Ethnomedicinal uses, phytochemistry and pharmacological aspects of the genus Premna: a review

Abstract Context: The genus Premna (Lamiaceae), distributed throughout tropical and subtropical Asia, Africa, Australia and the Pacific Islands, is used in folk medicine primarily to treat inflammation, immune-related diseases, stomach disorders, wound healing, and skin diseases. Objectives: This review exhaustively gathers available information on ethnopharmacological uses, phytochemistry, and bioactivity studies on more than 20 species of Premna and critically analyzes the reports to provide the perspectives and directions for future research for the plants as potential source of drug leads and pharmaceutical agents. Methods: A literature search was performed on Premna species based on books of herbal medicine, major scientific databases including Chemical Abstract, Pubmed, SciFinder, Springerlink, Science Direct, Scopus, the Web of Science, Google Scholar, and ethnobotanical databases. Results: More than 250 compounds have been isolated and identified from Premna species, comprising of diterpenoids, iridoid glycosides, and flavonoids as the most common secondary metabolites, followed by sesquiterpenes, lignans, phenylethanoids, megastigmanes, glyceroglycolipids, and ceramides. Many in vitro and in vivo studies have been conducted to evaluate the biological and pharmacological properties of the extracts, and isolated compounds of Premna species with antimicrobial, antioxidant, anti-inflammatory, immunomodulatory, antihyperglycaemia, and cytotoxic activities. Conclusion: The bioactive compounds responsible for the bioactivities of most plants have not been well identified as the reported in vivo pharmacological studies were mostly carried out on the crude extracts. The isolated bioactive components should also be further subjected to more preclinical studies and elaborate toxicity study before clinical trials can be pursued.


Introduction
The genus Premna was previously classified within the family Verbenaceae (Munir 1984), but has been transferred into the family Lamiaceae, subfamily Viticodeae (Harley et al. 2004;Olmstead 2010Olmstead , 2012. Currently, this genus contains 200 species which are mainly distributed throughout tropical and subtropical Asia, Africa, Australia, and the Pacific Islands (Harley et al. 2004). There are 46 species recognized in the Flora of China (Tan & Li 2014) and 14 species occurring in the Flora Malesiana area (de Kok 2013). The word 'Premna' is derived from the Greek 'premnon', meaning tree stump, which refers to the short and twisted trunks of P. serratifolia L., the first collected species of this genus. Based on the shape and number of calyx lobes, the genus Premna has been subdivided into five sections: Holopremna Briq. (consisting of two subsections: Thyrsoideae and Corymbiferae), Odontopremna Briq., Gumira (Rumph. ex Hassk.) Briq., Premnos Briq., and Holochiloma Briq. (de Kok 2013).
Morphologically most species in the genus Premna are small trees or shrubs and rarely found as lianas (P. trichostoma Miq.) and pyroherbs (P. herbacea Roxb.). Some species have young twigs with a series of small decussate triangular scales at the base which will fall off once the branch is older. The leaves are usually decussate and hairy. A ridge is often present between the petioles. There are two shapes of calyx types. The first one has four isomorphic lobes, the shape remaining largely intact when the flower develops and when the fruits are formed. The second type has 0-5 lobes, usually heteromorphic. There are also two fruit types: a globose drupe-like fruit consisting of four fleshy mericarps with one seed each, and a clavoid, almost single-seeded, drupe-like and consisting of one fleshy mericarp (de Kok 2013).
