Introduction

Poacynum pictum from Apocynaceae family is widely distributed in Xinjiang China. There are more than 6 ecotypes of the species that can be distinguished from morphological and biological characteristics (Gao et al. 2015). P. pictum are perennial herbaceous plants and their heights can be more than 4 m. Plants leaves are used as a traditional Chinese medicine for the treatment of hypertension, hepatitis, and depression (Zhang 2004; Xie et al. 2012) and teas from the leaves of P. pictum have been used as a nutritional supplement in North American and East Asian health food markets (Shi et al. 2011; Song and Zhou 2015). The fiber obtained from the stem of the two species has become the interest in the textile industry (Gong et al. 2017).

The leaves of Apocynaceae plants contains high flavonoids content (Zhou et al. 2015) which has been considered to play a major role in Chinese traditional medicines and tea (Nishibe et al. 2001; Ma et al. 2003), e.g., flavonoids isolated from Apocynum venetum have significant anti-depressant activities for mice (Yan et al. 2015) and inhibit the progress of atherosclerosis in rats via the AMPK/mTOR pathway (Lü et al. 2017). Flavonoids have also been noted to play a role in suppressing the growth of bacteria (Kang et al. 2014), such as Escherichia coli, and highly decrease the ability of the bacteria to initiate an infection (Nguyen et al. 2016; Shang et al. 2017). In tissues, the substances may act as antioxidants and significantly contribute to scavenging of free radicals produced via metabolic processes in plants. The leaves of Apocynaceae plants contain metals such as Ca, Fe, Zn and Na which play important role in both plant and animal metabolism (Yokozawa et al. 2002). Rust caused by Melampsora apocyni is the major disease affecting A. venetum and P. pictum, and all ecotypes have been found to be affected by this species causing rust (Gao et al. 2017a). The disease has been reported in Russia (Tranzschel 1891), Japan (Hiratsuka 1939), Kazakhstan (Nevodovskii 1956), China (Tai 1979), Bulgaria (Denchev 1995) and Turkey (Kirbağ 2004). In Xinjiang China, we studied the rust on leaves of wild and cultivated A. venetum plants from 2009 until now and this study is ongoing. The rust occurrence could reach up to 100% in field conditions cause the loss of leaves and also the death of plants in severe situations, ultimately causing large economic losses (Gao et al. 2017a).

Rust diseases have been reported to reduce the crop yield include A. venetum (Gao et al. 2017a), which decrease the crop protein and amino acid content in alfalfa and Vicia sativa (Nan 1986, 1990) and decrease the feed value of forage plants. Reports have revealed that rust diseases could decrease plant amino acid content by changing the metabolism of nitrogen in host plants (Nan 1986, 1990). Other reports found that rust pathogens such as species of Melampsora, Phakopsora, Puccinia and Uromyces can uptake amino acids in plants through intracellular haustoria, thus decrease amino acid content (Struck 2015). Rust disease has been reported to decrease the Ca and P concentration in alfalfa plants and these metals also play role in the plant disease defense system (Fones et al. 2010; Pinheiro et al. 2011). However, rust infection has been reported to enhanced the concentration of plant flavonoids (Miranda et al. 2007; Lu et al. 2017).

Two ecotypes of P. pictum, one with red stems with medium leaves referred to as ecotype RSM and another with green stems with fine leaves that is referred to as ecotype GSF, are essential cultivated plants in Altay for tea and fiber production (Gao et al. 2015). Rust has been found to be easily infect these two ecotypes, and the average disease incidence could reach up to 70% in cultivated fields. In this study, the effect of rust severity, based on a 6–step rating scheme and on series of parameters related to the quality of P. pictum for tea production was analyzed. We hypothesized that the infection of rust (M. apocyni) will decrease the biomass, amino acid content and metal concentration and increase the total flavonoids of these two ecotypes of P. pictum.

Materials and Methods

Experimental details and methods

Samples cultivation: The two ecotypes of P. pictum, Eco–RSM and Eco–GSF (Fig. 1), used in this study were basically obtained as selections from local naturally-growing P. pictum plants. They were cultivated in the Alakak Township in Altay Prefecture of the Xinjiang Uyghur Autonomous Region, China (Alakak field, 47°42′N, 87°33′E, at an altitude of 492 to 547 m, area 5.23 ha), using plants grown from seeds which were collected from locally growing plants. The soil was a sandy loam having pH of 7.2 to 7.5, and plant rows were spaced 3 meters apart while within the rows space was 1 m apart. Emergence of P. pictum stems began in the middle of April. P. pictum plants were irrigated by drip irrigation system after every 6 days and no disease control measures were applied.

Rust disease survey

For each ecotype of P. pictum, four 20 × 20 m plots were used to assess the disease incidence of the rust species. From each plot 200 leaves were sampled non–destructively for each evaluation, and the disease incidence was recorded on leaves based on the presence of uredinia. The rust severity was recorded as a 6–step rating scheme by visually calculation of the percentage of observed leaves that were covered by uredinia: 0 for no signs of infection; 1 for 0.1 to 5% of leaf area covered with uredinia; 2 for 5.1 to 20%; 3 for 20.1 to 50%; 4 for 50.1 to 75%; and 5 for >75.1%. For each rust severity of the two P. pictum ecotypes (Fig. 2), 100 leaves of similar size from similar branches of the plants were sampled for dry weight, and total flavonoids, amino acid and metal content were measured.

