Calea ternifolia Kunth, the Mexican “dream herb”, a concise review

In Mexico C. ternifolia and its members are widely known under the common names of “sacachichie”, “sacatechichi”, “zacachichi”, “zacatechi”, “zacatechichi”, or “techichic”, among other names listed in the Supplementary Material, Table S2². In Central America, the species is known as “simonillo”. The scientific name most widely used to recognize its members is Calea zacatechichi Schltdl., a name proposed by Schlechtendal in 1835 based on material supposedly collected somewhere in Colombia. Considering the geographic distribution of the plant, this account is likely an error because the species is distributed from northeastern Mexico (Nuevo León, Tamaulipas) and western Mexico (Nayarit and Jalisco) to Central America (Costa Rica) (Fig. 1).

Calea ternifolia is an earlier name (published in 1820) that taxonomically preceded that of C. zacatechichi (published in 1835). Accordingly, the proper name for its members is the former. In addition, due to its morphological variability along its area of distribution, the species includes a wide list of synonyms (see Supplementary Material, Table S1²).

Two subspecific taxa were recognized for the species: Calea ternifolia var. calyculata (B.L.Rob.) Wussow, Urbatsch & G.A. Sullivan (based on Calea zacatechichi var. calyculata B.L.Rob.) and Calea ternifolia var. hypoleuca (B.L.Rob. & Greenm.) B.L.Turner (based on Calea hypoleuca (B.L.Rob. & Greenm.). There is no agreement in recognizing subordinated taxa in the species. Most authors recognize a single although variable species; however, other authors place each one of them as different species, the former as the species Calea albida A.Gray (Pruski 2018) and the latter as C. hypoleuca.

Calea oliveri B.L.Rob. & Greenm. (published in 1901) is one additional name applied to specimens of C. ternifolia. This species is distributed in South America (Guyana, Venezuela) and its confusion with C. ternifolia probably comes from the original description, published by Oliver as C. ternifolia in 1886; however, as a later homonym, it is a rejected name substituted by C. oliveri.

Formally, the species was described by Kunth and published in the series published by Humboldt and Bonpland describing many American species. Calea ternifolia Kunth in Humb., Bonpl. et Kunth, gen. et sp. nov. folio ed. 4: 231 (1820[1818]). TYPE: Mexico, state unknown, Humboldt y Bonpland s.n. (Isotype: US!). In this work, we consider all specimens as part of the same species (C. ternifolia).

A formal description of the species (based on Wussow et al. 1985) is as follows: shrubs, 0.5–3 m tall; stems erect to lax and spreading, glabrous to pilose-tomentose; leaf blades broadly ovate, lance-ovate, or elliptic, 0.5–12 cm long, 0.5–7 cm wide, basally rounded or obtuse to acute, apically obtuse to acuminate, abaxially tomentose to glabrescent, usually resin-dotted, adaxially scabrous to glabrous, sometimes resin-dotted, trinerved, the margins coarsely crenate, serrate, or nearly entire; petioles 1–10 mm long, glabrous to densely pubescent and resin-dotted; capitulescence of axillary or terminal cymose or corymbose clusters of heads; primary peduncles 1–10 mm long, tomentose to sparsely pubescent; secondary peduncles lacking to ca. 3 cm long; capitula discoid or obscurely radiate, 5–15 flowered; involucres cylindric, 4–6(–8) mm tall, 2–4 mm wide, with phyllaries 4–6 seriate; outermost phyllaries broadly ovate or elliptic to suborbicular, 1–6 mm long, 0.5–3.5 mm wide, herbaceous or scarious throughout or herbaceous tipped, densely hirsute to essentially glabrous, usually resin dotted; inner phyllaries grading from ovate to narrowly elliptic, 2.5–6 mm long, 1–3 mm wide, scarious, apically rounded to acute, the margins and apices often ciliate and membranous, sometimes purplish; receptacle conic, 0.5–1 mm long and wide; paleae oblong, 5–6 mm long, 2–3 mm wide, scarious, conduplicate, apically blunt and laciniate, slightly yellowish; ray flowers mostly absent or 1–3, inconspicuous; ligules cream-colored to fairly bright yellow, shallowly to deeply 2–3(–4) lobed, ca. 2 mm long (not exceeding the disc florets in length), often resin-dotted; disc flowers 4–12(–20); corolla cream-colored; tube 2–3 mm long, usually glandular-dotted, flaring slightly at the base; throat ca. 0.5 mm long, dilated at the base; lobes acute, 1–2 mm long; anthers exserted; achenes (ray and disc similar) cylindric to somewhat prismatic, 2–3.5 mm long, glabrous to hirsute; light-colored carpopodium present; pappus of 8–12(–16) generally oblanceolate to occasionally lanceolate or oblong, stramineous squamellae, (0.5–)1–1.5(–2) mm long, apically rounded to acute, sometimes purplish at the base or throughout, the margins minutely laciniate; n = 19 (Fig. 2).

