Trypsin Inhibitor of Ricinus communis L. (Euphorbiaceae) in the Control of Spodoptera frugiperda (Smith, 1797) (Lepidoptera: Noctuidae)

Abstract

The fall armyworm, Spodoptera frugiperda (Lepidoptera: Noctuidae), is a polyphagous insect of various agricultural crops. The methods used for its control have led to the selection of resistant insect populations, which justifies the search for new alternatives for the management of this insect. Accordingly, in the present study, trypsin inhibitors present in the leaf extract of Ricinus communis (Euphorbiaceae) were investigated for their activity against S. frugiperda. Chemometric optimization methods were developed for the extraction, purification, identification, and structural characterization of the inhibitors. In addition, the effect of R. communis extract on S. frugiperda development was evaluated. Inhibitor extraction was chemometrically optimized, yielding an extract with an antitryptic activity of 94,837.14 mUIT g⁻¹. The R. communis extract was purified and found to contain two compounds (adenosine and ricinine) exhibiting trypsin inhibitor activity. However, after purification, only ricinine inhibited S. frugiperda trypsin in vitro (103.21 mUIT mg⁻¹). The extract was added to the diet of S. frugiperda larvae, resulting in reduced digestion, increased protein in the feces (control = 12,571 μg protein/mg feces; 1818.2 g mL⁻¹ = 16,867 μg protein/mg feces), and insect mortality. At the highest concentration, the treatment led to an LT50 of 15.9 days and a cumulative survival rate of 18.5%. Based on the results of this study, it is possible that ricinine binds to the catalytic site of trypsin, causing the mortality of S. frugiperda larvae.

1. Introduction

The fall armyworm, Spodoptera frugiperda (Smith, 1797) (Lepidoptera: Noctuidae), is a polyphagous and cosmopolitan insect that causes damage to several crops of economic importance. Control is typically performed using synthetic chemical insecticides, which have led to the selection of resistant insect populations [1,2,3]. Insecticides and genetically modified Bt crops are the primary methods for controlling the fall armyworm, S. frugiperda, but since its spread to regions such as Africa, the Far East, and Australia, where Bt crops are scarce, increased insecticide use has led to reduced effectiveness and resistance, especially in areas with underdeveloped agrochemical industries and limited access to proper equipment, further compromising control efforts and accelerating resistance development [4]. Another control method is the use of genetically modified plants; however, S. frugiperda was the first insect to show resistance to this technology under field conditions [5,6,7,8].

Considering these issues, research aiming to obtain substances of natural origin that can be used for the control of S. frugiperda has intensified [8,9,10,11,12]. Digestive enzyme inhibitors stand out among plant metabolites that affect insect herbivory [13]. This group of defense molecules can be found in plants from different botanical families, with the most common being serine proteinase inhibitors, showing distinct structures and classifications [14]. Since the discovery of the activity of protein inhibitors in the defense against herbivory, numerous attempts have been made during plant breeding to select plants containing high protein inhibitor concentrations [15]. Additionally, the use of these inhibitors from plant extracts for the control of agricultural pests has been reported in several studies [16,17,18,19,20,21,22].

Digestive proteinase inhibitors act by forming a complex in equilibrium with digestive proteinases, which limits protein hydrolysis, reducing the absorption of nutrients essential to insect development [15,22,23]. The formation of complexes with digestive proteinases affects insect development due to the lower absorption of essential amino acid residues, increasing the metabolic cost for these organisms.

Previous studies conducted by our research group revealed the presence of nonprotein enzyme inhibitors in the castor oil plant Ricinus communis L. (Euphorbiaceae) [18,19,24]. However, the substance responsible for inhibition and the interaction mechanism with the trypsin enzyme of S. frugiperda has not been elucidated.

Therefore, the aim of the present study was to obtain the extract of R. communis leaves (using chemometric methods), purify, and identify the trypsin inhibitor present in this extract using in vitro inhibition assays with S. frugiperda trypsin. In addition, the toxicity of the castor oil plant leaf extract against S. frugiperda was evaluated.

