Tissue Antigens ISSN 0001-2815

Evaluation of a competitive enzyme-linked immunosorbent assay for measurements of soluble HLA-G protein M. Rasmussen1,† , M. Dahl1,† , S. Buus2 , S. Djurisic1 , J. Ohlsson2 & T. V. F. Hviid1 1 Department of Clinical Biochemistry, Centre for Immune Regulation and Reproductive Immunology (CIRRI), Copenhagen University Hospital (Roskilde) and Roskilde Hospital, Roskilde, Denmark 2 Department of International Health, Immunology and Microbiology, The Faculty of Health Sciences, The Panum Institute, Copenhagen, Denmark

Key words competitive enzyme-linked immunosorbent assay; human leukocyte antigen; human leukocyte antigen class Ib; soluble human leukocyte antigen-G Correspondence Thomas Vauvert F. Hviid, MD, PhD, DMSc, Associate Professor Department of Clinical Biochemistry Centre for Immune Regulation and Reproductive Immunology (CIRRI) Copenhagen University Hospital (Roskilde) and Roskilde Hospital 7-13 Køgevej DK-4000 Roskilde Denmark Tel: +45-4732-5622 Fax: +45-4632-1615 e-mail: [email protected] or [email protected] Received 28 October 2013; revised 24 March 2014; accepted 25 March 2014 doi: 10.1111/tan.12357

Abstract The human leukocyte antigen (HLA) class Ib molecule, HLA-G, has gained increased attention because of its assumed important role in immune regulation. The HLA-G protein exists in several soluble isoforms. Most important are the actively secreted HLA-G5 full-length isoform generated by alternative splicing retaining intron 4 with a premature stop codon, and the cleavage of full-length membrane-bound HLA-G1 from the cell surface, so-called soluble HLA-G1 (sHLA-G1). A specific and sensitive immunoassay for measurements of soluble HLA-G is mandatory for conceivable routine testing and research projects. We report a novel method, a competitive immunoassay, for measuring HLA-G5/sHLA-G1 in biological fluids. The sHLA-G immunoassay is based upon a competitive enzyme-linked immunosorbent assay (ELISA) principle. It includes a recombinant sHLA-G1 protein in complex with β2-microglobulin and a peptide as a standard, biotinylated recombinant sHLA-G1 as an indicator, and the MEM-G/9 anti-HLA-G monoclonal antibody (mAb) as the capture antibody. The specificity and sensitivity of the assay were evaluated. Testing with different recombinant HLA class I proteins and different anti-HLA class I mAbs showed that the sHLA-G immunoassay was highly specific. Optimal combinations of competitor sHLA-G1 and capture mAb concentrations were determined. Two versions of the assay were tested. One with a relatively wide dynamic range from 3.1 to 100.0 ng/ml, and another more sensitive version ranging from 1.6 to 12.5 ng/ml. An intra-assay coefficient of variation (CV) of 15.5% at 88 ng/ml and an inter-assay CV of 23.1% at 39 ng/ml were determined. An assay based on the competitive sHLA-G ELISA may be important for measurements of sHLA-G proteins in several conditions: assisted reproduction, organ transplantation, cancer, and certain pregnancy complications, both in research studies and possibly in the future also for clinical routine use.

Introduction

The human major histocompatibility complex (MHC) includes human leukocyte antigen (HLA) class Ia and HLA class Ib genes. The latter consists of HLA-E, -F, and -G. Increasing attention has been given to the HLA-G gene and protein and its role in immune modulation during pregnancy and in relation to pregnancy complications such as preeclampsia and recurrent spontaneous abortions (1–5). HLA-G may also play an important role in in vitro fertilization (IVF), as well as in organ transplantation, autoimmune disease, cancer, asthma, and septic shock (6–11). By alternative splicing of HLA-G

† These

two authors contributed equally to this work.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

mRNA transcripts, seven HLA-G isoforms are generated: four membrane-bound isoforms (-G1 to -G4) and three soluble isoforms (-G5 to -G7) (12–14). The full-length soluble HLA-G protein is actively secreted, while soluble HLA-G1 (sHLA-G1) occurs after cleavage from the cell surface by the activity of metalloproteinases (15, 16). HLA-G5 and sHLA-G1 are associated with β2-microglobulin (β2m). Furthermore, soluble HLA-G exists in dimeric forms (17, 18). A number of studies indicate that HLA-G has a range of immunomodulatory functions. Through interactions with Ig-like transcript 2 (ILT-2) killing inhibitory receptor, ILT-4, and the HLA-G-specific Killer Ig-like receptor 2 (KIR2DL4), HLA-G has been shown to inhibit natural killer (NK)-mediated cell lysis (19–21). However, the interaction of the KIR2DL4 receptor and HLA-G 1

