Identification and characterization of jute LTR retrotransposons:

Abstract

Long Terminal Repeat (LTR) retrotransposons constitute a significant part of eukaryotic genomes and play an important role in genome evolution especially in plants. Jute is an important fiber crop with a large genome of 1,250 Mbps. This genome is still mostly unexplored. In this study we aimed at identifying and characterizing the LTR retrotransposons of jute with a view to understanding the jute genome better. In this study, the Reverse Transcriptase domain of Ty1-copia and Ty3-gypsy LTR retrotransposons of jute were amplified by degenerate primers and their expressions were examined by reverse transcription PCR. Copy numbers of reverse transcriptase (RT) genes of Ty1-copia and Ty3-gypsy elements were determined by dot blot analysis. Sequence analysis revealed higher heterogeneity among Ty1-copia retrotransposons than Ty3-gypsy and clustered each of them in three groups. Copy number of RT genes in Ty1-copia was found to be higher than that of Ty3-gypsy elements from dot blot hybridization. Cumulatively Ty1-copia and Ty3-gypsy may constitute around 19% of the jute genome where two groups of Ty1-copia were found to be transcriptionally active. Since the LTR retrotransposons constitute a large portion of jute genome, these findings imply the importance of these elements in the evolution of jute genome.

Introduction

Mobile genetic elements or transposons are an important component of eukaryotic genomes. Among the transposable elements, retrotransposons are the largest group.¹ They are ubiquitous in plants, though their abundance varies widely. For example, retrotransposons constitutes only 5.5% of the genomic content of Arabidopsis thaliana but more than 50% of Zea mays.^2,3 Based on the mode of transposition and propagation, retrotransposons belong to class I element of transposons.⁴ Long Terminal Repeat (LTR) retrotransposons are a subclass of retrotransposons which is sub-divided into two groups: Ty1-copia and Ty3-gypsy. Both Ty1-copia and Ty3-gypsy contain pol gene which consists of four domains: protease, integrase, reverse transcriptase and ribonuclease H (RNase H).⁵ The order of the protease, integrase and reverse transcriptase domains is the distinguishable feature between Ty1-copia and Ty3-gypsy elements. In Ty1-copia the order is protease, integrase and reverse transcriptase whereas for Ty3-gypsy it is: protease, reverse transcriptase and integrase.⁶ It is worth noting that, transposition or proliferation of LTR retrotransposons via error-prone reverse transcription is one of the key factors for the size of plant genomes and their evolution.^7,8

To date, Ty1-copia retrotransposons have been studied and characterized widely in plants like barley, tobacco, tomato, potato, rice, maize, arabidopsis, wheat, soybean, rye, citrus, strawberry and so on.⁹ A few reports have also demonstrated Ty3-gypsy elements in maize, tomato, rice, citrus and cotton genomes.^1,3,10–13 For their role in chromosomal rearrangement, fragmented gene movements and alteration of gene regulation and functions, retrotransposons are important players in the evolution of plant genomes.¹⁴ Although majority of plant retrotransposons are believed to remain transcriptionally inactive, there are reports of their expression at different developmental stages and under different biotic and abiotic stress conditions.^2,9,15

Jute (Corchorus spp.) is an important source of natural fibre, well-known for its high tensile strength, biodegradability and heat resistance. Despite its great agronomic importance, research on jute at the molecular level is insignificant. This is reflected by very little documentation on its molecular biology and so far only 1,210 sequences have been deposited in the GenBank.¹⁶ In this backdrop we have analyzed the abundance, heterogeneity and transcriptional activity of Ty1-copia and Ty3-gypsy in the jute genome. This study, in future, will pave the way to improve knowledge of major genomic components of jute better than our present level of understanding.

Results

Isolation and confirmation of Ty1-copia and Ty3-gypsy retrotransposons in jute genome.

