14.1 Introduction

The transcriptome is a total pool of RNA transcripts present in a cell at a particular moment. Transcriptomics assumes that cells react to environmental changes by adjusting a number of specific transcripts, such as mRNAs, transcripts of noncoding genomic regions, and regulatory RNA molecules. The ideal case of transcriptomic analysis is the genome-wide detection of transcript levels in control and study samples, followed by functional annotation of transcripts that are deemed as differentially regulated. There are several methodologies enabling researchers to study changes in the transcriptome. Most popular currently are DNA microarrays, serial analysis of gene expression (SAGE) (Velculescu et al. 1995), suppression subtractive hybridization (SSH), and RNA sequencing. Data obtained with SSH, DNA microarray technology, and RNA-seq are currently available and will be discussed here. SSH is a method developed in the 1980s and 1990s (Lisitsyn et al. 1993; Diatchenko et al. 1996, 1999). It provides a tool for comparative analyses of two populations of mRNA. In general, both populations are first converted into cDNA libraries, which are then hybridized. Hybrids are removed, whereas unhybridized cDNA fragments are regarded as differentially regulated transcripts. Those transcripts are cloned, sequenced, and annotated. Similarly, DNA microarray and RNA-seq methods rely on isolation of total RNA – which may be followed by mRNA enrichment step – and reverse transcription to create a library of cDNA molecules. DNA microarray analysis requires preparation of a chip with immobilized DNA fragments to which fluorescently labeled sample cDNA is hybridized and a fluorescent signal is read and analyzed (Grunstein and Hogness 1975; Gergen et al. 1979). It is thus necessary to have prior knowledge regarding expected transcript sequences before the experiment. This approach limits the analysis to sequences immobilized on the chip with no possibility of additional information. A gene may have more than one complementary probe on a gene chip. Thus the number of responsive probe sets might not always be equal to identified responsive genes. RNA sequencing emerged with the development of sequencing-by-synthesis technology (Ronaghi et al. 1998; Margulies et al. 2005). This approach allows analysis of theoretically all the RNA molecules present in a cell at the time of the experiment, with the limitations set by the need of reversed transcription step (lower efficiency of reversed transcription of fragments with high GC content or long homopolymer stretches). RNA-seq offers many advantages to microarray analysis: it does not require prior knowledge of analyzed sequence, and it allows whole transcriptome analysis and provides a direct measure of transcript abundance.

Although today transcriptomics is a well-established experimental approach to study cell or organism stress response to the variable environment, there are only a few reports presenting transcriptomic data on arsenic (As)-treated plants. It is because of the large size and complex organization of plant genome and the lack of sequenced and annotated reference genomes for many plant species. Here, we discuss studies conducted on rice (Oryza sativa), Arabidopsis thaliana, barrel clover (Medicago truncatula), purple willow (Salix purpurea), Indian mustard (Brassica juncea), and Abyssinian kale (Crambe abyssinica). With the exception of A. thaliana, these are economically important species, with rice forecasted paddy production of 754.6 million tonnes (500.8 million tonnes, milled basis) in 2017 (FAO 2017).

14.2 Transcriptomics of Arsenic in Rice

Majority of As-toxicity studies are performed on rice (O. sativa L.). Rice is a staple food for a large percentage of the human population and major crop accounting for one-fifth of the calories consumed by people worldwide (Smith 1998). Rice is grown in flooded fields what makes inorganic As readily available for plant intake and accumulation. Arsenic is accumulated in rice grains. Prolonged ingestion of contaminated rice can cause intoxication leading to severe health issues. This is a serious problem especially in Southeast and East Asia.

In a study by Chakrabarty and co-workers, gene expression in rice (O. sativa ssp. indica cultivar IR64) under As(III) and As(V) challenge was examined by DNA microarray (Chakrabarty et al. 2009). Ten-day-old seedlings, grown with or without 25 μM As(III) or 250 μM As(V), were used (Table 14.1). Study found that 72 genes were differentially regulated by As(V), while 27 genes were differentially regulated by As(III), although the toxic effect on seed germination was stronger in the case of As(III)-treated plants, as previously reported (Fitz and Wenzel 2002; Meharg and Hartley-Whitaker 2002; Abedin and Meharg 2002). As(V)-regulated genes are involved mainly in cell wall metabolism and primary and secondary metabolic processes, while As(III)-regulated genes are involved in hormonal and signaling processes. Both forms of As caused differential regulation of genes involved in photosynthesis, plant defense, and signal transduction. Altogether around 1% of genes on the array chip was differentially regulated (Chakrabarty et al. 2009). As(V) challenge led to increased transcription of a number of genes commonly implicated in xenobiotic stress, heavy metal response, and detoxification. Among them are ten glutathione S-transferase genes, one glutaredoxin-encoding gene, ten cytochrome P450-related genes, four metallothionein genes, 11 heat shock protein genes, and a gene encoding a germin-like protein with both superoxide and oxalate oxidase activities. Another group of genes upregulated upon As(V) treatment were genes encoding various types of transporters: one sulfate transporter, two metal transporters, two glutathione-related transporters, two multidrug and toxic compound extrusion (MATE)-efflux family proteins, one zinc-iron transport family protein, and a multidrug resistance protein. All these proteins are potentially capable of active efflux of inorganic As, as well as its organic methyl or glutathione conjugates. Transcription factors comprise another group of genes undergoing increased expression when rice seedlings are challenged with As(V). Considering the number of genes differentially regulated in such conditions, this is hardly a surprise. Some F-box, U-box proteins, and protein kinases possibly involved in signal transduction were also upregulated in As(V)-treated rice seedlings. The glycine-rich cell wall structural protein 2 precursor exhibited highest fold change in transcription; its gene was transcribed 146 times more in As(V) challenge conditions than in control seedlings. At the same time, As(V) challenge led to downregulation of many genes, often belonging to the same families as abovementioned upregulated genes. Among them were two glutathione-S-transferase genes, peroxidases, one MATE-efflux protein, one phosphate: H+-symporter, two amino acid transporters, one ATP-binding cassette transporter, one zinc transporter, F-box proteins, U-box proteins, and protein kinases. Rice seedlings exposed to As(III) upregulated transcription of one glutathione-S-transferase gene, three glutaredoxin genes, ten peroxidase family-related genes, two metallothioneins, three heat shock proteins, some transporter-encoding genes (sulfate transporter, two metal transporters, multidrug resistance protein), a number of transcription factors, F-box protein genes, U-box protein genes, and some protein kinases. Similarly to the effect of As(V), downregulation of a number of genes encoding proteins of the same groups took place upon As(III) treatment.

