Molecular Phylogenetics and Evolution Vol. 19, No. 3, June, pp. 479 – 485, 2001 doi:10.1006/mpev.2001.0929, available online at http://www.idealibrary.com on

Parallel Acceleration of Evolutionary Rates in Symbiont Genes Underlying Host Nutrition Jennifer J. Wernegreen,* ,† ,1 Aaron O. Richardson,† and Nancy A. Moran† *Josephine Bay Paul Center for Comparative Molecular Biology and Evolution, The Marine Biological Laboratory, 7 MBL Street, Woods Hole, Massachusetts 02543; and †Department of Ecology and Evolutionary Biology, University of Arizona, Tucson, Arizona 85721 Received October 17, 2000

The overproduction of essential amino acids by Buchnera aphidicola, the primary bacterial mutualist of aphids, is considered an adaptation for increased production of nutrients that are lacking in aphids’ diet of plant sap. Given their shared role in host nutrition, amino acid biosynthetic genes of Buchnera are expected to experience parallel changes in selection that depend on host diet quality, growth rate, and population structure. This study evaluates the hypothesis of parallel selection across biosynthetic pathways by testing for correlated changes in evolutionary rates at biosynthetic genes of Buchnera. Previous studies show fast evolutionary rates at tryptophan biosynthetic genes among Buchnera associated with the aphid genus Uroleucon and suggest reduced purifying selection on symbiont nutritional functions in this aphid group. Here, we test for parallel rate acceleration at other amino acid biosynthetic genes of Buchnera–Uroleucon, including those for leucine (leuABC) and isoleucine/valine biosynthesis (ilvC). Ratios of nonsynonymous to synonymous substitutions (d N/d S) were estimated using codon-based maximum-likelihood methods that account for the extreme AT compositional bias of Buchnera sequences. A significant elevation in d N/d S at biosynthetic loci but not at two housekeeping genes sampled (dnaN and tuf) suggests reduced host-level selection on biosynthetic capabilities of Buchnera–Uroleucon. In addition, the discovery of trpEG pseudogenes in Buchnera–U. obscurum further supports reduced selection on amino acid biosynthesis. © 2001 Academic Press 1

INTRODUCTION Bacterial endosymbionts are found in several insect groups, including aphids and other members of the Supplementary data for this article are available on IDEAL (http:// www.idealibrary.com). 1 To whom correspondence should be addressed. Fax: (508) 4574727. E-mail: [email protected].

sap-feeding Sternorrhyncha (mealybugs, whiteflies, psyllids), tsetse flies, carpenter ants, and cockroaches, among others (Buchner, 1965; Moran and Telang, 1998; Douglas, 1989). These endosymbionts are characterized by their location within specialized host cells (bacteriocytes), their strict maternal transmission through host lineages, and their requirement for host survival and reproduction (Baumann et al., 1998a; Moran and Telang, 1998). The symbiosis between Buchnera aphidicola and aphids is particularly well characterized genetically and physiologically. Buchnera is known to provide essential amino acids that are lacking in aphids’ diet of plant sap, as evidenced by poor growth and reproduction of aphids that are cured of symbionts, by a rescuing effect of an artificial diet containing essential amino acids (Douglas, 1998), and by the presence of genes for essential amino acid pathways (Baumann et al., 1995). In Buchnera associated with the fast-growing Aphididae, plasmid amplification of genes for the biosynthesis of leucine (leuABCD) and tryptophan (trpEG) is presumed to result from a strong effect of host-level selection for the overexpression of these loci (Lai et al., 1994; Rouhbakhsh et al., 1996; Baumann et al., 1998b, 1999). The symbiont loci underlying nutritional provisioning of hosts might be expected to show correlated patterns of evolution relative to other symbiont loci, for two general reasons. First, because aphid species encounter different ecological conditions affecting both nutritional demands and supply, hosts may vary in their dependence on symbiont provisioning, causing differences in intensity of selection on the relevant symbiont loci. The profile of essential amino acids in phloem tends to be similar among plants and resembles the profile defined for nutritional requirements and proteins of insects (Sandstro¨m and Moran, 1999). However, representation of essential amino acids within total amino acids of phloem sap is usually less than the optimal for animal nutrition and varies among plant species, developmental stages, and tis-

