Based on endogenous gene expression, five soybean Kunitz trypsin inhibitor (KTI) genes KTI1 (Glyma.01G095000.1), KTI2 (Glyma.09G155500.1), KTI3 (Glyma.08G341500.1), KTI5 (Glyma.08G341700.1), KTI7 (Glyma.08G341300.1) and one Bowman-Birk inhibitor gene BBI5 (Glyma.01G096200.1) were selected for overexpression studies. The coding region corresponding to each of these six genes was amplified by PCR from soybean cv. ‘Williams 82’ using primers incorporating specific restriction enzymes (SpeI and AscI) ( Supplementary Table S1 ). The amplified gene products were purified using a DNA Clean & Concentrator kit (Zymo Research, Irvine, CA, USA) according to the manufacturer’s protocol, and then digested with restriction enzymes. The binary vector pB2GW7 (Karimi et al., 2002) was used to generate the individual TI gene construct. The pB2GW7 binary vector consists of a Nos promoter::Bar::Nos terminator cassette for plant selection. The vector pB2GW7 was also digested with SpeI and AscI restriction enzymes and purified using the DNA Clean & Concentrator kit (Zymo Research). After purification of the vector and gene products, ligation reactions were performed to clone the individual TI gene downstream of the CaMV 35S promoter ( Supplemental Figure S1 ) followed by the transformation in E. coli with spectinomycin as bacterial selection. Each soybean TI gene was also cloned under the control of a green tissue-specific promoter. Soybean green-tissue specific promoter (1775 bp) of ribulose-1,5-bisphosphate carboxylase small subunit gene SRS4 (rbcs-SRS4) (Cui et al., 2015) was amplified by PCR from soybean cv. ‘Williams 82’ using primers incorporating specific restriction enzymes (SacI and SpeI) ( Supplementary Table S1 ). The amplified promoter fragment was purified and cloned upstream of the TI gene via restriction enzymes (SacI and SpeI) in the binary vector of pB2GW7 ( Supplementary Figure S1 ). The insertions were confirmed by sequencing. Two types of constructs were created for each TI gene. One set was produced using the CaMV 35S promoter and the other set utilized the rbcS-SRS4 promoter. All constructs were individually transferred into Agrobacterium tumefaciens strain EHA101 via the heat-shock method. The insertion into Agrobacterium was confirmed by colony PCR using the gene-specific primer sets. The primer sequences used in this study are listed in Supplementary Table S1 .

Protein sequence of KTI1, KTI2, KTI3, KTI5, KTI7 and BBI5 were obtained from Phytozome database v13 (https://phytozome-next.jgi.doe.gov/). Protein sequence alignments ( Supplemental Figure S1C ) were performed in Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/).

To evaluate transient TI gene expression, A. tumefaciens containing individual CaMV 35S-TI constructs were used in N. benthamiana agroinfiltration experiments. Agrobacterium-containing individual gene constructs were added to YEP media (5 g/L NaCl, 10 g/L peptone, and 10 g/L yeast extract) with 50 mg/L kanamycin and 200 mg/L spectinomycin and grown overnight in a sterile flask placed in an incubating shaker set at 28 °C and 225 rpm. Acetosyringone (100 µM) was added to the culture for 1 h before removing the flasks from the shaking platform. Cultures were centrifuged at 3000 × g for 15 min. The pelleted cells were resuspended with infiltration buffer (10 mM MgCl2, 10 mM MES, 100 µM acetosyringone, pH 5.6) to an optical density OD₆₀₀ of 0.5. Agrobacterium solution was incubated in the dark at room temperature for 3 h before infiltration. Four-week-old N. benthamiana plants were grown in a growth chamber (Percival Scientific Inc. Perry, IA, USA) at a day/night temperature of 23 °C, 300 µmol/m² s light intensity, and photoperiod of 16/8 h light/dark cycle for vacuum infiltration. The plants were immersed in the Agrobacterium solution and placed into a 20 L vacuum chamber (Best Value Vacs, Naperville, IL, USA). Vacuum pressure of approximately -84 kPa was applied three times for 1 min. For mock control treatments, a 10 mM MgCl2 solution was used for the vacuum infiltration of plants. After infiltration, filter paper was used to remove the excess bacterial solution. The infiltrated plants were covered with a clear plastic lid and returned to the growth chamber. Six independent biological replicates were used for each gene construct.

