Characterization of the Heavy-Metal-Associated Isoprenylated Plant Protein (HIPP) Gene Family from Triticeae Species

Heavy-metal-associated (HMA) isoprenylated plant proteins (HIPPs) only exist in vascular plants. They play important roles in responses to biotic/abiotic stresses, heavy-metal homeostasis, and detoxification. However, research on the distribution, diversification, and function of HIPPs in Triticeae species is limited. In this study, a total of 278 HIPPs were identified from a database from five Triticeae species, and 13 were cloned from Haynaldia villosa. These genes were classified into five groups by phylogenetic analysis. Most HIPPs had one HMA domain, while 51 from Clade I had two, and all HIPPs had good collinear relationships between species or subgenomes. In silico expression profiling revealed that 44 of the 114 wheat HIPPs were dominantly expressed in roots, 43 were upregulated under biotic stresses, and 29 were upregulated upon drought or heat treatment. Subcellular localization analysis of the cloned HIPPs from H. villosa showed that they were expressed on the plasma membrane. HIPP1-V was upregulated in H. villosa after Cd treatment, and transgenic wheat plants overexpressing HIPP1-V showed enhanced Cd tolerance, as shown by the recovery of seed-germination and root-growth inhibition by supplementary Cd. This research provides a genome-wide overview of the Triticeae HIPP genes and proved that HIPP1-V positively regulates Cd tolerance in common wheat.


Introduction
In recent decades, heavy-metal contamination in the environment has increased due to the rapid development of industry and the use of pesticides [1,2]. In some parts of the world, such as Ziyang County, China [3], soil conditions are detrimental to the growth of many or most plant species due to the increasing level of certain inorganic ions [4]. Sometimes, even low concentrations of heavy metals such as Cd, Pb, Al, and Hg can cause great damage to plants [5][6][7][8]. Heavy metals generally produce toxic effects in plants, including chlorosis, inhibition of growth and photosynthesis, low biomass accumulation, imbalance of nutrient assimilation and water, and senescence, which eventually lead to plant death [9]. Cd is a nonessential heavy metal and is ranked among the top 20 toxins [10], and it can affect many processes of plant growth and development. Contamination of soil with 5 µM Cd can significantly affect the seed germination and seedling growth of barley [11]. Different concentrations of Cd can also affect the root growth of wheat and rice [12,13]. Photosynthetic indices 3 of 19 In this study, genome-wide identification of the HIPP gene family in Triticeae species was performed by searching the published sequences. A phylogenetic tree of the HIPP genes was constructed and the evolutionary relationships were analyzed. Chromosome distribution and protein structure were further studied to gain a better understanding of HIPP genes in wheat. The potential functions of the common wheat HIPP genes were predicted through their expression profiles based on in silico analysis. Haynaldia villosa L. (2n = 2x = 14, VV) is a diploid wild relative of wheat. Previous studies have shown that H. villosa is resistant to various wheat diseases such as stripe rust and powdery mildew, and also possesses the characteristics of resistance to drought, cold, and salt; therefore, it provides excellent material for wheat genetic improvement [44][45][46]. As H. villosa has not been sequenced, we cloned H. villosa's HIPPs through homology-based cloning. One of the HIPP genes, HIPP1-V, was transformed into wheat to study its role in Cd tolerance. Our results will support the understanding of the evolution and diversification of HIPPs in Triticeae species at a genome-wide scale.

Genome-Wide Identification, Phylogenetic Analysis, and Protein Structure Analysis of the HIPP Gene Family in Triticeae Species
In total, 278 HIPP genes from five Triticeae species were identified from the public database. Of these, 114 were from common wheat (T. aestivum), 33 from Triticum urartu, 40 from Aegilops tauschii, 58 from Triticum dicoccoides, and 33 from Hordeum vulgare. By using the homology cloning technique, 13 HIPPs from H. villosa were cloned ( Figure 1). To understand the phylogeny of HIPP genes, the evolutionary relationships of the above 291 Triticeae HIPPs (278 from five Triticeae species and 13 from H. villosa), along with 40 from Brachypodium distachyon, 59 from rice, and 45 from Arabidopsis, were phylogenetically analyzed and a neighbor-joining tree was built ( Figure 2 and Supplementary  Table S1). The HIPPs were divided into five clades (Clades I-V) according to the classification established for Arabidopsis. Clade IV was the smallest clade with 11 members, while Clade II was the largest with 128 members. Apart from H. villosa, the other analyzed Triticeae species possessed genes in each of these clusters. In the same clade, HIPP genes from different Triticeae species were clustered together, and HIPP genes from the same species also tended to aggregate (Figure 2), showing differences between the HIPP genes of different species and the conservation of HIPP genes within the same species. TaPCS1 [43]. In this study, genome-wide identification of the HIPP gene family in Triticeae species was performed by searching the published sequences. A phylogenetic tree of the HIPP genes was constructed and the evolutionary relationships were analyzed. Chromosome distribution and protein structure were further studied to gain a better understanding of HIPP genes in wheat. The potential functions of the common wheat HIPP genes were predicted through their expression profiles based on in silico analysis. Haynaldia villosa L. (2n = 2x = 14, VV) is a diploid wild relative of wheat. Previous studies have shown that H. villosa is resistant to various wheat diseases such as stripe rust and powdery mildew, and also possesses the characteristics of resistance to drought, cold, and salt; therefore, it provides excellent material for wheat genetic improvement [44][45][46]. As H. villosa has not been sequenced, we cloned H. villosa's HIPPs through homology-based cloning. One of the HIPP genes, HIPP1-V, was transformed into wheat to study its role in Cd tolerance. Our results will support the understanding of the evolution and diversification of HIPPs in Triticeae species at a genome-wide scale.

