A Biofilm-State Bacillus thuringiensis Formulation Drives Midgut Structural Disruption and Transcriptomic Reprogramming in Ectropis grisescens

TL;DR

A biofilm-state formulation of Bacillus thuringiensis increases insecticide effectiveness by causing gut damage and altering gene activity in a tea pest.

Contribution

Demonstrates that biofilm-state Bt, combined with specific inducers, significantly enhances insecticidal efficacy and alters host gut responses.

Findings

Biofilm-state Bt with composite inducers achieved 2.88-fold higher insecticidal efficacy than planktonic Bt.

Biofilm-state Bt caused severe midgut damage and triggered detoxification and stress-response pathways in Ectropis grisescens.

Findings suggest Bt physiological state is a key factor in formulation effectiveness.

Abstract

Bacillus thuringiensis (Bt) is one of the most extensively used microbial insecticides, attributed to the action of insecticidal crystal proteins (ICPs), primarily Cry toxins, which mediate damage to the insect midgut epithelium. Recent evidence suggests that Bt toxicity is also strongly influenced by its physiological state and interactions with the host gut environment. Biofilm formation represents an important adaptive strategy that enhances bacterial stress tolerance and may modulate insecticidal performance, although the underlying mechanisms remain unclear. However, it is still unclear how Bt in the biofilm state alters host responses at the structural and transcriptomic levels. Using the tea plantation pest Ectropis grisescens as a model, we systematically evaluated the insecticidal efficacy of biofilm-state Bt formulations and their synergistic effects with a biofilm inducer system composed of Tween-80, tea saponin, matrine, and tea polyphenols. Bioassays showed that the biofilm-state Bt supplemented with composite inducers achieved the highest corrected mortality and reduced the LC50 against neonate larvae by 2.88-fold compared with conventional planktonic Bt. Histopathological, biochemical, and transcriptomic analyses further revealed that biofilm-state Bt caused more severe midgut damage and induced extensive remodeling of detoxification- and stress-response-related pathways. These findings highlight Bt physiological state as a critical determinant of formulation efficacy and provide a novel framework for Bt optimization through microbial physiological regulation.

Figures

Full text

1. Introduction

Bacillus thuringiensis (Bt) is a Gram-positive entomopathogenic bacterium that represents not only a cornerstone of sustainable pest management in agriculture but also a classical model for studying interactions between pathogenic microorganisms and their hosts [1,2,3,4]. Despite its extensive application, the field performance of Bt formulations is often constrained by poor environmental stability and low colonization efficiency, which ultimately results in reduced virulence against late-instar larvae [5,6]. Traditional perspectives have largely focused on the insecticidal activity of Bt insecticidal crystal proteins (ICPs), while overlooking the physiological traits of Bt as a living microorganism. In natural environments, Bt does not exist exclusively in a free-living planktonic state [7]; rather, the complexity of its life cycle suggests that its pathogenic potential may depend on specific multicellular physiological forms [8].

Biofilm formation represents one of the primary survival strategies by which bacteria adapt to environmental stress [9,10,11,12,13]. Through the secretion of extracellular polymeric substances (EPSs), bacterial cells become embedded within structured communities, creating a physical barrier that confers protection against abiotic stresses such as ultraviolet radiation and desiccation [14,15,16]. More importantly, recent microbiological evidence indicates that the biofilm state may reshape pathogen-associated physiological processes by promoting gut colonization and enabling the sustained release of virulence factors [17]. Accordingly, the artificial induction of biofilm formation in Bt using exogenous natural inducers, such as plant secondary metabolites or surfactants, has been proposed as a potential strategy to enhance its insecticidal performance [18,19]. However, the biological consequences of this strategy at the host level remain poorly understood. In particular, it remains unclear how biofilm-induced physiological states of B. thuringiensis influence host midgut integrity, detoxification-related responses, and genome-wide transcriptional reprogramming following ingestion, and whether insect hosts perceive biofilm-state Bt as a distinct pathogenic stimulus compared with planktonic cells [20].

To address these questions, we selected Ectropis grisescens, a representative lepidopteran pest in tea plantation ecosystems in China, as a model insect [21,22]. From the novel perspective of “microbial physiological state–host pathological response”, this study aimed to elucidate the mechanisms underlying the enhanced efficacy of biofilm-induced Bt formulations. Rather than relying solely on conventional mortality assays, we integrated histopathological analysis to characterize midgut damage dynamics, enzymatic assays to monitor immune and detoxification metabolism, and comparative transcriptomics to uncover genome-wide remodeling of host stress-response networks [23,24].

Specifically, we (i) evaluated the regulatory effects of four natural inducers [25,26], including Tween-80 [19,27] and tea saponin [28,29], on Bt biofilm formation and virulence; (ii) characterized the cytological basis of aggravated midgut pathology caused by biofilm-state Bt; and (iii) identified key host response genes through transcriptomic analysis and further validated the gene silencing efficiency of selected candidate genes using RNA interference (RNAi), providing molecular evidence supporting their potential relevance.

2. Materials and Methods

2.1. Insect Rearing and Sample Preparation

The E. grisescens population used in this study was kindly provided by the Institute of Applied Ecology, Fujian Agriculture and Forestry University (Fuzhou, China). The insects were continuously reared for multiple generations in an artificial climate chamber under controlled conditions (photoperiod 16 h light/8 h dark, temperature 26 ± 1 °C, relative humidity 70 ± 5%). For bioassays, healthy larvae at uniform developmental stages were selected, including newly hatched first-instar larvae and fourth-instar larvae with no significant differences in body length or weight.

2.2. Bioassay Design for E. grisescens Larvae

The Bacillus thuringiensis (Bt) formulation—G033A WP (Wuhan Kenuo Biotechnology Co., Ltd., Wuhan, China) [30] was prepared at six concentrations (5, 2.5, 1.25, 0.625, 0.3125, and 0.1563 g/L) and supplemented with a composite biofilm inducer. The inducer components were added in proportions such that their final concentrations in the Bt–inducer mixture were consistent with those listed in Table 1. Two treatment modalities were established:

Planktonic treatment:

A total of 200 μL of the Bt–inducer mixture was evenly spread onto the surface of artificial diet, air-dried at room temperature, and then used to feed synchronized first-instar larvae.

Biofilm-state treatment:

The Bt–inducer mixture was incubated statically in 12-well cell culture plates at 30 °C for 48 h to induce biofilm formation. Subsequently, 200 μL of the biofilm preparation was coated onto the artificial diet prior to larval exposure, as previously described [31].