Total flavonoids extraction

According to this method, total flavonoids content (TFC) was calculated by acid hydrolysis method (Zhao et al. 2016) with little changes. For each rust severity, 100 leaves were dried in an oven at 65°C for 24 h, the dry leaves were ground into fine powders and mixed thoroughly. One-gram sample was precisely weighed and added to conical flask containing 25 mL of 70% ethanol in a water bath with 65°C constant temperature for 1h for reflux extraction. After cooling, 70% ethanol was added to the 25 mL and was thoroughly mixed and filtered through quantitative filter paper. One mL of the filtered solution was transferred to an evaporating dish, 1 g polyamide powder was added and the dish was kept in the water bath to remove the ethanol. The sample was then transferred into a chromatographic column, 20 mL of petroleum ether was added to remove impurities, the column was then washed with 20 mL of methyl alcohol and the eluent was collected. This eluent was allowed to dry. The residue was dissolved in methanol and transferred into a 10 mL volumetric flask. Two mL of the solution was added to 10 mL volumetric flask, and then added 5 mL of 30% ethanol and 0.3 mL of 5% sodium nitrite solution, and the solution was allowed to stand for 5 min. Then 0.3 mL of 10% nitric acid aluminum solution was added and the sample kept for another 6 min. Finally, 2 mL of 1.0 mol L^-1 sodium hydroxide solution was added. The absorbance was measured at 510 nm by ultraviolet visible absorption spectrophotometer (UV-7502c, China).

Amino acid extraction

According to this method, amino acid extraction was measured by the Aluminium nitrate colorimetric method with minor modification (Ksenofontov et al. 2017). The fine powder of P. pictum leaves as described in above method was used for amino acid extraction. For each rust disease severity, 30 mg powdered samples were weighed and added accurately to a hydrolysate tube in which 10 mL of 6 mol L^-1 hydrochloric acid was added. The hydrolysate tube was evacuated by a vacuum pump and then filled with nitrogen; this procedure was repeated 3 times and then the tube was sealed with an alcohol blast burner. The tubes with samples were kept for 22 h in a constant temperature oven at 110ºC to hydrolyze. The hydrolysate solutions were shifted to volumetric flasks and made up to constant volume of 50 mL with deionized water. One mL of the solution was added into a 5 mL of beaker and left to dry in glass desiccator. The dried sample was dissolved with 1 mL of acetic acid buffer having 5.5 pH, and used to detect the amino acids presence by use of an automatic amino acid analyzer (Hitachi 835, Japan).

Tryptophan was determined by alkaline hydrolysis method. Powdered samples (30 mg) were added accurately to a poly tetra fluoroethylene tube in which 3 mL of 5 mol L^-1 sodium hydroxide was added. The tubes with samples were kept for 22 h in a constant temperature oven at 110℃ to hydrolyze. The hydrolysate solutions were shifted to 25 mL volumetric flasks and adjust pH to 7.6 after cooling, made up to constant volume of 25 ml with deionized water. Finally, 1 ml of the solution was shifted to 10 mL tube, made up to constant volume of 10 mL with 4 mol L^-1 (pH 11) urea solution, and used to detect the Tryptophan presence by use of a spectrofluorophotometer (Shimadzu RF-5000, Japan).

Metal evaluation

For mineral content, 0.2 g of the dried samples were precisely weighed and dissolved in HNO₃:HClO₄:H₂SO₄ (8:1:1) mixture, and heated on a digestion furnace at 420°C in a fume hood. After the digestion and subsequent cooling, the digested samples were added to volumetric flasks and were diluted by the addition of ultrapure water to 100 mL. A flame atomic absorption spectrophotometer (Thermo ICE 3300, Germany) was used to measure metal elements (Zhang 2004).

Fig. 1: Eco RSM-red stem with medium leaves (left) and Eco GSF-green stem with fine leaves (right) of P. pictum

Fig. 2: The rust severity visually estimating the percentage of observed leaves that were covered by uredinia: 0 for no signs of infection; 1 for 0.1% to 5% of leaf area covered with uredinia; 2 for 5.1% to 20%; 3 for 20.1% to 50%; 4 for 50.1% to 75%; and 5 for >75.1%. For each rust severity of the two P. pictum ecotypes

Statistical analysis

All data are presented as means and standard errors of means for four replicates. The significance of differences at a 5% level between averages was determined by one-way ANOVA using Tukey's HSD test.

Results

Rust disease occurrence, biomass production

In the research sites, the rust was firstly found in late July, and the two ecotypes of P. pictum had similar rust disease incidence during the growth period. The disease incidence was kept under 20% until the middle of August, and then there was a sharp increase from 20–60% in a week. Subsequently the development of rust disease slowed down but reached up to 70% in two weeks, and this persisted until leaves were lost and the growth period of the plants ceased (Fig. 3). The infection of rust had no effect on leaf dry weight of P. pictum plants of the GSF ecotype, however infection significantly decreased the dry weight of leaves of the RSM ecotype at rust severity 4 (P < 0.05). There was no significantly difference between rust-infected leaves with different rust severity (Fig. 4). Our hypothesis that rust will decrease the biomass of the two ecotypes of P. pictum was partly supported (Table 1).