Genome-Wide Identification, Phylogenetic Analysis, and Protein Structure Analysis of the HIPP Gene Family in Triticeae Species
In total, 278 HIPP genes from five Triticeae species were identified from the public database. Of these, 114 were from common wheat (T. aestivum), 33 from Triticum urartu, 40 from Aegilops tauschii, 58 from Triticum dicoccoides, and 33 from Hordeum vulgare. By using the homology cloning technique, 13 HIPPs from H. villosa were cloned ( Figure 1). To understand the phylogeny of HIPP genes, the evolutionary relationships of the above 291 Triticeae HIPPs (278 from five Triticeae species and 13 from H. villosa), along with 40 from Brachypodium distachyon, 59 from rice, and 45 from Arabidopsis, were phylogenetically analyzed and a neighbor-joining tree was built (Figure 2 and Supplementary Table  S1). The HIPPs were divided into five clades (Clades I-V) according to the classification established for Arabidopsis. Clade IV was the smallest clade with 11 members, while Clade II was the largest with 128 members. Apart from H. villosa, the other analyzed Triticeae species possessed genes in each of these clusters. In the same clade, HIPP genes from different Triticeae species were clustered together, and HIPP genes from the same species also tended to aggregate (Figure 2), showing differences between the HIPP genes of different species and the conservation of HIPP genes within the same species.  The conserved domains of the HIPP proteins were evaluated using the Conserved Domain Database. All HIPP proteins contained 1 or 2 HMA domains (pfam00403) and a C-terminal isoprenylation CaaX motif (Supplementary Figure S1). Significant differences in the genes' functional domains were observed between different clades (Supplementary Figure S1); for example, the 51 HIPPs in Clade I had two HMA domains and the others all had one. Members of the same clade resembled The conserved domains of the HIPP proteins were evaluated using the Conserved Domain Database. All HIPP proteins contained 1 or 2 HMA domains (pfam00403) and a C-terminal isoprenylation CaaX motif (Supplementary Figure S1). Significant differences in the genes' functional domains were observed between different clades (Supplementary Figure S1); for example, the 51 HIPPs in Clade I had two HMA domains and the others all had one. Members of the same clade resembled one another, e.g., the eight genes in Clade IV were similar in size, ranging from 139 to 175 amino acids (aa), and they all had an HMA domain and an isoprenylation motif (Supplementary Figure S1d). Minor differences within the same clades were observed, e.g., the sizes and functional domains of Clade V were more diversified; the coding sequence (CDS) lengths of the members of this

Chromosomal Distribution of HIPPs
Based on the genome databases data for common wheat, T. dicoccoides, T. urartu, Ae. tauschii, and H. vulgare, the chromosome constitution of the HIPP family was investigated ( Figure 3 and Table 1). The HIPPs were distributed on all of the chromosomes of the diploid genomes and polyploid subgenomes, indicating that the HIPP genes were distributed on the chromosomes of each homologous group. However, the number of HIPP genes of different homeologous groups varied greatly, with a minimum of 1-3 (the sixth and seventh homoeologous groups) and a maximum of 8-11 (the second homoeologous group) ( Table 1). For homoeologous chromosomes from different genomes of wheat and its ancestral/related species, most of the corresponding HIPP orthologs were present in the syntenic genome regions. For example, HIPP1, HIPP7, HIPP16, HIPP20, HIPP27, and HIPP29 were on the homoeologous Group 2 chromosomes in Triticeae species, while HIPP37 and HIPP39 were on the Group 3 chromosomes (Figure 3a-d). However, there were some exceptions; for example, HIPP4 and HIPP6 were present on Chromosome 2D, but not on Chromosome 2A or 2D in T. aestivum (Figure 3a). Furthermore, HIPP21, HIPP30, and HIPP12 were on 4AL of T. aestivum and T. dicoccoides (Figure 3a,b), but on 4AS of T. urartu (Figure 3d), indicating that Chromosome 4A of tetraploid and hexaploid experienced structural rearrangement during its evolution. The HIPP genes of the other chromosomes were potential orthologs that showed a good collinearity ( Figure 3).