Negative controls consisted of Bt formulation prepared in the corresponding planktonic or biofilm state without the addition of inducers, while sterile distilled water was used as a blank control. All bioassays were conducted under controlled environmental conditions (26 ± 1 °C, 65 ± 5% relative humidity). Each treatment included nine biological replicates, with ten larvae per replicate. Larval mortality was recorded 24 h after exposure. Corrected mortality was calculated, and the median lethal concentration (LC_50_) was estimated. To improve clarity of the experimental workflow, a schematic illustration of the Bt formulation and biofilm induction, larval feeding bioassays, and subsequent midgut sampling and analyses is provided in Supplementary Figure S1.

2.3. Bioassays at Different Larval Instars and Time Points

Based on the LC_50_ value determined for newly hatched larvae (Section 2.2), the Bt formulation at this concentration was mixed with the composite biofilm inducer (Table 1) and applied to the diet following the procedures described above. To evaluate the independent and combined effects of the inducer and biofilm formation, four treatment groups were established:

FC group: Planktonic Bt formulation without inducer,

F group: Planktonic Bt formulation with inducer,

BC group: Biofilm-state Bt formulation without inducer,

B group: Biofilm-state Bt formulation with inducer.

Sterile distilled water served as a blank control. Larvae at the second, third, and fourth instars with uniform development were selected for bioassays. Each treatment consisted of eight biological replicates, with twelve larvae per replicate. Mortality was recorded at 12, 24, and 36 h post-treatment, and corrected mortality rates were calculated using Abbott’s formula [32].

2.4. Histopathological Observation

To assess histopathological alterations in the midgut of E. grisescens, fourth-instar larvae from different treatment groups (blank control, planktonic Bt, planktonic Bt + inducer, biofilm-state Bt, and biofilm-state Bt + inducer) were dissected at 12, 24, and 36 h post-treatment (n = 3 per time point). Midgut tissues were collected, rinsed with phosphate-buffered saline (PBS), and fixed in 4% paraformaldehyde. Samples were subjected to routine paraffin embedding procedures, including dehydration, clearing, and paraffin infiltration. Serial sections (7 μm) were prepared using a rotary microtome and stained with hematoxylin and eosin (H&E). The stained sections were dehydrated, cleared, mounted with neutral balsam, and scanned using a digital slide scanner for subsequent pathological analysis [33].

2.5. Determination of Immune-Related Enzyme Activities

Midgut tissues were collected from fourth-instar larvae subjected to different treatments (blank control, planktonic Bt, planktonic Bt + inducer, biofilm-state Bt, and biofilm-state Bt + inducer) at 12, 24, and 36 h post-treatment. Each treatment included three biological replicates, with ten larvae per replicate. Samples were immediately frozen in liquid nitrogen and stored at −80 °C until analysis.

Enzyme activities were determined using commercial assay kits following the manufacturers’ instructions.

2.5.1. Carboxylesterase (CarE) Activity

Samples were removed from −80 °C storage and thawed on ice. Carboxylesterase (CarE) activity was assayed using a commercial CarE assay kit (Fujian Herui Biotechnology Co., Ltd., Fuzhou, China) according to the manufacturer’s instructions. Samples were homogenized on ice with assay buffer at a ratio of 1:10 (w/v) and centrifuged at 12,000× g for 30 min at 4 °C. The supernatants were collected and kept on ice for analysis.

Prior to measurement, the working solution was equilibrated at 37 °C for 30 min, and the microplate reader was preheated and set to a wavelength of 450 nm. For the blank, 5 μL of distilled water and 200 μL of working solution were mixed, and absorbance values at 10 s and 190 s were recorded as A1 and A2, respectively. For the sample reaction, 5 μL of supernatant and 200 μL of working solution were mixed, and absorbance values at 10 s and 190 s were recorded as A3 and A4, respectively.

CarE activity was calculated using the following equation:

$$[eqn]$$

2.5.2. Glutathione S-Transferase (GST) Activity

Samples were removed from −80 °C storage and thawed on ice. Glutathione S-transferase (GST) activity was determined using a commercial GST assay kit (Fujian Herui Biotechnology Co., Ltd., Fuzhou, China) according to the manufacturer’s instruc-tions. Samples were homogenized on ice with assay buffer at a ratio of 1:10 (w/v) and centrifuged at 8000× g for 10 min at 4 °C. The supernatants were collected and kept on ice for analysis.

Prior to measurement, the working solution was equilibrated at 25 °C for 30 min, and the microplate reader was preheated and set to a wavelength of 340 nm. The reac-tion mixture consisted of 20 μL of supernatant, 180 μL of reagent II, and 20 μL of reagent III. After rapid mixing, absorbance was recorded at 10 s (A1) and after 5 min (A2).

GST activity was calculated using the following equation:

$$[eqn]$$

2.5.3. Peroxidase (POD) Activity

Samples were removed from −80 °C storage and thawed on ice. Peroxidase (POD) activity was determined using a commercial POD assay kit (Fujian Herui Biotechnology Co., Ltd., Fuzhou, China) according to the manufacturer’s instructions. Samples were homogenized on ice with assay buffer at a ratio of 1:10 (w/v) and centrifuged at 8000× g for 10 min at 4 °C. The supernatants were collected and kept on ice for analysis.

Prior to measurement, the working solution was equilibrated at 25 °C for 30 min, and the microplate reader was preheated and set to a wavelength of 470 nm. The reac-tion mixture consisted of 10 μL of supernatant and 190 μL of reagent II. After rapid mixing, absorbance was recorded at 1 min (A1) and after 5 min (A2).

POD activity was calculated using the following equation:

$$[eqn]$$

2.6. Transcriptomic Analysis

Healthy fourth-instar larvae of E. grisescens with uniform developmental status were selected for transcriptomic analysis. Larvae were randomly assigned to four treatment groups: FC (planktonic Bt), F (planktonic Bt supplemented with a biofilm inducer), BC (biofilm-state Bt), and B (biofilm-state Bt supplemented with a biofilm inducer). The Bt wettable powder was applied at a concentration of 3 g/L, and the composition and proportion of the biofilm inducer followed those described in Table 1.

For the BC and B treatments, Bt suspensions were incubated statically at 30 °C for 48 h to allow biofilm formation. A leaf-dip bioassay was conducted by immersing fresh tea leaves in the corresponding treatment solutions for 30 s, followed by air-drying at room temperature before feeding to larvae. Prior to bioassay, larvae were starved for 24 h to ensure uniform feeding.

