A new sensitive and specific enzyme-linked immunosorbent assay for IgD

Journal of Immunological Methods 313 (2006) 74 – 80 www.elsevier.com/locate/jim

Research paper

A new sensitive and specific enzyme-linked immunosorbent assay for IgD David E. Mosedale a,⁎, Manjinder S. Sandhu b,c , Jian'an Luan c , Margaret Goodall d , David J. Grainger e a b

Translational Research Unit, Papworth Hospital NHS Trust, Papworth Everard, Cambridge Cambs CB3 8RE, UK Department of Public Health & Primary Care, Institute of Public Health, University of Cambridge, CB2 2SR, UK c MRC Epidemiology Unit, Institute of Public Health, University of Cambridge, UK d Division of Immunity and Infection, The Medical School, University of Birmingham, B15 2TT, UK e Department of Medicine, Box 157, Addenbrooke's Hospital, Hills Road, Cambridge CB2 2QQ, UK Received 24 June 2005; received in revised form 13 March 2006; accepted 23 March 2006 Available online 3 May 2006

Abstract We have developed a new highly specific ELISA for IgD, and then used it to measure levels of circulating IgD in the serum of 480 un-selected patients from the East Anglia region of UK. The assay is both extremely sensitive and specific, with a minimum detected IgD concentration of 30 pg/ml and more than 10,000-fold specificity for IgD over all other human immunoglobulins. The assay shows linear dilution characteristics with both purified IgD and human serum, and spiking of purified IgD into either purified immunoglobulins or human serum shows c. 100% recovery. Furthermore, intra-assay and inter-assay coefficients of variation for repeated measurements of the same samples are below 10% and 15% respectively. Measurement of IgD levels on the un-selected patient population showed levels to range from b 300 pg/ml to over 100 μg/ml, with a geometric mean of 8 μg/ml. The distribution is approximately normal after log transformation. Levels of circulating IgD were higher in men than in women. There was a significant negative correlation between levels of IgD and age in women, but not in men. Moreover, after adjustment for age and sex, there were statistically significantly higher levels of circulating IgD in male (but not female) smokers, compared to their non-smoking counterparts. These results highlight the care that needs to be taken to control for age, sex and cigarette smoking when examining levels of circulating IgD in future studies. © 2006 Elsevier B.V. All rights reserved.

1. Introduction Although IgD was first discovered in 1965 (Rowe and Fahey, 1965), relatively little is known about its function in vivo, compared with the other immunoglobulin isotypes. It is produced in two variants, a membrane-bound Abbreviations: ELISA, Enzyme-linked immunosorbent assay; Ig, Immunoglobulin; RT, Room temperature. ⁎ Corresponding author. Tel.: +44 1480 364470; fax: +44 1480 364902. E-mail address: [email protected] (D.E. Mosedale). 0022-1759/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jim.2006.03.012

form, which is a major component of the B lymphocyte Bcell receptor (Preud'homme et al., 2000) and a secretory form, whose function is largely unknown. Deletion of the gene encoding IgD in the mouse results in a phenotype that is not materially different from wild type animals (Nitschke et al., 1993; Roes and Rajewsky, 1993), suggesting that under some conditions IgD is not essential. However, IgD may, in some circumstances, be able to perform some of the functions of IgM, as deletion of the gene encoding IgM in a mouse model leads to IgD replacing membrane-bound and secretory IgM during Bcell development (Lutz et al., 1998).

