Multiple MicroRNAs are Involved in Regulating Peanut (Arachis hypogaea L.) Resistance to Sclerotium rolfsii at the Early Stage

Stem rot, caused by soilborne pathogen Sclerotium rolfsii, is one of the most destructive diseases of peanut (Arachis hypogaea L.) worldwide. Although microRNAs (miRNAs) are indispensable regulators for plant defense, the miRNA species have not been explored for peanut immunity against the soilborne pathogen Sclerotium rolfsii. Here, we report a miRNA comparative analysis of the durably resistant peanut variety Rizhaohua-1 (Rzh) and the susceptible peanut variety Jitian-1 (JT) in response to infection by Sclerotium rolfsii. We identified a group of known and novel miRNAs that were differentially expressed upon S. rolfsii infection. The predicted target genes included receptor kinases, transcription factors, and genes involved in the production and transport of metabolites. In this study, we have shown that miRNAs regulate a broad range of genes to respond to the pathogen, and eventually establish a systemic defense network to combat disease.


Introduction
Peanut (Arachis hypogaea L.) is an allotetraploid crop (2n = 4X = 40, AABB genome), which is widely cultivated in tropical and warm temperature regions. Peanut is one of the most important oil crops worldwide. Apart from oil, peanuts are also used for the production of peanut butter, snack products, roasted peanuts, and desserts. Peanut growth area and production reached nearly 30 million ha and more than 48 million tons respectively in 2019 and brought over 22 billion US dollars in 2018 worldwide according to data from FAO (FAOSTAT 2019). However, the peanut cultivation is challenged by abiotic stresses (e.g. high temperature, water deficit, high salinity, and heavy metal) and biotic stresses (e.g. viruses, bacteria, insects, fungi, and weeds) that result in a limitation of productivity and quality in many regions.
Stem rot, or Southern blight, caused by the soilborne pathogen Sclerotium rolfsii Sacc., is a damaging disease in peanuts worldwide. When peanut is infected with southern blight disease, typically 5-20% yield loss was reported, however, up to 80% yield loss was recorded in the extreme situation (Cui et al. 2020). The pathogen S. rolfsii has numerous hosts, and it is difficult to eliminate it from infested soil (Karthikeyan et al. 2006). The infection of peanuts by S. rolfsii initiates from the lower stem parts and spread between plants when the environment is suitable (Sconyers et al. 2005). Also, the fungus can infect pegs and pods at harvest, resulting in yield loss. Peanut production is threatened by the disease every year, and prevention of disease in peanuts is a major concern. However, the management of soilborne disease increases input costs due to the difficulty of eliminating fungus from the soil (Thiessen and Woodward 2012). Utilization of fungicide or other agronomic chemicals is the major strategy to treat S. rolfsii infection, however, this method becomes costly due to the pathogen resistance and government restrictions on various chemicals (Thiessen and Woodward 2012). Currently, peanut researchers and breeders focus on developing peanut resistance to S. rolfsii. Until recently, although no peanut genotype has yet been reported completely resistant to S. rolfsii, some genotypes showed tolerance in the field (Dodia et al. 2019;Thirumalaisamy et al. 2014). To develop peanut resistance to S. rolfsii infection, a fundamental understanding of peanut response to this To understand the mechanism of peanut resistance to S. rolfsii, information on gene expression changes is required. Recently, the release of peanut whole genome sequences (Bertioli et al. 2016;Bertioli et al. 2019;Zhuang et al. 2019) facilitates the understanding of peanut resistance mechanisms in response to S. rolfsii. Transcriptome sequencing or RNA-seq is an effective technology, providing useful profiling of transcripts in better depth and width (Kukurba and Montgomery 2015). MicroRNAs are a subset of short (range between 20 and 24 nucleotides) non-coding RNAs that post-transcriptionally regulate the degradation of target mRNAs or repress gene translation (Bartel 2004). In plants, the miRNAs are highly conserved, and they were found to play important roles in development and stress responses (Reinhart et al. 2002;Jones-Rhoades et al. 2006;Chuck et al. 2009;Sun 2011). The functions of miRNAs in abiotic stress responses have been extensively studied. Many miRNAs are reported to be involved in plant stress responses. For example, miR398 is down-regulated in responding to high light, high Cu 2+ , ozone, and salinity stress (Sunkar et al. 2006;Jagadeeswaran et al. 2009). Overexpressing miR169 in tomato enhanced drought tolerance (Zhang et al. 2011). Besides abiotic stresses, miRNAs were found to play vital roles in plant innate immunity regulation (Padmanabhan et al. 2009). In the studies of plant resistance against pathogens, many miRNAs showed their importance in responding to stresses. For example, overexpression of miR160a and miR398b enhanced resistance to Magnaporthe oryzae in transgenic rice (Li et al. 2014). In tomatoes, miR482 and miR2118 were also reported to be involved in the NBS-LRR defense system (Shivaprasad et al. 2012).
However, to date, only a few studies about the basis of plant resistance to S. rolfsii are available. Wu et al. (2016) used RNA-seq transcriptome analysis to reveal genes involved in responding to S. rolfsii infection in Brassica. Jogi et al. (2016) also identified a set of genes involved in response to S. rolfsii in peanut. Bera et al. (2016) identified a major QTL qstga01.1 for resistance to stem rot disease in peanut. Recently, microRNAs involved in response to Aspergillus flavus growth in peanut seeds have been reported (Zhao et al. 2020) and study of the mechanisms of resistance to Sclerotium rolfsii in peanuts by using RNA-seq was also published (Bosamia et al. 2020). Benefiting from the new techniques and increasing information about the peanut genome, more studies on revealing the mechanisms in response to abiotic and biotic stresses have been reported.
In this study, we compared the miRNAs that respond to infection of S. rolfsii from the S. rolfsii susceptible peanut cultivar line Jitian-1 (JT) and tolerant line Rizhaohua-1 (Rzh). The different expression patterns of miRNAs associated with infection response from two varieties might uncover the mechanism of Southern blight response and tolerance.

