A Comprehensive Interaction Network Constructed Using miRNAs and mRNAs Provides New Insights into Potato Tuberization under High Temperatures

High temperatures delay tuberization and decrease potato (Solanum tuberosum L.) yields. However, the molecular mechanisms and regulatory networks underlying tuberization under high temperatures remain largely unknown. Here, we performed the mRNA and miRNA sequencing of leaves and stems to identify genes and regulatory networks involved in tuberization under high temperatures. A total of 2804 and 5001 differentially expressed genes (DEGs) under high-temperature stress were identified in leaves and stems, respectively. These genes were significantly enriched in gene ontology terms regarding meristem development, the sucrose biosynthetic process, and response to heat. Meanwhile, 101 and 75 differentially expressed miRNAs (DEmiRNAs) were identified in leaves and stems, respectively. We constructed an interaction network between DEmiRNAs and DEGs, identifying 118 and 150 DEmiRNA–DEG pairs in leaves and stems, respectively. We found three miRNA–mRNA candidate modules involved in tuberization under high temperatures, including stu-miR8030-5p/StCPY714, stu-miR7981f-p5/StAGL8a, and stu-miR10532A/StAGL8b. Our study constructed an interaction network between miRNAs and target genes and proposes candidate miRNA–gene modules that regulate tuber formation under high temperatures. Our study provides new insights for revealing the regulatory mechanism of the high-temperature inhibition of tuberization and also provides gene resources for improving the heat tolerance in potatoes.


Introduction
Potato (Solanum tuberosum L.) is a critical tuber crop worldwide and is rich in nutrients such as starch, vitamins, and protein.Temperature is one of the most important environmental factors affecting tuber development in potatoes.The 14-22 • C range suits tuber initiation [1].High temperatures delay or even impede tuber development, particularly tuber formation, causing a decline in tuber yield [2][3][4][5].
A recent study proposed that post-transcriptional regulation plays a significant role in the early stages of tuberization [5].In miR156-overexpressed potato plants, which present increased aerial tubers, StBRC1b expression was inhibited.In contrast, the loss of StBRC1b function leads to an upregulated miR156 expression, indicating that the inhibition of aerial tuberization might be regulated through the negative regulation of the StBRC1b expression by miR156 [12].At elevated temperatures, the expressions of StSP5G and StCOL1 are upregulated, suppressing the expression of StSP6A [5,16,17].miRNA suppressing the expression of SP6A (SES), which targets StSP6A, was induced during high-temperature stress.Interfering with the function of SES promotes tuberization under high temperatures [4].miR172 is a stimulating tuberization factor [18,19]. miR156-resistant SQUAMOSA PROMOTER BINDING-LIKE (StSPL9) overexpression lines increase miR172 abundance, suggesting a miR156/StSPL9/miR172 regulatory module in tuberization [19].Therefore, post-transcriptional regulation also plays an important role during tuber formation in addition to transcriptional regulation.
miRNAs play essential roles in regulating the protein accumulation of target transcripts during high-temperature stress in plants.A large number of high-temperature-related miRNAs have been reported, such as miR156, miR172, and miR824 [20][21][22][23].In alfalfa, the overexpression of miR156 improves heat tolerance, paralleling the downregulation of its target gene, Squamosa Promoter-Binding Protein-Like 13 (SPL13) [21,24].The abundance of miR172 is elevated during heat stress, inducing the expression of FT and flowering in Arabidopsis [20,25].The MADS-box transcription factor AGAMOUS LIKE 16 (AGL16), a flowering negative regulator, is negatively regulated by miR824, leading to a mild depression of FT [22,26].However, the function of miRNAs in tuberization during high-temperature stress remains unclear.
In this study, high-temperature stress was induced in vitro to investigate the hightemperature response mechanism during tuberization in potatoes.We performed mRNA and miRNA sequencing and identified differentially expressed genes and miRNAs in leaves and stems between normal and high temperatures.An enrichment analysis identified regulatory pathways during tuberization under high temperatures.A comprehensive interaction network was constructed under high temperatures between miRNAs and target genes for tuberization.We identified key candidate miRNA-target gene pairs involved in tuberization under high-temperature stress.Our study shows a comprehensive understanding of tuberization under high temperatures in potatoes.