The species flowers almost all year round, but its flowering peak is reached between June and January. It shows a wide climatic tolerance, being found in tropical and temperate environments from semideciduous or deciduous tropical forests to temperate forests of pine, pine–oak, or cloud forests. The species also tolerates human disturbances, thriving along roads and open areas.

Most phytochemical studies have been carried out with plant material (aerial parts) collected in Mexico (Ortega et al. 1970; Chávez-Soto 1977; Quijano et al. 1979; Lee et al. 1982a, 1982b; Fischer et al. 1984; Ober et al. 1986; Martínez et al. 1987a; 1987b; Escandón-Rivera et al. 2012, 2017a, 2017b), although some investigations have been conducted with plants collected in Honduras (Herz and Kumar 1980), Guatemala (Bohlmann and Zdero 1977), El Salvador (Köhler et al. 2002), or purchased commercially (Wu et al. 2011; Sałaga et al. 2016). These works resulted in the isolation and characterization of a few specialized metabolites from C. ternifolia including 37 sesquiterpene lactones (1–37) (Ortega et al. 1970; Bohlmann and Zdero 1977; Quijano et al. 1979; Herz and Kumar 1980; Lee et al. 1982a, 1982b; Fischer et al. 1984; Ober et al. 1986; Martínez et al. 1987a; 1987b; Köhler et al. 2002; Wu et al. 2011; Escandón-Rivera et al. 2012, 2017a, 2017b), the sesquiterpene calicein II (38) (Chávez-Soto 1977), 2 triterpenoids (39 and 40) (Quijano et al. 1977), 4 chromenes (41–44) (Quijano et al. 1977; Escandón-Rivera et al. 2012, 2017a, 2017b), 2 simple coumarins (45 and 46) (Escandón-Rivera et al. 2017a, 2017b), 9 flavonoids (47–55), 6 flavones (47–52) (Herz and Kumar 1980; Martínez et al. 1987b; Martínez et al. 1987a, 1987b; Köhler et al. 2002; Escandón-Rivera et al. 2012, 2017a, 2017b), and 3 flavonols (53–55) (Escandón-Rivera et al. 2012, 2017a), 2 phenolic acids (56 and 57) (Escandón-Rivera et al. 2012, 2017a; Sałaga et al. 2016), 4 acetylenes (58–61) (Bohlmann and Zdero 1977; Quijano et al. 1977), and 2 common fatty acids (62–63) (Köhler et al. 2002). Table 1 summarizes the compounds isolated until now from the plant including their category, and chemical or common names as they appeared in the original publications. The corresponding structures of 1–61 are depicted in Figs. 3, 4, and Supplementary Fig. S1².

According to the data summarized in Table 1, the most relevant group of metabolites of the species are the sesquiterpene lactones. The isolated lactones are based on skeletons of heliangolides (1–23), germacranolides (24–36), and guaianolides (37). Most of these possess the lactone ring trans-fused at C-6 and C-7 and are acylated at C-8 with different acid moieties. These compounds were isolated using standard phytochemical procedures and were characterized mostly by spectroscopic and X-ray analyses. The references provided in Table 1 provide the details of such procedures.

Only one study of the essential oil of the plant has been carried out. The essential oil was obtained by hydrodistillation and the constituents were analyzed by gas chromatography coupled to mass spectrometry. The essential oil’s major constituents (see Supplementary Fig. S1²) were camphor, curcumene, spathulenol, caryophyllene oxide, β-eudesmol, and caleochromene A (41) (Escandón-Rivera et al. 2017b).

