- Email: [email protected]
Plasma vitamin D-binding globulin in vitamin D deficiency, pregnancy and chronic liver disease
359
Clinica Chimica Acta, 87 (1978) @ Elsevier/North-Holland
359-365 Biomedical Press
CCA 9560
PLASMA VITAMIN D-BINDING GLOBULIN IN VITAMIN PREGNANCY AND CHRONIC LIVER DISEASE
J.M. BARRAGRY J.D. MAXWELL
*, D. CORLESS, and S. SWITALA
JENNIFER
AUTON,
D DEFICIENCY,
N.D. CARTER,
R.G.
LONG,
Unit of Metabolism and Endocrinology, The London Hospital Medical College; Department of Geriatric Medicine, The Eastern, Hackney and St. Bartholomew’s Hospitals; Department of Medicine, The Royal Free Hospital; Departments of Child Health and Medicine, St. George’s Hospital Medical School, London (U.K.) (Received
February
25th,
1978)
Summary Plasma concentrations of vitamin D-binding globulin were measured by radial immunodiffusion in healthy subjects, pregnanky, and during oestrogen therapy. Subjects with disorders of vitamin D metabolism (dietary deficiency, malabsorption, anticonvulsant therapy, chronic liver disease) were also studied. Neither sex nor age influenced the plasma vitamin D-binding globulin concentration in healthy subjects, but there was a significant increase in concentration during pregnancy and oestrogen therapy. Elevated levels were found in vitamin D deficient elderly but not younger subjects, while levels in subjects with chronic liver disease were significantly reduced. Normal levels of vitamin D-binding globulin were present in hypervitaminosis D and no vitamin D-binding globulin was detected in human milk. No correlation was observed between plasma 25-hydroxycholecalciferol levels and plasma vitamin D-binding globulin concentrations.
Introduction Vitamin D and its metabolites circulate in the plasma bound to a specific carrier protein, vitamin D-binding globulin (VDBG, Gc protein) [1,2]. This is an cr-globulin which. is synthesised in the liver and has a molecular weight of approximately 59 000 [ 31. VDBG has highest affinity for 25-hydroxycholecalciferol (25(OH)D,) with a maximum binding capacity of one molecule of
* Correspondence should be addressed to: Dr. J.M. Barragry, Department of Metabolism and Endocrinology, The London Hospital. Whitechapel, London El 1BB. U.K.
360
25(OH)D3 per molecule of protein [4]. However the combined plasma content of vitamin D metabolites occupies less than 5% of the available high-affinity binding sites on VDBG [3]. It is of interest to establish what factors are concerned in the regulation of the plasma concentration of VDBG and whether an alteration in plasma VDBG concentration might be associated with an increased susceptibility to vitamin D disorders. We report here the plasma levels of VDBG in a variety of physiological states including pregnancy, and also in vitamin D deficient states and chronic liver disease. The relationship between plasma VDBG and 25(OH)D3 levels has been examined. Materials and methods Su bjec ts
Twenty-two healthy men and women ranging in age from 20 to 67 years were studied. No subject was taking any drug and all had an estimated dietary vitamin D content of at least 100 I.U. (6.25 nmol) daily and normal exposure to sunlight. Venous blood samples were also drawn from twenty-one elderly female subjects (mean age 81.5 years) who were known to have plasma 25(OH)D3 levels within the control group range; from a further sixteen elderly females in a longstay geriatric ward (mean age 83 years) who had a dietary vitamin D content of 40 I.U. (2.5 nmol) daily and little exposure to sunlight; from six subjects from the Indian subcontinent whose diets were grossly vitamin D deficient (<20 I.U./ day); from five Caucasian subjects with an untreated malabsorption syndrome; from seventeen epileptic subjects regularly taking at least two anticonvulsant drugs for a minimum of two years and from forty-five subjects with chronic liver disease viz. primary biliary cirrhosis (n = 16), alcoholic cirrhosis (n = 14) and active chronic hepatitis (n = 15). Thirty pregnant Caucasian subjects were studied. All had a normal diet and normal exposure to sunlight. Blood was also drawn from seven pregnant Asian subjects (daily vitamin D intake <40 I.U.) and from nine women receiving a 50 pg oestrogen preparation. Plasma VDBG levels were measured in two subjects with vitamin D intoxication. The VDBG content of human milk was also estimated. Methods