479

1055-7903/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

480

WERNEGREEN, RICHARDSON, AND MORAN

sues (Sandstro¨m and Moran, 1999). Furthermore, other ecological factors will influence the need for protein production and will have parallel effects on the demand for different essential amino acids. For example, if a fastgrowing aphid lives on a poor diet, strong selection will favor efficiency of all enzymes in amino acid production; if a slow-growing aphid lives on a rich diet, selection will be generally relaxed across these loci. A second reason for correlated evolution of the set of loci underlying amino acid biosynthesis is that the efficiency of selection depends on different factors for genes selected at the host level and those selected at the symbiont level (Rispe and Moran, 2000). The amino acid biosynthetic genes represent a set of loci for which selection is primarily or exclusively between hosts rather than between symbionts, in contrast to the large set of loci involved in cell growth and replication. In particular, host-selected genes experience more efficient selection if a tight bottleneck in symbiont numbers occurs in the course of the transfer of symbionts from mother to progeny, whereas the opposite is true for genes with substantial selection acting within hosts. Thus, a change in the infection process within mothers can have differential effects on the efficiency of selection on the amino acid biosynthetic loci. A reduction in the intensity or the efficiency of selection of amino acid biosynthetic genes within an aphid lineage might be expected to result in faster sequence evolution, especially at nonsynonymous sites, and possibly in the loss of gene function as evidenced by the transformation of loci into pseudogenes. This pattern might be expected, for the above reasons, to apply across loci underlying amino acid provisioning and to be absent from loci for bacterial housekeeping functions. Several recent molecular studies suggest reduced selection on biosynthetic functions in Buchnera associated with Uroleucon (Baumann et al., 1997; Rouhbakhsh et al., 1997), a recent radiation of aphids that feed primarily on host plants in the Asteraceae (Moran et al., 1999). Compared to related Buchnera lineages, Buchnera–Uroleucon show elevated rates of nonsynonymous substitutions at genes for anthranilate synthase (TrpEG), the rate-limiting step in tryptophan biosynthesis (Rouhbakhsh et al., 1997). Pseudogenes at trpEG of Buchnera of Uroleucon sonchi further support reduced selection on protein function in this symbiont group (Baumann et al., 1997). In this study, we test for reduced selection at genes encoding isoleucine and valine biosynthesis (ilvC) and leucine biosynthesis (leuABC) in Buchnera–Uroleucon. We also include two housekeeping genes, the ␤ subunit of DNA polymerase III (dnaN) and elongation factor Tu (tuf), to test for genome-wide acceleration in sequence evolution in this Buchnera clade. Changes in the ratio of nonsynonymous (d N) to synonymous (d S) substitutions are indications of a change

