Ecological genomics of tropical trees: how local population size and allelic diversity of resistance genes relate to immune responses, cosusceptibility to pathogens, and negative density dependence.

In tropical forests, rarer species show increased sensitivity to species-specific soil pathogens and more negative effects of conspecific density on seedling survival (NDD). These patterns suggest a connection between ecology and immunity, perhaps because small population size disproportionately reduces genetic diversity of hyperdiverse loci such as immunity genes. In an experiment examining seedling roots from six species in one tropical tree community, we found that smaller populations have reduced amino acid diversity in pathogen resistance (R) genes but not the transcriptome in general. Normalized R gene amino acid diversity varied with local abundance and prior measures of differences in sensitivity to conspecific soil and NDD. After exposure to live soil, species with lower R gene diversity had reduced defence gene induction, more cosusceptibility of maternal cohorts to colonization by potentially pathogenic fungi, reduced root growth arrest (an R gene-mediated response) and their root-associated fungi showed lower induction of self-defence (antioxidants). Local abundance was not related to the ability to induce immune responses when pathogen recognition was bypassed by application of salicylic acid, a phytohormone that activates defence responses downstream of R gene signalling. These initial results support the hypothesis that smaller local tree populations have reduced R gene diversity and recognition-dependent immune responses, along with greater cosusceptibility to species-specific pathogens that may facilitate disease transmission and NDD. Locally rare species may be less able to increase their equilibrium abundance without genetic boosts to defence via immigration of novel R gene alleles from a larger and more diverse regional population.

Application of qRT-PCR and RNA-Seq analysis for the identification of housekeeping genes useful for normalization of gene expression values during Striga hermonthica development.

Striga is a root parasitic weed that attacks many of the staple crops in Africa, India and Southeast Asia, inflicting tremendous losses in yield and for which there are few effective control measures. Studies of parasitic plant virulence and host resistance will be greatly facilitated by the recent emergence of genomic resources that include extensive transcriptome sequence datasets spanning all life stages of S. hermonthica. Functional characterization of Striga genes will require detailed analyses of gene expression patterns. Quantitative real-time PCR is a powerful tool for quantifying gene expression, but correct normalization of expression levels requires identification of control genes that have stable expression across tissues and life stages. Since no S. hermonthica housekeeping genes have been established for this purpose, we evaluated the suitability of six candidate housekeeping genes across key life stages of S. hermonthica from seed conditioning to flower initiation using qRT-PCR and high-throughput cDNA sequencing. Based on gene expression analysis by qRT-PCR and RNA-Seq across heterogeneous Striga life stages, we determined that using the combination of three genes, UBQ1, PP2A and TUB1 provides the best normalization for gene expression throughout the parasitic life cycle. The housekeeping genes characterized here provide robust standards that will facilitate powerful descriptions of parasite gene expression patterns.

The genome of black cottonwood, Populus trichocarpa (Torr. & Gray).

We report the draft genome of the black cottonwood tree, Populus trichocarpa. Integration of shotgun sequence assembly with genetic mapping enabled chromosome-scale reconstruction of the genome. More than 45,000 putative protein-coding genes were identified. Analysis of the assembled genome revealed a whole-genome duplication event; about 8000 pairs of duplicated genes from that event survived in the Populus genome. A second, older duplication event is indistinguishably coincident with the divergence of the Populus and Arabidopsis lineages. Nucleotide substitution, tandem gene duplication, and gross chromosomal rearrangement appear to proceed substantially more slowly in Populus than in Arabidopsis. Populus has more protein-coding genes than Arabidopsis, ranging on average from 1.4 to 1.6 putative Populus homologs for each Arabidopsis gene. However, the relative frequency of protein domains in the two genomes is similar. Overrepresented exceptions in Populus include genes associated with lignocellulosic wall biosynthesis, meristem development, disease resistance, and metabolite transport.

To B or Not to B a flower: the role of DEFICIENS and GLOBOSA orthologs in the evolution of the angiosperms.

