Quantitative Proteomics-Based Analysis Reveals Molecular Mechanisms of Chilling Tolerance in Grafted Cotton Seedlings

Abstract

Proteome analysis of grafted cotton exposed to low-temperature stress can provide insights into the molecular mechanistic of chilling tolerance in plants. In this study, grafted and non-grafted cotton plants were exposed to chilling stress (10 °C/5 °C) for 7 d. After the stress, rootstock and scion samples were labeled by 8-plex iTRAQ (isobaric Tags for Relative and Absolute Quantification), followed by two-dimensional liquid chromatography separation and tandem mass spectrometry identification. In total, 68 differential proteins were identified that were induced by low-temperature stress and grafting, and these proteins regulate physiological functioning. Under low-temperature stress, in the cotton seedlings, the proteins responded to the MAPK signaling pathway and calcium signaling pathway enhanced, the metabolisms of carbohydrate, lipid, nucleotide, and amino acid had a tendency to intensify, the proteins related to protein folding and degradation were activated, along with the system of antioxidant enzymes to offset cellular oxidative damage. In contrast, chilling stress reduced oxidative phosphorylation, photosynthesis, and carbon fixation. These data indicated that the physiological changes in cotton seedlings comprise a complex biological process, and the ability of plants to resist this stress can be improved after grafting onto a vigorous rootstock, although this was not obvious in the young plants. Further studies of low-temperature stress and/or graft-related differences in proteins could lead to the identification of new genes associated with chilling tolerance in plants. These data provide the basis for further studies on the molecular mechanism of chilling tolerance and the relationship of grafting and chilling tolerance in cotton.

1. Introduction

Low-temperature stress, including chilling and freezing injury, is a global problem for many field crops. Since 1980, crops on approximately 3.3 million hectares are annually affected by low temperatures in China, accounting for 7.4% of the country’s cropping area and up to 10% yield losses [1]. Cotton (Gossypium spp.) originates from the tropics and subtropics and is a cold-sensitive plant [2]. The optimum temperature for cotton growth and development is 22–32 °C. When the daily temperature falls below 15 °C, it affects the growth and development of cotton through chilling injury [3]. Low temperature disrupts normal plant growth and developmental processes, reducing overall lint yield and quality [4,5]. By comparing cotton cropping systems in Xinjiang, it was found that expansion of cotton-growing areas in the higher altitudes is restricted by low-temperature [6].

Low temperature is a potent abiotic stress and adversely affects seedling germination and establishment, resulting in inflicting a substantial yield penalty on many important field crops. Cotton is extremely sensitive to low-temperature stress, particularly during vegetative development and reproduction [7]. Cold stress thermodynamically depresses the kinetics of many physiological and biochemical processes, i.e., delays germination, seedling vigor reduces starch metabolism and respiration rate in plants, leading to substantial yield losses [8].

Under low-temperature stress, the cell membrane is the earliest to respond, i.e., increased cell permeability and cell membrane fluidity [9]. Low temperature also affects cell physiology, imbalances DNA stability and RNA secondary structures, impairs protein biosynthesis, enzyme reactivity, and photosynthetic performance [10,11,12]. Proteomic approaches are powerful tools to study plant responses to environmental stresses. Particularly, proteomics techniques combined with mass spectrometry (MS) can detect translational and post-translational regulation of different proteins [13,14]. Plants respond to environmental stress by modulating the expression or synthesis of proteins [15,16,17]. Thus, changes in plant proteome are a response to abiotic stresses. Proteomics techniques have widely been used in several crops for understanding their responses to abiotic stresses [18,19]. e.g., low-temperature stress tolerance in corn and sunflower crops [20,21]. Physiological and biochemical changes in cotton under low-temperature stress are complex, and proteins usually act together in the context of cellular networks rather than displaying their functions in an isolated manner [22].

Currently, cotton grafting is primarily performed for improving crop growth and yield under stressful environments [23], including low-temperature stress [24,25,26]. Our previous research [27] indicated that chilling tolerance in annual cotton could be achieved by grafting them onto chilling tolerant perennial cotton species. This chilling tolerance in the scion of grafted cotton was positively correlated with total soluble protein contents [28]. In this study, we used iTRAQ (isobaric Tags for Relative and Absolute Quantification), a high-throughput protein quantification technology to analyze the differential expression of proteomes in the scion and rootstock of grafted cotton seedlings under low-temperature stress, with an aim to understand the molecular mechanism of induced chilling tolerance.

2. Materials and Methods

2.1. Plant Materials, Growth and Sampling

The perennial Gossypium barbadense var. ‘arboreal cotton 113′ with a strong cold tolerance was used as the rootstock, and G. hirsutum cv. ‘Kangdi 99F₁′ (Bred by Xinjiang Kangdi Agricultural Technology Development Co., Ltd., Wulumuqi, China), a relatively cold-sensitive cultivar, was used as the scion. Seedlings of the above two materials and grafted seedlings of Kangdi 99F₁/arboreal cotton 113 (scion/rootstock) are presented in Figure 1.

Disease-free and uniformly sized seeds of the selected cotton materials ‘arboreal cotton 113′ and ‘Kangdi 99F_1′ were sown in plastic pots. Fourteen-day-old seedlings (~15 cm tall) were grafted. Two weeks after the grafting, fifteen seedlings were taken from each of the three materials (grafted seedlings, rootstock seedlings, and scion seedlings) using a randomized three blocks experiment design before treating them at day/night temperatures of 25 °C/20 °C (the controlled group) and 10 °C/5 °C (the treated group) for 7 d in two separate climate chambers, with a daily photoperiod of 12 h/12 h. After the treatment, the stem bark of three representative seedlings was taken directly from the self-rooted (S1, S1′, S2, and S2′), and grafted seedlings (S3, S3′, S4, and S4′). The samples were divided into rootstock and scion sampling (

Table 1. Two groups of stem bark samples and their serial numbers.

Stem Bark	Scion of Non-Grafted Seedling	Rootstock of Non-Grafted Seedling	Scion of Grafted Plant	Rootstock of Grafted Plant
Chilling stress	S1	S2	S3	S4
Control	S1′	S2′	S3′	S4′

) with 2 replicates for protein quantification [29,30].

2.2. Protein Extraction

The main stems of the scion and rootstock of grafted and non-grafted seedlings were harvested and the leaves removed. After washing with tap water, the gauze was wiped, and the leaves were washed with deionized water 3 times. The dried gauze was placed in a petri dish on an icebox.

2.2.1. Grinding

Polyvinylpolypyrrolidone (PVPP) was added into the stem bark by 10% of its weight, and then they were ground into powder in liquid nitrogen.

2.2.2. RIPA (Rapid Immunoprecipitation Assay) Lysis

Rapid immunoprecipitation assay solution was prepared by adding an appropriate amount of protease inhibitor phenylmethylsulphonyl fluoride (PMSF) into Beyotime^TM RIPA buffer (50 mM Tris-HCl pH 7.4, 1 mM EDTA, 150 mM NaCl, 1% Triton X-100, 1% Sodium deoxycholate, and 1% SDS) to make 1 mM PMSF solution. Then, 0.5 g of each sample was added to 1 mL RIPA, and the samples were repeatedly eddy-oscillated and then centrifuged for 20 min at 4 °C, 30 000× g. The supernatant was transferred to a clean tube and set it aside.

