Characterization of the aquaporin gene family in Pinus massoniana and functional analysis of PmNIP2;1 involved in high-temperature and drought stress

Abstract

Background

Aquaporin (AQP) in plants can regulate the transport of water and various small solutes and play important roles in plant development and abiotic stress responses. The aquaporin gene family in Pinus massoniana has not yet been described.

Results

In this study, 30 PmAQP genes were identified in P. massoniana seedlings. Based on phylogenetic analysis, these genes were categorized into four groups: 12 plasma membrane intrinsic proteins (PIPs), 11 tonoplast intrinsic proteins (TIPs), 5 NOD26-like intrinsic proteins (NIPs), and 2 small basic intrinsic proteins (SIPs). Within each group, the conserved motifs, tertiary structures, and aromatic/arginine selectivity filter (ar/R) presented high similarity. Subcellular localization analysis in Nicotiana benthamiana epidermal cells revealed a diverse and broad array of PmPIPs with various subcellular locations. Expression profiling revealed that 14 PmAQPs in leaves presented specific expression patterns under high-temperature stress, drought stress and combined high-temperature and drought stress. These findings suggest that PmAQP family genes may be involved in the response to high-temperature, drought and the combined stress of high temperature and drought. The overexpression of PmNIP2;1 promoted the root development of transgenic tobacco and improved the tolerance of transgenic tobacco to high-temperature, drought and the combination of high-temperature and drought by regulating the expression of stress resistance genes and changes in physiological and biochemical substances in transgenic tobacco.

Conclusions

This study provides a comprehensive transcriptome-wide analysis of the aquaporin gene family, shedding light on its potential roles in plant growth and stress responses.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-026-08262-x.

Introduction

High temperature and drought are pervasive abiotic stresses in nature, which severely constrain plant growth and development [1, 2]. Under the influence of climate change, extreme heat increases the risk of drought-induced mortality in plants and may trigger widespread forest die-off [3, 4]. Under combined high-temperature and drought stress, plants attempt to cool their leaves by increasing transpiration, yet drought conditions trigger stomatal closure to minimize water loss. Consequently, plants must strike a balance between preventing excessive water loss and avoiding lethal overheating [5]. Plants possess a highly sophisticated system for regulating water transport and maintaining internal homeostasis, with aquaporins (AQPs), as integral membrane proteins, playing an essential role in this process [6].

As members of the major intrinsic protein (MIP) superfamily, AQPs function as selective channels that facilitate the permeation of water and small, neutral solutes (e.g., urea, ammonia, glycerol, silicic acid, boric acid, CO_2, H₂O₂) across biological membranes [7–10]. Since their first discovery in the 1990s, AQPs have been found in almost all living organisms [11]. On the basis of sequence similarity and subcellular localization, AQPs are divided into five distinct subfamilies in higher plants: plasma membrane intrinsic proteins (PIPs), tonoplast intrinsic proteins (TIPs), NOD26-like intrinsic proteins (NIPs), small basic intrinsic proteins (SIPs), and uncategorized X intrinsic proteins (XIPs).

Aquaporins exist in the membrane as homo- or hetero-tetramers, which are assembled via hydrogen bonding interactions involving the monomer loops. Each monomer within the tetramer functions as an independent water channel [12]. Each AQP protein monomer contains six α-helical transmembrane domain (TMD1-TMD6), five loops (LA-LE), and two additional half-helices (HB, HE). HB and HE contain highly conserved Asn‒Pro‒Ala (NPA) motifs that form pores that allow single-file passage of water molecules [13]. Their stereo configuration and distance from each other determine the size-selective barrier of the various permeabilities. The aromatic arginine (ar/R) region formed by each residue from TMD2 and TMD5 and two residues from LE (LE1 and LE2) represents the other constriction, defining the substrate specificity and permeability [14].

Plants harbor highly abundant and divergent AQPs. In the last 5 years, the identification and expression analysis of AQP gene families in dozens of plant species have been reported, with 27 in Passiflora edulis [15], 34 in Kandelia obovata [16], 58 in Ligustrum vicaryi [17], 33 in Prunus armeniaca [18], and 33 in Betula pendula [19], 39 in Cucumis sativus [20]. Moreover, AQPs play versatile physiological roles in combatting high-temperature and drought stresses, which has been supported by analyses of transgenic plants overexpressing various aquaporins or by analyses of aquaporin transcript abundance. For example, AQP genes are differentially expressed in the leaves of Setaria italica under high-temperature and drought stress [21]. Similarly, the relative expression of AQPs in the desert Artemisia sieberi peaked under compound stress [22]. It has been demonstrated that the overexpression of AQP family genes improves plant drought tolerance by increasing the osmoregulatory material content and antioxidant enzyme activity and reducing ion leakage and lipid peroxidation [23–27].

Pinus massoniana is a major coniferous tree widely distributed in southern China [28]. Drought and high temperatures severely affect the normal growth and development of P. massoniana in the main production areas. Therefore, mining the stress resistance-related functional genes of P. massoniana, identifying their functions, and analysing their mechanisms of action are very important for improving the varieties of P. massoniana. To date, only one AQP gene, PmPIP1, from P. massoniana has been cloned, and its expression is induced by drought stress [29]. The aquaporin gene family in P. massoniana has not yet been described. Given the potential value of aquaporins in improving stress tolerance, identifying and structurally characterizing aquaporin genes and mining the key AQP genes that control high-temperature tolerance and drought tolerance in P. massoniana are necessary.

