Generating High Variability of B Chromosomes in Eyprepocnemis Plorans (Grasshopper)

Heredity 71 (1993) 352—362 Received 5 January 1993 Genetical Society of Great Britain

Generating high variability of B chromosomes in Eyprepocnemis plorans (grasshopper)

M. D. LOPEZ-LEON, J. CABRERO, M. C. PARDO, E. VISERAS, J. P. M. CAMACHO* & J. L. SANTOSI Departamento de Genética, Facu/tad de Ciencias, Un/versidad de Granada, E- 18071 Granada, Spa/n and tDepartamento de Genét/ca, Facultad de B/o/ogIa, Un/vers/dad Comp/utense de Madrid, E-28040 Madrid, Spa/n

Twenty-eightprogeny analyses (PAs) performed on specimens of E. plorans collected from four natural Iberian populations have been informative about the transmission of rare B chromosome types or the de novo origin of some of them. At least ii rare B-types have been found in addition to the predominant ones: B1 in Daimuz, B2 in Jete and Salobreña, and B5 in Fuengirola. The presence in two controlled crosses of one embryo carrying a B-type which was absent in the parents suggests that these B variants (B20 and B )haveoriginated de novo. Eleven other PAs suggest that new B derivatives are recurrently arising in these populations. The most frequent B chromosome mutation was centromere misdivision that originated four different B-types (B2m1,B110,B210 and Bmjnj). Other rearrangements were pericentric inversions (B211, B212 and B213), inverse tandem fusion (B211), centric fusion (B11) and deletions (B2d1andB2d2).Thefour B derivatives produced by centromeric misdivision are significantly eliminated during sexual transmission, most probably owing to deficiencies in the control of chromosome movement by their hemicentromeres. Those derived from translocations showed Mendelian transmission but deletion B variants showed a tendency to elimination. Our results suggest that B chromosome substitution of B1 by B2 in the Salobreña and Jete populations could be achieved by differences in relative transmission efficiency, as in one controlled cross, where the female carried 1 B1 plus 1 B2, B2 was significantly overtransmitted and B1 eliminated.

Keywords:Bchromosomes, coevolution, drive, genomic parasitism.

Introduction Arana, 1990) although three types are prevalent: B1, B2 and B5. B1 is the most frequent in all the populations Bchromosomes constitute one of the most intriguing of the eastern (from Tarragona to AlmerIa) and west- polymorphisms present in a large number of natural ern zones (Cádiz and Huelva) of distribution, and for populations of plant and animal species (Jones & Rees, this reason it has been considered the original B in the 1982). Some properties of Bs have been investigated, Iberian Peninsula (Henriques-Gil et al., 1984). B2 is the mainly from the cytological point of view, although main type in the province of Granada and eastern questions about their origin and their maintenance in Málaga, and B5 is restricted to western Málaga. natural populations remain largely unsolved. The evolution of Bs implies their origin from the One of the most striking B chromosome poly- chromosomes of the standard complement and subse- morphisms described to date is that shown by the quent differentiation to their present form. Further- grasshopper Eyprepocnemis plorans. Bs are present in more, in populations of a given species different but almost all natural populations from the entire geo- related B variants can also be generated and the graphical distribution in the Iberian Peninsula, i.e. the establishment and maintenance of these variants will Mediterranean and Atlantic coastal regions from depend on their transmission and their effects on the Huelva to Tarragona. About 30 different B variants fitness of B-carriers. Most Bs investigated to date show have been distinguished according to different accumulation mechanisms that may be enough to morphology and C-banding patterns (Henriques-Gil & explain their maintenance even though the Bs are detri- mental to carrier fitness (Nur, 1977). The three main *Correspondence. types of Bs of E. plorans constitute an exception to this VARIABILITY OF B CHROMOSOMES 353 general rule as they are transmitted in a Mendelian remaining one, both males were placed simultaneously ratio in populations where they are the most frequent B with the female and the frequency of mating exhibited variant: B1 at Daimuz (Valencia), B2 at Jete and by each male recorded, for which purpose one of the Salobreña (Granada) and B5 at Fuengirola (Malaga). males was marked on the pronotum with nail varnish. The possible causes of maintenance of this widespread The other 20 PAs came from females collected in the polymorphism have been discussed by López-León et field when they were gravid. The effective paternity a!. (1992). The present paper deals with the transmis- displacement existing in E. plorans (López-León et aL, sion rates of rare B variants that have appeared in the 1993)means that gravid females may be considered as populations mentioned above and demonstrates the de controlled crosses because the constitution of the novo origin of some of them and their possibilities of parental male can be inferred from the offspring with a maintenance in subsequent generations. high degree of confidence on the basis of Mendelian transmission exhibited by B1, B2 and B5 chromosomes. Materialsand methods PAs are coded as C (single crosses), DC (double crosses) or 9 (gravid females), followed by the code Specimensof the grasshopper E. plorans were number, the initial of the population (D, J, S or F) and collected at Daimuz (Valencia), Jete and Salobreña the year of sampling. (Granada) and Fuengirola (Málaga). A large number of progeny analyses (PAs) were performed on controlled Results crosses (single or double) and gravid females, from which 28 showed the transmission mode of rare B Figure1 shows the C-banding pattern of all types of Bs chromosome types or else the de novo origin of new B found in the progeny analyses (PAs) performed in this variants. All methods employed in the present investi- work. They are in addition to the standard complement gation are similar to those previously described of 22+ Xci/22 + XX9chromosomes. (Ldpez-León etal., 1991, 1992). The three main Bs show the following charac- Seven progeny analyses were obtained by means of teristics: single (2) or double (5) crosses in which a female was (1) B1 shows two similar sized C-bands on the long mated to one or two males, respectively. Two double arm, the light distal region being conspicuously larger crosses were made by substitution so that they are than the two C-bands as a whole. separated into two independent PAs, one for each (2) B2 in Daimuz also shows two C-bands of similar male, according to Ldpez-Ledn et al. (1991). In the size on the long arm but in the Granada populations

