Processing and filtering of transcriptome data
RNA-seq was used to compare S. sclerotiorum gene expression between samples taken during infection of two C. arietinum lines and during growth in vitro. Between 1.8 % to 62 % of sequence reads derived from the infected moderately resistant (MR) line samples collected between 6 – 72 hpi mapped to the reference genome of S. sclerotiorum, while between 0.7 % to 80 % of sequence reads derived from infected susceptible line samples collected between 6- 72hpi mapped back to the S. sclerotiorum genome (Table 1). At 72 hpi, the average percentage of reads mapping to the fungal genome in the S line was significantly higher (69%) than in the MR line (54%), suggesting that the S line tissues were more heavily colonised than those of the MR line (Table 1). Similar observations were previously reported during infection of susceptible and resistant soybean varieties [19]. On average, approximately 18.85 million reads per sample from the in vitro treatment mapped uniquely to the S. sclerotiorum genome. The number of reads assigned to S. sclerotiorum increased later in infection. Such differences have been reported in previous S. sclerotiorum transcriptome studies [5, 19, 18].
Table 1: Summary of the Illumina sequence reads generated by RNA – seq obtained from inoculation of a moderately resistant (MR) chickpea line PBA HatTrick and a susceptible (S) chickpea line Kyabra. The values for each time point are the averages of the three biological replicates.
Host
|
Hours post inoculation
(hpi)
|
Total raw
read pairs
|
Trimmomatic reads retention (%)
|
BBSplit reads separation
|
S. sclerotiorum
|
C. arietinum
|
MR
|
6
|
67,354,385
|
98.7
|
1,207,201
|
66,147,184
|
|
12
|
68,680,857
|
98.76
|
5,929,006
|
62,751,851
|
|
24
|
61,114,985
|
98.91
|
28,078,637
|
33,036,348
|
|
48
|
56,616,306
|
98.85
|
40,566,767
|
16,049,538
|
|
72
|
63,109,260
|
98.92
|
39,012,318
|
24,096,941
|
S
|
6
|
58,025,893
|
98.4
|
414,371
|
57,611,521
|
|
12
|
72,896,961
|
98.4
|
1,851 043
|
71,045,918
|
|
24
|
54,049,381
|
98.4
|
18,414,027
|
35,635,354
|
|
48
|
60,727,165
|
98.5
|
36,714,863
|
24,012,301
|
|
72
|
57,636,636
|
98.5
|
39,273,084
|
18,363,551
|
In vitro
|
0
|
56,566, 082
|
96.8
|
NA
|
NA
|
The similarity of the three biological replicates and the accuracy of the RNA-seq analysis was demonstrated using classic multidimensional scaling (MDS), which shows a multidimensional scaling plot of distances between gene expression profiles (Fig.1). The MDS showed a distinct grouping of samples at early (6-12 hpi), the mid (24 hpi) and late (48-72 hpi) stage of infection. Of the early time point samples, 6 hpi was the most different to in vitro with successive time points becoming more similar to in vitro (Fig 1). There was a clear distinction between the S. sclerotiorum transcriptomes at 24 and 48-72 hpi, an indication of the significant differences in the types of genes expressed at these time points.
Validation of RNA-seq data using reverse transcription-quantitative PCR
To validate the accuracy of the RNA-seq data, five upregulated genes and one downregulated gene in both chickpea lines at 12 hpi (early infection stage), and 48 hpi (late infection stage) were quantified using reverse transcription quantitative polymerase chain reaction (RT-qPCR). The five upregulated genes were sscle_05g041810, sscle_11g084430, sscle_08g067130, sscle_04g033880 and sscle_01g003110 and the downregulated gene was sscle_16g108230 were randomly selected for validation. A list of the genes used for validation, their putative functions and the primer sequences are in Table S1. The expression patterns for each gene, as compared to in vitro, were similar to the RNA-seq data (Fig. 2 A and B). These results thus show a correlation between our qPCR and RNA-seq data.
