Role of Long Noncoding RNAs ZlMSTRG.11348 and UeMSTRG.02678 in Temperature-Dependent Culm Swelling in Zizania latifolia

Temperature influences the physiological processes and ecology of both hosts and endophytes; however, it remains unclear how long noncoding RNAs (lncRNAs) modulate the consequences of temperature-dependent changes in host–pathogen interactions. To explore the role of lncRNAs in culm gall formation induced by the smut fungus Ustilago esculenta in Zizania latifolia, we employed RNA sequencing to identify lncRNAs and their potential cis-targets in Z. latifolia and U. esculenta under different temperatures. In Z. latifolia and U. esculenta, we identified 3194 and 173 lncRNAs as well as 126 and four potential target genes for differentially expressed lncRNAs, respectively. Further function and expression analysis revealed that lncRNA ZlMSTRG.11348 regulates amino acid metabolism in Z. latifolia and lncRNA UeMSTRG.02678 regulates amino acid transport in U. esculenta. The plant defence response was also found to be regulated by lncRNAs and suppressed in Z. latifolia infected with U. esculenta grown at 25 °C, which may result from the expression of effector genes in U. esculenta. Moreover, in Z. latifolia infected with U. esculenta, the expression of genes related to phytohormones was altered under different temperatures. Our results demonstrate that lncRNAs are important components of the regulatory networks in plant-microbe-environment interactions, and may play a part in regulating culm swelling in Z. latifolia plants.


Introduction
The smut fungus Ustilago esculenta infects the host Zizania latifolia plant causing formation of culm gall [1][2][3]. Apart from culm hypertrophy, which represents the characteristic symptom associated with infection of this fungus, no apparent change appears in Z. latifolia plants [2]. In maize plants, Ustilago maydis generally induces tumours once it invades plant tissues [4], however tumours initiation by U. esculenta is temperature dependent, and tends to occur in seasons with temperature ranging from 21 • C to 27 • C under field conditions. When plants grow at temperatures above 30 • C or below 15 • C, swollen galls do not develop though U. esculenta infection [5,6]. To date, the underlying mechanism responsible for this obvious difference in gall formation dynamics between thermal conditions is unknown.
As a fungus endophyte, U. esculenta completes its life cycle in plant tissues, leading to formation of culm gall [3,7]. In terms of the 'disease triangle,' disease outbreaks occur only under proper environmental conditions following pathogen colonisation and proliferation in host plants. Among the environmental factors, temperature, which generally influences the physiology, ecology, and growth of hosts and microorganisms, is considered to be integral to the promotion of disease outbreaks in plants [8].
Moreover, temperature can significantly affect the interactions between plants and microbes [9]. For instance, in Arabidopsis thaliana, cold leads to derepression of salicylic acid (SA) immunity under pathogen attack, thereby accumulating SA and pathogenesis-related proteins, a response that becomes repressed at optimal temperatures [10,11]. Further, altered plant performance in response to temperature ranges in biotic environments can alter the disease severity and even microbe phenotype [8]. Interestingly, the behaviours of microorganisms may differ across temperatures in plant-microbe interaction due to dimorphic transition of fungi for pathogenesis [12][13][14], inducing non-pathogenic attacks and decreased virulence protein stability [15,16]. Although pronounced changes regarding plant-microbe interactions in response to temperature have been described, the underlying mechanisms remain elusive.
Amino acids, constituents of proteins, function in many biological processes [17]. Some amino acids have been shown to play important roles in tumour progression. For example, serine and glycine contribute to tumour growth [18], asparagine and serine to breast cancer [19,20], glutamine to the growth and metabolism of tumour cells [21]; arginine, lysine, leucine, methionine and isoleucine are reportedly required for tumour formation induced by U. maydis in maize plants [22].
Amino acids also fuel the function of microorganisms during interactions with their host. For instance, methionine is required for Magnaporthe oryzae in appressoria formation to facilitate penetration of host leaves [23]. Additionally, amino acids take part in nitrogenfixing symbiosis [24][25][26][27], while, threonine, isoleucine, and methionine are involved in response to heat shock stress in fungi [28]. However, the mechanisms driving these responses remain largely unclear. The precise role played by amino acids in plant-microbeenvironment interactions requires further investigation.
Long noncoding RNAs (lncRNAs) are RNAs longer than 200 bp without apparent coding potential. They are classified as long intergenic noncoding RNAs, intronic noncoding RNAs and natural antisense transcripts. LncRNAs regulate gene expression through sequence complementarity or homology with RNAs, forming molecular frames for macromolecular complexes [29]. In plants, lncRNAs function in plant responses to abiotic and biotic stresses and are reportedly involved in modulating salt stress [30], immunity in A. thaliana against Fusarium oxysporum [31] and Pseudomonas syringe [32], plant cold acclimation [33], as well as plant response to high temperature stress [34]. A striking regulatory role for lncRNAs in tumour formation has also been extensively documented in animals and humans, including roles in oral squamous cell carcinoma tumorigenesis [35] and gastric cancer [36]. Indeed, lncRNAs have been evaluated in the progression of several cancers in animals; however, no study has yet reported on whether lncRNAs are directly involved in plant tumour formation.
The formation of U. esculenta-induced culm gall in Z. latifolia is accompanied by specific biochemical and physiological changes, such as decreased activities of catalase and superoxide dismutase [37], enhanced photosynthesis [2], altered sugar metabolism [38], and induced cytokinin and indole acetic acid production [3]. Similarly, in tumours induced by U. maydis in maize, cytokinin production and nitrogen transport are promoted [4,39]. Although U. maydis-maize interactions have been investigated in the context of tumour formation [4], the molecular mechanisms associated with temperature effects on plant tumour formation still remain unknown.
Herein, we aimed to analyse the whole transcriptome of Z. latifolia infected with U. esculenta at different temperatures, resulting in the identification of lncRNAs and genes relevant to culm gall formation. These findings might provide new insights into how lncRNAs contribute to plant tumour formation, and an inventory of lncRNA for future studies in plant-microbe-environment interactions.