Plasma 25(OH)D3 assays were performed by the competitive protein binding method of Edelstein, Charman, Lawson and Kodicek (1974) [5] using rat kidney binding protein [6] by a modification not requiring column chromatographic separation [ 71 using P-lipoprotein as solubilizing agent. In subjects with chronic liver disease plasma 25(OH)D3 levels were assayed using the method of Skinner and Wills (1977) [8]. Plasma VDBG levels were measured using M-Partigen immunodiffusion plates (Pharma-Behring Diagnostics, Hoechst U.K. Ltd.) containing specific anti-Gc serum. Five microlitre aliquots of sample were applied to the plate and diffusion allowed to proceed overnight at 37°C. Precipitin ring diameter was measured using calipers and the square of the radius was calculated. This value was
361
converted to mg VDBG by comparison with the value for the ring of a known standard applied in duplicate to each plate. Results Control group In the control group the mean plasma 25(OH)D3 level was 85.8 2 10.6 nmol/l. Plasma VDBG concentrations ranged from 18.7 to 34.5 mg/dl (mean + SE. 27.6 + 0.95 mg/dl). Within this group the mean value 2 S.E. among males -. was 26.2 of 1.5 mg/dl and among females 28.7 & 1.2 mg/dl (P > 0.0s). No relationship was observed between plasma 25(OH)D3 and plasma VDBG concentrations (r = 0.19). Vitamin D deficient states In subjects on anticonvulsant therapy (Fig. 1) the mean plasma 25(OH)D3 level i S.E. of 51.0 + 13.5 nmol/l (range 6.3 to 195 nmol/l) was significantly lower than that of the control group (P < 0.01); the mean plasma VDBG level 2 SE., 27.8 ? 1.7 mg/dl, (range 15.8 to 39.5 mg/dl) was similar to the control group (P > 0.05). In vitamin D deficient subjects (inadequate diet, malabsorption syndrome) whose mean plasma 25(OH)D3 level + S.E. was 5.6 2 2.4 nmol/l (range 0 to 15 nmol/l (P < O.Ol), the mean plasma VDBG level + S.E. was 25.5 + 1.5 mg/dl (range 16.6 to 31.8 mg/dl (P > 0.05) Plasma 25(OH)D3 levels in the group of elderly subjects known to have a normal vitamin D status ranged from 20 to 157 nmol/l and the mean value f S.E., 61.5 ? 7.9 nmol/l, was considerably lower than the control group mean
Plasma
VDBG
m/d1 40-
35 -
30 -
.. ... -
-1-
25 -
..
.
...
... ..
.
‘.
-
-7.* .. . .
.‘:
20 -
!. .. ..
15 -
10 -
_ 5-
PCO.01
. .
O-
NS
.
PcO.05 PCO.01
Fig. 1. Plasma VDBG concentrations low plasma 25(OH)D3 levels.
.
(mg/dl)
in control subjects (young and elderly) and in subjects with
362
(P < 0.01). However there was no significant difference in the plasma VDBG in this group (mean + S.E. 26.0 it 0.8 mg/dl, range 17.4 to 33.0 mg/dl) (P > 0.05). In the group of elderly vitamin D deficient subjects (mean plasma 25(OH)D3 + S.E. 5.9 2 1.5 nmol/l, range 0.0 to 19.7 nmol/l, (P < 0.01)) plasma VDBG level ranged from 25.6 to 39.2 mg/dl and the mean value k S.E. 30.7 t 0.9 mg/dl was significantly greater than that of the control group (P < 0.05) and that of the vitamin D replete elderly group (P < 0.01). In subjects with chronic liver disease (mean plasma 25(OH)D3 + S.E. 32.7 k 4.3 nmol/l, range 5.0 to 81.8 nmol/l), the mean plasma VDBG concentration 1 S.E. 21.6 i 0.9 mg/dl (range 8.1 to 34.3 mg/dl) was lower than that of the control group (P < 0.01). Within the group the mean plasma VDBG level k SE. among subjects with alcoholic liver disease (21.9 k 1.4 mg/dl), active chronic hepatitis 22.8 f 1.9 mg/dl) and primary biliary cirrhosis (20.3 + 1.2 mg/dl) were similar (P > 0.05). In a further four subjects with extra hepatic biliary obstruction the mean plasma VDBG concentration -CS,E_, 25.3 k 5.6 mg/dl, was not significantly lower than the control group mean (P > 0.05). kvels
Pregnancy
and oestrogen
therapy
Plasma 25(OH)D3 levels decreased during pregnancy. However the mean value + SE. in the first trimester (99,2 * 29.0 nmol/l), second trimester (80.2 t 11.4 nmol/l) and third trimester (59.0 k 8.3 nmol) did not differ significantly from the control group mean (P > 0.05). Plasma VDBG levels increased during Plasma
VDBG
mg/dl 60-
. .