Genotype-specific and genotype non-specific differential gene expression during Sclerotinia sclerotiorum infection of chickpea
We hypothesised that S. sclerotiorum would upregulate genes during infection of chickpea irrespective of the variety susceptibility level. We identified upregulation of 2,150 and 3,593 and downregulation of 7,341 and 6,894 S. sclerotiorum genes during MR and S line infection, respectively (Additional file: Table S2). During infection, 450, 667, 690, 745 and 715 genes were significantly upregulated in the MR line, and 549, 684, 765, 721 and 874 in the S line respectively at 6, 12, 24, 48 and 72 hpi (Fig.2 and Additional file: Table S2). Of the upregulated genes, 171 were common across time points in MR (Fig. 3A) and 230 in S line (Fig. 3B).
The most upregulated gene throughout infection in both lines, except at 6 hpi in the S line, was an alcohol oxidase (SsAOX; sscle_03g024060). An alcohol oxidase in Cladosporium fulvum has been suggested to be a key component in the detoxification of antifungal compounds released from the plant cell wall during infection [20]. Two putative hydrophobic cell surface proteins were found to be highly upregulated exclusively at 6-12 hpi in both lines relative to in vitro.
The first gene, sscle_12g091650 (logFC = 9.6 – 12.5), contains a hydrophobic surface binding protein A (HsbA) domain (PF12296) which was originally identified in Aspergillus oryzae as a surface protein that plays a key role in both the adhesion to and degradation of hydrophobic surfaces [21]. The second gene, sscle_09g070510 (LogFC = 7.3 – 8.6), contains a repeated fasciclin domain (PF02469) which has been reported in M. oryzae to be important in adhesion and binding to hydrophobic surfaces [22]. Our findings suggest these two genes might have a role during the S. sclerotiorum biotrophic phase on chickpea.
Gene Ontology term enrichment analysis of upregulated genes identifies multiple biological and molecular functions associated with infection
Gene Ontology (GO) enrichment analysis is a powerful technique for analysing differential gene expression data to gain insight into the broader biological processes (BP), molecular functions (MF) and cellular components (CC) of genes. S. sclerotiorum upregulated genes in the current study were significantly enriched with a wide range of GO categories (Additional file: Table S3). The BPs highly enriched during the early stage of infection included oxidation-reduction process (GO:0055114), protein metabolic process (GO:0019538), proteolysis (GO:0006508), cellular response to stimulus (GO:0051716) signal transduction (GO:0007165), carbohydrate metabolic process (GO:0005975) and metabolic processes (GO:0008152) (Additional file: Table S3). The BP category oxidation-reduction process (GO:0055114) was highly enriched exclusively in genes upregulated in the MR line at 6 hpi and 48 hpi, suggesting that S. sclerotiorum focuses on regulating the environment redox status during MR line infection to counter host resistance responses. GO term enrichment analysis also provided an insight into the temporal aspects of the S. sclerotiorum-chickpea interaction. Genes involved in cellular communication (GO:0007154), signalling (GO;0023052), response to stimulus (GO:0050896), and signal transduction (GO:0007165) were enriched in genes upregulated at the early stage of infection (6-24 hpi) indicating the importance of rapid adaptation to in planta growth (Fig.3). Among genes upregulated at the late stage of infection (48-72 hpi), the enriched GO categories included carbohydrate metabolic process (GO:0005975), and metabolic process (GO:0008152) among others, an indication of the importance of utilisation of energy sources during the necrotic phase of S. sclerotiorum infection. The most significantly enriched GO categories in the current study grouped into carbohydrate-active enzymes and affiliated proteins (CAZymes), transporters, transcription factors and other secondary metabolites. Genes were categorised based on their functions and predicted roles to simplify the study, as discussed below.
