AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 61:453-466 (1983) Cytogenetic Comparison and Phylogeny of Three Species of Hylobatidae PETER VAN TUINEN AND DAVID H. LEDBETTER Department of Anthropology, University of Wisconsin, Madison, Wisconsin, 53706 (I? VT), Department of Cell Biology, University of Texas System Cancer Center, M.D. Anderson Hospital and Tumor Institute (F? V T ) and Department of Medicine, Baylor College of Medicine, Houston, Texas, 77030 (D.H. LJ KEY WORDS Hylobatidae, Primate cytogenetics ABSTRACT Representatives of three subgenera of Hylobatidae, Hylobates (Symphalangus) syndactylus, H. (Nomascus) concolor and H. (Hylobates) agilis were compared karyotypically by G-banding, C-banding, and silver staining. A greater degree of similarity (30-55%) was found among these groups than previous reports suggest; however, these figures are still considerably lower than chromosome similarities characteristic of all other catarrhine groups. Inversion, translocation, fission, and fusion have all played a role in restructuring hylobatid chromosomes since a common hominoid ancestor. H. syndactylus and H. concolor show the greatest G-band correspondence, and in addition share a n unusual C-band distribution and a n extremely rare nucleolar organizing region placement (on the Y chromosome). The latter two are probably shared derived traits, suggesting that these two species shared a common ancestor not shared by other hylobatids. These data sugest a branching order for these three hylobatid groups different from those derived by other morphological and biochemical methods. The gibbons and siamang (family Hylobatidae) constitute a sister taxon to the great apes (Pongidae) and humans (Hominidae) in the primate superfamily Hominoidea. While few fossil specimens have been accepted a s ancestors of Hylobatidae, many other aspects of gibbon biology are well studied with regard to their taxonomic position. Amino acid sequencing, immunodiffusion, and DNA hybridization (Darga et al., 1973; Benveniste and Todaro, 1976) all corroborate morphological and phenotypic analyses of hylobatids (Groves, 1972)and demonstrate their natural grouping with other Hominoidea. By contrast, karyotypic studies of the Hylobatidae have been noteworthy in revealing karyotypic heterogeneity within this group and surprisingly little karyotypic similarity to other anthropoid apes and humans (Hamerton et al., 1963; Chiarelli, 1963; Dutrillaux et al., 1975a; Warburton et al., 1975; Tantravahi et al., 1975; Myers and Shafer, 1979; Couturier et al., 1982). This is unexpected since the great apes (and monkeys as remotely related to humans as Macaca, Papio, 0 1983 ALAN R. LISS. INC. and Cercopithecus) share extensive chromosome and gene map similarity with each other and with humans (Stock and Hsu, 1973; Finaz et al., 1977; Dutrillaux et al., 1978a,b). An analysis of conventionally stained chromosome preparations by Chiarelli (1972) indicated that the hylobatids had differentiated into a t least three karyotypic groups: H y b bates syndactylus (2n = 501, H. concolor (2n= 521, and most other Hylobates (lar, agilis, and klossi) (2n = 44). Hylobates pileatus Warkvong, 1973)and H. muelleri (Wolkin and Myers, 1980) also have 2n = 44. These 44 chromosome species (with the exception of H. klossi) and H. moloch constitute the socalled “lar group.” The identification of a fourth group occupied by H. hoolock (2n = 38) has recently been made (Prouty et al., 1983). The conventionally stained karyotypes of the 44-chromosome species appear to be very similar or identical (Chiarelli, 1972). This Received December 1,1982; accepted April 4, 1983. 454 P. VAN TUINEN AND D.H. LEDBETTER identity has been confirmed by G-banding (Tantravahi et al., 1975) between H. lar and an animal termed H. moloch, the latter probably a misidentified H. muelleri (Alan Mootnick, personal communication). However, changes in diploid number have been accompanied by a surprising degree of structural rearrangement, as shown in comparisons between banded preparations of H. pileatus and H. concolor (Dutrillaux et al., 1975a) where only seven autosomal homologies were observed. Only one chromosome was thought to be shared by H. muelleri and H. syndactylus (Myers and Shafer, 19791, a n unusual finding in species this closely related. The structural differences between these three groups were confirmed by Couturier et al. (1982),where only four to six autosomes were similar between any two groups. While both H. syndactylus and H. concolor have appeared karyotypically dissimilar to the 2n = 44 group, the similarity in diploid number of the former two suggests that they might share considerable chromosome homology with each other. Ledbetter (1981) noted that H. syndactylus possesses a nucleolar organizing region (NOR) on the Y chromosome, a rare finding in mammals and the first reported case in a primate. DeBoer and Van Oostrum-Van der Horst (1975) suggested that the Y chromosome of H. concolor participated with other small acrocentrics in “satellite associations,” but the presence of an NOR was not confirmed by more precise methods, e.g., silver staining or in situ hybridization. In this study we employ G-banding, Cbanding, and silver staining in a systematic comparison of the karyotypes of three major groups of Hylobatidae-H. syndactylus, H. concolor, and H. agilis-in order to reevaluate the degree of karyotypic divergence within this family. In addition, the mechanisms of chromosomal evolution in Hylobatidae are investigated, and several criteria are considered by which the phylogenetic relationships among the three groups can be established. MATERIALS AND METHODS Cytogenetic studies were performed on one male and one female from each of the three hylobatid species. Lymphocyte cultures from a male and female H. syndactylus, a male H. concolor, and a female H. agilis were grown in RPMI 1640 20% fetal calf serum (FCS) and harvested after 72 hours by standard + procedures. Fibroblast cultures from the same male H. syndactylus, a female H. concolor, and a male H. agilis