Effect of Cultivation Conditions on Components of Ephedra sp. Using Liquid Chromatography-Mass Spectrometry and Multivariate Analysis

Abstract

In this study, we investigated the correlation between the cultivation conditions and chemical composition of Ephedra sinica and E. sp. (denoted EP-13, which has been grown at the National Institutes of Biomedical Innovation, Health, and Nutrition for many years). The total contents of ephedrine and pseudoephedrine are specified in the Japanese Pharmacopoeia; therefore, we investigated the changes in their content under different cultivation conditions, including varying soil conditions and fertilization or the lack of fertilization. Poor growth due to low soil nutrition and lack of sunlight caused decrease of the alkaloid content. As expected, the plants accumulated proline, although the proline content varied considerably with cultivation location. The proline concentration correlated with the content of methanoproline. Moreover, a new compound, namely N,N-dimethyl-p-hydroxyphenylethylamine-O-[β-D-glucopyranosyl-(1→3)-α-L-rhamnopyranoside], was isolated from E. sinica but was absent in EP-13. This study on the correlation between cultivation methods and the alkaloid content in Ephedra is expected to assist in the future production of quality Ephedra herb.

Introduction

Plants have highly diverse chemical compositions, which can be partly explained by the differences in their growth environment. The composition of wild plants is dependent on the local climate, temperature, and soil conditions. Cultivated plants can be sensitive to the cultivation conditions employed, such as the fertilizers and pesticides or the planting density.

Most crude drugs are derived from plants, and their production can involve the processing of various parts of the plant (roots, leaves, fruits, bark, flowers, and seeds). Crucially, the chemical composition of the drug is highly dependent on that of the plant, which can differ considerably between plants of different species (e.g., Glycyrrhiza. uralensis or G. glabra in the case of Glycyrrhiza), place of origin, and quality or grade. Despite this variation, crude drugs are often marketed using similar names, which can result in drugs of the same name having significantly different properties.

Furthermore, because crude drugs are natural products, the final product varies significantly in quality. Hence, the Japanese Pharmacopoeia (JP) gives broad descriptions of crude drugs rather than describing them in narrow terms.

Plants are multicomponent systems; therefore, it is difficult to describe their properties with respect to one or two components. Multivariate analysis can be used to detect and analyze correlations and causal relationships between multiple variables when multiple experimental observations are obtained from a sample or samples. The theory of multivariate analysis was first developed in the late 19th century and has since been applied in many fields, though it has only recently been applied to plant components using LCMS and GCMS data. In particular, multivariate analysis, specifically, principal component analysis (PCA) of these data enables complex correlations between growth conditions and chemical composition for different plant varieties to be visualized in lower-dimensional space. A range of techniques can be used to obtain data for the multivariate analysis of phytoconstituents, including GCMS, LCMS, NMR, near infrared (NIR) spectroscopy, and HPLC; although, NMR¹⁾ and LCMS²⁾ data are most often reported. The disadvantage of GCMS³⁾ is that the detection is limited to volatile and low-polarity components. When using NMR data, the range is expanded to include highly polar components such as glycosides, although these can be difficult to identify in small quantities. LCMS is capable of detecting a relatively wide range of components, including those present in trace amounts, allowing for systems to be analyzed based on many components. However, the alignment process for LCMS data is important, especially when the mobile phase is used under gradient conditions because subtle shifts in retention times can occur.

We previously reported the identification of nitric oxide production inhibitors in ginger and Scutellaria root by multivariate discriminant analysis using individual differences in the components of these crude drugs.^2,4)

In this study, we focus on the crude drug Ephedra herb, which is comprised of dried Ephedra sp. plants. In the 18th edition of JP (JP18), Ephedra herb is defined as the terrestrial stem of Ephedra sinica Stapf, E. intermedia Schrenk et C. A. Meyer, and E. equisetina Bunge (Ephedraceae), though the most marketed products are derived from E. sinica. According to the JP18, Ephedra herb must also contain a total alkaloid (ephedrine [EP] and pseudoephedrine [PEP]) content of at least 0.7%.⁵⁾

For Ephedra herb, the relationship between the alkaloid content and growing conditions is not simply a factor of the soil nitrogen content, but it also involves a combination of factors that alter the growing environment during cultivation. The relationship between the alkaloid content and cultivation conditions has been investigated previously,⁶⁾ and Mikage and colleagues have published⁷⁾ several reports on crude drugs derived from Ephedra. Importantly, to promote the domestic cultivation of Ephedra herb, which is not native to Japan, the alkaloid content must meet the requirements of the JP; thus, detailed studies on the correlation between the alkaloid content and cultivation conditions are needed.

