Arabinogalactan present in the mountain celery seed extract potentiated hypolipidemic bioactivity of coexisting polyphenols in hamsters

Abstract

Context: Previously, we showed the essential oils (EO) of the mountain celery [Cryptotaenia japonica Hass (Umbelliferae)] seeds (MCS) to be a prominent hypolipidemic agent.

Objective: We hypothesized the aqueous extract (AE) of its seeds could also exhibit a comparable nutritional effect.

Materials and methods: Experiments were carried out for compositional analysis, antioxidant assay, and hypolipidaemic assay with AE in hamsters.

Results: AE contained soluble arabinogalactan (AGal) with molecular weight (MW) 878 kDa. AE also was enriched in polyphenolics and flavonoids, reaching 30.4 and 2.20 mg/100 g, respectively. AGal consisted of eight monosaccharides (in mols %), galactose (28.75), arabinose (24.84), glucose (17.91), mannose (6.93), ribose (6.03), fucose (5.83), xylose (5.30), and rhamnose (4.41), with average MW 878 kDa. In vitro, AE showed potent ferrous chelating and DPPH scavenging effects but only moderate H₂O₂ scavenging capability. In hamsters, AE exhibited promising hypolipidemic bioactivity, in particular, the HDL-C and hepatic unsaturated fatty acid (UFA) biosynthesis regarding oleic, linoleic, and arachidonic acids.

Discussion and conclusion: The presence of AGal enhanced the hypolipidemic and antioxidative bioactivity of MCS. MCS is feasibly beneficial to the hepatic de novo UFA synthesis and the hypolipidemics as evidenced by hamster model.

Keywords::

• mountain celery seeds
• aqueous extract
• LDL-C/HDL-C
• hypolipidemic
• polysaccharides
• polyphenols and flavonoids

Introduction

Moutain celery [Cryptotaenia japonica Hass (Umbelliferae)] (MC) for hundreds of years served as common daily vegetable dish in many East Asian countries, now has popularly become an artificial cultivar in Middle Taiwan. It has been highly reputed for its folkloric herbal medicines as a hypotensive, a hypocholesterolemic, and an antiobesity agent (Cheng et al., 2008).

Previously, our laboratory had identified 109 compounds in the seed essential oils (EO), which included categories (number of compounds) of monoterpenoids (9), sesquiterpenoids (31), and alcohols (22) (Cheng et al., 2008). We also demonstrated that all MC-compounded diets except the EO-added prominently exhibited strong hypolipidemic, especially, HDL-C elevating capability (Cheng et al., 2008).

Earlier, Saleh et al. (1985) identified a number of active compounds. 3-n-Butylphthalide was demonstrated to be a hypotensive as well as a hyocholesterolemic (Le & Elliott, 1991). Furthermore, five anticarcinoma components in celeries were identified to possess a glutathione S-transferase (GST) activating activity (Zheng et al., 1993; Tsi et al., 1995). More recently, by preliminary survey we found that the aqueous soluble fraction of MC seeds (AE) contained huge amount of soluble dietary fibers (SDF), polyphenols, and flavonoids. Since the documented data about its soluble dietary fiber are still lacking, considering the easier handling of these seed dishes, we hypothesized that AE could exhibit similar nutritional bioactivity comparable with its essential oils. To confirm this, we performed simultaneously both the in vitro and in vivo animal experiments.

Materials and methods

Chemicals and reagents

Linoleic acid, 1,1-diphenyl-2-picrylhydrazyl radical (DPPH), ferrous chloride (FeCl₂), ferrozine, nitroblue tetrazolium (NBT), dihydronicotinamide adenine dinucleotide (NADH), phenazine methosulfate (PMS), ethylene diaminetetraacetate (EDTA) (disodium salt), H₂O₂ (30%), peroxidase, phenol red, n-butylhydroxyanisole (BHA), n-butylhydroxy toluene (BHT), ascorbic acid, and citric acid were products of Sigma Chemical Co. (St. Louis, MO). All other reagents and chemicals used were of reagent grade.