A. tumefaciens containing individual CaMV 35S-TI constructs was used for Arabidopsis transformation. Wild-type Arabidopsis thaliana (ecotype Col-0) plants were grown in the growth chamber at a day/night temperature of 24 °C under a photoperiod of 16/8 h (light/dark) with 140 µmol/m² s light intensity. Eight-week-old Arabidopsis plants were used for transformation. Primary inflorescence buds were clipped to allow the formation of multiple inflorescence buds. A. thaliana transformation was individually carried out for each CaMV 35S-TI construct via the Agrobacterium floral dip method (Zhang et al., 2006). Briefly, multiple inflorescences were submerged in the Agrobacterium solution for 10 s, and then were allowed to grow in the growth chamber until seed maturity. Three independently transformed plants were selected for each CaMV 35S-TI construct. T₁ generation seeds from individual T₀ plants were harvested and sterilized in 70% ethanol for 1 min, washed with sterile water followed by 50% bleach for 8 min then washed with sterile water for seven times. Sterilized seeds were plated on half-strength MS-medium containing 6 mg/L glufosinate-ammonium. On each plate, 50 seeds were germinated. The transgenic lines were based on single copy lines, with a segregation ratio of 3:1 (resistant: susceptible) in the T₂ generation, and homozygous lines were presumed if there was no segregation in the T₃ generation (n > 100). Selected seedlings were transplanted into the pot containing potting mix (Sun Gro Horticulture, Agawam, MA, USA) for further growth. Transgenic Arabidopsis were allowed to self and homozygous T₃ seeds were selected for further experiments.

Strains of A. tumefaciens containing individual CaMV 35S-TI constructs or individual rbcS-SRS4-TI constructs were used for soybean cv. ‘TN15-5007’ transformation. The binary constructs were introduced into soybean cotyledons by Agrobacterium-mediated transformation (Li et al., 2017). All explants were cultured in a growth chamber at 24 °C (light/dark) under a photoperiod of 16/8 h (light/dark) with 140 µmol/m² s light intensity. Shoots were generated on a selective medium containing 6 mg/L glufosinate-ammonium. After rooting, putative transgenic plantlets were transferred to potting mix (Sun Gro Horticulture, Agawam, MA, USA). Transgenic T₀ soybean plants were confirmed for each construct by painting Finale^® herbicide with glufosinate-ammonium as an active ingredient (Bayer CropScience, Research Triangle Park, NC, USA) on the upper surface of the leaf (Paz et al., 2006). Wild-type soybean leaf was also painted as a negative control. Transgenic T₁ progeny of the individual lines was also confirmed using Finale^® herbicide with a segregation ratio of 3:1 (resistant: susceptible). T₂ seeds were harvested from self-pollinated T₁ progeny. T₂ progeny were screened for herbicide selection and those that showed 100% resistance to Finale^® herbicide were selected. Independent homozygous T₃ lines for each promoter construct were selected for further analysis. A chi-squared test was conducted to determine whether observed segregation ratios were significantly different from expected ratios.

