Background

Rice (Oryza sativa) is an important staple crop that serves as a primary food source for over half of the global population. However, the persistent threat of pathogen attacks, ranging from viruses, bacteria, to fungi, poses a significant challenge, leading to reduced yields and jeopardized food safety [1]. Extensive research conducted over the past three decades has shed light on the mechanisms underlying rice immunity in response to pathogen infections, resulting in the identification of numerous genes encoding resistance receptors, signaling molecules, and defense-related proteins [2,3,4]. Additionally, small RNAs, including microRNAs (miRNAs), have been characterized as playing essential roles in rice immunity [5,6,8]. Unlike mRNAs, lncRNAs tend to be less evolutionarily conserved, with lower but more pronounced tissue- and condition-specific expression patterns [9]. lncRNAs were thought to be by-products of transcriptional noise. However, more and more evidence has demonstrated that they play critical roles in diverse biological processes, exerting gene regulation at multiple levels, including histone modification, transcriptional regulation, and post-transcriptional regulation [10]. lncRNAs are often categorized into different subtypes based on their genomic locations in relation to protein-coding genes. Among these, long intergenic non-coding RNAs (lincRNAs) were the first discovered lncRNAs because of their widespread transcription in the regions that do not encode proteins [11].

Several well-characterized lncRNAs have been identified in plants, such as cold-assisted intronic noncoding RNA (COLDAIR), cold-induced long antisense intragenic RNA (COOLAIR), and long-day-specific male-fertility-associated RNA (LDMAR) [12,13,14]. COOLAIR and COLDAIR mediate the flowering process in Arabidopsis thaliana (A. thaliana), while LDMAR regulates photoperiod-sensitive male sterility in rice. Additionally, a leaf-expressed lncRNA named Xanthomonas oryzae pv. oryzae (Xoo) induced lncRNA 1 (ALEX1) has been found to enhance rice resistance to bacterial blight by activating the jasmonate (JA) signaling pathway [15]. In contrast to plants, numerous lncRNAs have been implicated in innate immunity in humans or mice, such as long intergenic noncoding RNA erythroid prosurvival (lincRNA-EPS), lnc13, and Mirt2 [16,17,18,19]. Despite the identification of numerous plant lncRNAs facilitated by RNA sequencing (RNA-Seq) techniques in recent decades, the functions of the majority of these lncRNAs remain elusive.

One of the primary roles of lncRNAs is their ability to regulate gene expression either in cis or trans. lncRNAs can regulate the expression of their neighboring protein-coding genes in cis through direct interactions with regulatory proteins or binding to specific motifs on the DNA sequence of the nearby target genes. A notable example is the lncRNA known as homeobox A (HOXA) transcript at the distal tip (HOTTIP). Positioned in proximity to the HOXA genes cluster, HOTTIP regulates the expression of HOXA genes in cis [20]. HOTTIP directly binds to the adaptor protein WD repeat-containing protein 5 (WDR5) and recruits the WDR5-myeloid/lymphoid or mixed-lineage leukemia (MLL) histone methyltransferase complexes to the HOXA locus. This recruitment leads to the deposition of histone H3 lysine 4 trimethylation (H3K4me3) on the promoter regions of HOXA genes, ultimately activating their transcription. Conversely, certain lncRNAs, such as TCF21 antisense RNA inducing demethylation (TARID), can form an RNA-DNA-DNA triplex (R-loop) with its nearby gene TCF21 on the promoter region [21]. Consequently, the growth arrest and DNA damage inducible-alpha protein (GADD45A) recognizes the R-loop and recruits the ten-eleven translocation 1 (TET1) protein to facilitate the demethylation of the DNA on the promoter region of TCF21, leading to increased transcription of TCF21.

In addition to their cis-regulatory functions, lncRNAs also regulate the expression of distant genes in trans through mechanisms such as chromatin interaction or post-transcriptional regulation. In A. thaliana, the lncRNA auxin-regulated promoter loop (APOLO) activates the expression of trans targets via an R-loop-induced regulation [22]. Upon auxin stimulation, the APOLO transcript is expressed and inserted into the promoter region of its trans targets through sequence complementarity, forming R-loops. These R-loops disrupt the binding of polycomb factors, such as heterochromatin protein 1 (LHP1), resulting in the deposition of the repressive mark histone H3 lysine 27 trimethylation (H3K27me3) and subsequent activation of the target genes. In addition to regulating gene expression through chromatin interaction, lncRNA plays a vital role in the post-transcriptional regulation of their trans targets. For instance, the lncRNA terminal differentiation-induced ncRNA (TINCR) can bind to the 25-nt ‘TINCR box’ motifs within target mRNAs, leading to the recruitment of Staufen homolog 1 (STAU1) and subsequent stabilization of the mRNAs [23].

MicroRNAs are typically ~ 21–22 nts long, single-strand RNAs that play important roles in diverse biological processes [24]. Through sequence complementarity, miRNAs can interact with target mRNAs, wherein the complementary sequences are referred to as miRNA response elements (MREs). Depending on the degree of complementarity between miRNAs and their target mRNAs, miRNAs can induce mRNA decay or translational repression of the target mRNAs [25]. lncRNAs containing MREs can function as competing endogenous RNAs (ceRNAs) or miRNA sponges, thereby reducing the availability of miRNAs for target mRNAs and regulating their expression. For instance, in tomato, lncRNA39026 acts as a ceRNA by sequestering miR168a, leading to the up-regulation of the resistance gene solanum lycopersicum argonaute1 (SIAGO1) and enhanced resistance against Phytophthora infestans infection [26].

Accumulating evidence has indicated that lncRNAs are involved in the response of rice to various pathogens. For instance, a transcriptome analysis on rice black-streaked dwarf virus (RBSDV) infected rice plants identified 1273 lncRNAs, with 22 showing differential expression in response to RBSDV infection [27]. In another study focused on rice lncRNAs during Magnaporthe oryzae (M. oryzae) infection, 161 lncRNAs were found to be differentially expressed following pathogen inoculation [28]. Furthermore, a comprehensive examination revealed 567 rice lncRNAs with distinct expression patterns in response to Xoo [15].

OsRpp30 is a subunit of Ribonuclease P (RNase P), an enzyme complex primarily responsible for the cleavage and maturation of tRNAs [29,30,31]. Our previous studies have demonstrated that OsRpp30 enhances disease resistance against M. oryzae and Xoo in rice [S1, Additional File 1). Overall, 948 highly confident novel lncRNA transcripts from 794 lncRNA gene loci were predicted in congruence by both LncDC [33] and Pfam [34] programs (Fig. 1b and Additional File 2).