D.E. Mosedale et al. / Journal of Immunological Methods 313 (2006) 74–80

Levels of secretory IgD have been measured in the circulation previously (see Preud'homme et al., 2000; Vladutiu, 2000 for reviews). Levels of circulating IgD differ throughout life, with most studies reporting low levels of IgD at birth (c. 0.2 μg/ml), rising during childhood until early adulthood, and then declining with age thereafter (Leslie et al., 1975; Zegers et al., 1975; De Greef et al., 1992). In adults, circulating IgD has been shown to be either higher in men than in women (Stoica et al., 1980), in women than in men (Leslie et al., 1975) or the same in both sexes (Dunnette et al., 1978; Levan-Petit et al., 2000). All studies agree that there is considerable variation in the levels of human IgD in normal populations (typically from 0.1–300 μg/ml), although the shape of the distribution has been variously described as unimodal (Levan-Petit et al., 2000), bimodal (Kholmogorova and Stefani, 1982) or trimodal (Dunnette et al., 1978). The reported geometric mean concentration of IgD in healthy adults also varies, from around 8 to 40 μg/ml (Peng et al., 1991; Levan-Petit et al., 2000; Preud'homme et al., 2000). Differences in circulating IgD have been found between volunteers of different ethnic backgrounds and Gm haplotypes (Litwin et al., 1985). A major environmental influence, smoking, has also been shown to be associated with increased levels of circulating IgD in one small study (Bahna et al., 1983). IgD antibodies have been shown to be specific for antigens from various infectious agents (Sewell et al., 1978; Mortensen et al., 1989), allergens (Zhang et al., 1994) and auto-antigens (Luster et al., 1976). Recently, we have shown that IgD antibodies specifically bind various carbohydrates, especially those related to the α-gal moiety (Galα1-3Galβ1-4GlcNAc) that is believed to be the major antigen acting as an obstacle for xenotransplantation (Mosedale et al., 2006). Various methods have been used to measure IgD in the studies described above, and this variation may contribute to the differences in results from one study to another. The majority of measurements have been made using radioimmuno assays, which do not readily lend themselves to be used to measure hundreds of samples. More recently, an ELISA has been described which has a sensitivity greater than most radioimmunodiffusion assays. However, the report describing this assay only showed it to be specific for IgD over other immunoglobulins by a factor of 1000. This is insufficient to guarantee specificity of the assay versus other immunoglobulins, as the normal range of human IgG concentrations in serum is up to about 20 mg/ml. We have generated and optimised a novel ELISA for IgD that is highly specific and sensitive for IgD. We validated the assay by measuring the levels of total IgD in over 400 healthy subjects from the East Anglia region of the UK,

75

one of the largest cohorts of healthy volunteers assayed for IgD. As expected from previous studies, levels of IgD differed widely within the normal population, from b30 pg/ml to N100 μg/ml. We found that the distribution of IgD was approximately log-normal. On average, levels of IgD were higher in men than in women. We also found an inverse association between IgD levels and age in women but not in men. IgD levels were also associated with smoking status in men but not in women. However, we found no statistically significant evidence to indicate that sex modifies the association between smoking and IgD levels. 2. Materials and methods 2.1. Purification of monoclonal anti-IgD from JA11 cell culture supernatant JA11 hybridoma cells were maintained in RPMI 1640 media (Sigma, Poole, UK) containing 10% foetal calf serum, 3 mM glutamine, 60 μg/ml penicillin, 100 μg/ml streptomycin and 25 μg/ml gentamycin at 37 °C in an atmosphere containing 5% CO2. Cells were subcultured when confluent, approximately every 3–4 days and reduced to 2 × 105 cells/ml. When collecting tissue culture supernatant for purification of the JA11 antibody, cells were maintained for extended periods of time in the same media — up to 14 days with regular supplementation with 1/25 volume of 625 mM HEPES, pH 7.2 containing 25% (w/v) glucose. A protein L column (Perbio Science, Helsingborg, Sweden) was equilibrated with binding buffer (100 mM phosphate containing 150 mM NaCl, pH 7.2). The tissue culture supernatant medium was diluted 1/1 (v/v) with binding buffer and applied to the column. Following extensive washing with binding buffer the bound immunoglobulin was eluted with 10 ml of 0.1 M glycine, pH 3 and immediately returned to pH 7.2 with 1.1 ml of 1 M Tris, pH 7.2. The column was regenerated by washing with 10 ml of 0.1 M glycine, pH 2.5, followed by at least 30 ml of binding buffer. The purified immunoglobulin was buffer-exchanged into phosphate-buffered saline (PBS: 8 mM Na2HPO4, 1.5 mM KH2PO4, 137 mM NaCl, 2.7 mM KCl, pH 7.2) and concentrated to 1 mg/ml using Ultrafree centrifugal concentrators (Millipore, Bedford, MA) for storage. 2.2. IgD ELISA ELISA plates (Nunc Maxisorp, Roskilde, Denmark) were coated overnight at room temperature (RT) in the dark with 0.5 μg of mouse anti-IgD (JA11; prepared as above) in 200 μl of 50 mM Na2CO3, pH 9.6. Following 3

