Use of Chimeric Antibodies as Positive Controls in an Enzyme-Linked Immunosorbent Assay for Diagnosis of Scrub Typhus (Infection by Orientia tsutsugamushi)

The use of human sera collected from individuals of known infectedand noninfected status is necessary for the validation of diagnosticassays and for the determination of cutoff values. However,the routine inclusion of pooled human sera from infected individualsfor use as positive controls in commercial assay kits has manydisadvantages. Sufficient quantities of sera can be difficultto obtain, and there are ethical and safety issues to be considered.Additionally, each batch of control material requires standardization,as each will differ in antibody titer. We have genetically engineeredchimeric immunoglobulin G (IgG), IgM, and IgA antibodies consistingof mouse-derived variable regions and human constant regionsderived from peripheral blood lymphocytes. The chimeric natureof these antibodies allows the desired antigen specificity createdthrough mouse immunization and hybridoma technology while retaininga human constant region required for recognition by the enzyme-conjugatedantihuman signal antibody. We have investigated the potentialuse of chimeric IgG with specificity for the major surface antigenof Orientia tsutsugamushi as an alternative positive controlfor inclusion in a commercial enzyme-linked immunosorbent assaykit for the diagnosis of rickettsia scrub typhus (caused byinfection with O. tsutsugamushi). Chimeric IgG was expressedin stably transfected CHO cells, allowing production of unlimitedquantities. The purified protein was found to have a much greaterspecificity for the scrub typhus antigen than the serum-derivedcontrols. The methods described could be applied to other assaykits for the detection of antibodies against infectious agents.

INTRODUCTION

Top

INTRODUCTION

Enzyme-linked immunosorbent assay (ELISA) kits for the detectionof antibodies against infectious agents use pooled human seraas reference material to establish cutoff values and to routinelyconfirm test integrity (3). The sera are then included in commercialkits as positive and cutoff calibrator controls. Sera are collectedfrom individuals with confirmed cases of specific infectiousdiseases and pooled, but the routine use of this material hasseveral disadvantages. Material for rare or exotic diseasescan be difficult to obtain, particularly if immunoglobulin M(IgM) is required, as high titers occur only early in diseaseprogression, often before diagnosis. Ethical problems arisefrom the collection of blood from individuals who are unwell,from children, or in societies where blood collection is unacceptable.There are safety concerns dealing with potentially infectioushuman sera, and each finite batch of material collected requiresstandardization due to differences in antibody titer (2).

An alternative, non-serum-derived positive control that overcomesthese disadvantages would be greatly beneficial for routineinclusion in commercial ELISA kits. Such a reagent must fulfilltwo essential criteria: it must specifically bind to a particularantigen, and it must contain epitopes recognized by the enzyme-conjugatedantihuman antibody used to generate the assay end point.

Chimeric antibodies containing variable regions cloned fromspecific mouse hybridomas and constant regions obtained fromhuman peripheral blood lymphocytes satisfy these requirements.Chimeric antibodies were shown to be useful positive controlsin an ELISA for Toxoplasma gondii (2). We have furthered thisconcept using our chimerization techniques reported previously(4) and show here its application in an ELISA for the diagnosisof rickettsia scrub typhus.

Rickettsia scrub typhus results from infection with Orientiatsutsugamushi, a bacterium transmitted to humans by bites fromLeptotrombidium mite larvae. Patients with this infection presentwith the nonspecific symptoms of fever, headache, muscle aches,and rash, while the characteristic eschars (scabbing) are notseen in all patients (7). Thus, the disease is often misdiagnosedin patients with diseases with similar symptoms and is incorrectlytreated with penicillin, to which O. tsutsugamushi is resistant,allowing progression of the disease to more serious conditions.Therefore, assays which specifically diagnose scrub typhus arevery useful, and an antibody ELISA is a practical format foruse in regions where facilities and trained personnel may belimited.

O. tsutsugamushi expresses a 56-kDa protein in its cell envelopewhich is recognized by antibodies present in the serum of patientsand of experimentally immunized animals (6). A recombinant versionof this protein (r56) (1) is used as the capture antigen inthe commercially available rickettsia scrub typhus group IgGand group IgM ELISA kits (Panbio, Brisbane, Australia). Therefore,the chimeric antibody that we have tested in this assay containsr56-specific mouse-derived variable regions and human constantregions for recognition by the enzyme-conjugated signal antibody.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

DNA constructs for chimeric antibodies. A mouse-human chimeric IgG construct was produced as describedpreviously (4). Briefly, variable heavy and kappa chain cDNAswere amplified by reverse transcription-PCR from RNA isolatedfrom mouse hybridoma cells secreting antibody against r56. Similarly,human constant-region heavy and kappa chain cDNAs were amplifiedfrom peripheral blood lymphocyte mRNA isolated from a humanblood sample. The mouse and human fragments were chimerizedby splice overlap extension PCR (4) and were cloned into a bicistronicexpression vector, pBudCE4.1 (Invitrogen, Melbourne, Australia).This vector was transfected into CHO-S cells, a derivative ofthe CHO cell line adapted to serum-free growth (Invitrogen).A stable cell line secreting chimeric IgG was produced by selectionwith 400 µg/ml phleomycin (Zeocin; Invitrogen).