were initiated in 15%FCS followed by convenHam’s F-10 tional harvest. All animal specimens were obtained from the Houston Zoo, with the exception of the male H. agilis, which was obtained from the Vilas Park Zoo, Madison, Wisconsin. G-banding (Seabright, 19711, C-banding (Arrighi and Hsu, 1971) and silver staining (Bloom and Goodpasture, 1976) were performed by standard procedures. The size order and percentage total chromosomal length (%TCL) of G-banded chromosomes were determined from caliper measurements taken of five well-spread metaphases of moderate contraction. The following abbreviations were used in accordance with the Paris Conference (19711, Supplement (19751, when referring to chromosomes of particular species: HSY = Hylobates syndactylus, HCO = Hylohates concolor, HAG = Hylohates agilis. The three banding methods employed were treated as independent cytogenetic traits in the assessment of phylogenetic relationships among the Hylobatidae. The functions of chromosomal rearrangements, as they are reflected in alteration of G-banding patterns, are not known, but they may serve to reinforce reproductive barriers between isolated, diverging populations. In any case, G-banding permits the tracing of chromosomally homologous regions that have undergone rearrangement during evolution; thus it can be used to compare the distributions of particular rearrangements in different taxa. Gbanding can be used to determine overall similarity of karyotypes of different taxa, but in addition, and more importantly, it is used in the present study to determine both ancestral and derived chromosome banding patterns. The advantage of this cladistic approach to primate phylogenetic reconstruction from comparative banding has recently been pointed out by Marks (1982). C-banding stains what is generally regarded to be genetically inert chromosomal material (heterochromatin) (John and Miklos, 19791, and as such minor quantitative variation in C-band blocks is usually of little phylogenetic significance within and between taxa. For this reason such variation is usually disregarded in cytogenetic comparisons in evolutionary studies (Stock and Hsu, 1973). However, major qualitative differences in the chromosomal distribution or lo- + CYTOGENETICS AND HYLOBATIDAE cation of C-band material has been used to distinguish mammalian species independently of other banding methods (Stock, 1981); when ancestral and derived distributions of heterochromatin can be determined, then common possession of unusual patterns may be indicative of close phylogenetic relationship. Similarly, quantitative variation in the intensity of silver staining of nucleolar organizing regions (NORs), where ribosomal RNA genes reside, is a common finding, while major differences in the location of these regions often exist between primate species (Tantravahi et al., 1976). Unusual, shared derived patterns of NOR distribution may indicate descent from a common ancestor with a derived distribution. The significance of comparative chromosome banding analysis to the study of phylogeny was evaluated using the ability of banding to indicate ancestral (plesiomorphic) and derived (apomorphic) chromosomal configurations. Briefly, chromosomal character states were considered to be symplesiomorphic (shared, ancestral) if they occurred generally in both Hylobatidae and other Anthropoidea. Synapomorphies (shared derived traits), on which the grouping of species is based, were defined a s those traits that had a restricted distribution in two or more species, and that, empirically, were found too rarely in other anthropoid primates to be the result of convergent evolution. It is assumed throughout this analysis that Hylobatidae shared with Pongidae and Hominidae a n ancestor not also shared by any Cercopithecoidea. Although this can be demonstrated karyotypically, it is our purpose only to make those out-group comparisons that are relevant in determining phylogeny within the Hylobatidae. RESULTS Hylobates (Symphalangus) syndactylus (2n = 50) The G-banding pattern of H. syndactylus (Fig. l a ) appears to be identical to that reported by Myers and Shafer (1979). One or both terminal regions of all pairs, including the X and Y, are darkly stained after Gbanding, and several pairs are obviously dimorphic for the size of this region (e.g., 5 and 17). A comparison to the C-bands (Fig. l b ) shows that these terminal regions are in fact heterochromatic (Myers and Shafer, 19791, a distribution like that in some chromosomes 455 of the chimpanzee and gorilla (Miller et al., 1974; Miller, 1977). By contrast, centromeric or juxtacentromeric C-bands are faint or nearly absent in many chromosomes of H. syndactylus, unlike the situation in humans, great apes, and Old World monkeys. The long arm of 12 bears a n interstitial block of heterochromatin (shown by arrows in Fig. lb). Interstitial heterochromatin is infrequently found in catarrhine primates but is reported in the gorilla and chimpanzee (Miller et al., 1974; Miller, 1977; Seuanez, 1979). Silver staining (Fig. lc, arrows) demonstrates the unusual distribution of NORs in H. syndactylus on a n autosomal pair of acrocentrics (Klinger, 1963; Henderson et al., 1976) and on the Y chromosome in the male as reported by Ledbetter (1981)for this same animal. In the male these three chromosomes participate in NOR association, while in the female only the two autosomal acrocentrics bear NORs and participate in associations. Hylobates (Nomascus) concolor (2n = 52) The G-banding pattern of H. concolor is shown in Figure 2a. Comparison to the Q- or R-banded preparations of Dutrillaux et al. (1975a) is difficult, but there appears to be good correspondence between the results of these two studies. The C-banding (Fig. 2b) shows that, like H. syndactylus, H. concolor possesses terminal heterochromatin on most pairs and small quantities of centromeric heterochromatin. Three autosomal pairs have interstitial C-bands (Fig. 2b, arrows). Figure 2c shows the presence of five NORs (shown by arrows for the autosomes) in a male H. concolor. The