Total flavonoids concentration

The concentration of flavonoids in healthy leaves of the two ecotypes is 2.4–2.8 g 100 g^-1, which is 5.2–9.4% higher than rust–infected leaves. Compared with healthy leaves, the infection of the rust pathogen M. apocyni only significantly decreased flavonoids when the rust severity were 4 and 5 and the flavonoids concentrations were 36.4 and 25.5% lower than healthy leaves, respectively, for the RSM ecotype (Fig. 5). There was no significant difference in flavonoid concentrations among leaves of all severities of leaves of the GSF ecotype (Fig. 5 and Table 1). Rust disease had no significant effects on flavonoids concentration in leaves of the GSF ecotype.

Table 1: ANOVA result of the effects of rust disease caused by Melampsora apocyni on the listed variables to two ecotypes of Poacynum pictum

Variables	Eco RSM -red stem with medium leaves			Eco GSF-green stem with fine leaves
	F values	DF	P values	F values	DF	P values
Leaf dry weight	6.3478	5	0.0015	1.5728	4	0.2325
Disease incidence	18.6747	3	0.0001	22.0470	3	0.0001
Flavonoids concentration	3.9519	5	0.0136	0.4540	4	0.7681
Calcium concentration	4.4805	5	0.0079	5.2512	4	0.0076
Copper concentration	13.3846	5	0.0001	21.8334	4	0.0001
Zinc concentration	7.4581	5	0.0006	0.8233	4	0.5304
Iron concentration	3.7604	5	0.0166	2.3700	4	0.0991
Alanine	24.1229	5	0.0001	11.3179	4	0.0002
Argnine	34.1459	5	0.0001	16.0278	4	0.0001
Aspartic	21.9949	5	0.0001	5.3065	4	0.0072
Cystine	3.1365	5	0.0329	11.3432	4	0.0002
Glutamine	22.0609	5	0.0001	11.9762	4	0.0001
Glycine	26.2400	5	0.0001	9.6809	4	0.0004
Hlstidine	26.3829	5	0.0001	11.2136	4	0.0002
Isoleucine	24.0377	5	0.0001	9.7925	4	0.0004
Leucine	28.0698	5	0.0001	13.4100	4	0.0001
Lysine	21.4042	5	0.0001	11.5913	4	0.0002
Methionine	2.6925	5	0.0550	6.1619	4	0.0039
Phenylalanine	25.9132	5	0.0001	11.2994	4	0.0002
Proline	24.5499	5	0.0001	11.1099	4	0.0002
Serine	15.9173	5	0.0001	4.7319	4	0.0114
Threonine	22.2670	5	0.0001	9.9547	4	0.0004
Tryptophan	6.0371	5	0.0019	2.5727	4	0.0806
Tyrosine	44.6389	5	0.0001	14.4998	4	0.0001
Valine	24.2808	5	0.0001	10.0280	4	0.0004

Amino acid concentration

Compared with healthy leaves, amino acid concentration was significantly affected (mainly decreased) by the rust disease except for methionine and cystine in the RSM ecotype, and tryptophan in the GSF ecotype. The decreases in the concentration of these three amino acids happen under all rust disease severity (Table 1 and 2). P. pictum leaves with rust severity 4 and 5 had the lowest amino acid concentration for the RSM ecotype. For the GSF ecotype the content of 10 and 16 of 18 amino acids was reduced by rust disease at rust severity 3 and by the rust disease at severity 4, respectively, rust severity 4 had significantly higher amino acid content, 16 of 18 except Methionine and Trytophan, than that of rust severity 1. When rust severity were 4 and 5, rust disease significantly decreased the concentration of 15 amino acids in leaves of the RSM ecotype, Except for Cystine, Methionine and Trypothan, which given the same value under the different rust severity (Table 2).

The correlations of rust disease severity and the amino acids depicts that amino acid content is negatively correlated with rust disease severity for the two ecotypes, except for Cystine, Methionine and Trytophan (Table 3). The amino acid response to rust disease differed in the two ecotypes. With the GSF ecotype the content of Methionine and Trytophan had no correlation with rust disease severity. Glutamine is the most sensitive amino acid to rust disease, followed by Leucine, while Trytophan is the least sensitive amino acid to rust disease with both ecotypes of P. pictum (Table 2). Our hypothesis that rust will decrease the amino acid of the two ecotypes of P. pictum was upheld.

