High-throughput sequencing of black pepper root transcriptome

Black pepper (Piper nigrum) is one of the most popular and oldest spices in the world, with culinary uses and more recently as a food preservative [1]. Due its medicinal properties, it is used in traditional medicine for its antioxidant, anti-inflammatory and anticancer properties [2, 3]. This species is the second most widely traded spice in the world, accounting for a value of 929 million US dollars from a production of 445900 tons in 2009 [4].

P. nigrum is a perennial, climbing vine indigenous to the southwestern region of India. Its production is limited to rainy tropical regions such as those found in Asia and South America. Tropical climates favor the development of diseases, and crop losses due to pests and diseases are a major constraint in the production of black pepper [1]. Resistance against the two major pathogens, Phytophtora capsici and Fusarium solani f. sp. piperis, has not been achieved by classical breeding of the germplasm of cultivated black pepper and remains a major challenge that may be amenable to improvement using plant biotechnology [1]. Despite the agricultural and economic importance of black pepper, knowledge of its genetics is presently very limited [5].

For non-model plants such as black pepper with little or no molecular information available, next-generation sequencing (NGS) technologies offer a great opportunity for the rapid access to genetic information [6–9]. In plant biology, due to large sizes of genomes and high levels of ploidy, the characterization of transcriptomes is a powerful tool for the identification of proteins. RNA-seq can sequence, in an extremely high-throughput and quantitative manner, the transcriptome of an organism or tissue [9]. Next-generation sequencing technologies generate large numbers of reads with high sampling rates of cDNA libraries, providing a deeper and more complete view of transcriptomes [8]. This transcriptome-wide description of RNA sequences identifies the coding sequence, single nucleotide polymorphisms (SNPs), other polymorphisms, splicing variants and relative levels of expression by counting reads, a method known as digital gene expression (DGE) [9, 10].

An RNA-seq analysis of the black pepper transcriptome can greatly impact our knowledge of this species. Black pepper belongs to the Piperaceae family of the Magnoliid subclass, know as basal angiosperms, a large and important family for which little genomic data is available (Figure 1). Transcriptomic data could have many ecological and biotechnological repercussions, from exploring and characterizing the evolution of this family, to defining markers such as microsatellites that are useful in breeding programs, to establishing a catalog of expressed genes related to plant defense and metabolism.

In this work, we used RNA-seq technology to analyze the transcriptome of black pepper. About 71 millions reads were generated and 22363 de novo assembled transcripts were obtained from tissue of the stem-root. The transcripts were functionally annotated and represent the first sequences derived from the transcriptome of P. nigrum. About 257 new simple sequence repeats (SSRs) were described. These results, the first dataset of sequences of the Magnoliid group, will open new perspectives for studies of diversity and ecology and will benefit breeding and genomic programs aiming to understand plant-pathogen interactions in the root, the organ that is central in the process of infection by F. solani and P. capsici, the two major pathogens of black pepper.

Biotechnological methods are being successfully employed in plant breeding [16] to confer, for example, increased resistances to pathogens. To use biotechnology in breeding programs, however, prior genetic and genomic knowledge is needed to isolate genes and characterize genotypes and phenotypes [16]. Plant genomics, especially for non-model plants, is always a challenge due to the high levels of ploidy, large genome sizes, low genome complexities and large proportions of repeat sequences. Instead of sequencing the complete genome, sequencing the transcriptome is a good alternative for rapidly and efficiently accessing the expressed genes and for characterizing phenotypes [17].

The characterization of transcriptomes can use hybrid- or sequenced-based technologies, but all hybrid-based technologies require prior knowledge of the transcriptome. In plant biology, Expressed Sequence Tags (ESTs) based on the sequencing of cDNA libraries have been a great advance and have contributed significantly to the study of plant genomics. The sequencing of cDNA libraries, however, has many limitations, such as the cost in time and money involved in cloning and Sanger sequencing. The serial analysis of gene expression (SAGE) and the improved variant SuperSAGE have been used to increase throughput and sampling, but the reliability of the identification of transcripts is one of their limitations [18]. The second generation of sequencing (2-GS) technology circumvents these limitations and generates a large number of reads or datasets of short reads in a shorter time and with a lower reagent cost per base than Sanger sequencing. 2-GS technologies have opened great prospectives for genetic studies of non-model plants [17].

The genome of black pepper is very poorly characterized, with only 184 sequences deposited in Genbank (access 3/11/2011). Previous genetic studies have focused on the diversity, phylogeny and taxonomy of cultivated black pepper and its wild relatives [19–22]. Despite the agricultural and economic importance of this spice, only DNA sequences of phylogenetics markers are available in molecular databases (Figure 1). Another difficulty for the genomic study of black pepper is the lack of a sequenced genome from the Magnoliid group, describe as basal angiosperms group [23, 24] (Figure 1). All full-sequenced genomes available in databases are from monocot or dicot plants, plants phylogenetically distant from black pepper, as show in Figure 1[23, 24].