Plant Culm Morphology
The appearance and morphology of plant culms are shown in Figure 1. Z. latifolia plants infected by U. esculenta (JB) produced a swollen culm gall at 25 • C (JB25). The other plants, including Z. latifolia plants without U. esculenta infection (CK) grown at 25 • C (CK25), as well as CK and JB plants grown at 35 • C (CK35, JB35), did not produce swollen culm gall, and exhibited a significant decrease in length, weight, perimeter, and volume in culms compared to JB25 (p < 0.05). These results clearly demonstrate that culm gall can form in JB plants grown at 25 • C, while gall induction was completely suppressed by a temperature of 35 • C.

Identification and Characterisation of lncRNAs in Z. latifolia and U. esculenta
To develop a comprehensive catalogue of lncRNAs, whole transcriptome strandspecific RNA sequencing was performed for JB and CK plants grown at 25 • C and 35 • C. Overall, 3194 and 173 lncRNAs were identified in Z. latifolia and U. esculenta, respectively. The length of the transcript and exon counts as well as the expression level of the putative lncRNAs were compared with those of mRNA identified in Z. latifolia and U. esculenta ( Figure S1). The length distribution of the identified lncRNAs was shorter than that of the mRNAs, while the exon number of lncRNAs was fewer than that of the mRNAs. The expression level of each transcript was then estimated using fragments per kilobase of exon per million mapped reads (FPKM), and the overall expression level of lncRNAs was lower than that of the mRNA.  Table S1).

