Alkaloid Contents in Epichloë Endophyte-Infected Elymus tangutorum Sampled along an Elevation Gradient on the Qinghai-Tibetan Plateau

: Alkaloids produced by endophytic fungi can have an important inﬂuence on agricultural ecology, and can often be a ﬀ ected by climatic factors. At present, there are no studies that have assessed the relationship between alkaloid production and elevation or climatic factors in the Qinghai-Tibetan Plateau. To address this knowledge gap, we explored ergot alkaloid and peramine production in Epichloë -infected (E + ) Elymus tangutorum collected from the Qinghai-Tibetan plateau and assessed the relationship between the concentration of these alkaloids and elevation. The e ﬀ ects of temperature and precipitation on these relationships were also investigated. The concentrations of ergonovine, ergine, and peramine ranged from 0.47–0.84, 0.35–1.72 and 9.18–13.00 µ g · g − 1 , respectively. Fitted cubic regression curves describing an arc-pattern across the elevational series were statistically signiﬁcant for ergine and peramine concentrations. The elevational trend in peramine concentration was linked to mean daily temperature, while the ergine elevational trend was linked to mean precipitation. Our results provide a new understanding of the climatic factors that inﬂuence the alkaloid production of endophytic fungi at di ﬀ erent elevations in the Qinghai-Tibetan plateau. These results suggest that the di ﬀ erent alkaloid proﬁles relate to di ﬀ erent climate factors at the di ﬀ erent and increased peramine at warmer sites may reﬂect greater insect Our not only the knowledge gap to and elevation e ﬀ ects on alkaloid production in the also provides a new of the alkaloid production of endophytic under varying climatic conditions with di ﬀ erent elevations in the QTP. Further to explore how other environment factors such as soil nutrient status impact on alkaloid by Epichloë symbionts could be


Introduction
Interactions between plants and endophytes are a widespread phenomenon in nature and are crucial to the establishment and maintenance of populations, influencing community dynamics and even ecosystem function, particularly under environmental stress [1][2][3][4]. Between 20-30% of cool-season grass species can form symbiotic associations with Epichloë endophytes [5]. Asexual endophyte species of the genus Epichloë, formerly known as Neotyphodium, usually form symbioses with cool-season grasses [6]. Asexual Epichloë are vertically transmitted through host plant seeds, and do not normally cause symptoms in the host grass [7][8][9].
Elymus spp. not only provide high quality herbage for livestock grazing [36], but they are also important grass species for ecological restoration of grasslands and for reducing land desertification in western China [22,37]. They often contain the endophytic fungus Epichloёbromicola [38], and the E. bromicola associations with Elymus spp. are host specific [39]. Endophyte-infected Elymus dahuricus was found to produce peramine in 21 sites across China [40]. No cases of toxicity to livestock grazing on Elymus spp. have been reported, although ergot alkaloid gene sets have been identified in E. bromicola [41]. There are many Elymus spp. in China and little is known concerning their ergot alkaloid and peramine production potential, especially in the Qinghai-Tibetan plateau (QTP) where natural grasslands play an important role in the ecology and economy of the region.
The QTP covers a large area that includes complex landscapes, climatic gradients and different elevations, and contains rich biodiversity [42]. Elevation often has an influence on the ecosystems of the QTP because it alters a range of climate factors including temperature, precipitation and atmospheric pressure that shape the evolutionary adaptation of plant species. [43,44]. Prior research on the QTP has investigated the effects of elevational gradients on diversity [45], evolutionary history [44,46], underlying adaptation [47], geographical distribution [48] and the chemical composition of plants [49,50], and fungal assemblages [51,52]. However, there are no studies that assess the relationship between the alkaloid production of endophyte-infected cool-season grasses and elevation and climate factors across the QTP.
Interestingly, previous studies have reported that climate factors (mainly temperature, precipitation and CO 2 ) affect fungal alkaloid concentration [53][54][55][56][57][58][59][60][61][62]. For example, McCulley et al. [56] found that ergot alkaloids of tall fescue (Festuca arundinacea) significantly increased at higher temperature in transition zone pastures in the U.S. Drought stress can increase the production of ergovaline in Epichloë-infected grass [54,59]. Ryan et al. [58] found that the alkaloid concentration of endophyte-infected tall fescue decreased under elevated CO 2 . The majority of studies on the influence of climate factors on the alkaloid production of Epichloë endophytes have been conducted under controlled conditions, and relevant studies conducted in natural conditions are lacking.
To address this knowledge gap, we explored ergot alkaloid and peramine production in Epichloë-infected (E+) E. tangutorum collected from 25 sites across the QTP and assessed the relationship between the concentration of these alkaloids and elevation. The climate factors that affect the relationships were also investigated.

