sign in sign up
<<
Home , Autosome, B chromosome

Heredity 57 (1986) 247-254 The Genetical Society of Great Britain Received 16 December 1985

Continuous variation in Y- structure of Rumex acetosa

A. S. Wilby and School of Biological Sciences, J. S. Parker Queen Mary College, Mile End Road, London El 4NS.

The dioecious angiosperm Rumex acetosa has an XXIXY1Y2sex-chromosomesystem. Each V-chromosome is heterochromatic except for a minute terminal euchromatic pairing segment. The Vs are constant in size but have a variable position. The can be located anywhere within the central 40 per cent of the chromosome but are excluded from the two distal 30 per cent regions. In a sample of 270 males from 18 different populations 68 distinct variants have been identified on the basis of V-morphology. All populations are highly polymorphic with a minimum of four variants in a sample of ten males. The origin and significance of this massive variability is considered in this paper. Increased rate of the Ys may be implicated in maintenance of this variation.

I NTRO DUCTI ON these "inert" Ys has been described (Vana, 1972) variation in their structure has been overlooked. -determinationin animals is usually genic and Extensive heterochroinatic content is a charac- frequently associated with visibly-differentiated teristic of many Y- and W-. Indeed, sex-chromosomes. Sex expression in plants, some have argued that the process of hetero- however, is usually more plastic, and is subject to chromatinisation itself was implicated in the initial environmental influences such as temperature and phase of sex-chromosome differentiation (Jones, photoperiod (Heslop-Harrison, 1957). Addition- 1984). Highly-heterochromatic sex-chromosomes ally, even strictly dioecious plants seldom have exhibit hypervariability in many groups of organ- visibly-differentiated sex-chromosomes and con- isms and this is normally quantitative in character siderable controversy exists concerning the partic- (Hamerton, 1971; Mengden 1981; Vorontsov eta!., ular chromosomes involved in sex-determination 1980). In the Y, for example, the long arm in as diverse as the gymnosperm Ginkgo varies more or less continuously with respect to its biloba and the angiosperm Spinacia oleracea C-band content while a single population of the (Lewis and John, 1968). rodent Nesokia indica contained three distinct Y- Highly-differentiated sex-chromosomes have morphs in equilibrium proportions (Rao et a!., been found in a few plant species such as Melan- 1983). drium rubrum and Rumex acetosa. In M. rubrum Structural variation in sex-chromosomes a pair of X-chromosomes is found in pistillate independent of C-band content is unusual, but (female) plants, a single X and a larger Y in a remarkable case has been described in the W- the staminate (male) plants. Detailed studies of chromosomes of the gekkonid lizard Gehyra pur- this active-Y system have been carried out by purascens. Six W-morphs which differ by para- and Westergaard (1958). The subgenus to which Rumex pericentric inversions have been identified in a acetosa belongs is characterised by an aneuploid sample of only 30 females from 13 populations sex-chromosome system, XX in females and (Moritz, 1984). In Rumex acetosa we have found XY1Y2 in males (Kihara and Ono, 1925). Several that the two Y-chromosomes, which are almost studies have shown this to be a dosage compensa- completely heterochromatic, show massive struc- tion system analogous to that in Drosophila (Ono, tural but not quantitative variability. Centromere 1935). Although the heterochromatic nature of location in each Y varies more or less continuously 248 A. S. WILBY AND J. S. PARKER such that in 270 males 68 recognisably-distinct 1 and 6 sometimes carry heterochromatic segments complements have been found. The nature of this terminal on their short arms. distinctive sex-chromosome system is reported in The X-chromosome is metacentric. It is the this paper. largest member of the complement (7 pm), nearly twice the length of . During mitotic prophase the X-chromosomes and coil MATERIALSAND METHODS regularly, displaying no conspicuous blocks of . The two X-chromosomes in Theplants used in this study were grown from females have identical coiling patterns. seed collected in 18 natural populations of The Y1- and Y2-chromosomes, by contrast, are R. acetosa.Inthree populations—Navestock highly heterochromatic. They are distinguishable Heath, Essex, Torver, Cumbria and Hendra, Corn- from the rest of the complement on the basis of wall—seeds were collected from 40 females. In the coiling and staining at all stages of the mitotic other 15 populations, scattered widely through cycle except and anaphase. In inter- Britain, 10 seed samples were obtained. From each phase nuclei, one large or two smaller female parent a single male offspring was chromocentres are present. These may either examined giving a total of 270 males. The precise appear as dense masses or as extended chro- locations of these populations will be given in a mosome-like structures. Plants can be reliably later paper. sexed by observations of interphase nuclei. During For chromosome studies roots were pretreated mitotic prophase the Ys are uniformly dark- with 0.05 per cent colchicine for 2 hours, fixed in staining with no evidence of intercalary euchro- 1:3 acetic alcohol and stained by the Feulgen matin. is present, however, and is method. was examined in PMCs fixed in limited to a very short terminal region on each Carnoy and stained in 2 per cent acetocarmine. Y-chromosome (fig. 3). These segments on Y1 and Y2 are probably the pairing segments, homologous with the distal regions of the X arms. RESULTS The Y-chromosomes are intermediate in size between the autosomes and the X-chromosome. The The Y1 is on average 83 per cent of the X-chro- mosome length at metaphase, while the Y2 is 74 Pistillate (female) plants of Rumex acetosa have per cent of the X. Most frequently both Y-chro- 2n =14chromosomes comprising six pairs of auto- mosomes are metacentric. The position of the somes and a pair of X-chromosomes. The centromeres, however, is remarkably variable and staminate (male) plants carry a single X-chromo- the pattern of this variation is described below. some and two distinguishable Y-chromosomes, designated Y1 and Y2, giving a somatic number of orientation 2n =15. Sex-trivalent The six pairs of autosomes are individually Maintenanceof a 1: 1 sex ratio in R. acetosa recognisable on the basis of length, arm-ratio and depends on the segregation of both Y-chro- secondary constrictions (figs. 1-2). Chromosomes mosomes to the same pole at anaphase-I. This is fr '1 'sf1,

