Differential gene expression as an indicator of nitrogen sufficiency in field-grown potato plants

Plant Soil (2011) 345:387–400 DOI 10.1007/s11104-011-0793-z

REGULAR ARTICLE

Differential gene expression as an indicator of nitrogen sufficiency in field-grown potato plants Bernie J. Zebarth & Helen Tai & Sainan Luo & Pete Millard & David De Koeyer & Xiu-Qing Li & Xingyao Xiong

Received: 17 November 2010 / Accepted: 8 April 2011 / Published online: 27 April 2011 # Springer Science+Business Media B.V. 2011

Abstract Use of an in-season measure of crop N sufficiency to guide fertilizer management is one approach to match the supply of N to the crop N demand. This study examined use of gene expression in leaf tissue of field-grown potatoes for use in assessment of potato N sufficiency. Potato cultivar ‘Shepody’ was grown with six fertilizer N rates (0– 250 kg N ha–1). Leaf disks were collected weekly for quantification of the expression of N uptake/transport, N assimilation, and amino acid metabolism genes in leaf tissue by nCounter. Many of the genes evaluated were responsive to crop N supply, but the response Responsible Editor: A. C. Borstlap. B. J. Zebarth (*) : H. Tai : D. De Koeyer : X.-Q. Li Agriculture and Agri-Food Canada, Potato Research Centre, PO Box 20280, Fredericton, NB, Canada E3B 4Z7 e-mail: [email protected] S. Luo Institute of Horticulture, Hunan Academy of Agricultural Sciences, Changsha, Hunan, China 410025 X. Xiong College of Horticulture and Landscape, Hunan Agriculture University, Hunan( Changsha, China 410128 P. Millard Macaulay Institute, Craigiebuckler, Aberdeen, UK AB15 8QH

varied widely among sampling dates. The exception was an ammonium transporter gene (AT1) which was highly expressed, was relatively consistent across sampling dates, was closely related to root zone soil nitrate concentration across N rates and sampling dates, and was highly negatively correlated with total tuber yield. The level of expression of AT1 in leaf tissue was as good as or better than conventional chemical or optical measures of potato N sufficiency in the current study. Keywords Nitrate reductase . SPAD . Ammonium transporter . Petiole nitrate concentration . Solanum tuberosum

Introduction Fertilizer nitrogen (N) management plays a critical role in achieving both economic and environmental objectives in potato production. Insufficient N supply results in economic losses due to decreases in tuber yield and size (Millard and Marshall 1986; Zebarth and Rosen 2007). Excessive N supply can reduce tuber quality through decreased tuber specific gravity (Long et al. 2004), increased tuber nitrate concentration (Zebarth et al. 2004a) and increased accumulation of asparagine which can contribute to acrylamide formation during cooking (Lea et al. 2007). Environmental losses of N through nitrate leaching (van Es et al. 2002) and nitrous oxide emissions (Zebarth et al.

388

2008) increase rapidly as fertilizer N rate is increased above the optimum rate. One of the most effective ways to meet both economic and environmental objectives is to match the supply of N to the crop N demand (Zebarth et al. 2009). While the concept is simple, selection of the optimal fertilizer N rate is difficult due to significant variation in both soil N supply and crop N demand among fields and among years (Olfs et al. 2005; Scharf et al. 2005). This has led to dynamic approaches where the in-season management of fertilizer N is modified based on measures of crop N sufficiency (Goffart and Olivier 2004; Goffart et al. 2008). A number of chemical and optical means of assessing crop N sufficiency have been evaluated (Olfs et al. 2005). Petiole nitrate concentration (Porter and Sisson 1991; Westcott et al. 1991) and relative leaf chlorophyll concentration as measured by light transmittance using a SPAD-502 meter (Vos and Bom 1993; Minotti et al. 1994; Goffart et al. 2008) are the most commonly used chemical and optical measures of potato N sufficiency. These methods are generally effective, but provide indirect measures of crop N sufficiency. Another disadvantage is that they may be influenced by the availability of water or other nutrients by soil and environmental conditions (Olfs et al. 2005) and stage of crop development (Millard and Mackerron 1986). One possible alternative is development of a gene expression based approach to assess crop N sufficiency. Plant responses to abiotic stresses are mediated through changes in gene expression (Hazen et al. 2003). Thus, quantification of gene expression may provide a more direct measure of plant N sufficiency. Stress-specific plant gene expression profiles have been identified in response to single and combined abiotic stresses and nutrient deficits (Hazen et al. 2003; Amtmann et al. 2005; Bohnert et al. 2006; Swindell 2006; Schachtman and Shin 2007). It may, therefore, be possible to use this approach to identify and distinguish among multiple abiotic stresses using a gene expression based approach. For example, Tamaoki et al. (2004) used stress-specific gene expression profiles to develop a stress indicator test using a cDNA macroarray with Arabidopsis thaliana genes that distinguished among ozone, drought and wound stresses. Short term gene expression in response to withdrawal of N supply has been examined in Arabidopsis

