McDonaldKreitman1991

McDonaldKreitman1991 - l LETTERS[0 NAl UHl 8 Shinomura Y...

Info iconThis preview shows pages 1–3. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: l, LETTERS [0 NAl UHl: 8, Shinomura. Y , Eng, 1., Rattan, S. C. & Yalow, R. 5. Comp, Blochem. Physiol 968, 239442 (1990) 9, Beintema, J, J. & Neuteooom. B, J, molec. Evol. 19. 145—152 (1983), 10. Sarkar. G., Koeberl, D, D. & Sommer, S. S. Genomics 6, 133443 (1990). 11. Fan, Z.-W,, Eng, J., Shaw, G & Yalow, R, S. Peptides 9, 429-431 (1988). 12. Smith, A I, at al. J. Endocr. 115, REAR8 (1987), 13, Wolfe, P. 8, & Cabra, J. J. Molec, lmmun. 17, 1493-1505 (1980) 14. Eng, J., Du. B,-H.. Rautman, J.-P, & Yalow, R. S. Peptides 7. Suppl, 1, 17—20 (1986), 15. Easteal, S. Genet/cs 124. 165M173 (1990). 16. Li, W.-H., Gouy. M., Sharp, P, M., O’Huigln, C, & Yang, Y,-W, Proc. natn Acad SC: USA, 87, 670343707 (1990), 17. Fitch, W, M, Am, Nat. 16, 111al20 (1980). 18, Felsenstem. J. Syst, Zool. 34, 152—161 (1985). 19. Li. W,-H. & Gouy, M, Meth. Enzym, 183, 645-659 (1990). 20. Fitch, W, M, & Belntema, J, J, Moles. Biol. Evol. 7, 438—443 (1990), 21, Sarich, V, M. & Wilson, A. C, Science 158, 1200—1203 (1.973). 22. Shosham, 1., Goodman, M, Czelusniak, J,‘& Braunltzer, G. in Evolutionary Relationships among Rodents/1 MultidiSCIplinary Analysis (eds Luckett, W, P. & Hartenberger. J.-L,) 191-210 (Plenum, New Yorx. 1985). 23, Vu, J.~H.. Eng. J., Rattan, S, & Yalow, R. S. Peptides 10. 1195—1197 (1989), 24 Li, W.-H., Wu, C,-|, & Luo. C.-C. Moles. Biol. Evol, 2, 1504174 (1985). 25. Felsenstein, J. Syst, 2001. 27. 4014110 (1978) ACKNOWLEDGEMENTS, We thank M, Gouy, R. Honeycutt, H. Mendelssohn, T. Mendelson, D, Eilam, Z. Sever, and L. Jacobs for help and suggestions and J. Felsenstein for sending us the PHYLlP package, This study was supported by NIH (W ~H.L.) and the USmlsrael Binational Science Foundation (0.6.), Adaptive protein evolution at the Adh locus in Drosophila John M. McDonald 8: Martin Kreitman Department of Ecology and Evolutionary Biology, Princeton University, Princeton, New Jersey 08544, USA PROTEINS often differ in amino-acid sequence across species. This difference has evolved by the accumulation'of neutral muta- tions by random drift, the fixation of adaptive mutations by selec- tion, or a mixture of the two. Here we prepose a simple statistical test of the neutral protein evolution hypothesis based on a com— parison of the number of amino—acid replacement substitutions to synonymous substitutions in the coding region of a locus. If the observed substitutions are neutral, the ratio of replacement to synonymous fixed differences between species should be the same as the ratio of replacement to synonymous polymorphisms within species. DNA sequence data on the Adh locus (encoding alcohol dehydrogenase, EC 1.1.1.1) in three species in the Drosophila melanogaster species subgroup do not fit this expectation; instead, there are more fixed replacement differences between species than expected. We suggest that these excess replacement substitutions result from adaptive fixation of selectively advantageous muta— tions. Consider a set of alleles from more than one species and assume that there is no recombination. The alleles are then connected by a single phylogenetic tree, which can be divided into two parts: between-species branches and within—species branches. Within~species branches connect all the alleles within each species to their most recent common ancestor. Between- species branches connect these common ancestors to the com- mon ancestor ofthe whole phylogeny. A mutation on a between— species branch will appear in all the descendant alleles and thus will be a fixed difference between species, whereas a mutation on a within-species branch will be a polymorphism within a species. Nucleotide substitutions in a coding region can also be divided into replacement substitutions and synonymous substitutions, Of M possible mutations in a coding region, let Mr be the number of possible neutral replacement mutations and Ms be the number of possible neutral synonymous mutations. All remaining mutations, M —(M,+ M,), are deleterious. Let p. be the nucleotide mutation rate per nucleotide site, so that the mutation rate