DEFICIENS (DEF) and GLOBOSA (GLO) function in petal and stamen organ identity in Antirrhinum and are orthologs of APETALA3 and PISTILLATA in Arabidopsis. These genes are known as B-function genes for their role in the ABC genetic model of floral organ identity. Phylogenetic analyses show that DEF and GLO are closely related paralogs, having originated from a gene duplication event after the separation of the lineages leading to the extant gymnosperms and the extant angiosperms. Several additional gene duplications followed, providing multiple potential opportunities for functional divergence. In most angiosperms studied to date, genes in the DEF/GLO MADS-box subfamily are expressed in the petals and stamens during flower development. However, in some angiosperms, the expression of DEF and GLO orthologs are occasionally observed in the first and fourth whorls of flowers or in nonfloral organs, where their function is unknown. In this article we review what is known about function, phylogeny, and expression in the DEF/GLO subfamily to examine their evolution in the angiosperms. Our analyses demonstrate that although the primary role of the DEF/GLO subfamily appears to be in specifying the stamens and inner perianth, several examples of potential sub- and neofunctionalization are observed.

Disintegration of the scrophulariaceae.

A molecular systematic study of Scrophulariaceae sensu lato using DNA sequences of three plastid genes (rbcL, ndhF, and rps2) revealed at least five distinct monophyletic groups. Thirty-nine genera representing 24 tribes of the Scrophulariaceae s.l. (sensu lato) were analyzed along with representatives of 15 other families of Lamiales. The Scrophulariaceae s.s. (sensu stricto) include part or all of tribes Aptosimeae, Hemimerideae, Leucophylleae, Manuleae, Selagineae, and Verbasceae (= Scrophularieae) and the conventional families Buddlejaceae and Myoporaceae. Veronicaceae includes all or part of tribes Angelonieae, Antirrhineae, Cheloneae, Digitaleae, and Gratioleae and the conventional families Callitrichaceae, Globulariaceae, Hippuridaceae, and Plantaginaceae. The Orobanchaceae include tribes Buchnereae, Rhinantheae, and the conventional Orobanchaceae. All sampled members of Orobanchaceae are parasitic, except Lindenbergia, which is sister to the rest of the family. Family Calceolariaceae Olmstead is newly erected herein to recognize the phylogenetic distinctiveness of tribe Calceolarieae. The Calceolariaceae are close to the base of the Lamiales. The Stilbaceae are expanded by the inclusion of Halleria. Mimulus does not belong in any of these five groups.

Purifying selection detected in the plastid gene matK and flanking ribozyme regions within a group II intron of nonphotosynthetic plants.

In a striking contrast, matK is one of the most rapidly evolving plastid genes and also one of the few plastid genes to be retained in all nonphotosynthetic plants examined to date. DNA sequences of this region were obtained from photosynthetic and nonphotosynthetic plants of Orobanchaceae and their relatives. The resulting plastid DNA phylogeny was congruent with that recently obtained from analyses of rps2 and provided much better resolution. This phylogeny was then used to examine the relative degrees of evolutionary constraint of both the matK gene and the non-protein-coding regions that flank it inside the trnK intron. The method of subtree contrasts was introduced to compare levels of constraint. matK has evolved with a low but significant level of constraint on its amino acid sequence in both photosynthetic and nonphotosynthetic plants. Constraint is greater in photosynthetic than in nonphotosynthetic plants of this group. Domain X, thought to contain the active site of the protein, is not significantly more constrained than the rest of the protein. The portions of the flanking regions that are thought to form paired stem structures also show constraint, but in this case, there is no significant difference in degree of constraint between photosynthetic and nonphotosynthetic plants.

Independent and combined analyses of sequences from all three genomic compartments converge on the root of flowering plant phylogeny.

Plant phylogenetic estimates are most likely to be reliable when congruent evidence is obtained independently from the mitochondrial, plastid, and nuclear genomes with all methods of analysis. Here, results are presented from separate and combined genomic analyses of new and previously published data, including six and nine genes (8, 911 bp and 12,010 bp, respectively) for different subsets of taxa that suggest Amborella + Nymphaeales (water lilies) are the first-branching angiosperm lineage. Before and after tree-independent noise reduction, most individual genomic compartments and methods of analysis estimated the Amborella + Nymphaeales basal topology with high support. Previous phylogenetic estimates placing Amborella alone as the first extant angiosperm branch may have been misled because of a series of specific problems with paralogy, suboptimal outgroups, long-branch taxa, and method dependence. Ancestral character state reconstructions differ between the two topologies and affect inferences about the features of early angiosperms.

Phylogeny of seed plants based on all three genomic compartments: extant gymnosperms are monophyletic and Gnetales' closest relatives are conifers.

