Chloroplasts are semi-autonomous organelles believed to have developed from free-living photosynthetic bacteria 12 They are members of a diverse and flexible family of organelles called plastids that are responsible for photosynthesis plus other essential biosynthetic pathways and cellular functions. Plastids have maintained a small remnant genome with a species-specific number of genes mostly involved in photosynthesis and gene expression. The full function and development of all the plastid types, however, requires thousands of nuclear encoded gene products.

Maize is an agriculturally important monocot grass that has served as a genetic model system for decades 3, is the focus of a major genome project 45, and is especially valuable for the study of chloroplast biology 6. Other members of the grass family (Poaceae) also offer unique opportunities to study differential plastid gene expression. The presence of closely related species with either C3 or C4 photosynthetic capabilities enables comparison of dimorphic C4 chloroplast development to monomorphic C3 species. In addition, grasses such as switchgrass (Panicum virgatum) and miscanthus (Miscanthus sinensis) are the subjects of an increased focus on cellulosic ethanol production 7.

To date, most Poaceae chloroplast gene expression studies have concentrated on proteomic or physiologic assays or focused on post-transcriptional modification and regulation of transcripts, 8910. While these approaches elucidate end point development of transcribed genes, they rarely illuminate the transcription activity of a particular gene. q2(RT)PCR is a maturing tool sensitive enough to detect the existence of small amounts of nucleic acid 11. This allows for in-depth, comprehensive investigations into transcript abundance and offers a useful tool to help elucidate the relationship between transcription, translation, and expression.

q2(RT)PCR is dependent upon the amplification and quantification of a single amplicon that makes primer design and amplification conditions key to the success of an experiment. Substantial time and resources may be spent in the design, testing, and subsequent reworking of primers for optimal efficiency 12.

In this paper we describe the development and empirical optimization of primer pairs to amplify each rRNA and mRNA from maize plastids. Primer sequences, optimal annealing temperatures, and extension times are reported. In addition, each primer set was tested, in silico, against published plastome sequences and in vitro against switchgrass and miscanthus transcripts. Using the conditions optimized for maize, all primer pairs successfully produced a single amplicon for these two grass species.

One of the challenging and time consuming aspects of q2(RT)-PCR is the design and optimization of primer pairs which yield single amplicons. The aim of this study was to design and optimize a comprehensive set of plastid specific primers for q2(RT)PCR specific enough to yield robust steady state transcript data from maize yet flexible enough to recognize transcripts from multiple members of the family Poaceae. To accomplish this, primers were designed against highly homologous protein coding regions from the maize plastome, as well as rRNA genes. Primers were designed to produce amplicons 75–150 bp in length, have similar annealing temperatures, and were carefully evaluated for favorable melting temperatures to insure a lack of intra-molecular folding. Each primer set was qualitatively examined and optimized with endpoint RT-PCR as seen in Figure 1(A–E). These optimal conditions were then utilized in q2(RT)-PCR for each primer pair to verify the amplification of single amplicons via melt curves and their ability to produce quantitative data as seen in figure 1(F) and 1(G). The complete set of primer pairs yielding single amplicons and their optimized conditions is listed in Additional File 1.

Throughout the design and optimization process, we found it useful to categorize the primers based on their ease of use and optimal amplification conditions. For example, primer pairs that produced a single amplicon over a relatively large range of annealing temperatures and elongation times as shown in figure 1A were designated category 1 primers. Category 2 primer pairs produced multiple products at non-optimal annealing temperatures or elongation times and are shown in figure 1B,1C,1D, and 1E; category 3 primer pairs produced multiple amplicons due to the presence of multiple organellar genomes or primers having a high affinity to unknown additional nucleotide sequences. Category 1 primer pairs consistently were the most forgiving of strict protocol adherence. Elongation and annealing temperatures could be adjusted with no visible endpoint differences. Category 2 primer pairs required stricter elongation times and annealing temperatures for optimal single product results. Subcategory 2a includes primer pairs where an elongation time greater than 30 seconds resulted in multiple erroneous amplicons. Subcategory 2b includes primer pairs requiring specific annealing temperatures either to generate the most product as seen in Figure 1D or to produce only one product as shown in Figure 1E. One lesson learned is that some primer pairs initially in the 2b subcategory could be moved to category one if the primers were re-designed such that their individual melting temperatures were more similar. Figure 1B illustrates inconsistent and nonspecific product bands from the first pair of primers designed for psaB that had individual primer melting point temperatures that differed by 4°C. The subsequent set of primers designed for psaB have more similar melting temperatures and optimized end point results can be seen in Figure 1C. The third category of primer pairs consist of gene sequences where a highly similar homolog exists either in the nucleus or mitochondria (4.5S, 16S, atpE, rpl2, and 23S) and\or genes where multiple attempts to design primer pairs produced unacceptable products (rps12 and rps7). Two category 3 examples, 4.5S and 16S, are shown in Figure 2. For 4.5S, an end-point reaction utilizing suboptimal primers appeared to produce a single band but a melt peak curve clearly showed three very similar but distinct products. A subsequent redesign of the 3' primer for 4.5S and 23S resulted in single amplicons verified by their respective melt curve signatures and are reported as the 3' primers in Additional File 1. 16S, on the other hand, should also produce multiple bands according to screens of the maize genomes but neither end-point nor q2(RT)PCR can distinguish separate products. We could not successfully design some category 3 primer pairs. They either repeatedly produced multiple amplicons or could not be verified as amplified from the chloroplast genome. These primer pairs have been excluded from Additional File 1 or do not have annealing temperatures associated with them.