Genes involved in the degradation of the host cuticle
The plant cuticle is the first physical barrier to pathogen invasion and is composed of a lipid-derived polyester and cuticular waxes [23]. In the current study, S. sclerotiorum genes encoding cutinases and lipases were upregulated throughout infection. Interestingly, four S. sclerotiorum genes encoding lysophospholipase (sscle_02g020060), carboxylesterase (sscle_03g027590), GDSL-lipase-acylhydrolase (sscle_01g004820), and triacylglycerol lipase (scle_01g008640) were significantly upregulated at the late stage of infection specifically in the S line (Additional file: Table S4). This suggests the induction of lipolytic enzymatic activity in S. sclerotiorum may depend on the immunity of the host. Lipases were also reported to act as virulence factors in the fungal phytopathogen B. cinerea [24], suggesting S. sclerotiorum lipases may play a role in virulence.
Table 2: The number of in planta upregulated S. sclerotiorum genes involved in the cell wall and cuticle degradation.
Substrate
|
CWDE category
|
Number of upregulated genes in the category
|
Lipid/cutin
|
Cutin
|
14
|
Polysaccharides
|
Cellulose
|
19
|
|
Arabinogalactan
|
6
|
|
Hemicellulose
|
16
|
|
Mannan
|
7
|
|
Pectin
|
16
|
|
Starch
|
3
|
Proteins/peptides
|
Protein
|
17
|
Genes involved in the degradation of the host cell wall
As a necrotroph, degradation of the host cell wall is important during S. sclerotiorum infection to achieve the required plant cell death for growth and development [25]. A portion of the numerous cell wall degrading enzymes (CWDEs) identified in the S. sclerotiorum genome [11] including those involved in the degradation of lipids, cellulose, arabinogalactan, hemicellulose, mannan, pectin, starch and proteins were upregulated during infection of chickpea (Table 2, Additional file: Table S4). After breaching the cuticle, polygalacturonases (PGs) are often the first lytic enzymes produced by a pathogen [26]. A putative exo-PG (sscle_05g046840, LogFC=3.2-8.2) was the most upregulated relative to in vitro in the current study in both chickpea varieties relative to in vitro throughout the infection (Additional file: Table S4). Four previously characterised PGs: endo-PGs Sspg1 (sscle_16g108170) and Sspg3 (sscle_09g070580), and exo-PGs Ssxpg1 (sscle_02g018610) and Ssxpg2 (sscle_04g035440) were also upregulated in the current study, relative to in vitro (Additional file: Table S4). Infiltration of purified endo-PG into plant leaf tissues causes rapid loss of cell wall integrity followed by cell death, [27] suggesting the importance of Sspg1 and Sspg3 in tissue maceration during S. sclerotiorum infection. Orthologs of Ssxpg1 and Ssxpg2 in B. cinerea (BcPG1 and BcPG2) showed necrosis inducing activities, and disruption of either of the genes reduced virulence [28, 29] an indication of the significant role exo-PGs play in lesion development and host colonisation.
Proteases are hydrolytic enzymes which act as important virulence factors in many fungal plant pathogens by degrading host proteins that are involved in the immune response [30]. The in planta upregulation relative to in vitro of non-aspartyl acid protease (acp1; sscle_11g082980) was observed at all time points, peaking in expression at 24 hpi in both lines (LogFC= 7.2-7.9) (Additional file: Table S4). Several factors control acp1 induction, including glucose levels, nitrogen starvation and acidification [16]. Previous studies found upregulation of acp1 at a later stage of S. sclerotiorum infection in H. annuus cotyledons [16], G. max petioles [19], and B. napus leaves [5], suggesting that acp1 has a possible role in virulence on multiple plant species and that it responds to cues present at different infection stages in different hosts. Another gene encoding an aspartyl protease, sscle_07g058540, was upregulated at all stages of infection in the current study, with a peak expression relative to in vitro at 24 hpi (Additional file: Table S4). The gene sscle_07g058540 is a homologue of several aspergillopepsin-like proteins (cd06097) in aspergillosis of humans which act as a cofactor for the persistence of colonisation [31]. Putting this all together, sscle_07g058540 may be a catalyst that assists S. sclerotiorum growth and development during infection.