55 -
50
l
45
: .
l
-
. . .
40 -
35 -
30
-
.. m
I-. . i:;.:
. . 7
JL .
. l
T+-
8
25 -
20 -
? . .
.
15 -
lo-
Fig. 2. Plasma VDBG concentrations (mg/dl) in control and pregnant subjects and in subjects on OeStKOgen therapy.
363
pregnancy (Fig. 2), with a first trimester mean value + S.E. of 31.8 & 2.8 mg/dl, second trimester 36.1 & 2.5 mg/dl and a third trimester value of 41.0 + 2.7 mg/ dl. The latter two values were significantly greater than the control group mean (P < 0.05 and P < 0.01). In seven pregnant Asian subjects the plasma 25(OH)D3 values were below the respective trimester ranges of the pregnant Caucasian subjects while their plasma VDBG levels were within the Caucasian ranges. Plasma VDBF levels in subjects on longterm oestrogen therapy ranged from 27.0 to 45.8 mg/dl and the mean value ? S.E. 37.0 % 2.2 mg/dl was greater than that of the control group (P< 0.01). Other subjects
studied
In two subjects with hypervitaminosis D (plasma 25(OH)D3 levels 350 nmol/l and 450 nmol/l) the plasma VDBG levels, 28.64 mg/dl and 34.0 mg/dl were within the control group range. No VDBG was detected in a sample of human milk. No correlation was noted between plasma 25(OH)D3 concentration and plasma VDBG concentration in any of the groups studied. Discussion The mean plasma VDBG concentration in healthy subjects in this series is lower than that observed by Haddad and Walgate [9] using radioimmunoassay and by other workers using radial immunodiffusion [ 101. The reason for this difference is not apparent. In the present study there was no evidence of a significant influence of sex on plasma VDBG concentration. Age likewise seems not to be a determinant of VDBG level: the mean plasma VDBG level in the control group (mean age 43.5 years) was almost identical to that of a group of elderly subjects of mean age 81.5 years. Moreover Haddad and Walgate (1976) [9] have shown that the ,VDBG concentration in cord blood of full-term infants is similar to that of adults. The relationship between vitamin D status and plasma VDBG concentration is more obscure. The mean plasma VDBG level in the vitamin D deficient elderly subjects was significantly greater than the control group and that of the age-matched group which was vitamin D replete. Furthermore, while the vitamin D replete elderly group had a lower mean plasma 25(OH)D3 level than that of the control group, both groups had levels within the normal range and the plasma VDBG levels were very similar. It is reasonable to infer therefore that it is vitamin D deficiency, rather than age, which is associated with the increased VDBG concentration in the plasma of elderly subjects and this is in agreement with the observation that a vitamin D deficient diet appears to induce increased synthesis of VDBG in rats [ 111. However, the plasma VDBG levels in younger Asian subjects on vitamin D depleted diets and in subjects with the malabsorption syndrome were within the normal range and the plasma VDBG levels in epileptic subjects on chronic anticonvulsant therapy were normal despite low 25(OH)D3 concentrations. Moreover normal plasma VDBG levels were observed in subjects with hypervitaminosis D. Similar findings have been reported by other workers [ 91. The plasma level of VDBG increased during pregnancy and the second and
364
third trimester values were significantly greater than those of the control group. There was also a significant increase in VDBG levels in the plasma of subjects on oestrogen therapy. These findings are similar to those of Haddad and Walgate [9] and suggest that, as with other transport proteins, e.g. thyroxinebinding globulin, cortisol-binding globulin, etc. [12] VDBG synthesis is enhanced by hyperoestrogenaemia. However by analogy with these compounds an increase in plasma 25(OH)D3 during pregnancy might have been anticipated whereas the opposite trend was seen and this was particularly noticeable in Asian subjects. This latter finding may reflect an inadequate dietary vitamin D intake in the presence of a foetus entirely dependent on the placental transfer of maternal 25(OH)D3 [13] or perhaps an induction of hepatic microsomal enzymes during pregnancy [ 141. Subjects with parenchymal and cholestatic liver disease have low serum