After 24 h of bioassay, larvae were anesthetized on ice until completely immobilized. Midgut dissection was performed on an RNase-free dissection platform under cold conditions. The head capsule and terminal abdominal segments were carefully removed using RNase-free forceps, and the midgut was gently pulled out through a small incision in the abdominal cuticle. Gut contents, Malpighian tubules, and attached fat body tissues were carefully removed. The isolated midgut tissues were gently rinsed with precooled RNase-free water to eliminate residual debris. Cleaned midguts were immediately transferred into RNAlater RNA stabilization solution for animal tissues and stored at 4 °C temporarily before long-term storage at −80 °C. Midguts from 20 larvae were pooled as one biological replicate, with three biological replicates per treatment.

Total RNA quality was assessed by RNA-specific agarose gel electrophoresis and an Agilent 2100 Bioanalyzer (Santa Clara, CA, USA) to evaluate RNA concentration, purity, and integrity. High-quality RNA samples were used for library preparation. Poly(A) mRNA was enriched using oligo(dT)-conjugated magnetic beads and fragmented into approximately 300 bp fragments. First-strand cDNA synthesis was performed using random hexamer primers, followed by second-strand cDNA synthesis (Fujian Herui Biotechnology Co., Ltd., Fuzhou, China). The resulting cDNA libraries were amplified by PCR and quality-checked using an Agilent 2100 Bioanalyzer. Qualified libraries were sequenced on the Illumina NovaSeq 6000 platform (San Diego, CA, USA) in paired-end mode.

Raw sequencing data were subjected to stringent quality filtering to obtain clean reads. After removal of rRNA sequences, clean reads were aligned to the E. grisescens reference genome using HISAT2. Differentially expressed genes (DEGs) were identified using edgeR (version 3.36.0) with the criteria |log_2_ fold change| > 0.75 and q < 0.05. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were subsequently performed. Selected DEGs were validated by RT-qPCR (Bio-Rad Laboratories, Hercules, CA, USA) using β-actin as the internal reference gene. Primers were designed using NCBI Primer-BLAST(http://www.ncbi.nlm.nih.gov/BLAST/, accessed on 12 July 2025) (Supplementary Table S1), and relative gene expression levels were calculated using the 2^−ΔΔCt^ method.

2.7. Spatiotemporal Expression Analysis

Samples representing different developmental stages of E. grisescens were collected, including eggs (2 days post-oviposition), first- to fifth-instar larvae (2 days post-ecdysis), pupae (2 days post-pupation), and 1-day-old adults of both sexes. All samples were immediately frozen in liquid nitrogen and stored at −80 °C. Three biological replicates were prepared for each developmental stage, consisting of 50 eggs per replicate, 20 individuals for the first and second instars, five individuals for the third instar, and pooled samples for later stages. Total RNA extraction, cDNA synthesis, and RT-qPCR analyses were performed as described above.

2.8. RNA Interference Assay

An RNA interference (RNAi) system was established to evaluate the silencing feasibility and validate the molecular responsiveness of candidate genes identified from transcriptomic analysis in E. grisescens. Gene-specific double-stranded RNAs (dsRNAs) were designed to target three candidate genes, including EgCYP6B2 (cytochrome P450 family 6 subfamily B member 2), EgArylα (arylphorin alpha subunit), and EgPGPEP1 (pyroglutamyl peptidase 1). Double-stranded RNA targeting the green fluorescent protein gene (dsGFP) was used as a negative control [34].

RNAi target sequences were designed based on the E. grisescens genome information. The length of each target fragment was approximately 300 bp, with GC content controlled between 45% and 55%. All candidate sequences were aligned against the E. grisescens genome using NCBI BLAST (http://www.ncbi.nlm.nih.gov/BLAST/, accessed on 23 October 2025) to exclude significant sequence homology and thereby minimize off-target effects. Target regions located within the 5′ and 3′ untranslated regions (UTRs) were avoided during dsRNA design. For in vitro transcription, T7 promoter sequences were added to both ends of each target fragment during primer design.

PCR amplification was performed using cDNA synthesized from E. grisescens larvae as the template and a high-fidelity DNA polymerase. The amplified products were verified by electrophoresis on a 1.5% agarose gel, excised, and purified using a commercial gel extraction kit (Fujian Herui Biotechnology Co., Ltd., Fuzhou, China) according to the manufacturer’s instructions. The purified PCR products were then used as templates for dsRNA synthesis [35].

Double-stranded RNA (dsRNA) was synthesized using the T7 RiboMAX^TM^ Express RNAi System (Promega Corporation, Madison, WI, USA) following the manufacturer’s protocol. Transcription reactions were incubated at 37 °C for 6 h, followed by heat treatment at 70 °C for 10 min and gradual cooling to room temperature to allow dsRNA annealing. After transcription, RNase and DNase were added to remove residual single-stranded RNA and DNA templates, ensuring the integrity of the double-stranded RNA.

The synthesized dsRNA was purified by chloroform substitute/isoamyl alcohol extraction, followed by precipitation with sodium acetate and absolute ethanol. The pellet was washed with 75% ethanol, air-dried, and dissolved in RNase-free H_2_O. dsRNA integrity was assessed by electrophoresis on a 1% agarose gel, and dsRNA concentration was determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA, USA).

Third-instar E. grisescens larvae with uniform growth status were selected and starved for 12 h prior to injection. Each larva was injected abdominally with 1 μL of dsRNA solution at a concentration of 1500 ng/μL using a microsyringe. Larvae injected with dsGFP at the same concentration served as the negative control. Samples were collected at 12 h, 24 h, and 48 h post-injection, with five larvae pooled as one biological replicate and three replicates per treatment. Collected larvae were immediately frozen in liquid nitrogen and stored at −80 °C until further analysis.

Total RNA was extracted from larvae using the TRIzol method and reverse-transcribed into cDNA for quantitative real-time PCR (RT-qPCR) analysis. RNAi efficiency was evaluated by measuring the relative expression levels of the target genes, with β-actin used as the internal reference gene. RT-qPCR primers were designed to avoid the dsRNA-targeted regions. Relative gene expression levels were calculated using the 2^−ΔΔCt^ method. The designed primer sequences are shown in Table S2.

2.9. Statistical Analysis

All experiments were performed with at least three independent biological replicates. Data were analyzed using one-way or two-way analysis of variance (ANOVA) in GraphPad Prism 10, followed by Tukey’s multiple comparison test. Differences were considered statistically significant at p < 0.05. Statistical significance is indicated by different lowercase letters or asterisks (* p < 0.05; ** p < 0.01; *** p < 0.001).