Total genomic DNA was isolated from 1 g of fresh leaves of three-week-old Arabidopsis and soybean plants using the CTAB extraction method (Stewart and Via, 1993). The insertion of the transgene (Nos promoter driving Bar gene) and (2×35S promoter driving TI genes), and (rbcS-SRS4 promoter driving TI genes) was confirmed by PCR using T₃ transgenic Arabidopsis and soybean genomic DNA as a template ( Supplemental Figures S2–S4 ). Genomic DNA was diluted to 100 ng/µL for PCR. PCR conditions for the Bar gene were as follows: 98°C for 2 min followed by 30 cycles at 98°C for 30 s, 64°C for 30 s, and 72°C for 30 s, and a final extension at 72°C for 5 min. PCR conditions for the TI genes were as follows: 98°C for 2 min followed by 30 cycles at 98°C for 30 s, 60°C for 30 s, and 72°C for 30 s, and a final extension at 72°C for 5 min. Testing for Agrobacterium contamination in transgenic lines was performed via PCR using a primer set from the Agrobacterium backbone. To amplify the Agrobacterium chvA (chromosomal virulence gene A) gene, as a control for the Agrobacterium contamination, the PCR conditions were as follows: 98°C for 2 min followed by 30 cycles at 98°C for 30 s, 65°C for 30 s, and 72°C for 30 s, and a final extension at 72°C for 5 min ( Supplemental Figures S2B, S3B, S4B ). PCR products were visualized on 0.8% agarose gels containing ethidium bromide. Primers used for the genotypic analysis of transgenic lines are provided in the Supplementary Table S2, S3 .

The trypsin inhibition analysis was performed as described by Malefo et al. (2020). Total protein was extracted from six-week-old T₃ transgenic and non-transgenic Arabidopsis and soybean plants for each gene construct. Arabidopsis rosettes or soybean leaves were ground and resuspended in protein extraction buffer (0.15 M NaCl, 50 mM sodium phosphate pH 6.2, and 2 mM EDTA pH 8.0) for 1 h at 4 °C. The content was centrifuged at 8000 g for 15 min and the supernatant was transferred to fresh tubes. Total protein content was determined using the Qubit Protein Assay Kit (Thermo Fisher Scientific, Pittsburgh, PA, USA). In vitro inhibitory activity of transgenic plant protein was tested against commercial bovine TPCK-treated trypsin (EC 3.4.21.4) (≥10,000 BAEE units/mg, Millipore Sigma, Saint Louis, MO, USA) and bovine TLCK-treated chymotrypsin (EC 3.4.21.1) ((≥40 units/mg, Millipore Sigma). Specific activities were assayed with the synthetic substrate; Nα-Benzoyl-DL-arginine 4-nitroanilide hydrochloride (BApNA) (Millipore Sigma) for trypsin and N-Succinyl-Ala-Ala-Pro-Phe p-nitroanilide (SAAPFpNA) (Millipore Sigma) for chymotrypsin. Enzyme activity was performed in 0.2 mL reactions in 96-well black non-stick microtiter plates. Protein extracts (20 µg) from transgenic and non-transgenic control lines were incubated with 100 ng of commercial trypsin in buffer (50 mM Tris-HCl, pH 8.2) for 10 min, and then the substrate was added to a 0.2 mM final concentration and incubated for 1 h at 28 °C. The absorbance of the reaction product, ρ-nitroaniline, was measured at 405 nm using a Synergy H1 multi-detection microplate reader (Bio-Tek Instruments Inc., Santa Clara, CA, USA). A standard curve was generated using set quantities of soybean trypsin inhibitor (Millipore Sigma) or chymotrypsin inhibitor (Millipore Sigma). Inhibitory activity was calculated and expressed as a percentage of inhibition of transgenic protein relative to that wild-type protein. Three independent transgenic lines were used for each gene construct and six replicate plants were used for each line. All assays were carried out in triplicate for each plant.

The effect of each TI gene overexpression on insect leaf defoliation was evaluated in transgenic Arabidopsis and soybean lines through detached leaf punch bioassays (Arnaiz et al., 2018) with corn earworm (H. zea) larvae. Corn earworm eggs were purchased from Benzon Research (Carlisle, PA, USA) and hatched in the growth chamber at 26 °C on a 16/8 h (light/dark) photoperiod, light intensity (150 µmol/m² s). Leaves were removed from six-week-old transgenic Arabidopsis or soybean plants and placed in a plastic bag with water spray to keep moistened for a few hours. Leaf punches (1 cm²) were cut and transferred to individual cells of 128-cell plastic bioassay trays (Frontiers Agricultural Sciences, Newark, DE, USA). Each cell was filled with 1 mL of 2% agarose solution to preserve turgidity of the leaf punch. Six-week-old wild-type plant leaf punches were used as a control. A single 1^st instar was tested per cell. During the assay, the old leaf punches were replaced with fresh leaf punches every three days. Three independent transgenic lines were used for each gene construct and six replicate plants were used for each line. Sixteen cells were filled with leaf punches from a single plant. All bioassay trays were kept in the same conditions as above mentioned. Larval weight was recorded after eight days of feeding.

