Complete Chloroplast Genome Sequence of a White Spruce (Picea glauca, Genotype WS77111) from Eastern Canada

Over tens of millions of years, conifers such as the white spruce (Picea glauca) have evolved to cope with adverse environmental conditions (1, 2), such as prolonged drought and increased pressure from forest insect pests (3). Plants have three different genomes, namely, a nuclear, a mitochondrial, and a plastid (i.e., chloroplast) genome. In general, chloroplast genomes are derived from the ancestral genomes of the microbial endosymbiont from which these organelles originated (4). The nuclear genome of P. glauca (genotype WS77111) was published in 2015 (5).

A P. glauca (genotype WS77111) needle tissue sample was collected in southeastern Ontario (44°19′48″N, 78°9′0″W; elevation, 250 m). Genomic DNA was extracted from 60 g of tissue by Bio S&T using an organelle exclusion method yielding 300 μg of high-quality purified nuclear DNA, as previously described (6). The sample was sequenced at Canada’s Michael Smith Genome Sciences Centre (GSC). Here, we report on the assembled and annotated chloroplast genome sequence of this genotype.

To sequence the sample, genomic DNA libraries were constructed according to the plate-based and paired-end library protocols at the GSC on a Microlab Nimbus liquid-handling robot (Hamilton, USA). Briefly, 1 μg of genomic DNA was sonicated (Covaris LE220) in 62.5 μl to 400 bp and purified with PCRClean DX magnetic beads (Aline Biosciences). Illumina sequencing adapters were ligated overnight at 16°C. Pooled libraries were sequenced with paired-end 250-bp reads on an Illumina HiSeq 2500 instrument in rapid mode. Using this protocol, four libraries were generated, sequencing approximately 400 million reads from each one.

To assemble this genome sequence, we generated various random subsamples of read pairs from one lane of one library (i.e., 42,881,319 read pairs), producing subsets with 21×, 43×, 88×, 172×, 345×, 711×, 1,219×, and 5,619× coverage of the chloroplast genome. Each subset was assembled with ABySS v2.1.0 (7) (using the parameters k = 128 and kc = 3). Due to the large number of chloroplasts per cell, the chloroplast genome would be sequenced at a very high coverage over a full lane of data. Therefore, by subsampling the full data set, the coverages of the nuclear and mitochondrial genomes were lowered to a level where these sequences do not assemble well, while the coverage of the chloroplast genome was still sufficient enough for a high-quality assembly. The 43×, 88×, and 172× subsets produced the best ABySS assemblies (N₅₀ lengths, 3,692, 1,313, and 949 bp, respectively), as determined by a QUAST analysis (v5.0.0) (8). For comparison, we used the white spruce admix (PG29) chloroplast genome (NCBI GenBank accession number NC_028594) (9), the published chloroplast genome that is most closely related to the WS77111 genotype. The use of this admix as a reference was established previously (10), as it is a naturally occurring ingress of Picea glauca, Picea engelmannii, and Picea sitchensis (5). We then performed additional ABySS assemblies with various k and kc parameters using these three subsets (k = 96, 112, 128, 144, and 160; kc = 3 and 4). The assembly with the fewest aligning contigs (n = 14) and fewest misassemblies (43×; k = 96; kc = 3) was chosen for further scaffolding with the PG29 chloroplast genome, using LINKS v1.8.5 (11), thereby joining the contigs into one piece. We then used Sealer v2.1.0 (12) to close the scaffold gaps. To be consistent with previously published chloroplast genomes when reporting gene annotations, we adjusted the start position of our assembly using BLAST v.2.7.1 (13) and polished the final assembly with Pilon v1.22 (14), using BWA v0.1.7 (15) for read alignment.

The complete genotype WS77111 chloroplast genome is 123,421 bp long, with a G+C content of 38.74%. Using GeSeq v1.65 (16) with several Picea sp. chloroplast genomes as references (9, 10), we annotated 114 genes, namely, 74 protein-coding, 36 tRNA-coding, and 4 rRNA-coding genes. Five genes (rps12, petB, petD, rpl16, and psbZ) required manual annotation. The genome map in Fig. 1 was generated using OrganellarGenomeDRAW v1.2 (17).

The assembly of this new chloroplast genome will enable further analysis of the phylogeny and genetics of Picea spp.

This work was supported by funds from Genome Canada, Genome BC, and Genome Quebec as part of the Spruce-Up (www.spruce-up.ca) (243FOR) and AnnoVis (281ANV) projects.