Global Analysis of miRNA Sequencing
In this study, miRNAs of two peanut cultivars with or without S. rolfsii infection were sequenced. Stem samples were collected from pre-inoculation (JT_0 and Rzh_0), 36 h after inoculation (JT_A and Rzh_A, 36 HAI), 3 days after inoculation (JT_B and Rzh_B, 3 DAI) and 6 days after inoculation (JT_C and Rzh_C, 6 DAI) of two peanut cultivars (Fig. S1c). Based on the sequencing results from 3 replicates, 22.79 M clean reads were measured for each sample on average, and the mapping rate was above 87.67% on average ( Fig. 1a and b, Table S1). Interestingly, the mapping rate in the JT group displayed a significant decrease at the early stage of S. rolfsii infection (90% at the pre-inoculation, 74% at the 36 h of inoculation) and gradually increased to a level similar to that of JT_0 at a later stage (Fig. 1b). The length of miRNAs from all sequenced samples showed the highest abundance at 24 nt, then followed by 21 nt (Fig. 1c). In plants, 24-nt small interfering RNAs (siRNAs) are known to be related to RNA-directed DNA methylation (RdDM) (Matzke and Mosher 2014). The ratio of 24nt/21nt in both susceptible and tolerant peanuts might reflect the degree of DNA methylation during S. rolfsii infection. Notably, the ratio in the JT group increased along with infection time, however, Rzh, the resistance group showed a stable ratio during the infection period (Fig. 1d).
To identify miRNAs in each sample, total sRNAs were classified into 13 groups (Fig. S2). Results showed that sRNAs mapping to the intergenic region constituted the majority (49-66%); introns and exons made up 7-12%, and the miRNAs (mature) made up 3-8%. The rest part of the sRNAs including rRNAs, tRNAs, snRNAs, snoRNAs, unmapped and repeats, etc. filled the remaining classification. After alignment against miRBase, a total of 203 including 29 known and 174 novel miRNAs were detected from all samples (Table S2). To further identify miRNAs in samples, the average reads of miRNAs from 3 replicates were calculated, and those with less than 20 reads in all samples were removed. Finally, 18 known and 142 novel miRNAs were identified (Table S2).