Figure la and b' Chromosome complements of Rumexacetosa (a) 9, 2n = 14+1B;X-chromosomesarrowed. (b) d, 2n= 15+2Bs; X,Y1 and Y2 arrowed 'lIar in all figures represents 10 p.m. V-CHROMOSOME STRUCTURE OF RUMEX ACETOSA 249

(b) Y-chromosomes In conventionally-stained preparations chro- mosome morphology is defined by centromere position alone. In the Y-chromosomes of R. acetosa,however,the euchromatic pairing segment defines one so centromere position in acrocentrics can be given a polarity. In about 40 per cent of the males (115/270) both Y1 and Y2 were metacentric while in the rest at least one Y was acrocentric. Four major classes can be defined on the basis of centromere position: both Ys meta- I centric (mm), Y1 or Y2 acrocentric (am and ma respectively), both Y1 and Y2 acrocentric (aa). In addition Y-polarity enables two classes to be dis- tinguished in am and ma and four classes in aa. We designate centric location nearer the pairing segment "+"andaway "—"(seefig. 3). All nine classes have been found in this sample of 270 males.

Figure 2 Karyotype of Rumexacetosa, showingthe extreme Y1 and Y2 variants. The Ys are drawn with the pairing segments uppermost C,

ensured by regular production of a linear sex- trivalent by two terminal chiasmata and its sub- •0 sequent orientation in a convergent manner (fig. 4). The system is extremely accurate. In 238 meta- .4,,. phase-I cells six trivalents in linear orientation (fig. 5)andthree bivalent+ univalent configurations were found (38 per cent abnormal PMCs) and at Figure 3 Mitotic prophase in male Rumexacetosa. Notethe anaphase-I only five PMCs out of 371 (1.3 per heterochromatic nature of the Ys. The euchromatic pairing cent) were XY: Y. Failure of regular segregation segments are arrowed should give numerical sex-chromosome variants in the progeny and in 1095seedgrown plants two Changes in the apparent position of the centro- XXY individuals were detected. meres could result from addition or of chromosomal material. This can be excluded by comparison of chromosome length in meta- and Sex-chromosomestructural variants acrocentric Y-chromosomes. For Y1 the meta- (a)X-chromosome centrics average 822 per cent of X and acrocentrics The X-chromosome is very stable and few variants 839 per cent, for Y2 735 per cent and 747 per have been detected. Two females carrying single cent respectively. Rather than addition/deletion, acrocentric Xs and a male with an acrocentric X, rearrangements such as pericentric inversions are all of standard length, have been found, presum- implicated in centric transposition. ably resulting from pericentric inversion. In addi- Centromere position is defined by long-arm tion two different X/ interchanges have length as a proportion of total length. Thus a been detected: one female was heterozygous for metacentric has an arm-ratio of 05 and a telo- an X/3 interchange while two females and a male centric 1.0. The smallest deviation from meta- from the same population carried an X/2 centricity which can be reliably assigned a polarity exchange. with respect to the euchromatic segment gives an 250 A. S. WILBY AND J. S. PARKER