S I I $ I

B1 82 B B Bj 'L1.1 B1 B, 'IC) B BmBjjj B B B,

Fig. I Three main types of B chromosomes in E. plorans: B1 from Daimuz (Valencia), B2 from Daimuz (the first), B2 from Salo- brefla and Jete (Granada) (the second) and B5 from Fuengirola (Málaga) together with 13 other B variants found in these populations. Bmjnj of at least two different sizes were found, which indicates that this is a heterogeneous category. 354 M. D. LOPEZ-LEON ETAL. the interstitial band is larger than the proximal one. (7) B111 is the result of a centric fusion between two B1 The main difference between B2 and B1 is the smaller chromosomes. size of its light distal region. (8) B210 is an isochromosome derived from the B2 by (3) B5 is an acrocentric chromosome similar in size to means of centromere misdivision and subsequent but with a much larger C-band near the centromere chromatid non-disjunction. and smaller light distal region. (9) B2111 is the result of an inverse tandem fusion In addition to these three main types, 12 other between two B2 chromosomes with the loss of the derivatives have appeared in the four populations centromere and the short arm in one of them. sampled. The number of B variants is very high (more Finally, one B variant appeared in Fuengirola than 40, with those described in the present paper) so besides B5 and Bmjnj: B51 which shows the C-banded that it is necessary to codify their names properly in region located at the end of the long arm as a result of order to avoid confusion. We propose that each B an intrachromosomal translocation. variant should be named by the number of the main B from which it is presumed to have been derived, followed by the initial of the chromosomal mutation Inheritanceof rare B chromosome variants from which it arose and an order number. For example, Theresults of all the progeny analyses are shown in B212 means that this B type is derived from the B2 Table 1. Details concerning them are the following. through an inversion and that it is the second B variant risen from this type of rearrangement. B1, B21711 and B,,0. Previous reports (Henriques-Gil et Besides B1 and B2, in Daimuz two other B variants al., 1984; Henriques-Gil & Arana, 1990) indicated appeared: that B1 was apparently absent from the provinces of (1) B110, which is an iso B1 derived from centromere Granada and eastern Málaga. The two populations misdivision and subsequent chromatid non-disjunc- analysed by us in this area (Jete and Salobreña) have tion. This B type was originally described as B12 by shown the presence of this type of B although in a low Henriques-Gil etal. (1984). number of individuals. Thus, seven of our PAs have (2) Bmjnj, which is small, about the size of the short arm permitted the analysis of the mode of inheritance of plus a hemicentromere, and has also arisen through this chromosome in four of which it was transmitted centromere misdivision. This B-type must be hetero- simultaneously with another B variant (B2m1 or B2) geneous as mini Bs can be derived from many other B through the same parent. variants, specifically from any Bs possessing a short In 943J90, the female lacked Bs and the male pre- arm. sumably possessed lB1 which was slightly, but not Nine other B-types have appeared in Granada significantly, undertransniitted (transmission rate, populations besides B1, B2 and Bmini: K=0.364). The 931J90 female carried 1 B2 and the (1) B2m1 is a B2 that has lost the short arm as a conse- inferred male lB1 +1B2,both Bs being transmitted quence of centromere misdivision. It could also have according to a Mendelian ratio. The 924S90 was the arisen from deletion of the short arm but the propen- third progeny analysis in which the male parent pre- sity of B2 to suffer centromere misdivision generating sumably carried a B1 accompanied, in this case, by the B210 and Bmjnj chromosomes with low transmission B2m1. Both Bs were significantly eliminated. rates, as we will see later, indicates that B2m1 is most The four remaining progeny analyses showed B1 probably derived from misdivision as it also shows a transmission by the female side. The female 96S87 marked tendency to elimination. carried lB2that was binomially transmitted (2) B211 shows part of the light distal zone of the long (K =0.559). Its progeny also showed the presence of arm located on the short arm as the result of a pericen- one B210 in 35.6 per cent of the embryos so that it may tric inversion. be inferred that the male parent probably carried 1 B2110 (3) B212 shows the entire light distal region of the long which was significantly undertransmitted. The three arm located on the short arm. remaining progeny analyses correspond to two double (4) B23 in which the proximal C-band is very small and crosses in which the female parent carried lB1 plus the remaining long arm has changed its order, C band lB2. In DC1OJ89a, the male parent carried 1B2, so that material being distally located and the light zone being the mean numbers of B2s (1.024) and B1s (0.595) in the interstitially located. offspring were consistent with Mendelian expectations (5) B2d1isa B2 without the light distal region which has (1 and 0.5,respectively).In DC1OJ89b the same been deleted. female as in the preceding PA, now crossed to a OB (6) B2d2 has lost the interstitial C-band and part of the male, again transmitted her B1 and B2 in rates consist- light region. ent with binomial expectations (0.523 and 0.538, Table I Transmission and de novo origin of 12 B chromosome variants derived from the three main types (B1, B2 and B5) in three Spanish naturalpopulations of the grasshopper Eyprepocnemisplorans