Plant Materials and Growth Conditions
We used diploid RH89-039-16 (RH, Solanum tuberosum L.) in this study, which was preserved at the Institute of Vegetable and Flower Research, Chinese Academy of Agricultural Sciences.The plants were propagated in vitro using single-node stems on MS medium supplemented with 30 g/L sucrose and 7 g/L plant agar under 16 h light/8 h dark at 20 • C. For the tuberization assay in vitro, the single-node stems of RH plants were cultured in MS medium supplemented with 60 g/L sucrose and 7 g/L plant agar under 16 h light/8 h dark at 20 • C for three weeks.Then, the plants were transferred to 8 h light/16 h dark at 20 • C and 27 • C, respectively.The light intensities in all conditions were 2000 lx.Each experiment was designed with three biological replicates of at least 24 plantlets (3 plantlets in each culture bottle).Seedlings with similar plant heights and stem diameters were selected for further culture to avoid the influence of individual plant differences.

mRNA Library Construction and Sequencing
Potato tuberization was delayed about 8 days at 27 • C compared to that at 20 • C in our study.Therefore, leaves and stems of 7-week-old plants with tubers cultured at 20 • C and without tubers cultured at 20 • C and 27 • C were selected for RNA sequencing.The leaves and stems of plants with tubers at 27 • C were not used in this study.The samples of leaves of plants with tubers at 20 • C, leaves of plant without tubers at 20 • C, stems of plant with tubers at 20 • C, stems of plant without tubers at 20 • C, leaves of plant without tubers at 27 • C, and stems of plant without tubers at 27 • C were abbreviated as TL_20, UTL_20, TS_20, UTS_20, UTL_27, and UTS_27, respectively.Each sample was designed with three biological replicates.One sample of UTL_20 was abolished due to its poor uniformity with the other two samples in the principal component analysis.
Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's procedure.The RNA of each sample was quantified using the Agilent Bioanalyzer 2100 system.Poly(A) RNA was purified using Dynabeads Oligo(dT) (Thermo Fisher, Waltham, CA, USA) and then fragmented into small pieces using a Magnesium RNA Fragmentation Module (NEB, Boston, USA).The RNA fragments were reverse transcribed to create cDNA using SuperScript™ II Reverse Transcriptase (Invitrogen, cat.1896649, USA).Then, the U-labeled second-stranded DNAs were synthesized with DNA polymerase I (NEB, cat.m0209,USA), RNase H (NEB, cat.m0297,USA), and dUTP Solution (Thermo Fisher, cat.R0133, USA).The blunt ends of each strand were added an A-base.The fragments were ligated with single-or dual-index adapters and underwent size selection with AMPureXP beads.Finally, we performed sequencing on an Illumina Novaseq™ 6000 (LC-BIO Technology CO., Ltd., Hangzhou, China) following the manufacturer's protocol.

miRNA Library Construction, Sequencing, and Analysis
The samples for miRNA analysis were isolated from the total RNA used for RNA sequencing.MiRNAs were isolated from the total RNA via agarose gel electrophoresis.The library construction was conducted according to the standard steps provided by Illumina Company.TruSeq Small RNA Sample Prep Kits (Illumina, San Diego, CA, USA) were used to prepare the libraries.The constructed libraries were sequenced using Illumina Hiseq2500 at LC-BIO (Hangzhou, China).MiRNA sequences were identified using ACGT101-miR (LC Sciences, Houston, TX, USA).Unique sequences with 18-25 nucleotides in length were mapped to potato precursors in miRBase 22.0 using BLAST search to identify known miRNAs and novel 3p-and 5p-derived miRNAs.In the alignment, length variation at both 3 ′ and 5 ′ ends and one mismatch within the sequence were allowed.Unique sequences that matched mature miRNAs from specific species in the hairpin arm were identified as known miRNAs.Unique sequences that matched the other arm of the known precursor hairpin arm of a specific species (opposite to the annotated arm containing mature miRNAs) were considered novel 3p-and 5p-derived miRNAs.The naming of miRNAs refers to the miRBase database (What's in a name?http://www.mirbase.org/blog).For the miRNAs that did not have an annotation before in potato, we named them according to the homologous miRNAs in other species.The expressions of miRNAs were calculated as reported by Li et al. 2016 [28].The differentially expressed miRNAs (DEmiRNAs) were screened out with parameters of a p-value < 0.05 and q-value < 0.05.

Principal Component, Gene Enrichment, and Hierarchical Clustering Analysis
A principal component analysis was performed using the vegan package of R. A gene ontology enrichment analysis, KEGG pathway enrichment analysis, and hierarchical clustering analysis of the DEGs and DEmiRNAs were performed using TBtools (version 1.0.99)[29].

miRNA-mRNA Interaction Network Generation
To identify the potential miRNA-mRNA network for tuberization during high temperatures, we used the DEmiRNAs and their target DEGs for the interaction network.The putative target genes of miRNAs were predicted using psRobot (http://omicslab.genetics.ac.cn/psRobot/, v1.2) with parameters of score ≤ 2.5.The graphic network was created using Cytoscape 3.9.1 [30].

