Heterobasidion annosum sensu lato is a serious pathogen causing root and stem rot to conifers in the northern hemisphere and rendering the timber defective for sawing and pulping. In this study we applied next-generation sequencing to i) identify transcriptional responses unique to Heterobasidion-inoculated Norway spruce and ii) investigate the H. annosum transcripts to identify putative virulence factors. To address these objectives we wounded or inoculated 30-year-old Norway spruce clones with H. annosum and 454-sequenced the transcriptome of the interaction at 0, 5 and 15 days post inoculation. The 491860 high-quality reads were de novo assembled and the relative expression was analysed. Overall, very few H. annosum transcripts were represented in our dataset. Three delta-12 fatty acid desaturase transcripts and one Clavaminate synthase-like transcript, both associated with virulence in other pathosystems, were found among the significantly induced transcripts. The analysis of the Norway spruce transcriptional responses produced a handful of differentially expressed transcripts. Most of these transcripts originated from genes known to respond to H. annosum. However, three genes that had not previously been reported to respond to H. annosum showed specific induction to inoculation: an oxophytodienoic acid–reductase (OPR), a beta–glucosidase and a germin-like protein (GLP2) gene. Even in a small data set like ours, five novel highly expressed Norway spruce transcripts without significant alignment to any previously annotated protein in Genbank but present in the P. abies (v1.0) gene catalogue were identified. Their expression pattern suggests a role in defence. Therefore a more complete survey of the transcriptional responses in the interactions between Norway spruce and its major pathogen H. annosum would probably provide a better understanding of gymnosperm defence than accumulated until now.
Citation: Lundén K, Danielsson M, Durling MB, Ihrmark K, Gorriz MN, Stenlid J, et al. (2015) Transcriptional Responses Associated with Virulence and Defence in the Interaction between Heterobasidion annosum s.s. and Norway Spruce. PLoS ONE 10(7): e0131182. https://doi.org/10.1371/journal.pone.0131182
Editor: Roberto Papa, Università Politecnica delle Marche, ITALY
Received: January 8, 2015; Accepted: May 31, 2015; Published: July 7, 2015
Copyright: © 2015 Lundén et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files. The sequence reads are deposited in SRR346081 in NCBI’s Sequence Read Archives (SRA, http://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi).
Funding: Financial support was received from the Swedish Research council Formas, grant brg0312st2 (http://www.formas.se/en/), and from the Swedish Foundation for Strategic Research (SSF), grant number R8b08-0011 (www.stratresearch.se). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Norway spruce [Picea abies (L.) Karst.] is a dominating tree species in much of northern Eurasia. It relies on both induced and constitutive defences to restrict the spread of invading fungi and insects. The bark is the first line of defence that provide an efficient defensive barrier . When this barrier-defence is breached, inducible defences remain essentially the only protection against invading pathogens . It has been indicated that the defence responses in conifers lack specificity in their responses to pathogens [3–5]. The induced transcriptional responses to pathogenic fungi are similar to those induced by wounding in Norway spruce [1, 4, 6–8]. Both responses involve expression of genes encoding pathogenesis related- (PR) proteins and components of well-described defence signal transduction pathways but the expression levels are generally significantly higher after pathogen infection.
The single most important biotic stress factor to Norway spruce is the root rot fungus Heterobasidion annosum sensu lato (s.l.) . H. annosum s.l. is a complex of five closely related species [10–12] that have different yet sometimes overlapping host ranges. In Scandinavia two species are present: H. annosum sensu stricto (s.s.) and H. parviporum, both pathogenic on Norway spruce . Recently, the first genome of H. annosum s.l. was published, of the North American H. irregulare . Transcriptional analyses of H. irregulare under different conditions suggest that living host tissues triggers an expanded metabolic repertoire involving genes associated with, for example, toxin production, pectin degradation, protection against plant defences, handling of low oxygen pressure and other abiotic stresses .
H. annosum s.l. causes root and stem rot and renders the timber defective for sawing and pulping. H. annosum s.l. infection also reduces the tree’s growth and increases the risk of wind-throw. Economic losses to the Scandinavian forest industry attributed to H. annosum s.l. infections exceed € 90 million annually . Identification of resistant host genotypes and understanding the mechanisms behind such resistance would enable selection of more resistant host material for replanting, which could reduce the economic losses. In an earlier study of the transcriptional responses of genes in the phenylpropanoid biosynthetic pathway in the Norway spruce genotypes we showed that those genes were highly induced in response to wounding and infection by H. annosum s.s. . In the current study, we expanded the analysis of the responses in these Norway spruce genotypes beyond genes in the phenylpropanoid pathway. Our objectives were to i) identify transcriptional responses unique to Heterobasidion-inoculated Norway spruce and ii) investigate the H. annosum s.s. transcripts and compare them to the expression patterns predicted from H. irregulare in Pine bark  to identify common virulence factors in H. annosum s.l.
Material and Methods
Plant material and sampling
In this study, we used 30-year-old trees of eight Norway spruce genotypes: S21K7822405, S21K7825237, S21K7827398, S21K7828590 S21K7823178, S21K7823340, S21K7825278 and S21K7828397. The trees are part of a Swedish clonal forestry program grow in a stand situated at Årdala, Sweden, (59°01' N, 16°49' E) . we contacted the land owner, to obtain permission to conduct the study described in this manuscript, and was granted permission by the Forest Manager. The field studies did not involve endangered or protected species. The inoculation and sampling procedures have described in detail in Danielsson et al. . Briefly, three ramets per genotype, and two roots per ramet, were used in the experiment. On one root, a 5 mm diameter woody plug colonized by H. annosum s.s. (isolate Sä 16–4)  was attached to an artificial wound on the root surface with Parafilm. The other root on the tree was wounded only and sealed with parafilm. The samples used in this study were harvested at the start of the experiment (0 days post inoculation) and at 5 and 15 days post inoculation (dpi) and preserved in RNAlater (Ambion) for subsequent RNA extraction.