smallest element of the karytoype, the Y, is silver positive and participates in NOR associations with two other acrocentric pairs, 24 and 25. This is only the second report of a n NOR-bearing Y chromosome in a primate. The G-banding patterns of 24 and 25 closely resemble those of human 22 and 21, respectively, and the fact that they are silver positive (like human 21 and 22) seems to confirm their homology with these human chromosomes, as proposed by Dutrillaux et al. (1975a). Hylobates (Hylobates) agilis (2n = 44) The G-banding patterns of a female H. agilis (Fig. 3a) are virtually identical to the published karyotypes of H. lar and H. muelleri (Warburton et al., 1975; Tantravahi et al., 1975; Myers and Shafer, 19791, confirm- 456 P. VAN TUINEN AND D.H. LEDBETTER Fig. 1. The banded chromosomes of male H. syndactylus. a. G-banding demonstrates dark terminal staining and heteromorphism for terminal heterochromatin in pairs 5 and 17 (long arms). b. C-banding showing terrni- nal heterochromatin and an interstitial region on pair 12 (arrows). c. Silver staining showing nucleolar organizing regions (NORs) on the Y chromosome and pair 21 (arrows). CYTOGENETICS AND HYLOBATIDAE Fig. 2. The banded chromosomes of H. concolor a. Gbanding of a male reveals dark terminal (heterochromatic) staining of most chromosome arms, with heteromorphism for the terminal region of pair 14. b. C-banding 457 in a female showing interstitial bands on three pairs (arrows). Y chromosome from the male is shown in inset. c. Silver staining in the male showing an NOR on the Y chromosome and two autosomal pairs (arrows). 458 P. VAN TUINEN AND D.H. LEDBETTER Fig. 3. The banded chromosomes of H. ugilis. a. Gbanding of the female. Pairs 8 and 15 from the male are represented (inset) to demonstrate heteromorphism, by inversion or complex rearrangement, for two different pairs in this animal. b. C-banding in the male showing only centromeric locations, and the Y chromosome being barely visible due to lack of heterochromatin. c. Silver staining showing NORs on pair 13 (arrows) and absence of NOR on the Y chromosome. CYTOGENETICS AND HYLOBATIDAE ing the close relationship within this group. With two exceptions, the G-banding patterns of our female are identical to those of our male. As Figure 3a (inset, from a single cell) shows, homologs of pairs 8 and 15 in the male are consistently found to match poorly, and it is evident that at least two independent rearrangements are represented in this animal. It is interesting that one of the rearranged elements (8)is apparently homologous to a chromosome involved in pericentric inversion in H. muelleri (Tantravahi et al., 1975); the rearrangements appear identical and could constitute a polymorphism. C-bands in H. agilis, unlike H. syndactylus and H. concolor, are present only a t the centromeres of many chromosomes, a distribution like that found in H. lar and H. muelleri (Tantravahi et al., 1975; Myers and Shafer, 1979; Fig. 3b). The secondary constriction of the marker chromosome (pair 13) often appears heterochromatic in the female, as noted by Tantravahi et al. (1975) in H. muelleri, while in our male it was most often achromatic. The NOR region was found a t the predicted location, i.e., at the secondary constriction of pair 13 (Fig. 3c, arrows), known to carry rRNA genes in H. lar (Warburton et al., 1975) and H. muelleri (Tantravahi et al., 1976).The Y chromosome in the male was not silver positive nor did it form associations with members of pair 13. In both the male and female frequent association was noted between the 13s. Comparison A composite illustration of all proposed Gband homologies is shown in Figure 4, while one example in which the ancestral form can be determined and the subsequent rearrangement traced is illustrated in Figure 5. As a group, the three species show significant banding pattern conservation for only six autosomal pairs (Fig. 4a). Two of these (HSY 1/HCO 4/HAG 6 and HSY 14MCO 16/ HAG 16) have been retained intact, while HSY 7MCO 15/HAG 15; HSY 9MCO l/HAG 1;HSY 11MCO 2MAG 5; and HSY 21/HCO 22/HAG 3 have diverged by chromatin loss or addition. It is of interest that the X and Y have undergone rearrangement in the Hylobatidae. While H. syndactylus and H. agilis possess the typical catarrhine primate X chromosome, H. concolor is unique in possessing a n X with a pericentric inversion, which has placed the proximal region of the 459 short arm onto the long arm. As noted earlier the Y chromosomes of both H. syndactylus and H. concolor contain heterochromatin in addition to a n NOR on the short arm, and their morphology, a s shown in Figure 4a, suggests a pericentric inversion difference between them. The comparatively small size of the H. agilis Y chromosome is due to the absence of both heterochromatin and NOR material. In addition to chromosomes shared by all three species, there are additional similarities revealed when comparisons between two species are made. Eight chromosomes of H. syndactylus and H. concolor show some homology (Fig. 4b), three of these without modification (HSY 12/HCO 12; HSY 13MCO 14; and HSY 15/HCO 17). Two others (HSY 3/ HCO 11and HSY 4/HCO 7) show identity for their short arms only. HSY 18 is identical t o HCO 8 except for the lack, in HSY 18, of the terminal region of the long arm of HCO 8. HSY 5 and HCO 6 share distinctive, lightly stained regions and differ by a pericentric inversion and the absence, in HCO 6, of the terminal region of the short arm of HSY 5. The banding pattern of HSY 22 is consistent with its being the fusion product of HCO 23 and HCO 24, with concomitant loss of the NOR region of HCO 24. H. concolor and H. agilis share three autosoma1 pairs not found in H. syndactylus (Fig. 4c). H. syndactylus and H. agilis share only three autosomal pairs not also found in H. concolor (Fig. 4d). A discrepancy in terminal staining is seen between H. syndactylus and H. agilis, and it is important to note that the proposed homologies of these two species are more clearly evident when the terminal dark (heterochromatic) bands of