Table 2: Amino acid concentration of two ecotypes of Poacynum picttum under different rust disease severity caused by Melampsora apocyni %

Amino acid	Rust disease severity of Ecotype red stem with medium leave (Eco-RSM)						Rust disease severity of Ecotype green stem with fine leaves (Eco-GSF)
	0	1	2	3	4	5	0	1	2	3	4
Alanine	1.22±0.05 a^*	1.24±0.05 a	1.28±0.01 a	1.11±0.06 a	0.82±0.05 b	0.77±0.04 b	1.31±0.07 a	1.23±0.05 ab	1.04±0.04 bc	1.08±0.06 ab	0.84±0.05 c
Argnine	1.16±0.05 ab	1.19±0.05 ab	1.34±0.06 a	1.01±0.05 b	0.73±0.04 c	0.68±0.03 c	1.20±0.07 a	1.10±0.05 ab	0.93±0.01 bc	0.92±0.03 bc	0.73±0.05 c
Aspartic	1.76±0.08 a	1.86±0.08 a	1.97±0.05 a	1.68±0.08 a	1.24±0.07 b	1.18±0.05 b	1.87±0.10 a	1.73±0.07 a	1.55±0.07 ab	1.60±0.10 ab	1.32±0.10 b
Cystine	0.58±0.06 a	0.55±0.02 a	0.49±0.01a	0.47±0.05 a	0.39±0.03a	0.46±0.03a	0.62±0.02 a	0.56±0.00ab	0.51±0.01bc	0.48±0.03bc	0.46±0.02 c
Glutamine	2.41±0.08 a	2.44±0.13 a	2.39±0.01 a	2.21±0.13a	1.59±0.07 b	1.53±0.07 b	2.46±0.12 a	2.26±0.09 ab	1.97±0.03 bc	1.97±0.08 bc	1.65±0.10 c
Glycine	1.04±0.05 a	1.08±0.05 ab	1.15±0.02 ab	0.95±0.04 b	0.70±0.04 c	0.66±0.03 c	1.12±0.06 a	1.04±0.05 ab	0.90±0.02 bc	0.93±0.05 abc	0.74±0.04 c
Hlstidine	0.45±0.02 ab	0.46±0.02 ab	0.51±0.02 a	0.41±0.02 b	0.30±0.02 c	0.28±0.02 c	0.48±0.02 a	0.44±0.02 ab	0.38±0.01 bc	0.40±0.02 ab	0.30±0.02 c
Isoleucine	0.96±0.04 ab	1.00±0.04 ab	1.07±0.02 a	0.89±0.04 b	0.67±0.04 c	0.64±0.03 c	1.00±0.06 a	0.96±0.04 ab	0.82±0.02 bc	0.83±0.03 abc	0.69±0.04 c
Leucine	1.73±0.08 ab	1.81±0.08 ab	1.96±0.04 a	1.56±0.08 b	1.14±0.07 c	1.07±0.06 c	1.88±0.11 a	1.75±0.08 ab	1.47±0.03 bc	1.48±0.05 bc	1.19±0.07 c
Lysine	1.21±0.04 a	1.23±0.05 a	1.18±0.02 a	1.09±0.06 a	0.80±0.05 b	0.76±0.04 b	1.23±0.07 a	1.15±0.05 ab	1.01±0.01 bc	1.02±0.03 bc	0.83±0.04 c
Methionine	0.21±0.06 a	0.08±0.00 a	0.09±0.00 a	0.08±0.01 a	0.12±0.04 a	0.10±0.02 a	0.13±0.00 ab	0.16±0.02 ab	0.18±0.02 a	0.13±0.01 b	0.11±0.00 b
Phenylalanine	1.10±0.05 ab	1.16±0.05 ab	1.29±0.04 a	1.03±0.05 b	0.75±0.04 c	0.70±0.04 c	1.20±0.07 a	1.13±0.05 ab	0.97±0.02 bc	0.98±0.04 bc	0.80±0.04 c
Proline	0.92±0.03 a	0.95±0.04 a	1.00±0.02 a	0.86±0.04 a	0.63±0.03 b	0.60±0.03 b	0.98±0.05 a	0.92±0.04 ab	0.82±0.03 bc	0.80±0.03 bc	0.68±0.03 c
Serine	0.81±0.04 a	0.83±0.03 a	0.80±0.01 a	0.74±0.04 a	0.58±0.03 b	0.55±0.03 b	0.81±0.04 a	0.83±0.03 a	0.80±0.01 a	0.74±0.04 a	0.58±0.03 b
Threonine	0.91±0.04 a	0.96±0.04 a	0.96±0.00 a	1.68±0.04 a	0.63±0.03 b	0.59±0.03 b	0.98±0.05 a	0.90±0.04 ab	0.79±0.02 bc	0.81±0.04 bc	0.66±0.04 c
Tryptophan	0.13±0.01ab	0.17±0.01 a	0.14±0.01 a	0.16±0.03 a	0.07±0.01 b	0.12±0.02 ab	0.13±0.00 a	0.12±0.00 a	0.18±0.02 a	0.16±0.02 a	0.17±0.01a
Tyrosine	0.68±0.01 a	0.65±0.02 a	0.68±0.00 a	0.53±0.03 b	0.39±0.03 c	0.37±0.01 c	0.70±0.04 a	0.65±0.03 ab	0.53±0.01 bc	0.54±0.02 bc	0.42±0.03 c
Valine	1.21±0.04 a	1.28±0.06 a	1.35±0.03 a	1.13±0.05 a	0.84±0.05 b	0.80±0.04 b	1.28±0.07 a	1.21±0.06 ab	1.06±0.04 bc	1.05±0.04 bc	0.87±0.05 c

Note:* Data marked by the same lowercase letter in the same row do not differ significantly between rust severity for the same ecotype of Pocynum pictum.