Source of mountain celery seeds

Fresh MC seeds were purchased from a local farm located in Nan-Tou, a middle Taiwan county. The seeds were stored immediately in the dark at −20°C before treatment (Cheng et al., 2008). The plant has been authenticated by the Research Centre for Chinese Herbal Medicines in Taichung City (Taiwan).

Preparation of desiccated powder of mountain celery seeds

Fresh MC seeds were desiccated at 60°C for 4 h as previously cited (Cheng et al., 2008). The crushed particles having size ranging within mesh #40-60 were collected (MCSP) (Table 1).

Table 1.  Ingredients of experimental animal diets.^a,b

Ingredients	C	H	MP1	MP2	MP3	AE1	AE2	AE3
Casein	20	20	20	20	20	20	20	20
Sucrose	15	15	15	15	15	15	15	15
Corn starch	50	45	45	45	45	45	45	45
Corn oil	2.5	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Lard	2.5	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Mx^c	3.5	3.5	3.5	3.5	3.5	3.5	3.5	3.5
Vx^d	1	1	1	1	1	1	1	1
Choline	0.2	0.2	0.2	0.2	0.2	0.2	0.2	0.2
Methionine	0.3	0.3	0.3	0.3	0.3	0.3	0.3	0.3
α-Cellulose	5	5	5	5	5	5	5	5
MP	0	0	0.5	1.0	3.0	—	—	—
AE	—	—	—	—	—	0.05	0.10	0.50

^aThis formulation was based on the formulation of AIN-76.

^bC: the regular diet for control group; H: high lipid diet; MP: desiccated MC seed powder, and MP1-MP3 denote the diet H plus, respectively, the desiccated MC seed powder 0.5, 1.0, and 3.0% w/w. AE: the dried aqueous extract of MC sees, and AE1-AE3 denote the diet H plus, respectively, 0.05, 0.1, and 0.5% w/w of dried AE.

^cMx (Mineral premix): CaHPO₄ 2H₂O, NaCl, K₃C₆H₅O₇, K₂SO₄, MgO₃, MnO₃, Fe-citrate, ZnCO₃, CuCO₃, KI, NaSeO₃, K₂SO₄ Cr₂(SO₄)₃ 24H₂O.

^dVx (Vitamin premix): thiamin hydrochloride, pyridoxine hydrochloride, riboflavin, nicotinic acid, vitamin B₁₂, retinyl palmitate, vitamin D₃, vitamin E, vitamin K.

Preparation of aqueous extract of mountain celery seeds

MCSP (100 g) was transferred into a 1-L reaction vessel, to which 1000 mL of double distilled water was added. The mixture was refluxed at 100°C for 1 h and filtered. The extraction was repeated twice. The filtrates were combined and evaporated with a 5000-mL rotary evaporator to dryness. The total desiccated extract was weighed, and the percentage yield was taken and stored at −20°C for further use.

Analyses

Total phenolics

The total polyphenolic content was assayed according to Taga et al. (1984). The amount of polyphenols was calculated against the calibration curve established and expressed as the mg of gallic acid equivalent (mgGAE) per 100 g of the desiccated.

Total flavonoids

The method of Quettier-Deleu et al. (2000) was followed to determine the total flavonoid content. The content of isoflavonoids was calculated against the calibration curve and expressed in milligrams of quercetin equivalent (mg QE) per 100 g of the desiccated.

Dietary fibers (DF), insoluble dietary fibers (IDF), and SDF

The official procedure of Association of Official Analytical Chemists (AOAC) (1995) for ingredients analysis (method 991.43) was adopted to determine the DF, the IDF, and the SDF.

Molecular weight of SDF

Sample SDF (10 mg) was accurately weighed and placed in a reaction vessel. To the wighed sample, 1 mL of NaOH (1 N) was added. The remaining procedures were conducted according to Ker et al. (2005). A total of 50 tubes were collected. The optical density was measured simultaneously at 490 and 280 nm.

Polysaccharides in SDF

The content of polysaccharides in SDF was determined by the method described by Ker et al. (2005). The optical density was measured at 490 nm.