The Tanegashima Division of the Research Center for Medicinal Plant Resources (RCMPR), National Institutes of Biomedical Innovation, Health, and Nutrition (NIBIOHN) is located in the center of Tanegashima Island in Kagoshima Prefecture, Southern Japan. Large fields covering 108693 m² are available for cultivation, and the island has a mild climate with an annual average temperature of 19.5 °C. A sample of Ephedra labeled EP-13 (Registry No. of RCMPR: No. TS0381-79) has been preserved for many years at RCMPR and was introduced from Washington D.C. as E. distachya; however, its origin was initially doubted due to its morphology and other factors. EP-13 is easy to cultivate and propagate and has been used for cultivation research into Ephedra at many research institutes in Japan.^8–10) Ando et al. performed a gene-based analysis of EP-13 and proposed that it exists as a hybrid of E. likiangensis and E. gerardiana.¹⁰⁾

In this study, we examined the relationship between the cultivation environment and alkaloid content using E. sinica and EP-13 under three different soil and light conditions on Tanegashima Island. Plants propagated from the same parent can have different components owing to differences in their growth environment; thus, we investigated the origin of these differences using the multivariate analysis of LCMS data.

Results and Discussion

Cultivation Conditions of Ephedra Samples

Table 1 shows the specific conditions of the three fields located on Tanegashima Island used for the study. The fields differed in their soil properties, particularly water retention and nutrient content, and sunlight conditions. Plants of E. sinica and EP-13 were cultivated for either three or four years in each of the three fields (Fig. 1, Table 2), and all samples were harvested between 26 October and 2 November 2020. In addition, samples of Ephedra marketed as crude drugs (JP grade, Table 3) were used for comparison. Harvested samples were analyzed using LCMS.

Table 1. Special Features of Each Field in the Tanegashima Division of NIBIOHN

Experimental Field	Daylight	Soil Texture	Drainage	Water Retention	Soil Nutrition
A	Good	Clay	Poor	Poor	Rich in exchangeable K+, Poor SOM*
B	Good	Loam	Good	Good	Low Mg2+, Ca2+
C	Poor	Sand	Good	Poor	Poor nutrition

*SOM: Soil organic matter.

Fig. 1. Cultivation Fields in Tanegashima Division, RCMPR, NIBIOHN Used for This Study

Table 2. Ephedra Samples for Analysis

2017 Planting E. sinica and EP-13 (Cultivated for 4 Years)
No.	Species	Cultivation Field	Additional Fertilizer*	No.	Species	Cultivation Field	Additional Fertilizer*
1	E. sinica	A	—	28	EP-13	A	—
2			—	29			—
3			—	30			—
4			July	31			July
5			July	32			July
6			July	33			July
7			Sep.	34			Sep.
8			Sep.	35			Sep.
9			Sep.	36			Sep.
10		B	July	37		B	July
11			July	38			July
12			July	39			July
13			Sep.	40			Sep.
14			Sep.	41			Sep.
15			Sep.	42			Sep.
16			—	43			Sep.
17			—	44			—
18			—	45			—
19		C	July	46			—
20			July	47		C	July
21			July	48			Sep.
22			—	49			Sep.
23			—	50			—
24			—
25			Sep.
26			Sep.
27			Sep.

* Additional fertilizer of Urea; July and Sep. are added an additional fertilizer in July and September, respectively. Multiple plants were sampled at each fertilizer application season in each field. — means no additional fertilizer was added.

2018 Planting E. sinica and EP-13(Cultivated for 3 Years)
No.	Species	Cultivation field	Additional fertilizer*	No.	Species	Cultivation field	Additional fertilizer*
72	E. sinica	A	—	105	EP-13	A	—
73			—	106			—
74			July	107			—
75			July	108			July
76			July	109			July
77			July	110			July
78			July	111			Sep.
79			Sep.	112			Sep.
80			Sep.	113			Sep.
81			Sep.	114		B	July
82			Sep.	115			July
83			—	116			July
84		B	July	117			—
85			July	118			—
86			July	119			—
87			July	120			Sep.
88			—	121			Sep.
89			—	122			Sep.
90			—	123		C	—
91			Sep.	124			July
92			Sep.	125			July
93			Sep.	126			July
94			Sep.	127			—
95		C	July	128			—
96			July	129			Sep.
97			July	130			Sep.
98			—	131			Sep.
99			—
100			—
101			—
102			Sep.
103			Sep.
104			Sep.

* Additional fertilizer of Urea; July and Sep. are added as additional fertilizer in July and September, respectively. Multiple plants were sampled at each fertilizer application season in each field. — means no additional fertilizer was added.

