Accumulation of lipid in rat liver was induced by vitamin B₆ deficiency and was ameliorated by supplemental phosphatidylcholine in the diet

Abstract

We investigated the efficacy of supplementing the diet with pteroylmonoglutamic acid (PGA), choline, or phosphatidylcholine (PC) in ameliorating the lipid accumulation in rat liver that is induced by vitamin B₆ (B₆) deficiency. In Experiment 1, male Wistar rats were fed a control, B₆-deficient, or PGA-, choline-, or PC-supplemented (10 mg, 4 g, and 6.3 g/kg of diet, respectively) B₆-deficient diet containing l-methionine at 9 g/kg of diet for 35 days. In Experiment 2, rats were fed a control, B₆-deficient, or PC-supplemented (at 3.15, 6.3, or 12.6 g PC/kg of diet) B₆-deficient diet for 35 days. Choline or PC supplementation ameliorated liver lipid deposition and returned plasma lipids to normal. Judging from these results, it appeared that B₆ deficiency decreased the synthesis of PC in the liver, thereby decreasing the secretion of very low-density lipoproteins, and in consequence producing lipid accumulation in the liver and reductions of plasma lipids.

Vitamin B₆ deficiency caused lipid accumulation in liver owing to decreased secretion of VLDL which was reduced by decreased synthesis of PC by disturbed methionine metabolism.

Key words:

• vitamin B₆ deficiency
• liver lipids
• phosphatidylcholine
• very low-density lipoprotein (VLDL)
• methionine metabolism

Introduction

Vitamin B₆ (B₆) participates in methionine (Met) metabolism as a coenzyme (Fig. 1). Previous studies have shown that deficiency of B₆ increased the levels of homocysteine (Hcy) in the plasma and liver of rats, the former being an independent risk factor for cardiovascular disease.^1,2) Vitamin B₆ is involved in the catabolism of Hcy as pyridoxal 5′-phosphate (PLP), which is a coenzyme of cystathionine β-synthase (EC 4.2.1.22; CBS) and cystathionine γ-lyase (EC 4.4.1.1; CGL) (Fig. 1). Thus, deficiency of B₆ reduces the activities of CBS and CGL, leading to increases of Hcy in liver and plasma.

Table 1. Composition of the experimental diets (g/kg diet) (Exps. 1 and 2).

Ingredients	Group
	Exps. 1 and 2		Exp. 1			Exp. 2
	C^a	D^b	PGA10D^c	Ch4D^d	PC6D^e	PC3D^f	PC6D^g	PC12D^h
Vitamin-free casein	200	200	200	200	200	200	200	200
Soybean oil	56.3	56.3	56.3	56.3	50	53.15	50	43.7
Phosphatidylcholine	–	–	–	–	6.3	3.15	6.3	12.6
Sucrose	487.693	487.7	487.692	485.7	487.7	487.7	487.7	487.7
Gelatinized cornstarch	150	150	150	150	150	150	150	150
Cellulose powder	50	50	50	50	50	50	50	50
Vitamin B6-free vitamin mixture	10	10	10	10	10	10	10	10
Pyridoxine hydrochloride	0.007	–	–	–	–	–	–	–
Folic acid	–	–	0.008	–	–	–	–	–
AIN-76 mineral mixture	35	35	35	35	35	35	35	35
l-Methionine^i	9	9	9	9	9	9	9	9
Choline bitartrate	2	2	2	4	2	2	2	2

^a C: Control

^b D: B₆-deficient

^c PGA10D: 10 mg of pteroylmonoglutamic acid (PGA)/kg diet; the content of PGA was 5 times higher than the C diet

^d Ch4D: 4 g of choline bitartrate/kg diet; the content of choline bitartrate was 2 times higher than the C diet

^e PC6D: 6.3 g of soybean phosphatide extract (phosphatidylcholine; PC)/kg diet; the content of choline in PC6D was same as in the Ch4D diet

^f PC3D: 3.15 g of PC/kg diet

^g PC6D: 6.3 g of PC/kg diet; the composition was same as the PC6D diet (Exp. 1)

^h PC12D: 12.6 g of PC/kg diet

ⁱ The content of l-methionine was 3 times higher than the AIN-76 diet.

