Anti-stress effects of rosemary (Rosmarinus officinalis L.) leaf extract on intestinal goblet cells and immobility of forced-swimming test in BALB/c mice

ABSTRACT

We investigated the anti-stress effect of rosemary (Rosmarinus officinalis L.) leaf extract (RLE) on restraint-stressed mice and found that RLE alleviated decreases in the number of intestinal goblet cells and amount of hepatic triglycerides. It also decreased the immobility time in the forced-swimming test and activation of microglia in the brain, suggesting that RLE has beneficial effects on stress-induced dysfunctions.

KEYWORDS:

• Rosemary
• goblet cells
• stress
• hepatic triglyceride
• forced-swimming test

The term “stress” was proposed by Hans Selye in 1956, denoting the series of biological reactions against strong external stimuli, including neurological, endocrinological, and immunological responses. Excessive stress is a recognized risk factor for various diseases, including depression, cancer, diabetes, and hypertension. Certain foods and their components that have been shown to modulate stress responses and may be effective for maintaining health. Some food components, such as theanine, were reported to exert anti-stress effects on the brain function and behavior of mice [1]. Additionally, several herbs have been widely used in aromatherapy as aids for improving relaxation regimens implemented upon stress conditions. Rosemary (Rosmarinus officinalis L.), an herb belonging to the Lamiaceae family, exerts anti-depressive-like effects in mice [2,3].

Rosemary is frequently employed in cooking and aromatherapy because of its characteristic flavor and aroma. Rosemary leaf extract (RLE) and its functional components, including carnosic acid (CA), have been shown to exert antioxidant [4], neuroprotective [5] and anti-obesity [6–8] effects. It was also previously reported that CA possessed anti-angiogenic [9] and cytoprotective properties [10] both in vitro and ex vivo. Although RLE reportedly has anti-depressive-like effects [2,3], its anti-stress properties have only been partially evaluated in physical disorders. Thus, this study was conducted to evaluate the effect of RLE on goblet cells as intestinal parameter and concentrations of hepatic lipids in stressed mice. Furthermore, the anti-depressant-like effect of RLE intake at a similar dose was evaluated using the forced-swimming test (FST). RLE was found to rescue the number of intestinal goblet cells and content of hepatic triglycerides (TG) under stress, as well as improve the immobility time of mice in the FST and activation of microglia in the brain.

RLE was extracted from dried rosemary leaves as previously described [2]. Dried leaves (10 g, Mitsuse Farm, Saga, Japan) were ground in a mortar, after which 70%-ethanol was added to a volume of 100 mL and extracted for 2 weeks at room temperature in the dark. The liquid fraction was collected by filtration through a 0.22 μM filter (Sartorius AG, Gettingen, Germany), and then concentrated by centrifugation using a freeze trap (VA-500 R, TAITEC Corporation, Saitama, Japan). Dried RLE (26.4 ± 0.9 mg/mL) was collected from liquid fractions. Dried RLE was dissolved in a solution of equal volumes (1:1) of dimethyl sulfoxide and ethanol before use. Mice were maintained according to the guidelines of the Ethical Committee on experimental animal care (permit No. 27–055). Male BALB/c mice weighing 21.9–24.8 g (8-week-old; Japan SLC, Inc., Shizuoka, Japan) were individually housed in polycarbonate cages in a temperature-controlled (24°C) room under a 12-h light-dark cycle. Mice were divided into 3 groups of 6 mice each: normal, no RLE/no stress; stress, no RLE/stress; and stress-RLE, RLE/stress. They were allowed free access to certified diet pellets (CE-2, CLEA Japan, Inc., Tokyo, Japan) and water containing 0.45 mg/mL dried RLE or vehicle. Intakes of food and water were measured every day and drinking water was changed once two days. After the mice followed this diet for 15 days, stress was induced in test group mice by restraining them for 3 h for 5 consecutive days using a flat-bottom rodent holder (KN-469, Natsume Seisakusho Co., Ltd., Tokyo, Japan). During restraint, all animals were restricted from feeding and drinking, and their fecal samples were collected on the first day of the restraint period. Thereafter, the mice were sacrificed on the day following the final restraint after having fasted for 6 h. The small intestine (within 1–3 cm proximal to the pylorus) and liver were collected for histochemical and lipid analyses, respectively. To quantify damage to the small intestine, the number of goblet cells per villus (left side) was counted in periodic acid-Schiff-stained intestinal sections. Hepatic lipids were extracted as described by Folch et al. [11] and measured as previously described [12–14]. Next, the FST was conducted according to the method of Porsolt et al. [15] with slight modifications. This standardized test detects depressive-like behavior; a reduced immobility time is considered to indicate an anti-depressive-like effect. Male BALB/c mice (6 weeks old; Japan SLC, Inc.) were divided into control and RLE groups. A mouse not subjected to restraint treatment was forced to move within an open cylindrical container (diameter 14.5 cm, height 19 cm) with water at 25 ± 1°C to a depth of 14 cm and forced to swim for 10 min. Immobility time was measured using a SCANET MV-40 AQ apparatus (MELQUEST Co., Ltd., Toyama, Japan). The duration of immobility was evaluated during the last 8 min of the 10-min test [15]. The brain was collected, sectioned, and used for ionized calcium-binding adapter molecule-1 (Iba-1) analysis as previously described [16]. Cells with extensive cytoplasm and short processes were counted as activated microglia.