RNA extraction, cDNA synthesis and sequencing
Total RNA was isolated according to Chang et al. . As previously described  poly(A)-RNA from four Norway spruce genotypes (2405, 3178, 3340 and 7398) were purified, amplified and prepared for 454-sequencing on a GS FLX (Roche, 454) at the Norwegian Sequencing Centre (http://www.sequencing.uio.no). Sequence reads and quality scores for sequences were obtained from the Norwegian Sequencing Centre. The reads were deposited at NCBI as SRR346081. Four additional genotypes (5237, 5278, 8397 and 8590) were used for analysis by quantitative reverse transcribed-PCR (qRT-PCR), these samples were also subjected to poly(A)-RNA isolation and mRNA amplification as previously described .
Assembly, annotation and statistical analyses
The bioinformatics and statistics pipeline are outlined in S1 Fig. The retrieved sequences were assembled with the sequence assembler software Newbler v2.6 (Roche) (www.454.com), with default settings for cDNA assembly and the sff-files as input file. The sequence assembly was carried out on the publicly available Bioportal (www.bioportal.uio.no). Sequences from all treatments were assembled into the gene-equivalent isogroups and into the plausible splice variants isotigs. The assembled sequences, excluding the singletons, were used as reference file and were annotated with the software Blast2GO ; where the sequences were annotated to blastx homologies and gene ontology (GO) terms. The taxonomic origin of the assembled sequences and singletons were evaluated based on the taxonomic origin of the corresponding blastx homolog by analysing the reference file with MEGAN4 with standard settings 
The best reciprocal blast hits of the assembled sequences to Norway spruce gene models and transcripts in the P. abies 1.0 release of the Norway spruce genome (http://congenie.org/) were identified using blastn 2.2.27 . Blastn was wrapped in a python script which established the best reciprocal matches between the two sequence sets. To identify H. annosum s.s. transcripts the sequences were queried against H. irregulare sequences in Genbank using blastn .
To estimate the relative gene expression between the libraries, individual reads from each library were mapped to the reference file built from the isotigs in the assembly. The mapping was done in Newbler 2.6 at the Bioportal (www.bioportal.uio.no) and count data was generated that was then imported into R v.2.15.0. The count data were normalized toon the overall sum of the counts in each library in the R package DESeq v.1.83 . To reduce the effect of multiple testing on the detection power, filtering independent of the test statistics was applied . Before the test statistic, 40% of the lowest quantile of the overall sum of the counts were removed. The differential expression of genotype, time and treatment dependent expression was tested with the condition based negative binominal distribution test (nbinomTest) in DESeq v.1.83  and adjusted for false discovery rate with the Benjamini-Hochberg procedure. The normalized count data underwent variance-stabilizing transformation to homoscedastic data in DESeq and clustered with JMP 9 by Ward’s hierarchical cluster. To determine if there were statistically over-represented GO categories between treatments or clusters we used Fisher's exact test with the Benjamini-Hochberg false discovery rate correction available in the Blast2GO software.
Verification of gene expression by qRT-PCR
Purified aRNA (1μg) from genotypes 5237, 5278, 8397 and 8590 were reverse transcribed with the iScript cDNA synthesis kit (Bio-Rad). The cDNA synthesis was diluted in deionized water, 1:1, and an aliquot of cDNA equivalent of 25 ng of aRNA was used per 20 μL of PCR reaction using SSoFast EVAGreen Supermix (Bio-Rad) and a final concentration of 0.5 μM of each primer. Primers were designed from isotig sequences using the Primer3 software (http://primer3.wi.mit.edu/) with a melting temperature (Tm) between 58°C and 60°C, and amplicon length between 95–183 bp (S1 Table). The thermal-cycling condition parameters, run on a iQ5 Multicolor Real-Time PCR Detection System (Bio-Rad), were as follows: 95°C for 30 s; 40 cycles of 95°C for 5 s, 58 or 60°C for 20 s. Each run was followed by a melt curve analysis to validate the specificity of the reaction. A linear plasmid standard curve was used to measure the PCR efficiency and primer pairs with efficiency lower than 95% were discarded. Two technical replicates were prepared for each sample.
Transcript abundance was normalized to the reference genes phosphoglucomutase , eukaryotic translation initiation factor 4A (elF4A)  and elongation factor 1-α (ELF1α). The relative expression was calculated using the Pfaffl-method . Kruskal-Wallis and Mann-Whitney U-tests were performed on the qRT-PCR data using the GraphPad Prism5 software (GraphPad Inc.).
Kumar and Blaxter  reported that Newbler 2.3 generates assemblies with significantly lower total lengths than newer versions of Newbler. Therefore we chose to reassemble our data with Newbler 2.6 (Roche, www.454.com). We had a limited improvement compared to our previous assembly , with an increase in the number of assembled reads and detected genes (isogroups), and a reduction of plausible splice variants compared to with Newbler 2.3 (S2 Table). We did, however, get a reduction in the isotig N50 length.
The taxonomic assignment of the assembled isotigs was estimated with the MEGAN4 software. The absolute majority of the isotigs had plant homologs, and of these most of the blastx hits were to sequences originating from species within the Pinaceae, and in particular from Picea sitchensis (Table 1). The analysis suggested that very few assembled isotigs were of fungal origin; only 393 isotigs gave a hit to taxa in the fungal kingdom (Table 1).