2. Materials and Methods

2.1. Castor Oil Plant Leaf Meal (CLM)

Ricinus communis leaves were collected from the campus of the Federal University of Lavras (Universidade Federal de Lavras—UFLA), Minas Gerais, Brazil (21°22′56″ South; 44°97′68″ West). The stalk and main veins were removed, and the remaining material was dried in a forced air oven at 45 °C for 72 h. The material was ground in a Tecnal TE-631 mill (Tecnal, Piracicaba-SP, Brazil) to obtain CLM [25].

2.2. Insects

An artificial diet [26] was fed to S. frugiperda larvae. Adults were fed an aqueous honey solution (0.01 mL mL⁻¹). For the trypsin activity inhibition assays, 6th instar larvae were used and kept under laboratory conditions. The in vivo assay was conducted with the second instar larvae of S. frugiperda, obtained from the second egg-laying cycle of the specimens. The insects were maintained under controlled laboratory conditions (temperature, 25 ± 2 °C; RH, 60 ± 10%; and photoperiod, 12 h light: 12 h dark).

2.3. In Vitro S. frugiperda Trypsin Inhibition Assays

Spodoptera frugiperda trypsin inhibition assays were used to evaluate the efficacy of the chemometric procedures and biodirected purification of compounds that inhibit trypsin activity [19]. To perform the assays, a homogenate, as a source of trypsin, was obtained from the digestive tubes of S. frugiperda reared in the laboratory. The larvae were immobilized at −20 °C for 10 min, and their digestive tubes were removed. The digestive tubes were macerated in a Potter homogenizer at a ratio of 1 digestive tube to 4 mL of deionized water at 4 °C. The crude extract was centrifuged at 10,000× g at 4 °C for 30 min. The supernatant, considered the enzyme extract, was stored in a freezer at −20 °C until use [19].

Trypsin activity inhibition was evaluated using a kinetic enzymatic assay performed at 4 reaction time periods (30, 60, 90, and 120 min) using the substrate N-benzoyl-DL-arginine-p-nitroanilide ((98% BAPNA, Sigma-Aldrich Brasil Ltda, São Paulo-SP, Brazil) according to [19,27]. For the analysis of HPLC fractions, a nonkinetic inhibition assay was performed, comparing the blank with the sample after 30 min of reaction.

The control, i.e., enzyme with no inhibitor, was obtained by replacing the sample with distilled water. The reaction product was quantified in a spectrophotometer at a wavelength of 410 nm. The inhibition values were expressed in mUIT (milliunits of inhibited trypsin).

2.4. Optimization of the Extraction of S. frugiperda Trypsin Inhibitor Present in CLM

2.4.1. 2⁴ Factorial Design

The extraction of S. frugiperda trypsin inhibitors present in CLM was optimized using a factorial design with 24 as the central point (24 FACT). For the design (Table S1), 2 levels and the midpoint were used for each of the 4 evaluated parameters: sample/solvent proportion (P), expressed in weight/volume of CLM extracted with the solvent designated for each experiment: 1/30 = 0.0333, 1/60 = 0.0166, and 1/40 = 0.025, solvent type (So), expressed in % ethanol, with 30, 60 and 45% used in each experiment, the number of re-extractions (Re), performed 1, 2 and 3 times as necessary for each experiment, and extraction time (T), with 60, 90 and 120 min of extraction for each given experiment. The inhibitory activity against S. frugiperda trypsin was used as a determining factor in the choice of the best conditions to extract the inhibitor present in CLM [28].

To perform the experiments, a CLM sample (100 mg) was added to a solvent (30, 45 or 60% ethanol/water). Ultrasound-assisted extraction (UAE) was conducted in an ultrasound bath at 35 kHz and 40 °C for 60, 90, or 120 min. Next, the samples were centrifuged (5000× g for 5 min), and the supernatant was stored until analysis. The procedures were repeated with the precipitate by performing re-extractions (1, 2, or 3 times) according to the 24 FACT design (Table S1). The supernatants from the re-extractions of each experiment were pooled for subsequent kinetic quantification of trypsin inhibition, which was used as a parameter to determine which factors (P, So, Re, and T) had greater effects on extracting the S. frugiperda trypsin inhibitor present in CLM.