No. of Bs in parents Progeny analysis 9 d Embryo/eggs Embryo offspring analysed Total x8 z

C5J88 lB2 1B2+ lB2d2 42/54 5(OB):9(1B2):2(1B2d2):11(2B2):6(1B2+ 1B2d2): 41 B2:1.293 1.88 8(2B2+ 1B2d2) B2d2:0.390 —1.41

C26S89 2B2 lB2 + lB212 62/67 1(OB):4(1B2):14(2B2):8(1B2+ 1B212):8(2B2 + 1B212): 39 B2: 1.744 1.01 3(3B2 + 1B212):1(3B2) B212: 0.487 —0.16

DC4J88a 1B2+ lB2dI OB 94/96 9(OB):12(1B2):17(lB2dl):14(1B2+ lB2dl) 52 B2: 0.500 0 B2d1: 0.596 1.39 DC4J88b 1B2+ lB2dl lB2 55/57 7(OB):10(1B2):8(lB2dl):17(1B2+ lB2dl):3(2B2): 50 B2: 0.860 —0.99 5(2B2 + lB2dl) B2d1: 0.600 1.41

DC1OJ89a lB2 + lB1 lB2 48/50 3(OB):1 1(1B2):5(1B1):5(2B2):1(2B1):1 1(1B1 + 1B2): 42 B2: 1.024 0.16 5(2B2 + 1B1):1(2B1 + 1B2) B1: 0.595 1.23 DC1OJ89b 1B2+ lB1 OB 88/99 8(OB):23(1B2):21(1B1):12(1B1 + 1B2):1(1R210+1B1) 65 B2:0.538 0.61 B1: 0.523 0.37

DC3J9O lB1 + lB2 OB 81/86 19(OB):14(1B1):27(1B2):12(1B1 + 1B2):9(2B2) 81 B2: 0.704 lB2 B1: 0.321

Parthenogenic 36/54 5(OB):8(1B1):14(1B2):7(1B1+ 1B2):1(1B111+ 2B2)j 35 B2: 0.629 1.53 B1: 0.457 —0.51 Total 117/140 24(OB):22(1B1):41(1B2):19(1B1 + 1B2):9(2B2): 116 B2:0.629 2.77* — 1(1B111 + 2B2) B1: 0.362 2.97*

96S87 lB2 1B21, 63/66 19(0B):18(1B2):1(2B2):8(1B210):13(1B2+1B210) 59 B2: 0.559 0.91 B21: 0.356 _2.21* > 93S90 lB2 2B2 40/70 4(OB):12(1B2):12(2B2):8(3B2):1(1B2+1B210) 37 B2: 1.649 0.60

924S90 lB2 lB1 + lB2mi 73/76 18(OB):24(1B2):5(1B1):6(lB2mi):9(1B2+ lB2mi): 70 B2: 0.614 1.91 6(1B1 + 1B2):2(2B2) B1: 0.157 574* — B2mi: 0.214 479* B1 + B2mi: 0.37 1 _2.16*

938S90 3B2 2B2 25/33 2(1B2):10(2B2):1 1(3B2):1(1B2 + 1B2111) 24 B2: 2.333 —0.32

943S90 2B2 lB2 or 2B2 38/38 l0(1B2):20(2B2):3(3B2):l(2B2+ 1B20) 34 B2: 1.794 1.14 —0.60

962S90 2B2 2B2 + lB2dl 30/30 4( 1B2):1 3(2B2):5(3B2):2(4B2):1( lB2dl): 30 B2: 2.133 0.36 — 2(2B2+ lB2dl):2(3B2+ lB2dl):1(1B2 + 1B211) B2d1: 0.167 3.65*

927S91 lB2 1B2+ lB213 64/67 8(OB):14(1B2):11(1B213):16(1B2+ 1B213): 61 B2: 0.885 —0.89 9(2B2):3(2B2 + 1B213) B213: 0.492 —0.13 Table 1 Continued (9

No. of Bs in parents Progeny analysis 9 d Embryo/eggs Embryo offspring analysed Total Z p 0'I- -i:J 39 1.154 0.96 rn 97J90 lB2 lB2 40/40 5(OB):23(1B2):9(2B2):1(3B2):1(1B2+ 1B212) B2: NI + + 36 0.944 —0,34 931J90 lB2 lB1 lB2 37/37 4(OB):11(1B2):3(1B1):4(2B2):13(1B1 1B2): B2: 0'm 1(2B2+1B1) B1:0.472 —0,34 942J90 iBmini lB2 20/20 7(OB):9(1B2):1(lBmjn) 17 B2:0.529 0.24 —I — Bmini: 0.059 3.64* 943J90 OB lB1 35/35 21(OB):12(1B1) 33 B1: 0.364 —1.56 947J90 2B2 lB2 39/39 3(0B):16(1B2):15(2B2):4(3B2):1(1B2+1B210) 39 B2: 1.513 0.05

