IVD manufacturers must consider various factors when choosing autoantigens for their immunoassays.
Throughout the history of immunodiagnostics, new assay designs and strategies resulted in changing the requirements for one of the key components: the antigens. For example, in thyroid autoimmunity, the move from immunofluorescence via hemagglutination to enzyme-linked immunosorbent assays (ELISA) was made possible by identifying the microsomal antigen. For the next few years, molecular biological techniques enabled the identification of thyroid peroxidase (TPO) as the molecular target; subsequently, the recombinant protein replaced the crude native antigen preparation.
Today, recombinant antigens are often the first and only choice for such standard immunoassays as ELISA, multiplex bead systems, immunodots, and microarrays for detecting many autoantibodies. However, due to the complexity of biology, recombinant technologies cannot yet meet certain antigen demands. This article will discuss all of the critical aspects in choosing the proper autoantigens for use in commercial autoimmune tests.
Background and History
Autoimmune disease is the generic term for a panel of pathological clinical conditions in which specialized white blood cells and autoantibodies that are generated by a patient's immune system attack the cells and tissues. Well-known examples include autoimmune rheumatic disorders such as systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA), and organ-specific disorders such as autoimmune thyroiditis and type 1 diabetes.
Establishing the presence of autoantibodies as an important clinical diagnostic tool for autoimmune conditions was first achieved 60 years ago when researchers described the microscopic phenomenon of LE cells.1 Today, this first antinuclear antibody (ANA) diagnostic is considered the birth of the concept of serological IVDs in systemic autoimmune diseases. Ten years later, other researchers presented the first analysis of autoimmune antibodies associated with organ-specific autoimmune diseases.2 It took another decade before indirect immunofluorescence microscopy on selected tissue slices could be applied in clinical diagnostics to identify systemic diseases on a large scale.3
Twenty-five years ago, an important advancement was made with the introduction of ELISAs for routine autoimmune diagnostics. ELISA and its adaptation for use in antibody detection led to significant improvements in the sensitivity and specificity of autoantibody determination.4,5 Even today, the ease of use, high sensitivity, and rapid sample throughput of ELISA makes it a useful and indispensable tool for autoantibody testing in clinical laboratories.6
In the early 1980s, ELISA replaced the previously routine hemagglutination assay for detecting microsomal antibodies, now known as TPO. However, it became apparent that ELISA's significantly improved resolution, discrimination capacities, and reproducible quantification could only be sufficiently used for screening clinical serum samples if the immunological quality and biochemical purity of the antigens kept up with ELISA's high analytical standards.
By 1990, classic nuclear autoantigens such as Sm, RNP-Sm (U1-snRNP complex), DNA topoisomerase I (Scl 70), SS-A/Ro, and histidyl-tRNA synthetase (Jo-1) were already being used in assays, even though they were unreliable and not reproducibly pure. Furthermore, small lot sizes did not allow for their ubiquitous availability. Since the majority of such native antigens are obtained from animal sources (e.g., calf thymus), even with satisfactory purity and quantity, their diagnostic value was limited.7
At that time, a recombinant strategy was attempted to eliminate such restrictions and difficulties, albeit only with moderate success. Even though the field of molecular biology had developed into a highly dynamic discipline, early attempts at producing recombinant antigens (usually as small fragments or fusion proteins expressed in E. coli) were unsuccessful. Unspecific background signal noise caused by contaminating E. coli components led to false-positive results.
In the early 1990s, the growing demand for defined single autoantigens that met the rising requirements in terms of quality, purity, lot-to-lot reproducibility, and consequently increased production yield and lot sizes led to a severe bottleneck in the supply of commercially available antigens. This bottleneck also increased the pressure on commercial production processes to develop innovative solutions to alleviate the problems caused by the greater demand for reliable antigens.