Identification and Analysis of Differentially Expressed miRNAs (DEMs)
To understand how miRNAs respond to S. rolfsii infection, the fold changes in the expression of miRNAs under treatment were calculated both within and between JT and Rzh groups. Also, to understand the difference between JT and Rzh cultivar, the fold changes of miRNAs expression in JT and Rzh control groups (pre-inoculation, JT_0, and Rzh_0) were compared. According to the result of differentially expressed miRNA (|fold change|> 1.5) between the JT_0 and Rzh_0 (JT_0 was used as control), a total of 61 miRNAs (1 upregulated and 60 downregulated) were detected (Fig. 2a). Since the background of the JT and Rzh variety is different, the DEMs contained both common and Rzh unique fungi infection responsive miRNAs, however, the DEMs between Rzh and JT might uncover the secret of resistance to S. rolfsii infection in the Rzh variety. To further identify the miRNAs responding to S. rolfsii infection, the DEMs within the JT or the Rzh group were investigated as well. The result showed that the DEM was 36 (6 upregulated, 30 downregulated), 35 (5 upregulated, 30 downregulated), and 56 (8 upregulated, 48 downregulated) in 36 HAI, 3 DAI, and 6 DAI samples respectively (Fig. 2a). Interestingly, when we investigate the DEMs in the Rzh group, more upregulated miRNAs were detected while the downregulated miRNAs were the majority in the JT group. In the Rzh group, 23 (20 upregulated, 3 downregulated), 28 (21 upregulated, 7 downregulated), and 34 (19 upregulated, 15 downregulated) DEMs were found in 36 HAI, 3 DAI, and 6 DAI samples respectively (Fig. 2a).
Although DEMs in individual samples reflected the involvement of miRNAs in responding S. rolfsii infection in peanuts, differentially expressed miRNAs in both peanut varieties might have a higher correlation with defense response. Based on this hypothesis, we examined DEMs in the JT and the Rzh groups with different combinations to dig out S. rolfsii infection defense responding miRNAs. There were 29 and 12 miRNAs that showed overlaps during infection in the JT_A/B/C and the Rzh_A/B/C DEM respectively (Fig. 2b, c). Then we investigated miRNAs from JT_A/B/C vs JT_0, Rzh_0 vs JT_0, and Rzh_A/B/C vs Rzh_0. Subsequently, the control group DEM (Rzh_0 vs JT_0) was overlapped with JT and Rzh to identify predicted Rzh unique expressed miRNAs. Among the 61 DEMs, 14 DEMs were considered Rzh unique (Fig. 2d). 27 miRNAs that were differentially expressed in both JT and Rzh groups were considered S. rolfsii infection responsive (Fig. 2d). Total sRNA reads distribution in 8 samples after removing poorquality data. Numbers are in million, mean values were merged from 3 biological replicates. b). The percentage of total sRNAs mapping to peanut (Arachis hypogaea) genome for each sample. c). The length distribution of 18-26 nt small RNAs in 8 samples. All samples show the highest abundance at 24 nt and the second highest at 21 nt. d). The ratio of 24/21 nt sRNAs during infection in JT and Rzh genotype peanut. 3 biological replicates were merged into the mean value. JT genotype shows an increasing trend, and Rzh shows a stable pattern

Identification of miRNAs Involved in S. rolfsii Infection Resistance
Since the Rzh cultivar is resistant to S. rolfsii infection (Fig. S1c), the DEMs between the JT and the Rzh cultivars might be the miRNAs involved in establishing S. rolfsii resistance in the Rzh peanut variety. Here we analyzed those miRNAs differentially expressed in Rzh_0 compared to JT_0 samples. The expression data of 61 DEM in all samples were plotted, and an expression heatmap was shown ( Fig. 3a). According to the expression pattern, the heatmap was classified into four types (type I-IV). The type I (in red box) miRNAs showed low expression levels, the type II (in black box) contained the miRNAs with high expression levels in all samples, the type III (in gray box) was miRNAs with moderate-low expression and the type IV (in blue box) miRNAs have a moderate-high expression (Fig. 3a). Further analysis focused on identifying the miRNAs involved in S. rolfsii infection resistance. 14 Rzh unique DEMs were removed since they might be only related to Rzh background.  However, 18 DEMs that were identified from the cluster containing both Rzh_0 vs JT_0 and JT groups were considered S. rolfsii infection responsive and responsible for enhanced resistance in the Rzh variety. Notably, 12 out of 18 miRNA families showed low abundance (type I and III) in Rzh variety while they showed significant fold changes (Fig. 3b). According to the expression data (Table S2), these miRNAs were repressed by S. rolfsii infection. In addition, they were less expressed in the Rzh variety, therefore these miRNAs might regulate the genes that are involved in pathogen resistance establishment.

