Adaptive Strategies of Cyrtorhinus lividipennis (Hemiptera: Miridae) to Short-Term High Temperature Stress: Insights from Physiological and Transcriptomic Responses

Simple Summary

As climate change intensifies, it is becoming increasingly critical to understand how insect natural enemies of crop pests tolerate rising temperatures. Here, we focus on the green mirid bug, a beneficial predator of the major rice pest, the brown planthopper, and examine how it copes with short-term extreme heat. Our experiments show that heat exposure induces the accumulation of protective sugars and lipids and alters energy metabolism by shifting metabolic pathways toward stress defense and energy supply, thereby improving survival. We further find that high temperature induces the expression of several stress-responsive genes, including those that help safeguard cellular structures from damage. Together, these results elucidate the mechanisms underlying short-term extreme heat tolerance in this natural enemy and provide a mechanistic basis for future studies to develop and evaluate field-release strategies in rice systems.

Abstract

Cyrtorhinus lividipennis, a key natural enemy of the brown planthopper, Nilaparvata lugens, has been observed to tolerate short-term high-temperature exposure; however, the physiological and molecular mechanisms underlying this heat tolerance remain unclear, which may hinder its effective conservation and utilization. Here, we combined physiological and biochemical assays with transcriptome sequencing to elucidate the physiological and molecular mechanisms of heat tolerance in C. lividipennis following 1 h exposure to three temperatures: 26 °C (control), 33 °C (moderate heat stress), and 40 °C (severe heat stress). At 40 °C, sorbitol, trehalose, lipid, and glycogen contents increased significantly, whereas glycerol levels declined. Transcriptomic profiling revealed temperature-dependent DEGs enriched in starch and sucrose metabolism, galactose metabolism, glycerolipid metabolism, oxidative phosphorylation, and protein folding, sorting, and degradation, with pronounced temperature-dependent upregulation of heat shock protein (HSP) gene families. Together, these results demonstrate that C. lividipennis coordinates its heat stress response through soluble polyol accumulation, which is known to act as a compatible osmolytes that help stabilize proteins and membranes and mitigate thermal damage, energy metabolic reprogramming, and HSP-mediated proteostasis, thereby providing a theoretical basis for its conservation and utilization in sustainable paddy agroecosystems.

1. Introduction

The green mirid bug Cyrtorhinus lividipennis Reuter is an important predatory natural enemy of the brown planthopper (BPH), Nilaparvata lugens Stål, a major rice pest, and contributes to biological control in rice ecosystems [1,2]. C. lividipennis is also a long-distance migrant and can co-migrate with BPH into the rice-growing regions of South China, followed by a northward spread along the migration route [3]. The Guangxi Zhuang Autonomous Region lies within this migration route and may act as an important area influencing subsequent population build-up and biological control services of C. lividipennis in other rice regions [4].

During midsummer (July–August), afternoon temperatures in Guangxi frequently exceed 35 °C, yet net-house and field observations indicate that C. lividipennis populations can remain relatively high, suggesting tolerance to short-term episodes of extreme heat. In natural environments, temperatures are inherently dynamic rather than static. Insects can exhibit significant physiological stress responses even to brief, transient exposures to high temperatures; consequently, these fluctuating thermal effects are indispensable for accurately predicting insect responses to climate warming [5]. Temperature strongly constrains insect activity and fitness [6] and exposure to extremely high temperatures can reduce survival and reproduction and disrupt development and physiology [7,8,9,10,11,12]. Given the increasing frequency and intensity of extreme heat events [13,14], clarifying how C. lividipennis responds to short-term high-temperature peaks is important for assessing its potential resilience and biological control function under future heat extremes.

Insects deploy multilayered responses to cope with heat stress. At the physiological and biochemical levels, shifts in metabolites (e.g., trehalose, sorbitol, glycogen, glycerol, and lipids) can mitigate heat-induced damage [15,16,17]. At the transcriptional level, RNA-seq coupled with GO and KEGG enrichment is widely used to characterize differentially expressed genes (DEGs) and associated pathways under heat stress [18,19]. Heat shock proteins (HSPs) are among the most rapidly induced DEGs in response to high temperature [20,21,22] and function as molecular chaperones that stabilize proteins and promote correct folding, thereby limiting cellular damage [23,24]. In addition, alternative splicing is an important post-transcriptional mechanism underpinning insect heat responses, expanding proteomic diversity without increasing gene number [25]. Notably, several HSP genes can themselves be regulated by alternative splicing, providing further regulatory flexibility during thermal challenge [26].

Despite its ecological importance, the heat-stress response of C. lividipennis remains poorly characterized, and integrated analyses linking metabolic adjustment with transcriptional and post-transcriptional regulation are limited. In this study, we aimed to characterize the physiological and molecular responses of C. lividipennis to short-term high-temperature exposure that mimics midday/afternoon thermal peaks in the field. We quantified changes in heat-tolerance-related metabolites, identified heat-responsive DEGs with an emphasis on HSP families, and profiled heat-associated alternative splicing events. These results provide candidate pathways and genes for future functional validation and help evaluate the resilience of this predator under increasing heat extremes.

2. Materials and Methods

2.1. Rearing of C. lividipennis

A laboratory colony of C. lividipennis was established from adults collected in 2023 from experimental paddy fields at the Guangxi Academy of Agricultural Sciences. The insects were maintained in a net house under controlled conditions of 26 ± 1 °C, 80 ± 10% relative humidity, and a 12:12 h (L:D) photocycle. Gravid BPH females were introduced onto TN1 plants at the tillering stage and allowed to oviposit for 5 days. The egg-bearing plants were then removed and placed in rearing cages as host plants to support all developmental stages of C. lividipennis. These host plants were replaced every 2 days to ensure a continuous and adequate supply of BPH eggs for predation. The experimental C. lividipennis population was obtained from the F3 generation of a laboratory colony maintained under these cage conditions.