in selective intensity or efficiency within particular lineages (Ohta, 1992). We tested for elevated d N/d S in Buchnera–Uroleucon specifically at biosynthetic loci. Codon-based maximum-likelihood methods allow explicit consideration of the genetic code structure and different base frequencies at codon positions and therefore provide the most realistic evolutionary model for sequences with biased base compositions, such as the AT-rich sequences of Buchnera (Goldman and Yang, 1994; Yang and Nielsen, 2000). These likelihood-based methods have been used to detect changes in dN/dS along branches of phylogenies (e.g., Yang, 1998) and were used here to detect changes in the strength or efficiency of selection in Buchnera of Uroleucon. METHODS Aphid Samples and DNA Sequencing Buchnera samples include endosymbionts of six Uroleucon species plus Macrosiphoniella ludovicianae, referred to inclusively as “Uroleucon” for brevity. Collection data and DNA extraction were described previously (Moran et al., 1999). TrpEG, leuABC, and dnaN were sequenced for all taxa, and ilvC and tuf were sequenced for a subset of taxa. GenBank accession numbers are given for sequences obtained here and in previous studies (Table 1). Molecular Methods The five gene regions of Buchnera were amplified with PCR using standard reaction conditions described previously (Moran et al., 1999). DnaN (1107 bp) was amplified with primers dnaN1F (TACGTWGTWATGCCWATG) and gyrBR (TACATATAWCCWCTATGWCTA), tuf (861 bp) was amplified with EfTu104F (CAGAAGAAAAAGCWAGAGGTA) and EfTu965 (GAACCTGTTACATCASTAGTT), leuABC (3919 bp) was amplified with leuA10F (AAGTTATTATTTTGATACSACCYTAC) and leuC1337R (GCWGCCATAATAGGACTAACTA), trpEG (1767 bp) was amplified with trpEUrol.151F (AAGYATYATGATYATYGAYAGTGC) and trpGUrol.507R (ATAGATTCRGGRTGAAATTG), and ilvC (1163) was amplified with ilvC112F (AAAATWGTTATTGTAGGTTGTGG) and ilvC1275 (AAWATAGGATATGCTGMTTCAG), each with annealing temperatures between 50 and 60°C. Internal sequencing primers were developed as necessary for specific groups. PCR products were cleaned with Concert Rapid PCR purification columns (Gibco BRL), and DNA sequences were obtained (as in Moran et al., 1999) directly from PCR products or TA clones of PCR fragments (Invitrogen). The accuracy of the Buchnera–U. obscurum trpEG pseudogene sequence was confirmed by performing four replicate PCR amplifications from aphid genomic DNA, TA-cloning the PCR

481

ACCELERATED EVOLUTIONARY RATES IN Buchnera–Uroleucon

TABLE 1 GenBank Numbers of Buchnera Loci Sequenced in This Study (Boldface) and Previously Aphid host

Abbreviation

dnaN

“Uroleucon clade” Uroleucon rudbeckiae U. ambrosiae U. aeneum U. sonchi U. helianthicola U. rurale

Urud Uamb Uaen Uson Uhel Urur

AF197882 AF197884 AF197885 AF197888 AF197890 AF197891

Ucal Uerig Ml

AF197892 AF197893 AF197894

Ap Sg Rp

AF197895 AF008210 AF197896

U. caligatum U. erigeronense Macrosiphoniella ludovicianae Acyrthosiphon pisum Schizaphis graminum Rhopalosiphum padi

tuf

AF217548 AF217549 AF217550

AF217551

Y12307 Y12308

products, and sequencing the cloned inserts. The complete Buchnera–U. obscurum trpEG product was sequenced from two clones, and partial sequences were obtained from the other two clones. Analysis of DNA Sequence Data Translated protein sequences were aligned using the Clustal option of Megalign (DNAstar) and corrected by hand when necessary. Phylogenetic analysis was performed using the parsimony option of PAUP* (version 4.0b2 written by D. L. Swofford; branch and bound searching), and the strength of support for nodes was estimated by bootstrapping (1000 replicates). Likelihood-Based Estimates of d N/d S Ratios of d N to d S (termed “␻”) were calculated across Buchnera genealogies following the approach of Yang (1998) (using codeML; Yang, 1997). Sequence changes were mapped onto phylogenies with the basic codonsubstitution model described previously and estimating transition/transversion ratios from the data (Yang, 1998). Values of d N, d S, and ␻ were estimated along branches according to three likelihood models that impose different constraints on ␻: (A) a simple one-ratio model assuming the same ␻ across all branches of a given phylogeny, (B) a two-ratio model allowing different ␻ ratios for the Buchnera–Uroleucon clade (␻ u) versus a “background” ratio (␻ 0), and (C) a “free-ratio” model allowing different ␻ values for each branch. Models A and C were compared to test whether ␻ varies across branches, and models A and B were compared to test whether ␻ within the Uroleucon clade differs from the background ratio (␻ u versus ␻ 0). Differences in the likelihood scores of these models were evaluated with the likelihood ratio test statistic [2⌬l ⫽ 2(l 1 ⫺ l 0)]. We determined whether differences between models were significant by comparing the like-