Competitive ELISA for sHLA-G

has been disputed in a recent study (22). The expression of HLA-G can modulate secretion of cytokines in favor of a Th2 response (23, 24). Also, there might be functional links between sHLA-G and the expansion of FoxP3-positive regulatory T cells (Tregs) (25). As an increasing number of reports supports a role for sHLA-G as a potential biomarker in assisted reproduction, in certain complications of pregnancy, in various cancer forms, and in organ transplantation (5, 8, 26–28), accessibility of well-characterized, specific, and sensitive sHLA-G immunoassays are of great importance. Here, we report a new approach for an immunoassay that can measure HLA-G5/sHLA-G1 protein. The aim of developing the assay is to be able to measure accurate concentrations of HLA-G5/sHLA-G1 in biological fluids, in high as well as low amounts, as great individual differences in concentrations have been reported. The sHLA-G immunoassay is based upon a competitive enzyme-linked immunosorbent assay (ELISA) principle. It includes a recombinant sHLA-G1 protein in complex with β2m and a peptide as a standard, and a commercially available monoclonal antibody (mAb) as the capture antibody. The method is based on the principle of competitive inhibition. First, samples containing unknown concentrations of sHLA-G are allowed binding to an HLA-G-specific mAb coating the wells of a microtiter plate. Subsequently, a recombinant HLA-G protein coupled to biotin is added that will attach to the remaining vacant mAb in the well. After incubation, a streptavidin-horse radish peroxidase (HRP) conjugate binding biotin is added, followed by the addition of HRP-substrate, which will result in a quantitative chemiluminescence signal. In a competitive ELISA, in contrast to the more widely used sandwich-ELISA-method, there is an inverse relationship between the raw signal obtained and the concentration of the analyte in the sample, i.e. the more analyte (sHLA-G), the lower the signal. This should favour accurate measurements of low sHLA-G concentrations. The specificity and the sensitivity of the sHLA-G competitive immunoassay were evaluated in various ways.

Materials and methods Samples

Blood samples were obtained from healthy donors in EDTA-sampling tubes and centrifuged for 10 min at 4000 g to separate plasma from the blood cells, and plasma samples were stored at −80∘ C until use. Conditioned cell culture medium from the choriocarcinoma cell line JEG-3 was obtained by establishing and maintaining JEG-3 cell cultures in Eagle’s Minimum Essential Medium with 10% fetal bovine serum (FBS) as described by the recommendations from ATCC (ATCC, Manassas, VA). Furthermore, conditioned cell culture medium from the human melanoma cell line IGR-1 that expresses HLA-G was obtained by establishing and 2

M. Rasmussen et al.

maintaining IGR-1 cell cultures in Dulbecco’s Modified Eagle’s medium with 10% FBS as described by the supplier [Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ), Braunschweig, Germany] (29). Preparation of biotinylated and unbiotinylated recombinant soluble HLA-G1

Recombinant HLA-G molecules were produced as previously described (30). Briefly, DNA sequences encoding the outer domains (excluding exons 5, 6, and 8, encoding the transmembrane and intracytosolic domains) of HLA-G1 were generated synthetically (GenScript, Piscataway, NJ), inserted into the Escherichia coli expression vector, pET28a (MerckMillipore, Hellerup, Denmark), and DNA sequenced (ABI 3100, Life Technologies, Carlsbad, CA). The validated recombinant vectors were cloned into an E. coli cell line BL21 (DE3) containing the BirA with pACYC184 plasmid and grown in a labscale fermentor (Labfors; INFORS, Laurel, MD). To maintain the pET28 and pACYC184-derived plasmids, the media was supplemented with kanamycin (50 μg/ml) and chloroamphenicol (20 μg/ml), respectively, throughout the expression cultures. To induce HLA-G and BirA protein expression, IPTG (1 mM) was added, the induction media was supplemented with biotin (Sigma #B4501, St. Louis, MO, 125 μg/ml), and the culture was continued for an additional 3 h at 42∘ C. Samples were analyzed by reducing sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) before and after IPTG induction. At the end of the induction culture, protease inhibitor (PMSF, 80 μg/l) was added, and cells were lysed in a cell disrupter (Constant Cell Disruptor Systems set at 2300 bar) and the released inclusion bodies were isolated by centrifugation (Sorval RC6, 20 min, 17,000 g). The inclusion bodies were washed twice in PBS, 0.5% NP-40 (Sigma), and 0.1% deoxycholic acid (Sigma) and extracted into a urea–Tris buffer (8 M urea, 25 mM Tris, pH 8.0), and any contaminating DNA was precipitated with streptomycin sulfate (1% w/v). The dissolved MHC-I proteins were purified by Ni2 +/IDA metal chelating affinity column chromatography followed by Q-Sepharose ion exchange column chromatography, hydrophobic interaction chromatography, and eventually by Superdex-200 size exclusion chromatography. Fractions containing MHC-I heavy chain molecules were identified by A280 absorbance and SDS-PAGE and pooled. Throughout purification and storage, the MHC-I heavy chain proteins were dissolved in 8 M urea to keep them denatured. Note that the MHC-I heavy chain proteins at no time were exposed to reducing conditions. This allowed purification of highly active pre-oxidized moieties as previously described (29). Protein concentrations were determined by bicinchoninic acid assay. The degree of biotinylation (usually >95%) was determined by a gel-shift assay (30). The pre-oxidized, denatured proteins were stored at −20∘ C in Tris-buffered 8 M urea. Recombinant human β2m was expressed and purified as described elsewhere (29). © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

M. Rasmussen et al.

Competitive ELISA for sHLA-G

Luminometer

Chemiluminescense detection (LumiGLO Reserve) Biotinylated recombinant soluble HLA-G1

Streptavidin-HRP biotin

biotin

Coating

MEM-G/9 mAb (mouse): Conformatory β2-microglobulin-associated epitope

β2-microglobulin Soluble HLA-G in sample

Figure 1 The principle of the competitive enzyme-linked immunosorbent assay (ELISA) method for measurement of soluble HLA-G in biological fluids.

