Introduction

The generation of reactive oxygen species (ROS) is particularly enhanced when animals and plants are subjected to abiotic and/or biotic stresses. ROS are chemical species that are produced by incomplete reduction of oxygen; they include the superoxide anion (O2−), hydrogen peroxide (H2O2), singlet oxygen (1O2) and the hydroxyl radical (HO·). ROS participate in a wide range of important signaling processes including growth, development, the cell cycle, acclimation to stress and programmed cell death1,2,3,4. A set of H2O2-decomposing enzymes like, catalases (CATs) and peroxidases cope with cells to uncontrolled oxidation status1,5,6. Peroxidases may be heme or non-heme peroxidises, the heme peroxidases having a cofactor in their active site (such as ascorbate peroxidases, APXs) or, a redox active cysteine (Cys) or selenocysteine (Sec) residues respectively7. Thiol peroxidases such as thioredoxin (Trx) peroxidases or peroxiredoxins (Prxs) and glutathione peroxidases (GPXs) are belongs to family of non-heme peroxidases. Moreover, GPXs, glutathione S-transferases (GSTs) with GPX activity and Prxs are also able to decompose alkyl hydroperoxides in addition to H2O28,9. For decades, GPXs by using glutathione (GSH) or other reducing equivalents as a reductant, have been known to catalyse the reduction of H2O2 or other organic hydroperoxides in to water or the corresponding alcohols10.

The GPXs were the first solenoenzyme that was discovered from mammals11,12. The reaction takes place at a single redox centre with Sec as the redox-active residue in selenocysteine-GPXs (Sec-GPXs). The catalytic centre of Sec-GPXs was first characterised as a triad composed of Sec or Cys, glutamine (Gln) and tryptophan (Trp)13, but later turned out to be a tetrad with an additional asparagine (Asn)14,15. In contrast, the peroxidative Sec was replaced by a Cys and function via a second redox centre that contains a resolving Cys in GPXs of most non-vertebrates. The former type of enzyme is more or less specific for GSH, while the latter is reduced by “redoxins”. The common denominator of the GPX family is the first redox centre comprising (seleno) Cys, Trp, Asn and Gln15,16; this kind of GPXs are shown to be reduced by redoxins, namely, thioredoxins or related proteins with a CXXC motif in plants, yeast, insects and protists15,17.

GPX genes from a range of plant species, such as Nicotiana sylvestris18, Citrus sinensis19,20, Avena fatua21, Spinacia oleracea22, Brassica campestris23, Helianthus annuus24, Pisum sativum25, Lycopersicon esculentum26, Arabidopsis thaliana27,28,29, Oryza sativa30, Triticum aestivum31, Panax ginseng32, Populus trichocarpa33,34 and Thellungiella salsuginea35, have been isolated and characterised. Unfortunately very few reports on genome-wide identification and characterisation of GPX family proteins to date have been reported. Furthermore it has been reported that the plant GPX family not only consists of multiple GPXs with distinct subcellular localisations and functions but also exhibited differential tissue-specific expression patterns, responses to environmental stress, functioning in co-ordination against ROS, immune responses, regulating the growth and development of plant root, and stomatal movement27,28,36,37 based on studies of four GPX family proteins from A. thaliana36,37, Lotus japonicus38, P. trichocarpa5 and T. salsuginea35. Therefore, these results also highlighting the importance of GPX study at the genome-wide level.

Cotton has a long cultivation history and is currently the main source of natural textile fibre and an important crop for oil and biofuel39. There are four commercially grown cotton species, Gossypium hirsutum, Gossypium barbadense, Gossypium herbaceum and Gossypium arboretum40. G. hirsutum and G. barbadense are the two most commonly cultivated species and producing ~98% of the textile fibre worldwide. Moreover, it requires to mention that there are no reports providing insight on the expression profiling of GPXs in G. hirsutum under abiotic stress conditions. In the present study, we have identified 13 GPX genes from G. hirsutum and not only their gene structure and promoter sequences were analysed, but also their potential subcellular locations were predicted. We examined the expression of GhGPX transcripts from the leaves and roots of G. hirsutum under short-term exposure to salt, osmotic and abscisic acid (ABA)-induced stresses to address their role in these stresses. Further exploring their role under abiotic stresses, gpx3Δ (H2O2-sensitive mutant) of Saccharomyces cerevisiae were complemented with the GhGPXs, and it suggests that GhGPXs have similar function to GPX3 in yeast, revealing their participation in the oxidative stress response.