2.4.2. Optimization of the Extraction Parameters for the S. frugiperda Trypsin Inhibitor Present in CLM Using a 2² Central Composite Design (2² CCD)

With the results obtained with the 24 FACT design, a 2² CCD was performed using proportion (P) and solvent type (So) (Table S2). The CLM samples (100 mg) were subjected to solvent extraction (23.79, 30, 45, 60, or 66.21% ethanol/water) at ratios of 1/37.4, 1/40, 1/48, 1/60, or 1/66.9 (w/v, CLM/solvent). The UAE was conducted at 40 °C for 60 min. After this period, the material was centrifuged (5000× g for 5 min). Re-extraction under the same conditions was performed, and the supernatants from each treatment were pooled. At the end of the experiments, a standardized CLM extract was obtained.

2.5. Purification and Identification of Substances

Part of the standardized CLM extract (1 g) was eluted on C₁₈ silica (Sigma-Aldrich, St. Louis, MO, USA) using water/methanol (0, 10, 20, 30, 40, 50, 60, and 70% v/v) as the eluent. The fractions obtained were collected, and their trypsin inhibitory activity was evaluated. The fractions that promoted greater trypsin inhibition were pooled and dried in a rotary evaporator, resulting in 2 fractions, 4 and 5. These fractions were analyzed by HPLC-UV/PDA (Shimadzu, Kyoto, Japan) equipped with a degasser (DGU-20A3), 2 pumps (LC 20AT), an autosampler (SIL-20A), a UV/PDA detector (SPD-M20A) and a column oven (CTO-20A). An Ascentis^® (Supelco, Bellefonte, PA, USA) C18 column (250 mm) (250 mm × 4.6 mm id, 5 μm) coupled to an Ascentis^® C₁₈ precolumn (15 mm × 3.2 mm id, 5 μm) was used for analyses in analytical mode (HPLC-AM). The injected volumes were 5 μL for the fractions obtained by flash chromatography on silica gel and 20 μL for the pooled fractions. The mobile phase was water + 0.01% acetic acid (A) and 100% methanol (B), and the elution was carried out in gradient mode with an initial A/B composition of 94:6 (v/v), increasing to 40% B in 30 min at a flow rate of 1 mL min⁻¹. The wavelength used for detection was 254 nm. This method was used to analyze the fractions obtained by flash chromatography on silica gel, both for pooled fractions and for fractions collected from preparative mode chromatography.

Preparative mode chromatography was performed using the same equipment but with a Luna^® (Phenomenex, Torrance, CA, USA) C₁₈ preparative column (150 mm × 21.20 mm id, 5 μm) at a flow rate of 10 mL min⁻¹ (HPLC-PM). The analyzed sample consisted of the fractions with the best inhibitory activity; these pooled fractions were obtained by flash chromatography on silica gel at a concentration of 50 mg mL⁻¹ and an injected volume of 1 mL.

To identify the isolated substances, ¹H and ¹³C NMR spectroscopy were performed on a Bruker Avance III 600 HD spectrometer at 600 and 150 MHz, respectively. One- and two-dimensional COSY (homonuclear correlation spectroscopy—¹H × ¹H), HMBC (heteronuclear multiple-bond correlation—¹H × ¹³C) and HSQC (heteronuclear single-quantum coherence—¹H × ¹³C) were performed. The samples were solubilized in dimethyl sulfoxide-d6 (DMSO-d6, Merck, (Sigma-Aldrich Brasil Ltd., Cajamar, SP, Brazil)).

The mass spectra were obtained using a Bruker Maxis Impact QTOF mass spectrometer (Bruker Daltonics Inc., Fremont, CA, USA) with the following settings: full scan mode from m/z 50 to 1500; electrospray ionization (ESI) in positive mode; active focus; N2 as the drying gas at a flow rate of 4 L min⁻¹ and 180 °C; nebulizer pressure of 0.3 bar; capillary voltage of 2900 V; charging voltage of 2000 V; end plate offset of 500 V.

2.6. Effects of the CLM Extract on S. frugiperda Development

The CLM extract was prepared according to the parameters optimized by means of chemometric analyses. For this purpose, CLM was subjected to extraction with 71.7% ethanol at a ratio of 1/48.6 (w/v CLM/ethanol) at 40 °C for 60 min in an ultrasound bath, and re-extraction was performed under the same conditions.