Fig. 1. Metabolic pathways linking methionine and lipid metabolism.

Notes: BHMT, betaine–homocysteine methyltransferase; CBS, cystathionine β-synthase; CDP–choline, cytidine diphosphate–choline; CGL, cystathionine γ-lyase; CK, choline kinase; CPT, choline phosphotransferase; CT, phosphocholine cytidylyl transferase; Hcy, homocysteine; Met, methionine; MS, methionine synthase; 5-MTHF, 5-methyltetrahydrofolate; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PEMT, phosphatidylethanolamine N-methyltransferase; PLP, pyridoxal 5′-phosphate; SAH, S-adenosylhomocysteine; SAM, S-adenosylmethionine; THF, tetrohydrofolate.

Hcy can be remethylated to Met by vitamin B₁₂-dependent methionine synthase (EC 2.1.1.13; MS) (Fig. 1). The methyl group in 5-methyltetrahydrofolate (5-MTHF) is firstly transformed on cobalamin and secondarily transformed to Hcy. During the reactions, the complex of MS–cobalamin is oxidized and converted to inactive form. The inactive form is reactivated by the presence of NADPH. In our previous experiments, pteroylmonoglutamic acid (PGA) supplementation ameliorated the hyperhomocysteinemia caused by B₆ deficiency, presumably as a consequence of enhanced removal of Hcy by MS.³⁾ In addition, PGA supplementation tended to suppress the increase in liver weight to which lipid deposition contributes.

It has been reported that B₆ deficiency was sometimes accompanied by fat deposition in the livers of rats fed a 70% casein diet.⁴⁾ This was not consistently observed in our B₆-deficient rats. However, Yamamoto et al.⁵⁾ found that supplemental l-Met (9 g/kg of diet) during B₆ deficiency caused an accumulation of lipids in the liver (pair-fed control: 52.1 ± 1.1 mg/g liver, B₆ deficiency: 189.1 ± 30.6 mg/g liver (p = 0.0039; using Mann–Whitney U-test)) with concomitant severe hyperhomocysteinemia. Therefore, it was speculated that disturbed Met metabolism as a consequence of B₆ deficiency might be a factor in the deposition of lipids in the livers of B₆-deficient rats. However, the details of the mechanism and its reduction by dietary manipulation are not clear.

If disturbed metabolism of Met, especially Hcy accumulation, is involved in the lipid accumulation in the livers of B₆-deficient rats, PGA, choline, and phosphatidylcholine (PC) might alleviate it, because an alternative pathway for Hcy metabolism is provided by the betaine–homocysteine methyltransferase (EC 2.1.1.5;BHMT) pathway, in which betaine acts as the methyl donor (Fig. 1). Betaine is a metabolite of choline, which animals obtain primarily from the diet or the catabolism of PC.⁶⁾ Therefore, in this study, we examined the effects of PGA, choline, or PC supplementation on liver lipid accumulation in rats fed a B₆-deficient l-Met-supplemented diet to investigate the mechanism of liver lipid accumulation caused by B₆ deficiency in relation to Met metabolism.

Materials and methods

Animals

The study was approved by the Committee for Animal Research and Welfare of Gifu University on the Proper Use of Laboratory Animals. Male 4-week-old Wistar rats weighing 80–100 g were obtained from SLC (Hamamatsu, Japan). The rats were housed individually in stainless steel wire-bottomed cages in an animal room at 23 ± 1 °C in an inverted 12-h light–dark cycle (light on at 06:00 h) and were acclimatized to an AIN-76 diet for 5 days before the start of each feeding experiment.

Diets and experimental procedures

In Experiment 1, the rats were randomly divided into five groups (n = 7 each) that were fed one of the following experimental diets: an AIN-76 diet-based control diet (C), a B₆-deficient diet (D), and B₆-deficient diets supplemented with PGA (10 mg of PGA/kg of diet; PGA10D), choline (4 g of choline bitartrate/kg of diet; Ch4D), or PC (6.3 g of soybean phosphatide extract/kg of diet; PC6D). The PGA10D diet contained fivefold the amount of PGA, and the Ch4D diet contained twofold the amount of choline, compared to the C diet. The PC6D diet contained the same amount of choline in PC as did the Ch4D diet. All diets contained 9 g of l-Met per kg. The compositions of the experimental diets are shown in Table 1. The D group had free access to the D diet, and the other groups were pair-fed to the D group. Body weight was measured daily at about 10:00 h.

