Vitamin D deficiency is prevalent today not only among the elderly but pervasively throughout all ages of life.This is due, in part, to systemic diseases that affect vitamin D metabolism, to changes in lifestyle, such as insufficient exposure to sunlight, and to increased use of sunscreen. Apart from the obvious effects of vitamin D deficiency on skeletal metabolism, the problem is assuming even greater significance because observational and interventional studies have linked vitamin D deficiency to cardiovascular disease, cancer, and diabetes.
This book addresses a variety of important issues that have emerged from this fast-moving area of clinical medicine. The topics include assays of vitamin D and its binding protein, effects on aging and associated complications, primary and secondary states of altered parathyroid hormone secretion, vitamin D in the growing years of children and adolescents, nutritional requirements, and vitamin D as it relates to systemic disorders such as diabetes mellitus.
Vitamin D in Clinical Medicine aims to offer new insights, in an evidence-based way, on important issues related to vitamin D. It is written for general practitioners and internists, as well as for endocrinologists, nutritionists, pulmonologists, cardiologists, and oncologists.
14 - 30: Vitamin D Assays
-
Published:2018
-
Subject Area: Cardiovascular System , Endocrinology , Further Areas , Immunology and Allergy , Nutrition and Dietetics , Oncology , PneumologyBook Series: Frontiers of Hormone Research
Daniel D. Bikle, 2018. "Vitamin D Assays", Vitamin D in Clinical Medicine, A. Giustina, J.P. Bilezikian
Download citation file:
Abstract
The number of requests for vitamin D metabolite measurements has increased dramatically over the past decade leading commercial laboratories to develop rapid high throughput assays. The measurement of 25-hydroxyvitamin D (25[OH]D) and to a lesser extent 1,25-dihydroxyvitamin D (1,25[OH]2D) dominates these requests, but requests for multiple metabolite measurements in the same sample are also increasing. The most commonly used methods include immunoassays and liquid chromatography/mass spectrometry (LC-MS). Each method has its advantages and disadvantages, but with improvements in technology, especially in LC-MS, this method is gaining ascendance due to its greater precision and flexibility. The use of standards from the National Institutes of Standards and Technology has substantially reduced the variability from laboratory to laboratory, thereby improving the reliability of these measurements. Although the current demand is for measurement of total vitamin D metabolite levels, these metabolites circulate in blood tightly bound to vitamin D binding protein (DBP) and albumin with less than 1% free. The free concentration may be a more accurate indicator of vitamin D status especially in individuals with DBP levels that deviate from the normal population. Thus, methods to measure the free concentration at least of 25(OH)D are becoming available and may supplement if not replace measurements of total levels.
Introduction
Nearly all, if not all, cells express the vitamin D receptor (VDR) at some stage in their development or activation, and many of these cells are also able to convert vitamin D to its active metabolites. As the appreciation that vitamin D affects numerous physiologic processes other than bone and mineral metabolism, and that these physiologic processes may have different optimal levels of vitamin D [1], interest in the measurement of the levels of vitamin D and its metabolites has soared. Moreover, disorders of vitamin D metabolism can be diagnosed by accurate measurement of these metabolites, and potential differences in ratios of vitamin D metabolites even in otherwise normal individuals can be predictive of differences in responses to dietary intakes of vitamin D in food and/or in supplements due to differences in metabolism [2]. However, the measurement of vitamin D metabolites is not trivial. These are lipophilic materials circulating in low concentrations tightly bound to proteins, vitamin D binding protein (DBP) and albumin in particular, making their measurements difficult. If one considers that only the free (i.e., non-protein bound) metabolite enters cells and is the biologically important concentration to consider [3], the requirements for sensitivity of measurement increase by several orders of magnitude. Moreover, distinguishing between the different metabolites that may differ only modestly chemically but with substantial differences biologically and quantitatively likewise contributes to the difficulties developers of assays have in providing a fast throughput assay at reasonable cost to meet the increasing demand for these measurements. In this review, the most common assays on the market today are described reviewing their advantages and limitations with a discussion of newer technologies including the development of assays intended to measure the free concentrations.
Vitamin D Production and Metabolism
Before discussing the different assays, it is important to review both the production of vitamin D and its subsequent metabolism to its active metabolites, the measurement of which is the major focus of this review.