Results

Identification and characterisation of GhGPX genes

We searched the GhGPX sequences from the Cotton Genome Project (CGP) database (http://cgp.genomics.org.cn/page/species/index.jsp)39, which is the recent release of the first version of the G. hirsutum genome by using the coding sequences (CDSs) of GPX1–8 from A. thaliana (AtGPX). 12 GhGPX genes had been identified and which were represented putative GhGPX1, GhGPX2, GhGPX3, GhGPX4, GhGPX5, GhGPX6, GhGPX7, GhGPX8, GhGPX11, GhGPX12, GhGPX13 and CotAD_39521. Moreover, we cloned CDSs of the 12 GhGPXs in at least three independent experiments and it was found that a nucleotide sequence of 45-bp inserted between nucleotides 373 and 374 bp of the GhGPX2 CDS, leading to a 15-amino-acid (GFLGSRIKWNFTKFL) insertion between 124S (Ser) and 125 V (Val). However, this 45-bp nucleotide sequence was considered as part of an intron on CGP database39. Thus it could be possible that GhGPX2 has two transcripts in young seedlings. We also found a 39-bp nucleotide sequence insertion behind the 441-bp of the CDS of GhGPX4 (marked with yellow in the Supplementary Data 1), and this 39-bp nucleotide sequence was not exist in the intron of GhGPX4 from the database. Therefore, we had cloned a part of the DNA sequence of GhGPX4 and found an error in the sequence of GhGPX4 that was reported on the database (the sequence between two yellow marked sequences of the cloned GhGPX4 in Supplementary Data 1). 72% identity with AtGPX6 to the 43-225-amino-acid sequence of the N terminus of CotAD_39521 realized it was a specific GPX, however, its amino-acid sequence (226–545) had 72% identity with a cysteinyl-tRNA synthetase (AT5G38830). Moreover, the cloned gene showed different CDS (Supplementary Data 1) to CotAD_39521, when we cloned CotAD_39521 by using two sets of primers (Supplementary Table S1); Gh_A12G2084 and Gh_D12G2260 were found to have high homology with CotAD_39521, when the protein sequence of CotAD_39521 was used for a Blast search with another genome database of the G. hirsutum41. This suggested us an error in the gene splicing of CotAD_39521. The CDSs of Gh_A12G2084 and Gh_D12G2260 were also cloned by PCR and we found that there was a deletion of 13-bp nucleotide sequence behind the 675-bp nucleotide (from the start codon, marked with yellow in Supplementary Data 1) of Gh_A12G2084 and Gh_D12G2260, this frame shift error evoked the emergence of a termination codon (marked with red in Supplementary Data 1). Thus it is possible that both Gh_A12G2084 and Gh_D12G2260 have two transcripts, Gh_A12G2084 and Gh_D12G2260 represented putative GhGPX10 and GhGPX9, respectively. Therefore, the renewed GhGPX2, GhGPX4, GhGPX9 and GhGPX10 sequences were used in the following bioinformatic analysis of the GhGPX sequences. The CDSs of 13 GhGPXs were shown in the Supplementary data 2, and the genomic locations of 13 GhGPXs were shown in Table 1.

Table 1 The genomic locations of GhGPX genes.
Full size table

Notably, twice of some GPX genes were found in allotetraploid than those of diploid cotton. Three genes had one copy in the diploid, while five homologous genes, from the A and D subgenomes (GhGPX2, 4, GhGPX3, 5, GhGPX7, 11, GhGPX8, 12 and GhGPX9, 10) were found in the tetraploid G. hirsutum. The sequences of homologues genes were inconsistent, which might have been resulted from loss of gene during their individual evolutionary processes or from assembly error in partial chromosomal regions, although this requires confirmation. Although GhGPX6 and GhGPX9 were located in chromosome 8 and 12, respectively, the same part sequence of GhGPX9 (190–702 bp) was shared by CDS of GhGPX6 (513 bp) and the promoter sequence of GhGPX6 was also similar to that of GhGPX9. This data vividly suggested that GhGPX6 is another copy of GhGPX9 and they may have two different transcripts.

To confirm the potential functions of the 13 putative GhGPXs, we analysed the homology between GhGPX and AtGPX protein sequences using WU-BLAST2 (http://www.arabidopsis.org/wublast/index2.jsp). Using the protein sequences of GhGPXs, we found the highest homology of AtGPX to each GhGPX protein (Table 2), for example, the amino acid sequences of GhGPX1 and AtGPX1 had 79% identity, while GhGPX6 and AtGPX6 had 78% identity. This suggested that GhGPXs may have similar functions to those of their homologues in Arabidopsis.

Table 2 Homology analysis between GhGPXs and AtGPXs.
Full size table

Identification of intron-exon from GhGPX family genes

Open reading frames (ORFs) of all 13 GhGPX genes consisted of six exons and five introns (Fig. 1, and the sequences were shown in Supplementary data 3). Similar exon-intron organisations in GPX genes were also reported from various plants like eight AtGPXs36, the six GPX genes from P. trichocarpa (PtGPX)5, the five GPX genes from L. japonicus (LjGPX)38 and the eight GPX genes from T. salsuginea (TsGPX)35, as well as some GPX genes from citrus20 and wheat31, suggesting a high degree of conservation in the exon-intron organisation among plant species. The second, third and fourth exons showed same length were of 77, 62 and 119-bp respectively, showed a same length in all 13 GhGPXs. The length of the fifth exon of 13 GhGPXs was also similar, with a length of 180 bp in GhGPX6, GhGPX9 and GhGPX10, and other GhGPXs were 168 bp.