Our review of the genus Premna is based on ethnomedicinal uses, phytochemical investigations, and pharmacological attributes. This review is comprised of more than 20 species of Premna from 150 publications. It is noted that some species have recently been considered as synonyms based on current plant taxonomy (The Plant List 2013). For example: P. obtusifolia R.Br., P. integrifolia Willd., and P. corymbosa var. obtusifolia (R.Br.) H.R. Fletcher are synonyms to P. serratifolia; P. japonica Miq. is a synonym to P. microphylla Turcz.; P. latifolia Roxb. as a synonym to P. mollissima Roth. However, in order to avoid any confusion, we continue to use the species names as referred to by the author(s) of the original papers. The detailed information gathered and critically analyzed in this review should be useful as reference for phytochemists, pharmacologists, medicinal

Iridoid and iridoid glycosides
Iridoids are monoterpene lactones which usually occur in plants as glycosides and sometimes are known as monoterpene alkaloids. They can be found in dicotyledone angiosperms within the superorders Corniflorae, Gentianiflorae, Lamiiflorae and Loasiflorae (Ghisalberti 1998). Their structures are based on cyclopentan[c]pyran skeleton represented as iridane (cis-2-oxabicyclo[4.3.0]nonane) and seems to be biosynthesized via alternative cyclization of geranyl diphosphate (Sampaio-Santos & Kaplan 2001). The name 'iridoid' itself comes from iridodial and related compounds isolated from the defense secretion of Iridomyrmex species (Tietze 1983). Classification of naturally occurring iridoids involves large groups, yet there are four distinguish classes i.e. the non-glycosidic iridoids, iridoid glycosides, iridoid acetal esters, and secoiridoid glycosides. Our current review has identified more than 53 iridoid glycosides within nine species of Premna (Table 2). Most of the isolated iridoids are catalpol derivatives  although mussaenosidic acid, epiloganic acid and gardoside derivatives [139][140][141][142][143][144][145][146][147]169] also could be identified in quite a great number. Majority of the iridoids are linked to their glycosides at C-1 though in catalpol, the glycoside could have linked to C-6. Interesting structure was displayed by compound 168, with two catalpol glycosides formed an ester to truxinic acid. Piscrosin D [148] was the only non-glycoside iridoid isolated from P. japonica (Otsuka et al. 1991b) and P. serratifolia , respectively. Figure 2 shows the structures of some of the iridoid and iridoid glycosides.

Flavonoids, xanthones and chalcones
The occurrence of these flavonoids was reported from 13 species ( Table 2) Figure 3. The skeleton structure resemblance of the flavonoids (C 6 -C 3 -C 6 ), xanthones (C 6 -C 1 -C 6 ) and chalcones (C 6 -C 3 -C 6 , without a heterocylic C-ring in the three-carbon a,b-unsaturated carbonyl system) suggested they shared a similar shikimate pathway via phenylpropanoid pathway in their biosynthesis whereas xanthones, in particular, might represent the modified shorthened forms of the C 6 -C 3 system (Dewick
Compound 48 (Salae et al. 2012) appeared to have potent antibacterial activity with most of their MICs were <5 lg/mL, except for P. aeruginosa. Interesting broad spectrum antibacterial and antifungal activities were also showed by compound 72 (MIC 5-10 lg/mL), isolated from the roots of P. herbacea (Murthy et al. 2006). Another study by Lirio et al. (2014) evaluated antitubercular activity against Mycobacterium tuberculosis of the leaves of P. odorata and its constituents. Although the extract showed relatively weak inhibitory activity, the fractions exhibited strong activity which eventually led to isolation of the active compound 4 (MIC 90 8 lg/mL whilst rifampin 0.05 lg/mL and isoniazid 0.23 lg/mL).
The insecticidal activity of different extracts and essential oil of P. latifolia was tested against Spodoptera litura larvae, a polyphagus crop pest, by using leaf-dip method. The essential oil showed the highest growth reduction (56.83%) followed by chloroform, hexane and butanol fractions (43.93, 26.01 and 23.69%, respectively) (Kumar et al. 2011). Recent study on P. angolensis and P. quadrifolia evaluated the insecticidal and repellent effects of its essential oils against Sitoroga cerealella, an insect pest of rice stocks, using olfactometer and contact toxicity test (Adjalian et al. 2015). The results showed that both essential oils have insecticidal and repellent activities as indicated by rate of death of S. cerealella, percentage of repulsion, number of rice attacked and loss of weight of rice. The leaf extract of P. serratifolia showed strong activity against Leishmania donovani (IC 50 4.4 lg/mL) but showed weak and/or no effect against Trypanosoma brucei brucei, Trichomonas vaginalis and Caenorhabditis elegans (Desrivot et al. 2007). It has been reported previously that clerodane diterpenes [28 and 29], isolated from P. oligotricha and P. schimperi, showed potent antileishmanial effects towards axenically cultured amastigotes of L. aethiopica (IC 50 1.08 and 4.12 lg/mL, respectively). Both compounds also exhibited high selectivity towards L. amastigotes than the permissive host cell line, THP-1 cells or the promastigotes stage of the parasites (Habtemariam 2003).
