Improvement of Bacillus thuringiensis Protein Contents with Increased Nitrogen Fertilizer Application in Gossypium hirsutum

Abstract

The insect resistance expression of Bacillus thuringiensis (Bt) cotton (Gossypium hirsutum L.) is unstable due to temporal and spatial variations in the Bt protein content in different organs and growth stages. The aim of this study was to improve the Bt protein content in cotton flowers and investigate the underlying physiological mechanism using biochemical analytical methods. In this study, a split-plot design with three replications was used. The main plots included two Bt cotton cultivars (a conventional cultivar, Sikang1 (S1), and a hybrid cultivar, Sikang3 (S3)), while five soil nitrogen application levels (CK (control check): normal level; N1: 125% of the CK; N2: 150% of the CK; N3: 175% of the CK; N4: 200% of the CK) constituted the subplots. The Bt protein content and related nitrogen metabolism parameters were measured. We found that the Bt protein content increased and then decreased with increasing nitrogen rates. It reached its maximum at N3, with significant increases of 71.86% in 2021 and 39.36% in 2022 compared to the CK. Correlation analysis indicated that the Bt protein content was significantly positively related to the soluble protein and free amino acid contents, as well as the GPT (glutamic pyruvic transaminase), GOT (glutamic oxaloacetic transaminase), GS (glutamine synthetase) and GOGAT (glutamate synthetase) activities. On the other hand, negative correlations were found between the Bt protein content and protease and peptidase activities. In addition, stepwise regression and path analysis indicated that the increased Bt protein content was mainly due to the enhanced GS and GOGAT activities. In summary, appropriately increasing nitrogen fertilizer application is a practical way to increase flower Bt protein content and insecticidal efficacy of Bt cotton. These findings provide an actionable agronomic strategy for sustaining Bt expression during the critical flowering period.

1. Introduction

Cotton (Gossypium hirsutum L.) is the most important fiber crop and is commercially grown worldwide, including in China [1]. According to the National Bureau of Statistics of China (NBSC), China’s cotton acreage reached 2.84 × 10⁶ hectares in 2024, with a total output of more than 6 million tons in that year. Transgenic Bacillus thuringiensis cotton (Bt cotton) is a cotton cultivar with insect-resistant characteristics, obtained by introducing the Bt toxin protein gene from Bacillus thuringiensis species into cotton [2]. Bt cotton is grown commercially worldwide due to the insecticidal properties of the entire plant, including the vegetative and reproductive organs [3,4]. Its application is conducive to reducing not only the cost of cotton cultivation but also the use of chemical pesticides, thereby enhancing labor safety and reducing environmental pollution [5,6,7].

However, the expression of insect resistance in Bt cotton varies significantly with environmental conditions [8,9], and an association has been shown between reduced cotton insect resistance and reduced Bt protein content [10,11]. A favorable temperature is crucial for crop growth, and low- or high-temperature stress not only affects the growth of Bt cotton but also the expression of Bt protein. Low-temperature stress at the square stage significantly inhibited the expression of Bt protein in cotton squares and their subtending leaves; this inhibition intensified with a decreasing temperature and more prolonged stress, with a maximum decrease in the Bt protein content of 30% recorded [7]. Similar findings were obtained when studying the effects of low temperatures on cotton bolls’ Bt protein content [12]. High temperatures cause a decrease in cotton’s Bt protein content as well. According to Wang, high temperatures below 38 °C resulted in a slight decrease in the Bt protein content in squares, but high temperatures above 38 °C resulted in a significant decrease in Bt protein [13]. Liu et al. [14] found that the longer cotton was exposed to high-temperature stress, the longer it took for its Bt protein concentration to return to normal levels. Bt cotton that had suffered water deficits showed reduced Bt protein content in its leaves, flowers and bolls [15], while a field soil water content below 60% of the field capacity caused a significant reduction in boll shells’ Bt protein content [16]. The combined stress from drought and high temperatures significantly reduced the Bt protein content in cotton bolls, with drought having a greater effect on the Bt protein concentration and insect resistance in Bt cotton [17]. Salt stress also significantly decreased Bt cotton’s insecticidal protein content [18,19]. It has been reported that this decrease in the Bt protein content under adversity may be related to a decrease in the soluble protein and total nitrogen content and NR, GS, GOT and GPT activities [20,21].

In addition to environmental factors, temporal and spatial variations in the Bt protein content in different organs and growth stages also result in changes in the cotton’s insect resistance [22,23]. In detail, temporally, it was found that the Bt protein content tended to decrease during the growth process and was lowest at the flowering and boll setting stage [11,24]; spatially, the Bt protein content was higher in vegetative than reproductive organs and the highest in leaves [22,25].

Therefore, it is essential to improve the stability of the Bt protein in Bt cotton. Several studies have previously been conducted by researchers to explore ways of enhancing the Bt protein content. According to these studies, agronomic measures, including fertilizer application, modulating the planting density, removing early fruit branches and applying plant growth regulators, can regulate the Bt protein content of Bt cotton [25,26,27,28]. For example, nitrogen application affects the activities of GOT and GPT, reducing the protein synthesis capacity in cotton plants and, in turn, affecting their Bt protein content and insect resistance [29]. Pettigrew and Adamczyk [30] concluded that the use of nitrogen fertilizer increased the chlorophyll and Bt protein content in leaves. Guo et al. [26] reported that moderate nitrogen fertilizer application was beneficial in promoting cotton plant development and Bt insecticidal protein expression.

