Analysis of the symptom and features associated with infection by A. protococcarum
Microscopic observation of the infection progress revealed that A. protococcarum was an endoparasite that can occupy the microalgal host cell and replace the host cytoplasm (Fig. 1a). In the infection test, the infection rates of G. emersonii by A. protococcarum have gradually increased over time (Fig. 1b). In contrast, healthy and uninfected G. emersonii cells did not exhibit any symptoms during cultivation. Infection stages were determined based on the infection rate. Briefly, healthy and uninfected G. emersonii cells were used as controls (GA). At 3 dpi, the infection rate of G. emersonii reached approximately 5% and was defined as early stage infection (ES). At 4 dpi, the infection rate reached approximately 50% and was defined as medium stage (MS). At 5 dpi, the infection rate reached approximately 90% in G. emersonii cells and was defined as late stage (LS). Samples of GA, ES, MS, and LS were collected and stored in liquid nitrogen for subsequent dual transcriptomic analysis.
The ROS content and Fv/Fm values of G. emersonii cells during infection were measured, and the results are shown in Figs. 1c and 1d. With the increase in culture time, ROS levels gradually increased. At 1, 2, and 3 dpi, the values of Fv/Fm were not significantly decreased compared to those of the control. At 5 dpi, Fv/Fm values decreased significantly.
A global view of transcriptomic analysis of microalga and endoparasite during their interaction
To characterize the gene expression profiles in the host microalga G. emersonii and endoparasite A. protococcarum during infection, we performed a dual RNA-seq analysis at different infection stages. A total of 102 Gb of clean bases were generated from 12 samples, and sequencing data quality is summarized in Additional File 1. For the microalgal host, 1843, 1750, and 2164 differentially expressed genes (DEGs) were upregulated, whereas 1818, 1168, and 1139 DEGs were downregulated in ES, MS, and LS compared with GA, respectively (Fig. 2a). We found that more genes were upregulated, while fewer genes were downregulated in MS. As the stages of infection progressed, the ratio of upregulated genes to the whole DEGs increased from 50.3 % at ES to 65.5 % at LS. Figure 2a illustrated that only 608 DEGs overlapped with ES, MS, and LS, suggesting that the microalgal response due to infection is different at the molecular level in three stages. For fungal parasites, a total of 11352 and 17819 upregulated DEGs were detected in MS and LS compared to the ES group (Fig. 2b). These genes may play a key role in the progression of infection. In addition, pearson correlation coefficients of the RNA-Seq 12 samples are displayed in Fig. 2c.
Gene expression profile of G. emersonii during infection
To further investigate the gene expression profile of G. emersonii during infection, the gene expression of the host G. emersonii was clustered using the short time-series expression miner (STEM). A total of 23248 genes were divided into 50 profiles, but only 22 highly significant expression profiles were chosen for further functional analysis (P < 0.001) (Additional file 2). The 22 profiles were then grouped into seven major clusters based on their expression patterns (Fig. 3a). Compared with the GA, cluster I comprised profiles 27, 28, and 41, where the genes were generally upregulated in both ES and MS stages and were not affected in LS stages. Genes in cluster II, combining profiles 18, 7, and 5, were downregulated in the ES stage but upregulated in both the MS and LS stages. Cluster III (profiles 21, 30, 29, and 20) included genes whose expression was generally upregulated but fluctuated. Genes in cluster IV showed an overall downregulation. The genes in cluster V were downregulated but experienced some variability. The majority of genes in cluster VI were upregulated or were stable in the ES and MS stages and downregulated in the LS stage. The genes in cluster VII were generally upregulated in ES and stable in the MS and LS stages.
To elucidate the biological function in each cluster, GO term enrichment and KEGG pathway analyses were conducted for significant enrichment. The top six GO terms for each cluster are shown in Fig. 3b. Cluster I, which peaked in expression at the MS stage, was significantly enriched for genes involved in hydrolase activity. The DEGs in cluster II were enriched with many ligase activity-related GO terms, cluster VI was enriched with numerous oxidoreductase-related terms, while cluster IV showed no significant results.