Efforts to resolve Darwin's "abominable mystery"-the origin of angiosperms-have led to the conclusion that Gnetales and various fossil groups are sister to angiosperms, forming the "anthophytes." Morphological homologies, however, are difficult to interpret, and molecular data have not provided clear resolution of relationships among major groups of seed plants. We introduce two sequence data sets from slowly evolving mitochondrial genes, cox1 and atpA, which unambiguously reject the anthophyte hypothesis, favoring instead a close relationship between Gnetales and conifers. Parsimony- and likelihood-based analyses of plastid rbcL and nuclear 18S rDNA alone and with cox1 and atpA also strongly support a gnetophyte-conifer grouping. Surprisingly, three of four genes (all but nuclear rDNA) and combined three-genome analyses also suggest or strongly support Gnetales as derived conifers, sister to Pinaceae. Analyses with outgroups screened to avoid long branches consistently identify all gymnosperms as a monophyletic sister group to angiosperms. Combined three- and four-gene rooted analyses resolve the branching order for the remaining major groups-cycads separate from other gymnosperms first, followed by Ginkgo and then (Gnetales + Pinaceae) sister to a monophyletic group with all other conifer families. The molecular phylogeny strongly conflicts with current interpretations of seed plant morphology, and implies that many similarities between gnetophytes and angiosperms, such as "flower-like" reproductive structures and double fertilization, were independently derived, whereas other characters could emerge as synapomorphies for an expanded conifer group including Gnetales. An initial angiosperm-gymnosperm split implies a long stem lineage preceding the explosive Mesozoic radiation of flowering plants and suggests that angiosperm origins and homologies should be sought among extinct seed plant groups.

The effect of relaxed functional constraints on the photosynthetic gene rbcL in photosynthetic and nonphotosynthetic parasitic plants.

The photosynthetic gene rbcL has been lost or dramatically altered in some lineages of nonphotosynthetic parasitic plants, but the dynamics of these events following loss of photosynthesis and whether rbcL has sustained functionally significant changes in photosynthetic parasitic plants are unknown. To assess the changes to rbcL associated with the loss of functional constraints for photosynthesis, nucleotide sequences from nonparasitic and parasitic plants of Scrophulariales were used for phylogeny reconstruction and character analysis. Plants in this group display a broad range of parasitic abilities, from photosynthetic ("hemiparasites") to nonphotosynthetic ("holoparasites"). With the exception of Conopholis (Orobanchaceae), the rbcL locus is present in all parasitic plants of Scrophulariales examined. Several holoparasitic genera included in this study, including Boschniakia, Epifagus, Orobanche, and Hyobanche, have rbcL pseudogenes. However, the holoparasites Alectra orobanchoides, Harveya capensis, Harveya purpurea, Lathraea clandestina, Orobanche corymbosa, O. fasciculata, and Striga gesnerioides have intact open reading frames (ORFs) for the rbcL gene. Phylogenetic hypotheses based on rbcL are largely in agreement with those based on sequences of the nonphotosynthetic genes rps2 and matK and show a single origin of parasitism, and loss of photosynthesis and pseudogene formation have been independently derived several times in Scrophulariales. The mutations in rbcL in nonparasitic and hemiparasitic plants would result in largely conservative amino acid substitutions, supporting the hypothesis that functional proteins can experience only a limited range of changes, even in minimally photosynthetic plants. In contrast, ORFs in some holoparasites had many previously unobserved missense substitutions at functionally important amino acid residues, suggesting that rbcL genes in these plants have evolved under relaxed or altered functional constraints.

Evolution of plastid gene rps2 in a lineage of hemiparasitic and holoparasitic plants: many losses of photosynthesis and complex patterns of rate variation.

The plastid genomes of some nonphotosynthetic parasitic plants have experienced an extreme reduction in gene content and an increase in evolutionary rate of remaining genes. Nothing is known of the dynamics of these events or whether either is a direct outcome of the loss of photosynthesis. The parasitic Scrophulariaceae and Orobanchaceae, representing a continuum of heterotrophic ability ranging from photosynthetic hemiparasites to nonphotosynthetic holoparasites, are used to investigate these issues. We present a phylogenetic hypothesis for parasitic Scrophulariaceae and Orobanchaceae based on sequences of the plastid gene rps2, encoding the S2 subunit of the plastid ribosome. Parasitic Scrophulariaceae and Orobanchaceae form a monophyletic group in which parasitism can be inferred to have evolved once. Holoparasitism has evolved independently at least five times, with certain holoparasitic lineages representing single species, genera, and collections of nonphotosynthetic genera. Evolutionary loss of the photosynthetic gene rbcL is limited to a subset of holoparasitic lineages, with several holoparasites retaining a full length rbcL sequence. In contrast, the translational gene rps2 is retained in all plants investigated but has experienced rate accelerations in several hemi- as well as holoparasitic lineages, suggesting that there may be substantial molecular evolutionary changes to the plastid genome of parasites before the loss of photosynthesis. Independent patterns of synonymous and nonsynonymous rate acceleration in rps2 point to distinct mechanisms underlying rate variation in different lineages. Parasitic Scrophulariaceae (including the traditional Orobanchaceae) provide a rich platform for the investigation of molecular evolutionary process, gene function, and the evolution of parasitism.