Plant Soil (2011) 345:387–400

thaliana, tomato and rice (Wang et al. 2001; Wang et al. 2003; Scheible et al. 2004; Bi et al. 2007). Ruzicka et al. (2010) quantified changes in gene expression in tomato roots grown in soil 53 h after exposure of roots to differential additions of ammonium. Lu et al. (2005) used a microarray to identify changes in gene expression in wheat grain in response to different rates and sources (i.e. organic vs mineral) of N fertilization. The genes identified in these studies are related to N metabolism, carbohydrate metabolism, photosynthesis and stress responses. Li et al. (2010) assessed the potential to use gene expression to quantify N sufficiency of potato plants grown in a hydroponic system in the greenhouse. Plants from three potato cultivars were grown under N abundant conditions and, while in vegetative growth, exposed to abundant, limited or deficient N supplies in nutrient solution. Gene expression in the last fully expanded leaf was quantified for genes encoding nitrate reductase, nitrite reductase, asparagine synthetase and ammonium transporter using real-time PCR. This study demonstrated that it was possible to use gene expression to quantitatively assess a change in potato N sufficiency within a few days of imposition of N deficiency stress. Although a nitrate reductase gene was identified as a good candidate gene for diagnosis of N sufficiency, the evaluation was done under somewhat artificial conditions where plants were grown in the greenhouse and exposed to an abrupt and substantial reduction in N supply (Li et al. 2010). There is now a need to examine the potential to use gene expression to quantify potato N sufficiency under more realistic growing conditions. The purpose of this experiment was to examine the potential to quantify potato N sufficiency under field conditions using gene expression indicators. Gene expression was quantified in leaf tissue, as would be appropriate for development of a practical diagnostic tool for assessment of potato N sufficiency. The experiment used the cultivar Shepody, which had been shown by Li et al. (2010) to be responsive under greenhouse conditions, grown under a series of six fertilizer N rates ranging from deficient to excessive. Expression of a range of genes involved in N uptake/ transport, N assimilation and amino acid metabolism was compared with relative leaf chlorophyll content and petiole nitrate concentration, the most common optical and biochemical measures of N sufficiency in potato plants, in order to determine the utility of

Plant Soil (2011) 345:387–400

monitoring gene expression for assessing crop N sufficiency.

Materials and methods The experiment was conducted at the Potato Research Centre of Agriculture and Agri-Food Canada, Fredericton NB, Canada in 2009. Soils at the experimental site belong to the Research Station soil association, are formed in coarse loamy morainal ablational till over coarse loamy morainal lodgement till, and are classified in the Canadian soil classification system as Orthic Humo-Ferric Podzols (Rees and Fahmy 1984). Soil properties for the 0 to 15 cm depth were: sand content 561 g kg–1, silt content 344 g kg–1, clay content 95 gkg–1 (pipette method), soil pH 5.9 (1:1 water), and soil organic C content 18.4 g kg–1 (dry combustion). The preceding crop was spring wheat (Triticum aestivum L.). No irrigation was applied as is common in this production region. The experiment used a randomized complete block design with six treatments and four blocks. Treatments were six fertilizer N rates (0, 50, 100, 150, 200 and 250 kg N ha–1) banded at planting as ammonium nitrate (34-0-0). The recommended fertilizer N rate for this field was 165 kg N ha–1 (Zebarth et al. 2007), and these N rates were chosen to range from deficient to excessive. All plots also received 150 kg ha–1 of P2O5 and K2O banded at planting. Plots were six rows (5.46 m) by 8 m in size where the outer rows were guard rows. The experiment was planted on May 25 using 0.91 m row spacing and 0.3 m within-row spacing. A modified planter was used to band the fertilizer treatments and open the rows. Hand-cut 57 g seedpieces of cultivar ‘Shepody’ were hand-planted and imidacloprid applied to control Colorado potato beetle, and the seed-pieces covered using discs. One hill of cultivar Chieftain was planted at the end of each row to avoid edge effects. Sampling of leaf tissues was performed weekly for seven sampling dates from July 7 until August 18. On each date the last fully expanded leaf, usually the fourth leaf from the top of the plant, was sampled from each of 20 randomly selected plants in each plot. One leaf disk (8 mm diameter) was sampled from the terminal leaflet of each plant, placed in a 2.0 ml Lysing Matrix D

