not analyzed by this method. j- The genotype of each mutant includes the assignment of a single gad allele, whether or not the presence of such an allele has been demonstrated in crosses. duces plasmodia more rapidly than CH486 (mt2gud-8) , but can conclude that CH589 produces plasmodia more rapidly than CHI at 30". The results in Table 5 reveal impressive phenotypic variation among the mutants. The phenotypes of several strains are worth noting. Strain CH478 (mtl gad-12) is the only strain that formed plasmodia earlier at 30" and 26'. Nevertheless, it formed plasmodia later at 30" than any other strain at 26". For two strains, CH486 (mt2gad-8) and CH487 (mt2 gad-9) ,both of which produced plasmodia earlier at 26" than CHI (mth)at 26", the time for the appearance of plasmodia in cultures was the same at 26O and 30". Strain CH496 (mt3 gad-ll) showed a remarkable temperature sensitivity. ,4t 26" it gave a T(50) of 4.5 days, while at 30°, even after 4 weeks, plasmodia had fornied iii only two of the ten replicate cultures. In a previous study, the kinetics of plasmodium production by mth amoebae was measured by a method that involves washing the cells off a plate of growing amoebae and assaying the numbers of viable amoebae and plasmodia in the resulting suspension (YOUNGMANet al. 1977). The same method was applied to the study of plasmodium production in GAD mutants. Typical results are shown in Figure 3. Several f-eatures of the data deserve mention. In all cases, the amoebae grew at a normal or near normal rate (doubling time 7.5-8 hr) . As in the earlier experiments with mth amoebae, there was a lag time (tl) before plasmodium production commenced. Strains judged temperature sensitive by the method used initially (Table 5) were clearly temperature sensitive by the present method. Strains varied both in tl and in the maximum number of plasmodia produced, and the value of T(50) at a given temperature increased with tl and decreased 41 0 P. N. ADLER AND C. E. HOLT

Y) - 1 I 'I'd' I'I I'I al - n " io7- CH487 -- CH478 A' 0 0 0- L 0 i

I I -1 Io= t

lo k oL 0 40 00 120 160 200 Time (hours) FIGURE3.-Kinetic experiments with GAD mutants. Replicate cultures of the strains shown were incubated at 26" or 30". Individual plates were harvested at the times shown and the number of amoehae (open symbols) and plasmodia (closed symbols) determined by biological assay. 0,26"; A, 30". with the maximum number of plasmodia. For two of the mutants, CH478 (mtl gad-12) and CH484 (mt4 gad-6), differentiation at 30" began only after the amoeba1 density had reached a plateau. The data for CH496 (mt3 gad-11 ) show more than IO5 plasmodia at 26" at a time when none were detectable at 30" (Figure 3). Mating specificity of GAD mutants: The mating specificity of those GAD mutants that formed plasmodia relatively slowly at 26" or 30" was tested by comparing the mean time for the appearance of macroscopic plasmodia in cultures of mutant amoebae only and cultures comprised of equal numbers of mutant and wild-type amoebae of different mating types. Five mutant strains were tested in this way (Table 6), and all gave the same result. Cultures comprised of mutan! DIFFERENTIATION MUTANTS IN PHYSARUM 41 1 TABLE 6 Mating specificity of GAD mutants*

Temperature Strain(s) of test CH478 (mtl gad-12) 26" 13.7 CH478 + CH269 (mt2) 26" 14.3 CH478 + CH326 (mt4) 26 7.4 CH496 (mt3 gad-21) 301" >21t CH496 + CH344 (mU) 30" >21f CH496 + CH326 (m24) 30" 2.7 CH484 (mt4 gad-6) 30" 7.4 CH484 + CH329 (mt4) 30" 7.5 CH484 + CH343 (mt3) 30" 2.5 CH480 (mt2gad4 30" 8.1 CH480 + CH393 (77222) 30" 7.8 CH480 + CH329 (mt4) 30" 3.5 CH404 (mt2 gad-3) 30" 8.0 CH4W + CH393 (mtl) 30" 8.4 CH404 + CH329 (mt4) 30" 3.0

*All cultures were inoculated with 2x104 amoebae. Cultures were scored and the T(50) determined as in Figure 2. t After 21 days only 3 cultures of CH496 amoebae and one containing a mixture of CH496 and CH344 amoebae had produced plasmodia.