To generate peptide-HLA-G complexes, denatured biotinylated recombinant HLA-G heavy chains were diluted into a renaturation buffer containing β2m and the HLA-G binding peptide and incubated at 18∘ C for 48 h allowing equilibrium to be reached. The competitive soluble HLA-G enzyme-linked immunosorbent assay

The overall principle of the competitive sHLA-G ELISA is shown schematically in Figure 1. In general, the competitive ELISA procedure includes the following steps: (i) Coating: Nunc polystyrene Maxisorp plates with 96 wells (cat. no: 436110, Nunc, Roskilde, Denmark) were coated with 100 μl MEM-G/9 mAb (0.5 μg/ml) (cat. no: 11-292-M001, EXBIO Praha, a.s., Vestec, Czech Republic) or control mAb diluted in coating buffer B (cat. no: CB01100, Invitrogen, Taastrup, Denmark), 100 μl in total, at 4∘ C overnight for later capture of sHLA-G in the samples and recombinant biotinylated sHLA-G1; or recombinant unbiotinylated sHLA-G1 and recombinant biotinylated sHLA-G1 for the standard curves. MEM-G/9 binds HLA-G5 and sHLA-G1 in complex with β2m. (ii) Blocking: The plates were washed three times with 300 μl wash buffer (cat. no: WB01, Invitrogen), after which, unoccupied sites were saturated or blocked with 300 μl buffer containing bovine protein (cat. no: DS98200, Invitrogen, Denmark) and incubated for 1 h at room temperature, then washed three times with 300 μl wash buffer (cat. no: WB01, Invitrogen, Denmark). (iii) Samples, standards, and controls: Samples, or recombinant unbiotinylated sHLA-G1 for the standard curves, were added to the wells diluted in bovine protein-containing buffer (Assay Buffer) to obtain a concentration twice the desired © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

final concentration in the respective well, 50 μl was added to each well, followed by incubation at 300 rpm for 15 min at room temperature allowing binding of sHLA-G in the samples to the mAb. (iv) Biotinylated recombinant sHLA-G1: The biotinylated recombinant sHLA-G1 was diluted in Assay Buffer to obtain a concentration of twice the desired concentration in the well; 50 μl of the recombinant biotinylated sHLA-G1 (50 ng/ml) was added to each well in order to obtain a final total reaction volume of 100 μl per well. (v) Incubation, washing, and detection: Plates were incubated at 300 rpm for 2 h at room temperature, then washed three times with wash buffer. Thereafter, 100 μl of Streptavidin-HRP conjugate (0.5 μg/ml) (cat. no: SNN2004, Invitrogen, Denmark) was added to the wells and incubated at 300 rpm for 1 h at room temperature. Following five washes with wash solution containing Tween-20 (cat. no: 50-63-04, KPL), 100 μl of LumiGLO chemiluminescent substrate was added to each well (cat. no: 54-61-01, KPL, Gaithersburg, MD), and the chemiluminescence signal was measured on a Modulus Microplate luminometer (Promega Turner BioSystems, Madison, WI). If not otherwise stated this was the standard protocol for the results presented in the figures and in the text. Determination of the optimal combination of concentrations of monoclonal antibody and biotinylated recombinant sHLA-G1 together with the generation of a standard curve with unbiotinylated sHLA-G1

In order to obtain a high sensitivity in the competitive ELISA, it is very important that concentrations of the analyte, the biotinylated recombinant protein, and the capturing mAb are 3

M. Rasmussen et al.

Competitive ELISA for sHLA-G

Assessment of the specificity of the monoclonal antibody and determination of coefficient of variation

To assess the specificity of the capture MEM-G/9 mAb (mouse IgG1), for sHLA-G, a number of different experiments were performed substituting MEM-G/9 with either an isotype mAb, MEM-E/8 (recognizes HLA-E; mouse IgG1), or W6/32 (recognizes HLA class Ia and HLA-G; mouse IgG2a). In other control experiments, biotinylated recombinant sHLA-G1 was substituted with either biotinylated recombinant HLA-A, HLA-B, or HLA-C, respectively, in combination with either MEM-G/9 mAb or W6/32 mAb. In one experiment, a full standard curve was tested with biotinylated recombinant HLA-A instead of 4

biotinylated recombinant sHLA-G1. Furthermore, the ability of human blood plasma to outcompete the biotinylated recombinant sHLA-G1 in the assay was tested in a number of titration experiments, in which various amounts of blood plasma were added to the biotinylated sHLA-G1 before analysis. One EDTA blood plasma sample diluted 1:10 was measured in 15 replicates for calculation of the intra-assay coefficient of variation (CVs; ratio of the standard deviation to the mean) for the competitive sHLA-G immunoassay. In another experiment, the same 39 individual blood plasma samples were measured on two independent ELISA plates to calculate an interassay CV.

Standard curve

(A)

50 ng MEM-G/9 and 2.5 ng biotinylated recombinant soluble HLA-G1 per well

100

B/B0 (%)

80 60 40 20 0 0

20 40 60 80 100 Recombinant soluble HLA-G1 (ng/mL)

Standard curve

(B)