for any one of the three possible mutations at a site is ,u/3. Under the neutral theory, the expected number of fixed replacement substitutions in a set of alleles is Tb(u/3)M,, where Tb is the total time on the between—species branches. The expected number of fixed synonymous substitutions is 652 * - \ Th(u/3)MS. For a particular phylogeny and mutation rate, the number of replacement substitutions is independent of the number of synonymous substitutions. Therefore, the expected ratio of replacement to synonymous .fixcd substitutions is Tb(,u/3)M,: Tb(u/3)M,, which reduces to Mr: Ms. If TW is the total time on the within-species branches of the phylogeny, the expected ratio of replacement to synonymous polymorphisms is Tw(p./3)Mr: TW(u/3)M,,which also reduces to Mr: M,. Thus if protein evolution occurs by neutral processes, the ratio oi“ replacement to synonymous fixed substitutions should be the same as the ratio of replacement to synonymous polymorphisms, A G-test of independence‘ can be used to test this null hypothesis. This test does not require many assumptions. Unlike mom tests of the neutral theory, it does not require that p0pulations have reached equilibrium. Recombination could cause different segments ofa locus to have different phylogcnies, but this would only cause a serious bias in the above test if the time to a common ancestor ofa genomic segment was correlated with the ratio of neutral replacement to synonymous sites in that segment. There is no reason to expect such a correlation. Similarly, variation in nucleotide mutation rate among regions of a locus would only be important if nucleotide mutation rate was cone» lated with the ratio of neutral replacement to synonymous sites, which does not seem likely. We have assumed, for simplicity, that all nucleotide mutations are equally likely. A difference in mutation rate between transitions and transversions, coupled with the greater proportion of transitions that are synonymous, would change the ratio of replacement to synonymous substitu- tions, But the ratio for fixed substitutions would still equal the ratio for polymorphisms. If there are more synonymous sites than replacement sites with two or three possible neutral muta- tions, multiple substitutions at a single nucleotide site will make the ratio ofreplacement to synonymous fixed difterences a slight overestimate of Mr: MS. For closely related species the eficct is trivially small and even in the most extreme case of complete saturation of neutral mutablc sites with substitutions, the over- estimation from multiple substitutions is not large enough to account for the results presented below. We have compared DNA sequences of the coding region of the Adh locus in three species in the mephila melanogaster species subgroup (Table 1). There is substantial evidence that the threoninc/lysine polymorphism in the alcohol dehydro- genase of D. melanogaster is subject to natural selection“, but there has been no evidence to indicate whether any of the replacement differences between species were adaptive. We limited the comparison to closely related species to minimize the number of multiple mutations at single nucleotide sites. The ratio of replacement substitutions to synonymous substitutions that are fixed between species is significantly greater than the ratio of replacement to synonymous polymorphisms (Table 2). About 29% of the fixed difierences between species are replace- ment substitutions, but only 5% of the polymorphisms are replacement substitutions. This is not what would be expected { if the observed replacement and synonymous substitutions were selectively neutral. It would be expected, however, if most of the observed replacement substitutions were due to adaptive fixation of selectively advantageous mutations. A substitution which becomes fixed by selection will be a polymorphism for less time than a substitution that becomes fixed by random drift4 and so an adaptive substitution will be less likely to appear as a polymorphism than will a neutral substitution. An alternative explanation for the excess number of fixed replacement differences could be a combination of many slightly deleterious replacement mutations and population sizes WhiCh