GC/MS analysis of polysaccharide composition in SDF—Hydrolysis, reduction, derivatization of monosaccharides

Similar method described by Ker et al. (2005) was followed to determine the polysaccharide composition in SDF. Briefly, the final dehydrated product was transferred into a 1-mL reaction vessel, lyophilized, and analyzed with GC/MS as described (Ker et al., 2005). After the polysaccharide fraction was separated from the Sephadex G-100 column, each 1 mL of the polysaccharide fraction was added with 1 mL of phenol solution (5%); then it was thoroughly mixed with 5 mL of sulfuric acid and shaken well. After the solution was cooled to room temperature, the absorbance [optical density (OD)] was measured at 490 nm.

Antioxidative capability of the aqueous extract

Chelating capability for ferrous ions

According to Lin et al. (2008a), 3.7 mL of methanol and 0.1 mL of ferric chloride solution (2 mM) were added to 1 mL of aqueous extract (0.5-20 mg/mL). The mixture was left to stand for 30 sec to facilitate the reaction. To the mixture, 0.2 mL of ferrozine (5 mM) was added and then left to stand for 10 min to facilitate the reaction. When reaction was completed, the absorbance of the colored reaction mixture was taken at 562 nm using a Hitachi U-2001 Spectrophotometer. Standard solutions of EDTA and citric acid, each 0.5∼5 mg/mL, were used as the controls.

Scavenging capability for DPPH free radicals

Lin et al. (2008a) was followed to determine the DPPH scavenging capability. Briefly, the dried AE was redissolved in 75% ethanol to make a concentration of 20 mg/ mL to serve as a stock. For analysis, the stock solution was diluted with ethanol to 0.5 to 20 mg/mL [aqueous ethanol extract (AEE) solution]. To 4 mL of AEE (0.5-20 mg/ mL), 1 mL of freshly prepared methanol 1,1-diphenyl-2-picrylhydrazyl (DPPH) free radicals was added. The mixture was left to stand in dark for 30 min. The absorbance was read at 517 nm using a Hitachi U-2001 Spectrophotometer. BHA and L-ascorbic acid were used as the positive controls and citric acid as the negative control.

Scavenging capability for H₂O₂

The testing procedures were conducted as previously described by Cheng et al. (2008). The optical density was measured at 610 nm using a Hitachi U-2001 Spectrophotometer. L -ascorbic acid and BHT were used as reference controls.

Hamster diets, serum, and hepatic lipid extract

Sixty-four male Syrian hamsters, aged 5 to 6 weeks, were purchased from the National Laboratory Animal Centre. All studies performed with this hamster model were approved by the Hungkuang University Ethic Committee in accordance with Helsinki Declaration in 1975. For the first 2 weeks, the hamsters were acclimated by supplying only regular market diet Fu-Sow (Longevity) Brand. Then the hamsters were randomly grouped by body weight into 8 groups (8 hamsters in each group, 2 in each stainless cage) and fed on different diets: the control group C fed on regular diet; the high-lipid diet (Group H). The groups MP1 through MP3 were fed diet H plus 0.5, 1.0, and 3.0% w/w, respectively, and the groups AE1 through AE3 were fed diet H plus 0.05, 0.10, and 0.50% w/w, respectively (Table 1) (Anonymous, 1977).

The animal room was conditioned at 24 ± 1°C with a relative humidity maintained at 40% to 60%. Light cycle was changed every 12 h. Water and meal takings were ad libitum. Body weight and amount of diets consumed were recorded every 2 days until the end of the experiment. After 10-week feeding, the hamsters were fasted for 12 h before being euthenized with CO₂. The hamsters were bled from the abdomen artery. The whole blood was centrifuged at 4°C at 1800g for 10 min. The sera were separated and stored at −70°C for lipid determination. The livers were dissected immediately after euthenizing and washed twice with ice-cold saline (150 mM). After the adhered water was drained off and their weights taken, the livers were transferred into a zippered bag and stored at −70°C for analysis of hepatic lipid content according to Fang et al. (2004). The assay methods used for analysis of serum triglycerides, serum total cholesterol, serum LDL-C, and serum HDL-C were in majority adopted from The Manual of Combined Agents provided by the Randox Laboratories Ltd.

