Anti-inflammatory activity of Penstemon gentianoides and Penstemon campanulatus^*

Abstract

Context: Penstemon gentianoides (Kunth) Poir. and Penstemon campanulatus (Cav.) Willd. (Plantaginaceae) are important medicinal plants in Mexico used by indigenous people for their anti-inflammatory effects and to also reduce rheumatic pains.

Objective: In addition to radical scavenging activity, the anti-inflammatory activity of the extracts, fractions and compounds of these plants were investigated and reported here for the first time.

Materials and methods: The anti-inflammatory activities of MeOH, CH₂Cl₂, and ethyl acetate extracts and iridoid, flavonoids, and phenylpropanoids from Penstemon gentianoides and P. campanulatus were studied in the TPA-induced mouse ear edema model. In addition, antioxidant activity against DPPH, crocin and β-carotene were investigated.

Results: All extracts were tested and a selection of known compounds significantly (p <0.05) inhibited mouse ear edema. The results showed that CH₂Cl₂ extracts of roots and stems from P. gentianoides and ethyl acetate extracts of leaves from P. gentianoides and P. campanulatus, as well as luteolin, diosmetin, penstemide and verbascoside produced the most positive results. Of all substances tested, the CH₂Cl₂ extract of P. gentianoides roots was the most powerful inhibitor (ED₅₀ = 0.07 mg/ear), with activity comparable to that of indomethacin. These extracts, compounds purified, as well as known compounds, inhibited oxidation of β-carotene and crocin.

Discussion and conclusion: These findings showed that the iridoid monoterpenes, flavonoids and phenylpropanoids present in these plants species may all contribute to the observed anti-inflammatory activity. Additionally, the observed antioxidant activity is correlated with the anti-inflammatory activity of these plants and the phytochemicals derived from them.

Keywords::

• Plantaginaceae
• radical scavenging
• antioxidants
• iridoid monoterpenes
• flavonoids
• phenylpropanoids
• crocin
• β-carotene
• hydroxyl radicals

Introduction

As a continuation of our systematic search for biological activities of compounds from the Mexican flora, we wish to report the isolation and anti-inflammatory activities of a series of iridoid monoterpenes, flavonoids, and phenylpropanoids from Penstemon gentianoides (Kunth) Poir. and Penstemon campanulatus (Cav.) Willd. (Plantaginaceae). Both species of Penstemon are shrubs 1 to 3 m in height that grow in mountainous regions of Guatemala, Mexico, and the US. These plants are used by indigenous people of Mexico under the common names of “jarritos”, “parrito”, and “campanitas” for their anti-inflammatory, emollient, balsamic, and laxative effects, as well as to treat rheumatic pains (Domínguez et al., 2005). Previous studies of these plants reported the presence of iridoid monoterpenes (Jiménez-Estrada et al., 2006) such as catalpol, aucubin, and penstemide. Extracts of these plants had pronounced antioxidant activity (Domínguez et al., 2005). One of these compounds, penstemide, exhibited antitumor activity and laxative effects (Jiménez-Estrada et al., 2006). In previous studies, iridoid monoterpenes have been found to have other biological activities such as antimicrobial activity (Li et al., 2009), as feeding stimulants for butterfly larvae, and as compounds that offer protection against avian predators (Bowers, 1983; Seigler, 1998). Large foliar concentrations of iridoid monoterpenes may provide specialist herbivores with defensive compounds that they sequester from the plants for their own defense and make the plants unavailable to generalist herbivores (Peñuelas et al., 2006).

In this paper, we report the anti-inflammatory activity of MeOH, CH₂Cl₂, EtOAc extracts from roots and aerial parts, as well as the purified compounds catalpol 1, penstemide 2, pensteminoside 3, plantarenaloside 4, globularisicin 5, martinoside 6, verbascoside 7, luteolin 8 and diosmetin 9 isolated from P. campanulatus and P. gentianoides. The anti-inflammatory effect was determined by means of the 12-O-tetradecanoylphorbol acetate (TPA)-induced mouse ear edema test (Tubaro et al., 1985; De Young et al., 1989; Paya et al., 1993; Recio et al., 2000).