have recently undergone dramatic expansion in all three Species. In the ancestral small populations, mildly deleterious replace- ment mutations are effectively neutral and would therefore accumulate as fixed differences. In recently expanded popula- - tions, slightly deleterious replacements are selected against and NATURE - VOL 351 ~ 20 JUNE 1991._ l LETTERS TO NATURE TABLE 1 Variable nucleotides from the coding region of the Adh locus in D. melanogaster. D. simulans and D. yakuba D, melanogaszer Con. a b c d e f g h i j k l T T T T T T T T T T T T 781 c 789 T 808 A 816 o 834 T 859 c 867 c 8’70 0 950 c 974 c 983 T 1019 c 1031 c 1054 T 1043 c 1068 c 1089 c 1101 G 1127 T 1131 c 1160 T 1175 T 1178 c 1184 c 1190 c 1196 o 1199 c 1202 T 1203 c 1229 T 1232 T 1235 c 1244 c 1265 c 1271 A 1277 T 1283 c 1298 c 1304 c 1316 c 1425 c 1431 T 1443 c 1452 c 1490 A 1504 c 1518 c 1524 T 1527 c 1530 c 1545 T 1548 c 1551 c 1555 c 1557 c 1560 c 1573 c 1581 c 1584 c 1590 c 1596 c 1611 A 1614 c 1635 c 1657 A 1665 c - - A A — A ~ « ~ ~ — A D. simulans a b c d e f a b o d e f g h i j k 1 l | l l l | l ! l l I H l l D.yakuba . . _ . . . _ _ . . _ — - . Repl. Fixed Fixed Fixed Poly. Poly. Fixed 2 Poly. Fixed Poly. Poly. Fixed Poly. Poly. Poly. Poly. Poly. Fixed Fixed Fixed Poly. Fixed Fixed Poly. Fixed Poly. Poly. Poly. Fixed Poly. Poly. Fixed Poly. Poly. Fixed Poly. Fixed Poly. Fixed Poly. Poly. Poly. Poly. Poly. Poly. Poly. Fixed Poly. Fixed Poly. Poly. Fixed Poly. Poly. Poly. Poly. Poly. Fixed Fixed Poly. Poly. - « — — . v n — — a ~ — — — — Syn. Poly. — — T T T T T T T T T T T T Syn. Fixed — — - — — — — ~ T — n < — Syn. 2 Poly. — — ~ a a - — Syn. Poly. T T T T T T T Repl. Fixed Syn. Poly. The first column contains the nucleotide positions. numbered according to ref. 6. so that the first base of the coding region of exon 2 is 778. of exon 3 is 942 and exon 4 is 1.417. The second column contains the consensus (con) nucleotide for each site. Nucleotides identical to the consensus are shown as a clash. Several D. yakuba individuals were heterozygous; sites which are underlined contained both the nucleotide shown and the consensus nucleotide. The D. melanogaste/ sequences are six Adhs (a-f) and five AdhF (g—k) alleles6 and one Adhm‘o allele”. D. simulans sequences a and h were also from the literaturen‘n. D. simu/ans sequences of were of cloned alleles from flies from Australia. Canada. France and the Congo. They were sequenced by the dideoxy chain termination method directly from supercoiled plasmids”. The D yakuba alleles of single flies from Brazzaville. Congo were sequenced directly from products of amplification by the polymerase chain reaction1"15. Each of these sequences is from one fly from a separate isofemale line. Every allele was completely sequenced in both directions. Each substitution relative to the consensus nucleotide at a site is classified as either fixed or polymorphic. A substitution that is fixed in one species and polymorphic in another is classified as a polymorphism. Because we do not know whether a substitution that is polymorphic in more than one species represents one or two mutations. it is classified as a single polymorphism. Syn. synonymous; Rent, replacement; Poly, polymorphic. would not appear as polymorphisms. This explanation requires that many replacement mutations have selection coefficients against them that make them neutral in each small population but selected against in each expanded population. It also requires that each population has been large for long enough that slightly deleterious polymorphisms have been eliminated by selection, but not for so long that slightly advantageous bachmutations have replaced the slightly deleterious fixed sub- stitutions. This slightly deleterious model requires so many assumptions about selection coefficients, population sizes, and the times of population expansions that we prefer the simpler explanation, the occasional fixation of an adaptive mutation. NATURE - VOL 351 ~ 20 JUNE 1991 The rate of neutral synonymous mutation at the Adh locus in Drosophila has been estimated to be about 8 X 10‘9 per syn— onymous site per years. There are about 192 efiectlvely synony- mous sites in the coding regions of the Adh ldcusé, so that the sum of the lengths of the between—species branches of the phylogeny connecting D. melanogaster, D. simulans and D. yakuba, with 17 synonymous substitutions, is about 11 million years (ignoring multiple hits for this rough calculation). If all seven fixed replacement substitutions were adaptive, there was an average of one adaptive fixation every 1.6 million years. It has been suggested that for a plausible value of genetic load, one mutation could be fixed by selection every 300 generations7. 