Table 14.1 List of plant species, arsenic source, and treatment conditions for presented transcriptomic studies
Full size table

Similar technology was employed in a study comparing an effect of As(V) on gene expression of two rice varieties: cultivars japonica (Azucena) and indica (Bala) (Norton et al. 2008a). Gene expression in roots of hydroponically grown plants exposed to As(V) challenge was analyzed (Table 14.1). Expression of 44% of genes was confirmed, with 1604 of probes significantly upregulated and 1828 probes downregulated in Azucena and 909 probes upregulated and 935 probes repressed in Bala plants. Of these 576 probes (460 genes) were significantly upregulated, and 622 probes (523 genes) were significantly downregulated in both Azucena and Bala samples. Three genes were found to be regulated in the opposite direction in the two plant variants: permease 1 gene was upregulated in Bala while downregulated in Azucena, and Bowman–Birk-type bran trypsin inhibitor precursor and cytochrome P450 CYP99A1 were both downregulated in Bala while upregulated in Azucena in response to As(V). Gene ontology categories were assigned to probe sets using WEGO tool (Ye et al. 2006). Upregulated genes comprised of genes encoding proteins involved in heat and toxin responses, the toxins biotransformation, sulfur, amines, organic acids, macromolecules and cellular catabolism as well as nitrogen biosynthesis categories. Downregulated genes were involved in transport; regulation of metabolism and cell size; metabolism of phosphorus, phenylpropanoids, and aromatic compounds; cellular morphogenesis; cell growth; and responses to auxin. In the transporter category, the phosphate: H+ symporter gene OsPT2 and an inorganic phosphate transporter were dramatically downregulated, as well as the transporters of chloride, ammonium, nitrate, sugars, amino acids, and peptides. Five sulfate transporter genes were upregulated under As(V) stress. Seven MATE transporter genes were upregulated, while two other MATE genes were repressed. A glutathione conjugate transporter gene, a gene annotated as a multidrug resistance-associated protein MRP2, and other possible ATP-binding cassette (ABC) family vacuole pumps were upregulated. The expression of a number of metal transporter genes was differentially regulated by As(V) treatment, including two Nramp1 genes, two potassium transporter genes, and a ZIP zinc/iron transporter gene. Genes encoding TIP and NIP types of aquaporins were found to be downregulated, while five major facilitator superfamily protein genes were upregulated. Glutathione-S-transferases-encoding genes, the majority of them belonging to tau subfamily, were differentially expressed in As-exposed rice. Some of them were up to 30-fold upregulated compared to control plants. Three methyltransferase-encoding genes were found to increase their expression upon As stress, while some heat shock proteins and chaperones exhibited differential expression in these conditions. Gene-encoding peroxidases were mostly repressed, while cytochrome P450 genes were differentially expressed with 8 being upregulated and 12 genes being downregulated. Expression of genes encoding proteins involved in cell growth and cell cycle was downregulated by As(V), including expansins, tubulin, actin, and microtubule genes.

The effect of As(V) exposure on expression of early response genes in rice roots was studied by DNA microarray (Huang et al. 2012). 1690 genes were found to be upregulated after 3 h of treatment, and 698 were found to be downregulated. Gene ontology annotations showed that the most upregulated biological processes were response to heat, regulation of transcription, toxin catabolic process, secondary metabolic process, cellular lipid metabolism, and jasmonic acid- and ethylene-dependent systemic resistance. Among the most upregulated molecular functions in GO terms were glutathione transferase activity, transcription factor activity, ion binding, calcium ion binding, and oxidoreductase activity. Secondary cell wall metabolism was identified as a downregulated biological process. Sixty-six transporter-related genes were upregulated in As-treated cells, including ABC transporters and tellurite resistance/dicarboxylate transporters. Two phosphate transporters OsPT4 and OsPT19 were upregulated, whereas citrate transporter OsLsi2, aquaporin OsNIP2;1, and two sulfate transporter genes were repressed during early response to As(V) exposure. Transcripts of genes involved with oxidative stress response were identified among As-regulated genes. Among them genes encoding GSTs, glutaredoxins, alternative oxidases, monodehydroascorbate reductases, thioredoxins, peroxiredoxin, and respiratory burst oxidase homolog were upregulated, while class III peroxidase genes were differentially regulated. In the group of phytohormone-related genes, jasmonic acid, abscisic acid, cytokinin, and ethylene biosynthesis and signaling pathways showed significant upregulation. A number of signal transduction-associated genes were identified as As responsive. Ninety-five genes showed increase and 39 genes repression of expression. Receptor-like kinases were the major upregulated group. One MAPK and seven MAPKKK genes were also found to be upregulated under As stress. Among genes involved in calcium regulation calcium-dependent protein kinases, calmodulins and calmodulin-like proteins were upregulated. Two hundred thirty-one transcription factor genes were identified as regulated by As(V). They belong to several TF families, such as AP2/ERF (APET-ALA2/ethylene response factors), HSF (heat shock factors), ZIM (zinc finger proteins expressed in meristem), and MYB and WRKY. Cell wall metabolism was found to be the biological process downregulated at the transcript level in the early response of rice roots treated with As. Twenty-seven genes involved in cell wall biogenesis were repressed, while 13 were upregulated. Downregulated genes included genes encoding cellulose and beta-mannose synthase-like proteins, xyloglucan galactosyltransferases, xyloglucan xylosyltransferases, galactomannan galactosyltransferases, xyloglucan hydrolases, glycosyl transferases, beta-galactosidases, glycoside hydrolases 9, and polygalacturonases. This work showed that As stress provokes dramatic changes in cell transcriptome in a very short time. Such a profound early response suggests significant toxicity of this metalloid toward rice root cells.