3. Results

3.1. Synergistic Promotion of Insecticidal Activity of Bt Formulation by Biofilm Induction

3.1.1. Bioassay of E. grisescens Larvae

The bioassay of neonate larvae using the film-coated artificial diet method, as summarized in Table 2, classified E. grisescens into four treatment groups. The results demonstrated that the LC_50_ value of planktonic Bt formulation (hereafter referred to as Bt) was 3.005 g/L. However, when a biofilm induction was added, the LC_50_ significantly decreased to 1.411 g/L, yielding a synergy ratio of 2.129. In contrast, the LC_50_ of biofilm-state Bt was 2.514 g/L. After treatment with the biofilm induction, the LC_50_ was further reduced to 1.044 g/L, corresponding to a synergy ratio of 2.878. The regression models for all treatment groups exhibited R^2^ values greater than 0.95, indicating a good fit of the dose–response relationships. These findings suggest that both the combination of Bt formulation with biofilm inductions and the cultivation of Bt in biofilm state enhanced its insecticidal activity against E. grisescens. Notably, the highest promotive effect on Bt toxicity was observed when the Bt was supplemented with the biofilm induction and subsequently cultured into the biofilm state.

Based on the comparison of LC_50_ values and synergy ratios, the biofilm inducer significantly enhanced the insecticidal activity of Bt under both planktonic and biofilm-state conditions. Notably, the promotive effect was substantially stronger when Bt was first cultured into the biofilm state and subsequently supplemented with the inducer. In contrast, the transition from planktonic to biofilm-state Bt alone resulted in only a moderate increase in toxicity. These observations suggest that the inducer does not simply act as an additive toxic factor, but rather amplifies Bt pathogenicity by modulating its physiological state and its interaction with the host midgut, thereby leading to enhanced lethality.

3.1.2. Bioassay of E. grisescens at Different Larval Instars and Time Points

To further evaluate the impact of different Bt formulation treatments on the stage-specific toxicity to E. grisescens larvae, the corrected mortality of second-, third-, and fourth-instar larvae at 12 h, 24 h, and 36 h was analyzed through bioassays. The results are shown in Figure S1 and Figure 1A. The corrected mortality across all larval instars exhibited a consistent trend, with insecticidal efficacy ranked as Group B > Group F > Group BC > Group FC (Bt + biofilm induction biofilm state > Bt + biofilm induction planktonic state > Bt biofilm state > Bt planktonic state). This trend was also consistent with the bioassay results of newly hatched larvae (Table 2).

For second-instar larvae, the differences among treatments were most pronounced at 12 h, and these differences diminished over time. Group B consistently showed the highest mortality rates (72.92–97.92%) across all time points (Figure 1(Aa)). For third-instar larvae, significant differences among treatments were observed at both 12 h and 24 h, while the differences narrowed by 36 h. Group B still exhibited the highest mortality, reaching 92.22% at 36 h (Figure 1(Ab)). For fourth-instar larvae, at 12 h, Group B showed a significantly higher mortality rate (62.12%) compared to the other groups, while no significant differences were observed among the remaining treatments. The differences among groups increased at 24 h and further expanded at 36 h. Both the biofilm state (46.97%) and the addition of the biofilm induction (Group F: 57.07%, Group B: 62.12%) resulted in markedly higher mortality compared to the pure Bt group (25.76%) (Figure 1(Ac)).

Larval instar-dependent differences in susceptibility to Bt formulations were clearly observed, with younger larvae showing higher sensitivity and older larvae exhibiting reduced mortality, consistent with the general enhancement of cuticular barriers and detoxification capacity during insect development. Nevertheless, even in fourth-instar larvae, treatment with biofilm-state Bt supplemented with the inducer resulted in significantly higher mortality compared with the planktonic Bt treatment. Moreover, this difference became more pronounced with prolonged exposure. These results indicate that the biofilm induction strategy partially overcomes the increased defensive capacity of older larvae, allowing Bt to maintain elevated insecticidal efficacy under physiologically less favorable conditions.

3.2. Histopathological Observation of Midgut of E. grisescens

To elucidate whether the enhanced insecticidal activity observed in different Bt formulations was associated with differential midgut damage, histopathological analyses of the midgut of E. grisescens larvae were conducted. As ICPs exert their lethal effects primarily through disruption of midgut epithelial integrity, comparative examination of midgut histopathology provides critical insight into the tissue-level basis underlying the observed differences in toxicity. Midgut tissues were therefore collected at 12, 24, and 36 h after treatment with planktonic Bt, planktonic Bt supplemented with the biofilm inducer, biofilm-formed Bt, and biofilm-formed Bt supplemented with the inducer, and subjected to paraffin sectioning and hematoxylin–eosin (H&E) staining.

As shown in Figure 1(Ba), the midgut of control larvae exhibited a normal histological appearance, with an intact epithelial layer, clearly identifiable goblet cells (GCs), and well-organized microvilli (Ms). The nuclei (Ns) of epithelial cells were regularly shaped and evenly distributed, indicating a stable cellular architecture.

At 12 h post-treatment, all Bt-treated groups exhibited early histopathological alterations, with changes most prominently observed in goblet cells. In these groups, GCs displayed varying degrees of deformation and irregular morphology [36] (Figure 1(Bb–e)). suggesting an early cellular response to Bt exposure. In treatments supplemented with the biofilm inducer, microvilli became partially disorganized and less densely packed (Figure 1(Bc,e)), indicating an early compromise of the epithelial surface. In addition, epithelial cells in larvae treated with biofilm-formed Bt showed increased cytoplasmic vacuolization, while nuclei remained visible but appeared less uniformly arranged, reflecting the onset of cellular stress.

By 24 h post-treatment, histopathological damage became more pronounced across all Bt-treated groups. Goblet cells exhibited further deformation and irregular distribution (Figure 1(Bg–j)), and the epithelial layer appeared progressively loosened. Microvilli were markedly reduced in continuity and density, indicating substantial impairment of the absorptive surface. Although nuclei were still detectable in most epithelial cells, their spatial organization was clearly disturbed compared with the control group, suggesting progressive loss of cellular structural integrity [37].

At 36 h post-treatment (Figure 1(Bk–o)), severe midgut damage was evident in all Bt-treated larvae. Microvilli were largely absent, and goblet cell structures were extensively disrupted or no longer clearly distinguishable. The epithelial layer showed extensive disorganization, accompanied by pronounced cellular lysis. Notably, biofilm-formed Bt treatments, particularly when combined with the biofilm inducer, resulted in markedly more extensive tissue damage than planktonic Bt treatments. In the biofilm-formed Bt plus inducer group, widespread epithelial collapse was observed, together with nuclear condensation and fragmentation and abundant cellular debris within the gut lumen (Figure 1(Bo)), indicating advanced cellular destruction. Overall, these observations demonstrate that biofilm-state Bt, especially in combination with the inducer, accelerates and intensifies midgut damage compared with planktonic Bt.