levels of 25(OH)D3 [15) possibly because of impaired hepatic 25-hydroxylation of vitamin D-3 [ 161, vitamin D deficiency [ 171 or impaired intestinal absorption of vitamin D [18]. The liver is believed to be the site of Gc protein (i.e. VDBG) synthesis [19] and low levels were observed in those subjects with chronic parenchymal liver disease although not in subjects with extrahepatic biliary obstruction. Reduced levels have also been reported in cirrhotic patients who are hypoproteinaemic [ 91. The vitamin D present in human milk is contained mainly in the aqueous phase as the sulphate [20]. This observation prompted an analysis of human milk for VDBG content and none was detected. One of the functions of VDBG is to transport the water-insoluble metabolites of vitamin D to the liver, kidney, gut, bone, etc. and it is appropriate that in an environment where vitamin D is already present in a water-soluble form no VDBG exists. No relationship was observed between serum 25(OH)D3 concentration and serum VDBG concentration either in healthy subjects or in subjects with a disturbance of either or both 25(OHjD3 and VDBG. VDBG is present in human plasma in considerable molar excess compared to the combined plasma concentration of the principal vitamin D metabolites. It seems probable that marked fluctuations in plasma 25(OH)D3 levels may occur without any alteration in VDBG concentration. Conversely minor changes in VDBG level may occur without a corresponding change in plasma 25(OH)D3 level. Nevertheless it is likely that gross reduction in VDBG levels (due to failure of synthesis or abnormal losses) would eventually lead to reduced circulating levels of 25(OH)D3. Thus the low plasma levels of VDBG seen in subjects with extensive liver disease might contribute to their vitamin D deficient state and gross vitamin D deficiency has been reported in the nephrotic syndrome in which there is loss of VDBG in the urine [ 211. Acknowle.dgements The authors wish to acknowledge the cooperation of Professor R.D. Cohen and Professor S. Sherlock. They also thank Dr. R.K. Skinner for the assay of plasma 25(OH)D3 in patients with liver disease.
365
References 1 Peterson, 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
P.A. (1971)
J. Biol. Chem.
246, 7748-7754
Edelstein, S., Lawson, D.E.M. and Kodicek. E. (1973) Biochem. J. 135, 417426 Haddad, J.G. and Walgate, J. (1976) J. Biol. Chem. 251, 48034809 Imwari, M., Kida. K. and Goodman, D.S. (1976) J. Clin. Invest. 58, 514-523 Edelstein, S., Charman. J.. Lawson, D.E.M. and Kodicek, E. (1974) Clin. Sci. Mol. Med. 46, 231-240 Haddad, J.G. and Chyu, K.J. (1971) J. Clin. Endocrinol. 33,992-995 Belsey, H., De Luca, H.F. and Potts, J.T. (1974) J. Clin. Endocrinol. 38, 1046-1051 Skinner, R.K. and MiUs, M.R. (1977) Clin. Chim. Acta 80. 543-554 Haddad, J.G. and W&ate. J. (1976) J. Clin. Invest. 58, 1217-1222 Schmidt-Gayk, H., Schmitt. W.. Grawunder, C.. Ritz, E.. Tschiipe, W.. Pietsch, V.. Andrassy, K. and Bouillon, R. (1977) Lancet ii, 105-108 Belsey, H., De Luca, H.F. and Potts, J.T. (1972) Excerpta Medica 256, p. 25 Westphal, U. (1971) Steroid-Protein Interactions, Springer-Verlag. Berlin, Heidelberg, New York Hillman, L.S. and Haddad. J.G. (1975) J. Pediatr. 86, 928-935 Davis, M., Simmons, C.J.. Dordoni, B., Maxwell, J.D. and Williams, R. (1973) J. Obstet. Gynecol. Br. Commw. 80.690-694 Long, R.G., Skinner, R.K.. Wills, M.R. and Sherlock, S. (1976) Lancet ii, 650-652 Wagonfeld. J.B., Nemchausky. B.A., Bolt, M., Bayer. J.L. and Rosenberg, I.H. (1975) Gastroenterology 69,878 Skinner, R.K.. Long, R.G., Sherlock, S. and Wills. M.R. (1977) Lancet i, 720-721 Compston, J.E. and Thompson, R.P.H. (1977) Lancet i. 721-724 Prunier, J.H., Beam, A.G. and Cleve, H. (1974) Proc. Sot. EXP. Biol. Med. 115.1005-1009 Lakdawala, D.R. and Widdowson, E.M. (1977) Lancet i, 167-168 Barragry, J.M.. France, M.W., Carter, N.D., Auton, J.A., Beer, M., Boucher. B.J. and Cohen, R.D. (1977) Lancet ii. 629432
Clinica Chimica Acta, 87 (1978) @ Elsevier/North-Holland
359-365 Biomedical Press
CCA 9560
PLASMA VITAMIN D-BINDING GLOBULIN IN VITAMIN PREGNANCY AND CHRONIC LIVER DISEASE
J.M. BARRAGRY J.D. MAXWELL
*, D. CORLESS, and S. SWITALA
JENNIFER
AUTON,
D DEFICIENCY,
N.D. CARTER,
R.G.