Metal concentration

Rust show different effects on calcium (Ca), copper (Cu), iron (Fe) and zinc (Zn) concentrations of the two ecotypes of P. pictum (Table 1). For the GSF ecotype, rust disease had no effect on Fe and Zn concentration, while in case of severe rust infection (severity 4) significantly (P < 0.05) increased Ca and Cu concentrations. For the RSM ecotype, when compared with healthy leaves, the concentration of Ca was significantly (P < 0.05) decreased at rust severity 5, and with Cu at rust severity 4 and 5. Healthy leaves had the same concentration of Zn and Fe as with rust-infected leaves, but there was significant difference among rust-infected leaves with varying rust severity. Rust severity 2 and 3 had higher Zn concentrations than the other severities of rust infection. Rust severities of 5 and 3 had the highest and lowest Fe concentration among rust-infected leaves (Fig. 6). Our hypothesis that rust will decrease the metal of the two ecotypes of P. pictum was partly upheld.

Discussion

Fig. 3: Rust disease incidence of two ecotypes Eco RSM-red stem with medium leaves and

Eco–GSF green stem with fine leaves of P. pictum in cultivated field. The same lowercase letters up the bars means there is no significantly different by Turkey’s HSD at P < 0.05 for Eco RSM; The same uppercase letters up the bars means there is no significantly different by Turkey’s HSD at P < 0.05 for Eco GSF

Fig. 4: Leave dry weight of two ecotypes Eco RSM-red stem with medium leaves and Eco

GSF-green stem with fine leaves of P. pictum in cultivated field. The same lowercase letters up the bars means there is no significantly different by Turkey’s HSD at P<0.05 for Eco RSM; the same uppercase letters up the bars means there is no significantly different by Turkey’s HSD at P<0.05 for Eco GSF

Fig. 5: Flavonoids of two ecotypes Eco RSM-red stem with medium leaves and Eco GSF-green stem with fine leaves of P. pictum in cultivated field.

The same lowercase letters up the bars

Means there is no significantly different by Turkey’s HSD at P < 0.05 for Eco RSM; the same Uppercase letters up the bars means there is no significantly different by Turkey’s HSD at P < 0.05 for Eco GSF

Our previous research showed that infection by M. apocyni causes the leaves of A. venetum to turn yellow, wither and prematurely fall, and results in significant yield loss in the field (Gao et al. 2015). In a greenhouse experiment, infection by M. apocyni had a slight effect on photosynthesis of A. venetum during early disease development, but drought stress was more damaging than for non-inoculated plants in later disease development, leading to a great decrease in the net photosynthetic rate This reduction, however, did not cause a significantly decrease in the aboveground biomass of A. venetum plants between the rust-infected and non-infected treatments (Gao et al. 2017b). The difference in leaf biomass of the two ecotypes of P. pictum following rust infection indicates the diversity of plant responses to this pathogen.

Table 3: Correlations of rust disease severity caused by Melampsora apocyni and the listed variables to two ecotypes of Poacynum hendersonii

Variables	Eco RSM-red stem with medium leaves			Eco GSF-green stem with fine leaves
	Coefficient	Regression equation	SE	Coefficient	Regression equation	SE
Leaf dry weight	0.710**	Y=0.721-0.038x	0.067	0.453	Y=0.387-0.010x	0.030
Flavonoids concentration	0.621**	Y=2.563-0.168x	0.378	0.150	Y=2.177-0.044x	0.435
Alanine	0.820**	Y=1.336-0.105x	0.131	0.819**	Y=1.319-0.109x	0.114
Argnine	0.790**	Y=1.314-0.118x	0.164	0.879**	Y=1.202-0.113x	0.092
Aspartic	0.757**	Y=1.974-0.144x	0.222	0.722**	Y=1.895-0.122x	0.174
Cystine	0.589	Y=0.566-0.030x	0.074	0.852**	Y=0.606-0.039x	0.036
Glutamine	0.836**	Y=2.601-0.203x	0.238	0.851**	Y=2.444-0.192x	0.177
Glycine	0.796**	Y=1.163-0.093x	0.126	0.804**	Y=1.118-0.086x	0.095
Hlstidine	0.772**	Y=0.503-0.041x	0.060	0.812**	Y=0.477-0.039x	0.042
Isoleucine	0.775**	Y=1.066-0.079x	0.115	0.817**	Y=1.010-0.076x	0.080
Leucine	0.790**	Y=1.956-0.164x	0.227	0.856**	Y=1.884-0.165x	0.148
Lysine	0.850**	Y=1.304-0.103x	0.114	0.837**	Y=1.234-0.094x	0.092
Methionine	0.332	Y=0.147-0.013x	0.068	0.350	Y=0.158-0.008x	0.333
Phenylalanine	0.748**	Y=1.257-0.100x	0.159	0.837**	Y=1.206-0.096x	0.093
Proline	0.794**	Y=1.020-0.038x	0.106	0.846**	Y=0.982-0.071x	0.067
Serine	0.827**	Y=0.087-0.061x	0.074	0.613**	Y=0.860-0.051x	0.098
Threonine	0.810**	Y=1.006-0.076x	0.098	0.814**	Y=0.977-0.074x	0.079
Tryptophan	0.387	Y=0.154-0.009x	0.040	0.446	Y=0.132+0.010x	0.029
Tyrosine	0.895**	Y=0.726-0.071x	0.063	0.861**	Y=0.704-0.067x	0.059
Valine	0.776**	Y=1.353-0.101x	0.146	0.829**	Y=1.289-0.098x	0.098
Caconcentration	0.555*	Y=4.727-0.323x	0.864	0.610*	Y=8.563-0.910x	1.763
Cu concentration	0.816**	Y=0.186-0.025x	0.032	0.639*	Y=0.132+0.070x	0.126
Zn concentration	0.384	Y=0.790-0.055x	0.238	0.374	Y=2.142-0.180x	0.667
Fe concentration	0.235	Y=0.051+0.007x	0.049	0.594*	Y=0.166-0.022x	0.045