Quantitative Real-Time PCR Analysis
The mRNA and miRNA validation were performed according to the manufacturer's protocol (Hieff ® qPCR SYBR ® Green Master Mix, YEASEN; miRNA Universal SYBR qPCR Master Mix, Vazyme, Nanjing, China), and the template-specific primers were designed using Primer3.0 (https://primer3.ut.ee/) and miRNA Design V 1.01 software (Vazyme).All reactions were conducted in a Light Cycler 480 Real-Time PCR System (Roche Life Science, Basel, Switzerland).Reaction specificity was assessed using a melting curve, and qPCR data were analyzed using the 2 −∆∆Ct method.The relative expression level was calculated relative to EF1α and U6 snRNA.All primer sequences for the real-time PCR analysis are listed in Table S2.

Suppression of Potato Tuberization In Vitro under High Temperatures
To explore the mechanism of potato tuberization under high temperatures, we induced high-temperature stress in vitro for a phenotypic analysis.Plants were cultured at a normal temperature of 20 • C for three weeks under long days (LDs) (16 h light/8 h dark).Then, the plants were transported to 27 • C and 20 • C under short days (SDs) (8 h light/16 h dark) for subsequent growth.We observed and determined the tuberization time to evaluate the effect of high-temperature stress.The average initiation tuberization time under 27 • C was 55 days, which was significantly delayed compared to 20 • C (47 days; Figure 1).

Transcriptome Analysis of Tuberization under High-Temperature Stress
RNA sequencing was performed to investigate the molecular basis of tuberization under high-temperature stress in potatoes.Considering that the expressions of many regulators for tuberization were controlled in leaves, but the proteins were transported to the stolon through stems to regulate tuberization, leaves and stems were both sequenced.At least two biological replicates of each tissue were used to construct the cDNA libraries.A total of 87.28 Gb valid data were generated after conducting a quality control of the raw sequencing reads, with an average mapped ratio of 96.29%.The Q30 was 98.42%, and the GC content was 43.03% (Table S1).We randomly selected 10 genes for real-time PCR (RT-PCR) to verify the accuracy of the transcriptome data (Table S2).The R 2 between RNA-seq and RT-qPCR was estimated to be 0.8184, which illustrates the reliability of our transcriptome (Figure S1).
The principal component analysis of the gene expression dataset showed high uniformity between biological replicates (Figure 2a).The clusters of leaves and stems at 20 °C (TL_20, leaves of plants with tubers at 20 °C; UTL_20, leaves of plant without tubers at 20 °C, TS_20, stems of plant with tubers at 20 °C; UTS_20, stems of plant without tubers at 20 °C) were distinguished clearly from those at 27 °C (UTL_27, leaves at 27 °C; UTS_27, stems at 27 °C).These results were further supported through the hierarchical clustering analysis, which showed similar gene expression profiles at the same temperatures but significantly distinct gene expression profiles between different temperatures (Figure 2b).
For the stem, 5001 and 1445 DEGs were identified by comparing UTS_27 vs. TS_20 and UTS_20 vs. TS_20, respectively (Figure 2c).In total, 1106 genes overlapped between these two comparisons (Figure 2c) and showed similar expression patterns between UTS_27 and UTS_20 (Figure S3a).These genes were significantly enriched regarding biological process gene ontology (GO) terms related to cell wall biogenesis-related processes (Figure S3b).

Transcriptome Analysis of Tuberization under High-Temperature Stress
RNA sequencing was performed to investigate the molecular basis of tuberization under high-temperature stress in potatoes.Considering that the expressions of many regulators for tuberization were controlled in leaves, but the proteins were transported to the stolon through stems to regulate tuberization, leaves and stems were both sequenced.At least two biological replicates of each tissue were used to construct the cDNA libraries.A total of 87.28 Gb valid data were generated after conducting a quality control of the raw sequencing reads, with an average mapped ratio of 96.29%.The Q30 was 98.42%, and the GC content was 43.03% (Table S1).We randomly selected 10 genes for real-time PCR (RT-PCR) to verify the accuracy of the transcriptome data (Table S2).The R 2 between RNA-seq and RT-qPCR was estimated to be 0.8184, which illustrates the reliability of our transcriptome (Figure S1).
The principal component analysis of the gene expression dataset showed high uniformity between biological replicates (Figure 2a).The clusters of leaves and stems at 20 • C (TL_20, leaves of plants with tubers at 20 • C; UTL_20, leaves of plant without tubers at 20 • C, TS_20, stems of plant with tubers at 20 • C; UTS_20, stems of plant without tubers at 20 • C) were distinguished clearly from those at 27 • C (UTL_27, leaves at 27 • C; UTS_27, stems at 27 • C).These results were further supported through the hierarchical clustering analysis, which showed similar gene expression profiles at the same temperatures but significantly distinct gene expression profiles between different temperatures (Figure 2b).
For the stem, 5001 and 1445 DEGs were identified by comparing UTS_27 vs. TS_20 and UTS_20 vs. TS_20, respectively (Figure 2c).In total, 1106 genes overlapped between these two comparisons (Figure 2c) and showed similar expression patterns between UTS_27 and UTS_20 (Figure S3a).These genes were significantly enriched regarding biological process gene ontology (GO) terms related to cell wall biogenesis-related processes (Figure S3b).