LONG,
Unit of Metabolism and Endocrinology, The London Hospital Medical College; Department of Geriatric Medicine, The Eastern, Hackney and St. Bartholomew’s Hospitals; Department of Medicine, The Royal Free Hospital; Departments of Child Health and Medicine, St. George’s Hospital Medical School, London (U.K.) (Received
February
25th,
1978)
Summary Plasma concentrations of vitamin D-binding globulin were measured by radial immunodiffusion in healthy subjects, pregnanky, and during oestrogen therapy. Subjects with disorders of vitamin D metabolism (dietary deficiency, malabsorption, anticonvulsant therapy, chronic liver disease) were also studied. Neither sex nor age influenced the plasma vitamin D-binding globulin concentration in healthy subjects, but there was a significant increase in concentration during pregnancy and oestrogen therapy. Elevated levels were found in vitamin D deficient elderly but not younger subjects, while levels in subjects with chronic liver disease were significantly reduced. Normal levels of vitamin D-binding globulin were present in hypervitaminosis D and no vitamin D-binding globulin was detected in human milk. No correlation was observed between plasma 25-hydroxycholecalciferol levels and plasma vitamin D-binding globulin concentrations.
Introduction Vitamin D and its metabolites circulate in the plasma bound to a specific carrier protein, vitamin D-binding globulin (VDBG, Gc protein) [1,2]. This is an cr-globulin which. is synthesised in the liver and has a molecular weight of approximately 59 000 [ 31. VDBG has highest affinity for 25-hydroxycholecalciferol (25(OH)D,) with a maximum binding capacity of one molecule of
* Correspondence should be addressed to: Dr. J.M. Barragry, Department of Metabolism and Endocrinology, The London Hospital. Whitechapel, London El 1BB. U.K.
360
25(OH)D3 per molecule of protein [4]. However the combined plasma content of vitamin D metabolites occupies less than 5% of the available high-affinity binding sites on VDBG [3]. It is of interest to establish what factors are concerned in the regulation of the plasma concentration of VDBG and whether an alteration in plasma VDBG concentration might be associated with an increased susceptibility to vitamin D disorders. We report here the plasma levels of VDBG in a variety of physiological states including pregnancy, and also in vitamin D deficient states and chronic liver disease. The relationship between plasma VDBG and 25(OH)D3 levels has been examined. Materials and methods Su bjec ts
Twenty-two healthy men and women ranging in age from 20 to 67 years were studied. No subject was taking any drug and all had an estimated dietary vitamin D content of at least 100 I.U. (6.25 nmol) daily and normal exposure to sunlight. Venous blood samples were also drawn from twenty-one elderly female subjects (mean age 81.5 years) who were known to have plasma 25(OH)D3 levels within the control group range; from a further sixteen elderly females in a longstay geriatric ward (mean age 83 years) who had a dietary vitamin D content of 40 I.U. (2.5 nmol) daily and little exposure to sunlight; from six subjects from the Indian subcontinent whose diets were grossly vitamin D deficient (<20 I.U./ day); from five Caucasian subjects with an untreated malabsorption syndrome; from seventeen epileptic subjects regularly taking at least two anticonvulsant drugs for a minimum of two years and from forty-five subjects with chronic liver disease viz. primary biliary cirrhosis (n = 16), alcoholic cirrhosis (n = 14) and active chronic hepatitis (n = 15). Thirty pregnant Caucasian subjects were studied. All had a normal diet and normal exposure to sunlight. Blood was also drawn from seven pregnant Asian subjects (daily vitamin D intake <40 I.U.) and from nine women receiving a 50 pg oestrogen preparation. Plasma VDBG levels were measured in two subjects with vitamin D intoxication. The VDBG content of human milk was also estimated. Methods
