The question on the patterns and limits of reduction of plastid genomes in nonphotosynthetic plants and the reasons of their conservation is one of the intriguing topics in plant genome evolution. Here, we report sequencing and analysis of plastid genome in nonphotosynthetic orchids Epipogium aphyllum and Epipogium roseum, which, with sizes of 31 and 19 kbp, respectively, represent the smallest plastid genomes characterized by now. Besides drastic reduction, which is expected, we found several unusual features of these “minimal” plastomes: Multiple rearrangements, highly biased nucleotide composition, and unprecedentedly high substitution rate. Only 27 and 29 genes remained intact in the plastomes of E. aphyllum and E. roseum—those encoding ribosomal components, transfer RNAs, and three additional housekeeping genes (infA, clpP, and accD). We found no signs of relaxed selection acting on these genes. We hypothesize that the main reason for retention of plastid genomes in Epipogium is the necessity to translate messenger RNAs (mRNAs) of accD and/or clpP proteins which are essential for cell metabolism. However, these genes are absent in plastomes of several plant species; their absence is compensated by the presence of a functional copy arisen by gene transfer from plastid to the nuclear genome. This suggests that there is no single set of plastid-encoded essential genes, but rather different sets for different species and that the retention of a gene in the plastome depends on the interaction between the nucleus and plastids.

Plastids are the semiautonomous organelles that possess their own genome inherited from the cyanobacterial ancestor. The primary function of plastids is photosynthesis so the structure and evolution of plastid genomes are extensively studied in photosynthetic plants. In contrast, little is known about the plastomes of nonphotosynthetic species. In higher plants, plastid genome sequences are available for only three strictly nonphotosynthetic species, the liverwort Aneura mirabilis and two flowering plants, Epifagus virginiana and Rhizanthella gardneri. We report here the complete sequence of a plastid genome of nonphotosynthetic mycoheterotrophic orchid Neottia nidus-avis, determined using 454 pyrosequencing technology. It was found to be reduced in both genome size and gene content; this reduction is however not as drastic as in the other nonphotosynthetic orchid, R. gardneri. Neottia plastome lacks all genes encoding photosynthetic proteins, RNA polymerase subunits but retains most genes of translational apparatus. Those genes that are retained have an increased rate of both synonymous and nonsynonymous substitutions but do not exhibit relaxation of purifying selection either in Neottia or in Rhizanthella.

Plastid genomes of nonphotosynthetic plants represent a perfect model for studying evolution under relaxed selection pressure. However, the information on their sequences is still limited. We sequenced and assembled plastid genome of Petrosavia stellaris, a rare mycoheterotrophic monocot plant. After orchids, Petrosavia represents only the second family of nonphotosynthetic monocots to have its plastid genome examined. Several unusual features were found: retention of the ATP synthase genes and rbcL gene; extensive gene order rearrangement despite a relative lack of repeat sequences; an unusually short inverted repeat region that excludes most of the rDNA operon; and a lack of evidence for accelerated sequence evolution. Plastome of photosynthetic relative of P. stellaris, Japonolirion osense, has standard gene order and does not have the predisposition to inversions. Thus, the rearrangements in the P. stellaris plastome are the most likely associated with transition to heterotrophic way of life.

Although plastid genomes of flowering plants are typically highly conserved regarding their size, gene content and order, there are some exceptions. Ericaceae, a large and diverse family of flowering plants, warrants special attention within the context of plastid genome evolution because it includes both non-photosynthetic and photosynthetic species with rearranged plastomes and putative losses of “essential” genes. We characterized plastid genomes of three species of Ericaceae, non-photosynthetic Monotropa uniflora and Hypopitys monotropa and photosynthetic Pyrola rotundifolia, using high-throughput sequencing. As expected for non-photosynthetic plants, M. uniflora and H. monotropa have small plastid genomes (46 kb and 35 kb, respectively) lacking genes related to photosynthesis, whereas P. rotundifolia has a larger genome (169 kb) with a gene set similar to other photosynthetic plants. The examined genomes contain an unusually high number of repeats and translocations. Comparative analysis of the expanded set of Ericaceae plastomes suggests that the genes clpP and accD that are present in the plastid genomes of almost all plants have not been lost in this family (as was previously thought) but rather persist in these genomes in unusual forms. Also we found a new gene in P. rotundifolia that emerged as a result of duplication of rps4 gene.

Although most plant species are photosynthetic, several hundred species have lost the ability to photosynthesize and instead obtain nutrients via various types of heterotrophic feeding. Their plastid genomes markedly differ from the plastid genomes of photosynthetic plants. In this work, we describe the sequenced plastid genome of the heterotrophic plant Rhopalocnemis phalloides, which belongs to the family Balanophoraceae and feeds by parasitizing other plants. The genome is highly reduced (18,622 base pairs vs. approximately 150 kbp in autotrophic plants) and possesses an extraordinarily high AT content, 86.8%, which is inferior only to AT contents of plastid genomes of Balanophora, a genus from the same family. The gene content of this genome is quite typical of heterotrophic plants, with all of the genes related to photosynthesis having been lost. The remaining genes are notably distorted by a high mutation rate and the aforementioned AT content. The high AT content has led to sequence convergence between some of the remaining genes and their homologs from AT-rich plastid genomes of protists. Overall, the plastid genome of R. phalloides is one of the most unusual plastid genomes known.