Figure 1: Genomic organisation of the GhGPX gene family.
figure 1

The exons are depicted in yellow boxes and introns by black lines. All exon and intron lengths are drawn to scale.

Full size image

Analysis of putative cis-acting regulatory elements from GhGPXs promoters

The plant GPX genes are well known to be involved in responses to various environmental stresses29,31,32,36. Therefore, we analysed the promoter regions of all 13 GhGPX genes to find their putative cis-acting regulatory elements involved in stress and hormone signaling (Table 3) by using PlantCARE software42. Multiple different cis-acting regulatory elements were found in each GhGPX promoter region and their number varied with the individual promoter (Table 3 and Supplementary data 4). Almost all GhGPX promoters contained cis-acting elements that were responsive to hypoxia (GhGPX1–3, 5–11), methyl jasmonate (MeJA) (GhGPX1–5, 7, 8, 11, 12) and drought (GhGPX1–3, 5–12); many promoters of GhGPX contained elements that were responsive to salicylic acid (SA) (GhGPX1, 3, 5, 7, 8, 10–13), gibberellin (GhGPX1, 2, 5, 6, 9–11, 13) and auxin (GhGPX1, 4–6, 8–10); whereas fewer GhGPX promoters contained elements that were responsive to ABA (GhGPX6, 7, 9, 10, 11), ethylene (GhGPX3, 8, 12, 13) and low temperature (GhGPX2, 4, 8, 13). These findings indicate that GhGPX genes may also be involved in responses to various environmental stresses and plant hormones.

Table 3 Putative cis-acting regulatory elements related to stress and hormone response in GhGPXs.
Full size table

Predicted properties of GhGPX proteins

To analyse the properties of all 13 GhGPX proteins, the amino acid sequences of GhGPXs were subjected to alignment. The alignment result revealed that all the protein sequences contained three conserved Cys residues (Fig. 2, marked with triangles) present in most of the plant GPXs, as well as three conserved domains: GPX signature 1, GPX signature 2 and a domain WNF(S/T)KF (Fig. 2, marked with yellow boxes) found in GPXs of most plant and mammalian43, besides this the other conserved sequences were also presented, such as (L/I)Y(E/D/N)(K/Q)YK(D/N)(Q/K)G(F/L), C(T/S/K)(R/C/M)(F/Y)(K/R)(A/S)E(Y/F) PIF and P(V/L/I)(Y/F)(K/Q)(Y/F)LK (Fig. 2, marked with red boxes). The comparing between the amino acid sequences of GhGPXs and previously reported GPXs of other plant species revealed that redox centre embraced with a Cys residue (known to be a catalytic site) as well as Trp, Asn and Gln conserved residues15 (Fig. 2, marked with red triangles).

Figure 2: Amino acid sequences of GHGPX proteins.
figure 2

Multiple GhGPX sequences alignment using Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/). The three conserved domains found in most animal and plant GPXs are marked in yellow boxes; the amino acid residues that form part of the proposed catalytic site of the GPXs are marked with red triangles, including the strictly conserved Cys (C), Gln (Q), Asn (N) and Trp (W); other conserved domains are marked with red boxes. The cysteine residue marked with a red triangle is replaced by Sec in mammalian phospholipid hydroperoxide GPXs; the other two conserved Cys residues are marked with green and purple triangles. Amino acid residues that are identical (or have similar properties) in all 13 sequences are marked with a black asterisk below the sequences. Amino acid residues that have similar properties are marked with tandem points below the sequences. Amino acid residues that are identical at least in five sequences are marked with a point below the sequences.

Full size image

The 13 GhGPX proteins can be divided into four major groups: GhGPX6, 8–10, 12 were classified into group 1; GhGPX1, 3, 5 into group 2; GhGPX7, 11 into group 3 and GhGPX2, 4, 13 into group 4 (Fig. 3a). Moreover, based on the phylogenetic tree analysis of GPXs from A. thaliana, G. hirsutum, O. sativa and P. trichocarpa, the GPXs could also be divided into four clades. Specifically, the GPXs of A. thaliana, G. hirsutum, O. sativa and P. trichocarpa were divided into each clade, which indicated that each GPX has a homologue in the other three kinds of plant, and thus they might have similar functions in different kinds of plant. These data suggested that the conserved nature of evolution of GPX family protein among the four plant species, it is consistent with the same number of exons and introns found in GPXs of different plants species (Fig. 1). GhGPX6, 8–10, 12 were grouped into clade I, GhGPX1, 3, 5 into clade II and the GPXs in sub-classification of clade II, including GhGPX1, PtGPX1, AtGPX1, 7 and OsGPX6, 8 were all predicted to have the subcellular localisation of the chloroplast or mitochondria (Table 4). Clade III composed of GhGPX7, 11, while GhGPX2, 4, 13 were in clade IV (Fig. 3b). These results suggested that GhGPXs have similar characteristics like AtGPXs, OsGPXs and PtGPXs.

