In the Mollusca, As Far As I Am Aware, Only Two Instances of Twq Populations Found on the North Downs in Surrey; One at a Locali

Home , B chromosome

SUPERNUMERARY CHROMOSOMES IN WILD POPULATIONS OF THE SNAIL HELIX POMATIA L. H. J. EVANS M.R.C. Radiobiologica! Research Unit, Harwel!, Didcot, Berks Receivedi6.ii.6o 1.INTRODUCTION THEpresence of chromosomes additional to, and not homologous with, the chromosomes of the normal diploid complement has been noted in a large number of species, both plant (Darlington, 1956) andanimal (White, 1954). The reports of undisputed supernumerary chromosomes in animal populations have been confined to invertebrate species and principally to one group, the Insecta, with the notable exception of the recording by Melander (1950) of supernumerary chromosomes in two species of Turbellarian worms. In the Mollusca, as far as I am aware, only two instances of variation in chromosome number within species are known. In the Prosobranchiate Purpura lapillus, Staiger (ig) found that the haploid chromosome number varied from n =13to n =r8,and in the snail Triodopsisfraudulenta Husted and Burch (1946) found individuals having up to eight chromosomes additional to the normal diploid complement of 2fl =58.The numerical variation found in Purpura lapillus is of the type first described by Robertson (1916),wherethe number of chromosome arms is constant between individuals, but the number of chromosomes varies because one or more of the chromo- somal elements may be present either as a single metacentric or as two acrocentric chromosomes. The extra chromosomes found in Triodopsis fraudulenta are of a quite different nature, and were inter- prcted by Husted and Burch as being polysomatic as they appeared to Le homologous with chromosomes of the regular complement. The present report is concerned with the occurrence and behaviour of supernumerary chromosomes in populations of the Roman snail, Helix pomatia L. which were initially studied in an attempt to obtain polyploid animals (cf. White, 1954) for use in radiobiological work.

2. THEPOPULATIONS InBritain H. pomatia L. has a limited distribution and is confined to chalky areas in the south: the populations themselves have a very localised distribution and tend to occur in small pockets. For the present study individuals were collected during May and June from twQ populations found on the North Downs in Surrey; one at a locality known as Farthing Downs (population F) and tile other about 5 miles away at White Hill, between Merstham and Caterham population K). Although the populations are in themselves quite 130 H. J. EVANS distinct, it is possible that they are not truly isolated because of the possibility of transfer of individuals by bird predators and human interference. In addition to the studies on H. pomatia L., fifteen individuals of H. aspersa Mull which were collected at Abingdon, Berks, were also studied. 3.METHODS Ovo-testes weredissected out and fixed in acetic-alcohol (i : 3) for up to 24 hours. The tissue was then stained using the Feulgen procedure and 4 to 6 slides made from the teased tissues by the squash technique. After squashing, the cover gIasse were separated from the slides by the dry-ice method (Conger and Fairchild, I953 and the preparations made permanent.

4. CYTOLOGICALOBSERVATIONS Observationswere made principally on meiosis in primary and secondary spermatocytes. In a few instances mitosis in spermatogonial cells was studied, but unfbrtunately oocytes at metaphase were not observed. All the slides made from each ovo-testis were studied and a total of about 50 cells were scored from each individual. Chromo- some counts were made on cells at diakinesis or at first metaphase of meiosis, and only unbroken cells with well spread chromosomes were scored. (i)The normal chromosome complement Insome of the individuals studied the chromosome number was. determined as 2fl =54,a value confirming counts made by Perrot and Perrot (i937). Studies of spermatogonial mitosis showed that all the chromosomes have median or sub-median centromeres and at meiosis 27 bivalents are regularly formed. At the metaphase stage the bivalents appear as rods or rings, the smallest bivalent being about 1.5 p long and the largest about 2.5 p long (plate, fig. i). No polyploid animals were found among the twenty-five individuals. of H. pomatia L. which were studied, or among the fifteen individuals of the related species H. aspersa Mull. In most individuals of H. pomatia L. about i to 2 per cent, of the primary spermatocytes were found to be tetraploid. In these tetraploid cells no multivalents were observed (plate, fig. 2) and the 54 bivalents were invariably orientated on a common spindle. The absence of quadrivalents might be regarded as indicating that polyploidy results from the fusion of nuclei, either within a binucleate cell or from separate cells, which occurs sometime after the completion of chromosome pairing, but prior to the formation of the spindle. (ii)The supernumerary chromosome Inboth populations of H. pomatia L. individuals were found which had from one to six chromosomes additional to the normal diploid chromosome complement. The morphology and behaviour of the supernumerary or B chromosomes indicate that there is only one type of B chromosome and that this chromosome is common to both SUPERNUMERARY CHROMOSOMES IN A SNAIL 131

