Analysis of gene expression and physiological responses in three Mexican maize landraces under drought stress and recovery irrigation

Abstract

Background: Drought is one of the major constraints for plant productivity worldwide. Different mechanisms of drought-tolerance have been reported for several plant species including maize. However, the differences in global gene expression between drought-tolerant and susceptible genotypes and their relationship to physiological adaptations to drought are largely unknown. The study of the differences in global gene expression between tolerant and susceptible genotypes could provide important information to design more efficient breeding programs to produce maize varieties better adapted to water limiting conditions.

Methodology/principal findings: Changes in physiological responses and gene expression patterns were studied under drought stress and recovery in three Mexican maize landraces which included two drought tolerant (Cajete criollo and Michoacán 21) and one susceptible (85-2) genotypes. Photosynthesis, stomatal conductance, soil and leaf water potentials were monitored throughout the experiment and microarray analysis was carried out on transcripts obtained at 10 and 17 days following application of stress and after recovery irrigation. The two tolerant genotypes show more drastic changes in global gene expression which correlate with different physiological mechanisms of adaptation to drought. Differences in the kinetics and number of up- and down-regulated genes were observed between the tolerant and susceptible maize genotypes, as well as differences between the two tolerant genotypes. Interestingly, the most dramatic differences between the tolerant and susceptible genotypes were observed during recovery irrigation, suggesting that the tolerant genotypes activate mechanisms that allow more efficient recovery after a severe drought.

Conclusions/significance: A correlation between levels of photosynthesis and transcription under stress was observed and differences in the number, type and expression levels of transcription factor families were also identified under drought and recovery between the three maize landraces. Gene expression analysis suggests that the drought tolerant landraces have a greater capacity to rapidly modulate more genes under drought and recovery in comparison to the susceptible landrace. Modulation of a greater number of differentially expressed genes of different TF gene families is an important characteristic of the tolerant genotypes. Finally, important differences were also noted between the tolerant landraces that underlie different mechanisms of achieving tolerance.

Figure 1. Physiological parameters of maize plants under drought stress and recovery irrigation.

(A) Soil water potential, (B) Leaf water potential, (C) Photosynthetic rate, (D) stomatal conductance. Data are from three measurements from different samples with standard error.

Figure 2. Sugar and proline content under 10 and 17 days of drought stress and recovery irrigation.

(A) Glucose content, (B) Myo-inositol content, (C) Proline content. Data are the means of three different samples with standard error.

Figure 3. Venn diagrams of up- and down-regulated transcripts under drought stress and on recovery irrigation.

(A) Differentially expressed genes at 10 days stress, (B) Differentially expressed genes at 17 days stress, (C) Differentially expressed genes at recovery irrigation. Number of genes with at least 2 fold change and FDR ≤0.5 are shown for each landrace identified by the name above the circle.

Figure 4. Overview of differentially expressed transcripts involved in different metabolic processes under stress and recovery irrigation.

(A) Genes at 17 days stress in 85-2, (B) Genes at 17 days stress in CC, (C) Genes at 17 days stress in M21, (D) Genes at RI in 85-2, (E) genes at RI in CC, (F) Genes at RI in M21. Gene transcripts that are induced or repressed are shown in red or green coloring respectively as shown in the color bar in each panel. The MapMan sotware was used to show the different functional categories involved. (CC: Cajete criollo, M21: Michoacán 21).

Figure 5. Differential expression of genes involved in photosynthesis under drought stress and at recovery irrigation.

(A) Genes differentially expressed at 17 days stress in 85-2, (B) Genes differentially expressed at 17 days stress in CC, (C) Genes differentially expressed at 17 days stress in M21, (D) Genes differentially expressed at RI in 85-2, (E) Genes differentially expressed at RI in CC, (F) Genes differentially expressed at RI in M21. Gene transcripts that are induced or repressed are shown in red or green colouring respectively as shown in the color bar in each panel. (CC: Cajete criollo, M21: Michoacán 21).

Figure 6. Functional classification of abiotic stress genes under 17 days of drought stress and recovery irrigation.

(A): Up-regulated genes under 17 days of drought stress, (B): Down-regulated genes under 17 days of drought stress, (C): Up-regulated genes under recovery irrigation, (D): Down-regulated genes under recovery irrigation.

Figure 7. General view of gene expression responses in the three maize landraces under drought stress.