Gene ontology Annotation and Pathway Enrichment Analysis of DEGs
We performed GO term enrichment analysis on the total DEGs in comparison of UTL_27 vs. TL_20 and UTS_27 vs. TS_20 to comprehensively classify the functional categories of the DEGs.The DEGs, in comparison of UTL_27 vs. TL_20, were significantly enriched in biological process (BP) terms related to photosynthesis, cell division, meristem development, tissue development, and gibberellin metabolic process (Figure 3a).In the molecular function (MF) category, GO terms, such as microtubule moto activity, glucosyltransferase activity, and enzyme inhibitor activity, were significantly enriched (Figure 3a).In the cellular component (CC) category, the DEGs were enriched regarding the chloroplast thylakoid, chloroplast thylakoid, and cell wall (Figure 3a).The DEGs in comparison of UTS_27 vs. TS_20 were enriched in the developmental process, sucrose biosynthetic process, and response to heat (Figure 3b).In the MF category, GO terms, such as glycosyltransferase activity, calmodulin binding, and the activity of ubiquitin-protein transferase

Gene Ontology Annotation and Pathway Enrichment Analysis of DEGs
We performed GO term enrichment analysis on the total DEGs in comparison of UTL_27 vs. TL_20 and UTS_27 vs. TS_20 to comprehensively classify the functional categories of the DEGs.The DEGs, in comparison of UTL_27 vs. TL_20, were significantly enriched in biological process (BP) terms related to photosynthesis, cell division, meristem development, tissue development, and gibberellin metabolic process (Figure 3a).In the molecular function (MF) category, GO terms, such as microtubule moto activity, glucosyltransferase activity, and enzyme inhibitor activity, were significantly enriched (Figure 3a).In the cellular component (CC) category, the DEGs were enriched regarding the chloroplast thylakoid, chloroplast thylakoid, and cell wall (Figure 3a).The DEGs in comparison of UTS_27 vs. TS_20 were enriched in the developmental process, sucrose biosynthetic process, and response to heat (Figure 3b).In the MF category, GO terms, such as glycosyltransferase activity, calmodulin binding, and the activity of ubiquitin-protein transferase activator, were significantly enriched (Figure 3b).In the CC category, the DEGs were enriched regarding the plastoglobule, cell periphery, and cell wall related terms (Figure 3b).
Plants 2024, 13, x FOR PEER REVIEW 7 of 15 activator, were significantly enriched (Figure 3b).In the CC category, the DEGs were enriched regarding the plastoglobule, cell periphery, and cell wall related terms (Figure 3b).We performed a pathway enrichment analysis based on the KEGG database to further analyze the molecular response under high-temperature stress.Leaf DEGs were significantly enriched in amino acid, fructose, and mannose metabolism related terms (Figure S4a).Stem DEGs were significantly enriched in 27 pathways, for instance, membrane trafficking, MAPK signaling pathway, amino acid metabolism, environmental information processing, and signal transduction (Figure S4b).

Analysis of Typical Tuberization Regulators under High Temperatures in Potatoes
Tuberization is regulated by a series of tuberigen formation factors, such as StSP6A, StBEL5, StSP5G, StCEN1, and StCOL1, and forms a regulatory pathway centered around StSP6A [31].We found that the expressions of StSP6A (RHC05H1G2713) and StBEL5 (RHC06H2G2699) were significantly downregulated at 27 • C compared to 20 • C. The expressions of StSP5G (RHC05H2G2263), StCEN1 (RHC03H1G1254), and StCOL1 (RHC02H2G2633) were not obviously altered at 27 • C (Figure 4), which is similar to that reported by Park et al. (2022) [5].These results suggest that the regulatory mechanisms of tuberization at high temperatures might differ from those at normal temperature.
We performed a pathway enrichment analysis based on the KEGG database to further analyze the molecular response under high-temperature stress.Leaf DEGs were significantly enriched in amino acid, fructose, and mannose metabolism related terms (Figure S4a).Stem DEGs were significantly enriched in 27 pathways, for instance, membrane trafficking, MAPK signaling pathway, amino acid metabolism, environmental information processing, and signal transduction (Figure S4b).