In 1990–1991, our former company, eLias Diagnostics (Freiburg, Germany) had difficulties producing enough high-quality human TPO from native sources that were suitable for IVDs. (Among the reasons were included growing reservations from clinical ethics committees.) Furthermore, irresolvable problems were encountered while attempting to manufacture this membrane-bound antigen recombinantly in E. coli. Any attempts, if at all practically manageable, to recombinantly express antigens in mammalian cells would be too expensive. Therefore, an alternative approach was taken, which received little attention in clinical diagnostics. A strategy was established to recombinantly produce human autoantigens using a eukaryotic transient cell expression system based on recombinant baculoviruses (Autographa californica multinuclear polyhedrosis virus) and insect cells (Spodoptera frugiperda, or Sf9).8,9
In 1992, with this Sf9-based expression system, Diarect achieved a breakthrough for TPO, which was significant because its immunological reactivity is conformation dependent.10 TPO provided the proof of principle for the suitability of the baculovirus expression system. During the next five years, several target antigens were produced, including a number of ANA/ENA antigens, which were needed for routine clinical diagnostics and ELISA test systems. At that time, Diarect reviewed the pros and cons of the various cell expression systems (e.g., E. coli, baculovirus/Sf9, mammalian and yeast cells), and whether they were appropriate for expressing diagnostically relevant targets with respect to different assay systems (see Table I).11
Applying Recombinant Autoantigens
Figure 1. (click to enlarge) Assay of five independent lots of Ro/SS-A (52 kDa) in standard ELISA tests. Antigen was coated at 1 µg/ml in PBS and blocked with StabilCoat. Sera were diluted 1/100 in StabilCoat.
Such critical assessments regarding the adequacy of various cell expression systems still hold true today. In addition to the routine cell expression systems previously used, today's technologies provide interesting alternatives by means of transient mammalian cell systems, which can produce significantly higher yields.12 However, the unrestricted suitability of such systems for clinical diagnostics should be carefully evaluated. The extensive knowledge and experience gained during the past decade have led to the following conclusions with respect to sensible and goal-oriented applications of recombinant autoantigens in routine commercial clinical diagnostics.
Figure 2. (click to enlarge) Lot-to-lot consistency of purified recombinant tTG. tTG was isolated from baculovirus-infected insect cells and purified by a two-step procedure. The lots were the result of separate expression and purification runs. Purity was assessed to be greater than 95%.
Reliability. Biotechnologically produced autoantigen targets provide reliable alternatives to native antigens, sometimes even surpassing their native counterparts (i.e., TPO), based on observations of the increasing number of test kits that use them.
Uniqueness. Several of the routinely applied autoantigens can only be produced recombinantly (e.g., centromere proteins B and A). Because of the insuperable difficulties in extracting other autoantigens from natural sources, they are predominantly or even exclusively commercially available as recombinant proteins (e.g., alanyl/threonyl-tRNA synthetases [PL-12, PL-7], formiminotransferase cyclodeaminase [LC1], cytochrome P450 2D6 [LKM1], and Ro/SS-A [52 kDa] proteins).
Lot-to-lot reproducibility of large lot sizes. The high quality of the recombinant autoantigen products contributes to the development of fine IVD products. One of the valuable features of recombinant antigens produced with sophisticated expression and purification systems is large lot sizes with high lot-to-lot consistency (see Figures 1 and 2).
Figure 3. (click to enlarge) Dot blot analyses of individual components of (a) M2, (b) ribosomal phosphoprotein, and (c) U1-snRNP.
The proteins were expressed in baculovirus-infected insect cells and purified to at least 90% homogeneity. The antigens were dotted on nitrocellulose strips and incubated with primary sera and secondary AP-antibody conjugates, followed
by development with NBT/BCIP. Data courtesy of Prof. R. L. Humbel, Luxembourg.
Analysis of complex antigens. A number of autoantigen targets can be complex in nature and consist of different single polypeptides. In order to differentiate between serological subspecies, discrete and ideally recombinant antigenic target domains should be examined. Figure 3 shows the results of analyses based on individual recombinant antigens for U1-snRNP antibodies (68/70 kDa, A, and C polypeptides), mitochondrial M2 antigen (full-length and individual polypeptides of PDC-E2, BCOADC-E2, OGDC-E2), and the ribosomal phosphoproteins (P0, P1, P2).