Vitamin D Production
Vitamin D3 (D3) (cholecalciferol) is produced from 7-dehydrocholesterol through a 2 step process in which the B ring is broken by ultraviolet light (UVB spectrum 280-320 nm) in sunlight, forming pre-D3 that isomerizes to D3 in a thermo-sensitive but a noncatalytic process. Both UVB intensity and skin pigmentation level contribute to the rate of D3 formation [4]. Melanin in the skin blocks UVB from reaching 7-dehydrocholesterol in the lower portions of the epidermis, thus limiting D3 production, as do clothing and sunscreen. The intensity of UVB from sunlight varies according to season and latitude, so the further one lives from the equator, the less time of the year one can rely on solar exposure to produce D3[5]. Vitamin D is also obtained from the food consumed. Most foods with the exception of fatty fish contain little vitamin D unless fortified. The vitamin D in fish is D3, whereas that used for fortification is often D2 (ergocalciferol). D2 is produced by UVB irradiation of ergosterol in plants and fungi (e.g., mushrooms). It differs from D3 in having a double bond between C22-C23 and a methyl group at C24 in the side chain. These differences from D3 in the side chain lower its affinity for DBP resulting in faster clearance from the circulation, alter its conversion to 25-hydroxyvitamin D (25[OH]D) by at least some of the 25-hydroxylases to be described, and alter its catabolism by the 24-hydroxyase (CYP24A1) [6,7,8]. Moreover, a number of immunoassays do not recognize the D2 metabolites as well as the D3 metabolites. However, the biologic activities of D2 and D3 metabolites are comparable, and if no subscript is used, both forms are meant.
Vitamin D Metabolism
The 3 main steps in vitamin D metabolism, 25-hydroxylation, 1α-hydroxylation, and 24-hydroxylation are all performed by cytochrome P450 mixed function oxidases (CYPs) located either in the endoplasmic reticulum (e.g., CYP2R1) or in the mitochondrion (e.g., CYP27A1, CYP27B1, and CYP24A1).
25-Hydroxylase. The liver has been established as the major if not sole source of 25(OH)D production from vitamin D. Initial studies of the hepatic 25-hydroxlase found activity in both the mitochondrial and microsomal (endoplasmic reticulum) fractions. Subsequent studies demonstrated a number of CYPs with 25-hydroxylase activity. CYP27A1 is the only mitochondrial 25-hydroxylase. It was initially identified as a sterol 27-hydroxylase involved in bile acid synthesis. This CYP is widely distributed in the body, not just in the liver. It hydroxylates D3 but not D2. Moreover, its relevance to vitamin D metabolism has been questioned when its deletion in mice actually resulted in increased blood levels of 25(OH)D [9]. Moreover, inactivating mutations of CYP27A1 in humans cause cerebrotendinous xanthomatosis with abnormal bile and cholesterol metabolism but not rickets [10]. More recently, CYP2R1 was identified in the microsomal fraction of mouse liver [11]. This enzyme 25-hydroxylates both D2 and D3 with comparable kinetics, unlike CYP27A1. Its expression is primarily in the liver and testes. When CYP2R1 is deleted from mice, blood levels of 25(OH)D fall over 50% but not to zero [9]. Even the double deletion of CYP2R1 and CYP27A1 does not reduce the blood level of 25(OH)D to zero, and actually has little impact on blood levels of calcium and phosphate [9] suggesting compensation by other enzymes with 25-hydroxylase activity. However, mutations in CYP2R1 have been found in humans presenting with rickets, and these mutations decrease 25-hydroxylase activity when tested in vitro [12]. Although other enzymes including the drug metabolizing enzyme CYP3A4 have 25-hydroxylase activity and may have roles in different tissues or in different clinical conditions, CYP2R1 appears to be the major 25-hydroxylase contributing to circulating levels 25(OH)D. Regulation of vitamin D 25-hydroxylation is modest at best with production being primarily substrate dependent such that circulating levels of 25(OH)D are a useful marker of vitamin D nutrition. Moreover, it circulates in concentrations well above that of other metabolites facilitating its measurement.