trpEG

trpB

leuABC

AF197464 AF197460 AF197459 AD001677 AF197462 L81122

AF058439 AF058431 AF058432 AD001677 AF058434 L81149

L81124 L81123 AF197458

L81150 L81151 AF058428

AF200469 AF197454 AF197455 AF197448 AF197451 AF200468, AF201382-3 AF197453 AF197452 AF197456

L43555 Z21938 L43551

L46355 Z19055 L46358

AF197457 AF041836 X71612

ilvC

AF217553 AF217554 AF217555

AF217556 AF217557 AF217558 AF008210

lihood ratio test statistic to the ␹ 2 distribution. Degrees of freedom were equal to the number of branches across which d N/d S was estimated in the alternative hypothesis, minus 1 (as Yang, 1998) [df ⫽ 1 for A vs B; df ⫽ (number of branches in tree) ⫺ 1 for A vs C]. RESULTS Pseudogene at trpEG of Buchnera–U. obscurum A mixed signal for Buchnera–U. obscurum trpEG sequences obtained directly from PCR products indicated multiple and different trpEG copies in this taxon. Independent amplification and cloning of four trpEG PCR products from Buchnera–U. obscurum yielded identical sequences showing strong homology to trpEG but with multiple frameshifts that introduce stop codons (see alignment in Supplementary Data). This agreement among independent products confirms the presence of a pseudogene and suggests that it is the dominant trpEG copy in the Buchnera–U. obscurum genome. Like Buchnera–D. noxia (Lai et al., 1996; Wernegreen and Moran, 2000) and Buchnera–U. sonchi (Baumann et al., 1997), Buchnera–U. obscurum may possess multiple copies of a trpEG pseudogene but also retain one or more functional trpEG copies (though no functional copy was detected here). Phylogenetic analysis of trpEG functional and pseudogene copies confirms the independent origins of pseudogenes in Buchnera–U. obscurum and –U. sonchi (tree not shown). The Buchnera–U. sonchi pseudogenes group closely with the functional trpEG gene of this taxon, while the Buchnera–U. obscurum pseudogene is basal to the clade consisting of U. sonchi, U. rudbeckiae, and U. ambrosiae, consistent with published phylogenies of this group based on aphid and symbiont loci (Clark et al., 2000; Moran et al., 1999).

482

WERNEGREEN, RICHARDSON, AND MORAN

inherited endosymbiont (Rouhbakhsh et al., 1997; Funk et al., 2000; Wernegreen and Moran, 2001). The phylogeny common to all loci is shown in Fig. 1. The only ambiguity in the phylogeny of each gene region was poor resolution of the placement of U. caligatum and U. erigeronense. These two taxa were therefore excluded from the phylogeny-based estimation of ␻, which requires a fully resolved genealogy. FIG. 1. Phylogeny common to all Buchnera loci sampled in this study. Relationships were well-resolved at all loci (⬎90% bootstrap support) except for the placement of U. caligatum and U. erigeronense (not shown).