Serum total cholesterol, serum low-density lipoprotein cholesterol (LDL-C), serum high-density lipoprotein cholesterol (HDL-C), serum total triglycerides (TG)

For all serum lipid analyses, similar procedures were conducted according to Cheng et al. (2008).

Extraction of hepatic lipids

The lipid extraction from hepatic tissues was conducted following the method of Lin et al. (2008). The finally obtained hepatic lipid extract (HLE) was stored at −20°C for further analysis.

Determination of hepatic triglycerides, total hepatic cholesterol (CT)

The method described by Lin et al. (2008b) was used for determination of hepatic triglycerides and total hepatic cholesterol. Briefly, for analysis of hepatic lipids, hamsters were euthenized on completion of feeding experiment and the livers were excised. After rinsed with Ringer solution, livers were wiped gently with tissue papers to get rid of any moisture adsorbed on the surface before their weights were measured. Preparation of hepatic lipids extracts was performed according to Folch et al. (1957). The final chloroform–methanol (2:1) extracts were combined and made to a volume of 10 mL with the same extraction solvent and stored at −20°C for further analysis. Hepatic lipid extracts (10 μL) was transferred into a 1.5 mL centrifuge tube. Similar procedures were performed as mentioned in the serum lipid analyses.

Determination of total hepatic phospholipids

Similarly, Lin et al. (2008b) was followed to determine the total hepatic phospholipids.

Hepatic fatty acid extract

Method of Folch et al. (1957) was followed to extract the hepatic fatty acids. To 3 mL of Folch extractive agent, 2 mL of the HLE was added (Folch et al., 1957). The solution was dried under nitrogen blowing. To the residue, 4 mL of methanol KOH (0.5 N, KOH/CH₃OH) was added and refluxed at 100°C for 15 min on the water bath. A solution of BF₃/CH₃OH (5 mL) was added and the reflux was continued for 2 min. In the final reflux step, n-hexane (3 mL) was added and refluxed for another 1 min. On cooling to room temperatue, a saturated solution of NaCl (2 mL) and the internal standard (methyl n-pentadecanoate/n-hexane, 0.5%, 1 mL) was added. After agitated for 3 min, the mixture was left to stand for separation. The organic layer was transferred into the reaction vessel, added with anhydrous sodium sulfate, shaken well to completely dehydrate the water content in the solution, and filtered. The filtrate was used for assay of hepatic fatty acid content by GC and GC/MS analyses. GC operational conditions and GC/MS qualitative analysis were similarly perfomed according to Ker et al. (2005).

Statistical analysis

Data obtained in the same group were analyzed by Student’s t-test with computer statistical software SPSS 10.0 (SPSS, Chicago, IL). Statistical Analysis System (2000) software was used to analyze the variances and Scheffe multiple comparison test was used to test their significances of difference between paired means. Significance of difference was judged by a confidence level of P < 0.05.

Results and discussion

The characteristics of AE and MP

AE occupied a total content of 16.2% in the mountain celery seeds. AE contained much higher contents of total polyphenolics and total flavonoids than MP, reaching 9.21- and 9.16-fold over MP, respectively (Table 2). As often cited, polyphenols (1.2 μM of gallic acid or 5 μM of quercetin) are effective LDL antioxidant and strong transition metal–chelating agents (Srivastava et al., 2006).

Table 2.  Main bioactive constituents of the pulverized mountain celery seeds (MP) and its aqueous extract (AE).^a

Sample	Total polyphenols (mg/100g)	Total flavonoids (mg/100g)	SDF (mg/100g)	IDF (mg/100g)	Yield (%)
MP	3.3 ± 0.7	0.2 ± 0.1	41 ± 4	447 ± 39	—
AE	30.4 ± 2.3	2.2 ± 0.4	301 ± 21	0	16.2 ± 3.9

^aMP: the pulverized mountain celery seeds. AE: the aqueous extract of mountain celery seeds. SDF: the soluble dietary fibers. IDF: the insoluble dietary fibers.