Although widely used traditionally in malarial treatment by the Philippines, the ethanol extract of P. angolensis barks only showed weak antiplasmodial activity (IC 50 180-500 lg/mL) towards both chloroquine sensitive and resistant strains of Plasmodium falciparum (do C eu de Madureira et al. 2002). However, the leaf extract of P. chrysoclada revealed high activity against chloroquinone sensitive and resistant strains of P. falsiparum (IC 50 7.75 and 9.02 lg/mL) while the root extract only showed moderate activity (IC 50 27.63 and 52.35 lg/mL). Further investigation also revealed that the leaf extract (dose 250 mg/mL) has strong ability to reduce the parasitized erythrocyte (9.26% parasitaemia) and to inhibit the parasite growth (65.08% chemo suppression) in Plasmodium berghei infected mice (Gathirwa et al. 2011).

Antioxidant, anti-inflammatory and immunomodulatory activities
Premna species are known to have high-antioxidant capacity, such as P. cordifolia Roxb. (Mustafa et al. 2010;Mohd Nazri et al. 2011), P. esculenta Roxb. (Mahmud et al. 2011), P. integrifolia (Gokani et al. 2011;Nguyen & Eun 2011), P. microphylla  and P. serratifolia (Rajagopal et al. 2014) (Table 3). The wide distribution of flavonoids and phenolics within this genus seems to contribute to this activity. Various methods were used to measure the antioxidant capacities such as radical scavenging (diphenylpicrylhydrazyl (DPPH), superoxide, nitric oxide NO, hydroxyl radicals), ferric reducing ability of plasma (FRAP), ferric thiocynate (FTC), lipid peroxidation, erythrocyte membrane stabilizing and b-carotene bleaching assays. Most of the radical scavenging capacity of the extracts has been correlated to their phenolic contentsthe higher the phenolic content, the higher the antioxidant capacity. The presence of hydroxyl group (OH) and/or unsaturated bond are suggested to play the main role in capturing the radical oxygen species (ROS).
Secondary metabolites such as flavonoids, xanthones, chalcone and other phenolic compounds with high-hydroxyl group substitution are hypothetically contributing to the high antioxidant activity of the plant. For example, two flavone glycosides [213,214] from P. latifolia leaves significantly inhibited oxidation of DPPH (IC 50 22.5 and 16.0 lg/mL, respectively) (Ghosh et al. 2014). Furofuran lignans [208,209] and iridoid glycosides [150,154,161,165] might contribute to antioxidant activity of the stem bark of P. integrifolia when evaluated with radical scavenging (DPPH and NO) and ferric reducing antioxidant power (FRAP) assays (Yadav et al. 2013). Compounds 165 and 154 possessed maximum radical scavenging activity (IC 50 0.29 and 0.37 lM) in DPPH assay, followed by compound 209; while compounds 150 and 161 exhibited maximum reducing power in FRAP assay. Aldehyde derivatives [186 and 187] and icetexane diterpenes [81,82] were thought to be potential free radical scavenger constituents from P. tomentosa (Ayinampudi et al. 2012;Ayinampudi 2013). The higher number of hydroxyl group in compound 82 (IC 50 7.01 lg/mL) than compound 81 (IC 50 24.80 lg/mL) reflected the higher antioxidant capacity of the former. Interestingly, this rule was not applied for compound 187 (IC 50 20.58 lg/mL) which has three hydroxyl moieties, in comparison to compound 186 (IC 50 20.83 lg/mL) which only has one hydroxyl moiety. Potential antioxidant activities were also exhibited by a series of icetexanes [73-76] from P. tomentosa towards DPPH, NO and superoxide scavenging assays, of which compound 76 demonstrated superior activities than the others and also on par with the standards (Naidu et al. 2014). Recent study also identified an aromatic diterpene [53] as antioxidant constituent from P. serratifolia with IC 50 of 20.4 ± 1.3 lM towards DPPH assay (Habtemariam & Varghese 2015).