Elevated ␻ at Biosynthetic Loci of Buchnera–Uroleucon

Phylogeny Estimations

Values of ␻ ( ⫽ d N/d S) were estimated across Buchnera genealogies under the (A) one-ratio, (B) two-ratio, and (C) free-ratio models (Fig. 2, Table 2). Under model B, ␻ u was higher than ␻ 0 at every gene tested. Comparisons of model A versus B with the likelihood ratio

Genealogies of all Buchnera genes were similar, reflecting the lack of recombination in this vertically

FIG. 2. Phylogenies of Buchnera loci under codon-based maximum likelihood model. The “Uroleucon clade” distinguished for estimates of ␻ u (d N/d S) includes Uroleucon spp. and Macrosiphoniella ludovicianae and is marked with boldface branches. Abbreviations for aphid host species are given in Table 1. Branches are drawn in proportion to the expected numbers of substitutions per codon, as estimated under the free-ratio model. Estimates of ␻ under the free-ratio model are given along branches. Values for ␻ are not given at branches with extremely short lengths, due to high standard errors for these ␻ estimates. Trees were left unrooted for the likelihood analyses, but are drawn according to the known rooting.

483

ACCELERATED EVOLUTIONARY RATES IN Buchnera–Uroleucon

TABLE 2 Comparison of Codon-Based Likelihood Models Estimating d N/d S (ⴝ␻) along Branches of Buchnera Phylogenies Likelihood model trpEG

leuABC

ilvC

trpB

dnaN

tuf

(A) (B) (C) (A) (B) (C) (A) (B) (C) (A) (B) (C) (A) (B) (C) (A) (B) (C)

One-ratio Two-ratios Free-ratios One-ratio Two-ratios Free-ratios One-ratio Two-ratios Free-ratios One-ratio Two-ratios Free-ratios One-ratio Two-ratios Free-ratios One-ratio Two-ratios Free-ratios

␻a

Likelihood ratio test results

⫺ln L

␻0

␻u

⫺9944.6850 ⫺9938.4365 ⫺9926.1163 ⫺15786.4709 ⫺15784.4708 ⫺15773.2092 ⫺4076.6368 ⫺4072.2966 ⫺4066.6441 ⫺2765.6572 ⫺2764.4117 ⫺2755.4719 ⫺5379.3045 ⫺5378.7889 ⫺5372.0315 ⫺1559.83271 ⫺1559.0676 ⫺1556.5962

0.073 0.053

⫽␻ 0 0.085

A vs B A vs C

12.5 37.1

*** **

␹ 2, df ⫽ 1 ␹ 2, df ⫽ 16

0.046 0.041

⫽␻ 0 0.05

A vs B A vs C

4.0 26.5

* *

␹ 2, df ⫽ 1 ␹ 2, df ⫽ 14

0.0529 0.0355

⫽␻ 0 0.0664

A vs B A vs C

8.7 20.0

*** n.s.

␹ 2, df ⫽ 1 ␹ 2, df ⫽ 12

0.042 0.033

␻0 0.048

A vs B A vs C

2.5 20.4

n.s. n.s.

␹ 2, df ⫽ 1 ␹ 2, df ⫽ 16

0.111 0.089

⫽␻ 0 0.119

A vs B A vs C

1.0 14.5

n.s. n.s.

␹ 2, df ⫽ 1 ␹ 2, df ⫽ 16

0.0073 0.0054

⫽␻ 0 0.0107

A vs B A vs C

1.5 6.5

n.s. n.s.

␹ 2, df ⫽ 1 ␹ 2, df ⫽ 6

2⌬l b

␻ 0, “background” ratio; ␻ u, ratio across Uroleucon clade; ␻’s for free-ratio model are given along branches in Fig. 2. 2⌬l, Likelihood ratio test statistic. * P ⬍ 0.05. ** P ⬍ 0.005. *** P ⬍ 0.001. a b

test show that this elevation in ␻ u is significant at biosynthetic loci trpEG, leuABC, and ilvC (i.e., model B is significantly better than model A), but not at trpB, dnaN, or tuf. DISCUSSION Our discovery of trpEG pseudogenes in Buchnera–U. obscurum demonstrates a second, phylogenetically independent origin of pseudogenes in this genus of hosts (in addition to those of Buchnera–U. sonchi; Baumann et al., 1997) and further supports reduced effects of selection on genes for tryptophan biosynthesis in this clade. Furthermore, the higher d N/d S values observed within Buchnera–Uroleucon for trpEG, ilvC, and leuABC indicate changes in the strength or efficiency of selection (Ohta, 1992), and these changes appear to be restricted to amino acid biosynthetic enzymes since d N/d S values were not significantly elevated for dnaN and tuf. Relaxed purifying selection on genes for amino acid biosynthesis might result from enhanced nutrition of hosts or from other changes in nutritional ecology that cause these amino acids to be present in excess (Wernegreen and Moran, 2000). Data on phloem composition for hosts of Uroleucon are few, but one study on the sap of Sonchus oleraceus ingested by U. sonchi