653 LETTERS TO NATURE TABLE 2 Number of replacement and synonymous substitutions for fixed differences between species and polymorphisms within species Fixed Polymorphic Replacement 7 2 Synonymous 17 42 A G-test of independence (with the Williams correction for continuity)l was used to test the null hypothesis, that the proportion of replacement substitutions is independent of whether the substitutions are fixed or polymorphic. (327.43, P=0.006. More realistic models of selection suggest that adaptation could occur much more quicklyg'g, but if we accept the conservative figure and assume there are 10 generations per year in Drosophila, it would mean that 50,000 loci could be undergoing adaptive evolution as quickly as Adh. Thus, the rate of adaptive protein evolution seen at the Adh locus may not be unusually high, and selective fixation of adaptive mutations may be a viable alternative to the clocklike accumulation of neural] mutations as an explanation for most protein evolution. (3 Received 20 March; accepted 13 May 1991. Sokal. R. R. & Rohlf, F. J. Biometry (Freeman. San Francrsco. 1981). Lewontin. R. C. A. Rev. Genet. 19, 81—102 (1985). Kreitman, M. & Hudson, R. R. Genetics 127. 565—582 (1991). Nei, M. Molecular Evolutionary Generics (Columbia University Press, New York, 1987), Sharp. P. M. & Li. W.»H. J. molec. Evol. 28. 398—402 (1989). Kreitman, M. Nature 304. 412-417 (1983). Haldane, J. B. S. J. Genet 57, 511—524 (1957). Maynard Smith, J. Nature 219. 1114—1116 (1968). Sved, J, A. Am. Nat. 102. 283493 (1968). 10. Collet, C. 1 molar). Evol. 27. 142—146 (1988). 11. Bodmer, M. & Ashburner, M. Nature 309. 425—430 (1984). 12. Cohn. V. H. & Moore, G. P. Molec. Biol. Evol. 5, 154-166 (1988). 13. Zagursky, R. .l., Baumeister, K.. Lomas, N. & Berman, M. L. Gene Ana/yr, Tea/tn. 2. 8984 (1985), 14. Saiki, R. K., Bugawan, T. L.. Horn. (3. T, Mullis. K. B. & Eriich. H. A, Nature 324. 1634.66 (1988), 15. Higuchl, R. G. & Ochman, H. Nucleic Acids Res. 17. 5865 (1989). SDPPT‘P’FJ‘PPN!‘ ACKNOWLEDGEMENTS. We thank S. El-Gogary for cloning the D. simulans alleles and J. R. Davtd to; the isotemale lines of D. yalruba. This work was supported by the NIH and the NSF. DNA sequences have been deposited in GenBank, accession numbers X57361—X57364 (D. simulans) and X57365. X57376 (D. yakuba). WM Development of the light response in neonatal mammalian rods _G. M. Ratio, D. W. Robinson, B. Yan* & P. A. McNaughton* Physiological Laboratory, Downing Street, Cambridge CB2 3EG, UK THE sensitivity to light is low in many neonatal mammals when compared with that in the adult. In human infants at one month of age, for example, the dark—adapted sensitivity for detection of large stimuli is 50 times lower than in the adultlT‘, and in rats the overall sensitivity of the neonatal retina is also low compared with the adult“. This low sensitivity in the neonate has been attributed to a number of factorss‘“, but the possibility that the photoreceptors themselves might be an important limitation on the overall visual sensitivity has not so far been clearly established. Here we record the light response of single neonatal rat rods and find that the sensitivity is considerably lower than in the adult. The response to a single photoisomerization is normal in the neonate, and the sensitivity deficit can therefore be attributed to a low level of functional rhodopsin. Opsin, the protein component of rhodopsin, must be present in normal amounts, as the sensitivity can be restored to adult levels by treating the retina with 9-cis retinal, an active homologue of the native chromophore ll-cis retina]. The low sensitivity of photoreceptors in the neonate can therefore be attributed mainly to a low concentration of ll—cis