populationsK and F. The extra chromosome appears to have a sub-median centromere and may be distinguished from the normal chromosome complement by its small size, being a little over i long at meiotic metaphase, and by the fact that it does not pair with any of the normal chromosomes. The supernumerary chromosomes in many plants and animals are heterochromatic, but the additional chromosome in the snail does not show any heterochromatin, although heterochromatin is present in many of the normal chromosomes. Apart from a few observations made on spermatogonial cells, the behaviour of the B chromosome(s) at mitosis was not studied, although, as will be discussed later, certain inferences about their behaviour at mitosis may be drawn from the frequency distribution of supernumerary chromosomes found between spermatocytes within individuals. The behaviour of the extra chromosome(s) at meiosis is character. istic.If only one supernumerary chromosome is present, in the early stages of meiosis its behaviour is indistinguishable from that of the normal chromosomes, except that it does not pair with any other members of the chromosome complement. At metaphase the extra univalent chromosome does not usually congress onto the metaphase plate, but lies between the equatorial region and one of the spindle poles (plate, fig. 3) becoming incorporated into one of the two telephase nuclei, and dividing mitotically at the second anaphase of meiosis. Occasionally the supernumerary chromosome undergoes congression with, or a little later than, the normal bivalent chromosomes and is then usually located at the edge of the metaphase plate (plate, fig. 3). In this latter case the extra chromosome undergoes a mitotic division at the first anaphase and does not divide at the second division (plate, fig. io). Both types of meiotic behaviour of the univalent chromosome result in 50 per cent, of the spermatids containing one supernumerary chromosome; no instances were observed where the extra chromosome failed to be included in the daughter cell nuclei (cf. Rees and Jamieson, 1954). Inprimary spermatocytes containing two extra chromosomes these chromosomes were associated as bivalents in 83 per cent, of the cells scored (table i),eachbivalent having a single chiasma (plate, fig. 5). When more than two B chromosomes are present in a cell the extra chromosomes either appear as univalents or are associated as bivalents. In the 220 cells scored which contained three or more B chromosomes no true multivalent associations were observed (plate, figs. 6-9), although at the early stages of meiosis the B chromosomes. are often seen to be closely associated. The supernumerary bivalents behave normally at meiosis, the half bivalents separating to opposite poles of the cell: no meio tic non-disjunction such as is shown by the acrocentric supernumerary chromosomes of certain Acrididze (White, '954) was found. The frequency of bivalent formation between B chromosomes in cells 132 H.J. EVANS containing up to six extra chromosomes is shown in table i.In cells containing more than one supernumerary chromosome, cells having odd numbers would have a greater probability of producing their maximum possible number of bivalents than cells having even numbers of supernumerary chromosomes. Such an expectation appears to be borne out by the results as shown in table i. TABLE Bivalent frequency amongst

Possible number of supernumerary bivalents per cell

No. of supernumeraries . . o i 2 3

Possible arrangemen . . . 7112711+112711+2 28112711+311 28"+ i

Observed frequencies . . . 482 303 70 253 5 90

Total possible numberofsuper- ...... 95 numerary bivalents

Observed frequencyofsuper- ...... 783 947 numerary bivalents per cent, per cent.

(iii) The frequency distribution of supernumerary chromosomes within and between individuals and populations Fourof the eleven individuals sampled from population K contained one or more supernumerary chromosomes. In the first of these individuals which were studied (K1—see table 2) about 8o per cent. of the cells contained a single supernumerary chromosome (plate, fig. 4) the same chromosome was present in the other three individuals K2, K6 and K7. Within each of these four individuals the number of supernumerary chromosomes per cell is not constant. In K1, K6 and K7 most of the spermatocytes contain a single B chromosome and the range of variation between cells is small and is similar in the three individuals (table 2). In K2, however, there is a much greater range in chromosome number and a higher modal value, 40 per cent. of the cells containing 5 supernumerary chromosomes (plate, fig. 8). The chromosome constitutions of population K and F were found to be rather similar, the same B chromosome is present in both populations, but fewer normal diploid individuals were found in population F (see tables 2 and 3). The actual frequency of individuals not containing B chromosomes in population K is 7 in I I, and in population F, i in 14, a difference which, using Fisher's (1950) exact test, is significant at the i per cent, level. SUPERNUMERARY CHROMOSOMES IN A SNAIL As in population K, in the individuals in population F which contain B chromosomes the number of extra chromosomes may vary between primary spermatocytes within an individual. This variation within individuals must arise through the unstable behaviour of the supernumerary chromosomes during the development of the organism. thesupernumerarychromosomes

2 3

4 5 6

2711+41 28"H-2' 29112711+51 2811+31 2911+ i2711+61 28+4 2911+21 3011

I 19 62 ... 2 32 ...... 6

164 68 27

872 971 889 per cent, per cent, per cent.