To determine whether RLE induces intestinal changes during stress in restrained mice, we examined its effects on the number of intestinal goblet cells, which are important components of the intestinal epithelial barrier. Intakes of food and water did not significantly differ among groups, and the estimated dose of dried RLE intake was 95.2 ± 8.9 mg RLE/kg body weight (b.w.) per day in the stress-RLE group. Notably, the number of intestinal goblet cells decreased moderately in the stress group but not in the stress-RLE group, compared to that in the normal group (Figure 1(a), (b)). Although fecal weight was increased in both the stress and stress-RLE groups during the restraint period (normal, 0.073 ± 0.070; stress, 0.320 ± 0.082; stress-RLE, 0.250 ± 0.139 g/3 h), its increase was more significant in the stress group (stress, p = 0.001; stress-RLE, p = 0.016, Dunnett’s test). Both of these parameters are known to be altered under stress [17–20], and our results suggest that RLE exerts certain improvements in stress-induced intestinal disorders. It remains unclear whether this effect is related to fecal mucin excretion, expressions of gel-forming mucin MUC2, and anti-microbial peptides produced by Paneth cells, which play a major role in protecting the small intestine. Therefore, these parameters should be evaluated to clarify the effect of RLE. Next, we conducted an FST to examine the anti-depressive-like effect of RLE in mice using a similar dose as that administered in the restraint stress experiment. As a result, the increase in immobility time observed in the control group was abolished following RLE intake (Figure 2(a)). This finding agreed with the tail suspension test performed in ICR mice at doses of 50 and 100 mg RLE/kg b.w. per day [2] (similar to the dose in our study). Thus, these results suggest that RLE has beneficial effects in stress-induced physical and psychological disorders.

Figure 1. Effects of RLE on the number of intestinal goblet cells and concentration of hepatic triglycerides in restraint-stressed BALB/c mice.

Representative images of goblet cells (black arrow) in small intestine are shown (a). Values represent the average number of goblet cells counted in 3 villi per mice (b). Concentrations of hepatic triglycerides are shown (c). Values are presented as the means ± SD (n = 5, 6). Significance compared to the normal (Cont) group was determined by Dunnett’s test (*p < 0.05).

Figure 2. Effects of RLE on immobility time by FST and the activation of microglial cells in BALB/c mice.

The graphs show the immobility time on the day before (Pre), day 7 (D7), and day 14 (D14) of sample intake (a), and activation of microglial cells in the brain. Statistical significance compared to respective Pre values (^†p < 0.05) or to Cont group (*p < 0.05, **p < 0.01) was determined using the t-test. Values are the means ± SD (n = 6, 7).

Because RLE except for CA also contains functional components, such as rosmarinic acid (RA), carnosol, and fragrance components, the observed effect of RLE may be due to the combined effect of several components. Although the CA content of rosemary leaf depends on seasonal factors and growing conditions, it has been reported that RLE extracted by the same method contains 1.3% CA and 4% RA [2]. During the stress response, corticotropin releasing factor (CRF), a major regulator of the hypothalamic-pituitary-adrenal (HPA) axis is known to affect the functions of digestive organs, including changes in the tight junction-related permeability of the intestinal epithelium [21], and decrease in the numbers of intestinal goblet cells [20]. CRF also stimulates large intestinal transit and increases fecal excretion [19]. CA and RA have been shown to protect nerve cells against toxicity induced by corticosterone [2], released post activation of HPA axis. Noted, CA and its derivatives have been reported to be present at substantial concentrations in the brain of rats several hours after oral administration of RLE [22], suggesting that their presence in the brain helps moderate disorders of the digestive organs by protecting against the HPA axis response.