Purification of chimeric IgG. To harvest chimeric IgG for purification, four flasks of 30-mlstable cell line cultures at 2 x 10⁵ cells/ml were incubatedwith shaking for 7 days at 37°C. The supernatants were collectedafter centrifugation at 250 x g for 3 min at ambient temperatureand were then passed through a 0.45-µm-pore-size filter.Purification of chimeric IgG was performed with a 1-ml HiTrapprotein G Sepharose column (Amersham Biosciences, Sydney, Australia),according to the manufacturer's instructions.

Quantitation of chimeric IgG. The purified IgG preparation was quantitated by an in-houseELISA. Microtiter plates coated with a goat anti-human IgG antibodywere obtained from a dengue virus capture IgG ELISA kit (Panbio).A standard curve was produced for each assay run by using knownconcentrations of purified IgG from human serum (Sigma, St.Louis, MO) as the standard. Standards and samples were detectedby using horseradish peroxidase-conjugated sheep anti-humanIgG and tetramethylbenzidine substrate (both obtained from therickettsia scrub typhus group IgG ELISA kit [Panbio]). Afterquantitation, the purified preparation was diluted twofold incontrol serum diluent (CSD; Panbio) to enhance its stability.

Titration of chimeric IgG. The chimeric IgG preparation was diluted in CSD to produce aseries of concentrations down to 0.1 µg/ml. These wereassayed, following the manufacturer's instructions, by usinga rickettsia scrub typhus group IgG ELISA kit (Panbio), whichuses immobilized r56 antigen to capture anti-r56 antibodies.Detection uses horseradish peroxidase-conjugated sheep anti-humanIgG and tetramethylbenzidine. The assay results were comparedwith those obtained with the kit's positive and cutoff calibratorcontrols. The purpose was to obtain preparations of the chimericIgG which return results equivalent to those for the serum-derivedcontrols, which could then be used as alternative controls.

Serial dilutions of both the serum-derived controls and thealternative chimeric controls were then prepared in CSD andassayed by using the rickettsia scrub typhus group IgG ELISAkit to determine if the dilutions of a non-serum-based preparationbehave the same as dilutions of the serum reagents in the assay.

Cross-reactivity of serum standards and chimeric IgG. The positive control preparation of chimeric IgG was testedfor its specificity toward the r56 antigen. Antigen-coated platesand their corresponding serum-derived controls were obtainedfrom commercial ELISA kits used for the diagnosis of severalinfectious diseases (Panbio). Each plate was incubated withits corresponding control, with the rickettsia scrub typhusserum-derived controls, and with the chimeric anti-r56 IgG.Each control and the chimeric IgG were processed according tothe ELISA kit instructions, and assay results were obtainedas absorbance values at 450 nm.

RESULTS

Top

RESULTS

Purification and titration of chimeric IgG. Purification of chimeric IgG from culture supernatants yielded0.5 ml of 50.8 µg/ml antibody in the peak fraction. Thiswas diluted to 25.4 µg/ml in CSD. Serial dilutions ofthe purified chimeric IgG were assayed by the rickettsia scrubtyphus ELISA. The absorbance results were plotted against thelogarithm of the antibody concentration, and the data were fittedto a sigmoidal dose-response curve by using Prism (version 3.00)software for Windows (GraphPad Software, San Diego, CA) (Fig.1). The equation for the curve was used to determine that 22.8µg/ml and 5.0 µg/ml of the purified chimeric IgGgave absorbance results equivalent to those for the serum-derivedpositive and cutoff calibrator controls, respectively. Theseconcentrations of chimeric IgG were then used for subsequentexperiments.

View larger version (13K):
[in this window]
[in a new window]

FIG. 1. Dilution series of purified chimeric IgG assayed by the rickettsia scrub typhus group IgG ELISA (Panbio) to determine concentrations which give assay results equivalent to those for the kit's serum-derived positive and cutoff calibrator controls. These were calculated by using the equation for the sigmoidal dose-response curve and the A₄₅₀ values of the kit controls.

Figure 2 shows the parallelism between the dilution curves forthe chimeric IgG controls (positive at 22.8 µg/ml andthe cutoff calibrator at 5.0-µg/ml starting concentrations)and the serum-derived controls by the rickettsia scrub typhusgroup IgG ELISA. The dilution curves were each fitted to a sigmoidaldose-response curve, and a t test analysis with GraphPad Prismsoftware revealed no significant difference in the curve parametersbetween the serum-derived and the chimeric positive controlsor between the serum-derived and chimeric cutoff calibrators.

View larger version (14K):
[in this window]
[in a new window]

FIG. 2. Dilution curves for rickettsia scrub typhus group IgG ELISA results for the current serum-derived kit controls and the alternative chimeric IgG controls. The chimeric IgG was prepared at initial concentrations which return ELISA results equivalent to those for the serum-derived positive and cutoff calibrator controls.