amoebae only and cultures comprised of mutant amoebae and wild-type amoebae of the same mating type as the parent of the mutant yielded macroqcopic plasmodia at the same time. The plasmodia that Pormed in these mixed cultures had the plasmodial fusion typ2 (see MATERIALSAND METHODS)expected of a plasmodium. formed asexually by the mutant amoebae, and not the fusion type expected if crossing had occurred. Cultures comprised of mutant amoebae and wild- type amoebae of a mating type that was difiercnt from that of the mutant parent produced plasmodia much earlier than cultures containing only mutant amoebae. In this case, the plasmodia had the fusion type expected if crossing had occurred. Thus. the mutant amoebae appear to have retained the mating specificity of their parental amoebae. Wc have also tested sewral combinations of GAD mutants derived from mt2 strains to see whether or not these strains, which now differ by two mutations, could cross. No speedup in the time of macroscopic plasmodium appearance was seen by co-culturing various combinations of mt2-derived mutants. When several of the mt2-derived mutants were co-cultured with CH496 (mt3 gad-ll), a speedup in the time of macroscopic plasmodium appearance was seen, and the plasmodia had the plasmodial fusion type expected from a cross. Genetics of gad mutations: Eight of the G4D mutants have been crossed to heterothallic strains and the progeny analyzed. When possible [for all crosses except CH484 (mt4 gad-&) x CH438 (mt3)], progeny were tested for recom- bination between mt and genes controlling plasmodial fusion type (fusA or fusC) . Recombinants between mt and fusA or fusC were found in all cases. which 412 P. N. ADLER AND C. E. HOLT TABLE 7 Genetic analysis of GAD mutants

Progeny Cross CPF' mti mt2 mt3 mt4 CH405 (mt2gad-1) X CH351 (mt3) 7- - 13 - CH403 (mi2gad-2) x CH347 (mt3) 12 - - 8- CH4M (mt2gad-3) X CH348 (mt3) 13 - - 2- CH480 (mt2gad-4) X CH348 (mt3) 10 - - 8- CH479 (mt3ga.d-5) x CH269 (ma) 11 9- - - CH484 (mt4gad-6) X CH348 (mt3) 11 - - 9- CH496 (mt3gad-11) X CH326 (mt4) 11 - - - 10 CH478 (mt1gad-12) X CH273 (mt4) 8+ 3 - - 6

* The progeny that produce plasmodia in clones are presumed to carry a gad mutation, except for some of the CPF progeny of CH404 X CH348. See text. t2 mt1 gad-12 and 6 mt4 gad-12.

shows that the plasmodia formed by the GAD amoebae and heterothallic amoebae were indeed crosses. Six of these crosses yielded, in approximately equal numbers, CPF progeny and heterothallic progeny of the same mating type as the heterothallic parent used in the cross (Table 7). Thus, these six mutants contain a mutation (gad) linked to mt. Progeny of the cross CH404 (mt2 gad-3) x CH348 (mt3) were all either CPF or mt3. The two types were not found in equal numbers, however, as most (13/15) of the progeny were CPF. Most of the CPF progeny formed plasmodia more rapidly than the CH404 parental amoebae and are presumed to be heterozygous for mt. Amoebae of CH404 form fuzzy plaques; thus, the strain is likely to be aneuploid (ADLERand H~LT1974a) ,which could explain the unusual results. The results are consistent with gad-3 being linked to mt, but the subject deserves further investigation. Four types of progeny were obtained from the cross CH478 (mtl gad-12) X CH273 (mt4).Nine heterothallic progeny were obtained, thrw of which were mtl and six mt4. Eight progeny exhibited the CH478 phenotype with regard to the time of macroscopic plasmodium formation at 26" and 30". Six of these had mt4 mating specificity and two had mtl specificity. Thus, it appears that gad-12 is unlinked to mt and does not require a specific mt allele for expression. The fusA and fusC phenotypes of the 8 CPF progeny were determined. Both fusA and fusC segregated 1: 1 and independently o€gud-12. A mt4 gad-12 strain was crossed to CH9 (mth aptAl) (aptAl prevents asexual plasmodium formation by mth amoebae; WHEALS1973). Based on the phenotypes of the progeny, it appears that mt, gad-12 and uptA are all unlinked, that uptA1 is epistatic to gad-12, and that amoebae with the genotype mth gad-12 form plasmodia more rapidly than mth strains at both 26" and 30". Most of the GAD mutants that wme not crossed form plasmodia so rapidly that only asexually formed plasmodia have been obtained from attempted crosses. It should be possible to cross them after first selecting variants that form plasmodia more slowly (DAVIDOW,personal communication). DIFFERENTIATION MUTANTS IN PHYSARUM 413