955J90 3B2 OB 36/36 5(OB):11(1B2):17(2B2):2(3B2):1(1B2+ 1Bmnj) 36 B2: 1.444 —0.23 986190 OB 2B2+1B211 40/40 1(0B):10(1B2):2(2B2):6(1B21):9(1B2+1B211): 33 B2:0.939 —0.36 4(2B2+ 1B211):1(1B210+ iBmini) B21: 0.576 0,87 9109J90 iBmini OB 51/51 41(OB):6(lBmn) 47 B0.128 5.10' 9115J90 2B2 2B2 45/45 3(OB):10(1B2):24(2B2):7(3B2):1(2B2+ 1B210) 45 B2: 1.800 —0.67 — 9116190 lB2 + lB211 OB 45/45 3(OB):15(1B2):19(1B21):5(1B2 + 1B21) 42 B2: 0.476 0,31 B21: 0.571 0.92

923F88 lB5 lB511 35/36 7(0B):10(1B5):10(1B51):8(1B5+ 1B51) 35 B5: 0.514 0.17 B511:0.514 0.17

912F89 OB lB5 48/48 21(OB):26(1B5):1(lBm,n,) 48 B5: 0.542 0.56 918D89 013 1B1+1B2+ 88/90 58(OB):6(1B1):5(1B2):7(lBijso):9(lBmjnj): 88 B1:0.080 — 1B110 + 1Bminj 1(1B1 + 1B2):1( 1B110+ lBmini):1(1B2 + 1B1151,) B2: 0.080 B110:0.102 — — Bmini:O.ll4 925D89 lB1 lB1 34/34 10(0B):15(1B1):8(2B1):1(1B110) 34 B1:0.912 —0.51

Z was calculated by the approximation of binomial to normal distribution. For this purpose, the transmission rate per B chromosome (the quotient between XB for a B variant and the total number of Bs of such a type present in the parents) was the parameter tested, and the expected value was 0.5, as is predicted by Mendelian segregation. B variants presumably arisen de novo are noted in bold-type letters. XB. mean numberof B chromosomes. tIn the parthenogenetic offspring of DC3J9O, frequencies of ova produced by the female (as deduced from the haploid cells in the embryos) are expressed instead of embryo frequencies, with the only exception of the last embryo which was diploid and mosaic for the presence of 2B1 + 2B2 or else 1B1f1 + 2B2 (see Figs 2 and 3). For calculations, however, this embryo was considered as derived from an ovum with lB1 + lB2. tThis figure results from the assumption that females transmitted B2 to all their ova in the ratio observed for parthenogenetic embryos. §In this PA it is uncertain whether the male parent possessed lB2 (Z 1.14, P>0.05) or 2B2 (Z —0.60, P>0.05). §Male parent was presumably a rare mosaic for the four types of B chromosomes, so that it is not clear what is the appropriate hypothesis to calculate Z. VARIABILITY OF B CHROMOSOMES 357 respectively). The progeny reflects exactly the compo- and B2d1 were consistent with Mendelian expectations sition of the ova produced by the female: 8 (OB): 21 (Table 1). The other progeny analysis was 962S90 in (1B1): 23 (1B2): 13 (lB1 + 1B2) as the male lacked Bs. which the female carried 2B2 and, on the basis of The last class included one embryo with lB + 1B20, offspring karyotypes, the male was inferred to possess which will be considered later. These figures allow us 2B2 plus lB2dl. Thus, the transmission rate (0.167) of to infer the segregation of these B-types during female B2d1wassignificantly lower than expected. meiosis. If they segregate at random, similar frequencies of the four types of embryos would be expected B2d2.Theonly progeny analysis for this B chromosome (16.25). A Chi-square test showed that B1 and B2 did type was performed by means of a single cross, C5J88, not segregate randomly during female meiosis in which the female carried lB2 and the male lB2 plus (z2=9.°3, d.f.=3, P=0.01—0.05), owing to an excess lB2d2. This cross showed a tendency to an accumula- of migrations to opposite poles (44 embryos with lB1 tion of B2 and elimination of B2d2 although the results or 1 B2) to the detriment of migrations to the same pole did not reach statistical significance (Table 1). (21 embryos with lB1 + lB2 or without Bs). This indicates that B1 and B2 segregate in 67.69 per cent of the B211. Two PAs, 986J90 and 9116J90, showed first meiotic divisions, as would be expected if B1 and Mendelian transmission of this B chromosome variant B2 formed a bivalent in 35.38 per cent of cells, and this through both sexes. determined their migration to opposite poles. The last progeny analysis in which B1 transmission B22. One PA, C26S89, showed that the transmission could be analysed is DC3J9O. As Table 1 shows, B1 mode of this B type by the male side was consistent was carried exclusively by the female and the mean with Mendelian expectations. number of B1 chromosomes among the progeny (0.362) showed significant elimination through the B213. One PA, 927S91, showed Mendelian trans- female(Z =—2.97,P<0.05). B2, however, was carried mission rate of this B variant by the male side. by the female and also by one of the two males placed with her. Both males copulated with the female and so B511. The female 923F88 carried lB5 that was bino- paternal B2 inheritance cannot be analysed with confi- mially transmitted. In the offspring one B511 also dence. Maternal B2 transmission, however, could be appeared, so it can be inferred that the male parent ascertained because the two latter egg-pods laid by the carried it, and that this B variant was transmitted female contained 35 haplo-diploid embryos produced according to Mendelian rules. by accidental thelytoky. The number of B2 chromosomes in the haploid cells of these embryos should De novoorigin of B chromosome variants correspond to the number in ova (0.629, see Table 1). If this figure were assumed to be valid for the 116 ova ThirteenPAs have demonstrated the appearance of B produced, then this transmission ratio showed signifi- chromosome types that were not present in the cant B2 accumulation (Z=2.77, P<0.05). Hence, this parents, or else their presence was unlikely. Two of female showed significant tendencies to B2 accumula- them were double crosses in which the new type was tion and elimination. absent from the female and the two males crossed with her. Thus, one embryo out of 107 yielded by the Bmjnj. Two progeny analyses (942J90 and 9109J90) DC1OJ89 female carried 1B2,0 chromosome (plus comprising a gravid female and her offspring showed 1B1) which most likely arose by centromere misdivi- the transmission of the Bmini chromosome. Both sion and chromatid non-disjunction of the maternal B2. females, but a significant minority of her offspring, A female embryo mosaic showing cells with either two carried the Bminichromosome(Table 1). As a whole, B1 chromosomes (plus 2B2) (Fig. 2) or with lB111 (plus only 7 of 64 embryos analysed from both females 2B2) (Fig. 3) was found among the offspring produced carried the Bmjnj so that its mean transmission ratio by parthenogenesis in DC3J9O. B111 had a larger centro- (0.109) was significantly lower than the expected one mere region than the B210 (compare both in Fig. 1), (Z= —6.26, P<0.05). suggesting an origin by centric fusion of two B1 chromosomes. This embryo presumably came from an B2d1.Threeprogeny analyses showed the transmission ovum with IB! plus lB2 which was subsequently mode of this chromosome. Two of them corresponded diploidized, the centric fusion arising in one of the first to the DC4J88 cross in which the female carried lB2 diploid cells produced. plus lB2dl, the first male lacked B chromosomes and We can be confident in these two PAs of the de novo the second male carried lB2. Transmission rates of B2 originof both B types whereas in the remaining 11 this 358 M. D. LOPEZ-LEON ETAL.