In this study, members of the AQP gene family of P. massoniana were screened and identified on the basis of data from the P. massoniana seedling transcriptome. Through the combination of high-temperature and drought stress, the PmAQP genes related to high-temperature and drought stress were further screened. The physicochemical properties and expression localization of AQP gene family members in P. massoniana were studied via bioinformatics and subcellular localization. To further explore the mechanism by which the P. massoniana aquaporin gene PmNIP2;1 responds to high-temperature and drought stress, tobacco plants overexpressing PmNIP2;1 were constructed and subjected to high-temperature and drought stress. Transcriptome-wide identification and primary functional characterization of P. massoniana aquaporins provide insight into the complexity of the AQP gene family and its biological functions in P. massoniana.

Materials and methods

Identification of PmAQP family members and subfamily classification

PmAQP candidates were initially screened from previous transcriptome data under drought stress [30]. Thirty-five AQP amino acid sequences of Arabidopsis thaliana were downloaded from the Phytozome database. The protein sequences of the AQPs in the model plant A.thaliana were used as queries to carry out local BLASTp searches against the P. massoniana transcriptome database, with a cut-off E-value cutoff of 1e^{− 10}. Furthermore, the resulting candidate sequences were checked for the presence of the MIP domain via the SMART and NCBI-CDD web servers. Sequences encoding the complete MIP domain and two NPA motifs were considered putative AQP genes. The aquaporin amino acid sequences from P. massoniana, B. pendula, and (A) thaliana were aligned using the ClustalW program.The sequences for (B) pendula were obtained from Supplementary Figure S1 of Venisse et al. [19].

A neighbor joining (NJ) algorithm phylogenetic tree was constructed via MEGAX software, with 1000 bootstrap tests. The online program EvolView was used to optimize the evolutionary tree. PmAQPs were named according to their percentage of identity to the 35 AQP proteins from A. thaliana [31].

Bioinformatics analysis of conserved motifs and characteristics of AQPs

Various physicochemical properties of the PmAQP proteins were predicted using the ProtParam tool. The variables included the length of the amino acid sequence, molecular weight, isoelectric point, aliphatic index, hydrophobicity, and instability coefficient. The number of transmembrane domain was detected via DeepTMHMM, TOPCONS, and CCTOP. The subcellular location was predicted with two different prediction software programs: ProtComp 9.0 and WoLF PSORT. PmAQP protein sequences were submitted to the protein modelling server Phyre2 for the construction of 3D models of proteins. Unless specified otherwise, the tools used in this study were performed with default parameters. Conserved motifs in the AQP protein sequences were identified via the MEME program, with default settings, and the motif number was set at a value of 10. The prediction and visualization of conserved motifs were performed through the Visualize Gene Structure program via TBtools software. Conserved functional region analysis of NPA, such as ar/R filters (H2, H5, LE1 and LE2) and Froger’s positions, was performed via alignment with BpeAQPs.

Plant materials and treatments

One-year-old P. massoniana seedlings obtained from a state-owned forest farm in Dushan, Guizhou Province, China, were used in this study. Seedlings with uniform growth were subjected to different stress treatments, including (a) control (CK) (normal irrigation, 16/8 h light/dark photoperiod and 25 °C/25°C day/night temperature), (b) drought stress (D) (16/8 h light/dark photoperiod, 25 °C/25°C day/night temperature, with irrigation stopped after water was poured and flooded), (c) high-temperature stress (H) (normal irrigation,16/8 h light/dark photoperiod, 35 °C/25°C day/night temperature), (d) high-temperature and drought stress (HD) (16/8 h light/dark photoperiod, 35 °C/25°C day/night temperature, with irrigation stopped after water was poured and flooded) [32–34]. The relative water content (RWC) of the soil treated with a and c was always maintained at 70%-80%. The upper needles of P. massoniana were collected after continuous treatment for 9 days (L), 14 days (M) or 16 days (S). Similarly, the RWC of the soils treated with b and d was 55%-60%, 40%-45% and 30%-35%, respectively. The samples were then immediately frozen in liquid nitrogen and subsequently stored at − 80 °C for further use.

Expression analysis of PmAQPs

Total RNA was extracted from collected leaf samples using the RNAprep Pure Plant Plus Kit (Polysaccharides & PolyphenoIics-rich) (Tiangen, Beijing, China). cDNA was subsequently synthesized via FastKing gDNA Dispelling RT SuperMix (Tiangen, Beijing, China). Gene-specific primers were designed with Primer Premier 6 software (Supplementary Material 2.). The expression of PmAQPs was analyzed by real-time quantitative PCR (RT‒qPCR) using Talent qPCR PreMix (SYBR Green) (Tiangen) on a real-time PCR system. Ubiquitin-conjugating enzyme (UBC) was used as an internal control for RT‒qPCR. The thermal cycling conditions were as follows: initial denaturation at 95 °C for 3 min, followed by 45 cycles of denaturation at 95 °C for 5 s, annealing at 57 °C for 10 s, and extension at 72 °C for 15 s. Three technical replicates were maintained for each sample. The relative expression levels were calculated using the 2^−ΔΔCt method [35]. The resulting data were visualized as a heatmap using the HeatMap tool in TBtools software.