Bioassays on transgenic soybean plants were performed with plants grown in a polyester-mesh cage under greenhouse conditions. The environmental conditions of the greenhouse were 16/8 h (light/dark) photoperiod and 25 °C temperature with fluctuations from a minimum of 22 °C to a maximum of 28 °C. In whole plant feeding bioassays, plants overexpressing the two best-performed genes (KTI7 and BBI5) were selected based on the results from detached leaf feeding bioassays for CaMV 35S-TI or rbcS-SRS4-TI. The mesh cage contained six-week-old T₃ transgenic as well as wild-type control plants. Three independent transgenic lines for each gene and six replicate plants per line were used in the experiments. Corn earworm eggs were hatched and larvae reared on artificial diet (beet armyworm diet, Frontiers Agricultural Sciences, Newark, DE, USA) for four days in a growth chamber at 26 °C on a 16/8 h (light/dark) photoperiod, light intensity (150 µmol/m² s). A total of ten 2^nd instar larvae were added to the bottom, middle, and top trifoliate leaf surface for each plant. After 10 days of feeding, percent leaf defoliation was assessed for each plant. The percentage of leaf defoliation area was calculated in ImageJ software, as described (Ortega et al., 2016).

Three independent transgenic soybean lines and six replicate plants for each line of individual TI gene construct under the control of CaMV 35S promoter were used for bioassays with SCN. The assay protocol was adapted according to Mazarei et al. (2011). Transgenic and wild-type soybean seeds were germinated in the potting mix for four days. Roots from each seedling were rinsed with water gently to remove potting mix and transferred to the pot containing sand: clay (in 3:1 ratio) mixture. One day after repotting, each seedling was inoculated with SCN HG type 0 (race 3). One hole, 1 cm deep, was made closure to the roots of seedlings. Each seedling was inoculated with 1 mL of inoculum with SCN eggs (approximately 2500). Plants were maintained in the growth chamber at 25 °C on a 16/8 h (light/dark) photoperiod with 140 µmol/m² s light intensity. After five weeks, plant roots were individually washed with a strong jet of water to dislodge the cysts and females. These were counted under a stereomicroscope. The female index was calculated for each soybean line where the average number of mature female nematodes and cysts on the transgenic soybean line was divided by the average number of mature female nematodes and cysts on the susceptible soybean control line and multiplied by 100 (Lin et al., 2016).

The effect of TI gene overexpression under the control of CaMV 35S promoter or rbcS-SRS4 promoter on soybean plant growth and development was evaluated under greenhouse conditions. The environmental conditions of the greenhouse were 16/8 h (light/dark) photoperiod and 25 °C temperature with fluctuations from a minimum of 22 °C to a maximum of 28 °C. Based on the results from insect feeding bioassays, plants overexpressing the two best-performing TI genes (KTI7 and BBI5) for each promoter type were chosen for growth and development studies with no herbivore application. Three independent T₃ homozygous transgenic lines for each gene type were used in the experiment. A total of 40 plants were grown for each gene type (ten plants per transgenic and ten per non-transgenic control lines) to compare plants growth and development. Plants were started from seed in the greenhouse and grown until the seed was set. Several growth parameters including plant height, pods number, dry aboveground biomass yield, and total seed weight were analyzed.