Identification of S. rolfsii Infection Responsive miRNAs
Besides the predicted resistance miRNAs, DEMs from both peanut varieties during infection were highly likely responsible for infection response. To identify these miRNAs, we investigated the DEMs in both JT and Rzh samples. Along the infection time, a total of 36, 35, and 56, and 23, 28, and 34 DEMs were identified at 36 h, 3 days and 6 days infected samples in JT and Rzh varieties respectively. The number of DEMs including known and novel miRNAs displayed an increasing trend (Fig. 4a). Interestingly, the JT, susceptible variety, has more affected miRNAs while the Rzh has only 2/3 of that in JT. This result might be due to the protection mechanism in the resistant variety. Among the DEMs, we identified 10 known miRNAs, and the expression fold changes of these miRNAs were shown (Fig. 4b). Most of the known miRNAs showed upregulated expression levels in the Rzh group even though they were downregulated in the JT group (Fig. 4b). Moreover, the maximum expression difference during the infection course was detected earlier in the Rzh variety than in the JT variety, indicating that the defense response activation in the Rzh variety is faster than in the JT variety, resulting in enhanced resistance to S. rolfsii infection.

KEGG Pathway Enrichment and miRNA Target Gene Prediction
According to the study of miRNA functions, miRNAs are considered post-transcriptional regulators of their messenger RNA (mRNA) targets via mRNA degradation and/or translational repression (Catalanotto et al. 2016). The expression change of miRNAs directly affects the target gene expression or function. To understand the possible biological function or mechanistic pathways of miRNAs that are responsible for S. rolfsii infection response in peanuts, the DEMs from JT and Rzh were used to predict the target genes. To get a comprehensive understanding of pathways involved in S. rolfsii infection response, KEGG pathway terms with predicted candidate gene numbers larger than 100 were selected and plotted (Fig. 5). The result showed that the immune response, MAPK signal, and phytohormone signal transduction were the main driving force to counter S. rolfsii infection in peanuts. Together with the previously identified S. rolfsii infection responsive miRNAs, target genes were further predicted with the online software psRNA target (https:// www. zhaol ab. org/ psRNA Target/ analy sis). A prediction was done to discover all 82 miRNA target genes (Table S3). Consistent with the KEGG pathway prediction, most miRNA target genes were related to stress sensing, defense activating, immune response, and phytohormone signal transduction (Table 1). In the Rzh control group, the expression of known miRNAs (ahy-miR167-3p, ahy-miR398, and ahy-miR408-3p) in peanut were found significantly repressed compared to that of in JT control, and they were identified target on the auxin response factor, las1-like family protein and blue copper protein-like respectively. Notably, in the Rzh group, novelahy-miR5-3p was highly repressed, though the expression was more stable in the JT group during infection. The target gene prediction of this miRNA went to chlorophyll synthase. This might result in higher chlorophyll content in the Rzh genotype. In addition, the novel-ahy-miR101-5p was highly repressed in Rzh groups compared to JT groups and the prediction showed it was responsible for regulating glycerophosphoryl diester phosphodiesterase (GDGP) which is related to plant cell wall organization. Together with the predicted higher chlorophyll content, these results indicated that the Rzh genotype has better resistance to fungal infection due to higher chloroplast and cell wall activities.