Table 3. Commercial Ephedra Crude Drugs

No.	Registry No. of NIBIOHN	Remark	Produced area	Original plant*
133	NIB-0053	—	Inner Mongolia	E. sinica
134	NIB-0088	Cultivated	Inner Mongolia	E. sinica
135	NIB-0104	—	Inner Mongolia	E. sinica
136	NIB-0141	Wild	Inner Mongolia	E. sinica
137	NIB-0166	—	Inner Mongolia	—
138	NIB-0209	—	Inner Mongolia	E. sinica
139	NIB-0210	—	Gansu	E. intermedia
140	NIB-0216	—	Inner Mongolia	—

*http://mpdb.nibiohn.go.jp

Effects of Species, Growth Conditions, and Age on Alkaloid Content in Ephedra Samples

The EP and PEP content in each sample was quantified by HPLC, as shown in Fig. 2, after being dried, ground and extracted with 50% methanol following the JP18 standard performed at half scale. Generally, the alkaloid content in plants of E. sinica was lower than that in EP-13, which is in agreement with previous reports that the alkaloid content is lower in E. sinica than in other species.¹¹⁾ Furthermore, most E. sinica samples did not meet the minimum content of 0.7% specified in JP18, and the EP and PEP contents of the plants grown in each of the three fields decreased in the order: B > A > C. For EP-13, the plants from fields A and B had very similar contents of EP and PEP, and their total alkaloid content was above 0.7%. However, similarly to those of E. sinica, EP-13 plants from field C had the lowest alkaloid contents, with neither the 3-year, nor the 4-year plants of either species reaching the total alkaloid content of 0.7%. Field C had the poorest soil nutrition and sunlight conditions, resulting in the stunted growth of the plants; thus, these results are consistent with previous reports that alkaloid content is reduced under poor growth conditions.¹¹⁾ No notable change in alkaloid content was observed between the 3-year and 4-year plants, which was perhaps slightly unexpected since the alkaloid content is thought to decrease with increasing age owing to lignification.¹²⁾ However, lignification was not observed in the 4-year samples, and it is possible that a longer cultivation period is needed to discern the full effects of this phenomenon. The extended cultivation period was also suggested by Hiyama et al., who found no notable change in the alkaloid content of E. sinica plants that were cultivated over an eight year period.¹²⁾

Fig. 2. Total Alkaloid Contents of Cultivated Samples (Sample Name Abbreviation: Species–Field–Years Old; CCD: Commercial Crude Drug)

(Multiple plants were sampled at each fertilizer application season in each field (see Table 2). J: Additional fertilizer (Ceracoat R50) in July. S: Additional fertilizer (Ceracoat R50) in September. n: no additional fertilizer. Sample Nos. correspond to Table 2. n = 1–5, mean ± S.D. for n = 3 or more.)

Multivariate Analysis of Ephedra Samples Using LCMS Data

PCA of all Samples: Effect of Cultivation Period and Location on Chemical Composition and the Presence of Proline

Using same extracts for determination of EP and PEP contents, the 3- and 4-year cultivated Ephedra samples were analyzed by LCMS and PCA was performed on the obtained data. In this study, we used a pentafluorophenyl (PFP) column was used with a combination of 10 mM ammonium acetate and methanol as a mobile phase, as previously reported for EP and PEP separation and analysis.¹³⁾ The LCMS total ion chromatogram (TIC) shows notably different profiles for the samples of different species and cultivation locations (Fig. 3). In particular, the PCA score plot (Fig. 4A) shows good separation between Ephedra samples of different species and age, and between the cultivated samples and commercial products (commercial crude drug, CCD). The loading plot (Fig. 4B) shows that CCD and EP-13 are separated relative to E. sinica because of their high EP contents. Further, two peaks were identified in the MS spectrum for E. sinica, 1 (m/z 474.23, t_R 9.95 min (see below)) and 2 (m/z 116.07, t_R 4.16 min), which could be considered among the components that characterize the species. Notably, both 1 and 2 were less abundant in EP13 and CCD.

Fig. 3. TIC of LCMS Data for Cultivated and Commercial Samples

Fig. 4. PCA Score Plot (A) and Loading Plot (B) for all Samples (Abbreviations: CCD: Commercial Crude Drug; t_R: Retention Time; EP: Ephedrine)

PCA and Discriminant Analysis (Orthogonal Partial Least Squares Discriminant Analysis, OPLS-DA) of E. sinica

The PCA score plot shows good separation between the E. sinica plants cultivated in different fields. Therefore, we performed PCA on the data for E. sinica only (Fig. 5).

Fig. 5. PCA Score Plot (A) and Loading Plot (B) of all E. sinica Samples (t_R: Retention Time; EP: Ephedrine)

The PCA score plot for E. sinica shows distinct groups for each field, and the loading plot (Fig. 5B) shows that the peak for compound 2 was lower for the plants grown in field C than those grown in the other fields. Therefore, the peak areas for compound 2 in the extracted ion chromatogram (XIC) were compared (Fig. 6). Clearly, 2 was more abundant in the Ephedra plants grown in fields A and B than those grown in field C, which is consistent with the PCA results. Based on the high-resolution mass spectrometry (HRMS) analysis, the molecular formula of 2 is C₅H₉O₂N, and it is assumed to be a cyclic amino acid based on its degree of unsaturation. In the LCMS comparison of the spectrum with that of the standard product, the retention time and tandem MS (MS/MS) data (116→70) confirmed that 2 is proline. In addition, the data for all samples were grouped for each field, and orthogonal projections to latent structures discriminant analysis (OPLS-DA) was performed (Fig. 7). In this case, the respective values for R2VY[2](cum) and Q2VY[2](cum) (cumulative fraction for the variation of the Y variable, where values close to 1.0 indicate reliable data) were 0.902 and 0.71 for field A, 0.859 and 0.61 for field B, and 0.98 and 0.89 for field C, respectively. Thus, the correlations between the growth location and components were all highly reliable.