76

D.E. Mosedale et al. / Journal of Immunological Methods 313 (2006) 74–80

quick washes with 400 μl of PBS containing 0.05% Tween-20 (wash buffer) wells were blocked with 350 μl of PBS containing 5% (w/v) sucrose and 5% (v/v) Tween-20 for 1 h at RT with shaking at 500 rpm. Following blocking, wells were washed three times quickly with wash buffer and 200 μl of the samples (run singly) and standards (run in duplicate) added to the wells. Both samples and standards were diluted in wash buffer and incubated in the wells for 2 h at RT with shaking. The standard used was IgD purified from normal human serum (not IgD myeloma protein; 16-16-090704, Athens Research and Technology). Wells were washed five times before addition of 200 μl of the detection antibody, rabbit anti-IgD (code A0093, Dako, Glostrup, Denmark) diluted to 0.5 μg/ml in wash buffer. Following a further 1 h incubation (as above) wells were washed three times with wash buffer and the secondary antibody added to the wells. Donkey anti-rabbit IgG peroxidase (711-035-152; Jackson ImmunoResearch, West Grove, PA) was diluted 1/64,000 to 6.25 ng/ml in wash buffer and incubated in the wells for 1 h at RT with shaking. Following three further washes with wash buffer 200 μl of K-blue substrate (Skybio, Wyboston, UK) was added to the wells as a colour reagent to develop the signal for approximately 5 min. Production of further colour was inhibited by the addition of 50 μl of 2 M H2SO4. The resulting colour was read in a plate reader (Molecular Devices, Sunnyvale, CA) at 450 nm. 2.3. Participants and protocol The volunteers in this study were all participants of the Ely Study, a continuing population-based cohort study in Ely, Cambridgeshire, United Kingdom. The detailed design of the study has been described previously(Wareham et al., 1999). The original sample, comprising of 1122 people without known diabetes, was recruited between 1990 and 1992 at random from a population-based sampling frame consisting of all people in Ely between 40 and 65 years of age in 1990. The initial response rate was 74%. These individuals attended a morning clinic and underwent a standard 75g oral glucose tolerance test, having fasted since 10 pm the previous evening. Blood samples were taken at fasting and 120 min during the OGTT. All samples were permanently stored at − 70 °C within 4 h. The present study is based on a subset of 480 samples taken at time 0 of the oral glucose tolerance test. 2.4. Analysis of data from Ely samples As described in the Results section, inter-assay reproducibility of the data depends upon careful storage of

a single batch of standard IgD preparation. The data presented in this work have been obtained using a single aliquot of purified human IgD standard (stock concentration 0.753 mg/ml), with all of the samples diluted 10,000-fold. However, insufficient standard from the same aliquot was available to re-assay those samples whose values were above and below the standard curve. Of the 480 samples that were assayed, 14 (2.9%) samples were above the top standard after dilution (sample concentrations above 100 ng/ml) and 102 (21%) were below the bottom standard (sample concentrations below 1.5 μg/ml). Of the low values, 63 samples (13%) were below the sensitivity limit of the assay (30 pg/ml) when diluted 1/10,000 (i.e. samples contained less than 300 ng/ml). The data described in this paper are from the remaining 364 data points, except for the histogram shown in Fig. 4, where IgD concentrations from the high and low values have been extrapolated from beyond the limits of the standard curve. Analyses of association among circulating IgD levels, age and smoking were performed using linear regression. 3. Results Prior to the generation of the IgD ELISA we tested several commercially available antibodies against IgD to identify those with high specificity for IgD over other immunoglobulin isotypes (data not shown). The best of the antibodies were then used to develop a sandwich ELISA for IgD, with a monoclonal anti-IgD as the capture reagent and a polyclonal antibody as the detection reagent (Fig. 1 and Materials and methods). The assay has a usable range from approximately 0.1 to 10 ng/ml IgD (c. 0.7 to 67 pM). The sensitivity of the assay, calculated by adding two standard deviations to repeated measurement of diluent alone, was 30 pg/ml (0.2 pM). Replacing any step of the assay with diluent alone abrogated the signal completely. We then checked the specificity of the ELISA by assaying purified preparations of other human immunoglobulins (Fig. 2). Stock preparations (1 mg/ml) of the immunoglobulins were diluted 500-fold and assayed using the IgD assay. The assay was over 10,000-fold specific for IgD over IgM, and more than 20,000-fold specific for IgD over all other human immunoglobulin classes and subclasses. Furthermore, spiking a large excess (2 μg/ml) of each of the immunoglobulin classes into a sample containing 2.5 ng/ml IgD did not affect the measurement of the IgD. The mean recovery of IgD when spiking in each of the other immunoglobulin classes was 103.1% ± 7.1% (mean ± SD). Therefore, the

D.E. Mosedale et al. / Journal of Immunological Methods 313 (2006) 74–80

1

0.1

0.01 0.1

1 IgD concentration (ng/ml)

10

Fig. 1. Standard curve of purified IgD in IgD ELISA. The standard curve was run on eight separate occasions, and the values of the lower standards normalised to the optical density of the top standard. Values shown are means ± 95% confidence intervals.