In Experiment 2, the rats were randomly divided into five groups (n = 7 each) that were fed one of the following diets supplemented with l-Met at 9 g/kg of diet: an AIN-76-based control diet (C), a B₆-deficient diet (D), and three B₆-deficient diets supplemented with graded levels of PC (3.15, 6.3, and 12.6 g of PC/kg of diet, designated respectively as PC3D, PC6D, and PC12D). The compositions of the diets are shown in Table 1. This experiment was conducted under the same protocol as Experiment 1.

In both experiments, after 35 days of feeding, the rats were anesthetized with diethyl ether, and blood was collected via the abdominal aorta with a heparinized syringe. Livers were immediately excised, rinsed in ice-cold 0.25 M sucrose, blotted on paper towels, and immediately frozen in liquid nitrogen. Plasma was obtained from the heparinized whole blood by centrifugation at 2000 × g (4 °C) for 15 min. The plasma and liver were stored at −20 °C until analyzed.

Materials

Vitamin-free casein, gelatinized starch, cellulose powder, and AIN-76 mineral mixture were purchased from Oriental Yeast (Tokyo, Japan). Choline bitartrate and l-Met were purchased from Wako Pure Chemical Industries (Osaka, Japan). Retinol acetate was purchased from Sigma-Aldrich Japan (Tokyo, Japan). Soybean phosphatide extract (PC 95%, Avanti Polar Lipids, Alabama, USA) was purchased from Funakoshi (Tokyo, Japan). S-Adenosylhomocysteine (SAH), S-adenosylmethionine (SAM), and L-α-phosphatidylcholine from egg yolk (Type XI-E) and L-α-phosphatidylethanolamine from egg yolk (Type III) were purchased from Sigma-Aldrich (St. Louis, MO, USA). N-Acetyl-l-cysteine, 4-fluoro-7-sulfobenzofurazan ammonium salt, and 1-heptanesulfonic acid sodium salt were purchased from Wako Pure Chemical Industries (Osaka, Japan). Methanol and acetonitrile (HPLC grade) were from Kanto Chemical (Tokyo, Japan). The other reagents were purchased from Nacalai Tesque (Kyoto, Japan), unless otherwise indicated.

Measurement of lipid concentrations in plasma and liver

Liver total lipids were extracted by the method of Folch et al.⁷⁾ The concentrations of total lipids, total cholesterol (TC), and triglyceride (TG) in the extract were measured by a conventional gravimetric method, the Zak Henly method,⁸⁾ and the Triglyceride G-test Wako (Wako Pure Chemical Industries, Osaka, Japan), respectively.

The concentrations of PC and phosphatidylethanolamine (PE) in liver total lipid extract were measured by isocratic HPLC with ultraviolet detection (210 nm).⁹⁾ The analytical conditions were as follows: column, TSKgel Silica-60 (4.6 mm I.D. × 25 cm); mobile phase, acetonitrile/methanol/85% phosphoric acid = 180/4/3; flow rate, 1.0 mL/min; standard substances, L-α-phosphatidylcholine from egg yolk (Type XI-E); and L-α-phosphatidylethanolamine from egg yolk (Type III).

Concentrations of plasma TC, high-density lipoprotein cholesterol (HDL-C), TG, and phospholipid (PL) were determined by enzymatic methods, using the Cholesterol E-test Wako, HDL cholesterol E-test Wako, Triglyceride G-test Wako, and Phospholipid C-test Wako (Wako Pure Chemical Industries, Osaka, Japan), respectively. The difference between TC concentration and HDL-C concentration was regarded as the very low-density lipoprotein (VLDL) + low-density lipoprotein (LDL)-cholesterol ((VLDL + LDL)-C) concentration.