Functional Analysis of lncRNAs
To analyse the functions of lncRNAs, potential cis-targets of lncRNAs were predicted. In Z. latifolia, 126 potential cis-target genes of DE lncRNAs were identified (Table S2), and 91 potential cis-target genes were identified for lncRNAs uniquely expressed in JB25 (Table S3), whereas only four potential cis-target genes of DE lncRNAs were identified in U. esculenta (Table 1, Figure 3a).  Further, the 126 DE lncRNAs in Z. latifolia and their potential cis-target genes were divided into three groups (G1-G3) according to their expression patterns under different temperatures (Figure 3b). In group G1, there were 76 lncRNAs differentially expressed between JB25 and JB35, indicating that these lncRNAs, and their potential cis-target genes, may be involved in temperature-dependent gall formation. Meanwhile, in group G2, 31 DE lncRNAs were found between CK25 and CK35, implying that they, along with their potential cis-target genes, are not associated with temperature-dependent gall formation. In addition, in group G3, 19 DE lncRNAs were identified between both JB25/JB35 and CK25/CK35, suggesting that some of these lncRNAs and their potential cis-target genes may be involved in temperature-dependent gall formation.
To validate the putative relationship between lncRNA ZlMSTRG.30807 and ZlRPM1, their expression levels were examined by quantitative reverse-transcription PCR (qRT-PCR) ( Figure 4c). qRT-PCR results were consistent with the RNA-seq data for expression of lncRNA and mRNA ( Figure S2). Additionally, expression levels of the pattern-triggered immunity (PTI) marker genes ZlWRK29 and ZlFRK1, as well as UePEP1, which encodes the effector pep1 in U. esculenta, were also examined by qRT-PCR. Through qRT-PCR analysis, we observed downregulation of ZlWRK29 and ZlFRK1 in JB25 compared to JB35, and slight difference between CK25 and CK35. Moreover, increased expression of UePEP1 was observed in JB25 compared to JB35 (Figure 4d,e).
Additionally, the expression pattern of ZlRPM1 in Z. latifolia differed from that in A. thaliana [40]. Thus, sequence alignment and phylogenetic analysis were further performed ( Figure 5). Phylogenetic analysis showed that ZlRPM1 is homologous with RPM from other plant species, including Oryza brachyantha, Oryza sativa, and Sorghum bicolor. Further, results of amino acid sequence alignment revealed that a portion of the ZlRPM1 amino acid sequence was lost compared with that of RPM1 from other plant species.
To further investigate the role of ET in temperature-dependent gall formation, qRT-PCR was performed to measure the expression of ZlMSTRG.09543, ZlETF1, and ZlACO in Z. latifolia (Figure 6c). The qRT-PCR results were consistent with our RNA-seq data ( Figure S2). ZlACO gene showed a significantly lower expression in JB25 compared to JB35, while slightly lower expression in CK25 compared to CK35.
Furthermore, DE genes involved in cytokinin metabolism were identified in Z. latifolia ( Table 2). Two of three genes (Zl.15021 and Zl.20687) encoding isopentenyltransferase, the key enzyme in cytokinin biosynthesis, were upregulated in JB25 compared to JB35. In addition, two of three genes (Zl.05044 and Zl.12768) encoding cytokinin dehydrogenase, an enzyme responsible for cytokinin degradation, were downregulated in JB25 compared to JB35; however, the expression of genes associated with cytokinin metabolism in CK25 differed only slightly from those in CK35.

Amino Acid Metabolism
To assess whether amino acids involve in the interaction between U. esculenta and its host plant Z. latifolia, lncRNAs with its cis-target genes related to amino acid metabolism were identified. Among these lncRNAs, ZlMSTRG.11348 was uniquely expressed in Z. latifolia in JB25, its cis-target gene may be bark storage protein A, ZlBSP (Table S3, Figure 7a). We also found ZlBSP was downregulated in JB25 compared to JB35, while no significant difference between CK25 and CK35. Moreover, 194 DE genes involved in amino acid metabolism were identified in Z. latifolia, among which 18, 11, and four genes were involved in glutamate, glycine, and histidine metabolism, respectively (Table S6).
In U. esculenta, DE lncRNAs UeMSTRG.02678 was also identified, and predicted to interact with the gene encoding major facilitator superfamily transporter, UeMFS (Table 1, Figure 7b). Both UeMSTRG.02678 and UeMFS were upregulated in JB25 compared to JB35.
To further investigate how amino acids influence the growth of U. esculenta under different temperatures, in vitro culture of U. esculenta with glutamate, glycine, and histidine were performed (Figure 7f). At 35 • C, the growth of U. esculenta in basic medium (BM) added with glycine and histidine was similar to that in control. Notably, the growth of U. esculenta in BM exhibited a significantly slower following addition of glutamate from 12 h to 72 h compared to control. However, at 25 • C, the growth of the fungus in BM with addition of amino acids was significantly faster from 48 h to 72 h compared to control.