The treatments consisted of an artificial diet supplemented with CLM extract at concentrations of 113.6, 227.3, 454.5, 909.0, or 1818.2 g mL⁻¹. Diet plus distilled water was used as a negative control. The experimental design was completely randomized, with 120 replicates. Each replicate consisted of 1 insect (2nd instar at the beginning) kept alone with an ad libitum diet. The following variables were analyzed: survival during the larval stage evaluated every 24 h; larval period duration; pupal weight; pupal period duration.

Upon adult emergence, couples from each treatment with a maximum of 24 h of age difference were placed individually in cylindrical cages measuring 10 cm × 10 cm (height × diameter). The insects were fed an aqueous honey solution (0.01 mL mL⁻¹). The following biological characteristics were evaluated: adult longevity, preoviposition and oviposition periods, and number of eggs per female.

2.7. Protein Concentration in S. frugiperda Feces

Feces from larvae fed CLM extract, as described in the previous section, were collected after the end of the larval period. Fecal samples were dried in a forced air oven at 45 °C. Distilled water (200 μL) and 1 mol L⁻¹ perchloric acid (200 μL) were added to the fecal samples (15 mg). This mixture was homogenized and placed in an ice bath for 10 min. The samples were centrifuged at 5000× g, and the supernatant was discarded. Next, 0.1 M NaOH (200 μL) was added to the precipitate, and after homogenization, the sample was centrifuged again. Bradford reagent (1 mL) [29] was added to an aliquot of supernatant from each sample (100 μL). The protein concentration was determined spectrophotometrically at 595 nm using a standard curve of bovine serum albumin (BSA; 2 to 20 μg).

2.8. Statistical Analysis

Chemoface software 1.67 [30] was used to analyze the data obtained in the 24 FACT screening experiments and in the 2² CCD. The inhibition assays were performed in triplicate each time, and the mean was used to calculate mUIT and % inhibition. Data for larval stage duration, pupal weight, pupal period duration, adult longevity, preoviposition period, oviposition period, and the number of eggs produced per couple were analyzed using analysis of variance, and the means were compared using the Scott–Knott test at the 5% significance level in the laercio package of R software 4.4.1 [31]. The protein concentration in the feces was analyzed using regression analysis in R software [31]. Insect survival over time was analyzed using the Weibull model and the survival package of R software [31].

3. Results and Discussion

3.1. Optimization of CLM Trypsin Inhibitor Extraction

The inhibition of S. frugiperda trypsin activity caused by the extracts obtained in the experiments (Table S3) was used as a response for the 2⁴ FACT design calculations. The data on the effects (absolute values) that each parameter caused on the extraction of trypsin inhibitors found in castor oil plant leaves and that were significant at 5% are shown in the Pareto chart (Figure S1). The P, T, and So parameters were significant at their lowest levels because the values of the first-order effects were negative. Additionally, second-, third-, and fourth-order interactions were significant, indicating that the factors affect each other during the extraction of the trypsin inhibitor present in CLM.

In the Pareto chart (Figure S1), effects that went beyond the p = 0.05 significance line on the right were considered significant for the extraction of trypsin inhibitors from castor oil plant leaves. X1 = Time; X2 = Re-extraction; X3 = Solvent; X4 = Proportion. Using the significant linear regression coefficients shown in Table S4, a linear regression equation was generated for the response surfaces (Equation (1)).

$$\begin{gathered} Inhibition=8.91\times {10}^4-311.98X_1-1.72\times {10}^4{X}_2-359.98X_3-1.66\times {10}^6{X}_4+1.02\times {10}^4{X}_1{X}_4+ \\ 242.61X_2{X}_3+2.78\times {10}^5{X}_2{X}_4 \end{gathered}$$

(1)

The analysis of variance (Table S5) for Equation (1) was not significant (Fcalc = 0.8509; p = 0.5983). The parameters used in the extraction strongly affected each other, as observed in Figure S2, constituting a complex system to be analyzed. The proposed equation only partially predicted how the factors influenced the extraction of the trypsin inhibitor.