Fig. 6. Comparison of the Areas of the Peak at m/z 116.07 (t_R 4.1 min) in the XICs of Cultivated and Commercial Samples

(Sample name abbreviation: species–field–years old. CCD: commercial crude drug) (Multiple plants were sampled at each fertilizer application season in each field (see Table 2). J: Additional fertilizer (Ceracoat R50) in July. S: Additional fertilizer (Ceracoat R50) in September. n: no additional fertilizer. Sample Nos correspond to Table 2. n = 1∼5, mean ± S.D. for n = 3 or more.)

Fig. 7. OPLS-DA Score Plot (A) and Loading Plot (B) for all Samples Cultivated in Field A–C

Effects of Cultivation Location on the Proline Andmethanoproline in Ephedra Samples

The OPLS-DA results show that the proline content was low in the plants from field C, in E. sinica and in EP-13, indicating that the concentration of this compound is sensitive to the cultivation conditions. Proline is an important amino acid in plants and participates in water retention and osmotic adjustment.¹⁴⁾ For example, when plants are exposed to desiccation, proline accumulates inside the cells to prevent water loss. We have previously grown clones of Angelica acutiloba at several locations in Hokkaido and found that the proline content is inversely proportional to the annual precipitation at each location (Fuchino et al., unpublished data). The water retention properties of the soil in each field are listed in Table 1. As indicated, fields A and C exhibited poorer water retention owing to their soil textures; therefore, the plants grown in these fields were expected to accumulate proline. However, we unexpectedly found that the plants in field C contained the least amount of proline. In terms of the water potential throughout the growing season in each field (depth = 10 cm), field C consistently exhibited the highest value. Although the soil in field C was composed of sand and the surface was dry, the movement of water from the deep layer to the surface layer was limited, and the area at a depth of 10 cm was consistently wet throughout the season (Fig. 8). It should be noted here that when the soil has a low water potential, it is difficult for plants to absorb water since the water is strongly held by the soil. Therefore, the differences in proline content would be attributed to the different water potentials.

Fig. 8. Soil Water Potential of Each Experimental Field during the Period from April to December 2018

In 2001, Caveney et al. compared the secondary metabolites of Ephedra species from around the world (although there was no mention of E. sinica in the report) and reported the abundance of kynurenate, quinokine-2-carboxylic acid, and non-protein amino acids, such as cis-3,4-methanoproline and 2-(carboxycyclopropyl) glycine.¹⁵⁾ Further, they focused on amino acids, which are alkaloid precursors and reported that cis-3,4-methanoproline, a derivative of proline, is found in large amounts in the aboveground parts and seeds of several Ephedra species. In this study, the EP and PEP contents were high in Ep-13 and low in E. sinica, while the proline content showed the reverse trend: low in Ep-13 and high in E. sinica. Since methanoproline was not available, the content ratio was determined from the m/z of the pseudo-molecular ion peak estimated from its molecular formula. Namely, the area ratios obtained from the XIC peak at m/z 128.14, which is the peak corresponding to [M + H]⁺ estimated from the molecular weight of cis-3,4-methanoproline, were calculated for all samples, as shown in Fig. 9.

Fig. 9. Comparison of the Areas of the Peaks at m/z 128.14 (t_R 4.2 min) in the XIC for Cultivated and Commercial Samples (Sample Name Abbreviation: Species–Field–Years Old CCD: Commercial Crude Drug)

(Multiple plants were sampled at each fertilizer application season in each field (see Table 2). J: Additional fertilizer (Ceracoat R50) in July. S: Additional fertilizer (Ceracoat R50) in September. n: no additional fertilizer. Sample Nos correspond to Table 2. n = 1–5, mean ± S.D. for n = 3 or more.)

As shown by these results, the cis-3,4-methanoproline content is inversely proportional to the proline content. Methanoproline has been isolated from the seeds and stems of several Ephedra species^16,17); however, its role and significance are not clearly understood. Since amino acids are alkaloid precursors, they may be accumulated as components of alkaloid synthesis. It is also worth noting that a compound with m/z 183 was abundant in the Ephedra plants grown in field C. A compound having this molecular weight has not been reported in Ephedra plants, therefore we plan to investigate this compound.