Do nonasterid holoparasitic flowering plants have plastid genomes?

Past work involving the plastid genome (plastome) of holoparasitic plants has been confined to Scrophulariaceae (or Orobanchaceae) which have truncated plastomes owing to loss of photosynthetic and other genes. Nonasterid holoparasites from Balanophoraceae (Corynaea), Hydnoraceae (Hydnora) and Cytinaceae (Cytinus) were tested for the presence of plastid genes and a plastome. Using PCR, plastid 16S rDNA was successfully amplified and sequenced from the above three holoparasites. The sequence of Cytinus showed 121 single base substitutions relative to Nicotiana (8% of the molecule) whereas higher sequence divergence was observed in Hydnora and Corynaea (287 and 513 changes, respectively). Secondary structural models for these 16S rRNAs show that most changes are compensatory, thus suggesting they are functional. Probes constructed for 16S rDNA and for four plastid-encoded ribosomal protein genes (rps2, rps4, rps7 and rpl 16) were used in Southern blots of digested genomic DNA from the three holoparasites. Positive hybridizations were obtained using each of the five probes only for Cytinus. For Smal digests, all plastid gene probes hybridized to a common fragment ca. 20 kb in length in this species. Taken together, these data provide preliminary evidence suggestive of the retention of highly diverged and truncated plastid genome in Cytinus. The greater sequence divergence for 16S rDNA and the negative hybridization results for Hydnora and Corynaea suggests two possibilities: the loss of typically conserved elements of their plastomes or the complete absence of a plastome.

Alternate paths of evolution for the photosynthetic gene rbcL in four nonphotosynthetic species of Orobanche.

We have determined the nucleotide sequence for the Rubisco large subunit from four holoparasitic species of Orobanche. Intact open reading frames are present in two species (O. corymbosa and O. fasciculata), whereas the remaining species (O. cernua and O. ramosa) have rbcL pseudogenes. Sequences for rbcL 5'-UTRs from species of Orobanche have few changes in the promoter and ribosome binding sites compared to photosynthetic higher plants. Comparison of rbcL 3'-UTR sequences for Nicotiana, Ipomoea, Cuscuta, and Orobanche reveal that nucleotide sequences from parasitic plants have regions capable of forming stem-loop structures, but 56-69 nt are deleted upstream of the stem-loop in the parasitic plants compared to their photosynthetic relatives. Although rbcL pseudogenes of O. cernua and O. ramosa have many large and small deletions, few indels are shared in common, implying that their common ancestor probably had an intact rbcL reading frame. Intact rbcL reading frames in O. corymbosa and O. fasciculata retain a bias of synonymous over nonsynonymous substitutions and deduced protein sequences are consistent with potentially functional Rubisco large subunit proteins. A conservative model of random substitution processes in pseudogene sequences estimates that the probability is low (P < 0.028) that these sequences would retain an open reading frame by chance. Species of Orobanche have either had recent photosynthetic ancestors, implying multiple independent losses of photosynthesis in this genus, or the rbcL gene may serve an unknown function in some nonphotosynthetic plants.

Using restriction-site variation of PCR-amplified cpDNA genes for phylogenetic analysis of tribe Cheloneae (Scrophulariaceae).

Data from restriction-site variation of three PCR-amplified chloroplast genic regions (trnK, rps2, and rbcL) were used to assess the utility of PCR-based methodology for phylogenetic reconstruction. Seventeen genera from tribe Cheloneae s.l. (Scrophulariaceae), and one genus each from Solanaceae, Acanthaceae, and Bignoniaceae, representing 32 taxa, were sampled. Phylogenetic reconstruction, based on a combined data set of 138 variable restriction sites, revealed a monophyletic clade of North American Cheloneae, which were not inconsistent with a polyphyletic Scrophulariaceae. Separate analyses of individual genie regions were unable to completely resolve the phylogeny, but were adequate for resolving relationships of major clades among the taxa sampled. We suggest that analysis of PCR-product restriction-site variation is useful for phylogenetic reconstruction above the species level.