H. syndact-ylus are disregarded in the comparison. Finally, autosomes of the three species that have no match in the group, even ignoring heterochromatin, are presented (Fig. 4e, H. syndactylus; 4f, H. concolor; 4g, H. agilis). At least two facts are evident from these figures: (1)Chromosomes with no match are distributed throughout the size range, and (2) a relatively large proportion (%TCL) of the genome of H. agilis remains unmatched. In order to further quantify and evaluate these pairwise comparisons of G-bands, we have calculated the percentage of total chromosomal length (%TCL) represented in the similar chromosomes of these three species. From caliper measurements of five G-banded cells from each animal, we estimate that H. 460 P. VAN TUINEN AND D.H. LEDBETTER Fig. 4. Haploid composite showing proposed G-banded pairings between species (a-d) and remaining unpaired chromosomes (e-g). a. Chromosomes common to all three species are ordered H. syndactylus (HSY; left), H concc lor (HCO; middle), and H. agilis (HAG; right). b. Similarities between HSY (left) and HCO (right). Note fusion of HCO 23 and 24 to form HSY 22. c. Three chromosomes are unique to HAG (left) and HCO (right). d. Excellent matching is evident between three chromosomes of HSY (left) and HAG (right) when terminal heterochromatin in HSY 19,23,and 24 is disregarded. Chromosomes from H. syndactylus (el, H. concolor (0, and H. agilis (g) that have no interspecific match in the Hylobatidae. (See Discussion for explanation of differences in banding patterdsize of proposed pairings.) syndactylus and H. concolor show conservation of G-banded chromosomes for 55-60% of their TCL. For H. syndactylus and H. agilis this figure drops to about 30-35%, while H. concolor and H. agilis show obvious banding pattern conservation for about 45% of the total length of G-banded chromosomes. While we have not presented the individual %TCL figures for each chromosome, we have found excellent agreement between the proposed homologies based on G-band patterns and the comparison of TCL for all such homologous chromosomes. ships among hylobatids and between them and other catarrhine primates have heretofore permitted only tentative statements. The present study is the first to our knowledge to present a comparison of G-banded karyotypes of three major hylobatid karyotypic groups. This study reveals a degree of intergroup conservation not previously detected, while at the same time confirming previous reports of relatively extensive karyotypic restructuring in the Hylobatidae. In addition, several interesting cytogenetic features unique among catarrhine primates have been found in this group. DISCUSSION While the karyotypic relationships among the hominoids (excluding hylobatids) and the Old World monkeys have been well characterized, karyotypic data regarding relation- Taxonomy of hylobatidae Chiarelli (1963, 1968) allied the Hylobatidae with the Colobinae, a leaf-eating subfamily of Cercopithecoid monkeys, on the basis 461 CYTOGENETICS AND HYLOBATIDAE of resemblance of unbanded karyotypes. This view has been criticized by many authors (Groves, 1972; Andrews and Groves, 1976), and has been dismissed on cytogenetic grounds by comparison of R-banded preparations between Colobus uellerosus, C. palliatus, and hylobatids (Dutrillaux et al., 1981). In a review of the biology and taxonomy of the Hylobatidae, Groves (1972)confirms their hominoid status and argues for subgeneric distinction between the three recognized ma- Fig. 5. Proposed scheme of derivation of HCO 1 and HSY 9 from the ancestral element HAG 1. a. One arm of HAG 1 (left) shows identity to a hominoid chromosome (orangutan equivalent of human chromosome 7, shown right in the brackets) that is ancestral to extant pongids and man, thereby establishing this as the ancestral configuration in these hylobatids. b. HCO 1 (right) can be derived from HAG 1 (left) by a single deletion (translocation ?) from this ancestral arm, with the break point indicated by a dash. c. HSY 9 (right) shares this break with HCO 1, and in addition is differentiated from the latter by a deletion (indicated by dash) of the other arm, which occurred after these two lineages diverged. a b Fig. 6. Four alternative phylogenetic schemes for three major hylobatid groups (A, H. ugilis and other “lar group” gibbons; C, H. concolor; S, H. synductylus). a. Three-way split, suggested by Chivers (1977) and Lucotte e t al. (1982). b. Common branch shared by “lar group” and H. concolor, suggested by Creel and Preuschoft (1976) and Bruce and Ayala (1979). c. Common branch shared by “lar group” and H. syndactylus, suggested by Groves (1972) and Haimoff et al. (1982). Two jor groups (the designation adopted here). In attempting to discern ancestral relationships from morphological traits he tentatively ordered the divergence of hylobatids as follows: H. concolor was the first to split from the ancestral hylobatid stock, after which H. syndactylus and the 44-chromosomegibbons split in the remaining branch. In this scheme, then, H. syndactylus and H. agilis are somewhat more closely related to each other than either is to H. concolor; that is, they share a more recent common ancestor (Fig. 6c). The same result has recently been obtained by Haimoff et al. (1982)using character compatibility analysis of 55 morphological and behavioral traits. However, the results of Creel and Preuschoft (1976) based on cranial morphometric traits seem to suggest that H. syndactylus was the first to split from the ancestral stock, after which H. concolor and the 44-chromosome species shared a common ancestor for a short time (Fig. 6b). Two biochemical studies that have encompassed all three groups likewise differ in their conclusions. The results of electrophoretic comparison of 23 blood proteins (Bruce and Ayala, 1979) suggest that H. lar shows the greatest genetic identity to H. concolor (Fig. 6b). However, in a comparison of 18 blood proteins Lucotte et al. (1982) found no especially close relationship between any two of the three species, suggesting all three were equidistant from each other (Fig. 6a). Thus no consensus on the exact branching order of C d phenotypic traits, length of 0s penis (op) and scrota1 sac morphology (ss) support this scheme in character compatibility analysis (Haimoff et al., 1982). d. Common branch shared by H. concolor and H. syndacty1u.s (this study). Qualitative traits of terminal heterochromatin (h),Y chromosome NOR (n), and translocation involving ancestral hylobatid chromosome HAG 1 (t) support this branching order. 