Subcellular localization analysis of the candidates

The PmPIP1;5, PmPIP2;1 and PmPIP2;2 genes were cloned and fused to green fluorescent protein (GFP) at the N-terminus via the pBWA(V)HS vector. The primers used are listed in Supplementary Table S1. The vector was subsequently introduced into Agrobacterium tumefaciens strain GV3101. Transient expression was performed in the leaves of Nicotiana benthamiana according to the methods of Li [36]. The GFP activity of the PmPIP2;2, PmPIP1;5 and PmPIP2;1 proteins was subsequently monitored in tobacco leaf epidermal cells via a confocal laser scanning microscope. A 488-nm excitation wavelength and a 510-nm emission wavelength were used.

Plant transformants and stress resistance assessment

The stable transformation of N. tabacum with the PmNIP2;1 H and PCAMBIA2300 empty vectors was performed via Agrobacterium-mediated leaf disc transformation. Trans transgene integration was confirmed via PCR, and the expression of PmNIP2;1 was assessed via RT-qPCR. A transgenic line generated via selection with neomycin was grown in humus soil in a greenhouse (25 °C, 60% relative humidity, 12 h light/dark photoperiod), and seeds were collected. After positive screening, the T1 generation plants were used for follow-up experiments. The PmNIP2;1 transgenic and wild-type (WT) tobacco seedlings growing for 8 weeks were subjected to different stresses. The specific treatment is the same as in Sect. Plant materials and treatments. To facilitate differentiation, they are named CK1, D1, H1, and HD1. After a duration of 10 d, the RWC of the soil treated with CK1 and H1 was always maintained at 70%–80%. The RWC of the soils treated with D1 and HD1 was 55 ± 5%.We measured the activity of superoxide dismutase (SOD) and the concentrations of malondialdehyde (MDA), proline (PRO) and H₂O₂ in both the transgenic tobacco and wild-type samples. All parameters were quantified using commercial assay kits (SuZhou MengXi Biomedical Technology Co., Ltd., China) according to the manufacturer’s instructions.The expression levels of tobacco stress resistance genes (NtAQP, NtHSF, WRKY) in the transgenic tobacco were detected. The sequences of primers used were obtained from the literature (Supplementary Table S1), and the tobacco NtActin gene was selected as the internal reference.

Statistical assessments

All experiments were performed with three biological replicates. Statistical analyses were conducted using Microsoft Excel 2021. The significance of differences between control and treatment groups was assessed using SPSS software (version XX; SPSS Inc., Chicago, IL, USA), specifically with Student’s t-test (paired, where applicable).Significance levels are denoted by asterisks: *P < 0.05, **P < 0.01.

Results

Identification of the PmAQP family and phylogenetic analysis

In this study, a total of 30 PmAQPs were identified from the P. massoniana seedling drought transcriptome data. On the basis of sequence cluster analyses of P. massoniana, B. pendula and A. thaliana, the 30 PmAQP genes were classified into four subfamilies (Fig. 1): PIPs, TIPs, NIPs and SIPs, which contained 12, 11, 5 and 2 genes, respectively. The PIP subfamily amino acid sequence similarity was 61.53%, including the PIP1 and PIP2 subgroups, with 6 PIP1s and 4 PIP2s. Among several major subfamilies, 11 PmTIPs presented the highest amino acid sequence similarity, at 67.34%. The TIP subfamily was divided into four subgroups: four TIP1s, four TIP2s, two TIP3s and one TIP4. The NIP subfamily amino acid sequence similarity was 40.6%, containing three subgroups: three NIP1s, one NIP2 and one NIP3. The amino acid sequence similarity of the SIP subfamily was 34.69%, and it contained two subgroups: PmSIP1 and PmSIP2.

Fig. 1.

The neighbor-joining phylogenetic tree of AQPs gene family among P. massoniana, A. thaliana and B. pendula

Physicochemical properties of the PmAQP gene family proteins

The results of the analysis of the physicochemical properties of the PmAQP gene family proteins revealed that the length of the proteins encoded by the PmAQP genes ranged from 185 (PmPIP1;3) to 325 aa (PmNIP1;2), with molecular weights (MWs) ranging from 19.331 to 34.625 kDa (Table 1). The theoretical isoelectric points (pIs) of the PmAQP gene family protein sequences ranged from 5.13 (PmTIP2;4) to 10.17 (PmSIP2;1), and 17 of the 30 PmAQP proteins presented a theoretical pI greater than 7.5, indicating that they are basic proteins; 9 of the 30 PmAQP proteins presented a theoretical pI of less than 6.5, indicating that they are acidic proteins. Compared with other subfamily proteins (pI 8.69), TIPs generally presented a relatively lower pI (5.98). The aliphatic indices of the PmAQP gene family protein sequences ranged from 92.19 to 114.85. The average hydropathicity of the PmAQP gene family proteins was positive, revealing that they were hydrophobic proteins.

Table 1.

