GENETIC ANALYSIS OF MATING TYPE DIFFERENTIATION IN

PARAMECIUM TETRAURELIA. 11. ROLE OF THE MICRONUCLEI I N MATING-TYPE DETERMINATION Y. BRYGOOf, T. M. SONNEBORNt, A. M. KFLLERt, R. V. DIPPELLS M. V. SCHNELLERS

AND

+ Centre de Gdndtique MoEculaire du C.N.R.S. 91190 Gif-sur-Yvette, France $. Department of Biology, Indiana University, Bloomington, Indiana 47401 Manuscript received August 7, 1979 Revised copy received November 28,1979 ABSTRACT

The two complementary mating types, 0 and E, of Paramecium tetrauretia are normally inherited cytoplasmically. This property has generally been interpreted to indicate the presence of cytoplasmic factors that determine macronuclear differentiation towards 0 or E . In these macronuclear-cytoplasmic interactions, the micronuclei were held to be unbiased and the determination to be established in the course of macronuclear development. In order to ascertain whether the micronuclei were actually neutral, amicronucleate clones were needed and a method to produce them was developed. In crosses between amicronucleate clones and normal micronucleate clones, we have observed regular deviations from cytoplasmic inheritance: the commonest deviation is that most 0 amicronucleate cells become E when they receive a micronucleus from an E partner. The data can be interpreted by assuming that the micronuclei are predetermined and that the apparent “cytoplasmic” inheritance of the two mating types is due, in E cells, to E-determining factors present in the cytoplasm and in the nucleus; and, in 0 cells, to 0-determining factors present only or mainly in the nucleus.

T has long been known that mating type in Paramecium tetraurelia is cytoplasmically inherited (for review, see SONNEBORN 1977). Although each cell is genetically competent to express both complementary mating types, called odd (0) or even ( E ) , it is, in fact, determined f o r either 0 or E . The 0 or E phenotype is clonally inherited via the parental cytoplasm. Concerning the basis of this cytoplasmic inheritance, it is generally agreed that ( 1 ) the determination towards 0 or E is a property of the macronucleus, (2) the determination takes place at each sexual process (conjugation or autogamy) during the rapidly repeated rounds of DNA synthesis in the newly developing macronucleus, and (3) the “cytoplasmic factors” that control macronuclear determination are direct or indirect gene products emanating from the old parental 0 or E macronucleus. In these macronuclear-cytoplasmic interactions, the micronuclei ordinarily seem to be unbiased: neither the gametic nuclei nor the diploid fertilized nucleus Genetics 94: 951-959 April, 1980.

952

Y. BRYGOO

et al.

(from which macronuclei arise) seems to be determined. However, some previous results (SONNEBORN 1954; BRYGOO1977) led to the hypothesis that gametic nuclei might transmit some predetermination for mating type through conjugation. SONNEBORN(1954) briefly stated (without detailed supporting data) that, in crosses between amicronucleate cells and normal diploids, the off spring of the amicronucleate parent tended to escape the rule of cytoplasmic inheritance and to acquire the mating type of the parent from which it received a gametic nucleus. In order to verify whether the gametic nuclei and the micronuclei that develop from them were predetermined, SONNEBORN, DIPPELL and SCHNELLER’S unpublished experiments (cited by SONNEBORN 1954) were repeated and the new results compared with their data. As will become apparent, the two sets of data are in remarkable agreement. We propose an hypothesis as to how such predetermination of gametic nuclei and micronuclei can lead to apparent “cytoplasmic” inheritance of mating type in normal diploid x diploid crosses. MATERIALS A N D METHODS

