| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
University Department of Medicine, Manchester Royal Infirmary, (E.B.M., M.D., P.E.S.), Manchester; and the Department of Clinical Biochemistry, St. Bartholomew’s and the Royal London School of Medicine and Dentistry (H.L.J.M., N.J.S.), London, United Kingdom; the Department of Biochemistry, Queen’s University (G.J., V.B.), Kingston, Ontario, Canada; and Bone Care International (C.W.B., J.C.K.), Madison, Wisconsin 53713
Address all correspondence and requests for reprints to: Prof. E. B. Mawer, Department of Medicine, Manchester Royal Infirmary, Manchester, United Kingdom M13 9WL.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
,25-dihydroxyvitamin D2
[1,25-(OH)2D2], and
24,25-(OH)2D2 as metabolites in human sera
(5).
As functional vitamin D status is usually assessed by assaying these
common metabolites, particularly 1,25-(OH)2D, care has
been taken in the past to select an assay method that will not
discriminate between the D2 and D3 forms during
both the purification and detection steps of the procedure. Where a RIA
has been used as the detection technique, antibodies have been chosen
that will not react preferentially with either
1,25-(OH)2D2 or
1,25-(OH)2D3 (6, 7), because in many methods
the two forms are not resolved before assay. Although special care has
been taken to accurately measure
1,25-(OH)2D2, the current assay technology
is based upon the assumption that the metabolism of the two forms of
vitamin D is exactly parallel. By focussing on the conventional
metabolites of vitamin D2, particularly
1,25-(OH)2D2, the analysis has precluded
observing any further active forms of vitamin D2 that may
be present in plasma.
Studies with animals have recently provided evidence that the subtle
differences in the chemistry of the side-chain between vitamins
D2 and D3 result in differences in the site of
hydroxylation and, hence, differences in hydroxylated products,
particularly when large doses of vitamin D are administered (8). Of
particular relevance is the direct 24-hydroxylation of vitamin
D2 to produce 24OHD2. This metabolite has been
observed in the blood of at least three species, including rat (9),
bovine (10), and human (11) and is presumed to be of hepatic origin
in vivo (12). Several other vitamin D2
metabolites have been reported, either as unique products of the
further metabolism of 24OHD2, such as
1,24-(OH)2D2 (10) or
24,26-(OH)2D2 (13), or the catabolites of
1,25-(OH)2D2 (e.g.
1,24,25-trihydroxyvitamin D2,
1,24,25,26-tetrahydroxyvitamin D2, and
1,24,25,28-tetrahydroxyvitamin D2) (14) or of
1,24-(OH)2D2 (e.g.
1,24,26-trihydroxyvitamin D2) (15).
Of these vitamin D2 metabolites,
1,24-(OH)2D2 is the most interesting, partly
because of its potent antiproliferative properties combined with low
calcemic activity (10, 15, 16) and partly because it can be formed
either from vitamin D2 by successive 24- and
1-hydroxylations (10) or from 1OHD2 by hepatic
24-hydroxylation (17). This latter 24-hydroxylation step is unusual, in
that it appears to be carried out by the liver cytochrome P450, CYP27,
which is currently believed to be the hepatic vitamin
D3-25-hydroxylase (18). Although
1,24-(OH)2D2 has been observed in animals
given large doses of vitamin D2 and in human hepatoma cell
incubations in vitro, it has not been shown to be formed in
humans given vitamin D2.
In this study, we set out to use some of the recent advances in assay
technology [cartridges, high performance liquid chromatography (HPLC),
and RIA] to search for some of these unique vitamin D2
metabolites, found previously in the blood of animals and generated
in vitro, in the blood of patients treated with large doses
of vitamin D2 for various clinical conditions. We were able
to measure in such human serum samples the usual 25-hydroxylated
metabolites represented by the mono- and dihydroxylated forms:
25OHD2, 1,25-(OH)2D2, and
24,25-(OH)2D2. We also detected the metabolites
24OHD2 and 1,24-(OH)2D2,
suggesting the existence of a unique 24-hydroxylation pathway of
vitamin D2. As might be expected from such a sequence of
metabolism involving first 24-hydroxylation and then renal
1-hydroxylation, we also found that the synthesis of
1,24-(OH)2D2 was stimulated by the infusion
of a potent 1-hydroxylase inducer, PTH, and correlated, although not
quite significantly, to the concentration of the substrate,
24OHD2.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Details of subjects are given in Table 1
.
|
Vitamin D-deficient subjects given a single large dose of vitamin D2 (subjects 1–7)
Seven vitamin D-deficient patients (initial serum concentration of 25-hydroxyvitamin D, 11.25 ± 1.64 nmol/L) were each given 65 µmol vitamin D2, orally (106 IU). Blood samples were taken before treatment and at intervals for up to 90 days after treatment, and serum was analyzed for vitamin D metabolites. In one of these patients, a larger volume of blood (40 mL) was extracted for definitive identification of certain metabolites.