Chromosomal Distribution of HIPPs
Based on the genome databases data for common wheat, T. dicoccoides, T. urartu, Ae. tauschii, and H. vulgare, the chromosome constitution of the HIPP family was investigated ( Figure 3 and Table 1). The HIPPs were distributed on all of the chromosomes of the diploid genomes and polyploid subgenomes, indicating that the HIPP genes were distributed on the chromosomes of each homologous group. However, the number of HIPP genes of different homeologous groups varied greatly, with a minimum of 1-3 (the sixth and seventh homoeologous groups) and a maximum of 8-11 (the second homoeologous group) ( Table 1). For homoeologous chromosomes from different genomes of wheat and its ancestral/related species, most of the corresponding HIPP orthologs were present in the syntenic genome regions. For example, HIPP1, HIPP7, HIPP16, HIPP20, HIPP27, and HIPP29 were on the homoeologous Group 2 chromosomes in Triticeae species, while HIPP37 and HIPP39 were on the Group 3 chromosomes (Figure 3a-d). However, there were some exceptions; for example, HIPP4 and HIPP6 were present on Chromosome 2D, but not on Chromosome 2A or 2D in T. aestivum (Figure 3a). Furthermore, HIPP21, HIPP30, and HIPP12 were on 4AL of T. aestivum and T. dicoccoides (Figure 3a,b), but on 4AS of T. urartu (Figure 3d), indicating that Chromosome 4A of tetraploid and hexaploid experienced structural rearrangement during its evolution. The HIPP genes of the other chromosomes were potential orthologs that showed a good collinearity ( Figure 3).   Comparing the A genomes of common wheat, T. dicoccoides, and T. urartu, four HIPPs were found on Chromosome 3A and three on Chromosome 6A in all three species (Table 1). However, there were differences on the other chromosomes, especially in the number of HIPPs on Chromosome 1A-five in common wheat, two in T. dicoccoides, and two in T. urartu. One possible reason for this is that HIPP has replicated and formed gene clusters over the course of the evolution of common wheat (Table 1). Moreover, the number of HIPP genes found on Chromosome 4A was also very different between the three species, with seven in common wheat, five in T. dicoccoides, and four in T. urartu (Table 1). Comparing the B subgenomes of common wheat and T. dicoccoides, it was found that except for Chromosome 7B (three HIPPs), the number of HIPPs was different on the other chromosomes, of which 1B had the largest difference, with four in common wheat and one in T. dicoccoides, and the number of HIPP genes on the remaining chromosomes differed by one (Table 1). Comparing the D genomes of common wheat and Ae. tauschii, the total number of HIPP genes was 42 and 40, respectively. The numbers of genes on Chromosomes 1D, 6D, and 7D were the same, with five, three, and two, respectively. There was one more HIPP gene on Chromosomes 2D, 3D, and 5D in common wheat than in Ae. tauschii, and one less HIPP gene on Chromosome 4D in common wheat than in Ae. tauschii. This shows that the number of HIPP genes in the D genome has been relatively stable over the course of its evolution (Table 1).

The Expression Pattern of the HIPP Gene Family in Common Wheat
To predict the potential functions of the identified HIPPs, we investigated their expression in different tissues or in responses to various biotic and abiotic stresses through in silico expression profiling. The expression patterns of wheat HIPPs in different tissues (e.g., roots, leaves, spikes, and grains) under two biotic stresses (e.g., stripe-rust pathogen CYR31 and powdery-mildew pathogen E09) and two abiotic stresses (e.g., drought and heat) were investigated using the wheat RNA-seq data from publicly available databases. The expression level was measured as tags per million (TPM), and the Z-score-normalized values of the TPM were used to generate a heat map in R Studio ( Figure 4). Tissue-specific expression analysis showed that 44 genes had the highest expression levels in the roots (e.g., TaHIPP43-A), 23 genes in the leaves (e.g., TaHIPP2-A), 42 genes in the spikes (e.g., TaHIPP29-B), and only TaHIPP20-A, TaHIPP20-B, TaHIPP21-A, TaHIPP21-D, and TaHIPP36-B in the grains (Figure 4a). Most HIPP genes showed the highest expression levels in the roots and the lowest in the grains, and the numbers of HIPP genes highly expressed in the spikes and leaves were somewhere in between. Different genes showed various expression patterns under biotic (e.g., Bgt and Pst) and abiotic (e.g., drought and heat) stresses. In response to Bgt and Pst infection, 52 and 69 genes were significantly upregulated, 31 and 13 genes were significantly downregulated, and 31 and 32 genes remained unchanged, respectively (Figure 4b,c). In response to drought and heat induction, 44