Note:SE=standard error; *P < 0.05, **P < 0.01; x=rust disease severity.

Fig. 6: Metal concentration of two ecotypes Eco RSM-red stem with medium leaves and

Eco GSF-green stem with fine leaves of P. pictum in cultivated field. The same lowercase letters up the bars means there is no significantly different by Turkey’s HSD at P<0.05 for Eco RSM; The same uppercase letters up the bars means there is no significantly different by Turkey’s HSD at P<0.05 for Eco GSF

Several research studies have found that the infection by pathogens increased plant flavonoids concentration in plant tissues. An example of this was that Miranda et al. (2007) using the Populus 15.5K cDNA microarray, found that genes encoding enzymes required for synthesis of the flavonoid proanthocyanidin were up regulated dramatically. Phytochemical analysis confirmed that in late infection, proanthocyanidin levels increased in infected leaves. Lu et al. (2017) also found that the amount of flavonoid compounds, especially anthocyanin and catechin, were significantly increased in rust-infected symptomatic tissue. The expression levels of structural genes and MYB transcription factors related to flavonoid biosynthesis were one to seven-fold higher in the tissue infected by rust.

The present study indicates that the two ecotypes of P. pictum had different responses to rust infection, as rust had no effects on flavonoids content of leaves of the GSF ecotype |but resulted in decreased flavonoids concentration in the leaves of the RSM ecotype. This decrease with the RSM ecotype is opposite to previous reports about the effects of rust disease on plant flavonoids concentration. The dry weight and flavonoids of the two ecotypes of P. pictum had the same response to rust. The accumulation of carbohydrates in rust-infected plants may be associated with the flavonoid biosynthesis pathway (Wan et al. 2015) and also, those carbohydrates which are a component of osmotic regulation during pathogen infection, may contribute to the accumulation of flavonoids (Lu et al. 2017). This new finding that rust decreases P. pictum flavonoids is important supplementary knowledge of the effects of rust disease on plant flavonoids as well as the evaluation of the resulting loss. It helps the understanding of physiological effects of this rust species as found in our previous study in a greenhouse that showed that the infection of rust changed activity of peroxidase, polyphenol oxidase and phenylalanine ammonialyase in A. venetum leaves (Gao et al. 2017b). Our hypothesis that rust will increase the total flavonoids of the two ecotypes of P. pictum was not upheld.

Many reports depicted that rust disease decreases the amino acid content of plants. For example, Nan found that the infection of rust (Uromyces onobrychidis) in Onobrychis viciaefolia, U. orobi in Vicia sativa and U. striatus in alfalfa (Medicago sativa) and U. baeumlerianus in Melilotus albus, decreased total crude protein and 16 amino acids by more than 30% (Nan 1986, 1990). Rust fungi only can complete their life cycle on living hosts where they grow through the leaf tissue by developing an extended network of intercellular hyphae from which intracellular haustoria are involved in suppressing host defense responses and acquiring nutrients. Three amino acid transporter genes of the rust fungi, Uf–AAT1, Uf–AAT2 and Uf–AAT3, are closely related with intracellular haustoria of rust fungi. AAT1 and AAT3 are expressed very early during rust development and are strongly up-regulated in haustoria (Struck et al. 2002, 2004) while AAT2 was shown to be strictly haustorium specific (Hahn and Mendgen 1997). The decrease in amino acid content with severe infection of rust disease in the two ecotypes of P. pictum may partly be due to the regulation of the expression of these three transporter genes by the rust pathogen, as well as the consumption, metabolism or the storage of amino acid by the pathogens (Hahn and Mendgen 1997; Struck et al. 2002, 2004). There is also research found that pathogens such as Pseudopeziza medicaginis in alfalfa decrease the content of amino acids in host plants (Morgan and Parbery 1980). Even a low level of infection by Drechslera siccans or Rhynchosporium spp. significantly reduced in vitro dry matter digestibility, and water-soluble carbohydrate and the total amino acid content of Italian ryegrass (Lolium multiflorum) and tall fescue (Festuca arundinacea), and the decreases are correlated with the nitrogen metabolism and transportation in the plant–pathogen system.