Midgut fluids from H. zea larvae (fourth instar) were used to assay trypsin inhibitory activity of transgenic plant extracts using the synthetic substrate BApNA. Midgut fluids were extracted from actively feeding 4th instar larvae. Ten larvae were chilled on ice and the guts containing the food bolus were carefully dissected and placed into a microcentrifuge tube. Immediately after collection of the 10 gut contents, the collection tube was submerged in liquid nitrogen and stored at -80 °C until used. Prior to assays, the gut content samples were thawed on ice and 500 μl of deionized water was added to each tube, vortexed for 2 min and centrifuged at 15,000 rpm for 20 min at 4 °C. The supernatants were filtered through 0.2 μm filters into new microcentrifuge tubes and considered as soluble proteinase samples. Protein concentrations were determined using the Qubit protein kit (Invitrogen) and adjusted to be 1 mg/ml using deionized water. The method of digestive enzyme inhibition analyses was conducted by following a previously described method. Total protein from six-week-old T3 transgenic and non-transgenic soybean plants was used for each gene construct. Three independent transgenic lines were used for each gene construct and six replicate plants were used for each line. All assays were carried out in triplicates from each plant.

TRI reagent (Zymo Research) was used to extract total RNA from six-week-old transgenic Arabidopsis and soybean plants according to the manufacturer’s protocol. Total RNA was quantified using a NanoDrop™ spectrophotometer ND-1000 (Thermo Fisher Scientific) and RNA integrity was checked by agarose gel electrophoresis. Total RNA was treated with DNaseI, and column purified with the RNA Clean & Concentrator™ kit (Zymo Research) to remove genomic DNA contamination. Total RNA (1 µg) was used to synthesize first-strand cDNA in a 20 µL reaction volume containing 1 µL of 50 µM oligo dT primer and 1 µL of 10 mM dNTP mix, 2 µL of 10×RT buffer, 4 µL of 25 mM MgCl2, 2 µL of 0.1 M DTT, 1 µL of RNaseOUT™ (40 U/µL), and 1 µL of SuperScript^® III RT (200 U/µL). The TI gene-specific primers ( Supplementary Table S1 ) were designed for qRT-PCR using Primer3 (v 0.4.0). Real-time PCR was conducted in a 15 µL reaction volume containing, 7.5 µL of Power SYBR Green 2X Master Mix (Applied Biosystems, Foster City, CA, USA), 1 µL of cDNA (12.5 ng), 0.375 µL of each primer (10 µM), and 5.75 µL of H₂O. The real-time PCR was carried out on a QuantStudio™ 6 Flex Real-Time PCR System (Applied Biosystems). Expression analyses were performed using a standard curve method for relative expression normalized to the Arabidopsis actin gene (ACT) for Arabidopsis and soybean ubiquitin gene (GmUBI3) for soybean plants.

Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) assays were used to determine the levels of expression for TI genes in leaves, stems, seeds, and roots of non-transgenic wild-type soybean plants. The expression of KTI1, KTI3, and BBI5 were the highest in seeds, whereas the expression of KTI2 was very low and KTI5, and KTI7 were undetectable in seeds ( Figure 1 ). The expression of KTI1 and KTI3 were undetectable in leaves, stems, and roots. The expression of KTI5 was low and detected only in leaves. Similarly, the expression of KTI7 was low and detected only in stems ( Figure 1 ). The expression of KTI2 and BBI5 were low and detected in leaves ( Figure 1 ). There was no detectable expression for any of the TI genes in roots ( Figure 1 ).

Agroinfiltrated N. benthamiana leaves containing the individual 35S::TI gene construct were evaluated for transient expression. The relative qRT-PCR expression levels for KTI2, KTI3, and KTI7 were the highest (3.75, 3.86, and 3.35, respectively), and for KTI1, KTI5, and BBI5 were low (0.21, 0.12, and 0.73, respectively) ( Supplemental Figure S5 ).

Six independent T₁ transgenic Arabidopsis and soybean lines were generated for each TI gene construct under the control of either 35S or rbcS-SRS4 promoters. Subsequently, five independent homozygous T₃ transgenic Arabidopsis and soybean lines were obtained and screened by PCR using genomic DNA to confirm the presence of transgenes Bar and TI ( Supplemental Figures S2–4 ). For further analysis, three independent homozygous T₃ transgenic Arabidopsis and soybean lines were selected to represent a range of transgene expression for each TI gene.