Signal as percentage of IgD

presence of other immunoglobulins does not affect the reading of IgD in this assay. We next determined the properties of the assay when using serum rather than the purified proteins used in the experiments described above. Several serum samples were used, with IgD concentrations ranging from about 3 1000 100 10 1 0.1 0.01 0.001

IgG1 IgG2 IgG3 IgG4 IgA

IgD

IgE

IgM

Immunoglobulin isotype

Immunoglobulin alone Spike of 2.5 ng/ml IgD

Fig. 2. Specificity of IgD ELISA. Values without spiked-in IgD are shown as percentage cross-reactivity with each of the other immunoglobulins, by calculation from an IgD standard curve. Spiked samples had an additional 2.5 ng/ml purified IgD added to each of the immunoglobulins, and data shown as percentage recovery versus 2.5 ng/ml IgD alone. Immunoglobulins had been shown to be pure by testing with a panel of World Health Organisation certified antibodies (Skybio, Wyboston, UK).

to 100 μg/ml. Firstly, to assess linearity of dilution, each serum sample was diluted to various extents and the IgD concentration assayed. The assay showed linear dilution characteristics of serum IgD compared with purified IgD (Fig. 3), with a mean coefficient of variation across serum samples for the different dilutions of 8.5%. Recovery of purified normal IgD spiked into serum samples was tested by spiking three different concentrations of purified human IgD into each of three different dilutions of three different serum samples. The mean recovery of the IgD was 100.8% (range 91.9% to 109.2%). These results suggest that there are no interactions between human IgD and other serum proteins that interfere with the assay. Repeated measurement of a sample over several days enabled us to calculate the intra-assay and inter-assay precision. Intra-assay precision was determined by measurement of three samples of differing IgD concentration (c. 3.5, 1.5 and 0.3 ng/ml after dilution) 20 times on a single assay. Precision was good at the two moderate concentrations of IgD (CV of 6.8% and 6.0% respectively), but not as good at the lowest measured concentration of IgD (CV 13.9%). The mean inter-assay precision, determined by measuring nine samples in triplicate on three separate occasions, was 12.4%. The significantly larger inter-assay variability than intraassay variability was due to the high dilution from stock of the IgD standard (1/75,300 dilution of stock) required. However, we have found that storage of the IgD standard at high concentrations and at temperatures of − 20 °C to be necessary to maintain recovery of the protein. The stability of the detected serum IgD was investigated by repeatedly freezing and thawing a serum sample. Measured IgD concentration (µg/ml)

Absorbance @ 450 nm (arbitrary units)

10

77

1000

100

10

1 100

1000

10000

100000

1000000

Serum dilution factor

Fig. 3. Linearity of dilution. Serum samples were diluted with PBS containing 0.05% Tween-20, as described in the Materials and methods. Each serum sample was diluted individually, the IgD concentration determined from a standard curve of purified IgD and the nominal IgD concentration calculated. Only data from samples that fell within the range of the standard curve following dilution are included in the figure.

78

D.E. Mosedale et al. / Journal of Immunological Methods 313 (2006) 74–80

60

Frequency

50 40 30 20 10 0 0

1

2

3

4

5

Log IgD concentrations (10x mg/ml) Fig. 4. Distribution of IgD in a population of healthy adults. Circulating IgD was measured in serum prepared from 480 adults using the IgD ELISA, with all samples diluted 10,000-fold. For the figure, all data were transformed by taking the logarithm to base 10. Note that in this graph some of the data shown (all values above 5 (100 μg/ml) and below 3.2 (1.56 μg/ml)) are from regions where the IgD level has been determined by extrapolation beyond the standard curve.