Frequency of GAD mutants: The frequency of GAD mutants among heterothallic amoebae has been estimated. The average frequencies of formation of plasmodia in cultures of heterothallic amoebae (Table 2) werz multiplied by the appropriate ratios of GAD mutant plasmodia to total plasmodia analyzed (Table 4). The averag.2 of' these products is about 1 %, that is, a GAD mutant plasmodium appears about once in every 100 cultures of heterothallic amoebae. The number of amoebae in cultures when they produce such a plasmodium is about 5 X IO7 (Figure 1). Thus, the frequency of recovered GAD mutants is approximately 2 x In order to be able to convert this frequency to the actual frequency of GAD mutants, the efficiency of the selective system was estimated in reconstruction experiments with two mutants, CH486 (mt2 gad-8), which dis- plays later plasmodium production (Table 5). In each case, a large number of nonmutant mt2 amoebae were mixed with a small number of mutant amoebae, and the efficiency of recovering the mutants as plasmodia was measured. With CH486, four dPRM agar plates inoculated with 12 CH486 amoebae and 1.2 X lo7 CH364 (mt2) amoebae each showed at least five loci of plasmodium formation after one week of incubation at 26". Similar results were obtained at 30". Thus, the efficiency of recovering this mutant is at least 5/12 and is probably close to 100%. With CH404, three dRPM agar plates inoculated with 10 CH404 amoebae and 1.2 X lo7CH634 amoebae showed only one plasmodium on one of the three plates after two weeks incubation at 26". This result gives a recovery frequency of 1 in 57 at 26". At 30", plates with 190 CH404 amoebae and 6 X lo5 CH634 amoebae gave no plasmodia after 2 weeks. Although there are uncertainties in all of these measurements, one can conclude that the frequency of extreme mutants such as CH486 is very low, of order to and the frequency of less extreme mutants such as CH404 is considerably higher, perhaps of order The fact that most of the mutants are of the more extreme type reflects their high efficiency of recovery. All plasmodia that formed in the above experiments were tested to determine their fusion type. Each of the plasmodia had the fusion type expected of a plasmodium that had been formed asexually by the GAD mutant. Thus, even under these extreme conditions, the GAD mutants still show the mating specificity of their parental strains. Characterization of ALC mutants General features: Seven mutants capable of going through an amoeba-less life cycle were isolated (Table 4). Microscopic observations demonatrating that plasmodia issue directly from spores were published earlier (ADLER:DAVIDOW and HOLT1975). The spore germination frequencies for the mutants werre low, run- ning from medians of 4 x €or mutant CH5004 to lo-* for mutant CH5006. Each of the ALC mutants was carried through the sequmce plasmodium-spore- plasmodium at least twice. In all cases, the ALC characteristic was retained in successive generations and thus appeared to have a genetic basis. This inference was confirmed in detail for mutant CH5001. Mutant CH5001: Spores of this mutant, like those of four other ALC mutants, gave rise to amoebae as well as plasmodia. The overall spore germination fre- 414 P. N. ADLER AND C. E. HOLT quency was about 2 X and the fraction of total viable spores giving rise to amoebal plaques varied from l-lO%, depending on the experiment. Amoebae were picked from 27 such plaques, recloned, and characterized with regard to amoebal growth and asexual plasmodium formation at 26" and 30". The results revealed two classes of amoebal progeny from CH5001 spores. For 25 of the CH5001 progeny, growth was reduced and asexual plasmodium formation increased at 30" relative to 26". Two of these "class A" progeny were studied by the kinetics method described above. The results for one of the progeny are shown in Figure 4A; the results for the other were virtually identical. The amoebae grew somewhat more slowly (9.8 hr doubling time) than wild-type amoebae (7-8 hr) at 26". At 30", the mutant amoebae grew slowly (17 hr) for about 4 doublings and then ceased growing. Nearly 100-foldmore plasmodia were produced at 30" relative to 26" and plasmodium production by these amoebae. unlike that by gad or mth amoebae, seemed to proceed without a time lag. Plas- modia produced by class A amoebae were sporulated, and the spores germinated. Plasmodia and amoebae were released from the spores in the same approximate frequencies as from spores of the original mutant plasmodium. Thus, class A amoebae seem to contain the mutation present in the original mutant plasmodium. The phmotype of the two class B amoebal strains is described in Figure 4B.