1α-Hydroxylase (CYP27B1). Unlike 25-hydroxylation, there is only one enzyme recognized to have 25-OHD 1α-hydroxylase activity, and that is CYP27B1. Although the kidney is the main source of circulating 1,25-dihydroxyvitamin D (1,25[OH]2D), a number of other tissues also express the enzyme, and the regulation of the extrarenal CYP27B1 differs from that of the renal CYP27B1(review in [13]). The renal 1α-hydroxylase is tightly regulated primarily by 3 hormones: parathyroid hormone (PTH), FGF23, and 1,25(OH2D itself. PTH stimulates, whereas FGF23 and 1,25(OH)2D inhibit CYP27B1. Elevated calcium suppresses CYP27B1 primarily through the suppression of PTH; elevated phosphate suppresses CYP27B1 primarily by stimulating FGF23, although these ions can have direct effects on renal CYP27B1 [14,15]. One major extrarenal location of CYP27B1 is in epithelia including epithelial cells of the epidermis, intestine, mammary gland, lung, and prostate [16]. In epidermal keratinocytes tumor necrosis factor-α [17] and interferon-γ [18 ]are the major inducers of CYP27B1 activity although PTH also stimulates CYP27B1 but not through cAMP mediated mechanisms as in the kidney [19]. Immune cells likewise express CYP27B1 especially when activated, and like the keratinocyte CYP27B1 is induced by tumor necrosis factor-α and interferon-γ [20]. In these cells, PTH and calcium have little impact on CYP27B1 activity, and feedback regulation by 1,25(OH)2D is mediated indirectly by induction of CYP24A1 expression [21], a mechanism that is blunted in macrophages [22]. Thus, the measurement of 1,25(OH)2D is useful not only in renal disease and in diseases associated with too little or too much PTH, FGF23, calcium and phosphate but also in identifying diseases of extrarenal tissues in which production of 1,25(OH)2D is activated.
24-Hydroxylase. CYP24A1 is the only established 24-hydroxylase involved with vitamin D metabolism. This enzyme has both 24-hydroxylase and 23-hydroxylase activity, the ratio of which is species dependent [23]. The enzyme in humans has both capabilities, but the rat enzyme is primarily a 24-hydroxylase [24]. Mutating ala 326 to gly 326 in the human CYP24A1 shifts the profile from one favoring 24-hydroxylation to one favoring 23-hydroxylation [25]. The 24-hydroxylase pathway results in the biologically inactive calcitroic acid, whereas the 23-hydroxylase pathway produces the biologically active 25(OH)D-26,23-lactone and 1,25(OH)2D-26,23 lactone. All steps are performed by one enzyme [24]. 1,25(OH)2D is the preferred substrate relative to 25(OH)D, but both are 23 or 24-hydroxylated. Like the lactones, 1,24,25(OH)3D has substantial affinity for the VDR and so has biological activity. There may also be a physiologic role for 24,25(OH)2D in the growth plate in that both 1,25(OH)2D and 24,25(OH)2D appear to be required for optimal endochondral bone formation [26]. Inactivating mutations in CYP24A1 have been found in children with idiopathic infantile hypercalcemia and more recently in adults who present with severe hypercalcemia, hypercalciuria, and nephrocalcinosis with decreased PTH, low 24,25(OH)2D, and inappropriately normal to high 1,25(OH)2D [27,28]. Measuring the ratio of 24,25(OH)2D:25(OH)D has proven useful in diagnosing these cases. Regulation of CYP24A1is the reciprocal of that of CYP27B1 at least in the kidney in that PTH inhibits but FGF23 stimulates its expression. However, in osteoblasts, PTH enhances 1,25(OH)2D induction of CYP24A1 transcription [29], illustrating the fact that regulation of these vitamin D metabolizing enzymes is cell specific. That said in essentially all cells in which it is expressed, CYP24A1 is strongly induced by 1,25(OH)2D, and often serves as a marker of 1,25(OH)2D response in that cell.
3-Epimerase. 3-Epimerase activity was first identified in the keratinocyte, which produces large amounts of the C-3-epi form of 1,25(OH)2D [30]. It has also been identified in a number of other cells but not in the kidney [31]. The enzyme per se has not yet been purified and sequenced, so it is not clear that one gene product is involved. The 3-epimerase isomerizes the C-3 hydroxy group of the A ring from the alpha to beta orientation of all natural vitamin D metabolites. This does not restrict the action of CYP27B1 or CYP24A1. However, the C-3 beta epimer of 25(OH)D has reduced binding to DBP relative to 25(OH)D, and the C-3 beta epimer of 1,25(OH)2D has reduced affinity for the VDR relative to 1,25(OH)2D [32], thus reducing its transcriptional activity and most biologic effects [32]. Surprisingly, however, it is equipotent to 1,25(OH)2D with respect to PTH suppression [33]. Clinically, interest in the C-3 epimerase arises because the C-3 beta epimer of the vitamin D metabolites is not readily distinguished from their more biologically active alpha epimers by LC/mass spectrometry unless special chromatographic methods to separate the epimers prior to mass spectrometry are employed. Thus, the measurement of these metabolites using standard LC/mass spectroscopic procedures results in a value increased above true levels of the C-3 alpha epimers to the extent that the sample contains the C-3 beta epimer. Immunoassays by and large do not recognize the C-3 beta epimer and so are not affected [34]. This issue is particularly important in assessing 25(OH)D levels in infants where levels of the C-3 beta epimer of 25(OH)D can equal or exceed that of the C-3 alpha epimer of 25(OH)D [31]. However, levels in adults can also be substantial [31]. Given that the C-3 beta epimer does have biologic activity and that the epimers can be separated prior to mass spectrometry, there may be justification for measuring both epimers to provide a more complete picture of vitamin D status at least in future research protocols or when assessing infant samples. At this point it is not clear whether the cost/benefit of such additional effort justifies its application to adult samples measured routinely. Unless otherwise stated, reference to the vitamin D metabolites without stipulating which epimer implies the C-3 alpha epimer.