3.3. Impact of Bt Formulation Combined with Biofilm Inducers on the Immune Enzyme Activities of E. grisescens

3.3.1. Dynamic Regulation of Carboxylesterase (CarE) Activity

To investigate the dynamic regulatory effects of the biofilm inducer and Bt formulations on carboxylesterase (CarE) activity in E. grisescens, CarE activities were measured in the FC, F, BC, and B groups at 12, 24, and 36 h post-treatment. As shown in Figure 2A and Figure S2, CarE activity in the F and B groups reached the highest levels at 12 h, with values of 9.49 and 10.35 U/g, respectively, both exceeding the peak activities observed in the FC (8.00 U/g) and BC (8.13 U/g) groups at 24 h.

At 12 h, CarE activity in the B group reached 10.23 U/g, which was significantly higher than that in the FC group (4.22 U/g) and the BC group (5.88 U/g). At the same time point, the F group also exhibited elevated CarE activity (9.50 U/g), significantly exceeding that of the FC and BC groups. By 24 h, CarE activity in the FC and BC groups increased to 8.00 and 8.13 U/g, respectively, which were significantly higher than the corresponding values in the F (5.47 U/g) and B (6.67 U/g) groups. At 36 h, CarE activity in the B group further declined to 4.19 U/g and was significantly lower than that in the F group (5.65 U/g).

Overall, the addition of the biofilm inducer markedly accelerated and intensified the early CarE response, whereas prolonged exposure resulted in a decline in enzyme activity. Notably, the B group exhibited clear suppression of CarE activity at 36 h, indicating that the interaction between the biofilm-state Bt state and the inducer induces a non-linear, time-dependent modulation of detoxification enzyme activity.

3.3.2. Sustained Inductive Effect on Glutathione S-Transferase (GST) Activity

To elucidate the dynamic regulatory effects of the biofilm inducer and Bt formulations on glutathione S-transferase (GST) activity in E. grisescens, GST activities were measured in the FC, F, BC, and B groups at 12, 24, and 36 h post-treatment. As shown in Figure 2B and Figure S3, GST activity in all treatment groups increased with prolonged exposure over the 36 h period. Among them, the B group consistently exhibited higher GST activity than the other groups at all time points (24 h: 12.50 nmol/s/g vs. 7.70–9.16 nmol/s/g; 36 h: 15.25 nmol/s/g vs. 11.46–13.37 nmol/s/g).

From 12 h to 36 h, GST activity increased by 62.2% in the FC group, while the increases in the B and BC groups reached 60.3% and 44.8%, respectively. Overall, GST activity increased progressively with exposure time across all Bt treatments, with the B group consistently exhibiting the highest activity levels, indicating that biofilm-state Bt combined with the inducer elicits a more sustained GST-related detoxification response. This trend is consistent with previous reports of GST induction following Bt infection [38].

3.3.3. Temporal Regulation of Peroxidase (POD) Activity

To clarify the dynamic regulatory effects of the biofilm inducer and Bt formulations on peroxidase (POD) activity in E. grisescens, POD activities in the FC, F, BC, and B groups were measured at 12, 24, and 36 h post-treatment. As shown in Figure 2C and Figure S4, POD activity in all treatment groups exhibited a temporal pattern characterized by an initial increase followed by a subsequent decline.

At 12 h, POD activity in the B group reached the highest level (21.29 U/g), which was significantly higher than that in all other treatment groups at the same time point. By 36 h, POD activity in the B group declined sharply to 3.61 U/g, which was significantly lower than that in the FC and BC groups, but not significantly different from that in the F group or the control group. At 24 h, both the FC and F groups maintained relatively high POD activity levels (19.54 U/g and 19.40 U/g, respectively), with the F group showing significantly higher activity than the B group, while no significant difference was observed between the FC and BC groups. However, by 36 h, POD activity in the F group decreased significantly to 10.17 U/g, which was lower than that in the FC group (17.36 U/g).

At 24 h, POD activity in the BC group was 12.60 U/g, which was numerically higher than that in the B group (11.76 U/g), although the difference was not statistically significant. Intergroup comparisons further indicated that POD activity in the B group was significantly higher than that in all other groups at 12 h, significantly lower than that in the FC and F groups at 24 h, and significantly lower than that in the FC and BC groups at 36 h.

Overall, the combination of the biofilm inducer with biofilm-state Bt induced a rapid but transient increase in POD activity, followed by a pronounced suppression at later stages. In contrast, Bt treatment alone elicited a more sustained POD activity response over time.

3.4. Transcriptomic Analysis of Bt Formulation Combined with Biofilm Induction

3.4.1. Principal Component Analysis (PCA)

Clear separation among treatment groups was observed at the transcriptomic level (Figure 3A). Consistent clustering patterns and data robustness were further supported by additional analyses provided in the Supplementary Materials (Figure S5 and Tables S3 and S4), with the results clearly indicating that in the experiments with E. grisescens, when the insects fed on the Bt and the Bt-compounded biofilm induction, the gene sample groups of the midgut tissues were separated from each other. Similarly, under the biofilm state, the gene sample groups of the midgut tissues of E. grisescens fed with the Bt and the Bt-compounded biofilm induction also showed mutual separation. This phenomenon reveals that gene expression regulation mechanisms differ between the two sample groups. These differences may involve the activation of distinct signaling pathways and the regulation of specific transcription factors, ultimately resulting in significant variations in gene expression profiles (i.e., the relative levels and types of genes expressed).

3.4.2. Analysis of Differentially Expressed Genes

To control the false positive rate, differentially expressed genes (DEGs) were identified based on a combination of q value and log_2_(Fold Change), using the criteria of |log_2_(Fold Change)| > 0.75 and q value < 0.05. The overall distribution of DEGs between sample groups was visualized using volcano plots, in which red dots represent up-regulated genes, blue dots represent down-regulated genes, and gray dots represent genes that did not meet the filtering thresholds.

Consistent with the clear group separation observed in the PCA, substantial transcriptional differences were detected between treatment groups (Figure 3B; Table 3). In the FC vs. F comparison, a total of 370 DEGs were identified in the midgut tissue, among which 258 genes were up-regulated and 112 genes were down-regulated in the FC group relative to the F group. In contrast, under biofilm-state conditions (BC vs. B), 333 DEGs were identified, with 73 genes up-regulated and 260 genes down-regulated in the BC group compared with the B group.

Notably, the BC vs. B comparison was characterized by a predominance of down-regulated genes, and the magnitude of down-regulation was generally greater than that observed in the FC vs. F comparison, indicating a more extensive transcriptional repression associated with biofilm induction under biofilm-state Bt exposure.

