Planta Med 2016; 82(05): 388-406
DOI: 10.1055/s-0035-1568274
Georg Thieme Verlag KG Stuttgart · New York

Rosmarinic Acid – Pharmaceutical and Clinical Aspects

Solomon K. S. Amoah
Postgraduate Program in Pharmacy, Universidade Federal de Santa Catarina – UFSC, Florianópolis, SC, Brazil
Louis P. Sandjo
Postgraduate Program in Pharmacy, Universidade Federal de Santa Catarina – UFSC, Florianópolis, SC, Brazil
Jadel M. Kratz
Postgraduate Program in Pharmacy, Universidade Federal de Santa Catarina – UFSC, Florianópolis, SC, Brazil
Maique W. Biavatti
Postgraduate Program in Pharmacy, Universidade Federal de Santa Catarina – UFSC, Florianópolis, SC, Brazil
› Author Affiliations
Further Information


Maique W. Biavatti
Laboratório de Farmacognosia, CIF/CCS/UFSC, Campus Universitário/Trindade
88040–900 Florianópolis, SC
Phone: +55 48 37 21 34 93   

Publication History

received 05 March 2015
revised 20 November 2015

accepted 07 December 2015

Publication Date:
04 February 2016 (online)



The biosynthesis and biotechnological production of Rosmarinic acid, a phenolic ester that is widespread in the plant kingdom, has been widely investigated. This compound has shown many remarkable biological and pharmacological activities, which have led to its pharmaceutical and analytical development, as well as clinical studies, which are summarized and analyzed here for the first time. This review compiles data from the Pubmed, Scopus, Scifinder, Web Of Science, and Science Direct databases published between 1990 and 2015, restricting the search to works with the keywords “Rosmarinic acid” in the title. The initial search identified more than 800 articles; after an initial screening and removal of duplicate works, the search was further refined, resulting in approximately 300 articles that were scrutinized and comprise this review. The articles were organized to describe extraction and isolation, analytical methods, pharmaceutical development, and biological and pharmacological activities [divided into nonclinical (in vitro, in vivo) and clinical studies], pharmacokinetic studies, and stability studies.



6-OHDA: 6-hydroxydopamine
CCL11: C-C motif chemokine 11
CC: column chromatography
DAD: diode array detector
DPPH: 2,2-diphenyl-1-picrylhydrazyl
ELISA: enzyme-linked immunosorbent assay
ESIMS: electrospray ionization mass spectrometry
HPLC: high performance liquid chromatography
(HP)TLC: (high-performance) thin-layer chromatography
IDO: indoleamine 2,3-dioxygenase
IKK-β : IκB kinase β
LC: liquid chromatography
LPS: lipopolysaccharide
MPLC: medium performance liquid chromatography
MS: mass spectrometry
NF-κB: nuclear factor-kappa B
NMR: nuclear magnetic resonance
PDA: photo diode array
PK: pharmacokinetic
PLE: pressurized liquid extraction
RA: rosmarinic acid
SCORAD: severity scoring of atopic dermatitis
SFE: supercritical fluid extraction
t-BOOH: tert-butylhydroperoxide
TEWL: transepidermal water loss
TNF-α : tumor necrosis factor-α
UAE: ultrasonic-assisted extraction
UPLC: ultra performance liquid chromatography
UV-Vis: ultraviolet-visible spectroscopy


The natural compound RA [(R)-(+)-rosmarinic acid, [Fig. 1] is a common ester derived from caffeic acid and (R)-(+)-3-(3,4-dihydroxyphenyl)lactic acid that can accumulate in high amounts in many plant species. RA is abundant in several medicinal plants of the Lamiaceae family, such as rosemary (Rosmarinus officinalis L.), spearmint (Mentha spp.), and lemon balm (Melissa officinalis L.), and also in plants used in traditional Chinese medicine, such as Perilla frutescens (L.) Britton, Salvia miltiorrhiza Bunge, and Rabdosia rubescens (Hemsl.) H. Hara. Many studies have reported the role of RA in the biological activity of these plants as well as its pharmaceutical and biotechnological applications. Concerning its aspects in the biotechnology field, many reviews have been published in the last decade [1], [2], [3], [4], [5], [6]. The present review covers predominantly analytical, pharmaceutical, and biological aspects of RA, including clinical trials.

Zoom Image
Fig. 1 (R)-(+)-rosmarinic acid.


Extraction, Isolation, and Analytical Methods

Several extraction protocols [7], [8], [9], [10], [11], separation [12], [13], [14], and chromatographic methods have been developed to optimize the extraction and identification of RA [7], [12], [15]. The type of solvents, temperature, extraction time, particle size of the material to be extracted, and ratio of the solvents are some of the parameters studied [16], [17], [18], [19]. Sometimes solvent modifiers were used, such as acids, bases, and salts, and ionic solvents have also been explored [8], [20], [21]. Moreover, various methods and techniques used to extract polyphenols from plant materials have ended up obtaining RA as one of the principal components [22], [23], [24], [25], [26]. [Table 1] shows a summary of selected techniques for RA extraction. Versatile extraction of RA requires a good choice of protic solvents, temperature, and sufficient time to avoid degradation in analytical studies or optimization processes. Response surface methodology has also been used to optimize the isolation of RA [27], [28], [29], [30], [31].

Table 1 Summary of extraction and isolation studies for rosmarinic acid.

Plant studied (Family), location, (ref)

Extraction method

Quantity crude material

Extraction solvents

Temperature (°C)

Time of extraction

Concentration of RA in the crude extract (mg/g)

Method of separation/quantification

Orthosiphon stamineus leaves, (Lamiaceae) Malaysia [209]


10 g



2 h 4 h 8 h

5.40 6.61 6.71

HPLC equipped with a LiChrosorb RP-18 column (250 × 4.6 mm, 5 µm) and H2O-MeOH-THF (45 : 50 : 5) as the mobile phase


2 h 4 h 8 h

5.41 6.48 7.99

H2O-MeOH 1 : 1

2 h 4 h 8 h

9.03 9.78 9.54

(CH3)2CO-H2O (3 : 7)

2 h 4 h 8 h

7.35 7.43 7.72

Orthosiphon stamineus leaves, (Lamiaceae) Germany [210]


100 g

EtOH-H2O (1 : 1) EtOH-H2O (3 : 7)

Room temperature

5 days

0.45 0.58

HPLC equipped with a guard column and LiChrosphere RP-18 column (125 × 4 mm, 5 µm). Eluent: gradient of H3CCN-H3PO4 (99.9 : 0.1)/H2O-H3PO4 (98 : 2)

Salvia officinalis leaves, (Lamiaceae) USA [15]


30 kg

EtOH-H2O (95 : 5)

Room temperature

2 weeks


Liquid-liquid extraction water-(Hex, EtOAc, n-BuOH), CC on silica gel CHCl3-MeOH gradient condition, Sephadex LH-20 column eluted with MeOH, Lichroprep RP-18 column using MeOH-H2O (3 : 7), and EtOAc-MeOH-H2O (12 : 1 : 1)

Salvia officinalis leaves, (Lamiaceae) Brazil [67]


10 g

Aqueous ethanol



Extraction with (Et)2O fractionation with Chromatotron (centrifugal thin-layer chromatograph) eluted with EtOAc-hexane-AcOH (3 : 6 : 1)

Rosmarinus officinalis leaves, (Lamiaceae) Spain [56]

Pressurized liquid extraction

1 g



20 min

12.37 ± 0.38

UPLC-ESIMS Hypersil Gold column (50 × 2.1 mm) using a gradient of CH3CN-H2O (0.1 % formic acid).