Cloning and Subcellular Localization Analysis of the HIPPs from H. villosa
The 13 cloned HIPPs from H. villosa were assigned the names HIPP1-V to HIPP13-V, which were classified into four clades. HIPP7-V, which encodes a protein of 151 aa, had the shortest CDS length, e.g., 456 bp. The longest one was HIPP2-V at 1251 bp, which encodes 416 aa. The predicted isoelectric points of the HIPPs-V ranged from 5.21 (HIPP13-V) to 10.04 (HIPP10-V) ( Table 2). Investigating the subcellular localization of a protein provides clues towards the elucidation of its function. To verify where the HIPPs-V proteins were localized in vivo, GFP (green fluorescent protein)-tagged fusion proteins were expressed and observed in the leaves of Nicotiana Benthamiana via transient expression using the Agrobacterium method. The GFP signal of the fusion proteins could be detected for all 13 cloned HIPP-V proteins, with different localization signals ( Figure 5). Compared to the control, which had an even distribution of GFP fluorescence (Figure 5a), all HIPP-V proteins were localized on the plasma membrane (PM), but with different fluorescence intensities, e.g., the fluorescence intensities of HIPP2, -4, -5, -7, -9, and -11-V were weaker than those of the others (Figure 5c,e,f,h,j,l). Some HIPPs also had obvious signals in the cytoplasm, such as HIPP2-V and HIPP9-V (Figure 5c,j). These different subcellular localizations revealed that they may be involved in multiple biological processes and may have distinct functions.

HIPP1-V Positively Regulates Cd Tolerance in Common Wheat
Previous studies have demonstrated the involvement of HIPPs in cadmium (Cd) homeostasis. Quantitative reverse-transcription PCR (qRT-PCR) analysis was used to investigate HIPP1-V expression in response to Cd. Following the treatment of H. villosa at the two-leaf stage with 1 mM Cd 2+ , HIPP1-V expression in the roots rapidly increased 1 h after inoculation (hai), reached a peak level at 12 hai, and remained at a higher level from 24 hai (Figure 6a). In the stems, HIPP1-V expression increased at 2 hai, declined at 6 hai, increased at 12 hai, and remained at a high level for all of the tested time points (Figure 6b). In the leaves, HIPP1-V expression increased at 2 hai, reached a peak level at 12 hai, and declined at 24 hai (Figure 6c). These results showed that HIPP1-V was upregulated after Cd induction. Although the expression trend was different in different tissues of H. villosa, they were all upregulated.