Measurement of methionine metabolites in plasma and liver

Plasma and liver total Hcy were measured by the method of Yamaguchi et al.¹⁰⁾ which uses an isocratic HPLC system equipped with a fluorescence detector (Ex. 380 nm; Em. 525 nm). Liver SAM and SAH contents were measured by the method of She et al.¹¹⁾ using isocratic HPLC with ultraviolet detection (254 nm).

Analysis of vitamin B₆ in plasma and liver

PLP levels in liver and plasma were determined by isocratic HPLC with fluorescence detection (Ex. 320 nm; Em. 420 nm), using the method of Tsuge.¹²⁾

Statistical analysis

Data are expressed as means ± standard error (SE). Statistical analyses were performed using Excel Toukei 2006 for Windows (Social Survey Research Information, Tokyo, Japan). Differences between means were considered to be significant at p < 0.05, according to one-way ANOVA followed by Tukey’s multiple comparison test.

Results

Growth parameters (Experiment 1)

Final body weight, total food intake, and liver weight did not differ among the five experimental groups (Table 2).

Table 2. Body weight gain, total food intake, liver weight, and liver and plasma lipid concentrations of rats fed the experimental diets (Exp. 1).

	C	D	PGA10D	Ch4D	PC6D
Initial body weight (g)	119 ± 2	120 ± 2	120 ± 2	121 ± 3	121 ± 3
Final body weight (g)	182 ± 7	173 ± 4	176 ± 6	175 ± 7	172 ± 4
Body weight gain (g/35 d)	62.8 ± 5.8	52.9 ± 3.8	55.9 ± 4.9	54.3 ± 4.5	51.5 ± 2.4
Total food intake (g/35 d)	330 ± 8	329 ± 8	323 ± 12	321 ± 11	328 ± 7
Liver weight (g/100 g B.W.)	5.61 ± 0.23	5.92 ± 0.27	6.04 ± 0.23	6.2 ± 0.2	5.96 ± 0.28
Liver
Total lipids (mg/g)	84.0 ± 5.0	106 ± 8	113 ± 14	93.4 ± 4.4	86.0 ± 3.7
Triglyceride (mg/g)	11.1 ± 1.5^b	26.4 ± 3.8^a	33.3 ± 6.2^a	18.4 ± 2.2^ab	14.5 ± 2.0^ab
Total cholesterol (mg/g)	3.89 ± 0.23^b	6.68 ± 0.77^a	6.39 ± 0.61^a	5.85 ± 0.45^ab	4.98 ± 0.33^ab
Plasma
Total cholesterol (mg/dL)	79.8 ± 2.1^a	59.4 ± 4.5^b	66.9 ± 3.2^ab	68.5 ± 4.4^ab	77.2 ± 4.0^a
HDL cholesterol (mg/dL)	53.4 ± 1.3	49.3 ± 3.1	53.4 ± 3	49.7 ± 2.7	49.9 ± 1.9
LDL+VLDL cholesterol (mg/dL)	26.3 ± 3.0^a	10.1 ± 1.7^b	13.5 ± 2.2^b	18.8 ± 2.4^ab	27.3 ± 3.4^a

Notes:

Each value is mean ± SE (n = 7).

Means in a row not sharing the same superscript are significantly different at p < 0.05 as determined by one-way ANOVA followed by Tukey’s multiple comparison test.

Liver and plasma lipid profiles (Experiment 1)

Contents of TG and TC in the livers of the D group were significantly greater than those of the C group (Table 2). Contents of liver total lipids were also greater in the D group than in the C group, although the difference was not statistically significant. In contrast, concentrations of plasma TC and (LDL + VLDL)-C in the D group were significantly lower than those in the C group. Liver lipids in the Ch4D and PC6D groups were decreased relative to the D group, although this was not statistically significant. Concentrations of TC and (LDL + VLDL)-C in the plasma of the PC6D group were significantly increased relative to the D group, and those in the Ch4D group were also increased compared with the D group, although not by an amount that was statistically significant. Lipid profiles in liver and plasma were not remarkably different between D and PGA10D groups.