16.00 ± 0.98


16.78 ± 0.64


9.12 ± 0.34



9.81 ± 0.89


14.19 ± 0.85


9.85 ± 0.38


8.60 ± 1.16

Cordia americana leaves, (Boraginaceae) Brazil [211]

Extraction assisted by heating

1140.94 g




CC on Sephadex LH-20 and MeOH used as the eluent, successive flash chromatography over an RP-18 eluted with a gradient of MeOH-H2O


HPLC, LiChrosphere RP-18 column (5 × 100 mm; 5 µm), CH3CN-H2O gradient (0.1 % formic acid)

Thunbergia laurifolia leaves, (Acanthaceae) Thailand [212]


20 g

EtOH-H2O (95 : 5)

Room temperature

3 times 72 h

HPLC equipped with a 4.6 × 250 mm Cosmosil 5C18-AR-II column, isocratic elution with H2O-MeOH-AcOH (65 : 35 : 0.1)

Nakhon Pathom


Nonthaburi Prakan


Nakhon Sawan


Prachin Buri




Samut Prakan


Mentha piperita leaves, (Lamiaceae) Iraq [16]

Maceration assisted by ultrasound

1 g

EtOH-H2O (4 : 1)


1 h


HPLC instrument, C18 column (4.6 × 250 mm, 5 mm), isocratic MeOH-H2O + 0.1 % AcOH (4 : 1)

Mentha longifolia leaves, (Lamiaceae) Iraq


Osimum basilicum leaves, (Lamiaceae) Iraq


Maranta depressa leaves, (Marantaceae) Germany [213]

Extraction assisted by sonication

50 mg

EtOH 70 %


10 min


HPLC on a Hypersil ODS column (5 mm, 280 mm × 40 mm) and 45 % MeOH/0.01 % H3PO4 used as the eluent

Maranta leuconeura “Fascinator” leaves, (Marantaceae) Germany


Maranta leuconeura var. kerchoviana leaves, (Marantaceae) Germany


Maranta leuconeura var. massangeana leaves, (Marantaceae) Germany


Canna edulis leaves, (Cannaceae) Germany


Canna indica leaves, (Cannaceae) Germany


Perilla frutescens leaves, (Lamiaceae) Japan [18]


650 g


Room temperature


Acidification, amberlite column eluted with acidified H2O and MeOH, droplet countercurrent chromatography eluted with CHCl3-n-BuOH-MeOH-H2O (9 : 2 : 12 : 8) and CC (Diaion HP20AG) eluted with H2O-MeOH in gradient conditions

Melissa officinalis leaves, (Lamiaceae) Slovak [19]


iPrOH-H2O (39 : 61)


1 h

72.6 ± 3.1

HPLC analysis RP18, 250 × 4 mm, 5 µm, gradient condition of CH3CN-H2O, 0.1 % CF3CO2 H

Anthurium versicolor leaves, (Araceae) Ecuador [214]


195 g


Room temperature


Chromatography on Sephadex LH-20, RP-HPLC on a C18 Bondapack column (30 cm × 7.8 mm) eluent isocratic MeOH-H2O (1 : 1)

Tournefortia sarmentosa (Boraginaceae) Taiwan [215]


20 kg

EtOH-H2O (85 : 15)



Diaion HP-20 column chromatography using a gradient of MeOH-H2O and Sephadex LH-20 using MeOH 80 %

Salvia officinalis (Lamiaceae) New Zealand [216]


50 g

Me2CO-H2O (7 : 3)

Room temperature


Diaion HP-20 column, eluent H2O then MeOH, Sephadex LH-20 column MeOH-H2O (7 : 3), HPLC equipped with an RP-18 column elution gradient of CH3CN-H2O (2 % AcOH)

Sanicula europaea aerial parts, (Apiaceae) Turkey [217]


940 g

EtOH–H2O (1 : 1)


1 h.


Column chromatography on silica gel 60 G mixed with silica gel 60 HF 254 and elution performed with CHCl3-MeOH-H2O (61 : 32 : 7)

Perilla frutescens seeds, (Lamiaceae) Japan [218]

Extraction assisted by heating

800 g



2 h


Liquid-liquid extraction Hex, EtOAc-H2O, EtOAc. Fraction purified by CC silica gel, Sephadex LH-20 and HPLC equipped with an ODS column, MeOH-H2O (1 : 1) used as the eluent

Nepeta cadmea aerial parts, (Lamiaceae) Turkey [14]


1.25 kg


Room temperature

30 days


Liquid-liquid extraction Hex-MeOH, H2O-(EtOAc, n-BuOH). n-BuOH fraction purified with CC on Dianion HP-20 MeOH-H2O Gradient. CC on silica gel eluted with CHCl3-MeOH and HPLC equipped with a C-18 column eluted with a gradient of MeOH-H2O

Pulsatilla koreana roots, (Ranunculaceae) South Korea [219]


1.2 kg

MeOH and CH2Cl2

Room temperature


Vacuum flash chromatography on RP-18 using MeOH and H2O as the eluent in gradient condition. Further purification took place with reversed-phase HPLC (YMC-ODS column; H2O-MeOH gradient)

Hyssopus cuspidatus whole plant, (Lamiaceae) China [220]

Maceration assisted by ultrasonication

1.5 kg



30 min


Diaion HP-20 resin column eluted with a gradient of MeOH-H2O acetone, silica gel CC eluted with CHCl3-MeOH-H2O reversed-phase HPLC with mobile phase MeOH-H2O (2 : 6) and reversed-phase HPLC CH3CN-H2O (1 : 9) as the mobile phase

Baccharis chilco aerial parts, (Asteraceae) Colombia [10]

Extraction assisted by heating

446 g

CH2Cl2 and EtOH


2 h


CC on silica gel with CH2Cl2-EtOH used as the eluent. H2O (alkaline)-EtOAc separation and HPLC purification of the organic phase using isocratic the condition of MeOH-H2O (2 : 3)

Helicteres isora fruits, (Sterculiaceae) Indonesia [221]

Extraction assisted by heating

4 kg


Reflux 60

7 h


Liquid-liquid extraction CHCl3-MeOH-H2O (5 : 3 : 8) and then with EtOAc and n-BuOH. CC on Sephadex LH-20 using MeOH-H2O (2 : 1) as th eluent. A second purification on Sephadex LH-20 eluted with MeOH-H2O (2 : 1), followed by successive purification on an MPLC silica gel column and an RP-18 column eluted with EtOAc-MeOH-H2O (60 : 25 : 15) and H2O-MeOH (3 : 1), respectively

Hedyotis scandens (Rubiaceae) [222]


9 kg

EtOH 95 %

Room temperature


Liquid-liquid extraction H2O, petroleum ether, EtOAc, and n-BuOH. CC on silica gel eluted with a gradient of CHCl3-MeOH. Further CC on silica gel eluted with a gradient of CHCl3-MeOH

Melissa officinalis leaves, (Lamiaceae) [223]