Characterization of the Study Area
The QTP ranges from the eastern edge of the Hengduan Mountains to the western boundary of the Pamir Mountains, and from the northern edge of the Kunlun Mountains to the southern edge of the Himalayan Range [63], with latitude 26 • 00 -39 • 47 N and longitude 73 • 19 -104 • 47 E. Our sampling sites were all distributed on permanent grassland of the QTP in Qinghai and Gansu Provinces; the elevation ranged from 2100 m to 3920 m, and the longitude ranged from 99 • 21 47 ' to 102 • 07 19", while the latitude ranged from 34 • 32 56" to 38 • 57 35" (Figure 1). A highland cold climate prevails in this region [64]. In winter, the mean daily temperature remains below 0 • C for nearly six months, while the mean daily temperature in summer typically ranges from 10-20 • C [65]. The mean annual precipitation is about 410 mm, with approximately 140 mm in the cold months (November-January), 90 mm in early spring (February-April), and the remainder from May to October [66].
Agronomy 2020, 10, x FOR PEER REVIEW 2 of 14 precipitation is about 410 mm, with approximately 140 mm in the cold months (November-January), 90 mm in early spring (February-April), and the remainder from May to October [66].

Plant Material
From late August to early October in 2017, mature individual plants of E. tangutorum with fully ripened seed-heads were collected from the 25 different sites distributed on permanent grassland of the QTP in Qinghai and Gansu Provinces ( Figure 1). For sites located at lower altitude (below ca. 3800 m), 20-40 plants were sampled at each site. Fewer plants (7)(8)(9)(10)(11)(12) were sampled at the remaining sites due to low E. tangutorum abundance. For each site, tillers of individual plants were cut approximately 3 cm from the soil surface. Each plant sample was packed separately into envelopes and returned to the laboratory for analysis. The geographic coordinates and altitude of each site were taken during sampling. The endophyte-infected (E+) E. tangutorum was identified from populations by using the aniline blue staining method of Cheplick [67] on culms of each individual plant. E+ E. tangutorum plants were identified in each of the 25 sites.

Determination of Ergot Alkaloid and Peramine Concentration
Standards of ergonovine, ergine and peramine were provided by Dr Wade Mace, AgResearch Limited, Grasslands Research Centre in New Zealand. Aboveground parts of individual plants were used to determine the alkaloid concentration.
A 200 mg subsample of dried plant material collected from each site was used to measure the ergot alkaloid concentration. Ergot alkaloid concentrations were determined by using a HPLC method adapted from Zhang et al. [68]. An Agilent 1100 HPLC (Agilent, Santa Clara, CA 95051, USA), fitted with a C18 column (Eclipse XDB-C18, 250 mm × 4.6 mm, 5 µm) was used to quantify the ergot alkaloids. The mobile phases used were (A) 0.1 M NH4OAc, and (B) CH3CN: 0.1 M NH4OAc, 3:1. The flow rate was 1 mL·min −1 . Detection was performed with an ultraviolet wavelength spectrophotometric detector (Agilent G1314A, Santa Clara, CA, USA) set to 312 nm. The quantity of

Plant Material
From late August to early October in 2017, mature individual plants of E. tangutorum with fully ripened seed-heads were collected from the 25 different sites distributed on permanent grassland of the QTP in Qinghai and Gansu Provinces ( Figure 1). For sites located at lower altitude (below ca. 3800 m), 20-40 plants were sampled at each site. Fewer plants (7)(8)(9)(10)(11)(12) were sampled at the remaining sites due to low E. tangutorum abundance. For each site, tillers of individual plants were cut approximately 3 cm from the soil surface. Each plant sample was packed separately into envelopes and returned to the laboratory for analysis. The geographic coordinates and altitude of each site were taken during sampling. The endophyte-infected (E+) E. tangutorum was identified from populations by using the aniline blue staining method of Cheplick [67] on culms of each individual plant. E+ E. tangutorum plants were identified in each of the 25 sites.