Although the adopted model was not significant, response surfaces were created to assess which evaluated factors tended to increase or decrease the inhibition of S. frugiperda trypsin. These data indicated a linear increase in inhibition with the effect caused by T, P, and So (Figure S2), and in all observed interactions, the factors at their lowest levels showed increased inhibition. In the Pareto chart (Figure S1), T, P, and So were significant at the first-order level and in second-, third- and fourth-order interactions. The Refactor, although not significant at the first order in the extraction, showed better results according to the response surfaces at its lowest value in the interactions with R and So, but its interaction with T was the opposite (Figure S2A,D,F); its value for subsequent extraction was fixed at 1 re-extraction due to this observation. Factor T was significant; however, analysis of the response surfaces indicated that its effect on the extraction of the CLM trypsin inhibitor was only more significant than Re (Figure S2F), and it was therefore fixed at its lowest value (60 min) for subsequent extractions.

The inhibition of S. frugiperda trypsin activity caused by the extracts obtained in experiments 1 to 13 (Table S6) was used as a response for the 2² CCD calculations. Data on the effects (absolute values) that were significant at 5% are shown in the Pareto chart (Figure S3). The So parameter was significant, whereas P and the interaction between the parameters were not significant.

Using the significant linear regression coefficients shown in Table S7, a polynomial regression equation was generated for the response surfaces (Equation (2)).

$$Inhibition=-3.84\times {10}^5+3.58\times {10}^7{X}_1+1.63\times {10}^3{X}_2-8.72\times {10}^8{X}_1^2-11.41X_2^2$$

(2)

The analysis of variance (Table S8) for Equation (2) was significant (Fcalc = 5.217 p = 0.0258). The Fcalc for the lack of fit was only 173.2, a relatively no significant value close to the pure error of 5.75 × 106. The R² value did not reach a high degree of fit between the observed and predicted data; however, it was higher than that for the 2⁴ FACT design. Figure 1 shows the response surface generated by Equation (2), where the experimental values did not reach the maximum point of the curve, but the calculated maximum points via the first derivative of Equation (2) were So = 71.7% and P = 0.02057292 (1/48.6 w/v), i.e., the extraction was more efficient under these conditions.

The optimization of the extraction parameters showed that the ideal extraction conditions were 1 g of sample in 48.6 mL of 71.7% ethanol extracted for 60 min in an ultrasonic bath at 40 °C, with 1 re-extraction under the same conditions. Extraction under these conditions resulted in an extract with trypsin inhibitory activity of 94,837.14 mUIT g⁻¹ CLM.

The results found in the present study are more promising than those described in other studies that also evaluated the trypsin inhibition capacity of castor oil plants. One example is the study by [18], who extracted S. frugiperda trypsin inhibitors by macerating castor oil plant leaves and extracting them with distilled water at a ratio of 1:3 (w/v) at 4 °C under stirring for 30 min. The extract was centrifuged, and the supernatant showed trypsin inhibition of 2490 mUIT, 38 times lower than that obtained in the present study. In another study, it was possible to obtain S. frugiperda protein inhibitors from castor oil plant leaves using extracted S. frugiperda inhibitors found in castor oil plant leaves using ethanol at a ratio of 1:7 (w/v), with the ethanol remaining in contact with the leaves for 20 d and then filtered [24]. This extract was fractionated by normal-phase chromatography, and the highest inhibitory activity was found in the fraction eluted with ethanol at 2000 mUIT, 47 times lower than that obtained in the present study [19], where extracted S. frugiperda trypsin inhibitors from partially defatted castor oil plant cake for 30 min with water at a 1:40 (w/v) ratio. The trypsin inhibition of the extract was 21.23 mUIT g⁻¹ seed, and after fractionation by flash silica chromatography, trypsin inhibition of 8906.12 mUIT was found (10.7 times lower than that obtained in the present study).

3.2. Chromatographic Purification

From the extract obtained under standardized conditions and with greater trypsin inhibitory activity, fractions were obtained by flash silica chromatography and then evaluated for their inhibitory potential with a non-kinetic trypsin inhibition assay. The results, expressed as percent inhibition (Table S9), indicated that fractions 2 to 7 contained trypsin-inhibiting compounds. Due to their inhibitory potential, 5 μL of fractions 2 to 7 was eluted by analytical HPLC (Figure S5).