Extraction assisted by stirring

100 g

Acidified H2O


45 min


Liquid-liquid extraction with Et2O, Sephadex LH-20 using MeOH-H2O (7 : 3). Recrystallization at 4 °C

Zostera noltii leaves, (Zosteraceae) France [224]

Extraction assisted by heating

500 g




Liquid-liquid extraction CHCl3, EtOAc and H2O, then flash chromatography on a silica gel using EtOAc-MeOH (95 : 5) as the eluent

Collection July

8.10 ± 0.11

HPLC equipped with a Hypersil GOLD C8 column (5 µm particle size, 250 × 4.6 mm) and MeOH-H2O (0.1 %TFA) used as the eluent in gradient conditions

Collection December

2.17 ± 0.27

Collection January

8.65 ± 0.11

Along with conventional extraction methods, for instance hydrodistillation, decoction, and maceration, other methods, such as PLE [32], UAE [29], [33], [34], SFE [35], microwave-assisted extraction [23], [36], [37], [38], [39], accelerated solvent extraction [40], and enzyme-assisted extraction [41], [42], have also been employed to extract polyphenols, of which RA was identified as one of the major compounds. Nanofiltration [43], [44], [45] and solid-phase extraction [46], [47] are examples of techniques used to improve the amount of RA extracted from plant materials.

Interestingly, in a study that used infusion, decoction, microwave, and ultrasonic methods to optimize the extraction of phenolic compounds from Plectranthus spp., the highest amount of RA was found in the decoction extract of Plectranthus ecklonii Benth. (Lamiaceae) (293 µM) [48]. Cabana and collaborators [49] prepared hydrolate, decoction, and ethanolic extracts from Satureja parvifolia (Phil.) Epling (Lamiaceae) and demonstrated that the hydrolate extract presented the highest amount of RA.

In some studies, conventional soxhlet extraction was compared with other methods. Dracocephalum moldavica L. (Lamiaceae) was extracted using a soxhlet apparatus with several solvents, of which n-butanol extract presented the highest RA content (114.54 ± 24.70 mg/g dry extract) [50]. Soxhlet extraction was compared with the maceration method on unripe, half ripe, and full ripe Cuminum cyminum L. (Apiaceae) seeds [51], demonstrating that the maceration method produced a higher RA content than the soxhlet extraction. By contrast, a comparison between soxhlet and SFE of Satureja hortensis L. showed that soxhlet extraction with ethanol provided the highest RA content [52].

With SFE of R. officinalis (rosemary), comparison of CO2 with dimethyl ether demonstrated that dimethyl ether provided the highest RA yield [53]. The same plant leaves treated with supercritical fluid-CO2 at 5000 psi and 40, 60, and 80 °C showed that extraction at 60 °C furnished the highest amount of RA [54]. However, comparative studies on rosemary leaves using SFE, the hydrodistillation method, and solvent extraction with different solvents showed that the RA content was highest in the solvent extraction with ethanol [55]. Again in rosemary leaves, PLE using water and ethanol as solvents, compared to SFE using neat CO2 and supercritical CO2 modified with ethanol, showed that PLE using ethanol at temperature (150 °C) produced the highest RA yield [56].

From these comparative studies, it cannot be concluded that there is an ideal solvent or method to obtain RA, as it depends on the combination of chosen parameters.

To provide one-step isolation and purification of RA from S. miltiorrhiza, an online adsorption chromatography-countercurrent chromatography was carried out after UAE, adsorbing the plant extract on the resin and immediately submitting it to high-speed countercurrent chromatography (HSCCC), which led to the isolation of RA with a purity of about 88 % [12].

A template-assisted process using the supramolecular technique has also been used to extract RA from P. frutescens leaves. A P. frutescens extract, separated by supramolecular formation using flavocommelin and magnesium acetate aqueous solution followed by solvent extraction (ethyl acetate and water), yielded RA with 62.9 ± 4.5 % purity [58].

Analytical techniques used to determine RA include capillary electrophoresis [59], electromechanical [60], spectroscopical [61], [62], [63], [64], and chromatographic methods [65], [66], [67], [68], [69], [70], [71].

Some unusual electrochemical techniques have been developed, such as that reported by Franzoi et al. [72], which uses a biosensor based on laccase from the fungus Aspergillus oryzae and the ionic liquid 1-n-butyl-3-methylimidazolium hexafluorophosphate (BMIPF6) to determine RA in M. officinalis by square wave voltammetry. Additionally, a synthetic heterodinuclear Fe(III)Zn(II) complex, a biomimetic complex of purple acid phosphatases, has also shown to be effective in determining RA in plant extracts [60]. The amount of RA determined by these techniques was in accordance with capillary electrophoresis data. Another electrochemical technique using dual-channel coulometric detection coupled with liquid chromatography has also been used to quantify RA in Prunella vulgaris L. (Lamiaceae), animal feed, and pig plasma, with minimum preseparation steps, offering good selectivity and sensitivity [69].

Standard chromatographic techniques are the most frequent form of analysis used in the determination of RA. These can be grouped into TLC densitometry or HPTLC, and LC methods. TLC/HPTLC methods are the most accessible ones, but provide variable quantitative results since some methods possess methodological weakness. For example, integrated in the development process of a new method, a validation step is mandatory to demonstrate its reliability and the suitability. These methods used diverse stationary phases, such as normal silica gel [65], [73], [74], [75], [76], [77] and chemically modified silica gel with amino, cyano, or octadecyl groups [70], [78], [79]. Using densitometry, Janicsak et al. [76] showed the RA content in some species of Lamiaceae ranging from 0.01 to 9.30 mg/g.

Regarding LC methods, they have been most widely used for the separation, detection, and quantification of RA in plant extracts, providing a wide range of concentrations of RA in several Lamiaceae species. UV-Vis or DAD/PDA are the most widely used detectors [66], [67], [68], [75], [80], followed by an MS detector [67], [71], [75], [81]. The majority of the developed methods use UV-Vis or DAD/PDA detectors for the determination of RA in plants and extracts, while an MS detector is preferable for determination of RA in formulations and pharmacokinetic studies. Bandoniene et al. [80] used single ion monitoring mode at m/z 383 [M + Na] to quantify RA in several Salvia spp. and Borago officinalis L. (Boraginaceae), while Berhow et al. [75] used m/z 359, corresponding to the [M – H] mass ion of RA, to quantify RA in Ocimum canum Sims (Lamiaceae). Moreover, LC/tandem mass spectrometry has also been exploited to determine RA and some other components in beagle dog plasma [82] during pharmacokinetic studies. Here, the deprotonated molecule [M – H] was also used, since the sensitivity in the negative mode was higher than that in the positive mode. The limit of detection was found to be 1.0 ng/mL. Meanwhile, an HPLC-DAD-DPPH method has been used for online detection of the radical scavenging activity of RA [80]. After HPLC separation and UV detection at 280 nm, the analytes reacted post-column with the DPPH at a concentration of 50 mg/mL in methanol. Trute and Nahrstedt [83] developed a laborious two-step derivatization gas chromatography method to analyze RA enantiomers.