The numbers of lncRNAs varied across different chromosomes, with the largest number (419) on chromosome 1 and the smallest number (171) on chromosome 9, likely due to the variation of chromosomal length (Fig. 1c). The numbers of intronic lncRNAs (code ‘i’) and sense lncRNAs (code ‘o’) were similar between the newly identified (novel) and reference lncRNAs. Among the intronic lncRNAs, 69 were newly identified and 64 were reference lncRNAs. Similarly, we found 316 novel and 238 reference sense lncRNAs. There were more long intergenic non-coding RNAs (lincRNA, classification code ‘u’) from the reference lncRNAs (737) compared to the novel lncRNAs (432). Additionally, we noticed that there were many more reference antisense lncRNAs (code ‘x’, 1024) than the novel antisense lncRNAs in our dataset (Fig. 1d) (131). In general, both the novel lncRNAs we identified and the reference lncRNAs were found in intergenic regions or overlap** with genes on the opposite strand.

We also examined the length distribution of different types of lncRNAs in rice (Fig. 1e). The average length of the novel lncRNAs we identified was 1261 nts, which was significantly longer than the average length of reference lncRNAs (902 nts, p-value = 1.16 × 10− 24, two-tailed Student’s t-test). This longer average length of novel lncRNAs is likely due to the fact that we removed the low-confidence single exon transcripts during the prediction process. Among the different categories of lncRNAs, intronic lncRNAs were the shortest on average, as they are located within introns and cannot exceed the length of the corresponding intron. In contrast, sense lncRNAs were generally longer than other types of lncRNAs, particularly among the predicted novel lncRNAs. Both lincRNAs and antisense lncRNAs had similar average lengths although they were shorter than sense lncRNAs.

Next, we examined whether these novel lncRNAs we identified were present in other rice lncRNA databases, such as NONCODE, CANTATAdb, and RiceLncPedia [35,36,37]. We found that 160, 341, and 353 lncRNAs showed alignment hits with identities of 90% or higher and alignment lengths of 100 nts or longer in NONCODE, CANTATAdb, and RiceLncPedia, respectively (Fig. 1f). These aligned lncRNAs shared conserved sequences with the lncRNAs found in the databases, suggesting that they likely belong to the same lncRNA family. We further focused on the predicted novel lncRNAs that not only had a 90% identity but also exhibited a high coverage rate when aligned with the lncRNAs from the databases. Specifically, we considered those lncRNAs where at least 90% of the sequences showed 90% or higher identity with the lncRNAs in the databases. In total, we found 48, 120, and 76 predicted novel lncRNAs that displayed high similarity to the lncRNAs in the NONCODE, CANTATAdb, and RiceLncPedia databases, respectively (Fig. 1g and Additional File 3).

Furthermore, we then observed that 47 of the predicted novel lncRNAs showed similarity to the lncRNAs present in both the NONCODE and CANTATAdb databases. Similarly, eight lncRNAs were confirmed by both RiceLncPedia and NONCODE, and 22 lncRNAs were confirmed by both RiceLncPedia and CANTATAdb. Remarkably, a total of eight predicted novel lncRNAs exhibited high similarity to lncRNAs in all three rice lncRNA databases. These eight lncRNAs are MSTRG.10686.3, MSTRG.12637.1, MSTRG.13742.1, MSTRG.16892.1, MSTRG.19604.1, MSTRG.24992.1, MSTRG.25244.1, and MSTRG.8373.1 (See Supplementary Table S2, Additional File 1).

Fig. 1
figure 1

Identification and characterization of rice lncRNAs. a The pipeline for the prediction of novel lncRNAs in rice. b The number of novel lncRNAs predicted by both Pfam and LncDC. c The number of predicted novel (novel) and annotated reference (ref) lncRNAs in different rice chromosomes. d The number of predicted novel and annotated reference lncRNAs from different categories. e The average length of different types of lncRNAs for all rice lncRNAs (all), predicted novel lncRNAs, and annotated reference lncRNAs. f The number of predicted novel lncRNAs with > = 90% identity and > = 100 nt alignment length against lncRNAs in various rice lncRNA databases. g The number of predicted novel lncRNAs with > = 90% identity and > = 90% length coverage rate against lncRNAs from different rice lncRNA databases

Full size image

Differential expression analysis reveals potential roles of lncRNAs in rice disease resistance mediated by OsRpp30

The RNA expression profiles of lncRNAs and mRNAs from different rice lines were analyzed using principal component analysis (PCA). The results revealed that the samples from the same rice line were clustered together, indicating that the replicates of the same genotype exhibited similar RNA expression profiles (See Supplementary Figure S1a, Additional File 1). We further compared the expression levels of lncRNAs among the samples from the three lines, WT, OsRpp30-KO (susceptible), and OsRpp30-OE (resistant). A total of 57 DElncRNAs, including 41 up-regulated and 16 down-regulated lncRNAs, were identified between OsRpp30-KO and WT plants (Fig. 2a and Additional File 4). To visualize the expression pattern of these DElncRNAs between OsRpp30-KO and WT samples, we normalized the expression of DElncRNAs using Fragments Per Kilobase of transcript per Million mapped reads (FPKM) and generated a heat map (Fig. 2b). Notably, the lncRNA MSTRG.10312.1 exhibited high expression in OsRpp30-KO samples (average FPKM of 17) but was less expressed in WT samples (average FPKM of 6). This sense lncRNA is located on the reverse strand of chromosome 2: 7,323,916-7,325,579 and partially overlaps with the protein-coding gene Os02g0230300 on the same strand.

In addition, we identified 29 up-regulated and 10 down-regulated lncRNAs between OsRpp30-OE and WT samples (Fig. 2c and Additional File 4). One interesting newly identified lncRNA, MSTRG.4676.1, was found to be a sense lncRNA located on the forward strand of chromosome 10: 11,079,933 − 11,087,253. MSTRG.4676.1 displayed minimal expression in WT samples but was highly expressed in OsRpp30-OE samples (average FPKM of 3.6) (Fig. 2d). Interestingly, MSTRG.4676.1 overlapped with the reference lncRNA transcript Os10t0360600-01, sharing similar exons, suggesting that MSTRG.4676.1 may represent a novel transcript isoform of Os10t0360600-01. Notably, although MSTRG.4676.1 displayed differential expression, no expression of Os10t0360600-01 was detected in any of the samples.