Materials and methods

Plant material

Plant material of P. campanulatus (4.9 kg) was collected in Desierto de Los Leones National Park, and that of P. gentianoides (10.1 kg) in Los Dinamos National Park, both in the mountains to west of Mexico City at 2890 m in March 2006. Samples were identified by Francisco Ramos, Instituto de Biología at UNAM, Ciudad Universitaria, Coyoacán, Mexico City. Voucher samples have been deposited in the Ethnobotanical Collection of the Herbarium of the Universidad Nacional Autónoma de Mexico, Instituto de Biología. Voucher: Francisco Ramos: MEXU, Gen: 7508, G-No. 7. All plant material was dried in shade.

Animals

Male Wistar rats (200–250 g body weight; 10 weeks old) were obtained from the Instituto de Fisiología Celular of the Universidad Nacional Autónoma de Mexico. Prior to the experiments the rats were fed with standard food and water ad libitum. All procedures and experiments reported here were carried out following the guidelines stated in Guide for the Care and Use of Laboratory Animals (NIH, 1985; NOM, 1999).

Chemicals and solvents

All reagents used were either analytical reagent (including TPA) or chromatographic grade, purchased from Sigma, Toluca, Mexico. Methanol, CH₂Cl₂, CHCl₃, KCl, CuSO₄, silica gel GF₂₅₄ analytical chromatoplates, silica gel grade 60, (70-230, 60A°) for column chromatography, n-hexane, and ethyl acetate were purchased from Merck, Mexico City. Indomethacin (Sigma-Aldrich, Toluca) and ovatifolin obtained during a previous work (Céspedes et al., 2001) were used as standard bioactive compounds.

Apparatus

¹H-NMR spectra were recorded at 300 MHz, ¹³C-NMR at 75 at MHz, on a Varian VXR-300S spectrometer, chemical shifts (ppm) are relative to (CH₃)₄Si as internal reference, CDCl₃ and acetone-d₆ from Aldrich, Toluca, were used as solvents, coupling constants are quoted in Hz. IR spectra were obtained in KBr on a Perkin Elmer 283-B and an FT-IR Nicolet Magna 750 spectrophotometers. UV spectra of pure compounds were determined on a Shimadzu UV-160 spectrophotometer. A Spectronic model Genesys 5 spectrophotometer was used for measurement of biological activities. Melting points were obtained on a Fisher-Johns melting point apparatus and are uncorrected.

General experimental procedures

HPLC was performed on a Waters Model 600E chromatograph, equipped with Bondapack RP 18 column, 250 mm × 8 mm, at a flow rate of 1.5 mL/min, paper speed 5 mm/min, UV detector 280 nm, mobile phase MeOH/H₂0 7:3 v/v. Analytical TLC was performed on silica gel 60 F₂₅₄ Merck plates (Mexico City). The plates were visualized by spraying with a 10% solution of H₂SO₄, followed by heating at 110°C.

Extraction and isolation of iridoid monoterpenes, flavonoids and phenylpropanoids

Plant material of P. gentianoides and P. campanulatus was separated into various morphological parts (roots, stems, leaves and flowers), which were dried (in shade), milled and extracted with methanol. The samples of plant material studied were based on previous reports concerning the popular use of the plants as ethnomedicines (Domínguez et al., 2005, 2007; Jiménez-Estrada et al., 2006). As an example, leaves (of P. gentianoides) (1.2 kg) were extracted with methanol. The methanol extract (180 g) (A) was dried, redissolved in MeOH-H₂O (6:4), and subsequently partitioned with n-hexane (C), CH₂Cl₂ (D) and ethyl acetate (E). The remaining liquid yielded a residue (B) (Domínguez et al., 2007).