lo7

2 I06 -P Q 5 L IO 0) Q - lo4 a,

r lo3 0 IO2 n: E 3 10

Time (hours) FIGURE4.-Kinetic experiments with amoebae from CH5001 spores. A, Class A amoebae; B, cIass B amoebae. Symbols as defined in Figure 3. DIFFERENTIATION MUTANTS IN PHYSARUM 415 Amoebal growth was slower at 30" than at 26" but, unlike the growth of class A amoebae, continued indefinitely at 30". As before, plasmodium production was enhanced by the temperature increase. It appears that there was some time lag before the onset of plasmodium production, but the small numbers of plasmodia counted make this conclusion uncertain. The basis for thc difference between the class A and class B amoebae is not known. In order to study the mutation in CH5001, class A amoebae were crossed with heterothallic amoebae (Table 8). The spores of the crossed plasmodia were germinated and found to give rise to amoebae only. Amoeba1 progeny strains were established and analyzed. The results suggest that the mutation, designated alc-1, is unlinked to mt and is expressed with mtl and mt2 as well as mt3 (Table 8). Progeny with the proposed genotypes mtl alc-l, mt2 alc-1, and mt3 alc-l were allowed to form plasmodia asexually, the plasmodia were sporulated and the spores analyzed. A total of 23 such progeny from the above and other crosses were examined in this fashion. All 23 gave ALC spores as predicted by the proposed genotypes. The fact that all the progeny of mt3 alc-l by mtl or mt2 are amoebae implies that either the lesion in CH5001 is recessive with regard to its effect on the cells released on spore germination, or the lesion is not expressed in a plasmodium that is heterozygous for mt. To test the lattm possibility, a mtl alc-l progeny and ;1 mt2 alc-2 progeny were each crossed to a mt3 alc-l progeny. The resulting plasmodia had the plasmodial fusion types expected of crosses. Spores obtained from the plasmodia were germinated and some gave rise to plasmodia. Thus, the lesion can b2 expressed in a plasmodium heterozygous for mt, and ale-l may be described as recessive. Spores from these crossed plasmodia released 2-4 times as many amoebae as plasmodia; thus, it appears that the expression of alc-l is quan-

TABLE 8 Crosses of amoebae from CH5001 with heterothallic amoebae*

Phenotype+ Clonal Mating Amoebal plasmodium Number Genotype type growth formation in class CH587(mt3 alc-1) x CH394(mt2) mt3 no,rmal - 5 mt3 + mt3 TS + 4 mt3 alc-1 mt2 normal - 5 mt2 + mt2 TS + 5 mt2 alc-l CH588(mt3 alc-1) x CH269(mtl) mt3 normal - 7 mt3 + mt3 TS + 4 mt3 alc-l mtl normal - 3 mtl + mtl TS 3- 4 mtl alc-l