CYP11A1. Recently an alternative pathway for vitamin D metabolism at least in keratinocytes has been identified, namely, 20-hydroxylation of vitamin D by CYP11A1, the side chain cleavage enzyme essential for steroidogenesis [35]. The product, 20OHD, or its metabolite 20,23(OH)2D, appears to have activity similar to that of 1,25(OH)2D at least for some functions [35]. At this point, the measurement of these metabolites is not commercially available and not further discussed in this review
Assays for Vitamin D and Its Metabolites
Vitamin D
Vitamin D is seldom measured in the blood clinically, but methods for its analysis have been developed for the food industry. Vitamin D is the most lipophilic of the compounds we will consider. Relative to 25(OH)D it has reduced affinity for DBP, and is cleared within hours from the blood, presumably deposited into fat tissues. Typical methods include saponification of the sample with organic extraction and high performance liquid chromatography (HPLC) to separate D2 and D3. These peaks are detected and quantitated by UV absorption with a diode array detector (LC-UV) or mass spectrometry (LC-MS) [36]. The latter detection system is more specific and less sensitive to interfering substances, but ionization of vitamin D required for MS is limited thus affecting precision. This plus lack of a universal standard has resulted in wide variation between laboratories in measuring vitamin D. More recently, however, the situation has improved with the use of a triple quadrupole tandem mass spectrometer equipped with an atmospheric pressure photo ionization source that enhances the ionization of D versus earlier chemical methods (atmospheric pressure chemical ionization or APCI and electrospray ionization [ESI]) [37]. This method was used to measure D2, D3, 25(OH)D2, and 25(OH)D3 in the same sample with limits of quantitation (LOQ) for each D of 2 ng/mL and for each 25(OH)D of 1 ng/mL [38].
25-Hydroxyvitamin D
The accurate measurement of 25(OH)D for the assessment of vitamin D status has been the major goal for most commercial laboratories measuring vitamin D metabolites for a good reason. Levels of 25(OH)D in the blood are higher than those of any other vitamin D metabolite, and most of the 25(OH)D in the body is found in the blood stream with limited distribution into less accessible depots like fat (unlike vitamin D). Its level in blood is the best indicator of vitamin D nutritional status because of its relatively long half life in the blood stream and first-order kinetics in which the rate of 25(OH)D production is dependent on vitamin D levels. Testing for 25(OH)D levels has soared over the past several years [39] driven by the appreciation that vitamin D deficiency may be contributing to a number of disease states [1]. This has generated the need for high throughput assays done in specialized laboratories. However, the disparity of results from one laboratory to the next, or one method to another has been a problem. But now many laboratories are part of the Vitamin D External Quality Assessment Scheme in which these laboratories report their results using defined standards from the National Institute of Standards and Technology (NIST). These standards currently include known concentrations of 25(OH)D2, 25(OH)D3, 3-epi 25(OH)D3, 24S,25(OH)2D3, and 24R,25(OH)2D3[34]. There are 3 general types of assays currently in use today: competitive protein binding assay (CPBA) and immunoassays, LC-UV and LC-MS. LC-MS is becoming the gold standard and is gradually replacing the CPBA and immunoassays, although immunoassays remain the dominant method in use today [40]. However, each method has its advantages and disadvantages.
CPBA. This was the first method developed for 25(OH)D measurements, and was published in 1971 [41]. It used DBP as the binder and 3H-25(OH)D as tracer. Although the method was largely abandoned when efforts to streamline the assay by eliminating the extraction and purification procedures proved unsatisfactory, Roche Diagnostics has introduced an automated CPBA. The sample is incubated with ruthenium red labeled DBP to which 25(OH)D conjugated with biotin is then added to bind the free DBP. Streptavidin coated beads are added to bind the 25(OH)D biotin conjugate, the beads captured magnetically, and chemiluminescence induced. The concentration of 25(OH)D in the sample is inversely proportional to the chemiluminescence signal.