389

Tube (MP Biomedicals LLC) and the tube immediately placed in liquid nitrogen, and stored at −80°C until analysis. These samples were used for RNA extraction and gene expression analyses as described below. Leaf chlorophyll index was determined using a Minolta SPAD-502 meter. One reading was made on each terminal leaflet from which a leaf disk was sampled as described by Zebarth et al. (2003). Petioles were then collected for determination of petiole nitrate concentration. Each leaf from which a leaflet disk had been collected was removed from the plant, the leaflets removed, the petioles dried at 55°C and ground to pass a 2 mm screen. A 0.2 g subsample of petiole tissue was extracted with 40 ml distilled water and a 15 min shaking time, and NO3-N concentration of the extract determined colorimetrically as described by Zebarth et al. (2003). In addition, soil samples were collected on each sampling date for determination of nitrate and ammonium concentrations within the root zone. In each plot, six soil cores 2.5 cm in diameter were collected from 0–30 depth in the ridge adjacent to the potato plants, and combined to form a single composite soil sample per plot. Soil samples were frozen at −20°C until analysis. Soil was thawed and a 20 g subsample of moist soil was extracted with 2 M KCl using a 1:5 soil:extractent ratio and 30 min shaking time. Concentrations of NO3-N and NH4-N in the extract were determined colorimetrically as described by Zebarth and Milburn (2003). Whole plant samples (i.e. vines plus tubers) were collected on August 25, prior to significant vine senescence, to estimate dry matter and N accumulation. Four adjacent plants from each of two rows were harvested, and plant tissues partitioned into tubers, vines, and stolons plus readily recoverable roots. Dry matter and N accumulation of each plant component was determined as described by Zebarth and Milburn (2003). Vine desiccation was performed using diquat on September 15 and total tuber yield from two rows was determined on October 14. Tuber size distribution was assessed as the percentage of small tubers (284 g) and the mean tuber weight of tubers greater than 5 cm diameter. Tuber specific gravity was determined using the weight-in-water and weight-in-air method (Kleinschmidt et al. 1984).

390

RNA extraction and quantification of gene expression Leaf tissue was ground in the Lysing Matrix D Tubes using a Geno Grinder 2000 bead mill (Glen Mills) for 2 min at 1500 strokes/min in liquid nitrogen cooled blocks. 750 μl RNA lysis buffer (200 mM sodium borate decahydrate (Borax), pH 9.0, 30 mM ethylene glycol bis (β-aminoethyl ether)-N-N’-tetraacetic acid (EGTA), 1% (w/v) sodium dodecyl sulfate (SDS), and 10 mM dithiothreitol) plus 1% (w/v) sodium deoxycholate and 2% (w/v) polyvinylpyrrolidone (PVP) (Mr 40,000)) (Wang and Wilkins 1994) was added to ground leaf tissue in each tube. The lysate was returned to the bead mill and mixed for 2 min at 1500 strokes/min. The homogenate was centrifuged for 5 min at 10,000 rpm to clear the lysate. 200 μl of lysate supernatant was used for RNA extraction using the Agencourt RNAdvance Tissue Kit (http://www.agencourt.com/products/spri_ reagents/rnadvancetissue/) according to manufacturer’s instructions for liquid samples. Multiplex analysis of the expression of 27 genes was done using the nCounter Digital Analyzer (Nanostring Technologies, Inc.). Details of nCounter multiplex gene expression analysis are described elsewhere (Geiss et al. 2008). The 27 genes included 5 housekeeping genes and 22 experimental genes (Table 1). The five housekeeping genes were adenine phosphoribosyl transferase (aprt), tubulin, cyclophilin (Nicot et al. 2005; Tai et al. 2009), elongation factor 1-α(ef1α) (Nakane et al. 2003) and COX1 (Quiñones et al. 1995; Li et al. 1996). The test genes are described in Table 1. Sequences for probes used in nCounter analysis are reported in Luo et al. (2011). Sequence data used to design probes were from National Centre for Biotechnology Information (NCBI) and the Canadian Potato Genome Project (CPGP) EST database (www.cpgp.ca). The nCounter data was adjusted using the manufacturer-provided spiked positive and negative controls according to manufacturer’s instructions. The geometric mean of the five housekeeping genes was calculated and used to normalize the experimental genes on a per sample basis (Vandesompele et al. 2002). Statistical analyses Data were tested for normality using the KolmogorovSmirnov test and log10 transformation was performed if