HIPP1-V Positively Regulates Cd Tolerance in Common Wheat
Previous studies have demonstrated the involvement of HIPPs in cadmium (Cd) homeostasis. Quantitative reverse-transcription PCR (qRT-PCR) analysis was used to investigate HIPP1-V expression in response to Cd. Following the treatment of H. villosa at the two-leaf stage with 1 mM Cd 2+ , HIPP1-V expression in the roots rapidly increased 1 h after inoculation (hai), reached a peak level at 12 hai, and remained at a higher level from 24 hai (Figure 6a). In the stems, HIPP1-V expression increased at 2 hai, declined at 6 hai, increased at 12 hai, and remained at a high level for all of the tested time points (Figure 6b). In the leaves, HIPP1-V expression increased at 2 hai, reached a peak level at 12 hai, and declined at 24 hai (Figure 6c). These results showed that HIPP1-V was upregulated after Cd induction. Although the expression trend was different in different tissues of H. villosa, they were all upregulated. The function of HIPP1-V in Cd tolerance was validated using stable transformation. The expression vector pBI220: HIPP1-V was constructed by cloning the HIPP1-V gene into pBI220 driven by the 2 × 35S promoter. The vectors pBI220: HIPP1-V and pAHC20 (carrying a Bar gene as a selection marker) were co-transformed into Yangmai158 via particle bombardment, and five positive transgenic plants were obtained (Supplementary Figure S3). OE-HIPP1-T0-17-3 and OE-HIPP1-T0-27-3 were selected for further analysis due to their higher HIPP1-V expression levels (Supplementary Figure S3b). Their derived T3 lines, OE-HIPP1-T3-1 and OE-HIPP1-T3-2, were identified as positive lines and used for Cd treatment. When treated with 0 μM Cd, there were no differences in plant height and root length between receptor Yangmai158 and OE-HIPP1-T3 (Figure 6d-f). As the concentration of Cd increased, the root length and plant height of both plants gradually decreased, but the decrease was more pronounced in Yangmai158 plants (Figure 6d,e). When the concentration of Cd reached 100 μM, the differences in root length and plant height were the greatest; Yangmai158 and OE-HIPP1-V-T3 were 8.19 cm and 11.06 cm in height (Figure 6d,g) and 3.03 cm and 5.75 cm in root length, respectively (Figure 6e,g). As the concentration of Cd continued to increase, the growth status of Yangmai158 and OE-HIPP1-V-T3 tended to be consistent (Figure 6d,e). When the concentrations of Cd reached 10 mM and 20 mM, the seeds of OE-HIPP1-V-T3 were still able to germinate, but the germination of Yangmai158 was completely suppressed (Figure 6h,i). These results indicate that HIPP1-V could significantly improve the tolerance of common wheat to Cd. The function of HIPP1-V in Cd tolerance was validated using stable transformation. The expression vector pBI220: HIPP1-V was constructed by cloning the HIPP1-V gene into pBI220 driven by the 2 × 35S promoter. The vectors pBI220: HIPP1-V and pAHC20 (carrying a Bar gene as a selection marker) were co-transformed into Yangmai158 via particle bombardment, and five positive transgenic plants were obtained (Supplementary Figure S3). OE-HIPP1-T 0 -17-3 and OE-HIPP1-T 0 -27-3 were selected for further analysis due to their higher HIPP1-V expression levels (Supplementary Figure  S3b). Their derived T 3 lines, OE-HIPP1-T 3 -1 and OE-HIPP1-T 3 -2, were identified as positive lines and used for Cd treatment. When treated with 0 µM Cd, there were no differences in plant height and root length between receptor Yangmai158 and OE-HIPP1-T 3 (Figure 6d-f). As the concentration of Cd increased, the root length and plant height of both plants gradually decreased, but the decrease was more pronounced in Yangmai158 plants (Figure 6d,e). When the concentration of Cd reached 100 µM, the differences in root length and plant height were the greatest; Yangmai158 and OE-HIPP1-V-T 3 were 8.19 cm and 11.06 cm in height (Figure 6d,g) and 3.03 cm and 5.75 cm in root length, respectively (Figure 6e,g). As the concentration of Cd continued to increase, the growth status of Yangmai158 and OE-HIPP1-V-T 3 tended to be consistent (Figure 6d,e). When the concentrations of Cd reached 10 mM and 20 mM, the seeds of OE-HIPP1-V-T 3 were still able to germinate, but the germination of Yangmai158 was completely suppressed (Figure 6h,i). These results indicate that HIPP1-V could significantly improve the tolerance of common wheat to Cd.

The Cd Tolerance of HIPP1-V Could Expand Wheat Breeding Resources
Wheat is an important crop worldwide and is known to be a leading dietary source of Cd [47]. Improving wheat tolerance to Cd and reducing Cd accumulation in grains are issues that need to be addressed in current wheat breeding. Some Cd-tolerance-related genes in wheat have been reported; including HMA-containing genes [29]. This HMA domain has a highly conserved CysXXCys (X refers to any amino acid) motif comprising a beta-alpha-beta-alpha fold shape for binding heavy metals, such as Cd, Cu, or Zn [48]. There have been two types of HMA-domain-containing proteins found in plants to date, namely HPPs (HMA plant proteins) and HIPPs (HMA isoprenylated plant proteins), which are a group of metal-binding metallochaperones [23,26]. These proteins play crucial roles in metal homeostasis and detoxification in plants. The wheat heavy-metal ATPase 2 (TaHMA2) can transport Zn 2+ and Cd 2+ across membranes, and overexpression of TaHMA2 in Arabidopsis increases the root length and fresh weight and enhances Zn 2+ /Cd 2+ root-to-shoot translocation compared to that of the wild-type [40,49]. TaHIPP1 from common wheat contains an HMA domain, and overexpression of TaHIPP1 in yeast significantly increases the cell growth rate under Cu 2+ stress [29]. The cell number regulator 2 (TaCNR2) from common wheat is also involved in regulating heavy-metal translocation; the expression of TaCNR2 in wheat seedlings increases under Cd, Zn, and Mn treatments, and overexpression of TaCNR2 in Arabidopsis and rice plants enhances their tolerances to Cd, Zn, and Mn stresses. Using new gene resources from alien species is a very useful approach for breeding new cultivars [50]. In this study, HIPP1-V, which contains an HMA domain, was upregulated in the roots, stems, and leaves of H. villosa, induced by Cd. The overexpression of HIPP1-V was able to increase the tolerance of common wheat to Cd. This result indicates that HIPP1-V may represent a gene resource for improving wheat tolerance to Cd stress.