Venn diagram analysis revealed 128 differentially expressed genes (DEGs) shared between the FC vs. F and BC vs. B comparisons (Figure S6). These shared DEGs were identified under both planktonic and biofilm-state conditions in response to the addition of the biofilm inducer, indicating that a subset of genes exhibited consistent transcriptional responses regardless of the physiological state of Bt. This overlap suggests that biofilm induction triggers a core transcriptional response in the midgut of E. grisescens following Bt exposure.

3.4.3. GO Enrichment Analysis of Differentially Expressed Genes

Gene Ontology (GO) annotation was performed by comparing differentially expressed genes between the two sample groups against the GO database. GO functional classification includes three major categories: molecular function, cellular component, and biological process.

The results (Table 4) showed that differentially expressed genes between the FC and F groups were annotated to a total of 2239 GO terms, including 136 cellular component terms, 29 molecular function terms, and 48 biological process terms.

In contrast, Gene Ontology enrichment analysis of the FC vs. F comparison revealed a distinct functional profile. Differentially expressed genes were primarily enriched in GO terms related to pattern-recognition-receptor-mediated signaling, detection of external biotic stimuli, innate immune response activation, and regulation of chemotactic responses (Figure 3C).

Enriched terms such as cell surface pattern recognition receptor signaling pathway, innate immune-response-activating cell surface receptor signaling pathway, detection of biotic stimulus, and response to molecules of fungal origin indicate that the addition of the inducer under planktonic Bt conditions mainly enhanced microbial recognition and immune sensing in the midgut. This transcriptional profile reflects a defense-oriented and relatively regulated immune response, rather than widespread oxidative or structural stress [39].

Consistent with this interpretation, GO terms associated with detoxification-related functions and metal ion transport (e.g., iron and zinc ion handling) were moderately enriched, corresponding to the sustained induction of GST and CarE activities observed under planktonic conditions. Importantly, the absence of strong enrichment in oxidative stress- or membrane-disruption-related GO categories aligns with the relatively intact midgut architecture and lower mortality observed in these treatments.

Overall, the GO enrichment patterns clearly distinguish two modes of transcriptional response: an adaptive immune-recognition-dominated response under planktonic Bt exposure, and a stress-driven, membrane-disruptive response under biofilm-state Bt exposure, providing a mechanistic framework linking gene expression changes with enzymatic activity dynamics, midgut pathology, and insect mortality.

Under biofilm-state treatment conditions, differentially expressed genes between the BC and B groups were annotated to 2159 Gene Ontology (GO) terms, including 170 cellular component (CC) terms, 342 molecular function (MF) terms, and 1647 biological process (BP) terms (Table 4). In the comparison between the BC and B groups, Gene Ontology enrichment analysis revealed that differentially expressed genes were predominantly associated with oxidative stress regulation, membrane-related cellular components, immune-associated recognition processes, and polysaccharide-binding functions (Figure 3C).

Specifically, enriched biological process terms such as oxidoreductase activity acting on peroxide as an acceptor, cytokine-related activity, and polysaccharide or β-glucan binding indicate that midgut tissues experienced intensified oxidative stress and immune stimulation following exposure to biofilm-state Bt combined with the inducer. The enrichment of peroxide-related oxidoreductase activity provides a direct molecular explanation for the pronounced dynamics of peroxidase (POD) activity observed in physiological assays, characterized by an early sharp increase followed by a marked decline, reflecting a transition from antioxidant induction to functional exhaustion.

At the cellular component level, significantly enriched terms included microvesicle, dense core granule membrane, and cell surface, all of which are closely linked to epithelial membrane integrity, secretion, and transport functions in the midgut. These transcriptional alterations are highly consistent with histopathological observations showing epithelial vacuolization, cellular disintegration, and loss of microvilli, providing molecular evidence for extensive disruption of the midgut membrane system.

In addition, enrichment of polysaccharide-binding and immune-related molecular functions suggests enhanced pattern recognition of biofilm-state extracellular matrix components. However, under conditions of severe epithelial damage and metabolic imbalance, such immune activation is unlikely to confer effective protection and may instead exacerbate physiological stress, thereby contributing to the markedly increased mortality observed in larvae exposed to biofilm-state Bt.

3.4.4. KEGG Enrichment Analysis of Differentially Expressed Genes

KEGG pathway enrichment analysis revealed that differentially expressed genes between the FC and F groups were mainly enriched in pathways related to metabolic regulation, nutrient absorption, signal transduction, and transmembrane transport (Figure 4 and Table 5). Prominent pathways included the PPAR signaling pathway, fat digestion and absorption, insulin resistance, and several lipid-metabolism-related pathways.

The enrichment of the PPAR signaling pathway and fat digestion and absorption suggests that, under planktonic Bt conditions, the addition of the inducer primarily triggers metabolic adjustment and energy redistribution in the midgut. Such transcriptional changes are indicative of adaptive metabolic regulation aimed at maintaining intestinal homeostasis rather than causing overt tissue damage. This interpretation is consistent with the relatively intact midgut morphology and the absence of pronounced epithelial disintegration observed in histopathological analyses.

Notably, transporter-related pathways showed prominent enrichment, exhibiting the highest gene counts among the identified KEGG categories, indicating enhanced transmembrane transport activity. These pathways may contribute to toxin efflux, metabolite exchange, and ionic balance maintenance, thereby buffering Bt-induced stress. This transcriptional signature corresponds well with the sustained induction of detoxification-related enzymes, particularly GST and CarE, under planktonic conditions, suggesting effective coordination between gene expression regulation and enzymatic detoxification capacity.

Additional pathways, including quorum sensing, sulfur metabolism, and the sulfur relay system, were also enriched. These pathways likely reflect host responses to bacterial signaling molecules or metabolic by-products; however, their moderate enrichment indicates that such responses do not escalate into severe oxidative stress or irreversible cytotoxicity. Although pathways associated with cell damage, such as ferroptosis, were detected, their relatively limited enrichment further supports the notion that planktonic Bt combined with the inducer does not strongly activate terminal cell death pathways.

Collectively, KEGG enrichment analysis of the FC vs. F comparison indicates that the inducer under planktonic Bt exposure mainly elicits adaptive metabolic and transport-related responses that support intestinal function and stress tolerance, rather than promoting extensive epithelial damage or systemic physiological collapse.

In contrast, KEGG pathway enrichment analysis of the BC vs. B comparison revealed significant enrichment of pathways associated with lipid metabolism, oxidative stress regulation, immune signaling, and metabolic homeostasis (Figure 4 and Table S5). These enrichment patterns indicate that the addition of the inducer under biofilm-state Bt conditions drives the midgut from localized stress responses toward a state of systemic physiological dysregulation.