Structural Determination of Compound 1

Thereafter, we examined the compositional differences between the Ephedra species. Discriminant analysis by OPLS-DA was performed on E. sinica and EP-13 (Fig. 10). The loading s-plot shows that the EP content in EP-13 tends to be higher than that in E. sinica, indicating that EP is a contributing component of EP-13, whereas compound 1 is suggested to be a contributing component of E. sinica. To identify this compound, MS/MS analysis of 1 was performed. From the HRMS results, the molecular formula of compound 1 is estimated to be C₂₂H₃₅NO₁₀. MS/MS analysis shows that the product ion at m/z 312.15, with a mass loss of 162 mass units, is observed first. This suggests the presence of a hexose molecule at the terminal end of this molecule. Further MS/MS analysis of the peak at m/z 312.15 shows that the product ion at m/z 166.07 had lost 146 mass units, suggesting the loss of deoxyhexose. In addition, MS/MS analysis of the peak at m/z 166.07 shows a product ion at m/z 121.01 with a loss of 45 mass units.

Fig. 10. OPLS-DA Score Plot (A) and Loading s-Plot (B) for E. sinica and EP-13

The compound at m/z 474 was purified from a methanol extract of E. sinica using various chromatographic methods (silica-gel column chromatography and HPLC), and its presence in the fractions was confirmed by LCMS. However, the isolation of the compound was extremely difficult owing to the low quantity present in the fractions. Nevertheless, compound 1 was obtained as a pure compound (< 1 mg) by HPLC separation, and its structure was deduced by NMR (Table 4). The sugar moieties are estimated to be β-glucose and α-rhamnose, based on the correlation of the observed signals with the anomeric hydrogens in the one dimensional (1D)-total correlation spectroscopy (TOCSY) spectra, and their coupling constants and ¹³C-NMR chemical shifts. From the MS/MS results, glucose was presumed to bind to the terminal end and rhamnose to the interior. The position of the linkage between the sugars was determined to be between the 3-position of rhamnose and the 1-position of glucose because the anomeric hydrogen atom of glucose (4.61 ppm, d) correlates with C-3 of rhamnose (81.4 ppm) in the heteronuclear multiple bond correlation spectrum, and there is a nuclear Overhauser (NOE) effect correlation between the glucose anomeric H and rhamnose H-3 (3.95 ppm, dd). The aglycon substructure was considered to have a molecular formula of C₁₀H₁₅NO based on the MS/MS results and the molecular formula of the remaining C₂₂H₃₅NO₁₀ unit excluding the sugar moieties. This molecular formula corresponds to EP or PEP, but this assignment was ruled out by NMR. Therefore, based on the molecular formula, we propose that this unit is phenylethylamine. More specifically, based on the presence of a singlet (H 2.38 ppm) corresponding to N(CH₃)₂, and two doublets (H 7.18, d, J = 8.3 Hz; 7.03, d, J = 8.3 Hz), characteristic of a p-substituted benzene ring, we identified this compound as a rhamnosylglucoside of N,N-dimethyl-p-hydroxyphenylethylamine, which was verified by two-dimensional NMR data (Fig. 11). Thus, compound 1 is identified as N,N-dimethyl-p-hydroxyphenylethylamine-O-[β-D-glucopyranosyl-(1→3)-α-L-rhamnopyranoside] (Fig. 12). This compound has not been reported previously in Ephedra plants. The aglycon part, N,N-dimethyl-p-hydroxyphenylethylamine, is commonly known as hordenine and has been isolated from E. aphylla, grown in Egypt.¹⁸⁾ However, its glycosides have yet to be reported in Ephedra.

Table 4. ¹H- and ¹³C-NMR Chemical Shifts for 1

No.	13C-NMR	1H-NMR
1	154.8
2, 6	116.3	7.03 d, J = 8.3 Hz
3, 5	129.3	7.18 d, J = 8.3 Hz
4	129.3
7	32.9	2.78 m
8	60.8	2.60 m
9, 10	43.8	2.38 s
1′	98.2	5.42 br s
2′	70.0	4.30 br s
3′	81.4	3.96 dd, J = 9.0, 3.4 Hz
4′	71.2	3.65 t, J = 9.0 Hz
5′	68.8	3.69 m
6′	16.7	1.25 d, J = 5.8 Hz
1″	104.5	4.61 d, J = 7.6 Hz
2″	74.0	3.34 t, J = 7.6 Hz
3″	76.4	3.35 t, J = 7.6 Hz
4″	69.6	3.41 m
5″	76.3	3.42 m
6″	60.8	3.87 d, J = 10.9 Hz
		3.76 dd, J = 10.9, 3.8 Hz

Measured in methanol-d₄.

Fig. 11. Key Correlations of 2D-NMR Spectra for 1

Fig. 12. Chemical Structure of 1

The ratio of 1 (m/z 474.23) in the XIC for each sample (Fig. 13) shows that EP-13 does not contain this compound. Previous genetic analysis showed that most CCDs were derived from E. sinica, though E. intermedia was also present (see Table 3). As mentioned, EP-13 was introduced as E. distachya and propagated from seeds imported from abroad; however, EP-13 was proposed to exist as a hybrid of E. likiangensis and E. gerardiana.¹⁰⁾

Fig. 13. Comparison of the Areas of the Peak at m/z 474.23 (t_R 9.9 min) in the XICs of All Samples

(Sample name abbreviation: species–field–years old. CCD: commercial crude drug) (Multiple plants were sampled at each fertilizer application season in each field (see Table 2). J: Additional fertilizer (Ceracoat R50) in July. S: Additional fertilizer (Ceracoat R50) in September. n: no additional fertilizer. Sample Nos. correspond to Table 2. n = 1–5, mean ± S.D. for n = 3 or more.)