Vitamin D-deficient subjects treated with small daily doses of vitamin D2 (subjects 8–10)
Two vitamin D-deficient subjects were treated with 400 IU (26 nmol)/day vitamin D2 for 14 days, followed by 4000 IU (260 nmol) vitamin D2/day. In addition, archival samples were reanalyzed from a patient given 1000 IU (65 nmol)/day each of vitamins D2 and D3.
Subjects treated with large daily doses of vitamin D2 (subjects 11 and 13–21)
Archival serum samples were reanalyzed from five patients with X-linked hypophosphatemic osteomalacia (XLH) who had been treated for several years with vitamin D2 [50,000 IU (3.25 µmol)/day] and from one additional XLH patient treated with 50,000 IU each of vitamins D2 and D3. Samples were also examined from two patients with idiopathic hypoparathyroidism receiving 50,000 IU vitamin D2/day and from one patient with Crohn’s disease given 50,000 IU each of vitamins D2 and D3. Archival samples were analyzed for vitamin D metabolites from two of the XLH patients who had received a PTH infusion. The patients had received two infusions of 200 U human PTH-(1–34) (Parathar, Rhone Poulenc-Rorer, Paris, France), iv, over 10 min at 1000 and 1600 h on the day of study. Blood samples were taken immediately before the first infusion and 12, 24, 48, and 72 h after the first infusion.
Vitamin D sterols
Standard vitamin D compounds were donated as follows; 25OHD3, 24,25-(OH)2D3, and 1,25-(OH)2D2, Dr. Milan Uskokvic; 24OHD2, 1,24-(OH)2D2, and 1,25-(OH)2D3, Bone Care International (Madison, WI); 25OHD2, Dr. M. A. Maestro, University of Coruna (Coruna, Spain); 24,25-(OH)D2, Dr. T Kobayashi, Kobe Pharmaceutical University (Kobe, Japan); and 25,26-(OH)D2, Dr. Y. Mazur, Weizman Institute (Tel Aviv, Israel). [26,27-Methyl-3H]-25OHD3 (6.5 tetrabecquerels/mmol) and [26,27-methyl-3H]1,25-(OH)2D3 (6.7 tetrabecquerels/mmol) were obtained from Amersham International (Aylesbury, UK).
Extraction, chromatography, and assay of metabolites
Sequential serum samples were extracted and analyzed for vitamin
D metabolites as described in Fig. 1A.
The scheme used was a development of a previously published method
(19). Elution positions of standard compounds were determined in the
preparative HPLC system used, namely a Zorbax-SIL 3-µm particle size
column eluted with a mobile phase of
n-hexane:propan-2-ol:methanol (30:1:1, vol/vol/vol) at a
flow rate of 2 mL/min (Fig. 1B).
|
-hydroxylated
metabolites and establish their elution positions (Fig. 1C
).
Competitive protein binding assay using human plasma as a source of
vitamin D-binding protein (21) was used to assay a metabolite eluting
in the position of 24,25-(OH)2D2. A novel RIA
using monoclonal antibody (mAb) 5F2 (8) was devised for
1,24-(OH)2D2. As no radiolabeled vitamin
D2 metabolites were available, the assumption was made that
recovery from serum was similar to that of comparable vitamin
D3 compounds, which was assessed by the use of isotopically
labeled metabolites. The validity of using archival samples that had been stored at -20 C for up to 10 yr was checked by comparing the values obtained on reanalysis with those originally obtained. In most cases the methodology was identical, but in some early samples 1,25-(OH)2D2 and 1,25-(OH)2D3 had been measured by RIA using polyclonal antibodies (19) instead of mAb 5F2 (20) as used in in-house assays since 1989. However, the reference ranges and assay characteristics for the two methods are extremely similar and do not introduce any systematic bias.
Cartridges incorporating mAb 1G7 (7) (IDS, Tyne and Wear, UK), were
used for immunoextraction of 1-hydroxylated metabolites (22) as part
of the validation procedure for the assay of
1,24-(OH)2D2.
Assay of PTH was performed using a Nichols Institute (Saffron Walden, UK), Allegro immunoradiometric assay kit for PTH-(1–84).