did indicate very high proportions of essential amino acids (45% of total amino acids compared to 15–30% for host plants of other aphids; Sandstro¨m and Moran, 1999). In addition to potentially enhanced diet, Uroleucon species commonly possess secondary symbionts (Sandstro¨m et al., 2001) that might provide nutritional benefits and decrease host dependency on Buchnera. Demand for essential amino acids might also be reduced if growth rate is lowered due to ecological circumstances or to limitation by some other nutrient. If an enzyme for one step in a pathway for essential amino acid biosynthesis is lost or defective, selection on the genes underlying all of the other steps and pathways might be relaxed since protein synthesis in the host will be limited by the defective gene. A similar case has been described for Buchnera of aphids in the genus Diuraphis, some of which show patterns of accelerated sequence evolution and pseudogenes at amino acid biosynthetic loci (Wernegreen and Moran, 2000). This model of relaxed selection on Buchnera biosynthetic genes would parallel the case of rbcL in chloroplasts of nonphotosynthesizing, parasitic plants; in these taxa, rbcL shows elevated rates of evolution especially at nonsynonymous sites and pseudogenes (Wolfe and DePamphilis, 1998). Alternatively, reduced efficiency of selection may specifically affect biosynthetic genes due to a change in

484

WERNEGREEN, RICHARDSON, AND MORAN

population structure. The benefits of amino acid biosynthetic genes are improved host nutrition and these genes are selected at the level of the host aphid. Increased inoculum size (e.g., number of bacteria transmitted to host offspring) would decrease the efficiency of selection at the level of the host and increase efficiency of selection among symbionts within hosts (Bergstrom and Pritchard, 1998; Rispe and Moran, 2000). No data currently exist for the inoculum size of Buchnera–Uroleucon. Notably, we observed no rate acceleration at trpB of Buchnera–Uroleucon, in contrast to the other biosynthetic loci sampled. This result agrees with previous relative rates tests that show a significant acceleration at trpEG in Buchnera–Uroleucon, but not at trpB (Rouhbakhsh et al., 1997). Several mechanisms may account for the lack of acceleration at this biosynthetic gene. Under the relaxed selection hypothesis, reduced demand for tryptophan in Uroleucon may affect sequence evolution more strongly at trpEG, the ratelimiting step for tryptophan production, than at trpB, which is not rate-limiting. In addition, selective constraints at different loci may vary due to functional considerations. For example, higher selective coefficients against changes at trpB may prevent any rate acceleration even if this gene experiences a strong effect of drift in Buchnera–Uroleucon. Recent studies have presented several kinds of evidence for high levels of genetic drift within Buchnera, which experiences an absence of recombination and small population size compared to other bacteria (e.g., Moran, 1996; Lambert and Moran, 1998; Clark et al., 1999; Wernegreen and Moran, 1999; Funk et al., 2001). We addressed the distinct question of reduced selection specifically at biosynthetic genes within Buchnera of Uroleucon. The observation of a correlated increase in d N/d S across biosynthetic loci suggests a change in either the nutritional ecology of Uroleucon or the population structure of their Buchnera. Data on the nutrition of Uroleucon species and on the transmission of Buchnera between mother and progeny could reveal the basis of the correlated patterns of sequence evolution observed for amino acid biosynthetic genes. ACKNOWLEDGMENTS The authors thank J. Sandstro¨m for the Uroleucon samples, P. Baumann for DNA extractions, and H. Dunbar for confirming the Buchnera–U.obscurum pseudogene sequence. We also thank D. Funk, E. Bernays, and other members of the Moran lab for helpful discussion and W.-H. Li and two anonymous reviewers for valuable suggestions. This work was supported by the Josephine Bay Paul and C. Michael Paul Foundation, a National Institutes of Health postdoctoral training grant in Molecular Insect Science (Center for Insect Science, University of Arizona) to J.J.W., and a National Science Foundation grant (DEB-9815413) to N.A.M.