Determination of Ergot Alkaloid and Peramine Concentration
Standards of ergonovine, ergine and peramine were provided by Dr Wade Mace, AgResearch Limited, Grasslands Research Centre in New Zealand. Aboveground parts of individual plants were used to determine the alkaloid concentration.
A 200 mg subsample of dried plant material collected from each site was used to measure the ergot alkaloid concentration. Ergot alkaloid concentrations were determined by using a HPLC method adapted from Zhang et al. [68]. An Agilent 1100 HPLC (Agilent, Santa Clara, CA 95051, USA), fitted with a C 18 column (Eclipse XDB-C 18 , 250 mm × 4.6 mm, 5 µm) was used to quantify the ergot alkaloids. The mobile phases used were (A) 0.1 M NH 4 OAc, and (B) CH 3 CN: 0.1 M NH 4 OAc, 3:1. The flow rate was 1 mL·min −1 . Detection was performed with an ultraviolet wavelength spectrophotometric detector (Agilent G1314A, Santa Clara, CA, USA) set to 312 nm. The quantity of extracted sample injected into the injection port was 20 µL. Ergot alkaloid concentration was quantified using external standard curves.
For peramine analysis, a 50 mg subsample of freeze-dried plant material was used. Peramine was extracted following the methods of Zhang and Nan [40]. The HPLC machine, column, flow rate and detector used to quantify the peramine concentration were as described above for the determination of ergot alkaloids. For peramine detection, the mobile phases were (A) 1.8 g L −1 guanidine carbonate and (B) acetonitrile, and the spectrophotometric detector was set to 280 nm. The quantity of extracted sample was 25 µL. Peramine concentration was quantified using an external standard curve.

Collection of Climate Data
Climatic data was collected from the National Meteorological Data Center (http://data.cma.cn). According to the latitude, longitude, and elevation of the sampling sites, we used the thin-plate smoothing spline algorithm implemented in the Anusplin package (Version 4.4, Canberra, Australia; http://fennerschool.anu.edu.au/files/anusplin44.pdf) for interpolation to obtain the mean daily temperature (MDT) and mean precipitation (PCP) in the growing season of each site for the period 2006-2015.

Statistical Analysis
Concentrations (average ± standard error) of the Epichloë alkaloids ergonovine, ergine and peramine were calculated using SPSS (Version 24.0, Chicago, IL, USA). Regression analysis was employed to assess the relationships between the elevation and ergot alkaloid and peramine concentrations. The Pearson correlation coefficients were determined as a first assessment of the relationship between elevation, MDT and PCP. Canonical correlation analysis (CCA) using SPSS 24.0 was used to further explore the relationship between these three measures of plant environment and to examine their association with the observed concentrations of Epichloë alkaloids in the foliage of the E. tangutorum host. CCA explores the relationship between two groups of variables, in this case the concentration of ergonovine, ergine and peramine on the one hand, and the site environment factors, elevation, mean daily temperature and mean precipitation, on the other hand. With three input variables in each group, three canonical correlations with progressively decreasing information content are available. Each of these canonical correlations is for a pair of canonical variates that are linear functions of the alkaloid and environment variables, respectively, formed from canonical coefficients identified in the CCA. Here we refer to the three canonical variates derived from the alkaloid data by CCA as v 1 , v 2 , and v 3 ; and the three canonical variates derived from the environment data as u 1 , u 2 , and u 3 . Data were standardized prior to performing CCA.

Climate Factors and Relationships with Elevation
The latitude, longitude and elevation of the 25 sampling sites, and the interpolated mean daily temperature and mean precipitation are presented in Table 1. In general, across the sampling sites, precipitation was lower in the east, while altitude was higher to the south. Within the elevation range of the study (2100-3920 m), the mean daily temperature and mean precipitation ranged from 9.1-15.6 • C and 182-471 mm, respectively. Not unexpectedly, mean daily temperature was significantly negatively correlated with elevation (r = −0.715, p < 0.001), while precipitation displayed a marginally significant positive correlation with elevation (r = 0.392, p = 0.053) ( Table 2).  Table 2. Pearson correlation (R) between elevation and mean temperature and rainfall during the growing season across the 25 sites in Figure 1.

The Relationship between Alkaloid Concentration and Elevation
Ergine concentrations were highest at mid altitudes (Figure 2b), while peramine concentration was observed to decline with increasing elevation above 3000 m (Figure 2c). Both of these patterns were found to be statistically significant on fitting of cubic regression curves across the elevational series where E is elevation. In contrast, ergonovine concentration displayed no elevational trends (r 2 ergonovine = 0.018, p = 0.817) (Figure 2a).

