Indole Glycosides from Calanthe discolor with Proliferative Activity on Human Hair Follicle Dermal Papilla Cells

Abstract

A methanol extract from the underground part of Calanthe discolor Lindl. (Orchidaceae) demonstrated significant proliferative activity on human hair follicle dermal papilla cells (HFDPC, % of control: 120.8 ± 0.2%) at 100 µg/mL against HFDPC. Through bioassay-guided separation of the extract, a new indole glycoside named 6′-O-β-D-apiofuranosylindican (1) was isolated along with six known compounds (2–7) including three indole glycosides. The stereostructure of 1 was elucidated based on its spectroscopic properties and chemical characteristics. Among the isolates, 1 (110.0 ± 1.0%), glucoindican (3, 123.9 ± 6.8%), and calanthoside (4, 158.6 ± 7.1%) showed significant proliferative activity at 100 µM. Furthermore, the active indole glycosides (1, 3, and 4) upregulated the expression of vascular endothelial growth factor (VEGF) and fibroblast growth factor-7 (FGF-7) mRNA and protein in HFDPC, which could be the mechanism of their proliferative activity.

Introduction

Calanthe discolor Lindl. in the Orchidaceae family¹⁾ is a perennial plant widely distributed in Japan (ebine in Japanese), China, and Korea. It is a source of popular folk medicines and is cultivated as a horticultural plant. In traditional Chinese medicine, the dried whole plant of C. discolor has been used for anti-inflammatory, antibacterial, and antitoxic purposes.^2,3) During the course of our studies on the characteristics of bioactive constituents from natural sources that are useful for cosmetics, such as melanogenesis inhibitory,^4–13) collagen synthesis-promoting, collagenase and hyaluronidase inhibitory,^14–19) and 5α-reductase inhibitory²⁰⁾ activities, we previously found that the methanol extract of the underground part of C. discolor and its constituents showed hair-restoring activity in C3H mice and promoted blood flow in the dorsal skin of rats.³⁾ Herein, to continue our study of the bioactive constituents from this plant material used for cosmetics, we show that the methanol extract exhibits proliferative activity on human hair follicle dermal papilla cells (HFDPC). Further purification of the extract allowed us to isolate a new indole glycoside named 6′-O-β-D-apiofuranosylindican (1) along with six known compounds (2–7), including three indole glycosides. Further isolation and structure elucidation of 1 as well as examination of the effects of the isolates on HFDPC proliferation were also conducted.

Results and Discussion

Extraction and Isolation

Hair is a derivative of the epidermis and consists of two separate structures: the hair follicle in the skin and visible hair shaft on the body surface.²¹⁾ Hair is produced in hair follicles, and its growth is regulated by several cellular activities, including apoptosis, proliferation, and differentiation. Hair growth undergoes a cycle consisting of anagen (growth), catagen (transitional), and telogen (resting) phases.^21,22) Hair loss (alopecia) can lead to mental suffering that affects the QOL and emotional well-being of the person experiencing it and is a common problem for both men and women. The major causes of alopecia are known to be disease, mechanical stress, nutritional deficiencies, hormonal imbalances, and aging.^23,24) A variety of treatment options for alopecia have been proposed, but all have had limited effectiveness. Finasteride and minoxidil have been approved for androgenetic alopecia (AGA) treatment by the U.S. Food and Drug Administration (FDA).^25,26) Although these medicines have been used in clinical settings, their efficacy is limited, and their effects are transient due to unpredictable side effects.²⁷⁾ Therefore, new effective and safe pharmacological treatments for alopecia are necessary.

HFDPC are specialized mesenchymal components of hair that play important roles in hair morphogenesis and regeneration through development and growth by acting as reservoirs for multipotent stem cells, nutrients, and growth factors. Moreover, HFDPC are major regulators of the hair cycle given that they respond to external stimuli and signals that are delivered by cytokines and cell junctions.²²⁾ Therefore, HFDPC have recently attracted attention as model systems to screen potential hair growth molecules.²⁸⁾ It has been suggested that abnormalities in the function of the dermal papilla disrupt the hair cycle and result in hair loss.²⁹⁾ Further, in the anagen phase, the number of HFDPC has been found to increase, generating the signals that regulate the proliferation and differentiation of keratinocytes.³⁰⁾

To characterize the active constituents of C. discolor, the proliferative effects of its isolates on HFDPC were examined. Previously, we reported that several ent-kaurane-type diterpenes isolated from Isodonis Herba, such as enmein, showed potent proliferative activity toward HFDPC. Furthermore, we found that the mechanisms of action of enmein likely involved the activation of the Akt/glycogen synthase kinase 3β (GSK-3β)/β-catenin signal transduction pathway, resulting in enhanced vascular endothelial growth factor (VEGF) production.³¹⁾ As shown in Table 1, the methanol extract (5.85% from fresh material) of the fresh underground part of C. discolor also showed significant proliferative effects on HFDPC (% of control: 120.8 ± 0.2%) at 100 µg/mL. From the extract, a new indole glycoside 6′-O-β-D-apiofuranosylindican (1, 0.00016% from the fresh plant material) was isolated along with six known compounds, indican^2,3) (2, 0.00080%), glucoindican^2,3) (3, 0.00018%), calanthoside^2,3) (4, 0.00015%), quinazoline-2,4-dione³²⁾ (5, 0.00035%), tryptanthrin³³⁾ (6, 0.000048%), and loroglossol³⁴⁾ (7, 0.000035%) (Fig. 1).

Table 1. Effects of the Constituents (1–7) from C. discolor on Proliferation of HFDPC