The significance of the increasing importance of biotechnological autoantigen targets becomes clearer when the 10-year record of clinically relevant diagnostic antigens is compared with a recent overview of recombinant autoantigens (see Table II).11 This table lists only those autoantigens that are commercially available key components because they fulfill the following requirements:
Figure 4. (click to enlarge) Purification of LC1. Baculovirus-infected insect cell lysate (CL) was applied to a nickel-loaded chelating-sepharose column, washed (W), and eluted with buffers of increasing imidazole concentration (Fr. 1–Fr. 5). Markers are as indicated. Pure LC1 elutes as a 64-kDa protein with a purity of >95% (Fr. 4).
Since multiplex systems and protein biochips/microarrays are relatively new and highly innovative, they merit further discussion on the current state of their development. Because of the steadily growing and emerging availability of multianalyte technologies, the quality of validated recombinant antigens must be not only maintained but also increased in order to fulfill every requirement in future IVD applications. The antigens should be suitable for chemical coupling using virtually any chemistry and able to be spotted on many surfaces.
Figure 5. (click to enlarge) Gel electrophoresis and Western blot analysis of three independent lots of (a) centromere protein B, (b) Jo-1, and (c) PCNA. The proteins were separated on 6–20% gradient gels and transferred to PVDF membranes. Membranes were probed with patient sera and developed with AP-conjugated antihuman IgG antibody followed by NBT/BCIP.
Most of the antigens from Diarect's research and development pipeline have been used for developing and producing array or bead-type multiplex assays ever since their use was discussed in studies about autoantigen arrays for autoantibody analysis.14,15 Microarrays and increasingly more often bead-based multiplex techniques will complement genomic and proteomic approaches for discoveries in autoimmunity. Recent studies provide a greater understanding of these new technologies.16–18
Recent Autoantigen Developments
Figure 6. (click to enlarge) A retained sample of Jo-1 produced in 2001 was reanalyzed in 2007 to ascertain stability under suggested storage conditions. Electrophoresis and purity analysis was carried out identically to analysis in 2001, and newly established dot-blot analysis was performed compared with a recently produced lot.
During the past 10 years, the number of recombinant autoantigens has not changed much. However, some important antigens are now commercially available (e.g., PCNA, SRP54, PL-7, PL-12, and Mi2). For example, Mi2 had previously been used as three overlapping fragments; it is now available as a baculovirus-expressed full-length protein. In the area of autoimmune liver diseases, the soluble liver antigen (SLA) is an interesting new addition. SLA was previously described as the liver-pancreas antigen, which was characterized as a protein component of the tRNP(Ser)Sec complex.19, 20
Figure 7. (click to enlarge) Stability of M2-antigen components upon repeated freezing and thawing. Dot-blot analyses were carried out with original protein and protein that was frozen and thawed seven times. In addition, BCOADC-E2 was incubated for 30 hours at 37°C prior to dotting. The proteins are resistant to freezing and thawing, but a decrease in activity could be detected upon prolonged incubation at a higher temperature.
The most important newly added component to the recombinant autoantigen toolbox was tissue-type transglutaminase (tTG), which revolutionized serum-based clinical diagnosis of celiac disease. The recombinant antigens that were produced in both baculovirus and E. coli were equally well suited for clinical diagnostic purposes. Variants with an implemented point-mutation that inhibited enzymatic activity possessed the best assay characteristics.21,22 The availability of this antigen was a milestone in the IVD industry, compared with 10 years ago when only immunofluorescence diagnostics of the endomysium/reticulin antibody were possible. Raw protein extracts from intestinal preparations of guinea pigs were also the only technology available for the first generation of solid-phase assays (see Figure 8).
Figure 8. (click to enlarge) Comparative tTG/IgA ELISA performance using tTG from different sources. Serum samples were selected from a reference panel and tested for anti-tTG IgA using a standard test protocol. Endomysium/IFA was considered as the reference method for interpretation of the result.
Even though it is not exactly a recombinant antigen, an epochal new achievement in the antigen family is the cyclic citrullinated peptide (CCP). CCP is a synthetic peptide-antigen with excellent diagnostic explanatory power as an early-stage prognostic marker for RA.23
A special feature of CCP antibody analysis is that its presence is highly indicative of an erosive disease. This marker protein has enormous clinical potential in selecting patients for aggressive treatment strategies. However, although using such peptide antigens is valuable in some instances, their general application does not seem probable. The prerequisite for their use is identifying all potential epitopes that are recognized by the disease-specific autoantibodies. Such identification would be difficult, especially considering the size and conformation dependence of some autoantigens.