However, previous studies mainly focused on squares, bolls and their subtending leaves, neglecting to study flowers [20,25]. As a reproductive organ, the flower has a very low Bt protein content, which makes it more vulnerable to bollworms, especially as the flower is one of the main targets of bollworms and other insects [31]. Bollworms can cause serious damage, affecting pollination and boll formation and resulting in boll abscission and ultimately yield reduction [32]. Thus, this study focused on increasing the Bt protein content in cotton flowers, which were treated with varying levels of nitrogen fertilizer application to explore its effects on the flower Bt protein content and related physiological mechanisms.

2. Materials and Methods

This study was carried out on a farm at Yangzhou University (32°30′ N, 119°25′ E; shown in Figure 1), China, during the cotton growth seasons in 2021 and 2022 [14]. We used a 525 m² test plot and two Bt cotton (Gossypium hirsutum L.) cultivars, the conventional cultivar Sikang1 (S1) and the hybrid cultivar Sikang3 (S3). S1 was developed through the hybridization of Siyang 269 × Guokang 22, while S3 was derived from a Siyang 231 × Siyang 8201 cross. Both cultivars carried the introduced Cry1Ac Bt gene. Furthermore, these were the primary commercial cultivars in China’s Yangtze River Valley cotton region during the early 2000s. The experimental site had sandy loam soil (typical fluvaquents and Entisols (U.S. taxonomy)). Specific indicators of the soil’s physicochemical properties are shown in

Table 1. The physicochemical properties of the soil in 2021 and 2022.

	Sand Grain Content (%)	Silt Content (%)	Clay Content (%)	Bulk Density (g·cm−3)	Field Capacity (%)	Organic Matter (g·kg−1)	Available N (mg·kg−1)	Available P (mg·kg−1)	Available K (mg·kg−1)
2021	62.7	25.9	12.2	1.42	22.9	2.17	110.2	38.6	76.3
2022	63.2	27.1	11.7	1.47	22.3	2.13	112.7	36.9	78.8

. Seeds were sown in the greenhouse on 13 April 2021 and 15 April 2022, and seedlings were transplanted to the field on 15 May 2021 and 17 May 2022. The planting density was 37,500 plants per hectare for S1 and 27,000 plants per hectare for S3 in both years.

Five levels of nitrogen application were used: 300 (CK, normal level), 375 (N1, 25% increase), 450 (N2, 50% increase), 525 (N3, 75% increase) and 600 (N4, 100% increase) kg·ha⁻¹. The soil was top-dressed with nitrogen (urea) after transplantation, at early flowering and at peak flowering at proportions of 20%, 20% and 60%, respectively. The application of K (375 kg ha⁻¹ provided as KCl) and P (600 kg ha⁻¹ provided as single superphosphate) was kept the same for all treatments, with proportions of 50% used both after transplantation and at early flowering. Intertilling and weeding were carried out twice during the growing period; irrigation was carried out according to the drought conditions.

A split plot design with three replications was used. The main plots were assigned to the two cultivars (S1 and S3), while the five nitrogen application rates (CK, N1, N2, N3 and N4) constituted the subplots.

2.1. Measurements

At the peak flowering stage, the first flowers were collected from the first position of the tenth fruiting branches of plants in the middle two rows of each plot. For each replication, 10 plants were selected, and one flower was collected from each. The samples were stored at −80 °C until later measurements.

The Bt protein content was analyzed using an immunological analysis, ELISA, according to Chen [33]. The free amino acid concentration was estimated using the ninhydrin assay described by Yemm [34], while the soluble protein concentration was determined using Bradford’s Coomassie Blue dye-binding assay [35]. Following the procedure described by Chen [7], 0.2 g of fresh sample was ground in 1.5 mL of a Tris–HCl buffer (0.05 mol·L⁻¹) at 4 °C, and the homogenate was centrifuged at 20,000× g for 20 min. The supernatant was used to determine the glutamic oxaloacetic transaminase (GOT) and glutamic pyruvic transaminase (GPT) activities. We followed the procedure described by Liu [36] and Liu [37] to obtain supernatants for the estimation of the glutamine synthetase (GS) and glutamate synthetase (GOGAT) activities: fresh samples (0.2 g) were ground in 1.5 mL of a 0.1 mol·L⁻¹Tris–HCl buffer (1 mmol·L⁻¹ of MgCl₂, 1 mmol·L⁻¹ of LEDTA and 10 mmol·L⁻¹ of 2-mercaptoethanol) at pH 7.6 and 4 °C, and then the homogenates were centrifuged at 13,000 r·min⁻¹ for 25 min. To obtain supernatants for protease activity determination using Vance’s method [38], the samples (0.8 g) were homogenized at 4 °C in 1 mL of a β-mercaptoethanol extraction buffer (a mixture of ethylene glycol, sucrose and phenylmethylsulfonyl fluoride at pH 6.8). Fresh samples (0.5 g) were homogenized in 8 mL of a Tris–HCl buffer (4 mM of DTT, 4 mM of EDTA, 1% PVP, pH 7.5) at 4 °C to measure the peptidase activity [39].

To verify the actual insecticidal activity in the two Bt cotton cultivars examined, we conducted a feeding experiment using S1 and S3 flowers treated with the CK and N3 in 2022. This experiment was performed on the second instar larvae of bollworms in an artificial climate chamber at the Jiangsu Key Laboratory of Crop Genetics and Physiology. The eggs were purchased from Beijing Bloomnost Biotechnology Co. (Beijing, China). The conditions in the climate chamber were set at 25–28 °C, 80% relative humidity and a photoperiod of 16:8 h (L:D). Three replicates were set up for each treatment, and 48 s instar bollworm larvae were fed for each replicate. Fresh flowers from S1 and S3 treated with the CK and N3 were collected for the experiment, and the feed was replaced every two days. The dead larvae were counted on the third (3 d) and fifth days (5 d) after feeding began to calculate the mortality rate.