Regarding spectroscopic methods, a spectrophotometric analysis developed by Oztürk et al. [61], based on the complexation of RA with zirconium ions, was used for the quantification of RA in 11 Salvia species. Infrared spectroscopy has also been used to quantify RA in the Lamiaceae family [62], [64]. While Stehfest et al. [64] used the first derivative of the spectral region for quantification purposes, Saltas et al. [62] used the second derivative of the spectral region 1344–806 cm−1.

NMR spectroscopy has also been used to identify and quantify RA in methanol and ethanol extracts of plant species of the Lamiaceae family [63], [84], based on variable-temperature two-dimensional 1 H-1 H double quantum filter correlation spectroscopy (DQF-COSY), 1 H-13 C heteronuclear multiple-quantum coherence (HMQC), and 1 H-13 C heteronuclear multiple-bond correlation (HMBC) gradient NMR spectroscopy.


Pharmacological and Biological Activities

The pharmacological and biological activities of RA and RA-containing plant extracts have received considerable attention. Many scientific reports and patents have explored these properties, ranging in complexity from in vitro to clinical studies.

The exceptional antioxidant activity of RA, in parallel with the effects on cell signalling pathways and gene expression, contributes to the majority of the biological properties and technological applications covered by this review [28], [85], [86], [87], [88], [89], [90], [91]. As a result, RA antioxidant and pro-oxidant profiles [92], [93] have been comprehensively studied, through multiple assays [93], [94], [95], [96], [97], [98], [99], [100], [101]. The four phenolic hydrogens underwrite its ability to modulate free radical scavenging, in combination with two catechol moieties that provide the suitable polarity for RA to penetrate the lipid bilayers and protect them against oxidation without disturbing their structure [94], [96]. Electrochemical investigations have revealed that the first oxidation step is associated with the caffeic acid moiety, whereas the second oxidation step corresponds to the oxidation of the 3,4-dihydroxyphenyl lactic acid residue [94]. Due to the combination of these structural features, the antioxidant potential of RA is higher than that of the other hydroxycinnamic acid derivatives [96], [102], [103], [104].

Besides the antioxidant potential, RA has been reported to present a multitude of pharmacological and biological activities. An in-depth description of each would be beyond the scope of this review, but a compilation and analysis of these reports is presented below (studies reporting solely extracts results were not included, unless otherwise stated). It is divided into three sections: in vitro, in vivo, and clinical studies.

In vitro studies

RA exhibits a wide collection of biological activities in vitro, including antioxidant, anti-inflammatory, antimutagenic, antigenotoxic, cytotoxic, antimetastatic, antiangiogenic, neuroprotective, antimicrobial, immunomodulatory, melanogenic, and antivenom effects. The array of active concentrations reported in the studies covers different magnitudes, which raises concerns about specificity. Additionally, active concentrations in the milimolar, or even high micromolar range, are unlikely to translate into clinical effects due to limitations related to plasma levels and biodistribution in humans (see section “Pharmacokinetics in Humans”). [Table 2] provides a non-exhaustive list of in vitro findings obtained with RA.

Table 2 Summary of in vitro studies with isolated rosmarinic acid.

First author, year (ref)



Main findings

↑Increase; ↓decrease/inhibition; NA – not available; Most studies used multiple assays/endpoints for the pharmacological evaluation of RA; For a complete methodological assessment, please refer to the original article


Lembo et al., 2014, 2011 [225], [226]

Cell viability (HaCaT cells) Cytokine modulation by qRT-PCR

2.7–55 µM

prevention of UVB-induced inflammation

Ku et al., 2013 [124]

ELISA (primary human umbilical vein endothelial cells)

0.1–2 µM

inhibition of endothelial protein C receptor shedding

Moon et al., 2010 [227]

Cell viability (U937 human leukemia cells)

15–60 µM

↓ TNF-α-induced ROS generation ↓ NF-κB activation ↑ TNF-α-induced apoptosis

Huang et al., 2009 [228]

LPS stimulation (RAW 264.7 mouse macrophage cells)

2.67 µM

↓ LPS-induced prostaglandin E2 and nitric oxide production

Zdarilová et al., 2009 [229]

LPS stimulation (primary cultured human gingival fibroblasts)

1 µg/mL

↓ oxidative damage ↓ inflammation

Kim et al., 2008 [230]

LPS inflammatory stimulation (BMDCs bone marrow-derived dendritic cells)

0.1–200 µM

↓ cell maturation and migration ↓ monocyte chemoattractant protein-1 ↓ macrophage inflammatory protein-1α

Scheckel et al., 2008 [231]

Luciferase assay (HT-29 colon cancer and MCF10 A non-malignant epithelial cells)

5–20 µM

↓ AP-1-dependent activation of cyclooxygenase-2 in human cancer and normal cells

Lee et al., 2006 [232]

ELISA and Western blotting (human dermal fibroblast cells)

1–40 µM IC50 = 9.1–15.8 µM

↓ expression of CCL11 and C-C motif chemokine 11 receptor (CCR3) ↓ IKK-β activity

Antiproliferative and Antimutagenic

Zhang et al., 2011 [233]

Cell viability (HSC-T6 hepatic stellate cells)

1–16 µg/mL IC50 = 7.1 µg/mL

↓ cell proliferation ↑ apoptosis

Xu et al., 2010 [113]

Wound healing, adhesion and Transwell assays (Ls174-T human colon carcinoma cells)

5–300 µg/mL IC50 < 20 µg/mL

antimetastasis effect

Xu et al., 2010 [234]

Cell viability (MDA-MB-231BO human bone-homing breast cancer cells) and wound healing assays

1–300 µg/mL

↓tumoral cells migration (IC50 = 118 µg/mL)
bone protection in vitro

Kim et al., 2009 [119]

Cell viability and morphological changes (HRMEC human retina microvascular endothelial cells)

10–100 µM

antiangiogenic activity ↓ cell proliferation ↓ tube formation

Xavier et al., 2008, 2009 [235], [236]

Cell viability (HCT15 and CO115 human colorectal carcinoma-derived cells)

10–100 µM

↓ cell proliferation ↑ apoptosis

Hur et al., 2004, 2007 [237], [238]

Flow cytometry (peripheral blood mononuclear cells obtained from rheumatoid arthritis patients and Jurkat acute T cell leukemia cells)

3–100 µM

↓ cell proliferation ↑ apoptosis

Huang et al., 2006 [239]

Cell viability and migration assays (HUVEC human umbilical vein endothelial cells)

12.5–200 µM

↓ several steps of angiogenesis

Kolettas et al., 2006 [240]

Cell viability (Jurkat human T lymphoma cells)

10–150 µM

↓ cell proliferation ↑ apoptosis

Vattem et al., 2006 [241]

AMES and supercoiled DNA strand scission assay

500 µg/plate

antimutagenic effect

Yoshida et al., 2005 [242]

Cell viability (MK-1, HeLa, and B16F10 cells)

MK-1, GI50 = 119 µM HeLa, GI50 = 75 µM B16F10, GI50 = 16 µM

↓ cell proliferation


Alcaraz et al., 2014 [243]

Micronucleus (human lymphocytes) Cell viability (PNT2 and B16F10 cells)

10–40 µM

radioprotection in normal cells ↑radio-induced damage in melanoma cells

Braidy et al., 2014 [244]

Cell viability (primary cultured human neurons)

0.01–1 mg/mL

protection against ciguatoxin-induced neurotoxicity

Costa et al., 2013 [245]