Circulating IgD levels (µg/ml; geometric mean)

Following each freeze-thaw cycle an aliquot of the serum was removed. The replicate samples, differing only by number of freeze-thaws were then assayed for IgD using the standard protocol. There was no significant difference in the detected levels of IgD between the samples freezethawed once and samples freeze-thawed up to eight times, with variation between freeze-thawed aliquots no greater than between replicate samples (data not shown). A comparison was made between the current assay and a commercially available radioimmuno assay for IgD. Assay of seven samples for IgD was performed using the assay described above and as per the manufacturer's instructions for the commercially available assay. The levels of IgD in the serum measured by both assays

were very similar, with a correlation coefficient between the two assays of 0.973. Having fully characterised the assay, we proceeded to measure levels of IgD in a randomly selected population from the East Anglia area of England. Subjects were recruited as described in the Materials and methods section, blood samples taken and serum prepared. All samples were then assayed using the ELISA for IgD. Values ranged from b 300 pg/ml to N 100 μg/ml, over a 105 range. Such a large range for a biological variable is unusual, but has previously been noted for levels of circulating IgD (Levan-Petit et al., 2000; Preud'homme et al., 2000). We found the distribution of the levels of IgD to be approximately normal following log transformation (Fig. 4). Previous reports have suggested that the levels of IgD were different between men and women, and varied with age (see Preud'homme et al., 2000 for a review). In our large cohort of volunteers we found levels of IgD to be higher in men than in women (after adjustment for age and smoking the geometric mean values (95% CI) were 14 093 (11 945–16 627) for men and 9620 (8337– 11101) for women, p = 0.001). IgD levels were inversely correlated with age in women (r = − 0.18, p = 0.01), but not in men (r = − 0.06, p = 0.43). However, we found no statistically significant evidence to indicate that sex modifies the association between age and IgD levels (p for interaction = 0.32). We also investigated the effect of smoking on IgD levels in men and women separately. Fig. 5 shows that circulating IgD levels were associated with smoking in men ( p = 0.03). Specifically, current smokers had higher IgD levels than ex-smokers or non-smokers. By contrast, there was no statistically significant association between smoking and IgD levels in women ( p = 0.45; Fig. 5). However, we found no statistically significant evidence to indicate that sex modifies the association between smoking and IgD levels ( p for interaction = 0.21).

30000

Non-smoker Ex-smoker Current smoker

20000

10000

0 Men

Women

Fig. 5. Circulating IgD levels in men and women classified by cigarette smoking. Data from 364 adults of known cigarette smoking status. Values shown are age-adjusted geometric means ± 95% confidence intervals.

D.E. Mosedale et al. / Journal of Immunological Methods 313 (2006) 74–80

4. Discussion We have developed a highly sensitive, extremely specific ELISA for human IgD that has several excellent characteristics, including linear dilution of serum samples, and 100% recovery of purified IgD spiked into serum. We then used the assay to assess circulating IgD levels in a large cohort of volunteers drawn from the population in the East of England. Levels of IgD differed widely, and followed a log-normal distribution. There was a small, but statistically significant, inverse correlation between age and IgD in women, but not in men, and a positive correlation between smoking and IgD in men, but not in women. Despite the differences in these interactions between men and women, when models are built incorporating either age and sex or smoking and sex with circulating IgD levels, the interactions between age and sex or smoking and sex are not statistically significant. The ELISA for IgD demonstrated in this study has all of the characteristics required of a clinical assay. It is highly sensitive (down to 30 pg/ml) yet very specific for IgD over other immunoglobulin isoforms (at least 10,000fold). It has been shown to give linear dilution characteristics from serum dilutions of 1/400 to 1/800,000. This means that the assay has a useable range (taking into account dilution factor) of at least 6 orders of magnitude. Initially, the JA11 antibody we use in the ELISA as the capture reagent was only available as an ascites preparation. Ascites fluid is generally unsuitable as a capture reagent in ELISA techniques as it contains a significant proportion of other proteins, reducing the coating efficiency of the antibody of interest. We therefore purified JA11 from the tissue culture supernatant of the hybridoma cells. However, the JA11 anti-IgD antibody has recently become available commercially in a purified form (Skybio, Wyboston, UK or University of Birmingham, Division of Immunity, UK). Although all of the experiments described above were performed using JA11 purified in our laboratory, we expect the assay to perform similarly with the commercially available product. All of the other components of the assay are also commercially available, increasing its potential utility. It has been configured such that none of the incubation steps use a buffer containing any protein (except the antibodies being used). This is relatively unusual, but reduces any possibility of exogenous immunoglobulin being introduced into the system throughout the procedure. Furthermore, all steps of the assay are performed at room temperature, eliminating the requirement for a method of keeping the ELISA plate at 37 °C. Measurement of all of the samples in this study took place over a short time period, using fresh aliquots of the