Table I. (click to enlarge) Technologies and methods for the detection of autoantibodies.
Evolution of methods for the detection of autoantibodies emphasizing the
different degrees of complexity of each of the applied antigens.
The long-lasting development of recombinant antigens was not only marked by breakthroughs and decisive positive findings but also allowed a realistic assessment of what could be achieved by recombinant technologies and what is not yet possible. For example, expressing SmD antigen in its fully functional state has not been possible. The reason for this inadequacy is deeply rooted in a molecular peculiarity. The polypeptides D1 and D3 contain symmetrical dimethylarginine residues, which constitute a major autoepitope.24
Despite exhaustive biotechnological efforts to compensate for this deficiency by using host cells engineered for parallel expression of the corresponding methyl transferase, this problem has not yet been solved. However, a new generation of high-level expression mammalian cell lines has been reported. It is not unlikely that such optimized cell expression tools may aid in overcoming some of the unsolved examples of recombinant expression.12
Similarly for the production of proteinase 3 and myeloperoxidase (MPO), the main targets of antineutrophil cytoplasmic antibody clinical diagnostics, no real alternatives to native proteins currently exist. MPO can be extracted and purified with high purity and yield from conventional sources. Considering the highly complicated processing and hetero/homo-dimeric structure of MPO, a shift to a recombinant counterpart in the near future will not be likely. While coexpressing the mature polypeptides is possible, as has been done for Ku (p70, p80), the individual subunits do not form an adequately reactive antigen.
Table II. (click to enlarge) Recombinant human autoantigens currently available for routine autoimmune diagnostics.
The solubility of a recombinantly expressed antigen may pose some problems. Some proteins are produced in an insoluble form in E. coli (inclusion bodies). If the epitopes that are detected by the autoantibodies are linear, the proteins can be purified under denaturing conditions in the presence of urea or guanidine. If conformational epitopes are required and the protein must be present in a soluble form, more work has to be done. The simplest way is to determine expression conditions that yield higher percentages of soluble protein (e.g., temperature, medium, inducer concentrations, expression systems, etc.). If the protein remains insoluble, refolding will be required. Refolding success depends highly on the protein's tertiary structure.
With infectious-disease diagnostics (e.g., viral, bacterial, and protozoon/parasitic antigens), the patent situation often becomes unclear because of a close link to vaccination or pharmacological approaches, which prohibits more-rapid development of improved antigens. However, the majority of autoimmune targets are freely available for clinical use. This is primarily because they were already described in molecular detail as essential, housekeeping polypeptides or nucleoprotein complexes (e.g., spliceosomes) before being defined as autoantigens. If any patent claims do exist, it is often due to the fact that the use of such targets for clinical diagnostics is not as widespread as it could be (e.g., GAD65, IA2, LP/SLA), and only a small number of IVD companies may be using such antigens.
The quality of every immunoassay depends almost exclusively on the antigens. For the detection of autoantibodies, the requirements for antigens are the structure, immunologic reactivity, stability, and reproducible production of large lot sizes. Such requirements have made recombinant proteins the antigens of choice for modern assay systems. Since the introduction of such proteins into autoimmune diagnostics in the early 1990s, they have exemplified the importance of biotechnology to patients with severe autoimmune disorders.
Bodo Liedvogel, PhD,
Heinz Haubruck, PhD,
Richard Kneusel, PhD,
1. MM Hargraves, H Richmond, and R Morton, “Presentation of Two Bone Marrow Elements: The Tart Cell and the L.E. Cell,” Mayo Clinic Proceedings 23 (1948): 25–28.
2. IM Roitt, PN Campbell, and D Doniach, “The Nature of the Thyroid Auto-Antibodies Present in Patients with Hashimoto's Thyroiditis (Lymphadenoid Goitre),” The Biochemical Journal 69 (1958): 248–256.
3. NM Muna, JL Verner, and DF Hammond, “Fluorescent Antibody Technique as a Routine Procedure in the Diagnosis of Lupus Erythematosus Using Stored Tissue Culture Cells,” American Journal of Clinical Pathology 45 (1966): 117.