KEGG pathway analysis showed that seven clusters mapped significantly with 20 pathways (Fig. 3c). Notably, cluster I was enriched in the ubiquitin-mediated proteolysis pathway, endocytosis, fatty acid biosynthesis, and fatty acid metabolism. Multiple clusters were enriched with endocytosis, proteasome, protein processing in the endoplasmic reticulum, and spliceosome.
Ubiquitin Mediated Proteolysis And Endocytosis In Response To Infection
Given the above-mentioned enriched pathways in the microalga G. emersonii during infection, we focused on the pathways of ubiquitin-mediated proteolysis and endocytosis. In the transcriptome data, the expression of genes encoding enzyme E1 (gene-Cem18475, gene-Cem15045) was upregulated by 2.2- and 1.1-log2fold changes, respectively (Fig. 4a). The expression of genes encoding E2 (gene-Cem18885 gene-Cem21008, gene-Cem11452 and gene-Cem01030) also experienced a significant increase. In contrast, the expression of several genes encoding E3 was downregulated. Moreover, the genes encoding other key enzymes, such as heat shock proteins (HSPs), which are also involved in the endocytosis pathway, were induced.
Potential pathogen receptors and putative R proteins in G. emersonii
In G. emersonii, genes that contained typical pattern-recognition receptor (PRR) functional domains were searched using Pfam annotation. Epidermal growth factor (EGF)-containing, lectin-containing genes, leucine-rich repeat (LRR)-containing genes, and LysM-containing genes with upregulated expression between one or more infection stages are summarized in Fig. 4b. Among these genes, EGF (gene-Cem19681), lectin-containing genes (gene-Cem18265, gene-Cem18266), and LRR-containing genes (gene-Cem02208, gene-Cem04600, gene-Cem11344, gene-Cem14526, gene-Cem14758, gene-Cem16818) were predicted as putative transmembrane proteins using TMHMM v. 2.0. At the same time, LRR-containing genes (gene-Cem08139, gene-Cem08774, gene-Cem12703, gene-Cem13414, and gene-Cem16818) significantly increased in both ES and MS during infection. The LRR-containing gene (gene-Cem02208) was upregulated in ES by 5.6-fold. Moreover, four genes encoding putative R proteins with an LRR domain (gene-Cem07770, gene-Cem16366, gene-Cem22200, and Novel00913) were found, but none were significantly upregulated at the three infection stages. No genes encoding the TIR or NBS domain were detected in the G. emersonii transcripts.
Ros-related Genes In Response To Infection
The generation of superfluous ROS due to pathogens may cause oxidative stress and cell damage. G. emersonii can protect themselves by various ROS-scavenging enzymes. Our results demonstrated that the expression of genes encoding glutathione (gene-Cem18750), ferritin (gene-Cem21474), and CAT (gene-Cem18215) were upregulated at ES and maintained at high levels in MS and LS (Fig. 4c). The genes encoding glutaredoxin (gene-Cem06407, gene-Cem11714) were not significantly modified at ES, but were upregulated in MS and LS. Notably, the gene encoding for ferritin (gene-Cem21474) was upregulated by 7.9-log2fold change at ES, 8.4-log2fold change at MS, and 8.1-log2fold change at LS. The expression of genes encoding glutathione (gene-Cem18750) was also dramatically increased by 5.3-log2fold change at ES, 6.6-log2fold change at MS, and 5.4-log2fold change at MS. These transcriptional results confirmed that the upregulation of genes encoding ROS-scavenging enzymes is one of the defense strategies of G. emersonii during infection.
Some heat shock proteins (HSPs) and transcription factors (TFs) involved during infection
The results of this study showed that 12 genes encoding heat shock proteins were differentially regulated during infection (Fig. 4d). Compared with the uninfected control group, five HSP genes (four HSP90 and one HSP70) were upregulated in one or more infection stages.