In this study we amplified the reverse transcriptase domains of the Ty1-copia and Ty3-gypsy like retrotransposons of jute genome using degenerate primers. Length of the PCR products obtained were 280 bp for Ty1-copia and 290 bp for Ty3-gypsy retrotransposons in the jute genome (Fig. 1). After cloning of the PCR products, 36 colonies for Ty1-copia and 25 colonies for Ty3-gypsy were randomly selected and plasmids were isolated followed by sequencing. Homology based searching (BLASTx and BLASTn) revealed that all the 36 Ty1-copia sequences were very similar to those of other plant species. Among these 36 sequences 4 different types of sequences were found to be repeated twice or more. From these repeated sequence groups, a random sequence of each group was selected for further analysis. The total number of Ty1-copia sequences thus analyzed was 30. The same 30 Ty1-copia RT sequences of jute genome were named as JTEC1–JTEC30 and submitted to GenBank (GU734714–GU734743). Translation of these PCR amplified sequences implied that among the 30 Ty1-copia RT sequences, 13 (approximately 43%) sequences contained in frame stop codon(s) which inferred that these sequences do not have potential functional reverse transcriptase fragment. These stop codon containing sequences are JTEC3, JTEC10, JTEC13, JTEC14, JTEC16, JTEC20, JTEC21–JTEC24, JTEC26, JTEC28 and JTEC29. The remaining 17 amplified sequences (approximately 57%) were accounted as partial Ty1-copia reverse transcriptase sequences of jute with potential functional reverse transcriptase domain as they lack any in-frame stop codon. Among the 25 PCR amplified Ty3-gypsy clones, 17 sequences matched with Ty3-gypsy elements of other plant species. 12 of these 17 sequences were distinct from each other in their putative amino acid sequences. The sequences were named as JTEG1 to JTEG12 and were submitted to GenBank (GU592183–GU592188 and GU592190–GU592195). Eight of these 12 sequences (approximately 67%) contained in-frame stop codon(s). Stop codon containing sequences are JTEG3, JTEG5–JTEG10 and JTEG12. Due to high variability in the nucleotide sequences, multiple sequence alignments were performed on amino acid sequences using ClustalW. Although sequence motifs obtained from alignments (Figs. 2 and 3) were lacking similarity in few sequences of Ty1-copia and Ty3-gypsy retrotransposons of jute, homology based searches (BLASTn and BLASTx) confirmed these as sequences of respective groups.

Figure 1.

PCR amplified products of RT domains of Ty1-copia and Ty3-gypsy retrotransposons of jute genome. Lane 1: Ty3-gypsy PCR products, Lane 2: Ty1-copia PCR products, Lane 3: Control PCR without template, Lane 4: 1 Kb⁺ ladder. Genomic DNA was isolated from mature leaves of jute variety O-9897 and after PCR amplification, the PCR amplified products were run in 2% agarose gel.

Figure 2.

Multiple Sequence Alignments (MSA) of predicted amino acid sequences of the reverse transcriptase domains of Ty1-copia retrotransposons of jute and other species (Table 1). (a) is the Group I, (b) is the Group II and (c) is group III according to the phylogenetic tree.

Figure 3.

Multiple Sequence Alignments (MSA) of predicted amino acid sequences of the reverse transcriptase domains of Ty3-gypsy retrotransposons of jute and other species (Table 2). (a) is the Group I, (b) is the Group II and (c) is Group III according to the phylogenetic tree.

Transcriptional analysis and phylogenetic tree construction.

To identify whether the LTR retrotransposons are transcriptionally active or not in jute genome, reverse transcription-polymerase chain reaction (RT-PCR) was carried out using the degenerate primer pairs. Designing specific primers for each group of Ty1-copia and Ty3-gypsy was not possible because of the small size of reverse transcriptase domains and high similarities. RT-PCR revealed that only Ty1-copia retrotransposon is transcriptionally active in jute genome (Fig. 4). Followed by cloning and sequencing, homology based search with 18 of the RT-PCR positive cloned sequences found matches with the Ty1-copia retrotransposons of other plant species. Among these, 8 were unique in terms of their putative amino acid sequences. These sequences were named as JTECrt1 to JTECrt8.

Figure 4.

RT-PCR products of jute Ty1-copia and Ty3-gypsy retrotransposons. Lane 1: PCR from isolated RNA only as a control, Lane 2: RT-PCR of Ty1-copia, Lane 3: RT-PCR of Ty3-gypsy. RNA was isolated from the mature leaves of jute variety O-9897.