Methionine metabolites and vitamin B₆ status in liver and plasma (Experiment 1)

Liver and plasma Hcy were significantly increased in the B₆-deficient (D, PGA10D, Ch4D, and PC6D) groups compared with the C group (Table 3). Contents of Hcy in the livers of the PGA10D and PC6D groups were about half of that in the D group, although statistical significance was not observed. Contents of liver SAH in the D and Ch4D groups were significantly greater than in the C group. However, those in the PGA10D and PC6D groups were not significantly increased compared with the C group. Liver SAM/SAH ratios in the D, PGA10D, and Ch4D groups were significantly lower than that in the C group. Liver and plasma PLP were significantly lower in the four B₆-deficient groups compared with the C group (Table 3).

Table 3. Status of methionine metabolites and PLP in liver and plasma of rats fed with the experimental diets (Exp. 1).

	C	D	PGA10D	Ch4D	PC6D
Liver
Hcy (nmol/g)	4.87 ± 0.38^b	109 ± 55^a	30.5 ± 9.6^a	119 ± 60^a	51.8 ± 29.8^a
SAM (nmol/g)	41.1 ± 6.6	118 ± 55	61.1 ± 7.2	172 ± 82	71.1 ± 16.6
SAH (nmol/g)	33.3 ± 5.3^b	274 ± 131^a	138 ± 31^ab	405 ± 172^a	174 ± 75^ab
SAM/SAH ratio	1.27 ± 0.12^a	0.63 ± 0.13^b	0.54 ± 0.10^b	0.48 ± 0.13^b	0.70 ± 0.22^ab
PLP (nmol/g)	5.59 ± 0.82^a	2.31 ± 0.20^b	2.00 ± 0.23^b	1.75 ± 0.29^b	1.85 ± 0.24^b
Plasma
Hcy (μM)	11.0 ± 1.1b	322 ± 110^a	257 ± 45^a	382 ± 100^a	236 ± 97^a
PLP (nM)	63.7 ± 16.8^a	3.15 ± 0.42^b	3.25 ± 0.41^b	3.61 ± 0.74^b	2.96 ± 0.26^b

Notes:

Each value is mean ± SE (n = 7).

Means in a row not sharing the same superscript are significantly different at p < 0.05 as determined by one-way ANOVA followed by Tukey’s multiple comparison test.

Growth parameters (Experiment 2)

Final body weight and total food intake did not differ among the five experimental groups (Table 4). The liver weight in the D group was greater than those in the other four groups, although the differences were not statistically significant.

Table 4. Body weight gain, total food intake, liver weight, and liver and plasma lipid concentrations of rats fed with the experimental diets (Exp. 2).

	C	D	PC3D	PC6D	PC12D
Initial body weight (g)	123 ± 2	123 ± 2	124 ± 2	124 ± 2	124 ± 2
Final body weight (g)	170 ± 6	169 ± 4	166 ± 4	164 ± 4	161 ± 4
Body weight gain (g/35 days)	47.2 ± 5.0	45.6 ± 3.4	42.0 ± 4.2	40.7 ± 4.1	38.2 ± 2.8
Total food intake (g/35 days)	316 ± 7	315 ± 8	314 ± 6	314 ± 6	317 ± 7
Liver weight (g/100 g B.W.)	5.62 ± 0.44	6.41 ± 0.49	5.68 ± 0.27	5.85 ± 0.45	5.51 ± 0.34
Liver
Total lipids (mg/g)	46.1 ± 2.0	124 ± 31	40.6 ± 2.9	45.6 ± 9.8	37.6 ± 4.1
Triglyceride (mg/g)	7.99 ± 1.39^b	19.1 ± 2.8^a	11.2 ± 1.2^ab	10.7 ± 2.5^b	9.56 ± 1.77^b
Total cholesterol (mg/g)	5.04 ± 0.31	13.4 ± 2.8	5.88 ± 0.28	6.76 ± 1.30	5.36 ± 0.24
Phosphatidylcholine; PC (mg/g)	23.7 ± 3.3	27.0 ± 3.2	24.5 ± 2.5	26.2 ± 2.6	26.2 ± 4.3
Phosphatidylethanolamine; PE (mg/g)	11.5 ± 0.7^b	15.8 ± 1.0^a	13.3 ± 0.7^ab	13.7 ± 0.4^ab	13.2 ± 0.7^ab
PC/PE	2.03 ± 0.20	1.72 ± 0.20	1.84 ± 0.15	1.93 ± 0.21	1.94 ± 0.22
Plasma
Total cholesterol (mg/dL)	59.0 ± 3.3	47.4 ± 6.4	52.8 ± 4.0	60.0 ± 3.0	57.2 ± 2.5
HDL cholesterol (mg/dL)	34.1 ± 2.8	34.9 ± 3.5	32.9 ± 2.1	36.2 ± 3.2	34.5 ± 1.9
LDL+VLDL cholesterol (mg/dL)	24.9 ± 2.0^a	12.5 ± 3.5^b	20.0 ± 2.6^ab	23.8 ± 1.9^a	22.7 ± 1.2^a
Triglyceride (mg/dL)	77.5 ± 25.8	29.0 ± 5.5	44.4 ± 5.7	40.7 ± 7.6	68.2 ± 14.6
Phospholipid (mg/dL)	122 ± 8^a	96 ± 9^b	107 ± 6^ab	118 ± 4^ab	118 ± 4^ab