	Cell proliferation (% of control)
	0 µg/mL	3 µg/mL	10 µg/mL	30 µg/mL	100 µg/mL
MeOH extract	100.0 ± 3.1	107.3 ± 4.1	118.1 ± 5.0	121.5 ± 5.1*	120.8 ± 0.2**
	Cell proliferation (% of control)
	0 µM	3 µM	10 µM	30 µM	100 µM
6′-O-β-D-Apiofuranosylindican (1)	100.0 ± 3.1	111.2 ± 1.8	107.9 ± 2.5	104.3 ± 1.3	111.0 ± 1.0*
Indican (2)	100.0 ± 2.1	98.4 ± 5.2	97.6 ± 1.6	93.5 ± 0.3	102.1 ± 5.4
Glucoindican (3)	100.0 ± 7.1	102.4 ± 3.7	100.6 ± 2.2	100.4 ± 2.2	123.9 ± 6.8*
Calanthoside (4)	100.0 ± 4.3	108.1 ± 7.9	107.3 ± 2.9	112.5 ± 2.9	158.6 ± 7.1**
Tryptanthrin (5)	100.0 ± 3.3	102.1 ± 1.8	79.1 ± 2.4	—	—
Quinazoline-2,4-dione (6)	100.0 ± 4.0	102.7 ± 1.5	101.9 ± 4.3	99.9 ± 0.8	110.8 ± 6.5
Loroglessol (7)	100.0 ± 2.6	101.3 ± 5.2	99.4 ± 1.5	107.1 ± 2.2	—
	Cell proliferation (% of control)
	0 µM	2.5 µM	5 µM	10 µM	20 µM
Enmein31)	100.0 ± 2.2	103.8 ± 3.2	116.6 ± 2.3**	137.3 ± 2.5**	160.9 ± 3.0**
	Cell proliferation (% of control)
	0 µM	12.5 µM	25 µM	50 µM	100 µM
Sinapic acida)	100.0 ± 2.3	107.9 ± 2.8	121.2 ± 1.1**	130.3 ± 2.5**	179.2 ± 6.5**
Minoxidil sulfatea)	100.0 ± 6.1	100.3 ± 1.0	103.4 ± 1.8	105.5 ± 0.7	108.0 ± 5.2

Each value represents the mean ± standard error of the mean (S.E.M.) (N = 3). Significantly different from the control, * p < 0.05, ** p < 0.01. a) Commercial sinapic acid and minoxidil sulfate were purchased from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan).

Fig. 1. Isolates from the Underground Part of C. discolor

Structure of 6′-O-β-D-Apiofuranosylindican (1)

Compound 1 was obtained as an amorphous powder with negative optical rotation ([α]_D²⁴ −5.6 in MeOH). In the positive-ion electrospray ionization (ESI)-MS profile, a quasimolecular ion peak was observed at m/z 450 [M + Na]⁺, and high-resolution ESI-MS analysis revealed the molecular formula to be C₁₉H₂₅NO₁₀. The UV spectrum exhibited absorption maxima at 224 and 281 nm, while the IR spectrum showed absorption bands at 3383 and 1072 cm⁻¹ suggested the presence of a glycosyl moiety. The ¹H- and ¹³C-NMR spectra (Table 2, CD₃OD) of 1, which were assigned using distortionless enhancement by polarization transfer (DEPT), ¹H–¹H homonuclear correlation spectroscopy (COSY), heteronuclear multiple quantum coherence (HMQC), and heteronuclear multiple-bond correlation (HMBC) experiments, showed signals assignable to an indole ring [δ 6.98 (1H, ddd, J = 0.8, 7.1, 8.0 Hz, H-5), 7.07 (1H, ddd, J = 1.0, 7.1, 8.2 Hz, H-6), 7.08 (1H, s, H-2), 7.26 (1H, dd, J = 0.8, 8.2 Hz, H-7), and 7.68 (1H, dd, J = 1.0, 8.0 Hz, H-4)], a β-D-apiofuranosyl moiety [δ 5.02 (1H, d, J = 2.4 Hz, Api-H-1″)], and a β-D-glucopyranosyl moiety [δ 4.67 (1H, d, J = 7.8 Hz, Glc-H-1′)]. The ¹H- and ¹³C-NMR spectroscopic properties of 1 were quite similar to those of glucoindican (3)^2,3) as shown in Table 2. In the HMBC analysis of 1, long-range correlations were observed between the following proton and carbons pairs: Glc-H-1′ and C-3 (δ_C 139.0), and Api-H-1″ and Glc-C-6′ (δ_C 68.9), as shown in Fig. 2. Treatment of 1 with 1 M hydrochloric acid (HCl) liberated D-glucose and D-apiose, which were identified by HPLC with an optical rotation detector.^35–37) On the basis of the abovementioned evidence, the stereostructure of 1 was determined to be 6′-O-β-D-apiofuranosylindican.

Table 2. ¹H-NMR (800 MHz) and ¹³C-NMR (200 MHz, CD₃OD) Data for 6′-O-β-D-Apiofuranosylindican (1) and Glucoindican (3)

Position	1		3
	δH	δC	δH	δC
2	7.08 (s)	112.5	7.15 (s)	112.2
3		139.0		138.9
4	7.68 (dd, 1.0, 8.0)	118.7	7.67 (dd, 1.0, 8.0)	118.6
5	6.98 (ddd, 0.8, 7.1, 8.0)	119.5	6.97 (ddd, 0.8, 7.1, 7.9)	119.5
6	7.07 (ddd, 1.0, 7.1, 8.2)	122.8	7.07 (ddd, 1.0, 7.1, 8.1)	122.8
7	7.26 (dd, 0.8, 8.2)	112.3	7.26 (dd, 0.8, 8.2)	112.3
8		135.3		135.3
9		121.5		121.3
Glc-1′	4.67 (d, 7.8)	105.8	4.74 (d, 7.8)	105.5
2′	3.48 (m)	75.1	3.51 (dd, 7.8, 8.8)	75.1
3′	3.43 (dd, 9.0, 9.0)	78.2	3.44 (m)	80.0*
4′	3.37 (d, 9.0)	71.8	3.58 (m)	71.6**
5′	3.50 (m)	77.0	3.83 (m)	77.4
6′	3.64 (dd, 6.6, 11.0)	68.9	3.85 (dd, 5.9, 11.6)	69.9
	4.04 (dd, 2.1, 11.0)		4.19 (dd, 2.1, 11.6)
Api-1″	5.02 (d, 2.4)	111.1
2″	3.94 (d, 2.4)	78.1
3″		80.5
4″	3.75 (d, 9.7)	75.0
	3.96 (d, 9.7)
5″	3.58 (2H, s)	65.7
Glc-1″			4.40 (d, 7.8)	104.7
2″			3.22 (m)	75.2
3″			3.33 (m)	80.0*
4″			3.83 (m)	71.6**
5″			3.83 (m)	80.0*
6″			3.58 (dd, 2.3, 11.9)	62.7
			3.65 (dd, 5.9, 11.9)

*^,** May be interchangeable within the same column.

Fig. 2. ¹H–¹H COSY and HMBC Correlations of Compound 1

(Color figure can be accessed in the online version.)