The results obtained by HPLC-AM, when analyzed with the fraction inhibition results, suggest that the peak with a retention time of 13.09 min may be responsible for trypsin inhibition. This peak was present in fractions 4, 5, and 6, and antitrypsin activity was observed in all these fractions. Although inhibition was also found in fraction 6, the observed peak showed some smaller peaks close to its base, indicating the presence of other compounds that would hinder purification because they elute at the same time. Therefore, fractions 4 and 5 were pooled.

Before pooling, aliquots (20 μL) of fractions 4 and 5 were eluted, and the peaks with the best resolutions were collected; their inhibitory potentials were evaluated in a non-kinetic inhibition assay (

Table 1. Spodoptera frugiperda trypsin inhibitory activity for the peaks of fractions 4 and 5.

Fraction 4		Fraction 5
Peak Retention Time	% Inhibition	Peak Retention Time	% Inhibition
7.00	70.69	13.01	62
10.92	68.68	15.73	43.18
13.38	69.35	17.07	63.87
14.30	70.47	-	-
16.93	61.07	-	-
21.33	62.98	-	-
24.61	68.68	-	-

). All peaks collected from both fractions showed trypsin inhibition.

After pooling fractions 4 and 5 (Fr 4-5), 20 µL was eluted to evaluate changes in the chromatographic profile and resolution of the main peaks (Figure S6). Pooling of the fractions allowed for the separation of the most expressive peaks because the profile did not change in either HPLC-AM or HPLC-PM, with an injection volume of 1 mL at a concentration of 50 mg mL⁻¹ (Figure S7), facilitating the collection of two apparently pure peaks with good resolution for identification and characterization.

3.3. Identification and Characterization by Nuclear Magnetic Resonance Spectroscopy and Mass Spectrometry

The chemical shifts and coupling constants obtained from the ¹H, COSY, HMBC, and HSQC NMR spectra for the compound collected after chromatographic separation (tr = 13.79 min) are shown in Tables S12–S15 and Figures S8–S11.

The observed values and the established correlations showed the presence of an adenine moiety in the structure under analysis, which corroborates literature data [32,33]. In the mass spectrum, the formed precursor ion (M + H)⁺ had a mass-to-charge ratio of m/z 268.1045 (Figure S12). The chemical shifts and the precursor ions obtained are consistent with the literature for the compound adenosine (Figure 2A), whose elemental composition is C₁₀H₁₃N₅O₄ [32,33]. This is the first report of the isolation, identification and characterization of adenosine from R. communis.

The chemical shifts and coupling constants obtained from the ¹H and ¹³C NMR spectra for the major compound collected after chromatographic separation (tr = 19.16 min) are shown in Tables S16–S18 and Figures S13 and S14.

The mass spectrum in ESI mode showed the peak of the precursor ion formed (M + H)⁺ at m/z 165.0664 (Figure S15). The set of information obtained with the shifts in the NMR analyses and the high-resolution mass spectra compared with the literature indicated that the major compound collected at tr = 19.16 min was a major alkaloid present in the castor oil plant, ricinine (Figure 2B), whose molecular formula is C₈H₈N₂O₂. The ¹H and ¹³C NMR shifts are shown in Table S11, and the structure of the molecule is presented in Figure 2B [34,35].

Ricinine is a neurotoxic alkaloid found in both castor oil plant leaves and seeds, and the symptoms of ingestion include liver and kidney damage, vomiting, and convulsions, which can lead to death. Therefore, several studies have attempted to understand and evaluate how ricinine acts on the body [34,35,36]. Some studies have also used ricinine as a possible drug for the treatment of amnesia and anti-acetylcholinesterase activity [37,38,39] and in the control of pests such as Spodoptera exigua (Hübner) (Lepidoptera: Noctuidae) and S. frugiperda [16,40,41].

One hypothesis is that the inhibition of S. frugiperda trypsin activity by ricinine is directly related to the catalytic site of this enzyme or the S1 pocket, responsible for the specificity of trypsin to substrates containing lysine or arginine, which are characterized by being positively charged. A triad formed by the residues aspartate 102, histidine 57, and serine 195 characterizes the catalytic site of trypsin. The catalytic mechanism in Figure 3 shows that the initial step of peptide bond cleavage involves the removal of the serine 195 proton by histidine 57 in conjunction with aspartate 102, leaving the serine 195 oxygen atom with a free negative charge to form a tetrahedral intermediate with the substrate [42].