Some plastid genes with intra-cellular homologs could be specifically amplified when primers were designed to take advantage of small non-similar regions within the genes. This was achieved by comparing the homology between the genes and designing primers to recognize unique regions within chloroplast genes. Primers were then compared to other competing homologs, if the primer homology was less than 50% and there were no 3' matching nucleotides, then the design process was completed and the primers synthesized. The similarity estimation was based on the BLAST 13 percent identity score and not the E value which can be inflated when sequence sizes are small. One exception to this design strategy was the atpA gene primers which have >50% homology and should anneal to mitochondrial homologous regions. The possible mitochondrial amplicon however was predicted to be 125,630 base pairs long. An experimental end-point check of the atpA primers amplified the desired 77 base pair chloroplast amplicon and the single product was verified by a q2(RT)PCR melt curve. Multi-genomic genes exhibiting high homology but whose primer annealing sites and subsequent amplicons are plastid specific when analyzed in this manner are reported in Additional File 1 as category 1 or category 2 genes. Multi-genomic genes with homologies below 50% and whose primer pairs were determined to amplify only chloroplast genomic material are also reported in Additional File 1 as category 1 or category 2 genes.

The homology between plastomes from closely related plant species is well documented 141516. This homology has been previously utilized to design PCR primer pairs that would produce amplicons across genera 1718. Based on this, we hypothesized that our primer pairs should amplify similar regions of the same genes from other members of the grass family (Poaceae). To test this, an in silico comparison of the primers to grasses with completed plastomes was conducted. The possibility of a successful reaction was judged based on overall percent homology and the presence of mismatched bases at the 3' end of the primers. This comparison suggested that primer homology and subsequent amplification success would closely follow established taxonomic designations (Figure 3 and Additional File 2). Members of the PACCAD clade (S. officinarum and S. bicolor), to which maize belongs, have the greatest possibility of successful amplifications for all primer pairs while there is a noticeable drop in predicted success among members of the BEP clade (H. vulgare, T. aestivum, L. perrene). Arabidopsis was included as an outgroup, with only 22% of the primers predicted to successfully anneal to the intended target gene.

The in silico results were tested in vitro with two species in the PACCAD clade with unsequenced (or unpublished) plastomes – switchgrass (Panicum virgatum) and miscanthus (Miscanthus. sinensis). As predicted, we found that the primers and their empirically determined optimal conditions worked well across these genera and produced reliable quantitative data [Figure 4]. Representative data from the chloroplast Large Single Copy area (rpoC1 and matK), Small Single Copy area (psaC), and the Inverted Repeat (16S) areas are illustrated in Figure 5. As shown, these primer sets and amplification conditions eliminated the need for additional primer design for these species.

As seen in Figure 5, there are obvious differences in expression amounts and amplification slopes between maize (A), switchgrass (B) and miscanthus (C). Although equal amounts of total RNA were added for each round there was no attempt to normalize the data to account for any differences that may exist between species. It also must be stressed that the primer pairs have been optimized for the maize plastome and although all were checked and verified to produce single amplicons for both switchgrass and miscanthus, the reported annealing temperatures were not optimized to ensure the greatest signal was detected for these two species. Individual reagents, specific types of equipment and reaction volumes may require some user-specific modifications for optimal performance for species specific work.

Primers were designed and empirically optimized to quantify plastid transcripts from Zea mays. Due to the highly conserved nature of plastomes within the family Poaceae, this set of primers works equally as well for the majority of targeted genes of other species in the Poaceae family, particularly members of the PACCAD clade. This primer set and the conditions published in this report should reduce the time and cost associated with primer design and testing for future plastid q2(RT)PCR experiments.

RT-PCR: reverse transcriptase PCR; q2(RT)PCR: quantitative real time reverse transcriptase PCR

The authors declare that they have no competing interests.

RMS designed the primers, performed the q2(RT)PCR reactions, interpreted reaction results and prepared the manuscript, SND processed and performed the q2(RT)PCR for the majority of the switchgrass samples, ABC was involved in the conception and design of the project, revised and edited the manuscript, and is the PI of the laboratory.