Extraction and chromatography for identification of metabolites
A 40-mL volume of serum from one subject was extracted with acetonitrile (23) and used to confirm the characterization of 24OHD2, 25OHD2, 24,25-(OH)2D2, and 1,24-(OH)2D2 by comparison with standards. Retention times of HPLC-purified vitamin D metabolites prepared from serum by the above method were compared to those of authentic standards in a further HPLC system consisting of a Zorbax-SIL 3-µm particle size column eluted with n-hexane:propan-2-ol:methanol (91:7:2 or 94:5:1, vol/vol/vol) at a flow rate of 1 mL/min (24). Eluting peaks were monitored using a photodiode array spectrophotometric detector (model 990, Waters Associates, Milford, MA) set to scan in the UV (200–400 nm) range.
In additional experiments, 200 mL serum were extracted to obtain metabolite peaks for analysis by gas chromatography-mass spectrometry (GC-MS).
Mass spectrometry
Two serum samples were extracted and purified by straight phase
HPLC as described above, but with an additional step in that the
fraction corresponding to 1,24-(OH)2D2 was
rechromatographed in the same system, and the
1,24-(OH)2D2 fraction was collected. One sample
came from subject 17 (Table 1) who was taking large daily doses of
vitamin D2 and had an assayed concentration of
1,24-(OH)2D2 of 34 pg/mL (79 pmol/L). The other
sample came from a control subject with no detectable vitamin
D2 metabolites in serum, to which standard
1,24-(OH)2D2 (20 pg/mL; 47 pmol/L) had been
added. To the final extract from each sample, 500 pg (1.2 pmol)
1,25-dihydrotachysterol [1,25-(OH)2DHT3] were
added in a small volume of ethanol; this was used as an internal
standard during GC. These extracts were subjected to an additional
straight phase HPLC separation (Lichrospher Si 60, Merck, Darmstadt,
Germany) using a solvent system hexane:isopropanol:methanol (92:4:4) at
a flow rate of 1.2 mL/min. Solvent eluting between 11–14 min was
collected [this fraction should contain
1,24-(OH)2D2 (11.56 min) and the internal
standard 1,25-(OH)2DHT3 (12.83 min)]; in this
system 1,25-(OH)2D3 has a retention time of
14.85 min. Pertrimethylsilyl ethers were made using
trimethylsilylimidazole and were separated on Lipidex 5000 columns as
previously described (25). The samples were dissolved in a small volume
of N-methyl-N-trimethylsilyl-trifluoroacetamide
(20 µL) before injection into the GC-MS system. Chromatography, using
an HP6890 GC linked to an Autospec high resolution mass spectrometer
(Micromass, Manchester, UK), was carried out on a WCOT capillary column
(Hewlett Packard, Bracknell, United Kingdom) (HP1, cross-linked methyl
silicone gum, 12-m x 0.2-mm x 0.33-µm film thickness).
The end of the GC column was inserted directly into the ion source of
an Autospec high resolution mass spectrometer (Micromass, Manchester,
UK). Ions previously identified by analysis of standard
1,24-(OH)2D2 at a resolution of 10,000 [m/z
513.3584, m/z 554.3975, and m/z 601.3929, representing the ions at
(M-132)+, (M-90)+ and (M-43)+]
were monitored.
Statistical analysis
Statistical analysis was undertaken using Instat instant statistics software (GraphPad, San Diego, CA). Results were expressed as the mean ± SE. Associations between variables were examined using Pearson’s correlation for normally distributed data and Spearman’s rank correlation for nonparametric data.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Vitamin D metabolites, isolated from serum using an
acetonitrile extraction, were rechromatographed using a Zorbax-SIL
column, with various strengths of the ternery solvent system
n-hexane:propan-2-ol:methanol from 91:7:2 to 94:5:1
(vol/vol/vol) required to achieve optimal resolution of peaks. In the
results shown in Fig. 2, each quadrant
represents the comparison of purified vitamin D2
metabolites from human serum with its chemically synthesized standard.
In each case the putative metabolite comigrates exactly with its
standard, providing strong evidence of correct identification.