For phylogenetic relationship of jute retrotranspososn sequences with that of other species, Ty1 and Ty3 retroelements, classified at GyDB. (The Gypsy Database, http://gydb.org/index.php/Main_Page) were used (Tables 1 and 2). The consensus RT domains at the GyDB were used as query to find the nucleotide sequences of the RT domains from the NCBI. As the Ty1-copia and Ty3-gypsy elements of jute comprise a portion of the RT domain, core RT sequences obtained from GyDB were manually refined and used to infer the phylogenetic relationship. Phylogenetic analysis with 38 sequences of jute Ty1-copia (8 from transcriptionally active- and 30 from genomic Ty1-copia sequences), clustered them in 3 distinct groups (Fig. 5A). According to sequence similarity and GyDB classification, Group I sequences may be considered as “osser” like retroelements. Group II sequences are found as mixed type of retroelements: sire, tork, PyRE and orycro. Group II did not match with any specific type of retroelements and few Ty1-copia sequences of jute also remain ungrouped. Since both Group I and Group III contain RT-PCR positive jute Ty1-copia sequences, therefore it can be said that Group I and Group III sequences are transcriptionally active in jute.

Table 1.

List of Ty1-copia retrotransposon extracted from GyDB which were used for phylogenetic analysis

Element	Accession	Host
Oryco	ABA99612	Oryza sativa
Oryco	AC079604.5	Arabidopsis thaliana
Osser	X69552	Volvox carteri
PyRE1G1	AB371726	Porphyra yezoensis
PyRE10G	AB286055	Porphyra yezoensis
Retrofit	AAX96193	Oryza sativa
Sire	AAG52949	Arabidopsis thaliana
Sire	AF053008	Glycine max
Tork	TOBAA	Nicotiana tabacum
pCreto	AAZ28935	Phanerochaete chrysosporium

Table 2.

List of Ty3 retroelement sequences extracted from GyDB used for phylogenetic analysis

Element	Accession	Host
Skipper	AAC39021.1	Dictyostelium discoideum
Maggy	AF202956.1	Aspergillus fumigatus
Maggy	AAA33420.1	Magnaporthe grisea
Maggy	Y14976.1	Neurospora crassa
Maggy	AJ272171.1	Podospora anserina
Pyret	CAA77891.1	Passalora fulva
Pyret	AAG24792.1	Colletotrichum gloeosporioides
Pyret	AAA88791.1	Fusarium oxysporum
Smut	AC114899.5	Ustilago hordei
CRM	AC012327.6	Arabidopsis thaliana
CRM	AC026757	Arabidopsis thaliana
CRM	BH562012.1	Brassica oleracea
CRM	AY040832.1	Hordeum vulgare
CRM	AC131249.44	Medicago truncatula
CRM	AC022352.5	Oryza sativa
CRM	AAM94350.1	Zea mays
Geladriel	BH658798.1	Brassica oleracea
Geladriel	AF119040.1	Lycopersicon esculentum
Geladriel	AF143332.1	Musa acuminata
Geladriel	AJ508603.1	Nicotiana tomentosiformis
Reina	AL049655.2	Arabidopsis thaliana
Reina	AF541963.1	Glycine max
Reina	AP009654.1	Lotus japonicus
Reina	AC136141.25	Medicago truncatula
Reina	AJ004945.1	Pinus radiata
Reina	AF123535.1	Zea mays
Tekay	AC006837.16	Arabidopsis thaliana
Tekay	AP004965.1	Lotus japonicus
Tekay	AC119148.2	Oryza sativa
Tekay	AF448416.1	Zea mays
Athila	AF378081.1	Arabidopsis thaliana
Athila	AF378080.1	Arabidopsis thaliana
Athila	AF378079.1	Arabidopsis thaliana
Athila	AF378078.1	Arabidopsis thaliana
Calypso	AF186182.1	Glycine max

Figure 5.

(A) Phylogenetic tree of amino acid sequences of RT domain of Ty1-copia retrotransposons of jute and other species. The phylogenetic tree was constructed using NJ method and displayed using MEGA4. The bootstrap consensus tree was inferred from 1,000 replicates showing bootstrap values higher than 50%. Among the three groups based on this phylogenetic tree, Group I and Group III contain jute genomic Ty1-copia sequences and transcribed Ty1-copia sequences respectively. The transcribed sequences are denoted by ‘JTECrt’ whereas the genomic sequences are denoted by ‘JTEC’. (B) Phylogenetic tree of amino acid sequences of RT domain of Ty3-gypsy retrotransposons of jute and other species. The phylogenetic tree was constructed using NJ method and displayed using MEGA4. The bootstrap consensus tree was inferred from 1,000 replicates showing bootstrap values higher than 50%. According to this phylogenetic tree, Group I, Group II and Group III are like athila, CRM and reina type Ty3-gypsy retroelements.