The ethyl acetate partition (46.22 g) exhibited strong antioxidant activity. The amorphous residue was further separated into eight fractions (I-VIII), by open column chromatography on silica gel (type G, 10-40 μm, Sigma-Aldrich, Toluca) (Domínguez et al., 2005). The fractions were eluted with different ratios of n-hexane-ethyl acetate (from n-hexane 100% increasing in 10% with ethyl acetate until 100% of ethyl acetate and finally addition of MeOH (up to 100%) to increase the polarity of the gradient. Each fraction was analyzed by TLC using ceric sulfate as the visualization system.

A series of iridoid monoterpenes, flavonoids, and phenylpropanoids were isolated from various fractions by conventional phytochemical procedures and collected for bioassay. In each case, the compounds were separated by vacuum chromatography column with silica gel (type G, 10-40 µm, Sigma-Aldrich) using a CH₂Cl₂-Me₂CO elution system (6:4). Another series of compounds were isolated from fractions by chromatography on Sephadex LH-20 with n-hexane-CH₂Cl₂-MeOH 2:1:1. Compounds 3 (110 mg), 6 (60 mg) and 7 (40 mg) were obtained from this type of column with CH₂Cl₂-MeOH (8:2). Compound 2 was isolated from crude methanol extract by column chromatography on Sephadex LH-20 using mixtures of EtOAc-MeOH (8:2) as the elution system.

The structural elucidation of each of these compounds has been previously reported (Domínguez et al., 2007; Jiménez-Estrada et al., 2006). All data were compared to literature values, those of our data base, and with authentic samples.

Anti-inflammatory activity

The TPA-induced mouse ear edema assay was based on the described method (Tubaro et al., 1985; Merlos et al., 1991; Della Loggia et al., 1996). Groups of 5 male CD-1 mice (25–30 g, were used, 5 units due to their existence in the laboratory) were anesthetized with Imalgen^®. A solution of 12-O-tetradecanoylphorbol-13-acetate (TPA, 2.5 μg) in acetone (10 μL) was applied topically to both faces (5 μL each face) of the right ear of the mice. The left ear faces received only acetone. Solutions of the extracts, compounds 1-7, diosmetin, luteolin, ovatifolin and indomethacin (as standard compounds) were prepared at concentrations of 0.05, 0.1 and 0.5 mg in 20 μL of acetone. These solutions were applied to both faces of the right ear (10 μL each face) 10 min after TPA treatment. The ears of control animals received only acetone. Four hours later, the animals were killed by cervical dislocation. A 9 mm diameter plug was removed from each ear. Swelling was assessed as the difference in weight between the right and left ear plugs. The percentage inhibition of edema was calculated by the equation: % = (edema A – edema B/edema A) × 100, where edema A = edema induced by TPA alone and edema B = edema induced by TPA plus sample.

Reduction of the 2,2-diphenyl-1-picrylhydrazyl radical

Extracts and partitions were chromatographed on TLC and examined for antioxidant effects by spraying the TLC plates with 2,2-diphenyl-1-picrylhydrazyl radical (DPPH) reagent. Specifically, the plates were sprayed with 0.2% DPPH in methanol. Quercetin and α-tocopherol were used as standards (Domínguez et al., 2005).

Reduction of crocin radical

Each of the solutions above was placed under UV₂₅₄ light. Following the decrease of absorbance, bleaching of crocin and fluorescence emission at 440 and 470 nm were monitored each 0.5 min (Céspedes et al., 2008; Domínguez et al., 2005).

Reduction of hydroxyl radical

Briefly, measurement of hydroxyl radical was based on the reduction of the absorbance of β-carotene treated with OH radicals generated from the reaction between hydrogen peroxide and ferrous ions (Soares et al., 2003).

Statistical analysis

Data shown in Table 1 consists of the mean results obtained from means of five animals and are presented as the percentage (%) of mean ± standard errors of the mean (SEM). Data were subjected to analysis of variance (ANOVA) with significant differences between means identified by GLM procedures and Student’s t-test analysis. The results are given in the text as probability values, with p <0.05 adopted as the criterion of significance, differences between treatment means were established with a Dunnett’s test. The EC₅₀ values for each activity were calculated by Probit analysis on the basis of the percentage of inhibition obtained at each concentration of the samples. EC₅₀ is the concentration producing 50% inhibition. All statistical analyses were performed by means of the MicroCal Origin 8.0 statistical and graphs PC program.