S. sclerotiorum secondary metabolite synthesis and detoxification enzymes
Secondary metabolite (SM) polyketide synthases (PKSs) and non-ribosomal peptide synthases (NRPSs) were the major enzymes associated with SM synthesis in S. sclerotiorum and make up to 47.2% of the upregulated SM biosynthesis clusters in the current study (Additional file: Table S5). The SM biosynthesis gene expressed at the highest level (LogFC = 7.6-9.2) was a gene encoding the PKS responsible for dihydroxy naphthalene (DHN) melanin biosynthesis (PKS13; sscle_03g031520) at 6-12 hpi as compared to the in vitro control, indicating a possible role in penetration during chickpea infection (Additional file: Table S5). In a previous study, disruption of S. sclerotiorum genes involved in melanin biosynthesis showed no change in pathogenicity; however, slower development of mycelial and hyphal branching was observed [32]. The current results indicate the importance of melanin to aid appressoria mediated penetration of S. sclerotiorum.
Table 3: Sclerotinia sclerotiorum detoxification enzymes upregulated (LogFC) in planta relative to in vitro.
|
|
MRa line hpi*
|
Sb line hpi*
|
Gene ID
|
Description
|
6
|
12
|
24
|
48
|
72
|
6
|
12
|
24
|
48
|
72
|
sscle_01g003110
|
Glutathione S-transferase
|
3.6
|
3.5
|
4.9
|
3.6
|
4.1
|
-
|
2.5
|
5.2
|
3.1
|
3.8
|
sscle_01g005000
|
Glutathione S-transferase
|
-
|
-
|
3.7
|
3.1
|
2.9
|
-
|
-
|
3.2
|
-
|
-
|
sscle_08g067590
|
Glutathione S-transferase
|
-
|
-
|
-
|
2.8
|
2.9
|
-
|
-
|
-
|
-
|
-
|
sscle_02g021570
|
Laccase
|
4.3
|
-
|
-
|
-
|
-
|
4.6
|
4.5
|
-
|
-
|
-
|
sscle_01g005590
|
Cytochrome P450
|
-
|
-
|
-
|
3.3
|
2.9
|
-
|
-
|
-
|
2.9
|
3.1
|
sscle_04g033880
|
Cytochrome P450
|
4.1
|
4.2
|
5.6
|
4.9
|
5.9
|
-
|
2.2
|
2.3
|
2.6
|
5.7
|
a moderately resistant line; b susceptible line; * hpi = hours post-inoculation.
Glutathione S-transferases (GSTs) play critical roles in detoxification of xenobiotic chemicals in fungi by reducing them to glutathione [33]. The S. sclerotiorum GST most upregulated relative to in vitro in this study was a UDP-glucosyltransferase (Ssbgt1; sscle_01g003110, LogFC = 3.6-5.2) (Table 3). Ssbgt1 plays a role in the degradation of the antimicrobial compound brassinin through glycosylation and is induced in response to the presence of a variety of plant phytoalexins [34].
Other GSTs induced in planta in the current study were sscle_01g005000 (logFC = 2.9-3.7) in both lines and sscle_08g067590 (logFC = 2.8-2.9) in the MR line, at 24-72 hpi (Table 3). Disruption of GST genes AbGSOT1 and AbUre2pB1 in the host generalist brassica pathogen Alternaria brassicicola led to reduced virulence [35]. This indicates the importance of xenobiotic compound detoxification during fungal infection. The greatest upregulation of GSTs was observed in the MR line, possibly a reflection of host resistance exerted by the release of host defence-related antifungal compounds during infection.
Benzoic acid derivatives are aromatic compounds arising from the plant defence β- ketoadipate pathway [36]. The CYP enzyme, benzoate 4-hydroxylase, from Aspergillus niger, was reported to play a role in the hydroxylation of benzoic acid to 4-hydroxybenzoate [37]. An S. sclerotiorum CYP gene, sscle_01g005590, encoding benzoate 4-hydroxylase, was upregulated at 48-72 hpi in both chickpea lines (Table 3), which may suggest higher pressure from host defence toxins at the late stage during S. sclerotiorum-chickpea interaction.