A reciprocal blastn analysis between our dataset and the transcript assembly (Trinity contaminant free) in the Norway spruce genome (P. abies 1.0 release)  showed that 11908 of the assembled isotigs had at least one significant hit in the P. abies 1.0 transcript assembly (Table 1, S3 Table). Out of the 1096 orphan isotigs, isotigs without significant hits in the Trinity database, some were members of isogroups which had one or more other isotigs variants with significant hits in the transcript assembly and other orphan isotigs had homologs in the Pinaceae. Two hundred five orphan isotigs were of fungal origin (Table 1).
Analysis of the transcriptome
Given the limited depth in our study we filtered away 40% of the lowest quantile of isotigs with mapped read before the test statistic, leaving 7342 isotigs in the analyses of differential expression patterns. Using this dataset we analysed first, the transcriptional changes in Norway spruce in response to inoculation and wounding and secondly a two-way clustering and GO category enrichment analysis was made on the 150 most highly expressed isotigs.
Transcriptional changes in Norway spruce in response to inoculation and wounding.
Compared to untreated bark, transcriptional changes after wounding and inoculation were generally small, and the majority of the assembled isotigs did not show significant changes (padj > 0.05) in expression level between the treatments. When comparing the wounded and inoculated samples to the control samples with the nbinom test fewer than 40 Norway spruce transcripts were found to be significantly regulated (padj < 0.05) in either treatment (Table 2). Ten of the most highly induced genes in the comparison between inoculation and control were found in cluster 3 (Fig 1). A general pattern seen in Table 1 is that genes differentially expressed in response to inoculation were induced while wounding alone resulted in significant down-regulation of genes (Table 2). Only four genes were significantly up-regulated in both comparisons; isotig08156, a protein of unknown function, one class IV chitinase genes (isotig01680), one protein kinase (isotig00518) and a thaumatin-like protein (isotig01982). Several transcripts with similarity to PR-protein gene families were strongly induced in response to inoculation. Besides the already mentioned class IV chitinase (isotig01680) and thaumatin-like protein (isotig01982) genes, PR1-like (isotig05256), peroxidase (isotig01055) and beta–glucosidase (isotig1778 and isotig01779) genes were differentially expressed.
A two–way clustering of normalized 454 sequence count data using Ward’s hierarchical cluster algorithm. The increasing intensity of red in the heat map visualise counts from white (no counts in the particular condition) to red. Clusters are indicated by numbers and the enriched GO categories are noted for each cluster. Clusters containing significantly regulated genes are highlighted in colour, with orange for cluster 3, containing genes significantly induced in inoculated samples and blue for clusters with genes more highly expressed in control than in wounding. Please refer to S4 Table for information about specific isotigs found in the clusters.
Nineteen of the strongly regulated genes (Table 2) were also found to display a significant temporal variation (Fig 2). Irrespective of treatment, more transcripts were differentially regulated at 5 dpi compared to 15 dpi (Table 2). For instance beta–glucosidase (isotig1778 and isotig01779) and isotig04668 encoding PaGLP2, Picea abies Germin-like protein 2 [28, 29] were significantly up-regulated in response to inoculation with H. annosum at 5 dpi. Isotig02279 (MA_277805g0010) encoding a transcript without known function is significantly induced at both time points and treatments (Table 2).
A Venn diagram showing the differentially expressed genes (P-value <0.05 after FDR [Benjamini-Hochberg procedure]) at 5 dpi inoculation with H. annosum (green) or wounding (yellow) and 15 dpi inoculation with H. annosum (red) or wounding (blue).
In the comparison between the inoculation and wounding treatments, irrespective of time point, 26 isotigs were found to be significantly up-regulated after inoculation with H. annosum s.s. compared to wounding (Table 3). Fifteen of these differentially expressed isotigs were of Norway spruce origin (Table 3). Five differentially expressed isotigs (isotig06294, isotig06826, isotig08690, isotig01014 and isotig12672) had significant matches in the Norway spruce (v1.0) high confidence gene catalogue, but neither the isotig nor the gene model had significant alignment to any annotated protein in Genbank. A transcript differentially expressed between inoculation and wounding but not detected in the previous analyses, was isotig00380. This transcript has similarity to 12-oxophytodienoic acid–reductase and it was induced 8 times in inoculated samples compared to wounded samples (Table 3). No significant differences in expression patterns between wounding and inoculation could be detected when the two time points were analysed separately (data not shown).
GO category enrichment among highly expressed genes.
Among the 150 most expressed transcripts (basemean > 7.7) after filtering of the transcripts 75% had blastx homologs in Picea spp. and 3% had blastn homologs in the H. irregulare genome. Norway spruce transcripts associated with translation (GO:0006412), ribosome (GO:0005840), cellular biosynthetic process (GO:0044249) and gene expression (GO:0010467) were enriched (Fisher's Exact Test; P <0.05 after correction for multiple testing) compared to the data set as a whole. A two-way clustering of these most highly expressed transcripts displayed seven clusters (Fig 1, S5 Table). Cluster 6, together with clusters 1, 2, and 4, showed similar expression levels between treatments. Fisher's Exact Test for enrichment of any GO category (P <0.05 after correction for multiple testing) showed an overrepresentation of transcripts in GO categories ribosome (GO:0005840) and organic substance biosynthetic process (GO:1901576) in cluster 6. Clusters 1, 2 and 4 showed no enrichment of GO categories.