RNA-seq analysis revealed 774 differentially expressed TFs in G. emersonii (see Additional file 3). SET, SNF2, GNAT, MYB, AP2-EREBP, and TRAF were the top six differentially expressed TFs. Among them, the identified 5 TFs that were upregulated during the infection included two SNF2, one PLATZ, one FHA, and one orphan (Table 1).
Table 1
Transcriptional factors of G. emersonii that were continuously upregulated because of infection.
No.
|
Gene ID
|
Types
|
Log2FC(ESvsGA)
|
Log2FC(MSvsGA)
|
Log2FC(LSvsGA)
|
1
|
gene-Cem00854
|
PLATZ
|
1.65
|
3.61
|
3.06
|
2
|
gene-Cem04263
|
SNF2
|
1.53
|
1.31
|
1.65
|
3
|
gene-Cem15036
|
FHA
|
2.91
|
2.45
|
1.58
|
4
|
gene-Cem16321
|
SNF2
|
2.08
|
3.57
|
2.51
|
5
|
gene-Cem19050
|
Orphans
|
1.41
|
2.68
|
3.18
|
Log 2FC=log 2 Fold change |
Gene expression profile of A. protococcarum during infection
STEM analysis was carried out to elucidate the gene expression profile of A. protococcarum during infection. In total, 16 profiles were generated, but only seven highly significant expression profiles were chosen for further functional analysis (P<0.001) (see Additional file 4). The seven profiles were subsequently grouped into three clusters based on their expression patterns (Fig. 5a). Cluster I comprised profile 7, 2, 3 and 0, where the genes were generally downregulated compared to those in ES. Genes in cluster II, which combined profiles 13 and 6, were upregulated. Genes in cluster III were upregulated and then downregulated.
GO term enrichment and KEGG pathway analyses were also conducted for the three clusters of A. protococcarum. The top 10 GO terms for each cluster are shown in Fig. 5b. The genes involved in macromolecular complex, intracellular organelle, and organic substance biosynthetic processes tended toward downregulation were related to cluster I. The genes associated with the regulation of GTPase activity and response to host immune that showed an upregulated trend were in cluster II. The genes in cluster III were enriched in the establishment of localization in cells, cellular localization, and vesicle-mediated transport GO terms. These GO terms were both necessary and significant to A. protococcarum infection.
KEGG pathway analysis demonstrated that the three clusters in A. protococcarum were mapped with 25 pathways (Fig. 5c). Interestingly, cluster II, featuring an upregulated trend over time, was enriched in glycosylphosphatidylinositol (GPI)-anchor biosynthesis, endocytosis, phagosome, ubiquitin-mediated proteolysis, SNARE interactions in vesicular transport, MAPK signaling pathway, and amino sugar and nucleotide sugar metabolism.
Expression of CAZymes and pathogen-host interaction genes in A. protococcarum during infection
To assess the potential of A. protococcarum to depolymerize the cell walls of microalgal hosts, CAZymes-containing genes were searched using the dbCAN2 meta server (HMMER). There were 69 putative genes encoding CAZymes differentially expressed and identified in cluster II of A. protococcarum, which exhibited an upregulated temporal trend (see Additional file 5). CAZymes were classified into glycoside hydrolases (GHs), glycosyltransferases (GTs), and auxiliary activities (AAs) super families. Among them, two genes (Cluster-8366.7755 and Cluster-8366.85495) encoding α-glucosidase were upregulated by 6.2- and 10.7-log2fold changes in LS, respectively. Two genes (Cluster-8366.11810 and Cluster-8366.64947) encoding trehalase and one gene (Cluster-8366.6141) encoding cellulase experienced a dramatic increase as well. The expression of genes encoding mannosyltransferase, galactosyltransferase, and glucosyltransferase was significantly upregulated during infection. It should be emphasized that several CAZymes genes associated with chitin binding and chitin synthase (AA15 and GT2) were also significantly upregulated, which might play a role in the chitin synthesis of A. protococcarum.
To predict key genes involved in infection, the pathogenicity genes of A. protococcarum were predicted using BLASTp against the Pathogen-Host Interaction database (PHI-base). In total, 1269 genes associated with key virulence and pathogenicity, accounting for 5.3% of the total predicted genes, were identified (see Additional file 6). Most genes (1231.97%) were expressed in the transcriptome of A. protococcarum during infection. Among them, 72 genes associated with increased virulence (Additional file 7) and 15 genes associated with effectors were significantly expressed at all infection stages (Table 2). Furthermore, 11 putative secretory protein-encoding genes were identified in pathogen-host interaction genes to seek for candidate virulence and effectors in A. protococcarum. However, only a candidate hypervirulence gene (Cluster-8366.11022) was upregulated by 2.1-, 5.9-, and 3.5-log2fold change during infection. Two candidate effectors (Cluster-9881.0, Cluster-8366.6798) were upregulated by 2.4- and 2.5-log2fold changes, during infection.
Table 2
Fifteen genes associated with effector in A. protococcarum were significantly up-regulated in all the infection stages.
No.
|
Gene ID
|
ProteinID
|
PHI ID
|
Gene name
|
Phenotype
|
log2FC(MSvs ES)
|
log2FC(LS vs ES)
|
log2FC(LS vs MS)
|
1
|
Cluster-8366.85734
|
A0A0H3HVK0
|
PHI:5335
|
clpV-5
|
Effector
|
1.729
|
3.8033
|
1.8221
|
2
|
Cluster-8366.8318
|
Q8RP09
|
PHI:981
|
hopI1
|
Effector
|
2.0296
|
6.6465
|
4.3176
|
3
|
Cluster-8366.82670
|
Q79LY0
|
PHI:992/PHI:7237/PHI:7265
|
hopPtoD2/hopAO1/HopPtoD2
|
Effector
|
1.6504
|
4.9986
|
3.0935
|
4
|
Cluster-8366.80929
|
P17778
|
PHI:6101/PHI:6824/PHI:6830
|
YopM
|
Effector
|
6.0254
|
11.636
|
3.4623
|
5
|
Cluster-8366.7571
|
C5BD30
|
PHI:6294
|
EseM
|
Effector
|
2.5059
|
7.5318
|
4.6832
|
6
|
Cluster-8366.72579
|
Q8PC98
|
PHI:7945
|
pip
|
Effector
|
4.4021
|
9.7346
|
2.9476
|
7
|
Cluster-8366.6633
|
Q8PC98
|
PHI:7945
|
pip
|
Effector
|
3.5808
|
10.373
|
4.6194
|
8
|
Cluster-8366.43382
|
Q8XTK9
|
PHI:5119
|
RSp0099
|
Effector
|
1.7086
|
4.8665
|
2.9109
|
9
|
Cluster-8366.14241
|
P17778
|
PHI:6101/PHI:6824/PHI:6830
|
YopM
|
Effector
|
1.6503
|
5.6023
|
3.6416
|
10
|
Cluster-8366.11799
|
Q8PI08
|
PHI:2703
|
xac3090
|
Effector
|
1.5342
|
5.0159
|
3.2262
|
11
|
Cluster-8366.11711
|
Q8XZN9
|
PHI:5173
|
RSc1356
|
Effector
|
2.971
|
6.0476
|
2.8247
|
12
|
Cluster-8366.11710
|
Q8XZN9
|
PHI:5173
|
RSc1356
|
Effector
|
2.3996
|
5.6692
|
3.0096
|
13
|
Cluster-8366.11150
|
Q8XZN9
|
PHI:5173
|
RSc1356
|
Effector
|
3.3433
|
7.5956
|
3.9299
|
14
|
Cluster-8366.11149
|
Q8XZN9
|
PHI:5173
|
RSc1356
|
Effector
|
3.8592
|
11.787
|
3.2499
|
15
|
Cluster-8366.11148
|
Q8XZN9
|
PHI:5173
|
RSc1356
|
Effector
|
3.1417
|
7.4909
|
4.0076
|