Basic information on AQP genes identified in P. massoniana

ID	Gene name	Length(aa)	MW(kDa)	pI	Instability index	GRAVY	Aliphatic index
Cluster-17498.50806	PmNIP1;1	308	33.181	8.45	26.29	0.455	107.44
Cluster-33388.0	PmNIP1;2	325	34.625	8.74	39.91	0.371	101.29
Cluster-21412.0	PmNIP1;3	315	34.215	7.68	20.25	0.308	94.79
Cluster-51085.0	PmNIP2;1	306	32.182	9.34	30.78	0.335	93.14
Cluster-17498.29551	PmNIP3;1	279	29.549	8.36	35.24	0.639	108.03
Cluster-42323.0	PmPIP1;1	285	30.527	9.16	31.22	0.419	97.62
Cluster-48208.0	PmPIP1;2	257	26.997	7.64	31.42	0.389	94.2
Cluster-46824.1	PmPIP1;3	185	19.331	9.44	32.17	0.567	100.22
Cluster-46824.2	PmPIP1;4	282	29.991	8.82	34.42	0.473	96.95
Cluster-17498.10160	PmPIP1;5	285	30.702	6.3	36.18	0.365	94.14
Cluster-17498.30451	PmPIP1;6	288	30.701	6.64	30.05	0.411	92.19
Cluster-17498.31169	PmPIP1;7	288	30.862	8.18	29.58	0.345	92.88
Cluster-17498.29372	PmPIP2;1	281	30.334	9.25	31.94	0.425	94.84
Cluster-17498.32025	PmPIP2;2	282	30.24	8.86	31.98	0.494	97.98
Cluster-17498.29707	PmPIP2;3	275	30.09	9.12	36.43	0.421	101.45
Cluster-17498.25631	PmPIP2;4	188	19.996	9.8	32.03	0.446	98.67
Cluster-17498.39566	PmPIP2;5	241	25.927	9.61	29.94	0.408	104.44
Cluster-17498.32106	PmSIP1;1	238	25.207	9.64	23.18	0.741	114.45
Cluster-17498.11198	PmSIP2;1	241	26.492	10.17	19.88	0.567	114.85
Cluster-46300.0	PmTIP1;1	253	25.792	5.32	29.82	0.834	112.25
Cluster-17498.47949	PmTIP1;2	253	25.942	5.59	27.87	0.711	104.94
Cluster-17498.5	PmTIP1;3	252	25.94	5.93	28.47	0.751	109.25
Cluster-46755.0	PmTIP1;4	256	26.173	5.91	19.95	0.886	112.93
Cluster-45777.0	PmTIP2;1	248	24.812	5.58	24.45	0.981	111.41
Cluster-46975.0	PmTIP2;2	249	25.007	5.89	21.91	0.975	112.13
Cluster-17498.27958	PmTIP2;3	250	25.346	6.16	26.2	0.877	111.6
Cluster-17498.48021	PmTIP2;4	221	22.776	5.13	26.5	0.872	112.08
Cluster-48192.0	PmTIP3;1	254	27.077	7.13	38.94	0.579	104.61
Cluster-48701.0	PmTIP3;2	255	27.075	6.54	37.2	0.604	107.18
Cluster-17498.15795	PmTIP4;1	204	21.278	6.56	29.94	0.815	110.39

The number of transmembrane helices in PmAQPs and their subcellular localization are tabulated in Table 2 Protein transmembrane domain (TMD) prediction analysis of PmAQPs via three different programs is indicated, and the majority (24, 80%) of the identified putative PmAQPs contained six TMDs. PmPIP1; 2, PmPIP2; 5, PmTIP2; 4, PmTIP4; 1 with 5 TMDs. Two different programs were used to predict the subcellular localization of PmAQPs, as shown in Table 2. On the basis of the results of the subcellular localization prediction, all PmAQPs likely localized to the plasma membrane. Moreover, the subcellular localizations predicted by ProtComp 9.0 were diverse and not always in agreement with those predicted by WoLF PSORT. Most PmAQPs exhibit multipositional localization and are simultaneously localized in the plasma membrane, cytoplasm and Golgi.

Table 2.