Note: All contents are expressed in mg/g. Data state as Mean ± SD from triplicate determinations (significance level P < 0.01).

Dietary fiber contents

Mountain celery seeds contain quite a few amounts of DF. The SDF content was 41 and 301 mg/100g in MP and AE, respectively. The difference reached 7.34 fold. In contrast, MP had higher content of IDF, which was totally not found in AE (Table 2).

As often cited, SDF at 1% has been considered to be very efficient in suppressing plasma lipid levels, especially hepatic TG and TC (Frias & Sgarbieri, 1998). Literature elsewhere indicated that for human health, a dose of 20 g SDF/day would be sufficient (Knopp et al., 1999). In present study, the GC/MS profile revealed the presence of eight monosacharides (moles %) including galactose (28.75), arabinose (24.84), glucose (17.91), mannose (6.93), ribose (6.03), fucose (5.83), and xylose (5.30), with MW average at 878 kDa (Figure 1, Table 3), hence named arabinogalactan. Xu et al. (2006) demonstrated that the abundant occurrence of galactose, arabinose, and xylose were derived from galactoarabinoxylans, whereas glucose was mostly derived from β-glycan.

Table 3.  The monosaccharide profile in arabinogalactan obtained from the mountain celery seeds and its retention time in GC/MS spectrum.

Peak no./Monosacharide	Mole % in galactoarabinosan	Retention time (min)
8. Galactose	28.75	37.354
4. Arabinose	24.84	25.368
7. Glucose	17.91	32.941
6. Mannose	6.93	31.264
3. Ribose	6.03	23.919
2. Fucose	5.83	22.996
5. Xylose	5.30	28.857
1. Rhamnose	4.41	22.191
Average molecular weight	878 kDa	—

Note: The peak no. disignates the number assignd in the GC/MS spectrum shown in Figure 1.

Figure 1.  GC/MS profile of monosaccharides in the hydrolysate of the glucoarabinogalactan obtained from the mountain celery seeds. Monosaccharide was assingned by peak number as: 1 = rhamnose. 2 = fucose. 3 = ribose. 4 = arabinose. 5 = xylose. 6 = mannose. 7 = glucose. 8 = galactose.

Ferrous ion chelation capability (FICC)

AE showed potent FICC and reached an 82%-89% chelation at concentrations of 0.1-5.0 mg/mL comparing with EDTA and citric acid (Figure 2). Correspondingly, the FICC may mimic the anticatalytic oxidative power occurring in many biosystems including LDL, implicating the possible role of AE in antiatherosclerogenesis (Chait et al., 1993). Prevously, we reported that the MC seed essential oils exhibit moderately low antioxidant bioactivity (Cheng et al., 2008), a fact now evidenced by AE.

Figure 2.  Ferrous ion chelating capability of the aqueous extract of mountain celery seeds.

DPPH free radical scavenging capability (FRSC)

AE at concentrations 0.1-5.0 mg/mL exhibited very efficient DPPH FRSC. Comparable effects were seen in AE and BHA (83%-93% vs. 95.35%) but far exceeding that of ascorbic acid (Figure 3).

Figure 3.  DPPH free radical scavenging capability of the aqeous extract of mountain celery seeds.

Hydrogen peroxide scavenging capability (HPSC)

AE showed only moderate antioxidant capability against hydrogen peroxide. Even at 5 mg/mL, AE only exhibited 73% of HPSC, being far lower than ascorbic acid yet amazingly higher than BHT (Figrue 4). AE was more effective than the seed essential oil (Cheng et al., 2008).

Figure 4.  H₂O₂ scavenging capability of the aqueous extract of mountain celery seeds.

Serum and hepatic lipids

The body weight gain and fed efficiency were seen comparable among all groups (Table 4). The weight of liver and kidney appeared almost unchanged among all groups. However, the lipid patterns varied more apparently depending on different diet feeding.