Discussion
To understand how peanut manages to optimize their response to biotic stress such as infection of S. rolfsii, susceptible and resistant peanut genotypes were selected to evaluate the miRNAs change during S. rolfsii infection. Both peanut materials were successfully inoculated and showed S. rolfsii susceptible and resistant phenotypes (Fig. S1) respectively. The miRNA from samples pre-and post-inoculation were sequenced. Then miRNA expression data was collected and analyzed in this study. By using the pre-inoculation samples (JT_0 and Rzh_0) as control, we compared the expression of miRNAs within and between the JT/Rzh groups. The differentially expressed miRNAs along the infection in JT and Rzh groups represented the infection responsive miR-NAs, then the comparison between JT_0 and Rzh_0 indicated the background differences. Adding them together, we analyzed the infection responsive miRNAs and predicted their target genes. Based on the prediction data, we made the following conclusion: 1) our S. rolfsii resistant genotype Rzh has a higher pre-activated miRNA activity that increased fungal infection resistance; 2) phytohormone transduction plays important role in fungal infection response.
miRNAs were considered to play important role in regulating mRNA post-transcriptionally (Bartel 2004), moreover, the RdDM, which involves 24 nt and 21 nt siRNAs Zhao et al. 2016), is also a key factor regulating gene expression. The crucial roles of DNA methylation in regulating gene expression and development in plants have been revealed recently (He et al. 2022). The changes of 24/21 nt in JT and Rzh samples might reflect the frequency of genomic DNA methylation during S. rolfsii infection, thus representing the miRNA activities in peanut plants under biotic stresses. In our study, the S. rolfsii susceptible and resistant peanut genotypes showed different patterns of a 24/21 nt ratio during infection (Fig. 1d). The increasing ratio of JT genotype during infection might be caused by higher responsive activities. However, the stable 24/21 nt ratio in Rzh genotype might be due to high adaption to biotic stresses.
Two studies about microbe interaction against S. rolfsii revealed that Induced Systemic Resistance (ISR) provided enhanced protection against S. rolfsii in peanut plants (Figueredo et al. 2017;Rodríguez et al. 2018). In addition, plants have developed a complex, multilayered defense system that is activated following pathogen perception and directs the establishment of immunity. Interestingly, in the Rzh genotype, we observed several repressed miRNAs that are responsible for establishing S. rolfsii resistance. Notably, most differentially expressed miRNAs in peanuts from this study were novel, however, the known miRNAs such as ahy-miR167, ahy-miR398, and ahy-miR408 were detected in this study as well. Studies have revealed that these known miRNAs are involved in abiotic stresses, yield, and photosynthesis (Hang et al. 2020;Ma et al. 2015;Zhang et al. 2017). Besides the known miRNAs, target genes were predicted among those novel miRNAs, and the cell wall organizationrelated target genes were detected. Glycerophosphoryl diester phosphodiesterase (Hayashi et al. 2008) and pectinesterase inhibitor (Liu et al. 2018;Jolie et al. 2010) were demonstrated associated with plant cell wall organization and disease resistance, the miRNAs which repress expression of these genes were detected highly repressed in Rzh samples. In addition, asparagine synthetase, receptor-like kinase, and disease resistance protein (TIR-NBS-LRR class) family were predicted from Rzh_0 miRNAs, which showed significant repression compared to the JT_0 group. These predicted gene families were proven associated with pathogen defense response (Hwang et al. 2011), cell wall thickening (Liang and Zhou 2018), and immune signaling (Huang et al. 2018). Moreover, when we examine the miRNA expression in peanuts during the infection, the transporters (e.g. MATE efflux family proteins) showed continuously active that extrude secondary metabolites and regulate phytohormone transport and plant development and senescence (Upadhyay et al. 2019).
Phytohormone plays an important role from sensing to activating downstream transcription factors in regulating plant development to adapt to adverse environmental conditions. In this study, a few critical stress-associated transcription factor families were predicted. Auxin, GA, and ethylene have been demonstrated to regulate plant response and tolerance to abiotic and biotic stresses in studies. The transcription factors, such as ERF, ARF, and SCL, were the target genes of affected miRNAs in this study. The ERF genes superfamily is a huge and important family in regulating abiotic and biotic stresses in plants. It is not surprising to detect such transcription factors in peanuts with S. rolfsii infection. The studies of overexpressing ERF-type transcription factors have conferred biotic and abiotic stress tolerance. For example, overexpression of GmERF113 (Zhao et al. 2017) and OsERF83 (Tezuka et al. 2019) genes, in response to ethylene mediated signal transduction pathway, were believed to be effective against pathogen infection. A study of ARFs in responding to bacterial and yellow curl leave virus (YCV) infection has shown most of the SIARFs were down-regulated in tomatoes (Bouzroud et al. 2018).
The SCL, a member of the GRAS family, is involved in GA mediated signaling pathway. It was shown to be associated with volatile terpene biosynthesis and glandular trichome development in tomatoes (Yang et al. 2021), chlorophyll biosynthesis under light conditions , and phytochrome A signal transduction (Torres-Galea et al. 2013).
In another study of MicroRNA roles in response to pathogen infection, the authors indicated that many miRNA target genes such as hypersensitive-induced response protein, leucine-rich repeat (LRR) receptor-like serine/threonine-protein kinase, GRAS, aquaporin, lipid transfer protein, ARF, MYB transcription factors, and MLP-like protein are involved in peanut disease resistance (Zhao et al. 2015). Here in this study, in addition, to uncovering the disease-responsive genes in peanuts, we also revealed a molecular mechanism established by pathogen-resistant peanuts. With higher cell wall activity, efficient metabolites transport, and delayed senescence, peanut combat diseases, and makes itself tolerant to pathogen infection (Fig. 6).