Effects of Constituents on Proliferation of HFDPC

The proliferative effects of the isolates (1–7) on HFDPC were examined. As shown in Table 1, 6′-O-β-D-apiofuranosylindican (1, % of control: 111.0 ± 1.0%), glucoindican (3, 123.9 ± 6.8%), and calanthoside (4, 158.6 ± 7.1%) exhibited significant proliferative activity at 100 µM. However, the activities of these active indole glycosides (1, 3, and 4) were moderate compared with that of enmein, which was previously investigated by our research group.³¹⁾

Effects of Active Indole Glycosides (1, 3, and 4) on Testosterone 5α-Reductase

Male sex hormones (i.e., androgens) play a crucial role in the development, growth, and function of the prostate and other androgen-sensitive peripheral tissues. In the prostate gland, androgens are involved in benign prostatic hyperplasia and prostate cancer, in addition to being implicated in skin disorders, such as acne, seborrhea, AGA, and hirsutism. Among the androgens, testosterone is the most abundant in serum and is primarily secreted by the testicles and ovaries. The enzyme steroid 5α-reductase catalyzes the conversion of testosterone to the most potent natural androgen, 5α-dihydrotestosterone.^38–40) Therefore, the inhibition of testosterone 5α-reductase could be useful for the treatment and prevention of the aforementioned diseases. To date, three types of 5α-reductases, chronologically named types 1, 2, and 3 5α-reductases, have been described.^38,40,41) Finasteride, a type 2 and 3 5α-reductase inhibitor, is currently marketed worldwide as a drug for benign prostatic hyperplasia and is used in the treatment of hair loss^42,43) and in the prevention of prostate cancer.⁴⁴⁾ Type 2 5α-reductase is an important factor of AGA, and a variety of inhibitors have been reported.⁴⁵⁾ We examined the effects of 6′-O-β-D-apiofuranosylindican (1), glucoindican (3), and calanthoside (4), which showed relatively potent proliferative activities on HFDPC (vide supra), on rat prostatic 5α-reductase activity. As shown in Table 3, none of the tested compounds showed inhibitory activity. This suggests that 5α-reductase inhibition was not involved in the mechanisms of action of these indole glycosides (1, 3, and 4) on HFDPC proliferation.

Table 3. Inhibitory Effects of the Constituents from C. discolor on Testosterone 5α-Reductase

	Inhibition (% of control)
	0 µM	3 µM	10 µM	30 µM	100 µM
6′-O-β-D-Apiofuranosylindican (1)	0.0 ± 12.2	5.5 ± 8.1	−1.6 ± 11.9	−0.2 ± 5.8	17.4 ± 10.1
Indican (2)	0.0 ± 12.8	5.2 ± 10.7	7.0 ± 5.5	3.7 ± 5.3	−4.7 ± 3.8
Glucoindican (3)	0.0 ± 6.0	5.4 ± 6.9	0.5 ± 6.2	9.7 ± 5.7	27.0 ± 5.1
Calanthoside (4)	0.0 ± 5.9	2.5 ± 5.3	0.8 ± 9.7	8.1 ± 3.8	16.7 ± 9.1
Quinazoline-2,4-dione (5)	0.0 ± 12.7	5.5 ± 11.8	0.4 ± 16.9	3.8 ± 12.4	11.5 ± 17.3
Loroglessol (7)	0.0 ± 11.2	−2.0 ± 14.8	10.4 ± 13.8	16.7 ± 14.8	25.3 ± 11.5
	Inhibition (% of control)					IC50 (µM)
	0 µM	0.1 µM	0.3 µM	1 µM	3 µM
Finasteridea)	0.0 ± 6.4	48.5 ± 10.9**	61.8 ± 11.4**	76.0 ± 9.8**	91.3 ± 13.1**	0.12

Each value represents the mean ± S.E.M. (N = 3). Significantly different from control; ** p < 0.01. a) Commercial finasteride was purchased from Sigma-Aldrich.

Effects of Active Indole Glycosides (1, 3, and 4) on VEGF and FGF-7 Production in HFDPC

To promote cell growth and cell survival, HFDPC can produce VEGF and fibroblast growth factor-7 (FGF-7), which act as mitogens.^46,47) Since hair follicles need sufficient nutrients for vigorous cell division during the anagen phase, the role of VEGF as an endothelial cell-specific mitogen in inducing angiogenesis is important to ensure that sufficient nutrients are available for the hair follicle at this phase. The increase in follicle vascularization above physiological levels leads to an increase in follicle size above normal levels, and it has been suggested that angiogenic blood vessels directly influence the growth of the follicle epithelium in addition to improving tissue perfusion.⁴⁸⁾ Furthermore, it is known that VEGF secreted from hair papilla cells promotes their proliferation by an autocrine mechanism.^46,49) FGF-7, also known as keratinocyte growth factor (KGF), regulates epidermal proliferation and differentiation via a paracrine mechanism and has been shown to stimulate wound healing and hair growth.^50,51) In addition, FGF-7 decidedly impacts hair growth through cell multiplication to prolong the anagen phase.^52,53) Therefore, the effects of the indole glycosides, 6′-O-β-D-apiofuranosylindican (1), glucoindican (3), and calanthoside (4), on the expression of these growth factors in HFDPC were examined by quantitative RT-PCR (RT-qPCR) and enzyme-linked immunosorbent assay (ELISA). As shown in Table 4, the VEGF and FGF-7 mRNA levels were significantly increased in the HFDPC 24 h after treatment with 1 (relative mRNA expression, VEGF: 1.27 ± 0.05 and FGF-7: 1.27 ± 0.05, both at 30 µM), 3 (VEGF: 1.91 ± 0.36 at 100 µM and FGF-7: 1.26 ± 0.05 at 30 µM), and 4 (VEGF: 1.62 ± 0.01 and FGF-7: 2.24 ± 0.10, both at 30 µM). Next, to evaluate if these indole glycosides (1, 3, and 4) can enhance the production of VEGF and FGF-7 in HFDPC, the cells were incubated for 24 h with 1, 3, and 4, and the levels of VEGF and FGF-7 in the culture medium were measured using ELISA kits. As shown in Table 5, the secretion levels were significantly increased in the HFDPC 24 h after treatment with 1 (% of control, VEGF: 168.0 ± 15.0% at 100 µM and FGF-7: 182.5 ± 27.1% at 30 µM), 3 (VEGF: 140.4 ± 5.0% at 100 µM and FGF-7: 120.8 ± 11.0% at 30 µM), and 4 (VEGF: 134.9 ± 13.9% and FGF-7: 166.6 ± 16.9%, both at 100 µM). These results suggest that the indole glycosides (1, 3, and 4) induce the secretion of growth factors that stimulate cell proliferation through an autocrine mechanism.