2.2.3. Protein Concentration Detection

Fifty μL of the collected supernatant from each sample was used for quantifying protein concentration using a BCA (bicinchoninic acid assay) Protein Assay Kit (Beyotime Biotechnology Company, Shanghai, China), and the process was repeated 3 times. A standard curve of protein content was generated through this BCA (Figure 2).

2.2.4. TCA/Acetone Precipitation

In this step, 150 μL of 100% TCA and 10 μL of β-mercaptoethanol (β-ME) were added to 1 mL of the collected supernatant of each sample, and then mixed and precipitated at −20 °C for 1 h. The mixture was centrifugated at 4 °C, 100,000× g for 15 min. The supernatant was discarded, and the precipitate was washed with 1 mL pre-cooled acetone at 4 °C and centrifuged at 100,000× g for 15 min. These washing steps were repeated 3 times and then the acetone was removed. Finally, the tubes were turned upside down on a filter paper to drain the liquid and covered tightly to prevent drying particles.

2.3. Off-Line Strong Cation Exchange Chromatography

The iTRAQ^TM labeled peptides were concentrated by vacuum dryer, mixed, and acidified to obtain a total volume of 2.0 mL. These samples were then injected into an Agilent 1100 HPLC (High-Performance Liquid Chromatography) system (Palo Alto, CA, USA) with a Zorbax 300-SCX column (4.6 ID ×250 mm) (Agilent, Waldbronn, Germany). Solvent A contained 25% ACN (pH = 3.0) and 5 mM KH₂PO₄, whereas solvent B was 350 mM KCl in solvent A. Peptides were eluted from the column with a linear solvent B from 0% to 100% in 40 min. A total of 30 fractions (1 mL each) were collected and the samples were dried in a Speed-Vac concentrator before LC-MS/MS analysis.

2.4. On-Line Nano-LC ESI QqTOF MS Analysis

A Nanobore LC system (Dionex, Sunnyvale, CA, USA) connecting a QSTAR XL QqTOF mass spectrometer to a Nanospray ion source (Applied Biosystems, Foster City, CA, USA) was used for tandem mass spectrometry. A Magic-C18 column with 75 µm inner diameter and 15 cm length (Michrom Bioresources Inc. Auburn, CA, USA) was packed with 5 µm particles with 100 Å pore size in house. Solvent A contained 5% ACN, 0.1% formic acid (FA) and 0.01% trifluoracetic acid (TFA), while solvent B contained 85% ACN, 10% isopropanol, 0.075% FA and 0.0075% TFA. At a flow rate of 250 nL/min, the peptide mixture (reconstituted in 120 µL of 5% FA) was injected and eluted from the column with a mobile phase solvent B gradient in 110 min (5% B for 5 min, 5–18% B for 10 min, 18–30% B for 65 min, 30–60% B for 10 min, 60–90% B for 10 min, and 90% for 5 min). The mass spectrometer was run in the information-dependent acquisition (IDA) mode to query MS data (m/z 350–1500) using a 1 s survey scan. Ions were selected for MS/MS analysis based on ion intensity (>25 counts/million) and charge state (+2, +3, and +4). Three production scans (2, 3, 3 s each) were set for each survey scan. The rolling collision energies were automatically chosen according to the m/z and charge state of the selected precursor ions. The information dependent acquisition Extensions II script was set to one repetition before dynamic exclusion [31].

2.5. Protein Identification and Quantification

A plant proteome database (PPD), which formed the theoretical searching database, was generated by searching “plant” in the protein database of NCBI. Protein identification and quantification were performed using ProteinPilot^TM software (SCIEX, Washington, DC, USA) integrated with PPD. The digestion agent trypsin was specified for all the searches. Only one missing cleavage site was allowed (default settings were used). Peptide identification confidence was assessed by a query mass list (experimental) matched against the digested peptide under the hypothesis that the query mass arises from either a y or b ion. The ProtScore was used and calculated from the percent confidence of a matched peptide, ProtScore -log (1-Percent Confidence/100). The ProtScore value of 99% confidence peptide was 2 and the ProtScore of the 66% confidence peptide was 0.47. The protein ProtScore was added from the ProScore of a set of peptides with strict criteria as follows: (i) only the maximal confidence instance of each distinct sequence counted toward ProtScore; (ii) Each peptide contributed not more than 2.0 (99% confidence) to the ProScore; (iii) Only tryptic peptide contributed to the ProScore. There were two types of ProtScore reported for each protein. The unused ProtScore was calculated using only peptides from spectra that were not used by winning proteins with higher scores. The total Protscore was calculated using all of its peptides. It did not indicate the confidence percent for the identification of a protein. The cutoff value of confidence (ProtScore) was 1.3. The Pro Group Report was used to eliminate redundancy. Only the winning proteins in the group were reported. The false-positive rate was estimated to be 3.87%, which was determined from the decoy (reversed) database searching (

Table 2. ID Statistics (Protein-threshold) based on plant protein database.

Unused (Conf) Cutoff	Proteins Detected	Proteins Before Grouping	Distinct Peptides	Spectra Identified	% Total Spectra
>2.0 (99%)	50	6207	562	3885	8.1
>1.3 (95%)	89	11,242	748	4659	9.7
>0.47 (66%)	143	22,092	951	5419	11.3
Cutoff Applied: >0.05 (10%)	768	37,568	2140	8006	16.7

) [31].

For protein relative quantification, only the specific MS/MS spectrum that was unique to a specific protein and the sum of signal-to-noise ratios for all peak pairs greater than 9 could be used for quantification (the default settings of the software). The relative quantification ratios of the identified proteins were calculated, averaged, and any systematic errors in the labeling of the iTRAQ^TM peptides were corrected. The accuracy of each protein ratio was given by the “error factor” calculated in the software, and a p-value to evaluate whether there was significant differential expression of proteins. The “error factor” represented the 95% uncertainty range (95% confidence error) of the reported ratio, where the 95% confidence error was the weighted standard deviation of the weighted mean of log-ratios multiplied by degree of freedom n-1 of Student’s t-factor, where n was the number of peptides that contribute to protein relative quantification. The p-value was determined by calculating Student’s t-factor by dividing the weighted mean of Log Rations–log Bias and the weighted standard deviation. To identify the expression differences between proteins, each experimental run was initially considered separately. Moreover, in order to be identified as significantly differential expressed, a protein must be quantified with at least three spectra generating a p-value of <0.05, and have a ratio of fold change >1.2 and <0.8 in both experiments (

Table 3. ID Statistics (Protein-threshold) based on protein database of Malvaceae.

Unused (Conf) Cutoff	Proteins Detected	Proteins before Grouping	Distinct Peptides	Spectra Identified	% Total Spectra
>2.0 (99%)	23	288	429	2458	5.1
>1.3 (95%)	30	326	490	2670	5.6
>0.47 (66%)	41	607	616	3091	6.4
Cutoff Applied: >0.05 (10%)	54	650	736	3393	7.1

) [32].