Figure 3: Phylogenetic analysis of GPX proteins in plants.
figure 3

Only complete sequences were considered. (a) Phylogenetic analysis of GhGPXs. (b) Phylogenetic analysis of GPX in G. hirsutum, A. thaliana, O. sativa and P. trichocarpa. The tree was constructed using the neighbour-joining method of ClustalW, with 1000 bootstraps, and the bar indicates 0.1 substitutions per site. Different species of GPX proteins are denoted by different colours: red, G. hirsutum; blue, P. trichocarpa; pink, O. sativa; grey, A. thaliana. Abbreviations of plant species: Gh, G. hirsutum; At, A. thaliana; Os, O. sativa; Pt, P. trichocarpa. Accession numbers for G. hirsutum genes are indicated in Table 1 and those of other plant species are as follows (in brackets): AtGPX1 (At2g25080); AtGPX2 (At2g31570); AtGPX3 (At2g43350); AtGPX4 (At2g48150); AtGPX5 (At3g63080); AtGPX6 (At4g11600); AtGPX7 (A4g31870); AtGPX8 (At1g63460); OsGPX1 (Os04g0556300); OsGPX2 (Os06g0185900); OsGPX3 (Os02g0664000); OsGPX4 (Os03g0358100); OsGPX5 (Os11g0284900); OsGPX6 (P0568D10.7); OsGPX7 (Os04g0556300); OsGPX8 (Os06g0185900); OsGPX9 (LOC_Os11g18170); PtGPX1 (ABK96776); PtGPX2 (DT518382); PtGPX3.1 (ABK96047); PtGPX3.2 (ABK94488); PtGPX4 (ABK95195); PtGPX5 (2P5Q_A).

Full size image
Table 4 Predicted properties of GhGPXs.
Full size table

The GhGPX proteins can be divided into two categories based on the difference between the length of amino acid sequences: the first category comprised GhGPX1 (242 amino acids, 26.65 kDa), GhGPX9 (233 amino acids, 25.49 kDa), and GhGPX10 (233 amino acids, 25.56 kDa); while the second category comprised with the other GhGPXs having 166–171 amino acids (Table 4). Protein subcellular localisation prediction programmes suggested most of the GhGPX proteins localised in the cytoplasm, while GhGPX1 in chloroplasts and GhGPX9, 10 in chloroplasts or mitochondria (Table 4). The putative homologues of GhGPXs in Arabidopsis and poplar are also shown in Table 4.

Although GhGPX9 and GhGPX10 (consisted of ~233 amino acids) were clustered in clade I along with GhGPX6, 8, 12, which each had about 166–170 amino acids. The GhGPX1 comprised of 242 amino acids and a chloroplast transit peptide signal (1–30 amino acids) was at the N-terminal; in contrast, the 1–30 amino acids residues of GhGPX9, 10 may be a mitochondrial or chloroplastic targeting peptide, as predicted by iPSORT and TargetP, which were used to predict the subcellular localisation of eukaryotic proteins (http://ipsort.hgc.jp/index.html#predict; http://www.cbs.dtu.dk/services/TargetP/).

Differential expression pattern and subcellular localisation of GhGPX products

To determine which tissues normally express GhGPX genes, gene-specific primers were used for transcription analysis in roots, stems, flowers, cotyledons and leaves of G. hirsutum. We had designed six specific primers for real-time quantitative PCR (qRT-PCR): GhGPX1–3, 5, 8, 13, owing to the high homology between GhGPX genes.

As shown in Fig. 4a, the GhGPX1 gene encoding a putative chloroplast protein not only showed expression pattern in roots, stems, flowers, leaves and cotyledons, but was also with most strongly expression level in flowers, and high transcript level in leaves. However, it was nearly absent in roots; the possible explanation to this may be due to the predicted subcellular localisation of GhGPX1 being the chloroplast. The Arabidopsis chloroplastic GPXs (AtGPX1 and AtGPX7) play a role in scavenging of active oxygen species under photo-oxidative stress28, and AtGPX7 also shows the highest expression level in flowers (http://bar.utoronto.ca/). In addition, the evolutionary relationship among AtGPX1, AtGPX7 and GhGPX1 is close (Fig. 3b), indicating the role of GhGPX1 in protection against ROS in response to photosynthesis. GhGPX2 transcript was high in roots, and with the lowest expression level in leaves (Fig. 4b). GhGPX3 and GhGPX5 were homologous from the A and D subgenomes, the transcripts of them showed similar patterns, with low levels in stems and high expression level in flowers and leaves (Fig. 4c,d), which is consistent with their close evolutionary relationship (Fig. 3a). GhGPX8 transcripts were present at high levels in flowers and stems, and the region of the promoter had many cis-elements, including SA, gibberellins and ethylene-responsive elements. This suggested that GhGPX8 may regulate the seedling growth and development (Fig. 4e). The level of GhGPX13 transcript was high in flowers and low in other tissues; compared with roots, the transcript in flowers was found to be nearly 30 times (Fig. 4f). This data indicates that GhGPXs play roles in specific tissues and in different growth periods.