Immunoassays.These fall into 3 main categories.
Radioimmunoassays were introduced in the 1980s [42,43]. Acetonitrile separation of 25(OH)D from DBP simplified sample preparation. 125I-25(OH)D serves as the tracer. DiaSorin continues to offer this assay using their goat polyclonal antibody, and it correlates well with LC-MS methods. However, this antibody like others crossreacts with 24,25(OH)2D, 25,26(OH)2D, and 25(OH)D3-26,23 lactone, which unless these other metabolites are separated out (generally not done) could increase the apparent level of 25(OH)D measured. This antibody is claimed to recognize both 25(OH)D3 and 25(OH)D2 equally, but that is not always the case for other immunoassays [44].
Enzyme linked immunosorbent assays (ELISA) introduced commercially by IDS use a sheep polyclonal antibody to coat micro-titre wells, which then are incubated with a sample in which 25(OH)D has been dissociated from DBP in competition with biotin labeled 25(OH)D. After washing the wells are incubated with streptavidin conjugated with horseradish peroxidase that enables cleavage of tetramethylbenzidine to produce a chromogenic product. The amount of 25(OH)D in the sample is inversely proportional to the color formed. The IDS antibody is less efficient in detecting 25(OH)D2 than in identifying 25(OH)D3[45].
Chemiluminescent assays were first commercialized by DiaSorin, although a number of such assays are now on the market. In these assays, the antibody is bound to a solid-phase material and sample 25(OH)D competes for binding with 25(OH)D conjugated to a chemiluminescent label. After washing the light signal is induced and quantitated. The amount of light produced is inversely related to the amount of 25(OH)D in the sample.
Immunoassays share a number of advantages and problems relative to LC-MS. As will be discussed, immunoassays generally do not detect the C3-beta epimer of 25(OH)D. However, as mentioned earlier, the different antibodies used have variable ability to measure 25(OH)D2 relative to 25(OH)D3. The expedited extraction procedures do not separate 25(OH)D from other metabolites such as 24,25(OH)2D, which can be found at levels that are 10-15% of 25(OH)D [44]. Increased levels of DBP in the sample can reduce recovery [46]. These assays can show substantial bias when compared to LC-MS (i.e., they do not parallel the results obtained from LC-MS measurements over their range) [47]. This is of particular importance at the lower 25(OH)D levels when a decision to treat or not to treat depends on accurate results. However, as noted above the role of the Vitamin D External Quality Assessment Scheme program with certified NIST standards is improving the variability between laboratories and methods [48].
LC-UV. The use of HPLC for the separation of the vitamin D metabolites was developed in the 1970s [49]. Although the original columns used were silica based, reverse phase C18 columns are more widely used today. Detection is done primarily using UV at 265 nm although electrochemical detection can also be employed. This method readily separates the D3 and D2 metabolites and can separate the C3-beta epimers from their C3-alpha epimers. It is quite accurate, correlating well with LC-MS, but for metabolites circulating at concentrations well below that of 25(OH)D, unacceptably high sample volumes are required. Moreover, samples with high lipid content can alter the elution pattern, requiring careful sample preparation to avoid erroneous results [50]. Because of the large sample requirements and slow sample throughput, LC-UV is being replaced in many laboratories by LC-MS.
LC-MS. As noted earlier, LC-MS is becoming the gold standard for 25(OH)D assays, and is being developed for measurement of other vitamin D metabolites. In this discussion, I use the term LC-MS to refer to HPLC for initial separation of the vitamin D metabolites followed by tandem mass spectrometry. Tandem mass spectrometry, involving several stages of MS, although more expensive and complex than single stage mass spectrometry, is substantially more sensitive with less matrix interference (i.e., fewer interfering substances in the injected sample) [51]. MS does not distinguish the C3-beta epimer from the C3-alpha epimer of 25(OH)D, so requires a preceding chromatographic step that separates these epimers. Pentafluorophenylpropyl columns are often used for this purpose in current methods. The contribution of the C3-beta epimer to total 25(OH)D measurements (if not separated) is substantial in infants (mean level 18 nM but up to 61% of total) but also can be significant in adults (mean levels 4.3 nM, but up to 47% of total 25(OH)D [52]). Moreover, the concentration (and % of total) of the C3-beta epimer increases with vitamin D supplementation. Following chromatographic separation, the metabolites must be ionized. The vitamin D metabolites are lipophilic and so ionization can be a limit to sensitivity. As mentioned previously in discussing vitamin D measurements, atmospheric pressure photo ionization appears to be more efficient in ionizing these metabolites than APCI or ESI resulting in greater sensitivity with lower limits of detection [38,53]. Increased sensitivity can also be obtained with the use of readily ionizable derivatives such as 4-phenyl-1,2,4-triazoline-3,5 dione (PTAD). This modification enabled the measurement of 25(OH)D in saliva [54], presumably the free fraction. I will be discussing the measurement of free (i.e., non-protein bound) vitamin D metabolites subsequently, but the free concentration of 25(OH)D is well below 0.1% of total in individuals with normal DBP and albumin levels.