Strains and culture conditions: The strains used by SONNEBORN,DIPPELL and SCHNELLER (cited in SONNERORN 1954) were standard stock 51 and its derivatives grown in a baked lettuce infusion, as described by SONNEBORN(1950). All the strains used by BRYGOOand KELLERderived from stock d4-2, also a derivative of stock 51 of Paramecium tetraurelia. In the course of the genetic analysis, two mutant strains were used: the temperature-sensitive mutant t s l l l of BEISSON and ROSSIGNOL(1969) and the mutant hr of AUSTIN(SONNEBORN1975), which has very high and long-lasting sexual reactivity. The latter strain was, therefore, also used for mating-type tests. Cells were grown as described by SONNEBORN(19870) in a Cerophyl infusion inoculated the day before utilization with Klebsiella pneumoniae and supplemented with 8-sitosterol 0.4 mg/l. All the cultures were maintained a t 27” except when heat sensitivity was tested, in which case the cells were placed at 36”. Genetic analysis and mating iype tests: Crosses, genetic analysis and mating-type tests were performed according to the methods reviewed by SONNEBORN(1970). In these experiments, all matjng-type tests were made on “caryonides”, i.e., on clones derived from one product of the first post-conjugal cell division. A caryonide is classified as 0 or E on the basis of its reaction with solely E or solely 0 testers, respectively. Sometimes cells of both mating types arise within a caryonide. Such caryonides are called selfers (S). In the progeny of normal crosses between micronucleate cells, selfer caryonides are uncommon (NANNEY1957a). Obtaining amicronucleate cells: In the experiments of SONNEBORN,DIPPELL and SCHNELLER (see SONNEBORN1954), the amicronucleate cells were formed after exposure to UV (3300 ergw’mmz) of I00 to 200 cells in 0.1 ml of culture fluid (in the form of a disc 2 cm diameter, 0.16 mm deep). The exposed cells were kept, usually at 31”, in the dark for 24 h r to avoid photoreactivation. In addition to amicronucleate cells, lineages with grossly abnormal micronuclei were also used because they proved to be functionally amicronucleate, i.e., they were unable to give rise to functional gametic nuclei. In BRYGOO and KELLER’Sexperiments, the cells were exposed to an antimicrotubule drug, vinblastine sulphate (E. Lilly and CO). The drug was kept frozen at -20” as a sto8cksolution (500 pg/ml) and diluted to the required concentration (51) pg/ml) just prior to use. Cells from an exponentially growing culture maintained at 27” were concentrated to a final concentration of about 3000 cells/ml by filtration over a nylon cloth with pore size of 01.5 P (Nitre1 C” 48360, Panissi&re, France). Then, they were transferred to a test tube, the drug was added to

MATING-TYPE DETERMINATION I N PARAMECIUM

953

the desired final concentration rand the cultures placed at 22”. After incubation for 10 min, the cells were rapidly washed (by filtration over a similar Nitre1 filter) with fresh culture fluid, transferred to depression slides and observed under a dissecting microscope. During the next 20 min, all the cells in early stages of division were isolated into fresh medium. After completion of their division, the two fission products were put into separate depressions, and both were cultured. Detailed data on obtaining these amicronucleate clones and on their morphological features will be described elsewhere (KELLERand BRYGOO,in preparation). The presence of micronuclei was determined by checking stained cells that were flattened under a coverslip and observed under phase-contrast microscopy, using a 100 x objective. The cells were stained by the lacto-orcein technique used by BEALEand JURAND(1966) for the detection of symbionts with the following modification: the fixation was not made by osmic acid vapors but by the acetic acid vapors of the stain. Evidence for the amicronucleate state: In addition to the routine cytological assay for the presence of micronuclei, two types of data, cytological and genetical, permit the conclusion that a clone is indeed functionally amicronucleate. The significant cytological data are provided by observation of stages of nuclear reorganization i n the amicronucleate clones. Normally, at autogamy and a t conjugation, the fertilization nucleus is derived from micronucIei and gives rise to two new macronuclei, which are readily identified as different from other nuclear structures found; hence, if micronuclei are absent or nonfunctional, neither a fertilization nucleus nor the two distinctive new macronuclei can be formed and nuclear reorganization results only in macronuclear fragmentation. In all the clones that appeared amicronucleate during vegetative growth, reorganizing cells were observed to be devoid of macronuclear anlagen and micronuclei. Although these amicronucleate celIs can undergo macronuclear regeneration, they do not survive because they cannot regenerate a comin preparation). plete oral apparatus (KELLERand BRYGOO, Genetic evidence for the amicronucleate state is provided by the use of gene markers. In all crosses of amicronucleate x normal, the micronucleate partner was genetically marked by either t s l l l or hr. The t s l i l mutation causes lethality at 36“ and the hr mutation determines a high and long-lasting sexual reactivity and a smaller cell size. Under these conditions, if a n amicronucleate cell conjugates with a diploid cell and if a gametic nucleus from the latter gets into the originally amicronucleate cell, then the latter can form both micronuclei and new macronuclei. Since the donor also retains a gametic nucleus and the two gametic nuclei are mitotic products of a single haploid nucleus, the two partners will have the same genotype as that of the nucleate parent, a fact confirmed by the use of recessive gene markers. The use, in some experiments, of cytoplasmic markers such as E”, (ADOUTTEand BEISSON1970) permitted us to distinguish pairs of amicronucleate x 2n conjugants from possible pairs formed by “selfing” in the 2n clone. RESULTS