50 ng MEM-G/9 and 2.5 ng biotinylated recombinant soluble HLA-G1 per well

100 80 B/B0 (%)

adjusted to avoid saturation of the capturing mAb, assuring equal competition between the analyte and the biotinylated recombinant protein. Also of importance is the need to keep within the optimal measuring range of the detection system. In this experiment, the optimal concentrations were initially calculated on the basis of the molar quantities of physiologically relevant sHLA-G concentrations, equimolar amounts of biotinylated recombinant sHLA-G1, and equimolar number of binding sites of the coating mAb. To obtain the optimal standard curves in two versions of the sHLA-G competitive immunoassay, cross-titration of biotinylated recombinant sHLA-G1 and capture MEM-G/9 mAb was performed in duplicate and in independent experiments. (A) First version of the assay that generated a relatively wide dynamic range for measurements of sHLA-G concentrations in, e.g. diluted blood plasma samples and (B) another version is that to explore the sensitivity of the competitive sHLA-G assay. In version (A), biotinylated sHLA-G1 in the concentrations 50 ng/ml (2.5 ng per well) and 100 ng/ml (5.0 ng per well) was cross-titrated with MEM-G/9 mAb in the concentrations 0.50 μg/ml (50 ng per well), 0.37 μg/ml (37 ng per well), and 0.25 μg/ml (25 ng per well). A twofold titration standard curve with unbiotinylated recombinant sHLA-G1 concentrations of 100.0, 50.0, 25.0, 12.5, 6.3, 3.1, and 1.6 ng/ml in Assay Buffer (final concentration in well) was added to a series of wells in duplicate in a volume of 100 μl for each well in these experiments. In version (B), cross-titration of biotinylated sHLA-G1 in the concentrations 0–100.0 ng/ml (0–5 ng per well; twofold titration) and MEM-G/9 mAb in the concentrations 62.5 ng/ml (6.25 ng per well) or 125.0 ng/ml (12.50 ng per well) was evaluated. In this modified version of the ELISA, sHLA-G was replaced with 50 μl Assay Buffer. To determine whether a detection system with a wider measuring range will enable a higher sensitivity to be achieved, the sHLA-G standard curve was tested using low concentrations of MEM-G/9 (6.25 ng per well) and biotinylated recombinant sHLA-G (0.79 μg/ml), and replacing LumiGlo with 30 μl per well of Quantum Western Bright (cat. no: K12042-C20). A standard curve in the range of 1.6–12.5 ng/ml was included in these experiments.

60 40 20 0 0.0 0.5 1.0 1.5 2.0 LOG(Recombinant soluble HLA-G1) (ng/mL)

Figure 2 Standard curve for the competitive soluble HLA-G1 ELISA (enzyme-linked immunosorbent assay). (A) The chemiluminescence signals are scaled to set B0 = 100%. (B) Standard curve with log10 transformation of the standard concentrations and sigmoidal dose–response with variable slope curve fitting for the interpolation of unknowns. Each data point is the mean of three replicates.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

M. Rasmussen et al.

Controls for the monoclonal capture antibody

Specificity of the MEM-G/9 anti-HLA-G monoclonal antibody 2.0×10 8

2.0×107 1.5×107 1.0×107 5.0×106

Chemiluminisence signal (RLU)

Chemiluminescence signal (RLU)

2.5×10

7

Competitive ELISA for sHLA-G

1.5×10 8 1.0×10 8 5.0×10 7

Monoclonal antibody used for coating and capture of soluble HLA-G1/HLA-G5

Figure 3 Controls for the coating and capture monoclonal antibody MEM-G/9 [RLU, relative light units; mean ± SEM of four replicates; concentration of MEM-G/9: 0.5 μg/ml, and concentration of biotinylated recombinant sHLA-G1 (competitor): 50 ng/ml].

Results Standard curve

The cross-titration experiments showed that a combination of 50 ng of the coating MEM-G/9 mAb and 2.5 ng biotinylated recombinant sHLA-G1 per well was the optimum for version (A) of the assay with a dynamic range from 3.1 to 100.0 ng/ml sHLA-G (Figure 2). Monoclonal antibody specificity

To test for the specificity of HLA-G binding to the MEM-G/9 antibody, plate wells were coated with either MEM-G/9, an IgG1 mAb negative isotype control produced against a synthetic hapten not present in humans or animals, or MEM-E/08 specific for HLA-E. Furthermore, some wells were not coated with antibody at all, and in some wells coated with MEM-G/9 no biotinylated recombinant sHLA-G1 was added. As shown in Figure 3, only the wells coated with MEM-G/9 produced a signal. To test further for cross-reactivity with other HLA molecules, experiments were performed, where wells were coated with MEM-G/9 or W6/32 (an anti-HLA-class I mAb reacting with HLA-A, -B, -C, and -G), respectively, and added one of the biotinylated HLA class I recombinant proteins: HLA-A, HLA-B, HLA-C, or HLA-G. The results are shown in Figure 4. In the wells coated with MEM-G/9, only HLA-G produced a © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

0

W 6/ 32 ,b W io 6/ 32 tinH ,b LA W io -A 6/ 32 tinH W ,b LA 6/ io 32 -B , b tinM H E M iot LA -C -G in-s H /9 M LA EM , b io -G tin G1 /9 M H EM , b LA io -A -G tin M /9 EM H ,b LA -G io -B /9 , b tinH io LA tin -C -s H LA -G 1

M

EM -G /9 ,n N o o bi M ot EM in -s -G H / Is 9, LA ot b -G yp io 1 t e in m -s A H b, LA bi -G ot M 1 EM in -s -E H /8 LA ,b -G io 1 M tin EM -s -G H LA /9 ,b -G io 1 tin -s H LA -G 1

0

Combination of anti-HLA class I monoclonal antibody and biotinylated recombinant HLA class I-protein Figure 4 Specificity of the capture monoclonal antibody, MEM-G/9 [RLU = relative light units; mean ± SEM of four replicates, for HLA-G two replicates; 5 ng recombinant HLA class I protein per well; mAb concentration: 0.5 μg/ml (50 ng per well)].