Analysis of Typical Tuberization Regulators under High Temperatures in Potatoes
Tuberization is regulated by a series of tuberigen formation factors, such as StSP6A, StBEL5, StSP5G, StCEN1, and StCOL1, and forms a regulatory pathway centered around StSP6A [31].We found that the expressions of StSP6A (RHC05H1G2713) and StBEL5 (RHC06H2G2699) were significantly downregulated at 27 °C compared to 20 °C.The expressions of StSP5G (RHC05H2G2263), StCEN1 (RHC03H1G1254), and StCOL1 (RHC02H2G2633) were not obviously altered at 27 °C (Figure 4), which is similar to that reported by Park et al. ( 2022) [5].These results suggest that the regulatory mechanisms of tuberization at high temperatures might differ from those at normal temperature.

Expression Dynamics of miRNAs during High-Temperature Stress
Post-transcriptional regulation plays an essential role at the early stage of tuberization, such as microRNA (miRNAs, miR) regulation [5].We performed miRNA sequencing to identify the miRNAs involved in tuber formation during high temperatures.The miRNA expression was validated using RT-PCR and was consistent with miRNA sequencing, suggesting the reliability of miRNA sequencing (Figure S5 and Table S2).The principal component analysis showed that clusters of leaves and stems at 20 °C were distinguished clearly from those at 27 °C, similar to the RNA sequencing analysis (Figure 5a).These results were supported by the hierarchical clustering analysis (Figure 5b).
A total of 463 miRNAs belonging to 75 families were identified (Figure S6 and Table S3), among which 200 miRNAs were known in potato (Table S4).In comparison of UTL_27 vs. TL_20, 104 differentially expressed miRNAs (DEmiRNAs) were identified, including 45 up-and 59 downregulated miRNAs (Figure 5c).A total of 1913 target genes were predicted, most of which were expressed differently between TL_20 and UTL_27 (Figure S7a).These 1,913 target genes were enriched in BP terms regarding the defense response, response to stress, and cell death (Figure S7b).Similarly, 75 DEmiRNAs in comparison of UTS_27 vs. TS_20 were identified, including 29 upregulated and 46 downregulated miR-NAs (Figure 5d), and 1275 target genes were predicted and expressed differently between

Expression Dynamics of miRNAs during High-Temperature Stress
Post-transcriptional regulation plays an essential role at the early stage of tuberization, such as microRNA (miRNAs, miR) regulation [5].We performed miRNA sequencing to identify the miRNAs involved in tuber formation during high temperatures.The miRNA expression was validated using RT-PCR and was consistent with miRNA sequencing, suggesting the reliability of miRNA sequencing (Figure S5 and Table S2).The principal component analysis showed that clusters of leaves and stems at 20 • C were distinguished clearly from those at 27 • C, similar to the RNA sequencing analysis (Figure 5a).These results were supported by the hierarchical clustering analysis (Figure 5b).
A total of 463 miRNAs belonging to 75 families were identified (Figure S6 and Table S3), among which 200 miRNAs were known in potato (Table S4).In comparison of UTL_27 vs. TL_20, 104 differentially expressed miRNAs (DEmiRNAs) were identified, including 45 upand 59 downregulated miRNAs (Figure 5c).A total of 1913 target genes were predicted, most of which were expressed differently between TL_20 and UTL_27 (Figure S7a).These 1913 target genes were enriched in BP terms regarding the defense response, response to stress, and cell death (Figure S7b).Similarly, 75 DEmiRNAs in comparison of UTS_27 vs. TS_20 were identified, including 29 upregulated and 46 downregulated miRNAs (Figure 5d), and 1275 target genes were predicted and expressed differently between TS_20 and UTS_27 (Figure S7c).All the target genes were significantly enriched in BP terms related to the cellular response to stimulus, the regulation of hydrolase activity, and anther morphogenesis (Figure S7d).TS_20 and UTS_27 (Figure S7c).All the target genes were significantly enriched in BP terms related to the cellular response to stimulus, the regulation of hydrolase activity, and anther morphogenesis (Figure S7d).