Lipid-metabolism-related pathways, including glycerophospholipid metabolism, glycerolipid metabolism, ether lipid metabolism, and cutin, suberin and wax biosynthesis, were prominently enriched. These lipid classes are essential structural components of epithelial cell membranes and microvilli. Dysregulation of these pathways therefore provides a direct molecular explanation for the extensive epithelial disintegration and microvillar loss observed in histopathological analyses, which are key determinants of enhanced Bt toxicity.

The significant enrichment of the peroxisome pathway further highlights profound remodeling of reactive oxygen species (ROS) metabolism. Given the central role of peroxisomes in lipid oxidation and ROS detoxification, disruption of this pathway is consistent with excessive lipid peroxidation and secondary membrane damage. Importantly, this transcriptional signature aligns closely with the physiological dynamics of POD activity, characterized by an early induction followed by a marked decline, indicating that prolonged oxidative stress ultimately overwhelms the antioxidant defense system.

In addition, enrichment of immune-related pathways such as the Toll and Imd signaling pathway, together with amino sugar and nucleotide sugar metabolism, suggests intensified immune recognition of biofilm-state Bt. Considering the polysaccharide-rich extracellular matrix of biofilms, this immune activation is likely driven by enhanced pattern recognition. However, under conditions of severe epithelial damage and metabolic imbalance, sustained immune signaling may impose substantial energetic costs and further exacerbate physiological deterioration rather than confer effective protection.

Consistently, enrichment of pathways associated with insulin resistance and longevity regulation indicates profound disruption of metabolic control systems in the midgut. These transcriptional changes suggest that biofilm-state Bt combined with the inducer not only damages intestinal structure but also induces metabolic exhaustion, thereby providing a mechanistic explanation for the significantly elevated mortality observed in the B group.

3.5. Validation of the Reliability of Transcriptomic Data Expression Levels

The expression patterns of the selected genes obtained by RT-qPCR were generally consistent with those derived from RNA-Seq analysis (Figure 5). Among the validated genes, only Eg011054 exhibited an inconsistent expression trend in the FC vs. F comparison, whereas all other genes showed concordant expression patterns between the two methods.

The overall consistency between RT-qPCR results and transcriptomic data indicates that the RNA-Seq analysis reliably reflected gene expression changes under the experimental conditions.

3.6. Spatiotemporal Expression Analysis of Candidate Resistance-Associated Genes in E. grisescens

Candidate resistance-associated genes were screened from the commonly and significantly differentially expressed genes (DEGs) identified in both the FC vs. F and BC vs. B transcriptomic comparisons. Given that midgut-associated resistance in insects is typically associated with multiple physiological processes, including detoxification metabolism, epithelial maintenance, and the regulation of digestive- and immune-related enzymes, this analysis focused on genes functionally annotated to detoxification-related metabolism, midgut tissue function, and peptidase activity.

Based on functional annotation and consistent differential expression patterns, three genes (EgCYP6B2, EgArylα, and EgPGPEP1) were selected as candidate genes potentially associated with resistance for subsequent spatiotemporal expression profiling. It should be noted that the classification of these genes as resistance-associated is based on transcriptomic evidence, and their direct involvement in resistance requires further functional validation.

3.6.1. Expression Profile of EgCYP6B2 Gene

As shown in Figure 6(Aa), EgCYP6B2 expression was lowest in eggs, increased progressively from the first to the fourth larval instars, decreased markedly at the fifth instar and pupal stages, and increased again in adults. Expression levels in adult males were higher than those observed in adult females.

The gradual increase in expression during larval development coincides with active feeding stages, whereas reduced expression during the pupal stage corresponds to metabolic reorganization during metamorphosis. Elevated expression in adults suggests that EgCYP6B2 maintains transcriptional activity beyond larval development.

3.6.2. Expression Profile of EgArylα Gene

As shown in Figure 6(Ab), EgArylα expression was relatively low in eggs, pupae, and adults, increased significantly during the first and second larval instars, gradually declined during the third and fourth instars, and showed a pronounced increase at the fifth instar [40,41].

The elevated expression at the fifth instar corresponds to a developmental stage characterized by intensive physiological preparation for pupation, whereas lower expression levels in adults indicate a reduced association with larval growth-related processes.

3.6.3. Spatiotemporal Expression Profile of the EgPGPEP1 Gene

As shown in Figure 6(Ac), EgPGPEP1 expression was relatively high in eggs, remained at low levels from the first to fourth larval instars, increased slightly at the fifth instar, remained comparable in pupae, and was markedly elevated in adults, with higher expression observed in females.

High expression in eggs suggests potential involvement during early embryonic development. In contrast, low expression during larval feeding stages indicates limited transcriptional activity during this period. The increased expression observed in adults, particularly in females, suggests sustained expression of EgPGPEP1 in later developmental stages.

3.7. Evaluation of RNA Interference Efficiency in Selected Candidate Genes of E. grisescens

3.7.1. RNAi Silencing Efficiency and Screening of Optimal Interference Fragments

To determine whether the candidate genes identified from transcriptomic analyses could be effectively manipulated in vivo, RNA interference (RNAi) assays were performed to evaluate the silencing efficiency of different double-stranded RNA (dsRNA) fragments in E. grisescens larvae.

Silencing Efficiency and Screening of Optimal Interference Fragments for EgCYP6B2

As shown in Figure 6B and Figure S7, larvae treated with dsEg003536-1 exhibited a continuous decrease in EgCYP6B2 expression at 12 h, 24 h, and 48 h compared with the dsGFP control group. A significant reduction was observed at 24 h (p < 0.05), with a silencing efficiency of 64%.

In contrast, dsEg003536-2 and dsEg003536-3 did not induce significant changes in EgCYP6B2 expression at any of the examined time points relative to the dsGFP group (p > 0.05), indicating limited or no detectable silencing effects. Among the tested fragments, dsEg003536-1 exhibited the most pronounced and consistent suppression of EgCYP6B2 expression.

Silencing Efficiency and Screening of Optimal Interference Fragments for EgArylα

As shown in Figure 6B and Figure S8, larvae treated with dsEg015698-1 exhibited a significant reduction in EgArylα expression at 48 h compared with the dsGFP control group (p < 0.001), corresponding to a silencing efficiency of 63%. The dsEg015698-2 treatment resulted in sustained suppression of EgArylα expression, with relative transcript levels decreasing to 0.41 and 0.32 at 24 h and 48 h, respectively (p < 0.001), corresponding to silencing efficiencies of 59% and 68%.

In contrast, larvae treated with dsEg015698-3 showed an elevated expression level at 24 h and returned to levels comparable to the dsGFP control by 48 h, with no significant differences detected at any examined time point. Among the tested fragments, dsEg015698-1 and dsEg015698-2 effectively reduced EgArylα transcript levels, with dsEg015698-2 exhibiting more consistent silencing across multiple time points.