Another phylogenetic tree was constructed using 12 JTEG sequences only (Fig. 5B). From this tree Ty3-gypsy sequences were clustered into three groups. As per GyDB classification, these three groups: Group I, Group II and Group III sequences resembled athila, CRM and reina type sequences.

Generally sequence similarity within and between clades of a phylogeny tree are indicated on a threshold of 80% sequence identity.¹⁷ Unfortunately, the recommendation of Wicker et al. can not be applied in our case. It is evident from the multiple sequence alignment that the partial Ty1 and Ty3 reverse transcriptase domains are highly heterogeneous and diverse. And the similarity drops even more when it comes to nucleotide sequences. However, as few sequences harbor stop codons, we could not use amino acid sequences for the construction of phylogenetic trees and classification of retrotransposons. In this regard we relied on the approach followed out by other groups.^1,18

LTR retrotransposons copy number in jute.

To determine the copy numbers of Ty1-copia and Ty3-gypsy elements of jute genome two heterogenous populations of 280 bp Ty1-copia and 290 bp Ty3-gypsy whole PCR products were used as probes for dot blot hybridization. Densitometric method was used to measure the intensities of hybridization signals. Serial dilutions of heterogeneous PCR products were used as standards against total jute genomic DNA which is approximately 1,250 Mb in length.¹⁹ Based on signal intensities copy numbers of RT genes in Ty1-copia and Ty3-gypsy elements of jute were estimated to be approximately 31,298 and 3,621 respectively per diploid jute genome (Fig. 6). Since whole PCR products were used for dot blot probe preparation and the current study focuses only on partial reverse transcriptase domains, assuming average sizes for Ty1-copia (7 kb) and Ty3-gypsy (10 kb) LTR retrotransposons,²⁰ these two groups of LTR retrotransposons, may comprise up to 18.91% of the jute genome (Ty1-copia = 16.23%; Ty3-gypsy = 2.68%).

Figure 6.

Determination of the copy number of Ty1-copia and Ty3-gypsy sequences in jute genome where (a) is for Ty1-copia and (b) is for Ty3-gypsy. For Ty1-copia 10 µg, 8 µg and 5 µg of genomic DNA corresponding to 100 ng, 50 ng, 40 ng and 25 ng of PCR products were immobilized on membrane followed by probing with non-radioactive labeled PCR probe. For Ty3-gypsy 20 µg and 10 µg of genomic DNA corresponding to 200 ng, 100 ng and 50 ng of PCR products were immobilized on membrane followed by probing with non-radioactive labeled PCR probe.

Substitution pattern of Jute Ty1-copia and Ty3-gypsy retrotransposons.

For substitution pattern analysis, jute LTR retrotransposon (Ty1-copia and Ty3-gypsy) sequences having potentially active functional RT domain (without stop codons), were included to estimate the synonymous and nonsynonymous substitution patterns. According to the estimates of synonymous and nonsynonymous substitution patterns of Ty1-copia and Ty3-gypsy elements, the amplified RT domains of Ty1-copia and Ty3-gypsy has dN:dS ratios of 2.453 and 1.463, respectively (Table 3); suggesting the presence of positive Darwinian selection in these two lineages. Also, the Z-test for the Ty1-copia elements indicates the presence of positive selection for most of the pair wise comparisons (73.2%) (Table 4).

Table 3.

Synonymous divergence (dS), nonsynonymous divergenece (dN) and dN:dS ratios of Ty-1 copia and Ty-3 gypsy elements of jute

Sequences	N	S	dN (mean)	dS (mean)	dN/dS
Ty1	83.092 ± 1.169	33.908 ± 1.178	0.682 ± 0.101	0.278 ± 0.057	2.453
Ty3	96.125 ± 1.430	38.875 ± 1.413	1.127 ± 0.155	0.770 ± 0.159	1.463

Table 4.

