Identification of pathogens associated with apple mosaic symptom

Abstract: 【Objective】Apple(Malus domestica Borkh.) is one of the most widely grown fruit crops worldwide. China is the largest producer of apple in the world, in terms of both production and planting area.However, viral diseases can significantly reduce plant growth, development, yield, and even fruit quality.Apple trees with mosaic disease develop pale yellow to bright cream-colored irregular spots or bands along major veins on young leaves as they expand in the spring, and the symptomatic leaves drop prematurely. Retardation of growth and general declining of apple orchards have been noted, and bud set is severely affected in some cultivars. Apple mosaic disease has been known to inhibit plant growth and reduce yield. Apple mosaic virus(Ap MV) was named after the symptoms in apple, the first host in which the disease was described. Ap MV is usually regarded as the causative agent of apple mosaic disease during diagnosis. Wei did a series of experiments on apple mosaic disease in China, showing it was viruscaused disease. However, the identity of the virus(es) causing this disease remains unclear. More recent-ly, unbiased detection methods have been published based on deep sequencing technology. These methods provide a powerful and generic front-line screen that is especially well suited to virus detection.This type of approach does not require the purification of virions or specific viral nucleic acids, which are the steps that limit the range of viruses that can be discovered by conventional means. Mining the RNA sequence data can help to identify the dominant variants of viral species and give an indication of the types and frequency of viruses present in infected tissues. Here, we describe the deep sequencing of two small RNA(s RNA) libraries to examine the occurrence of viruses in symptomatic and asymptomatic leaf tissues sampled from apple orchards, in an effort to identify and characterize the causal agent(s) of mosaic disease symptoms in apple. 【Methods】1. Sample collection. Four major production areas in northern China with a continental monsoon climate, including Yan'an and Weinan of Shaanxi province,Linyi of Shandong province, and Beijing, were selected to collect samples. To identify the viral pathogens associated with apple mosaic disease, one-hundred and forty samples(111 samples with mosaic symptoms and 29 samples with no visible symptoms) were collected. Two apple leaves with and without apple mosaic symptoms were selected from the later samples for deep sequencing and analysis. 2. s RNA library construction and sequencing, read processing, and sequencing analysis. Total RNA was extracted using TRIzol?(Invitrogen, Carlsbad, CA) using the manufacturer's recommended protocol. RNA was quantified by Nanodrop?(Thermo scientific, Wilmington, DE), and its integrity was verified by gel electrophoresis. si RNA library construction was performed following the protocol of NEBNext Multiplex Small RNA Library Prep Set for Illumina?. Deep sequencing was done on the Illumina Solexa platform(San Diego, CA) using the standard protocol. s RNA sequence processing, assembly, and virus genome identification were performed using a custom bioinformatics pipeline. Briefly, raw Illumina s RNA reads were first processed to remove adaptors and low-quality sequences, and the reads were then assigned to individual samples based on the index sequences they contained. Reads between 18 and 26 nt were selected for further analysis. s RNAs were mapped to the apple genome using Bowtie to eliminate the apple genome sequences. The retained s RNAs were identified by Bowtie searching against the database of viral genomic reference sequences. Also, the remaining s RNAs from each sample were then assembled de novo using Velvet with a wide range of 12–22 k-mer in order to optimize the length and the number of the contigs, and 17 k-mer(setting for the software) turned out to be the appropriate parameter. The assembled contigs derived from Velvet were used as queries to search the Gen Bank nt, nr, and Gen Bank Virus Ref Seq databases, both nucleotide and protein sequences(version: viral.1.1.genomic.fna, viral.1.protein.faa), using the BLAST program. Contigs with significant similarity to known plant virus sequences in the databases were identified as candidate virus sequences. RT-PCR, cloning, and sequencing.The primers for each individual virus were designed by Primer PREMIER 5.0. To confirm the identity of the amplified products, PCR products were analyzed on 2% agarose/TAE gels and stained with DNAGREEN(TIANDZ, Beijing, China). The amplified DNA fragments were purified using a PCR purification kit(Tiangen) and then cloned into the p MD18-T vector(Ta Ka Ra). Positive clones were isolated from Escherichia coli DH5 a and sequenced using an automated DNA sequencer(ABI Prism 3730 XL DNA Analyzer; ABI, Carlsbad, CA, USA), and the identities of the viral sequences were verified by a BLAST search against the NCBI nucleotide databases(http://blast.ncbi.nlm.nih.gov/Blast).