Cluster 3, with the highest levels of transcript accumulation in inoculated bark (Fig 1), had an overrepresentation of transcripts associated with GO:0003824 (catalytic activity). The transcript accumulation patterns in clusters 5 and 7 indicated higher expression levels in untreated bark (Fig 1). Fisher's Exact Test indicated that Cluster 5 contained transcripts associated with protein metabolic process (GO:0019538) and cellular glucan metabolic processes (GO:0006073) was overrepresented in cluster 7.
qRT-PCR validation of expression patterns of Norway spruce transcripts.
A small subset of transcripts was selected for verification of the RNAseq data by qPCR analysis in independent Norway spruce genotypes. As predicted by the RNAseq data, isotig01779 and isotig04668 showed higher expression in the inoculated samples than wounded alone at both 5 and 15 dpi (Fig 3a and 3b). RNAseq data predicted a significant induction of contig00509 and isotig00380 in response to inoculation at 15 dpi; this was was observed for both transcripts (Fig 3c and 3d). However, isotig00380 also showed significant induction at 5 dpi (Fig 3c) and contig00509 was induced in response to wounding at 15 dpi (Fig 3d). According to the RNAseq data, isotig06030 (type1 nsLTP) was highly expressed but showed no differential expression between treatments, which was validated by qPCR (Fig 3e). Isotig06890 (Hin1-like) was expected to be expressed at low levels but without significant regulation, and was corroborated by qPCR data at 5 dpi. However at 15 dpi significant induction were seen in both wounded and inoculated samples (Fig 3f).
Isotig01779, beta-glucosidase (a); isotig04668 GLP2 (b); isotig00380, OPR-like (c); contig00509, isoflavone reductase-like (d); isotig06030, ns-LTP (e) and isotig06890, hin1-like (f). The columns indicate relative expression levels over the control, determined with qPCR, for each time point and treatments. The bars indicate the standard error (SE) and superscript letters indicate significant differences between treatments (P<0.05, Kruskal-Wallis test with Dunn´s post test), N = 3. The circles indicate the basemean data reported by DEseq, where shaded symbols were significantly different from un-treated bark.
Heterobasidion annosum s.s. transcripts accumulating in inoculated bark
Among the significantly induced transcripts, ten transcripts showed significant alignment to H. irregulare TC32-1 sequences in Genbank (Table 3). Three delta-12 fatty acid desaturase transcripts i.e. isotig04483, isotig04652, isotig05463 (ETW77095.1, ETW77100.1 ETW81752.1 respectively) were found among the significantly induced transcripts. Other significantly induced transcripts were isotig10242 (weakly similar to CRISPR-associated protein 1, ETW81403.1), isotig04522 (mitochondrial carrier protein, ETW86881.1 ) and isotig05435 (Clavaminate synthase-like, ETW83498.1). Among the predicted Heterobasidion transcripts isotig03463 is the only transcript that had any mapping reads from the wounding treatment. The corresponding H. irregulare gene (ETW80796.1) encodes a highly conserved protein present in all Dikarya. The two isotigs/genemodels isotig03399/ETW81018.1 and isotig09146/ETW83069.1 showed no significant alignments outside the taxon of Heterobasidion in Genbank but were highly expressed in inoculated samples (basemean 13.7 and 14.7, respectively).
Thereafter, we searched for expressed transcripts (Basemean >1) with a significant alignment to the H. irregulare gene models available in Genbank and 54 sequences specifically expressed in inoculation treatments were identified (S6. Table). Most of the identified transcripts match genes in primary- (GO:0044238), single-organism- (GO:0044710), cellular- (GO:0044237) and organic substance- (GO:0071704) metabolic processes. The most highly expressed transcript was a delta-12 fatty acid desaturase (isotig04483) (Table 3, S3 Table). Probes corresponding to 49 of the identified H. irregulare gene models were included on the micro-array made by Olson and co-workers  (S5 Table). Only seven transcripts identified in the H. annosum s.s.-Norway spruce bark interaction were among the 250 most highly expressed H. irregulare genes in necrotic bark tissue after the filtering for potential cross-hybridisation by Pine transcripts . However, generally the H. irregulare homologs to the identified 49 H. annosum s.s. transcripts showed very high relative expression after inoculation in Pine (relative expression values > 10000) on the microarray (S5 Table). Irrespective of potential cross-hybridization with Pine, seven gene models showed higher relative expression on necrotic bark than in pure culture of H. irregulare (S5 Table) and high levels of expression in the H. annosum s.s.-Norway spruce interaction. Two of these were isotig10242 (weakly similar to CRISPR-associated protein 1, ETW81403.1) and isotig05435/ETW83498.1 (Clavaminate synthase-like) (Table 3, S4 Table).
The main objective of this study was to characterize the transcriptional changes associated with induced defence responses in Norway spruce to the necrotrophic root rot fungus H. annosum s.l. however, the statistical analyses of the transcriptome data revealed relatively few differentially expressed transcripts. One obvious reason to this observation is that with smaller datasets, such as ours, the accuracy in the analyses of genes with a low expression level decreases more than the accuracy in the analyses in genes with higher expression levels . In order to improve the accuracy we filtered our dataset to remove the 40% least abundant transcripts, even so only a handful of genes were differentially expressed between inoculation and wounding treatments. However, also previous work report an extensive overlap in the responses to wounding and H. annosum s.l. inoculation in this material  and in other Norway spruce materials [4, 5, 7, 8].
Several of the genes that were differentially expressed between inoculated bark and control have previously been reported to be strongly up-regulated in response to biotic stress in Pinaceae e.g. class IV and II chitinases , PR1-like,  peroxidase  germin-like protein , thaumatin-like protein  and dirigent protein . Most of these were also found in cluster 3, with the most highly expressed transcripts in inoculated bark.