Although LC-MS is a very versatile method, and unlike immunoassays readily measures multiple metabolites in a single sample, it is not without problems some of which I have already discussed (e.g., inability to distinguish epimers). These include ion suppression by interfering substances (so called matrix effects) [55] and mass spectral overlaps with isobaric compounds with comparable m/z ratios (e.g., 7α-hydroxy-4 cholestene-3-one) [56]. The problem with mass spectral overlaps is in part due to the standard use of nonspecific transitions (e.g., loss of H2O) used for multiple reaction monitoring. These problems can be mitigated by the use of an internal standard, namely, deuterated 25(OH)D as a control for ionization efficiency [57], better sample preparation including an LC step to separate the epimers and potential isobars, and high resolution MS (and tandem MS) to distinguish potential spectral overlaps.
1,25(OH)2D
The measurement of 1,25(OH)2D is more challenging than that of 25(OH)D because it circulates in blood at levels that are nearly 1/1,000 that of 25(OH)D. As for the measurement of 25(OH)D, there are 3 basic methods: CPBA, immunoassay, and LC-MS. LC-UV does not have the sensitivity to measure 1,25(OH)2D so is not included in this discussion.
CPBA.The first assay for 1,25(OH)2D was a CPBA using a chicken intestinal extract containing the VDR and 3H-1,25(OH)2D as tracer [58]. The intestinal extract was subsequently replaced by calf thymus VDR [59]. These assays required substantial sample preparation by HPLC to avoid interfering substances, and have subsequently been replaced by immunoassays.
Immunoassays.The most common immunoassay in use today is the RIA available in kit form from DiaSorin and IDS among others with antibody (e.g., sheep polyclonal) and 125I-1,25(OH)2D as tracer. This assay does not require HPLC preparation of the sample and uses serum samples as standards without an internal standard to monitor recovery [60]. However, the IDS kit has been reported to have less than 100% recovery for 1,25(OH)2D2[61]. ELISA kits are also available from IDS and Immunodiagnostik. The IDS assay uses solid phase immunoextraction in terms of their RIA and colorimetric detection as described for 25(OH)D, but again may underestimate 1,25(OH)2D2. Less is known about the performance of the Immunodiagnostik kit. Recently, a fully automated chemiluminescent assay for 1,25(OH)2D has been introduced (DiaSorin Liason XL) [62]. This method uses the ligand binding domain of VDR as the capture molecule, reaction conditions favoring the binding of 1,25(OH)2D to VDR versus DBP in the sample but retaining the preferential binding of 25(OH)D, 24,25(OH)2D, and 25,26(OH)2D to DBP, and using a monoclonal antibody that selectively detects the VDR conformation induced by ligand binding. This antibody is attached to magnetic beads enabling nonbound materials to be washed away and then followed by a monoclonal antibody conjugated with a chemiluminescent label and specific for an epitope in the ligand binding domain. After washing, the chemiluminescent signal is triggered and quantitated, the strength of which is directly proportional to the amount of 1,25(OH)2D in the sample. This assay was compared to 2 LC-MS assays using immuno enrichment as a preliminary step and found to have a correlation coefficient around 0.92-0.94, a slope approximately of 1, a mean bias of 2.4-15.5%, and an intercept of approximately 2-4. In this regard, it outperformed the earlier DiaSorin immunoassay. Its LOQ was reported as 2 pg/mL using 75 microliter samples.
An important problem for all immunoassays for 1,25(OH)2D with the possible exception of the above-described chemiluminescent assay is that the polyclonal antibodies generally employed show some cross reactivity with 25(OH)D and 24,25(OH)2D, which circulate at much higher concentrations in blood than does 1,25(OH)2D [63]. This can be prevented by careful preparation of the sample to separate these metabolites from 1,25(OH)2D, but this is not always done.