In P . tetraurelia, conjugation takes place between two cells with complementary, odd (0) and even ( E ) , mating types. At the end of conjugation, both partners contain identical zygotic nuclei resulting from the fusion of a haploid stationary gametic nucleus and a haploid migratory gametic nucleus from the partner. Under normal conditions, the zygotic diploid nucleus appears to play no role in mating-type determination. Cytoplasmic inheritance is observed: the F, and F, clones express the mating type previously expressed in each cytoplasmic parent. To analyze the possible role of the micronucleus in mating-type determination, we have carried out two types of crosses: normal diploid cells of mating type 0 by amicronucleate E cells and normal diploid cells of mating type E by amicronucleate 0 cells. In such crosses. both exconjugants develop a new macronucleus from an identical gametic nucleus provided by the diploid

954

Y. BRYGOO

et al.

partner. This situation enables a comparison to be made of the influence of the same nucleus in the two different cytoplasms. For each cross, 50 to 200 pairs were isolated. The progeny of each pair were analyzed f o r mating-type inheritance only when at least one caryonide (product of the first post-conjugal division) per exconjugant survived. Depending on the experiment, 50 to 100% of pairs were eliminated. In the majority of cases, this lethality affected the descendants of the amicronucleate cells. The majority of the observed viable F, caryonides contained micronuclei that were smaller than the normal diploid micronuclei; they were probably haploid. Most F, clones were 100% inviable at the following uniparental nuclear reorganization. However, in a few cases, all F, cells survived at autogamy, an observation that was interpreted as indicating an early diploidization in the F, clone. Table 1 (A and B) summarizes our observations on mating-type heredity in these crosses. The unpublished data obtained by SONNEBORN, DIPPELLand SCHNELLER, briefly referred to in SONNEBORN (1954), are given in parallel with the recent data of BRYGOO and KELLER.It is worth pointing out the remarkable consistency of results obtained some 26 years apart. Three situations are compared: crosses between noma1 micronucleate cells O(2n) X E(2n), crosses between E amicronucleate cells and normal 0 cells, O(2n) x E(o) and crosses between 0 amicronucleate cells and normal E cells 0 (0) x E (2n). To simplify the presentation, the results obtained in F, clones derived, respectively, from the E parent and from the 0 parent are presented separately in parts A and B of Table 1. Three phenotypes, E, S and 0, were recorded, as described in MATERIALS AND METHODS. The S phenotype corresponds to caryonides that contain cells of both mating types. This S phenotype can be interpreted as a n intermediate state between pure 0 and pure E. In Table IA, we observe in the E(2n) x O(2n) and E(2n) X O ( o ) crosses that the mating type in the lineage from the E parent nearly always remains E, i.e., it still follows the rule of cytoplasmic inheritance. In other words, in the descendants of E micronucleate parents, the determination of mating type is independent of whether or not the partner supplies a gametic nucleus. In the third case, [E(o) X O(2n)], when the E parent is amicronucleate, although there are many more exceptions, more than 3/4 of its descendant caryonides follow the rule of cytoplasmic inheritance. In those crosses in which the 0 parent is micronucleate (Table lB), O(2n) x E(2n) and O(2n) x E ( o ) , the mating type observed shows cytoplasmic inheritance: more than 90% of the descendant caryonides are pure 0. Thus, the mating type O f the 0 micronucleate parent is maintained by its descendant caryonides regardless of whether the partner supplies a gametic nucleus. In contrast, when the 0 parent is amicronucleate, 0 (0) x E (2n), the mating type of its descendant caryonides changes to S o r E in more than 75 % of the cases. DISCUSSION