Regulatory Networks between DEGs and DEmiRNAs under High Temperatures
We constructed interaction networks using DEmiRNAs and DEGs to identify the potential miRNAs and target genes correlated with tuberization under high temperatures.In total, 118 DEmiRNA-DEG pairs were identified in the leaf (Figure 6 and Table S5).Among these pairs, five StSPLs were determined to be targeted by stu-miR156f-5p and stu-miR156a_L, of which their expressions were reduced under high temperatures (Figures 6  and S8).Among these five StSPLs, the expressions of StSPL6-like1 (StSPL6-L1), StSPL6-like2 (StSPL6-L2), StSPL16-like1 (StSPL16-L1), and StSPL16-like2 (StSPL16-L2) in leaves increased under high temperatures (Figure S9).Two MADS translation factors RHC06H2G2343 and RHC06H1G2441, named StAGL8a and StAGL8b, respectively, were found to be targeted by stu-miR7981f-p5 and stu-miR10532A (Figure 6).The orthologous genes of these factors in Arabidopsis and tomato play crucial roles in floral organ development and fruit ripening [32][33][34].The expression levels of these two genes at 27 °C were

Regulatory Networks between DEGs and DEmiRNAs under High Temperatures
We constructed interaction networks using DEmiRNAs and DEGs to identify the potential miRNAs and target genes correlated with tuberization under high temperatures.In total, 118 DEmiRNA-DEG pairs were identified in the leaf (Figure 6 and Table S5).Among these pairs, five StSPLs were determined to be targeted by stu-miR156f-5p and stu-miR156a_L, of which their expressions were reduced under high temperatures (Figures 6 and S8).Among these five StSPLs, the expressions of StSPL6-like1 (StSPL6-L1), StSPL6-like2 (StSPL6-L2), StSPL16-like1 (StSPL16-L1), and StSPL16-like2 (StSPL16-L2) in leaves increased under high temperatures (Figure S9).Two MADS translation factors RHC06H2G2343 and RHC06H1G2441, named StAGL8a and StAGL8b, respectively, were found to be targeted by stu-miR7981f-p5 and stu-miR10532A (Figure 6).The orthologous genes of these fac-tors in Arabidopsis and tomato play crucial roles in floral organ development and fruit ripening [32][33][34].The expression levels of these two genes at 27 • C were significantly reduced compared with those at 20 • C (Figure S9).In addition, a P450 monooxygenase RHC03H2G1398.2,named StCPY714, was found to be a target of stu-miR8030-5p and downregulated in leaves under high temperatures (Figures 6 and S9).The homologous genes of StCPY714 in Arabidopsis and rice catalyze the deactivation of bioactive gibberellins, which have been reported to be negative regulators during tuberization [35].These DEmiRNA-DEG pairs were considered candidate regulators for tuberization under high temperatures.In addition, we found that the miR172d_5p expression increased under high temperatures and targeted two endoglucanase genes (RHC04H1G3081 and RHC04H2G2838; Figures 6 and S8).
Plants 2024, 13, x FOR PEER REVIEW 10 of 15 significantly reduced compared with those at 20 °C (Figure S9).In addition, a P450 monooxygenase RHC03H2G1398.2,named StCPY714, was found to be a target of stu-miR8030-5p and downregulated in leaves under high temperatures (Figures 6 and S9).The homologous genes of StCPY714 in Arabidopsis and rice catalyze the deactivation of bioactive gibberellins, which have been reported to be negative regulators during tuberization [35].These DEmiRNA-DEG pairs were considered candidate regulators for tuberization under high temperatures.In addition, we found that the miR172d_5p expression increased under high temperatures and targeted two endoglucanase genes (RHC04H1G3081 and RHC04H2G2838; Figures 6 and S8).For stems, 150 pairs of DEmiRNA-DEGs were identified, among which 13 pairs comprising 5 DEmiRNAs and 9 DEGs overlapped with those in leaves, including stu-miR156f-5p and stu-miR156a_L (Figure 7 and Table S6).However, only two StSPLs (StSPL1 and StSPL6-L2) targeted by miRNA156 were identified in the stem due to the similar expression levels of other StSPLs at 20 °C and 27 °C (Figure S9), indicating different regulation mechanisms between leaves and stems in response to high temperatures.For stems, 150 pairs of DEmiRNA-DEGs were identified, among which 13 pairs comprising 5 DEmiRNAs and 9 DEGs overlapped with those in leaves, including stu-miR156f-5p and stu-miR156a_L (Figure 7 and Table S6).However, only two StSPLs (StSPL1 and StSPL6-L2) targeted by miRNA156 were identified in the stem due to the similar expression levels of other StSPLs at 20 • C and 27 • C (Figure S9), indicating different regulation mechanisms between leaves and stems in response to high temperatures.

Discussion
Potato tuberization is regulated by environmental signals and internal factors, su as temperature and hormones [16,35].In this study, we demonstrated high-temperat delays in tuberization in vitro.RNA sequencing was performed to investigate the regu tory pathways for tuberization under high temperatures.We identified 2804 DEGs leaves under high temperatures, most of which were enriched in GO terms related to te perature stress-related processes, development-related processes, and hormone respo processes (Figure 3a).Meanwhile, 5001 DEGs were identified in stems, which were sign icantly enriched in processes related to sucrose metabolism, hormone response, and tis development (Figure 3b).These results comprehensively reveal the pathways of tuberi tion under high-temperature stress.