Table 4. Effects of 6′-O-β-D-Apiofuranosylindican (1), Glucoindican (3), and Calanthoside (4) on Relative mRNA Expression of VEGF and FGF-7 in HFDPC

	VEGF mRNA/β-Actin mRNA
	0 µM	10 µM	30 µM	100 µM
6′-O-β-D-Apiofuranosylindican (1)	1.00 ± 0.04	1.43 ± 0.07**	1.27 ± 0.05*	1.11 ± 0.26
Glucoindican (3)	1.00 ± 0.13	1.15 ± 0.05	1.44 ± 0.09	1.91 ± 0.36*
Calanthoside (4)	1.00 ± 0.20	1.32 ± 0.13	1.62 ± 0.01*	0.89 ± 0.14
	FGF-7 mRNA/β-Actin mRNA
	0 µM	10 µM	30 µM	100 µM
6′-O-β-D-Apiofuranosylindican (1)	1.00 ± 0.06	1.42 ± 0.15	1.21 ± 0.05*	0.82 ± 0.04
Glucoindican (3)	1.00 ± 0.05	1.03 ± 0.02	1.26 ± 0.05**	0.96 ± 0.12
Calanthoside (4)	1.00 ± 0.17	1.60 ± 0.24	2.24 ± 0.10**	0.68 ± 0.02

Each value (relative mRNA expression) represents the mean ± S.E.M. (N = 3). Significantly different from the control, * p < 0.05, ** p < 0.01.

Table 5. Effects of 6′-O-β-D-Apiofuranosylindican (1), Glucoindican (3), and Calanthoside (4) on Production of VEGF and FGF-7 Proteins in HFDPC

	VEGF content (% of control)
	0 µM	10 µM	30 µM	100 µM
6′-O-β-D-Apiofuranosylindican (1)	100.0 ± 5.0	134.4 ± 6.4	149.5 ± 7.7*	168.0 ± 15.0**
	(100.0 ± 2.2)	(91.1 ± 0.6*)	(91.7 ± 2.2*)	(82.4 ± 0.2)
Glucoindican (3)	100.0 ± 5.0	173.3 ± 2.2**	156.3 ± 14.1*	140.4 ± 5.0*
	(100.0 ± 2.2)	(98.5 ± 2.8)	(104.3 ± 0.4)	(100.1 ± 1.5)
Calanthoside (4)	100.0 ± 5.0	128.6 ± 8.1	117.0 ± 11.0	134.9 ± 13.9
	(100.0 ± 2.2)	(99.7 ± 1.3)	(98.0 ± 2.1)	(92.6 ± 3.9)
	FGF-7 content (% of control)
	0 µM	10 µM	30 µM	100 µM
6′-O-β-D-Apiofuranosylindican (1)	100.0 ± 7.9	175.0 ± 15.2*	182.5 ± 27.1*	163.6 ± 27.3
	(100.0 ± 2.2)	(92.1 ± 1.6)	(90.2 ± 3.1*)	(83.6 ± 1.2**)
Glucoindican (3)	100.0 ± 7.9	133.9 ± 30.9	120.8 ± 11.0	126.5 ± 13.3
	(100.0 ± 2.2)	(98.5 ± 2.8)	(104.3 ± 0.4)	(100.9 ± 2.2)
Calanthoside (4)	100.0 ± 7.9	157.9 ± 27.0	149.9 ± 12.0*	166.6 ± 16.9**
	(100.0 ± 2.2)	(100.0 ± 1.4)	(97.6 ± 2.1)	(91.6 ± 3.4)

Each value represents the mean ± S.E.M. (N = 3). Significantly different from the control, * p < 0.05, ** p < 0.01. Values in parentheses indicate cell viability (%) in MTT assay.

Conclusion

We found that the methanol extract of the underground part of C. discolor showed proliferative activity on HFDPC. From the extract, a new indole glycoside, 6′-O-β-D-apiofuranosylindican (1), was isolated along with six known compounds (2–7). Among the isolates, 1, glucoindican (3), and calanthoside (4) showed significant proliferative activities on HFDPC. Furthermore, 1, 3, and 4 upregulated VEGF and FGF-7 mRNA expression and protein secretion, which could be the mechanism of HFDPC proliferative activity. Thus, these indole glycosides (1, 3, and 4) appear to be novel candidates for haircare with stimulative effects on hair growth. Further studies on the structural requirements of these indole glucosides for the stimulation of HFDPC proliferative activity and their mechanisms of action are in progress.

Experimental

The following instruments were used to obtain spectroscopic data: specific rotation, JASCO P-2200 polarimeter (JASCO Corporation, Tokyo, Japan, l = 5 cm); UV spectra, Shimadzu UV-1600 spectrometer; IR spectra, IRAffinity-1 spectrophotometer (Shimadzu Co., Kyoto, Japan); ¹H-NMR spectra, JNM-ECA800 (800 MHz), JNM-LA500 (500 MHz), JNM-ECS400 (400 MHz), and JNM-AL400 (400 MHz) spectrometers; ¹³C-NMR spectra, JNM-ECA800 (200 MHz), JNM-LA500 (125 MHz), JNM-ECS400 (100 MHz), and JNM-AL400 (100 MHz) spectrometers with tetramethylsilane as an internal standard; ESI-MS and high-resolution ESI-MS, Exactive Plus mass spectrometer (Thermo Fisher Scientific Inc., MA, U.S.A.); HPLC detectors, Shimadzu SPD-10A UV-VIS detector and Shodex OR-2 optical rotation detector (Showa Denko Co., Ltd., Tokyo, Japan); HPLC columns, Cosmosil 5C₁₈-MS-II (Nacalai Tesque, Inc., Kyoto, Japan), 4.6 mm i.d. × 250 mm and 20 mm i.d. × 250 mm for analytical and preparative purposes, respectively. A Kaseisorb LC NH₂-60-5 column (Tokyo Kasei Co., Ltd., Tokyo, Japan, 4.6 mm i.d. × 250 mm) was used to identify the sugar moieties.

The following experimental conditions were used for column chromatography (CC): highly porous synthetic resin, Diaion HP-20 (Mitsubishi Chemical Co., Tokyo, Japan); normal-phase silica gel CC, silica gel 60N (Kanto Chemical Co., Ltd., Tokyo, Japan; 63–210 mesh, spherical, neutral); reversed-phase octadecyl silica (ODS) CC, Chromatorex ODS DM1020T (Fuji Silysia Chemical, Ltd., Aichi, Japan; 100–200 mesh); TLC, pre-coated TLC plates with silica gel 60F₂₅₄ (Merck, Darmstadt, Germany, 0.25 mm) (normal-phase) and silica gel RP-18 WF_254S (Merck, 0.25 mm) (reversed-phase); reversed-phase HPTLC, pre-coated TLC plates with silica gel RP-18 WF_254S (Merck, 0.25 mm). Detection was carried out by spraying with 1% Ce(SO₄)₂ in 10% aqueous H₂SO₄ followed by heating.