REFERENCES Baumann, L., Clark, M. A., Rouhbakhsh, D., Baumann, P., Moran, N. A., and Voegtlin, D. J. (1997). Endosymbionts (Buchnera) of the aphid Uroleucon sonchi contain plasmids with trpEG and remnants of trpE pseudogenes. Curr. Microbiol. 35: 18 –21. Baumann, L., Baumann, P., Moran, N. A., and Sandstro¨m, J. (1999). Genetic characterization of plasmids containing genes encoding enzymes of leucine biosynthesis in endosymbionts (Buchnera) of aphids. J. Mol. Evol. 48: 77– 85. Baumann, P., Baumann, L., Lai, C.-Y., Rouhbakhsh, D., Moran, N., and Clark, M. (1995). Genetics, physiology, and evolutionary relationships of the genus Buchnera: Intracellular symbionts of aphids. Annu. Rev. Microbiol. 49: 55–94. Baumann, P., Baumann, L., Clark, M. A., and Thao, M. L. (1998a). Buchnera aphidicola: The endosymbiont of aphids. ASM News 64: 203–209. Baumann, P., Baumann, L., Clark, M., and Thao, M. (1998b). Genetic properties and adaptations of Buchnera aphidicola to an endosymbiotic association with aphids. ASM News 64: 203–208. Bergstrom, C., and Pritchard, J. (1998). Germline bottlenecks and the evolutionary maintenance of mitochondrial genomes. Genetics 149: 2135–2146. Buchner, P. (1965). “Endosymbiosis of Animals with Plant Microorganisms.” Wiley–Interscience, New York. Clark, M. A., Moran, N. A., and Baumann, P. (1999). Sequence evolution in bacterial endosymbionts having extreme base compositions. Mol. Biol. Evol. 16: 1586 –1598. Clark, M. A., Moran, N. A., Baumann, P., and Wernegreen, J. J. (2000). Cospeciation between bacterial endosymbionts (Buchnera) and a recent radiation of aphids (Uroleucon) and pitfalls of testing for phylogenetic congruence. Evolution 54: 517–525. Douglas, A. (1998). Nutritional interactions in insect–microbial symbioses: Aphids and their symbiotic bacteria Buchnera. Annu. Rev. Entomol. 43: 17–37. Douglas, A. E. (1989). Mycetocyte symbiosis in insects. Biol. Rev. Cambridge Philos. Soc. 64: 409 – 434. Funk, D., Wernegreen, J. J., and Moran, N. A. (2001). Intraspecific variation in symbiont genomes: Bottlenecks and the aphid–Buchnera association. Genetics 157: 477– 489. Funk, D. J., Helbling, L., Wernegreen, J. J., and Moran, N. A. (2000). Intraspecific congruence among multiple symbiont genomes phylogenetic. Proc. R. Soc. London B 267: 2517–2521. Goldman, N., and Yang, Z. (1994). A codon-based model of nucleotide substitution for protein-coding sequences. Mol. Biol. Evol. 11: 725– 736. Lai, C.-Y., Baumann, L., and Baumann, P. (1994). Amplification of trpEG: Adaptation of Buchnera aphidicola to an endosymbiotic association with aphids. Proc. Natl. Acad. Sci. USA 91: 19 –23. Lai, C.-Y., Baumann, P., and Moran, N. A. (1996). The endosymbiont (Buchnera sp.) of the aphid Diruaphis noxia contains plasmids consisting of trpEG and tandem repeats of trpEG pseudogenes. Appl. Environ. Microbiol. 62: 332–339. Lambert, J. D., and Moran., N. A. (1998). Deleterious mutations destabilize ribosomal RNA in endosymbiotic bacteria. Proc. Natl. Acad. Sci. USA 95: 4458 – 4462. Moran, N. (1996). Accelerated evolution and Muller’s ratchet in endosymbiotic bacteria. Proc. Natl. Acad. Sci. USA 93: 2873–2878. Moran, N. A., and Telang, A. (1998). Bacteriocyte-associated symbionts of insects. Bioscience 48: 295–304. Moran, N. A., Kaplan, M. E., Gelsey, M. J., Murphy, T. G., and Scholes, E. A. (1999). Phylogenetics and evolution of the aphid