Discussion
Temperature is a primary environmental factor that affects the growth and development of plants. At elevated temperatures (usually ≥ 30 • C), both developmental and growth parameters are affected [41], and growth of microbes in plants and outcome of plant-microbe interactions may be modified [7]. Temperature influences the formation of culm gall caused by U. esculenta in Z. latifolia [5]; however, the regulatory network that controls the expression of genes during culm gall development is not well studied. Whether lncRNAs act as a developmental regulator of culm gall remains unknown. In this study, we examined whether temperature play a role in activation or repression of genes associated with culm gall formation, and the role of lncRNAs in controlling differential expression of numerous genes during gall formation. Our results have revealed that a number of plant defence response, phytohormone, and amino acid metabolism-related genes are potentially controlled by temperatures in host and the fungus U. esculenta.

Culm Gall Formation Is Regulated by Temperature
Gall formation in Z. latifolia is significantly influenced by temperature, and it can initiate when temperatures fluctuate between 21 • C and 27 • C in the field [5]. In this study, we noticed a culm gall formation in Z. latifolia plants infected with U. esculenta at 25 • C; however, gall was not observed in control plants without U. esculenta infection, indicating that U. esculenta is the causal agent for culm gall formation ( Figure 1). Interestingly, when U. esculenta-infected plants were grown at a high temperature (35 • C), no swollen gall was produced, demonstrating that gall formation in U. esculenta-infected plants appears to depend on temperature (Figure 1).
LncRNAs have been recognised as important regulators of gene expression in plant cell growth [42] and play a role in plant-microbe interactions [32]. Therefore, temperature may regulate the expression of lncRNAs and further modulate genes involved in plant physiological processes. For instance, heat alters the expression pattern of lncRNAs and regulate the expression of their downstream genes involved in specific biological pathways responding to stresses in Brassica juncea [34]. In this study, we, therefore, performed RNA sequencing of culm samples from Z. latifolia plants infected with or without U. esculenta infection grown at 35 • C or 25 • C. A total of 3194 and 173 lncRNAs were identified in Z. latifolia and U. esculenta, respectively ( Figure 2).

Temperature Alters lncRNA-Mediated Genes Related to Plant Defence Response and Hormone
We could not find DE lncRNAs associated with cell division in Z. latifolia; however, we were able to detect DE lncRNAs that is involved in plant defence response ( Figure 4). Specifically, in JB plants, lncRNA ZlMSTRG.30807 and its cis-target gene ZlRMP1 were downregulated in JB25 at low temperature (25 • C) compared to JB35 at high temperature (35 • C). Contrarily, in CK plants, no significant difference in the expression of lncRNA ZlMSTRG.30807 and its cis-target gene ZlRMP1, was observed between CK25 and CK35. In addition, genes involved in plant defence responses were differentially expressed at different temperatures (Table S4). Activation of plant defence response might possibly contribute to the low plant growth at higher temperature [43].
We observed DE genes involved in effector-triggered immunity (ETI) and PTI, including marker genes of PTI ZLWRKY29 and ZlFRK1, indicating that ETI and PTI were largely compromised and that programmed cell death, hypersensitive response, ROS burst, and activation of the plant stress hormone regulatory network, were inhibited in JB under cooler temperature ( Figure 4; Table 2) [44][45][46]. Consistently, the expression of ZlAIL5, regulated by DE lncRNA ZlMSTRG.09415, as well as genes involved in ET biosynthesis, was lower in JB25 than JB35 (Figure 6), which may cause suppression of ET biosynthesis [47] and downregulation of ET-regulated defence-related gene expression [48], further leading to inhibition of the ET-dependent disease resistance response in JB25 [49]. Moreover, upregulation of effector genes in U. esculenta may contribute to the inhibition of the defence response in JB under cooler temperatures (Figure 4, Table S1) [50][51][52][53]. These results were consistent with a previous study [3], while different slightly from a report on A. thaliana [40]. The loss of a portion of the ZlRPM1 amino acid sequence caused by the long-term cultivation and artificial selection of this crop may account for this difference.
Our results showed that high temperature induces increased expression of ZlRPM1 regulated by lncRNA ZlMSTRG.30807 and activation of plant defence response, possibly leading to suppressed fungal cell proliferation, whereas under cooler temperature, downregulation of ZlMSTRG.30807 results in an inhibition of defence response, which would not only provide the opportunity for U. esculenta to proliferate in plant tissues but also induces plant cell division [43]. Upregulation of genes associated with cytokinin biosynthesis and downregulation of cytokinin degradation genes were observed in JB25, suggesting that greater cytokinetic activity might involve in gall formation [3]. These results indicated that lncRNA differential expression driven by temperature variation might regulate the plant defence response in Z. latifolia infected with U. esculenta. However, the functional analysis of lncRNA ZlMSTRG.30807 in the regulation of plant defence responses requires further investigation.