Plant Material

The fresh underground part of C. discolor was collected from the Miyazaki prefecture, Japan in May 2014. The plant material was identified by one of the authors (M.Y.). A voucher specimen (2014.05. Nomura-01) of this plant is on file in our laboratory.

Extraction and Isolation

The fresh underground part of C. discolor (12.0 kg) was extracted three times with methanol under reflux for 3 h. Evaporation of the combined extracts under reduced pressure yielded the methanol extract (702.6 g, 5.85%). Isolation and identification of the constituents (1–7) from the methanol extract were carried out as follows.

An aliquot (401.2 g) of the methanol extract was partitioned into an EtOAc–3% aqueous tartaric acid (1 : 1, v/v) mixture to furnish an acidic EtOAc-soluble fraction (74.1 g, 1.89%) and an acidic aqueous solution. The aqueous solution was adjusted to pH 9 with saturated aqueous Na₂CO₃ and then extracted with CHCl₃. Removal of the solvent under vacuum yielded a CHCl₃-soluble fraction (6.98 g, 0.18%). The aqueous layer was extracted with n-BuOH, and removal of the solvent under vacuum yielded an n-BuOH-soluble fraction (53.8 g, 1.37%).

An aliquot (65.6 g) of the acidic EtOAc-soluble fraction was subjected to normal-phase silica gel CC [2.4 kg, n-hexane–EtOAc (50 : 1→30 : 1→5 : 1→1 : 1, v/v)→EtOAc→MeOH)] to give six fractions [fr. 1 (6.90 g), fr. 2 (2.96 g), fr. 3 (7.07 g), fr. 4 (3.28 g), fr. 5 (3.48 g), and fr. 6 (11.1 g)]. Fraction 3 (7.07 g) was subjected to reversed-phase silica gel CC [235 g, MeOH–H₂O (80 : 20→90 : 1, v/v)→MeOH)] to give loroglossol³⁴⁾ (7, 10.8 mg, 0.000035%).

An aliquot (6.83 g) of the CHCl₃-soluble fraction was subjected to reversed-phase silica gel CC [225 g, MeOH–H₂O (20 : 80→30 : 70→40 : 60→60 : 40→80 : 20→90 : 10, v/v)→MeOH)] to give eight fractions [fr. 1 (2.76 g), fr. 2 (925.8 mg), fr. 3 (394.5 mg), fr. 4 (212.1 mg), fr. 5 (338.6 mg), fr. 6 (544.4 mg), fr. 7 (954.8 mg), and fr. 8 (579.2 mg)]. Fraction 2 (925.8 mg) was further purified by HPLC [detection: UV (254 nm), MeOH–H₂O (1 : 1, v/v)] to give quinazolin-2,4-dione³²⁾ (5, 13.0 mg, 0.00035%). Fraction 4 (212.1 mg) was further purified by HPLC [detection: UV (254 nm), MeOH–H₂O (1 : 1, v/v)] to yield tryptanthrin³³⁾ (6, 1.8 mg, 0.000048%).

An aliquot (216.8 g) of the methanol extract was partitioned into an EtOAc–H₂O (1 : 1, v/v) mixture to furnish an EtOAc-soluble fraction (15.5 g, 0.42%) and an aqueous phase. The aqueous phase was subjected to Diaion HP-20 CC (3.0 kg, H₂O→MeOH) to give H₂O-eluted (169.1 g, 4.57%) and MeOH-eluted (8.96 g, 0.24%) fractions.

An aliquot (8.00 g) of the MeOH-eluted fraction was subjected to normal-phase silica gel CC [400 g, CHCl₃–MeOH (10 : 1→3 : 1→1 : 1, v/v)→CHCl₃–MeOH–H₂O (65 : 35 : 10, lower layer→6 : 4 : 1, v/v/v)→MeOH] to give seven fractions [fr. 1 (378.5 mg), fr. 2 (1.24 g), fr. 3 (3.72 g), fr. 4 (901.7 mg), fr. 5 (463.6 mg), fr. 6 (172.3 mg), and fr. 7 (12.9 mg)]. Fraction 4 (901.7 mg) was subjected to reversed-phase silica gel CC [29.6 g, MeOH–H₂O (20 : 80→90 : 10, v/v)→MeOH] to afford three fractions [fr. 4-1 (219.5 mg), fr. 4-2 (673.0 mg), and fr. 4-3 (40.8 mg)]. Fraction 4-1 (219.5 mg) was further purified by HPLC [detection: UV (254 nm), MeOH–H₂O (15 : 85, v/v)] to give 6′-O-β-D-apiofuranosylindican (1, 4.7 mg, 0.00016%), indican^2,3) (2, 23.6 mg, 0.00080%), glucoindican^2,3) (3, 5.2 mg, 0.00018%), and calanthoside^2,3) (4, 4.3 mg, 0.00015%). The known isolated compound (2) was unambiguously identified by comparison of its physical and spectroscopic data with those of commercially available sample. Other known isolates (3–7) were also identified by comparison of their spectroscopic properties with those of reported values.

6′-O-β-D-Apiofuranosylindican (1): Amorphous powder, [α]_D²⁴ −5.6 (c 0.20, MeOH); High-resolution positive-ion ESI-MS m/z: 450.1371 [M + Na]⁺ (Calcd for C₁₉H₂₅O₁₀NNa, 450.1371); UV [nm (log ε), MeOH]: 224 (4.33), 281 (3.64); IR (KBr): 3383, 1628, 1589, 1562, 1458, 1072 cm⁻¹; ¹H-NMR (800 MHz, CD₃OD) and ¹³C-NMR (200 MHz, CD₃OD): given in Table 2; positive-ion ESI-MS m/z: 450 [M + Na]⁺.