The relative qRT-PCR transcript abundance of each TI transgene (under the control of 35S promoter) was determined in three independent homozygous T₃ transgenic Arabidopsis lines. Considerable expression levels of KTI1, KTI2, KTI3, KTI5, KTI7, and BBI5 were observed in leaves of all three corresponding transgenic lines ( Supplemental Figure S6 ). In particular, the relative expression of KTI7 was found to be the highest (397.72 ± 63.56) among TI genes ( Supplemental Figure S6E ). The high expression levels of KTI7 were consistent among three transgenic lines. The expression of KTI3 and BBI5 were medium (15.35 ± 2.65 and 13.03 ± 2.69, respectively) ( Supplemental Figures S6C–F ), whereas the expression of KTI1, KTI2, and KTI5 were low (3.87 ± 0.56, 6.82 ± 2.36, and 1.72 ± 0.49, respectively) ( Supplemental Figures S6A, B-D ). No expression was detected in the non-transgenic (WT) plants ( Supplemental Figure S6 ).

The relative qRT-PCR expression of each TI transgene and total gene (under the control of the 35S promoter) was determined in three independent homozygous T₃ transgenic soybean lines. Similar to transgenic Arabidopsis, the relative expression of the transgene and total gene for KTI7 was the highest (transgene: 6.92 ± 0.96 and total gene: 15.68 ± 2.11) in soybean among TI genes ( Figures 2A, B ). The relative expression of the transgene and total gene for KTI3, KTI5, and BBI5 were medium with 2.4-, 4.80-, and 3.70-fold, respectively, lower transgene expression and 4.6-, 3.9-, and 2.2-fold, respectively, lower total gene expression compared to KTI7 ( Figures 2A, B ). The expressions of KTI1 and KTI2 were low (transgene: 6.6- and 13-fold lower compared to KTI7) (total gene: 11- and 14-fold lower compared to KTI7) in transgenic soybean lines ( Figures 2A, B ). The relative expression in all transgenic lines was significantly different compared to non-transgenic (WT) plants ( Figures 2A, B ). Similarly, the relative expression of each TI transgene and total gene (under the control of the rbcS-SRS4 promoter) had similar patterns in transgenic soybean lines ( Figures 2C, D ). The expression of the transgene and total gene for KTI7 was the highest (transgene: 3.74 ± 0.46 and total gene: 5.00 ± 0.59) in soybean lines among TI genes ( Figures 2C, D ). The expression of the transgene and total gene for KTI3, KTI5, and BBI5 were medium with 1.8-, 3-, and 2.5-fold, respectively, lower transgene expression and 2-, 3-, and 1.4-fold, respectively, lower total gene expression compared to KTI7 ( Figures 2C, D ). Whereas the expression of KTI1 and KTI2 were low (transgene: 3.8- and 7.5-fold lower compared to KTI7) (total gene: 3.9- and 8-fold lower compared to KTI7) in transgenic soybean lines ( Figures 2C, D ). The relative expression in all transgenic lines was significantly different compared to non-transgenic (WT) plants ( Figures 2C, D ).

In vitro enzyme inhibitory activity (trypsin and chymotrypsin) was determined using total protein from transgenic Arabidopsis and soybean lines overexpressing the individual TI gene construct. For transgenic Arabidopsis lines overexpressing the 35S::TI gene construct, transgenic lines of KTI7 (line 1: 60.75 ± 8.68%, line 2: 61.45 ± 8.36%, and line 3: 66.06 ± 4.31%) and BBI5 (line 1: 43.37 ± 8.53%, line 2: 47.96 ± 6.35%, and line 3: 40.45 ± 6.69%) showed higher inhibitory activity against commercial trypsin among all TI genes ( Figure 3A ). Transgenic Arabidopsis lines of KTI1 (an average of 2.5-fold), KTI2 (2.2-fold), and KTI3 (1.9-fold) showed moderate inhibitory activity, whereas KTI5 (3.3-fold) showed lower inhibitory activity ( Figure 3A ) compared to KTI7. Transgenic Arabidopsis lines of BBI5 showed higher inhibitory activity (27.73 ± 2.08%) against commercial chymotrypsin, whereas KTI2 (2-fold), KTI3 (1.7-fold), and KTI7 (1.6-fold) had moderately low activity compared to BBI5. The KTI1 (4-fold) and KTI5 (3-fold) had much lower inhibitory activity compared to BBI5 ( Figure 3A ).