462 P. VAN TUINEN AND D.H. LEDBETTER Hylobatidae has been achieved from previous studies of biochemical and morphological traits. However, as pointed out by Haimoff et al. (1982) these various phylogenetic schemes have in common an early separation of both H. syndactylus and H. concolor from a n ancestor of the 44-chromosome “lar group” (including H. agilis). Karyology and cladistics The cytogenetic data support this early separation but suggest a branching of H. syndactylus and H. concolor which is different in detail from those of previous studies. The results of G-banding, C-banding, and silver staining all suggest that H. concolor and H. syndactylus share several derived cytogenetic traits not present in H. agilis or in other 44-chromosome gibbons. To our knowledge this is the first suggestion of close phylogenetic relationship between H. concolor and H. syndactylus (Fig. 6d). The greatest correspondence of G-bands between any two species is that between H. syndactylus and H. concolor (55-60% TCL). While this is quantitatively greater than the relationship of H. agilis to either H. syndactylus (35% TCL) or H. concolor (45% TCL), it is difficult to assess the importance of these figures. Of course, a more cladistic method of assessing phylogenetic relationship would require interpretation of the nature of the rearrangements by which the karyotypes differ. Unfortunately chromosomal restructuring in the Hylobatidae has been so extensive as to make it impossible t o recognize neartotal homology with confidence or to trace all the changes from one species’ karyotype to the next. However, at least one G-banded chromosome, traceable in all three groups, is informative regarding the cladistic relationships among the Hylobatidae. As seen in Figure 4a, HSY 9 and HCO 1 are similar in sharing the same deletion from the long arm of HAG 1.The long arm of HAG 1 is probably ancestral for all hominoids as shown by R-banding (Dutrillaux et al., 1975a) and confirmed by us with G-banding as shown in Figure 5a. Therefore, the simplest scheme for the derivation of these three chromosomes is that HAG 1 is a conserved ancestral hylobatid chromosome, and that in a lineage common to H. syndactylus and H. concolor a deletion of a portion of long arm (probably by translocation) occurred (Fig. 5). After separation of H. concolor from H. syndactylus, this deriv- ative chromosome underwent a n additional translocation involving the other arm, resulting in the morphology seen in HSY 9. It is abundantly clear from many studies that heterochromatin is predominantly distributed a t or near the centromeres of chromosomes of most primates. Variation in this pattern as seen in C-banding of platyrrhine and catarrhine primates (Ma, 1981; Ponsa et al., 1981; Seuanez, 1979) is usually attributable to small but detectable quantitative differences in the amounts of this heterochromatin. The pattern of primarily centromeric C-band locations was probably ancestral in Catarrhini, and has been retained in H. agilis and other 44-chromosome gibbons. The C-band distributions in H. syndactylus and H. concolor, on the other hand, differ from this pattern in a major qualitative way. In these species C-bands are not only reduced in most centromeric regions, but more significantly, large blocks of terminal heterochromatin are present in both arms of all but a few chromosomes. The probable mechanism for deriving this pattern is not known, but the most parsimonious conclusion is that the C-bands in H. syndactylus and H. concolor were altered in a systematic fashion in a common ancestor not shared by other hylobatids. Terminal C-bands have also been found on many chromosomes of the chimp and gorilla, but there are differences apparent between C-bands of pongids and hylobatids. In these pongids the presence of terminal heterochromatin has not been associated with reduction of centromeric heterochromatin as in hylobatids. In addition, the terminal C-bands of Pan and Gorilla are not nearly so prominent in G-banded preparations as they are in H. syndactylus and H. concolor (Seuanez, 1979; our own observations). These differences, and the absence of terminal heterochromatin in Pongo and Homo, suggest that it was an independent evolutionary occurrence in pongids and hylobatids, and do not weaken the hypothesis that centromeric heterochromatin was the ancestral distribution in Hominoidea. The exclusive presence in H. syndactylus and H. concolor of a n NOR-bearing Y chromosome is a shared trait unknown in other primates. It is found very rarely in other mammals, where in several instances closely related species have conserved a commonly derived NOR-bearing Y chromosome (Pathak et al., 1982). There is no detectable sil- CYTOGENETICS AND HYLOBATIDAE ver staining of the Y chromosome in the male H. agilis reported here, nor is there evidence in other published reports for a n NOR-bearing Y chromosome in any 2n = 44 Hylobates. Here too, the most parsimonious explanation for this distribution is that the Y chromosome NOR arose in a common ancestor of H. syndactylus and H. concolor after its separation from the branch leading to 2n = 44 Hylo bates, It is pertinent to consider the disagreement between this and other phylogenetic studies of Hylobatidae. Of the contrasting techniques employed, it appears that neither the electrophoretic procedures (Bruce and Ayala, 1979; Lucotte et al., 1982) nor the craniometric comparison (Creel and Preuschoft, 1976) have determined rigorously the ancestral and derived states of the traits employed. Character compatibility analysis (Haimoff et