Number of transmembrane helices and subcellular localization of PmAQP genes

Gene	TMD (Deep TMHMM)	TMD (TOPCONS)	TMD (CCTOP)	Subcellular localization (ProtComp 9.0)	Subcellular localization (WoLF PSORT)
PmNIP1;1	6	6	6	plas	plas
PmNIP1;2	6	6	6	plas	plas
PmNIP1;3	6	6	6	plas	plas
PmNIP2;1	6	6	6	plas, cyto, Golg	plas
PmNIP3;1	6	6	6	plas	plas
PmPIP1;1	6	7	6	Plas, Golg	plas
PmPIP1;2	5	5	5	plas	plas
PmPIP1;3	4	4	4	Plas, ER	plas
PmPIP1;4	6	6	6	plas	plas
PmPIP1;5	6	7	6	plas	vacu
PmPIP1;6	6	6	6	plas	plas
PmPIP1;7	6	6	6	Plas, Golg	plas
PmPIP2;1	6	6	6	plas	plas
PmPIP2;2	6	6	6	Plas, cyto, Golg	plas
PmPIP2;3	6	6	6	plas	plas
PmPIP2;4	4	4	4	plas	plas
PmPIP2;5	5	5	5	plas	plas
PmSIP1;1	6	6	6	plas	plas
PmSIP2;1	6	7	6	plas	plas
PmTIP1;1	6	6	6	Plas, cyto, Golg	plas
PmTIP1;2	6	6	6	Plas, cyto, Golg	vacu
PmTIP1;3	6	6	6	Plas, cyto, Golg	plas
PmTIP1;4	6	6	6	Plas, cyto, Golg	-
PmTIP2;1	6	7	6	plas, cyto, Golg	plas
PmTIP2;2	6	7	6	plas, cyto, Golg	plas
PmTIP2;3	6	7	6	plas, cyto, Golg	plas
PmTIP2;4	5	6	5	plas, cyto, Golg	plas
PmTIP3;1	6	6	6	plas, cyto, Golg	plas
PmTIP3;2	6	6	6	plas, cyto, Golg	vacu
PmTIP4;1	5	6	5	plas, cyto, Golg	plas

Analyses of conserved motifs and homology modelling of AQP genes in P. massoniana

The distribution of 30 PmAQPs conserved motifs is shown in Fig. 2, and Table 3 shows that the PmAQP gene family contains 10 main conserved motifs. Motif 1 was found in all the PmAQPs, except for the SIP subfamily. With the exception of PmSIP2; 1, all the remaining members contained Motif 2. The protein sequences of the TIP subfamily members were highly similar. Some subfamilies contain several relatively specific motifs. For example, motifs 7, 8 and 9 were found only in the subfamily PmPIP, and motifs 6 and 10 were found only in the subfamily PmTIP. The 3D modelling results revealed that these aquaporins presented tertiary structure similarity (Fig. 3). Almost all PmAQP protein monomers contain a six α-helical transmembrane domain (TMD1-TMD6), five loops (LA-LE), and two additional half-helices (HB, HE). However, PmPIP1;2, PmPIP2;5, PmTIP2;4, and PmTIP4;1 lacked partial structures.

Fig. 2.

Conserved motif distribution map of PmAQP genes. Different color blocks represent different motifs

Table 3.

Information on the 10 conserved motifs of PmAQP genes

Motifs	Motif Sequence	Sites	Width
1	GISGGHVNPAVTFGLFLAGKISLPRAVLYWIAQCLGAICAAGLLKGFTGG	28	50
2	PLPIGFAVGLVILATGPITGTSMNPARSF	28	29
3	YDRYGGGANTVAHGYTKGDGLGAEIIGTFVLVYTVFSATDAKRSARDSHV	12	50
4	GSWDDHWIYWVGPLIGAALAALYY	22	24
5	SFYRALIAEIVATFLLLYTVYATAIDPKK	21	29
6	LAEFISTLJFVFAGEGSGIAYDKLTSDAALTPAGLVAVALAHAFALFVAV	10	50
7	SSSKCGSVGIQGIAWAFGGMIFALVYCTA	10	29
8	EQAQSQSKDYTEPPPAPLFEPGEFKSW	9	27
9	MEGKEEDVKLGANKFPERQPI	5	21
10	FGRFDEATHPDALRA	11	15

Fig. 3.

Predicted 3D protein models of PmAQP. Models were generated by using the Phyre2 server. Models were visualized by rainbow color from the N- to C-terminus

Conserved NPA amino acid residues of PmAQPs

Double NPA motifs are highly conserved in PmPIPs and include typical NPA residues of AQPs(Table 4). In PmSIP1;1, residue A (arginine) in the third position was replaced by a threonine (T). In PmSIP2;1, residue A (arginine) in the third position was replaced with valine (V). The amino acid residues on the selectivity filter (ar/R) revealed that PmPIPs presented variation only in LE2 (F-H-T-R/K). The amino acid sequences of PmPIP2;4 and PmPIP2;5 lack the first residue of the Ar/R selection filter. The positions LE1 and LE2 were highly variable in PmTIPs, whereas H1 and H2 were conserved (H-I-A/G-V/R/L). In PmNIP1, the ar/R selectivity filter is composed of the same F-C-A-R residues. In PmNIP2 and PmSIP, the amino acid residues on the selectivity filter (ar/R) are highly variable (S/W-I/V-G/A-R) and (L/M-T/K-P-N/T).

Table 4.

Amino acid composition of the NPA motifs, ar/R selectivity filters (H2, H5, LE1, LE2) and froger’s positions (P1-P5) of PmAQP genes