A noteworthy observation is the apparent repression of 29 transcripts in wounded bark compared to the control; most of these transcript show very low relative expression values already in the control and must be treated with caution. However three of the transcripts are highly expressed in the control, of which encodes putative dehydrin genes (isotig03505 and contig00136). Dehydrins play a fundamental role in plant response and adaptation to abiotic stresses [36–38]. They accumulate typically in maturing seeds or are induced in vegetative tissues following salinity, dehydration, cold and freezing stress. Despite all the research on dehydrins, their regulation in relation to their function is not fully understood. It has been described that different members of the dehydrin gene family in Norway spruce and Maritime pine (Pinus pinaster) differential regulation in development and in response to abiotic stress [37, 39, 40]. At least one Picea dehydrin gene has been reported to respond to wounding, MeJA and ethylene treatments . This indicates large functional differences between different dehydrin gene family members  and an overlap in the transcriptional responses to drought and wounding in the dehydrin gene family. Thus, it is possible that signalling cues other than drought stress can induce the suppression of the dehydrin genes isotig03505 and contig00136 after wounding.
Fifteen Norway spruce transcripts where differentially expressed between inoculation and wounding. All of the transcripts were more highly expressed after inoculation compared to wounding, which also agrees with our previous observations in other materials , namely that responses are more aggravated after inoculation than wounding. Three genes, not previously associated with defence to H. annosum s.l., showed specific induction in response to inoculation; one oxophytodienoic acid–reductase (OPR), one beta–glucosidase and one GLP2 (isotigs 00380, 01778, 01779 and 04668, Fig 3a–3c). OPRs catalyze the reduction of double bonds adjacent to an oxo- group in a,b-unsaturated aldehydes or ketones. Many OPRs are part of the octadecanoid pathway that converts linolenic acid to jasmonic acid (JA), but 12-oxo-phytodienoic acid can also act as a mediator in the defence against necrotrophic pathogens [41–43]. The induction of isotig00380 (OPR) in inoculated samples compared to wounded samples (Fig 3c), fits with the observation that JA-mediated signalling is a central pathway in the defence against H. annosum s.l. and highly induced in response to the fungus [4, 8].
Two germin-like genes GLP1 and GLP2 have previously been reported to be highly induced in Norway spruce seedlings in response to inoculation with Ceratobasidium bicorne [28, 29]. Our observation that GLP2 is specifically induced in response to H. annosum inoculation (Fig 3b) could indicate that at least GLP2 may respond to particular signalling cues from biotic stressors not present in the wounding treatment. However it has been reported that drought may influence the pathogen-induced expression GLP2 expression patterns and that the induction patterns in response to C. bicorne varies between roots and shoots of young in Norway spruce seedlings . Thus further studies of the expression pattern of GLP2 in response to H. annosum are needed.
Two putative splice variants of the beta–glucosidase gene MA_10394370g0010, isotig1778 and isotig01779, were both highly induced in inoculated bark (Tables 2 and 3). Beta–glucosidases, with capacity to hydrolyze coniferin (coniferyl alcohol 4-O-glucoside) into coniferyl alcohol, which can be used for cell wall reinforcement, are known from cambium and primary xylem in Pinus and Picea [44, 45]. It is interesting that the induction of these transcripts were specific to inoculated materials (Fig 3a), as the phenylpropanoid pathway and lignification are induced in response to both inoculation and wounding [4, 5]. The induction of this beta-glucosidase gene could indicate that a local release of precursors for lignification are activated to control the spread of H.annosum s.l. in the cambium and phloem.
A second objective was to identify H. annosum s.s. genes expressed in the pathogen´s interaction with bark of Norway spruce. However, the proportion of fungal reads was very low. Fewer than 400 putative Heterobasidion transcripts were assembled and of these a fourth was retained after filtering of the expression data. The vast majority of these expressed H. annosum s.s. transcripts encodes genes in the metabolic machinery sustaining fungal growth. This observation is supported by the comparison with the expression profiles of H. irregulare in necrotic pine  which showed that the corresponding H. irregulare genes generally were highly expressed both in necrotic bark and pure culture. The identification of three Heterobasidion delta-12 fatty acid desaturases (delta-12 FADs), which were highly expressed in inoculated bark are examples of such central genes probably sustaining fungal growth. Delta-12 FADs are known from many fungal species, and produce the major components of fungal cell membranes [46–50]. It has been shown that Aspergillus nidulans and A. parasiticus delta-12-FAD mutants have reduced growth, conidiation and impaired host colonization [51, 52]. The protein encoded isotig05435/ETW83498.1 (Clavaminate synthase-like), could be involved in secondary metabolite production as Clavaminate synthases belong to the non-heme iron, alpha-ketoglutarate-dependent oxygenases and catalyse the production of antibiotic compounds [53, 54].
In the comparison between our 454-dataset and the H. irregulare microarray analysis described by Olson and colleagues  it is obvious that most of the H. irregulare genes homologous with the H. annosum s.s. genes in our experiment were filtered out, due to cross-hybridisation with Pine in the microarray study. A handful of genes present in both datasets showed higher relative expression levels in bark compared to pure culture on the H. irregulare microarray  possibly representing H. annosum s.l. virulence factors. Three candidates are isotig10204/ETW808090.1 (Formate dehydrogenase), isotig12111/ETW81845.1 (hypothetical protein) and isotig10242/ ETW81403.1 (containing the pfam domain PF11327). The two first showed differential expression on the microarray and expression in the 454-data set, while isotig10242 (weakly similar to CRISPR-associated protein 1, ETW81403.1), showed high expression levels in H. annosum s.s. Isotig10204 has the GO annotation GO:0043581 (mycelium development), suggesting that it, just like the delta-12 FAD genes, is involved in host colonization. The function of isotig12111/ETW81845.1 and isotig10242/ ETW81403.1 cannot be accurately predicted using sequence similarity and GO annotation as neither transcript has GO annotations. Further studies of these two candidates are needed to establish their role in virulence.