al., 1982) is perhaps the most rigorous attempt to deduce hylobatid phylogeny from determination of ancestral and derived morphological and behavioural traits. Although that analysis suggested early divergence of H. concolor from a common ancestor of H. symphalangus and H. agilis, we would like to point out that the length of the apparent common trunk shared by the latter two is very short; only two character state transitions define this common trunk, whereas a t least 20 character state transitions define the common trunk leading to H. agilis and other 44 chromosome gibbons. We emphasize this to indicate that the results of their study and ours are actually fairly close in that both recognize a n early divergence of both H. syndactylus and H. concolor from 44-chromosome gibbons. Although we cannot critically evaluate character compatibility analysis, consideration of our study and of other modern karyotypic studies employing multiple banding methods permits a reevaluation of the four karyotypic traits among the 55 traits employed by Haimoff et al. (1982). Strictly speaking, it is not accurate to designate as primitive and derived the diploid numbers and autosomal arm numbers (their first two traits) found in hylobatids. Apparent tendencies in the same direction by two species may or may not be commonly derived, and this can only be determined by banding analysis. In fact, the assumptions by Haimoff et al. (1982) concerning primitive diploid number and arm number in hylobatids are now known to be in error in light of modern com- 463 parative banding analysis and have apparently led to spurious agreement between their karyotypic traits and their phylogeny. Their third trait, satellites (now known to be NORs), has been shown in the present study to be complex, and previously undetected NORs lead to a conclusion different from that of Haimoff et al. (1982). The presence or absence of acrocentrics (trait 4) is not independent of diploid number in these species, and for the same reasons should be excluded. We would like to consider the possibility that reanalysis of the data of Haimoff et al. (1982) with the inclusion of modern karyotypic data might bring the results of our two studies in closer agreement. Since only 29 of 55 characters employed by that study fully agree with their phylogeny, and the alternative phylogenies generated by the other 26 characters are not shown, we cannot estimate the likelihood that a phylogeny similar to ours would emerge from their data upon reanalysis. Since according to Haimoff et al. (1982) their analysis is easily repeated with new or modified data, it seems worthwhile to pursue this promising method. Similarly, additional karyotypic data should be sought as well from previously incompletely described hylobatids in order to expand the basis of karyotypic comparison. Mechanisms of rearrangement While karyotypic evolution in the Hylobatidae has been too extensive to allow identification of all chromosomal rearrangements in this group, several mechanisms of rearrangement have been identified. Pericentric inversion, the most frequent type of alteration encountered in hominoid evolution (DUtrillaux, 1975b; Yunis and Prakash, 19821, is found in several hylobatid elements. The most obvious of these involves the X chromosome of H. concolor (Dutrillaux, 1975a Fig. 4a), making it the the only exception to the typical catarrhine configuration for this element. Among the autosomes pericentric inversions account for a t least two differences among the hylobatids (HSY 5 B C O 6; HSY 21/HCO 22). In addition, paracentric inversion may explain the dissimilarity in the long arms of HSY 3/HCO 11and HSY 4HCO 7. The metacentric chromosome HSY 22 is apparently a tandem fusion of HC023 and HC024, the latter of which is not strictly acrocentric but possesses a euchromatic short arm. A true Robertsonian centric fusion has 464 P. VAN TUINEN AND D.H. LEDBETTER apparently occurred forming HAG 3. AS Figure 4a shows, one arm of this chromosome is a precise match to HSY 21, which as noted earlier is acrocentric and bears a n NOR. The other arm of HAG 3 is similar to, and may be homologous to, human (and pongid) 13 (Tantravahi et al., 19751, which is also acrocentric and bears NORs. It is reasonable to suggest that these two acrocentrics were present in ancestral hylobatids, and that centric fusion occurred in the lineage leading to H. agilis and other 2n = 44 species. The tendency of acrocentric NOR-bearing chromosomes to undergo fusion is well documented in human populations. The existence of homology between chromosomes of disparate size (HSY 5/HCO 6; HSY 7/HCO 15/HAG 15; HSY 11/HCO 2/ HAG 5; HSY 9/HCO l/HAG 1)suggests noncentric fission or fusion, or translocation. HSY 7 differs from its smaller counterparts (HCO 15/HAG 15) by a small chromatin addition, suggesting donation of a piece resulting from translocation. Conversely, HSY 11 differs from its larger counterparts (HCO 2/ HAG 5) by a small terminal deletion, with the smaller piece presumably incorporated elsewhere in the genome. It is unlikely that fission and centromeric activation alone can account for this change, since the smaller product is not apparently present as a discrete element. It should be noted here that the direction of change (i.e., deletion or addition) can be inferred from the fact that H. concolor and H. agilis, which are in different clades within the Hylobatidae, share identity for these elements, which must therefore be the ancestral forms for all three species. These instances of partial homology are the strongest evidence so far that translocation has been a factor in the differentiation of karyotypes in the Hylobatidae, as suggested by Dutrillaux et al. (1975a) and Couturier et al. (1982). Many other karyotypic differences that are not yet traceable are probably also due to translocation events, since the other classical modes of rearrangement such as inversion or fusiordfission do not readily explain these differences. This is a surprising observation, since only one instance of fixation of a reciprocal