Proc. Nat. Acad. Sci. USA Vol. 72, No. 1, pp. 200-204, January 1975 Slow Evolutionary Loss of the Potential for Interspecific Hybridization in Birds: A Manifestation of Slow Regulatory Evolution (anatomical evolution/protein evolution/chromosomal evolution/frogs/mammals/immunology) ELLEN M. PRAGER AND ALLAN C. WILSON Department of Biochemistry, University of California, Berkeley, Calif. 94720 Communicated by Nathan 0. Kaplan, October 11, 1974 ABSTRACT Birds have lost the potential for inter- hybridization potential and degree of protein sequence dif- specific hybridization slowly. This inference emerges from protein comparisons made on 36 pairs of bird species capa- ference among species. At first thought it might seem likely ble ofhybridization. Micro-complement fixation tests show that degree of protein similarity between parental species that hybridizable pairs of bird species differ by an average should be the major factor affecting the probability of suc- of 12 units of albumin immunological distance and 25 cessful development of an interspecific zygote. The more units of transferrin immunological distance. As these similar the proteins of two species, the more likely it is, one proteins evolve at a known and rather steady rate, it is inferred that the average hybridizable species pair diverged might suppose, that their genomes would be compatible from a common ancestor about 22 million years ago. The enough to permit development of viable hybrids. However, corresponding period for frog species pairs capable of our first study of this produced a result inconsistent with this hybridization is about 21 million years, while for hybrid- expectation. Mammals that can hybridize differ only slightly izable placental mammals it is only 2 to 3 million years. at the at Thus birds resemble frogs in having lost the potential for protein level, whereas frogs that differ substantially interspecific hybridization about 10 times as slowly as have the protein level hybridize readily (6). mammals. To explain this result, it has been proposed that the Birds have also been evolving very slowly at the anatomi- principal molecular barriers to interspecific hybridization are cal level, particularly within the last 25 million years, regulatory differences between the parental genomes and according to Simpson, Romer, and many other vertebrate zoologists. In this respect they resemble frogs and differ that placental mammals appear to have been undergoing from placental mammals, which have been undergoing more rapid regulatory evolution than frogs have (6). We unusually rapid anatomical evolution. Chromosomal consider regulatory differences to be differences in the pat- evolution is also thought to have proceeded very slowly in terns of gene expression. Since mammals have also under- both birds and frogs, relative to mammals. gone more rapid anatomical evolution than frogs it was The above observations are consistent with the hypoth- have, esis that evolutionary changes in regulatory systems, suggested that rapid regulatory evolution in mammals may that is, changes in the patterns of gene expression, provide account for both their rapid anatomical evolution and their the basis for both anatomical evolution and the evolution- rapid evolutionary loss of the potential for interspecific ary loss of hybridization potential. hybridization. In view of the observation that mammals have undergone rapid evolutionary changes in gene arrange- Several authors have suggested that in order to understand the organismal evolution one needs to focus attention on ment compared to frogs (7), it was further suggested that the control of gene expression rather than on the amino phenomenon of gene rearrangement may be an important acid sequences of proteins coded for by structural genes means of achieving new systems of regulation. (1-5). This article and the preceding ones in this series (6-8) To explain the observation that protein evolution has gone the oc- present evidence consistent with this suggestion. This on at the same rate in frogs as in mammals, despite currence of far more rapid organismal evolution in mammals evidence comes from studies of the relative rates of evolution evolution of anatomy, pro- than in frogs, it has been proposed that protein chromosomes, hybridization potential, and not be at the basis of evolution Thus teins in three major groups of vertebrates. The data on frogs may organismal (6, 7). and mammals have already been published (6, 7). We show there may be two types of molecular evolution: (1) protein evolution, which proceeds relentlessly in a predominantly here that similar conclusions may be drawn from analogous which studies on birds. time-dependent manner, and (2) regulatory evolution, The data presented in this article