* Two amoeba1 strains (CH587 and CH588) derived from spores of CHSO(E1 were crossed as shown. Amoebal progeny strains were established and analyzed. -f TS, grows at 26" but not 30" on LIA; normal, grows at both 26" and 30" on LIA; +, forms plasmodia in clones on dPRM agar; -, does not form plasmodia in clones. 41 6 P. N. ADLER AND C. E. HOLT titatively different in diploid plasmodia heterozygous for mt and haploid plasmodia. The kinetics of plasmodium formation was measured for amoebae of one mtl alc-1 strain and one mt2 alc-I strain. The results were similar to those for mt3 alc-1 (Figure 4A), except that mt2 aZc-I amoebae increased 100-fold rather than 20-fold at 30", and the doubling time for mtl alc-1 amoebae at 26" was 12.2 hr rather than 9.8 hr. This apparent dependence of alc expression on mt is intrigu- ing, but more progeny would have to be characterized before a systematic difference could be established. Other ALC mutants: Three of the ALC mutants could not be analyzed genetically, two (CH5003 and CH5007) because their spores did not yield any amoebae, and one (CH5006) because the only amoebae released appeared by both phenotypic and genotypic criteria to be revertants. Mutants CH5004 and CH5005 appear to be similar to CH5001. The majority of amoebae from these were like class A. amoebae (Table S), that is, they displayed reduced growth and enhanced asexual plasmodium production at 30". The mutants have not been crossed. The remaining mutant, CH5002, provided curious results. The mutant was isolated from a culture of mtl amoebae incubated at 30". Amoebae germinated from CH5002 spores only rarely, usually at a frequency several orders of mag- nitude lower than the plasmodial germination frequency. All (a total of 5) of the amwbae isolated from CH5OO2 are phenotypically revertants. that is, they are wild type with respect to mating specificity, frequency of asexual plasmodium formation, plaque size, and growth rate (for the one strain where this was testedj . One amoeba1 progeny of CH5002, CH5004, was crossed to CH343 (mt3) ancl spores were obtained from the resulting plasmodium. Platings of these spores yielded mtl amoebae, mt3 amoebae and ALC plasmodia with an amoebal-plasmodial plaque ratio of about 10: 1. This result is what might be expected if CH594 contained an unlinked suppressor and if the germination frequency of plasmodia- producing spores were inherently lower than that of amoeba-producing spores. The plasmodia fusion types of four of the six ALC plasmodia differed from that of the parental plasmodium. Thus, meiosis occurred in the formation of spores that gave rise to the ALC plasmodia. Fire of the 6 ALC plasmodia were induced to sporulate and the spores plated. Surprisingly, these yielded mtl amoebae, mt3 amoebae, and ALC plasmodia. Thus there was no segregation for mating type in the formation of the CH594 x CH343 ALC spores. Spores were obtained from 5 of these ALC plasmodia and thes also yielded mtl amoebae, mt3 amoebae, and ALC plasmodia. Although the data suggest that the original mutant contains B mutation unlinked to mt and that amoebae from the mutant contain an unlinked supprr2ssoras well, the presence of both mt alleles in progeny of ALC plasmodia remains unexplained. Since segregation for plasmodial fusion genes were ob- served, it does not seem reasonable to explain the lack of segregation for mt by lack of meiosis. Nuclear autonomy of alc mutations: Heterokaryons consisting of equal parts of an ALC mutant plasmodium and a plasmodium lormed asexually from either a mth strain or a strain carrying a gad mutation were made and induced DIFFERENTIATION MUTANTS IN PHYSARUM 41 7 to sporulate. Spores from these heterokaryons were plated and the plates examined for the presence of CPF amoebal clones and ALC plasmodia. For all seven of the mutants, spores from at least one heterokaryon gave rise to both CPF amwbae and ALC plasmodia. Thus, the alc mutations appear to be nuclearly autonomous in a heterokaryon. This could be due to the €unction of the alc gene being restricted to its own nucleus or to the need for aZc gene function after spore cleavage. Some caution must be noted about this conclusion for two reasons. One is that the results are based only on an examination of the spore plating and not on genotypic testing. Thus, it is possible that the mutations are not completely nuclearly auonomous (e.g., some of the ALC plasmodia could have a wild-type genotype). Second, the ratio of amoebal to plasmodial pIaques arising from platings of heterokaryon spores was only occasionally close to 1. In many spore platings there were 5-100 times more amoebal than plasmodial plaques and on two occasions this ratio was reversed. This could be due to selective spore germination, but other possibilities exist.