I 4 * p S ala'• 5;',' • a.a • as 4 55 S 4 Figs 2—4 Mitotic metaphase cells of S • S •2 • B2j80 V embryos of E. plorans. Fig. 2 Cell • S •. from a mosaic embryo carrying V. S • I 2B1 +2B2chromosomes. Fig. 3 Cell SS. • from the same embryo as in Fig. 2 showing 1B1f1 +2B2chromosomes. Fig. S 4 Cell from an embryo possessing a 4 1B2,() plus lBmini (arrowhead).

Table2 Frequency of different types of B chromosomes found in embryos yielded by gravid females collected at Jete (55 females) and Salobreña (29) populations. Mutation rates have been calculated assuming that a single embryo in a PA carrying a B type that was not present in the mother, has arisen de novo

Mean number of Bs in

B type Population Embryos N Adults N Mutation rate

B1 Jete 0.0139 2091 0.0182 110 — Salobreña 0.0119 921 0.0172 58 — B20 Jete 0.0014 2091 0 110 0.0014 Salobreña 0.0022 921 0.0172 58 0.0021

Bmini Jete 0,0042 2091 0.0182 110 0.0009 Salobreña 0 921 0 58 —

B21 Jete 0.0202 2091 0.0182 110 — Salobreña 0.0010 921 0 58 0.0010 B22 Jete 0.0005 2091 0 110 0.0005 Salobreña 0 921 0 58 — — B2d1 Jete 0 2091 0 110 Salobreña 0.0052 921 0.0172 58 — — B2111 Jete 0 2091 0 110 Salobrefia 0.0010 921 0 58 0.0010

B2m1 Jete 0 2091 0 110 — Salobreña 0.0098 921 0.0172 58 — possibility has to be inferred because we did not common origin by means of the same process of analyse the male parent. 93S90, 943S90, 947J90 and centromere misdivision and subsequent non-disjunc- 911 5J90 showed the presence of a single embryo with tion of long arm chromatids to constitute the B20. one B210. On the other hand, one B110 was found in a A single embryo carrying iBmini was observed in the descendant from 925D89 (Table 1). Significantly, in offspring from 955J90 and 912F89. B211 and B22 each 986J90 one embryo carried two very infrequent types appeared in a single embryo from 962S90 and 97J90, of Bs (1B20 plus IBmini) (Fig. 4). The probability of respectively. The case of 938S90 is remarkable where carrying both Bs from independent origins is one embryo carried a B variant (B2111) that has never 0.0014x0.0042=5.88x 10-6 (see Table 2), so that been found in adults, so it is likely to have arisen de the expected number of embryos (among the 2091 novo by telomere fusion of two B2 chromosomes analysed in Jete) carrying both types of Bs is 0.0 125. It during male or female gametogenesis. is likely that in this embryo B20 and Bmjnj have a VARIABILITY OF B CHROMOSOMES 359