Identification and Characterization of the HIPP Gene Family Can Help to Better Understand the Evolutionary Relationships in Triticeae Species
HIPP proteins contain an HMA domain and a conserved isoprenylated motif, and are found only in vascular plants [23,51]. HIPP-family genes have been reported mostly in Arabidopsis and rice [23,26]. de Abreu-Neto speculated that plant metallochaperone genes (including ATX, HPP, and HIPP) share a common ancestor in plants and algae species, and started diversification early in plant evolution [23]. The distribution of the HIPP genes on chromosomes and segmental duplication events highlight the HIPP family's expansion in the rice genome [26]; such gene duplication during evolution can help plants adapt to various environmental stresses [52]. In this study, a total of 291 HIPP genes were identified from five Triticeae species and H. villosa, and the HIPP gene family was divided into five clades; apart from H. villosa, the other Triticeae species possessed genes in each of these clusters. In Triticeae species, the number of HIPPs increased in fold accordingly in different ploidy species-the diploids T. urartu, Ae. tauschii, and H. vulgare had 33, 40 and 33, respectively, while the tetraploid T. dicoccoides had 58 and the hexaploid T. aestivum had 114. Chromosome location analysis indicated that the HIPP genes were distributed on the chromosomes of each homologous group, while protein structure analysis showed that the structures of the HIPP proteins in the same clade were similar. This indicates that most HIPP genes were formed before the differentiation of Triticeae species. Only a few genes might have evolved independently after the differentiation of Triticeae species, or this result might be due to the lack of genomic sequence information.

The HIPP Gene Family Might Regulate Plant Development and Various Responses to Stress in Common Wheat
Roles of HIPP proteins in heavy-metal homeostasis and/or detoxification [23], cold-and salt-stress responses [28], virus long-distance movement [35], plant-pathogen interactions, and plant development [25] have been demonstrated. HIPP20, HIPP22, HIPP26, HIPP27, AtFP6, AtCdI9, and ATX1 from Arabidopsis and OsHIPP16, OsHIPP28, OsHIPP34, OsATX1, OsHIPP60, and OsHIPP29 from rice can positively regulate tolerance to heavy metals [22][23][24]26,30,32,53]. Different genes can be induced by different heavy metals, which, in turn, increase plant tolerance to different heavy metals. AtCdI9 can be induced by Cd, Hg, Fe, and Cu [30], while OsATX1 and OsHIPP60 can be induced by Mn, Cd, and Cu [26]. OsHIPP34-expressing yeast cells grow better in medium containing 60 µM Cu, but show no difference in medium containing Mn, Cd, and Zn compared to the control. Meanwhile, OsHIPP60-expressing yeast cells grow better under Cd and Zn treatment, but show no difference under treatment with Mn and Cu compared to the control [26]. Some HIPP proteins can also bind to specific heavy metals: ATFP3 binds to transition-metal ions (e.g., Cu, Zn, and Ni) in a reversible manner [51], HIPP3 from Arabidopsis can combine with Zn [25], and AtFP6 can combine with some divalent cations, such as Pb 2+ , Cd 2+ , and Cu 2+ . The expression profiles of HvFP1 and HIPP26 are induced by cold and drought [28,33]. NbHIPP26 is expressed specifically in the vascular parenchyma and is induced by drought and PMTV (Potato Mop-Top Virus) infection-when NbHIPP26 is knocked down, the long-distance movement of the virus is suppressed [35]. HIPP3 acts as an upstream regulator of the salicylate-dependent pathway of pathogen response, and is also involved in abiotic-stress responses and seed and flower development [25]. Gene expression profiling analysis may provide important clues to gene functions. HIPP1-V from H. villosa was upregulated under Cd treatment, and overexpression of HIPP1-V indeed improved wheat tolerance to Cd stress. In this study, the HIPP genes showed different expression patterns in different tissues and different tolerances in wheat. TaHIPP16-B and TaHIPP7-B were highly expressed in the roots and were only upregulated in response to Bgt inoculation; TaHIPP24-D was highly expressed in the roots and was upregulated in response to Bgt inoculation and Pst infection; TaHIPP29-B was highly expressed in the spikes and was upregulated in response to drought and heat treatments. These results reveal the involvement of the different HIPPs in different biological processes; however, their specific roles need to be investigated by intensive functional analyses, such as the transgenic approach.