Effects of RNA editing and gene processing on phylogenetic reconstruction.

RNA editing is a ubiquitous phenomenon affecting most mitochondrial and chloroplast, and some nuclear genomes, where mutations in genomic DNA are "corrected" in the mRNA during transcriptional processing. Most editing in plants and animals corrects T-to-C substitutions at nonsynonymous first or second base positions, and the overall effect is an mRNA and protein sequence that differs from that predicted by the DNA. It has been suggested that genomic sequences that undergo editing should not be used in phylogenetics. We contend that editing will have little or no effect on DNA-based phylogenetic reconstruction because it is an intrinsic transcriptional process that does not affect the historical information in the DNA sequence. The only effect of editing on protein-coding DNA should be an increase in the rate of T-to-C transitions. Here we test the effects of RNA editing on phylogenetic reconstruction, using two data sets with high levels of editing, plant coxII and coxIII. Even with high levels of editing, phylogenies based on DNA and edited mRNA are virtually identical. The two types of sequences should not be used in the same analysis, however, because the particular forms of the gene will tend to group together. We also examine the effects of processed paralogs--a term proposed for mRNA sequences that are reverse transcribed and reinserted into the genome as intact gene sequences. Processed paralogs result in a distinct and under-appreciated source of conflict among gene trees because of RNA editing. Analyses with unidentified processed paralogs may yield incorrect phylogenies, and the sequences may evolve at different rates if the gene has been transferred from one genetic compartment (nuclear, mitochondrial, chloroplast) to another. Although RNA editing itself is not a problem in phylogenetic reconstruction, analyses should not combine mRNAs with DNAs, and processed paralogs should be either excluded or analyzed with caution.

Transcription, splicing and editing of plastid RNAs in the nonphotosynthetic plant Epifagus virginiana.

Expression of the vestigial plastid genome of the nonphotosynthetic, parasitic flowering plant Epifagus virginiana was examined by northern analysis and by characterization of cDNAs. Probes for each of 12 plastid genes tested hybridized to all lanes of northern blots containing total RNA prepared from stems and fruits of Epifagus and from leaves of tobacco. Certain transcript patterns in Epifagus plastids are highly complex and similar to those of tobacco operons. In contrast, genes such as rps2, which have become orphaned in Epifagus as a result of evolutionary loss of formerly cotranscribed genes, show simpler transcript patterns in Epifagus than in tobacco. Sizing and sequencing of cDNAs generated by reverse transcriptase-PCR for three genes, rps12, rpl2, and clpP, show that their transcripts are properly cis- and/or trans-spliced at the same five group II intron insertion sites used in photosynthetic plants. A single, conventional C-->U edit in rps12 was found among the total of 1401 nucleotides of cDNA sequence that was determined for the three genes. An octanucleotide sequence identical to a putative guide RNA of plant organelles and perfectly complementary to the rps12 edit site itself was identified just 200 bp upstream of the edit site. These data, together with previous results from the complete sequencing of the Epifagus plastid genome, provide compelling evidence that this degenerate genome is nonetheless expressed and functional. Analysis of the putative maturase MatK, encoded by the group II intron of trnK in photosynthetic land plants but by a freestanding gene in Epifagus, leads us to hypothesize that it acts 'in trans' to assist the splicing of group II introns other than the one in which it is normally encoded.

Sequence diversity within the reovirus S2 gene: reovirus genes reassort in nature, and their termini are predicted to form a panhandle motif.

To better understand genetic diversity within mammalian reoviruses, we determined S2 nucleotide and deduced sigma 2 amino acid sequences of nine reovirus strains and compared these sequences with those of prototype strains of the three reovirus serotypes. The S2 gene and sigma 2 protein are highly conserved among the four type 1, one type 2, and seven type 3 strains studied. Phylogenetic analyses based on S2 nucleotide sequences of the 12 reovirus strains indicate that diversity within the S2 gene is independent of viral serotype. Additionally, we found marked topological differences between phylogenetic trees generated from S1 and S2 gene nucleotide sequences of the seven type 3 strains. These results demonstrate that reovirus S1 and S2 genes have distinct evolutionary histories, thus providing phylogenetic evidence for lateral transfer of reovirus genes in nature. When variability among the 12 sigma 2-encoding S2 nucleotide sequences was analyzed at synonymous positions, we found that approximately 60 nucleotides at the 5' terminus and 30 nucleotides at the 3' terminus were markedly conserved in comparison with other sigma 2-encoding regions of S2. Predictions of RNA secondary structures indicate that the more conserved S2 sequences participate in the formation of an extended region of duplex RNA interrupted by a pair of stem-loops. Among the 12 deduced sigma 2 amino acid sequences examined, substitutions were observed at only 11% of amino acid positions. This finding suggests that constraints on the structure or function of sigma 2, perhaps in part because of its location in the virion core, have limited sequence diversity within this protein.