Figure 4: Expression of GhGPX genes in leaves, flowers, roots, stems and cotyledons.
figure 4

Relative expression of each gene in organs with respect to leaves, flowers, roots, stems and cotyledons. (a) GhGPX1, (b) GhGPX2, (c) GhGPX3, (d) GhGPX5, (e) GhGPX8, (f) GhGPX13. Cotton seedlings were grown in sterile culture pots under long-day conditions (16-h photoperiod) with 26/20 °C day/night temperatures for 25 days. The leaves were the first true leaves. Steady-state mRNA levels were normalised with respect to Ubiquitin7 (UBQ7, DQ116441) and are expressed relative to the values found in the roots, which were given an arbitrary value of 1. All data are means ± SD of six replicates.

Full size image

To test the subcellular localisations of GhGPXs, N-terminal GFP-fused GhGPX proteins were expressed in the protoplast of Arabidopsis leaves. As shown in Fig. 5, GhGPX1 was expressed in chloroplasts, while GhGPX3, 5–8 were located in the cytoplasm. This result is consistent with the prediction of the subcellular localisation of GhGPXs as shown in Table 4.

Figure 5: The subcellular localisation of GhGPX proteins in mesophyll protoplast of Arabidopsis.
figure 5

Localisation of GFP signals from GhGPX proteins fused with GFP. Epifluorescence, chloroplast autofluorescence, bright field and merged images of Arabidopsis mesophyll protoplasts transfected with constructs expressing different fusion proteins. Bars = 5 μm.

Full size image

Expression analyses of GhGPX genes in response to various stress conditions

It is widely accepted that diverse environmental conditions, such as exposure to high concentrations of salt and metals, or high temperatures, induce oxidative stress in plants. Many plant species have been shown to respond to environmental stress by increasing the expression of GPXs, which might protect the cells from damage19,27,28,29,35,44. However, there were no studies of whether GhGPXs could response to the multiple environmental stresses. Considering that multiple stress-response cis-acting elements were predicted in the promoters of GhGPX genes (Table 2), we used qRT-PCR to investigate the expression profiles of GhGPXs when the plants were under high-temperature, high-salinity, osmotic and heavy-metal stress.

Short-term exposure of G. hirsutum plants to heavy metal stress (400 mM ferrous sulphate) induced or suppressed the expression of some members of the GhGPX genes at different processing times and the effect varied with the tissue, such as leaves and roots (Fig. 6a,b). Compared with the control samples, the GhGPX1–3, 5, 8, 13 expression levels changed not more than four times. In leaves, the increased of transcript level could be seen within 3–6 h of treatment for GhGPX1, 3, 5, 8; however, within 12–24 h of treatment, the expression level decreased. However, the transcript level of GhGPX2 decreased when the seedlings were treated with ferrous sulphate for 6 h but increased after 12 h of treatment. In addition, the transcript level of GhGPX13 increased with ferrous sulphate treatment for 3 h, and then decreased after 6 h. In roots, the increased in transcript level could be seen within 3 h of treatment for GhGPX1–3, 5, 8, 13, with a decline after 6 h of treatment and a further increased after 12 h of treatment (Fig. 6b).

Figure 6: Effect of various stresses on the expression of GhGPX genes in leaves and roots.
figure 6

GhGPX gene transcripts in response to various stress conditions were analysed by qRT-PCR. Cotton seedlings were grown in sterile culture pots under long-day conditions (16-h photoperiod) with 26/20 °C day/night temperatures for 25 days and then treated with 400 mM FeSO4 (a,b), 400 mM NaCl (c,d) and a high temperature (35 °C) (e,f). mRNA levels were normalised with respect to GhUBQ7 and are expressed relative to the values at 0 h (control), which were given an arbitrary value of 1 for each gene. Data represent the means ± SD of at least three replicates.

Full size image

The expression patterns of GhGPX genes were also analysed in detail following the treatments of salt (Fig. 6c,d) and high-temperature (Fig. 6e,f). Salt stress with 400 mM NaCl treatment for 24 h increased the transcript levels for GhGPX1–3, 5, 8, 13 in roots. The expression level of GhGPX2 was induced in leaves by 24 h treatment with 400 mM NaCl, and the transcript levels of GhGPX1, 3, 5, 8, 13 decreased in leaves after treatment with 400 mM NaCl for 3 h (Fig. 6c). A high temperature (35 °C) induced expression of GhGPX1–3, 5, 13 after treatment for 24 h in leaves and roots (Fig. 6e,f); however, the transcript level of GhGPX8 increased in roots after 24 h treatment and in leaves at 3 h treatment (Fig. 6f).