Ty1-copia Z values (above) and significance P-values (below) for the codon based Z-test for positive selection

Discussion

In this study, partial reverse transcriptase domains of Ty1-copia and Ty3-gypsy retrotransposons of jute were amplified by degenerate primers which have been tested on many higher plants and were found to be successful (www.le.ac.uk/bl/phh4/retros.htm).²¹ Analyses of the amplified sequences confirmed the successful amplification of reverse transcriptase domains of jute LTR retrotransposons. The Ty1-copia reverse transcriptase sequences of jute were found to be highly heterogeneous. Similar results of sequence heterogeneity were also observed for other species.^5,22–27 On the other hand, the reverse transcriptase sequences of Ty3-gypsy elements of jute were found to be less heterogeneous. Few recently published articles also described similar findings.^28–30 High level of sequence heterogeneity may be the combined effect of a number of reasons viz.: (1) error-prone retrovirus like replication mechanism due to lack of fidelity and proof-reading activity of the reverse transcriptase,³¹ (2) mixture of active and defective populations of retrotransposons,² (3) pressure of copy number since the divergence of a retrotransposon in any population is proportional to their copy number,^22,32 (4) homologous recombination,^33,34 (5) horizontal and (6) vertical transmission.² In addition, substitution pattern analysis also shows that the dN:dS (the ratio of the number of non-synonymous nucleotide substitutions to the number of synonymous nucleotide substitutions) is higher in the Ty-1 copia lineage which could also be a reason for sequence heterogeneity.

Despite their low abundance in the genome and lower dN:dS value compared to their counterparts undergoing strong positive selection, most of the Ty3-gypsy elements in the current study contain stop codons. This could have arisen due to the fact that weakly selected amino acid sites saturate rapidly when nonsynonymous substitution rates reach a maximum. On the other hand, the synonymous substitution or divergence increases.^35,36 But in light of the current study, it can not be substantiated that Ty3-gypsy sequences harbor more weakly selected amino acids sites. However, a considerable number of Ty1-copia elements have also been found to contain stop codons. McAllister and Werren³⁷ showed that newly transposed retroelements rarely act as a source of new elements and accumulate mutations like pseudogenes. Although transcription does not necessarily mean active transposition, for Ty1-copia retrotransposons, this ‘Pseudogene effect’ could be the cause of their low abundance and their tendency to accumulate mutations. In that case, natural selection could act only at the genomic level, masking the tendency to accumulate mutations to become nonfunctional, conserving the retroelements.

Presence of repetitive elements is a recognized reason for the variation in genome size of many plant species where abundance of retrotransposon is quite high.³⁸ It has been reported that, copy number of Ty1-copia retrotransposons may vary from several hundred to one million³⁹ as for example, copy number of Tnt1, a tobacco copia like retrotransposon is 100 whereas BARE-1, a copia-like retrotransposons of barley is 100,000. Ty3-gypsy elements have also been reported to be present in high copy number in several plant genomes, in Vicia species Ty3-gypsy elements contribute 18% to 35% of their genome²⁰ and in apple genome Ty3-gypsy contributes 33.5%.¹⁸ In this study, it was found that, LTR retrotransposons constitute about 19% of jute genome having approximately 31,000 copies of Ty1-gypsy elements and 3,600 copies of Ty3-gypsy elements. This might be an indication of the importance of Ty1-copia retrotransposons in jute genome evolution and its large size. These copy numbers may be little bit skewed because the whole PCR product was used for dot blot probe preparation and which may create the chance of non-specific binding of the probe. On the other hand, the genome size of jute is quite large; it may harbor in its genome a large number of transposable elements which move through copy-paste method and part of which is reflected by the large copy number of LTR retrotransposons found in this study. It is worth mentioning that, number of retrotransposons is low in small genomes due to the pressure of rapid cell division⁴⁰ whereas it is just the opposite where the genome size is large and retrotransposon selection is relaxed³⁹ for example; jute genome.

Transcriptional activity of retrotransposons is mostly demonstrated for LTR subclasses and this property has been confirmed for Tnt1A, BARE-1 and Tto1.⁴¹ Plant retrotransposons are usually transcriptionally inactive during the developmental stages to lessen the detrimental effect on host, but may be activated upon enforcement of various biotic and abiotic stresses.¹ There has been a long evolutionary co-existence between the retrotransposons and the host genomes which profoundly influence the genome activity. The host genome has devised few surveillance strategies like methylation and RNAi based silencing to control its expression and activity. Recombination, also a host genome factor, is another mediator of copy number and expression pattern. Transcriptional activity of plant retrotransposons were reported in roots, tassels and leaves.^3,29,42–44 In our study, four groups (Group 1, 6, 8 and 9) of Ty1-copia retrotransposons were found to be transcriptionally active in the leaf tissue under normal conditions. Since this was the first ever work on jute retrotrasnposons, therefore mature leaves of farmer popular variety C. olitorius O-9897 grown under normal conditions were used in the study so as to build a standard for similar work in the future. The phylogenetic trees constructed in this study were an approach to clustering the jute LTR retrotransposons based on their reverse transcriptase domain sequences. Although transcriptional activity may be a reason for expanding genome size, it is however not well established because replication cycle of LTR retrotransposons consists of four steps: transcription, translation, reverse transcription and integration; regulation at any of these levels can limit the transposition of retrotransposons.⁴⁵