3. Results

3.1. The Overview of Differential Proteins Based on Categories of Pathway

The categories of 68 significantly different proteins according to their synthetic pathways, with reference to the KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway (Kyoto, Japan), are shown in

Table 4. Categories of all different proteins based on their pathways.

Biology Pathway		ID	Sum
Carbohydrate Metabolism	Glycolysis/Gluconeogenesis	ABW21688; BAD08850; ABK95670; ABK25148; BAA02729	18
	Citrate cycle (TCA cycle)	XP_002285356; XP_002263337
	Pentose phosphate pathway	CAB87149; XP_002271514; BAA02729
	Pentose and glucuronate interconversions	XP_002526594; XP_001759864; O04886
	Fructose and mannose metabolism	ABK95670; BAA02729; Q9SID0
	Galactose metabolism	XP_002526594; XP_001759864
	Starch and sucrose metabolism	O04886; Q9SID0; XP_002526594; XP_001759864
	Amino sugar and nucleotide sugar metabolism	ABK94952; XP_002526594; XP_001767949; XP_001759864; Q9SID0; NP_187685
	Pyruvate metabolism	CAO21229; XP_002285356
	Glyoxylate and dicarboxylate metabolism	XP_002285356; XP_002263337; ABE26954
	Inositol phosphate metabolism	ABK95670
Energy Metabolism	Oxidative phosphorylation	Q01859	16
	Photosynthesis	AAS59949; ABG74856; Q6ENJ7; ABK95197; ACG24344; Q8MAV7; NP_683825; CAN64486; ACJ85810; Q01859; CAO47122
	Carbon fixation in photosynthetic organisms	ABK95670; BAA02729; XP_002285356; ABE26954
	Methane metabolism	P85433
Lipid Metabolism	Glycerolipid metabolism	AAT01325	2
	Glycerophospholipid metabolism	ACG70839
Nucleotide Metabolism	Purine metabolism	Q96559	1
	Pyrimidine metabolism	Q96559
Amino Acid Metabolism	Aspartate metabolism	CAB81805	6
	Threonine metabolism	ABV27483
	Cysteine and methionine metabolism	Q42699; BAF28208
	Glutathione metabolism	ABO41849
	Phenylalanine metabolism	P85433
	Selenoamino acid metabolism	BAF28208
Folding and Degradation	Chaperonin and isomerase	CBC02991; ACG33102; ACB70177; ABY65001; ABX76300; O49886; ABK95670; EAZ30819; CAZ91440	11
	Ubiquitin mediated proteolysis	ABR16377; NP_568552;
Transport	Vesicular transport	XP_001780193; ABK92828; Q01111; BAB90396	8
	Annexin	1N00_A; CBC30882
	Phosphotransferase system	ABK92937
	Transfer protein	ACI26701
Signal Transduction	MAPK signaling pathway	ABW34390	6
	Calcium signaling pathway	1N00_A; ACG33102; ABY65001;Q7Y052; Q9ZSW9
Antioxidase	Superoxide dismutase	ACC93637	6
	Peroxidase	CBC66557; P85433
	Oxidoreductase	ABK25148; XP_002303520; CAO47501
Other Process	Cell cycle	XP_002446509	12
	Porphyrin metabolism	NP_171847
	Spliceosome	XP_002461966
	Chromosome scaffold	XP_001690796
	Ribosome	P41099
	Biosynthesis of alkaloids	ABW21688; BAD08850; BAA02729; XP_002285356; XP_002263337
	Biosynthesis of phenylpropanoids	ABW21688; BAD08850; BAA02729; XP_002285356; XP_002263337; P85433
	Biosynthesis of plant hormones	ABW21688; BAD08850; BAA02729; XP_002285356; XP_002263337; Q42699
	Biosynthesis of terpenoids and steroids	ABW21688; BAD08850; BAA02729; XP_002285356; XP_002263337

. The analysis indicates that these proteins are involved in at least 46 biological pathways, with some enzymes being associated with multiple pathways in the KEGG. Most of the differential proteins were assigned to nine categories according to the KEGG pathway, and the proteins with no classification or only in a single subclass of metabolism were assigned to other categories.

Among the 68 (duplicates involving different categories of the pathway were not counted) significantly different proteins, those involved in carbohydrate and energy metabolism had the most significant variation, accounting for 40%, of which 18 and 16 proteins were involved in carbohydrate and energy metabolism, respectively. In proportion, 12.94%, 9.41%, 7.06%, 7.06%, 7.06%, 2.35%, and 1.18% of the proteins were related to protein folding and degradation, transport, antioxidase, amino acid metabolism, signal transduction, lipid metabolism, and nucleotide metabolism, respectively. In addition, 12.94% of the proteins were related to other biological processes (Figure 3).

3.2. Induced Changes by Chilling Stress and Grafting in Cotton Proteins

Compared with control, chilling stress significantly increased the expression of total proteins in cotton bark under all conditions. Under non-stressed conditions, grafted rootstock of have significantly (p < 0.05) higher protein compared with non-grafted rootstock. However, there was no significant difference between the protein contents of the grafted and non-grafted seedlings after chilling stress (

Table 5. Changes induced by chilling stress and grafting in cotton proteins.

Group	Sample	Material	Protein Concent (mg/mL)	p < 5%	p < 1%
Control	S1′	Scion of non-grafted seedling	0.28 ± 0.02	d	D
	S2′	Rootstock of non-grafted seedling	0.41 ± 0.03	c	CD
	S3′	Scion of grafted plant	0.29 ± 0.03	d	D
	S4′	Rootstock of grafted plant	0.51 ± 0.03	b	BC
Chilling	S1	Scion of non-grafted seedling	0.55 ± 0.01	b	B
	S2	Rootstock of non-grafted seedling	0.93 ± 0.00	a	A
	S3	Scion of grafted plant	0.57 ± 0.06	b	B
	S4	Rootstock of grafted plant	0.89 ± 0.11	a	A

Note: Different lowercase letters in the same column indicate significant differences at the level of 0.05, and different uppercase letters in the same column indicate significant differences at the level of 0.01.

).

3.3. Pathways of Differential Enzymes Referred and the Key Enzymes

Enzymes are the key regulators of many metabolic pathways in living organisms. In this study, 18 differential enzymes (enzyme ID 4.2.1.11, 1.2.1.12, 5.3.1.1, 4.1.2.13, 1.1.1.37, 4.2.1.3, 2.7.7.9, 2.7.1.4, 3.1.1.11, 3.2.1.55, 4.1.1.39, 3.6.3.14, 1.11.1.7, 2.7.4.6, 2.1.1.14, 3.3.1.1, 5.2.1.8, 1.15.1.1) were found in cotton bark tissues using a Pathway Hunter Tool (http://pht.tu-bs.de/) accessed on 13 July 2018. The identified proteins mainly regulate metabolic pathways, e.g., carbohydrate and amino acid synthesis. In addition, the top 10 choke points for enzymes based on the metabolic pathways of Gossypium arboretum (tree cotton) (

Table 6. Analysis of top 10 choke points for enzymes based on the metabolic pathways of Gossypium arboretum.