ACCELERATED EVOLUTIONARY RATES IN Buchnera–Uroleucon genus Uroleucon based on mitochondrial and nuclear DNA sequences. System. Entomol. 24: 85–93. Ohta, T. (1992). The nearly neutral theory of molecular evolution. Annu. Rev. Evol. Syst. 23: 263–286. Rispe, C., and Moran, N. A. (2000). Accumulation of deleterious mutations in endosymbionts: Muller’s ratchet with two levels of selection. Am. Nat. 156: 425– 441. Rouhbakhsh, D., Lai, C.-Y., von Dohlen, C. D., Clark, M. A., Baumann, L., Baumann, P., Moran, N. A., and Voegtlin, D. J. (1996). The tryptophan biosynthetic pathway of aphid endosymbionts (Buchnera): Genetics and evolution of plasmid-associated anthranilate synthase (trpEG) within the Aphididae. J. Mol. Evol. 42: 414 – 421. Rouhbakhsh, D., Clark, M. A., Baumann, L., Moran, N. A., and Baumann, P. (1997). Evolution of the tryptophan biosynthetic pathway in Buchnera (aphid endosymbionts): Studies of plasmidassociated trpEG within the genus Uroleucon. Mol. Phylog. Evol. 8: 167–176. Sandstrom, J., and Moran, N. A. (1999). How nutritionally imbalanced is phloem sap for aphids? Entomol. Exp. Appl. 91: 203–210. Sandstro¨m, J. P., Russell, J. A., White, J. P., and Moran, N. A. (2001). Independent origins and horizontal transfer of bacterial symbionts of aphids. Mol. Ecol. 10: 217–228.

485

Wernegreen, J. J., and Moran, N. A. (1999). Evidence for genetic drift in endosymbionts (Buchnera): Analyses of protein-coding genes. Mol. Biol. Evol. 16: 83–97. Wernegreen, J. J., and Moran, N. A. (2000). Decay of mutualistic potential in aphid endosymbionts through silencing of biosynthetic loci: Buchnera of Diuraphis. Proc. R. Soc. London B 267: 1423– 1431. Wernegreen, J. J., and Moran, N. A. (2001). Vertical transmission of biosynthetic plasmids in aphid endosymbionts (Buchnera). J. Bacteriol. 183: 785–790. Wolfe, A. D., and dePamphilis, C. W. (1998). The effect of relaxed functional constraints on the photosynthetic gene rbcL in photosynthetic and nonphotosynthetic parasitic plants. Mol. Biol. Evol. 15: 1243–1258. Yang, Z. (1997). PAML, a program package for phylogenetic analysis by maximum likelihood. CABIOS 13: 555–556. Yang, Z. (1998). Likelihood ratio tests for detecting positive selection and application to primate lysozyme evolution. Mol. Biol. Evol. 15: 568 –573. Yang, Z., and Nielsen, R. (2000). Estimating synonymous and nonsynonymous substitution rates under realistic evolutionary models. Mol. Biol. Evol. 17: 32– 43.