Next, we compared the expression of lncRNAs between OsRpp30-OE and OsRpp30-KO samples and identified 61 DElncRNAs, including 34 up-regulated and 27 down-regulated in OsRpp30-OE samples (Fig. 2e and f and Additional File 4). Of particular interest was Os03t0268000-03, a reference lncRNA derived from the protein phosphatase 1(OsPP1) gene, which also has two other protein-coding isoforms, Os03t0268000-01 and Os03t0268000-02. Os03t0268000-03 exhibited high expression in OsRpp30-OE samples (average FPKM of 4.2), while its expression was lower in OsRpp30-KO samples (average FPKM of 0.87). In addition, we discovered Os12t0136500-01, a reference lincRNA that displayed up-regulation in OsRpp30-KO samples with an average FPKM of 117.3. Despite being down-regulated in the OsRpp30-OE samples, it maintained a relatively high average FPKM of 20.75, indicating that it may play an important biological role in rice.

Fig. 2
figure 2

Differential expression of lncRNAs between individual rice lines. a The volcano plot of DElncRNAs between OsRpp30-KO and WT. Red dots represent up-regulated lncRNAs and blue dots represent down-regulated lncRNAs with statistical significance (|log2(FC)| > 1, p-adjusted < 0.05). b The heat map of DElncRNAs between OsRpp30-KO and WT. The color from blue to red represents the expression level from low to high. c The volcano plot of DElncRNAs between OsRpp30-OE and WT. d The heat map of DElncRNAs between OsRpp30-OE and WT. e The volcano plot of DElncRNAs between OsRpp30-OE and OsRpp30-KO. f The heat map of DElncRNAs between OsRpp30-OE and OsRpp30-KO.

Full size image

In addition to the pairwise comparisons between OsRpp30-KO and WT, between OsRpp30-OE and WT, and between OsRpp30-OE and OsRpp30-KO, we also investigated lncRNAs with specific differential expression patterns in each line. Given that OsRpp30-OE exhibited the highest resistance while OsRpp30-KO displayed the highest susceptibility, we focused on the lncRNAs that display differential expression when comparing OsRpp30-KO samples to both OsRpp30-OE and WT samples (KO vs. OE&WT), as well as OsRpp30-OE samples to both OsRpp30-KO and WT samples (OE vs. KO&WT).

The expression levels of DElncRNAs from the comparison between KO and OE&WT are shown in Fig. 3a. We identified 14 up-regulated lncRNAs in OsRpp30-KO samples but down-regulated in OE&WT samples, with the majority of them belonging to lincRNAs (9). Conversely, nine lncRNAs were down-regulated in OsRpp30-KO samples but up-regulated in OE&WT samples. Additionally, we examined the expression levels of DElncRNAs from the comparison between OE and KO&WT (Fig. 3b). The results showed that 11 lncRNAs were up-regulated in OsRpp30-OE samples, including five lincRNAs and six sense lncRNAs. Only six lncRNAs were down-regulated in OsRpp30-OE samples compared to KO&WT samples, with four being lincRNAs. Lastly, we extracted DElncRNAs differentially expressed in at least two different rice lines, specifically, OsRpp30-KO vs. WT, OsRpp30-OE vs. WT, and OsRpp30-OE vs. OsRpp30-KO. In total, we identified 91 DElncRNAs that are likely crucial for OsRpp30-mediated disease resistance in rice (Fig. 3c).

Notably, we discovered several DElncRNAs that bear resemblance to the disease-responsive lncRNAs documented in other studies. For instance, TCONS_00004581 is a rice lncRNA down-regulated in response to Xoo infection, indicating a negative association with rice immunity [15]. Interestingly, we observed that MSTRG.1326.1 overlapped with TCONS_00004581 on the same strand and shared identical start positions. Sequence alignment revealed that they possess a 100% identity with 78% coverage, suggesting that they are likely isoforms derived from the same lncRNA gene locus. Moreover, similar to TCONS_00004581, MSTRG.1326.1 was down-regulated in OsRpp30-OE samples, further supporting its negative relationship with rice immunity. Another Xoo-responsive lncRNA, TCONS_00490059, is known to be down-regulated after Xoo infection [15]. MSTRG.15145.2 overlapped with TCONS_00490059 with 99.56% identity and 95% coverage. Corresponding to the negative association of TCONS_00490059 with rice immunity, MSTRG.15145.2 showed down-regulation in OsRpp30-OE samples.

Fig. 3
figure 3

Differential expression of lncRNAs in all three rice lines. a The heat map of DElncRNAs between OsRpp30-KO and both OsRpp30-OE and WT (OE&WT) samples. b The heat map of DElncRNAs between OsRpp30-OE and both OsRpp30-KO and WT (KO&WT) samples. c The heat map of all DElncRNAs among all three lines without including the DElncRNAs between WT and both OsRpp30-OE and OsRpp30-KO (OE&KO) samples

Full size image

Differential expression of mRNAs reveals insights into OsRpp30-mediated disease resistance in rice

We next performed mRNA expression analysis to investigate the differential expression patterns associated with OsRpp30-mediated disease resistance. Between OsRpp30-KO and WT samples, we identified 670 up-regulated and 377 down-regulated mRNAs, as depicted in Fig. 4a and Additional File 5. Similarly, in comparison between OsRpp30-OE and WT samples, we found 571 DEmRNAs, with 400 mRNAs up-regulated and 171 mRNAs down-regulated (Fig. 4b and Additional File 5). Additionally, we identified 1018 DEmRNAs between OsRpp30-OE and OsRpp30-KO samples, consisting of 523 up-regulated and 495 down-regulated mRNAs (Fig. 4c and Additional File 5). In total, we identified 1671 DEmRNAs, which are likely to play crucial roles in OsRpp30-mediated disease resistance in rice.

Our analysis revealed a substantial number of DEmRNAs associated with plant immunity. Notably, several genes including Pita, HDT701, Os11g0644700, Os06g0707350, Os01g0350300, Os05g0143550, Os08g0543500, OsCrRLK1L14 and Oschib1 were up-regulated in OsRpp30-OE but down-regulated in OsRpp30-KO. Among these genes, Pita, Os06g0707350, and Os08g0543500 encode proteins belonging to the nucleotide-binding site leucine-rich repeat (NBS-LRR) class, which are crucial for recognizing pathogen effectors and activating general immune responses [4]. Pita, in particular, is a rice blast resistance gene against M. oryzae [38, 39]. Additionally, we have identified two more NBS-LRR protein-coding genes, Os11gRGA5 [40] and BPH26 [41], which exhibited elevated expression levels in OsRpp30-OE samples compared to WT samples. Moreover, Os11g0644700, Os01g0350300, and OsCrRLK1L14 encode plant disease resistance response proteins, while Os05g0143550 encodes a protein similar to the blast and wounding-induced mitogen-activated protein kinase. Oschib1 produces chitinase, a pathogenesis-related protein involved in rice defense against fungal pathogens. Interestingly, we observed the up-regulation of HDT701 in OsRpp30-OE samples compared to OsRpp30-KO. HDT701 is a histone H4 deacetylase, known as a negative regulator in rice innate immunity, and directly interacts with OsRpp30 [32, 42].