Plasma 25(OH)D3 assays were performed by the competitive protein binding method of Edelstein, Charman, Lawson and Kodicek (1974) [5] using rat kidney binding protein [6] by a modification not requiring column chromatographic separation [ 71 using P-lipoprotein as solubilizing agent. In subjects with chronic liver disease plasma 25(OH)D3 levels were assayed using the method of Skinner and Wills (1977) [8]. Plasma VDBG levels were measured using M-Partigen immunodiffusion plates (Pharma-Behring Diagnostics, Hoechst U.K. Ltd.) containing specific anti-Gc serum. Five microlitre aliquots of sample were applied to the plate and diffusion allowed to proceed overnight at 37°C. Precipitin ring diameter was measured using calipers and the square of the radius was calculated. This value was
361
converted to mg VDBG by comparison with the value for the ring of a known standard applied in duplicate to each plate. Results Control group In the control group the mean plasma 25(OH)D3 level was 85.8 2 10.6 nmol/l. Plasma VDBG concentrations ranged from 18.7 to 34.5 mg/dl (mean + SE. 27.6 + 0.95 mg/dl). Within this group the mean value 2 S.E. among males -. was 26.2 of 1.5 mg/dl and among females 28.7 & 1.2 mg/dl (P > 0.0s). No relationship was observed between plasma 25(OH)D3 and plasma VDBG concentrations (r = 0.19). Vitamin D deficient states In subjects on anticonvulsant therapy (Fig. 1) the mean plasma 25(OH)D3 level i S.E. of 51.0 + 13.5 nmol/l (range 6.3 to 195 nmol/l) was significantly lower than that of the control group (P < 0.01); the mean plasma VDBG level 2 SE., 27.8 ? 1.7 mg/dl, (range 15.8 to 39.5 mg/dl) was similar to the control group (P > 0.05). In vitamin D deficient subjects (inadequate diet, malabsorption syndrome) whose mean plasma 25(OH)D3 level + S.E. was 5.6 2 2.4 nmol/l (range 0 to 15 nmol/l (P < O.Ol), the mean plasma VDBG level + S.E. was 25.5 + 1.5 mg/dl (range 16.6 to 31.8 mg/dl (P > 0.05) Plasma 25(OH)D3 levels in the group of elderly subjects known to have a normal vitamin D status ranged from 20 to 157 nmol/l and the mean value f S.E., 61.5 ? 7.9 nmol/l, was considerably lower than the control group mean
Plasma
VDBG
m/d1 40-
35 -
30 -
.. ... -
-1-
25 -
..
.
...
... ..
.
‘.
-
-7.* .. . .
.‘:
20 -
!. .. ..
15 -
10 -
_ 5-
PCO.01
. .
O-
NS
.
PcO.05 PCO.01
Fig. 1. Plasma VDBG concentrations low plasma 25(OH)D3 levels.
.
(mg/dl)
in control subjects (young and elderly) and in subjects with
362
(P < 0.01). However there was no significant difference in the plasma VDBG in this group (mean + S.E. 26.0 it 0.8 mg/dl, range 17.4 to 33.0 mg/dl) (P > 0.05). In the group of elderly vitamin D deficient subjects (mean plasma 25(OH)D3 + S.E. 5.9 2 1.5 nmol/l, range 0.0 to 19.7 nmol/l, (P < 0.01)) plasma VDBG level ranged from 25.6 to 39.2 mg/dl and the mean value k S.E. 30.7 t 0.9 mg/dl was significantly greater than that of the control group (P < 0.05) and that of the vitamin D replete elderly group (P < 0.01). In subjects with chronic liver disease (mean plasma 25(OH)D3 + S.E. 32.7 k 4.3 nmol/l, range 5.0 to 81.8 nmol/l), the mean plasma VDBG concentration 1 S.E. 21.6 i 0.9 mg/dl (range 8.1 to 34.3 mg/dl) was lower than that of the control group (P < 0.01). Within the group the mean plasma VDBG level k SE. among subjects with alcoholic liver disease (21.9 k 1.4 mg/dl), active chronic hepatitis 22.8 f 1.9 mg/dl) and primary biliary cirrhosis (20.3 + 1.2 mg/dl) were similar (P > 0.05). In a further four subjects with extra hepatic biliary obstruction the mean plasma VDBG concentration -CS,E_, 25.3 k 5.6 mg/dl, was not significantly lower than the control group mean (P > 0.05). kvels
Pregnancy
and oestrogen
therapy
Plasma 25(OH)D3 levels decreased during pregnancy. However the mean value + SE. in the first trimester (99,2 * 29.0 nmol/l), second trimester (80.2 t 11.4 nmol/l) and third trimester (59.0 k 8.3 nmol) did not differ significantly from the control group mean (P > 0.05). Plasma VDBG levels increased during Plasma
VDBG
mg/dl 60-
. .
