Retrotransposon expression in response to in vitro inoculation with two fungal pathogens of Scots pine (Pinus sylvestris L.)

Methods

For the LS inoculation 27 grafted 2-year-old pine ramets from Ja3(5), Sm3(13) and Sm9(9) were used. One plant from each of the three genotypes was treated with water as a control. RNA was isolated from the needles of the same seedling before inoculation, 3 days post inoculation (dpi), 14 dpi and 31 dpi with prominent signs of infection. Details on LS inoculation are provided in (Additional file 1). For the HA inoculation 101 seedlings of 12 pine plus trees were used. 6-day-old healthy seedlings were inoculated with a HA suspension details provided in (Additional file 2). 2–4 seedlings from each family were used as non-inoculated controls and harvested after 7, 14 and 21 dpi. For each seedling RNA was isolated from the roots and shoots separately. Samples from both experiments were stored at − 80 °C until extraction.

RNA was isolated using the method described by [20]. Treatment with the Turbo DNA-free kit (Thermo Fisher Scientific) was performed for all samples following the manufacturer’s instructions. RNA purity was tested by polymerase chain reaction (PCR) with an RNA stock solution as a template and oligonucleotides amplifying REs and APT1. RNA concentration was measured with a Qubit (Life technologies) and equated. The Taqman Reverse transcription kit (Applied Biosystems) was used for reverse transcription of 0.8 μg RNA with random hexamer oligos (Thermo Fisher Scientific). Produced cDNA was diluted 1:4 with nuclease-free water. Primers to 12 RE polyprotein sequences [29] were used, expression levels of the antimicrobial gene Pinosylvin synthase (PsBs) was evaluated to compare the induction of defence responses. Comparative CT real-time PCR was performed with the SyberGreen Maxima qPCR Master Mix (Thermo Fisher Scientific) standard protocol on a StepOnePlus thermocycler (Thermo Fisher Scientific). For each sample two technical replications per plate were used in order to analyse more biological replicates and to include three reference genes. The normalization coefficient was calculated and a correction applied for plates with identical sample sets [21], identical thresholds for each target were set. Primers for pine endogenous reference genes were designed using Primer3 v.4.1.0. software [22], (Additional file 3). NormFinder [23] and Bestkeeper [24] were used to evaluate the most stable genes. For the evaluation of the amplification efficiencies, the standard curve method was used: qPCR with six 1:10 dilutions of the experimental cDNA were performed in triplicate and the efficiency was calculated using StepOne Software v.2.2.2 (Applied Biosystems). The relative expression level (ΔΔCt) of REs before and after inoculation was calculated [25]. The amplification product melting curves provided in (Additional file 4). Multi-Response Permutation Procedures (MRPP) provided by PC-ORD v.5. statistical package [26] was used to investigate differences between expression responses after HA. Pearson correlation was calculated for the shoots and roots HA data on four RE loci and phenotype observed.

Results

Inoculation with LS

No PsBs or REs expression level changes were observed at all-time points studied in the healthy controls (Additional file 5a). Relative PsBs expression induction was detected at 3 dpi and expression levels increased as disease progressed (Additional file 5b). At 14 dpi and 31 dpi following inoculation with LS all tested families of REs were expressed at high levels, but PsBs and RE expression correlated (Additional file 6a). The three studied genotypes showed varying RE expression responses at the sampled time points. On average Sm3 ramets showed higher RE expression and earlier induction at all-time points (Fig. 1). REs Conagree and Appalachian were expressed more intensively in Sm3; Appalachian expression was higher in Sm9, but in Ja3 Riga-4 was expressed more intensively. An observed increase in RE expression level was confirmed by absolute transcript quantification of the Conagree element as described previously [27], (Additional file 5c).

Inoculation with HA

Following signs of infection were observed in inoculated seedlings compared to the control seedlings: reduced length of the inoculated roots, absence or reduced number of lateral rootlets, reduced number and length of secondary needles, and yellowing of needles and stem (Additional file 7a). In all seedlings at 7 dpi the enhancement of PsBs levels was observed mostly in the roots. At 14–21 dpi PsBs expression was increased in the shoots (Additional file 7b). The levels of PsBs expression in families M259, M248, M110, M241 were lower compared to other seedling families at each time point studied.