Cell viability (A172 human astrocyte cells)

30 µg/mL

protection against H2O2-induced oxidative damage

Jeon et al., 2013 [246]

Cell viability (primary cultured rat hepatocytes)

1–100 µM

cytoprotection against hypoxia-induced injury

Kim et al., 2013 [247]; Jeong et al., 2011 [248]

Cell viability (HEI-OC1 auditory cell line) Fluorescence microscopy (Rat Corti primary explants)

1–100 µM

protection against Cd2+-induced or cisplatin-induced ototoxicity
protection against cisplatin-induced destruction of hair cell arrays ex vivo

Yang et al., 2013 [249]

Cell viability (HepG2 human hepatoma cells)

10–26.84 µg/mL

protection against t-BOOH-induced cytotoxicity

Du et al., 2010 [250]; Ren et al., 2009 [251]

Cell viability (MES23.5 dopaminergic cells)

0.001–100 µM

protection against 1-methyl-4-phenylpyridinium (MPP+) and 6-OHDA-induced neurotoxicity

Furtado et al., 2010 [252]

Micronucleus and comet assay (V79 cells)

0.28–1.12 mM

protection against doxorubicin-induced genotoxicity

Fallarini et al., 2009 [253]

Cell viability [differentiated SH-SY5Y and SK-N-BE(2) human neuroblastoma cells]

10–100 µM EC50 = 0.9–3.7 µM

protection against t-BOOH-induced oxidative stress ↓ excitotoxicity ↓ ischaemia–reperfusion-induced neuronal death

Lee et al., 2008 [254]

Cell viability (SH-SY5Y human dopaminergic neuronal cells)

14–56 µM

protection against H2O2-induced neurotoxicity ↓ apoptosis

Salimei et al., 2007 [255]

Cytofluorimetric approach (K562, NPA and ARO cells)

25 µM

↓ sorbitol-induced apoptosis

Iuvone et al., 2006 [256]

Cell viability (PC12 adrenal medulla cells)

0.0036–36 µg/mL

β-amyloid-induced cell death

Psotova et al., 2006 [257]

Cell viability (HaCaT human keratinocyte cells)

0.9–18 µg/mL

photoprotection against UVA-induced damage

Yan et al., 2006 [258]

Cell viability (PC12 cells)

100 µM

protection against glutamate-induced cell death ↓ apoptosis

Gao et al., 2005 [259]

Cell viability (primary cultured rat astrocytes)

10–40 µM

protection against H2O2-induced cell death ↓ apoptosis

Kim et al., 2005 [260]

Cell viability (H9c2 cardiomyoblast cells)

1–20 µg/mL

protection against adriamycin-induced apoptosis ↓ apoptosis

Chlopcikova et al., 2004 [261]

Cell viability (primary cultured rat cardiomyocytes)

100–200 µM

protection against doxorubicin-induced cell death ↓ lipid peroxidation (IC50 = 8.17 µM)

Renzulli et al., 2004 [262]

Cell viability (Hep G2 human hepatoma-derived cells)

2.5–100 µM

protection against aflatoxin B1 and ochratoxin A-induced cytotoxicity


Lee et al., 2007 [263]

IFN-γ stimulation (BMDCs murine bone marrow-derived dendritic cells)

1–100 µM

↓ IDO-dependent T cell suppression ↓ functional expression of IDO

Ahn et al., 2003 [264]; Kang et al., 2003 [265]; Won et al., 2003 [266]; Yun et al., 2003 [137]

Binding assay Cell proliferation and differentiation

0.1–1000 µM IC50 = 5.6–14 µM

inhibition of lymphocyte cell-specific kinase Src-homology 2 domain binding ↓ T cell activation ↓T cell proliferation

Kang et al., 2003 [267]

Luciferase and Ca2+ mobilization assays (Jurkat T cells)

5–30 µM

↓ Ca2+-dependent pathways of T-cell antigen receptor-mediated signalling

Sahu et al., 1999 [268]; Peake et al., 1991 [269]

Binding assays C3 and C5 convertase assays

0.01–10 mM

↓ complement activation
binding to activated C3b C5 convertase inhibition


Abedini et al., 2013 [270]

Minimal inhibitory concentration/minimal bactericidal concentration by broth microdilution

0.0093–2.5 mg/mL

antimicrobial activity

Airoldi et al., 2013 [271]

NMR spectroscopy

1 mM

binding to β-amyloid oligomers (Alzheimerʼs disease)

Marcelo et al., 2013 [272]; Yin et al., 2008 [273]

NMR spectroscopy Molecular Modelling

333 µM 50 mM

binding to Acetylcholinesterase (Alzheimerʼs disease)

Slobodníková et al., 2013 [274]

Minimal inhibitory concentration/minimal bactericidal concentration by broth microdilution Biofilm regrowth technique

156–5000 µg/mL

antimicrobial activity
inactive as biofilm-eradicator

Yang et al., 2012 [161]

Cellular differentiation reversal (primary cultured hepatic stellate cells)

135–270 µM

antifibrotic effect ↓ epigenetic peroxisomal proliferator-activated receptor γ repression
reversal of activated cellular phenotypes

Dos Santos et al., 2010, 2011 [142], [143]; Ticli et al., 2005 [144]

Phospholipase2 enzymatic activity inhibition Molecular modelling Muscle-damaging and neuromuscular-blocking activities

13.7 µg/mL

antivenom effects in vitro and ex vivo

Lin et al., 2011 [275]

Enzymatic activity inhibition

0.01–0.4 mM

tyrosinase and α-glucosidase inhibition

Murata et al., 2011 [276]; Ippoushi et al., 2000 [277]

Enzymatic activity inhibition

0.03–1 mM IC50 = 309 µM

hyaluronidase inhibition

Falé et al., 2008, 2009 [278], [279]

Enzymatic activity inhibition

IC50 = 440 µg/mL

acetylcholinesterase inhibition (Alzheimerʼs disease)

Lee et al., 2007 [280]

Melanin content determination (B16 melanoma cells)

1–100 µM

↑ melanogenesis

Kang et al., 2004 [281]

Enzymatic activity inhibition

IC50 = 16.8 µM

tyrosinase inhibition

McCue et al., 2004 [282]

Enzymatic activity inhibition

0.07–0.42 mM

amylase inhibition

Makino et al., 2000 [151]

Cell viability (primary cultured murine mesangial cells)

1–25 µg/mL IC50 = 1.4–3.8 µg/mL

↓ cell proliferation

Simpol et al., 1994 [283]

Histamine release inhibition (rat mast cells)

IC50 = 18 µM

↓ histamine release

Studies that focus on the technological application of RA have reported that this phenolic acid is a good bioprotective agent against fungal infections in crops [105], and a good candidate for partial monolignol substitution via bioengineering approaches, aiming at more efficient utilization of plant fibers for biofuels or livestock production [106].