TABLE 2 The incidence of supernumerary chromosomes in individuals from population K *

No. of supernumerary chromosomes . Total no. Snasl of cells 0 I 2 3 4 6

K1 4 33 3 ...... 40 K0 ... 37 3 ...... 40 K7 8 i ...... 65 K2 ...... 7 13 19 20 1 50

* A total sample of eleven individuals was taken from the population ; seven of these did not possess supernumerary chromosomes.

It is not known whether mitotic instability leads to the elimination of B chromosomes from the somatic cells, as in Polycelis tenuis (Melander, 1950), but mitotic non-disjunction may occur either during the development of the organism as a whole or possibly only in the ovo- testes tissues. Mitosis in spermatogonial cells was studied but no examples of non-disjunction were found. The type of distribution of cell classes within individuals, see 134 H. J. EVANS tables 2 and 3, does not follow the same pattern for each individual. In two snails, F8 and F11, the frequency pattern fitted a binomial distribution for p= o5(F8, P4 =>o5and F11, P4 =>ox),whilst in other individuals irregular distributions and, in at least two cases (F4 and F10), bimodal distributions were found. In six out of the total of seventeen individuals which contained B chromosomes the

TABLE 3 The incidence of supernumerary chromosomes in individuals from population F *

No. of supernumerary chromosomes Total no. Snail of cells 0 I 2 3 4 6

F, 6 ...... 50 F1, 5 44 I ...... 50 F7 ... 50 ...... 50 F, 4 ... 46 ...... 50 F, ... 6 3 ...... 45 F1, ... 6 36 I ... , F, 4 4 37 4 1 ...... 50 F, 2 ... 58 8 2 ...... 70 F,4 i , 7 19 is 6 2 6o F11 2 8 14 19 7 I 3 54 F, I 2 12 23 14 5 I 58 F1, 4 2 26 5 i8 ... I 56 F4 ... I 27 2 19 I ... 50

* In one of the fourteen individuals studied, F,, no supernumerary chromosomes were observed. most frequent cell class consisted of cells having one extra chromosome; a mode at two supernumerary chromosomes per cell was found in seven individuals, and modes at three and five extra chromosomes in three and one individuals respectively (tables 2 and 3).