|
contain the
background-subtracted UV spectrum from the apex of the peak for each
biologically derived metabolite. Although the quality of these spectra
improves as the concentration of metabolite increases, in each case the
spectrum is consistent with a vitamin D compound possessing the
classical vitamin D chromophore (
max = 265 nm;
min = 228 nm). This provides further confidence that the
identification is correct and indicates that each metabolite is not the
previtamin D (max = 260 nm; min = 235 nm)
for the corresponding vitamin D3 metabolite, which
routinely runs close to the vitamin D2 metabolite
(e.g. 24OHD2 runs close to 25-hydroxyprevitamin
D3, both of which run ahead of 25OHD3 on
straight phase HPLC). Particular attention should be paid to Fig. 2D, because this HPLC peak, representing
1,25-(OH)2D2, was calculated to contain only
4.7 ng (10.8 pmol) solute. This corresponds to a concentration of
around 200 pg/mL (500 pmol/L) in the original serum, which agrees quite
well with the RIA result in this case. For the sample processed here
for identification purposes, the fraction containing
1,24-(OH)2D2 did not contain sufficient
material for detection by UV methodology. However, the application of
high resolution MS to the analysis of extracts from a patient taking
high doses of vitamin D2 was sufficiently sensitive to
detect the presence of 1,24-(OH)2D2 in this
extract (see Fig. 3). The identification
of this metabolite as 1,24-(OH)2D2 relies
upon two pieces of evidence. The metabolite had a relative retention
time corresponding to that of standard
1,24-(OH)2D2 that had either been injected
directly onto the mass spectrometer or had been added to a serum sample
which was then extracted in the same way as the sample from a patient
taking pharmacological doses of vitamin D2. The
identification also relies upon the fact that the ratio of the three
ions monitored during GC-MS was the same as that obtained from both the
standard and the extract of supplemented control serum; these data are
summarized in Table 2. Such criteria of
specificity are widely accepted as satisfactory evidence for the
presence of a particular compound in a biological fluid (26).
|
|
Assay systems were developed to measure those vitamin D2 metabolites not included in our routine assay procedure (20). 24OHD2 was measured by UV absorbance in an extension of our system for 25OHD2 and 25OHD3 (20). Standard 24OHD2 was provided as an unresolved mixture of the 24R- and 24S-diastereoisomers that eluted as a single peak at 10.06 min on HPLC using a Zorbax-SIL column and a mobile phase consisting of n-hexane:propan-2-ol:methanol (24:1:1, vol/vol/vol; data not shown). On examining the serum extracts from the vitamin D2-treated patients, a peak was identified in all samples that corresponded to the standard peak for 24OHD2. Quantitation of this peak in the patient samples was achieved by measuring the UV absorbance.
A peak corresponding to 1,24-(OH)2D2 was
measured by RIA after separation from
1,25-(OH)2D2 and
1,25-(OH)2D3 by HPLC. Efficient displacement of
[3H]1,25-(OH)2D3 was obtained
with 50% binding of 13 pmol/L (Fig. 4A),
enabling a standard curve to be established for the assay, which was
validated as follows. A serum extract was prepared from patient
12 (who had previously received vitamin D2) using
immunoextraction cartridges based on mAb 1G7 (22), which binds
1,25-(OH)2D2 and
1,25-(OH)2D3. These are retained on the column
and separated from monohydroxylated metabolites and dihydroxylated
metabolites that lack a 1-hydroxyl group and are readily removed
from the column. The cartridges, however, retain
1,24-(OH)2D2, so a combined 1,24- and
1,25-dihydroxylated fraction can be prepared. This was done, and the
fraction was then applied to a Zorbax-SIL column (4.6 mm x 25 cm)
eluted with n-hexane:propan-2-ol:methanol (30:1:1,
vol/vol/vol). One-minute fractions were collected across the whole
volume corresponding to the HPLC elution positions of standards for the
three metabolites. Clear baseline values were obtained in the fractions
collected between the positions of the three peaks, showing that the
values corresponding to 1,24-(OH)2D2 were a
discrete peak and not a shoulder of the
1,25-(OH)2D2 peak (Fig. 1C).
|
).
Further evidence for the identity of the material as
1,24S-(OH)2D2 came from a study in
which serum from a patient chronically treated with vitamin
D2 was extracted and the
1,24S-(OH)2D2 fraction prepared as
described in Fig. 1B on straight phase HPLC. This fraction was then
rechromatographed in a reverse phase system consisting of a Zorbax-ODS
column eluted with methanol:water (78:22), and 1-min fractions were
collected and assayed as described in Fig. 4A. A clear peak of
immunoassayable activity was observed in the position corresponding to
1,24S-(OH)2D2 in the reverse phase
system (see Fig. 4C).
Time course of vitamin D2 metabolism in vitamin D-deficient subjects given a single large dose of vitamin D2
The 30-day time course for the mean concentration of the
monohydroxylated metabolites and 24,25-(OH)2D2
from seven vitamin D-deficient subjects (Table 1, no. 1–7) is shown in
Fig. 5A, and the time course for the
three dihydroxylated metabolites with a 1-hydroxy function is shown
in Fig. 5B. Similar patterns of metabolism were observed in all seven
subjects, although the extent of the response varied. The rapid rise in
25OHD2 became a plateau between days 3–17, with a slow
decline thereafter (Fig. 5A); the level was still high (170 nmol/L) in
one subject sampled on day 90. In contrast, levels of
25OHD3 remained low and constant at about 10 nmol/L. The
metabolite that cochromatographed with standard 24OHD2
achieved its maximum level on day 1 with a value of 34 ± 8.4
nmol/L (mean ± SE) and declined slowly over 30 days,
although low values could still be detected up to 90 days.