Pharmacokinetic studies of RA using in vitro models have also been conducted by some research groups. Qiang et al. compared the permeability and phase II biotransformation of RA as a pure compound and also in herbal extracts using human intestinal epithelial Caco-2 cell monolayers. The apparent permeability coefficient (Papp) obtained for RA was 0.2 × 10− 6 cm/s, indicating a low permeability profile [107]. In agreement, Konishi and Kobayashi reported that in vitro intestinal absorption of RA was rather low. The majority of RA loaded on the apical side of Caco-2 cells was not transported, and the permeation occurred mainly via paracellular diffusion [108]. Moreover, RA appeared to be unsusceptible to chemical and enzymatic hydrolysis in Caco-2 cells and gastrointestinal models [108], [109]. Conversely, the addition of Lactobacillus johnsonii cells to the rosemary extract resulted in a substantial hydrolysis of RA [109]. Taken together, these results suggest that RA is further metabolized and degraded by gut microflora, in accordance with in vivo studies (see “Pharmacokinetics in Animals” section). RA metabolites are then efficiently absorbed and distributed by the monocarboxylic acid transporter within the body [108]. Qiang et al. demonstrated that RA permeability significantly increased after treatment with β-glucuronidase/sulfatase [107], and Falé et al. demonstrated that the presence of benzoic acid, a substrate of monocarboxylic acid transporter, or flavonoids (i.e., luteolin and apigenin) can decrease RA permeation [110], supporting the involvement of membrane transporters in the uptake of RA.

The interaction of RA with other pharmaceutical relevant transporters and enzymes has been studied in vitro. Lin et al. reported RA inhibits UDP-glucuronosyltransferase activity [111], and Li et al. showed that RA could decrease the transcription and expression of P-glycoprotein, promoting the reversal of multidrug resistance phenomena [112].


In vivo studies

By compiling the in vivo studies, it was found that numerous biological studies corroborate the in vitro hypothesis. A significant prevalence of anti-inflammatory, antitumoral, and tissue damage prevention studies was also found. These reports highlight, once again, antioxidant and radical scavenging activities as key factors in the in vivo outcomes reported for RA.

The search for antitumoral compounds also included investigations of RA against different types of cancer [113], [114], [115], [116], [117]. Karmokar et al. showed that chronic consumption of RA (0.3 % w/w chow supplementation) for 8 weeks inhibited colon carcinogenesis in mice, and produced quantifiable levels of the parent compound in the plasma and the intestinal tract [118]. Additionally, RA suppressed retinal neovascularization [119], and exhibited transcorneal penetration when the RA-containing polyherbal formulation was instilled into rabbit eyes [120].

No increase was found in the frequency of micronuclei in mice treated orally with 50–200 mg/kg of RA, indicating that RA does not affect chromosomal integrity [121]. In fact, RA showed antimutagenic effects, protecting animals against doxorubicin [121], ethanol [122], and γ ray-induced DNA damage [123], assessed by the comet and micronucleus assays using murine models.

Studies exploring the anti-inflammatory activity of RA are also frequent [124], [125], [126], [127]. Youn et al. showed that RA (daily intraperitoneal injections, 50 mg/kg) suppressed synovitis in a murine collagen induced arthritis model [128], while other groups reported the inhibition of allergic inflammatory reactions [129] and airway inflammation [130], [131] as well as atopic dermatitis [132], [133]. Jiang et al. reported that RA, administered intravenously (25 and 50 mg/kg), either alone or in combination with imipenem, protected rats against experimental sepsis by decreasing local and systemic levels of an extensive spectrum of inflammatory mediators [134]. Additionally, oral administration of RA (2 mg/body for 3 days) inhibited diesel exhaust particle-induced lung injury in mice by reducing the expression of proinflammatory molecules and antioxidative activities [135].

In regard to anti-infective potential, RA presented in vivo antiviral activity against the Japanese encephalitis virus, a member of the Flaviviruses family [136]. At 8 to 9 days postinfection, the mortality of mice treated intraperitoneally twice daily (25 mg/kg) was reduced in comparison to untreated animals. This effect was mediated by decreased viral loads, and also by reduced levels of proinflammatory cytokines [136].

RA was also shown to be immunosuppressant, improving skin [137] and islet allograft survivals in murine models [138]. Antihemorrhagic, antithrombotic, and antiplatelet effects, along with the direct neutralization of multiple snake venoms and fish toxins, have also been reported for RA or RA-containing extracts [139], [140], [141], [142], [143], [144], [145], [146], [147], corroborating for a potentially beneficial role of RA for the treatment of poisoning accidents, already described in the in vitro section ([Table 2]).

Protective effects of RA in tissue or organs have been demonstrated. Domitrovic et al. showed that oral treatment with RA (1–5 mg/kg) for two consecutive days improved cisplatin-induced kidney injury in mice by inhibiting oxidative stress, inflammation (TNF-α and NF-κB), and apoptosis [148]. Other groups reported similar profiles [149], [150], [151], [152], i.e., inhibition of diabetes-related renal damage [153] or inhibition of gentamicin sulphate-induced renal oxidative damage in rats treated orally with 50 mg/kg for 12 days [154], [155]. Also, memory protective effects of RA in an Alzheimerʼs model (β-amyloid neurotoxicity) was attributed to the radical scavenging potential of the molecule [156]. Cardioprotection [157], [158], inhibition of endothelial tissue damage [159], and reduction of multiple organ failure in thermal injury [160] were linked to the antioxidant and anti-inflammatory effects of RA in vivo. Liver protection effects against cholestatic fibrosis, ischemia-reperfusion injury, CCl4, and lipopolysaccharide-induced damage were also reported at oral doses ranging from 2.5 to 50 mg/kg [160], [161], [162], [163], [164], [165].

The neuroprotective effect of RA was studied in animal models of central nervous system diseases. Mushtaq et al. showed that oral treatment with RA (10 mg/kg) for 21 days significantly reduced the level of lipid peroxidation in multiple areas of the brain in diabetic rats, together with modulation of cholinergic neurotransmission [166]. At doses higher than 50 mg/kg, RA produced a significant anti-inflammatory effect in rats submitted to an experimental ischemic diabetic stroke model [167]. In a Parkinsonʼs disease model in mice, 6-OHDA-induced degeneration of the nigrostriatal dopaminergic system was reversed by 21 days of oral treatment with RA at 20 mg/kg [168]. A number of groups reported that administration of RA (orally or intraperitoneally) alleviated stress symptoms [169], [170], [171], [172], [173] and depressive-like behaviors [174], [175], [176], and produced anxiolytic-like effects, without exerting locomotor alterations or DNA damage in brain tissue [177]. Additionally, subchronic oral RA treatment of mice (up to 3 weeks) has a cognitive-enhancing effect [178], and, in a model of Amyotrophic lateral sclerosis, RA administered intraperitoneally (0.13 mg/kg, twice a week) significantly delayed motor dysfunction, attenuated motor neuron degeneration, and improved clinical outcomes and lifespan [179].

Finally, an unusual preclinical study in a Rhesus monkey model was conducted to determine the effect of the topical application of RA on the progression of plaque-induced gingivitis. Six young adult (4–6 years) male Rhesus monkeys (Macaca mulatta) randomly assigned to 3 groups of 2 each (A, B and C) received a topical application of vehicle alone, the drug ebselen (1 %), and RA (5 %) twice weekly for two weeks. To promote plaque retention, the chow was softened with water prior to feeding. Clinical evidence of gingivitis (erythema, edema) was apparent after 2–3 days, with established gingivitis (GI = 2) apparent in all animals after one week. Based on these preliminary studies, the authors concluded that at least in the short term, ebselen and RA are effective at reducing both gingival inflammation and plaque accumulation when topically applied [180].