Diversification of the Subcellular Localization Pattern of HIPP in H. villosa
Protein localization and functions are related [54]. Isoprenylation, also known as farnesylation, is a post-translational protein modification that involves the addition of a C-terminal hydrophobic anchor, which is important for the interaction of a protein with membranes or other proteins [23,55]. All HIPP proteins contain an isoprenylation motif, which is then modified by isoprenylation. To the best of our knowledge, except for HIPP3 from Arabidopsis, all reported HIPP proteins localized on PM, including AtFP6 from Arabidopsis [22], NbHIPP26 from N. Benthamiana [35] and OsHIPP29 from rice [32]. Mutation in the isoprenylation domain may cause changes in protein localization; for instance, AtIPT3 farnesylation directs its localization in the nucleus/cytoplasm, whereas the nonfarnesylated protein is located in the plastids [56]. The wild-type CdI19 (GFP-CdI19W) is localized on the PM, whereas the farnesylated motif mutant CdI19 (in which the Cys-389 residue is replaced with Gly, GFPCdI19M) is localized in the entire cytoplasm, as is GFP [30]. Protein farnesylation of HIPPs has direct impacts on their function. For example, the function of AtNAP1;1 depends on its farnesylation motif-the wild-type AtNAP1;1 develops smaller leaves and cotyledons, while these organs are enlarged in plants expressing the mutant AtNAP1;1C369S (C369S transit at the CaaX motif) [57]. In our study, all HIPP proteins from H. villosa localized on PM, probably due their farnesylation, dependent on the CaaX motif. Most HMA-domain-containing proteins have heavy-metal-binding activity [23]; the membrane-associated characteristics of HIPPs-V may facilitate their binding affinities to specific heavy metals. Beside PM subcellular localization, several HIPPs-V also localized in other organelles, e.g., HIPP1/2/8/9/11-V had spot-like GFP signals in the cytoplasm; and except for HIPP2/5/9-V, all other HIPPs had nucleus GFP signals. Some HMA-domain-containing proteins participate in transporting heavy metals to maintain heavy-metal homeostasis [58]. HIPP1-V subcellularly localized both on PM and in cytoplasm. Its function in improving wheat tolerance to Cd may mediated by the binding and transportation of Cd for its homeostasis. NbHIPP26 is also subcellularly localized both on the plasma membrane and in the motile vesicles [35]. NbHIPP26, which plays roles in potato mop-top virus resistance, is related to the sensing and long-distance movement of viruses; it also regulates plant responses to drought stress [35]. The significance of the diverse subcellular localization of different HIPPs for their biological roles need to be further investigated.

Plant Materials
H. villosa (genome VV, accession no. 91C43) from the Cambridge Botanical Garden, Cambridge, U.K., was used for gene cloning and expression analysis. Commercial wheat variety Yangmai158, cultivated in the Institute of Agricultural Sciences in Lixiahe District of Jiangsu Province, was used as the transgenic receptor. N. benthamiana (NC89) plants were used for subcellular localization analysis. All plants were grown in a greenhouse and the growth conditions were as follows: 14/10 h day/night cycle, 24/20 • C day/night temperature, and 70% relative humidity.

Identification of the HIPP Gene Family in Triticeae Species
The HIPP genes from Arabidopsis and Oryza were based on those previously identified [23] and download from the National Center for Biotechnology Information (NCBI, https://www.ncbi.nlm.nih. gov/). The "HMA protein" in the annotated proteins database of Hordeum vulgare (HH, 2n = 2x = 14) [59] was searched and entries containing gene ID and protein sequences were obtained. We then performed a conserved-domains (CD) search (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and reserved the protein sequences that contained the HMA domain pfam00403.6 and the C-terminal isoprenylation motif CaaX. The presence of the isoprenylation motif, which was not characterized in the databases, was confirmed using PrePS [60].

Cloning of HIPP Genes from Haynaldia villosa
Prediction of HIPP in H. villosa was carried out by comparing the transcriptomic data of H. villosa (data unpublished) with the HIPP protein sequences of H. vulgare, and the predicted proteins were processed according to the above method. According to the sequences of the HIPP genes from H. vulgare, the primers (Supplementary Table S2) were designed to clone the full-length cDNA of the HIPP genes from H. villosa using online software Primer3 (v. 0.4.0, University of California, Oakland, CA, USA). Mixed root, stem, and leaf tissue cDNA of H. villosa served as a template for the isolation. The specific primers for the HIPP genes from H. villosa were used for cloning. This was performed at 95 • C for 30 s, followed by 32 cycles of 95 • C for 30 s, 56 • C for 45 s, and 72 • C for 1 min, and then 5 min at 72 • C in Phanta Max Super-Fidelity DNA polymerase (Vazyme, Nanjing, China).

Subcellular Localization
Subcellular localization was performed as described by Zhao et al. [66] with the following modifications. The ORFs of the HIPP-V genes (without stop codons) were cloned into vector pCambia1305-GFP to produce the fusion expression construct 1305-GFP: HIPP-V. The plasma membrane (PM) marker construct (mCherry-SYP122) was provided by Prof. Yiqun Bao. The 1305-GFP fusion constructs and mCherry-SYP122 were transformed into N. benthamiana epidermal cells by Agrobacterium tumefaciens (strain GV3101) bacteria, incubated in darkness at 22 • C for 48-60 h, and the fluorescence signals were then observed by confocal microscopy (LSM780, Zeiss, Oberkochen, Germany).