55 -
50
l
45
: .
l
-
. . .
40 -
35 -
30
-
.. m
I-. . i:;.:
. . 7
JL .
. l
T+-
8
25 -
20 -
? . .
.
15 -
lo-
Fig. 2. Plasma VDBG concentrations (mg/dl) in control and pregnant subjects and in subjects on OeStKOgen therapy.
363
pregnancy (Fig. 2), with a first trimester mean value + S.E. of 31.8 & 2.8 mg/dl, second trimester 36.1 & 2.5 mg/dl and a third trimester value of 41.0 + 2.7 mg/ dl. The latter two values were significantly greater than the control group mean (P < 0.05 and P < 0.01). In seven pregnant Asian subjects the plasma 25(OH)D3 values were below the respective trimester ranges of the pregnant Caucasian subjects while their plasma VDBG levels were within the Caucasian ranges. Plasma VDBF levels in subjects on longterm oestrogen therapy ranged from 27.0 to 45.8 mg/dl and the mean value ? S.E. 37.0 % 2.2 mg/dl was greater than that of the control group (P< 0.01). Other subjects
studied
In two subjects with hypervitaminosis D (plasma 25(OH)D3 levels 350 nmol/l and 450 nmol/l) the plasma VDBG levels, 28.64 mg/dl and 34.0 mg/dl were within the control group range. No VDBG was detected in a sample of human milk. No correlation was noted between plasma 25(OH)D3 concentration and plasma VDBG concentration in any of the groups studied. Discussion The mean plasma VDBG concentration in healthy subjects in this series is lower than that observed by Haddad and Walgate [9] using radioimmunoassay and by other workers using radial immunodiffusion [ 101. The reason for this difference is not apparent. In the present study there was no evidence of a significant influence of sex on plasma VDBG concentration. Age likewise seems not to be a determinant of VDBG level: the mean plasma VDBG level in the control group (mean age 43.5 years) was almost identical to that of a group of elderly subjects of mean age 81.5 years. Moreover Haddad and Walgate (1976) [9] have shown that the ,VDBG concentration in cord blood of full-term infants is similar to that of adults. The relationship between vitamin D status and plasma VDBG concentration is more obscure. The mean plasma VDBG level in the vitamin D deficient elderly subjects was significantly greater than the control group and that of the age-matched group which was vitamin D replete. Furthermore, while the vitamin D replete elderly group had a lower mean plasma 25(OH)D3 level than that of the control group, both groups had levels within the normal range and the plasma VDBG levels were very similar. It is reasonable to infer therefore that it is vitamin D deficiency, rather than age, which is associated with the increased VDBG concentration in the plasma of elderly subjects and this is in agreement with the observation that a vitamin D deficient diet appears to induce increased synthesis of VDBG in rats [ 111. However, the plasma VDBG levels in younger Asian subjects on vitamin D depleted diets and in subjects with the malabsorption syndrome were within the normal range and the plasma VDBG levels in epileptic subjects on chronic anticonvulsant therapy were normal despite low 25(OH)D3 concentrations. Moreover normal plasma VDBG levels were observed in subjects with hypervitaminosis D. Similar findings have been reported by other workers [ 91. The plasma level of VDBG increased during pregnancy and the second and
364
third trimester values were significantly greater than those of the control group. There was also a significant increase in VDBG levels in the plasma of subjects on oestrogen therapy. These findings are similar to those of Haddad and Walgate [9] and suggest that, as with other transport proteins, e.g. thyroxinebinding globulin, cortisol-binding globulin, etc. [12] VDBG synthesis is enhanced by hyperoestrogenaemia. However by analogy with these compounds an increase in plasma 25(OH)D3 during pregnancy might have been anticipated whereas the opposite trend was seen and this was particularly noticeable in Asian subjects. This latter finding may reflect an inadequate dietary vitamin D intake in the presence of a foetus entirely dependent on the placental transfer of maternal 25(OH)D3 [13] or perhaps