LC-MS. The major problem to be overcome using LC-MS to measure 1,25(OH)2D is the limited sensitivity caused by the low circulating levels of 1,25(OH)2D and the poor ionization efficiency. The sensitivity problem can be addressed by using immunoaffinity extraction with an antibody to 1,25(OH)2D prior to HPLC and tandem mass spectrometry. This would eliminate the use of LC-MS to measure 1,25(OH)2D along with other vitamin D metabolites and does not lend itself to making this measurement in small samples. The second method is to use derivatization with polar groups such as PTAD [64] or adduct formation with ammonia [65] or lithium [66]. In these measurements, deuterated 1,25(OH)2D is used as an internal standard to calculate recovery. Combining immunoaffinity extraction with derivitazation has achieved an LOQ for 1,25(OH)2D3 of 3 pM (1.2 pg/mL) and for 1,25(OH)2D2 of 1.5 pM (0.6 pg/mL) [63]. The advantages and limitations of LC-MS over and above the sensitivity issue are similar to those described for the LC-MS measurement of 25(OH)D. However, at this point, there is no universally accepted NIST standard for 1,25(OH)2D by which labs can compare their results, or investigators and clinicians can compare one assay to another. Moreover, like 25(OH)D, the C3-beta epimer of 1,25(OH)2D is found in serum [67], and other dihydroxy vitamin D metabolites such as 23,25(OH)2D, 24,25(OH)2D, 25,26(OH)2D, and 4β,25(OH)2D (the product of CYP3A4 hydroxylation of 25[OH]D) need to be separated from 1,25(OH)2D prior to MS, as all exhibit the same molecular weight and m/z ratios. 4β,25(OH)2D was found in similar concentrations as 1,25(OH)2D in one study [68].
24,25(OH)2D
Although theoretically 24,25(OH)2D could be measured by CPBA using DBP because of its equivalent affinity for DBP compared to 25(OH)D or immunoassay using a 1,25(OH)2D antibody that cross reacts 24,25(OH)2D [69,] given its relatively high concentrations in the blood (0.7-24 nM) [70], modern assays for this metabolite use LC-MS exclusively. Frequently this is done as part of a multimetabolite profile. Like that for 25(OH)D and 1,25(OH)2D the C3-beta epimer of 24,25(OH)2D has been identified [71]. Moreover, 24,25(OH)2D exists as both the 24R,25(OH)2D and 24S,25(OH)2D epimer, but only the R epimer is biologically active [72]. Therefore, careful separation of these epimers as well as other dihydroxylated vitamin D metabolites prior to MS is required to obtain accurate results. As noted, 24,25(OH)2D measurement is often part of a multimetabolite profile. The approaches to such assays have recently been reviewed [40]. Most of these assays used ESI for ionization, triple-quadrupole instruments for MS, and nonspecific water loss transitions for monitoring. Derivitazation is often employed to increase sensitivity for the less abundant metabolites in the profile, but PTAD derivatization of 25(OH)D was found to interfere with the separation of C3-beta epi-25(OH)D from 25(OH)D [73]. Other methods of derivitazation have been developed that may circumvent this problem [74], and derivitazation may not be required if larger samples are used [75].
Free Vitamin D Metabolite Measurements
Up to this point, I have focused on measurements of total vitamin D metabolite levels. However, the free levels may be a better marker of vitamin D status. The vitamin D metabolites circulate in blood extensively bound to 2 liver-produced proteins, vitamin DBP and albumin. Approximately 85% of this binding is to DBP, 15% to albumin. Given the high affinity of these metabolites for DBP, and the abundance of DBP in normal individuals, free vitamin D metabolite levels are very low (approximately 0.03% of total for free 25(OH)D and 0.4% of total for free 1,25(OH)2D) [76,77]. However, in conditions such as liver disease (and likely nephrotic syndrome and protein losing enteropathies), DBP and albumin levels are low. The reverse is true during the 3rd trimester of pregnancy when DBP levels are increased. In these situations, the free level of 25(OH)D and 1,25(OH)2D is likely to provide a better assessment of vitamin D status than total level [76,77,78]. Prior to the development of a commercially available direct measure of free 25(OH)D, investigators have calculated the free level using the published affinity constants of 25(OH)D (and 1,25[OH]2D) for DBP and albumin. This method has several drawbacks in that it assumes that these affinity constants do not change with different physiologic states and are identical from one individual to the next. Neither assumption is valid [79,80,81]. Moreover, these calculations rely on accurate measurement of the levels of DBP and albumin. Although the measurement of albumin is standard from laboratory to laboratory, the measurement of DBP has been more problematic. DBP has a number of polymorphisms, which affect not only its affinity for the vitamin D metabolites [81], but also its measurement. A commercially available monoclonal antibody used in an ELISA from R&D Systems was found to underestimate the levels of the 1F allele of DBP, an allele common in Africans (Gambians) and African Americans, but seldom found in whites of European descent [82]. Using this ELISA resulted in substantially lower measurements of DBP levels in African Americans compared to Caucasians [83], results that were not substantiated when polyclonal antibodies were used either as ELISAs or with radial immunodiffusion [82]. When mass spectrometry was used to measure DBP, the results confirmed that DBP levels differed little on the basic of allelic differences, but measurements were substantially below that obtained with the polyclonal assay [84]. This disparity between the polyclonal immunoassays and mass spectrometry has not been resolved, but this disparity further points to the difficulty in calculating free vitamin D metabolite levels rather than measuring these levels directly. Centrifugal ultrafiltration was the first of the direct assays for the free metabolites, and provided much of the initial data on affinity constants and levels in different groups varying in DBP levels, but because of its labor intensive requirements it is no longer used. Instead, an immunoassay has been developed by Future Diagnostics that is now commercially available, and measurement of free 25(OH)D by LC-MS may soon be available.