The Relationship between Alkaloid Concentration and Elevation
Ergine concentrations were highest at mid altitudes (Figure 2b), while peramine concentration was observed to decline with increasing elevation above 3000 m (Figure 2c). Both of these patterns were found to be statistically significant on fitting of cubic regression curves across the elevational series (r 2 ergine = 0.482, p = 0.003; r 2 peramine = 0.726, p < 0.001). The respective equations were: where E is elevation. In contrast, ergonovine concentration displayed no elevational trends (r 2 ergonovine = 0.018, p = 0.817) (Figure 2a).

The Relationship between Alkaloid Concentration and Climatic Factors
When the data were submitted to CCA, the first two of the three available pairs of canonical variates (designated here as Canonical 1 and Canonical 2) displayed statistically significant canonical correlations (r = 0.690, p = 0.006; r = 0.603, p = 0.038, respectively), and between them explained 97.2% of the data variation (Table 3). Hence, these first two canonical correlations are reported in the results that follow and the third canonical correlation was discarded. Inspection of the canonical coefficients (Table 4) and the correlations between the original variables and their canonical variates (Table 5) shows that the first pair of canonical variates identifies a tendency for all three alkaloids (but especially ergine and peramine) to be present at higher concentrations at sites with lower elevation and precipitation (with correlations between the original data and canonical variate of −0.696 and −0.527, respectively). Temperature is not involved in this relationship as the correlation between mean daily temperature and the first climate canonical variate (designated u1 in Table 5) is only 0.034. Meanwhile, the second pair of canonical variates identifies a tendency for peramine in particular (r = 0.604) to be present at higher concentrations at sites of lower elevation (r = −0.710) with a higher mean daily temperature (r = 0.912).