In our analysis, we observed reduced expressions of key resistance-related genes in OsRpp30-KO samples, such as OsCPS2, WAK112, Os09g0481600, and HAP2E, compared to WT samples. OsCPS2 is an ent- copalyl diphosphate (CPP) synthase essential for the biosynthesis of natural antibiotic products, phytoalexins. Studies have shown that overexpression of OsCPS2 enhances rice defense against both fungal and bacterial pathogens [43]. WAK112, a wall-associated kinase, is known to be involved in rice blast resistance [44]. HAP2E, a rice heme activator protein (HAP) gene, is associated with rice resistance to pathogens and abiotic stress [45].

In contrast, we observed significantly elevated expression levels of OsERF6, OsERF922, Os01g0112800, GH3-2, OsMPK20-1, Os11gRGA3, OsCrRLK1L16, and Os07g0130100 in OsRpp30-KO samples compared to WT samples. Of particular interest, OsERF922, an apetela2/ethylene response factor (AP2/ERF) type transcription factor, functions as a negative regulator of rice defense against blast pathogens [46].

Furthermore, OsPR1-101, Os10g0398100, GLP8-3, GLP8-4, Os03g0126700, Os05g0440100, Oschib2 and CHT2 were up-regulated in OsRpp30-KO samples but down-regulated in OsRpp30-OE samples. Notably, while Oschib1, Oschib2, and CHT2 are all chitinases, we observed contrasting expression profiles between Oschib1 and Oschib2/CHT2, suggesting potential functional divergence or regulation between these closely related genes involved in rice defense mechanisms.

Fig. 4
figure 4

Differential expression of mRNAs between individual rice lines. a The volcano plot of DEmRNAs between OsRpp30-KO and WT. Red dots represent up-regulated mRNAs and blue dots represent down-regulated mRNAs with statistical significance (|log2(FC)| > 1, p-adjusted < 0.05). b The volcano plot of DEmRNAs between OsRpp30-OE and WT. c The volcano plot of DEmRNAs between OsRpp30-OE and OsRpp30-KO. d The Gene ontology (GO) terms enriched for up-regulated mRNAs between OsRpp30-OE and OsRpp30-KO. e The GO terms and KEGG pathways enriched for down-regulated mRNAs between OsRpp30-OE and OsRpp30-KO.

Full size image

In light of our previous findings demonstrating the direct interaction between OsRpp30 and the rice histone deacetylase HDT701, we aimed to identify any differentially expressed mRNAs involved in histone modification in rice [32]. Notably, we found that some histone-modifying writers or erasers were differentially expressed, including HDT701, Os05g0440100, Os02g0180400, JMJ716, OsSET11, OsSET18, OsSET30, and SDG714. Among these genes, Os02g0180400 is a GCN5-related N-acetyltransferase (GNAT) domain-containing protein with similarity to histone H4 acetyltransferase. HDT701 and Os05g0440100 are histone deacetylases. Interestingly, our data indicated that HDT701 was up-regulated in OsRpp30-OE samples, while Os05g0440100 and Os02g0180400 were up-regulated in OsRpp30-KO samples. OsSET11, OsSET18, OsSET30, and SDG714 are su(var)3–9, enhancer of zeste, and trithorax (SET) domain-containing methyltransferases. In particular, OsSET11 showed significantly higher expression in OsRpp30-OE samples than in both WT and OsRpp30-KO samples, whereas OsSET18 exhibited elevated expression in OsRpp30-OE samples compared to WT samples. Conversely, OsSET30 displayed reduced expression levels in OsRpp30-KO samples compared to WT samples. Additionally, we identified JMJ716, which shares similarities with a jumonji C (JmjC) domain-containing protein exhibiting histone demethylase activity [47]. Both JMJ716 and SDG714 were up-regulated in OsRpp30-KO samples. Together, several histone modification-related protein-coding genes were differentially expressed in our samples, suggesting their potential roles in OsRpp30-mediated disease resistance in rice.

To gain insights into the putative functions of the DEmRNAs, we performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses. Many GO terms were enriched for the protein-coding genes up-regulated in OsRpp30-OE samples (OsRpp30-OE vs. OsRpp30-KO) (Fig. 4d). In terms of the biological process category, GO terms related to stimulus-response, phosphorus metabolic, and phosphate-containing compound metabolic process were enriched, suggesting a role of OsRpp30 in rice disease response. GO terms underlying molecular functions included transferase activity, nucleotide binding, ribonucleotide binding, organic/heterocyclic compound binding, ATP binding, ion binding, and catalytic activity were enriched, all of which are closely associated with plant immunity. On the other hand, a few GO terms and one KEGG pathway were enriched for the down-regulated mRNAs in OsRpp30-OE samples (OsRpp30-OE vs. OsRpp30-KO) (Fig. 4e). Notably, GO biological processes, such as the hydrogen peroxide metabolic process, cellular detoxification, cellular response to toxic substances, and reactive oxygen species metabolic process, as well as GO molecular functions, including oxidoreductase activity, peroxidase activity, tetrapyrrole binding, and oxidoreductase activity were enriched in these down-regulated mRNAs.

We also performed GO and KEGG pathway analyses for DEmRNAs between OsRpp30-OE and WT (See Supplementary Figures S2a and S2b, Additional File 1) and between OsRpp30-KO and WT samples (See Supplementary Figures S2c and S2d, Additional File 1). We found that some stress response-associated GO biological processes were enriched for the down-regulated mRNAs in OsRpp30-KO samples compared to WT samples (OsRpp30-KO vs. WT), including response to stress, response to stimulus, response to abiotic stimulus, terpenoid metabolic process, terpenoid biosynthetic process, diterpenoid biosynthetic process, and others.

Distinct miRNA expression profiles across OsRpp30-associated samples

To identify miRNAs involved in the regulation of differentially expressed lncRNAs and mRNAs in response to fungal and bacterial pathogens, we sequenced small RNAs from the same rice samples (See Supplementary Table S1, Additional File 1). The miRNA expression levels of the samples from the same rice lines were clustered together, indicating similar miRNA expression profiles among replicates of the same genotype (See Supplementary Figure S1b, Additional File 1). We identified a total of 10 up-regulated and 11 down-regulated miRNAs between OsRpp30-KO and WT samples (Fig. 5a and b, and Additional File 6). Notably, the osa-miR1428 family miRNAs were up-regulated in OsRpp30-KO samples. Interestingly, the lncRNA Os03t0611100-01, which serves as a precursor of miR1428e and miR1428d, was also up-regulated in OsRpp30-KO samples. This finding further supports the notion that Os03t0611100-01 is the precursor of the miR1428 family miRNAs. Previous research has demonstrated that osa-miR169 acts as a negative regulator in rice immunity against M. oryzae by inhibiting nuclear factor Y-A (NF-YA) genes [6]. Our data showed that osa-miR169i-5p.2 was up-regulated in OsRpp30-KO samples, suggesting that it plays a critical role in OsRpp30-mediated disease resistance in rice. Interestingly, miR1320 has been implicated in miR164-mediated immunity enhancement in rice [48]. However, in our study, we found the up-regulation of osa-miR1320-5p in OsRpp30-KO samples, while miR164 did not show differential expression. Additionally, miR398b has been identified as a positive regulator in rice defense against rice blast, leading to increased H2O2 levels after M. oryzae attack [7]. In our study, miR398b exhibited reduced expression in OsRpp30-KO samples compared to WT samples.