Table 1.  Inhibitory effect of extracts and compounds 1-9 and indomethacin on the TPA-induced inflammation in a mouse model^a.

Samples	Dose(mg/ear)	Edema inhibition (%)	ED50 (mg/ear)^b
Indomethacin^**	0.05	26.12 ± 2.9a	0.11
	0.1	46.01 ± 3.1b
	0.3	71.22 ± 5.5c
	0.5	74.77 ± 5.3c
MeOH-P.gent-leaves^*	0.1	27.9 ± 1.7a	0.65
	0.5	45.96 ± 3.8b
CH2Cl2-P.gent-roots^*	0.1	63.28 ± 4.9c	0.07
	0.5	89.75 ± 6.7d
CH2Cl2- P.gent –stems^*	0.1	36.12 ± 2.5a	0.16
	0.5	73.14 ± 3.7c
EtOAc-P.camp-leaves^*	0.1	34.55 ± 2.7a	0.18
	0.5	69.76 ± 5.1c
EtOAc-P.gent-leaves^*	0.05	22.45 ± 2.1a	0.15
	0.1	37.89 ± 2.7a
	0.5	75.67 ± 7.7c
Diosmetin (8)^**	0.05	8.46 ± 1.9d	0.45
	0.1	12.76 ± 1.8d
	0.3	35.39 ± 2.4a
	0.5	54.77 ± 3.2b
Luteolin (9)	0.5	86.9	0.25
Plantarenaloside (4)	0.5	35.6	0.7
Verbascoside (6)	0.5	40.3	0.62
Penstemide (2)	0.5	44.3	0.57
Catalpol (1)	0.5	29.0	0.87
Pensteminoside (3)	0.5	ND
Globularisicin (5)	0.5	ND
Martinoside (7)	0.5	ND

MeOH-P.gent, methanol extract of Penstemon gentianoides; EtOAc-P.gent, ethyl acetate extract of P. gentianoides; EtOAc-P.camp, ethyl acetate extract of P. campanulatus; ND, not determined.

^aEffects on ear edema of female mice CD-1. Means of five animals in independent experiments. Data expressed as % of the mean ± SD of weight of ear. All data analyzed with Student’s t-test. Values followed by the same letter are not significantly different.

^bEach value corresponds to concentration that inhibits 50% of edema development during bioassay stage.

In all values the non-supercript letters indicate that values followed by the same letter are not significantly different.

^*p < 0.05; ^**p < 0.01.

Results and discussion

Structural determination of compounds 1-9

The identification of compounds was achieved by comparison of their spectroscopic and physical data with those previously published (Domínguez et al., 2007; Jiménez-Estrada et al., 2006).

Anti-inflammatory activity

TPA-induced inflammation determined with the 12-O-tetradecanoylphorbol-13-acetate mouse ear method was inhibited most strongly by CH₂Cl₂ extracts of roots and stems from P. gentianoides and EtOAc extracts of leaves from P. campanulatus and P. gentianoides, with IC₅₀ values of 0.07, 0.16, 0.18 and 0.15 mg/ear, respectively (Table 1, Figure 1). Luteolin, diosmetin, penstemide, catalpol, verbascoside and plantarenaloside exhibited an inhibition of 86.9, 54.77, 44.3, 29, 40.3 and 35.6% at 0.5 mg/ear, respectively. This bioassay was carried out at 0.1 and 0.5 mg/ear with all compounds. The CH₂Cl₂ extract from roots of P. gentianoides, luteolin, and diosmetin proved to be the most active samples, therefore a dose-response curve was made with these substances, which had IC₅₀ values of 0.07, 0.25 and 0.45 mg/ear, respectively. Luteolin, diosmetin, and CH₂Cl₂ extract possessed dose-dependent anti-inflammatory activity that was compared to that produced by the commercially available anti-inflammatory drug indomethacin and a natural compound, ovatifolin, previously isolated from Podanthus ovatifolius Lag. (Asteraceae) (Céspedes et al., 2001) (Table 1). Each of the compounds assayed inhibited TPA-induced inflammation at 0.5 mg/ear. The CH₂Cl₂ extract from roots of P. gentianoides was even more active than indomethacin, a selective cyclo-oxygenase (COX) inhibitor, at 0.1 mg/ear (Table 1). Decrease in the anti-inflammatory activity was observed when iridoid monoterpenes with sugar moieties at C-1, catalpol 1, penstemide 2, and plantarenaloside 4, were assayed. Surprisingly, penstemide, which also has a sugar moiety at C-1, had intermediate activity; at 0.5 mg/ear this iridoid glucoside showed 48.9% inhibition. A similar, but less potent activity was observed when diosmetin (with a 3-O-methoxy group) was used instead luteolin (which possesses a 3-O-hydroxyl group). The compounds pensteminoside 3 and globularisicin 5 were not examined in this assay due to minute amounts available (although they were examined in the accompanying antioxidant activity).