positive signal; whereas positive signals for all three HLA class Ia proteins and HLA-G were observed in wells coated with W6/32, confirming the binding of this antibody to an antigenic determinant common to HLA class I molecules in complex with β2m. To test for the specificity of the inhibition, plates were coated with MEM-G/9, and varying amounts of unbiotinylated recombinant sHLA-G1 were added to the same amount of either biotinylated sHLA-G1 or biotinylated soluble HLA-A. The results are shown in Figure 5. Inhibition of the signal is observed only in wells where recombinant HLA-G is added confirming the specificity. Exploration of the sensitivity of the competitive sHLA-G ELISA principle

It was possible to discriminate between chemiluminescence signals in wells, when 0 or 0.79 ng/ml recombinant biotinylated sHLA-G1, respectively, was added when using the Western Bright Quantum chemiluminescence system (Figure 6A). Furthermore, in version (B) of the assay, a standard curve in the range of 1.6–12.5 ng/ml could be obtained (Figure 6B, C). Analyses of blood plasma samples and conditioned cell culture media

On the basis of repeated measurements on blood plasma samples, the intra-assay CV was determined to 15.5% (at 88 ng/ml) and the inter-assay CV to 23.1% (at 39 ng/ml). 5

M. Rasmussen et al.

Competitive ELISA for sHLA-G

1.5×107

5 ng/well sHLA-G1 competitor

1.0×107

5 ng/well HLA-A competitor

5.0×106

0 3. 13 6. ng/ 25 m 12 ng L .5 /m 0 25 ng L .0 /m 0 50 ng L .0 /m 10 0 n L 0 . g/ 00 m ng L /m L 3. 13 n 6. g/ 25 m 12 ng L .5 /m 0 25 ng L .0 /m 0 50 ng L .0 /m 10 0 n L 0. g/ 00 m ng L /m L

Chemiluminescence signal (RLU)

Specificity of the HLA class I competitor Standard curves

Concentration of recombinant soluble HLA-G1

Figure 5 Specificity of the competitor recombinant biotinylated HLA class I protein and the standard curve. Recombinant biotinylated soluble HLA-G1 protein versus recombinant biotinylated HLA-A protein (RLU, relative light units; mean ± SEM of two replicates).

In titration experiments of human blood plasma added to recombinant biotinylated sHLA-G1 the signal was inhibited, showing that native sHLA-G competes with biotinylated sHLA-G1 for binding to the MEM-G/9 mAb (Figure 7A). Also, a titration of FBS is shown. A minor matrix effect may be observed but titration of human plasma shows the same titration curve as recombinant sHLA-G1 does. Finally, to further test that the assay detects native soluble HLA-G, twofold titrations were made of conditioned cell culture media from two human cell lines that express HLA-G, JEG-3, and IGR-1 (Figure 7B, C). The concentration of sHLA-G in JEG-3 conditioned medium was 25 ng/ml, which is in the same range as reported in the literature (16). Discussion

Accurate and sensitive measurements of sHLA-G are important because of increasing focus on this specific HLA class Ib molecule for its role in immunomodulatory mechanisms and pathways. Here, we present a new approach with the use of a competitive sHLA-G ELISA principle. The method is based on the principle of inhibition as presented and described in section Introduction. The specific concentrations of sHLA-G1/HLA-G5 in the samples can be determined from dilutions of a standard unbiotinylated recombinant HLA-G protein by the generation of a standard curve. The principle is summarized in Figure 1. Furthermore, in a competitive ELISA or immunoassay, there is an inverse relationship between the assay signal obtained and the concentration of the analyte to be measured in the sample material. The more the analyte, in this case sHLA-G1/HLA-G5 protein, the lower the signal. Theoretically, competitive ELISA 6

formats may provide a very accurate quantitation of the analyte. So, the competitive sHLA-G ELISA principle might have some advantages over the sandwich-ELISA principle used so far in all published in-house sHLA-G assays (27, 31–36). In this study we did not determine a precise limit of detection but the experimental results obtained in this first evaluation of the competitive sHLA-G assay support that the sensitivity may be below 1.0 ng/ml sHLA-G1/HLA-G5. However, it might be possible by further development of the competitive sHLA-G immunoassay principle to construct an ultra-sensitive assay for measuring sHLA-G in the range of 0.5–5.0 ng/ml accurately. This may be of special interest for measuring sHLA-G in conditioned media from embryo cultures, where the embryos are transferred to women undergoing IVF treatment. Several studies have shown a positive association between detection of sHLA-G in the media and successful IVFtreatment (37). Several publications have reported in-house assays for measuring sHLA-G in various biological fluids. To our knowledge, all assays have been based on a sandwich-mAb principle. As capture mAb, a mAb against sHLA-G1 and HLA-G5 in complex with β2m, typically MEM-G/9 but other mAbs not commercially available have been used; or rarely a mAb against HLA-G5 (typically the 5A6G7 mAb). As detection mAb, a mAb against β2m has typically been the choice, or the W6/32 mAb (27, 33, 35). Most researchers have used the ELISA format; however, a few have developed beads-based methods using fluorescence detection (38, 39). An increase in sensitivity has been obtained with fluorescence beads-based methods. For the ELISA format, typically colorimetric detection has been the method of choice; however, a few variants have used chemiluminescence or fluorescence with no obvious increase of sensitivity (35, 36). To our knowledge, we have for the first time systematically tested the MEM-G/9 mAb in an ELISA format for the specificity toward sHLA-G1 in contrast to soluble HLA class Ia molecules. We found a high degree of specificity (Figures 4 and 5). However, several reports have previously been published evaluating the specificity of the MEM-G/9 mAb, both for use in ELISAs and in flow cytometry experiments (31, 33, 40, 41). MEM-G/9 reacts with native HLA-G but not with denatured molecules. The MEM-G/9 mAb was made on the basis of extracellular domains of HLA-G renatured in the presence of β2m and peptides (31, 40). One study reported that MEM-G/9 does not react with the HLA-G2, -G3, and -G4 isoforms (40); however, this has been questioned by a recent study that reported binding of MEM-G/9 to HLA-G3 in K562-G3 transductants (41). This challenges the general view that MEM-G/9 only recognizes HLA-G in association with β2m as HLA-G3 lacks the α2 and α3 domains involved in binding of β2m. So, and although it is controversial, it cannot be ruled out on the basis of the existing literature that MEM-G/9 to some degree may bind to free HLA-G heavy chain (HC) without β2m. Then, the competitive ELISA presented in the current study may © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