Another study utilized RNA sequencing technology to examine the response in roots and shoots of 14-day-old rice seedlings exposed to 20 or 80 μM As(III) for 6 or 24 h (Yu et al. 2012). O. sativa ssp. japonica cultivar Nipponbare was used as experimental material (Table 14.1). Authors have found that 7865 genes were differentially regulated upon As challenge in either time-dependent or dosage-dependent manner. It was also shown that roots responded to As dose, while shoots responded to the time of treatment in a more profound manner. It can be explained in a way that roots are the organs where the intake of metalloid takes place, and so As levels in root cells increase relatively quickly even when treated with lower concentrations. Shoot cells, on the other hand, need As to be transported in xylem, so As levels in shoot cells build up gradually with time. Genes involved in transport, phytohormone biosynthesis, and signaling and lipid metabolism along with transcription factors were found to be differentially regulated by As(III) in this study. In root samples, 27 genes were found to be differentially regulated during treatment with lower As(III) concentration, while 72 genes showed altered transcription when treated with high concentration of As(III). Among them a number of ABC transporter family G genes, P-type ATPase genes, phosphate transporter gene, metal transporter gene OsZIP8, nodulin 26-like intrinsic membrane proteins (NIPs: OsNIP3;2 and OsNIP1;1)-encoding genes, and genes encoding P-type heavy metal ATPases (HMAs) – OsHMA5, a xylem loading protein, and OsHMA9 which is a metal efflux protein – were all upregulated. Some of the aquaporin genes were downregulated in As(III) stress, while citrate transporter genes were either not responsive to this condition or also downregulated. A number of genes involved in jasmonate biosynthesis were upregulated in As(III)-stressed roots, indicating that this phytohormone is accumulated and plays a role in plant response to the heavy metal challenge, as was proposed before (Maksymiec et al. 2005; Maksymiec 2007). Possible accumulation of auxins, cytokinins, and ethylene, but not brassinosteroids, was also suggested by the analysis. Forty-two genes encoding proteins involved in lipid metabolism, including both prokaryotic and eukaryotic lipid biosynthesis pathway genes, were found to be differentially regulated upon exposure to As(III) in roots and 32 genes in shoots. Expression analysis of transcription factor (TF) genes revealed that high concentration of As(III) induces expression of more TFs than low As(III) concentration and that more genes are upregulated after 6 h of treatment than after 24 h. Downregulation of TF genes was also found to depend on As(III) dose. NAC and WRKY families of TFs are likely to be responsible for regulation of transcriptional response to As(III) challenge in rice roots. Very interestingly, this study also presented a short analysis of regulator miRNA molecules differentially expressed upon As(III) challenge. They found 13 upregulated and 12 downregulated miRNAs in roots and 20 upregulated and 10 downregulated miRNAs in rice shoots. The analysis of the biological significance of possible miRNA-mRNA pairs indicated that there are 237 such pairs with possible significance in roots and 128 pairs in shoots, including transporter mRNA-miRNA pairs, lipid metabolism involved mRNA-miRNA pairs and jasmonate metabolism related mRNA-miRNA pairs. These results suggest that described changes in expression of certain genes might be regulated by coordinated actions of both transcription factors and miRNAs.

14.3 Transcriptomics of Arsenic Tolerance in Selected Dicot Plants

14.3.1 Arabidopsis thaliana

A massive amount of genomic and metabolomic data available for A. thaliana makes this species a perfect candidate to study the general metabolism of As in plant cells, although A. thaliana is neither economically nor industrially important plant. When whole-genome oligonucleotide DNA microarray technology was used to study changes in transcript levels of A. thaliana ecotype Columbia exposed to 100 μM potassium arsenate in in vitro cultures (Table 14.1), 46 genes were found to be upregulated in response to As, while 113 genes were deemed downregulated after applying stringent statistical criteria (Abercrombie et al. 2008). Gene ontology (GO) terms were assigned to identify genes using GO annotation bioinformatics tool available at the Arabidopsis Information Resource web. Most upregulated categories included unknown function, hydrolase, antioxidant, transferase, kinase, lyase, transporter, and binding activity. Downregulated gene ontology categories included unknown function, hydrolase, binding, transporter, kinase, transferase, and transcriptional regulator activity.