Notes:

Each value is mean ± SE (n = 7).

Means in a row not sharing the same superscript are significantly different at p < 0.05 as determined by one-way ANOVA followed by Tukey’s multiple comparison test.

Liver and plasma lipid profiles (Experiment 2)

The content of TG in the liver was significantly greater in the D group than in the C group, and those in the PC6D and PC12D groups were significantly lower than that in the D group (Table 4). Contents of total lipids and TC in the D group were double those in the other four groups, although the differences were not statistically significant. The content of PC in liver showed no statistically significant difference, but the content of PE was significantly greater in the D group than in the C group. Liver PC/PE ratio had no statistically significant difference.

By contrast, concentrations of TC, (LDL + VLDL)-C, and PL (PC) in the plasma of the D group were significantly lower than those of the C group. Although B₆ deficiency induced hypocholesterolemia in the D group, these concentrations were gradually increased as the dietary PC supplementation level increased. The concentration of HDL-C in plasma showed no statistically significant differences among the five experimental groups.

Methionine metabolites in liver and plasma (Experiment 2)

Levels of Hcy in liver and plasma, and SAH in liver were significantly greater in the D group than in the C group, but lower in the PC3D, PC6D, and PC12D groups than in the D group, although in the latter cases the differences were not statistically significant (Table 5). Liver SAM/SAH ratios were lower in the B₆-deficient groups (D, PC3D, PC6D, and PC12D) than in the C group, although again the differences were not statistically significant.

Table 5. Status of methionine metabolites in liver and plasma of rats fed with the experimental diets (Exp. 2).

	C	D	PC3D	PC6D	PC12D
Liver
Hcy (nmol/g)	8.19 ± 0.80^b	360 ± 146^a	35.1 ± 10.1^ab	85.2 ± 46.6^ab	18.5 ± 5.1^ab
SAM (nmol/g)	56.3 ± 4.0	216 ± 73	63.1 ± 7.6	105 ± 47	57.7 ± 10.3
SAH (nmol/g)	62.5 ± 10.0^b	469 ± 144^a	133 ± 28^ab	205 ± 81^ab	121 ± 31^ab
SAM/SAH ratio	1.04 ± 0.18	0.55 ± 0.10	0.67 ± 0.23	0.69 ± 0.21	0.59 ± 0.12
Plasma
Hcy (μM)	51.7 ± 18.5^b	519 ± 111^a	227 ± 57^ab	228 ± 92^ab	153 ± 47^ab

Notes:

Each value is mean ± SE (n = 7).

Means in a row not sharing the same superscript are significantly different at p < 0.05 as determined by one-way ANOVA followed by Tukey’s multiple comparison test.