Díaz (1979) reported the experiences of 12 healthy people after drinking infusions or smoking cigarettes made up with the dried leaves of the plant. The individuals reported mild augmentation of sensorial perceptions, imaginings, thought gaps, and retrieval problems; lethargy and a short sleep with lively dreams followed these events. Later, it was found that the organic extracts of the leaves (hexane and methanol at doses of 30 and 86 mg·kg^–1, respectively) produced an augmentation of the reaction period and time-lapse perception in healthy persons. In contrast, in a nap-controlled investigation, these extracts increased the superficial stages of sleep, the number of spontaneous awakenings, and the recollection of having experienced vivid dreams (Mayagoitia et al. 1986). These organic extracts (2, 4, 6, 8, and 10 mg·kg^–1) also induced somnolence-like behaviors in freely moving cats in correspondence with electroencephalographic evidence of sleep activity in limbic structures. Interestingly, these extracts did not reproduce the changes in the electric brain activity induced by hallucinogens or dissociative anesthetics such as ketamine, phencyclidine, and SKF-10047 (Mayagoitia et al. 1986). The extracts also caused ataxia and vomiting in the cats.

Sałaga et al. (2016) evaluated the neuropharmacological effects of an aqueous extract (200, 400, and 800 mg·kg^–1) of the plant purchased commercially using three seizure models of convulsions, the elevated plus-maze and locomotor activity tests, as well as antidepressant-like behavior (forced swim and muscular strength in a grip tests). They established that the aqueous extract does not modify the 6 Hz psychomotor, phenyltetrazole, or maximal electroshock seizure thresholds in mice. In addition, the aqueous preparation did not influence muscle strength, exploratory behavior, nor anxiety levels at the doses tested. However, at the doses of 400 and 800 mg·kg^–1, the extract extended the immobility time of mice. The authors concluded that the aqueous extract of C. ternifolia did not change fundamental neurological events, levels of anxiety, and (or) exploratory behavior in mice after oral administration.

More recently, Martínez-Mota et al. (2021) analyzed the anxiolytic- and antidepressant-like effects of a lyophilized aqueous extract from the aerial parts of C. ternifolia in rodents; the doses ranged from 0.5 to 50 mg·kg^–1. Furthermore, they examined the extract action on hippocampal activity in the rat sleep–waking cycle. The anxiolytic- and antidepressant-like effects were evaluated using classical behavioral assays in rodents including those used by Sałaga et al. (2016). The results revealed that the extract induced precise and significant anxiolytic- and antidepressant-like effects in mice and rats. The results were compared with those of benzodiazepine. The same extract at 100 mg·kg^–1 produced visible moderate sedative effects in rats, associated with a significant rise in slow-wave sleep episodes during a 6 h experiment and enhanced fast frequencies of the hippocampus during rapid eye movement sleep. Their results are consistent with those of Mayagoitia et al. (1986) but differ significantly from Sałaga et al. (2016). The group of Martínez-Mota considered that the discrepancies might be attributed to some methodological variations including the dosage scheme; however, the difference could also be due to the plant material source, since Martínez-Mota et al. (1986) collected their material in Mexico while Sałaga et al. (2016) acquired it commercially. In conclusion, the work of Martínez-Mota et al. (2021) provided the best pharmacological evidence of the effects provoked by the dream herb described by MacDougall (1968), Díaz (1979), and Mayagoitia et al. (1986).

Venegas-Flores et al. (2002) showed that an aqueous extract of C. ternifolia (100 and 10 mg·kg^–1) prevented edema formation after administration of carrageenan in rat paws. The extract decreased the number of neutrophils migrating into the peritoneal cavity by 44% — the effect was comparable to those of the positive controls, namely dexamethasone (1 mg·kg^–1) and indomethacin (10 mg·kg^–1).