Among upregulated genes encoding proteins involved in antioxidant response were copper superoxide dismutases (CuSOD), peroxidases, and peroxiredoxin Q. Upregulated were also genes encoding tau class GST, transporter proteins (plasma membrane intrinsic protein 2B and tonoplast intrinsic protein gamma), metal ion-binding protein (metallothionein-like protein IA and ferredoxin), 5′-adenylylsulfate reductase (APR3), nitrate reductase I (NR I), leucine-rich repeat transmembrane protein kinase and cyclin-dependent protein kinase, cell wall-localized glycine-rich protein, cytochrome b 561 family protein, and a number of unknown molecular function proteins (e.g., universal stress protein, pentatricopeptide repeat-containing protein, photoassimilate-responsive protein). Other transcript levels were found to be downregulated in As(V) challenge. Among them were genes encoding catalase 3, some peroxidases, FeSOD, lipoxygenase, cytochrome P450, two germin-like proteins, calcium-binding EF-hand family protein, calmodulin-related protein, ferritins 1 and 4, zinc finger proteins, glycosyl hydrolases, xyloglucan endotransglucosylases/hydrolases, lipase class 3 family protein, invertase/pectin methylesterase family protein, and others. Interestingly expression of allene oxide cyclase involved in jasmonic acid synthesis was also downregulated. A number of transcription factors were repressed during As exposure, among them three WRKY DNA-binding protein family members, two zinc-finger proteins, a NAC (NAM, ATAF, CUC) domain-containing protein, DRE (dehydration-responsive element)-binding protein, and an AP2 (APETALA2) domain-containing transcription factor. This study clearly exhibited that arsenate treatment causes downregulation of expression of genes involved in phosphate starvation, since As may utilize phosphate transporters to get into plant cells.

14.3.2 Medicago truncatula

DNA microarray technology was also used to examine As(III) effect on transcript abundance in the roots of M. truncatula (Gaertn.) (cultivar Jemalong) in control conditions or inoculated with nitrogen-fixing bacterium Ensifer medicae MA11 (Sinorhizobium medicae) (Lafuente et al. 2015). Arsenic was provided to pregerminated seedlings in the form of 25 μM sodium arsenite; the exposure continued for 5 days (Table 14.1). Compared to control conditions, non-inoculated roots challenged with As(III) exhibited upregulated expression of 263 genes, while the expression of 528 genes was repressed. Among the upregulated genes were abiotic stress-related genes, such as ATP5a (mitochondrial ATP synthase subunit alpha), glutathione-S-transferase, germin-like protein, and 1-pyrroline-5-carboxylate synthase (P5CS) genes. Genes encoding transporter proteins including ABC transporter B family member, sulfate high-affinity transporter, root-specific metal transporters, and phosphate transporter 2 were also found to be overexpressed in As(III)-treated roots. Genes related to sugar metabolism, such as glucosyltransferase-13 and probable mannitol dehydrogenase genes, as well as secondary metabolism genes, like terpenoid synthase gene, isoflavone-7-O-methyltransferase gene, glucosyltransferase gene, O-diphenol-O-methyltransferase gene, and naringenin chalcone synthase gene, were upregulated, suggesting that the general biotic and abiotic stress response mechanism is activated in root cells during exposure to As (Vogt 2010). Downregulated genes included genes involved in As uptake, namely, the gene encoding silicon aquaporin at the plasma membrane NIP2-1, and cell wall biosynthesis- and architecture-related genes, such as pectin-esterase inhibitor genes, genes encoding extensins, and cell wall-specific peroxidases. The gene encoding a small subunit of Rubisco was also found to be repressed in treated root cells. Interestingly, inoculation of Medicago roots with nodule-forming bacteria caused a significant decrease in the number of upregulated stress response genes and flavonoid biosynthesis-related genes. It has also reduced the downregulation of cell wall architecture genes, photosynthesis-related genes, and genes involved in carbohydrate metabolism. Altogether the results of this study indicate that inoculation of M. truncatula with E. medicae had a mitigating effect on the stress imposed by the presence of As in growth medium. It is likely that microorganisms may be able to deal with at least a part of the stressor alleviating its effect on plant cells.

14.3.3 Salix purpurea

Hydroponically cultured purple willow (S. purpurea, cultivar Fish Creek) was examined in another RNA sequencing study (Yanitch et al. 2017). Stem cuttings of normalized length were cultivated for 2 weeks in Hoagland medium before the addition of 5 mg/L sodium heptahydrate arsenate, after which the treatment continued for another 2 weeks (Table 14.1). The transcriptomic analysis was conducted separately for leaf, stem, and root samples. Differential expression of genes was found not only between control and treated plants but also between different plant organs. Genes were assigned gene ontology terms using a PANTHER tool (Thomas et al. 2006; Mi et al. 2013, 2017). It revealed that in root cells most upregulated gene categories were catalytic activity, transferase activity, biological regulation, carbohydrate metabolism, oxidoreductase activity, cell cycle, and peptidase activity. Most repressed gene categories in roots were protein metabolic processes, RNA-binding protein, translation, structural molecule activity, and ribosomal activity. Root cells can deal with As presence in the environment by decreasing its uptake via the decrease of the number of transporters and aquaporins capable of capturing the As ions and/or by activating intracellular detoxification mechanisms, such as efflux, conjugation, and storage in the vacuole. The transcription of a gene encoding phosphate transporter PHO1 and three genes encoding aquaporin NIP1;1 was upregulated, while the transcription of a gene encoding aquaporin TIP2 was repressed. A number of ABC transporter transcripts were found to be differentially regulated in As(V) challenge with 8 transcripts exhibiting increased abundance and 19 transcripts showing reduced abundance, compared to the control samples. Two vacuolar cation/proton exchanger 2 protein (CAX2) transcripts were upregulated. Glutathione synthase, GST, and phytochelatin synthetase transcripts were found among the ones increased by the treatment. Seven transcripts (from five genes) of S-adenosyl methionine-dependent methyltransferase (SAM-dependent methyltransferase) were upregulated, while six other transcripts (from three genes) were downregulated in this experiment.