Discussion
The production of ergot alkaloids in a number of Epichloë-infected grass species, throughout the world, is responsible for mammalian toxicoses [26]. In China, endophytic fungus-infected drunken horse grass (Achnatherum inebrians) can produce ergonovine and ergine, which can lead to livestock toxicity. The ergonovine and ergine concentrations in drunken horse grass can be as high as 120-280 µg·g −1 and 45-170 µg·g −1 , respectively [68]. In this study, the total ergonovine and ergine alkaloid concentration was less than 2.30 µg·g −1 (Figure 2). In the QTP, the absence of toxicity to livestock grazing on E. tangutorum may be due to the low level of ergot alkaloid production in the grass, although the toxicity threshold of ergine and ergonovine remains undefined in the literature. In this study, ergonovine, ergine and peramine were detected in Epichloë-infected E. tangutorum collected from different elevations. Peramine concentration in endophyte-infected E. tangutorum was much higher than that of ergonovine and ergine alkaloids ( Figure 2, Table 1).
Superficially, ergine and peramine concentrations displayed a statistically significant arc-shaped trajectory along the elevational gradient with the highest concentrations in E. tangutorum foliage observed at mid-elevation and the lowest concentrations at higher elevation (p < 0.05) (Figure 2). An elevation gradient involves associated changes in various climatic factors, especially temperature and moisture [69]. A highland cold climate prevails [64] in our sampling sites on the QTP, and the mean daily temperature during the growing season ranged from 9.1-15.6, while the mean precipitation ranged from 182-471 mm in the growing season (Table 1). These two climate factors were significantly correlated to elevation (Table 2). Canonical correlation analysis was able to resolve the superficial relationship between alkaloid concentration and elevation into two independent linear components that together explain 97.2% of the canonical data variation, with 31.4% of the standardized variance in the alkaloid data explained by the environment canonical variates (Tables 3-5). The first pair of canonical variates linked the increased concentration of all three alkaloids to sites at lower elevation with lower precipitation, independently of temperature. The second pair of canonical variates linked the increased peramine concentration to sites at lower elevation with a higher mean daily temperature, independently of precipitation. Many previous studies have examined the association between temperature and precipitation (an index of plant water supply) and alkaloid production (for example, [53][54][55][56][57][58][59][60][61][62]70,71]). In particular, the studies conducted by Repussard et al. [71] andŻurek et al. [62] were also carried out in natural grassland areas. Repussard et al. [71] found that ergovaline concentration of endophyte-infected Festuca arundinacea was positively correlated to cumulative temperature in the south of France. A similar conclusion was found in our study, in that peramine concentration was positively correlated to mean daily temperature. However, among the studies to date, there is no consensus as to the effects of the environment on Epichloë alkaloid concentration in the host grass. Zhou et al. [61] analyzed the relationship between temperature and the ergot alkaloid concentrations of Festuca sinensis and showed that ergot alkaloids significantly increased as temperature decreased. Interestingly, McCulley et al. [56] found ergot alkaloid concentration increased (by 30-40%) in Epichloë-infected tall fescue (F. arundinacea) under higher temperature in transition zone pastures of the U.S., but loline alkaloid concentration was not affected. Many studies have indicated that sufficiency of water is not conducive to the production of alkaloids [54,59,60,62].Żurek et al. [62] found that higher amounts of ergovaline produced by endohyte-infected tall fescue were much more frequent in regions of lower summer precipitation. Similar conclusions with regard to precipitation were also reached by Vazquez-de-Aldana et al. [60], who suggested that lower ergovaline production can be linked to higher precipitation. However, McCulley et al. [56] and Bourguignon et al. [53] concluded that precipitation had no effect on alkaloid levels.
To resolve the conflicting conclusions of various studies cited above, we hypothesize that the Epichloë-host relationship has evolved alkaloid concentration responses that maximize protection to the host from biotic stressors such as insect predation and attack by pathogenic fungi, while minimizing the metabolic cost of alkaloid production by limiting alkaloid synthesis when the plant is less exposed to biotic stress. Under this hypothesis, the findings from the present study that more ergine and peramine were produced at lower elevation sites with lower mean precipitation, and more peramine was produced at lower elevation sites with higher mean daily temperature are intuitively sensible and align with a number of the studies cited above. The results also indicate that future studies could include the collection of data on fungal pathogen and insect predation loads with a view to clarifying whether such factors are mechanisms by which climate factors influence Epichloë alkaloid concentrations in the host grass. Also, factors such as the effect of winter cold at higher elevations on the biotic stress load of the plant population could be assessed. Interestingly, a recent global analysis found that insect herbivory is reduced with increasing elevation [72,73], which could, in turn, select for reduced need for defense in high elevation plants [73]. Reduced insect herbivory might be a factor in the decreased concentration of peramine in host plants at high elevation sites.
Previous studies have indicated that in addition to temperature and precipitation, CO 2 concentration is also an important factor affecting alkaloid concentration [58,74,75]. CO 2 affects alkaloid production by producing carbohydrates for plant growth and the synthesis of alkaloids [75]. Elevated CO 2 reduces the concentrations of ergot alkaloid and loline produced by E+ tall fescue [74]. A similar result was found by Ryan et al. [58], where the alkaloid concentration of E+ tall fescue decreased under elevated CO 2 . However, Hunt et al. [75] found that elevated CO 2 had only a marginally positive effect on peramine and ergovaline production under high N conditions. Low air pressure is one of the plant environment characteristics of the QTP [63], with the air pressure at 3500 m elevation being approximately 70% of that at sea level. Carbon may be preferentially allocated to plant growth due to the carbon limited biosynthesis (under ambient CO 2 ) [75]. Therefore, reduced atmospheric partial pressure of CO 2 may be a possible explanation for the lower ergot alkaloid and peramine concentrations at higher elevation in this study. The impact of CO 2 concentration on ergot alkaloid and peramine concentration at high elevation could be assessed in the future in studies specifically designed for that purpose.
Beyond climate factors, soil nutrient status and symbiont genotype also have an influence on alkaloid production, and regional genetic variation in the host grass or Epichloë endophyte may also be responsible for alkaloid variation [75][76][77][78]. Further possible explanations for the different responses of the three measured alkaloids to different combinations of elevation, mean daily temperature and mean precipitation include differences in soil nutrient status, and genotype or genetic variation in the Epichloë symbiont or the E. tangutorum host. While further exploration of these factors is obviously highly relevant to building an improved understanding of the factors driving alkaloid production in the Epichloë-Elymus symbiosis, the present study did not collect relevant data so these must remain as points for future study.
To our knowledge, this is the first study to evaluate the elevational trends in alkaloid production in the native cool-season grass, E. tangutorum, on the QTP. There are a number of factors that co-vary with elevation in the QTP and cannot be easily disentangled. In this study, our main findings were that ergine alkaloid and peramine concentrations in endophyte-infected E. tangutorum were highest at mid-elevation and lowest at high elevation and that the peramine elevational trend in the QTP was driven by mean daily temperature while the ergine elevational trend was driven by mean precipitation. These results suggest that the different alkaloid profiles relate to different climate factors at the different sites, and increased peramine at warmer sites may reflect greater insect challenge. Our study not only addresses the knowledge gap relating to climate and elevation effects on alkaloid production by Elymus-Epichloë in the QTP, but also provides a new understanding of the alkaloid production of endophytic fungi under varying climatic conditions with different elevations in the QTP. Further research to explore how other environment factors such as soil nutrient status impact on alkaloid production by Epichloë symbionts could be worthwhile.