Expression of Amino Acid Metabolism and Sugar Metabolism Genes Influences Gall Formation
Previous studies have shown that amino acids participate in many tumour pathways. For instance, restriction of serine and glycine drives the accumulation of toxic sphingolipids within tumours and reduces tumour cell growth [18]. Further, histidine deprivation reduce the growth of Drosophila tumour cells [54]. In order to examine whether lncRNAs are possibly involved in the amino acid mechanism at different temperatures, we identified genes involved in the amino acid metabolism from our RNA-seq data. We identified only a few DE genes involved in the amino acid biosynthesis between CK plants grown at low temperature (25 • C) and high temperature (35 • C), while the upregulation of ZlPSO, ZlHID, and ZlGLS, and downregulation of ZlGNS and ZlADG was observed in JB plants at low temperature (25 • C) compared to high temperature (35 • C) (Figure 7), suggesting that variation in ambient temperature scarcely affected the amino acid mechanism in CK plants, but promoted the synthesis of glycine, histidine and glutamate in Z. latifolia infected by U. esculenta at a cooler temperature [55][56][57].
Besides this, we also found lncRNA ZlMSTRG.11348 was uniquely expressed in JB plants at 25 • C and, thus, may result in a lower expression of ZlBSP in JB25 compared to JB35, indicating that amino acids synthesis was induced in Z. latifolia infected with U. esculenta under cooler temperatures [58]. Hence, these synthesized amino acids may serve as nitrogen sources for plant cell division during gall formation at cooler temperature.
Amino acids also represent the most favourable nitrogen sources assimilated by fungi for growth inside the plant environment [4]. For example, glutamate [25], glycine [24], and histidine [26] are indispensable for the establishment and maintenance of the interaction between microorganisms and their hosts. Also, high glycine and glutamic acid contents were detected in tumours induced by U. maydis [59]. We, therefore, determined whether these amino acids fuelled U. esculenta cell proliferation under cooler temperature. Our qRT-PCR analysis demonstrated an increased expression of UeR40 and a decreased expression of UeGYD and UeGUD in U. esculenta from JB plants at 25 • C compared to 35 • C (Figure 7), together with higher growth rate of U. esculenta in BM added with amino acids compared to control at 25 • C, suggesting that U. esculenta has high amino acid requirements for cell proliferation at lower temperature.
However, genomic analysis of U. esculenta indicated that it lacks certain genes involved in amino acid metabolism [60]. In addition to amino acid synthesis in Z. latifolia infected with U. esculenta at cooler temperature, these results led us to the hypothesis that U. esculenta may be capable of altering amino acid mechanism of Z. latifolia and absorption of amino acids from host plant cells to provide nitrogen sources for fungal cell proliferation [7]. Interestingly, the upregulation of UeMSTRG.02678 and its target gene UeMFS in JB plant at 25 • C compared to 35 • C suggested that uptake of several amino acids was highly induced in U. esculenta under lower temperature [61]. Moreover, DE genes encoding amino acid transporters were identified in U. esculenta from JB plant at 25 • C compared to 35 • C (Table S1), implying more active exchange of substances between U. esculenta and Z. latifolia during culm gall formation under lower temperature.
Previous studies have demonstrated that fungus can influence sugar metabolism and transport in plants to facilitate absorption of sugars from host plants and to promote the cell proliferation during the gall formation [38,62]. In Z. latifoa plants, U. esculenta infection enhance photosynthesis of host plants and induce the synthesis of sugars [2]. Among the cis-target genes of lncRNA grouped in G1 and G3 (Table S2), as well as lncRNA uniquely expressed in JB plant at 25 • C, we identified genes involved in sugar metabolism (Table S3), indicating that temperature variation may affect the sugar metabolism of Z. latifolia infected with U. esculenta. However, high temperature causes a significant reduction in photosynthesis in Z. latifolia [6], and this may limit the supply of plant sugar to fungal cells and thus repress fungal cell proliferation.