DISCUSSION The mating-type locus, which was originally discovered because it controls mating specificity, is now seen to control asexual plasmodium formation as well. Earlier studies showed a striking difference in selfing frequency between mth and heterothallic strains, and the present study reveals marked differences among heterothallic strains of different mating type (Tables 2 and 3). Since nothing is known about the structure of the mating-type locus in P. polycephalum, little can be said about the number of cistrons involved in the control of these two functions. The mating specificity of four mutants that carry gad mutations linked to mt was measured and found to be identical to the mating specificity of the parent strains (Tables 6 and 7). This suggests that more than one cistron is involved. The recent finding that gad-ll is 12 map units from mt supports this suggestion (SHINNICKand HOLT1977). Four of the mutants, CH403, CH404, CH405 and CH526, €orm plasmodia in a similar fashion to mth strains (Table 5) , and we cannot ruIe out the possibility that they are identical to milz strains. The other mutants differ from mth strains either in the time of plasmodium formation (Table 5), mating specificity (Table 6), or both. The GAD mutants display a wide variation in their rate and temperature sensitivity of asexual plasmodium formation (Table 5; Figure 3). This variation, the relatively low frequency with which these mutants arise, and the fact that almost all of the gad mutations studied map at or near mt suggest that the mutations cause quantitative alterations in a function. Plasmodia germinate from spores produced by ALC mutants, and thus by definition the alc mutations affect the establishment of the amoebal state. Since the mutants were originally obtained from plasmodia formed asexually, the mutations must also affect the maintenance of the amoebal state. For the mutation alc-I, the maintenance of the amoebal state is indeed affected (Figure 4). Similar data for mutant strains CH5004 and CH5005 suggest that they too contain mutations that affect the maintenance of the amoebal state. The involvement 418 P. N. ADLER AND C. E. HOLT of alc mutations in both the establishment and maintenance of the amo--.balstate argues in favor of the mutations being in a gene (or genes) that is of central importance in the control of the differentiated state. Heterokaryons of aZc+ and aZc-1 nuclei differ strikingly from aZc+/aZc-1 heterozygotes: the former produce both normal and ALC spores. whereas the latter produce only normal spores. One possible explanation of the heterokaryon result is that the alc gene product is restricted to the nucleus in which it is produced. This explanation is compatible with the heterozygote result if one assumes that the product is made prior to the separation of alc+ and alc-l into separatc nuclei during meiosis. Another possible explanation of the heterokaryon result is that the alc gene does not function until plasmodial nuclei are separated into individual cells during sporulation. This explanation is also compatible with the heterozygote result, since meiosis follows the separation (cleavage) of nuclei (ALDRICH1974). Here again, one may assume that the gene functions prior to meiosis in order to explain the fact that ab2 spores from a heterozygote germinate normally. (Alternatively, one could assume that the gene functions after meiosis and that its product can diffuse among the products of a cingle meiosis only; but this seems unlikely.) In summary. no matter which explanation of the heterokaryon result is used, one comes to the conclusion that aZc makes its product prior to meiosis. The mutations isolat.cd in this study were spontaneous. The one attempt (see Table 2) to induce mutants by mutagenesis was not effective. Thic result must be viewed with skepticism, as only about lo6cells were mutagenized and no inde- pendent check of the effectiveness of mutagenesis was done in that experiment. Indeed, GORMAN(pcrsonal communication) has induced mutations that appear to be similar to the gad mutations described here. A small number of plasmodia formed in clones of heterothallic amoebae were sickly and could not be induced to sporulate and, in some cases, to grow to a size of more than a few cm2. The frequency of such plasmodia was increased under the conditions where fewer plasmodia appeared, and thus the plasmodia may repre2ent additional mutants. It should be possible to rescue such sickly plasmodia by fusing them with a wild-type plasmodium to form a heterokaryon and then inducing the heterokaryon to sporulate. This has not been done, however. Some bias was inherent in the procedure by which the asexually formed plasmodia were screened, as they were test.2d at 26" regardless of the temperature at which they formed. Thus, it is possible that some plasmodia were misclassified as nonmutant because of being tested at the wrong temperature. We believe, however, that misclassifications, if any, were likely to be few. For example, iI the 20 mutants found had been tested at 30" instead of 26", only one strain, CH496, would have been misclassified. Genes known to control plasmodium formation in P. polycephulum now include mt, aptA, npfA, the two apt and npf complementation groups tightly linked to mt, gad-12, alc-1, and the gad mutants linked to mt. Whether or not gad-12 and alc-1 are linked to one another and whether or not alc-l is linked to aptA or npfA, are not known. It does appear that gad-12 is unlinked to apiA. The DIFFERENTIATION MUTANTS IN PHYSARUM 419 other gad mutations studied genetically were linked to mt. At least one, gad-ll, is not as tightly link.Zd to mt as the mt-linked npf and apt mutations that have been studied (SHINNICKand HOLT1977). Although a great deal of additional work is needed, it is clear that it will be possible to build a detailed picture of genes that control differentiation in this organism. This work was supported by National Science Foundation Grant number BMS70-00763. P.A. was supported by Public Health Service Training Grant number TO-1-GM00710 to the Depart- men? of Biology. We thank DR. CHRISTINETRUITT and DR. ROGERANDERSON for helpful com- ments on the manuscript.