Plant Soil (2011) 345:387–400

appropriate. Analysis of Variance (ANOVA) was performed where the statistical model treated N rate treatment as a main plot and time as a sub-plot. Treatment means in main effects were compared using a protected LSD test. Treatment means in interactions were compared using LSmeans. All statistical analyses were performed using SAS (SAS Institute Inc., Cary, NC, version 9.2) with a critical P value of 0.05. Means presented in tables and figures were calculated using non-transformed data. Mean growing season (May to September) air temperature averaged 15.8°C, the same as the longterm (1971–2000) average (Environment Canada 2010). Growing season precipitation was 554 mm, 20% higher than the long-term average of 462 mm, with above-average rainfall occurring in every month except May.

Results Tuber total yield increased with increasing fertilizer N rate (Table 2). Maximum tuber yield was measured at the 200 kg N ha−1 rate; however there was no significant increase in tuber yield above the 150 kg N ha−1 rate. Increasing fertilizer N rate generally decreased the percentage of small tubers, increased the percentage of large tubers, and increased mean tuber weight. Tuber specific gravity averaged 1.094 and was not affected by fertilizer N rate. Increasing fertilizer N rate also increased plant dry matter measured prior to vine desiccation, reaching a maximum value at the 150 kg N ha−1 rate. Plant N accumulation increased with increasing N rate reaching a maximum value at the 250 kg N ha−1 rate, however, there was no significant increase in plant N accumulation for N rates above 150 kg N ha−1. There was a strong positive correlation between plant N accumulation and both dry matter accumulation (r=+0.98) and total tuber yield (r=+0.99). There was a significant fertilizer N rate by sampling date interaction on soil ammonium concentration for 0–20 cm depth in the potato root zone. Soil ammonium-N concentrations were low, less than 10 mg N kg−1 in all but one case (Fig. 1a). Soil ammonium concentrations were higher for the 200 and 250 kg N ha−1 rates on July 7, and higher for the 250 kg N ha−1 rate on July 14 and 21, compared with the unfertilized plots.

Plant Soil (2011) 345:387–400

391

Table 1 The 5 housekeeping and 22 experimental genes used for analysis of gene expression by nCounter. The nCounter Code is the name of the probe used, the Gene Name column contains abbreviated gene names used in this study, Accession numbers are either from NCBI or are contig names from the nCounter code

Gene name

Canadian Potato Genome Project (www.cpgp.ca), functional description for genes were taken from Solanum tuberosum gene annotation at NCBI or from the annotation of the top blast hit of the contig at www.cpgp.ca

Accession number

Functional description

Housekeeping genes Aprt

APRT

DQ284483.1

Adenosine phosphoribosyl transferase

TUBST1

TUBST1

Z33382.1

mRNA for beta-tubulin

cyclophilin

cyclophilin

AF126551.1

cyclophilin mRNA

EF-1-alpha

EF1a

AB061263.1

EF-1-alpha mRNA for elongation factor 1-alpha

cox 1-B

COX1

X83206.1

Mitochondrial cox1 gene

N uptake/ transport genes St_NT1

NT1

05Mar7_CPGP_1000_C1.1

Nitrate transporter similar to Arabidopsis thaliana NTP3 gene

St_NT2

NT2

05Mar7_CPGP_913_C1.1

Nitrate transporter similar to Nicotiana tabacum NtNRT1.1-t gene

St_AT1

AT1

05Mar7_CPGP_10374_C1.1

Ammonium transporter similar to Solanum lycopersicum LeAMT1;2 gene

St_AT2

AT2

05Mar7_CPGP_4203_C1.1

Ammonium transporter similar to Solanum lycopersicum LeAMT1;3 gene

NiT_St3

NiT

05Mar7_CPGP_3786_C3.1

Nitrite transporter similar to Cucumus staivus NiTR1 gene

N assimilation genes NR_St1

NR

AB062142.1

StNR5 mRNA for nitrate reductase

NiR_St9

NiR

05Mar7_CPGP_7970_C1.1

Nitrite reductase similar to Nicotiana tabacum nii2 gene

St_GS1

GS1

AF302115.1

Glutamine synthetase GS1 (gln) mRNA

St_GS2

GS2

AF302113.1

Glutamine synthetase GS2 (gln) mRNA

GSR_St3

GLU1

05Mar7_CPGP_5699_C1.1

Glutamate synthase similar to Glycine max ferridoxin-dependent glutamate synthase GLU1 gene