S -4 S

Figures4 and 5 Metaphase-I in PMCs of Rumex acetosa. (4) Sex-trivalent in convergent orientation. (5) Sex-trivalent in linear orientation.

arm-ratio of about 0-53. In the following analysis tion occurs from 056 to 058 (16/29). The number the numbers of different Y-chromosome arm-ratios of acrocentric types of Y1 and Y2 are about equal, have been established within populations and then 64 and 71 respectively. In Y1 positive and negative summed between populations. Thus chromosomes locations are equally frequent but Y2 shows a slight with the same arm-ratio occur in several popula- negative bias (42:29). tions. In view of the heterogeneity of the observa- The total number of morphologically distin- tions, each of these occurrences has been assessed guishable variants in all 270 males from 18 popula- as if it were independent. The actual numbers of tions was 68. In order to show whether Y1 and Y2 males have not been considered because of the variation is independent the variants can be plotted large difference in sample sizes between popula- on a simple cross-shaped diagram, one axis tions (10 vs 40). representing Y1, the other Y2 and the crossing In the Y1-chromosomes arm-ratios from 067 representing the mm type (fig. 8). to O66 have been recorded (figs. 6, 7). There is In general, Y1 and Y2 appear to vary indepen- a more or less continuous range in the positive dently of each other. Thus the four classes am, direction with no clustering. In the negative direc- am, ma and ma contain roughly equal numbers tion the majority of observations fall between 0.55 of variants (9, 8, 9 and 13 respectively). An and 0-57 (20/33). In Y2 similarly centromere posi- anomaly occurs, however, within the class aa. tion ranges from 068k to 0-71(figs. 6,7). In Twenty-eight morphologically-distinct variants contrast to Y1 ,thecentromere location in the nega- have been found, of which 13 are aa and 7 tive direction is continuously distributed (0.55— The classes aa and aTh are uncommon 071) while in the positive direction a concentra- with only four variants in each (fig. 8). I ae a. 4-f a a Sm -s as 01. 0101 Figure 6 Comparison of variation in centromere position in Y1 and Y2 chromosomes. Arm-ratios where centromeres are closer to the pairing segment are denoted '+'andfurther away V-CHROMOSOME STRUCTURE OF RUMEX ACETOSA 251 . .. . . ÷ S.. .• S Yl • .•...... • .. . ..S •• • ..•• ..• • .5 T.... .••. •••••••ST O•7 O6 05 O•6 O•7

• Y Y2: : ..:. : • ...•••...••• ..••. •S S •..S.....•S•••S ...... • I I I I I 0•7 06 0•5 06 07

Figure7 The location of centromeres in 270 Y, and Y2 chromosomes drawn from 18 natural populations. Variants were assessed in each population and all have been plotted. Variants with the same arm-ratio and polarity may or may not be identical. Chromosomes with arm-ratios less than 054 cannot be assigned a polarity and have been omitted.

Population structure extremely fluid, apparently moving along the cen- tral 40 per cent of the chromosome about 20 per Allthree of the large populations are extremely cent each side of the median position. There seem heterogeneous in Y-chromosome composition. to be some preferred locations for the centromere Seventeen of the 40 males in the Navestock Heath when non-median. In Y1 a concentration occurs population, for example, were mm while the 5-8 per cent further away from the terminal remainder carried one or two acrocentric Ys. Ten euchromatic segment, while in Y2 a peak occurs different acrocentric types could be defined by 6—8 per cent nearer the euchromatic segment. All arm-ratio and polarity. Four variants were detected other locations giving acrocentric Y-chromosomes each. once, and the remainder only 2-5times are about equally frequent. Similar patterns of population heterogeneity were In each Y there is a strict limit to free centro- found in the Hendra and Torver populations with mere movement at about 20 per cent from the seven and 10 acrocentric variants respectively; the median in both directions. Centromeres are maximum number of any one variant was four. excluded from the two distal regions, each com- Most of these variants are demonstrably unique to prising 30 per cent of the chromosome. It might the population. be argued that the distribution of centromere posi- The extreme Y-variability also characterises all tion is normal about the median so that chromo- 15 small populations. In the 10-male samples somes from the tails of the distribution would be drawn from these populations a minimum of four rare and hence undetected in our sample of 270 and a maximum of eight variants have been detec- males. If this were so, however, we would expect ted. Detailed analyses of population structure will the majority of acrocentric Y-chromosomes to have be reported in a later paper. arm ratios between 054 and 056 which is clearly not the case. DISCUSSION What then limits centric disposition in Rumex acetosa?Apossible mechanism is suggested by a TheY1- and Y2- chromosomes of Rumex acetosa functionally-similar system in mantid species in show a remarkable pattern of structural variation. which males contain an X1X2Y trivalent. Between- The location of the centromere in each Y is species differences in centromere location indicate 252 A. S. WILBY AND J. S. PARKER • • + vi