Discussion

This study was undertaken to understand the mechanism of liver lipid accumulation in rats fed a l-Met-supplemented B₆-deficient diet. In rats fed a B₆-deficient diet, total lipids, TG, and TC in the liver increased relative to those in the control rats (Tables 2 and 4). On the other hand, rats fed a choline or PC-supplemented B₆-deficient diet showed a smaller accumulation of these lipids in liver compared with the B₆-deficient rats (Tables 2 and 4). There was a difference in total lipid content in “C” of Exps. 1 and 2 (Tables 2 and 4), in the former being 84.0 ± 5.0 mg/g of liver and in the latter 46.1 ± 2.0 mg/g of liver. We adopted pair-feeding to minimize dispersion in dietary intake because B₆ deficiency caused decreased dietary intake. Although these two experiments were conducted on the same experimental protocol, food intake in the D group of Exp. 2 started to reduce at earlier stage of feeding period. This meant that B₆ deficiency begun earlier in Exp. 2 than in Exp. 1. So, dietary intake in the C group of Exp. 2 was low as compared to Exp. 1. We are thinking that this might be the reason why total lipid in liver of the C group was lower in Exp. 2 compared to Exp. 1. As for plasma lipids, rats fed a B₆-deficient diet had decreased concentrations of TC, (LDL + VLDL)-C, TG, and PC in plasma compared to the control rats (Tables 2 and 4). Although B₆ deficiency induced hypocholesterolemia, these reductions were ameliorated when the B₆-deficient diet was supplemented with choline or PC (Tables 2 and 4). The effect of PC supplementation was much greater than that of choline. The effective supplemental level of PC in the diet was estimated to be 6.3 g/kg diet or more (Table 4).

To determine the relationship between liver lipid deposition and disturbed Met metabolism in B₆ deficiency, we analyzed some of the Met metabolites. In the B₆-deficient rats, Hcy in the liver and plasma and SAH in the liver were significantly increased, and in the liver, the SAM/SAH ratio, the index parameter of methylation, decreased significantly (Tables 3 and 5) in agreement with the results of previous studies.^3,5,13–16) In one of the Hcy metabolic pathways, the trans-sulfuration pathway, Hcy is metabolized to cystathionine by CBS, which requires PLP as a coenzyme (Fig. 1). In B₆ deficiency, CBS activity was reduced because of decreased PLP; thus, Hcy metabolism to cystathionine was impaired.¹⁷⁾ As a result, Hcy accumulated in the liver, and this led to increased Hcy in the plasma. Moreover, accumulated Hcy caused an increase in liver SAH, because the equilibrium of Hcy and SAH was toward SAH.¹⁶⁾ We sometimes get annoyed by individual variability in the level of methionine metabolites. Content of SAH in liver was usually 30–60 nmol/g of liver; however, individual with high liver SAH in the C group of Exp. 2 increased mean of SAH with concomitant increase in SE compared to Exp. 1. This was also true for the D groups (Tables 3 and 5). Standard errors in these groups were larger in both Exp. 1 and Exp. 2. A little higher SAH in the D group of Exp. 2 might be ascribed to the progress of B₆ deficiency, which was severe in Exp. 2 as noted above in total lipid section.

Phosphatidylcholine can be synthesized in the liver via two pathways: the cytidine diphosphate–choline (CDP–choline) pathway, and the PE methylation pathway by phosphatidylethanolamine N-methyltransferase (EC 2.1.1.17; PEMT), both of which are linked to Met metabolism (Fig. 1).¹⁸⁾ The CDP–choline pathway requires choline, which is obtained either from the diet or from the metabolism of PC.¹⁹⁾ The CDP–choline pathway and Met metabolism are linked via choline, because choline is a precursor of betaine, which is an alternative methyl donor in Hcy methylation.²⁰⁾ The PE methylation pathway requires SAM as the methyl donor. This reaction is not only influenced by the availability of SAM, but is also inhibited by SAH.²¹⁾ Thus the SAM/SAH ratio regulates the PE methylation pathway.¹⁴⁾ In the present study, SAH increased and the SAM/SAH ratio decreased significantly in the livers of B₆-deficient rats (Tables 3 and 5). Moreover, PE in the liver increased significantly in B₆-deficient rats compared with control rats. It was suggested that accumulated SAH caused an increase in liver PE through an inhibition of PEMT. Moreover, She et al.^13,14) reported that the content of microsomal PC in the liver decreased with a concomitant increase of PE in B₆-deficient rats compared with the pair-fed control rats. In a similar fashion, Loo and Smith²²⁾ reported B₆ deficiency altered the methylation of PE and decreased the levels of microsomal PC. In the present study, we could not analyze PC and PE in microsome, but it was assumed that the PC supply via the PE methylation pathway became insufficient because of reduced PEMT activity in B₆ deficiency.