2.2. Statistical Analysis

A two-way ANOVA (ANOVA 2) was performed in SAS 9.4. The means were separated using an LSD test at the 5% significance level, and Pearson’s correlation coefficient was used to measure the relationships in the data after normalization. The tables and figures were processed and plotted with Excel 2019 and Origin 2021, while stepwise regression and path analysis were performed using SPSS 26. The experimental units used for data analysis were the plot means. In addition, Q-Q (Quantile–Quantile) plots were used to test the data’s normality, which was shown to be good, with the data points falling roughly around the diagonal of the Q-Q plot. In order to test the variance homogeneity, a Levene test was carried out. The results indicated variance homogeneity (p = 0.134), and, therefore, a standard ANOVA could be performed on the data.

After the data were normalized using Z-score normalization, stepwise regression analysis was performed to establish the optimal linear regression equations. We used the Bt protein content (Y) as the dependent variable and the soluble protein (X₁) and free amino acid content (X₂) and the GOT (X₃), GPT (X₄), GS (X₅), GOGAT (X₆), protease (X₇) and peptidase activities (X₈) as the independent variables.

Path analysis quantified the causal relationships between key predictors (X) identified via the stepwise regression and the response variable (Y; the Bt protein content). This method simultaneously partitions direct and indirect effects, establishing a robust foundation for inferential statistical decisions in agronomic trait studies. The computational framework followed the work of Liu [27] and Mahmoud [40] with adaptations for cotton physiology data.

The formulas used in the path analysis were as follows:

P_yi = b_is_i/s_y; P_yij = r_ij × P_yi

where P_yi and P_yij represent the direct and indirect path coefficients, respectively. The diameter coefficient r_ij is the correlation coefficient between the independent variable i and dependent variable j; b_i is the partial regression coefficient of the dependent variable y on the independent variable i; and s_i and s_y represent the standard deviations of the independent variable i and dependent variable y, respectively.

3. Results

3.1. Flower Bt Protein Content Under Increased Nitrogen Fertilizer Application

The flowers’ concentration of Bt protein was significantly affected by nitrogen fertilization but not by the cultivar or the interaction between the cultivar and nitrogen fertilization in both 2021 and 2022. The increase in the Bt protein content was greater in 2021 than in 2022. The Bt protein content showed an increasing trend at first, peaking at N3 and decreasing at nitrogen levels exceeding this. In comparison to the CK condition, under N3 treatment, the Bt protein content increased by 71.86% in 2021 and 39.36% in 2021 (

Table 2. Effects of increased nitrogen fertilizer application on flowers’ Bt protein content (ng·g^–1 FW).

Treatment		Year
		2021		2022
S1		170.56 ± 44.77 a		161.03 ± 19.02 a
S3		200.95 ± 35.93 a		153.74 ± 19.76 a
CK		138.57 ± 14.24 d		132.07 ± 8.47 d
N1		164.6 ± 36.56 c		149.34 ± 8.61 c
N2		210.4 ± 25.88 b		162.68 ± 10.36 b
N3		238.15 ± 20.92 a		184.05 ± 11.62 a
N4		177.08 ± 25.35 c		158.78 ± 7.83 bc
S1CK		126.81 ± 5.8 e		135.63 ± 10.22 ef
S1N1		131.45 ± 1.39 e		152.86 ± 7.76 cde
S1N2		198.18 ± 28.38 abc		164.53 ± 5.37 bc
S1N3		234.83 ± 21.43 ab		186.94 ± 14.32 a
S1N4		161.55 ± 10.09 cde		165.18 ± 3.77 bc
S3CK		150.33 ± 7.63 de		128.5 ± 6.07 f
S3N1		197.74 ± 6.57 abc		145.81 ± 9.38 def
S3N2		222.62 ± 20.52 ab		160.84 ± 15.14 cd
S3N3		241.46 ± 24.54 a		181.17 ± 10.38 ab
S3N4		192.61 ± 27.95 bcd		152.38 ± 4 cde
Sources of Variation		p Value
Year		<0.001
	df	F Value	p Value	F Value	p Value
S	1	12.29	0.07	9.8	0.09
N	4	27.72 *	<0.001	20.72 *	<0.001
S × N	4	2.2	0.12	0.16	0.95

FW, fresh weight; S1, cultivar Sikang 1; S3, Sikang 3; CK, normal nitrogen application level (300 kg·ha⁻¹); N1, nitrogen application increased to 125% of CK (375 kg·ha⁻¹); N2, nitrogen application increased to 150% of CK (450 kg·ha⁻¹); N3, nitrogen application increased to 175% of CK (525 kg·ha⁻¹); N4, nitrogen application increased to 200% of CK (600 kg·ha⁻¹); S, cultivar; N, nitrogen application level. Different letters in the same column represent significant differences at p < 0.05. * Significant correlation at α = 0.05.

).

Nitrogen application had significant effects on the mortality of bollworm larvae. The mortality was higher on the fifth day after feeding began than on the third day. Furthermore, mortality under N3 conditions was 40.80% higher than under the CK condition on the third day and 36.61% higher on the fifth day. Mortality under S1 conditions was slightly higher than that under S3 conditions, but this was not significant. In addition, the interaction between the cultivar and nitrogen fertilization did not have any remarkable effects on the mortality (

Table 3. The mortality of second instar bollworm larvae fed on flowers in the laboratory in 2022 (%).