Sclerotinia sclerotiorum signalling pathways are vital during chickpea infection
A total of 24 S. sclerotiorum transcription factors were upregulated in the current study (Additional File: Table S6). Two functionally characterised S. sclerotiorum TFs Pac1 (sscle_06g049830) [38], and Ssfkh1 (sscle_06g049780) [39] were upregulated in planta. Pac1 was upregulated at 12 hpi in S line only (LogFC = 2.8) during chickpea infection (Additional File: Table S6). Pac1 triggers oxalic acid (OA) biosynthesis in response to increased pH and reduces the ambient pH, which in turn causes an increase in Sspg1 and acp1 expression and promotes sclerotial development [38]. The upregulation of Pac1 on the S line may suggest that S line tissues were more alkaline than those of the MR line.
Fungal histidine kinases play a vital role in controlling signalling pathways that regulate osmotic and oxidative stress responses, cell cycle control and virulence [11]. We found that the two-component sensor histidine kinase Shk1 (sscle_16g107650) was upregulated at 12-72 hpi relative to inoculum in both lines (Table 4). A previous study showed disruption of Shk1 led to reduced and altered hyphal growth and failed sclerotia formation [40]. Although Shk1 mutants exhibited normal virulence, they showed sensitivity to osmotic stress and increased resistance to fungicides, which suggest that Shk1 likely works upstream of the MAPK cascade to control these processes.
A substantial portion of putative effectors are upregulated on both host varieties during infection
We compared the S. sclerotiorum expressed genes with the 523 secreted proteins identified in the S. sclerotiorum genome [10], to determine specific temporal changes in their regulation during chickpea infection. Of these, 173 were upregulated in both varieties, and 148 downregulated, with nine of the upregulated genes observed in the MR variety only and 27 in the S variety only (Additional File: Table S7). Of the identified S. sclerotiorum secreted proteins, 78 were predicted to be candidate effectors by Guyon et al. [41] and 70 by Derbyshire et al. [10]. Of these candidate effectors, 32 were upregulated, and 40 downregulated on both hosts during the current study (Additional File: Table S8).
In addition to putative candidate effectors, we also considered the expression of experimentally characterised S. sclerotiorum effectors. Previous studies showed that S. sclerotiorum small cysteine-rich secreted protein SsSSVP1 (sscle_01g003850) plays an essential role during infection by interfering with host respiration and inducing localised tissue necrosis [14]. In the current study, SsSSVP1 was upregulated at 48 hpi in the MR line and 72 hpi in S line (logFC = 5.1 and 5.7), respectively (Additional File: Table S8). S. sclerotiorum SsSSVP1 mutants showed reduced virulence in B. napus and A. thaliana [14]. Similarly, SsSSVP1 upregulation was previously reported during the late stage of infection in B. napus [5, 17] and at all-time points in G. max [19]. The earlier induction SsSSVP1 of in the MR line (48 hpi) compared to the S line (72 hpi) may suggest that temporal regulation of expression of SsSSVP1 may depend on host susceptibility level.
Two S. sclerotiorum necrosis and ethylene-inducing (NEP) proteins (SsNEP1 and SsNEP2) were characterised in a previous study on Nicotiana benthamiana and were reported to function as necrotrophic effectors [42]. The previous study showed upregulation of both genes at mid to later stages of infection with SsNEP2 expressed at a higher level than SsNEP1. In the current study, SsNEP1 (sscle_04g039420) was not differentially expressed, and SsNEP2 (sscle_12g090490) was upregulated at the later stages of infection (48 hpi in MR and at 48-72 hpi In S lines) relative to in vitro (Additional File: Table S8). Orthologs of these two genes in B. cinerea (BcNEP1 and BcNEP2) are both proteins capable of inducing necrosis in dicotyledonous but not monocotyledons host [43].
