conventional microcytotoxicity and all 36 were detected by the new IRA and all 106 B27^ individuals were also accurately assigned by IRA. The problem of ...
Immunol. Cell Biol. (1988)66, 215-219
Immunoradiometric assay for the rapid detection of HLA-B27 Joseph A. Trapani, Hilary A. Vaughan, Brian D. Tait* and Ian F. C. McKenzie Research Centre for Cancer and Transplantation, Department of Pathology, University of Melbourne, Parkville, Vic. 3052, and *Tissue Typing Laboratory, The Royal Me/bourne Hospital. Parkvi/le. I'ic. 3050. Australia (Submitted I December. 1987. Accepted for publication 25 February. 1988.) Summary A new method for the rapid detection ofthe HLA-B27 antigen is discussed which consists of a direct immunoradiometric assay (IRA) utilizing a '-^l-labelled. HLA-B27 specific monoclonal antibody to detect HLA-B27 in whole cilrated blood. Thus far. the assay has been used to assign HLA-B27 status in a population study involving 142 individuals; 36 subjects were HLA-B27+ by conventional microcytotoxicity and all 36 were detected by the new IRA and all 106 B27^ individuals were also accurately assigned by IRA. The problem of detection ofthe cross-reactive HLA antigen HLA-B7 was overcome by blockmg the reactive sites on HLA-B7 molecules with the HLA-B7 specific antibody, BB7 • 1. This new assay enables accurate detection of HLA-B27 within 30 min of venepuncture and should facilitate the assessment of HLA-B27 status in tissue typing laboratories.
INTRODUCTION Historically, only two methods for serological detection of HLA specificities have been popularly used since the introduction of routine tissue typing, almost two decades ago. The first of these, a leukagglutination technique (1) was replaced during the late 1960s by the more sensitive and reliable cytotoxicity assays which were developed as a micro-assay (2). Microcytotoxicity remains the standard test today as it enables the simultaneous assessment ofthe reaction of a patient's cells with a very large number of scarce HLA typing sera which can be used in minute quantities. Since the description ofthe close association of HLA-B27 with ankylosing spondylitis (AS) (3,4), a significant proportion ofthe labour of tissue typing laboratories has been directed at assessment of HLA-B27 status, regardless of the patient's full HLA phenotype. Requests for HLA-B27 typing by rheumatologists and other physicians have grown steadily despite warnings that HLAB27 status should rarely, if ever, be used as a
diagnostic criterion for AS, as 8% of the (Caucasian) population is HLA-B27+ and only a small minority of these ever develop AS (5,6). At the Royal Melbourne Hospital, for example, 45% of the 2500 requests for tissue typing during 1983 were solely for B27 status (B. D. Tait pers. comm.). HLAB27 typing has always been performed along routine tissue-typing lines, by testing a panel of anti-HLA-B27 sera on the patient's lymphocytes. This has proven costly in terms of materials (especially antisera) and escalating labour costs. Although the aim of a full panel of standardized monoclonal HLA typing sera is still a distant one, monoclonal antibodies may have much to offer in certain specific situations such as HLA-B27 typing. Accordingly, a procedure which utilizes an antiHLA-B27 monoclonal antibody has been developed, which should enable rapid, cheap and dependable assessment of HLAB27 status.
MATERIALS AND METHODS Abbreviations used in this paper: AS, ankylosing spondylitis; BSA, bovine serum aJbumin; IRA, immunoradiometric assay; ELISA, enzyme-linked immunosorbcnt assay; PBS, phosphate-buffered saline.
Cell lines Three EBV transformed lymphoblastoid coll lines produced in our laboratory were used (Table I).
216
J. A. TRAPANI iinlibodie.s
HE,VABC-m3, delects HEA-B27 (7); BB7.1 ,s an MLA-U7-specihc monoclonal antibody (8). These antibodies were used as ascites lluid. rurifualion of HL.l-.{BC-m3 bv protein A a/fmuy chromai>grnph To remove lipid material from ascites lluid prior to punhcation. an equal volume of Ficon (Aldrich Chemical ( onipans Inc., Milwaukee. WI, USA) was added, mixed, and the suspension centrifugcd. The recovered ascitcs was added dropwise to a Protein A-Sephaiose column prewashed with phosphate bullered salme (PBS). After further washing with PBS. antibody was clutcd with glycine-HCE. pH 2 S (Sigma Chemical Co.). To neutrah/e the pH of the eluate. 1 ml fractions were collected in tubes containing 5 ml oi' saturated Tris butler. The absorbance of each fraction was determined and tubes con tain ing the el uled protein retained. The column was washed with glycine-HCE, pH 2 8 until all protein was removed, t h e pooled protein sample of known concentration was diaiysed against bicarbonate butler (0-1 mol/l NaHCO3 pH 7 5) at 4°C overnight. RadiO'iodination of nioni'cloncil untihodv HEA-ABC-m3 antibody (100 ^g) was radiolabelled with '-^I (Amcrsham International)(9) with a final specific activity of approximately 1 8 pCi |ig protein: free '-^1 was removed by Sephadex G25 chromatography (Pharmacia Fine Chemicals, Uppsala, Sweden). Immunoradifynietric assays Plastic 3 ml test-tubes (Disposable Products, South Australia) were prepared and stored ready for use as follows: (i) test tubes were filled with PBS containing 1% (w. V) bovine serum albumin (PBS BSA) and incubated at room temperature for 30 min (to prevent nonspecific protein binding) then emptied by flicking: (ii) anti-HEA-B27 antibody (50|il, equivalent to 4X10^ ct/min) and blocking antibody (20 |il of BB7 I ascites fluid diluted L 10 with PBS/BSA) were then added, the tubes tightly capped and stored at 4 C (if sodium azide is added to a final concentration of 0 1% w/v, the tubes can be stored for at least one week): (iii) blood (250 |il citrated blood in triplicate) or cell suspensions to be tested (5X105 cells in 250^1 PBS/BSA) were then added to each tube and incubated at room temperature for 20 mm; (iv) the cells were washed twice with PBS/BSA, the pellet drained of all fluid and bound radioactivity was estimated.
RESULTS In developing a direct binding assay using a monoclonal antibody specific for HLAB27, it was anticipated that false positives might be encountered when testing those individuals whose cells express HLA determinants cross-reactive with HLA-B27. Previously, it had been shown in two large population studies that apart from HLAB27 + cells, only HLA-B7 + cells were able to bind significant quantities of radiolabelled anti-HLA-B27, the affinity of the antibody for this molecule being approximately onetenth of that for HLA-B27 (7). It was thus reasonable to presume thai HLA-B7 would cause the greatest problems with false positive assignment of HLA-B27 status, particularly as HLA-B7 is a commonly encountered allele in Caucasian populations and approximately 4% of these populations are HLA-B7+ homozygotes. Binding assays using '251 radiolabelled anti-HLA-B27 and a number of cell lines confirmed this suspicion. It was shown that an HLA-B7 homozygous cell line, LD-B bound almost as much anti-HLA-B27 as did the HLA-B27 heterozygous cell line. Bordin (Table 1). and that this binding persisted despite several washes. We then examined the binding of i25i_anti-HLA-B27 in a study using whole blood derived from 12 individuals; HLAB27+ heterozygotes (3), an HLA-B7+ homozygote (I), HLA-B7+ heterozygotes (4) and four individuals who lacked both specificities (Fig. 1). It was found that the HLA-B27+ heterozygous individuals bound 22-44 000 ct/min while the HLAB7+ homozygote bound 13 000 ct/min. By comparison, the HLA-B7+ heterozygotes bound