Among 15 most abundant gene ontology terms in stem cells, only ubiquitin-protein ligase and steroid metabolism were upregulated, while binding, kinase activity, protein phosphorylation, translation, protein kinase receptor, ribosomal protein, structural constituents of the ribosome, and cellular amino acid catabolic process were all repressed. Upregulated transcripts included a silicon transporter and CAX2 transporter, catalytic subunit 9 of cellulose synthase A, sucrose-phosphate synthase, and salicylic acid carboxyl methyltransferase. Downregulation of expression of 2 other CAX exchangers, 31 ribosomal proteins, 2 cellulose synthase catalytic subunits and a cellulose synthase-like protein, callose synthase, and proteins related to ethylene biosynthesis was reported.

In leaves most upregulated gene ontology categories were hydrolase activity, transporter activity, lipid metabolic process, biological regulation, transcription factor, regulation of nucleobase-containing compound metabolic process, and cellular amino acid biosynthesis. Downregulated groups of genes included RNA-binding protein, translation, structural molecule activity, and ribosomal protein. A number of transporter genes were found to be differentially regulated in leaves. PHO1 phosphate transporter; Na-dependent phosphate transporter; boron transporter; aquaporins NIP, NIP1;1, TIP1, and SIP1; three CAX2 vacuolar exchangers; and 20 ABC transporters were all upregulated, whereas PHO1-like transporter, one aquaporin of NIP subfamily, two CAX1 proteins, and four ABC transporters were downregulated. Cadmium-induced protein and cadmium resistance protein were also upregulated. Unlike in root cells, phytochelatin synthase transcripts exhibited decreased abundance in As-treated leaves. Very interesting is the differential expression of many genes involved in flavonoid biosynthesis in As-treated willow leaves. Transcripts of genes encoding chorismate mutase, cinnamate-4-hydroxylase, 4-coumarate: CoA ligase, chalcone synthase, chalcone-flavanone isomerase, flavanone-3-hydroxylase, flavonoid-3′,5′-hydroxylase, flavonoid-3′-hydroxylase, dihydroflavonol-4-reductase, anthocyanidin reductase ANR1–1, anthocyanidin synthase, and a leucoanthocyanidin reductase were all upregulated, while abundance of transcripts of chalcone-flavanone isomerase, dihydroflavonal-4-reductase, and flavonol synthase were downregulated. Such a massive upregulation of secondary metabolite biosynthesis in As-stressed leaves may indicate increased oxidative stress in these organs or induction of a general stress response mechanism.

14.3.4 Brassica juncea

A DNA microarray study was also conducted on a plant commonly used in soil phytoremediation to remove heavy metals, such as lead or cadmium, i.e., Indian mustard (B. juncea L.). Exposure to As(V) (Table 14.1) highlighted the importance of hormones and kinases in As-dependent signaling of this species (Srivastava et al. 2015). Differential expression of many transporter-encoding genes was additionally reported, namely, genes encoding major intrinsic protein family members (NIP2;1, TIP2, PIP1;2, PIP1;4, PIP2;1 PIP2;2), ABC transporter family proteins (ABCB4, ABCC4, ABCF4, ABCG27, and ABCG32), and mitochondrial transporters (phosphate transporter PHT3;2, dicarboxylate transporter 1, and dicarboxylate carrier 1). Metabolism of several phytohormones was found to be altered during As challenge. Jasmonic acid-related genes were differentially expressed. Allene oxide cyclase 4 and two jasmonate-zim-domain proteins (JAZ1 and JAZ5) were downregulated in roots, 12-oxophytodienoate reductases were upregulated in both roots and shoots, and jasmonate resistance one gene was downregulated in shoots. Genes involved in abscisic acid signaling were induced (ABA-induced PP2C1, ABA-insensitive 1, and ABA-interacting protein 2). Auxin-related differentially regulated genes included those of auxin-responsive proteins from shoots and roots and were upregulated. Two auxin efflux carrier proteins (PIN3 and PIN6) were downregulated in shoots. Farnesoic acid carboxyl-O-methyltransferase (FAMT), a gene from salicylic acid biosynthetic pathway, was upregulated in both roots and shoots. A large number of transcription factors representing various families were found to be differentially regulated by As. A total of 12 redox-related genes were described, of which seven were upregulated and five were downregulated, including monothiol glutaredoxin 17, glutathione peroxidase 6, monodehydroascorbate reductase, and copper/zinc superoxide dismutase 1 in roots and glutathione peroxidase 3 and iron superoxide dismutase 2 in shoots. Genes encoding proteins comprising mitochondrial electron transport chain showed significant upregulation in both roots and shoots, indicating that alteration of energy requirements imposed by cell need to respond to the presence of As puts electron transfer chain under significant stress.

14.3.5 Crambe abyssinica

The last study discussed here was conducted on C. abyssinica (cv. BelAnn), an oilseed-producing plant that was shown to be a heavy metal accumulator (Paulose et al. 2010). In this experiment 10-day-old crambe seedlings were exposed to 250 μM sodium arsenate. At this As(V) concentration plants showed a significant decrease in biomass, as compared to control plants, but no severe toxicity symptoms. After 24 h of treatment, plants were harvested, frozen, and used for further experimental steps (Table 14.1). One hundred five transcript clones were obtained that represented 38 unique coding transcript sequences. Identified proteins encoded by differentially regulated genes indicated As(V) effect on metabolic pathways related to oxidative stress, defense, ion transport, sulfur assimilation, signal transduction, photosynthesis, and metabolism.