Protein ID	Enzyme ID	Enzyme Name	Incoming Degree SP	Outgoing Degree SP
BAA02729	4.1.2.13	fructose-bisphosphate aldolase	57	57
Q96559	2.7.4.6	nucleoside-diphosphate kinase	31	31
BAD08850	1.2.1.12	glyceraldehyde-3-phosphate dehydrogenase	18	18
Q9SID0	2.7.1.4	Fructokinase	15	15
ACC93637	1.15.1.1	superoxide dismutase	8	8
XP_002526594	2.7.7.9	UDP glucose pyrophosphorylase	8	8
P85433	1.11.1.7	Peroxidase	7	7
BAF28208	3.3.1.1	Adenosylhomocysteinase	6	6
ABW21688	4.2.1.11	enolase, phosphopyruvate hydratase	6	6
XP_002263337	4.2.1.3	aconitate hydratase	6	6

) were used for analyzing the corresponding gene functions, expression levels, and regulation.

3.4. Functional Classification of Cold-Responsive Proteins

Differential analysis was performed to understand the relative change in the expression of cotton proteins in response to chilling stress and grafting.

3.4.1. Differential Proteins Related to Carbohydrate Metabolism

Sixty-eight significantly different proteins were analyzed by the KEGG pathway leading to the identification of 18 differential proteins related to carbohydrate metabolism. These proteins were mainly involved in 11 pathways (Table 4). A gene ontology analysis elucidated the molecular functions of 15 differential proteins related to carbohydrate metabolism (

Table 7. Differential proteins related to substance metabolism in the cotton stem bark induced by chilling stress and grafting.

Protein ID	Name	Molecular Function	Chilling Treat				Grafting
			S1/S1′	S2/S2′	S3/S3′	S4/S4′	S3/S1	S3′/S1′
Carbohydrate metabolism
ABW21688	Enolase	magnesium ion binding, phosphopyruvate hydratase activity	1.47	1.30	1.16	1.13	1.16	0.91
BAD08850	Glyceraldehyde-3-phosphate dehydrogenase	NAD or NADH binding, glyceraldehyde-3-phosphate dehydrogenase activity	1.33	1.47	1.43	1.33	0.79	0.85
ABK25148	Putative uncharacterized protein	pyruvate dehydrogenase (acetyl-transferring) activity	0.74	0.89	0.62	1.15	1.21	1.01
BAA02729	Fructose-bisphosphate aldolase	fructose-bisphosphate aldolase activity	2.52	1.57	1.22	2.54	1.52	0.73
XP_002285356	Malate dehydrogenase	L-malate dehydrogenase activity; binding	0.34	2.61	1.53	5.95	0.72	3.24
XP_002263337	Aconitate hydratase 1	4 iron, 4 sulfur cluster binding; aconitate hydratase activity	1.28	1.14	0.76	1.40	0.97	0.58
ABK94952	Putative uncharacterized protein	ribokinase activity	1.47	1.48	1.74	1.36	0.98	1.16
CAB87149	Transaldolase-like protein	transaldolase activity	NK	1.10	NK	1.11	NK	NK
XP_002271514	Chromosome chr1 scaffold_166	transaldolase activity	1.63	1.06	0.34	1.22	2.95	0.61
XP_001759864	Predicted protein	nucleotidyltransferase activity	1.97	1.68	1.75	1.37	1.14	1.01
XP_002526594	UTP-glucose-1-phosphate uridylyltransferase	UTP: glucose-1-phosphate uridylyltransferase activity	0.00	0.00	0.00	0.00	0.51	0.00
CAO21229	Chromosome chr10 scaffold_297	lactoylglutathione lyase activity; metal ion binding	1.15	1.37	0.95	0.88	1.05	0.87
Q9SID0	Probable fructokinase-1	ATP binding; fructokinase activity; ribokinase activity	1.03	1.19	1.66	0.90	0.79	1.27
O04886	Pectinesterase 1	aspartyl esterase activity; enzyme inhibitor activity; pectinesterase activity	0.86	1.29	1.01	2.00	0.97	1.14
ABE26954	Ribulose bisphosphate carboxylase large chain	monooxygenase activity; ribulose-bisphosphate carboxylase activity	1.43	NK	NK	NK	1.01	NK
Amino acid metabolism
CAB81805	Nucleoid DNA-binding-like protein	DNA binding; aspartic-type endopeptidase activity	0.71	1.02	0.76	1.56	0.51	0.55
ABV27483	Fasciclin-like arabinogalactan protein 12	cell adhesion	1.36	0.91	1.00	0.97	1.14	0.84
Q42699	Homocysteine methyltransferase	homocysteine S-methyltransferase activity; zinc ion binding	1.65	1.54	1.57	1.04	1.22	1.16
BAF28208	Adenosyl-homocysteinase	Hydrolase	1.18	1.12	1.07	1.11	1.21	1.09
ABO41849	Alcohol dehydrogenase A	oxidoreductase activity; zinc ion binding	0.77	1.07	1.27	0.75	0.89	1.47
Lipid metabolism
AAT01325	Os05g0518300 protein	hydrolase activity, acting on ester bonds	0.68	0.68	1.04	1.62	1.01	1.54
ACG70839	Phospholipase D alpha	calcium ion binding; phospholipase D activity	0.80	1.07	0.93	1.48	0.82	0.95
Nucleotide metabolism
Q96559	Nucleoside diphosphate kinase	ATP binding; magnesium ion binding; nucleoside diphosphate kinase activity	2.47	1.53	1.23	1.56	1.88	0.94

Note: S1: Scion seedling under chilling stress; S2: Rootstock seedling under chilling stress; S3: Scion of grafted plant under chilling stress; S4: Rootstock of grafted plant under chilling stress; S1′: Scion seedling; S2′: Rootstock seedling; S3′: Scion of grafted plant; S4′: Rootstock of grafted plant. NK means not known.

) and the regulatory pathways of these enzymes are shown in Figure 4.

According to the quantitative results (Table 7), under chilling stress, five differential proteins, namely ABW21688 (EC = 4.2.1.11), BAD08850 (EC = 1.2.1.12), BAA02729 (EC = 4.1.2.13), ABK94952, and XP_001759864, were consistently upregulated in stem bark of all the tested cotton plants, i.e., S1, S2, S3 and S4. All of these were listed among the 10 choke points for enzymes based on the metabolic pathways of Gossypium. Under chilling stress, expression of proteins such as ABK25148, XP_002285356, XP_002263337, XP_002271514, CAO21229, Q9SID0, and O04886 were up- or down-regulated; XP_002526594 was completely deficient; changes in CAB87149, XP_001767949, NP_187685, and ABE26954 were unclear in the tested cotton tissues.

Chilling stress caused maximum upregulation of the enzyme BAA02729 (EC = 4.1.2.13, fructose-bisphosphate aldolase, FBP) in the cotton stem bark tissues (across all the tested plants). The FBP contents in the stressed stem bark of S1, S2, S3 and S4 were 2.52-, 1.57-, 1.22-, and 1.54-fold higher than that in their respective non-stressed plants, respectively. This enzyme is also among the top 10 choke points for enzymes based on the metabolic pathways of Gossypium, and it requires further exploration. FBP (EC = 4.1.2.13) is involved in most metabolic reactions in carbohydrate metabolism and energy metabolism.