GSR_St4

GLU2

05Mar7_CPGP_7797_C1.1

Glutamate synthase similar to Arabidopsis thaliana NADH-dependent glutamate synthase gene

Amino acid metabolism genes St_As1

AS1

05Mar7_CPGP_4091_C1.1

Asparagine synthetase similar to Striga hermonthica AS gene

St_As2

AS2

05Mar7_CPGP_790_C1.1

Asparagine synthetase similar to Triphysaria versicolor AS gene

St_GLT1

GDH1

05Mar7_CPGP_601_C1.1

NADH-glutamate dehydrogenase similar to Solanum lysopersicum gdh gene

St_GLT2

GDH2

05Mar7_CPGP_11097_C1.1

NADH Glutamate dehydrogenase similar to Nicotiana plubaginifolia gdhA gene

A_St

ARG1

05Mar7_CPGP_2985_C1.1

Arginase 1 similar to Solanum lysopersicum LeARG1 gene

UR1

ureG

AJ272526.1

mRNA for urease accessory protein G

U1

ure

AJ308543.1

mRNA for urease (ure gene)

CDPK

AY098940.1

Calcium-dependent protein kinase-like protein

SOD

SOD

BE923620.1

Superoxide dismutase copper chaperone

St_CP

Cys-peptidase

CK264216.1

Cysteine-type peptidase

ABI2

ABI2

CV473245.1

ABA Insensitive 2

Stress response genes ST_CDPK

There was also a significant fertilizer N rate by sampling date interaction on soil nitrate concentration for 0–20 cm depth in the potato root zone. Soil nitrate

concentrations were higher and more responsive to fertilizer N rate (Fig. 1b) compared with soil ammonium concentrations (Fig. 1a). Elevated soil nitrate concen-

392

Plant Soil (2011) 345:387–400

Table 2 Effect of fertilizer N rate on total tuber yield, tuber size distribution and plant dry matter and N accumulation measured prior to significant vine senescence N rate kg N ha−1

Total tuber yield t ha−1

Small tubers %

0

15.5da

21.4a

50

23.0c

18.0a

100

29.2b

9.6b

150

36.6a

200

39.7a

250

38.0a

Large tubers %

Mean tuber weight g

Dry matter accumulation t ha−1

N accumulation kg N ha−1

3.7c

156c

4.42d

52c

6.7bc

168c

5.69cd

72c

11.4b

183b

7.17bc

109b

7.5b

22.6a

207a

9.12a

158a

8.1b

24.0a

213a

8.49ab

165a

7.1b

25.3a

209a

8.96ab

170a

*

*

*

*

Statistical significance N rate SEM (n=4; 15 df)

*

* 1.6

1.8

2.1

4.2

0.62

7.5

*P0.05) based on a protected LSD test

trations were measured for fertilizer N rates above 150 kg N ha−1 for the July 7, July 14 and July 21 sampling dates and for the 250 kg N ha−1 rate on most sampling dates (Fig. 1b). Petiole nitrate concentration increased with increasing fertilizer N rate with the maximum value measured at the highest fertilizer N rate on each sampling date (Fig. 2a). Petiole nitrate concentration generally decreased over time, but increased between the July 14 and July 21 sampling dates. There was a significant interaction between fertilizer N rate and sampling date on petiole nitrate concentration, and a strong positive correlation between total tuber yield and petiole nitrate concentration measured on every sampling date (Table 3). When all sampling dates and fertilizer N rates were considered, there was a curvilinear relationship between petiole nitrate concentration and soil nitrate concentration in the potato root zone (Fig. 3a). There was also a significant fertilizer N rate by sampling date interaction on SPAD reading. SPAD

-1

a)

N rate (kg N ha ) 0 150 50 200 100 250

b) 60

20

Nitrate (mg N kg-1)

-1

Ammonium (mg N kg )