The progeny from 91 8D89 are of special interest as B211 B2 four types of Bs (B1, B2, Bmjnj and B1,0) have been found (Table 1). The female parent was a non-B carrier so they must, undoubtedly, have been transmitted by the male side. The presence of lB1 plus lB2 in the same embryo indicates that both Bs must be present in Fl flb.H the same male parent. The same deduction can be ______rJBm made from the observation of one embryo possessing f 1B10 plus lBmjnj and from another embryo with lB2 Bz B2m1 plus 1B10. Consequently, it is likely that the four B B types were carried by a single mosaic male parent. In this case, it was not clear what transmission hypothesis must be tested because we did not know the proportion of germ cells in this male carrying each type of B Bidi chromosome. B21t1 Discussion Fig. 5 Chromosomal rearrangements involved in the generation of nine B chromosome variants derived from B2 Someprogenies shown in Table 1 constitute a clear in natural populations of the province of Granada. In all demonstration of de novo origin of B chromosome cases the origin of the arrows point to the presumed variants as they were carried by a single embryo in breakage points on the B2 chromosome, a: Centromere PAs whereas the parents did not carry them. These misdivision; b: chromatid non-disjunction; c and d: observations suggest both that new B-types are arising pericentric inversions; e: paracentric inversions; f: inverse tandem fusion; g and h: deletions. (i.e. B211 and B22 that were present only in embryos) and that some types of Bs are recurrently arising in each generation (i.e. B10, B210, Bmjj and B211 that were ficant that all chromosomal mutations affecting Bs present in both embryos and adults). This indicates depend, as a whole, on breakage at seven main points: that the B chromosome polymorphism in E. plorans is the short arm, the centromere, a region in the proximal being fed by the recurrent formation of new B variants C-band, the secondary constriction, the joining with rates ranging from 0.0005 (B22) to 0.0021 (B21,0) between the interstitial C-band and the light distal (Table 2). This is also supported by the appearance in a region, and the mid and distal zones of the latter (Fig. gravid female from the Salobreña population (927S9 1) 5). of a B variant (B213) that was not observed in extensive Several chromosome rearrangements have also been samplings of this same population in precedent genera- proposed to explain Bs' diversity in other species. For tions. In plants, the origin of new types of Bs from con- instance: (i) B derivatives could originate as deletions trolled crosses and experimental populations has also and centric misdivisions in the plants Centaurea been reported (Bougourd & Parker, 1979; Holmes & scabiosa (Frost, 1957), Secale cereale (Lima-de-Faria, Bougourd, 1991; M. Jiménez, personal communica- 1963), Aster ageratoides (Matsuda, 1970), Hypochoeris tion). maculata (Parker, 1976; Parker et al., 1982), Allium Henriques-Gil et at. (1984) suggested that deletions, schoenoprasum (Bougourd & Parker, 1979) and Crepis duplications and misdivisions are the most likely capillaris (Parker et at., 1989). Remarkably, the rate of mechanisms responsible for B chromosome poly- centromere misdivision giving rise to iso Bs reported morphism in E. plorans. In this work, the most frequent in C. scabiosa (1.76x l0 to 2.6X103) and H. chromosomal mutation yielding new types of Bs was maculata (1.33 x 10 to 1.86>< l0) was very similar centromeric misdivision that gave rise to B11,0, B21,0, to those we have observed in E. plorans (1.4 x 10 to Bmjnj and B2mi (Fig. 5). Also relevant are putative inver- 2.1 x 10 for B21,0; Table 2). (ii) Three break inser- sions (B21, B212 and B213), inverse tandem fusion (B211), tions which are indistinguishable from inversions and centric fusion (B1f1) and deletion (B2d1andB2d2). addition or deletion of bands could be responsible for This indicates that the Bs of E. plorans seem to the differences found among some of the 23 distinct be prone to undergoing any type of chromosomal G-banding B variants observed in the grasshopper rearrangement. Translocations between Bs and Chortoicetes terminifera (Webb & Nehaus, 1979). (iii) members of the standard chromosome complement C-bandingpatterns displayed by Bs of the rodents have been reported in natural populations (Henriques- Perognatus baileyi (Patton, 1977) and Uromys Gil et at., 1983; Cabrero et al., 1987). It is highly signi- caudimaculatus (Baverstock et at., 1982) are consistent 360 M. D. LOPEZ-LEON ETAL. with an evolutionary model in which additions or dele- The contrary situation (accumulation of B1 and elimi- tions in the C-banded regions could play an important nation of B2) has not been observed as the female role. As Jones & Rees (1982) have pointed out, the DC1OJ89 transmitted B1 and B2 at Mendelian rates. existence of different types of Bs within a given species Consequently, the substitution of B1 by B2 in the is not surprising on the grounds of their relative 'inert- Salobreña and Jete populations could be based on trans- ness' because we might expect a greater tolerance to mission efficiency differences between both variants, B2 different chromosome rearrangements. It has recently being favoured. If a B chromosome substitution is effect- been suggested (Green, 1990) that a mechanism like ively taking place, and populations are near the com- Muller's Ratchet could account for the evolution of plete substitution of B1 by B2, we can deduce that even some Bs as it would allow the accumulation of dele- hypothetical improvements of individual fitness exerted tions and heterochromatinizations. by B1 will not be enough to counteract its elimination At this point, we may speculate about the scarcity of during transmission. The evolution of this B chromosome B chromosome variants in natural populations, some polymorphism at Málaga and Granada provinces for instance B liso'B250and Bmjnj, which are recurrently will be the purpose of future study. formed in each generation. The maintenance of newly For B chromosome polymorphisms that appear to arisen Bs depends mainly on their efficiency during be at or near equilibrium, it is generally accepted that sexual transmission and their effects on carrier fitness. selective forces acting against individuals with a high Whereas the latter is unknown, it seems clear that Bs number of Bs are counteracted by cytological accumula- with a hemicentromere, as those which resulted from tion mechanisms (Shaw & Hewitt, 1990). In these situa- misdivision, show a very high tendency to elimination tions mutations decreasing the transmission of the B will probably from a loss in the efficiency of the hemi- be favoured. In fact, the existence of genes in the centromere to control chromosome movement at A genome suppressing meiotic drive of Bs has been meiotic anaphases during gametogenesis and even at reported in the grasshopper Myrmeteotettix macu/atus mitotic anaphase during zygote cleavage divisions. At (Shaw et at., 1985), the mealy bug Pseudococcus affinis present, the reason for the tendency to eliminate Bs (Nur & Brett, 1988) and rye (Romera etal., 1991). On with deletions affecting the light distal region on the these grounds, the B chromosome polymorphism dis- male side is not evident. It is conceivable that it may be played by Eyprepocnemis plorans is quite unusual related to some type of unknown function of the lost because the three main variants (B1, B2 and B5) showed chromatin as the relocation of this region on the short a Mendelian transmission ratio, at least in the central arm (B21 and B22) or on the end of the long arm (B511) areas of their distribution (López-León et at., 1992). does not affect the characteristic Mendelian inherit- One of the most feasible hypotheses to explain the ance of B2 and B5, respectively. The increase in fre- existence of the widespread B chromosome poly- quency of those B variants lacking accumulation morphism in E. plorans is the existence of drive in mechanisms is difficult unless they improve the fitness the beginning (López-León et at., 1992). If drive of individuals carrying them, a fact that is unknown at suppressor genes were to have evolved in this species, present. the current situation of the polymorphism might be The case of the B1 chromosome found in Granada explained by a majority implantation in different province merits special mention. Most observations populations of allelic variants in A chromosomes suggest that B1 is the original B chromosome in the responsible for the Mendelian transmission ratio Iberian Peninsula from which the remaining Bs have observed at present. In any case, a certain coadaptation derived, including some types (i.e. B2 and B5) that are between A and B genomes appears to have been much more frequent than B1 in certain localities achieved. Thus, data from the present work indicate (Henriques-Gil et at., 1984). This indicates that a sub- that differences in transmission rate could play an stitution of Bs must have taken place in at least two important role in the substitution of B1 by B2 in regions, namely in Fuengirola (western Málaga) where Granada populations, as postulated by Henriques-Gil B5 is the predominant type and the province of et al. (1984). Moreover, other different B variants are Granada where B2 has substituted B1 (see for discus- continuously arising, perhaps as a coevolutionary sion Henriques-Gil & Arana, 1990). In Granada pro- response to the challenges imposed by the host vince, showed a trend to elimination by both sexes genome, thus being a reflection of the relationships that although it was only significant in two of the seven PAs take place between A and B genomes. Whereas most of studied. The most remarkable fact is that in the female these new B variants will disappear from the popula- DC3J9O (lB1 +1B2),B2 was accumulated and B1 elimi- tions in which they originate, owing to their low trans- nated, although total B transmission by this female was mission rate (Table 3), the future of others showing that expected for a 2B carrier (0.362+0.629=0.991). VARIABILITY OF B CHROMOSOMES 361