Even with a small data set like ours, novel Norway spruce transcripts without significant alignment to any previously annotated protein in Genbank but present in the P.abies 1.0 High confidence gene catalog were identified. Predicting a role for these gene models in Norway spruce defence responses. Thus a more complete survey of the transcriptional responses in the interactions between Norway spruce and its major pathogen H. annosum s.l. would probably provide a better understanding of gymnosperm defence.
S1 Fig. The bioinformatics and statistics pipeline.
S2 Table. Transcriptome assembly and annotation statistics.
S4 Table. The 150 most highly expresses isotigs included in the cluster analysis and their cluster.
We thank Drs. Jenny Arnerup, Kerstin Dalman and Åke Olson for valuable discussions on analyses and results and Diem Nguyen for reviewing the language. We thank Dr. Gunilla Swedjemark and Dr. Bo Karlsson at Skogforsk, The Forestry Research Institute of Sweden, for sharing information about the clones and the site with us. The skilled assistance of Karolin Axelsson in the field is acknowledged. The Norwegian Sequencing Centre is acknowledged for the 454-sequencing.
Conceived and designed the experiments: KL MD ME JS FOA. Performed the experiments: KL MD ME KI MNG. Analyzed the data: KL ME MBD. Contributed reagents/materials/analysis tools: MBD. Wrote the paper: KL ME.
- 1. Danielsson M, Lunden K, Elfstrand M, Hu J, Zhao T, Arnerup J, et al. Chemical and transcriptional responses of Norway spruce genotypes with different susceptibility to Heterobasidion spp. infection. BMC Plant Biology. 2011;11:154(154). CABI:20113390730. pmid:22067529
- 2. Franceschi VR, Krokene P, Christiansen E, Krekling T. Anatomical and chemical defenses of conifer bark against bark beetles and other pests. New Phytologist. 2005;167(2):353–75. An 2005:923821. pmid:15998390
- 3. Adomas A, Asiegbu FO. Analysis of organ-specific responses of Pinus sylvestris to shoot (Gremmeniella abietina) and root (Heterobasidion annosum) pathogens. Physiological and Molecular Plant Pathology. 2006;69(4–6):140–52. WOS:000249190200004.
- 4. Arnerup J, Lind M, Olson A, Stenlid J, Elfstrand M. The pathogenic white-rot fungus Heterobasidion parviporum triggers non-specific defence responses in the bark of Norway spruce. Tree Physiology. 2011;31(11):1262–72. WOS:000297223100011. pmid:22084022
- 5. Fossdal CG, Nagy NE, Hietala AM, Kvaalen H, Slimestad R, Woodward S, et al. Indications of heightened constitutive or primed host response affecting the lignin pathway transcripts and phenolics in mature Norway spruce clones. Tree Physiology. 2012;32(9):1137–47. WOS:000308947900009. pmid:22899808
- 6. Deflorio G, Horgan G, Woodward S, Fossdal CG. Gene expression profiles, phenolics and lignin of Sitka spruce bark and sapwood before and after wounding and inoculation with Heterobasidion annosum. Physiological and Molecular Plant Pathology. 2011;75(4):180–7. WOS:000291484500008.
- 7. Yaqoob N, Yakovlev IA, Krokene P, Kvaalen H, Solheim H, Fossdal CG. Defence-related gene expression in bark and sapwood of Norway spruce in response to Heterobasidion parviporum and methyl jasmonate. Physiological and Molecular Plant Pathology. 2012;77(1):10–6. WOS:000300280800003.
- 8. Arnerup J, Nemesio-Gorriz M, Lundén K, Asiegbu FO, Stenlid J, Elfstrand M. The primary module in Norway spruce defence signalling against H. annosum s.l. seems to be jasmonate-mediated signalling without antagonism of salicylate-mediated signalling. Planta. 2013;237(4):1037–45. pmid:23223898
- 9. Bendz-Hellgren M, Lipponen K, Solheim H, Thomsen IM. The Nordic countries. In: Woodward S, Stenlid J, Karjalainen R, Hüttermann A, editors. Heterobasidion annosum: biology, ecology, impact and control. London: CAB International; 1998. p. 333–45.
- 10. Dalman K, Olson A, Stenlid J. Evolutionary history of the conifer root rot fungus Heterobasidion annosum sensu lato. Molecular Ecology. 2010;19(22):4979–93. WOS:000283986700013. pmid:20964759
- 11. Otrosina WJ, Garbelotto M. Heterobasidion occidentale sp nov and Heterobasidion irregulare nom. nov.: A disposition of North American Heterobasidion biological species. Fungal Biology. 2010;114(1):16–25. WOS:000277408300004. pmid:20965057
- 12. Niemelä T, Korhonen K. Taxonomy of the genus Heterobasidion. In: Woodward S, Stenlid J, Karjalainen R, Hüttermann A, editors. Heterobasidion annosum: biology, ecology, impact and control. London: CAB International; 1998. p. 27–33.
- 13. Swedjemark G, Johannesson H, Stenlid J. Intraspecific variation in Heterobasidion annosum for growth in sapwood of Picea abies and Pinus sylvestris. European Journal of Forest Pathology. 1999;29(4):249–58. WOS:000082165900001.