It is note worthy that although those studies showed some potential antioxidant capacities of some extracts of Premna species and its constituents, they do not necessarily reflect the molecular or in vivo activities. For example, the DPPH and Methods: egg albumin-induced paw edema (acute inflammation model) and cotton pelletinduced granuloma formation (chronic inflammation model); both in rats. Findings: The extract significantly inhibited the edema in acute inflammation model dose dependently while in chronic model the results indicated mild but significantly decreased granuloma formation (% inhibition 35.17% and 50.38% at doses 200 and 400 mg/kg, respectively). P. herbacea (Narayanan et al. 2000) EtOH ext.; roots Anti-inflammatory, Antipyretic, Antinociceptive Dose: 100, 200, 400 mg/kg Methods: carrageenan-induced paw edema (acute inflammation model) and cotton pelletinduced granuloma formation (chronic inflammation model); both in rats. Antipyretic: Typhoid-Paratyphoid A, B (TAB) vaccine-induced pyretic in rabbits. Antinociceptive: acetic acid-induced writhing and hot plate tests on mice. Findings: The extract significantly showed antipyretic and antinociceptive effects on particular animal models. The extract did not reduce edema's volume in the acute inflammation rat and only showing mild yet statistically significant anti-inflammation in chronic model. All, except antinociceptive activity on hot plate test, was shown to be dose dependent. P. integrifolia (Gokani et al. 2011) MeOH ext.; roots Anti-inflammatory Dose: 300 mg/kg Methods: In vivo: acute inflammation models (carrageenan-induced edema, histamineinduced wheal formation, formalin-induced edema, acetic acid-induced vascular permeability) and chronic inflammation model (cotton pellet-induced granuloma). In vitro: COX-1 inhibitory activity using spontaneous contractions of the rat's uterus and heat-induced hemolysis of rat's erythrocytes. Findings: The extract showed significant reduced both acute and chronic edema/granulation in inflammation models which were supported by significant prostaglandin synthase inhibition (% inhibition was 30.43%) on rat's uterus and stabilization of plasma membrane of rat's erythrocyte (conc 50, 100 and 150 lg/mL).
FRAP assays are mostly based on the simple chemical reaction (Benzie & Strain 1996;Molyneux 2004). These cell-free antioxidant assays do not support the cellular physiological conditions, do not include particular biological substrates that need to be protected, may not encounter the relevant types of antioxidant at molecular level, may not describe the partition coefficient of the compounds, or other cellular factors. Cell-based antioxidant assays are considered more relevant and accurate in representing the in vivo conditions since they involve several aspects such as uptake, metabolism, and target site where the compounds might potentially worked within cells (L€ u et al. 2010). Inflammatory reaction occurs due to pathogen invasion into the body or other types of body injury which can cause injury to the tissues or cells as well. At macroscopic level, inflammation is indicated by reddened, swollen, hot, pain, and loss of function of the inflamed area. The loss of function is usually referring to simple loss of mobility in a joint due to pain or edema, or the replacement of functional tissue by the scar tissue. This inflammatory event usually will be followed by the release of mediators from the cells or plasma which modify and regulate the immune response (innate/nonspecific and specific immunological response) (Punchard et al. 2004). Hence, several studies have been conducted to evaluate the anti-inflammatory effect of the extracts of Premna species (Table 3). In addition, an extensive study by Salae et al. (2012) identified several compounds from P. obtusifolia roots that exhibited potent anti-inflammation activity. Of 20 isolated compounds, four diterpenes [48,49,69,70] showed potent in vitro lipopolysaccharide (LPS) induced NO inhibitor (IC 50 6.1, 7.8, 1.7 and 6.2 lM) that were comparable to positive control, caffeic acid phenylester (IC 50 5.6 lM). Meanwhile, megastigmane [21] only showed weak anti-inflammatory activity. Further structure-activity relationship analysis suggested that the presence of a hydroxyl group in an orthonaphtoquinone skeleton provided stronger anti-inflammation activity. It was postulated that these active compounds might be responsible for the strong NO inhibitor activity of the hexane and dichloromethane extracts (IC 50 4.3 and 6.1 lg/mL, respectively). Another species, P. integrifolia, also showed significant  (Mahire et al. 2009) MeOH ext.; leaves Anti-inflammatory Dose: 125, 250 and 500 mg/kg Methods: carrageenan-induced paw edema, cotton pellet-induced granuloma, and acetic acid-induced vascular permeability models. Findings: The extract exhibited significant anti-inflammatory activity on those three animal models, dose dependently. P. latifolia (Kumari et al. 2011) Water ext.; leaves Anti-inflammatory Dose: 9 mL/kg Methods: carrageenan-induced paw edema in rats. Findings: The extract showed significant reduced in the edema after 60 min of the edema induction, and the findings showed better results than P. obtusifolia and on par with indomethacin. P. obtusifolia (Kumari et al. 2011) Water ext.; leaves Anti-inflammatory Dose: 9 mL/kg Methods: carrageenan-induced paw edema in rats. Findings: The extract showed significant reduced in the edema after 60 min of edema induction. P. obtusifolia (Salae et al. 2012) Hexane and CH 2 Cl 2 ext.; roots Anti-inflammatory Concentration: 0, 3, 10, 30 and 100 lg/mL. Methods: LPS-induced nitric oxide (NO) production by murine macrophage-like RAW 264.7 cells. The NO production was measured by using Griess assay. Findings: Both extracts significantly inhibited NO production that comparable to caffeic acid phenylester (positive standard, IC 50 5.6 lg/mL), with IC 50 4.3 (hexane) and 6.1 (CH 2 Cl 2 ) lg/mL. P. serratifolia (Rajendran & Krishnakumar 2010) EtOH ext.; woods Antiarthritis Dose: 300 mg/kg Methods: Freund's adjuvant-induced arthritis rats, where suspension of killed Mycobacterium tuberculosis (0.5%) in liquid paraffin was injected into the left hind paw, and the changes in paw edema were measured. Findings: The extract inhibited the edema by 68.32% after 21 days (indomethacin showed 74.87% inhibition). In hematological parameter, treatment with the extract significantly decreased the total whole blood count (WBC) and erythrocyte & sedimentation rate (ESR), but increased the red blood count (RBC) and hemoglobin (Hb) level. P. serratifolia (Rajagopal et al. 2014) MeOH ext.; flowers Anti-inflammatory Concentration: various, 10-1000 lg/mL Methods: in vitro HRBC membrane stabilization, with measured parameter was inhibition of HRBC membrane lysis. Findings: Starting at concentration 100 lg/mL, the extract showed an anti-inflammatory activity with percentage inhibition at 69.41 ± 0.12 lg/mL. The percentage inhibition appeared in linearity with concentration, and at 300 lg/mL, the extract exhibited inhibition at 97.30 ± 0.59 lg/mL. P. tomentosa (Alam et al. 1993) MeOH ext.; leaves Anti-inflammatory Dose: 100 mg/kg Methods: cotton pellet-induced granuloma in rats. Findings: The extract caused a reduction of granuloma by 32.21%, in comparison to phenylbutazone (positive control) which was 33.77%. There was also a decreased in serum protein, SGOT and SGPT.
Other activities such as antioxidant, antidiabetic/antihyperglycaemic, antihyperlipidemic, hepatoprotective and cardioprotective activities are discussed in the main article.
in vivo anti-inflammatory activity in both acute and chronic inflammation models; further in vitro study suggested inhibition of prostaglandin synthase and stabilization of plasma erythrocyte membrane might play role in the in vivo activity (Gokani et al. 2011).
Only one calculogenesis-related study has been carried out on Premna. The anticalculogenic activity of P. latifolia leaves and stems was evaluated in vitro by assessing oxalate crystal growth on gel medium in Hane's tubes via single diffusion method over period of 30 days at the concentrations of 20 and 200 mg/mL (Aravindakshan & Bai 1996). The extract effectively reduced the size of oxalate crystal in comparison to negative control and further analysis by using scanning electron microscope showed development of cracks in the crystal interior and rupture tendency. These results concluded chemolysis as an anticalculogenic mechanism of this extract.
Interesting immunostimulant activity was exhibited by P. pubescens Blume and P. tomentosa leaves. In their in vitro studies, Devi et al. (2003aDevi et al. ( , 2004a) used rat's splenic lymphocytes and J770 macrophage cell culture which has been induced by using chromium, Cr(IV), to provide immunosuppressant condition. The results showed P. tomentosa inhibited the apoptosis of the Cr(IV)-induced cells by preventing the proliferation of the lymphocytes and the macrophages. At the same time, the extract has significantly reduced the ROS level by increasing the levels of the endogenous antioxidant enzymes such as glutathione (GSH), glutathione peroxide (GPx) and superoxide dismutase (SOD) enzymes, and reducing malondialdehyde (MDA) level. Meanwhile, in vivo study by Restuati et al. (2014) in the antigen sheep red blood cell (SRBC)-induced immunostimulant rats, suggested that P. pubescens stimulated the immune response by increasing the number of leukocytes, immunoglobulin IgG and IgM, and lysozyme. In addition, the methanol extract of P. integrifolia roots also produced significant immunomodulatory activity in both specific and nonspecific immune responses following hemagglutinating antibody titer, plaque forming cell assay, delayed-type hypersensitive response, carbon clearance test (phagocytic activity) and E. coli-induced abdominal sepsis parameters (Gokani et al. 2007).