Figure 1.  Chemical structures of catalpol 1, penstemide 2, pensteminoside 3, plantarenaloside 4, globularisicin 5, verbascoside 6, martinoside 7, diosmetin 8 and luteolin 9.

Interestingly, all extracts produced a curve of inhibition of logarithmic form, sometimes called the Monod function (Dockery & Klapper, 2001) (Figure 2). The activity of the compounds follows a linear form (from 0 to 0.9 mg of sample), but at higher concentrations (>1 mg sample) the plot changes to a horizontal line. We are presently studying the kinetics of inhibition of these plant extracts and compounds (Figure 2).

Figure 2.  Percentage of inhibition of edema by extracts and compounds from Penstemon gentianoides and P. campanulatus. (▪) indomethacin, (•) CH₂Cl₂ extract from roots of P. gentianoides, (▴) MeOH extract from leaves of P. gentianoides, (▾) CH₂Cl₂ extract from stems of P. gentianoides, (♦) Ethyl acetate extract from leaves of P. gentianoides, (◂) ethyl acetate extract from leaves of P. campanulatus, (▸) diosmetin, (★) luteolin.

Antioxidant capacity radical scavenging properties

Radical scavenging properties of extracts and compounds 1-9 were evaluated against the DPPH radical, using DPPH as a TLC spray reagent. Compounds 6-9 (10 μM) appeared as yellow spots against a purple background, whereas the same amounts of compounds 1-5 did not react with the radical. Compounds 6-9 were also tested against DPPH in a spectrophotometric assay. This method confirmed that compounds 6-9 exhibited the strongest radical-scavenging activity (Table 2). Compounds 1-5 were less sensitive to oxidation than compounds 6-9.

Table 2.  Amounts of phenolic content (mg/L ± standard error) from extracts and compounds from P. gentianoides and P. campanulatus needed to inhibit oxidative damage by 50%^a.

Sample	DPPH^b	Hydroxyl radical OH^•
		Crocin^c	β-Carotene
MeOH-P.gent-leaves	40.05 ± 2.13d	13.4 ± 1.63a	33.4 ± 1.56d
MeOH-P.gent-roots	55.41 ± 1.75c	63.44 ± 4.55e	56.9 ± 1.34f
CH2Cl2-P.gent-roots	>1000	>1000	>1000
MeOH-P.gent-stems	46.28 ± 2.66c	55.99 ± 6.71c	44.8 ± 2.59c
CH2Cl2-P.gent-stems	259.28 ± 13.57g	279.12 ± 14.44g	278.7 ± 20.12g
MeOH-P.camp-leaves	41.01 ± 2.15c	50.77 ± 5.66f	35.9 ± 1.99d
EtOAc-P.camp-leaves	58.9 ± 3.9e	60.71 ± 5.89e	45.6 ± 2.6c
Diosmetin (8)	22.48 ± 0.87b	21.1 ± 0.99b	16.7 ± 0.54b
Luteolin (9)	66.2 ± 1.34e	60.9 ± 2.11e	35.3 ± 1.44d
Plantarenaloside (4)	WA	WA	WA
Verbascoside (6)	87.25 ± 1.34h	71.9 ± 2.34h	45.8 ± 1.99c
Martinoside (7)	58.13 ± 1.24c	50.9 ± 3.54c	39.8 ± 1.23d
Penstemide (2)	WA	WA	WA
Catalpol (1)	WA	WA	WA
Pensteminoside (3)	WA	WA	WA
Globularisicin (5)	WA	WA	WA
Myricetin	21.5 ± 0.98b	22.9 ± 0.79	6.6 ± 0.34a
Quercetin	19.9 ± 0.88b	21.0 ± 0.67	6.1 ± 0.22a
α-Tocopherol	11.9 ± 0.43a	10.1 ± 0.47a	17.9 ± 0.67b
Gallic acid	10.8 ± 0.21a	15.3 ± 0.54a	12.7 ± 0.43a