Under chilling stress, the second most consistently up-regulated protein in the cotton stem bark across all the studied plants was BAD08850 (EC = 1.2.1.12, glyceraldehyde-3-phosphate dehydrogenase, GAPDH), which ranked third out of 10 choke points for enzymes. Under chilling stress, the third most consistently up-regulated protein in the cotton stem bark of all the tested plants was ABW21688 (EC = 4.2.1.11, enolase, phosphopyruvate hydratase), which ranked 9th out of 10 choke points for enzymes. The protein XP_002526594 (EC = 2.7.7.9, UDP-glucose pyrophosphorylase, UDPGP) was completely deficient under chilling stress and was only found in the S1′, a control. In addition, under chilling stress, two significantly up-regulated differential proteins in the cotton stem bark across all the studied plants were XP_002285356 and O04886. Among these, the prominent expression of XP_002285356 (EC = 1.1.1.37, malate dehydrogenase, MDH) in S2, S3 and S4 was 2.61-, 1.53-, and 5.95-fold higher than that in their respective control plants, respectively.

Nine other differential proteins, namely ABW21688, ABK25148, BAA02729, XP_002285356, ABK94952, XP_002271514, XP_001759864, Q9SID0, and O04886 in cotton stem bark were also up-regulated in response to chilling stress. Among these, the differential proteins, XP_002285356, ABK94952, Q9SID0, and O04886 were up-regulated by more than 10% in the cotton stem bark from S3′ vs. S1′. The protein XP_002285356 (malate dehydrogenase) showed maximum up-regulation from S3′ was 3.24-fold higher than that in the cotton stem bark from S1′.

3.4.2. Differential Proteins Related to Energy Metabolism

There were 16 significantly differential proteins related to energy metabolism. These proteins were allied in at least four pathways, including oxidative phosphorylation, photosynthesis, carbon fixation in photosynthetic organisms, and methane metabolism, and 11 of them are involved in photosynthesis (Table 4). Three of the enzymes (BAA02729; XP_002285356; ABE26954) were involved in the carbon fixation in photosynthetic organisms (Table 7). Furthermore, the molecular functions of the remaining four enzymes are unclear.

According to the quantitative results (

Table 8. Differential proteins related to energy metabolism in the cotton stem bark induced by chilling stress and grafting.

Protein ID	Name	Molecular Function	Chilling Treat				Grafting
			S1/S1’	S2/S2’	S3/S3’	S4/S4’	S3/S1	S3’/S1’
AAS59949	Photosystem II CP47 chlorophyll apoprotein	chlorophyll binding	1.00	0.52	0.71	0.57	1.75	1.24
ABG74856	Cytochrome b559 subunit alpha	heme binding	NK	NK	2.62	8.76	4.82	NK
Q6ENJ7	Photosystem Q(B) protein; 32 kDa thylakoid membrane protein; hotosystem II protein D1	electron transporter, transferring electrons within the cyclic electron transport pathway of photosynthesis activity; iron ion binding	0.56	0.47	0.80	0.89	0.60	0.86
ABK95197	Light-harvesting complex II protein Lhcb6	chlorophyll binding; metal ion binding	1.00	0.80	1.00	0.49	0.99	0.99
ACG24344	Photosystem I reaction center subunit XI	NK	0.66	0.54	0.71	0.72	0.87	0.94
NP_683825	Photosystem II D2 protein (Photosystem Q(A) protein)	electron transporter, transferring electrons within the cyclic electron transport pathway of photosynthesis activity; iron ion binding	NK	NK	0.59	0.90	1.05	NK
CAN64486	CAAD domain-containing protein	NK	0.89	0.88	0.13	0.62	2.75	0.39
Q01859	ATP synthase subunit beta, mitochondrial	ATP binding Ref.1; ydrogen ion transporting ATP synthase activity, rotational mechanism; ydrogen-exporting ATPase activity, phosphorylative mechanism; roton-transporting ATPase activity, rotational mechanism	0.75	0.73	0.94	0.75	0.81	1.01
P85433	Peroxidase 7	calcium ion binding; heme binding; peroxidase activity	0.67	0.97	0.81	1.14	0.86	1.04

NK—not known.

), under chilling stress, only one differential protein ABG74856 was mostly consistently up-regulated in the cotton stem bark of all the tested plants. This protein is a subunit of cytochrome b559 and it located in the photosynthetic electron transport chain. Function of proteins Q8MAV7, NP_683825, ACJ85810, and CAO47122 was unclear. Three proteins, namely AAS59949, ABG74856, and CAN64486, were up-regulated by more than 10% in the cotton stem bark of the scion in response to grafting in the stem bark of scion both under stressed and non-stressed conditions. Out of these, ABG74856 was the most responsive protein, showing 4.82-fold higher expression in the scion of then grafted than the non-grafted plants.

3.4.3. Differential Proteins Related to Lipid Metabolism

There were two significantly differential proteins related to lipid metabolism in the studied cotton tissues. These are involved in at least two pathways, including glycerolipid metabolism, and glycerophospholipids (Table 4).

Under a non-stressful environment, grafting induced 1.54-fold higher expression of AAT01325 in the stem bark of scion. Similarly, there was a significant increase in the expression of AAT01325 in stem bark of both scion and rootstock in response to grafting.

3.4.4. Differential Proteins Related to Nucleotide Metabolism

In this study, Q96559 is the only differential protein associated with nucleotide metabolism, including metabolism of cytidine triphosphate (CTP), guanosine triphosphate (GTP), and glucose pyrophosphorylase (UDP) significantly responded to chilling stress in the stem bark of cotton (Table 7). It is a nucleoside diphosphate kinase (EC = 2.7.4.6) located in the second position of the 10 choke points for enzymes. Chilling stress increases the expression of these enzymes by 2.47-, 1.53-, 1.23-, and 1.56-fold higher in the stem bark of S1′, S2′, S3′, and S4′, respectively.

3.4.5. Differential Proteins Related to Amino Acid Metabolism

There were six significantly differential proteins related to amino acid metabolism in the studied cotton tissues. These were involved in at least six pathways (Table 4).

Under chilling stress, two differential proteins were consistently up-regulated in the cotton stem bark namely Q42699 (EC = 2.1.1.14) and BAF28208 (EC = 3.3.1.1, S-adenosyl-L-homocysteine hydrolase, SAHH) (Table 7). Out of these, BAF28208 (EC = 3.3.1.1, S-adenosyl-L-homocysteine hydrolase, SAHH) was among the 10 choke points for enzymes (Table 6). Under chilling stress, CAB81805, ABV27483, and ABO41849 were up- or down-regulated (Table 7) and P85433 was down-regulated in the cotton stem bark of cotton (Table 8).