Clinical studies with rosmarinic acid containing extracts

As a widespread metabolite, several traditional plants have been clinically studied using RA as a chemical marker or as an active compound. Different clinical protocols designed to assess the anti-inflammatory effects and the antioxidant potential of plants containing RA have been employed.

Initially, links between oxidative stress and adverse health effects have been suggested; for example, Ranjbar et al. [181] investigated the antioxidant influence of an Echium amoenum Fisch & C. A. Mey (Boraginaceae) decoction in healthy volunteers using a cross-sectional before/after clinical trial. The authors considered E. amoenum, one of the most important medicinal plants in Iranian traditional medicine (despite its hepatotoxic pyrrolizidine alkaloids [182]) to be a rich source of RA and flavonoids. They randomly selected thirty-eight volunteers (18–25 years old), students at Arak University of Medical Sciences, who were interviewed by a specialized physician before receiving the E. amoenum flower decoction (7 mg/kg) twice daily (morning and evening) for 2 weeks. The authors found a significant decrease (p < 0.05) in lipid peroxidation (24.65 ± 11.33 to 19.05 ± 9.7 nmol/mL), an increase in total blood antioxidant capacity (p < 0.05; 1.46 ± 0.51 to 1.70 ± 0.36 mmol/mL), as well as an increase (p < 0.001) in total thiol molecules (0.49 ± 0.11 to 0.56 ± 0.12 mmol), after administration of the decoction. In their view, this is the mechanism by which RA and antioxidants protect the human body from various diseases. This study has a major limitation, as the authors administered a decoction that was not standardized in terms of RA content.

With a similar objective, a crossover randomized controlled trial study [183] detected, by means of LC-quadrupole time-of-flight, phenolic acids (among them RA) and aromatic compounds in human plasma 24 h after consumption of a blueberry drink containing 766 mg total polyphenols by healthy volunteers (n = 10). The primary outcome was the measure of flow-mediated dilation. Flow-mediated dilation increased after 1 h of consumption and then plateaued. According to the authors, increases in flow-mediated dilation were closely linked to increases in circulating phenolic metabolites and decreases in neutrophil NADPH oxidase activity at 1–2 and 6 h. The authors reported limitations when interpreting the trial datasets, such as the short time frame, causality, and the population studied.

Also focusing on antioxidant foods, a study was conducted with 11 healthy volunteers consuming two kinds of burger meat – seasoned or unseasoned with a spice blend [184]. The production of malondialdehyde in the burgers and the malondialdehyde concentration in plasma and urine after ingestion were measured, since the formation of malondialdehyde has implications for atherogenesis and carcinogenesis [185]. RA from oregano was monitored to assess the effect of cooking on the spice antioxidant content. Forty percent (19 mg) of the rosmarinic acid added remained in the spiced burger after cooking. There was a 71 % reduction in the malondialdehyde concentration (0.52 ± 0.02 µmol/250 g) in the meat of the spiced burgers compared with the malondialdehyde concentration (1.79 ± 0.17 µmol/250 g) in the meat of the control burgers. The plasma malondialdehyde concentration increased significantly in the control burger group compared to the baseline (p = 0.026). There was a significant time-trend difference (p = 0.013) between the two groups. Urinary malondialdehyde concentrations (µmol/g creatinine) decreased by 49 % (p = 0.021) in subjects consuming the spiced burgers compared with subjects consuming the control burgers. The authors concluded that cooking hamburgers with a polyphenol-rich spice mixture can significantly decrease the concentration of malondialdehyde, suggesting potential health benefits for atherogenesis and carcinogenesis.

Nasal polyposis is a mucosal inflammatory disease that was investigated in a double-blind placebo-controlled crossover trial using 1 cup of a mint tea high in RA (about 300 mg) versus 1 cup of a mint tea low in RA (about 20 mg per day). Each treatment period lasted 4 weeks, separated by a 4-week washout period, and the first treatment followed a 2-week baseline period. Twenty-two adult subjects completed the study, and the authors found no statistically significant difference between the treatments in nasal stuffiness, as recorded on daily diary cards, in peak nasal inspiratory flows measured twice daily, or in patientsʼ global assessment (including ability to smell or sleep) performed at the end of each treatment period [186].

A more recent randomized, parallel-arm, double-blind study (meeting the CONSORT statement) investigated the effect of RA in the management of knee osteoarthritis symptoms using tea brewed from the high-RA spearmint plant (130–150 mg of RA per cup, n = 22) and a commercially available spearmint tea (ca 13 mg of RA, n = 24). The subjects were instructed to consume two cups of tea per day from a 300-mL study mug provided to them, for a 16-week period. The outcome was measured using the Western Ontario and McMaster Universities Osteoarthritis (WOMAC) pain score, which is a validated, standardized 24-item questionnaire that assesses pain, disability, and joint stiffness associated with osteoarthritis. The study allowed the subjects to maintain their normal pain medication in order to explore the potential of the high-RA tea as a complementary therapy. The authors concluded that individuals who consumed 600 mL of high-RA spearmint tea daily showed a significant decrease in pain scores from weeks 0 to 16, inferring that adults with knee osteoarthritis may benefit from the inclusion of high-RA spearmint tea in their daily diet. No serious adverse events were reported during the study [187].


Clinical studies with isolated rosmarinic acid

RA was clinically investigated in atopical dermatitis due to its in vitro and in vivo anti-inflammatory effects, such as its ability to block complement fixation, inhibit lipoxygenase and cyclooxygenase activity, and suppress IKK-β downstream signaling in the TNF-α-induced upregulation of CCL11. Oil-in-water cream with or without rosmarinic acid (0.3 %) was applied to the elbow flexures of 21 subjects (14 women and seven men; 15.1 ± 3.1 years) twice daily, for 4 and 8 weeks. The subjects were clinically graded as having moderate atopic dermatitis, according to the guidelines of the SCORAD index. The evaluation methods included clinical assessments (SCORAD), instrumental assessments (TEWL) and self-assessments by questionnaire. The authors reported that RA cream caused no reactions in the patch test on patient, suggesting that RA can safely be applied to human skin. After treatment with RA, the authors observed a statistically significant reduction in the SCORAD score, a decrease in itching, and a decrease in TEWL, suggesting that RA is a possible atopical dermatitis-mitigating agent [188].

Likewise, considering that RA is a strong anti-inflammatory agent in several animal models, a 21-day randomized controlled trial was undertaken to determine whether oral RA supplementation [200 mg (n = 10) or 50 mg (n = 9)] is an effective intervention for patients with seasonal allergic rhinoconjuctivitis. The patients recorded their symptoms daily in a diary card, and profiles of infiltrating cells, concentrations of eotaxin, IL-1β, IL-8, and histamine in nasal lavage fluid, as well as serum IgE concentrations, were measured. The authors found that RA supplementation resulted in a significant increase in responder rates for itchy nose, watery eyes, itchy eyes, and overall symptoms (p < 0.05). Additionally, the treatment significantly decreased the numbers of neutrophils and eosinophils in nasal lavage fluid, compared to placebo supplementation. Patients reported no adverse events, and no significant abnormalities were detected in routine blood tests [189].