Plant Material and Disease Induction
To minimize the impact on seed quality, peanut seeds from Jitian-1 (JT) and Rizhaohua-1 (Rzh) with uniform size were selected and surface sterilized with 70% ethanol. Then peanut seeds were incubated in the growth chamber at 30 ℃ for 2 days to induce germination. After the emergence of the radicle, peanuts were transferred to a pot and placed in the growth room for 10 days with 50 μmol m −2 s −1 light intensity and 8/16 h light/dark cycle at 26 ℃.
The strain of Sclerotium rolfsii (BJ-1) was isolated from the Peanut Disease Nursery at Qingbaijiang, Sichuan Province of China. After isolation, the strain was inoculated on the potato dextrose agar (PDA) plate for 3 days to generate mycelium. Then the mycelium was inoculated on oats (Avena sativa L.) followed by 3 days of incubation at 28 ℃.
The infected oats with uniform size were selected and placed on the soil in contact with the base of the peanut Fig. 6 proposed mechanism summary of peanut response and tolerance to S. rolfsii infection at miRNA level. After pathogen infection occurs, the expression of miRNAs was altered therefore regulating target genes including cell wall-related genes, antioxidant enzymes, hormone-associated genes, and immune-related genes. In the tolerant genotype, biotic stress-associated genes were pre-regulated, thus resulting in a tolerance phenotype that builds vigor plants. In susceptible genotypes, responsive genes were triggered, thus resulting in a damaging phenotype that shows wilting plants. AS, asparagine synthetase; GDGP, glycerophosphoryl diester phosphodiesterase; ChlG, chlorophyll synthase; PDI, protein disulfide isomerase; SOD, superoxide dismutase; SAG, senescence-associated gene; ERF, ethyleneresponsive transcription factor; ARF, auxin response factor; SCL, scarecrow-like; DRP, disease resistance protein; MATE, MATE efflux family protein (Fig. S1a). Then peanut was grown in the growth room with 50 μmol m −2 s −1 light intensity, 16/8 h light/dark cycle, 75% humidity at 30 ℃ till sampling collection.
Five stems with size 0.5 cm each from peanut samples (0, 36 h, 3 days, and 6 days) were collected (Fig. S1b) and stored in liquid nitrogen for further use.

RNA Extraction and Sequencing
Total RNA of JT and Rzh (0, 36 h, 3 days, and 6 days) samples from above were extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the instructions. RNAs were purified with gel electrophoresis and the fragments of 18-30 nt were recovered. Then linkers were added to both 5'-and 3'-ends. The sRNA sequencing was conducted on DNBseq platform from BGI genomics (BGI, Shenzhen, China). A total of 24 samples were sequenced with 3 biological replicates for each treatment.

Analysis of Differentially Expressed miRNAs (DEMs)
Clean reads were obtained after the removal of low-quality reads, including reads with a length less than 18 nt, a poly-A sequence, a 5'-adaptor, no 3'-adaptor and no insert tag from raw reads. All sRNAs were aligned to the peanut genomic reference database (GCF_003086295.2_arahy.Tifrunner. gnm1.KYV3) by using Bowtie2 (Langmead and Salzberg 2012). The sRNAs were then aligned to a non-coding RNA database and Rfam using AASRA (Tang et al. 2021) and CMsearch (Cui et al. 2016) to annotate the rRNA, tRNA, scRNA, snRNA and snoRNA. miRNAs were predicted through miRbase and miRA (https:// github. com/ mhutt ner/ miRA). In addition, the clean reads for each miRNA were normalized using the TPM (transcript per million) formula: TPM = mapped reads/total reads * 10 6 , 3 biological replicates were merged into mean value. DEGseq (Wang et al. 2010) was used to analyze differentially expressed genes among sample groups with the value of log2(TPM1/TPM2). JT_0 and Rzh_0 samples were the control within the peanut variety, and the JT_0 was the control of DEG comparison between JT and Rzh varieties. The DEGs with fold changes larger than 1.5 were used in further analysis.

Prediction of Target Genes
Tapir and TargetFinder were used to predict miRNA target genes for general purposes. A further prediction for some important candidates was conducted using the software psRNA target (Dai et al. 2018, https:// www. zhaol ab. org/ psRNA Target/ analy sis). KEGG Pathway annotation classification on target genes of differential miRNAs was performed and enrichment analysis was conducted using phyper package in R software.