The present study showed that RLE decreased activated microglia (Figure 2 (b)). Furthermore, the number of activated microglia (activated cells/total) was positively correlated with an increase in the immobility time during the FST (D7/Pre, r = 0.727, p = 0.007; D14/Pre, r = 0.751, p = 0.05; n = 13). CA and RA are known to inhibit microglial activation in vitro or in vivo [23,24], and these functional components may contribute to the effect of RLE on microglia in this study. CA has been reported to accumulate in the brain, suggesting that CA can pass through the blood-brain barrier (BBB) and exert neuroprotective effects [5]. Although hydrophilic RA has a low ability to cross the BBB, it had anti-depressive-like effect in the FST in mice [25]. Interestingly, it has been reported that repeated psychological stress induces increased numbers of activated microglia and leads to both neuronal and behavioral changes through inflammation-related cytokines in mice [26]. We propose that RLE potentially exerts an anti-stress effect by suppressing neuroinflammation via a reduction of activated microglia. Further, psychosocial stress can accelerate the oxidative damage in the cerebral DNA, cerebral atrophy and behavioral depression in senescence-accelerated mice [1]. In addition, stress has been reported to induce brain microinflammation, which is also found in Alzheimer’s disease, and cause dysregulation of organ homeostasis [27]. It is necessary to quantify their concentrations in the RLE, but RLE containing both CA and RA is expected to alleviate stress-induced and age-related disadvantages by mediating and sustaining the healthy function of the brain.

We also found that the concentration of hepatic TG was decreased following restraint-induced stress (Figure 1(c)); similar results were obtained in our preliminary experiments (data not shown). Interestingly, this noted decrease was alleviated in the stress-RLE group. The concentrations of hepatic cholesterol (normal, 3.82 ± 0.49; stress, 3.86 ± 0.52; stress-RLE, 3.91 ± 0.46 mg/g liver) and phospholipids (normal, 36.0 ± 2.8; stress, 36.7 ± 2.0; stress-RLE, 36.0 ± 1.7 mg/g liver) did not differ among groups (p > 0.05). Similarly, food intake did not significantly differ among groups (normal, 3.84 ± 0.25; stress, 3.85 ± 0.25; stress-RLE, 4.04 ± 0.77 g/d). These results indicate that RLE intake can improve the metabolism of TG during stress. It has been reported that RLE suppresses the increase in hepatic TG concentrations in high-fat diet-fed mice [6]; thus, RLE may improve the metabolism of lipids under various conditions. In this study, stress was suggested to reduce the concentration of hepatic TG by disrupting, at least partially, the absorption of TG in the intestine. In contrast, there was no significant difference in the serum ratio of GOT/GPT (normal, 7.50 ± 3.24; stress, 6.12 ± 1.79; stress-RLE, 6.96 ± 3.42) and levels of TG (normal, 103 ± 32; stress, 96 ± 22; stress-RLE, 93 ± 17 mg/dL) between the normal and stress groups. These results suggest that the reduction in the concentration of hepatic TG due to stress was likely not caused by a dysfunction in the liver and elevation of blood TG under such conditions. Although stress is known to impair the intestinal absorption of bile acid [18] and elevate the plasma levels of TG by decreasing the rate of the metabolism of blood lipids [28], the metabolism of hepatic TG under stress is less well-understood under normal diets. Further studies are needed to identify the underlying mechanism by which RLE modulates the levels of hepatic TG and metabolism of lipids under stress conditions. CA rich-rosemary extracts (20% CA [7], 40% CA [8]) exert anti-obesity effects in mice fed a high-fat diet. The approximate effective dose at study initiation was estimated to be 700 mg extract/kg b.w. per day (280 mg CA/kg b.w. per day) in obese animals [8]. Taken together with the estimated dose of RLE intake in this study, the anti-obesity effect of rosemary leaf may be exerted at higher CA intake than its anti-stress effect.

In summary, we found that RLE attenuated stress-induced intestinal dysfunction in mice, suggesting that it exerts a beneficial effect on stress-induced physiological disorders. Additionally, RLE was shown to reduce the immobility time in mice subjected to the FST, likely by suppressing activated microglia in the brain. Our findings suggest that RLE containing specific doses of CA and RA are effective in protecting against stress-induced physical and psychological dysfunctions.

Author contributions

T.K., K.N., and K.M. conceived the experiments and wrote the manuscript. T.K., K.N., M.U., H.K., S.S., and K.M. also performed the experiments and analyzed the data. All the authors have read and approved the manuscript.

Acknowledgments

Mouse breeding and RLE extraction using a Freeze Trap were performed at the Analytical Research Center for Experimental Sciences, Saga University.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported in part by JSPS KAKENHI [19K02357] (T.K. and K.M.);Japan Society for the Promotion of Science [Grant-in-Aid for Scientific Research(C)/19K02357].