O•6- • • • S

05-

• S S S S • S • S

06 S S

0-7 0-6 0•5 0•6 0•7 Y÷2 Figure 8 Arm-ratios of the 67 morphologically-distinguishable Y1Y2 Variants found in the 270-mile sample, in which at least one Y was acrocentric. The mm variant is omitted.

coordination of the X1— Y and X2 —Yintercentric shifts of the centromeres in opposite directions distances, maintaining an effectively isosceles from the median position. The difference between triangle at metaphase-I to ensure regular disjunc- centromere positions in the two Ys, however, may tion (White, 1973). It could perhaps be argued that be considerable even in ma and am morphs, up regular disjunction of the XY1Y2 trivalent in R. to 21 per cent in ma. This explanation then is acelosa is dependent on more or less median cen- inadequate to explain the deficiency of aa and tric disposition in both Ys. However, disjunction aa classes in our sample. in males carrying the most extreme acrocentric Ys The Y-chromosomes of R. acetosa are uni- in combination with metacentrics is entirely regu- formly heterochromatic in conventionally-stained lar. If disjunction efficiency does limit centric dis- preparations. The boundaries between the centro- position, then breakdown must be a sudden thresh- mere-containing region and the distal regions, old event. It is also difficult to understand why therefore, must fall within a continuous stretch of centromere movement, particularly away from the heterochromatin. A finer analysis of the hetero- pairing segment, should adversely affect orienta- has yet to be undertaken, but others tion, especially so if both Y1 and Y2 are affected have found uniformity using fluorochromes and in concert, or at least in a stepwise progression. C-banding (Vana, 1972; Leeman and Ruch, 1983). The deficiency of aa and aa classes may Hypervariability of sex-chromosomes has been result from imbalance of the sex-trivalent due to noted in groups as diverse as mammals and V-CHROMOSOME STRUCTURE OF RUMEX ACETOSA 253 reptiles, in systems with both male and female tural variation of the Y-chromosomes then may be heterogamety (Moritz, 1984; Vorontsov et aL, neutral although a selective component may be 1980). Most of the variation encountered is quanti- implicated in defining the limits of centric move- tative. Structural variation, unaccompanied by ment. If Y-variation is neutral monomorphic popu- changes in C-band quantity, is less usual. The most lations might be expected as a result of random striking example is the W-chromosome of the gek- fixation of variants. None has been found in the konid lizard Gehyra purpurascens, in thirteen popu- 18 populations examined; indeed all are more or lations of which six W-morphs have been identified less equally variable with at lest four variants in (Moritz,1984). These W-chromosomes are random samples of 10 males. This massive vari- differentiated by pen- and paracentric inversions ation perhaps results from an enhanced Y-chro- and a centric shift, and an evolutionary scheme mosome mutation rate compared with the rest of has been proposed deriving them from the stan- the . dard Z-chromosomes. Only two populations with Sex-chromosomes in many groups of organ- more than three females have been studied, with isms are rich in particular DNA families named six and eight individuals, and both were mono- Bkm sequences originally isolated from the W- morphic. Only one population contained two chromosomes of snakes (Singh et a!., 1980). These morphs, thus the species appears polytypic with Bkm sequences have some of the properties of respect to the W-chromosome. Clearly, the highly transposable elements (Jones, 1984) and enhanced variable sex-chromosome system of Gehyra is rates of chromosome mutation have recently been qualitatively different from that in Rumex acetosa ascribed to transposable elements, as for example reported here. in maize (McClintock, 1978) and Drosophila All populations of R. acetosa are polymorphic (Kidwell eta!., 1977). It is not known whether Bkm with respect to the Y-chromosomes. Centric loca- or similar transposon-like sequences are present tion in the Ys is so fluid that very few individuals in enriched amounts on the Y-chromosomes of R. of any one variant can be found in a single popula- acetosa. If they do occur, however, it is tempting tion. The simplest mechanism to account for these to speculate that they are located preferentially in extensive centric shifts is pericentric inversion the central 40 per cent of the Y-chromosomes requiring two breaks. Both para- and pericentric through which the centromeres wander. Exclusion inversions are demonstrable in Gehyra by G-, C- of the centromere from the two terminal regions and N-banding patterns. In R. acetosa, however, then may reflect a lack of potential rearrangement the homogeneous nature of the constitutive hetero- sites due to the absence of transposon-like sequen- chromatin of the Ys precludes direct demonstra- ces. This proposal is testable by molecular tech- tion of pericentric inversion. Other mechanisms of niques. centric transposition require three or more simul- taneous breaks and such occurrences will be rarer Acknowledgements A. S. Wilby is supported by an S.E.R.C. than simple inversions. In Gehyra, for example, 9 research studentship. inversions have been implicated in the generation of the W-chromosomes, but only a single 3-break event (Moritz, 1984). If centric shifts in R. acetosa REFERENCES do result from inversion, then it is also likely that DARLINGTON,C. 0. 1958. Evolution of Genetic Systems. Oliver the mm class is a heterogeneous assemblage and and Boyd, Edinburgh. the extent of polymorphism is even greater than HAMERTON, J. L. 1971. Human . Academic Press, that recorded. New York and London. HESLOP-HARRISON. j.1957.The experimental modification of The centric shifts in the heterochromatic Y- sex expression in flowering plants. Biol. Rev., 32, 3 8-90. chromosomes probably have no mechanical sig- JONES, K. W. 1984. The evolution of sex chromosomes and their nificance at meiosis since the segregation of the consequences for the evolutionary process. Chromosomes trivalent is as regular in the most acrocentric Ys Today, 8. Eds. M. D. Bennett, A. Gropp and V. Wolf. George Allen and Unwin, London, 241-255. as in the more frequent metacentrics. In addition, KIDWELL, M. G., KIDWELL, J. F. AND SVED, J. A. 1977. Hybrid chiasma formation between the X and Ys is exclus- dysgenesis in : a syndrome of aber- ively terminal involving only the minute euchro- rant traits including mutation, sterility and male recombi- matic pairing segments. Inversions then cannot nation. , 86, 813—833. KIHARA, H. AND ONO,T. 1925.The sex chromosomes of Rumex fulfil a role in isolating blocks (Darlington, acetosa. Z. Indukt. Abstamm. U Vererb., 39, 1-7. 1958). Indeed it seems likely that the Ys carry few, LEEMAN. U. AND RUCH, F. 1983. DNA and base content in if any, major transcribed loci since they remain the nuclei and sex chromatin of Rumex acetosa. Bot. Helv., tightly coiled throughout the cell cycle. The struc- 93, 77—83. 254 A. S. WILBY AND J. S. PARKER