Subgroup	Protein name	NPA motifs		Ar/R selectivity filter				Froger’s position (P1-P5)
		LB	LE	H2	H5	LE1	LE2	P1	P2	P3	P4	P5
NIP	PmNIP1;1	NPA	NPA	F	C	A	R	N	D	R	G	V
	PmNIP1;2	NPA	NPA	F	C	A	R	Y	D	R	G	V
	PmNIP1;3	NPA	NPA	F	C	A	R	Y	D	R	G	I
	PmNIP2;1	NPA	NPV	S	I	G	R	F	T	A	Y	I
	PmNIP3;1	NPA	NPA	W	V	A	R	F	T	A	Y	M
PIP	PmPIP1;1	NPA	NPA	F	H	T	R	Q	S	A	F	W
	PmPIP1;2	NPA	NPA	F	H	T	R	E	S	A	-	-
	PmPIP1;3	NPA	NPA	F	H	T	R	H	S	A	-	-
	PmPIP1;4	NPA	NPA	F	H	T	R	G	S	A	F	W
	PmPIP1;5	NPA	NPA	F	H	T	K	Q	S	V	F	W
	PmPIP1;6	NPA	NPA	F	H	T	R	N	S	A	F	W
	PmPIP1;7	NPA	NPA	F	H	T	R	M	S	A	F	W
	PmPIP2;1	NPA	NPA	F	H	T	R	Q	S	A	F	W
	PmPIP2;2	NPA	NPA	F	H	T	R	Q	S	A	F	W
	PmPIP2;3	NPA	NPA	F	H	T	R	Q	S	A	I	Y
	PmPIP2;4	NPA	NPA	-	H	T	R	Q	S	A	F	W
	PmPIP2;5	NPA	NPA	-	H	T	R	Q	S	P	F	W
SIP	PmSIP1;1	NPT	NPA	L	T	P	N	F	A	A	Y	W
	PmSIP2;1	NPV	NPA	M	K	P	T	I	A	A	Y	W
TIP	PmTIP1;1	NPA	NPA	H	I	A	V	T	A	A	Y	W
	PmTIP1;2	NPA	NPA	H	I	A	R	T	A	A	Y	W
	PmTIP1;3	NPA	NPA	H	I	A	R	T	A	A	Y	W
	PmTIP1;4	NPA	NPA	H	I	A	V	T	A	A	Y	W
	PmTIP2;1	NPA	NPA	H	I	G	R	T	S	A	Y	W
	PmTIP2;2	NPA	NPA	H	I	G	R	T	S	A	Y	W
	PmTIP2;3	NPA	NPA	H	I	G	R	T	S	A	Y	W
	PmTIP2;4	NPA	NPA	H	I	G	R	T	S	A	Y	W
	PmTIP3;1	NPA	NPA	H	I	A	R	T	A	A	F	W
	PmTIP3;2	NPA	NPA	H	I	A	L	T	A	S	F	W
	PmTIP4;1	NPA	NPA	H	I	A	R	T	S	A	Y	W

In the PmPIP subfamily, the Froger’s positions were formed by Q/E/G/N/M-S-A/V/P-F/I-Y/W. The PmTIP subfamily members presented highly conserved residues at the P1 and P5 positions (T-A/S-A/S-Y/F-W). The PmNIP subfamily included N/Y/F at P1, D/T at P2, R/A at P3, G/Y at P4, and V/I/M at P5. Similarly, the P2–P5 positions in the PmSIP subfamily included A-A-Y-W, whereas the P1 position (F/I) included various residues. PmPIP1; 2 amino acid sequence missing the last two amino acid residues of Froger’s positions.

Expression profiles of candidate AQP genes

Fourteen PmAQP genes were subjected to RT‒qPCR analyses to evaluate their roles in the response to high-temperature stress, drought stress, combined high-temperature and drought stress(Fig. 4). Drought stress induced the expression of PmNIP2;1, PmPIP1;7, PmPIP2;1, PmPIP2;3, PmTIP1;2, PmTIP1;3. PmTIP2;3 and suppressed the expression of PmPIP2;5 and PmSIP1;1. The expression levels of 9 PmAQPs were upregulated by high-temperature stress: PmNIP3;1, PmPIP1;4, PmPIP1;6, PmPIP1;7, PmPIP2;1, PmPIP2;3, PmTIP1;2, PmTIP2;3, PmTIP4;1. Compared with those of the control, the expression of PmPIP1;4, PmPIP2;1, PmPIP2;3, PmTIP1;2, and PmTIP2;3 was upregulated in the P. massoniana seedlings subjected to both high-temperature and drought stress. PmPIP2;5, PmSIP1;1 expression is downregulated. The expression of PmPIP1;6 and PmPIP2;3 under high-temperature stress was greater than that under compound stress. The expression of PmTIP1;3 and PmTIP2;3 under compound stress was lower than that under single stress. With increasing stress duration and increasing degree of stress, the expression level of PmTIP1;2 continued to increase. With increasing stress duration and increasing degree of stress, the expression level of PmPIP1;6 first increased but then decreased.

Fig. 4.

Relative expression of PmAQPs in the leaf for the control (CK) and the drought (D), high- temperature (H), high- temperature and drought (HD) treatments. L (continuous treatment for 9 days), M (continuous treatment for 12 days), S (continuous treatment for 15 days).The relative expression was determined by RT‒qPCR. Values are the mean ± SD of three replications, and each replication included three technical replications. Different letters above error bars (mean ± SD) indicate significant differences among the CK, D, H, and DH treatments (p < 0.05, Duncan’s multiple range test)

Subcellular localization of candidates

The subcellular localization of PmPIP1;6, PmPIP2;1, and PmPIP2;2 was analysed via transient expression of the constructs in the epidermal cells of N. benthamiana. The fluorescence signals of PmPIP2;1 were predominantly observed in intracellular membrane compartments, whereas PmPIP1;6 and PmPIP2;2 displayed fluorescence at the cell periphery as well as within internal membrane structures(Fig. 5 ).

