2.1 DNA Isolation, genome sequencing and mapping of reads to reference plastid genome

Raw reads were mapped to the reference wheat plastid genome (accession number NC_002762). The number of reads mapped was 1,195,172, which represents 1.1% of the total reads. The read depth averaged 1,229X coverage across the genome. The assembled plastid genome of the Giza 168 wheat cultivar is 133,873 bp, which is 672-bp smaller than the published genome for the “Chinese Spring” cultivar [2]. Mapping of the Giza 168 genome to the “Chinese Spring” genome identified three DNA sequences of 332, 131 and 131 bp that are absent from the Giza 168 cultivar (Fig. 1). Other shorter sequences were also found only in the published reference plastid genome. The 332, 131 and 131 bp DNA sequences are located at positions 6,164-6,495, 83,918-84,048 and 130,960-131,090 of the reference genome, respectively. Alignment of these two genomes generated gaps within the plastid genome sequence of the Giza 168 cultivar (Fig. 1). BlastN analyses of the three extra sequences from the wheat reference plastid genome to the NCBI database indicated 100% sequence identity of these fragments to plastid genome of rice (Oryza sativa, Japonica group) as well as the published “Chinese Spring” reference wheat genome (Table 1). The next best Blast hits were to another rice species, Oryza rufipogon with 99% identity. Blast hits to plastid genomes of other cereals were not detected. To further confirm that the published wheat plastid genome sequence is contaminated with rice sequence, the sequences from the wheat GZ168 cultivar flanking the three gaps were blasted to the NCBI database (Table 2). These sequences are located in the plastid genome of cultivar GZ168 between 5,957-6,288 (332 bp), 83,385-83,515 (131 bp) and 130,192-130,322 (131 bp) bases (Fig. 1). The results indicated 100% sequence identity to plastid genome sequences of hexaploid wheat and other members of the Triticeae, while no Blast hits were detected to any of the rice plastid genome sequences available in the GenBank. Annotation of the corrected wheat plastid genome confirmed the gene content and order from the published genome (Fig. 2).

The cause of the contamination of the wheat plastid genome sequence with rice DNA is unknown, however, other cases of errors in published plastid genome sequences have been detected in the past. For example, in the plastid genome of tobacco [6], sequencing errors of 119 bp were reported seven years after it was published [7]. In the case of tobacco, 90 bp were missed because a small AluI restriction fragment was missed in the cloning and sequencing strategy. One possible explanation for the error in the previously published wheat plastid genome is that the genome was assembled using the rice genome as a reference, and after completion, the entire rice sequences were not removed. Another possibility is that during the cloning of the wheat plastid genome, there was contamination of rice DNA. No matter what the explanation is, we recommend that the corrected wheat genome sequence (accession number KJ592713) be used in all future studies.