Phylogenetic Analysis of the HIPP Gene Family
Phylogenetic analysis was performed as described by de Abreu-Neto et al. [23] with the following modifications. Multiple-sequence alignment was conducted using MUSCLE, which was integrated in MEGA X. The neighbor-joining analyses were performed in MEGA X using a p-distance parameter with 2000 bootstrap replications. Only the HMA domains were used to construct the phylogenetic tree.

Chromosome Distribution and Protein Structure Analysis of the HIPP Gene Family
Chromosomal information for the predicted HIPP genes from each species was obtained after using the cDNA sequences as a query sequence blasted to the genomic sequence to determine their chromosomal locations. Their locations were then drawn onto the physical map of each chromosome using Tbtools [67].
The domains of the wheat HIPP proteins were predicted using a CD search (https://www.ncbi.nlm. nih.gov/Structure/cdd/wrpsb.cgi). The corresponding evolutionary trees were constructed by MEGA X as described above. The formation of the protein structure was constructed using Tbtools [67].

RNA-seq Expression Analysis
The expression data were downloaded from expVIP to analyze the expression pattern of predicted wheat HIPP genes in different tissues (e.g., roots, leaves, spikes, and grains) under abiotic stress (e.g., drought and heat), and biotic stress (e.g., stripe rust and powdery mildew) [68], and the Z-score-normalized values of the TPM were used to construct the expression heat map using R version 3.5.1.

Transcription Abundance Analysis
H. villosa plants were grown in liquid Murashige and Skoog (MS) medium at 22-25 • C with a photoperiod of 12 h until the third-leaf stage. For expression analysis in response to Cd treatment, plants were inoculated with 1 mM CdCl 2 and the leaves, stems, and roots were collected at 0, 1, 2, 6, 12, or 24 h after treatment. All of the samples were rapidly frozen in liquid nitrogen, then stored in an ultrafreezer (−80 • C) until use.
Total RNA was extracted using TRIZOL reagent according to the manufacturer's protocol (Invitrogen, Carlsbad, CA, USA), and the sample was used to synthesize the first-strand cDNA using an RNA PCR Kit (TaKaRa, Shiga, Japan). Specific primers for HIPP1-V were used for qRT-PCR (Supplementary Table S2), and the tubulin gene was used as the internal control for normalization. qRT-PCR was performed with AceQ SYBR Green (Vazyme, Nanjing, China) using a Louts PCR 480 sequence detection system. The reactions were subjected to the following program: 95 • C for 1 min, 41 cycles of 95 • C for 10 s, and 60 • C for 30 s. The relative values of gene expression were derived from 2 −∆∆CT [69].

Genetic Transformation
HIPP1-V was cloned into the plant expression vector pBI220 (driven by the CAMV 35S promoter) to generate vectors pBI220: HIPP1-V. The method of stable transformation using bombardment was applied according to Xing et al. [70]. The vector pBI220: HIPP1-V, together with pAHC20 (carrying the Bar gene as a selectable marker), was co-transformed by bombardment into calli produced from young embryos of wheat cultivar Yangmai158. Regenerated plants were transplanted in the field and used for further characterization. Regenerated plants were analyzed via PCR using conjugated gene-specific primers CAMV35S-F and HIPP1-V-R (Supplementary Table S2) to identify positive transgenic plants.

Cd-Tolerance Analysis
For the Cd-tolerance tests, seeds of HIPP1-V transgenic plants and the receptor variety Yangmai158 were sterilized with 75% ethanol and 12% NaClO and washed five times with sterile water. To determine plant height and root length, the seeds were grown on 1/2 liquid Murashige and Skoog medium containing different concentrations of CdCl 2 (e.g., 0, 5, 10, 20, 50, 100, 200, 500, 1000, 5000, 10,000 and 20,000 µM) for 15 days. The growth conditions were as follows: 14/10 h day/night cycle, 24/20 • C day/night temperature, and 70% relative humidity. The plant height and root length were measured and analyzed. The plant height and root length averages were computed based on the means of three independent experiments, each based on at least 10 independent plants, and the significance analysis was performed using Excel software.

Conclusions
The present study identified 278 HIPP genes from five Triticeae species and cloned 13 HIPP genes from H. villosa. Evolutionary and protein structure analyses showed that the HIPP genes are conserved in Triticeae species. The expression analysis of the HIPP gene family in common wheat and subcellular localization analysis of HIPPs-V showed that the functions of the HIPP genes may be involved in multiple pathways of plant development. HIPP1-V was upregulated in the roots, stems, and leaves of H. villosa in response to Cd treatment, and overexpression of HIPP1-V enhanced the tolerance of common wheat to Cd. Our study systematically analyzed the function and evolution of the HIPP gene family in Triticeae species and demonstrated the important role of HIPP1-V in heavy-metal homeostasis and detoxification.

Conflicts of Interest:
The authors declare no conflict of interest.