- 14. Olson A, Aerts A, Asiegbu F, Belbahri L, Bouzid O, Broberg A, et al. Insight into trade-off between wood decay and parasitism from the genome of a fungal forest pathogen. New Phytologist. 2012;194(4):1001–13. WOS:000303435400014. pmid:22463738
- 15. Karlsson B, Högberg KA. Genotypic parameters and clone x site interaction in clone tests of Norway spruce (Picea abies (L.) Karst.). Forest Genetics 1998;5:21–30.
- 16. Stenlid J, Karlsson JO. Partial intersterility in Heterobasidion annosum. Mycological Research. 1991;95:1153–9. WOS:A1991GN83200001.
- 17. Chang S, Puryear J, Cairney J. A simple and efficient method for extracting RNA from pine trees. Plant Molecular Biology Reporter. 1993;11(2):113–6.
- 18. Conesa A, Götz S, García-Gómez JM, Terol J, Talón M, Robles M. Blast2GO: A universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics. 2005;21(18):3674–6. pmid:16081474
- 19. Huson DH, Mitra S, Ruscheweyh HJ, Weber N, Schuster SC. Integrative analysis of environmental sequences using MEGAN4. Genome Res. 2011;21(9):1552–60. Epub 2011/06/22. pmid:21690186; PubMed Central PMCID: PMCPMC3166839.
- 20. Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, Miller W, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research. 1997;25(17):3389–402. WOS:A1997XU79300002. pmid:9254694
- 21. Anders S, Huber W. Differential expression analysis for sequence count data. Genome Biology. 2010;11(10). R106 WOS:000287378900008. pmid:20979621
- 22. Bourgon R, Gentleman R, Huber W. Independent filtering increases detection power for high-throughput experiments. Proceedings of the National Academy of Sciences of the United States of America. 2010;107(21):9546–51. WOS:000278054700015. pmid:20460310
- 23. Vestman D, Larsson E, Uddenberg D, Cairney J, Clapham D, Sundberg E, et al. Important processes during differentiation and early development of somatic embryos of Norway spruce as revealed by changes in global gene expression. Tree Genetics & Genomes. 2011;7(2):347–62. WOS:000288656800011.
- 24. Palovaara J, Hakman I. Conifer WOX-related homeodomain transcription factors, developmental consideration and expression dynamic of WOX2 during Picea abies somatic embryogenesis. Plant Molecular Biology. 2008;66:533–49. pmid:18209956
- 25. Pfaffl MW, Horgan GW, Dempfle L. Relative expression software tool (REST (c)) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Research. 2002;30(9). e36 WOS:000175591100030. pmid:11972351
- 26. Kumar S, Blaxter ML. Comparing de novo assemblers for 454 transcriptome data. BMC Genomics. 2010;11. 571 WOS:000283662000001. pmid:20950480
- 27. Nystedt B, Street NR, Wetterbom A, Zuccolo A, Lin Y-C, Scofield DG, et al. The Norway spruce genome sequence and conifer genome evolution. Nature. 2013;497(7451):579–84. WOS:000319556100035. pmid:23698360
- 28. Johnk N, Hietala AM, Fossdal CG, Collinge DB, Newman MA. Defense-related genes expressed in Norway spruce roots after infection with the root rot pathogen Ceratobasidium bicorne (anamorph: Rhizoctonia sp.). Tree Physiology. 2005;25(12):1533–43. WOS:000234019900007. pmid:16137939
- 29. Fossdal CG, Nagy NE, Johnsen O, Dalen LS. Local and systemic stress responses in Norway spruce: Similarities in gene expression between a compatible pathogen interaction and drought stress. Physiological and Molecular Plant Pathology. 2007;70(4–6):161–73. WOS:000253485700008.
- 30. Karlsson M, Olson A, Stenlid J. Expressed sequences from the basidiomycetous tree pathogen Heterobasidion annosum during early infection of scots pine. Fungal Genetics and Biology. 2003;39(1):51–9. WOS:000182941400005. pmid:12742063
- 31. Haas BJ, Chin M, Nusbaum C, Birren BW, Livny J. How deep is deep enough for RNA-Seq profiling of bacterial transcriptomes? BMC Genomics. 2012;13.: 734 WOS:000314607900006. pmid:23270466
- 32. Hietala AM, Kvaalen H, Schmidt A, Johnk N, Solheim H, Fossdal CG. Temporal and spatial profiles of chitinase expression by Norway spruce in response to bark colonization by Heterobasidion annosum. Applied and Environmental Microbiology. 2004;70(7):3948–53. WOS:000222758600021. pmid:15240268
- 33. Koutaniemi S, Warinowski T, Karkonen A, Alatalo E, Fossdal CG, Saranpaa P, et al. Expression profiling of the lignin biosynthetic pathway in Norway spruce using EST sequencing and real-time RT-PCR. Plant Molecular Biology. 2007;65(3):311–28. WOS:000249460400008. pmid:17764001
- 34. Urano D, Jones AM. Heterotrimeric G protein-coupled signaling in plants. Annual review of plant biology. 2014;65:365–84. MEDLINE:24313842. pmid:24313842
- 35. Ralph SG, Jancsik S, Bohlmann J. Dirigent proteins in conifer defense II: Extended gene discovery, phylogeny, and constitutive and stress-induced gene expression in spruce (Picea spp.). Phytochemistry. 2007;68:1975–91. pmid:17590394
- 36. Richard S, Morency MJ, Drevet C, Jouanin L, Seguin A. Isolation and characterization of a dehydrin gene from white spruce induced upon wounding, drought and cold stresses. Plant Molecular Biology. 2000;43(1):1–10. WOS:000088085300001. pmid:10949369
- 37. Yakovlev IA, Asante DKA, Fossdal CG, Partanen J, Junttila O, Johnsen O. Dehydrins expression related to timing of bud burst in Norway spruce. Planta. 2008;228(3):459–72. WOS:000257594800009. pmid:18493789
- 38. Hanin M, Brini F, Ebel C, Toda Y, Takeda S, Masmoudi K. Plant dehydrins and stress tolerance: versatile proteins for complex mechanisms. Plant signaling & behavior. 2011;6(10):1503–9. MEDLINE:21897131.