Glutathione-S-transferase transcripts comprised the largest group of differentially expressed genes identified in crambe cells. Tau and phi GST subfamily members were present in studied samples. Additionally, transcripts of other genes encoding proteins that may work together with GST in response to As(V) challenge were found to be differentially regulated, namely, monodehydroascorbate reductase (MDAR), adenosine phosphosulfate kinase (APSK) and adenosine phosphosulfurylase reductase (APR), and sulfite reductase (SiR). Expression of ABC transporter proteins was also found to be differentially regulated including multidrug-resistant proteins (MRPs) and yeast cadmium factor1 (YCF1). Other membrane transporter gene transcripts identified in SSH experiment were MATE family drug transporter and putative cation transporter-associated protein (ChaC). Transcripts of genes encoding proteins involved in oxidative stress response aldo/keto reductase (AKR) and peptide methionine sulfoxide reductase (PMSR) were found to have altered expression in As(V)-treated plants, as well as a gene encoding oxophytodienoate reductase (OPR) involved in jasmonate synthesis and serine palmitoyl transferase (SPT) implicated in sphingolipid biosynthesis. Three proteins connected to ubiquitin-mediated protein degradation pathway were deemed differentially regulated between control and study sample, namely, 20S proteasome beta subunit, ubiquitin 14 (UBQ14), and an ubiquitin-associated (UBA)/TS-N domain-containing protein. Other identified transcripts encoded glucosidases, heat shock proteins, defense-related protein, pathogenesis-related protein, iron ion-binding oxidoreductase, and a number of proteins of unknown function.

14.4 Plant Response Mechanisms as Seen by Transcriptomic Analyses

Transcriptomic attempts to unravel the mechanism of plant resistance in the presence of As species are presented in this chapter. Plants used as model organisms in these studies are classified in different families (Poaceae, Brassicaceae, Salicaceae, Fabaceae). Some of them, like rice and crambe, are effective As accumulators, whereas others (i.e., A. thaliana, willow, Indian mustard, and Medicago) do not accumulate this metalloid. Despite the differences, there is a rough pattern of gene expression in examined plants that indicate the involvement of a general biotic and abiotic stress response mechanism (Fig. 14.1). Similar mechanisms are engaged when plants are exposed to other nonessential metals, excessive amounts of essential metals, or xenobiotics.

Fig. 14.1
figure 1

Schematic representation of cell processes differentially regulated by arsenic challenge. Particular proteins encoded by up- or downregulated genes are listed when feasible

Full size image

14.4.1 Arsenic Transport

Arsenic enters the protoplasm through roots, and these organs are the first to activate their protection arsenal, as shown by Yu et al. (2012). As(III) and As(V) differ in the influx mechanism. As(III) is most probably taken up by silicon transporters, while As(V) utilizes the phosphate route (Ma et al. 2006, 2007, 2008). To prevent As uptake, root cells can downregulate the expression of plasma membrane transporters capable of introducing As into the cytoplasm (Hartley-Whitaker et al. 2001) and upregulate different types of efflux pumps (Lee et al. 2007; Kim et al. 2007; Yu et al. 2012). The expression of transporter-encoding genes was found to be differentially regulated by As in all of the presented transcriptomic studies. Downregulation of NIP2;1, similar to Lsi1 rice transporter, was reported by Lafuente and Huang (Huang et al. 2012; Lafuente et al. 2015), while the same transporter was found to be upregulated during As treatment by Srivastava (Srivastava et al. 2015). Differential regulation of many aquaporin-encoding genes was shown (Yu et al. 2012; Huang et al. 2012; Srivastava et al. 2015; Yanitch et al. 2017) as well as upregulation of MATE extrusion pumps was found (Norton et al. 2008b; Paulose et al. 2010; Srivastava et al. 2015). Another group of transporters transcriptionally regulated by As is the ABC transporters that may be involved in metal export or vacuolar sequestration (Kim et al. 2007). ABC transporter involvement in plant response to As stress was described on transcriptional level (Norton et al. 2008a; Yu et al. 2012; Lafuente et al. 2015; Srivastava et al. 2015; Yanitch et al. 2017). From the root As can be transported to plant aboveground organs (root-to-shoot transport) (Deng et al. 2013). One of the proteins responsible for xylem loading, OsHMA5, was found to be upregulated in rice (Yu et al. 2012).