Segura-Cobos et al. (2010) established the antinociceptive effects of a methanol extract (25 and 50 mg·kg^–1) from C. ternifolia and its fractions (30 and 50 mg·kg^–1) using the acetic acid writhing test in mice. This effect was later corroborated by Sałaga et al. (2015, 2016), who tested an aqueous extract (200 mg·kg^–1) and a dichloromethane (200 mg·kg^–1) fraction in the writhing induced by acetic acid and mustard oil, respectively, in mice. However, when the extract was tested at the same dose in the hot plate test, also in mice, no activity was observed. In Sałaga’s studies, only one dose was assayed, and no positive control was included. Segura-Cobos et al. (2010) also demonstrated that the methanol extract (150 and 300 mg·kg^–1) and its chloroform fraction (50 mg·kg^–1) inhibited the edema induced by intraplantar injection of carrageenan in rats. One of the subfractions (2 mg·mL^–1) from the chloroform fraction inhibited the production of prostaglandin E2 (PGE2), a critical mediator of the inflammatory process, in macrophages stimulated by lipopolysaccharide to a similar extent to indomethacin (10 μg·mL^–1). The authors concluded that this fraction was rich in flavonoids and was responsible for the pharmacological effects.

Another investigation related to the anti-inflammatory action of C. ternifolia was carried out by Bork et al. (1997). They showed that the ethanol extract (100 µg·mL^–1) of the plant inhibited the nuclear factor-κΒ (NF-κΒ), which is critical for regulating cellular inflammation.

In mouse models, a dichloromethane fraction prepared from the methanol extract obtained from the plant’s aerial parts, significantly inhibited the contractility of the mouse colon in vitro (IC₅₀ = 17 μg·mL^–1). In vivo, the same fraction (200 mg·kg^–1, oral administration) increased gastrointestinal transit by 117 min with respect to the control (46 min, p = 0.0023); once more, in this investigation, no positive controls were used (Sałaga et al. 2015).

The first investigation on the plant’s antidiabetic action was reported by Ramos et al. (1992), who found that a decoction of the plant significantly decreased hyperglycemia during an oral glucose tolerance test in rabbits (50% dextrose solution, 4 mL·kg^–1).

Then, Ramírez et al. (2012) demonstrated that a hydroalcoholic extract (1 mg·mL^–1) of the plant inhibited the crude enzyme activity in Sprague Dawley rat intestines by 60%. In the same year, Escandón-Rivera et al. (2012, 2017a) reported that an aqueous extract (100 and 316 mg·kg^–1) from the aerial parts of C. ternifolia attenuated postprandial hyperglycemia in hyperglycemic mice (nicotinamide/streptozotocin, NA/STZ; 40 and 100 mg·kg^–1) during oral glucose and sucrose tolerance tests. The effects were compared with glibenclamide and acarbose, respectively; at the highest dose tested, the activities were better than those of the positive controls. The extract also showed hypoglycemic action in an acute experiment; at the dose of 316 mg·kg^–1, its effect was much better than that of glibenclamide. Finally, the extract showed inhibitory activity (IC₅₀ = 0.169 vs. 1.12 mg·mL^–1 for acarbose) against yeast-α-glucosidase. These pharmacological results demonstrated that C. ternifolia effectively controlled fasting and postprandial blood glucose levels in animal models. Bioassay-guided fractionation of the active extract using the α-glucosidase inhibitory assay led to the identification of 6-hydroxyacetyl-5-hydroxy-2,2-dimethyl-2H-chromene (43), calein C (26), acacetin (47), isorhamnetin (53), and quercetin (55) as the most active α-glucosidase inhibitors, with IC₅₀ = 0.42, 0.28, 0.16, and 0.53 mmol·L^–1, respectively, in comparison to acarbose (IC₅₀ = 1.7 mmol·L^–1). Kinetic analysis revealed that the chromene 43 behaved as a non-competitive inhibitor. Caleochromene A (41) and caleins A (24) and C (26) (3.16–31.6 mg·kg^–1 per oral), the major components of the infusion, were able to control postprandial glucose levels during an oral sucrose tolerance test (3 g·kg^–1) in normal and NA/STZ (40 and 100 mg·kg^–1) hyperglycemic mice. The effects were comparable to those of acarbose (5 mg·kg^–1). The essential oil (31.6, 100, and 316 mg·kg^–1 per oral) from C. ternifolia was also active in the same in vivo assay. The presence of caleochromene A (41) at 20% in the oil was consistent with this good activity in vivo (Escandón-Rivera et al. 2017b).