Serine 195, in its anionic form, can bind to ricinine through the carbonyl carbon atom, which has a positive partial charge due to a double bond with oxygen, and the resonance of the aromatic ring stabilizes the additional negative charge of serine 195 (Figure 4).

The S1 pocket (Figure 5) has a negatively charged aspartate residue at the bottom, and therefore, trypsin specificity occurs through substrates containing positively charged groups. In addition to lysine and arginine, the other positively charged amino acid is histidine, but it is too large to fit into the S1 pocket because of the ring on the side chain. The binding and stabilization of ricinine in the S1 pocket occurs in the same way as that proposed for the catalytic site [43].

The two hypotheses regarding how ricinine binds and inhibits trypsin are feasible, and as shown, the inhibition of S. frugiperda trypsin by ricinine would be competitive because the bond formed is in equilibrium and, therefore, reversible.

Antitrypsin activity was only observed with ricinine, with 103.21 mUIT mg⁻¹ ricinine after purification. Adenosine showed no trypsin inhibitory activity.

3.4. Effect of Castor Oil Plant Leaf Extract on S. frugiperda Development

The survival of S. frugiperda larvae was reduced only by treatment with the highest concentration of CLM extract, 1818.2 μg mL⁻¹, which showed an LT₅₀ of 15.9 d and cumulative survival of 18.5% at the end of the evaluation period, as observed in Group 2 of Figure 6.

A positive relationship was observed between an increase in the concentration of CLM extract in the diet and the concentration of protein in the feces of S. frugiperda larvae (Figure S16). Thus, the increase in the protein concentration in feces may be directly related to the reduced ability of larvae to absorb protein, which corroborates the inhibitory effect of S. frugiperda trypsin activity observed in vitro.

The CLM extract at concentrations of 113.6, 227.3, and 454.5 μg mL⁻¹ did not cause changes in the assessed biological characteristics of S. frugiperda. However, at the CLM extract concentration of 909.0 μg mL⁻¹, there was an increase in the duration of the larval and prepupal periods and a reduction in pupal weight. Adult females that were fed 909.0 μg mL⁻¹ CLM extract exhibited decreased fecundity. Notably, at the highest CLM extract concentration, 1818.2 μg mL⁻¹, the highest percentage of mortality was detected (Figure 6), and there was also an increase in the duration of the larval and prepupal periods, in addition to a reduction of 22.3% in pupal weight. The high mortality caused by the 1818.2 μg mL⁻¹ CLM extract prevented the evaluation of biological characteristics in adult insects (Table S21).

The results obtained in the in vivo assay were not as promising as those obtained in the in vitro assay. This discrepancy can be attributed to the complexity of plant-insect interactions, as evidenced by the adaptive mechanisms of S. frugiperda to plant protease inhibitors [44,45]. However, the in vitro assay is promising for use in the selection of active metabolites, mainly because it can be performed in a shorter time interval and with less material.

Notably, throughout evolution, insects have developed different adaptation strategies to circumvent the negative effect of plant proteinase inhibitors [46,47,48]. The lower concentrations of CLM extract may have activated defenses in the insects and, therefore, did not negatively affect their development; however, the metabolic cost in insects treated with the highest concentrations was evident because of the reduced survival, increased prepupal period, and reduced number of eggs per couple.

The defense mechanisms that insects have against the secondary metabolites produced by plants seem to be associated with the overexpression of the inhibited proteinase, which increases its activity against the inhibitor and the synthesis of several isoenzymes, and some are not inhibited. These defense mechanisms cause damage to the insect, as they redirect amino acids essential for development toward the synthesis of digestive enzymes, thus causing reduced growth and development.

4. Conclusions

The extraction parameters optimized by chemometric methods increase the inhibitory activity of the extract. Adenosine and ricinine are present in the CLM extract. The peak containing adenosine has no antitryptic activity. Ricinine inhibits trypsin activity in S. frugiperda. The CLM extract negatively affects the development of S. frugiperda starting at a concentration of 900 μg mL⁻¹. In summary, the results of the present study demonstrate the potential of the R. communis extract to be used in management programs for this important pest.