Treatment		Days After Feeding Began
		3 d		5 d
S1		44.44 ± 9.56 a		66.32 ± 12.18 a
S3		41.67 ± 9.95 a		61.81 ± 12.48 a
CK		35.76 ± 6.1 b		53.47 ± 5.84 b
N3		50.35 ± 5.5 a		74.65 ± 3.59 a
S1CK		37.5 ± 5.51 bc		55.56 ± 4.34 b
S1N3		51.39 ± 7.31 a		77.08 ± 2.09 a
S3CK		34.03 ± 7.32 c		51.39 ± 7.31 b
S3N3		49.31 ± 4.34 ab		72.22 ± 3.18 a
Sources of Variation	df	F Value	p Value	F Value	p Value
S	1	3.37	0.21	10.59	0.08
N	4	10.56 *	0.03	232.77 *	<0.001
S × N	4	0.02	0.88	0.06	0.82

S1, cultivar Sikang 1; S3, Sikang 3; CK, normal nitrogen application level (300 kg·ha⁻¹); N3, nitrogen application increased to 175% of CK (525 kg·ha⁻¹); S, cultivar; N, nitrogen level; 3 d, third day after start of feeding; 5 d, fifth day after start of feeding; S, cultivar; N, nitrogen application level. Different letters in each panel represent a significant difference at p < 0.05. * Significant correlation at α = 0.05.

).

3.2. Nitrogen Metabolism in Flowers Under Increased Nitrogen Fertilizer Application

The interaction between the cultivar and nitrogen fertilization had significant effects on the flowers’ soluble protein content in 2021 but not in 2022. The soluble protein content showed a trend of increasing and then declining with increased nitrogen fertilizer application in both 2021 and 2022. Compared to the results for S1CK, the soluble protein content was enhanced by 102.50% under treatment S1N3 and 67.57% under treatment S1N2 in 2021. However, in 2022, the soluble protein content was not significantly affected by the interaction between the cultivar and nitrogen fertilization or the cultivar, despite being remarkably affected by nitrogen fertilization. The highest value was observed for N3 and was 53.49% higher than the value for the CK (

Table 4. Effects of increased nitrogen fertilizer application on flowers’ soluble protein and free amino acid content.

Treatment		Soluble Protein Content (mg g−1 FW)				Free Amino Acid Content (mg g−1 FW)
		2021		2022		2021		2022
SK-1		0.65 ± 0.18 a		0.54 ± 0.11 a		5.47 ± 1.36 a		2.71 ± 0.82 b
SK-3		0.49 ± 0.15 b		0.55 ± 0.15 a		5.35 ± 1.03 a		3.66 ± 1.19 a
CK		0.38 ± 0.04 b		0.38 ± 0.08 c		4.43 ± 0.41 c		2 ± 0.6 c
N1		0.58 ± 0.13 a		0.47 ± 0.11 c		4.68 ± 0.32 c		2.66 ± 0.36 b
N2		0.65 ± 0.14 a		0.62 ± 0.05 ab		6.32 ± 0.54 b		4.24 ± 0.94 a
N3		0.64 ± 0.21 a		0.68 ± 0.11 a		7.03 ± 0.64 a		4.27 ± 0.86 a
N4		0.59 ± 0.23 a		0.58 ± 0.06 b		4.59 ± 0.71 c		2.74 ± 0.29 b
S1CK		0.4 ± 0.04 e		0.43 ± 0.01 bc		4.58 ± 0.03 cde		1.46 ± 0.11 e
S1N1		0.56 ± 0.12 cde		0.45 ± 0.07 abc		4.87 ± 0.31 cd		2.42 ± 0.31 d
S1N2		0.69 ± 0.06 abc		0.59 ± 0.03 abc		6.33 ± 0.63 b		3.41 ± 0.14 bc
S1N3		0.81 ± 0.11 a		0.66 ± 0.16 ab		7.54 ± 0.15 a		3.61 ± 0.3 b
S1N4		0.78 ± 0.14 ab		0.58 ± 0.05 abc		4.01 ± 0.15 e		2.63 ± 0.1 d
S3CK		0.37 ± 0.04 e		0.34 ± 0.1 c		4.29 ± 0.59 de		2.54 ± 0.14 d
S3N1		0.6 ± 0.16 bcd		0.49 ± 0.16 abc		4.49 ± 0.21 cde		2.9 ± 0.22 cd
S3N2		0.62 ± 0.2 abcd		0.65 ± 0.05 ab		6.31 ± 0.56 b		5.07 ± 0.32 a
S3N3		0.47 ± 0.12 de		0.7 ± 0.02 a		6.51 ± 0.44 b		4.93 ± 0.66 a
S3N4		0.4 ± 0.02 e		0.58 ± 0.09 abc		5.16 ± 0.51 c		2.85 ± 0.4 cd
Sources of Variation		p Value				p Value
Year		0.47				<0.001
	df	F Value	p Value	F Value	p Value	F Value	p Value	F Value	p Value
S	1	188.16 *	0.01	0.03	0.88	1	0.42	161.38 *	0.01
N	4	4.74 *	0.01	12.71 *	<0.001	46.03 *	<0.001	55.54 *	<0.001
S × N	4	3.68 *	0.03	0.81	0.53	5.23 *	0.01	4.7 *	0.01

FW, fresh weight; S1, cultivar Sikang 1; S3, Sikang 3; CK, normal nitrogen application level (300 kg·ha⁻¹); N1, nitrogen application increased to 125% of CK (375 kg·ha⁻¹); N2, nitrogen application increased to 150% of CK (450 kg·ha⁻¹); N3, nitrogen application increased to 175% of CK (525 kg·ha⁻¹); N4, nitrogen application increased to 200% of CK (600 kg·ha⁻¹); S, cultivar; N, nitrogen application level. Different letters in each panel represent a significant difference at p < 0.05. * Significant correlation at α = 0.05.

).

The interaction between the cultivar and nitrogen fertilization had significant effects on the flowers’ free amino acid content in both 2021 and 2022. The free amino acid content increased with the nitrogen fertilizer levels until they exceeded 525 kg·ha⁻¹ (treatment N3), at which point it declined. Compared to the results for S1CK, the free amino acid content was enhanced by 64.63% under treatment S1N3 and 51.75% under S3N3 in 2021. Moreover, similar trends were recorded in 2022 (Table 4).