3.4.6. Differential Proteins Related to Protein Folding and Degradation

There were 11 significantly differential proteins related to protein folding and degradation in the studied cotton tissues. These proteins were involved in two pathways, including chaperonin and isomerase, and ubiquitin-mediated protein hydrolysis.

Chilling stress consistently upregulated four differential proteins, namely ACG33102 (14-3-3-like protein), ACB70177 (70 kDa heat shock protein, HSP70) (

Table 9. Family of molecular chaperones.

Family	Function
HSP10	HSP60 of auxiliary chaperonin, Helping the substrate of HSP60 to fold to bind to HSP60
sHSP	Including a variety of proteins, depend on ATP for their function binding to unnatural state proteins
HSP40	Auxiliary chaperonin, regulating the activity of HSP70, Some of them can bind to non-natural state proteins
HSP60	Helping 15–30% of cellular proteins to fold through ATP
HSP70	Preventing the adhesion and aggregation of unfolded polypeptide chains, Depolymerizes multifolded proteins, participating in protein transport, and regulating heat shock response
HSP90	Acting together with some kinases and steroid receptors in signal transduction pathways, they may also play the role of some “typical” molecular chaperones
HSP100	Depolymerization of protein polyplexes and aggregates
HSP110	Highly homologous to HSP70, function unknown

), ABR16377 (Ubiquitin), and NP_568552 (Polyubiquitin)in the cotton stem bark (

Table 10. Differential proteins related to folding and degradation, transport, signal transduction, antioxidation, and other processes in the cotton stem bark induced by chilling stress and grafting.

Protein ID	Name	Molecular Function	Chilling Treat				Grafting
			S1/S1′	S2/S2′	S3/S3′	S4/S4′	S3/S1	S3′/S1′
Protein folding and degradation
CBC02991	Unnamed protein product	ATP binding	NK	NK	NK	NK	NK	NK
ACG33102	14-3-3-like protein	protein domain specific binding	1.99	1.22	1.21	1.16	1.57	0.95
ACB70177	70 kDa heat shock protein	ATP binding	1.19	1.51	1.52	1.37	0.75	0.96
ABY65001	14-3-3b protein	protein domain specific binding	1.45	1.16	1.12	0.91	0.92	0.71
ABX76300	Heat shock protein 60	ATP binding;protein binding	0.37	0.55	0.86	1.00	0.90	2.08
EAZ30819	Putative uncharacterized protein	isomerase activity	1.79	0.84	0.80	1.12	2.46	1.11
O49886	Peptidyl-prolyl cis-trans isomerase	peptidyl-prolyl cis-trans isomerase activity	0.92	1.02	0.93	0.86	0.90	0.91
CAZ91440	Calreticulin	unfolded protein binding	0.97	1.00	0.82	1.19	1.42	1.21
ABR16377	Uncharacterized protein (Ubiquitin)	mRNA binding	1.34	1.19	1.06	1.00	1.09	0.86
NP_568552	Polyubiquitin	mRNA binding; ubiquitin protein ligase binding	1.68	1.32	1.39	1.28	1.33	1.10
Substance transport
ABK92828	Predicted protein	GTP binding	0.70	0.66	0.72	0.87	0.88	0.90
1N00_A	Annexin	calcium ion binding; calcium-dependent phospholipid binding	0.70	1.84	3.44	1.36	0.27	1.33
CBC30882	Anx1	calcium ion binding; calcium-dependent phospholipid binding	0.53	0.96	0.95	1.40	0.65	1.17
ABK92937	Predicted protein	adenosine kinase activity	0.03	0.00	0.00	0.00	1.11	0.00
ACI26701	Non-specific lipid-transfer protein	lipid binding	0.65	0.86	1.09	0.80	0.70	1.17
Signal transduction
ABW34390	Glutathione S-transferase 5	transferase activity	NK	NK	NK	0.98	NK	NK
Q7Y052	Calmodulin (CaM)	calcium ion binding	1.09	0.96	0.82	0.96	1.17	0.88
Q9ZSW9	tumor protein homolog (TCTP)	calcium ion binding	1.55	1.22	0.93	1.18	1.41	0.85
Antioxidation
ACC93637	Superoxide dismutase	copper ion binding; zinc ion binding	1.20	1.28	0.88	1.79	1.32	0.97
CBC66557	Class III peroxidase	calcium ion binding; electron carrier activity; heme binding	1.10	1.98	2.55	2.29	0.56	1.30
XP_002303520	Predicted protein	copper ion binding; oxidoreductase activity	0.91	NK	NK	NK	NK	NK
CAO47501	Chromosome chr3 scaffold_8	copper ion binding; oxidoreductase activity	0.77	0.98	0.92	0.87	0.87	1.04
Other processes
XP_002446509	ultraviolet-B receptor UVR8	ultraviolet-B receptor	0.52	0.45	0.27	0.58	1.70	0.88
XP_002461966	28 kDa ribonucleo protein	nucleic acid binding; nucleotide binding	0.84	0.80	0.80	0.77	1.15	1.10
XP_001690796	Predicted protein	snRNA processing	0.23	1.35	0.43	3.52	0.46	0.87

Note: S1: scion seedling under chilling stress; S2: rootstock seedling under chilling stress; S3: scion of grafted plant under chilling stress; S4: rootstock of grafted plant under chilling stress; S1′: scion seedling; S2′: rootstock seedling; S3′: scion of grafted plant; S4′: rootstock of grafted plant. NK—known.

). Further, chilling stress up- or down-regulated ABY65001, EAZ30819, O49886, and CAZ91440, and consistently down-regulated ABX76300 (heat shock protein 60, HSP60) but the regulation of ABK95670 in the cotton stem bark was unclear (Table 7).

Grafting also up-regulated six differential proteins in the stem bark of scion, namely ACG33102, ABX76300, EAZ30819, CAZ91440, ABR16377, and NP_568552 under a non-stressful environment. Out of these, ABX76300 (HSP60) was the most responsive protein, showing 2.08-fold higher expression in the stem bark of grafted than non-grafted scion.

3.4.7. Differential Proteins Related to Cellular Substance Transport

In this study, eight differential proteins related to cellular material transport in the studied cotton tissues. They were involved in four pathways, including membrane vesicle transport, annexin, phosphotransferases, and transporter proteins (Table 4).

Chilling stress did not up-regulate any differential proteins in the cotton stem bark in the studied samples (Table 10). However, chilling up- or down-regulated 1N00_A, CBC30882, and ACI26701 u; ABK92828 and consistently down-regulated ABK92937. Although regulation of XP_001780193, Q01111, and BAB90396 under chilling stress was unclear. Among these, ABK92937, a protein with adenosine kinase activity, was deficient under chilling stress and only detected in S1′ tissues.

Grafting also up-regulated four differential proteins, namely 1N00_A, CBC30882, ABK92937, and ACI26701in the cotton stem bark. Among these, three differential proteins were up-regulated by more than 10%, namely 1N00_A, CBC30882, and ACI26701.

3.4.8. Differential Proteins Related to Signal Transduction

There were six significantly differential proteins related to signal transduction in the studied cotton tissues. They are involved in two pathways, i.e., the MAPK signaling pathway and the calcium signaling pathway (Table 4).