Induced expression patterns of GPX genes under multiple phytohormone treatments in G. hirsutum

Previous results indicated that plant hormones play a direct role in regulation of a rice GPX gene45. AtGPX3 might mediate ABA and oxidative signaling through physical interaction with the 2C-type protein phosphatase ABA INSENSITIVE2 (ABI2) and, to a lesser extent, with ABI127. AtGPX1 and AtGPX7 were also shown to regulate ROS and SA pathways, which affect leaf development, light acclimation, basal defence and cell death programmes28. RNA was isolated from roots and leaves of 10-day-old G. hirsutum plants for further study of the effects of signaling molecules on regulation of the expression level of the GhGPX gene family. The leaves were then sprayed with SA, jasmonic acid (JA) and ABA, and the roots were dipped in ABA. We performed qRT-PCR analyses to examine the expression levels of the GhGPX genes in response to the three phytohormones. ABA induced the expression of GhGPX2 in both leaves and roots with respect to controls (Fig. 7a,b). A slight increased in expression of GhGPX3 and GhGPX5 were observed in leaves and roots, while GhGPX3 and GhGPX5 showed similar expression patterns under ABA treatment. The expression of GhGPX1, GhGPX8 and GhGPX13 were also induced by ABA in leaves; however, in roots, ABA induced the expression of GhGPX1 and GhGPX13, while no obvious change was found for GhGPX8. We previously reported that the enhanced production of H2O2 in guard cells on treatment with ABA, resulted stomatal closure46. In addition, it was shown that AtGPX3 might play dual and distinctive roles in H2O2 homeostasis, acting as a general scavenger and specifically relaying the H2O2 signal as an oxidative signal transducer in ABA and drought stress signaling27. Hence, ABA induced expression of GhGPX genes may involve enzymatic responses to oxidative stress. JA treatment specifically increased the expression levels of GhGPX8 among the six analysed GhGPX genes (Fig. 7c). SA treatment induced slight up-regulation of the expression level of GhGPX1 and GhGPX8, and down-regulation of GhGPX3 (Fig. 7d). These results indicated that individual members of the GhGPX gene family respond differentially against abiotic stress and plant hormones, and these findings may provide a clue to explain the differential regulation of transcripts under stress conditions.

Figure 7: The effects of different plant hormones on the steady-state transcript level of GhGPX genes by qRT-PCR.
figure 7

Leaves or roots of G. hirsutum exposed to 400 μM ABA (a,b), 100 μM JA (c) and 500 μM SA (d). mRNA levels were normalised with respect to GhUBQ7 and are expressed relative to the values at 0 h (control), which were given an arbitrary value of 1. Data represent the means ± SD of at least three replicates.

Full size image

Functional complementation of the H2O2-sensitive yeast growth by transforming GhGPXs in gpx3Δ mutant yeasts

The GPX homolog YIR037W/HYR1/GPX3 was found in the Saccharomyces Genome Database, and a previous study indicated that gpx3Δ mutant was hypersensitive to peroxides, and with decreased ~57% GPX activity in gpx3Δ mutants with respect to wild type (BY4741)47. Plant GPXs contain Cys instead of Sec at their active site and some of them have both GPX and thioredoxin peroxidase activities, and the thioredoxin regenerating system is much more efficient in vitro than the GSH system. At present, the functions of these enzymes in plants are not addressed. To test whether the GhGPXs can detoxify H2O2, the CDSs of GhGPX1, GhGPX2, GhGPX3, GhGPX5, GhGPX6, GhGPX7 and GhGPX8 were cloned into the pYES2 vector and the recombinant plasmids were transformed into gpx3Δ cells. As shown in Fig. 8, the gpx3Δ cells were sensitive to H2O2 when OD600 was <0.01, and few yeasts were seen under 1 mM H2O2 treatment compared with that of controls, in terms of the number of wild-type BY4741 yeast cells. The gpx3Δ cells, which had been transformed with GhGPX1, GhGPX3, GhGPX5, GhGPX6, GhGPX7 and GhGPX8, were insensitive to H2O2 compared with the gpx3Δ cells when OD600 was 0.1; however, the GhGPX2-transformed gpx3Δ was found to be sensitive to H2O2. This suggested that GhGPX1, GhGPX3, GhGPX5, GhGPX6, GhGPX7 and GhGPX8 can scavenge H2O2 in yeast, with functions similar to that of GPX3 in yeast. The gpx3Δ cells, which had been transformed with GhGPX6, GhGPX7 and GhGPX8, could even grow in medium containing 1 mM H2O2, although their OD600 were 0.001 (Fig. 8). This indicated that GhGPX6, GhGPX7 and GhGPX8 could be more effective scavengers of H2O2 than other GhGPXs, which could not grow in medium containing 1 mM H2O2 when the OD600 was 0.001.

Figure 8: Some GhGPXs can functionally complement the H2O2-sensitive morphology in gpx3Δ.
figure 8

Growth of yeast strains on solid minimal medium containing 1 mM H2O2 (MOCK represents equal volume of H2O). Strain BY4741 (gpx3Δ, WT) transformed with plasmid pYSE2 and pYSE2-GhGPXs was incubated at 30 °C. Pictures were taken 4 d later.