Only one up-regulated and 11 down-regulated miRNAs were identified between OsRpp30-OE and WT samples (Fig. 5c and d, and Additional File 6). Moreover, we identified 11 up-regulated and 15 down-regulated miRNAs between OsRpp30-OE and OsRpp30-KO samples (Fig. 5e and f, and Additional File 6). One of note is miR1846, which acts as a negative regulator of 1-aminocyclopropane-1-carboxylic acid oxidase (ACO), an enzyme involved in ethylene (ET) biosynthesis [49]. ET has a dual function in plant immunity, as it can either promote disease or enhance resistance [50]. In our study, we observed the up-regulation of osa-miR1846a-5p and osa-miR1846b-5p in OsRpp30-OE samples, while their mRNA target ACO4 was down-regulated. This finding suggests a potential mechanism for OsRpp30-mediated disease resistance in rice by suppressing ET synthesis. Previous studies have shown that miR528 negatively regulates rice resistance against viruses [51, 52]. In our study, osa-miR528-5p showed significantly higher expression in both OsRpp30-OE and WT samples compared to OsRpp30-KO samples. The miR171 family, including osa-miR171, is highly conserved and plays an essential role in stress responses, such as salt and drought tolerance, as well as pathogen resistance in rice [53,54,55]. In particular, the expression level of osa-miR171 in rice increases after sheath blight disease pathogen infection [56]. Our data revealed the down-regulation of osa-miR171e-5p and osa-miR171d-5p in OsRpp30-OE samples. Additionally, osa-miR166e-5p, osa-miR166k-3p, and osa-miR166l-3p were down-regulated in OsRpp30-OE samples. The osa-miR166 family miRNAs are known to participate in stress response in rice. For instance, under salt conditions and mineral deficiencies, osa-miR166k-3p and osa-miR166l-3p are down-regulated in both panicles and shoots [57]. Moreover, osa-miR166 family miRNAs have been implicated in rice resistance against fungal and viral pathogens, including rice blast fungus, rice dwarf virus, and rice stripe virus [58, 59].

Next, we examined the expression levels of miRNAs between OsRpp30-KO and OE&WT samples. Strikingly, we did not find any up-regulated miRNAs in OsRpp30-KO samples but identified six down-regulated miRNAs. Similarly, between OsRpp30-OE and KO&WT samples, we identified ten down-regulated miRNAs, while no miRNAs were up-regulated. Finally, we identified a total of 41 DEmiRNAs that may play important roles in OsRpp30-mediated disease resistance in rice.

Fig. 5
figure 5

Differential expression of miRNAs between individual rice lines. a The volcano plot of DEmiRNAs between OsRpp30-KO and WT. Red dots represent up-regulated miRNAs and blue dots represent down-regulated miRNAs with statistical significance (|log2(FC)| > 1, p-adjusted < 0.05). b The heat map of DEmiRNAs between OsRpp30-KO and WT. The color from blue to red represents the expression level from low to high. c The volcano plot of DEmiRNAs between OsRpp30-OE and WT. d The heat map of DEmiRNAs between OsRpp30-OE and WT. e The volcano plot of DEmiRNAs between OsRpp30-OE and OsRpp30-KO. f The heat map of DEmiRNAs between OsRpp30-OE and OsRpp30-KO.

Full size image

lncRNAs regulate disease resistance-related genes in trans

To shed light on the potential regulation of lncRNAs on protein-coding genes, we investigated both trans and cis targets of DElncRNAs. lncRNAs can modulate the expression of distant genes through trans mechanisms, involving chromatin remodeling or post-transcriptional regulation [10]. For trans targets, we calculated Pearson’s correlation coefficients between DElncRNAs and DEmRNAs, using a similar approach as described in Zhang et al. [27]. Given that lncRNAs can regulate their trans targets through various mechanisms, including direct lncRNA-mRNA binding through sequence complementarity, we utilized LncTar [60] to predict putative sequencing binding events between DElncRNAs and their trans targets. In total, we identified ten DElncRNA-DEmRNA co-expression pairs between OsRpp30-KO and WT samples, three pairs between OsRpp30-OE and WT samples, and ten pairs between OsRpp30-OE and OsRpp30-KO samples. Given that the OsRpp30-OE and OsRpp30-KO rice lines exhibited the most distinct phenotypes in terms of resistance and susceptibility, our primary focus was on the trans-targeting pairs between these two rice lines.

To further explore the relationship between DElncRNAs and their trans targets, we constructed a co-expression network. For instance, we found that the lincRNA MSTRG.5751.1 and its three mRNA targets, Os01t0113450-00, Os07t0153150-02, and Os03t0835600-01 were down-regulated in OsRpp30-OE samples (Fig. 6a and Supplementary Table S3, Additional File 1). LncTar analysis confirmed the direct RNA-RNA interaction between MSTRG.5751.1 and Os01t0113450-00 (See Supplementary Figure S3a, Additional File 1), as well as MSTRG.5751.1 and Os03t0835600-01 (See Supplementary Figure S3b, Additional File 1), suggesting potential regulation through sequence complementarity. It is worth noting that Os03t0835600-01 is a transcript of the gene OsACBP6, which encodes the Acyl-CoA-binding protein 6. OsACBP6 is involved in various biological processes, including peroxisomal beta-oxidation and lipid metabolism. Interestingly, a previous study has indicated that OsACBP6 mutants exhibit up-regulated levels of JA, a product of beta-oxidation in peroxisomes, and JA is involved in regulating plant responses to abiotic and biotic stresses [61]. In our analysis, we found that MSTRG.5751.1, MSTRG.4506.1, and Os05t0475550-00 act as trans-regulators of OsACBP6 (See Supplementary Figure S3c, Additional File 1), suggesting the potential involvement of OsACBP6 as an essential lncRNA trans target in OsRpp30-mediated disease resistance in rice.