For transgenic soybean lines overexpressing the 35S::TI gene construct, transgenic lines of KTI7 and BBI5 showed the highest trypsin inhibitory activity (an average of 45.44 ± 4.55% and 37.53 ± 2.92%, respectively) among TI genes ( Figure 3B ). Transgenic soybean lines of KTI1 (3.3-fold), KTI2 (2.2-fold), KTI3 (1.6-fold), and KTI5 (2.6-fold) showed lower trypsin inhibitory activity compared to KTI7 lines ( Figure 3B ). Transgenic soybean lines of BBI5 showed the highest chymotrypsin inhibitory activity (an average of 20.78 ± 2.85%), whereas KTI1, KTI2, KTI3, KTI5, and KTI7 showed an average of 7-, 3-, 2.9-, 6-, 1.6-fold respectively lower inhibitory activity compared to BBI5 ( Figure 3B ).

For transgenic soybean lines overexpressing the rbcS-SRS4::TI gene construct, transgenic lines of KTI7 showed the highest trypsin inhibitory activity (an average of 37.97 ± 2.23%) ( Figure 3C ). In contrast, transgenic soybean lines of KTI1, KTI2, KTI3, KTI5, and BBI5 showed an average of 2.8-, 1.8-, 1.5-, 2.3-, and 1.4-fold, respectively lower inhibitory activity compared to KTI7 ( Figure 3C ). The chymotrypsin inhibitory activity was found to be higher in transgenic soybean lines of BBI5 (an average of 21.63 ± 3.40%), whereas transgenic lines KTI1, KTI2, KTI3, KTI5, and KTI7 showed an average 4.7-, 3-, 2.9-,4-, 1.9-fold, respectively lower inhibitory activity compared to transgenic BBI5 lines ( Figure 3C ).

Homozygous T₃ transgenic Arabidopsis and soybean lines overexpressing the individual TI gene construct were used to determine the effect of corn earworm by feeding leaf punches. Leaf punches from transgenic Arabidopsis lines overexpressing the 35S::TI gene and non-transgenic Arabidopsis were fed by corn earworm first instar larvae ( Supplemental Figures S7A, B ). Upon eight days of feeding, the representative larval size was shown in Supplemental Figure S7C . Larval weight was recorded and compared between transgenic lines of each TI gene and non-transgenic plants. Notably, all transgenic lines of each TI gene showed a significant reduction of larval weight compared to non-transgenic (WT) plants ( Supplemental Figure S7 ). Particularly, transgenic KTI7 and BBI5 lines had the greatest reduction in corn earworm larval biomass, with an average of 8.55 ± 1.20 mg and 10.09 ± 1.58 mg larval weight, respectively, compared to non-transgenic plants (an average of 23.63 ± 0.82 mg) ( Supplemental Figure S7D ). Transgenic Arabidopsis lines of KTI1, KTI2, KTI3, and KTI5 showed a significant reduction of larval weight, an average of 14.87 ± 1.07 mg, 11.72 ± 1.4 mg, 11.09 ± 1.31 mg, and 12.93 ± 1.49 mg respectively ( Supplemental Figure S7D ).

Transgenic soybean lines overexpressing the 35S::TI gene were also used for leaf punch feeding assays ( Figures 4A–C ). Transgenic soybean lines of KTI7 and BBI5 had a significant reduction of larval weight (45.80 ± 3.16 mg and 51.42 ± 2.73 mg, respectively) compared to non-transgenic control plants (85.88 ± 3.71 mg) ( Figure 4D ). Larvae fed on other transgenic soybean lines of KTI1, KTI2, KTI3, and KTI5 showed no consistent significant differences in the larval weight compared to non-transgenic control plants ( Figure 4D ).