2.2. Sampling of C. lividipennis

The experiments were conducted in three separate batches to ensure biological independence. In each batch, groups of adult C. lividipennis (2 days post-emergence) were exposed to three distinct temperature regimes in controlled-environment chambers. A temperature of 26 °C was used as the control to represent optimal growth conditions, 33 °C was set to impose moderate heat stress, and 40 °C was selected to represent severe heat stress. These regimes were chosen based on preliminary experiments indicating that prolonged exposure to 40 °C resulted in complete mortality. After 1 h at the designated temperatures, 100 adults (50 males and 50 females) were collected per treatment and pooled to form one biological replicate. Males and females were combined to minimize sex-specific bias in heat stress responses and to better capture the overall response to each treatment. This entire procedure was repeated three times, providing three independent biological replicates for each temperature condition. All samples were immediately flash-frozen in liquid nitrogen for subsequent physiological, biochemical, and transcriptomic analyses.

2.3. Quantification of Key Physiological and Biochemical Substances in C. lividipennis

C. lividipennis adults subjected to the temperature treatments (Section 2.2) were used for all biochemical assays. For each treatment, three biological replicates were analyzed, with each replicate consisting of a pooled sample of 100 insects. Due to differences in the physicochemical properties of the target metabolites, sample preparation and assay procedures were performed according to the respective manufacturer’s protocols. Quantification for all assays was based on standard curves generated using the standards provided with each kit.

The sorbitol content was quantified using a sorbitol assay kit (BC2525, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). For each treatment group, a pooled sample of 100 adults was weighed. The insects were homogenized in 1 mL of distilled water in a new microcentrifuge tube (using a sample-to-buffer ratio of approximately 1:10, w/v). The homogenate was boiled in a water bath for 10 min, cooled to room temperature, and centrifuged at 8000× g for 10 min at 25 °C. The supernatant was transferred to a new 1.5 mL microcentrifuge tube. For the assay, the kit-provided reagents were added to the assay tube according to the manufacturer’s instructions, followed by 230 µL of the sample supernatant. After thorough mixing, the mixture was incubated at room temperature for 15 min and then centrifuged again at 8000× g for 10 min at 25 °C. Then, 200 µL of the supernatant was transferred to a 96-well plate, and absorbance was measured at 655 nm using a microplate reader (Multiskan FC, Thermo Fisher Scientific, Waltham, MA, USA). The sorbitol concentration was determined based on a standard curve prepared with the kit-provided standard.

The trehalose content was determined using the Trehalose Content Assay Kit (BC0335, Beijing Solarbio Science & Technology Co., Ltd.), which is based on the anthrone-sulfuric acid method and includes an acidic extraction buffer for trehalose isolation. Working solution and standard (0.05 mg/mL) were prepared as instructed. For each treatment group, 100 adults were pooled and weighed. Insects were homogenized in 1 mL of extraction buffer at a 1:10 (w/v) ratio on ice. The homogenate was incubated at room temperature for 45 min with intermittent shaking (3–5 times), cooled, and centrifuged at 8000× g for 10 min at 25 °C. The supernatant was transferred to a new microcentrifuge tube. For the assay, 60 µL of the sample supernatant, standard, or distilled water was added to the assay tube, standard tube, and blank tube, respectively. Then, 50 µL of working solution and 190 µL of concentrated sulfuric acid were added sequentially to each tube. The mixtures were vortexed thoroughly, sealed, and heated in a 95 °C water bath for 10 min. After cooling, 200 µL of the reaction mixture was transferred to a 96-well plate. Absorbance was measured at 620 nm using a microplate reader (Multiskan FC, Thermo Fisher Scientific, USA), and the values were recorded.

For lipid determination, insects were rapidly ground into a fine powder in liquid nitrogen. The powder was transferred to a Soxhlet extractor (250 mL, YZB, Thermo Fisher Scientific, USA), and lipids were extracted by adding a chloroform–methanol mixture (2:1, v/v) as the solvent. The extraction was conducted in a 60 °C constant-temperature water bath (HWS-24, Yiheng Scientific Instrument Co., Ltd., Shanghai, China) for 8 h. After extraction, the solvent was removed using a distillation apparatus. When the distillation flask had cooled to room temperature, the total mass of the flask and extract was weighed. The lipid content (%) was calculated as (m₂ − m₁)/m × 100, where m₁ is the mass of the empty distillation flask (mg), m₂ is the mass of the flask plus extracted lipids (mg), and m is the mass of the sample before extraction (mg). A solvent blank was included to correct for potential interference. All procedures were conducted under anhydrous conditions to ensure complete solvent removal.

The glycogen content was measured using a Glycogen Assay Kit (BC0345, Beijing Solarbio Science & Technology Co., Ltd.). The assay involves glycogen hydrolysis in a strong alkaline extraction buffer, followed by quantification using the anthrone-sulfuric acid method. Approximately 0.1 g of insect tissue was homogenized directly in 0.75 mL of extraction buffer. The mixture was boiled in a water bath for 20 min and shaken thoroughly every 5 min during heating. After the water bath, when the sample was fully dissolved, the tube was cooled and then the volume was adjusted to 5 mL with distilled water. After thorough mixing, the mixture was centrifuged at 8000× g for 10 min at 25 °C. The supernatant was transferred to a new microcentrifuge tube for assay. For the assay, 60 µL of sample supernatant, standard, or distilled water was added to the assay tube, standard tube, and blank tube, respectively. Then, 50 µL of working solution and 190 µL of concentrated sulfuric acid were added sequentially to each tube. The mixtures were vortexed thoroughly and heated in a 95 °C water bath for 10 min. After cooling, 200 µL of the reaction mixture was transferred to a 96-well plate. Absorbance was measured at 620 nm using a microplate reader (Multiskan FC, Thermo Fisher Scientific, USA), and the absorbance values were recorded.