Gene expression was monitored at 10 and 17 days stress and differences in gene expression levels were observed between the landraces. Transcripts encoding signal transduction, transcription factors, HSP, detoxification enzymes and aquaporins are shown. Gene identifiers correspond to the accession numbers from the corresponding databases as reported in Maize Oligonucleotide Array Annotation GAL files version 1.13 (http://www.maizearray.org/maize_annotation.shtml). Microarray data were visualized using the FiRe 2.2 Excel macro . A ≥2 fold change is shown in red, a fold change ≤0.5 in green and no change in black (FDR ≤0.05). Left column: 85-2, middle column: CC and the right column: M21. PS: photosystem, TPI: triosephosphate isomerase, FBPase: fructose-1,6-bisphosphatase, GAP: glyceraldehyde 3-phosphate dehydrogenase, Rib5PI: ribose 5-phosphate isomerase-related, Rubisco SU: ribulose bisphosphate carboxylase small subunit, PSII: photosystem II, MAP kinase: mitogen-activated protein kinase, AP2/EREBP: AP2/Ethylene-responsive element binding protein family, bHLH: Basic Helix-Loop-Helix family, C2C2: C2 domain-containing protein, similar to zinc finger and C2 domain protein, C2H2: C2 domain-containing protein, similar to zinc finger and C2 domain protein, C3H: zinc finger (CCCH-type) family protein, HB: Homeobox transcription factor, Putative DNA BP: putative DNA binding protein, zf-HD: zinc finger homeobox, APX & GLU: ascorbate peroxidase and glutathione related, TRX: thioredoxin, PRX: peroxiredoxin, PX: peroxidase, HSP17: 17 kDa class I heat shock protein, HSP70:heat shock protein 70, HSC70: heat shock cognate 70 kDa protein, HSP22: 22.0 kDa ER small heat shock protein, DNA J HSP: DNAJ heat shock protein, HSP18: 18.1 kDa class I heat shock protein, LEA: late embryogenesis abundant protein, NIP: NOD26-like membrane integral protein, PIP: Plasma membrane intrinsic protein, TIP: tonoplast intrinsic protein.

Figure 8. General view of gene expression in the three maize landraces on recovery.

Gene expression was monitored at 10 and 17 days stress and differences in gene expression levels were observed between the landraces. Transcripts encoding signal transduction, transcription factors, HSP, detoxification enzymes and aquaporins are shown. Gene identifiers correspond to the accession numbers from the corresponding databases as reported in Maize Oligonucleotide Array Annotation GAL files version 1.13 (http://www.maizearray.org/maize_annotation.shtml). Microarray data were visualized using the FiRe 2.2 Excel macro . A ≥2 fold change is shown in red, a fold change ≤0.5 in green and no change in black (FDR ≤0.05). Left column:85-2, middle column: CC and the right column: M21. TPI: triosephosphate isomerase, FBPase: fructose-1,6-bisphosphatase, GAP: glyceraldehyde 3-phosphate dehydrogenase, PGK: phosphoglycerate kinase, PRK: phosphoribulokinase, Rib5P Iso: ribose 5-phosphate isomerase-related, Rubisco SU: ribulose bisphosphate carboxylase small subunit, LHC-I: Light harvesting chlorophyll a/b binding protein of PSI, PSI: photosystem I, LHC-II: Light harvesting chlorophyll a/b binding protein of PSII, PSII: photosystem II, MAP kinase: mitogen-activated protein kinase, AP2/EREBP: AP2/Ethylene-responsive element binding protein family, bHLH: Basic Helix-Loop-Helix family, C2C2: C2 domain-containing protein, similar to zinc finger and C2 domain protein, C2H2: C2 domain-containing protein, similar to zinc finger and C2 domain protein, CCAAT: CCAAT-binding transcription factor, HB: Homeobox transcription factor, MIP:myo-inositol-1-phosphate synthase, Susy: sucrose-phosphate synthase, TPS: trehalose-6-phosphate synthase, SPP: sucrose-phosphatase, IP: nositol monophosphatase, TPS/TPP: trehalose-6-phosphate synthase/phosphatase, MO: myo-inositol monophosphatase, TPS: trehalose-6-phosphate synthase, ASC & GLU: ascorbate and glutathione related, GRX: glutaredoxins, PRX: peroxiredoxin, TRX: thioredoxin, HSC70: heat shock cognate 70 kDa protein, HSF7: heat shock factor protein 7, HSP70:heat shock protein 70, HSP83: heat shock protein 83, HSP17: 17 kDa class I heat shock protein, HSP25: 25.3 kDa small heat shock protein, LEA: late embryogenesis abundant protein, NIP: NOD26-like membrane integral protein, PIP: Plasma membrane intrinsic protein, TIP: tonoplast intrinsic protein.