Expression of genes related to oxalic acid production and reactive oxygen species regulation
Oxalic acid (OA) has roles in weakening the host cell wall, activating hydrolytic enzymes, suppressing the oxidative burst and intensifying programmed cell death (PCD) leading to full colonisation [44]. A gene encoding oxalate decarboxylase (Ss-odc2: sscle_09g069850) was highly upregulated at the very early stage (6 - 24 hpi) of infection (LogFC = 6.5-8.4) and showed lower expression at the later stage (LogFC = 3.8-4.2) of infection relative to in vitro, in both chickpea lines (Table 4). Ss-odc2 protects the pathogen cells by preventing excess accumulation of OA [45]. A previous study suggested that an alternative route of OA biosynthesis may be utilised during S. sclerotiorum early stages of infection [46]. Expression of Ss-odc2 without the induction of Ssoah1 in the current study may support previous findings that OA is not the only source of acidification or determinant of S. sclerotiorum virulence expression [47], or alternatively, the host tissue was already acidic enough for growth.
An S. sclerotiorum gene, sscle_09g069850, with a bicupin domain, was previously reported to be a possible oxalate oxidase enzyme [48]. This gene was highly upregulated at 6-12 hpi (logFC= 7.6-8.7) and expression decreased at 48-72 hpi (logFC = 3.8-4.2) with no expression at 24 hpi, in both chickpea lines, relative to in vitro (Table 4). A previous study suggested that oxalate oxidases play a role in inducing programmed cell death [44]. The pattern of sscle_09g069850 expression in the current study may suggest involvement in both the biotrophic and necrotic stages during chickpea interaction.
Catalases and peroxidases are also important S. sclerotiorum ROS scavengers [49]. Three catalases, sscle_03g026200 (Sscat1), sscle_04g037170, and sscle_15g104430, were upregulated during the late stage of infection (48 -72 hpi) in the MR and 72 hpi in the S line (Table 4). The most upregulated catalase during the current study was the previously characterised Scat1 (sscle_04g037170). Scat1 mutants show slower radial growth, a higher number of small sclerotia and hypovirulence [50]. The upregulation of catalases and peroxidases was observed at an early stage in the MR line and later stage in the S line during infection, suggesting S. sclerotiorum induces ROS scavengers depending on the host speed of employing defence responses.
Table 4. Sclerotinia sclerotiorum reactive oxygen scavenging (ROS) enzymes upregulated (LogFC) during chickpea infection relative to in vitro.
|
|
MRa line hpi*
|
Sb line hpi*
|
Gene ID
|
Description
|
6
|
12
|
24
|
48
|
72
|
6
|
12
|
24
|
48
|
72
|
sscle_15g104430
|
Catalase
|
-
|
-
|
-
|
5.1
|
5.7
|
-
|
-
|
-
|
-
|
4.5
|
sscle_04g037170
|
Catalase
|
-
|
-
|
-
|
5.6
|
5.8
|
-
|
-
|
-
|
-
|
4.7
|
sscle_03g026200
|
Catalase
|
-
|
-
|
-
|
-
|
-
|
-
|
-
|
-
|
-
|
2.9
|
sscle_09g069850
|
Oxalate decarboxylate
|
8.4
|
6.5
|
-
|
4.1
|
4.2
|
8.2
|
8.1
|
-
|
-
|
3.8
|
sscle_04G035020
|
Peroxidase
|
-
|
3.5
|
3.5
|
4.3
|
5.1
|
-
|
-
|
5.3
|
5.4
|
4.7
|
sscle_15g102360
|
Peroxidase
|
3.6
|
-
|
-
|
-
|
-
|
4.1
|
3.8
|
-
|
-
|
-
|
sscle_09g070530
|
Peroxidase
|
4.1
|
-
|
-
|
-
|
-
|
-
|
3.6
|
-
|
-
|
3.6
|
sscle_08g065740
|
Peroxidase heme-thiolate
|
4.5
|
-
|
-
|
-
|
-
|
4.6
|
-
|
-
|
-
|
-
|
a moderately resistant line; b susceptible line; * hpi = hours post-inoculation.