an induction of hepatic microsomal enzymes during pregnancy [ 141. Subjects with parenchymal and cholestatic liver disease have low serum levels of 25(OH)D3 [15) possibly because of impaired hepatic 25-hydroxylation of vitamin D-3 [ 161, vitamin D deficiency [ 171 or impaired intestinal absorption of vitamin D [18]. The liver is believed to be the site of Gc protein (i.e. VDBG) synthesis [19] and low levels were observed in those subjects with chronic parenchymal liver disease although not in subjects with extrahepatic biliary obstruction. Reduced levels have also been reported in cirrhotic patients who are hypoproteinaemic [ 91. The vitamin D present in human milk is contained mainly in the aqueous phase as the sulphate [20]. This observation prompted an analysis of human milk for VDBG content and none was detected. One of the functions of VDBG is to transport the water-insoluble metabolites of vitamin D to the liver, kidney, gut, bone, etc. and it is appropriate that in an environment where vitamin D is already present in a water-soluble form no VDBG exists. No relationship was observed between serum 25(OH)D3 concentration and serum VDBG concentration either in healthy subjects or in subjects with a disturbance of either or both 25(OHjD3 and VDBG. VDBG is present in human plasma in considerable molar excess compared to the combined plasma concentration of the principal vitamin D metabolites. It seems probable that marked fluctuations in plasma 25(OH)D3 levels may occur without any alteration in VDBG concentration. Conversely minor changes in VDBG level may occur without a corresponding change in plasma 25(OH)D3 level. Nevertheless it is likely that gross reduction in VDBG levels (due to failure of synthesis or abnormal losses) would eventually lead to reduced circulating levels of 25(OH)D3. Thus the low plasma levels of VDBG seen in subjects with extensive liver disease might contribute to their vitamin D deficient state and gross vitamin D deficiency has been reported in the nephrotic syndrome in which there is loss of VDBG in the urine [ 211. Acknowle.dgements The authors wish to acknowledge the cooperation of Professor R.D. Cohen and Professor S. Sherlock. They also thank Dr. R.K. Skinner for the assay of plasma 25(OH)D3 in patients with liver disease.
365
References 1 Peterson, 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
P.A. (1971)
J. Biol. Chem.
246, 7748-7754
Edelstein, S., Lawson, D.E.M. and Kodicek. E. (1973) Biochem. J. 135, 417426 Haddad, J.G. and Walgate, J. (1976) J. Biol. Chem. 251, 48034809 Imwari, M., Kida. K. and Goodman, D.S. (1976) J. Clin. Invest. 58, 514-523 Edelstein, S., Charman. J.. Lawson, D.E.M. and Kodicek, E. (1974) Clin. Sci. Mol. Med. 46, 231-240 Haddad, J.G. and Chyu, K.J. (1971) J. Clin. Endocrinol. 33,992-995 Belsey, H., De Luca, H.F. and Potts, J.T. (1974) J. Clin. Endocrinol. 38, 1046-1051 Skinner, R.K. and MiUs, M.R. (1977) Clin. Chim. Acta 80. 543-554 Haddad, J.G. and W&ate. J. (1976) J. Clin. Invest. 58, 1217-1222 Schmidt-Gayk, H., Schmitt. W.. Grawunder, C.. Ritz, E.. Tschiipe, W.. Pietsch, V.. Andrassy, K. and Bouillon, R. (1977) Lancet ii, 105-108 Belsey, H., De Luca, H.F. and Potts, J.T. (1972) Excerpta Medica 256, p. 25 Westphal, U. (1971) Steroid-Protein Interactions, Springer-Verlag. Berlin, Heidelberg, New York Hillman, L.S. and Haddad. J.G. (1975) J. Pediatr. 86, 928-935 Davis, M., Simmons, C.J.. Dordoni, B., Maxwell, J.D. and Williams, R. (1973) J. Obstet. Gynecol. Br. Commw. 80.690-694 Long, R.G., Skinner, R.K.. Wills, M.R. and Sherlock, S. (1976) Lancet ii, 650-652 Wagonfeld. J.B., Nemchausky. B.A., Bolt, M., Bayer. J.L. and Rosenberg, I.H. (1975) Gastroenterology 69,878 Skinner, R.K.. Long, R.G., Sherlock, S. and Wills. M.R. (1977) Lancet i, 720-721 Compston, J.E. and Thompson, R.P.H. (1977) Lancet i. 721-724 Prunier, J.H., Beam, A.G. and Cleve, H. (1974) Proc. Sot. EXP. Biol. Med. 115.1005-1009 Lakdawala, D.R. and Widdowson, E.M. (1977) Lancet i, 167-168 Barragry, J.M.. France, M.W., Carter, N.D., Auton, J.A., Beer, M., Boucher. B.J. and Cohen, R.D. (1977) Lancet ii. 629432