5.DISCUSSION Thehaploid chromosome number of n =27in individuals which do not contain B chromosomes, confirms the counts of Perrot and Perrot (i7) and of White andconforms with their contention that the species does not exhibit polyploidy. However, the presence of supernumerary chromosomes in the populations studied, results in a chromosome polymorphism which exists at two levels, between individuals and between cells within individuals. The significant difference between the low frequency of individuals which contain B chromosomes in population K and the high frequency in population F is interesting, especially as the supernumerary chromosome is the same in both populations. The difference between the populations may be the result of a number of causes such as differences SUPERNUMERARY CHROMOSOMES IN A SNAIL 135 in the time of introduction of the supernumerary into the populations and differences in selection pressure. No morphological differences could be detected between normal diploid individuals and individuals possessing B chromosomes, and we have no information indicating that the population difference can be ascribed to a particular factor or factors. The range of variation in the number of B chromosomes found between cells within individuals reveals a number of interesting features, perhaps the most important of which is the low frequency or often complete absence of normal diploid cells.It should be pointed out that the number of cells classed as normal (class o in tables i and 2) is a maximum estimate, for the B chromosome is small and although only cells with well separated chromosomes were scored, the extra chromosome could be obscured by the larger chromosomes of the complement. In animals which show inter-cell variation in chromosome number it seems plausible to interpret the chromosome constitution of the most frequent, or modal, cell class in each individual as representing the original chromosome constitution of the fertilised egg. In the absence of cell selection the range in variation found within individuals, and the frequency of the non-modal classes, would reflect both the time and the number of occasions on which mitotic non-disjunction occurred.However, the situation does not appear to be as simple as this as some of the evidence is not in contradiction with the suggestion that cell selection does in fact occur. For example in individuals F4 and F10, which show bimodal distributions, if 2fl+2B chromosomes represents the original chromosome complement of the fertilised egg, or the young ovo-testes, then to produce cells having 4 chromosomes would require at least two non-disjunctional events either (a) at one mitosis, producing daughter cells with o and 4 B chromosomes; or (b) at separate mitoses, the first non-disjunction producing cells with i and 3 B chromosomes and the second producing daughter cells with and 2 B chromosomes. The derivation of cells with 4 B chromosomes from cells which contain 2 B chromosomes, or vice- versa, might involve more than two non-disjunctional events, but in all cases if all cells had equal chances of becoming spermatocytes then there would be no markedly bimodal cell distributions of the type found in F4 and F10, where 8o to 90 per cent, of the cells have either 2 or B chromosomes. That the distribution of the various cell classes does not fit a simple pattern is not entirely unexpected, for factors such as proximity in relation to blood supply and the spatial arrangement of cells will be important to the mitotic potential and the mitotic rate of the various cell types. It is thus difficult to draw any firm conclusions about cell lineage from the data on the frequency of the various cell classes at the spermatocyte stage and it is not clear whether the modal cell class in each sample represents the original zygotic chromosome 12 136 H. J. EVANS complement or a selected line, although it should be noted that four different modal values have been found in the populations. Perhaps the most regular feature of the data is the low frequency or absence of normal cells in individuals which contain B chromosomes. Evidently such cells must be produced as a result of non- disjunction in a number of individuals, especially in those individuals which show a wide variation in cell type. The paucity of normal cells indicates that the presence of one or a few B chromosomes in a cell is not detrimental, but may in fact confer an advantage on the cell. What sort of advantage is gained is not known, but the presence of B chromosomes might for instance result in a decrease in the duration of the mitotic cycle giving a rate of proliferation which is greater than that of the normal diploid cell. Provided that the presence of B chromosomes in an individual does not place it at a disadvantage, then B chromosomes will be maintained in the population by virtue of their advantage over normal cells in the development of the gonads. A system such as this would be somewhat similar to that described by Roman (1948) in Maize, who found that B chromosomes showed a selective advantage at one stage in the life cycle of the organism, namely at fertilisation. H. pomatia L. is confined to chalky areas on the North Downs in this country whereas on the Continent the species is widely dis- tributed and is not confined to calcareous soils; the populations in this country are in fact at the periphery of the geographical range of the species. Although hermaphrodite, the species is cross-fertilising, but because of its very localised distribution in this country true outbreeding must be severely limited and a large degree of inbreeding must occur. It is possible that the supernumerary chromosome has arisen from chromosome structural change following instability of the genotype due to inbreeding. Such instability is known to occur in inbred lines of a normally outbreeding species, for example in Rye (Rees, 1955) and can be inferred from the development of structural hybrids which are found in normally outbreeding populations of plants (Walters, 1942; Cleland, '949) and animals (White, i'; Lewis and John, 1957) which have been forced to inbreed. No supernumerary chromosomes have been found in continental populations of H. pomatia L. (Perrot and Perrot, 1937; and White, 1954) and the occurrence of B chromosomes in the present populations may be the result of changes due to inbreeding. In this connection it is also of interest to note the work of Darlington (g), who suggested that differences found between the frequency of supernumerary chromosomes in wild populations and in mass cultures of Cimex lectularius might be attributed to inbreeding. Changes in the chromosome system which result in the maintenance of variation would be an advantage to a population existing at the margins of its geographical Plate All figs. except fig. 3 are X 1500. Fig. 3 is X I i6o. FIG. .—Meiotic metaphase, normal spermatocyte, 2711. FIG. 2.—Polyploid cell, 5611 including two supernumerary bivalents. Fio. 3.—Side VieW of Mi showing two cells each with a supernumerary chromosome. The B chroniosome in each cell is unpaired and lies off the metaphase plate. Fio. 4.—M I cell with one B chromosome, 2711 + . Fia.5.—Two cells at diakinesis each containing two B chromosomes. Left-hand cell shows a supernumerary bivalent, 2811 ; in the right-hand cell the B's are unpaired, 2711+21. Fio. 6.—Three B chromosomes, 2811 + i1. Fic. 7.—Four B chromosomes, 2g. Fin. 8.—Five B chromosomes, 2911 + . Fin.g.—Six B chromosomes, 2811+41. Fm. io.—M2 in a secondary spermatocyte. Note the single structure of the B chromosome indicating that the parent B chromosome underwent a mitotic separation at A1 of meiosis. IF,

ea S•e I-. s % ,'e.ta ate 4$ 'p. e1 s.3 j0

I 2 3 1 r0 t w % •. 0jS • • a S.. S S. •• S C 'S w5 S S • 4' 'S S 4 S p5• I 'S S 1' • S • S.. IC 'S... 6 8

a •$ 4!u! ir 'U.S. %. .,...