In both years, the GOT activity was significantly affected by the cultivar and nitrogen fertilization, but not by their interaction. In 2021, the GOT activity of S3 was markedly higher than that of S1, with an enhancement of 50.40%, while, in 2022, the GOT activity of S3 was markedly lower than that of S1, with a reduction of 20.93%. The GOT activity increased with the level of nitrogen fertilization but decreased at levels exceeding 525 kg·ha⁻¹ (treatment N3). In comparison to the CK, the GOT activity for N3 was 80.96% higher in 2021 and 66.57% higher in 2022 (

Table 5. Effects of increased nitrogen fertilizer application on flowers’ GOT and GPT activities.

Treatment		GOT Activity (μmol g−1 FW h−1)				GPT Activity (μmol g−1 FW h−1)
		2021		2022		2021		2022
SK-1		6.23 ± 1.49 b		4.73 ± 0.89 a		6.36 ± 0.77 a		5.11 ± 1.15 b
SK-3		9.37 ± 1.55 a		3.74 ± 1.19 b		6.65 ± 0.93 a		6.36 ± 2.6 a
CK		5.99 ± 1.65 d		3.38 ± 0.7 c		5.28 ± 0.13 d		3.19 ± 0.85 d
N1		7.09 ± 1.6 c		3.59 ± 0.86 c		6.71 ± 0.66 b		4.53 ± 0.52 c
N2		8.48 ± 1.41 b		4.62 ± 1.17 b		6.71 ± 0.6 b		7.61 ± 1.17 a
N3		9.74 ± 1.75 a		5.63 ± 0.86 a		7.51 ± 0.28 a		7.36 ± 1.89 a
N4		7.7 ± 2.78 bc		3.96 ± 0.62 bc		6.31 ± 0.41 c		5.99 ± 1.45 b
S1CK		4.57 ± 0.65 e		3.96 ± 0.42 bcd		5.29 ± 0.19 e		3.82 ± 0.7 e
S1N1		5.66 ± 0.39 e		4.33 ± 0.07 bc		6.16 ± 0.22 cd		4.6 ± 0.55 de
S1N2		7.23 ± 0.37 d		5.02 ± 0.86 ab		6.24 ± 0.47 cd		6.6 ± 0.51 bc
S1N3		8.27 ± 0.55 d		6 ± 0.62 a		7.45 ± 0.41 a		5.75 ± 0.71 cd
S1N4		5.42 ± 1.08 e		4.36 ± 0.63 bc		6.66 ± 0.11 bc		4.78 ± 0.86 de
S3CK		7.4 ± 0.61 d		2.81 ± 0.22 d		5.27 ± 0.06 e		2.57 ± 0.4 f
S3N1		8.53 ± 0.3 cd		2.84 ± 0.44 d		7.26 ± 0.37 ab		4.45 ± 0.6 e
S3N2		9.73 ± 0.33 bc		4.22 ± 1.49 bcd		7.18 ± 0.06 ab		8.62 ± 0.31 a
S3N3		11.21 ± 0.95 a		5.25 ± 1.02 ab		7.58 ± 0.12 a		8.96 ± 0.82 a
S3N4		9.98 ± 1.6 ab		3.56 ± 0.3 cd		5.96 ± 0.2 d		7.21 ± 0.27 b
Sources of Variation		p Value				p Value
Year		<0.001				<0.001
	df	F Value	p Value	F Value	p Value	F Value	p Value	F Value	p Value
S	1	469.56 *	<0.001	18.98 *	0.05	6.17	0.13	46.42 *	0.02
N	4	16.65 *	<0.001	8.85 *	<0.001	54.89 *	<0.001	52.39 *	<0.001
S × N	4	1.37	0.29	0.27	0.89	11.32 *	<0.001	13.02 *	<0.001

GOT, glutamic oxaloacetic transaminase; GPT, glutamic pyruvic transaminase; FW, fresh weight; S1, cultivar Sikang 1; S3, Sikang 3; CK, normal nitrogen application level (300 kg·ha⁻¹); N1, nitrogen application increased to 125% of CK (375 kg·ha⁻¹); N2, nitrogen application increased to 150% of CK (450 kg·ha⁻¹); N3, nitrogen application increased to 175% of CK (525 kg·ha⁻¹); N4, nitrogen application increased to 200% of CK (600 kg·ha⁻¹); S, cultivar; N, nitrogen application level. Different letters in each panel represent a significant difference at p < 0.05. * Significant correlation at α = 0.05.

).

The activity of GPT was significantly affected by the interaction between the cultivar and nitrogen fertilization in both years. It was enhanced with an increased level of nitrogen fertilization but decreased at levels exceeding 525 kg·ha⁻¹ (treatment N3). In 2021, treatment S1N3 showed an increase of 40.83% compared with the value for S1CK, and treatment S3N3 showed an enhancement of 43.83% compared with the value for S3CK. Similar trends were observed in 2022 (Table 5).

In 2021, the activity of GS was markedly affected by the interaction between the cultivar and nitrogen fertilization. Compared with the activity of GS under treatment S1CK, an increase of 98.80% was observed under S1N3. In comparison with the results for treatment S3CK, an enhancement of 83.33% was recorded under S3N3. However, in 2022, the cultivar and nitrogen fertilization significantly affected the activity of GS, but their interaction did not. In 2022, the GS activity of S1 was 6.2% higher than that of S3. Moreover, the GS activity increased with the level of nitrogen fertilization but decreased at levels exceeding 525 kg·ha⁻¹ (treatment N3). The value for N3 was 113.75% higher than that for the CK, a significant difference. Similar trends were shown in the GOGAT activity as well (

Table 6. Effects of increased nitrogen fertilizer application on flowers’ GS and GOGAT activities.