Acid Hydrolysis of Compound 1

A solution of compound 1 (1.0 mg) in 1 M HCl (1.0 mL) was stirred at 80 °C for 1 h. After the solution cooled, it was neutralized with Amberlite IRA-400 (OH⁻ form) and the resin was removed by filtration. After removal of the solvent from the filtrate under reduced pressure, the residue was partitioned in EtOAc–H₂O (1 : 1, v/v), and the solvent was removed under vacuum from both the EtOAc and aqueous phases. The aqueous layer was subjected to HPLC analysis under the following conditions: HPLC column, Kaseisorb LC NH₂-60-5, 4.6 mm i.d. × 250 mm (Tokyo Kasei Co., Ltd.); detection, optical rotation [Shodex OR-2 (Showa Denko Co., Ltd.)]; mobile phase, CH₃CN–H₂O (85 : 15, v/v); flow rate, 0.8 mL/min. Identification of D-apiose and D-glucose present in the aqueous phase was carried out by comparing their retention times and the optical rotations with those of authentic samples [D-apiose, t_R: 6.6 min (positive optical rotation) and D-glucose, t_R: 13.9 min (positive optical rotation)].^35–37)

Bioassays

Cell Culture

The HFDPC were originally purchased from TaKaRa Bio Inc. (Shiga, Japan), and the cells were grown in Dulbecco’s modified eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), penicillin (100 units/mL), and streptomycin (100 µg/mL) at 37 °C in 5% CO₂/95% air. The cells were harvested by incubation in phosphate-buffered saline (PBS) containing 1 mM ethylenediaminetetraacetic acid (EDTA) and 0.25% trypsin for approximately 5 min at 37 °C and were used for the subsequent bioassays.

Measurement of Cell Proliferation

The HFDPC (1.0 × 10⁴ cells/mL) were seeded in 96-well plates in serum-free DMEM (100 µL/well) and cultured for 24 h. Afterward, 100 µL of the test compounds at various concentrations dissolved in serum-free DMEM were added to each well, followed by incubation for 4 d. After the incubation, 20 µL of Cell Counting Kit-8 (CCK-8, Dojindo Laboratories, Kumamoto, Japan) reagent was added to each well, followed by incubation for 1 h at 37 °C. The absorbance was recorded at 450 nm (absorbance at 650 nm as reference) on a microplate reader (MTP-900Lab; HITACHI, Japan). The results are expressed as mean percentages of the control ± standard error of the mean (S.E.M.).

Cell proliferation was calculated using the following equation:

  

where A is the absorbance at 450 nm (650 nm for the test sample) and B is the absorbance at 450 nm (650 nm for the control).

Assay for Testosterone 5α-Reductase Inhibitory Activity

The experiment was performed in accordance with previously reported methods^20,45,54,55) with slight modifications. In brief, the assay was performed in 48-well microplates (Sumitomo Bakelite Co., Ltd., Tokyo, Japan). The reaction solution was pre-incubated with or without test sample (5 µL/well, dissolved in dimethyl sulfoxide (DMSO)) in potassium phosphate buffer (40 mM, pH 6.5, 490 µL/well) containing substrate (0.35 nmol of testosterone; Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) and reduced nicotinamide adenine dinucleotide phosphate (NADPH) (10 nmol; Oriental Yeast Co., Ltd., Tokyo, Japan) at room temperature (25 °C) for 20 min. The enzymatic reaction was carried out by the addition of rat liver S9 fraction (10 µL/well, dissolved in the same phosphate buffer, 20.6 µg/well; Oriental Yeast Co., Ltd., lot no. 109031513) and incubation at 37 °C for 30 min. After incubation, the mixture was immediately heated in boiling water for 2 min to stop the reaction. Then, the assay solution in each well was transferred to a microtube and extracted with 500 µL of EtOAc. After the microtube was centrifuged (10000 rpm, 5 min), an aliquot of each EtOAc phase (300 µL) was transferred to another tube. The solvent in the tube was evaporated and the residue was dissolved in 30 µL of acetonitrile containing the internal standard (I.S.) fludrocortisone acetate (20 µg/mL; Sigma-Aldrich, Co., LLC, St. Louis, U.S.A.). A 2-µL aliquot was injected into an LC-20A Prominence HPLC system (Shimazdu Co., Kyoto, Japan) under specific conditions: [detection: UV (254 nm); column: Cosmosil 5C₁₈-MS-II (Nacalai Tesque Inc.; 5 µm particle size, 2.0 mm i.d. × 150 mm); column temperature: 40 °C; mobile phase: MeOH–H₂O (60 : 40, v/v); flow rate: 0.2 mL/min; retention time: 13.5 min for testosterone and 8.0 min for the I.S.]. A similar procedure was used for the control tubes. The 5α-reductase inhibitory activity was determined from the equation below using the peak area ratios (r = testosterone/I.S.). Experiments were performed in triplicate or quadruplicate, and the IC₅₀ values were determined graphically. The 5α-reductase inhibitor finasteride (Tokyo Chemical Industry Co., Ltd.) was used as a reference compound. Inhibition (%) was determined with the following equation:

  

where C is the Control [enzyme (+), test sample (−)], T is the Test [enzyme (+), test sample (+)], and B is the Blank (B) [enzyme (−), test sample (+)].

Analysis of Growth Factor Levels (RT-qPCR)

RT-qPCR was used to investigate effects on the expression of VEGF-related genes in HFDPC. The cells were seeded in 6-well plates (1.2 × 10⁵ cells/well) in DMEM containing 10% FBS (2 mL/well) and cultured for 4 d. After removing the culture medium, 1 mL of serum-free DMEM was added to each well, followed by incubation for 24 h. Then, 1 mL of the test compounds at various concentrations were dissolved in serum-free DMEM and added to each well, followed by incubation for 24 h. Total RNA was extracted from the cells using an RNeasy Mini Kit (Qiagen, Dusseldorf, Germany) according to the manufacturer’s instructions. The concentration and purity of the RNA were determined by measuring absorbance at 260 nm and determining the ratio of the readings at 260 and 280 nm. cDNAs were synthesized from 0.2 µg total RNA using iScript™ RT Supermix for RT-qPCR (Bio-Rad, CA, U.S.A.) according to the manufacturer’s instructions. The resulting template cDNAs were incubated with gene-specific primers and with SsoAdvanced™ Universal SYBR^® Green Supermix (Bio-Rad) in a CFX Connect Real-Time PCR Detection System (Bio-Rad). The relative abundance of each gene product was calculated, with values for the target genes normalized to that of β-actin mRNA. The thermal cycling program used an initial denaturation step (95 °C for 30 s) and then 40 cycles of denaturation (95 °C for 15 s) and annealing/extension (60 °C for 30 s). The primer pairs were as follows: VEGF primers, 5′-TCA CCA AAG CCA GCA CAT AG-3′ and 5′-AAA TGC TTT CTC CGC TCT GA-3′⁵⁶⁾; FGF-7 primers, 5′-ATC AGG ACA GTG GCA GTT GGA-3′ and 5′-AAC ATT TCC CCT CCG TTG TGT-3′⁵⁷⁾; and β-actin primers, 5′-TGG ATC AGC AAG CAG GAG TA-3′ and 5′-TCG GCC ACA TTG TGA ACT TT-3′.⁵⁸⁾