M. Rasmussen et al.

Competitive ELISA for sHLA-G

Chemiluminescence signal (RLU)

(A) Sensitivity of binding of recombinant biotinylated sHLA-G1 to the MEM-G/9 anti-HLA-G monoclonal antibody 12.50 ng MEM-G/9 per well

1.5×10 9

6.25 ng MEM-G/9 per well 1.0×10 9

5.0×10 8

0. 0 n 7 g 1. 9 n /m 56 g L 3. n /m 1 g L 6. 3 n /m 12 25 g/mL . n 25 50 g/mL .0 ng L 5 0 0 /m .0 ng L 0 /m ng L /m 0 0. n L 79 g 1. n /m 5 g L 3. 6 n /m 13 g L 6. n /m 12 25 g/mL . n 25 50 g/mL . n 50 0 0 g /m L . 0 ng L 0 /m ng L /m L

0

Concentration of biotinylated recombinant soluble HLA-G1

(B)

Standard curve

(C)

Standard curve 6.25 ng MEM-G/9 per well and 0.79 ng/mL biotinylated recombinant soluble HLA-G1

100

100

80

80

B/B0 (%)

B/B0 (%)

6.25 ng MEM-G/9 per well and 0.79 ng/mL biotinylated recombinant soluble HLA-G1

60 40 20

60 40 20

0 0

5 10 15 Recombinant soluble HLA-G1 (ng/mL)

0 0.0 0.5 1.0 1.5 LOG(Recombinant soluble HLA-G1) (ng/mL)

Figure 6 Exploration of the sensitivity of the competitive sHLA-G ELISA concept. Western Bright Quantum was used as chemiluminescence substrate in these experiments. (A) Twofold dilution of recombinant biotinylated sHLA-G1 protein in wells coated with two different concentrations of the anti-HLA-G monoclonal antibody MEM-G/9. The signal is above background at the lowest concentration of sHLA-G1 corresponding to 0.79 ng/ml. (B, C) Standard curves in the range of 1.6–12.5 ng/ml sHLA-G1. The chemiluminescence signals are scaled to set B0 = 100%. Combination of 6.25 ng MEM-G/9 per well and 0.79 ng/ml recombinant biotinylated sHLA-G1 (RLU = relative light units; mean ± SEM of two replicates).

have the potential to measure a slightly different sHLA-G pool than the commercial Exbio assay (EXBIO Praha, a.s., Vestec, Czech Republic), and similar in-house assays, as an anti-β2m mAb is used for detection in these assays. However, this would also mean that ELISAs as the Exbio assay may bind sHLA-G molecules that are not detected leading to a lower signal in samples with free sHLA-G heavy chains compared with samples with only β2m-associated sHLA-G. The current approach based upon a competitive ELISA can also be used for the implementation of an immunoassay specific for HLA-G5 and -G6, e.g. with the use of the 5A6G7 mAb raised against a peptide corresponding to the last C-terminal 22 amino acid sequence specific for the HLA-G5 and -G6 isoforms (EXBIO Praha, a.s., Vestec, Czech Republic). © 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

A main problem with the only currently commercial sHLA-G kit available manufactured by EXBIO Praha is that the quantitative measurements are based upon a manufacturer-specific sHLA-G standard calibrator that is in arbitrary ‘units’. No international and independent laboratory or organization supervises the standardization and calibration of this ‘unit’ making comparisons of studies and specific results difficult. In conclusion, this new approach presented in the current study based upon a competitive sHLA-G1/HLA-G5 ELISA principle, including a recombinant sHLA-G calibrator protein and with a validation of the binding specificities of the used mAb (MEM-G/9) in relation to other HLA class I molecules, may be an interesting alternative to existing ELISAs for use in basic research and diagnostic procedures related to measuring 7

M. Rasmussen et al.

Competitive ELISA for sHLA-G

(A)

Two-fold titration of human blood plasma and of fetal bovine serum

Chemiluminescence signal (RLU)

1.5×10 7

1.0×10 7

5.0×10 6

50 . 25 0 μ . L 12 0 μ .5 L 6. μL 3 3. μL 13 1. μL 6 μ 50 L . 25 0 μ . L 12 0 μ .5 L 6. μL 3 3. μL 13 1. μL 6 μL

0

Human blood plasma (B)

Fetal bovine serum Two-fold titration of recombinant sHLA-G1 (same experiment as JEG3 medium)

Two-fold titration of JEG3 conditioned cell culture medium

2.0×10 0 6

Chemiluminescence signal (RLU)

Chemiluminescence signal (RLU)