79

same standard. Unfortunately, we found that storage of the standard at high concentrations is required for its stability. This leads to the requirement for a large dilution to the top concentration used in the assay (approximately 1/100,000) and ultimately to the introduction of a higher inter-assay variation than intra-assay variation. Fortunately, the large variation from person to person makes this small increase in inter-assay variation essentially irrelevant. The 480 samples used in this study were collected over 10 years prior to assay for the determination of circulating IgD levels using the ELISA procedure. Although it has been reported that IgD is susceptible to proteolysis and has a short half life in serum in vivo (Rogentine et al., 1966), ex vivo serum IgD has been shown to be remarkably stable at 4 °C and even higher temperatures (Dunnette et al., 1977; Drenth et al., 1994; Levan-Petit et al., 2000). Since the samples used in this study were stored at − 70 °C and the levels of IgD found using this new ELISA were similar to previous studies, we conclude that there was no significant loss of IgD in the storage of the samples prior to assay. Previous studies have measured levels of circulating IgD in healthy adults, and found levels of IgD to be higher in men than in women (Stoica et al., 1980), higher in women than in men (Leslie et al., 1975) or not different between the sexes (Dunnette et al., 1978; LevanPetit et al., 2000). We believe that the current study is the largest to measure IgD levels in healthy adults, and we found levels of IgD to be higher in men than in women. Taken together with the published data, it is not clear if there are any consistent differences between the sexes. Following adjustment for smoking status, we found a small inverse correlation between IgD levels and age in women, but not in men. However, the age range of the volunteers in our study was relatively narrow (45 to 65), reducing our power to detect such correlations. Previous studies have shown levels of IgD to decrease throughout adulthood (De Greef et al., 1992). Levels of IgD have previously been shown to be elevated in volunteers who smoked compared to their non-smoking controls (Bahna et al., 1983). Moreover, in this previous report, ex-smokers were found to have levels of IgD similar to non-smokers whereas our data suggests that levels of IgD in ex-smokers are intermediate between non-smokers and current smokers. However, the previous study examined only 83 subjects, and their power to detect small changes will have been much less. Furthermore, we have preliminary evidence for a possible difference in levels of IgD between male and female smokers, which was not noted in the previous report. However, the difference between men and

80

D.E. Mosedale et al. / Journal of Immunological Methods 313 (2006) 74–80

women did not reach statistical significance and further work will be needed to clarify this observation. Acknowledgements We would like to thank Nicholas Wareham for the access to the Ely study samples, the Ely study participants and the staff of the St. Mary's Street Surgery (Ely, United Kingdom), and H. Shannasy, S. Curran, S. Hennings and J. Mitchell for the help with the fieldwork for the Ely study. We would also like to thank Sri Aitken for the technical support, Prof. Roy Jefferis for providing the JA11 hybridoma cell line and Annik Panicker for the critical reading of this manuscript. References Bahna, S.L., Heiner, D.C., Myhre, B.A., 1983. Changes in serum IgD in cigarette smokers. Clin. Exp. Immunol. 51, 624. De Greef, G.E., Van Tol, M.J., Van Den Berg, J.W., Van Staalduinen, G.J., Janssen, C.J., Radl, J., Hijmans, W., 1992. Serum immunoglobulin class and IgG subclass levels and the occurrence of homogeneous immunoglobulins during the course of ageing in humans. Mech. Ageing Dev. 66, 29. Drenth, J.P., Haagsma, C.J., van der Meer, J.W., 1994. Hyperimmunoglobulinemia D and periodic fever syndrome. The clinical spectrum in a series of 50 patients. International Hyper-IgD Study Group. Medicine (Baltimore) 73, 133. Dunnette, S.L., Gleich, G.J., Miller, R.D., Kyle, R.A., 1977. Measurement of IgD by a double antibody radioimmunoassay: demonstration of an apparent trimodal distribution of IgD levels in normal human sera. J. Immunol. 119, 1727. Dunnette, S.L., Gleich, G.J., Weinshilboum, R.M., 1978. Inheritance of low serum immunoglobulin D. J. Clin. Invest. 62, 248. Kholmogorova, G.T., Stefani, D.V., 1982. Levels of IgD in patients with rheumatoid arthritis. Allergol. Immunopathol. (Madr) 10, 211. Leslie, G.A., Lopez Correa, R.H., Holmes, J.N., 1975. Structure and biological functions of human IgD. IV. Ontogeny of human serum immunoglobulin D (IgD) as related to IgG, IgA and IgM. Int. Arch. Allergy Appl. Immunol. 49, 350. Levan-Petit, I., Cardonna, J., Garcia, M., Migeon, J., Corbi, C., Preud'homme, J.L., Lecron, J.C., 2000. Sensitive ELISA for human immunoglobulin D measurement in neonate, infant, and adult sera. Clin. Chem. 46, 876. Litwin, S.D., Tse-Eng, D., Butler Jr., V.P., Pernis, B., 1985. Studies on human low serum IgD phenotype and Gm markers. Clin. Genet. 27, 134. Luster, M.I., Leslie, G.A., Bardana, E.J., 1976. Structure and biological functions of human IgD. VII. IgD antinuclear antibodies