Under chilling stress, three differential proteins were consistently up-regulated in the cotton stem bark, namely 1N00_A, ACG33102, ABY65001 (Table 10). Further, chilling stress up- or down-regulated Q7Y052 and Q9ZSW9 but impact of chilling on ABW34390 was unclear.

Four differential proteins were up-regulated in the cotton stem bark of scion, namely ACG33102, 1N00_A, Q7Y052, and Q9ZSW9, in response to grafting under non-stressed conditions, with 1N00_A showing more than 10% up-regulation.

3.4.9. Differential Proteins Related to Antioxidation

There were six significantly differential proteins related to antioxidation in the studied cotton tissues. They are involved in three pathways, including superoxide dismutase, peroxidase, and oxidoreductase (Table 4).

Chilling stress consistently up-regulated CBC66557 in the cotton stem bark. Proteins such as ABK25148, P85433, and ACC93637 were up- or down-regulated in response to chilling and the regulation of XP_002303520 was unclear (Table 10). Further, the stressed tissues, e.g., S1, S2, S3, and S4 had 1.10-, 1.98-, 2.55-, and 2.29-fold higher CBC66557 (peroxidase III) compared with their respective non-stressed controls.

Grafting treatment also up-regulated five differential proteins (ABK25148, P85433, ACC93637, CBC66557, and CAO47501) in the scion of stem bark.

3.4.10. Differential Proteins Related to Other Processes

Our results showed were 12 significantly differential proteins related to other processes in the studied cotton tissues. They are involved in nine pathways (Table 4).

Quantitative analysis showed that chilling stress consistently up-regulated the differential proteins in the cotton stem bark. Further, chilling stress up- or down-regulated XP_001690796, consistently down-regulated XP_002446509 and XP_002461966 and its impact on NP_171847 and P41099 was unclear. XP_002446509 and XP_ 002461966 are speculative proteins; their authenticity has yet to be verified, so they are not analyzed here.

Two differential proteins, namely XP_002446509 and XP_002461966, were up-regulated in the scion of cotton stem bark in response to grafting. Among these, XP_002461966, a chromosomal scaffolding protein, was up-regulated by 1.10 fold higher in non-stressed scion.

4. Discussion

4.1. Induced Changes in Cotton Proteins and Biology Pathway by Chilling Stress and Grafting

In this study, low-temperature stress, increased carbohydrate metabolism, nucleotide metabolism, lipid metabolism, and amino acid metabolism in cotton seedlings showed more up-regulated enzymes. Particularly, enzymes associated with carbohydrate metabolism e.g., BAA02729 (FBP), BAD08850 (GAPDH), and ABW21688 (enolase) were significantly up-regulated (Table 7). These enzymes had a significant impact on glycometabolism and the corresponding energy metabolism. One example is enolase s, one of the glycolytic enzymes, also known as phosphopyruvate hydratase, which cleaves 2-phosphoglycerate to produce phosphoenolpyruvic acid, and thus regulates carbohydrate and energy metabolisms in living organisms [33]. Similarly, GAPDH is a key enzyme in all prokaryotic and eukaryotic processes of glycolysis and gluconeogenesis, and it is necessary to catalyze oxidation and phosphorylation of D-glyceraldehyde-3-phosphate [34]. This suggests that upregulation of GAPDH under low-temperature stress increases ATP production and thus supports plant growth. Consistent with the high concentrations of glucose, fructose, and sucrose in Arabidopsis thaliana induced cold stress tolerance after cold acclimation [35].

Grafting also increased the relative amounts of some proteins in the scion, especially the key enzymes such as MDH and SAHH. MDH is commonly found in various organisms, including animals, plants, and bacteria, and is a key enzyme in carbohydrate metabolism, catalyzing the reversible conversion between malate and oxaloacetate. MDHs regulate various physiological activities in cells, including energy and reactive oxygen species metabolisms in plants [36]. Similarly, SAHH is the only enzyme known that catalyzes the reversible hydrolysis of S-adenosyl-L-homocysteine (SAH) to homocysteine and adenosine, and SAH is the product of all S-adenosyl-L-methionine (SAM)-dependent biological methyl transfer reactions [37]. SAH and recombinant adenosylhomocysteinase (EC 3.3.1.1) are associated with the Clock-Bmal1 cycle at chromatin loci, and they promote circadian transcriptional activity [38]. In general, proteins with increased expression at low temperatures, such as UDPGP and MDH, were associated with the defense and improvement of abiotic stresses or energy metabolism [39]. In this study, XP_002526594 (UDPGP) was completely deficient under chilling stress (Table 7). Low-temperature stress affected the carbohydrate metabolism of cotton. The deficiency of UDPGP may be related to the fact that some genes were not expressed due to chilling stress. According to Mohapatrs et al. [40], chilling stress induces expression of cold-regulated genes in plants, producing new mRNAs and cold-regulated proteins. This stress also down-regulates some key genes, resulting in a disappearance of the corresponding proteins. However, previous reports showed that exposure of plants to low-temperature stress led to an increase in the relative abundance of UDPGP and sucrose synthase 1 involved in the tricarboxylic acid cycle [16].

4.2. Photosynthesis-Related Proteins by Chilling Stress and Grafting

In this study, low-temperature stress attenuated oxidative phosphorylation, photosynthesis, and carbon fixation in cotton using light for energy synthesis. The effect of low-temperature treatment on saline plants was also mainly focused on photosynthesis, energy, and primary metabolism-related proteins [41]. Earlier studies indicated that chilling can inhibit FBP, an enzyme which regulates carbon assimilation and thus limits plant growth, particularly in chilling-sensitive plants [42]. In this study, 11 significantly differential proteins were involved in photosynthesis (Table 4). Among those, four significantly differential proteins were significantly down-regulated, namely Q6ENJ7, ACG24344, CAN64486 and Q01859 (Table 8). The molecular functions of ACG24344 and CAN64486 are unknown. Expression of several photosynthesis-related proteins, NADH dehydrogenase, exo-oxygenation-enhancing protein, and dehydroascorbate reductase, was down-regulated at the onset of cold acclimation [43]. However, some interspecific differences were found in the gene products of interest. In this study, proteomic analysis in the tested cotton plants suggests that the proteomic low-temperature stress response pathways are similar in other plants. Under low-temperature stress, CBF transcription factors closely interact with the expression of photosynthesis-related genes [16]. Overexpression of ectopic AtCBF1 in eggplant improves its cold tolerance [44]. Low temperature had a negative impact on photosynthesis; for example, it affects the protein involved in the electron transport chain of photosynthesis [16,45]. Photorespiration helped protect the photosynthetic process and played a beneficial role in stress regulation. Therefore, inhibition of photorespiration may be one of the main reasons for the sensitivity of plants to low-temperature stress. In this study, the photosynthesis-related proteins exhibited by grafted seedling were mostly down-regulated. However, the down-regulation degree of grafted plant protein was less over non-grafted plant protein in energy metabolism (Table 8). Grafting increased the relative amounts of some photosynthesis-related proteins in the scion, which may contribute to photosynthesis in scion and thus improve cold tolerance.