Transcriptional activity of LTR retrotransposons in jute leaves under typical conditions is an important finding which upon further study will shed light on the regulation of gene expression patterns in jute. In this regard, identification of the full length sequences of these transcriptionally active jute retrotransposons, study of the functionality of their promoters, detection of their locations in chromosomes and analyzing their expression pattern under different stress conditions would be the next steps of research.

This is a pioneering work in the study of molecular biology of jute. Isolation and characterization of the reverse transcriptase domains of jute LTR retrotransposons and their estimated copy number were the main focus of this study but the most outstanding discovery was identification of the transcriptionally active retrotransposons in jute leaves. This study has opened the door of a new horizon of jute research in many ways like, the development of retrotransposon based marker systems for jute which is more polymorphic than AFLP and RAPD,^46,47 phylogenetic analysis of jute to uncover its evolutionary history, to identify the stress responsive retrotransposons and to tag genes of agronomic importance.

Materials and Methods

Plant material; DNA and RNA isolation.

Jute (C. olitorius) variety O-9897 was used in this study. Mature leaves of jute were collected from the garden of Bangladesh Jute Research Institute (BJRI). Total DNA from fresh leaves was isolated following the protocol described by Haque et al. (2004),⁴⁸ and used for PCR and dot blot hybridization.⁴⁸ A treatment with RNaseA (Invitrogen) was performed at 37°C for 1 hour before DNA precipitation. Purity of the DNA was confirmed after running the genomic DNA in 1% agarose gel. Total RNA was also isolated from fresh mature leaves using Trizol (GIBCO) according to supplier's manual. Isolated RNA samples were treated with DNase I (Invitrogen) at 37°C for 30 minutes. Quality of RNA was assessed by running the samples in 1.3% denaturing agarose gel. Quantity of the genomic DNA and RNA were measured using NanoDrop^® (NanoDrop Technologies Inc., Wilmington, USA). Purity of RNA was determined by spectrophotometric analyses using the ratio of absorbance at 260 nm and 280 nm (A260/A280).

Amplification of copia and gypsy like retrotransposons from jute genome.

The reverse transcriptase domains of the Ty1-copia like retrotransposons of jute genome were amplified from the genomic DNA of C. olitorius O-9897 using degenerate primer pair J-Ty1-RT (F) 5′-ACN GCN TTY YTN CAY GG-3′ and J-Ty1-RT (R) 5′-ARC ATR TCR TCN ACR TA-3′.^22,49 The reverse transcriptase domains of Ty3-gypsy like retrotransposons were amplified using degenerate primer pair J-Ty3-Da (F) 5′-TAY CCN HTN CCN CGN ATH GA-3′ and J-Ty3-Da (R) 5′-ARC ATR TCR TCN ACR TA-3′.^22,49 In each of the amplifications, PCR was performed in 25 µl reaction mixture with 100 ng of genomic DNA, 1.0 U Taq DNA polymerase, 2.5 µl PCR reaction buffer (Invitrogen), 0.15 mM dNTPs (Invitrogen), 4.0 mM Mg₂Cl and 0.6 µM of each of the primers. Thermal cycler (Eppendorf) condition for Ty1-copia PCR was as: initial denaturation at 95°C for 5 minutes, then 35 cycles each with denaturation at 95°C for 45 seconds, annealing at 48.5°C for 45 seconds, elongation at 72°C for 1 minute followed by a final extension at 72°C for 5 minutes and finally held at 4°C. The PCR condition for Ty3-gypsy was same except for the annealing step which was 49°C for 50 seconds. The products were separated by electrophoresis in 2% agarose gel, stained with ethidium bromide and visualized under UV light.

Cloning, DNA sequencing and sequence analysis.

The PCR products were purified from the agarose gel by QIAquick Gel Extraction Kit (QIAGEN), cloned into pCR^®2.1 vector using TA cloning kit (Invitrogen) and transformed into DH5_α strain of chemically competent Escherichia coli according to supplier's manual. Plasmid DNA was isolated⁵⁰ and cloned DNA was sequenced in ABI PRISM 3730 (Applied Biosystems) using M13 primers and Big Dye terminator v3.1. Sequencing services were obtained from 1^st Base Laboratories, Malaysia.