It has long been known that determination of mating type i n P . tetraureliu occurs through interaction between the nuclei and a nucleus-determined cyto-

955

MATING-TYPE DETERMINATION I N PARAMECIUM

." Fi

al

g B 01 u )

u

g 2

.. J

i

7 U

- 2 % .9 2 & U

2 . ' .*o z

p $j

G +

B

.i

E 2 0

'5

k

B

W

E

3 P

g ;' f

E 2 U c 5

3

%z

2

0

r s l z

.$

!d

._ e

s u

-

U 2 . o U

2

..z 2 3

W

g

0

)

O

5 ,

B a ODs

2 5 % .$ .Es

g

&.

::

2 j

e R" - %

."c

U -

.F?-

0

:;

W

@

$

1 " a G

2

w

-

E

2 2

-

.s

2 7

v

.-e w

-5

F i L 0 , -

U -

Q

w" 8

-o 2 r

rn

w

.-6

n

g m

2 2 4 2c ss

U,

.g EI

F

-a.s

Y

0

lw

+?

12

B

.3

eU

u

@

8

3

6EZ

8

P

gsgs % v i

a,

5 1 1

EZ2 m"

,-Ed8 0 Pldg P 9 2 s W M - %

$!&a Em& p i s s g@Z

i= w z o

sg2

2msll-E E-.l4tt.u

A

956

Y. BRYGOO

et al.

plasmic state. Because determination of mating type is perpetuated by the macronucleus and generally can be switched only during nuclear reorganization (i.e., the development of a new macronucleus) at autogamy or conjugation, it is generally assumed that the nuclear changes underlying this determination concern only the macronucleus (see, for example, SONNEBORN 1977). However, the possibility that the micronucleus might exert an effect that is transmitted through conjugation was suggested by SONNEBORN (1954) on the basis of cited, but unpublished, results and by BRYGOO (1977). In order to test this possibility, two of us (BRYGOOand KELLER)repeated the SONNEBORN, DIPPELL and SCHNELLER study of mating-type inheritance in the progeny of crosses between an amicronucleate E or 0 parent and a normal 0 o r E parent, i.e., under conditions that make it possible to monitor the contribution of a single haploid gametic nucleus of 0 or E origin in the absence of a partner gametic nucleus of opposite origin. The results of both the earlier and later experiments can be summarized as follows: (1) The E or 0 msting type is maintained at conjugation through cellular continuity in the progeny of nearly all E o r 0 micronucleate parents. even when they do not receive a gametic nucleus from their partner. (2) When an amicronucleate cell of mating type 0 receives a micronucleus from its E partner, its mating type becomes E in the majority of cases. However, it remains 0 in 12 to 25.5% of the cases (Table 1B). ( 3 ) When an amicronucleate cell of mating type E receives a micronucleus from its 0 partner, its mating type tends to remain E. as in a normal cross between two micronucleate cells; however, it switches towards 0 in about 23% of the cases (Table IA), as compared to 1 4 % in normal crosses. These results suggest the following conclusion. The 0 or E cytoplasmic state corresponding to the activity of an 0- or E-determined parental macronucleus is not sufficient to maintain determination through the F, in the absence of a stationary gametic nucleus; the normally stable perpetuation of mating-type determination is destabilized. Furthermore, the fact that most amicronucleate parents of mating type 0 produce E F, progeny, after receiving a micronucleus of E origin, while, in the reverse situation, most amicronucleate cells of E mating type remain E suggests that there is little or no O-determining cytoplasmic factor of macronuclear origin in an 0 cell, whereas E cells contain a usually effective concentration of a cytoplasmic factor contributing to the determination towards E. The following hypothesis is proposed to explain mating-type inheritance in normal crosses between micronucleate cells and its disturbance in crosses of amicronucleate cells x micronucleate cells: (I) As previously assumed by SONNEBORN (1954, 1977). there are alternative stable genic states that characterize the 0 and E macronuclei; precondition toward these alternative states is present in the micronuclei of 0 and E cells. ( 2 ) The transformation from one genic state to the other is under the control of two kinds of factors: one involved in 0 determination and one in E determination. The first kind of factor, which characterizes the 0 state, is mostly confined within the micronucleus and macro-