Pharmacokinetics in Humans

Pharmacokinetic studies designed for healthy volunteers have been described. A randomized, open-label, single-dose study investigated the PK parameters of RA and other depside salts after intravenous infusion in 12 volunteers (6 male, 6 female) divided into two groups, randomly receiving either 100 or 200 mg of RA. The authors described no significant differences in the PK parameters between male and female subjects. Three undetermined metabolites were found in the plasma at low concentrations. The urinary excretion recoveries of RA were 25.21 % (20.61 %) for the 100 mg dose, and 20.11 % (10.50 %) for the 200 mg dose. No adverse events were reported by the subjects or found by the investigators [190]. In a crossover design, a PK study with six healthy male volunteers with a 10-day washout period between the administration of P. frutescens extract in a tablet containing 20 % RA and placebo was carried out. The authors found that RA reached a maximum concentration in plasma after 0.5 h, followed by a gradual increase in the plasma concentration of methyl-RA, which reached a peak by 2 h. Approximately 75 % of the total RA metabolites were then excreted in the urine within 6 h. These results showed that RA contained in P. frutescens extract was rapidly absorbed, subsequently methylated, and then excreted in the urine, and the majority of RA and its metabolites were present in the plasma as conjugated forms. The main metabolites found in the urine after consuming P. frutescens extract were sulfoglucuronide conjugates of RA and methyl-RA, indicating that conjugation of polyphenolic substances, either by glucuronidation and/or sulfation, occurs in human tissues such as the intestine and liver [191]. In comparative studies, Nakazawa and Ohsawa [192], [193] found differences between human and rat metabolites of RA, reporting that RA may be predominantly metabolized to trimethoxycinnamic acid monoglucuronide in humans through hydrolytic cleavage in the gut, whereas in rats, RA is mainly metabolized to sulfated forms of trans-caffeic, ferulic, and m-coumaric acids.


Pharmacokinetics in Animals

Pharmacokinetic studies have been conducted in rats. Azevedo et al. showed that RA can modulate the transit of the intestinal Na+/glucose cotransporter-1 (SGLT-1) to the brush-border membrane, an effect that may contribute to the control of plasma glucose levels in diabetic rats [194]. Additionally, Debersac et al. showed that supplementation of a rat diet with 0.5 % RA for 2 weeks did not increase the liver cytochrome P450 enzyme levels [195].

Konishi et al. compared PK parameters of caffeic acid and RA after oral administration in rats, indicating that the absorption efficiency of caffeic acid was higher than that of RA, and that the conjugation of both compounds occurs during permeation across the rat epithelium [196]. These findings are in accordance with Baba et al., which demonstrated that RA was rapidly absorbed (plasma values ca. 5 µmol/L) and metabolized into conjugated and/or methylated forms, and a large amount of the absorbed RA was degraded and metabolized as conjugated forms of caffeic, ferulic, and m-coumaric acids [197].

Two studies on the development of the LC-MS/MS method and its application to PK studies of RA in rats found similar values for RA PK parameters, such as Cmax (48.67 ± 11.24 [198] and 37.19 ± 13.85 ng/mL [199]) and Tmax (1.08 ± 0.38 and 0.74 ± 0.12 h [199]), despite using different mathematical methods and doses. In [198], Drug and Statistics (DAS) 2.0 software (non-compartmental), and in [199], the practical pharmacokinetic program version 87 (3P87) (survival square sum) were used, with a single administration of Herba Isodi Rubescentis extract containing RA 2.55 mg/kg in [198], and in [199], a mixture of the standards in saline, with RA at 6.39 mg/kg.


Pharmaceutical Development

Focusing on the potential antioxidant activity of RA, technological studies were conducted with the aim of improving stability and bioavailability, and verifying antioxidant interactions. The incorporation of RA into formulas as an active component was also proposed. In order to overcome the limitations of RA for cosmetic purposes, such as its low water solubility and discoloration, polycaprolactone microspheres loaded with RA were developed. RA-polycaprolactone microspheres with zwitterionic and non-ionic surfactants presented better loading efficiency when 10 mg of RA was incorporated. The microspheres showed better long-term stability in a cream formulation compared to RA alone [200]. In another study, solid lipid nanoparticles loaded with RA were prepared with the same objectives, resulting in better stability, which can be useful for applications in the food industry [201]. Also with the aim of increasing stability and water solubility, the cyclodextrin (β-CD) complexation of RA improved the properties and antioxidant capacity [202], [203]. To study the interactions with α-tocopherol, the antioxidant efficiency of RA and RA-esters in oil-water emulsions was measured. All combinations resulted in improved antioxidant activity when compared to isolated compounds, since the concentrations were higher, and the combination of α-tocopherol and RA presented a synergistic effect. According to the authors, the formation of caffeic acid from RA in the presence of α-tocopherol occurred, and this conversion provided additional antioxidant outcomes [204].

RA proposed as an active compound against Acne vulgaris has been incorporated in a niosomal gel tested in vivo (Swiss albino mice). Both plain and niosomal formulations inhibited inflammation, and only the niosomal gel reduced the bacterial multiplication rate 4 days after application, due to prolonged release [205]. Also, considering the anti-inflammatory potential of RA, a preparation of a polygalacturonic/RA biodegradable membrane was developed to prevent postoperative abdominal adhesion, and was evaluated in vitro and in vivo. The authors demonstrated that this membrane could effectively inhibit adhesion as well as acute and chronic inflammation in rats [206].

Focusing on the construction of biosensors with high analytical performance, RA was used as a model for the quantification of pharmaceutical samples in an optimized biosensor synthetized with gold nanoparticles in an ionic liquid phase, supported in a biopolymeric matrix. The developed biosensor offered good precision and accuracy for the determination of RA [207]. In another study, RA was used to generate and stabilize gold and silver nanoparticules, acting as an appropriate antioxidant, being able to reduce the metallic ions, and also acting as a surface-passivation agent [208].


Concluding Remarks

RA is a natural metabolite that has attracted the interest of researchers in various areas of knowledge due to its abundance and potential biological properties. Analytical, biological, and technological RA data are compiled here for the first time.

It has served as a model for a series of scientific proposals, and will continue to be a target of investigations and the production of knowledge. There is still much that needs to be done to ensure the data from nonclinical to clinical trials are translated into meaningful knowledge. There is a lack of connectivity between studies, and many gaps are yet to be filled in order to authenticate the role of RA in human health. In future investigations covering RA, translational research should be accomplished: the application of the findings generated during laboratory or preclinical research to the development of trials and studies in humans, furnishing promising new treatments with practical applications.



Relevant articles from the literature were collected by searching the main scientific databases including Pubmed, Scopus, Scifinder, Web of Science, and Science Direct from 1990–2015, limiting the search to the occurrence of the keywords “Rosmarinic acid” in the title. This preliminary search identified more than 800 articles, which after the initial screening and removal of duplicates was refined, resulting in 281 articles that were scrutinized to compose this review.



The authors are grateful to CNPq for the grants and scholarships.


Conflict of Interest

The authors declare no conflicts of interest.


Maique W. Biavatti
Laboratório de Farmacognosia, CIF/CCS/UFSC, Campus Universitário/Trindade
88040–900 Florianópolis, SC
Phone: +55 48 37 21 34 93   

Zoom Image
Fig. 1 (R)-(+)-rosmarinic acid.