About 20 Gb of the data can be mapped, corresponding to 71 million short reads of 50 bp. De novo assembly generated 22363 contigs, 10338 unigenes with an N50 of 168 bp. The N50 is slightly less than the 202 bp obtained in an Illumina 75 bp dataset for the Ipomoea batatas transcriptome and the 208 bp obtained in a 75 bp single- and paired-end dataset from Camellia sinensis[25, 26]. The number of contigs, however, is lower than in these other studies: 56516 and 127094 contigs for I. batatas and C. sinensis, respectively. Although that the read number is higher in our study, (71 million vs 59 and 35 million for I. batatas and C. sinensis), the contig number is inferior. This result could be explain by the read lengths were different (50 bp vs 75 pb) and SOLiD dataset is lower in quality, only 50–60% of reads mapped against a transcriptome reference [27, 28].

Estimating the amount of coverage of the transcriptome is complex due to the lack of genetic data from black pepper. Black pepper is tetraploid with 2n = 52. The size of its genome, estimated by cytogenetic studies, is about 1220 Mbp (1C = 1.25 pg), which is about ten-fold larger than the genome of Arabidopsis thaliana[29]. In A. thaliana, about 6% of the genome is transcribed, representing 41671 transcripts [18]. In the A. thaliana TGI database, however, two root-derived cDNA libraries have largely been sequenced and have identified 5609 and 5884 transcripts (11.3 and 9.3 Mbp), which seems to indicate that only 14% of predicted genes are expressed in the root [30]. Assuming a similar proportion of transcription in P. nigrum, the entire transcriptome is estimated to be 73.2 Mbp (6%) and the root transcriptome to be 10.3 Mbp (0.84%).

The cDNA library was sequenced with a high coverage per base, 62x on average, summing 3.8 Mbp. Shown as an example in Figure 4, the transcript NODE-882-2-0_930 has a good read coverage, up to 20x, and shows a high identity and a good co-linearity with an mRNA sequence from P. trichocarpa, indicating a good quality of sequencing and assembly. The comparison of predicted CDSs with the proteomes of other plants demonstrates that the percentage of sequences with significant homology is higher with dicots (51–53%) than with monocots (15–23.4%). This low degree of homology may be due to the phylogenetic distance between magnoliids and others groups (eucots and dicots) and to the different average transcript size (211 pb), as described during the sequencing of the I. batatas transcriptome [25]. The method of k-mer additive assembly is efficient in maximizing the number of transcripts, and interestingly, the proportion of transcripts with significant homology (22–35%), the N50 (129 to 236 bp) and the average length of transcript (156–214 bp) are relatively well conserved for each k-mer (19–43 bp) (Figure 2). Comparison with other studies of plant transcriptomes is difficult because the read length used with the Illumina platform is longer (75 bp) and the library is paired-end, which clearly facilitates the process of assembly, as demonstrated for I. batatas[25]. Finally, limitations of the 2-GS platform produce high numbers of short contigs, but our dataset has 2144 unigenes over 200 bp in length and about 72% of reads with a significant homology, which represent a significant advance in our biological knowledge of black pepper.

The functional annotation of the dataset of transcripts is informative for the physiology of stems and roots of black pepper. NGS datasets from plant root tissue are only available for ginseng and sweetpotato. The “response to stimulus” and “localization” categories of Figure 5 are highly represented in the transcriptomes from roots of both sweetpotato and black pepper [25]. The profile of molecular functions is conserved between the two root plants. In both, the function “transporter activity” is quite high. This feature may be explained by the fact that roots are very important in the absorption of microelements. This preliminary annotation of the transcriptome from black pepper is very important and should lead to the identification of new genes coding for transporters, transcription factors specific to root and stem tissue or proteins used for defence.

In model plants, NGS has increased the ability to generate and analyze the expression and detection of most transcripts, to identify genes and their forms of alternative splicing and to analyze DNA methylation and histone modifications on a genomic scale [6, 8]. The technology based on Sanger sequencing has generated much data from model species, and plants with sequenced genomes and transcriptomes have provided a robust database reflecting gene expression. The transcriptomic resources for non-model plants, particularly with NGS, however, are limited, especially for the analysis of expressed sequences, alternative splicing and SNPs.

The sequencing of P. nigrum using the SOLiD platform (short-reads) produced the first dataset of sequences for a non-model species belonging to the Magnoliid group and order Piperales. Assembly without a reference genome or transcriptome is considered difficult, but a partial transcriptome of the non-model species P. nigrum was assembled. This work was aided by the use of some de novo assemblies performed using NGS data from non-model plants such as chickpea Cicer arietinum L. [31], Eucalyptus grandis x Eucalyptus urophylla hybrid [32], rubber Hevea brasiliensis[33], sweetpotato Ipomoea batatas[25] and buckwheat Fagopyrum[34].

Analysis of SSRs in the data generated has great potential for applications in studies of genetic diversity, plant breeding and reproduction. The sequences produced in this study are very relevant, since little information about black pepper is available in the biological databases, and the data produced can be used in ecological studies and biotechnology.