MATING-TYPE DETERMINATION IN PARAMECIUM

95 7

nucleus, and the second kind of factor, which characterizes the E state, is present in both the cytoplasm and the nuclei. ( 3 ) During conjugation, the diploid nuclei formed from the fusion of one 0-predetermined and one E-predetermined gametic nucleus in each conjugant cell become homogeneously 0 or E determined in agreement with the parental type. We assume that this transformation involves an interaction between the two types of factors: in the absence of the E cytoplasmic factor, the 0 factor is “dominant” over the E factor; but in the presence of the E cytoplasmic factor, the 0 factor is less effective. The presence, solely in the nucleus, of the “dominant” factor that controls the determination of 0 explains, on the one hand, the fact that an 0 micronucleate parent cell produces 0 progeny at conjugation and, on the other hand, the fact that, in the absence of a micronucleus, the 0 cell changes its mating type when it receives the micronucleus of an E partner. The 25% of amicronucleate 0 cells that remain 0 despite having received an E micronucleus (Table IB) could indicate either the presence in the 0 cytopIasm of an 0-determining factor with a weak efficiency or that the E factor within the nucleus is less effective in the absence of an E cytoplasmic factor. The 23% of amicronucleate E cells that change their mating type when receiving the micronucleus of an 0 partner can be taken to mean that E cytoplasm has less effect on an 0 micronucleus than on a micronucleus formed by fusion of 0 and E micronuclei. To interpret his results in 1954, SONNEBORN proposed that the gamete nuclei are predetermined. but he assumed that the cytoplasmic factors 0 or E in normal conjugation tip the baIance between the oppositely determined gametic nuclei: LL When the gametic nucleus develops parthenogenetically, the balance of determinative action is already heavily weighted and is less readily reversed by cytopIasmic action on the developing macronucleus” (SONNEBORN 1954). As for the second point of our hypothesis, NANNEY (1957a) proposed a similar hypothesis for explaining the difference in mating-type inheritance between Paramecium species that display cytoplasmic inheritance of mating type (group B) and the species in which mating type is determined independently in each newly formed macronucleus (group A ) : “The group A system has only an intranuclear mechanism of perpetuation; the group B system has both an internuclear and an intranuclear communication system” ( NANNEY 1957b). Our hypothesis, which assumes that the 0 state is only nuclear, while the E state is nucleocytoplasmic, is jn good agreement with the conclusion drawn from the study of the O* phenotype (BRYGOO 1977). A major conclusion of this study was that locus mtD, identified in crosses between stock 51 ( m t D 5 l / m t D 5 1 ) and stock 32 (mtD?2/mtD32), was active in E cells and inactive in 0 cells of either genotype and that the activity o€ the m t D locus was induced in newly arising macronuclei when its product was present in the cytoplasm, as it is in E, but not in 0 cells. In this study (BRYGOO 1977), it was also concluded that the O* state was characterized by (a) a hereditary cytoplasmic state like the E state, and (b) a nuclear state different from the 0 state and revealed only by comparison of the results of the crosses O* (mtD5l/mtD51) X E(mtD32/mtD?2) and O ( m t D 5 l / m t D S I ) x E(mtD32/mtD32). In view of the results reported here,

958

Y. BRYGOO

et al.