The leaf punch feeding assays were also performed on transgenic soybean lines overexpressing the rbcS-SRS4::TI gene ( Figures 5A–C ). Similarly, transgenic soybean lines of KTI7 and BBI5 reduced larval weight (50.00 ± 7.76 mg and 53.54 ± 4.32 mg, respectively) compared to non-transgenic control plants (90.3 ± 10.06 mg) ( Figure 5D ). There were no consistent significant differences for other transgenic soybean lines of KTI1, KTI2, KTI3, and KTI5 ( Figure 5D ). Based on these results, two best performing types of transgenic plants overexpressing KTI7 and BBI5 genes under the control of either 35S or rbcS-SRS4 promoter were selected for further evaluation.

The corn earworm bioassays were performed on plants grown in a polyester-mesh cage under greenhouse conditions ( Figure 6 ). Homozygous T₃ transgenic soybean lines for KTI7 and BBI5 (under the control of 35S promoter) and non-transgenic control plants were inoculated with corn earworm second instar larvae ( Figures 6A, B ). After ten days of feeding, the percent leaf defoliation was assessed for each plant ( Figure 6C ). Representative defoliated transgenic, and non-transgenic plants are shown in Figure 6D . Both transgenic KTI7 and BBI5 lines showed a significant reduction in leaf defoliation (KTI7: line 1, 42.89 ± 6.00%; line 2, 41.20 ± 4.4%; line 3, 45.29 ± 3.07%) (BBI5: line 1, 50.37 ± 4.65%; line 2, 50.60 ± 5.83%; and line 3, 51.64 ± 4.84%), compared to non-transgenic control plants (79.53 ± 5.29%) ( Figure 6E ). Transgenic soybean lines for both KTI7 and BBI5 driving under the control of the rbcS-SRS4 promoter and non-transgenic control plants were also used for corn earworm feeding bioassay ( Figures 7A–C ). Representative defoliated plants for transgenic KTI7 and BBI5 lines, and non-transgenic control plants are shown in Figure 7D . Similarly, both KTI7 and BBI5 transgenic lines showed a significant reduction in leaf defoliation for KTI7 lines (line 1: 49.51 ± 5.10%, line 2: 47.66 ± 3.64%, and line 3: 47.60 ± 5.10%) and BBI5 lines (line 1: 52.79 ± 3.57%, line 2: 53.69 ± 4.78%, and line 3: 54.51 ± 4.48%), compared to non-transgenic control plants (79.74 ± 5.48%) ( Figure 7E ).

Homozygous T₃ transgenic soybean lines for KTI7 and BBI5 (under the control of 35S promoter) were used for SCN HG Type 0 (race 3) bioassay. Both transgenic KTI7 and BBI5 lines showed no significant differences in the female index compared to non-transgenic control plants ( Supplemental Figure S8 ).

Homozygous T₃ transgenic soybean lines for KTI7 and BBI5 (under the control of either 35S or rbcS-SRS4 promoter) and non-transgenic control plants were grown under greenhouse conditions to evaluate agronomic traits for plant height, number of pods, dry aboveground biomass, and seed yield per plant. The representative transgenic plants under the control of two different promoters were shown in Supplemental Figures S9A, B . There were no significant differences in the plant height, pod numbers, dry aboveground biomass, and seed yield per plant comparing transgenic and non-transgenic plants ( Supplemental Figures S9C–F ). Furthermore, there was no variation in agronomic traits among transgenic lines ( Supplemental Figures S9C–F ). There was also no negative effect on plant growth and development compared to non-transgenic plants.

The effect of trypsin inhibitor genes on the inhibition of gut enzymes has been performed. The results showed the highest inhibition of gut enzymes by KTI7 (line 1: 28.95 ± 2.65, line 2: 32.15 ± 2.64, and line 3: 29.03 ± 3.03) and BBI5 (line 1: 20.84 ± 1.52, line 2: 20.61 ± 2.08, and line 3: 19.67 ± 0.92) compared to non-transgenic control (1.00± 0.01) ( Supplemental Figure S10 ). The proteolytic inhibition by KTI1, KTI2, KTI3, and KTI5 showed an average of 4-, 3-, 3-, 7- fold, respectively lower compared to KTI7 ( Supplemental Figure S10 ).