in sera of patients with autoimmune disorders. Int. Arch. Allergy Appl. Immunol. 52, 212. Lutz, C., Ledermann, B., Kosco-Vilbois, M.H., Ochsenbein, A.F., Zinkernagel, R.M., Kohler, G., Brombacher, F., 1998. IgD can largely substitute for loss of IgM function in B cells. Nature 393, 797. Mortensen, J., Nielsen, S.L., Sorensen, I., Andersen, H.K., 1989. Specific serum immunoglobulin D, detected by antibody capture enzyme-linked immunosorbent assay (ELISA), in cytomegalovirus infection. Clin. Exp. Immunol. 75, 234. Mosedale, D.E., Chauhan, A., Schofield, P.M., Grainger, D.J., 2006. A pattern of anti-carbohydrate antibody responses present in patients with advanced atherosclerosis. J. Immunol. Methods 309 (1−2), 182−191. Nitschke, L., Kosco, M.H., Kohler, G., Lamers, M.C., 1993. Immunoglobulin D deficient mice can mount normal immune responses to thymus-independent and -dependent antigens. Proc. Natl. Acad. Sci. U. S. A. 90, 1887. Peng, Z., Fisher, R., Adkinson Jr., N.F., 1991. Total serum IgD is increased in atopic subjects. Allergy 46, 436. Preud'homme, J.L., Petit, I., Barra, A., Morel, F., Lecron, J.C., Lelievre, E., 2000. Structural and functional properties of membrane and secreted IgD. Mol. Immunol. 37, 871. Roes, J., Rajewsky, K., 1993. Immunoglobulin D (IgD)-deficient mice reveal an auxiliary receptor function for IgD in antigen-mediated recruitment of B cells. J. Exp. Med. 177, 45. Rogentine Jr., G.N., Rowe, D.S., Bradley, J., Waldmann, T.A., Fahey, J. L., 1966. Metabolism of human immunoglobulin D (IgD). J. Clin. Invest. 45, 1467. Rowe, D.S., Fahey, J.L., 1965. A new class of human immunoglobulins. J. Exp. Med. 121, 171. Sewell, H.F., Chambers, L., Maxwell, V., Matthews, J.B., Jefferis, R., 1978. The natural antibody response to E. coli includes antibodies of the IgD class. Clin. Exp. Immunol. 31, 104. Stoica, G., Macarie, E., Michiu, V., Stoica, R.C., 1980. Biologic variation of human immunoglobulin concentration. I. Sex–age specific effects on serum levels of IgG, IgA, IgM and IgD. Med. Interne 18, 323. Vladutiu, A.O., 2000. Immunoglobulin D: properties, measurement, and clinical relevance. Clin. Diagn. Lab. Immunol. 7, 131. Wareham, N.J., Byrne, C.D., Williams, R., Day, N.E., Hales, C.N., 1999. Fasting proinsulin concentrations predict the development of type 2 diabetes. Diabetes Care 22, 262. Zegers, B.J., Stoop, J.W., Reerink-Brongers, E.E., Sander, P.C., Aalberse, R.C., Ballieux, R.E., 1975. Serum immunoglobulins in healthy children and adults. Levels of the five classes, expressed in international units per millilitre. Clin. Chim. Acta 65, 319. Zhang, M., Niehus, J., Brunnee, T., Kleine-Tebbe, J., O'Connor, A., Kunkel, G., 1994. Measurement of allergen-specific IgD and correlation with allergen-specific IgE. Scand. J. Immunol. 40, 502.