one is led to ask whether the inheritance of the O* state is based upon a nuclear state, comparable to that of 0 cells. To answer this question, the cross O* (mtD52/mtD51) x O(mtD32/mtD32) was performed with the use of cilia from E reactive cells (BRYGOO, unpublished results). If the O* state, like the 0 state, is only a nuclear state, the two exconjugants are expected to have the same phenotype. On the contrary, we observed that the two exconjugants did not display the same phenotype: the 0’ parent became E and the 0 parent remained 0. Hence, the O* state cannot be solely a nuclear state. Moreover, the fact that the F, from the O* parent is E regardless of whether its mate provides an inactive mtD gene (when the mate is 0) o r a n active mtD gene when the mate is E [as in the cross O* (mtD51/mtD52) x E(mtD32/mtD32) J indicates that O* contains a cytoplasmic factor(s) that activates mtD32 in developing new macronuclei and, hence, acts like the E cytoplasmic factor inferred to be present in E cells on the basis of the results of the amicronucleate crosses reported here. Therefore, it can be concluded that the 0’ cells contain a part (quantitative or qualitative) of the E-determining cytoplasmic factor (s) , but it is still unknown whether O* cells contain an 0-determining nuclear factor. The existence of this system of predetermination in germ line nuclei (micronuclei) leading to stable determination of somatic nuclei (macronuclei) raises the question of how (if at all) these nuclear “states” differ from mutation. Further, this ambiguity leads to the suggestion that in other organisms some examples considered to be genic mutations may instead be stably determined alterations of genic activity. One thinks particularly of so-called mutable genes, those that have frequencies of change much higher than typical mutations, but are somatically stable. Among examples in ciliates are those reported for certain mating-type “alleles” in P. bursaria (SIEGELand COLE 1967) and in Tetrahymena (ORIAS1963). In summary, the main point of interest of the present paper is the demonstration that stable determination of somatic nuclei (macronuclei) is influenced not only by cytoplasmic factors, but also by factors in the nuclei themselves, including the germinal nuclei (micronuclei and gamete nuclei). By this means, nuclear factors bring about apparent “cytoplasmic inheritance.” The authors thank J. BEISSONfor critical reading of the manuscript and many fruitful discussions and J. BRIZARD for typing the manuscript. LITERATURE CITED

ADOUTTE, A. and J. BEISSON,1970 Cytoplasmic inheritance of erythromycin-resistant mutations in Paramecium aurelia. Molec. gen. Genet., 108: 70-77.

BEALE,G. H. and A. JURAND, 1966 Three different types of mate-killer (mu) particles in Paramecium aurelia (syngen 1).J. Cell Sci. 1: 31-34. BEISSON,J. and M. ROSSIGNOL, 1969 The first case of linkage i n Paramecium aurelia. Genet. Res. 13 : 85-90.

BRYGOO, Y., 1977 Genetic analysis of mating-type differentiation in Paramecium tetraurelia. Genetics 87: 633-653.

MATING-TYPE DETERMINATION IN PARAMECIUM

959

NANNEY, D. L.,1957a Mating type inheritance at conjugation in variety of 4 of Paramecium aurelia. J. Protozool. 4: 89-95. -, 1957b The role of the cytoplasm in heredity. pp. 134-64. In: The Chemical Basis of Heredity. Edited by W. D. MCELROY and B. GLASS. Johns Hopkins Univ. Press, Baltimore. ORIAS,E.,1963 Mating type determination i n variety 8, Tetrahymena pyriformis Genetics 48: 1509-1518. SIEGEL,R. W. and &LE, J., 1967 The nature and origin of mutations which block a temporal sequence for genic expression in Paramecium. Genetics 55: 607-617. SONNEBORN, T. M.,1950 Methods in the general biology and genetics of Paramecium aurelia. J. Exp. 2001. 113: 87-147. , 1954 Patterns of nucleo-cytoplasmic integration in Paramecium. Caryologia 6 (suppl.) 307325. -, 1970 Methods in Paramecium research. pp. 241-339. In: Methods in Cell Physiology, Vol. 4. Edited by D. M. PRESCOTT. Academic Press, New York. -, 1975 The Paramecium aurelia complex of fourteen sibling species. Trans. Am. Microsc. Soc. 94: 155-78. -, 1977 Genetics of cellular differentiation: stable nuclear differentiation in eucaryotic unicells. Ann. Rev. Genet. 11 : 349-367.

-