24,25-(OH)2D2 was initially undetectable, but
rose to about 10 nmol/L by day 17. Quantitatively,
1,25-(OH)2D2 was the major 1-hydroxylated
metabolite (Fig. 5B), with a peak by day 2 of 584 ± 205 pmol/L;
1,24S-(OH)2D2 was present at about
1/10th the concentration, again peaking on day 2 at 57.5 ± 19.9
pmol/L. The endogenous 1,25-(OH)2D3 declined
throughout the course of the study from an initial value of 72.7
± 14.8 pmol/L to 14.5 ± 4.2 pmol/L by day 30. To investigate
whether the metabolic response to vitamin D2 depended on
the degree of vitamin D deficiency, correlations were examined between
levels of the dihydroxylated metabolites and indices of vitamin D
deficiency. The peak level of 1,25-(OH)2D2,
but not that of 1,24-(OH)2D2, correlated
significantly and inversely with the initial serum calcium level
(P = 0.048; Spearman’s rank correlation,
rs, -0.786), but neither metabolite correlated with
initial serum 25OHD or PTH levels.
|
Linear correlations were examined between the principal
metabolites over the 30 days after treatment. No significant
relationships were found, however, for three sets of data; these were
reported as being "almost significant." The Pearson correlation
coefficient for 1,24S-(OH)2D2 and
24OHD2 was 0.699 (P = 0.053), that for
1,25-(OH)2D2 and 25OHD2 was 0.71
(P = 0.051), and that for
1,24S-(OH)2D2 and 25OHD2
was 0.66 (P = 0.073). The stronger relationships
between the 1-hydroxylated forms and their corresponding respective
24- or 25-hydroxylated compounds suggest a precursor-product
relationship.
Production of 1,24-(OH)2D2 in patients treated daily with large or small doses of vitamin D2
Data are presented in Table 3 for
vitamin D2 and D3 metabolite concentrations in
patients treated with large (50,000 IU) daily doses, usually of vitamin
D2 (subjects 13–17, XLH patients; subjects 20–21,
hypoparathyroid patients), but in two cases (subject 11, vitamin D
deficient; subject 18, XLH) they were given equimolar doses of each
form of the vitamin. Metabolites corresponding to
1,24S-(OH)2D2 and 24OHD2
were seen in most patients. A decline in both of these metabolites was
seen in patients 15 and 16, in whom therapy was discontinued in 1993
and 1987, respectively. Patient 18 was vitamin D deficient when she
started taking equimolar doses of vitamin D2 and
D3, but with vitamin D repletion,
1,24-(OH)2D2 became measurable in serum.
Patient 11 was a vitamin D-deficient patient with Crohn’s disease and
was given the same treatment; he responded in a similar fashion. Table 4 shows results for vitamin D-deficient
patients treated with near-physiological doses of vitamin
D2. Both 1,25-(OH)2D2 and
1,24-(OH)2D2 were detectable in the serum of
patient 9. Low concentrations of 1,24-(OH)2D2
were detectable even when this patient was taking only 400 IU/day
vitamin D2. In contrast, only trivial amounts of
1,24-(OH)2D2 were assayed in patient 8, whose
1,25-(OH)2D2 did not reach high levels with
either 400 or 4,000 IU doses; however, PTH concentrations were much
lower than those in patient 9. In the serum of the third patient
treated with low doses of vitamin D2 and D3
(1,000 IU/day), assayable amounts of
1,24-(OH)2D2, but not 24OHD2, were
detected.
|
|
Archival samples were available for two patients with XLH
(subjects 15 and 19) who were given an infusion of PTH (see
Materials and Methods). Serum samples were collected for 3
days, and the stored samples were analyzed for vitamin D metabolites
and PTH as described above. Results for subject 19 (shown in Fig. 6) demonstrate that changes in the
concentration of 1,24-(OH)2D2 mirror those in
1,25-(OH)2D2 and rise in response to the
increase in PTH. In patient 15, levels rose somewhat later in relation
to PTH, but the same basic pattern was seen.
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
,25-(OH)2D2, and
24,25-(OH)2D2, which are routinely measured in
clinical assays for vitamin D3 metabolites, there
exists a series of metabolites based upon 24-hydroxylation of the
D2 side-chain. These unique derivatives include
24OHD2 (9, 10, 11) and a biologically active form,
1,24-(OH)2D2, observed previously only in
animal studies and in experiments employing cultured cells in
vitro (10, 17).