It is known that VLDL exports lipids from the liver to peripheral tissues²³⁾ and that PC is an essential structural component of VLDL.^24,25) Endogenously synthesized lipids, particularly TG and cholesteryl ester, are incorporated into VLDL for transport via the blood circulation. Moreover, it was reported that the active synthesis of PC derived from PE was essential for the hepatic synthesis and secretion of VLDL.^24–27) However, in B₆-deficient rats, plasma TG, TC, and PL (as PC) concentrations as well as plasma (LDL + VLDL)-C concentration decreased significantly (Tables 2 and 4). Consequently, it was speculated that the accumulation of liver lipids induced by B₆ deficiency was accompanied by a reduction of VLDL secretion by the liver.

On the other hand, supplementation with choline or PC ameliorated the liver lipid accumulation, and returned the (LDL + VLDL)-C concentration in plasma toward normal (Tables 2 and 4). In addition, supplemental PC tended to recover the plasma TG and PL (as PC) concentrations as well (Table 4). These results lead us to speculate that the deposition of lipids in the liver induced by B₆ deficiency was effectively prevented by choline or PC supplementation, presumably because of recovery of VLDL secretion into the circulation by PC, which was impaired by B₆ deficiency and recovered by either choline or PC supplementation.

According to several reports, rats fed a 70% casein diet without B₆ showed marked accumulation of liver lipid.^4,28,29) In addition, serum PC concentration was significantly lower in B₆-deficient rats than in pair-fed controls.²⁹⁾ In an analogous fashion, Kaimoto et al.³⁰⁾ reported that about 50% of rats fed a diet supplemented with branched-chain amino acids during B₆ deficiency showed lipid deposition in the liver and tended to have decreased serum TG levels. Moreover, they found that rats with liver lipid deposition also had decreased TG and TC levels in VLDL. However, they could not determine the mechanism whereby lipid deposition occurred in B₆ deficiency.

In our present experiments, supplementation with PGA had little effect on the alleviation of lipid metabolism in the liver and plasma in comparison with that of PC (Table 2), although it was a little more effective than PC in alleviating the disturbance of Met metabolism (Table 3). This might be because PGA had little effect on PC biosynthesis in the liver. Even in the control group, a normal amount of PGA was present in the diet. However, actual MS activity was decreased, as 5-MTHF is reduced by B₆ deficiency.³⁾ Therefore, supplemental PGA had a small effect in alleviating abnormal Met metabolism,³⁾ but was not effective for maintaining PC biosynthesis. This is why PGA supplementation was less effective than PC supplementation.

In another remethylation pathway of Hcy, betaine is also an important substrate. Betaine is produced by oxidation of choline, which is derived from dietary PC.⁶⁾ Therefore, it was conceivable that supplemental PC produced betaine via choline and contributed to remethylation of Hcy by BHMT. However, addition of choline to the B₆-deficient diet had only a small effect on the disturbance of Met metabolism relative to PC (Table 3). Absorbed choline would be delivered to several tissues including the brain and used as a neurotransmitter.¹⁹⁾ So a smaller amount of choline reaches the liver as compared to PC. This might be the reason why supplemental choline was less effective than PC.

In conclusion, this study indicated that B₆ deficiency caused liver lipid deposition with a reduction of plasma lipids. Judging from the liver and plasma lipid profiles, our results could be explained by decreased secretion of VLDL as a consequence of reduced synthesis of PC via the PE methylation pathway owing to the disturbance of Met metabolism produced by B₆ deficiency. Choline or PC supplementation effectively stimulated the synthesis of PC from PE in the liver and thereby normalized lipid transport from the liver. In a future study, we intend to evaluate how VLDL secretion is affected by B₆ deficiency and dietary supplementation.

Authors’ contributions

EK, KY, and TH designed the study. EK drafted the manuscript. TN reviewed the manuscript. TY took part in a part of analysis. All authors reviewed and approved the final manuscript.

Disclosure statement

The authors disclose no potential conflicts of interest.