Fig. 6
figure 6

Analysis of DElncRNA trans and cis targets. a The co-expression network of DElncRNAs and their trans DEmRNA targets between OsRpp30-OE and OsRpp30-KO. b The network of DElncRNAs and their cis DEmRNA targets between OsRpp30-OE and OsRpp30-KO. c The GO terms enriched for all DElncRNA targets

Full size image

The sense lncRNA MSTRG.7147.1 exhibited up-regulation in OsRpp30-OE samples, along with its three trans mRNA targets, Os02t0185633-00, Os08t0543900-00, and Os02t0569400-01. Notably, all three trans targets were predicted to have direct RNA-RNA interactions with MSTRG.7147.1 (See Supplementary Figures S3d, S3e and S3f, Additional File 1). It is worth noting that MSTRG.7147.1 and its trans targets were exclusively expressed in OsRpp30-OE samples, with no detectable expression in WT or OsRpp30-KO samples. Specifically, Os08t0543900-00 is a transcript of the gene OsbZIP68, a member of the basic leucine zipper (bZIP) transcription factor family. bZIP68 is highly conserved among plants and plays a role in various stress response-associated biological processes, including cold and oxidative stress tolerance [62, 63]. On the other hand, Os02t0569400-01 corresponds to a transcript of the gene CYP76M8, which encodes a cytochrome P450 (CYP) family protein. CYP76M8 is involved in the biosynthesis of oryzalexin, an antimicrobial compound produced by rice in response to pathogen attacks [64]. To further investigate the interaction between DElncRNAs and their trans targets, we performed sequence alignments and identified a conserved ‘GCG’ tandem repeat motif shared between MSTRG.7147.1 and the trans targets Os02t0185633-00 and Os08t0543900-00, suggesting a potential interaction with the same motif binding protein (See Supplementary Figures S4a and S4b, Additional File 1).

lncRNAs regulate disease resistance-related genes in cis

In addition to their role in regulating trans targets, lncRNAs also exert influence on nearby genes, referred to as cis targets, by recruiting regulatory proteins to their promoters or by forming R-loops through direct DNA-RNA binding [10]. To identify cis targets of DElncRNAs, we adopted a similar approach previously described [15, 27]. By extracting protein-coding genes located within a 100 kb region up and downstream of DElncRNAs, we identified a total of 919 cis targets for all DElncRNAs, with 42 exhibiting differential expression in at least two different rice lines. Specifically, we identified 27 DElncRNA-DEmRNA cis-targeting pairs between OsRpp30-OE and OsRpp30-KO samples (Fig. 6b and Supplementary Table S4, Additional File 1).

Among the cis targets of the DElncRNAs, we identified several genes associated with plant immunity, including HIPP49, Os10t0370800-01, Os06t0129900-01, and Os10t0361000-01. Of particular interest, the reference lncRNA Os01t0678900-01 was found to cis targeting the heavy metal-associated isoprenylated protein 49 (HIPP49). HIPPs are known to participate in plant responses to heavy metal detoxification and abiotic stress, but certain HIPPs can act as susceptibility factors, negatively impacting disease resistance against pathogens [65]. In OsRpp30-OE samples, both Os01t0678900-01 and its cis target HIPP49 were downregulated, suggesting a potential mechanism of OsRpp30-mediated disease resistance in rice involving the negative regulation of HIPP49 by Os01t0678900-01. Furthermore, our analysis revealed that Os01t0678900-01 is located upstream (~ 3 kb) of the HIPP49 gene, indicating a plausible gene regulatory role of Os01t0678900-01 on the HIPP49 promoter region (See Supplementary Figure S5a, Additional File 1).

In addition, we observed that both MSTRG.4760.1 and MSTRG.4754.1 targeted Os10t0370800-01, a transcript derived from the gene Os10g0370800. This gene encodes a protein that shares similarities with an exo-1,3-beta-glucanase precursor, known for its ability to degrade the cell walls of various pathogens, including bacteria, viruses, and particularly fungi, thereby playing a crucial role in plant resistance against pathogens [66]. Interestingly, we observed a similarity of 26-nt fragment sequence between MSTRG.4754.1 and the second intron of Os10g0370800 (See Supplementary Figure S4d, Additional File 1). Moreover, the reference lncRNA Os06t0128200-03 demonstrated specific targeting of the mRNA Os06t0129900-01, which encodes a protein resembling a CYP family protein. CYPs are known to play vital roles in hormone signaling pathways and regulate plant response to both biotic and abiotic stresses [67].

Another interesting finding was the targeting of Os10t0361000-01, a lipoxygenase gene, by MSTRG.4676.1. Lipoxygenases are enzymes that catalyze polyunsaturated fatty acids into hydroperoxides, ultimately leading to the production of stress response-associated molecules such as JA [68]. A study in maize has demonstrated that the disruption of 9-lipoxygenase results in enhanced resistance against fungal pathogens [69]. Interestingly, a sequence alignment between MSTRG.4676.1 and Os10t0361000-01 revealed a shared short sequence fragment, ‘GCCGCCGGCCACGA,’ suggesting the presence of a potential binding site for the regulatory proteins (See Supplementary Figure S4e, Additional File 1).

In addition to Os01t0678900-01, we identified several DElncRNAs that are located in the promoter regions of their respective cis targets, including MSTRG.10312.1 and MSTRG.5751.1. Interestingly, MSTRG.10312.1 not only occupies the promoter region of its cis target Os02t0230300-01 but also functions as a sense lncRNA that partially overlaps with it (See Supplementary Figure S5b, Additional File 1). Conversely, MSTRG.5751.1, classified as a lincRNA, is positioned within the promoter region of Os10t0567900-01 (See Supplementary Figure S5c, Additional File 1).

To investigate the functions of DElncRNA targets, we performed a GO and KEGG pathway enrichment analysis for both trans and cis targets. Interestingly, we did not find any significantly enriched GO or KEGG terms among the targets of the down-regulated lncRNAs in the OsRpp30-OE samples. In contrast, we observed significant enrichment of several GO terms among the targets of up-regulated lncRNAs in the OsRpp30-OE samples (Fig. 6c). Notably, one of the enriched GO terms was lipid droplet, representing specialized organelles for lipid storage, primarily in the form of triacylglycerols. Proteins associated with lipid droplets, such as caleosins and steroleosins are involved in plant responses to abiotic and biotic stresses [70]. Additionally, four GO biological processes were enriched, including response to arsenic-containing substance, response to ethanol, triterpenoid biosynthetic, and triterpenoid metabolic processes. Triterpenoids are natural plant products that contribute to plant defense against pathogens and herbivores [71]. Finally, in terms of GO molecular function, we found enrichment in a single term: translation elongation factor activity.

lncRNAs regulate Disease resistance-related genes through a lncRNA-miRNA-mRNA competing endogenous RNA (ceRNA) network

In plants, miRNAs can target both lncRNAs and mRNAs, reducing their expression through RNA transcript cleavage. Additionally, lncRNAs can act as sponges, competing with mRNAs for MREs, thereby sequestering miRNAs and indirectly modulating the expression of mRNAs. To gain deeper insights into how lncRNAs orchestrate the regulation of mRNAs through this ceRNA mechanism, we constructed a ceRNA network that integrates the DEmiRNAs and their respective differentially expressed targets from the OsRpp30-OE and OsRpp30-KO samples (Fig. 7 and Supplementary Table S5, Additional File 1). The representative predicted DEmiRNA-DElncRNA and DEmiRNA-DEmRNA targeting pairs were shown in Supplementary Tables S6 and S7 of Additional File 1, respectively.