To summarize, our results indicate that high temperatures suppress the synthesis of amino acids and plant sugars in Z. latifolia, which further suppress the fungal cell proliferation of U. esculenta. Moreover, lower temperature alters expression of lncRNAs and genes in U. esculenta, this further regulate the synthesis of amino acids and plant sugars in Z. latifolia, which play important roles in providing plant cell nutrients required for cell division and growth. Alternatively, galls may serve as mobilising sinks to drive nutrients (amino acids, plant sugars), thereby improving the growth environment inside infected plant tissues allowing U. esculenta to absorb and utilise nutrients for efficient proliferation [7,62,63]. Plant cell division and fungal proliferation subsequently results in culm gall formation under lower temperature. Notably, it appears that some amino acids are dispensable for tumour formation [18]. However, little is known about the amino acid metabolism in fungi during the plant tumour formation, and their roles in plant-microbe-environment interactions. Therefore, further investigations are required to understand how amino acids participate in gall formation in Z. latifolia, as well as the substances exchange between Z. latifolia and U. esculenta. Based on our results, we created a model illustrating the potential mechanisms involved in temperature-dependent culm gall formation (Figure 8).
In conclusion, in Z. latifolia plants, we identified 293 DE lncRNAs between plants infected with U. esculenta grown at 25 • C and 35 • C, and 149 DE lncRNAs between CK plants without U. esculenta infection grown at 25 • C and 35 • C, while we only identified 10 DE lncRNAs in U. esculenta from JB plants grown at 25 • C and 35 • C. The potential cis-targets of these DE lncRNAs were further predicted, among which we found, in Z. latifolia, cis-target genes were involved in amino acid metabolism, plant defence response, and ET signalling, and in U. esculenta, cis-target genes were related to amino acid transport. Also, RNA-seq data exhibited altered expression of several genes related to amino acid metabolism, plant defence response, ET metabolism, and cytokinin metabolism in Z. latifolia; as well as those involved in amino acid metabolism, effectors, and protein synthesis in U. esculenta between plants infected with U. esculenta grown at 25 • C and 35 • C. These lncRNAs and their potential cis-targets, as well as DE genes, may play importmant roles in the regulatory networks of plant-microbe-environment interactions and temperaturedependent culm gall formation. and it may negatively regulate the expression of the gene encoding bark storage protein, which further induces amino acid biosynthesis under cooler temperatures. These amino acids may not only provide nutrients for plant cell division but also be obtained by U. esculenta through the major facilitator superfamily transporter, which is positively regulated by lncRNA UeMSTRG.02678, and further be used for protein synthesis in fungal cell proliferation. These events induce culm gall formation in Z. latifolia infected by U. esculenta under decreased temperatures. High temperature (35 • C) may inhibit the expression of lncRNA ZlMSTRG.11348 and induce the expression of lncRNA ZlMSTRG.30807 causing inhibition of amino acid metabolism and induced plant defence response in Z. latifolia. The expression of lncRNA UeMSTRG.02678 and its cis-target gene encoding major facilitator superfamily transporter is also suppressed in U. esculenta. These largely inhibits culm gall formation. Red and green indicate increased and decreased expression levels of genes or pathways, wherein most genes are up-regulated and down-regulated, in Z. latifolia infected by U. esculenta grown at 25 • C, respectively. Red arrows indicate the predicted regulation between lncRNAs and their cis-target genes.