In our studies, we were able to detect
1,24-(OH)2D2 by a combination of high
resolution HPLC and a sensitive RIA based upon an antibody known to
detect both 1,25-(OH)2D2 and
1,25-(OH)2D3 (7). Levels of
1,24-(OH)2D2 were measured in the low
picomolar range after the administration of large doses of vitamin
D2 to a variety of patients with disorders of calcium
homeostasis. In some patients receiving long term vitamin
D2 therapy, the levels of
1,24-(OH)2D2 exceeded or were equivalent to
the values reached by the alternative 1-hydroxylated metabolite,
1,25-(OH)2D2. The detection of
1,24-(OH)2D2 was in most cases accompanied
by a parallel increase in the level of 24OHD2 and appeared
to be PTH dependent, suggesting that an alternative pathway of
activation of vitamin D2 through 24OHD2 occurs
in humans.
Although the group of vitamin D-deficient subjects who received a large
single dose of vitamin D2 were predominantly of Asian
origin (i.e. from the Indian subcontinent), as this
immigrant group in the United Kingdom is particularly prone to vitamin
D deficiency, there is no reason to believe that there is a genetic
component to the response. We were fortunate in having access to stored
samples from patients with XLH and hypoparathyroidism dating from the
time when treatment of these conditions with large doses of vitamin
D2 was standard. These patients were all Caucasian, and
they showed qualitatively the same response as the vitamin D-deficient
Asians. In the study employing a single large dose of vitamin
D2, the level of 1,25-(OH)2D2
formed correlated with the level of serum calcium, one of the indices
of vitamin D deficiency, although not with the initial 25OHD level,
possibly because the range of the latter was small. In addition,
concentrations of 25OHD tend to be labile, changing quickly with small
differences in intake of the vitamin.
The identification techniques used in this paper have rarely been
applied to blood derived from clinical studies. For most of the
metabolites of vitamin D2, we were able to analyze the
compounds not only by chromatographic mobility with standards but also
by diode array spectrophotometry, which indicated that the compounds
have the correct chromophore for vitamin D. For the peaks containing
tens of nanograms of material, such as 25OHD2 and
24OHD2, this approach is relatively straightforward and
provides good confirmation of peak identity. For the analysis of the
1,25-(OH)2D2 peak, which contains many fewer
nanograms of material, we were able to obtain a UV spectrum, albeit of
lower quality.
Despite this remarkable sensitivity of the HPLC diode array detector,
we were unable to analyze 1,24-(OH)2D2 by
this method, and we, therefore, used a combination of high resolution
HPLC and a sensitive RIA. In the sample analyzed in this way, however,
1,24S-(OH)2D2 was present at only about 10%
of the concentration of 1,25-(OH)2D2. It
should be noted that no other known metabolite comigrates with
chemically synthesized 1,24S-(OH)2D2 on this
chomatographic system, and extreme care was taken not to confuse this
metabolite with closely migrating 1,25-(OH)2-previtamin
D2 or 1,25-(OH)2-previtamin D3
during collection. As previtamin D/vitamin D equilibrium ratios rarely
reach 1/10 in any vitamin D samples, and given that
1,24-(OH)2D2 values in some cases approached
those of 1,25-(OH)2D2 in our study, and
1,25-(OH)2D3 levels were usually very low,
it seems highly unlikely that the measured fraction simply represents
poorly resolved 1,25-(OH)2-previtamin D2 or
1,25-(OH)2-previtamin D3. To resolve the
question of whether 1,24-(OH)2D2 is present
in human serum after the administration of vitamin D2, we
used the highly sensitive technique of high resolution GC-MS of the
pertrimethylsilyl ether derivative. Although the concentrations of
1,24-(OH)2D2 were insufficient to obtain a
complete mass spectrum, monitoring three ions [one of which is
specific for the 1,24-(OH)2D2 structure]
gave a peak with the appropriate retention time with ion ratios that
corresponded with those found in standard
1,24-(OH)2D2. This technique provides
satisfactory specificity to allow the conclusion that
1,24-(OH)2D2 is present in human serum after
the administration of vitamin D2.