Treatment		GS Activity (U g−1 FW min−1)				GOGAT Activity (U g−1 FW min−1)
		2021		2022		2021		2022
S1		1.31 ± 0.29 a		1.29 ± 0.37 a		0.93 ± 0.25 b		0.74 ± 0.15 b
S3		1.33 ± 0.3 a		1.21 ± 0.33 b		1.25 ± 0.23 a		0.88 ± 0.25 a
CK		0.87 ± 0.05 d		0.8 ± 0.07 d		0.75 ± 0.19 c		0.61 ± 0.05 d
N1		1.26 ± 0.02 c		1.07 ± 0.16 c		1.03 ± 0.22 b		0.69 ± 0.14 cd
N2		1.59 ± 0.06 b		1.46 ± 0.14 b		1.27 ± 0.19 a		0.88 ± 0.07 b
N3		1.65 ± 0.04 a		1.71 ± 0.17 a		1.34 ± 0.17 a		1.11 ± 0.23 a
N4		1.25 ± 0.05 c		1.22 ± 0.16 c		1.08 ± 0.25 b		0.76 ± 0.11 bc
S1CK		0.83 ± 0.03 f		0.78 ± 0.1 f		0.59 ± 0.12 e		0.57 ± 0.05 e
S1N1		1.27 ± 0.02 c		1.17 ± 0.14 de		0.85 ± 0.12 d		0.6 ± 0.02 e
S1N2		1.54 ± 0.03 b		1.48 ± 0.18 b		1.15 ± 0.09 c		0.88 ± 0.07 bc
S1N3		1.65 ± 0.03 a		1.8 ± 0.06 a		1.21 ± 0.11 bc		0.91 ± 0.11 b
S1N4		1.28 ± 0.01 c		1.22 ± 0.21 cd		0.87 ± 0.1 d		0.72 ± 0.03 cde
S3CK		0.9 ± 0.05 e		0.81 ± 0.06 f		0.9 ± 0.06 d		0.65 ± 0.02 de
S3N1		1.25 ± 0.01 cd		0.96 ± 0.09 ef		1.21 ± 0.07 bc		0.77 ± 0.17 bcd
S3N2		1.64 ± 0.03 a		1.44 ± 0.13 bc		1.39 ± 0.19 ab		0.87 ± 0.1 bc
S3N3		1.65 ± 0.06 a		1.62 ± 0.21 ab		1.46 ± 0.12 a		1.31 ± 0.04 a
S3N4		1.21 ± 0.04 d		1.22 ± 0.14 cd		1.29 ± 0.12 abc		0.8 ± 0.15 bcd
Sources of Variation		p Value				p Value
Year		<0.001				<0.001
	df	F Value	p Value	F Value	p Value	F Value	p Value	F Value	p Value
S	1	14.77	0.06	29.45 *	0.03	133.97 *	0.01	51.16 *	0.01
N	4	444.73 *	<0.001	31 *	<0.001	28.71 *	<0.001	23.98 *	<0.001
S × N	4	5.28 *	0.01	0.73	0.58	0.76	0.56	3.86 *	0.02

GS, glutamine synthetase; GOGAT, glutamate synthetase; FW, fresh weight; S1, cultivar Sikang 1; S3, Sikang 3; CK, normal fertilizer application level (300 kg·ha⁻¹); N1, nitrogen application increased to 125% of CK (375 kg·ha⁻¹); N2, nitrogen application increased to 150% of CK (450 kg·ha⁻¹); N3, nitrogen application increased to 175% of CK (525 kg·ha⁻¹); N4, nitrogen application increased to 200% of CK (600 kg·ha⁻¹); S, cultivar; N, nitrogen application level. Different letters in each panel represent a significant difference at p < 0.05. * Significant correlation at α = 0.05.

).

The activity of protease was significantly affected by the cultivar and nitrogen fertilization, but not by their interaction in 2021. Compared with the value for S1, the value for S3 increased by 15.29% in 2021. The protease activity showed a decreasing trend, with the lowest value occurring under treatment N4 and being 17.15% lower than the value under CK conditions. Furthermore, similar trends were recorded in 2022 (Table 6).

In 2021, nitrogen fertilization had remarkable effects on the peptidase activity, but the cultivar and the interaction between the two did not. The peptidase activity decreased with an increasing nitrogen fertilizer level. When this level reached 600 kg·ha⁻¹ (treatment N4), the peptidase activity showed a significant reduction of 35.94% relative to that in the CK. Similar trends were apparent in 2022 as well (

Table 7. Effects of increased nitrogen fertilizer application on flowers’ protease and peptidase activities.