Plant Growth Conditions and Sampling
Zizania latifolia 'Chongjiao 1' plants were grown at the Experimental Farm of Zhejiang University, Hangzhou, China (120.2 • E, 30.3 • N). Genetically identical clone pairs differing in the presence (JB) or absence (CK) of U. esculenta in Z. latifolia plants were planted in greenhouse with controlled temperatures. Culm samples collected from JB and CK plants grown at 28 • C/19 • C (day/night), with an average temperature of 25 • C, were named JB25 and CK25, respectively. Meanwhile, culm samples collected from JB and CK plants grown at 38 • C/30 • C (day/night), with an average temperature of 35 • C, were named JB35 and CK35, respectively. Other growth management was carried out in accordance with that described by Yan et al. [2] with 13.5 h illumination duration. The culms were excised, frozen in liquid nitrogen, and stored at −80 • C until RNA extraction. Culm length, weight, and perimeter were measured. Culm volume was measured by the drainage method. Sampling and experiments were performed in triplicate.

Fungal Strains and Culture Conditions
The U. esculenta strains used in this study were UeMT09 and UeMT42, isolated from JB plants. Growth media were prepared as follows: BM basic medium, including 1 g of K 2 HPO 4 , 0.5 g of MgSO 4 ·7H 2 O, 0.02 g of FeSO 4 ·7H 2 O, 0.5 g of KCl, and 18 g of glucose, was dissolved in 1000 mL of distilled water and autoclaved for sterilisation. Amino acids (glutamic acid, glycine, and histidine) were added to BM to a final concentration of 20 mmol/L. The medium was filtered with 0.22 µm Millipore filters.
To investigate influence of amino acids on growth of U. esculenta under different temperatures, in vitro culture of U. esculenta were conducted at 25 • C and 35 • C. Fungal strains were grown in 100 mL of YEPS medium (2% sucrose, 2% tryptone, and 1% yeast extract; pH 7.3) for 24 h at 28 • C and 200 rpm. The strains were then centrifuged (10,000× g, 10 min, 25 • C), and 2 mL strain suspension in 50 mL of the prepared medium supplemented with different amino acids was incubated at 200 rpm and 25 • C or 35 • C. After 0, 12, 24, 36, 48, 60, and 72 h, the absorbance of the suspension solution was determined at 600 nm. BM and potato dextrose broth (PDB) medium were prepared as control and nutritious substrates, respectively.

RNA Extraction, Library Construction, and Sequencing
Total RNA was isolated from each culm sample using TRIzol reagent (Sangon Biotech, Shanghai, China) according to the manufacturer's protocol. The RNA concentrations were quantified by Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). The strand specific sequencing was then performed for the identification of lncRNAs. Total RNA was treated to remove rRNA, and transcriptome libraries were constructed using TruSeq Stranded Total RNA with Ribo-Zero Plant kit (Illumina, San Diego, CA, USA) for mRNA and lncRNAs sequencing. These libraries were run on an Illumina HiSeq X ten sequencer (Illumina), which generated paired-end reads of 150 bp.

Transcriptome Assembly and lncRNA Identification
Raw data were processed by removing the adaptor polluted reads and low-quality reads, trimming the reads with more than 5% N bases and removing rRNA using SOAP2 (version 2.21) [64]. The clean data were mapped to the Z. latifolia genome (http://ibi.zju. edu.cn/ricerelativesgd/download.php) (accessed at 24 May 2017) and U. esculenta [60] genomes using HISAT2 (version 2.0.4) [65]. StringTie (version 1.2.1) was used for recreating transcripts [66]. We then adopted five steps to identify lncRNAs: (1) transcripts with an overage ≥3 calculated by Cufflinks were selected; (2) known protein-coding transcripts were removed; (3) transcripts with length < 200 bp were removed; (4) transcripts that did not pass the protein-coding-score test using the Coding Potential Calculator (CPC) [67] and Coding-Non-Coding Index (CNCI) [68] were eliminated; (5) transcripts were aligned in the pfam HMM database to remove protein-coding domains by PfamScan (E value < 0.001) [69]. The flow chart on transcriptome assembly and putative lncRNA identification was shown in Figure S3.