4 I0 SUPERNUMERARY CHROMOSOMES IN A SNAIL 137 range, and the B chromosomes in the snail may well be of importance in this respect.

6.SUMMARY i.A cytological study of the snails Helix aspersa MUll and Helix pomatia L., 2fl =54,was carried out in an attempt to obtain polyploid individuals for use in a radiobiological study. No polyploid animals were found although 1-2 per cent, of the primary spermatocytes in H. pomatia L. were found to be tetraploid. 2. Some of the individuals of H. pomatia L. which were sampled from two wild populations were found to contain from one to six supernumerary chromosomes. The proportion of individuals containing B chromosomes differed between populations. 3. Differences in the frequencies of B chromosomes were found both between individuals and, in many cases, between primary spermatocytes within individuals. The relative frequencies of normal diploid cells and cells containing varying numbers of B chromosomes within an individual, indicates that cells with B chromosomes may have an advantage over cells without B chromosomes during the development of the ovo-testes. 4. It is suggested that the presence of supernumerary chromosomes may be the result of inbreeding forced upon small populations of a normally outbreeding species existing at the limits of its geographical range. Acknowledg7nents.—I am indebted to Mr T. R. L. Bigger for his part in co]lecting the snails used in this investigation and for his efficient technical assistance in the preparation of the slides and to Mr T. F. J. Hobson for his advice on statistical matters.

7. REFERENCES

CONGER,A. D., AND FAIRCHILD, L. M. 1953. A quick-freeze method for making smear slides permanent. Stain Tech., 28, 281-283. CLELAND, R. E.1949. Phylogenetic relationships in Enothera. Proc. 8th mt. Congr. Genet., StockhoJm. Hereditas, suppl. 35, 173-188. DARLINGTON, C. D. 1939. The genetical and mechanical properties of the sex chromosomes. V. Cimex and the Heteroptera. J. Genet.. 39, 10 I -138. DARLINGTON, C. ID. 5956. Chromosome Botany. Allen & Tjnwin Ltd., London. FISHER, R. A.1950. Statistical Methods for Research Workers, ith edition. Oliver & Boyd, Edinburgh. HIJSTED, L., AND BURCH, r'. R. 1946. The chromosomes of Polygrid snails. Amer. Xat., So, 4 10-429. LEWIS, K. R., AND JOHN, B. 1957. Studies on Periplaneta Americana. II. Interchange heterozygosity in isolated populations. Heredity, ii, I 1-22. MELANDER, Y. 1950. Accessory chromosomes in animals, especially in Polycelis tenuis. Hereditas, 36, 19-38. PERROT, J. L., AND PERROT, M. 1937. La formula chromosomique de 1'Helix pomatia. Rev. Suisse Zool., 44, 203-209. REE5, II.1955. Genotypic control of chromosome behaviour in rye.i. Inbred lines. Heredity, 9, 93-116. 138 H. J. EVANS

REES,H.,AND JAMIESON, A. 1954. A supernumerary chromosome in Locusta. Nature (Lond.), 173, 43-44. ROBERTsON, w. R. B. 1916. Chromosome studies.i. Taxonomic relationship shown in the chromosomes of Tettigid and Acrididm. V-shaped chromosomes and their significance in Acridid, Locustid and Gryllidm : chromosomes and variation. J. Morph., 27, 179-331. ROMAN, H. 1948. Directed fertilisation in maize. Proc. Nat. Acad. Sci., 34, 36-42. STAIGER, H. 1954. Der chromosomendimorphismus beim prosobranchier Purpura lapillus in bezeihung zur okologie der art. Chromosoma, 6, 419-478. WALTERS, j.L.5942. Distribution of structural hybrids in Paeonia ca4fornica. Amer. J. Bot., 29, 270-275. WHITE, as. j.D.1954. Animal Cytology and Evolution. Cambridge: University Press. WHITE, M. p D. 5957. Cytogenetics of the Grasshopper Moraba scurra. Austral. J. 7ool., 5, 305-337.