translocation has been noted in other hominoids, between chromosomes 5 and 17 in the gorilla (Yunis and Prakash, 1982).It is puzzling why karyotypic evolution in Hylobatidae should be characterized by a class of rearrangement so rarely found in other hominoids. That such translocation may be a frequent recurring event in hylobatids is evidenced by the finding of a H. concolor leucogenys x H. c. gabriellae subspecies hybrid that is heterozygous for a reciprocal translocation (Couturier et al., 1982). Finally, it can be stated confidently that fission played a role in the evolution of diploid numbers in H. syndactylus and H. concolor. It is fairly clear, based on cercopithecoidhominoid comparison (Dutrillaux, 1978a,b), that the last common ancestor of all living Catarrhini possesed 42-48 chromosomes. To generate diploid numbers of 50 and 52 would have required at least one or two fission events. The failure to ascertain fission products with certainty in these species may indicate rearrangement subsequent to such fission. Intraspecific variation The finding of a H. agilis with two apparent inversion heteromorphisms is noteworthy, given the small number of animals studied. One of the inversions in the present study is identical to that reported by Tantravahi et al. (1975) in a H. muelleri, the only other such report in 2n = 44 hylobatids. A review of the banded karyotypes of 2n = 44 gibbons in previous published reports suggests maintenance of two morphs of chromosome 8 in living populations. Three Hylobates lar (Tantravahi et al., 1975; Benirschke and Hsu, 1977; Warburton et al., 1975), three H. pileatus (one by Dutrillaux et al., 1975a; two by our own unpublished observations), and one termed “gibbon” (Pearson, 1973) are apparently homomorphic for the submetacentric form, while one H. agilis (present study) is homomorphic for the metacentric form. A H. agilis (present study) and a H. muelleri (Tantravahi et al., 1975) are heteromorphic for this pair-i.e., both forms are present. A H. muelleri abbotti, male parent to a siamang-gibbon hybrid (Myers and Shafer, 1979), was evaluated from G-banded photographs (kindly provided by David Shafer), and appears also to be heteromorphic for pair 8. Thus, individuals assigned to two H y l e bates species (H. muelleri and H. agilis) probably possess both forms of chromosome 8. Several explanations may be offered for this finding. Since these two groups are placed in H. lar a s subspecies by Groves (1972), the inversion may reflect a widespread polymorphism maintained in the various populations of H. lar. Alternatively, the CYTOGENETICS AND HYLOBATIDAE two forms of 8 may actually distinguish different subspecies of H. lar, with heteromorphism occurring mainly by hybridization in captivity. This is now known to be the situation between Sumatran and Bornean subspecies of orangutans, which have been geographically separated for at least 8,000 years and which differ by a n inversion (Seuanez, 1979). However, if H. muelleri and H. agilis are distinct species as Haimoff et al. (1982)recognize, then the maintenance of the same polymorphism in both is probably evidence of the relative recency of their separation. Many more gibbons of verifiable provenance will have to be studied in order to determine the relationship of chromosome variation to population structure. SUMMARY The degree of detectable karyotypic similarity between three major groups of Hylobatidae has been greatly increased with the aid of multiple chromosome banding methods. Considerable disparity still exists, however, between the banded karyotypes of these three groups, and there is now stronger evidence that translocation has been a major factor in karyotypic evolution in this family. The existence of two heteromorphisms in a single individual in this study is in accord with other studies showing structural heteromorphisms, suggesting that extant populations of hylobatids are karyotypically variable. Karyotypic differences between H. syndactylus, H. concolor, and H. agilis (and other “lar group” species) are consistent with previous assignments of these three to distinct subgenera. However, all three banding methods employed reveal shared derived traits that suggest the existence of a common ancestor to H. syndactylus and H. concolor not shared by H. agilis and other 44-chromosome species, a phylogenetic scheme which differs from other recent schemes based on morphological or biochemical comparisons. ACKNOWLEDGMENTS The authors would like to thank the Houston Zoological Gardens and the Henry Vilas Zoo, Madison, for their cooperation in obtaining tissue specimens used in this study. We would like to thank Dr. T.C. Hsu for much help and encouragement throughout the duration of this study. We thank Dr. Leonard Prouty and Dr. Elliott Haimoff for providing data prior to publication. The invaluable assistance of Chris Fuscoe in the preparation 465 of this manuscript is gratefully acknowledged. This work was supported in part by National Science Foundation grant # BNS 7924439 to P.V.T. LITERATURE CITED Andrews, P, and Groves, C (1976) Gibbons and brachiation. In D Rumbaugh (ed): Gibbon and Siamang, Vol. 4.Basel: Karger, pp. 167-218. Arrighi, FE, and Hsu, TC (1971) Localization of heterochromatin in human chromosomes. Cytogenetics 1023-86. Benirschke, K, and Hsu, TC (1977) Mammalian Chromosome Atlas, Vol. 10. Berlin: Springer Verlag, Folio 518. Benveniste, RE, and Todaro, GJ (1976)Evolution of type C viral genes: evidence for an Asian origin of man. Nature 261t101-108. Bloom, SE, and Goodpasture, C (1976)An improved technique for selective silver staining of nucleolar organizer regions in human chromosomes. Hum. Genet. 34t199-206. Bruce, EJ, and Ayala, F J (1979) Phylogenetic relationships between man and apes: electrophoretic evidence. Evolution 33: 1040-1056. Chiarelli, B (1963) Comparative morphometric analysis of primate chromosomes. 111. The chromosomes of the genera Hylohates, Colohus and Preshytis. Caryologia 16:637-648. Chiarelli, B (1968)Caryology and taxonomy of the catarrhine monkeys. Am. J. Phys. Anthropol. 