Differential Expression of lncRNAs and mRNAs
The collected RNA-seq data were used to determine the expression levels of both putative lncRNAs and mRNA transcripts. We analysed the expression distribution in terms of FPKM using RNA-Seq by Expectation-Maximisation (RSEM) software (version 1.2.12) (http://deweylab.biostat.wisc.edu/rsem) with default settings. We then performed differential expression analysis using DESeq2 [70]. Genes/lncRNAs with adjusted p-value < 0.05 and |Log 2 fold-change| > 1 were identified as differential expression.

Prediction and Functional Analysis of lncRNA Targets
To predict the function of lncRNAs, their cis-targets were predicted. We searched for coding genes 100 kb upstream and downstream of a lncRNA as the cis-target gene of the lncRNA. RNAplex was further used to identify the interaction between lncRNA and cis-target mRNA according to the minimum free energy based on the thermodynamic structure to predict the complementary base pairing [71]. Gene functions of cis-target mRNA were assigned according to the best match of the alignments (E-value < 10 −5 ) by BLASTp searching against the NCBI nucleotide database.

Quantitative Reverse-Transcription PCR
To determine the relative expression levels of selected lncRNAs and their target genes, qRT-PCR was performed using specific primers (Table S7). Total RNA was isolated from samples with the same procedures as described for RNA extraction. Using the M-MuLV First Strand cDNA Synthesis Kit (Sangon Biotech) with random primer p(dN) 6 , RNA was reverse-transcribed. The product was used as a template for qRT-PCR, which was carried out using the StepOne apparatus (Applied Biosystems, Foster City, CA, USA) with SGExcel UltraSYBR Mixture (Sangon Biotech). The reactions were as follows: amplification for 10 min at 95 • C, followed by 40 cycles of 15 s at 95 • C, 20 s at 60 • C, and 25 s at 72 • C. UeActin2 and ZlActin2 were used as internal control genes in these experiments. All reactions were conducted in triplicate for both technical and biological repetitions. The 2 −∆∆CT method was used to calculate relative gene expression levels [72,73].

Statistical Analysis
All data obtained were analysed using SPSS software with advanced models (SPSS 20.0, SPSS Inc., IBM, Armonk, NY, USA). Significant differences between treatments were analysed using Tukey's multiple comparison test (p < 0.05). All experiments were conducted with three biological replicates.  Figure S2 Correlation between RNA-seq and qPCR-derived expression data. Log 2 fold changes determined by qPCR plotted against the respective log 2 fold changes generated by DESeq2 RNA-seq analysis. A strong Pearson correlation of r = 0.90 was obtained. A linear trend line is depicted in black. Figure S3 Systematic representation of the approach for identification of lncRNAs in Zizania latifolia and Ustilago esculenta. Table S1 Differentially expressed genes in Ustilago esculenta between Zizania latifolia infected with U. esculenta grown at 25 • C (JB25) and 35 • C (JB35). Table S2 Potential cis-target genes of the differentially expressed lncRNAs in Zizania latifolia between Z. latifolia plants infected with Ustilago esculenta grown at 25 • C (JB25) and 35 • C (JB35), as well as between Z. latifolia plants without U. esculenta infection grown at 25 • C (CK25) and 35 • C (CK35). Table S3 Potential cis-target genes of lncRNAs uniquely expressed in Zizania latifolia infected with Ustilago esculenta grown at 25 • C (JB25). JB35 represents culms of Z. latifolia infected with U. esculenta grown at 35 • C. CK25 and CK35 represent culms of Z. latifolia without U. esculenta infection grown at 25 • C and 35 • C, respectively. Table S4 Differentially expressed genes involved in plant defence response in Zizania latifolia between Z. latifolia plants infected with Ustilago esculenta grown at 25 • C (JB25) and 35 • C (JB35), as well as between Z. latifolia plants without U. esculenta infection grown at 25 • C (CK25) and 35 • C (CK35). Table S5 Full names and abbreviations of transcripts. Table S6 Differentially expressed genes involved in amino acid metabolism in Zizania latifolia between Z. latifolia plants infected with Ustilago esculenta grown at 25 • C (JB25) and 35 • C (JB35), as well as between Z. latifolia plants without U. esculenta infection grown at 25 • C (CK25) and 35 • C (CK35). Table S7 Primers used for quantitative reverse-transcription PCR.