LITERATURE CITED

ADLER,P. N., 1975 Control of the differentiated state in Physcr-um polycephalum. PhBA Thesis, Massachusetts Institute of Technology. ADLEE,1’. N. and C. E. HOLT,1974a Change in properties of Physarum polycephalum amoebae during extended culture. J. Bacteriol. 120: 532-533. -, 1974b. Genetic analysis in the Colonia strain of Physarum polycephalum: Heterathallic strains that mate with and are partially isogenic to the Colonia strain. Genetics 78: 1051-1969. --, 1975 Matingtype and the differentiated state in Physarum polycephalum. Devel. Biol. 43 : 240-253. ADLER,P. N., L. S. DAVIDOWand C. E. HOLT,1975 Life cycle variants of Physarium polycephalum that lack the amoeba stage. Science 190: 65-67. ALDRICH,H. C., 1967 The ultrastructure of meiosis in three species of Physarum. Mycologia 59: 127-148. -, 1974 Spore cleavage and the development of wall ornamentation in two myxomycetes. Proc. Iowa Acad. Sci. 81 : 28-35. ANDERSON,R. W., D. J. COOKEand J. DEE, 3976 Apogamic development of plasmodia in the Myxomycete Physarum polycephalum: A cinematographic analysis: Protoplasma 89: 29-40. ANDERSON,R. W. and J. DEE, 1977 Isolation and analysis oE amoebal-plasmodial lransition mutants in the Myxomycete Physarum polycephalum. Genet. Res. 29: 21-34. COLLINS,0. R., 1975 Mating types in five isolates of Physarum polycephalum. Mycologia 67: 98-107. COLLINS,0. R. and H.-C. TANG,1977 New mating types in Physarum polycephalum. Mycologia 69: 421-423. COORE,D. J. and J. DEE, 1974 Plasmodium formation without change in nuclear DNA content in Physarum polycephalum. Genet. Res. 23: 307-318. -, 1975 Methods for the isolation and analysis of plasmodial mutants in Physarum polycephalum. Genet. Res. 24: 175-187. DAVIDOW,L. S. and C. E. HOLT,1977 Mutants with decreased differentiation to plasmodia in Physarum polycephalum. Molec. Gen. Genet. (In press.) DEE,J., 1966 Multiple alleles and other factors affecting plasmodial formation in the true slime mold Physarum polycephalurn Schw. J. Protozaol. 13 : 610-616. GRAY,W. D. and C. J. ALEXOPOULOS,1968 Biology of the Myromycctes. Ronald Press, ‘New York. HAUGLI,F. B., 1971 Mutagenesis, selection and genetic analysis in Physarum polycephalum. Ph.D. Thesis, Univ. of Vl’isconsin. LAANE,M. M. and F. B. HAUGLI,1976 Nuclear behavior during meiosis in the Myxomycete Physarum polycephalum. Norwegian J. of Botany 23: 7-21. LAFFLER,T. G. and W. F. DOVE,1977 The viability of Physarum polycephalum spores and thg ploidy of plasmodial nuclei. J. Bacteriol. 131: 473476. 420 P. N. ADLER AND C. E. HOLT

SHINNICK,T. M. and C. E. HOLT,1977 Analysis of an mt-linked mutation affecting asexual plasmodium formation in Physarum. J. Bacteriol. 131 : 247-250. WHEALS,A. E., 1970 A homothallic strain of the Myxomycete Physarum polycephalum. Ge- netics 66: 623-633. -, 1973 Development mutants in a homothallic strain of Phy- sarum polycephalum. Genet. Res. 21 : 79-96. YOUNGMAN,P. J., P. N. ADLER,T. M. SHINNICKand C. E. HOLT,1977 An extracellular inducer of asexual plasmodium formation in Physarum polycephnlum. Proc. Nat. Acad. Sci. U.S. 74: 1120-1124. Corresponding editor: S. L. ALLEX