Treatment		Protease Activity (μg g−1 FW h−1)				Peptidase Activity (μg g−1 FW h−1)
		2021		2022		2021		2022
S1		14.58 ± 1.62 b		12.27 ± 1.62 a		14.17 ± 2.47 a		9.3 ± 2.07 a
S3		16.81 ± 1.28 a		10.26 ± 1.28 b		14.35 ± 2.57 a		9.81 ± 2.71 a
CK		17.03 ± 1.66 a		13.75 ± 1.66 a		17.52 ± 0.55 a		12.77 ± 0.83 a
N1		16.58 ± 1.43 ab		12.64 ± 1.43 ab		15.22 ± 1.12 b		11.1 ± 1.55 b
N2		15.92 ± 1.09 b		11.19 ± 1.09 bc		14.1 ± 1.59 bc		8.93 ± 1.11 c
N3		14.82 ± 1.63 c		9.97 ± 1.63 c		13.69 ± 1.35 c		7.7 ± 0.94 cd
N4		14.11 ± 1.94 c		8.77 ± 1.94 c		10.78 ± 0.92 d		7.28 ± 1.24 d
S1CK		15.91 ± 0.4 ab		14.48 ± 0.4 a		17.64 ± 0.79 a		12.31 ± 0.37 a
S1N1		15.85 ± 1.58 ab		13.07 ± 1.58 ab		14.96 ± 1.46 ab		10.19 ± 0.88 bc
S1N2		15.24 ± 0.14 bc		11.97 ± 0.14 ab		13.51 ± 1.23 bcd		9.34 ± 0.86 cd
S1N3		13.41 ± 0.37 cd		11.92 ± 0.37 ab		13.44 ± 2.07 bcd		7.64 ± 0.34 de
S1N4		12.47 ± 1.04 d		9.89 ± 1.04 bc		11.3 ± 1.13 cd		7.02 ± 1.22 e
S3CK		18.15 ± 1.72 a		13.02 ± 1.72 ab		17.41 ± 0.31 a		13.24 ± 0.97 a
S3N1		17.31 ± 1.03 ab		12.2 ± 1.03 ab		15.47 ± 0.9 ab		12 ± 1.67 ab
S3N2		16.6 ± 1.25 ab		10.42 ± 1.25 bc		14.69 ± 1.94 ab		8.52 ± 1.35 cde
S3N3		16.24 ± 0.68 ab		8.01 ± 0.68 c		13.93 ± 0.32 bc		7.75 ± 1.44 de
S3N4		15.76 ± 0.49 abc		7.65 ± 0.49 c		10.26 ± 0.12 d		7.54 ± 1.48 de
Sources of Variation		p Value				p Value
Year		<0.001				<0.001
	df	F Value	p Value	F Value	p Value	F Value	p Value	F Value	p Value
S	1	20.27 *	0.05	84.55 *	0.01	0.07	0.82	4.83	0.16
N	4	11.06 *	<0.001	6.06 *	<0.001	26.35 *	<0.001	21.41 *	<0.001
S × N	4	1.34	0.3	0.52	0.72	0.78	0.55	0.93	0.47

FW, fresh weight; S1, cultivar Sikang 1; S3, Sikang 3; CK, normal nitrogen application level (300 kg·ha⁻¹); N1, nitrogen application increased to 125% of CK (375 kg·ha⁻¹); N2, nitrogen application increased to 150% of CK (450 kg·ha⁻¹); N3, nitrogen application increased to 175% of CK (525 kg·ha⁻¹); N4, nitrogen application increased to 200% of CK (600 kg·ha⁻¹); S, cultivar; N, nitrogen application level. Different letters in each panel represent a significant difference at p < 0.05. * Significant correlation at α = 0.05.

).

3.3. The Correlation Between the Flowers’ Bt Protein Content and Nitrogen Metabolism-Related Parameters

Significant positive correlations were detected between the flowers’ Bt protein content and soluble protein and free amino acid contents, with values of 0.36 (p = 0.306) and 0.91 * (p < 0.001) for their soluble protein content and 0.79 * (p = 0.006) and 0.66 * (p = 0.038) for their free amino acid content in 2021 and 2022, respectively. In addition, the flowers’ Bt protein level was significantly positively correlated with the activities of GPT, GOT, GS and GOGAT but negatively correlated with the activities of protease and peptidase. In 2021, the correlation coefficients for the Bt protein content and GOT and GS activities were 0.84 * (p = 0.002) and 0.85 * (p = 0.002), respectively; in 2022, they were 0.89 * (p = 0.001) and 0.96 * (p < 0.001), respectively (Figure 2 and Figure 3).

Stepwise regression analysis was performed to establish the optimal linear regression equations using the Bt protein content (Y) as the dependent variable and the soluble protein (X₁) and free amino acid content (X₂) and GOT (X₃), GPT (X₄), GS (X₅), GOGAT (X₆), protease (X₇) and peptidase activities (X₈) as the independent variables. The P values for all four models were less than 0.001, indicating that all four models were statistically significant. In addition, the AIC values of these models were calculated to determine which model was optimal. With the gradual introduction of the independent variables, the AIC values of the obtained regression equations increased (

Table 8. Statistics for regression equations regarding flowers’ Bt protein content.

Model	Regression Equation	p Value	AIC Value
1	Y = 4.138 × 10−16 + 0.881X6	<0.001	29.725
2	Y = 6.033 × 10−16 + 0.664X6 + 0.311X5	<0.001	26.698
3	Y = 7.235 × 10−16 + 0.259X6 + 0.386X5 + 0.412X3	<0.001	23.006
4	Y = 8.002 × 10−16 + 0.478X5 + 0.586X3	<0.001	22.632

). The lowest AIC value was achieved by Model 4. The regression equation using this model was statistically significant, and the selected indicators (X₅, X₃) were the main factors affecting the Bt protein content. The optimal linear regression equation, Y = 8.002 × 10⁻¹⁶ + 0.478X₅ + 0.586X₃, is shown in Table 8. In other words, an increased protein synthesis capacity, i.e., the increased GOT and GS activities, was the main reason for the increased Bt protein content (

Table 9. Coefficients resulting from stepwise regression equation.

Model	Independent Variable	Unstandardized Coefficients		β	t	p Value
		Sample Regression Coefficient	SE
4	Constant	0.001	0.089	-	0.012	0.991
	X5	0.479	0.106	0.479	4.527	<0.001
	X3	8.216	1.491	0.586	5.511	<0.001

X₅, GS activity; X₃, GOT activity. β indicates the standardized regression coefficient. t indicates the T-test used to indicate whether the regression equation was statistically significant.

).