2.0×10 0 6 1.5×10 0 6 1000000 500000

1.5×10 0 6 1000000 500000

0

10

0. 0

ng 50 /m L .0 n 25 g/m .0 L ng 1 2 /m L .5 ng / 6. mL 3n g/ 3. mL 1n g/ m L

μL

μL

1. 6

μL

μL

3. 1

6. 3

μL 12

.5

.0 25

50

.0

μL

0

Two-fold titration of IGR-1 conditioned cell culture medium and of medium only

(C)

Chemiluminescence signal (RLU)

1.5×10 7

1.0×10 7

5.0×10 6

50 . 25 0 μ . L 12 0 μ .5 L 6. μL 3 3. μL 1 1. μL 6 μ 50 L . 25 0 μ . L 12 0 μ .5 L 6. μL 3 3. μL 1 1. μL 6 μL

0

IGR-1 medium/10%FBS Only medium/10%FBS Figure 7 Twofold titration experiments of (A) human blood plasma and of fetal bovine serum, (B) conditioned cell culture medium from the human choriocarcinoma cell line JEG-3, and (C) conditioned cell culture medium from the human melanoma cell line IGR-1 (RLU, relative light units; mean ± SEM of two replicates).

8

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

M. Rasmussen et al.

sHLA-G in such diverse areas as assisted reproduction, pregnancy complications, organ transplantation, and cancer. Acknowledgments

The authors would like to thank Anja Steenfatt Torbensen for careful technical assistance. Support for this work was generously provided by grants from Region Zealand Health Sciences Research Foundation. JO was supported by a scholarship from The Danish Council for Independent Research (Medical Sciences). Conflict of interest

The authors have declared no conflicting interests. References 1. Geraghty DE, Koller BH, Orr HT. A human major histocompatibility complex class I gene that encodes a protein with a shortened cytoplasmic segment. Proc Natl Acad Sci U S A 1987: 84: 9145–9. 2. Kovats S, Main EK, Librach C, Stubblebine M, Fisher SJ, DeMars R. A class I antigen, HLA-G, expressed in human trophoblasts. Science 1990: 248: 220–3. 3. Hviid TV. HLA-G in human reproduction: aspects of genetics, function and pregnancy complications. Hum Reprod Update 2006: 12: 209–32. 4. Gonzalez A, Rebmann V, LeMaoult J, Horn PA, Carosella ED, Alegre E. The immunosuppressive molecule HLA-G and its clinical implications. Crit Rev Clin Lab Sci 2012: 49: 63–84. 5. Dahl M, Hviid TV. Human leucocyte antigen class Ib molecules in pregnancy success and early pregnancy loss. Hum Reprod Update 2012: 18: 92–109. 6. Fuzzi B, Rizzo R, Criscuoli L et al. HLA-G expression in early embryos is a fundamental prerequisite for the obtainment of pregnancy. Eur J Immunol 2002: 32: 311–5. 7. Lila N, Carpentier A, Amrein C, Khalil-Daher I, Dausset J, Carosella ED. Implication of HLA-G molecule in heart-graft acceptance. Lancet 2000: 355: 2138. 8. Deschaseaux F, Delgado D, Pistoia V, Giuliani M, Morandi F, Durrbach A. HLA-G in organ transplantation: towards clinical applications. Cell Mol Life Sci 2011: 68: 397–404. 9. Rizzo R, Farina I, Bortolotti D et al. HLA-G may predict the disease course in patients with early rheumatoid arthritis. Hum Immunol 2013: 74: 425–32. 10. Nicolae D, Cox NJ, Lester LA et al. Fine mapping and positional candidate studies identify HLA-G as an asthma susceptibility gene on chromosome 6p21. Am J Hum Genet 2005: 76: 349–57. 11. Monneret G, Voirin N, Krawice-Radanne I et al. Soluble human leukocyte antigen-G5 in septic shock: marked and persisting elevation as a predictor of survival. Crit Care Med 2007: 35: 1942–7. 12. Ishitani A, Geraghty DE. Alternative splicing of HLA-G transcripts yields proteins with primary structures resembling both class I and class II antigens. Proc Natl Acad Sci U S A 1992: 89: 3947–51.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