Based upon experience with vitamin D3, current dogma in the
vitamin D field suggests that 24-hydroxylation of the side-chain only
occurs after prior 25-hydroxylation (27). Furthermore, this
24-hydroxylation is renal or target cell based and involves the
cytochrome P450, CYP24 (28). Although this may well be true for vitamin
D3, it certainly does not appear to be the case for vitamin
D2, where a wealth of information gained from animal models
(8, 9, 10) suggests that direct 24-hydroxylation of the vitamin
D2 side-chain can occur, and 24OHD2 is a
significant product. From various clues gathered in these studies,
investigators have speculated that 24OHD2 is formed in the
liver (12, 29). Further support for 24-hydroxylation by the liver has
come from direct in vitro work with hepatoma cell lines,
HepG2 and Hep3B, and with the liver cytochome P450, CYP27, transfected
into COS-1 cells (17, 18). Thus, it seems that unlike vitamin
D3, for which 24-hydroxylation is mainly performed in
target cells and not the liver, the 24-hydroxylation of vitamin
D2 can occur in the liver. The 1-hydroxylation step for
24OHD2, however, appears to be carried out by the same
renal enzyme that is involved in the formation of
1,25-(OH)2D2 and
1,25-(OH)2D3. The observed pattern of PTH
stimulation of the biosynthesis of
1,24-(OH)2D2 is consistent with this
hypothesis, as PTH is a well characterized stimulator of the renal
enzyme. Work by Horst et al. (10) in the bovine species also
found evidence of the same pathway of 24-hydroxylation followed by
1-hydroxylation, which has profound implications when contemplating
the clinical use of vitamin D2.
Examination of the pattern of results obtained from patients treated
either with one large dose or repeated doses of vitamin D2
suggests that levels of both 24OHD2 and
1,24-(OH)2D2 are higher in the patients
receiving chronic therapy. This would support the theory that the
initial 24-hydroxylation of vitamin D2 occurs in the liver,
as vitamin absorbed from the diet and entering the circulation in
chylomicrons would undergo a type of first pass phenomenon; hence, the
concentration of this metabolite would remain high when there was daily
oral treatment of the parent vitamin. In contrast, a single large dose
would pass through the liver initially on absorption, but thereafter
would be largely redistributed into body tissues (30), from where it
would be mobilized slowly on the vitamin D-binding protein. The sharp
peak of 24OHD2 seen in contrast to the steadily maintained
level of 25OHD2 supports this hypothesis. As it is
postulated that 1,24-(OH)2D2 is formed from
24OHD2, then a similar pattern was expected for this
metabolite and was seen after the single dose. In contrast,
consistently raised levels of 1,24-(OH)2D2
and 24OHD2 are seen after chronic treatment,
e.g. in the XLH patient 17. An additional reason for this is
suggested by experiments using the murine model of XLH, which indicate
increased renal 24-hydroxylase messenger ribonucleic acid levels and
activity (31, 32); it is possible that a hepatic 24-hydroxylase could
also be up-regulated. Support for the renal origin of
1,24-(OH)2D2 is provided by the increase
seen in response to PTH infusion, as the renal 1-hydroxylase is
stimulated by PTH. The data in Table 3 suggest that serum
1,24-(OH)2D2 levels may be higher in XLH
patients than in those with hypoparathyroidism (who by definition had
low PTH levels), all of whom had the same intake of vitamin
D2, but unfortunately numbers were too small for a
definitive comparison. In general, in chronically treated patients,
concentrations of 1,24-(OH)2D2 seem higher
in relation to 1,25-(OH)2D2 than the ratio
of 24OHD2 to 25OHD2 would warrant. This may
indicate either that 24OHD2 is more readily
1-hydroxylated than 25OHD2 or that the turnover of
1,24-(OH)2D2 is slower than that of
1,25-(OH)2D2.
To the best of our knowledge, only one preliminary report has presented
evidence for the existence of the alternative monohydroxylated form,
24OHD2, in human plasma samples (11). This lack of
corroborative evidence reflects the rigidity of current analytical
approaches that focus only on conventional metabolites. Furthermore,
the S-isomer of 1,24-(OH)2D2,
which is also the natural isomer formed in hepatomas in
vitro (17) and is found in human blood, possesses a biological
activity approaching that of 1,25-(OH)2D3
(15, 17) in vitamin D-dependent transcription assays and cell
antiproliferation assays. Clinical vitamin D assay strategies (33),
therefore, should be designed to take these new forms of vitamin
D2 into account, especially where there is a heavy use of
vitamin D2 in the population. A reminder that assay
technology should be measuring all of the active forms of the given
analog of vitamin D has been sounded of late based on the realization
that certain active analogs may give rise to one (or more) biologically
active products. Examples of such analogs include
F6-1,25-(OH)3D3 and KH1060,
which are not only both active in their own right, but also are
converted to other stable and biologically active forms:
F6-1,23,25-(OH)3D3 and
26OHKH1060, respectively (34, 35). Thus, the concept that a vitamin D
analog, in this case vitamin D2, gives rise to two
biologically active products in the form of
1,24-(OH)2D2 and
1,25-(OH)2D2 is not without precedent.