Nucleotide sequence of each clone was compared with the deposited sequences in the non-redundant databases using BLASTn and BLASTx from the National Center of Biotechnology Information (NCBI) web portal (www.ncbi.nlm.nih.gov/BLAST/). Nucleotide sequences were converted to protein sequences using Transeq Translate tool (www.ebi.ac.uk/Tools/emboss/transeq/index.html). Along with the LTR retrotransposon sequences of other species (obtained from the Gypsy Database, GyDB),⁵¹ translated sequences of Ty1-copia and Ty3-gypsy of jute were aligned separately using ClustalW program (Thompson et al. 1994) from the EMBL EBI web page (www.ebi.ac.uk/Tools/clustalw/) and presented using BOXSHADE 3.21 (www.ch.embnet.org/software/BOX_form.html).

Phylogenetic trees were constructed using the NJ method as implemented in MEGA 4.1 ⁵² under the following conditions: uniform rates among sites and pairwise deletion of gaps. Branch support was calculated by bootstrap analysis consisting of 1,000 replicates.⁵³

Dot blot hybridization and copy number determination.

Genomic DNA and the heterogeneous PCR products of the reverse transcriptase domains of Ty1-copia and Ty3-gypsy were used for dot blotting. Both genomic DNA and PCR products were denatured by heating at 100°C for 10 min, and then immediately chilled on ice. Equal volume of 1 N NaOH was added and incubated at room temperature for 30 min and transferred onto Immobilon-Ny⁺ membrane (Millipore, Boston, MA USA). The denatured genomic DNA was diluted to 10, 8 and 5 µg. The Ty1-copia PCR products were diluted to 100, 50, 25 and 40 ng. For analysis with Ty3-gypsy, the denatured genomic DNA was diluted to 20, 10 and 5 µg and the PCR products were diluted to 200, 100 and 50 ng. The membrane was air dried and incubated in neutralization buffer (1 mM EDTA pH 8.0, 1.5 M Nacl and 0.5 M Tris-HCl pH 8.0) for 30 min. Finally the membranes were rinsed in 2X SSC followed by air drying and fixed by ultraviolet cross-linking. The Ty1-copia or Ty3-gypsy PCR products used as probes, were labeled with digoxigenin-dUTP according to manufacturer's instructions of DIG DNA Labeling and Detection Kit (Roche, Germany). Stringent hybridization was carried out at 65°C for 18 h, with washing conditions as recommended by the manufacturer. The hybridized probes were immunodetected with alkaline phosphatase conjugated anti-digoxigenin antibody fragments (F^ab) and colorimetric substrate NBT/BCIP. Intensity was analyzed with ImageJ 1.42 software (http://rsb.info.nih.gov/ij/). Copy numbers were calculated as described by Ma et al. (2008) using the following equation: Copy number = (the size of genome × average proportion of nuclear genomic DNA hybridizing to the probe)/size of probe element.

RT-PCR.

First strand cDNA was synthesized using Moloney-Murine Leukemia Virus (M-MLV) SuperScript™ III Reverse Transcriptase (Invitrogen) according to manufacturer's instructions, using oligo d(T)_12–18 primer. PCR amplification for both Ty1-copia and Ty3-gypsy were performed for the reverse transcribed products by using the same degenerate primers mentioned above. For negative control PCR amplification was carried out using RNA as the template instead of the RT-PCR product. Amplified products were purified from the gel, cloned into vector, transformed into chemically competent E. coli and then sequenced as described above. Sequences obtained were analyzed in the manner stated earlier. By using these reverse transcribed sequences along with the previously obtained RT domain sequences from genomic DNA, separate phylogenetic trees were constructed for Ty1-copia and Ty3-gypsy elements. Bootstrap Neighbor-joining method with 1,000 replication produced by MEGA 4.0 software⁵² was used to construct these trees.

Substitution pattern.

For substitution pattern analysis, Codon Based Z-test and dN:dS ratios using modified Nei-Gojobori⁵⁴ method with the Jukes-Cantor correction as implemented in MEGA4.1.⁵²

Financial Support

This work was financially supported by US Department of Agriculture (USDA).

Author Contributions

All the authors contributed to a similar extent overall.