Centrifugal Ultrafiltration.The centrifugal ultrafiltration assay consists of an inner vial capped on one end with dialysis membrane resting on filter pads at the bottom of an outer vial. The serum sample following incubation with freshly purified 3H-labeled vitamin D metabolite and 14C-labeled glucose as a marker of free water was placed in the inner vial and centrifuged at 37°C for 45 min. The ratio of 3H/14C in the ultrafiltrate to that in the sample determined the % free. The free concentration was then calculated by multiplying the % free times the total metabolite concentration [77]. This method is dependent on the purity of the labeled vitamin D metabolite requiring purification by HPLC before each assay, as degradation products could increase the fraction filtered. Also, the assembly of the ultrafiltration apparatus is labor intensive, as no suitable commercial equipment is available that does not bind to the filtered vitamin D metabolite. This assay is not commercially available.
Immunoassay.In this assay, developed by Future Diagnostics B.V., Wijchen, The Netherlands, antibodies reactive to 25(OH)D are immobilized in a microtiter well (solid phase). Standards, controls, and patient samples are added to the wells, binding to the solid-phase antibodies. The solid phase is then washed and a biotin-labeled analog of 25(OH)D is added to react with the remaining antibody in a second incubation. After washing, the wells are incubated with a streptavidin-peroxidase conjugate and bound enzyme is quantitated using a colorimetric reaction. Intensity of the signal is inversely proportional to the level of free 25(OH)D in the sample. This assay like all immunoassays is dependent on the specificity of the antibody. It is reported to modestly (70-90%) underestimate 25(OH)D2 relative to 25(OH)D3 (http://www.future-diagnostics.nl/).
The measurements of free 25(OH)D using this immunoassay are in agreement with the measurements obtained with centrifugal ultrafiltration in subjects with a large range of DBP and albumin concentrations [76,85].
LC-MS. As noted earlier, LC-MS has been used to detect 25(OH)D in saliva, which is expected to be free of DBP and albumin and so represents free 25(OH)D [54]. In this method, 1 mL of saliva was deproteinized with acetonitrile, purified using a Strata-X cartridge, derivatized with PTAD, ionized by ESI and subjected to LC-MS. The limits of detection were reported as 2 pg/mL. The range of values obtained in normal controls was between 3 and 15 pg/mL, correlating well with total serum 25(OH)D (10-30 ng/mL). The intercept was positive, but the free fraction in the mid range of the assay was approximately 0.04%, in line with the results from centrifugal ultrafiltration and the Future Diagnostics immunoassay.
Conclusions
The request for vitamin D metabolite measurements has exploded over the last several years due to the growing appreciation of the role of vitamin D in health maintenance. Much interest is shown in the measurement of 25(OH)D and to a lesser extent in the measurement of 1,25(OH)2D, but as the assays develop, requests for multiple metabolites in a single sample are increasing. Moreover, there is increased interest in the possibility that free vitamin D metabolite levels might be better markers of vitamin D status than total levels especially in individuals with altered DBP concentrations. Although immunoassays have been and remain the most prevalent assay in use today, LC-MS measurements are becoming more widespread. Each method has its advantages and disadvantages. However, as methods improve, especially for LC-MS, it is expected that LC-MS will become the more widely used method for most applications because it offers precision without the variability intrinsic to immunoassays with different antibodies. A major advance in reducing the variability between laboratories is the introduction of standards for many of the vitamin D metabolites provided by the NIST. Thus, we are approaching a time that the physician requesting these measurements from a certified laboratory can have confidence that the results are reliable in guiding clinical decision making.