LEWIS. K. R. AND JOHN, B. 1968. The chromosomal basis of SINGH, L., PURDOM, I. F. AND JONES, K. W. 1980. Sex-chromo- sex determination. mt. Rev. Cytol., 23, 277-379. some associated satellite DNA: Evolution and conserva- McCLINTOCK,B. 1978. Mechanisms that rapidly reorganise the tion. Chromosoma, 79, 137—157. genome. Stadler Symp., 10, 25-48. VANA, V. 1972. The localisation of heterochromatic segments MENGDEN, G. A. 1981. Linear differentiation of the C-band in the chromosomes of Rumex acetosa L. Preslia, 44, 316- pattern of the W.chromosome in snakes and birds. 326. Chromosoma, 83, 276-287. VORONTSOV, N. N., LYAPUNOVA, E. A., BORRISOV, Y. M. AND MORITZ,c.1984. Theevolutionof a highly-variable sex- DOVGAL, V. E. 1980. Variability of sex chromosomes in chromosome in Gehyra purpurascens (Gekkonidae). mammals. Genetica, 52/53, 361-372. Chro,nosoma, 90, 111-119. WESTERGAARD, M. 1958. The mechanism of sex-determination ONO, T. 1935. Chromosomen und sexualitat von Rumex acetosa. in dioecious flowering plants. 4dv. Gen., 9, 217-281. Sri. Rep. Tohuku Imp. Univ. IV, 10, 41-210. WHITE, M. J. D. 1973. Animal Cytology and Evolution. Cam- RAO, S. P.. V., VASANTHA, K., THELMA, B. K., JUYAL, R. C. AND bridge University Press, Cambridge. JHANWAR, S. C. 1983. Heterochromatin variation and variation in Nesokia indica: a population study. Cytogenet. Cell Genet., 35, 233—237.