Table 4.  Effects of mountain celery seed powder (MP) and its aqueous extract (AE) on serum and hepatic nutritional parameters in male hamsters.^a

Parameter	C	H	MP1	MP2	MP3	AE1	AE2	AE3
Food intake	8.0 ± 0.1	8.0 ± 0.1	8.0 ± 0.2	8.0 ± 0.4	8.1 ± 0.2	8.1 ± 0.4	8.0 ± 0.3	8.0 ± 0.6
body gain (g)	14.3 ± 3.7	14.2 ± 2.8	14.3 ± 2.3	14.3 ± 2.4	14.4 ± 2.6	14.2 ± 2.0	14.3 ± 2.1	14.2 ± 2.5
Feed efficiency	1.35 ± 0.07	1.34 ± 0.06	1.35 ± 0.10	1.34 ± 0.07	1.36 ± 0.09	1.34 ± 0.10	1.35 ± 0.08	1.34 ± 0.10
Organ wt Gain, g
Liver	4.3 ± 0.1	4.5 ± 0.2	4.3 ± 0.1	4.4 ± 0.1	4.3 ± 0.1	4.4 ± 0.2	4.4 ± 0.2	4.3 ± 0.2
Kidney	0.8 ± 0.0	0.8 ± 0.0	0.8 ± 0.1	0.8 ± 0.0	0.8 ± 0.1	0.8 ± 0.1	0.8 ± 0.1	0.8 ± 0.1
Serum
sTG	138.2 ± 4.8^a	190.9 ± 4.0^f	160.1 ± 3.1^d	154.4 ± 2.5^b	146 ± 2.6^b	171.1 ± 6.3^e	156.6 ± 3.9^c	138.2 ± 3.7^a
sTC	121.4 ± 4.6^a	167.0 ± 4.4^e	143.3 ± 7.8^c	144.3 ± 7.6^c	145.5 ± 8.7^c	146.7 ± 4.7^d	144.3 ± 3.2^c	126.0 ± 4.0^b
LDL-C	30.3 ± 3.9^a	67.1 ± 3.5^d	49.6 ± 2.3^c	49.2 ± 2.2^c	48.1 ± 3.0^c	51.9 ± 5.1^b	48.0 ± 4.8^c	34.9 ± 3.5^b
HDL-C	65.8 ± 1.9^b	62.9 ± 2.2^a	63.0 ± 1.6^a	63.4 ± 2.0^a	63.7 ± 6.9^a	69.6 ± 2.5^d	69.9 ± 1.8^d	67.0 ± 2.9^c
LDL/HDL-C	0.46 ± 0.1^a	1.07 ± 0.0^f	0.78 ± 0.1^e	0.77 ± 0.1^e	0.74 ± 0.0^d	0.75 ± 0.1^d	0.69 ± 0.0^c	0.52 ± 0.0^b
Liver (mg/g)
hTC	20.5 ± 0.8^a	25.1 ± 0.9^e	24.6 ± 1.9^d	24.5 ± 1.6^d	23.1 ± 1.5^c	23.2 ± 1.0^c	23.1 ± 0.8^c	21.5 ± 0.6^b
hTG	20.5 ± 0.5^a	23.4 ± 0.4^d	22.7 ± 0.8^c	22.6 ± 1.0^c	22.2 ± 1.1^c	21.5 ± 1.0^b	21.4 ± 0.8^b	20.7 ± 0.7^a
Phospholipids	75.9 ± 1.6	74.6 ± 3.5	75.0 ± 2.5	75.1 ± 2.2	74.7 ± 2.3	75.1 ± 2.7	74.8 ± 3.2	75.5 ± 3.2

^aC: the regular diet for control group; H: high lipid diet; MP: desiccated MC seed powder, and MP1-MP3 denote the diet H plus, respectively, the desiccated MC seed powder 0.5, 1.0, and 3.0% w/w; AE: the desiccated aqueous extract of MC seeds. And AE1-AE3 denotes the diet H plus, respectively, 0.05, 0.1, and 0.5% w/w of dried AE.

