Phylogenomics Of Orchidaceae Based On Plastid Genomes

Recent results indicated that Orchidaceae originated in the Late Cretaceous.It has been proved that in extinction events occurred 66 million years ago(Ma),orchids independently experienced the whole genome duplication(WGD)and emerged as a variety of life forms.The phylogenetic relationships have been proposed by many scientists based on different types of data,but still controversial.The chloroplast genomes are relatively conservative and moderate in substitution rate,which suggests that plastid genomes are very suitable for reconstruction of plant phylogeny.We sequnenced and assemblied 59 plastid genome using next generation high-throughput sequencing technology,combined with 56 chloroplast genome data released by NCBI,and 14 species chloroplast sequence data extracted from the matrix used in Givnish's research.We explored and reconstructed the phylogeny of Orchidaceae,through two methods(maximum likelihood method and Bayesian inference method),using 79chloroplast protein coding genes for 129 species of 19 tribes and 29 subtribes,representing all orchid subfamilies.The molecular dating of Orchidaceae was analyzed by using BEAST and integrating all 3 available fossils.Using Apostasia odorata as a reference species,114 orchids were selected for evolution rate analysis,the average synonymous substitution rate(d _S)and the average non-synonymous substitution rate(d _N)of five housekeeping genes(rpl23,rpl36,rps11,rps14,rps18)was calculated.The main results were summarized as follows:1.Results of phylogenomicsOur results indicated that Orchidaceae are subdivided into 5 subfamilies with high resolution.Apostasioideae is the first differentiated group,divided into Apostasia and Neuwiedia.Vanilloideae is a sister group of Cypripedioideae,Orchidoideae and Epidendroideae,divided into the Polanieae and the Vanilleae.Cypripedioideae is a sister group of Orchidoideae and Epidendroideae.Cypripedium diverged first in the phylogeny of the subfamily Cypripedioideae.In Orchidoideae,(Orchideae+Codonorchideae)are sister to(Cranichideae+Diurideae).The phylogenetic positions of seven tribes of Higher Epidendroideae has been well resolved.Coelia belongs to Calypsoinae.Neottieae is located in the base of Epidendroideae,while Thaieae is sister to Higher Epidendroideae.The three tribes of Lower Epidendroideae,i.e.,Sobralieae,Triphoreae and Nervilieae,are clustered into one group with high support.Our results suggested that more sampling is needed to resolved the intertribal relationships of Lower Epidendroideae.2.Diverging time estimatationThere were two long periods,i.e.,90.03-79.79 Ma and 76.31-67.9 Ma,during which there were no diversification of subfamilies of Orchidaceae.The diverging time of Apostasioideae from the remaing of Orchidaceae was 90.03 Ma(95%HPD,was 88.16-92.03 Ma),of Vanilloideae was 79.79 Ma(95%HPD,was 71.26-87.82Ma),of Cypripedioideae was 76.31 Ma(95%HPD,was 67.24-84.87 Ma),of Orchidoideae and Epidendroideae were 67.9 Ma(95%HPD,was 57.90-77.20 Ma).Diverging time of Orchidaceae of most subfamilies or subtribe agree with previous results.However,the crown group age of Vanilloideae was 67 Ma,which was about10 Ma younger or older than previous studies.Seven tribes rapidly diverged in Higher Epidendroideae from 42.01 to 33.37 Ma,during 8 Ma.3.Evolution rate and branch lengthOur results indicated that the substitution rate and length of branches of Pogonia yunnanensis(d _N=0.0641,d _S=0.2433)of Pogonia,and Vanilla aphylla(d _N=0.1285,d _S=0.3289)?V.planifolia(d_N=0.1626,d _S=0.4782)and V.pompona(d _N=0.1621,d _S=0.4596)of Vanilla and of Rhizanthella gardneri(d _N=0.1621,d _S=0.5380)of Rhizanthella are significantly higher than those of the others.The average nonsynonymous substitution rate(d _N)of Cypripedioideae accelerated with the stable speed.The average non-synonymous substitution rate(d _N)of most orchid plants in the largest orchid subfamilies(Epidendroideae and Orchidoideae)is between 0.04 and0.06,and the average synonymous substitution rate(d _S)between 0.2 and 0.3.According to the results of the phylogeny tree and the genes of the matrix,it is assumed that the length of branches is related to the number of housekeeping genes.