Analysis of Growth Factor Levels (ELISA)

The HFDPC were seeded in 24-well plates (8.0 × 10⁴ cells/well) in DMEM containing 10% FBS (500 µL/well) and cultured for 4 d. After removing the culture medium, 200 µL of serum-free DMEM was added to each well, followed by incubation for 24 h. Then, 200 µL of the test compounds at various concentrations were dissolved in serum-free DMEM and added to each well, followed by incubation for 24 h. The culture medium was then collected from all test groups, and the levels of secreted VEGF and FGF-7 were determined using ELISA kits [VEGF Human ELISA Kit (Thermo Fisher Scientific Inc.); FGF-7 (KGF) Human ELISA Kit (Thermo Fisher Scientific Inc.)], according to the manufacturer’s instructions. Data are expressed as the percent of control VEGF and FGF-7 content. In addition, cell proliferation was determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay. Thus, the medium was exchanged with 100 µL of fresh medium and 10 µL of MTT solution (5 mg/mL in PBS) was added. After incubation for 4 h, the medium was removed and i-PrOH (100 µL containing 0.04 M HCl) was added to dissolve the formazan formed in the cells. The absorbance of the formazan solution was measured using a microplate reader at 570 nm (reference: 655 nm). To produce a stock solution, each test compound was dissolved in DMSO (final DMSO concentration in the medium, 0.5%).

Statistics

Values are expressed as means ± S.E.M. One-way ANOVA, followed by Dunnett’s test was used for statistical analysis. Probability (p) values less than 0.05 were considered significant.

Acknowledgments

This work was supported in part by JSPS KAKENHI, Japan [Grant Number 18K06726 (TM)]. The authors gratefully thank the Division of Joint Research Center of Kindai University for the NMR and MS measurements.

Conflict of Interest

Y. I. and T. K. are employees of Nomura Co., Ltd. T. M., Y. M., F. L., H. F., K. N., O. M., and M. Y. have no conflicts of interest.

Supplementary Materials

The online version of this article contains supplementary materials.