- 39. Eldhuset TD, Nagy NE, Volarik D, Borja I, Gebauer R, Yakovlev IA, et al. Drought affects tracheid structure, dehydrin expression, and above- and belowground growth in 5-year-old Norway spruce. Plant and Soil. 2013;366(1–2):305–20.: WOS:000317988600022.
- 40. Velasco-Conde T, Yakovlev I, Pedro Majada J, Aranda I, Johnsen O. Dehydrins in maritime pine (Pinus pinaster) and their expression related to drought stress response. Tree Genetics & Genomes. 2012;8(5):957–73.: WOS:000309173200002.
- 41. Boettcher C, Pollmann S. Plant oxylipins: Plant responses to 12-oxo-phytodienoic acid are governed by its specific structural and functional properties. Febs Journal. 2009;276(17):4693–704. WOS:000268791300009. pmid:19663904
- 42. Raacke IC, Mueller MJ, Berger S. Defects in allene oxide synthase and 12-oxa-phytodienoic acid reductase alter the resistance to Pseudomonas syringae and Botrytis cinerea. Journal of Phytopathology. 2006;154(11–12):740–4. WOS:000242016900015.
- 43. Stintzi A, Weber H, Reymond P, Browse J, Farmer EE. Plant defense in the absence of jasmonic acid: The role of cyclopentenones. Proceedings of the National Academy of Sciences of the United States of America. 2001;98(22):12837–42. WOS:000171806100097. pmid:11592974
- 44. Dharmawardhana DP, Ellis BE, Carlson JE. A Beta-glucosidase from Lodgepole pine xylem specific for the lignin precursor coniferin. Plant Physiology. 1995;107(2):331–9. WOS:A1995QH87200003. pmid:7724669
- 45. Marjamaa K, Lehtonen M, Lundell T, Toikka M, Saranpaa P, Fagerstedt KV. Developmental lignification and seasonal variation in beta-glucosidase and peroxidase activities in xylem of Scots pine, Norway spruce and silver birch. Tree Physiology. 2003;23(14):977–86. WOS:000185466800005. pmid:12952784
- 46. Sakuradani E, Kobayashi M, Ashikari T, Shimizu S. Identification of Delta 12-fatty acid desaturase from arachidonic acid-producing Mortierella fungus by heterologous expression in the yeast Saccharomyces cerevisiae and the fungus Aspergillus oryzae. European Journal of Biochemistry. 1999;261(3):812–20. WOS:000080278800026. pmid:10215899
- 47. Huang YS, Chaudhary S, Thurmond JM, Bobik EG, Yuan L, Chan GM, et al. Cloning of Delta 12-and Delta 6-desaturases from Mortierella alpina and recombinant production of gamma-linolenic acid in Saccharomyces cerevisiae. Lipids. 1999;34(7):649–59. WOS:000081820300001. pmid:10478922
- 48. Minto RE, Blacklock BJ, Younus H, Pratt AC. Atypical Biosynthetic Properties of a Delta(12)/nu+3 Desaturase from the Model Basidiomycete Phanerochaete chrysosporium. Applied and Environmental Microbiology. 2009;75(4):1156–64. WOS:000263119100033. pmid:19088315
- 49. Zhang S, Sakuradani E, Ito K, Shimizu S. Identification of a novel bifunctional Delta 12/Delta 15 fatty acid desaturase from a basidiomycete, Coprinus cinereus TD#822–2. Febs Letters. 2007;581(2):315–9. WOS:000244188700023. pmid:17208234
- 50. Damude HG, Zhang HX, Farrall L, Ripp KG, Tomb JF, Hollerbach D, et al. Identification of bifunctional Delta 12/omega 3 fatty acid desaturases for improving the ratio of omega 3 to omega 6 fatty acids in microbes and plants. Proceedings of the National Academy of Sciences of the United States of America. 2006;103(25):9446–51. WOS:000238660400014. pmid:16763049
- 51. Calvo AM, Gardner HW, Keller NP. Genetic connection between fatty acid metabolism and sporulation in Aspergillus nidulans. Journal of Biological Chemistry. 2001;276(28):25766–74. WOS:000169823300019. pmid:11352908
- 52. Wilson RA, Calvo AM, Chang PK, Keller NP. Characterization of the Aspergillus parasiticus Delta(12)-desaturase gene: a role for lipid metabolism in the Aspergillus-seed interaction. Microbiology-Sgm. 2004;150:2881–8. WOS:000224151100010.
- 53. Paradkar AS, Jensen SE. Functional analysis ofthe gene encoding the Clavaminate synthase-2 isoenzyme involved in clavulanic acid biosynthesis in Streptomyces clavuligerus. Journal of Bacteriology. 1995;177(5):1307–14. WOS:A1995QJ43900025. pmid:7868606
- 54. Zhang ZH, Ren JS, Stammers DK, Baldwin JE, Harlos K, Schofield CJ. Structural origins of the selectivity of the trifunctional oxygenase clavaminic acid synthase. Nature Structural Biology. 2000;7(2):127–33. WOS:000085173000014. pmid:10655615