Note: Feed efficiency = [wt gain (g)/total diet intake (g)] × 100, after fed for 10 weeks. hTC: hepatic total cholesterol; hTG: hepatic total triglycerides; sTC: serum total cholesterol; sTG: serum total triglycerides; LDL-C: low-density lipoprotein-cholesterol; HDL-C: high-density lipoprotein-cholesterol; LDL-C/HDL-C: ratio of low density lipoprotein-cholesterol/high density lipoprotein-cholesterol. All data state as Mean ± SD from triplicate determinations. Values in the same row with different superscripts are significantly different (P < 0.05).

Serum triglycerides

Both MP and AE effectively reduced the TG levels in a dose-responsive manner. AE showed more effective and prominent suppression than MP (Table 4), as seen in AE3 group based on 10-week feeding duration (Table 4), implying the SDF of MC exhibiting a significant nutritional bioactivity to affect the in vivo lipid biosynthesis (P < 0.05), consistent with Frias and Sgarbieri (1998) and Knopp et al. (1999).

Serum LDL-C

Both MP and AE significantly reduced the LDL-C, but MP was less effective than AE in this respect. Speculatively, a dose of AE 0.50 g/100 g diet could be sufficiently enough to treat hyperlipidemia (Table 4).

Serum HDL-C

Comparing with group H, MP groups failed to raise any HDL-C levels. In contrast, AE effectively raised the HDL-C levels (Table 4), resulting in reduced ratio of LDL-C/HDL-C in MP groups to 0.74-0.78 and to 0.52-0.75 in AE groups, implicating that AE could be more efficient to prevent cardiovascular diseases (P < 0.05) (Table 4), consistent with our previous report with the essential oils of mountain celery seeds (Cheng et al., 2008). Literature often cited that HDL-C level is inversely related with coronary atheroscrotic diseases (CAD) (Moorjani et al., 1988). In this regard, AE can be an effective cardiovascular protective nutraceutic. As often cited, CAD prevention would rely more on the reduction of LDL-C/HDL-C ratio than simply on single parameter either LDL-C or HDL-C (Moorjani et al., 1988).

Hepatic triglycerides

Similar to the above, MP did not show any significant effect on hepatic TG. In this regard, unlike AE that exerted more efficient suppressive activity. All AE groups, especially the AE3, significantly recovered the hyperlipidaemic levels (Table 4).

Hepatic total cholesterol

As for hepatic TC reduction, AE was more efficient than MP (Table 4), and only higher dose of MP like MP3 was comparable with all AE groups.

Hepatic phospholipids

Both MP and AE were all ineffective to affect the hepatic phospholipid content in all groups. Literature indicated that the higher level of phospholipid would facilitate a greater hypocholesterolemic effect (Liu et al., 1995) (Table 4), an implication in phospholipid metabolism of hamsters totally unaffected by MC (Table 4).

Hepatic fatty acid profile

The hepatic acid pattern varied greatly depending on diet composition. Both AE and MP effectively suppressed the hepatic palmitic acid (16:0) and stearic acid (18:0) biosynthesis, but instead they elevated hepatic oleic acid (Δ18:1), linoleic acid (Δ18:2) (LA), and arachidonic acid (20:4) (ARA) levels (P < 0.05). Both oleic and linoleic acids are hypolipidemics and vasodilators (Hiley & Hoi, 2007; Jenko & Vanderhoek, 2008).

α-Linolenic acid (18:3) and LA accelerate the biosynthesis of docosahexaenoic acid (DHA) in many organs (Sarkadi-Nagy et al., 2004). DHA, a crucial nervous system n-3 PUFA, may be obtained in the diet or synthesized in vivo from dietary α-linolenic acid (DeMar et al., 2008). Otherwise, the LA is activated by ATP and CoA as cosubstrates to produce linoleic-CoA through the catalytic reaction of acyl CoA lygase (Stryer & Lubert, 2000). Then through a serial steps involving Δ⁶ desaturase to create site of unsaturation at C-6, the addition of malonyl CoA to increase the chain length by 2 carbons accompanied by the simultaneous release of CO₂ by mechanism similar to fatty acid biosynthesis, Δ⁵ desaturase to create site of unsaturation at C-5, and finally splitting off CoA by action of thiolnase to produce ARA (Stryer & Lubert, 2000). Apparently, these related enzymes were more upregulated by AE than MP (Table 5), resulting in higher conversion of arachidonic acid, provided the biosynthesis of linoleic acid is similarly upregulated (Table 5). With respect to the biosynthesis of ARA, AE exhibited more prominent effect than MP (Table 5).