【Results】1. Deep sequencing and composition of s RNAs. Samples of apple leaves showing mosaic disease symptoms and asymptomatic leaves were collected for deep sequencing and analysis. Libraries of total s RNAsfrom these two samples were generated and sequenced, yielding approximately 20 million sequence reads per library. s RNAs were then aligned to the apple genome sequence. This analysis showed that the majority of filtered reads were derived from the host genome, genomic s RNAs comprised 9 068 239 reads in the symptomatic sample, and 4 860 802 reads in the asymptomatic sample. The remaining s RNA reads that were not mapped to the apple genome were retained for further analysis. 2. Composition of s RNAs population and de novo assembly of s RNAs. The s RNAs that could not be mapped to the host genome were identified by Bowtie searching against the database of viral genomic reference sequences to obtain the candidate virus species. As a result, there were 15 575 and 7 711 s RNAs from the symptomatic and asymptomatic leaves, respectively, which could be mapped to viral genomic sequences in Ref Seq.Following analysis, sequences related to some phages, along with some insect virus sequences, were found to predominate over the plant virus sequences. Of the plant viral reads, 1 834 and 1 152 reads could be mapped to apple stem grooving virus(ASGV) in symptomatic and asymptomatic leaves, respectively, which had the highest relative representation. Apart from the reads that were mapped to ASGV,905 reads of the plant viral reads from the symptomatic leaves could be mapped to apple stem pitting virus(ASPV), 557 to apricot latent virus(ALV), 194 to prunus necrotic ringspot virus(PNRSV), 134 to the alfalfa mosaic virus(Al MV) genome whose genome sequence is highly homologous to Ap MV, especially in RNA1 and RNA2(Shiel and Berger 2000), and 22 reads could be mapped to apple chlorotic leaf spot virus(ACLSV). In the asymptomatic leaf library, the species of viruses present were similar to those in the symptomatic leaf library, but the number of s RNAs of certain viruses detected was smaller. Of the plant viral reads in the asymptomatic leaves, 98 reads were mapped to ASPV, 71 to ALV, 4 to PNRSV,21 to ACLSV, and only 1 to Al MV. The virus species detected included ASGV, ASPV, ACLSV, PNRSV,ALV, and Al MV initially based on the number of mapped reads in the two libraries, and further assays needed to be performed to verify the status of each virus species. After the assembling the 18-26 nt short reads into longer contigs by the Velvet software, two plant virus contigs were found to present in the symptomatic leaf assembly, ASPV with a length of 71 nucleotides, and ASGV with a length of 57 nucleotides. RT-PCR verification and field surveys for the presence of seven viruses. RT-PCR assays were used to verify the presence of the virus species in the two samples detected with deep sequencing. We obtained fewer virus types from the RT-PCR assays compared with the deep sequencing results. We detected Ap MV, PNRSV, ASGV, ASPV, ACLSV, ALV, and Al MV. It turned out to be positive for ASGV,ASPV, ACLSV, and PNRSV in the symptomatic sample, while positive only for ASGV and ASPV in the asymptomatic leaves. These results show that apple mosaic disease may result from a multiple virus infection that does not include Ap MV. We also investigated the field distribution of virus after verifying the deep sequencing result. This enabled us to determine the virus distribution in apple leaves showing mosaic disease symptoms in the major apple production areas in an attempt to uncover the relationship between mosaic symptom and virus distribution among samples. Also, the representativeness of the deep sequencing samples could be estimated. The results of field survey indicated mixed infection of two or more viruses was common in samples displaying mosaic symptoms. The infection rate of PNRSV in symptomatic samples(90%)was higher than that in the asymptomatic ones(7%). ACLSV and ASGV could be detected in some asymptomatic samples but at lower rates compared to the symptomatic samples. In contrast, the infection rate of ASPV in asymptomatic leaves was 41%, higher than in symptomatic leaves. The viral field distribution may declare the representativeness of the deep sequencing samples.