Silencing Efficiency and Screening of Optimal Interference Fragments for EgPGPEP1

As shown in Figure 6B and Figure S9, larvae treated with dsEg004577-1 exhibited significantly reduced EgPGPEP1 expression at 12 h and 24 h compared with the dsGFP control group, with relative expression levels of 0.45 and 0.36, respectively (p < 0.01), corresponding to silencing efficiencies of 55% and 64%. By 48 h, EgPGPEP1 expression in this group returned to levels comparable to the dsGFP control, with no significant difference detected.

In contrast, treatment with dsEg004577-2 did not induce significant changes in EgPGPEP1 expression at any of the examined time points relative to the dsGFP group (p > 0.05). These results indicate that dsEg004577-1 effectively suppressed EgPGPEP1 expression at early time points, whereas the silencing effect was not maintained at 48 h.

4. Discussion

The present study demonstrates that inducing Bacillus thuringiensis (Bt) into a biofilm-associated physiological state, particularly in combination with a composite biofilm inducer, markedly enhances its insecticidal efficacy against Ectropis grisescens. This enhancement is consistently reflected by increased larval mortality and substantially reduced LC_50_ values across multiple larval instars [8]. These results support the growing recognition that the insecticidal performance of Bt is not solely determined by ICPs, but is also strongly modulated by the physiological state of the bacterium. This concept is supported by recent experimental evidence showing that enhanced biofilm formation significantly improves Bt insecticidal activity and stress resilience under environmental challenges [42].

Histopathological analyses provide direct tissue-level evidence explaining this enhanced toxicity. Compared with planktonic Bt, biofilm-state Bt caused earlier onset and greater severity of midgut epithelial damage, including goblet cell deformation, microvillar disorganization, epithelial vacuolization, and eventual epithelial collapse [31]. Consistent with these observations, biofilm-state Bt phenotypes have been reported to exert stronger effects on host midgut morphology than planktonic forms, particularly through disruption of microvilli, which is indicative of aggravated epithelial injury [31,36]. These structural disruptions are consistent with established models of Bt intoxication, in which compromised epithelial integrity facilitates toxin penetration, osmotic imbalance, and larval death. Importantly, the most severe pathological alterations were observed in larvae exposed to biofilm-state Bt supplemented with the inducer, indicating that physiological modulation of Bt amplifies its capacity to disrupt midgut homeostasis.

Beyond structural damage, biofilm-state Bt elicited pronounced and temporally dynamic changes in detoxification- and antioxidant-related enzyme activities. Carboxylesterase (CarE), glutathione S-transferase (GST), and peroxidase (POD) exhibited distinct response patterns characterized by early induction followed by partial or complete suppression, particularly in larvae exposed to biofilm-state Bt with the inducer. Such non-linear dynamics suggest that initial activation of detoxification and antioxidant defenses is subsequently overwhelmed under sustained pathogenic stress. Consistent with these observations, previous studies have reported significant modulation of GST and CarE expression in Lepidopteran larvae following Bt exposure, highlighting the complexity of host metabolic responses to Bt-induced stress [43,44].

Importantly, transcriptomic analyses provide mechanistic insight into these physiological patterns. Enrichment of pathways related to oxidoreductase activity, peroxisome function, and lipid metabolism under biofilm-state Bt exposure offers a molecular explanation for the observed POD dynamics and membrane damage [45,46]. Peroxisomes play a central role in reactive oxygen species (ROS) homeostasis and lipid oxidation; therefore, transcriptional disruption of peroxisome-associated pathways is likely to exacerbate oxidative stress. This oxidative stress burden may further compromise detoxification systems, including cytochrome P450-mediated pathways involved in insect responses to xenobiotics [47], ultimately leading to antioxidant system exhaustion [48,49]. Similarly, enrichment of lipid and glycerophospholipid metabolism pathways aligns closely with histopathological observations of microvillar loss and epithelial membrane disintegration. This interpretation is further supported by previous transcriptomic studies showing that Cry9A- and Vip3A-induced transcriptional responses in insect midgut cells prominently involve genes associated with lipid metabolism and oxidation–reduction processes, implicating disruption of lipid and glycerophospholipid metabolic pathways in epithelial membrane damage and compromised gut integrity [50].

In contrast, under planktonic Bt conditions, transcriptomic enrichment was dominated by metabolic adjustment [36], transporter activity, and moderate immune recognition, corresponding to sustained GST and CarE activity and relatively preserved midgut architecture. Together, these findings establish a coherent link between transcriptional regulation, enzymatic activity, and tissue-level pathology, highlighting biofilm-state Bt as a driver of systemic physiological collapse rather than localized stress adaptation.

Transcriptomic analyses identified multiple genes involved in detoxification, metabolism [51], and midgut-associated functions that were consistently responsive to biofilm induction. Based on functional annotation and expression patterns, several candidate Bt-responsive genes were selected for further spatiotemporal expression analysis and RNA interference (RNAi) assays. Developmental expression profiling revealed distinct stage-specific transcriptional patterns, suggesting potential roles in feeding-stage physiology or post-larval processes.

RNAi experiments successfully demonstrated efficient and gene-specific transcript knockdown [35], confirming the technical feasibility of gene silencing in E. grisescens. However, it is important to emphasize that the RNAi assays in the present study were limited to validation of silencing efficiency and were not coupled with downstream phenotypic, histopathological, or physiological assessments following Bt exposure [52]. Consequently, the functional involvement of these genes in mediating Bt susceptibility or resistance remains inferential.

Accordingly, these genes are conservatively interpreted as candidate Bt-responsive or stress-associated genes rather than definitive resistance determinants. Future studies integrating RNAi with mortality assays, midgut pathology, and enzyme activity measurements will be required to establish causal links between gene function and Bt toxicity. Explicit recognition of this limitation avoids overinterpretation and ensures that conclusions remain firmly grounded in the available data.

Collectively, this study highlights microbial physiological state as a critical but often overlooked determinant of Bt formulation efficacy. By integrating bioassays, histopathology, enzymatic profiling, and transcriptomics, we demonstrate that biofilm-state Bt drives coordinated structural, molecular, and physiological disruption in the insect midgut. These findings provide a conceptual framework for Bt optimization through physiological regulation rather than reliance solely on toxin content.

From an applied perspective, biofilm-induction strategies may offer a promising avenue to overcome reduced susceptibility in late-instar larvae and improve the field performance of Bt-based biopesticides. At the same time, careful consideration of host stress responses and detoxification exhaustion will be essential to balance efficacy with ecological safety [43]. Overall, this work advances our mechanistic understanding of host–pathogen interactions and contributes to the rational design of next-generation microbial insecticides [53].