The glycerol content was determined using a Glycerol Assay Kit (BC6050, Beijing Solarbio Science & Technology Co., Ltd.). For each treatment group, a pooled sample of 100 adults was weighed and homogenized in extraction buffer on ice at a 1:10 (w/v) ratio. The homogenate was heated in a boiling water bath for 5 min, cooled to room temperature, and centrifuged at 12,000× g for 10 min at 25 °C. The supernatant was collected and kept on ice for assay. For the assay, 100 µL of the sample supernatant, standard, or distilled water was added to the assay tube, standard tube, or blank tube, respectively, followed by the sequential addition of 600 µL of Reagent One and 300 µL of working solution to each tube. The mixtures were vortexed thoroughly and incubated at 37 °C for precisely 15 min. Absorbance was then measured at 505 nm using a visible spectrophotometer (ND2000C, Thermo Fisher Scientific, USA) with a 1 mL glass cuvette, and the values were recorded. The glycerol content was calculated based on a standard curve.

2.4. RNA Extraction and Library Construction for Transcriptome Sequencing of C. lividipennis

Total RNA was extracted from C. lividipennis using the TRIzol reagent kit (Invitrogen, Carlsbad, CA, USA). RNA quality was assessed by agarose gel electrophoresis (DYY-8C, Beijing Liuyi Biotechnology Co., Ltd., Beijing, China) and quantified using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). RNA integrity was further validated using an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA); only samples with an RNA Integrity Number (RIN) ≥ 7.0 were used for library construction.

Poly(A)-tailed mRNA was enriched with Oligo(dT) magnetic beads and then randomly fragmented in fragmentation buffer at high temperature. Using the fragmented mRNA as a template and random oligonucleotides as primers, first-strand cDNA was synthesized with an M-MuLV reverse transcriptase system. The RNA strand was subsequently degraded with RNase H, and second-strand cDNA was synthesized using DNA polymerase I via a nick-translation reaction. The purified double-stranded cDNA was then subjected to end repair, A-tailing, and ligation of sequencing adapters. cDNA fragments of approximately 200 bp were selected with AMPure XP beads, followed by PCR amplification and purification to generate the final cDNA libraries. These libraries were sequenced on an Illumina NovaSeq X Plus platform (Gene Denovo Biotechnology Co., Ltd., Guangzhou, China) using a paired-end strategy.

2.5. Data Quality Control and Alignment

Raw sequencing data were processed with fastp (v0.18.0) [27] to remove adapters, reads with >10% unknown bases (N), and low-quality reads (>50% of bases with Q-score ≤ 20). The resulting clean reads were aligned to a ribosomal RNA database using Bowtie2 (v2.2.8) [28] to eliminate rRNA contamination. The remaining unmapped reads were then aligned to the C. lividipennis reference genome [29] using HISAT2 (v2.1.0) [30] with default parameters.

2.6. Transcript Assembly, Annotation, and Differential Expression Analysis

To optimize the existing genome annotation, transcripts were assembled and merged across all samples using StringTie (v1.3.1) [31], enabling the discovery of both known and novel transcripts. For comprehensive functional annotation, all assembled transcripts were searched against multiple public databases, including the NCBI non-redundant protein (Nr), Swiss-Prot, Pfam, and KOG databases using BLASTx (NCBI, https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 1 November 2023) with an E-value threshold of 1 × 10⁻⁵.

Gene expression levels were quantified as expected counts using RSEM [32]. Differential expression analysis was performed using DESeq2 [33], Genes with FDR < 0.05 and |log2FoldChange| ≥ 0.585 (1.5-fold change) were defined as differentially expressed genes (DEGs). This threshold was chosen to capture robust metabolic and regulatory responses under short-term heat stress. GO and KEGG enrichment analyses were performed via the Omicsmart platform (http://www.omicsmart.com, accessed on 1 December 2023).

Alternative splicing (AS) events were identified and quantified using rMATS (v4.1.2) in junction count (JC) mode. Five types of AS events, including skipped exon (SE), retained intron (RI), mutually exclusive exons (MXE), alternative 5′ splice site (A5SS), and alternative 3′ splice site (A3SS), were analyzed. Both known (annotated in the reference GTF) and novel AS events were identified. Significant differential AS events between treatments were determined based on an FDR < 0.05.

2.7. Phylogenetic Analysis of the HSP Family

To classify the HSP genes of C. lividipennis into subfamilies, candidate HSP amino acid sequences were combined with homologous sequences from other insects, including Drosophila melanogaster, Aphis gossypii, and Laodelphax striatellus. Homologous sequences from closely related species were included as outgroups for phylogenetic comparison. Multiple sequence alignment was performed using the MUSCLE algorithm in MEGA 11. A phylogenetic tree was then constructed with the neighbor-joining (NJ) method using the p-distance model for evolutionary distances and pairwise deletion for gap treatment. Tree robustness was evaluated with 1000 bootstrap replicates. The subfamily assignments of C. lividipennis HSP sequences were determined based on branch clustering patterns and bootstrap support values in the resulting phylogenetic tree.

2.8. Validation by Quantitative Real-Time Reverse Transcription PCR

To validate the reliability of the transcriptome data, key differentially expressed genes (DEGs) in C. lividipennis were selected for quantitative real-time reverse transcription PCR (qRT-PCR) analysis. First-strand cDNA was synthesized from 1.0 μg of total RNA using the HiScript III 1st Strand cDNA Synthesis Kit (+gDNA wiper; Vazyme Biotech Co., Ltd., Nanjing, China). Gene-specific primers were designed with Primer Premier 6.0 (Tm: 58 °C ± 2 °C; length: 18–25 bp; sequences listed in

Table 1. Primers for RT-qPCR.

Gene ID	Primer Sequence 5′-3′
β-actin	GCCCTCTTCCAACCCTCGTTCT GGCGTACAAGTCCTTCCTGATGTC
MSTRG.10015	CGTTACGGCTGGTTGGTTGTCA ACTCCTCTGTCCCTCGGCATTG
lividipennis05053	GGTGAGGATGAGGAAGAGGACGAA CCAGATGGGCTTGGTCTTGTTGAG
lividipennis08573	CCGTCGAAGGTGAACACGAAGAG CGGTGAGCGACAGGATTCCATC
lividipennis14108	CCTGCTCGGGCTATCTGGAACA TGAGTCTACGTCGGCGGTATGG
lividipennis07278	CGACGGCGTTCTCACTCTTCAG TGGAGAGCGGGTTGGTTGGTT
lividipennis07416	GCAACCTGATGGCACCACTAAGAA TGAACACAATGTCGGCTGGAATCT
lividipennis04442	TCCACCAACGGCGACACTCA TTTGGACGGCCCTGTTGTCTTTC

). Each qPCR reaction (20 μL) contained 1 μL of 10-fold diluted cDNA, 0.4 μL each of forward and reverse primers (10 μmol/L), 10 μL of 2× ChamQ Universal SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd.), and nuclease-free water to the final volume. Amplification was performed on a LightCycler^® 96 System (Roche, Basel, Switzerland) under the following conditions: 95 °C for 3 min; followed by 40 cycles of 95 °C for 20 s, 60 °C for 30 s, and 72 °C for 30 s. The β-actin gene was used as the internal reference, and relative expression levels of target genes were calculated using the 2^−△△Ct method.

2.9. Data Analysis and Experimental Instruments

The bioinformatic analysis of the transcriptome data, including quality assessment, genome alignment, transcript assembly, gene functional annotation, expression quantification, and differential expression analysis as described in Section 2.5 and Section 2.6, was carried out by Gene Denovo Biotechnology Co., Ltd. (Guangzhou, China) using the Omicsmart online platform. The specific software tools and analytical procedures employed (e.g., fastp, HISAT2, StringTie, DESeq2), along with their parameters, are detailed in the preceding sections.

For physiological and biochemical data, three independent biological replicates (n = 3) were analyzed per temperature treatment. Statistical analysis was performed using DPS software (v 9.50) [34]. Significant differences among temperature treatments for each metabolite were assessed by one-way analysis of variance (ANOVA), followed by Duncan’s new multiple range test (p < 0.05) for post hoc comparisons.

To integrate the physiological and transcriptomic results, genes involved in the metabolism of the target compounds (e.g., carbohydrate, polyol, and glycerolipid metabolism) were identified through GO and KEGG annotations. Short Time-series Expression Miner (STEM) analysis was then employed via the Omicsmart platform to identify significant gene expression trends across the temperature gradient. Genes exhibiting statistically significant expression patterns (p < 0.05) that were consistent with the observed metabolite concentration trends were selected for further investigation, including hierarchical clustering and functional enrichment analysis.

qRT-PCR data were analyzed using the 2^−ΔΔCt method to calculate relative expression levels; statistical analysis of inter-group differential expression (among Control, M, and H groups) was performed via one-way analysis of variance (ANOVA) followed by Bonferroni’s multiple comparison test (p < 0.05) using GraphPad Prism 8.3, which was also used to visualize the qRT-PCR results; image processing and final figure assembly were completed using Adobe Illustrator 2019.

3. Results

3.1. Changes in Physiological Parameters of C. lividipennis Under Short-Term High-Temperature Stress

We examined changes in physiological and biochemical parameters in C. lividipennis under short-term exposure to 26 °C, 33 °C, and 40 °C. As the temperature increased, the contents of sorbitol, trehalose, lipids, and glycogen increased significantly, and the glycerol content decreased significantly (Figure 1). Sorbitol and trehalose, both soluble sugars, showed clear temperature-dependent accumulation. The sorbitol content was 0.1687 ± 0.0066 mg/g at 26 °C, increasing significantly to 0.2117 ± 0.0017 mg/g at 33 °C and 0.2343 ± 0.0019 mg/g at 40 °C, with significant differences among all three groups (p < 0.05). A similar pattern was observed in the trehalose content, increasing from 1.3277 ± 0.0594 mg/g at 26 °C to 1.5670 ± 0.0615 mg/g at 33 °C and 1.7897 ± 0.0774 mg/g at 40 °C (p < 0.05).

Lipids and glycogen also increased with temperature. Under short-term heat stress, the lipid content rose significantly from 2.0926 ± 0.1725% at 26 °C to 3.2384 ± 0.1607% at 33 °C and 3.8408 ± 0.2359% at 40 °C (p < 0.05). The glycogen content likewise differed significantly among all temperature treatments, increasing from 33.3145 ± 2.3676 mg/g at 26 °C to 40.5510 ± 1.8450 mg/g at 33 °C and 47.7884 ± 1.3025 mg/g at 40 °C (p < 0.05). The glycerol content declined significantly with increasing temperature, decreasing from 951.5667 ± 10.7723 μmol/g at 26 °C to 489.2333 ± 26.0203 μmol/g at 40 °C (p < 0.05). These divergent physiological shifts, characterized by the accumulation of protective osmolytes and the adjustment of energy reserves, provide a critical biological context for investigating the underlying molecular mechanisms through transcriptomic profiling.

3.2. Quality Assessment of the Transcriptomic Data

To investigate the transcriptional responses of C. lividipennis to short-term heat stress (1 h), we established three temperature treatments: 26 °C as the control, 33 °C as the moderate temperature group (M), and 40 °C as the high-temperature group (H), each with three independent biological replicates (obtained from three separate experimental batches). The sequencing data generated in this study have been deposited in GenBase [35] at the National Genomics Data Center [36], Beijing Institute of Genomics, Chinese Academy of Sciences/China National Center for Bioinformation, under accession number CRA028009, and are publicly available at https://ngdc.cncb.ac.cn/gsa/ (data released on 18 July 2025).

Principal component analysis (PCA) was used to assess variation in gene expression among samples. The PCA results showed clear separation among the three treatment groups, with tight clustering of independent biological replicates within each group (Figure 2A). Pearson correlation coefficients for gene expression between samples ranged from 0.909 to 0.992 (Figure 2B), confirming high consistency across independent experimental runs and high experimental reproducibility. Quality control metrics demonstrated that all sequencing libraries were of high quality (

Table 2. Evaluation of the sequencing data.

Sample	Raw Data /Gb	Clean Data /Gb	Clean Reads	GC Content/%	Percent of Q20/%	Percent of Q30/%	Total Mapped/%
Control-1	5.94	5.83	37325078	44.32%	98.39%	95.22%	27,638,091 (74.05%)
Control-2	7.56	7.47	46624758	43.73%	98.94%	96.79%	36,025,040 (77.27%)
Control-3	9.84	9.75	61819400	44.13%	96.81%	91.30%	47,359,903 (76.61%)
M-1	7.51	7.39	47560804	44.69%	98.49%	95.55%	35,352,676 (74.33%)
M-2	7.02	6.92	43727268	44.65%	98.64%	95.92%	33,611,915 (76.87%)
M-3	8.67	8.52	54511084	44.42%	98.54%	95.70%	41,677,309 (76.46%)
H-1	8.07	7.93	48548664	43.51%	98.56%	95.77%	37,536,664 (77.32%)
H-2	8.74	8.58	53222180	44.08%	98.59%	95.88%	41,354,103 (77.70%)
H-3	7.17	7.05	44260312	44.56%	98.65%	95.97%	32,423,251 (73.26%)

Raw Data: Total number of bases in the raw sequencing data. Clean Data: Total number of high-quality bases after filtering. Clean Reads: Number of reads after rRNA removal, also known as valid reads. GC Content: GC percentage of bases in the filtered sequences. Percent of Q20: Number of bases with sequencing quality value ≥ Q20 and their percentage of Clean Data. Percent of Q30: Number of bases with sequencing quality value ≥ Q30 and their percentage of Clean Data. Total Mapped: Total number of reads mapped to the reference genome and their percentage of valid reads.

). The Control group yielded 145,769,236 clean reads, with the GC content ranging from 43.73% to 44.32% and Q30 scores between 91.30% and 96.79%. The M group produced 145,799,156 clean reads, with a GC content of 44.42–44.69% and Q30 scores of 95.55–95.92%. The H group generated 146,031,156 clean reads, with a GC content of 43.51–44.56% and Q30 scores of 95.77–95.97%. All samples had Q30 values above 91%, confirming the robust sequencing quality. Across all libraries, a total of 437,599,548 clean reads were obtained. Alignment rates to the reference genome ranged from 74.05% to 77.70%, indicating efficient mapping. Overall, the GC content ranged between 43.51% and 44.69%, and Q20 and Q30 scores exceeded 96.81% and 91.30%, respectively. Together, these metrics demonstrate the high integrity and reliability of the transcriptome sequencing data.

3.3. Analysis of DEGs

To characterize the dynamic transcriptional responses of C. lividipennis to short-term heat stress, we systematically analyzed DEGs across the three pairwise comparisons. As shown in Figure 3A (for details see Tables S1–S3), the number of DEGs varied markedly among comparisons: only 5 DEGs were detected in the Control vs. M group, all of which were up-regulated; 319 DEGs were identified in the Control vs. H group (156 up-regulated and 163 down-regulated); 98 DEGs were detected in the M vs. H group (33 up-regulated and 65 down-regulated). Notably, the five DEGs in the Control vs. M group were identified as four heat shock protein (HSP) genes (Hsp70Ba [represented by two transcripts: MSTRG.10015 and MSTRG.10016], and Hsp83 [represented by two transcripts: lividipennis02991 and lividipennis05053]) and one unannotated gene (MSTRG.1193). The Venn diagram (Figure 3B) illustrates the overlap of DEGs among the comparison groups. The Control vs. M group shared 4 DEGs with both the Control vs. H and M vs. H groups. A total of 84 DEGs were common to the Control vs. H and M vs. H groups (26 up-regulated and 58 down-regulated). All four DEGs shared by all three comparison groups were up-regulated. Collectively, these results indicate that the transcriptional response of C. lividipennis to high-temperature stress at 40 °C was markedly stronger than its response at 33 °C, with moderate heat stress inducing only a minimal, HSP-focused transcriptional change.

Hierarchical clustering analysis (Figure S1) revealed the transcriptional response patterns of C. lividipennis under the 26 °C (Control), 33 °C (M), and 40 °C (H) treatments. Biological replicates from the same temperature treatment clustered closely together, indicating high within-group consistency and suggesting that temperature was the primary factor driving variation in gene expression. Compared with the 26 °C control, the 40 °C treatment exhibited a denser distribution and larger amplitude of red/green signals in the heatmap, reflecting not only a larger number of upregulated and downregulated genes but also more pronounced shifts in fold-changes. In contrast, the 33 °C treatment showed weaker overall color variation and involved a narrower set of DEGs than the 40 °C group. These patterns indicate that the transcriptional response of C. lividipennis is amplified in a temperature-dependent manner, with both the intensity and breadth of the response at 40 °C clearly exceeding those at 33 °C.

3.4. GO Enrichment Analysis

To systematically investigate the response mechanisms of C. lividipennis under temperature stress, GO enrichment analysis was performed on DEGs identified under the three temperature treatments (26 °C, 33 °C, and 40 °C). A total of 221 DEGs were assigned to three main GO categories: Biological Process (BP), Molecular Function (MF), and Cellular Component (CC), which were further subdivided into 42 functional subcategories (Figure 4).

GO enrichment analysis revealed distinct changes in DEG enrichment profiles along the temperature gradient. In the M vs. H comparison, the range of enriched functional categories expanded and included Metabolic Process, Biological Regulation, Response to Stimulus, and Developmental Process, indicating activation of a wider array of adaptation-related functions under high-temperature stress.

In the cross-gradient Control vs. H comparison, the number of upregulated genes in each enriched category increased significantly, highlighting an intensified stress response. Beyond the categories shared with the previous comparisons, Catalytic Activity and Protein-containing Complex emerged as core enriched categories, highlighting reinforced adaptive mechanisms under extreme high temperature. This expansion of enriched functional categories potentially reflects the large-scale biological adjustments required to maintain cellular homeostasis as thermal stress intensifies.

3.5. KEGG Enrichment Analysis

To systematically elucidate the molecular response mechanisms of C. lividipennis under temperature stress, KEGG pathway enrichment analysis was performed on DEGs identified across the three temperature treatments (Figure 5). A total of 125 DEGs were mapped to 40 significantly enriched pathways, and the enrichment profiles shifted progressively along the temperature gradient.

In the M vs. H comparison, DEGs became newly enriched in Environmental Adaptation and the Global and Overview Maps pathway, indicating that high-temperature stress activates pathways related to stress defense and metabolic homeostasis.

In the Control vs. H comparison, the number of DEGs in these pathways increased markedly, and significant enrichment was observed in Energy Metabolism pathways. This pattern suggests that under extremely high temperatures, temperature signal transduction, protein homeostasis, stress defense, and metabolic coordination are all reinforced, with energy supply-related pathways additionally activated. Collectively, these pathways constitute the mature adaptive machinery that enables C. lividipennis to withstand high-temperature stress.

3.6. Expression and Analysis of Genes Related to Physiological Parameters in C. lividipennis Under Short-Term High-Temperature Stress

By integrating transcriptome data with trends in physiological parameters, we identified 1, 4, 2, 72, and 79 DEGs associated with sorbitol, trehalose, glycogen, glycerol, and lipid metabolism, respectively, whose expression patterns closely matched corresponding physiological changes (Figure 6). The coordination between gene expression and metabolite trends suggests that C. lividipennis actively modulates its metabolic pathways at the transcriptional level to support the observed physiological phenotypes. Among genes whose expression trends were consistent with the physiological indicators, sorbitol- and trehalose-related genes (e.g., Map2k4, tre1) were progressively up-regulated with temperature, mirroring increases in metabolite levels. Glycerol-related genes (e.g., AKR1B1, Gk) were down-regulated as temperature increased, consistent with the observed decline in the glycerol content. Lipid-related genes (e.g., Fasn, fabG) exhibited heterogeneous responses to high temperature overall, but their collective patterns were consistent with the increased lipid content. Glycogen-related genes (e.g., Pgm1, Mal-B1) showed no significant differences among temperature treatments, consistent with the absence of significant changes in the glycogen content. Together, these findings highlight the distinct responses of metabolic pathways to high-temperature stress, suggesting a strategic resource allocation to fuel the heat stress response.

KEGG enrichment analysis of genes associated with changes in sorbitol, trehalose, and glycogen (Figure 7A) showed that they were most strongly enriched in the “Starch and sucrose metabolism” pathway, as well as “Galactose metabolism.” As core components of carbohydrate metabolism, these pathways are directly involved in the synthesis and turnover of sorbitol, trehalose, and glycogen.

For genes associated with changes in glycerol (Figure 7B), KEGG analysis revealed significant enrichment in “Metabolic pathways” and “Glycerolipid metabolism.” The former provides the overall energetic and metabolic framework for glycerol utilization, whereas the latter directly mediates glycerolipid synthesis and interconversion. Enrichment was also detected in pathways such as “Glycosylphosphatidylinositol (GPI)-anchor biosynthesis” and “Inositol phosphate metabolism,” which are linked to processes including membrane remodeling and intracellular signal transduction.

Genes associated with changes in lipids (Figure 7C) were most significantly enriched in the “Non-alcoholic fatty liver disease” pathway, indicating that these genes are involved in lipid synthesis, transport, and maintenance of lipid homeostasis. Concurrent enrichment in “Oxidative phosphorylation” and “Diabetic cardiomyopathy” pathways highlights the tight coupling between lipid metabolism and energy production. In addition, enrichment in neurodegeneration-related pathways, such as “Parkinson’s disease” and “Alzheimer’s disease,” suggests that maintaining the lipid metabolic balance is an important component of the temperature-stress response.

3.7. Analysis of Alternative Splicing Changes in C. lividipennis Under Short-Term High Temperature

Alternative splicing analysis of the C. lividipennis transcriptome identified five types of alternative splicing events: skipped exon (SE), mutually exclusive exons (MXE), alternative 5′ splice site (A5SS), alternative 3′ splice site (A3SS), and retained intron (RI), with SE being the predominant form. Alignment to the reference genome revealed 1530 known alternative splicing events, and software prediction identified an additional 8820 novel events (Figure 8A).

We further compared alternative splicing profiles under the 26 °C (Control), 33 °C (M), and 40 °C (H) treatments (Figure 8B). SE events dominated in all temperature groups, with 7848, 7870, and 7847 events in Control, M, and H, respectively. MXE events numbered 874, 880, and 881 in Control, M, and H, respectively, showing a slight increasing trend; A5SS and A3SS event counts remained relatively stable across treatments, and RI events consistently occurred at low frequencies. These results indicate that short-term heat stress does not substantially alter the overall distribution of alternative splicing types but may modulate gene expression by subtly adjusting the frequency of specific events, such as MXE, in response to temperature stress. A Venn diagram integrating alternative splicing-related genes with DEGs (Figure 9) identified 112 overlapping genes, suggesting a direct functional connection between alternative splicing and differential gene expression under short-term high-temperature stress.

To further investigate the role of alternative splicing in metabolic regulation, we analyzed splicing patterns in genes associated with lipid, glycerol, sorbitol, glycogen, and trehalose metabolism that showed trends consistent with the physiological and biochemical data. Distinct distributions of metabolism-related genes were observed across splicing categories: SE events involved 1, 1, 7, 36, and 10 genes related to sorbitol, trehalose, glycogen, glycerol, and lipid metabolism, respectively; A5SS events were detected in 6 glycerol-metabolism genes; A3SS events involved 4 glycerol-metabolism genes and 1 lipid-metabolism gene; MXE events included 2 genes in glycogen metabolism, 4 in glycerol metabolism, and 2 in lipid metabolism; and RI events included 1 gene each in lipid and glycerol metabolism. Notably, one heat shock protein (HSP) gene was also identified within the MXE events, suggesting that alternative splicing may likewise participate in the post-transcriptional regulation of core heat stress response molecules (Figure 10, for details see Table S4).

3.8. Transcriptional Response and Validation of HSP Family Genes to Short-Term High Temperature in C. lividipennis

To investigate the effects of short-term high temperature on HSP gene expression in C. lividipennis, transcriptome data were used to screen differentially expressed HSP family genes and perform phylogenetic analysis. In the Control vs. M, M vs. H, and Control vs. H comparisons, 4, 12, and 15 differentially expressed HSP family genes were identified, respectively.

A maximum-likelihood phylogenetic tree constructed from the amino acid sequences of these 15 HSP genes (Figure 11) showed that they clustered into distinct HSP subfamilies: MSTRG.10015, MSTRG.10016, lividipennis05146, lividipennis07351, lividipennis04442, and lividipennis02663 grouped into the HSP70 family; lividipennis02865 into the HSP10 family; lividipennis07416 into the HSP40 family; lividipennis02991 and lividipennis05053 into the HSP90 family; and lividipennis08573, lividipennis07278, lividipennis00655, lividipennis13035, and lividipennis14108 into the small heat shock protein (sHSP) family.

Integrated analysis showed that both the number of differentially expressed HSP genes and the diversity of HSP subfamilies increased with rising temperature. Notably, MSTRG.10015, MSTRG.10016, lividipennis02991 and lividipennis05053 were differentially expressed in all three comparison groups, indicating that these genes may play central roles in the heat stress response of C. lividipennis.

To validate the reliability of the transcriptome data, seven highly expressed genes among the 15 candidate HSP genes were selected for RT-qPCR analysis, with β-actin as the reference gene. The expression patterns obtained by qPCR were highly consistent with the RNA-seq results (Figure 12). This validation reinforces the reliability of the transcriptomic findings and highlights the pivotal role of the HSP-mediated proteostasis network in thermal adaptation.

4. Discussion

Natural enemy insects are being exposed to increasingly severe heat stress under global warming. The ability of C. lividipennis, a key natural enemy of rice planthoppers, to survive short-term extreme high temperatures directly affects the stability of paddy ecosystems. In this study, we employed short-term (1 h) heat treatments to specifically mimic the acute midday and afternoon thermal peaks characteristic of heatwaves in South China’s rice-growing regions, such as Guangxi. It has been emphasized that research on the impacts of global warming on insect populations should prioritize extreme temperature events over gradual changes in mean temperature [37]. Under field conditions in these agroecosystems, extremely high-temperature events typically occur as transient peaks during the day [7]. Our preliminary experiments confirmed that continuous exposure to 40 °C for only 2 h resulted in 100% mortality in C. lividipennis. Therefore, investigating the rapid physiological and molecular responses under these short-term high-temperature conditions is essential for elucidating the resilience and survival mechanisms of this predatory natural enemy within its natural habitat.

The activities of insects are highly dependent on ambient temperature, and populations may face extinction under extreme heat if effective physiological and biochemical coping mechanisms are lacking [38]. Protective small molecules such as sorbitol, trehalose, and glycerol enhance thermotolerance by maintaining cellular osmotic pressure, stabilizing protein conformations, and reducing oxidative damage [39,40]. The sorbitol and trehalose contents in C. lividipennis were significantly higher under short-term exposure to 40 °C than under the 26 °C control (p < 0.05). Precisely, these soluble polyols act as compatible osmolytes that stabilize biological membranes and protein structures under thermal stress [41]. This pattern is consistent with findings in the silverleaf whitefly (Bemisia argentifolii), where sorbitol accumulation after exposure to 40 °C under high-sucrose feeding improved short-term survival at 46 °C [17], and in the oriental fruit fly (Bactrocera dorsalis), in which trehalose levels increased significantly under heat stress [42]. These findings further support the protective roles of these soluble polyols at high temperatures.

In contrast, the glycerol content in C. lividipennis decreased significantly with increasing temperature (p < 0.05), a pattern similar to that observed in adult Hylurgus ligniperda, where glycerol levels declined over time at 35 °C. This decline in glycerol content under increasing temperature indicates that it may serve as a metabolic substrate consumed to meet the high energy costs of heat stress responses [43], in contrast to its accumulation as a protectant during cold exposure [44,45]. Lipids and glycogen are core energy reserves in insects and are metabolized in a coordinated manner to maintain a stable ATP supply and support stress responses under high temperatures [46]. In C. lividipennis, the contents of both lipids and glycogen increased with treatment temperature, peaking at 40 °C, similar to the lipid peak observed in the bean bug (Riptortus pedestris) at 44 °C [47]. However, the “initial rise followed by decline” pattern of glycogen in C. lividipennis highlights interspecific differences in energy storage strategies under extreme heat. This accumulation of energy reserves during acute stress may reflect a strategic metabolic shift to support post-stress recovery.

The transcriptome data provided molecular support for these physiological changes. Among the 158 DEGs whose expression trends matched those of sorbitol, trehalose, glycogen, glycerol, and lipids, genes related to sorbitol, trehalose, and glycogen metabolism were significantly enriched in Starch and sucrose metabolism and Galactose metabolism pathways. This is consistent with findings in the lesser grain borer (Rhyzopertha dominica), which activates these pathways to promote trehalose synthesis under 40 °C heat stress [48], suggesting that these routes are broadly conserved in regulating soluble sugar synthesis at high temperatures in insects. Glycerol metabolism-related genes were mainly enriched in the “Glycerolipid metabolism” pathway and were overall down-regulated, similar to patterns in the silkworm (Glyphodes pyloalis) under heat stress, where this pathway was enriched and accompanied by a decline in the glycerol content [49]. This clarifies the regulatory mechanism underlying glycerol reduction in C. lividipennis at the molecular level. Lipid metabolism-related genes were significantly enriched in the “Oxidative phosphorylation” pathway, consistent with observations in ladybirds, where this pathway acts in concert with fatty acid metabolism under 40 °C heat stress [20]. This suggests that oxidative phosphorylation may participate in high-temperature energy regulation by enhancing mitochondrial respiratory chain activity. Overall, C. lividipennis appears to maintain energy homeostasis and osmotic balance under heat stress by coordinately modulating soluble sugar, glycerol, and lipid metabolic pathways. These regulatory patterns are highly consistent with known heat stress response mechanisms in other insects and enhance our understanding of the adaptation of insects to high temperatures.

Beyond these metabolite- and pathway-level adjustments, global GO and KEGG enrichment patterns further suggest that C. lividipennis may adopt a heat response strategy that is activated only beyond a critical temperature, shifting from minimal adjustment to comprehensive defense. The extremely limited transcriptional changes at 33 °C, centered on a few HSP genes, suggest that moderate short-term heat stress is primarily managed through post-translational or metabolic adjustments rather than de novo transcription. This aligns with findings in the parasitoid wasp Pteromalus puparum, where exposure to a similarly moderate temperature of 35 °C also resulted in limited transcriptional changes, primarily activating basal physiological pathways rather than a broad stress response [18]. As the temperature increases (40 °C), the scope of GO enrichment broadens to include functions such as Metabolic Process, Response to Stimulus, and Protein-containing Complex, while KEGG analyses reveal additional significant enrichment in pathways classified under Environmental Adaptation, Global and Overview Maps, and Energy Metabolism. Together, these changes demonstrate the activation of an active and coordinated defense mechanism only under severe heat stress. On the one hand, activation of Response to Stimulus and Environmental Adaptation pathways likely contributes to more specific defenses against high-temperature stress, whereas the upregulation of genes associated with Protein-containing Complex may help stabilize protein conformations and mitigate heat-induced protein denaturation and aggregation, which is considered a key basis for insect high-temperature tolerance. This view is supported by findings in the fall armyworm (Spodoptera frugiperda), in which GO terms such as Response to Stimulus, Protein Complex, and Metabolic Process are significantly enriched under high-temperature conditions [50]. On the other hand, enrichment of global metabolic regulation and Energy Metabolism pathways suggests coordinated regulation of detoxification metabolism and energy allocation, with ATP supplied for energy-demanding processes such as protein refolding and stress defense, thereby helping to reduce the risk of energy depletion under heat stress. This is consistent with the proposed mechanism in honeybees (Apis mellifera), where enhanced energy metabolism has been shown to contribute to improved heat tolerance [51]. Notably, the synergistic pattern of “metabolic coordination–protein homeostasis–energy supply” observed under 40 °C stress points to a species-specific adaptive strategy of C. lividipennis as a natural enemy insect, potentially enabling it to withstand high-temperature stress while maintaining its predatory capacity.

Among the molecular components underlying this coordinated heat response, heat shock proteins (HSPs) represent a central module. HSPs are highly conserved molecular chaperones in insect heat stress responses and are classified into subfamilies such as sHSPs, HSP40, HSP60, HSP70, HSP90, and HSP100 according to their molecular weight and function. They maintain intracellular proteostasis by mediating correct protein folding, preventing aberrant aggregation, and promoting the degradation of damaged proteins; their thermoprotective roles have been demonstrated in fruit flies, green mirid bugs, cotton bollworms, and other species [52,53,54]. Based on DEG screening and phylogenetic analysis of the transcriptome, 4, 12, and 15 differentially expressed HSP genes were identified in the Control vs. M, M vs. H, and Control vs. H comparison groups, respectively, which reflects a pronounced temperature-dependent increase. Phylogenetic analysis further revealed that these DEGs span core thermotolerance-related subfamilies, including HSP40, HSP70, HSP90, and sHSP. RT-qPCR validation of seven HSP-related genes confirmed that their expression levels increased with temperature, supporting the reliability of the transcriptome data.

This gradient response pattern has also been observed in other insects. For example, Zheng et al. (2022) reported that in the tobacco whitefly (Bemisia tabaci) exposed to 35 °C, 40 °C, and 45 °C for 1 h, the number of significantly up-regulated sHSP genes increased from 4 to 12 and that of HSP70 genes increased from 4 to 9, indicating that the number of HSP family members upregulated increased with temperature [55]. Similarly, Liu et al. (2025) found in the small hive beetle (Aethina tumida) that after 0.5 h of exposure at 38 °C, 42 °C, and 46 °C, the number of HSP DEGs increased in a temperature-dependent manner, with HSP70 significantly up-regulated across all heat treatments [56]. Collectively, these findings indicate that the gradient activation of HSP family genes represents a conserved strategy by which insects cope with escalating heat stress.

Alternative splicing, an important post-transcriptional mechanism, also plays a key role in insect heat adaptation. In this study, alternative splicing in C. lividipennis was dominated by SE events, and SE event counts were highest in heat-responsive metabolic pathways related to glycerol and lipids. This pattern is consistent with the findings of a previous study of the fruit fly Drosophila melanogaster, showing that SE events were the principal temperature-responsive splicing type across a 13–29 °C gradient [57]. Together, these findings support a conserved role of SEs in insect heat responses and will help clarify post-transcriptional regulatory mechanisms underlying heat adaptation in natural enemy insects.

Short-term high-temperature treatments were employed under controlled laboratory conditions, which are unable to fully reproduce field temperature fluctuations or ecological interactions. Future studies should incorporate ecological factors such as natural temperature gradients, stress duration, and predator–prey interactions to more comprehensively elucidate the heat adaptation mechanisms of C. lividipennis. Furthermore, while this study provides a foundational profile of alternative splicing at the gene level, future research will focus on in-depth differential expression analysis at the transcript level. This will include investigating the specific roles of alternatively spliced isoforms, particularly those involved in MXE and A5SS events that may regulate alternative transcription initiation sites, to provide a more granular understanding of post-transcriptional regulation under heat stress.

5. Conclusions

This study integrates physiological, biochemical, and transcriptomic analyses to characterize the stress response strategies of C. lividipennis under short-term extreme heat. At the physiological level, exposure to 40 °C significantly increased sorbitol, trehalose, lipid, and glycogen contents while decreasing glycerol, reflecting a systemic response involving osmoprotectant accumulation and energy reserve adjustment. At the molecular level, the stepwise increase in DEGs and the enrichment of pathways such as starch and sucrose metabolism and glycerolipid metabolism provide correlative molecular insights for the observed physiological changes. Additionally, the temperature-dependent expression of HSP genes and alternative splicing events highlight the transcriptional and post-transcriptional flexibility of the heat stress response. Collectively, this work advances our understanding of the thermal resilience mechanisms in this predator and provides a mechanistic basis for future studies aimed at its conservation and utilization in biological control within rice ecosystems under climate change.