Essential and non–essential heavy metals, such as Cu and Zn, are quantified in selected medicinal plants, including A. venetum and P. pictum, which are extensively used in the preparation of herbal medicines for heart disease and tonics for general human health. Rust disease has been shown decreased the content of Ca and P in alfalfa (Nan 1986), but the mechanism is not clear. The defensive properties of metals against diseases of plants has received much attention and support (Fones et al. 2010; Fones and Preston 2013), an example of which is the supply of Ca and K reduces soybean rust (Phakopsora pachyrhizi) area under the disease progress curve (PUDPCS) (Pinheiro et al. 2011). The accumulation of metals in plants is correlated with the activity of metal transporters, e.g., repeated duplication of the gene encoding the P–type ATPase, HMA4, which is responsible for xylem loading of Zn and cadmium (Cd) (Hanikenne et al. 2008).

Conclusion

Rust disease was caused by Melampsora apocyni widely occur in field of cultivated P. pictum, and the disease incidence could reach up to 70% for the two ecotypes. The severe occurrence of rust disease led to the decrease of leaf biomass and flavonoids in the RSM ecotype, reduced the amino acid content of the two ecotypes, and increased or decreased Ca and Cu for the GSF and RSM ecotypes, respectively. These changes due to rust infection decreased the value of P. pictum for tea and Chinese traditional medicine production. Control methods for this rust disease are urgently required in this region.

Acknowledgements

This research was financially supported by Key Project of Science and Technology Department of Xinjiang Autonomous Region, China (2016E02015, 2016A03006).

Author Contributions

Tingyu Duan and Peng Gao designed the experiment and analyzed the data; Yanru Lan performed the experiments; Tingyu Duan wrote the manuscript.

References

Denchev CM (1995). Bulgarian Uredinales. Mycotaxon 55:405‒465

Fones H, CA Davis, A Rico, F Fang, JA Smith, GM Preston (2010). Metal hyperaccumulation armors plant against disease. PLoS Pathog 9; Article e1001093

Fones HN, GM Preston (2013). Trade-offs between metal hyperaccumulation and induced disease resistance in metal hyperaccumulator plants. Plant Pathol 62:63‒71

Gao P, ZB Nan, YN Wu, QT Liu, FJ Meng, ZC Xiao, TY Duan (2017a). Identification of the pathogen causing rust disease of traditional Chinese herb Apocynum venetum. J Plant Prot 44:129‒136

Gao P, TY Duan, MJ Christensen, ZB Nan (2017b). The occurrence of rust disease, and biochemical and physiological responses on Apocynum venetum plants grown at four soil water contents, following inoculation with Melampsora apocyni. Eur J Plant Pathol 150:549‒563

Gao P, ZB Nan, YN Wu, QT Liu, FJ Meng, ZC Xiao, TY Duan (2015). Characteristics of photosynthetic physiology and growth with 8 Luobuma Ecotypes in the Apocynum and Poacynum from Altay of Xinjiang, China. Acta Bot Bor-Occident Sin 35:2069‒2077

Gong JX, QY Zhang, T Zhang, Z Li, JF Zhang (2017). Investigation on bio-recalcitrance in bio degumming of Apocynum. J Textil Res 38:83‒87

Hahn M, K Mendgen (1997). Characterization of in planta-induced rust genes isolated from a haustorium-specific cDNA library. Mol Plant Microb Interact 10:427‒443

Hanikenne M, IN Talke, MJ Haydon, C Lanz, A Nolte, P Motte, J Kroymann, D Weigel, U Krämer (2008). Evolution of metal hyperaccumulation required cis-regulatory changes and triplication of HMA4. Nature 453:391‒395

Hiratsuka N (1939). Miscellaneous notes on the East-Asiatic Uredinales with special reference to the Japanese species. J Jap Bot 14:33‒38

Kang NN, ST Fang, Y Liu, JH Wang, CY Yang, HC Xia (2014). Flavonoids from the halophyte Apocynum venetum and their antifouling activities against marine biofilm-derived bacteria. Nat Prod Res 12:928‒931

Kirbağ S (2004). New records of microfungi from Turkey. Pak J Bot 36:445‒448

Ksenofontov AL, AI Boyko, GV Mkrtchyan, VN Tashlitsky, AV Timofeeva, AV Graf, VI Bunik, LA Baratova (2017). Analysis of free amino acids in mammalian brain extracts. Biochemistry 82:1183‒1192

Lü L, D Zhang, B Sun, YW Hu, MT Yan, K Liu (2017). Apocynum leaf extract inhibits the progress of atherosclerosis in rats via the AMPK/mTOR pathway. Die Pharm Intl J Pharm Sci 72:41‒48

Lu YF, Q Chen, YF Bu, R Luo, SX Hao, J Zhang, YC Yao (2017). Flavonoid accumulation plays an important role in the rust resistance of Malus plant leaves. Front Plant Sci 8:1‒13