Competitive ELISA for sHLA-G

13. Kirszenbaum M, Moreau P, Gluckman E, Dausset J, Carosella E. An alternatively spliced form of HLA-G mRNA in human trophoblasts and evidence for the presence of HLA-G transcript in adult lymphocytes. Proc Natl Acad Sci U S A 1994: 91: 4209–13. 14. Hviid TV, Moller C, Sorensen S, Morling N. Co-dominant expression of the HLA-G gene and various forms of alternatively spliced HLA-G mRNA in human first trimester trophoblast. Hum Immunol 1998: 59: 87–98. 15. Park GM, Lee S, Park B et al. Soluble HLA-G generated by proteolytic shedding inhibits NK-mediated cell lysis. Biochem Biophys Res Commun 2004: 313: 606–11. 16. Rizzo R, Trentini A, Bortolotti D et al. Matrix metalloproteinase-2 (MMP-2) generates soluble HLA-G1 by cell surface proteolytic shedding. Mol Cell Biochem 2013: 381: 243–55. 17. Boyson JE, Erskine R, Whitman MC et al. Disulfide bond-mediated dimerization of HLA-G on the cell surface. Proc Natl Acad Sci U S A 2002: 99: 16180–5. 18. Shiroishi M, Kuroki K, Ose T et al. Efficient leukocyte Ig-like receptor signaling and crystal structure of disulfide-linked HLA-G dimer. J Biol Chem 2006: 281: 10439–47. 19. Navarro F, Llano M, Bellon T, Colonna M, Geraghty DE, Lopez-Botet M. The ILT2(LIR1) and CD94/NKG2A NK cell receptors respectively recognize HLA-G1 and HLA-E molecules co-expressed on target cells. Eur J Immunol 1999: 29: 277–83. 20. Ponte M, Cantoni C, Biassoni R et al. Inhibitory receptors sensing HLA-G1 molecules in pregnancy: decidua-associated natural killer cells express LIR-1 and CD94/NKG2A and acquire p49, an HLA-G1-specific receptor. Proc Natl Acad Sci U S A 1999: 96: 5674–9. 21. Rajagopalan S, Long EO. A human histocompatibility leukocyte antigen (HLA)-G-specific receptor expressed on all natural killer cells. J Exp Med 1999: 189: 1093–100. 22. Le Page ME, Goodridge JP, John E, Christiansen FT, Witt CS. Killer Ig-like receptor 2DL4 does not mediate NK cell IFN-gamma responses to soluble HLA-G preparations. J Immunol 2014: 192: 732–40. 23. Kapasi K, Albert SE, Yie S, Zavazava N, Librach CL. HLA-G has a concentration-dependent effect on the generation of an allo-CTL response. Immunology 2000: 101: 191–200. 24. McIntire RH, Morales PJ, Petroff MG, Colonna M, Hunt JS. Recombinant HLA-G5 and -G6 drive U937 myelomonocytic cell production of TGF-beta1. J Leukoc Biol 2004: 76: 1220–8. 25. Selmani Z, Naji A, Zidi I et al. Human leukocyte antigen-G5 secretion by human mesenchymal stem cells is required to suppress T lymphocyte and natural killer function and to induce CD4+CD25highFOXP3+ regulatory T cells. Stem Cells 2008: 26: 212–22. 26. Larsen MH, Hviid TV. Human leukocyte antigen-G polymorphism in relation to expression, function, and disease. Hum Immunol 2009: 70: 1026–34. 27. Rizzo R, Andersen AS, Lassen MR et al. Soluble human leukocyte antigen-G isoforms in maternal plasma in early and late pregnancy. Am J Reprod Immunol 2009: 62: 320–38. 28. Rouas-Freiss N, Moreau P, Ferrone S, Carosella ED. HLA-G proteins in cancer: do they provide tumor cells with an escape mechanism? Cancer Res 2005: 65: 10139–44.

9

Competitive ELISA for sHLA-G

29. Paul P, Rouas-Freiss N, Khalil-Daher I et al. HLA-G expression in melanoma: A way for tumor cells to excape from immunosurveillance. Proc Natl Acad Sci U S A 1998: 95: 4510–5. 30. Pedersen LE, Harndahl M, Rasmussen M et al. Porcine major histocompatibility complex (MHC) class I molecules and analysis of their peptide-binding specificities. Immunogenetics 2011: 63: 821–34. 31. Fournel S, Huc X, Aguerre-Girr M et al. Comparative reactivity of different HLA-G monoclonal antibodies to soluble HLA-G molecules. Tissue Antigens 2000: 55: 510–8. 32. Yie SM, Taylor RN, Librach C. Low plasma HLA-G protein concentrations in early gestation indicate the development of preeclampsia later in pregnancy. Am J Obstet Gynecol 2005: 193: 204–8. 33. Rebmann V, LeMaoult J, Rouas-Freiss N, Carosella ED, Grosse-Wilde H. Quantification and identification of soluble HLA-G isoforms. Tissue Antigens 2007: 69 (Suppl 1): 143–9. 34. Rudstein-Svetlicky N, Loewenthal R, Horejsi V, Gazit E. HLA-G levels in serum and plasma. Tissue Antigens 2007: 69 (Suppl 1): 140–2. 35. Sageshima N, Shobu T, Awai K et al. Soluble HLA-G is absent from human embryo cultures: a reassessment of sHLA-G detection methods. J Reprod Immunol 2007: 75: 11–22.

10

M. Rasmussen et al.

36. Tabiasco J, Perrier d’Hauterive S, Thonon F et al. Soluble HLA-G in IVF/ICSI embryo culture supernatants does not always predict implantation success: a multicentre study. Reprod Biomed Online 2009: 18: 374–81. 37. Vercammen MJ, Verloes A, Van de Velde H, Haentjens P. Accuracy of soluble human leukocyte antigen-G for predicting pregnancy among women undergoing infertility treatment: meta-analysis. Hum Reprod Update 2008: 14: 209–18. 38. Rizzo R, Dal Canto MB, Stignani M et al. Production of sHLA-G molecules by in vitro matured cumulus-oocyte complex. Int J Mol Med 2009: 24: 523–30. 39. Rebmann V, Switala M, Eue I, Schwahn E, Merzenich M, Grosse-Wilde H. Rapid evaluation of soluble HLA-G levels in supernatants of in vitro fertilized embryos. Hum Immunol 2007: 68: 251–8. 40. Menier C, Saez B, Horejsi V et al. Characterization of monoclonal antibodies recognizing HLA-G or HLA-E: new tools to analyze the expression of nonclassical HLA class I molecules. Hum Immunol 2003: 64: 315–26. 41. Teklemariam T, Zhao L, Hantash BM. Full-length HLA-G1 and truncated HLA-G3 differentially increase HLA-E surface localization. Hum Immunol 2012: 73: 898–905.

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd