FM affects the gall formation and yield of Z. latifolia
After treatment with different concentrations of FM, the timing of gall formation and yield of Z. latifolia was greatly affected. As Fig. 1A shows, the Z. latifolia began harvest from May 27 onward. Under the action of 3.6 g/L FM, the harvest time of Z. latifolia was markedly advanced. The peak harvest period of Z. latifolia under the 1.8, 0.9, and 0.45 g/L FM treatments was respectively delayed by 4, 17, and 19 days, when compared with that under the 3.6 g/L FM application. The control group maintained a lower harvest volume during the entire harvest period.
The FM concentration applied not only affected the timing of gall formation, but also their total yield from Z. latifolia (Fig. 1B). When treated with a higher concentration of FM, a higher yield of Z. latifolia was harvestable. Under the 1.8 g/L and 3.6 g/L FM treatments, the yields of Z. latifolia respectively were 46.07 kg and 44.86 kg, which corresponded to 59.5% and 55.3% increases over the 0.9 g/L FM treatment (28.88 kg). Moreover, when compared with the yield of the 0.45 g/L treatment group (12.09 kg), their increments were 2.8-fold and 2.7-fold greater, respectively.
The harvest timing of Z. latifolia was the latest in the control group, for which it never reached the peak period during the whole harvest period (Fig. 1A, B). Accordingly, the control group had the lowest yield, at only 1.61 kg. Since fertilizer and water management and pest control were identical among the treatment groups, weather conditions were monitored (Fig. 1C, D). Around May 7, the Z. latifolia entered the period of gall formation, about 20 days before the onset of the harvest period (From May 7 to 27). According to the information provided by the local meteorological bureau, from May 7 to 12, the average temperature was around 20°C, a suitable temperature for gall formation. After May 17, however, the experimental field experienced high-temperature weather exceeding 30°C whose daily average temperature was above 25°C for 6 days. This spell of high temperature happened to occur in the critical period of gall formation in the control group.
FM affects the sporidia growth of U. esculenta in vitro
A high concentration of FM completely inhibited sporidia growth of U. esculenta in the YEPS culture. The sporidia began growing so long as the concentration was below 0.09 mg/ml; corresponding inhibition rates of 0.009 mg/ml, 0.018 mg/ml, 0.027 mg/ml, and 0.036 mg/ml FM on sporidia growth were 9.5%, 18.6%, 29%, and 37.8%, respectively (Fig. 2A). Hence, the inhibition rate increased with a rising concentration of FM. The FM resistance was quantified as EC50 after exposure to FM irradiation. The EC50 was 0.042 mg/ml.
Seventeen relative genes possibly involved in the altered responses of sporidia to FM (0.018 mg/ml) versus control group were assessed for their transcript levels via qRT-PCR with paired primers (Table S1). Sporidia isolated from teliospores could be divided into two strains: one containing mfa1.2, mfa1.3, and pra1 alleles, while the other contained mfa2.1, mfa2.3, and pra2 alleles, that were α mating-type alleles in U. esculenta. The expression levels of all of them were significantly repressed, by 38.1%–85.2% (Fig. 2B). The drastic repression of mating relative genes indicated that FM impaired the conjugation formation, the initial step of the mating process in U. esculenta.
Under the stress of FM (0.018 mg/ml), the expression levels of cell metabolism-related genes were also significantly repressed (by 21.8%–89.8%) in U. esculenta (Fig. 2C). Remarkably, gpa3, kss1, and pkaC of U. esculenta were all downregulated more than 80%. Among five chitin synthase genes (chs1 through chs5) responding to FM (Fig. 2D), all were largely downregulated by 84.7%–89.2%. Transcriptional levels of these genes related to cell metabolism and cell wall disturbing were all repressed in response to FM; this strongly implied FM could considerably affect the normal metabolic growth of U. esculenta.
Genome-wide expression analysis of FM effects upon Z. latifolia and U. esculenta
To determine how FM affects the gall formation in Z. latifolia and the interaction between it and U. esculenta, a transcriptome analysis was performed. The swollen stem galls harvest from the control group (without FM) and experimental group receiving the 3.6 g/L FM treatment were used to collect transcriptional information for Z. latifolia and U. esculenta. Samples with three replicates were prepared from these two groups.
Deep RNA sequencing produced the 9.2×107 and 8.6×107 valid reads for the Z. latifolia, 2.4×107 and 2.4×107 valid reads for the U. esculenta libraries, in the control and experimental group, respectively. These sequence reads were mapped onto the genome of Z. latifolia and U. esculenta, resulting in the identification of 46 092 and 7347 genes derived respectively from the Z. latifolia and U. esculenta libraries. Comparative analysis between the control and experimental groups’ expression profiles revealed 663 and 912 transcripts upregulated and downregulated in Z. latifolia, and 34 and 24 transcripts upregulated and downregulated in U. esculenta, respectively (Fig. 3A).
GO enrichment analyses were used to annotate the function of differentially expressed genes (DEGs), which could be assigned to three major categories: molecular function (MF), biological process (BP), and cellular component (CC). In Z. latifolia, the unigenes were categorized into 45 GO terms, with most DEGs belonging to BP and MF (Fig. 3B). In the BP category, phosphorylation, protein modification, phosphorus- and phosphate-containing compound metabolic process, and macromolecule modification were the top four classes enriched by regulated transcripts. In the MF category, many genes were categorized as protein kinases and binding proteins. In U. esculenta, the number of DEGs was small (Table S3), and there was an average distribution in these three enriched GO categories.
As Fig. 3C shows, KEGG pathway enrichment analysis revealed a set of genes involved in galactose metabolism, butanoate metabolism, and starch and sucrose metabolism that were differentially expressed as enriched (corrected p-value < 0.2) in U. esculenta. In Z. latifolia, a fair number of DEGs (corrected p-value < 0.05) involved in plant–pathogen interaction, phenylpropanoid biosynthesis, zeatin biosynthesis, plant hormone signal transduction, flavonoid biosynthesis, biosynthesis of secondary metabolites, and biosynthesis of stilbenoid, diarylheptanoid, and gingerol were evidently enriched. Generally, those genes with a corrected p-value < 0.05 could be considered as an enriched item.
In Z. latifolia, the transcript levels of 77.7% of the DEGs (101 of 130) in the enrichment items were downregulated under the stress of FM (Table S2). Specifically, 94.4% (17 of 18), 73.7% (14 of 19), 83.3% (5 of 6), and 76.8% (43 of 56) of DEGs were downregulated dramatically in the process of plant–pathogen interaction (see also Fig. 4), phenylpropanoid biosynthesis, flavonoid biosynthesis, and biosynthesis of secondary metabolites. Notably, in the process of zeatin biosynthesis (7 genes) and stilbenoid, diarylheptanoid, and gingerol biosynthesis (4 genes), all of the DEGS were especially downregulated. For the process of plant hormone signal transduction, 55% (11 of 20) of the involved DEGs were downregulated (Table S2; see also Fig. 5) and those were related to auxin signaling, seed dormancy, stress response, and disease resistance.
A previous study identified 170 and 205 putative host genes, along with 53 and 71 U. esculenta genes involved in the initial and late stage of culm gall formation28. Compared with those genes, we found that 6 of 13 DEGs in Z. latifolia under FM stress had the same trend of change, in that they were all downregulated (Table 1). In U. esculenta, however, no DEGs with the same change pattern could be discerned by comparison in our experiment.