Conclusion

Two species of Ephedra (E. sinica and Ep-13) were cultivated in three different fields distributed over Tanegashima Island. The chemical contents were assessed by multivariate analysis of the LCMS data, and E. sinica and EP-13 were easily distinguished by PCA due to the characteristic differences between these species; specifically, in their total alkaloid contents and EP/PEP ratios. In addition, there were notable differences in composition between the plants cultivated at different sites, and soil nutrients and lighting conditions were identified as the major factors that influenced the EP and PEP contents. Intriguingly, the proline content was lowest in the plants grown in field C, which could be explained by the water potential of the soil. The XIC estimated from the molecular weight of methanoproline showed that the peak area was inversely proportional to that of proline, suggesting that Ephedra plants accumulate proline, which may be converted to methanoproline depending on the growth environment.

PCA of the E. sinica, EP-13, and CCD samples revealed 1 as a marker component of E. sinica, and structural determination revealed this to be a hordenine glycoside. In the future, we plan to confirm the presence of compound 1 as a biomarker for different species of Ephedra. In addition, we plan to compare the results with those of samples grown at multiple locations to determine whether cultivation in Tanegashima Island results in higher levels of this compound present within the plant.

We noted that plants of the same species cloned from the same parent strain differed significantly in their alkaloid content depending on the growth environment.

The correlation between the growing conditions and resulting plant components requires a detailed consideration of many other factors, such as the plant height, dry yield, solar radiation, and soil physicality; these considerations will be addressed in the near future.

Finally, the quantity and timing of nitrogen addition were varied in 2020 but had no effect on the alkaloid content for that year's samples. The effects of this fertilizer will be observed over the next few years.

Experimental

Instruments

LCMS was performed on an Orbitrap Elite (Thermo Fischer Scientific, MA, U.S.A.) equipped with an electrospray ionization (ESI) ion source. NMR spectra were recorded using a Bruker Ascend 600 spectrometer (600 MHz). HPLC was performed on a JASCO LC-2000 Plus series (pump: PU-2080 Plus, diode array detector (DAD): MD-2018 Plus, column oven: CO-2060 Plus) or Waters HPLC system (pump: 1525, UV detector: 2487, autosampler: 717 plus). Preparative HPLC was performed on a GLScience PLC-561 series (pump: PU714; UV detector: UV714; column oven: CO705). Recycling HPLC was performed on an LC-908W (Japan Analytical Industry Co., Ltd., Tokyo, Japan) with a serial connection to GS-320 and GS-310 columns.

The soil water potential was measured at 30 min intervals using an MPS-6 dielectric water potential sensor (Decagon Devices, Inc., WA, U.S.A.). The data shows the average of each month in each of the fields.

Chemical and Materials

Silica gel (60N, 100–210 μm, Kanto Chemicals Co., Ltd., Tokyo, Japan) was used for column chromatography. L-Proline was purchased from Wako Pure Chemical Corporation (Osaka, Japan). Ephedrine hydrochloride (Maruishi Pharmaceutical Co., Ltd., Osaka, Japan) and pseudoephedrine hydrochloride (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) were used for quantitative analyses.

Commercial materials (CCD, JP grade) were graciously provided by the Japan Kampo Medicines Manufacturers Association and Japan Natural Medicines Association. They are registered with NIBOHN, given a control number (NIB), and stored at 20 °C.

Magnesium lime (Tsukumi seiryu kudosekkai 1 go, Tsukumi Dolomite Industrial Co. Ltd., Oita, Japan), compost (Minamitane-town Government office), fused phosphate (Kumiai ryujo yorin, Minami Kyushu Kagaku Kogyo Co. Ltd., Miyazaki, Japan), and chemical fertilizer (Kumiai nyosoiri IB kasei S1 go, JCAM AGRI. Co., Ltd., Tokyo, Japan) were used as basal fertilizer. These were supplemented with chemical fertilizer (Yukiiri BB 888 go, JA Butsuryu Kagoshima Co. Ltd., Kagoshima, Japan) and nitrogen fertilizer (Cera-coatR50, Central Glass Co., Ltd., Tokyo, Japan).

Statistical Analysis

SIMCA 14.0 version 14.0.0.1359 (Umetrics) was used for all multivariate analyses. The scale type was set to Pareto.

Quantitative Analysis of EP and PEP (JP18)

Preparation of Sample and Standard Solutions

Approximately 250 mg of Ephedra powder was accurately weighed and placed in a centrifuge tube. Thereafter, 10 mL portion of 50% methanol was added, the sample was shaken for 30 min and centrifuged, and the supernatant was separated. This procedure was repeated twice with the residue using 10 mL portions of 50% methanol. Finally, the extracts were combined and topped up to 50 mL with 50% methanol to give the sample solution.