Among the up-regulated lncRNAs in OsRpp30-OE samples, MSTRG.24461.2 emerged as a target for several down-regulated miRNAs, including members of the osa-miR166 family (osa-miR166k-3p and osa-miR166l-3p) and osa-miR1882 family (osa-miR1882a, osa-miR1882b, osa-miR1882c, osa-miR1882d, osa-miR1882e-5p, osa-miR1882f, osa-miR1882g, and osa-miR1882h). Previous studies have demonstrated that osa-miR166 is involved in rice defense against fungi and viruses [58, 59]. In addition to MSTRG.24461.2, osa-miR166 and osa-miR1882 also targeted several OsRpp30-OE up-regulated mRNAs within the ceRNA network, such as Pita, Os03t0170200-01, and OsCSLA3. Both Pita and MSTRG.24461.2 were up-regulated in the OsRpp30-OE samples, competing for the limited number of osa-miR166 transcripts. On the contrary, Os03t0170200-01 and OsCSLA3 competed with MSTRG.24461.2 for osa-miR1882. Os03t0170200-01 is a transcript of the gene Os03g0170200, which shares similarities with the MADS-box transcription factor 30. MADS-box transcription factors are involved in different development processes and stress responses in plants [72]. For instance, the knockdown of OsMADS26 in rice plants enhances resistance against rice blast and bacterial blight pathogens [73]. OsCSLA3 is a member of cellulose synthases, which are crucial proteins involved in cell wall formation and plant disease resistance. Disruption of cellulose synthases A3 in A. thaliana has been shown to enhance pathogen resistance by activating JA and ET signaling pathways [74]. Furthermore, inhibiting cellulose synthases necessary for secondary cell wall integrity increases pathogen resistance in A. thaliana through the abscisic acid (ABA) pathway [75].

MSTRG.16012.6, an up-regulated lncRNA in the OsRpp30-OE samples, was found to be targeted by the down-regulated miRNA, osa-miR171d-5p. In the ceRNA network, several up-regulated mRNAs competed with MSTRG.16012.6 for osa-miR171d-5p. For instance, Os02t0783700-01 is a transcript of the bifunctional gene lysine ketoglutarate reductase/saccharopine dehydrogenase (OsLKR/SDH). LKR/SDH serves as a key enzyme in the saccharopine pathway for lysine catabolism, which is induced in response to abiotic or biotic stresses [76, 77]. Another example is OsTPS1, which encodes a trehalose 6-phosphate synthase responsible for the biosynthesis of trehalose. Trehalose plays a crucial role in abiotic stress response in plants, and studies have shown that overexpression of OsTPS1 increases trehalose concentration and enhances rice tolerance to cold, salt, and drought stress [78]. In addition, trehalose contributes to resistance against bacterial wilt disease in tomatoes [79].

Fig. 7
figure 7

lncRNA-miRNA-mRNA competing endogenous RNA (ceRNA) network between OsRpp30-OE and OsRpp30-KO. RNA transcripts that were up- and down-regulated in OsRpp30-OE samples are in red and blue colors, respectively. lncRNA, miRNA, and mRNA are represented by triangle, diamond, and ellipse, respectively

Full size image

MSTRG.17424.2 was identified as a target of the up-regulated osa-miR1846 family miRNAs, osa-miR1846a-5p and osa-miR1846b-5p. Our study revealed that several down-regulated mRNAs competed with MSTRG.17424.2 for osa-miR1846, many of which were associated with plant immunity. One of the targets, OsMATE2, encodes a multi-antimicrobial extrusion protein known to negatively regulate plant immunity in response to bacterial pathogens in A. thaliana [80]. The down-regulation of OsMATE2 in the OsRpp30-OE samples suggests its potential role in OsRpp30-mediated disease resistance. Another target, Os04t0107650-00 is a transcript of the gene Os04g0107650, which shares similarities with arginine decarboxylase (ADC). ADC plays a vital role in plant defense against bacterial pathogens, as evidenced by the up-regulation of salicylic acid (SA)- and JA-dependent pathogenesis-related genes in A. thaliana with down-regulated ADC2 [81]. OsNAC15 is a NAC (NAM, ATAF1/2, and CUC2) domain-containing protein-coding gene, belonging to a large transcription factor family that plays critical roles in plant stress response, including biotic infections and abiotic stresses [82]. Previous studies have demonstrated that overexpression of the NAC family gene OsNAC6 enhances rice resistance against blast disease [83]. Although the function of OsNAC15 in biotic stress response is currently not well understood, a recent study showed that OsNAC15 contributes to abiotic stress response in rice, such as zinc deficiency and cadmium stress [84]. EUI1 (elongated uppermost internode 1) is a Gibberellin (GA) deactivating enzyme and it is involved in rice defense against bacterial blight [85]. Notably, miR1846 negatively regulates ACO proteins, which regulate the biosynthesis of ET, thereby affecting the plant defense-associated ET signaling pathway [49]. The consistent down-regulation of ACO4 by osa-miR1846 suggests a putative regulatory network involving ET in OsRpp30-mediated disease resistance in rice.

Discussion

RNase P is a conserved ribonucleoprotein complex found in all three life kingdoms, with its primary function being the cleavage of the 5’ end of precursor tRNAs and the maturation of tRNAs [29,30,31]. It is composed of an RNA molecule and several protein subunits, ranging from one in bacteria to nine to ten in eukaryotes [86]. However, except for a few exceptional cases, such as human mitochondrial RNase P, only the RNA component is required for catalytic activity, while the protein subunits primarily serve auxiliary roles such as tRNA recognition and proper folding of the catalytic RNA [31, 87]. In addition to their role in precursor tRNA cleavage, RNase P protein subunits are involved in various biological processes, including chromatin assembly, transcriptional regulation, and DNA damage repair [31, 88, 89]. For instance, human Rpp29 and Rpp21 proteins are implicated in DNA double-strand break repair, whereas other RNase P subunits, such as Rpp14, Rpp25, and Rpp38, do not respond to DNA breaks [89]. These findings suggest that the protein subunits of RNase P may have functions beyond tRNA maturation. Indeed, as a conserved RNase P protein subunit, Rpp30 has been found to play essential roles, and dysregulation of Rpp30 has been linked to diseases, such as female sterility in Drosophila and cancers in humans [90,91,92]. Our previous study also indicated that the rice OsRpp30 positively regulates plant immunity, and the histone deacetylase HDT701 may serve as an upstream regulator of OsRpp30 [103]. (3) Many of the identified lncRNAs exhibited high similarity to lncRNAs present in existing databases, and some of the DElncRNAs we discovered were previously reported as disease-responsive lncRNAs in rice [15, 104]. Despite the absence of wet-lab validation, these factors collectively enhance the reliability and validity of our identified lncRNAs.