Fig. 1 Effect of fertilizer N rate on a) soil ammonium-N concentration and b) soil nitrate-N concentration for 0–20 cm depth in the potato root zone on seven sampling dates

readings generally increased with increasing fertilizer N rate (Fig. 2b). However on sampling dates until August 11, SPAD reading did not increase for fertilizer N rates in excess of 100 kg N ha−1. There was a strong positive correlation between total tuber yield and SPAD reading on all sampling dates, but the correlation was not statistically significant on the July 14 and July 21 sampling dates (Table 3). There was a significant effect of fertilizer N rate and sampling date on expression of four of the five N uptake/transport genes examined (Table 4). Expression of AT1 was very high, and decreased with increasing fertilizer N rate in a relatively consistent manner on all sampling dates (Fig. 4a). There was a significant negative correlation between expression of AT1 on all sampling dates and total tuber yield (Table 3). In addition, there was a curvilinear relationship between expression of AT1 and soil nitrate concentration in the root zone over all sampling dates and fertilizer N rates (Fig. 3b). In contrast to AT1, expression of AT2 was

±1 SE

15 10 5

50

±1 SE

40 30 20 10 0

0 1

15 July

1

15 August

31

1

15 July

1

15 August

31

Plant Soil (2011) 345:387–400

a) 40

40 ±1 SE

SPAD reading

±1 SE

30 20 10 0

35 30 25 20

1

15 July

1

very low and not affected by fertilizer N rate or sampling date (Table 4). Expression of NT2 generally decreased with increasing fertilizer N rate (Fig. 4b). There was a significant negative correlation between total tuber yield and expression of NT2 measured on Table 3 Pearson correlation coefficients between total tuber yield and measures of potato N status including petiole nitrate concentration, leaf chlorophyll using a SPAD-502 meter and gene expression on seven sampling dates

-1

N rate (kg N ha ) 0 150 50 200 100 250

b)

-1

Petiole nitrate (mg N kg )

Fig. 2 Effect of fertilizer N rate on a) petiole nitrate concentration and b) leaf chlorophyll measured using a SPAD-502 meter on seven sampling dates

393

Parameter

15 August

31

1

15 July

1

15 August

31

July 14, July 21, Aug. 6 and Aug. 12 (Table 3). In comparison, there was a significant negative correlation between total tuber yield and expression of NT1 only on July 14 (Table 3). Expression of NiT generally decreased with increasing fertilizer N rate on July 7, July 7

July 14 July 21 July 28 Aug. 6

Aug. 12 Aug. 18

Petiole nitrate concentration +0.97*

+0.97*

+0.95*

+0.96*

+0.96*

+0.95*

+0.88*

SPAD reading

+0.94*

+0.78

+0.77

+0.93*

+0.96*

+0.97*

−0.34

+0.87*

N uptake/ transport genes NT1

−0.19

−0.87*

−0.33

−0.44

−0.20

NT2

−0.35

−0.89*

−0.96*

−0.77

−0.86* −0.85*

−0.50 −0.67

AT1

−0.90* −0.96*

−0.97*

−0.95*

−0.90* −0.86*

AT2

−0.38

−0.92*

+0.19

+0.27

+0.48

−0.65

−0.64

−0.87*

NiT

−0.79

−0.91*

−0.74

−0.29

+0.54

−0.16

−0.49

+0.18

−0.93*

−0.93*

−0.19

+0.44

−0.48

−0.47 −0.01

N assimilation genes NR NiR

+0.89*

−0.88*

−0.85*

+0.60

+0.78

+0.72

GS1

+0.98*

+0.72

+0.59

+0.63

+0.72

+0.91*

+0.49

GS2

+0.88*

−0.52

−0.71

+0.71

+0.76

+0.34

+0.06

GLU1

+0.55

−0.68

−0.76

−0.03

+0.95*

+0.36

−0.25

GLU2

+0.68

+0.79

+0.48

+0.81

+0.63

+0.06

−0.40 −0.42

Amino acid metabolism genes AS1

+0.51

−0.69

−0.68

+0.09

−0.43

+0.13

AS2

+0.94*

+0.76

+0.55

−0.17

+0.26

+0.39

+0.22

GDH1

−0.95* −0.91*

+0.08

−0.78

−0.55

−0.89*

−0.78

GDH2

+0.20

−0.72

−0.79

+0.50

+0.23

−0.43

0.00

ARG1

−0.93* −0.77

+0.13

−0.54

−0.40

−0.86*

−0.87*

ureG

−0.05

−0.87*

+0.05

+0.08

−0.15

−0.59

ure

−0.92* −0.72

+0.41

−0.18

−0.01

+0.15

−0.19

−0.65

+0.41

−0.68

+0.27

−0.17

+0.39

−0.94*

Stress response genes CDPK

*P0.05) based on a protected LSD test.

*P0.05) based on a protected LSD test

*P