For Sm12 and M110 seedling families, the expression of nine REs was analysed (Additional file 7c). No RE expression enhancement in roots and a small induction of RE expression at 7 dpi was revealed. The M110 family displayed a lower induction of REs at 7 dpi and 14 dpi. At 21 dpi no RE expression was observed in either tissues. Since enhancement of different RE loci in one seedling correlates strongly (Additional file 6b), four representative REs and PsBs were selected to perform qPCR analyses on remaining seedlings. A more diversified response was observed following the analysis of 11 pine seedling families (Fig. 2). RE expression was enhanced in the shoots at 7 dpi in all seedling families tested. At 14 dpi in some seedlings RE expression was increased in the shoots, but at 21 dpi RE expression in the shoots was decreased while a small increase was observed in the roots. Overall levels of RE transcription were significantly lower after inoculation with HA when compared to the response after LS inoculation (Fig. 3).

Discussion

Although activation of RE in response to stress has been shown in many plants [28–31], usually transposition of autonomous elements were studied [32–34]. Meanwhile, non-autonomous REs are transposed in plants [35, 36], but transcription could also occur in disrupted sequences, which could contribute to non-coding RNA production [18, 37]. RE in conifers was characterized as more diverse and abundant when compared to angiosperm repeats [5]. For example, in the tomato genome the active TLC1.1 element is found in 900 copies and this RE is a dominating family [38]. In the genome of Pinus taeda about 1500 RE families were revealed [5]. Random relaxation of thousands of RE families found in the large genome would result in the neutral enhancement of the most distributed RE families. However, the results obtained in this study demonstrate no connection between RE copy number distribution in the genome and relative levels of transcriptional induction. The most abundant in terms of copy numbers per single diploid genome of Pinus sylvestris from RE studied was IFG (83,986 ± 7439) first detected in sugar pine [39]. Nevertheless, Appalachian, Riga-4 and Cumberland transcripts were found to be more abundant than IFG transcripts in both the HA and LS experiments. Appalachian was found to be propagated in the P. sylvestris genome with 22,885 ± 3158 copies; Riga-4 with 3916 ± 686 copies; Cumberland RE only with 165 ± 16.68 copies [27].

LS is an ascomycetous foliar inhabiting fungi and is a serious pathogen of the young trees of P. sylvestris, where it causes needle cast [40–43]. Pine sprouts with secondary needles are able to survive single LS attacks and develop new healthy shoots in the next season (Lagerberg 1913, cited by [40]). Most Lophodermium species are endophytes and pathogenicity has likely evolved via endophysm in this group [44]. L. nitrenes and L. pinastri were found to be propagated in needles as a slowly growing biotrophic fungi that avoids penetration into mesophyll cells during the early stages of infection [45, 46]. Scots pine originated from Latvian seed orchard Smiltene were carrying induced LS defence and recovery rate [47]. HA is an aggressive pathogen belonging to basidiomycete necrotrophic fungi that colonize stumps and roots [48]. Necrotrophic fungi secrete cell-wall-degrading enzymes and toxins, which they use for rapid propagation in dead cells through induction of host plant hypersensitive responses [49, 50]. Defence against biotrophs and necrotrophs in Arabidopsis involve different Jasmonic acid/Ethylene-dependent or Salicylic acid-dependent pathways [51]. For inoculation with HA, no significant increase in RE expression was evaluated, which was in contrast to the response after inoculation with LS. Differences in RE response were observed for the pine seedling families, while gene PsBs expression increased constantly with the spread of HA and LS. In the current study it was shown that sets of REs responded to both fungi pathogens, but the levels and patterns of their responses was considerably different and more prominent in likely biotrophic pathogen inoculations. In more aggressive necrotrophic HA infections, RE transcription levels were significantly lower at time points and tissues studied.

There is growing evidence regarding the role of non-coding RNA in host–pathogen interplay through different RNA interference mechanisms [18, 52]. TEs could be acquired by horizontal gene transfer [53] and pathogenic strategies could rapidly evolve. Advantageous flexible stress responsive gene networks could be formed using REs, contrary, some pathogens could acquire features that through the agency of host REs downregulates transcription of nearby genes by various mechanisms. Evolutionary change via recombination, transposition and/or fast degradation of influenced RE sequences result in high genetic diversity and could induce adaptability of populations that undergoes continuous and variable pressure of the natural selection.