Table 3 Transmission ratios observed for 12 B chromosome Table 3 Continued variants Transmitted Progeny Transmission Transmitted Progeny Transmission B type by analysis ratio B type by analysis ratio B2d2 d C5J88 0.390 B1 924S90 0.157 931J90 0.472 B210 d 96S87 0.356 943J90 0.364 B511 d 923F88 0.514 Total 0.33 1 9 DC1OJ89a 0.595 DC1OJ89b 0.523 DC3J9O 0.362 Mendelian transmission ratio is at present unpredict- Total 0.49 3 able. Another situation is worthy of mention in which B d+9 925D89 0.456 chromosomes could avoid the influence of the A B2 d 942J90 0.529 genome, namely their integration in this genome by 986J90 0.470 means of centric fusions with regular members of the Total 0.500 complement. This type of rearrangement has been 9 DC4J88a 0.500 reported to occur spontaneously in natural populations DC1OJ89b 0.538 of E. plorans (Henriques-Gil et a!., 1983; Cabrero et 96S87 0.559 a!., 1987) and has recurrently appeared in the last 924S90 0.614 years either in adult individuals or in single embryos 955J90 0.481 among offspring of crosses where the parents did not 9116J90 0.476 show such chromosomal mutation (M. D. López-León Total 0.528 et a!., unpublished data). In fact, Bs fused to an auto- C5J88 0.647 some do not show any cytologically detectable methyl- C26S89 0.581 ation, in contrast to free Bs, which produce the . DC4J88b 0.430 reactivationof a silent nucleolar organizer region 0.512 DC1OJ89a present in the Bs (López-León eta!., 1991). Therefore, 93S90 0.550 both the generation of new variants and the integration 938S90 0.476 962S90 0.533 in the A genome are the main evolutionary strategies 927S9 1 0.443 displayed by B chromosomes in natural populations of 97J90 0.577 E. plorans. 931J90 0.472 947J90 0.504 9115J90 0.450 Acknowledgements Total 0.515 Weare indebted to Mrs J. Stocco for corrections to the B5 d 912F89 0.542 English. This study was partially supported by grants from the Dirección General de Investigación CientIfica 9 923F88 0.514 y Técnica (no. PB9O-0865) and the Plan Andaluz de Bmjnj 9 942J90 0.059 Investigación, Grupo no. 3122 (Spain). 9109J90 0.128 Total 0.094