Discussion
Potato tuberization is regulated by environmental signals and internal factors, such as temperature and hormones [16,35].In this study, we demonstrated high-temperature delays in tuberization in vitro.RNA sequencing was performed to investigate the regulatory pathways for tuberization under high temperatures.We identified 2804 DEGs in leaves under high temperatures, most of which were enriched in GO terms related to temperature stress-related processes, development-related processes, and hormone response processes (Figure 3a).Meanwhile, 5001 DEGs were identified in stems, which were significantly enriched in processes related to sucrose metabolism, hormone response, and tissue development (Figure 3b).These results comprehensively reveal the pathways of tuberization under high-temperature stress.
The expression of StSP6A is controlled by the negative regulator StSP5G, which is positively regulated by StCOL1 [8].StSP6A, StFDL1, and St14-3-3 combine to form a tuberigen activation complex to induce tuber formation [10].StCEN suppresses tuberization by competitively binding StFDL1 through St14-3-3 in TAC [11].As previously reported [4,5,16,17], our study also found that the expression of StSP6A is significantly suppressed under high temperatures.However, the expressions of StCEN, StCOL1, and StSP5G were not altered (Figure 4).Additionally, StSP6A transcriptional activity is linked to the expression of StBEL5, another tuberization inducer [14].The StBEL5 expression was reduced under high temperatures (Figure 4).This result further supports the study by Park et al. (2022) [5] in that other negative regulators exited during tuberization under high temperatures.
miRNAs are essential in plant growth, development, environmental adaptability, and stress resistance.At present, many miRNAs have been reported to respond to temperature signals, such as miR156, miR172, miR824, and miR168 [20][21][22][23].In rice and cassava, the expression of miR156 is reduced under high-temperature stress [36,37].An elevated expression of miR156 in alfalfa enhances the response to high temperatures, paralleling the reduced expression of target gene SPL13 [21,24].In potatoes, the overexpression of miR156 suppresses tuberization, accompanied through reductions in miR172 levels and StSP6A expression and through the suppression of the expression of StSPLs [18,19,38].In our study, 104 DEmiRNAs were identified in leaves under high temperatures, including miR156 and miR17.In total, 1913 target genes of the DEmiRNAs were predicted, and 118 DEmiRNA-DEG pairs were identified (Figure 6 and Table S5).The stu-miR156a and stu-miR156f expressions were significantly downregulated under high temperatures in leaves and stems (Figure S8).Among the DEmiRNA-DEG pairs, five StSPLs targeted by stmiR156s were identified, four of which showed increased expressions under high temperatures (Figures 6 and S9).This result suggests a conservative function of miR156s/SPLs in plants under high-temperature stress.miR172 functions as a positive regulator in tuberization through upregulating the StBEL5 expression [18,19].However, only two endoglucanase genes targeted by stumiR172b_5p were identified (Figure 6), suggesting different regulatory mechanisms of stu-miR172 under high temperatures.For stems, 150 DEmiRNA-DEG pairs were identified, most of which differed from those in leaves.For instance, only two StSPLs targeted by stu-miR156f-5p and stu-miR156a_L overlapped with those in the leaf (Figures 6 and 7).Interestingly, miR156 overexpression reduces tuber yield [19,39].However, a reduction in the miR156 expression also displayed the suppression of tuberization under high temperatures.Further research is needed on the regulatory mechanism of miR156 in tuber formation.
Gibberellin is a negative regulator during tuberization by constraining cortical microtubule reorganization at the sub-apical region of the stolon [35,40,41].Our study identified that a gibberellin deactivator StCPY714 targeted by stu-miR8030-5p was downregulated under high temperatures.This result provides a new clue for the gibberellin involved in tuberization.Moreover, two MADS-box transcription factors, StAGL8a and StAGL8b, targeted by stu-miR7981f-p5 and stu-miR10532A, showed reduced expression levels under high temperatures (Figures 6 and S9).The orthologous genes in Arabidopsis and tomato regulate floral organ development and fruit ripening [32][33][34].Therefore, StAGL8a and StAGL8b might be related to the response to high temperature and tuberization.