Cytotoxic activities
Traditional use of P. herbacea by the Thai to treat cancer has led to the evaluation of the rhizome extract of this species towards several cancer cell lines such as COR-L23, LS-174 T and MCF-7 (Itharat et al. 2004). The results turned out to be negative. However, another study by Dhamija et al. (2013), showed that the root nodules extract had cytotoxic activity on brine shrimp lethality test (BSLT), Ehrlich ascites carcinoma (EAC) cells (trypan blue dye exclusion assay), and MCF-7 cell lines (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay). Ethanol extract and ethyl acetate fraction exhibited the most potent cytotoxic effect and further investigation on EAC-inoculated mice and Dalton's lymphoma ascites (DLA) mice (250 and 500 mg/kg, orally) led to significant elevation of the mean survival rate and reduction of the solid tumor weight and volume. These findings were supported by hematological and antioxidant parameters. The EAC-inoculated mice model was used to evaluate antitumor activity of the ethanol extract of P. integrifolia; the findings were found to be comparable to the standard, 5-flurouracil (20 mg/kg) (Sridharan et al. 2011).

Antidiabetic/antihyperglycaemic and antihyperlipidemic activities
So far, four species of Premna has been studied for their antidiabetic properties. The most common method was using a chemically-induced diabetic animal model. Alloxan-induced hyperglycemic rats have been used to evaluate antidiabetic activity of ethanol extract of P. integrifolia at a dose of 250 mg/kg, to confirm the hypolgycaemic activity of this herbal based on the Indian folk medicines (Kar et al. 2003). This activity was further evaluated by Mali (2013) using cafeteria diet induced mice (inbreed) through various parameters (body-mass index, blood glucose, lipid profile, histology valuation) and comparison with a standard drug (simvastatin). The findings indicated significant protective effect of the roots of P. integrifolia at doses of 200 and 400 mg/kg. P. corymbosa Rottler & Willd., also reduced blood glucose level in both normolgycaemic and alloxan-induced hyperglycemic rats, at doses of 200 and 400 mg/kg (Dash et al. 2005). Similar studies by Ayinampudi et al. (2012) and Ayinampudi (2013) successfully identified two diterpenes [81,82], and two aldehydes [185, 187] that were responsible for antihyperglycaemic activity of P. tomentosa root by inhibiting enzyme a-glucosidase in vitro (IC 50 values were 22. 58, 9.59, 18.41, and 12.11 lg/mL, respectively). One clinical study, based on the Ayurvedic system, evaluated the effectiveness of P. obtusifolia roots as an alternative treatment for diabetics (Ghosh et al. 2009). This 9-month study involved 50 patients with a history of obesity. The results showed significant reduction on body-mass index (BMI), atherogenic index and waist-hip ratio after 6 months while the uric acid and mid-triceps skin fold thickness were significantly reduced after 9 months.
The in vivo evaluation of antihyperlipidemic activity of herbal extract is normally done by determining the lipid profiles (LDL, HDL, triglycerides, cholesterol) and histology parameters. As mention earlier, P. tomentosa leaves extract showed antihyperlipidemic activity towards the animal model by improving lipid profile and reducing lipid metabolizing enzymes (Devi et al. 2004c). Meanwhile, Mali (2013) reported the effect of P. integrifolia roots on lipid profile parameters of caffeinated-diet mice. Additionally, the antihyperlipidemic effect of the leaves and roots of P. esculenta was evaluated in vivo by using Poloxamer 407-induced hyperlipidemis mice and rats (Mahmud et al. 2011). The study was designed for single dose (mice, 500 mg.kg, i.p) and repeated dose (rats, 4 days, 250 mg/kg, p.o), and the results suggested the extract significantly reduced the serum total triglycerides, total cholesterol, LDL and VLDL levels which were comparable to the standard drug, atorvastatin.