retinal in the developing retina. Rod outer segments begin to appear around the optic disk at about postnatal day 8 (P8) in the rat, and they grow rapidly towards the adult size over the next three weeks”. We have recorded the membane current of single rod outer segments, using the suction pipette method”, from animals ranging in age from P12 (the earliest time at which any response to light has been detected) to adult (age greater than two months). Figure 1 compares typical recordings obtained from rods from a one— month-old animal and a 13-day-old animal. There are two prin— cipal respects in which the responses from neonatal rods differ from those in more mature animals: the lightvsensitive current is smaller, and the sensitivity to light is less by more than an * Present addresses: Department of Microbiology. University of Texas Medical School at Houston, PO Box 20708. Houston, Texas 77225, USA (B.Y.): and Physiology Subject Group, Department of Biomedical Sciences, King’s College London, Strand, London WCZR 2L3, UK (PAMcN). 654 order of magnitude. As a qualitative illustration of the second point, the flash strength giving the smallest response in the P13 rod shown in Fig. 1b is the same as that giving the second—largest response in the P35 rod in Fig. 1a. The intensity—response relations for rods from albino rats of various ages are shown in Fig. 2. In each case the points have been fitted by an exponential functionM (see legend to Fig. 2), which provides a good fit to the intensity-response relation of other mammalian rods”. The sensitivity is inversely proportional to the light intensity required to produce a half-saturating response. Only a small fraction of rods at P13 responded to light and, of those that did, the sensitivity was on average 49 times less than in the adult. Rods from 17-day-old animals are seven times less sensitive than those from adults, and at P35, when the animals have been weaned and the outer segments are almost of adult size, the sensitivity is still three times less than in the adult. Similar results were obtained from pigmented rats (see Fig. 2 legend). There are several mechanisms which might account for the low sensitivity of the neonate. Some components of the light- sensitive pathway may not be fully expressed, resulting in a reduced response to a single photoisomerization. A second possibility is that fewer photons of light may be absorbed in the neonatal outer segments, either because ofa delay in expression of the gene coding for opsin, or because an insufficient supply of the light-sensitive chromophore ll—cis retinal is available. The outer segment length at P13 is one half of adult (see Fig. 2 legend), but this difference does not account for the low sensitivity (see Fig. 3 legend). The possibility that the action spectrum of neonatal rhodopsin might be different can also be discarded”. The response to a single photoisomerization has been deter— mined in the experiments shown in Fig. 3 by observing the responses to a series of dim flashes‘s'”. The single-photon response in this adult (Fig. 3, top) was 0.2 pA, or 4.2% of the maximum response, whereas in the P13 rod (Fig. 3, bottom), which was 56 times less sensitive, the single photon response was 0.42 pA, or 21% of the saturating response. In four experi— ments on animals aged between P13 and P17 the mean single— photon response was 0.58 pA, and in seven experiments on animals P30 or older, the mean response was 0.50 pA (for further details see legend to Fig. 3). The gain of the light-sensitive pathway is therefore not significantly different in the neonate, and the low sensitivity must be due to a reduced absorption of light in the immature outer segments. If there is a delay in the expression of the gene coding for opsin, then the absorption of light should be localized to the NATURE ~ VOL 351 - 20 JUNE 1991 ‘ r v{Wawwummmwmmwuwmwl :W‘n «View. m—v-xs-r—ar'm ...
View Full Document

This note was uploaded on 12/18/2010 for the course GENETICS 486 taught by Professor Hey during the Fall '10 term at Rutgers.

Page1 / 3

McDonaldKreitman1991 - l LETTERS[0 NAl UHl 8 Shinomura Y...

This preview shows document pages 1 - 3. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online