For the ephedrine and pseudoephedrine standards, 50 mg of ephedrine hydrochloride and pseudoephedrine hydrochloride were accurately weighed and dissolved in 50% methanol to a final volume of 20 mL. A 2 mL aliquot of this solution was then diluted in 50% methanol to exactly 100 mL.

The following HPLC conditions were used for the quantitative analysis of EP and PEP: Detection: UV (210 nm); column: TOSOH TSK-gel 80T_S (5 μm, 4.6 × 150 mm); column temperature: 40 °C. The mobile phase was prepared as follows: 350 mL of acetonitrile containing 5 g of sodium lauryl sulfate was mixed by shaking thoroughly, and 650 mL of water and 1 mL of phosphoric acid were then added. The mobile phase flow rate was 1.1 mL/min.

Multivariable Analysis by LCMS

The same sample whose EP and PEP contents were measured by HPLC was diluted 10-fold and applied to LC/MS profiling.

The LCMS conditions were as follows: ESI-Orbitrap MS (positive mode); Heater temperature: 500 °C, Capillary temperature: 250 °C, Sheath gas flow: 50 arb, Auc gas flow: 15 arb, column: SUPELCO Discovery HS F5-5 (5 μm, 4.6 × 150 mm); column temperature: 50 °C; and solvent: 10 mM ammonium acetate (adjusted to pH 3.0 with formic acid)/methanol (85/15). The flow rate was 0.5 mL/min, and the analysis time was 45 min.

All LCMS data were aligned using SIEVE (ThermoScientific) and then SIMCA 14.0 (Umetrics) was used for multivariate analysis.

Cultivation of Ephedra at Tanegashima Division of the Research Center for Medicinal Plant Resources

All Ephedra samples were prepared by plant division.

(1) Samples planted in 2017 (Nos. 1–50) were cultivated as follows:

2017 (1st year after planting)

Date of planting: April 13, 2017.

Planting density: 30 cm (planting distance), 80 cm (ridge spacing), 33 plants/row.

Fertilizer (basal fertilizer): magnesium lime 100 kg/10 areas (a), compost 1000 kg/10 a, fused phosphate 40 kg/10 a (8 kg/10 a) and chemical fertilizer 80 kg/10 a (8 kg/10 a of NPK each).

Place of planting: One row each of E. sinica and EP-13 were planted in fields A, B, and C of the Tanegashima Division of the Research Center for Medicinal Plant Resources.

History of each plant: E. sinica Stapf: 0003-07TN (date of introduction: February 23, 2007; place of introduction: Tsukuba Research Division (introduced by Tsumura). EP-13: 0099-99TN (Ephedra distachya VILL., (♀); date of introduction: April 1999; place of introduction: Tsukuba Experimental Station, National Institute of Health Sciences (NIHS). Prior to Tsukuba Experimental Station, it was preserved from 1956 to 1979 at Kasukabe Experimental Station, NIHS. According to the plant registry, it was introduced from Washington D.C., U.S.A. in 1956).

2018 (2nd year after planting)

Additional fertilizer: chemical fertilizer 100 kg/10 a (8 kg/10 a of NPK each) in early May.

2019 (3rd year after planting)

Additional fertilizer: chemical fertilizer 100 kg/10 a (8 kg/10 a of NPK each) in late April.

2020 (4th year after planting)

Additional fertilizer: chemical fertilizer 100 kg/10 a (8 kg/10 a of NPK each) in early May to all plants; nitrogen fertilizer (8 kg/10 a) in July or September to some plants.

Harvest: Between October 26 and November 2, 2020. The aboveground parts were cut off leaving a portion 5 cm above the ground; thereafter, the plants were dried at 50 °C.

(2) Samples planted in 2018 (Nos. 72–131) were cultivated as follows:

2018 (1st year after planting)

Date of planting: April 24, 2018.

The planting density, fertilizer, place of planting and history of each plant was the same as those for Ephedra planted in 2017; as was the additional fertilizer introduced in 2019 (2nd year after planting) and 2020 (3rd year after planting), and the harvesting period.