Lack of a functional plastid tRNA(Cys) gene is associated with loss of photosynthesis in a lineage of parasitic plants.

We recently reported that the gene for chloroplast tRNA(Cys)(GCA) is a pseudogene in the plastid DNA of Epifagus virginiana, a non-photosynthetic parasitic flowering plant in the family Orobanchaceae. Since this is the only tRNA(Cys) gene in the plastid genome, and since Epifagus appears to possess a functional plastid translational apparatus, it seems probable that nuclear-encoded tRNAs are imported into plastids to effect translation. In this study we have surveyed species closely related to Epifagus to establish how widespread the loss of this tRNA gene has been. We find that Conopholis americana, another non-photosynthetic parasite, lacks the gene altogether, but that seven closely-related photosynthetic plants (both parasitic and free-living) maintain an intact chloroplast tRNA(Cys) gene. Thus, the tRNA(Cys) gene appears to have become non-functional at the same time that photosynthetic ability was lost. This may be because the levels of putatively imported tRNAs are sufficient to meet the demands of plastid gene expression under nonphotosynthetic conditions only.

Plastid translation and transcription genes in a non-photosynthetic plant: intact, missing and pseudo genes.

The non-photosynthetic, parasitic flowering plant Epifagus virginiana has recently been shown to contain a grossly reduced plastid genome that has lost many photosynthetic and chloro-respiratory genes. We have cloned and sequenced a 3.9 kb domain of plastid DNA from Epifagus to investigate the patterns of evolutionary change in such a reduced genome and to determine which genes are still present and likely to be functional. This 3.9 kb domain is colinear with a 35.4 kb region of tobacco chloroplast DNA, differing from it by a minimum of 11 large deletions varying in length from 354 bp to 11.5 kb, as well as by a number of small deletions and insertions. The nine genes retained in Epifagus encode seven tRNAs and two ribosomal proteins and are coextensive and highly conserved in sequence with homologs in photosynthetic plants. This suggests that these genes are functional in Epifagus and, together with evidence that the Epifagus plastid genome is transcribed, implies that plastid gene products play a role in processes other than photosynthesis and gene expression. Genes that are completely absent include not only photosynthetic genes, but surprisingly, genes encoding three subunits of RNA polymerase, four tRNAs and one ribosomal protein. In addition, only pseudogenes are found for two other tRNAs. Despite these defunct tRNA genes, codon and amino acid usage in Epifagus protein genes is normal. We therefore hypothesize that the expression of plastid genes in Epifagus relies on the import of nuclear encoded tRNAs and RNA polymerase from the cytoplasm.

Loss of photosynthetic and chlororespiratory genes from the plastid genome of a parasitic flowering plant.

Photosynthesis is the hallmark of plant life and is the only plastid metabolic process known to be controlled by plastid genes. The complete loss of photosynthetic ability, however, has occurred on several independent occasions in parasitic flowering plants. Some of these plants are known to lack chlorophyll and certain photosynthetic enzymes, but it is not known to what extent changes have occurred in the genes encoding the photosynthetic apparatus or whether the plants even maintain a plastid genome. Here we report that the nonphotosynthetic root parasite Epifagus virginiana has a plastid chromosome only 71 kilobases in size, far smaller than any previously characterized land plant plastid genome. The Epifagus plastid genome has lost most, if not all, of the 30 or more chloroplast genes for photosynthesis and most of a large family of plastid genes, the ndh genes, whose products may be involved in a plastid respiratory chain. The extensive changes in Epifagus plastid gene content must have occurred in a relatively short time (5-50 x 10(6) yr), because Striga asiatica, a related photosynthetic parasite, has a typical complement of chloroplast genes for photosynthesis and chlororespiration. The plastid genome of Epifagus has retained transcribed ribosomal RNA and ribosomal protein genes, suggesting that it expresses one or more gene products for plastid functions not related to photosynthesis.