In order to further clarify the effects on the Bt protein content of the indicators selected through stepwise regression analysis, a path analysis (Figure 4) was carried out for X₅ and X₃ with the Bt protein content (Y) as the dependent variable. The direct path coefficients of X₅ and X₃ for the Bt protein content were 0.479 and 0.587, respectively. Their effects on the Bt protein content were all positive, with X₅ exhibiting a greater effect. In conclusion, the increase in the flowers’ GS and GOT activities under increased soil N fertilizer application was the main reason for the increase in the Bt protein content.

4. Discussion

4.1. Appropriate Increase in Soil Nitrogen Fertilizer Application Favors Increase in Flowers’ Bt Insecticidal Efficacy

Several studies have reported the effects of nitrogen fertilizer application on the Bt protein content of Bt cotton [4]. Under a nitrogen deficit, the square Bt protein content was enhanced gradually with increasing nitrogen application in two cultivars studied [41]. Zhou et al. [4] found that spraying the whole cotton plant with a urea solution significantly increased the concentration of Bt protein in boll shells. In our study, more than the conventional amount of nitrogen fertilizer was applied to the soil. We found that this increased the flowers’ Bt protein content, which reached its maximum with nitrogen fertilizer application at 175% of the normal rate (N3, 525 kg·ha⁻¹); N3 is a high fertilizer application level and might not be agronomically or environmentally sustainable in all conditions. Hence, an appropriate increase in nitrogen fertilizer application favored an increase in the flowers’ Bt protein content, which improved their insecticidal efficacy. This indicated that the increase in the Bt protein content could mainly be attributed to the increased nitrogen fertilizer application. Furthermore, increased nitrogen fertilizer application benefits the cotton yield. For example, Wang et al. [42] reported that the application of nitrogen fertilizer not only promotes the uptake of nitrogen, phosphorus and potassium, as well as the rational distribution of cotton, but also enhances the amount of nitrogen accumulation and biomass of cotton. These factors ultimately lead to an enhanced cotton yield. Therefore, the increased application of nitrogen fertilizer is a practical measure, as it improved both the Bt insecticidal efficacy and cotton yield in Bt cotton. However, applying too much nitrogen fertilizer might cause excessive vegetative growth in cotton, increasing the abscission of squares and bolls and reducing the yield [43]. Methods to promote the synergistic expression of flowers’ Bt insecticide resistance and the yield by increasing nitrogen fertilization need to be investigated further.

4.2. GS and GOT Activities as Decisive Factors Determining Flowers’ Bt Protein Content

A number of studies have investigated the factors and related mechanisms affecting the Bt protein content. One study demonstrated that cotton-derived tannins are capable of binding to Bt proteins, resulting in conformational changes that inactivate these proteins [44]. Additionally, a decrease in the expression levels of insecticidal proteins has been associated with the methylation of Bt genes’ promoter regions, causing the proteins to lose their insecticidal activity [45,46]. In addition, our other studies focusing on reproductive organs, such as squares and bolls, have shown that changes in the Bt protein content, as part of the total protein content, are closely related to nitrogen metabolism [19,29]. In this study, changes in the Bt protein content were directly related to nitrogen metabolism. Our results suggested that increased nitrogen fertilizer application primarily caused the flowers’ Bt protein content to increase due to the upregulation of protein anabolic enzymes and downregulation of catabolic enzymes. Moreover, the correlation analysis indicated that the soluble protein and free amino acid content and GOT, GPT, GS and GOGAT activities were significantly and positively correlated with the Bt protein content, whereas the protease and peptidase activities showed negative correlations with the Bt protein content (Figure 2 and Figure 3). This analysis further confirmed the close relationship between the Bt protein content and nitrogen metabolism, with increases in the Bt protein content mainly caused by the enhancement of protein synthesis and weakening of catabolism. In addition, stepwise regression and path analysis revealed significant changes in the activities of GS and GOT related to the Bt protein content, and both showed strong positive effects (Figure 4). In other words, among all the indicators measured in our study, the flowers’ GS and GOGAT activities were the main contributors to increases in the Bt protein content. This is different from Liu’s [29] results, as they believed that the free amino acid content and peptidase activity had the biggest effects on the Bt protein content. It is possible that the year in which they conducted their experiment and the different treatments used caused our results to differ from theirs. In future studies, we plan to further investigate the relationship between the genes associated with these two enzymes and the Bt gene at the molecular level. This will enable us to reveal the underlying mechanism by which they alter the Bt protein content.

5. Conclusions

Increased nitrogen fertilizer application significantly increased the flowers’ Bt protein content, which reached a maximum at N3 (a 75% increase), and this change was closely related to nitrogen metabolism. The soluble protein and free amino acid contents increased under increased nitrogen fertilizer application conditions; similarly, the GOT, GPT, GS and GOGAT activities were elevated, but the protease and peptidase activities decreased. Stepwise regression and path analysis indicated that, among all the indicators, the GS and GOT activities were the most essential in determining the Bt protein content. In addition, because the response of the Bt protein content to changes in nitrogen application did not differ significantly between the two varieties examined (the conventional cultivar S1 and the hybrid cultivar S3), it was feasible to increase the Bt protein content in both by enhancing the soil nitrogen fertilizer application. Based on the results of this study, we recommend that smallholders plant conventional cultivars when aiming to improve Bt cotton’s insect resistance through appropriate increases in nitrogen fertilizer application. This is because, in actual production, hybrid cultivar seeds are more expensive than those of conventional varieties, so planting conventional cultivars can not only improve insect resistance but also decrease the production cost, increasing smallholders’ income. To sum up, under specific agronomic conditions, appropriately increasing nitrogen fertilizer application may be a practical way to increase flowers’ Bt protein content and insecticidal efficacy.