24t155-170. Chiarelli, B (1972) The karyotypes of the gibbons. In D Rumbaugh (ed):The Gibbon and Siamang, Vol. 1. Basel: Karger, pp. 90-102. Chivers, DJ (1977)The lesser apes. In HSH Prince Rainier and GH Bourne (eds): Primate Conservation. New York: Academic Press, pp. 539-598. Couturier, J, Dutrillaux, B, Turleau, C, and de Grouchy, J (1982) Comparaisons chromosomiques chez quatre especes ou sous-especes de gibbons. Ann. Genet. 25t510. Creel, N, and Preuschoft, H (1976) Cranial morphology of the lesser apes. In D Rumbaugh (ed): Gibbon and Siamang, Vol. 4. Basel: Karger, pp. 219-303. Darga, LL, Goodman, M, and Weiss, ML, (1973)Molecular evidence on the cladistic relationships of the Hylobatidae. In D Rumbaugh (ed): The Gibbon and Siamang, Vol. 2. Basel: Karger, pp, 149-162. de Boer, LEM, and Van Oostrum Van der Horst, C (1975) A note on the karyotypes and chromosome associations of the Hylobatidae, with special reference to Hylobates (Nomascus) concolor. J. Hum. Evol. 4t559-564. Dutrillaux, B, Rethore, MO, Aurias, A, and Goustard, M (1975a) Analyse du caryotype de deux especes de Gibbons (Hylohates lar et H. concolor)par differentes techniques du marquage. Cytogenet. Cell Genet. 1581-91. Dutrillaux, B, Rethore, MO, and Lejeune, J, (1975133 Comparaison du caryotype de I’orang-outang (Pongo pygmaeus) a celui de l’homme, du chimpanze et du gorille. Ann. Genet. 18t153-161. Dutrillaux, B, Viegas-Pequignot, E, Dubos, C, and Masse, R (1978a) Complete or almost complete analogy of chromosome banding between baboon (Pupiopapio) and man. Hum. Genet. 43t37-46. Dutrillaux, B, Viegas-Pequignot, E, Couturier, J, and Chauvier, G (197%) Identity of euchromatic bands from man to Cercopithecidae. Hum. Genet. 45:283-296. 466 P. VAN TUINEN AN'D D.H. LEDBETTER Dutrillaux, B, Couturier, J, and Ruedi, D (1981) Le caryotype de Colohus vellerosus et de C. palliatus: Comparaison avec les Cercopithecidae et l'homme. Ann Genet. 2438-81. Finaz, C, Van Cong, N, Cochet, C, Frezal, J, and de Grouchy, J (1977) Fifty-million-year evolution of chromosome 1 in the primates: Evidence from banding and gene mapping. Cytogenet. Cell Genet. f8:160-164. Groves, C (1972) Systematics and phylogeny of gibbons. In D Rumbaugh (ed): The Gibbon and Siamang, Vol. 1. Basel: Karger, pp. 1-89. Haimoff, EH, Chivers, DJ, Gittins, SP, and Witten, AJ (1982) A phylogeny of gibbons based on morphological and behavioral characters. Folia Primatol. 39t213-237. Hamerton, JL,Klinger, HP, Mutton, DE, and Lang, EM (1963) The somatic chromosomes of the Hominoidea. Cytogenetics 2240-263. Henderson, AS, Atwood, KC, and Warburton, D (1976) Chromosomal distribution of rDNA in Pan paniscus, Gorilla gorilla heringei, and Symphalangus syndactylus: comparison to related primates. Chromosoma 59:147-155. John, B, and Miklos, GLG (1979) Functional aspects of satellite DNA and heterochromatin. Int. Rev. Cyt. 58:l-114. Klinger, H P (1963) The somatic chromosomes of some primates (Tupaia glis, Nycticehus coucnng, Tarsius hancan.us, Cercocehus aterrimus, Symphalangus s,yndactylus. Cytogenetics 2: 140-151. Ledbetter DH (1981) NOR-bearing Y chromosome in a primate, Hylohates (Symphalangusj syndoct.ylus. Cytogenet. Cell Genet. 29:250-252. Lucotte, G, Gautreau, C, and Smith, DGi1982) Distances electrophoretiques entre les trois sous-genres de gibbons. Ann. Genet. 25t11-13. Ma, NSF (1981) Chromosome evolution in the Owl Monkey, Aotus. Am J. Phys. Anthropol. 54:293-303. Marks, J (1982) Evolutionary tempo and phylogcnetic inference based on primate karyotypes. Cytogenet. Cell Genet. 34:261-264. Markvong, A (1973) Karyotypes of three species of gibbons. Mammalian Chromosome Newsletter f4: 149. Miller, DA, Firschein, IL, Dev, VG, Tantravahi, R, and Miller, O J (1974) The gorilla karyotype: Chromosome lengths and polymorphisms. Cytogenet. Cell Genet. f3:536-550. Miller, DA (1977) Evolution of primate chromosomes. Science 198: 1116-1124. Myers, RH, and Shafer, DA (1979) Hybrid ape offspring of a mating of gibbons and siamang. Science 205r308310. Paris conference (19711, Supplement (1975) Standardization i n Human Cytogenetics. Birth Defects: Original Article Series. New York: The National Foundation, Vol. 11, No. 9. Pathak, S, van Tuinen, P, and Merry, DE (1982) Heterochromatin, synaptonemal complex, and NOR activity in the somatic and germ cells of a male domestic dog, Canis familiaris. Cytogenet. Cell Genet. 34t112-118. Pearson, P (1973) The uniqueness of the human karyotype. In T Caspersson and L Zech (eds): Chromosome Identification. New York: Academic Press, pp. 145-151. Ponsa, M, Miro, R, Estop, AM, And Egozcue, J (1981) Banding patterns of the chromosomes of Erythrocehus patas compared to other primate species. Genetica 56t39-45. Prouty, LA, Buchanan, PD, Pollitzer, WS, and Mootnick, AR (1983)A presumptive new hylobatid subgenus with 38 chromosomes. Cytogenet. Cell Genet. 35:141-142. Seabright, M (1971) A rapid banding technique for hitman chromosomes. The Lancet 11:971-972. Seuanez, HN (1979) The phylogney of human chromosomes. Berlin: Springer Verlag, p. 173. Stock, AD (1981) Chromosomal variation and constitutive heterochromatin i n three porpoise species (Genus Stenella). Cytogenet. Cell Genet. 3ft91-100. Stock, AD, and Hsu, TC (1973) Evolutionary conservatism in arrangement of genetic material. Chromosoma 43t211-224. Tantravahi, R, Dev, VG, Firschein, IL, Miller, DA, and Miller, O J (1975) Karyotype of the ribbons Hylohates lar and H. moloch: inw&on in chromosome 3.Cytogenet. Cell Genet. f5:92-102. Tantravahi, R, Miller, DA, Dev, VG, and Miller, O J (1976) Detection of nucleolar organizer regions in chromosomes of human, chimpanzee, gorilla, orangutan, and gibbon. Chromosoma 56:15-27. Warburton, D, Henderson, AS, and Atwood, KC (1975) Localization of rDNA and Giemsa-banded chromosome complement of white-handed gibbon, Hylohates lar. Chromosoma 51r35-40. Wolkin, JR,and Myers, RH (1980) Characteristics of a gibbon-siamang hybrid ape. Int. J. Primatol. 1r203221. Yunis, JY, and Prakash, 0 (1982) The origin of man: A chromosomal pictorial legacy. Science 2f5: 1525-1530.