4. S Avrameas, “Coupling of Enzymes to Proteins with Glutaraldehyde: Use of the Conjugates for the Detection of Antigens and Antibodies,” Immunochemistry 6 (1969): 43–52.
5. E Engvall and P Perlmann, “Enzyme-Linked Immunosorbent Assay (ELISA): Quantitative Assay of Immunoglobulin G,” Immunochemistry 8 (1971): 871–874.
6. M Gripenberg and P Kurki, “Demonstration of Human Autoantibodies by Quantitative Enzyme Immunoassays,” Journal of Immunological Methods 92 (1986): 145–159.
7. PSC Leung and ME Gershwin, “Native Autoantigens versus Recombinant Autoantigens,” in Autoantibodies, ed. Y Shoenfeld, ME Gershwin, and PL Meroni (Amsterdam: Elsevier, 2007): 37–45.
8. YA Luckow and MD Summers, “Trends in the Development of Baculovirus Expression Vectors,” Biotechnology 6 (1988): 47–55.
9. LK Miller, “Baculovirus as Gene Expression Vectors,” Annual Review of Microbiology 42 (1988): 177–199.
10. H Haubruck et al., “Expression of Recombinant Human Thyroid Peroxidase by the Baculovirus System and Its Use in ELISA Screening for Diagnosis of Autoimmune Thyroid Disease,” Autoimmunity 15 (1993): 275–284.
11. S Kellermann et al., “Recombinant Antigens in Autoimmune Diagnostics,” Inmunologia 17 (1998): 37–59.
12. L Baldi et al., “Recombinant Protein Production by Large-Scale Transient Gene Expression in Mammalian Cells: State of the Art and Future Perspectives,” Biotechnology Letters 29 (2007): 677–684.
13. E Hochuli et al., “Genetic Approach to Facilitate Purification of Recombinant Proteins with a Novel Metal Chelate Adsorbent,” Biotechnology 11 (1988): 1321–1325.
14. WH Robinson et al., “Autoantigen Micro-arrays for Multiplex Characterization of Autoantibody Responses,” Nature Medicine 8 (2002): 295–301.
15. W Hueber et al., “Autoantibody Profiling for the Study and Treatment of Autoimmune Disease,” Arthritis Research 4 (2002): 290–295.
16. CG Fathman et al., “An Array of Possibilities for the Study of Autoimmunity,” Nature 435 (2005): 605–611.
17. SR Binder et al., “Computer-Assisted Pattern Recognition of Autoantibody Results,” Clinical and Diagnostic Laboratory Immunology 12 (2005): 1353–1357.
18. SR Binder, C Hixson, and J Glossenger, “Protein Arrays and Pattern Recognition: New Tools to Assist in the Identification and Management of Autoimmune Disease,” Autoimmunity Reviews 5 (2006): 234–241.
19. I Wies et al., “Identification of Target Antigen for SLA/LP Autoantibodies in Autoimmune Hepatitis,” Lancet 355 (2000): 1475–1476.
20. M Costa et al., “Isolation and Characterization of cDNA Encoding the Antigenic Protein of the Human tRNP(Ser)Sec Complex Recognized by Autoantibodies from Patients with Type-1 Autoimmune Hepatitis,” Clinical and Experimental Immunology 121 (2000): 364–374.
21. B Van Meensel et al., “Diagnostic Accuracy of Ten Second-Generation (Human) Tissue Transglutaminase Antibody Assays in Celiac Disease,” Clinical Chemistry 50 (2004): 2125–2135.
22. PG Hill and SA McMillan, “Anti-Tissue Transglutaminase Antibodies and Their Role in the Investigation of Celiac Disease,” Annals of Clinical Biochemistry 43 (2006): 105–117.
23. ER Vossenaar and WJ van Venrooij, “Anti-CCP Antibodies, a Specific Marker for (Early) Rheumatoid Arthritis,” Clinical and Applied Immunology Review 4 (2004): 239–262.
24. H Brahms et al., “The C-Terminal RG Dipeptide Repeats of the Spliceosomal Sm Proteins D1 and D3 Contain Symmetrical Dimethylarginines, which Form a Major B-cell Epitope for Anti-Sm Autoantibodies,” Journal of Biological Chemistry 275 (2000): 17122–17129.