14.4.2 Detoxification in the Cell

Once inside the cell, the majority of As(V) is reduced to As(III) by As reductase (AR) (Bleeker et al. 2006; Dhankher et al. 2006). Then As(III) may undergo conjugation with glutathione (GSH) and phytochelatins (PCs) (Grill et al. 1987; Sneller et al. 1999; Mendoza-Cózatl et al. 2005). Such conjugates are sequestered in the vacuole, as a final step of detoxification. GSH is synthetized in an ATP-dependent process by γ-glutamylcysteine synthetase and glutathione synthetase that were found to be differentially regulated by As (Huang et al. 2012; Yanitch et al. 2017). GSH is the cell largest reservoir of nonprotein thiol groups (Mendoza-Cózatl et al. 2005); thus its synthesis strongly relies on sulfur assimilation. Genes encoding proteins involved in sulfur metabolism were reported to be upregulated (Norton et al. 2008a; Paulose et al. 2010; Lafuente et al. 2015; Srivastava et al. 2015). Phytochelatins (PCs) are small cysteine-rich peptides synthetized by phytochelatin synthase from GSH, which are capable of metal chelation (Grill et al. 1985; Grill et al. 1989; Cobbett 2000; Schmöger et al. 2000). Synthase-encoding genes were found to be differentially transcribed in the presence of As (Huang et al. 2012; Yanitch et al. 2017). Both phytochelatin synthase and vacuolar transporters were shown to be upregulated to increase sequestration of metal ions, thereby removing it from the cytoplasm (Grill et al. 1987; Song et al. 2010; Tripathi et al. 2007). Another group of peptides that can bind metal ions are metallothioneins (MTs), relatively small cysteine-rich molecules (Hassinen et al. 2011). Their expression was shown to be correlated with metal accumulation in plant tissues (Hassinen et al. 2009; Zimeri et al. 2005) and with As treatment (Yu et al. 2012; Chakrabarty et al. 2009). Glutathione-S-transferases are a diverse family of enzymes that exhibit As-dependent expression in each of described experimental studies (Abercrombie et al. 2008; Norton et al. 2008a; Chakrabarty et al. 2009; Paulose et al. 2010; Yu et al. 2012; Huang et al. 2012; Lafuente et al. 2015; Yanitch et al. 2017); moreover in several studies, GST-encoding genes constituted the biggest group of detected transcripts. This protein family is divided into eight classes of which tau and phi are involved in detoxification of xenobiotics and are usually stress-responsive (Marrs 1996; Dixon et al. 2002; Dixon and Edwards 2015). One possible mechanism of action of GSTs in the presence of As is performing conjugation of the metalloid with glutathione (Pandey et al. 2015). Another way plant may use GSTs is to mitigate oxidative stress caused by As (Schutzendubel and Polle 2002). It was shown that metals can induce oxidative stress (Gallego et al. 1996; Hartley-Whitaker et al. 2001; Pinto et al. 2003) and that GSTs take part in cell response to oxidative stress (Cummins et al. 1999; Kilili et al. 2004). Transgenic plants overexpressing GSTs are more resilient to oxidative stress conditions (Roxas et al. 2000; Zhao and Zhang 2006; Ji et al. 2010); moreover tau class GST from tomato expressed in yeast conferred yeast resistance to oxidative damage (Kampranis et al. 2000). GSTs may also be responsible for the transfer of phytochemicals between cell compartments (Edwards et al. 2000). Arsenic methylates are found in some plant species, but it is controversial whether plants can methylate As themselves or if they take it up from soil microorganisms (Lomax et al. 2012). If plants were performing the methylation reaction, then a good candidate enzyme would be SAM-dependent methyltransferases, which expression was shown to be upregulated in the presence of As (Norton et al. 2008a; Srivastava et al. 2015; Yanitch et al. 2017). It was recently shown, however, that rice was only able to methylate As after transformation with fungal WaarsM methyltransferase gene from Westerdykella aurantiaca (Verma et al. 2018).

14.4.3 Oxidative Stress Response

Nonessential metals, such as As, present cells with oxidative stress (Singh et al. 2006). A number of genes encoding proteins typically involved in oxidative stress response were found to be differentially regulated by As in all examined plant species. For example, genes encoding CuSOD and ZnCuSOD were found to be upregulated by As challenge (Abercrombie et al. 2008; Srivastava et al. 2015), while genes encoding FeSOD were found to be upregulated in shoots of B. juncea (Srivastava et al. 2015) and downregulated in A. thaliana (Abercrombie et al. 2008) in similar conditions. It should be noted, however, that B. juncea experiment was carried on root and shoot material separately, whereas A. thaliana samples were prepared from whole plants (Table 14.1). As it was shown, plant organs respond with specific expression patterns and temporal regulation to As treatment (Yanitch et al. 2017). Germin-like proteins were also found to be differentially regulated in the presence of As (Abercrombie et al. 2008; Chakrabarty et al. 2009; Lafuente et al. 2015). These are metal-binding glycoproteins associated with extracellular matrix that usually have oxalate oxidase activity, often supplemented by SOD or phosphodiesterase activity (Bernier and Berna 2001; Nakata et al. 2004; Lu et al. 2010). They are known to accumulate in response to variety of stress conditions, such as bacterial, viral, fungal infections, parasite attacks, xenobiotic and chemical toxicity, and salt and drought stress (Hurkman et al. 1991; Hurkman et al. 1994; Schweizer et al. 1999; Lane 2002; Lou and Baldwin 2006; Zimmermann et al. 2006; Manosalva et al. 2009; Wang et al. 2013). Peroxidases, hydrogen peroxide scavenging enzymes, comprise another group of proteins differentially regulated by As treatment (Asada 1992; Yoshida et al. 2003). Expression regulation pattern seems to be very complex depending on plant species, organ examined, As dosage, and time of treatment (Abercrombie et al. 2008; Norton et al. 2008a; Chakrabarty et al. 2009; Huang et al. 2012; Lafuente et al. 2015; Srivastava et al. 2015). It might be connected with the fact that these enzymes are involved in a variety of cell processes including reactive oxygen species scavenging but also defense, auxin signaling, and cell wall metabolism (Kawano 2003; Passardi et al. 2004; Correa-Aragunde et al. 2015). Despite the complicity of plant response, it is clear that excess levels of As promote differential regulation of oxidative stress-responsive genes.