Ma M, CL Hong, SQ An, B Li (2003). Seasonal, spatial, and interspecific variation in quercetin in Apocynum Venetum and Poacynum Hendersonii, Chinese traditional herbal teas. J Agric Food Chem 51:2390‒2393

Miranda M, SG Ralph, R Mellway, R White, MC Heath, JC Bohlmann, CP Constabel (2007). The transcriptional response of hybrid poplar (Populus trichocarpa × P. deltoides) to infection by Melampsora medusae leaf rust involves induction of flavonoid pathway genes leading to the accumulation of proanthocyanidins. Mol Plant-Microb Interact 20:816‒831

Morgan C, DG Parbery (1980). Depressed fodder quality and increased oestrogenic activity of lucerne infected with Pseudopeziza medicaginis. Aust J Agric Res 31:1103‒1110

Nan ZB (1990). Effects of rust on the growth and nutritive composition of forage legumes. Acta Pratacult Sin 1:83‒87

Nan ZB (1986). The influence of rust on the growth and nutrients of Vicia sativa. Gansu Agric Sci Technol 8:22‒26

Nevodovskii GS (1956). Cryptogamic flora of Kazakhstan, Vol. 1, pp:227–229. Kazakhstan Academy of Sciences, Alma-Ata, Kazakhstan

Nguyen NH, JH Kim, J Kwon, CY Jeong, W Lee, D Lee, SW Hong, H Lee (2016). Characterization of Arabidopsis thaliana Flavonol Synthase 1 (FLS1) -overexpression plants in response to abiotic stress. Plant Physiol Biochnol 103:133‒142

Nishibe S, T Takenaka, A Kodama (2001). Flavonoids from leaves of Poacynum pictum. Nat Med 55:38‒38

Pinheiro JB, EA Pozza, AAA Pozza, ADS Moreira, MC Alves (2011). Effect of potassium and calcium supplied via nutrient solution on the severity of Asian soybean rust. Rev Ceres 58:43‒50

Shang YY, Q Qin, MS Li, HY Xu, YB Ge (2017). Flavonol glycosides from the leaves of Elaeagnus pungens. Nat Prod Res 31:1066‒1072

Shi JY, GL Li, HL Wang, J Zheng, YR Suo, JM You, YJ Liu (2011). One-step separation of three flavonoids from Poacynum hendersonii by high-speed counter-current chromatography. Phytochem Anal 22:450‒454

Song RJ, J Zhou (2015). Microemulsion liquid chromatographic method for simultaneous separation and determination of six flavonoids of Apocynum venetum leaf extract. J Chromatogr B 995:8‒14

Struck C (2015). Amino acid uptake in rust fungi. Front Plant Sci 6; Article 40

Struck C, E Mueller, H Martin, G Lohaus (2004). The Uromyces fabae UfAAT3 gene encodes a general amino acid permease that prefers uptake of in planta scarce amino acids. Mol Plant Pathol 5:183‒189

Struck C, M Ernst, M Hahn (2002). Characterization of a developmentally regulated amino acid transporter (AAT1p) of the rust fungus Uromyces fabae. Mol Plant Pathol 3:23‒30

Tai FL (1979). Sylloge fungorum sinicorum. In: Sylloge Fungorum Sinicorum, p:534. Science Press, Academia Sinica, Peking, China

Tranzschel WA (1891). New kinds fungi of Melampsora spp., Vol. 3, pp:137‒140. Botanical Notes issued upon the Botanical Garden of the Imperial Saint Petersburg University, Russian Empire

Wan HH, J Zhang, TT Song, J Tian, YC Yao (2015). Promotion of flavonoid biosynthesis in leaves and calli of ornamental crabapple (Malus spp.) by high carbon to nitrogen ratios. Front Plant Sci 6; Article 673

Xie W, X Zhang, T Wang, JJ Hu (2012). Botany, traditional uses, phytochemistry and pharmacology of Apocynum venetum L. (Luobuma): A review. J Ethnopharmacol 141:1‒8

Yan SX, JL Lang, YY Song, YZ Wu, MH Lv, X Zhao, YH Liu, CY Xu (2015). Studies on anti-depressant activity of four flavonoids isolated from Apocynum venetum Linn (Apocynaceae) leaf in mice. Trop J Pharm Res 14:2269‒2277

Yokozawa T, Y Kashiwada, M Hattori, HY Chung (2002). Study on the components of Luobuma with peroxynitrite-scavenging activity. Biol Pharm Bull 25:748‒752

Zhang ZJ (2004). Therapeutic effects of herbal extracts and constituents in animal models of psychiatric disorders. Life Sci 75:1659‒1699

Zhao CQ, XX Zhao, J Zhang, W Zou, Y Zhang, L Li, J Liu (2016). Screening of bacillus strains from sun vinegar for efficient production of flavonoid and phenol. Ind J Microbiol 56:498‒503

Zhou J, JB Sun, XY Xu, ZH Cheng, P Zeng, FQ Wang, Q Zhang (2015). Application of mixed cloud point extraction for the analysis of six flavonoids in Apocynum venetum leaf samples by high performance liquid chromatography. J Pharm Biomed Anal 107:273‒279