B2m1 d 924S90 0.214 References

B211 d 986J90 0.576 BAVERSTOCK, P.R., GELDER, M. AND JAHNKE, A. 1982. Cytogenetic 9 91 16J90 0.571 studies of the Australian rodent Uromys caudimaculatus, a species showing extensive heterochromatin variation. d C26S89 0.487 B212 Chromosoma, 84, 5 17—533. B23 d 927S91 0.492 BOUGOURD, S. M. AND PARKER, j.s.1979. The B chromosome system of Allium schoenoprasum. II. Stability, inheritance d 962S90 0.167 B2d and phenotypic effects. Chromosoma, 75, 369—383. 9 DC4J88a 0.596 CABRERO, J., ALCHE, J. D. AND CAMACHO, J. P. M. 1987. Effects of B DC4J88b 0.600 chromosomes of the grasshopper Eyprepocnemis plorans Total 0.598 on nucleolar organiser regions activity. Activation of a 362 M. D. LOPEZ-LEON ETAL.

latent NOR on a B chromosome fused to an autosome. NUR, u. 1977. Maintenance of a parasitic' B chromosome in Genome, 29, 116—121. the grasshopper Melanoplus femur-rubrum. Genetics, 87, FROST, S. 1957. The inheritance of the accessory chromo- 499—5 12. somes in Centaurea scabiosa. Hereditas, 43, 403—422. NUR, U. AND BRETF, B. H. L. 1988. Genotypes affecting the GREEN,D. M. 1990.Muller's Ratchet and the evolution of condensation and transmission of heterochromatic B supernumerary chromosomes. Genome, 33, 8 18—824. chromosomes in the mealy bug Pseudococcus affinis. HENRIQUES-GIL,N. ANDARANA, P. 1990. Origin and substitution Chromosoma, 96,205—212. of B chromosomes in the grasshopper Eyprepocnemis PARKER, .i. s. 1976. The B chromosome system of Hypochoeris plorans. Evolution, 44,747—753. maculata. I. B-distribution, meiotic behaviour and inherit- HENRIQUES-GIL, N., ARANA, P. AND SANTOS, J. L 1983. Sponta- ance. Chromosoma, 59, 167—177. neous translocations between B chromosomes and the PARKER, J. 5., JONES, G. H., EDGAR, L. AND WI-IITEHOUSE, c. 1989. normal complement in the grasshopper Eyprepocnemis The population cytogenetics of Crepis capillaris. II. The plorans. Chromosoma, 88, 145—148. stability and inheritance of B chromosomes. Heredity, 63, HENRIQUE5-GIL, N., SANTOS, J. L. AND ARANA, p1984.Evolution 19—27. of a complex polymorphism in the grasshopper Eyprepoc- PARKER, J. S., TAYLOR, S. AND AIN5wORTH, c. c. 1982. The B nemis plorans. Chromosoma, 89, 290—293. chromosome system of Hypochoeris maculata. III. Varia- HOLMES, D. S. AND BOUGOURD, S. M. 1991. B-chromosome selec- tion in B-chromosome transmission rates. Chromosoma, tion in Alliuni schoenoprasum. II. Experimental popula- 85, 229—3 10. tions.Heredity, 67, 117—122. PATrON, .J. L. 1977. B chromosome systems in the pocket JONES,R. N. AND REES,H. 1982. BChromosomes.Academic mouse, Perognathus baileyi: meiosis and C-band studies. Press, London. Chromosoma, 60,1—14. LIMA-DE-FARIA, A. 1963. The evolution of the structural ROMERA, F., JIMENEZ, M. AND PUERTAS, M. j.1991.Genetic control pattern in a rye B chromosome. Evolution, 17, 289—295. of the rate of transmission of rye B chromosomes. I. LOPEZ-LEON, M. D., CABRERO, J. AND CAMACHO, J. P. M. 1991. Effects in 2B x OB crosses. Heredity, 66,61—66. Meiotic drive against an autosomal supernumerary seg- SHAW, M. W. AND HEwITF, G. M. 1990. B chromosomes, selfish ment promoted by the presence of a B chromosome in DNA and theoretical models: where next? In: Futuyma, D. females of the grasshopper Eyprepocnemis plorans. and Antonovics, J. (eds) Oxford Surveys in Evolutionary Chromosoma, 100,282—287. Biology, vol. 7, pp. 197—223. LOPEZ-LEON, M. D., CABRERO, J., CAMACHO, J. P. M., CANO, M. I. AND SI-lAW, M. W., HEwIrr, G. M. AND ANDERSON, D. A. 1985. Poly- SANTOS, J. L. 1992. A widespread B chromosome polymorphism in the rates of meiotic drive acting on the B morphism maintained without apparent drive. Evolution, chromosome of Myrmeleotettix maculatus. Heredity, 55, 46,529—539. 6 1—68. LOPEZ-LEON, M. D., CABRERO, J., PARDO, M. C., VISERAS, E. AND WEBB, G. C. AND NEHAU5, i'. 1979. Chromosome organization in CAMACHO, J. P. M. 1993. Paternity displacement in the grass- the Australian plague locust Chortoicetes terminifera. II. hopper Eyprepocnemis plorans. Heredity, in press. Banding variants of B chromosome. Chromosoma, 70, MATSuDA, iT. 1970. On the accessory chromosomes of Aster. I. 205—2 12. The accessory chromosomes of Aster ageratoides. J. Sci. Hiroshima Univ. SerB Div. 2 (Bot.), 13, 1—63.