are based on studies of parallels organismal evolution. We now present the results of a study of protein species capable of interspecific hybridization. Studies of inter- of of new in- resemblance within pairs bird species capable hybridiza- specific hybridization have often provided valuable tion. Birds, like frogs, seem to have lost the potential for inter- sights into the mechanism of evolution (9). There are, of specific hybridization slowly. We also review evidence that course, many natural barriers to interspecific hybridization. birds have evolved slowly at both the anatomical and the Geographical, ecological, behavioral, and anatomical barriers levels. normally prevent contact between an egg of one species and chromosomal sperm of another. If these barriers are circumvented and fertilization occurs, the resulting interspecific zygote may MATERIALS AND METHODS develop into a viable hybrid organism. We have examined Avian Samples. Bird egg whites, sera, and tissue extracts the problem of what relationship, if any, exists between were obtained, prepared, and stored as described (10). Thirty-three different avian species capable of hybridizing Abbreviation: MY, million years. with one or more other species were studied. They were 200 Downloaded by guest on September 30, 2021 Proc.PNat.NAcad.c1RegulatorySci. USA 72 (1975) Evolution in Birds 201 TABLE 1. Immunological distances within avian species pairs chosen from three of the 27 bird orders: Rheiformes, contain- capable of hybridization ing the large, flightless rheas of South America; Anseriformes, made up of waterfowl; and Galliformes, consisting of game Immunological distance birds. The- first footnote to Table 1 gives further details re- Pair*t Albumin Transferrin garding the classification of the species examined. This con- stituted a representative sample of hybridizable species pairs, Rheiformes since more than half of all reported cases of avian inter- Rhea americana specific hybrids occur within the orders Anseriformes and X Pterocnemia pennata 0 NDJ Galliformes (11-13). We prepared antisera (see below) to Anseriformes albumin from six species and to transferrin from five species Anas platyrhynchos and made immunological comparisons of the proteins of 36 X Anas laysanensis 0 1 different avian species pairs capable of interspecific hybridiza- X Anas poecilorhyncha 0 2 X Anas castanea 0 3 tion. X Anas rubripes 0 4 Protein Purification. Eleven avian proteins-six serum X Anas bahamensis 2 3 albumins, two serum transferrins, and three ovotransfer- X Anas acuta 0 11 rins-were purified and used as immunogens. The purifica- X Aix galericulata 8 16 tion of common rhea (Rhea americana) albumin and chicken X Aix sponsa 8 29 X Cairina moschata 9 21 (Gallus gallus) ovotransferrin has been described (10). X Netta rufina 8 21 Serum albumin from Peking duck (Anas platyrhynchos), X Tadorna tadorna 13 34 chicken, ring-necked pheasant (Phasianus colchicus), peafowl X Anser cygnoides 10 65 (Pavo cristatus), and turkey (Meleagris gallopavo) was puri- X Anser anser 13 69 fied by Rivanol precipitation and subsequent regeneration of X Branta canadensis 12 61 the albumin (14) followed by preparative polyacrylamide gel Galliformes electrophoresis (15). Eight percent polyacrylamide gels and a Gallus galluw continuous buffer system of 0.1 M Tris plus 0.05 M glycine X Gallus sonnerati ND§ 0 were generally used. Occasionally 6.4% gels and a discon- X Gallus varius 11 6 tinuous buffer system (16) were employed. 8-anilino-l- X Lophophorus impeyanuw 18 33 naphthalene sulfonate (17) or simply the difference in re- X Lophura nycthemera ND§ 50 fractive index was used to detect the albumin band on the X Phasianus colchicus 291 35¶ gels. Elution of the albumin from the polyacrylamide discs X Chrysolophus pictus 26 32 X Pavo cristatus 241 34¶ (15) into isotris buffer (18) containing merthiolate (2.5 4g X Coturnix coturnix 23 35 per ml) was done at room temperature for 1 day and then at 40 X Numida meleagris 34 53 for two or more days with occasional shaking. The albumins X Meleagris gallopavo 185 *325 (except that of the chicken) were then concentrated by Phasianus colchicus vacuum dialysis, further purified by a repetition of the X Lagopu8 mutus ND§ 16 electrophoresis in polyacrylamide, and again eluted. In cer- X Lophura nycthemera ND§ 26 tain cases three polyacrylamide gel purification steps were X Syrmaticus reevesi 18 25 utilized. X Chrysolophus amherstiae 14 13 Duck and peafowl serum transferrin were purified es- X Chrysolophus pictus 12 12 sentially as described (19). A small amount of ferrous am- X Meleagris gallopavo 19w 15I Pavo cristatus monium sulfate was added prior to the Rivanol precipitation X Pavo muticus 0 0 and polyacrylamide gel electrophoresis steps to saturate