Hepatoprotective and cardioprotective activities
Premna tomentosa has been extensively studied for its hepatoprotective activity. Devi et al. (1998Devi et al. ( , 2004bDevi et al. ( , c, 2005 have evaluated the possible protection mechanisms of the extract of P. tomentosa leaves on acetaminophen-induced hepatoxicity in rats, which suggested via (i) reducing ROS and generating endogenous antioxidant enzymes in the liver (e.g. glutathione system, superoxide dismutase, catalase); (ii) improving lipid profile and reducing the activities of lipid metabolizing enzymes; (iii) decreasing the acetaminophen-induced membrane damage so that total membranebound ATPases would improve and eventually help maintaining active transport and balancing of Na þ , Ca 2þ and K þ in the liver and serum; and (iv) protecting the liver against mitochondrial damage as the mitochondrium contains enzymes that would catalyze the production of lipid peroxidation products and other toxic metabolites. Additionally, Hari Prasad et al. (2006) postulated the protective mechanism of P tomentosa towards dimethylnitrosamine (DMN)-induced hepatic fibrosis was through decreasing the activation of liver stellate cells and accumulation of collagen and other connective tissue proteins. Recently Naidu et al. (2014) reported that the in vitro (using HepG2 cells) and in vivo (using tBHP-induced hepatic damage mice) hepatoprotective activity of compound 76 increased the viability of hepatic cells and decreased the elevation of serum transferases (SGOT/ SGPT) and oxidative damage, including lipid peroxidation. P. corymbosa and P. serratifolia also showed protective activity on chemically induced (carbontetrachloride (CCl 4 ) and paracetamol, respectively) hepatic damage in rats (Karthikeyan & Deepa 2010;Singh et al. 2011).
Two species, P. mucronata Roxb. (Patel et al., 2012;Savsani et al., 2014) and P. serratifolia , are reported to have cardioprotective activity towards a myocardial infarction rat model. The extracts provided protection to the heart via several mechanisms, i.e., (i) decreasing injured cardiac marker enzymes; blood glucose; heart tissue protein; and heart tissue nucleic acids; as well as (ii) maintaining the electrocardiogram (ECG) pattern and hemodynamics changes, increasing myocardial glycogen and restoring antioxidant status. Further investigation has ruled out cardiac stimulant activities of P. serratifolia extracts by significantly supporting positive inotropic and negative chronotropic actions similar to that of b-adrenergic effect, decreasing membrane Na þ K þ ATPase and Mg 2þ ATPase and increasing Ca 2þ ATPase . There was only one study reporting the gastroprotective activity of P. serratifolia leaves on aspirin-induced ulcer rats (Jothi et al. 2010). The evaluation was carried out at doses of 200 and 400 mg/kg by looking at several parameters: lesion index, total-and free-acidity, and percentage of ulceration. The findings suggested that P. serratifolia exhibited significant antiulcer and anti-secretory activities in both applied doses.

Neuropharmacological acitivities
So far, two studies have evaluated the hypnotic and the neuropharmacological effects of Premna species on animal models. Devi et al. (2003b) evaluated the effects of the methanol extract of P. tomentosa leaves as a central nervous system (CNS) depressant using potentiation of phenobarbitone-induced hypnotic and locomotor activities on rats. At doses of 400 and 500 mg/kg orally, the extract decreased the locomotor activity and moderately increased the sleeping time, that were comparable to CNS depressant, chlorpromazine (10 mg/kg, i.p) yet significantly different to CNS stimulant, ephedrine hydrochloride (10 mg/kg, i.p). A recent study also evaluated the effect of P. integrifolia bark on locomotor activity of the rats in the open field and hole-cross tests (Khatun et al. 2014). The findings suggested that P. integrifolia significantly affected locomotor activity of the rats at the doses of 250 and 500 mg/kg, orally on both methods, therefore, might act as CNS depressant.

Discussion
This review summarizes the phytochemical work of more than 19 species (24 species once the synonyms are considered) of Premna with more than 250 secondary metabolites have successfully been isolated and identified. It comprises a high number of diterpenes, iridoid glycosides and flavonoids (glycosides and glycones), followed by sesquiterpenes, lignans, phenylethanoids, megastigmanes, glyceroglycolipids and ceramides. Xanthones and alkaloids were rarely identified though a few studies reported their presence in this genus. Meanwhile essential oils were reported in seven species. The distribution of identified secondary metabolites within the genus Premna is shown in Table 3.