Conclusions
In this study, we analyzed regulatory pathways and constructed a comprehensive miRNA-mRNA regulatory network for tuberization under high temperatures.We identified the conservative function of miR156/StSPL modules in response to high temperatures and suppressing tuberization.We found three candidate miRNA-mRNA pairs in regulating tuberization under high temperatures, including stu-miR8030-5p/StCPY714, stu-miR7981f-p5/StAGL8a, and stu-miR10532A/StAGL8b.Our study provides new insights into potato tuberization under high temperatures and will facilitate the identification of regulatory genes involved in the response to high temperatures.

Figure 2 .
Figure 2. Transcriptome analysis of leaves and stems under different temperatures.(a) Principal component analysis showing the relationships of samples in leaves and stems at 20 °C and 27 °C.The X-axis and Y-axis represent PCA1 and PCA2, respectively, explaining 46.1 and 43.8% of the total variance.(b) Hierarchical cluster analysis of DEGs.The columns and rows represent samples and DEGs, respectively.The color key on the right shows normalized FPKM values for relative gene expressions.(c) Venn diagram of DEGs at different temperatures.The upper and lower rows represent samples of leaves and stems, respectively.(d) Quantification of upregulated and downregulated genes at high temperature of 27 °C compared to normal temperature of 20 °C.TL_20, UTL_20, TS_20, UTS_20, UTL_27, and UTS_27 represent leaves of plant with tubers at 20 °C, leaves of plant without tubers at 20 °C, stems of plant with tubers at 20 °C, stems of plant without tubers at 20 °C, leaves at 27 °C, and stems at 27 °C, respectively.

Figure 2 .
Figure 2. Transcriptome analysis of leaves and stems under different temperatures.(a) Principal component analysis showing the relationships of samples in leaves and stems at 20 • C and 27 • C. The X-axis and Y-axis represent PCA1 and PCA2, respectively, explaining 46.1 and 43.8% of the total variance.(b) Hierarchical cluster analysis of DEGs.The columns and rows represent samples and DEGs, respectively.The color key on the right shows normalized FPKM values for relative gene expressions.(c) Venn diagram of DEGs at different temperatures.The upper and lower rows represent samples of leaves and stems, respectively.(d) Quantification of upregulated and downregulated genes at high temperature of 27 • C compared to normal temperature of 20 • C. TL_20, UTL_20, TS_20, UTS_20, UTL_27, and UTS_27 represent leaves of plant with tubers at 20 • C, leaves of plant without tubers at 20 • C, stems of plant with tubers at 20 • C, stems of plant without tubers at 20 • C, leaves at 27 • C, and stems at 27 • C, respectively.

Figure 3 .
Figure 3. Gene ontology (GO) enrichment analysis of DEGs under high-temperature stress.GO term enrichment analysis of DEGs in leaves (a) and stems (b).The colors of the box represent −log10 (pvalue).The size of the boxes represents the gene count for GO terms.Biological process, molecular function, and cellular component terms are presented in blue, red, and yellow, respectively.

Figure 3 .
Figure 3. Gene ontology (GO) enrichment analysis of DEGs under high-temperature stress.GO term enrichment analysis of DEGs in leaves (a) and stems (b).The colors of the box represent −log 10 (p-value).The size of the boxes represents the gene count for GO terms.Biological process, molecular function, and cellular component terms are presented in blue, red, and yellow, respectively.

Figure 4 .
Figure 4. Expression of tuberization-related genes.The FPKM statistics of genes related to tuberization in leaves at different temperatures.Error bars indicate ± SD. *** p < 0.001, Student's t test.

Figure 4 .
Figure 4. Expression of tuberization-related genes.The FPKM statistics of genes related to tuberization in leaves at different temperatures.Error bars indicate ± SD. *** p < 0.001, Student's t test.

Figure 5 .
Figure 5. miRNA sequencing analysis of leaves and stems under different temperatures.(a) Principal component analysis showing the relationships of samples in leaves and stems under different temperatures.The X-axis and Y-axis represent PCA1 and PCA2, respectively, explaining 24.4 and 14.9% of the total variance.(b) Hierarchical cluster analysis of total miRNAs.(c,d) Hierarchical cluster analysis of DEmiRNAs in leaves and stems, respectively.The columns and rows represent samples and miRNAs, respectively.The color key on the right of (b-d) shows normalized norm values for relative expression.

Figure 5 .
Figure 5. miRNA sequencing analysis of leaves and stems under different temperatures.(a) Principal component analysis showing the relationships of samples in leaves and stems under different temperatures.The X-axis and Y-axis represent PCA1 and PCA2, respectively, explaining 24.4 and 14.9% of the total variance.(b) Hierarchical cluster analysis of total miRNAs.(c,d) Hierarchical cluster analysis of DEmiRNAs in leaves and stems, respectively.The columns and rows represent samples and miRNAs, respectively.The color key on the right of (b-d) shows normalized norm values for relative expression.