Fig. 5.

Subcellular localization of PmPIP1;6, PmPIP2;1, PmPIP2;2 in epidermal cells of N. benthamiana. Bars = 20 μm.The red channel represents chlorophyll autofluorescence

Overexpression of PmNIP2;1 enhanced the stress tolerance of transgenic tobacco

Under normal growth conditions, the phenotypes of the aboveground parts of the WT tobacco and PmNIP2;1 transgenic tobacco did not significantly differ under normal growth conditions (CK) (Fig. 6). Compared with WT plants, PmNIP2;1 transgenic tobacco exhibited greater growth under high-temperature, drought and combined high-temperature and drought stress. Compared with WT tobacco, PmNIP2;1 transgenic tobacco presented significantly greater numbers of lateral roots under both CK and stress conditions.

Fig. 6.

Comparison of the phenotypes of transgenic tobacco and wild-type tobacco after 10 days of high-temperature and drought stress. The scale bar (1 cm) in the right-hand root scanning image

Compared with WT plants, PmNIP2;1 transgenic tobacco presented significantly lower MDA contents under stress, indicating that the overexpression of PmNIP2;1 reduced the degree of membrane damage in transgenic tobacco under stress (Fig. 7). Compared with WT tobacco, PmNIP2;1 transgenic tobacco presented a significantly greater proline content and greater SOD activity under drought stress, indicating that the overexpression of PmNIP2;1 may increase the osmoregulatory capacity by increasing the proline content of plant cells. These findings suggest that the overexpression of PmNIP2;1 may increase osmoregulation and antioxidant capacity in transgenic tobacco plants, thereby conferring drought tolerance. The changes in the expression of NTAQPs, NtHSFs and WRKYs between PmNIP2;1 transgenic tobacco and wild-type tobacco were analysed, with normal-grown wild-type tobacco used as a control(Fig. 7). The overexpression of PmNIP2;1 affects the expression of endogenous stress resistance genes in tobacco.

Fig. 7.

Comparison of the MDA、H₂O₂、SOD、Pro and expression analysis of stress responsive genes of transgenic tobacco and wild-type tobacco after 10 days of high-temperature and drought stress. A(MDA), (B)(SOD), (C)(H₂O₂), (D)(PRO), (E)(NtPIP1), (F)(NtPIP2;6), (G)(NtHSF24-like), (H)(NtPIP2), (I)(NtHSF2), (J)(WRKY33). CK1 (The RWC was 70%~80%, 16/8 h light/dark photoperiod and 25 °C/25°C day/night temperature), D1 (The RWC was 55 ± 5%, 16/8 h light/dark photoperiod, 25 °C/25°C day/night temperature), H1 (The RWC was 70%~80%, 16/8 h light/dark photoperiod, 35 °C/25°C day/night temperature), HD1 (The RWC was 55 ± 5%, 16/8 h light/dark photoperiod, 35 °C/25°C day/night temperature)

Discussion

Aquaporins play crucial roles in stress tolerance in plants by altering the permeability of various membranes to water and small solutes in response to drought stress [37]. However, little information is available on their evolutionary and functional roles in Pinus plant species. In the present study, we conducted transcriptome-wide identification of PmAQPs, with a focus on both drought stress and high-temperature stress, which will provide new clues for further research on the function of AQP genes and genetic improvement in P. massoniana. In this study, phylogenetic analysis revealed that these 30 PmAQP genes can be divided into four subfamilies: PIPs, TIPs, NIPs and SIPs. This phenomenon is the same as that of Picea glauca, which belongs to Pinaceae [38]. The physical and chemical properties of different subgroup members differ to some extent. The NIP subfamily possesses a relatively large number of amino acids. In addition, proteomic analyses revealed that TIPs had a much lower pI (average pI of 5.66) than other AQP proteins (average pI of 8.87) in Acacia auriculiformis [39]. Our results reflect the same findings, i.e., a lower pI for TIPs, with an average pI of 5.98, than for other AQP proteins, with an average pI of 8.59 (Table 1).

Sequence analysis revealed that the PmAQP protein members of the same subfamily presented high similarity in terms of conserved motifs and conserved amino residues. PIPs and TIPs are the most abundant and extensively studied AQPs under both normal and stress conditions. The PIP members in P. massoniana harbor a typical NPA motif and the hydrophilic ar/R selectivity filter F-H-T-R (Table 4), which has also been observed in the PIP family of AQPs in other plant species, such as Citrullus lanatus [40] and Coffea canephora [41]. In P. massoniana, TIP members presented conserved NPA motifs and H-I-A/G-V/R ar/R selectivity filters. The H-I-A-V residue forms a more hydrophobic ar/R filter than H-I-A/G-R does. In addition to transporting water, PIP is involved in CO₂ and H₂O₂ transport in plants, and TIP is involved in NH₃ and water‒ammonia transport in plants [42].