Although the Premna genus is rich in diterpenes and iridoid glycosides, they were not well distributed within the studies species. Diterpenes were abundant in three species such as P. mollissima, P. serratifolia, and P. tomentosa while iridoid glycosides were reported abundantly in P. serratifolia, P. subscandens and P. microphylla. On the contrary, flavonoids seem to be well distributed among 16 reported species despite of their low number in comparison to other groups. Only a few species such as P. serratifolia, P. microphylla, P. mollissima, P. fulva and P. subscandens, have been extensively studied for their secondary metabolites. Nonethless, a previous review (Taskova et al. 1997) endorsed terpenoids, iridoids, and flavonoids to be used as taxonomic markers in the family Lamiaceae based on their occurrence in 39 species of 25 genera such as Sideritis, Stachys, Lamium, Phlomis, Ballota, Salvia, Ajuga, Teucrium. Thus, diterpenoids (icetexane, abietane, labdane, pimarane types), iridoid glycosides (catalpol derivatives), and flavonoids (flavonols and flavones) can be very useful to characterize the taxonomic markers of the genus Premna (Taskova et al. 1997) and to provide the secondary metabolite fingerprint of each species through infrared (IR), thin layer chromatography (TLC), high performance liquid chromatography (HPLC), mass spectroscopy (MS), or nuclear magnetic resonance (NMR) analysis.
Some of the biological and pharmacological studies reported on the studied plants have suggested scientific evidence to justify the various plant uses in traditional medicine. However, adequate biological and pharmacological studies on most of the species in the genus Premna have not yet been performed because most, especially in in vivo studies, were carried out using their crude extracts (Table 4). For example, none of the bioactive molecules have been identified from the active antimalarial Premna species. Similarly, some Premna species showed potential in vivo antihyperlipidemic, cardioprotective, hepatoprotective, gastroprotective, and neuropharmacological activities which require further studies to determine the active compounds and possible mechanisms for a particular activity. While, numerous isolated compounds have been isolated and evaluated for related biological activites, they were limited to in vitro studies. No toxicological studies that have been carried out, although some species, such as P. serratifolia, have been used in Ayurvedic medicine for a long time.
There was no effort to qualitatively and quantitatively analyze the extracts used. Standardization of the extracts should be carried out to ensure consistency of the quantitative amounts of the active chemical markers in the plants of similar species collected from different locations. The variety and distribution of active secondary metabolites from this genus are useful as bioactive chemical markers for standardization and quality control purposes. Otherwise the work on biological activities may not be reproducible due to variations in the quantitative amounts of chemical constituents in the plants. These quantitative and qualitative differences in the chemical composition are related to responses of the plants to environmental factors or genetic adaptation of the populations growing at different altitudes to a specific environment (World Health Organization [WHO] 2000[WHO] , 2003.

Conclusions and future prospects
Further investigations are required to transform the experiencebased claims on the traditional uses of Premna species into evidence-based information. The present knowledge obtained mainly from experimental studies was critically assessed to provide evidence and justification for their traditional uses to propose future research prospects for this plant. Phytochemical studies on Premna species have led to characterization of diterpenoids, iridoid glycosides, and flavonoids as the charactetistic chemical composition of the genus. The in vitro and in vivo evaluation of biological properties of the extracts and isolates from various species of Premna on antimicrobial, antioxidant, anti-inflammatory, immunomodulatory, cytotoxic, antihyperglycaemic, and other activities should lead to further detailed investigations to identify the bioactive compounds and their mechanisms of action. The antimalarial, hepatoprotective, cardioprotective and gastroprotective effects of the plant extracts should encourage further studies on these plants for use as preventive agents. Toxicological evaluation should be conducted to address any adverse side effects which may occur. The roles and mechanisms of the bioactive compounds should be addressed appropriately to understand the contribution of individual compound to the activities as well as to become potential lead molecules for development into drug candidates. Attempts should be made to carry out more preclinical studies of the standardized extracts and bioactive compounds of Premna species, which include determination of modes or mechanisms of action in different animal models, bioavailability, pharmacokinetics and toxicological studies before submission of potential candidates to serious randomized human trials is possible. As more scientific evidences on therapeutic effects are discovered, Premna species will be recognized as a valuable source of drug leads and pharmaceuticals.
Disclosure statement