Full size image

Discussion

Most plant defence proteins are encoded by families of closely related genes that usually display specific structural and regulatory features48. In this study, we performed genome-wide identification of the GPX family genes in G. hirsutum. The gene expression profiles of the leaves and roots of GhGPXs were investigated under high-salinity and hormone treatments to analyse their differences. Our results showed that the different members of the GPX gene family were co-ordinately regulated in different tissues under specific environmental stress conditions. The complementation analysis using the gpx3Δ cells in this study therefore provided evidence that the GhGPX1, GhGPX3, GhGPX5, GhGPX6, GhGPX7 and GhGPX8 play crucial role in scavenging of H2O2 in response to different stresses.

Characterisation of GPXs from G. hirsutum and their evolution

Based on the scans of several plant genomes, GPX family genes have been systematically investigated from A. thaliana36,37, L. japonicas38, P. trichocarpa5 and T. salsuginea35. In this study, a total of 13 GPXs from G. hirsutum were identified and classified into four groups based on their phylogenetic clades (Fig. 3a). These 13 GhGPX proteins contained not only the three motifs present in most plant GPXs, but also three conserved cysteines (Fig. 2). GhGPX1 was shown to belong to clade 2, and is an enzyme localised to plastids (Table 4, Fig. 5), with 88% identity in its amino acid sequence compared with chloroplast-localised AtGPX1. The expression level of GhGPX1 was also high in flowers and leaves, and it is consistent with the high concentrations of GSH in chloroplasts1 and its low levels in roots and stems, which suggested an important role for GPXs. This matches with previous report describing the involvement of AtGPX1 and AtGPX7 in the protection against photosynthesis-induced oxidative damage28. The plant GPXs are ubiquitous enzymes that have been shown to be present in different plant tissues, subcellular compartments and developmental stages5,35,36,37,38. ROS were produced at high levels in endoplasmic reticulum and peroxisome, surprisingly, no GhGPX was reported to target in these cellular compartments, it could be a possibility that the APX, CAT and SOD are the main antioxidant enzymes in these organs. Most of the 13 GhGPX proteins in terms of the evolutionary relationships were found to be closer to those of P. trichocarpa, revealed by phylogenetic analysis and in the same clade, each GhGPX was closer to PtGPX (Fig. 3b, marked in red and blue).

The role of GPXs during plant growth and development

It had been reported that GPX genes were involved in not only plant growth and development, but also play very crucial role in the control of plant responses to abiotic stresses, biotic stresses and phytohormones37,49. For example, AtGPX1 and AtGPX7 contribute to cross-talk between photo-oxidative stress and immune responses, therefore they might be involved in defence against virulent pathogen infection28. AtGPX2 was also found to be part of a cytosolic protein complex involved in stress defence50. In addition, AtGPX3 plays a dual role in response to stresses: on one hand, it contributes to the maintenance of H2O2 homeostasis by the elimination of H2O2 and organic hydroperoxides; while on the other hand it works as a switch in the ABA and H2O2 signaling pathway in guard cells27; AtGPX8 plays an important role in the protection of cellular components including nuclear DNA against oxidative stress29. Moreover, Atgpx1, Atgpx4, Atgpx6, Atgpx7 and Atgpx8 mutants have been shown to have significantly larger lateral root density than the wild type, while the Atgpx2 and Atgpx3 mutants were embraced with significantly lower. The role of GPXs in the hormone-mediated (auxin, strigolactone and ABA) control of lateral root development has been documented37. In this study, we observed similar characters between GhGPXs and AtGPXs. First, the same number of exon-intron organisation, conserved Cys residues and GPX domains of GhGPXs and AtGPXs. Second, the different expression patterns of GPXs were also observed in cotton. The genes GhGPX1, 3, 5, 8, 13 were expressed at high levels in leaves, of which the expression of GhGPX1 being the highest; in addition, GhGPX8 was expressed at a high level in stems and cotyledons. High expression levels of GhGPX2, 3, 13, were also found in roots. The tissue- or organ-specific GPX expression patterns indicated their functional divergence during plant development and growth. Moreover, the expression of all GhGPX genes induced by stress- or phytohormone-related signal treatment (Figs 6 and 7). Taken together, these results showed that GhGPXs are generally responsive to stress-related signals and phytohormone treatments, and play an important role in the responses to environmental stresses and the regulation of plant growth and development.

The level of GhGPXs has antioxidant function

The plant GPX isoenzymes were demonstrated to detoxify H2O2 and organic hydroperoxides, and could be involved in maintenance of the thiol/disulphide or NADPH/NADP+ balance. Our results showed that the amount of GhGPXs could functionally complement a yeast mutant defective of GPX3 gene (Fig. 8). Besides GHGPX2, the other GhGPXs such as GhGPX1, GhGPX3, GhGPX5, GhGPX6, GhGPX7 and GhGPX8 could functionally complement the H2O2-sensitive morphology in gpx3Δ yeast. AtGPX isoenzymes were able to utilize Trx as a sole electron donor for the reduction of H2O2 or hydroperoxides51, thiol-dependent redox regulation is more diverse in plants than in animals, bacteria or fungi52. These findings are consistent with the function of GPXs in catalysing reduction of H2O2. Therefore, further work to clone Trxs in G. hirsutum, which may act as downstream components of GhGPXs to regulate redox homostasis, is crucial and of great interest.