In many studies, high-temperature stress modulates the transcription of plant AQPs, opening a new avenue for identifying P. massoniana AQPs involved in thermotolerance [43–45]. However, the relationship between PmAQPs and heat resistance in P. massoniana remains elusive. Further expression analyses also confirmed that high-temperature stress could significantly activate or inhibit the expression of candidate PmAQPs (Fig. 4). Higher expression of the PmAQP gene may be associated with increased stomatal aperture via changes in CO₂ transport in the plant leaf. This increases the effectiveness of Rubisco and can potentially influence transpiration efficiency [46]. The stress combination resulted in a new profile of transcript expression that could not be predicted by the study of each of the different stresses applied individually [47]. Interestingly, similar alteration patterns were observed for PmAQPs in response to high-temperature, drought stress and combined high-temperature and drought stress, and the differences were focused mainly on the degree of expression of PmAQPs.

Overexpression of aquaporin genes generally significantly enhances plant adaptation to various abiotic stresses [48]. In this study, heterologous overexpression of PmNIP2;1 in tobacco significantly promoted root development (Fig. 6), a phenotype consistent with previous findings on the regulation of root growth by aquaporins [49]. Existing research indicates that aquaporins play a crucial role in maintaining cellular osmotic homeostasis and regulating the accumulation of protective solutes [50]. Consistent with this, PmNIP2;1 overexpression lines exhibited higher levels of proline accumulation and SOD activity under stress conditions(Fig. 7), suggesting enhanced osmotic regulation and antioxidant defense systems, which effectively mitigated membrane lipid peroxidation damage. At the molecular level, PmNIP2;1 overexpression significantly reshaped the plant stress response regulatory network, maintaining higher expression levels of endogenous aquaporin genes under stress conditions and specifically inducing the activation of heat shock transcription factors. The latter may further enhance plant heat tolerance by maintaining protein homeostasis and intracellular environmental stability [51]. The synergistic regulatory mechanisms at the morphological, physiological, and molecular levels collectively explain the intrinsic reasons why PmNIP2;1 overexpression plants can maintain better growth conditions under drought, high temperature, and their combined stresses.

Conclusions

The present study revealed that the P. massoniana transcriptome harbors 30 genes encoding putative AQPs. On the basis of phylogenetic analysis, the 30 candidate PmAQP genes were divided into four subfamilies: 12 belonged to the PIP subfamily, 11 belonged to the TIP subfamily, 5 belonged to the NIP subfamily, and 2 belonged to the SIP subfamily. A comparison of the PmAQP sequences revealed great diversity in the NPA motifs and AR/R selectivity filters. Predictive analysis revealed that this diversity may facilitate the transport of water, CO₂, H₂O₂, and ammonia throughout the leaf. PmPIPs localize to multiple positions, including the endoplasmic reticulum, vacuole plasma membrane, and endomembrane. RT‒qPCR revealed that the relative expression levels of these 14 selected PmAQP genes responded differently to high-temperature, drought or stress combination treatment in P. massoniana. These findings could be exploited in future breeding efforts aimed at improving plant adaptation to climate change. The overexpression of PmNIP2;1 promotes the root development of transgenic tobacco and may improve the tolerance of transgenic tobacco to high temperature, drought and the combination of high temperature and drought by regulating the expression of stress resistance genes and changes in physiological and biochemical substances.

Abbreviations

AQP

Aquaporin

PIP

Plasma membrane intrinsic protein

NIP

Nod26-like intrinsic protein

TIP

Tonoplast intrinsic protein

SIP

Small and basic intrinsic protein

XIP

X-intrinsic protein

MIP

Major intrinsic protein

TMD

Transmembrane domain

NPA

Asparagine-proline-alanine

ar/R

Aromatic/Arginine region

MW

Molecular weight

pI

Isoelectric point

GFP

Green fluorescent protein

GRAVY

Grand average of hydropathicity

UBC

Ubiquitin-conjugating enzyme

NJ

Neighbor joining

RT‒qPCR

Real-time fluorescent quantitative polymerase chain reaction

RWC

Relative water content

WT

Wild type

SOD

Superoxide dismutase

MDA

Malondialdehyde

PRO

Proline

Authors’ contributions

Yan Li and Liangliang Li wrote the main manuscript text and ‌Investigation, Guijie Ding contributed to ‌ ‌Writing - review & editing‌ and ‌Resources‌. All authors reviewed the manuscript.

Funding

This work was supported by the Science Technology Major Project of Guizhou Province ([2024010]) and Science Technology Planning Project of Guizhou Province (ZK[2022]YIBAN274).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author (Guijie Ding, gjding@gzu.edu.cn) on reason able request. The nucleotide sequences and amino acid sequence of PmAQPs are provided in Supplementary Material 1.

Declarations

Ethics approval and consent to participate

The plant material utilized in this study does not involve any endangered or protected species. The ongoing experimental research and field study, including the collection of plant material, adhere to relevant institutional, national, and international guidelines and legislation for the purpose of research and development. Masson pine (Pinus massoniana) specimens were obtained from a state-owned forest farm located in Dushan County, Guizhou Province, within the Qiannan Buyi and Miao Autonomous Prefecture. All samples were accurately identified by Jiaguo Wang. The specimens were deposited in the Tree Herbarium of the College of Forestry at Guizhou University (GZAC), with a deposition number of GZAC0031640 (https://www.cvh.ac.cn/spms/detail.php?id=f01e49a7).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.