Table 5.  Change of hepatic fatty acid patterns in hamsters after having consumed the mountain celery seed powder and its aqueous extracts.^a

Fatty acid	C	H	MP1	MP2	MP3	AE1	AE2	AE3
16:0	5.2 ± 0.2^a	7.2 ± 0.6^e	6.3 ± 0.5^d	5.9 ± 0.4^c	5.9 ± 0.5^c	5.6 ± 0.2^b	5.9 ± 0.4^c	5.3 ± 0.3^a
18:0	2.8 ± 0.4^a	3.9 ± 0.3^d	3.4 ± 0.4^c	3.3 ± 0.3^c	3.9 ± 0.5^d	3.1 ± 0.4^b	3.0 ± 0.3^b	2.9 ± 0.4^a
18:1	8.1 ± 0.6^b	7.2 ± 0.6^a	7.1 ± 0.6^a	8.1 ± 0.8^b	8.5 ± 0.6^e	8.3 ± 0.7^c	8.1 ± 0.5^b	8.4 ± 0.8^d
18:2	11.4 ± 1.7^f	6.5 ± 0.2^a	8.1 ± 0.8^d	7.6 ± 0.7^c	8.0 ± 0.5^d	7.5 ± 0.4^c	7.3 ± 0.5^b	8.4 ± 0.6^e
20:4	3.1 ± 0.3^f	2.6 ± 0.3^c	2.4 ± 0.2^a	2.4 ± 0.3^a	2.6 ± 0.5^c	2.5 ± 0.2^b	2.7 ± 0.4^d	2.8 ± 0.3^e

^aC: the regular diet for control group; H: high lipid diet; MP: desiccated MC seed powder, and MP1-MP3 denote the diet H plus, respectively, the desiccated MC seed powder 0.5, 1.0, and 3.0% w/w; AE: the dried aqueous extract of MC seeds, and AE1-AE3 denote the diet H plus, respectively, 0.05, 0.1, and 0.5% w/w of dried AE.

Note: The amount of fatty acid is expressed in mg/g. All data state as Mean ± SD from triplicate determinations. Values in the same row with different superscripts are significantly different (P < 0.05).

The rate limiting reactions are well differentiated in vivo, for the biosynthesis of arachidonic acid may be at the level of the 18:3 to 20:3 elongase (Wynn & Ratledge, 2000). Speculatively, the increased fatty acid and triglyceride biosynthesis in rats and hamsters may be due to the activation of squalene synthetase activity (Ugawa et al., 2003). As seen, levels of serum TC and TG, and hepatic TC and TG were more effectively suppressed by AE (Tables 4 and 5), implicating AE was more actively affecting these enzymes.

Taken together, AE contains abundant amount of soluble arabinogalactan, polyphenolics, and flavonoids, which virtually contributed to its in vitro powerful ferrous chelating and DPPH scavenging effects. The high-soluble arabinogalactan and polyphenols in AE play a synergistic hypotriglyceridemic and hypocholesterolemic role in hamsters. AE is a very effective hypolidemic for elevation of HDL-C and lowering of LDL-C. Simultaneously, AE is a prominent activator for hepatic unsaturated fatty acid de novo synthesis involving oleic, linoleic, and the arachidonic acids.

Acknowledgment

The work was in part financially supported by Grant NSC 91- 2626-B-241-004, NSC 96-2320-B-241-006-MY3, NSC 97-2313-B-241-007-MY3 and NSC 97-2320-B-039-049-MY3 from the National Science Council, Taiwan.

Declaration of interest

All the authors do not have any conflict of interest in submitting this paper.