Materials and Methods

Plant materials and stress treatments

The seeds of cotton were wrapped in moist absorbent cotton and placed in a Petri dish, which were placed in an incubator for germinate with illumination at 25 °C for 3 days to germinate. The seedlings were grown in a sterile culture pots under long-day conditions (16 h/8 h light/dark photoperiod) with 26/20 °C day/night temperatures until the second true leaf appeared. Cotton seedlings were inoculated using the root dip method53. The root tissues were harvested at 24 h post-inoculation and >10 plants were collected per treatment.

Isolation and cloning of GhGPX genes

We searched most of the G. hirsutum GPX gene sequences in the Cotton Genome Project database (http://cgp.genomics.org.cn/page/species/index.jsp) using the CDS of GPX1–8 from A. thaliana with the BLAST tool, and searched GhGPX9, 10 from another database (http://mascotton.njau.edu.cn/html.Data/Genomefhsequence/2015/05/05/16ab0945-19e9-49f7-a09e-8e956ec866bf.html). The GhGPX genes were amplified by PCR (Fast Pfu DNA polymerase, Novoprotein) and the primers used for cloning and analysing the full-length CDS of GhGPX genes were designed by using Primer Premier 5 and their sequences were listed in Supplementary Table S1.

Sequence alignments, phylogenetic analysis and other bioinformatic analyses

Isoelectric point (pI) and molecular weight (MW) predictions for each deduced GhGPX were performed using the Compute pI/MW tool: EXPASY (http://web.expasy.org/compute_pi/). The deduced sequences of 13 GhGPXs were aligned with the DNAMAN program (version 4.0; Lynnon Corporation, Pointe-Claire, Quebec, Canada) to identify conserved domains. These GhGPXs sequences were also used together with the sequences of the homologous proteins from A. thaliana, O. sativa and P. trichocarpa, to construct an unrooted phylogenetic tree by the neighbour-joining method using the ClustalW2.0 (http://www.ebi.ac.uk/Tools/msa/clustalw2/) and MEGA 6.0.554 programs. For promoter analysis, 2 kb upstream region from translation start code ATG of each GhGPXs was from database of CGP39 and another database41, and then predicted cis-acting elements in (+) strand and (−) strand of each promoter of GhGPXs was found using the database of PlantCARE42 (Plant Cis-Acting Regualtory Element, http://bioinformatics.psb.ugent.be/webtools/plantcare/html/).

Gene expression analysis

Total RNA was extracted by using TRIzol reagent (Invitrogen) from G. hirsutum seedlings. The cDNA was synthesised using M-MLV Reverse Transcriptase (Promega) and was 30-fold diluted with deionised water before use as a template in qRT-PCR. The quantitative reaction was performed on a 7500 Fast Real-Time PCR System (Applied Biosystems) using GoTaq® qPCR Master Mix kit (Promega). Three independent biological and technical replicates were subjected to qRT-PCR analysis for each sample. The expression levels of GhGPXs were normalised against that of an internal reference gene, GhUBQ7 (ubiquitin 7, accession number: DQ116441). All primers used in this study are listed in Supplementary Table S1.

Subcellular localisation of GhGPXs

For green fluorescent protein (GFP) analysis, the final construct pHBT-GFP-GhGPXs and empty vector pHBT-GFP were transiently expressed in Arabidopsis mesophyll protoplasts55. Transgenic cells were examined under an FV1000 (Olympus) confocal microscope.

Fluorescence microscopy

Fluorescence was observed by using the laser scanning confocal microscope LSM FV1000 (Olympus). The fluorescence was excited at 488 nm and detected at 500–560 nm for GFP, while autofluorescence of chloroplasts was detected at 650–750 nm. Transmitted fields were collected simultaneously for use in merging images.

Yeast complementation analysis

Full-length complementary DNA of GhGPXs was cloned into the vector pYES2 and transformed into the wild type and gpx3Δ of Saccharomyces cerevisiae (genotype BY4741). Successfully transformed colonies were screened by complementation of leucine autotrophy. The presence of the correct plasmid was confirmed by PCR. The cell density of cultures was measured during exponential growth and was adjusted to OD600 1.0, was analysed with the Biophotometer No. 6131–21283 (eppendorf, Germany). The growth assays were conducted at 30 °C. Yeast strains were precultured in liquid and solid yeast peptone dextrose medium. Yeast colonies grew on medium that containing 10% galactose and 10% raffinose. Primer sequences used in the PCR are presented in Supplementary Table S1.

Additional Information

How to cite this article: Chen, M. et al. The Glutathione Peroxidase Gene Family in Gossypium hirsutum: Genome-Wide Identification, Classification, Gene Expression and Functional Analysis. Sci. Rep. 7, 44743; doi: 10.1038/srep44743 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.