^aValues expressed as μg/mL (ppm), Mean ± SD, n = 3. Different letters show significant differences at (p <0.05), using Duncan’s multiple-range test.

^bIC₅₀ for inhibition of DPPH radical formation.

^cIC₅₀ for bleaching of crocin. Values are expressed as μg/mL (ppm), See Methods for details. Mean ± SD, n = 3. Different letters show significant differences at (p <0.05), using Duncan’s multiple-range test.

WA, without activity, all these samples showed activity with concentrations greater than 500 ppm.

In all values the lowercase letters indicate that values followed by the same letter are not significantly different.

The antioxidant activity of the methylene chloride extract from roots of P. gentianoides and compounds 1-9 was also evaluated spectrophotometrically by an assay involving bleaching of the H₂O-soluble carotenoid crocin. Alkoxyl radicals were generated from t-BuOOH by UV photolysis of aqueous solutions containing 10 μM crocin and 1 mM t-BuOOH. t-BuOH (0.5 M) was added to scavenge the HO^• radicals produced. Gallic acid was added as a reference compound. Compounds 6-9 were all active and exhibited activities comparable to that of gallic acid (Table 2).

Finally, compounds 6-9 showed the greatest capacity to scavenge hydroxyl radicals generated by the interaction between hydrogen peroxide, ferrous ions and β-carotene as percentage inhibition of β-carotene oxidation (Table 2).

It should be noted that the criteria used for measurement of antioxidant activity were valid with concentrations from 0 to 500 ppm. For extracts greater than 500 ppm, anti-inflammatory activity was not correlated with antioxidant activity.

Conclusions

The CH₂Cl₂ extract of root of P. gentianoides was the best inhibitor of inflammation (ED₅₀ = 0.07 mg/ear) and had greater activity than the well-known COX inhibitor indomethacin (ED₅₀ = 0.11 mg/ear) (Table 1). Interestingly, this extract had the lowest antioxidant activity (I₅₀ >1000 ppm) (Table 2). We are presently examining the mechanism of action responsible for the anti-inflammatory activity. In this instance, we believe that the activity may be attributed to presence of steroidal compounds.

The flavonoids and phenylpropanoids that occur in these two Penstemon species have been considered as the active principles of other anti-inflammatory plants. Many iridoid monoterpenes and flavonoids have known inhibitory activity on the biosynthesis of nitric oxide, a compound implicated in physiological and pathological process such as chronic inflammation (Matsuda et al., 2000; Odontuya et al., 2005; López-Lázaro, 2009).

Acknowledgements

We thank M. Teresa Ramírez-Apan, Luis Velasco and Rocio Patiño for technical assistance; Chemistry Institute, UNAM [Universidad Nacional Autonoma de Mexico]. The correspondence author is indebted to David S. Seigler, Plant Biology Department, University of Illinois at Urbana-Champaign, for his useful comments and assistance in the revision of this paper. This work was supported in part by grant MEXUS-CONACYT, PAPIIT-DGAPA-UNAM Grants: IN243802-2 and IN211105-3. We also acknowledge and are grateful for an internal grant of the Departamento de Ciencias Básicas, Universidad del Bío-Bío, Chile.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.