References

• 1) “The Plant List.”: ‹http://www.theplantlist.org/tpl1.1/record/kew-30158›, cited 23 March, 2021.
• 2) Yoshikawa M., Murakami T., Kishi A., Sakurama T., Matsuda H., Nomura M., Matsuda H., Kubo M., Chem. Pharm. Bull., 46, 886–888 (1998).
• 3) Murakami T., Kishi A., Sakurama T., Matsuda H., Yoshikawa M., Heterocycles, 54, 957–966 (2001).
• 4) Morikawa T., Nakanishi Y., Ninomiya K., Matsuda H., Nakashima S., Miki H., Miyashita Y., Yoshikawa M., Hayakawa T., Muraoka O., J. Nat. Med., 68, 539–549 (2014).
• 5) Tanabe G., Sugano Y., Shirato M., Sonoda N., Tsutsui N., Morikawa T., Ninomiya K., Yoshikawa M., Muraoka O., J. Nat. Prod., 78, 1536–1542 (2015).
• 6) Morikawa T., Ninomiya K., Kuramoto H., Kamei I., Yoshikawa M., Muraoka O., J. Nat. Med., 70, 89–101 (2016).
• 7) Ninomiya K., Matsumoto T., Chaipech S., Miyake S., Katsuyama Y., Tsuboyama A., Pongpiriyadacha Y., Hayakawa T., Muraoka O., Morikawa T., J. Nat. Med., 70, 179–189 (2016).
• 8) Morikawa T., Kitagawa N., Tanabe G., Ninomiya K., Okugawa S., Motai C., Yoshikawa M., Lee I.-J., Muraoka O., Molecules, 21, 930 (2016).
• 9) Manse Y., Ninomiya K., Nishi R., Kamei I., Katsuyama Y., Imagawa T., Chaipech S., Muraoka O., Morikawa T., Bioorg. Med. Chem., 24, 6215–6224 (2016).
• 10) Manse Y., Ninomiya N., Okazaki A., Okada-Nishida E., Imagawa T., Imamura-Mizushima M., Yamano Y., Kaname K., Nakamura S., Morikawa T., Nat. Prod. Commun., 12, 1185–1188 (2017).
• 11) Manse Y., Ninomiya K., Nishi R., Hashimoto Y., Chaipech S., Muraoka O., Morikawa T., Molecules, 20, 2279 (2017).
• 12) Taira N., Katsuyama Y., Yoshioka M., Muraoka O., Morikawa T., Int. J. Mol. Sci., 19, 1144 (2018).
• 13) Tanabe G., Manse Y., Ogawa T., Sonoda N., Marumoto S., Ishikawa F., Ninomiya K., Chaipech S., Pongpiriyadacha Y., Muraoka O., Morikawa T., J. Org. Chem., 83, 8250–8264 (2018).
• 14) Morikawa T., Ninomiya K., Takamori Y., Niashida E., Yasue M., Hayakawa T., Muraoka O., Li X., Nakamura S., Yoshikawa M., Matsuda H., Phytochemisrtry, 116, 203–212 (2015).
• 15) Morikawa T., Manse Y., Koda M., Chaipech S., Pongpiriyadacha Y., Muraoka O., Ninomiya Y., J. Nat. Med., 72, 542–550 (2018).
• 16) Morikawa T., Nagatomo A., Kitazawa K., Muraoka O., Kikuchi T., Yamada T., Tanaka R., Ninomiya K., J. Oleo Sci., 67, 1271–1277 (2018).
• 17) Morikawa T., Okugawa S., Manse Y., Muraoka O., Yoshikawa M., Ninomiya K., Nat. Prod. Commun., 14, 1934578X19857834 (2019).
• 18) Morikawa T., Nakanishi Y., Inoue N., Manse Y., Matsuura H., Hamasaki S., Yoshikawa M., Muraoka O., Ninomiya K., Phytochemistry, 169, 112185 (2020).
• 19) Morikawa T., Inoue N., Nakanishi Y., Manse Y., Matsuura H., Okino K., Hamasaki S., Yoshikawa M., Muraoka O., Ninomiya K., Fitoterapia, 143, 104584 (2020).
• 20) Morikawa T., Luo F., Manse Y., Sugita H., Saeki S., Chaipech S., Pongpiriyadacha Y., Muraoka O., Ninomiya K., Front. Chem., 8, 199 (2020).
• 21) Buffoli B., Rinaldi F., Labanca M., Sorbelini E., Trink A., Guranziroli E., Rezzan R., Rodella L. F., Int. J. Dermatol., 53, 331–341 (2014).
• 22) Alonso L., Fuchs E., J. Cell Sci., 119, 391–393 (2006).
• 23) Shimomura Y., Christiano A. M., Annu. Rev. Genomics Hum. Genet., 11, 109–132 (2007).
• 24) Otchkarev V. A., Am. J. Pathol., 162, 709–711 (2003).
• 25) Burton J. L., Marshall A., Br. J. Dermatol., 101, 593–595 (1979).
• 26) Kaufman K. D., Olsen E. A., Whiting D., Savin R., DeVillez R., Bergfeld W., Price V. H., van Neste D., Roberts J. L., Hordinsky M., Shapire J., Binkowits B., Gomley G. J., J. Am. Acad. Dermatol., 39, 578–589 (1998).
• 27) Jain R., De-Eknamkul W., Opin. Ther. Targets, 18, 787–806 (2014).
• 28) Madaan A., Verma R., Singh A. T., Jaggi M., Int. J. Cosmet. Sci., 40, 429–450 (2018).
• 29) Lin C. M., Li Y., Ji Y. C., Keng H., Cai X. N., Zhang J. K., Arch. Dermatol. Res., 300, 531–535 (2008).
• 30) Yamana K., Labrie F., Luu-The V., Mol. Biol. Clin. Invest., 2, 293–299 (2010).
• 31) Manse Y., Luo F., Kato K., Okazaki A., Okada-Nishida E., Yanagida M., Nakamura S., Morikawa T., J. Nat. Med., 75, 326–338 (2021).
• 32) Hosoe T., Nazawa K., Kawahar N., Fukushima K., Nishimura K., Miyaji M., Kawai K., Mycopathologia, 146, 9–12 (1999).
• 33) Kesternich V., Pérez-Fehrmann M., Ortíz S., Verdugo F., Brito I., Bolte M., Cárdenas A., J. Chil. Chem. Soc., 58, 1817–1819 (2013).
• 34) Stoessl A., Stothers J. B., Org. Magn. Res., 20, 166–169 (1982).
• 35) Wang L.-B., Morikawa T., Nakamura S., Ninomiya K., Matsuda H., Muraoka O., Wu L.-J., Yoshikawa M., Heterocycles, 68, 1235–1242 (2009).
• 36) Morikawa T., Wang L.-B., Ninomiya K., Nakamura S., Matsuda H., Muraoka O., Wu L.-J., Yoshikawa M., Chem. Pharm. Bull., 57, 853–859 (2009).
• 37) Ninomiya K., Motai C., Nishida E., Kitagawa N., Yoshihara K., Hayakawa T., Muraoka O., Li X., Nakamura S., Yoshikawa M., Matsuda H., Morikawa T., J. Nat. Med., 70, 435–451 (2016).
• 38) Yamana K., Labrie F., Luu-The V., Horm. Mol. Biol. Clin. Invest., 2, 293–299 (2010).
• 39) Yao Z., Xu Y., Zhang M., Jiang S., Nicklaus M. C., Liao C., Bioorg. Med. Chem. Lett., 21, 475–478 (2011).
• 40) Azzouni F., Godoy A., Li Y., Mohler J., Adv. Urol., 2012, 530121 (2012).
• 41) Titus M. A., Li Y., Kozyreva O. G., Maher V., Godoy A., Smith G. J., Mohler J. L., Prostate, 74, 235–249 (2014).
• 42) Heinzl S., Med. Monatsschr. Pharm., 22, 124–127 (1999).
• 43) Tosti A., Piraccini B. M., J. Am. Acad. Dermatol., 42, 848–849 (2000).
• 44) Coltman C. A. Jr., Thompson I. M. Jr., Feigl P., Eur. Urol., 35, 544–547 (1999).
• 45) Matsuda H., Sato N., Yamazaki M., Naruto S., Kubo M., Biol. Pharm. Bull., 24, 586–587 (2001).
• 46) Thornton M. J., Hamada K., Messenger A. G., Randall V. A., J. Invest. Dermatol., 111, 727–732 (1998).
• 47) Morgan A. M., Harb. Perspect. Biol., 4, 1–14 (2014).
• 48) Yano K., Brown L. B., Detmar M., J. Clin. Invest., 107, 409–417 (2001).
• 49) Lachgar S., Moukadiri H., Jonca F., Charveron M., Bouhaddioui N., Gall Y., Bonafe J. L., Plöuet J., J. Invest. Dermatol., 106, 17–23 (1996).
• 50) Guo L., Degenstein L., Fuchs E., Genes Dev., 10, 165–175 (1996).
• 51) Booth C., Potten C. S., Invest. Dermatol., 114, 667–673 (2002).
• 52) Gentile P., Garcovich S., Cells, 8, 466 (2019).
• 53) Greco V., Chen T., Rendl M., Schober M., Pasolli H. A., Strokes N., Cruz-Racells J. D., Fuchs E., Cell Stem Cell, 4, 155–169 (2009).
• 54) Lee S. H., Lee D. H., Lee J., Lee W. Y., Chung B. C., Choi M. H., Mass. Spectrom. Lett., 3, 21–24 (2012).
• 55) Koseki J., Matsumoto T., Matsubara Y., Tsuchiya K., Mizuhara Y., Sekiguchi K., Nishimura H., Watanabe J., Kaneko A., Hattori T., Maemura K., Kase Y., Evid. Based Complement. Alternat. Med., 2015, 853846 (2015).
• 56) Zisa D., Shabbir A., Mastri M., Suziki G., Lee T., Am. J. Physiol. Regul. Integr. Comp. Physiol., 297, 1503–1515 (2009).
• 57) Cheon I. H., Bae S., Ahn J. K., J. Microbiol. Biotechnol., 29, 321–329 (2019).
• 58) Wang Y., Wang L.-Y., Feng F., Zhao Y., Huang M.-Y., Shao Q., Chen C., Sheng H., Chen D.-L., Zeng Z.-L., Xu R.-H., Li Y.-H., Oncol. Rep., 34, 2106–2114 (2015).