4.3. Differential Proteins Related to Protein Folding and Degradation by Chilling Stress and Grafting

Heat shock protein, 14-3-3 proteins, and ubiquitin, which were related to protein folding and degradation, were more active in cotton under low-temperature stress (circadian temperature), implying that these proteins are closely related to stress response. Overexpression of heat shock proteins has been associated with abiotic stress tolerance in plants [46]. Heat-stimulated proteins can be classified into several families, such as HSP110, HSP90, HSP70, HSP60, sHSP (15–30 kD), and ubiquitin, according to the apparent molecular weight size of SDS electrophoresis (Table 9) [47,48]. Although the main function of HSP was initially thought to be to improve heat tolerance in plants, a significant correlation between HSP and cold tolerance in plants has also been reported [49]. For example, long-term cold tolerance by tomato fruits of heat stress treatment was associated with the persistent maintenance of HSP in vivo [49]. Heat kinins can regulate cytoskeletal dynamics [47,50] and identified soybean cold response-related proteins, one of which is the heat stress protein HSP70. Inhibition of HSP70 and HSP90 accumulates oxidative stress, which could lead to autophagy in cotton ovules.

The 14-3-3 protein is a “bridge” for protein–protein interactions, interacting with more than 100 proteins and participating in almost all physiological reactions of living organisms, and therefore has received widespread attention [51]. Currently, research on 14-3-3 proteins in plants has made a great progress in seed germination, basic carbon and nitrogen metabolism, response to plant defense and stress, and control of proton pump and ion channel activity [51], which is a frontier in the field of signaling pathway research [52]. Ubiquitination ought to be a unique and widely conserved molecular mechanism for the degradation of proteins. Its main function is to label proteins that need to be broken down and removed to be hydrolyzed. Ubiquitin can also label transmembrane proteins, such as receptors. Covalent attachment of ubiquitin to target substrates involved the sequential action of three enzymes: ubiquitin-activating enzyme (E1), ubiquitin-binding enzyme (E2), and ubiquitin-ligase (E3) [21,53]. Ubiquitin can remove them from the cell membrane after covalent attachment. E3-mediated protein degradation regulated almost all cellular processes, including abiotic stress responses [21].

4.4. Differential Proteins Related to Cellular Material Transport by Chilling Stress and Grafting

In this study, low-temperature stress down-regulated the substance transport in cotton tissues, as shown by the absence of consistently up-regulated proteins in the four cotton materials and even the deficiency of individual proteins. In contrast, grafting increased the relative amounts of some proteins associated with substance transport, especially the relative amounts of annexin, implying that grafting has a potential to improve substance transport in the scion tissues. The annexin family is a superfamily of calcium-dependent phospholipid-binding proteins that are found in most eukaryotic cells and have a wide range of important physiological functions, such as cytoskeletal movement, cell growth, membrane ion channels, and cellular signal transduction [54,55]. The cell membrane is a critical site of damage in chilling-induced plants [56]. Proteomic analysis of chilling-induced rice root plasma membrane showed that most of the cold-responsive proteins are related to energy production, signal transduction, and protein synthesis [16]. In addition, altered cell membrane permeability, annexin can lead to an increase in intracellular cations under low-temperature stress. This further resulted in Ca²⁺-dependent annexin that has an important effect on function in regulating low temperature stress tolerance [57]. Annexin played an important role in the Golgi-mediated secretion of plant cell plasma membrane and wall material [58]. In this study, 1N00_A (annexin) was up-regulated in the scion of grafted plants under chilling stress (Table 10). In cotton [59], annexin was highly expressed in secretory cells. Fibrillar annexin could bind to membrane callus synthase and regulate its activity. Up-regulation of annexin may be a response to chilling-induced osmotic stress [58].

4.5. Differential Proteins Related to Signal Transduction by Chilling Stress and Grafting

The proteins related to signal transduction were significantly up-regulated in cotton under low-temperature stress, suggesting an enhanced in vivo sensing of low-temperature stress in cotton. Grafting up-regulated signal transduction. Signal transduction proteins, stress defense proteins, and reactive oxygen scavenger enzymes are also reported to increase under abiotic stresses in plant tissues. Further, low temperature can alter the expression of fructose diphosphate aldolase, rubisco, and ATPase intolerant and sensitive species [16,60]. Adaptation of metabolism and structure through signal transduction in tillering node cells of wheat to provide the energy required for plant development and resilience. ABW34390 (Glutathione S-transferase 5) was unclear (Table 4). Glutamate receptor (GLR) channels may be an important component of plant signal transduction pathways that affect the level of plant metabolism under adverse conditions [61].

MAPK is a category of serine/threonine kinases that consists of multiple isozymes [62]. Q7Y052 and Q9ZSW9 were involved in calcium regulation (Table 10). Elevated intracellular free calcium concentrations under low-temperature stress can trigger the MAPK signaling pathway, which in turn mediates the expression of different target genes through phosphorylation of transcriptional regulators [16,63]. Plant hypothermic signaling includes the calcium signaling pathway and other signaling pathways. In calcium signaling pathway, cytoplasmic calcium ions increased by chilling stress can be recognized and transduced by CDPK, phosphatase, and MAPK. Other signaling pathways are mainly associated with abscisic acid (ABA), and hypothermic signaling will eventually initiate CBF and non-CBF-mediated transcriptional regulation to improve chilling resistance in plants [64]. Variation in GhSAD1, a gene related to the synthesis of the phytohormone ABA, is the main factor causing differences in low temperature resistance among cotton varieties [65].

4.6. Differential Proteins Related to Antioxidation by Chilling Stress and Grafting

In this study, under low-temperature stress, the protein antioxidant enzymes content was regulated, while the effect of grafting on antioxidant enzymes was relatively weak, except peroxidase, which was induced to some extent both in response to chilling and grafting [66]. Peroxidases are mostly found in the peroxisomes of cells, with iron porphyrins as cofactors, and they can catalyze the oxidation of phenolic and amine compounds by hydrogen peroxide [67]. The balance of SOD, CAT, and APX activities is important for suppressing toxic ROS levels in cells and the importance of cooperative activity in reactive oxygen species scavenging systems in response to chilling stress [68,69]. XP_002303520 was predicted protein (Table 10). LOX is involved in the biosynthesis of jasmonic acid (JA) and several secondary metabolites and antioxidants that play specific roles in the development and response to stress [70,71].

5. Conclusions

In this study, proteomic analysis shows the physiological and biochemical pathways of cotton plants responding to low-temperature stress and grafting. Chilling stressed carbohydrate, lipid, nucleotide, and amino acid metabolism, with relatively more active pathways related to protein folding and degradation and activation of antioxidant enzyme system. However, oxidative phosphorylation, photosynthesis, carbon fixation, and other energy synthesis using light were weakened and material transport was slowed down. This indicates that the physiological and biochemical changes in cotton under low-temperature stress are very complex, involving multiple factors. Grafted plants improved the ability of scions to resist this stress, but the effect was not obvious in rootstock. This experiment lays the foundation for further research on differential proteins related to chilling stress or grafting, and the discovery of new cold tolerance genes.