Purification and Isolation of Compound 1

Dried E. sinica samples harvested in Tanegashima (81 g) were pulverized and extracted with methanol (800 mL) by refluxing for 2 h, followed by filtration and evaporation to give a residue (17 g) that was separated on silica gel with chloroform–methanol (gradient condition) as the eluent to afford 20 fractions (fr. 1–20). fr. 11–20 were combined and subjected to HPLC (Imtakt Unison UK C-18, 10 × 250 mm) with 0.1% trifluoroacetic acid (TFA)/water (A) and 0.1% TFA/acetonitrile (B) under gradient conditions (flow rate: 2.36 mL/min; gradient condition: 5% B (0–9.33 min), 5–70% B (9.33–51 min), 70–100% B (51–59.33 min), 100% B (59.33–64.33 min)). All fractions obtained by HPLC were analyzed by LCMS, and those containing 1 were combined and subjected to further recycling HPLC (coupled with columns (GS-320 and GS-310)) with 50% methanol/water to obtain eight fractions. Thereafter, the fractions containing 1 were subjected to HPLC (Imtakt Unison UK C-18, 10 × 250 mm) with 0.1% TFA/water (A) and 0.1% TFA/acetonitrile (B) under gradient conditions (flow rate: 2.36 mL/min; gradient condition: 5% B (0–9.33 min), 5–70% B (9.33–51 min), 70–100% B (51–59.33 min), 100% B (59.33–64.33 min)) to obtain a mixture with a high content of 1 (1.2 mg). Finally, the mixture containing 1 was further purified using HPLC (Imtakt Unison UK C-18, 4.6 × 250 mm) with 0.1% formic acid (FA) /water (A) and 0.1% FA/acetonitrile (B) under gradient conditions (flow rate: 0.5 mL/min; gradient condition: 5% B (0–5 min), 5–70% B (5–30 min), 70–100% B (30–35 min), 100% B (35–38 min)) to afford pure 1 (0.4 mg).

Analytical Data for Compound 1

ESI-orbitrap-MS (positive mode) m/z 474.23260 [M + H]⁺ (Calculated for C₂₂H₃₆O₁₀N: 474.23337). m/z 312.15 (MS/MS of 474.23; [M + H-162 (glucose)]⁺), m/z 166.07 (MS/MS of 312.15; [M + H-162-146 (rhamnose)]⁺), m/z 121.01 (MS/MS of 312.15; [M + H-162-146-45(NH(CH₃)₂)]⁺). ¹H-NMR and ¹³C-NMR data are shown in Table 4.

Acknowledgments

We thank Mr. Shiga and Mr. Kamada, and the staff members of the Tanegashima Division, for cultivating the E. sinica plants. We also thank the Japan Kampo Medicines Manufacturers Association and Japan Natural Medicines Association for providing the commercial crude drugs.

Conflict of Interest

The authors declare no conflict of interest.

References

• 1) Zou P., Song Y., Lei W., Li J., Tu P., Jiang Y., Acta Pharm. Sinica B, 7, 647–656 (2017).
• 2) Daikonya A., Fuchino H., Takahashi Y., Goda Y., Kawahara N., Jpn. J. Pharmacol., 67, 1–6 (2013).
• 3) Kobayashi S., Putri S. P., Yamamoto Y., Donghyo K., Bamba T., Fukusaki E., J. Biosci. Bioeng., 114, 232–236 (2012).
• 4) Daikonya A., Fuchino H., Arai R., Takahashi Y., Wada H., Goda Y., Kawahara N., Jpn. J. Pharmacol., 67, 35–40 (2013).
• 5) “The Japanese Pharmacopoeia,” 18th edition, Ministry of Health, Labour and Welfare, Japan, 2021, pp. 2060–2061.
• 6) Shibata S., Imaseki I., Ito H., Yakugaku Zasshi, 71, 806–8011 (1951).
• 7) Ando H., Ni S., Sasaki Y., Mikage M., Yakuyoshokubutsu Kenkyu, 38, 20–27 (2016).
• 8) “Yakuyoshokubutsu Saibai To Hinshitsu Hyoka, Part 9,” ed. by Committee for the Study of Cultivation of Medicinal Plants and Development of Quality Control Guidelines, Yakuji Nippo-sya, Tokyo, 2000, pp. 69–80.
• 9) Ando H., Kusaba D., Mikage M., Sasaki Y., Yakuyoshokubutsu Kenkyu, 38, 10–16 (2016).
• 10) Ando H., Kitamura M., Sasaki Y., Kitaoka F., Mikage M., Yakuyoshokubutsu Kenkyu, 38, 1–9 (2016).
• 11) Matsumoto M., Ph.D. Thesis, Graduate school of Natural Science and Technology, Kanazawa University, Ishikawa, 2014.
• 12) Hiyama H., Ph.D. Thesis, “Educational programs in Agro-Biological Resource Sciences,” University of Tsukuba, Ibaraki, 2021.
• 13) Fuchino H., Iida O., Kawano N., Kawahara N., Yoshimatsu K., J. Nat. Med., 75, 707–716 (2021).
• 14) Fukutoku Y., Yamada Y., Plant Cell Physiol., 22, 1397–1404 (1981).
• 15) Caveney S., Charlet D. A., Freitag H., Maier-Stolte M., Starratt A. N., Am. J. Bot., 88, 1199–1208 (2001).
• 16) Starratt A. N., Caveney S., Phytochemistry, 40, 479–781 (1995).
• 17) Stammer C. H., Tetrahedron, 46, 2231–2254 (1990).
• 18) Abdel-Kader M. S., Kassem F. F., Abdallah R. M., Nat. Prod. Sci., 9, 52–55 (2003).