14.4.4 Hormonal Regulation

Hormonal regulation is crucial for plant ability to acclimate to variable environmental conditions (Peleg and Blumwald 2011) such as nonessential metal exposure, as indicated by transcriptomic analyses. Differential regulation of expression of genes involved in jasmonate (JA) metabolism was reported in several studies (Abercrombie et al. 2008; Chakrabarty et al. 2009; Paulose et al. 2010; Yu et al. 2012; Huang et al. 2012; Srivastava et al. 2015). JA and methyl-JA are active metabolites with roles in cell wall metabolism, defense against herbivore attack and other biotic and abiotic stress factors, and induction of phytoalexin production (Rakwal et al. 1996; Tamogami et al. 1997; McConn et al. 1997; Thomma et al. 1998). It was suggested that JA might act as a regulator of sulfate assimilation pathways in order to enhance As complexation by phytochelatins (Srivastava et al. 2009). It was also shown that JA activates GSH biosynthesis genes in A. thaliana (Xiang and Oliver 1998). JA is known to act together with ethylene in the induction of plant defense mechanisms (Penninckx et al. 1998; Lorenzo et al. 2003). Genes encoding proteins involved in ethylene signaling were also found to be among the ones regulated by As. Downregulation of such genes was described in willow (Yanitch et al. 2017), whereas in rice they were found to be upregulated (Yu et al. 2012) or differentially regulated (Huang et al. 2012). The increase in expression of an ethylene-responsive transcription factor in rice upon exposure to As(V) was also reported (Chakrabarty et al. 2009; Huang et al. 2012). Salicylic acid (SA) is another phytohormone which seems to play an important role in plant response to As(V) challenge (Chakrabarty et al. 2009; Srivastava et al. 2015). It is not surprising given that it is known as a regulator of ion transport (Raskin 1992; Hayat et al. 2007). Transcriptome analyses showed that genes related to metabolism and physiological actions of abscisic acid, auxins, brassinosteroids, and cytokinins were also regulated by As. It is clear that the response to high As levels is coordinated at an organism level, not just a cell level.

14.4.5 Lipid Metabolism

There are several reports indicating As-dependent regulation of the expression of genes involved in lipid metabolism (Abercrombie et al. 2008; Paulose et al. 2010; Huang et al. 2012; Yu et al. 2012; Srivastava et al. 2015; Lafuente et al. 2015). However, not all transcriptomic studies were able to detect these alterations. Gene encoding serine palmitoyl transferase, an enzyme from sphingolipid biosynthetic pathway, was shown to be upregulated by As(V) in C. abyssinica. In A. thaliana repression of monogalactosyldiacylglycerol type 3 synthase was reported under As(V) stress. As(III) seems to have a stronger impact on lipid metabolism in plants. In M. truncatula two lipid-related genes were found to be regulated under As(III), namely, a non-specific lipid transfer protein was upregulated, while a gene involved in lipid metabolism during nodulation, MtEnod8.1, was repressed. Using RNA-seq technology, 59 lipid biosynthesis and metabolism genes differentially regulated by As(III) were identified in rice (Yu et al. 2012). The pattern of up- and downregulation remains quite complex, but the importance of adjustments of biological membrane composition upon As(III) challenge seems clear.

14.4.6 Transcription Factors

Transcription factors (TFs) comprise a significant number of differentially regulated genes identified by transcriptomic methods. It does not come as a surprise considering the number of genes involved in the systemic response to As exposure. General need to rearrange expression patterns during stress creates demand for the action of TFs. TF families bHLH (basic helix-loop-helix), BZIP (Basic Leucine Zipper Domain), MYB (myeloblast DNA-binding domain), WRKY (Wrky DNA-binding proteins), RAV (related to ABI3/VP1), ERN (ethylene-responsive), NAC (NAM, ATAF, and CUC), and WOX (WUS homeobox containing) were among the affected by As treatment.

14.4.7 Cell Wall Reorganization

The cell wall is able to bind certain metal cations present in the environment. For this purpose, it undergoes remodeling during challenge with metals and metalloids, including polysaccharide modifications (Krzesłowska 2011). Differential regulation of expression of genes related to cell wall and polysaccharide metabolisms, such as xyloglucan-related enzymes, pectin esterases, expansins, and glycine-rich proteins, was discovered by transcriptome analyses (Abercrombie et al. 2008; Chakrabarty et al. 2009; Huang et al. 2012; Lafuente et al. 2015). Some peroxidases, already mentioned as oxidative stress-responsive, are also involved in cell wall structural reorganization via decreasing the cross-linking of cell wall compounds (Passardi et al. 2005).

14.4.8 Interaction with Microorganisms

The most common resistance mechanisms that developed in bacteria in response to As challenge are As(V) to As(III) reduction and As efflux. These processes depend on the presence of ars genes on plasmid or chromosome (Kruger et al. 2013). Transcriptomic data from Rhizobium sp. NT-26 (Andres et al. 2013), Herminiimonas arsenicoxydans (Weiss et al. 2009; Cleiss-Arnold et al. 2010), and Geobacter strains (Dang et al. 2017) in the presence of As(III) and Enterobacteriaceae LSJC7 (Zhang et al. 2016) in the presence of As(V) demonstrated that this metalloid regulates expression of a number of genes involved in many metabolic processes. As it was previously reported, the interaction of plants with microorganisms in rhizosphere modulates metal toxicity (Fitz and Wenzel 2002; Wenzel et al. 2003). Transcriptomic data confirm this observation on the level of transcript accumulation (Lafuente et al. 2015). Inoculation of M. truncatula with Ensifer medicae led to decreased response to As in terms of gene expression modulation. When compared to non-inoculated plants, inoculated samples exhibited a reduced number of differentially regulated genes and suppressed susceptibility to As. It is supposed that microorganisms present in rhizosphere uptake and detoxify a pool of available metalloid, thus reducing the extent of adverse effect posed by the stressor.

14.5 Concluding Remarks

The picture of plant response to As on a transcriptome level seems to be the one of a systemic nature, involving a variety of mechanisms to fight metalloid itself, as well as the secondary effects of its presence in the cell (Fig. 14.1). This consists of changes in hormonal signaling, remodeling of transcription factor assemblage, activation of the oxidative stress response, and others. Transcriptomic analyses produce a huge amount of data of which only a fraction can be currently understood and appreciated. With the accumulation of such datasets, the need will grow for synthetic meta-analyses of co-regulated gene clusters but also for experimental verification of the significance of particular transcripts.