We conclude from these studies that the biological effects of vitamin
D2 in humans are different, perhaps only subtly, from those
of vitamin D3. Our conclusion builds logically on those of
other researchers who have reported different effects of the two
vitamins in other mammals (36, 37). It is also consistent with the
observed lower toxicity of vitamin D2 (38, 39, 40, 41) and
1OHD2 (42, 43, 44, 45, 46). Such a toxicity advantage appears likely
for the synthetic analog 1OHD2 and for
1,24-(OH)2D2 over their vitamin
D3 equivalents (8, 45, 46) and could make
1OHD2 and 1,24-(OH)2D2
attractive alternatives to the corresponding vitamin D3
analogs for future clinical use. Thus, although vitamin D2
has long been considered as equivalent to its natural homolog, the
current studies suggest that it may possess a sufficiently different
metabolic pattern so as to offer advantages in its toxicity profile,
advantages that may be carried over into the design of future
generations of vitamin D analogs, which can be used as clinical
therapies.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received December 12, 1997.
Revised February 3, 1998.
Accepted February 18, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
,25-dihydroxyvitamin D. Steroids. 46:741–754.[CrossRef][Medline]
-Hydroxylation of 24-hydroxyvitamin D2 represents a
minor physiological pathway for the activation of vitamin
D2 in mammals. Biochemistry. 29:578–582.[CrossRef][Medline]
,24(S)-dihydroxyvitamin D2 in normal
and immortalized human epidermal cells. Biochem Pharmacol. 52:133–140.[CrossRef][Medline]
,24-dihydroxyvitamin D2 in rats. Comparison with
1:24-dihydroxyvitamin D2, calcitriol and calcipotriol. Biochem Pharmacol. 53:829–837.[CrossRef][Medline]
,24(S)-Dihydroxyvitamin D2: a biologically
active product of 1-dydroxyvitamin D2 made in the human
hepatoma, Hep3B. Biochem J. 310:233–241.
-Hydroxyvitamin D2 is less toxic than
1-hydroxyvitamin D3 in the rat. Proc Soc Exp Biol Med. 178:432–436.[CrossRef][Medline]
-hydroxyvitamin D2
on calcium homeostasis in postmenopausal osteopenic women. J Bone
Miner Res. 9:607–614.[Medline]
-hydroxyvitamin
D2 in hemodialysis patients with moderate to severe
secondary hyperparthyroidism. Kidney Int. 51:317–323.[Medline]
(OH)-vitamin D2 in
suppressing 2° hyperparathyroidism in hemodialysis patients. Dialysis
Transplant. 26:583–630.
-Hydroxyvitamin D2 partially dissociates between
preservation of cancellous bone mass and effects on calcium homeostasis
in ovariectomized rats. Calcif Tissue Int. 60:449–456.
This article has been cited by other articles:
 |
|
 |
|
  K. J. Martin and E. A. Gonzalez Metabolic Bone Disease in Chronic Kidney Disease J. Am. Soc. Nephrol., March 1, 2007; 18(3): 875 - 885. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A Houghton and R. Vieth The case against ergocalciferol (vitamin D2) as a vitamin supplement. Am. J. Clinical Nutrition, October 1, 2006; 84(4): 694 - 697. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Maunsell, D. J. Wright, and S. J. Rainbow Routine Isotope-Dilution Liquid Chromatography-Tandem Mass Spectrometry Assay for Simultaneous Measurement of the 25-Hydroxy Metabolites of Vitamins D2 and D3 Clin. Chem., September 1, 2005; 51(9): 1683 - 1690. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. P. Gupta, Y. A. He, K. S. Patrick, J. R. Halpert, and N. H. Bell CYP3A4 Is a Vitamin D-24- and 25-Hydroxylase: Analysis of Structure Function by Site-Directed Mutagenesis J. Clin. Endocrinol. Metab., February 1, 2005; 90(2): 1210 - 1219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Upton, J. C. Knutson, C. W. Bishop, and L. W. LeVan Pharmacokinetics of doxercalciferol, a new vitamin D analogue that lowers parathyroid hormone Nephrol. Dial. Transplant., April 1, 2003; 18(4): 750 - 758. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Vieth, P.-C. R Chan, and G. D MacFarlane Efficacy and safety of vitamin D3 intake exceeding the lowest observed adverse effect level Am. J. Clinical Nutrition, February 1, 2001; 73(2): 288 - 294. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. J. Schroeder and J. Cunningham What's new in vitamin D for the nephrologist? Nephrol. Dial. Transplant., April 1, 2000; 15(4): 460 - 466. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
), 25OHD3
(
), and 24,25-(OH)2D2 (
). B,
Dihydroxylated metabolites, 1