Conclusions

This study represents a comprehensive investigation of the expression profiles of lncRNAs, miRNAs, and mRNAs in a collection of defense mutants in rice, shedding new light on their roles in OsRpp30-mediated disease resistance. The results underscore the significance of lncRNAs as crucial regulators in this context, as evidenced by their differential expression patterns across OsRpp30-OE, OsRpp30-KO, and WT rice samples. These DElncRNAs exert regulatory control over DEmRNAs that participate in pathogen recognition, hormone pathways, and other downstream responses in rice immunity by targeting them through cis, trans, or ceRNA networks. We anticipate that these findings will enhance our understanding of the functions of OsRpp30 and provide candidate lncRNA-miRNA-mRNA interaction pairs for future research aiming to validate the functional roles of these lncRNAs.

Methods

Plant materials and growth conditions

The generation of the OsRpp30-OE and OsRpp30-KO transgenic lines was described previously [107]. Transcript assembly was performed using StringTie v2.1.2 [108]. To ensure that the assembled transcripts were expressed in our rice samples, we removed transcripts with FPKM values less than 0.1. Additionally, to minimize false positives produced by StringTie for single-exon transcripts, we retained only transcripts with a minimum of two exons for further analysis [103]. In addition, given that lncRNAs are defined as RNAs longer than 200 nts, we excluded transcripts shorter than 200 nts from the analysis.

The selected transcripts were compared with known rice gene annotations obtained from RAP-DB using Cuffcompare [98, 109]. Transcripts with Cuffcompare generated classification codes ‘i’ (intronic transcript), ‘o’ (transcript overlapped with an exon of a gene), ‘x’ (transcript overlapped with an exon of a gene on the opposite strand), and ‘u’ (intergenic transcript) were selected as lncRNA candidates. These lncRNA candidates were then subject to prediction using both LncDC v1.3.3 and Pfam v35.0 [33, 34]. Only candidates predicted as lncRNAs by both programs were considered confident lncRNAs. BLAST v2.9.0 [110] was used to compare the lncRNAs against existing lncRNA databases, including NONCODE v6, CANTATAdb v2, and RiceLncPedia [35,36,37]. To maintain consistency in naming, the predicted lncRNAs were assigned the prefix ‘MSTRG’. Additionally, prefixes, such as ‘Os’, ‘NONOSA’, ‘CNT’, ‘Osa’, and ‘TCONS’ were used for lncRNAs from RAP-DB, NONCODE, CANTATAdb, RiceLncPedia, and the study by Yu et al. [15], respectively.

In addition to the predicted novel lncRNAs, we extracted 2,063 annotated reference lncRNAs obtained from RAP-DB [98] (biotype = nontranslating_CDS, length > = 200 nts) for further analyses.

Identification of miRNAs

The raw small RNA-Seq reads were subject to quality control using FastQC [105], followed by removing adapters, low-quality ends, reads containing poly-N bases, and reads shorter than 18 nts or greater than 30 nts. Subsequently, Bowtie v1.3.1 [111] was used to map the clean reads to the same rice reference genome IRGSP-1.0 [106]. The mapped reads were compared to known mature and precursor miRNAs from miRbase v22.1 [112] by miRDeep2 v0.1.3 [113]. miRDeep2 was also used for the quantification of miRNA expression levels in different samples.

Differential expression analysis of lncRNAs, mRNAs, and miRNAs

Stringtie v2.1.2 [108] was used to quantify the transcript abundance of lncRNAs and mRNAs in samples from the three rice lines. On the other hand, miRDeep2 v0.1.3 [113] was used to quantify the expression of miRNAs. The read counts data was then subject to DESeq2 [114] to identify DElncRNAs, DEmRNAs, and DEmiRNAs among the WT, OsRpp30-OE, and OsRpp30-KO samples. Differential expressions were determined based on the criteria of |log2(Fold Change)| > 1 and p-adjusted value < 0.05. PCA analysis and expression plots were generated using DESeq2. Volcano plots for DElncRNAs, DEmRNAs, and DEmiRNAs were created using ggplot2 [115]. Additionally, FPKM values were calculated for DElncRNAs across different rice lines and heatmaps were generated using pheatmap (https://cran.r-project.org/web/packages/pheatmap/index.html).

Target prediction of DElncRNAs and construction of lncRNA-mRNA interaction network

We performed predictions for both trans and cis mRNA targets of the DElncRNAs. The correlation between DElncRNAs and DEmRNAs across all six samples was assessed by calculating Pearson’s correlation coefficients using the R package psych [116]. Trans-targeting pairs were selected based on |R| > 0.9 and p-adjust value < 0.05. We employed LncTar to predict RNA-RNA interactions between DElncRNAs and their targeted DEmRNAs, with a criterion of normalized delta G (ndG) < -0.1 [60].

Additionally, we selected adjacent protein-coding genes within a 100 kb region upstream and downstream of DElncRNAs as their cis mRNA targets. The resulting interaction network, which encompasses DElncRNAs, their trans-targeted DEmRNAs, and cis-targeted DEmRNAs was visualized using Cytoscape [117].

Enrichment analysis

We conducted GO [118] and KEGG [119] pathway enrichment analyses for DEmRNAs and the targets of DElncRNAs using g:Profiler [120]. Three GO categories were considered in the GO enrichment analysis: molecular function, cellular component, and biological process. The GO terms with an adjusted p-value < 0.05 were considered significantly enriched terms.

Target prediction of DEmiRNAs and construction of lncRNA-miRNA-mRNA ceRNA network

We used psRNATarget and RNAhybrid (< -20 kcl/mol) to predict the interactions between DEmiRNAs and DElncRNAs as well as DEmiRNAs and DEmRNAs [121, 122]. Only the targeting pairs concordantly predicted by both programs were selected for constructing the lncRNA-miRNA-mRNA ceRNA network. The constructed network was visualized using Cytoscape (https://cytoscape.org/).