Cross-reactivity of serum standards and chimeric IgG. The cross-reactivity of the rickettsia scrub typhus kit controlsand the chimeric IgG at the positive control concentration weretested against antigens for several other infectious diseases.Table 1 shows the results for each of these assays, expressedas the ratios of the average absorbance to the average absorbanceof each cutoff calibrator control. The r56 serum-derived controlscross-reacted with the antigens of dengue virus, Epstein-Barrvirus, human herpesvirus 6, Leptospira, varicella-zoster virus,and West Nile virus, such that the A₄₅₀ result was greater thanor close to the result for the specific cutoff calibrator control,resulting in equivocal or positive ratios. For all antigenstested (except r56), the chimeric anti-r56 IgG showed no cross-reactivity.

View this table:
[in this window]
[in a new window]

TABLE 1. Cross-reactivities of the rickettsia scrub typhus group IgG ELISA controls and the r56-specific chimeric IgG with the immobilized capture antigens of several other infectious disease ELISA kits

DISCUSSION

Top

DISCUSSION

Serum-derived reference material must be used during the developmentand validation of a diagnostic assay for an infectious disease.Reference material is required to determine cutoff values forthe assay which will optimize the balance between assay sensitivityand specificity (3). Serum-derived material is also necessaryfor the continued monitoring of the assay performance. For example,the cutoff values may need to be optimized for different populationsor may drift with time. However, the controls included in theassay kit for distribution to the end user need not be serumderived, as long as the alternative performs equivalently inthe assay.

We have determined the concentrations of chimeric mouse-humananti-r56 IgG which gave assay results equivalent to those forthe serum-derived positive control and the cutoff calibratorand have shown that dilutions of these preparations made inCSD behave identically to the serum-derived reagents in theELISA. This validates the use of purified recombinant chimericantibodies as alternative positive control reagents, even thoughthe matrix is not serum and is therefore different from thesample matrix. Additionally, we have studied the degradationof the chimeric IgG at elevated temperatures and predict thatthe preparation should remain stable for the shelf-life of anELISA kit (nominally 15 months at 4°C for the rickettsiascrub typhus group IgG ELISA) (data not shown) (5).

Recombinant chimeric antibodies would be superior alternativesto serum-derived positive and cutoff calibrator controls forroutine inclusion in ELISA kits, as they eliminate many of thedisadvantages associated with serum-derived controls. The supplyof serum-derived controls is finite, and so new batches mustbe obtained regularly from regions of the world where the diseaseis endemic. The chimeric antibodies, however, are expressedfrom a stable cell line, allowing consistent and continuousproduction. This also avoids the ethical problems associatedwith the collection of serum from individuals who are unwelland safety issues involving the handling of potentially infectiousmaterial.

The serum-derived controls from the rickettsia scrub typhusgroup IgG ELISA kit were found to cross-react with antigensfrom several other infectious agents. The serum is collectedfrom several individuals with known O. tsutsugamushi infectionand pooled; therefore, the batch contains a mixed populationof antibodies reflecting the immune status of each individual.Therefore, it is conceivable that the positive control serumcontains antibodies against many other antigens. This cross-reactivitywill differ between each batch of control material collected.This is not a problem for the performance of each individualassay, as the controls are intended for use only with the correspondingkit. However, in a clinical laboratory that is processing severaldifferent assays at high throughput, the potential use of theincorrect standard may not be detected if it cross-reacts withthe antigen. This will cause errors in the assay cutoff values,potentially leading to false-positive or false-negative results.The chimeric anti-r56 IgG is a monoclonal preparation whichshowed no cross-reactivity with the other antigens tested. Therefore,the chimeric antibody is a more reliable, safe, and consistentpositive control.

By using the same principles described here, chimeric antibodieswith specificity toward many different antigens can be created,making the procedures described in this study applicable tothe production of alternative controls for other commercialELISA kits. The human constant regions are cloned and readyfor chimerization to specific variable-region cDNA, which canbe isolated from hybridomas secreting specific antibodies, ifthey are available. The manufacturers of diagnostic kits maywish to outsource the creation of chimeric antibody DNA constructsand stable cell lines. The chimeric antibodies can then be expressedand purified in-house, providing a constant supply of consistentcontrol reagents. Thus, the initial high cost of creating theantibodies will be offset by eliminating the high costs andrisks associated with sourcing, transporting, and standardizingserum-derived reagents.

ACKNOWLEDGMENTS

We thank Panbio for supplying the ELISA kits, standards, andreagents used in this study.

Portions of this study were jointly supported financially byan Australian Research Council Linkage grant and Panbio.

FOOTNOTES

* Corresponding author. Mailing address: School of Molecular and Microbial Sciences, University of Queensland, St. Lucia 4072, QLD, Australia. Phone: 61 7 3365 4635. Fax: 61 7 3365 4699. E-mail: martina.jones{at}uq.edu.au

Published ahead of print on 8 August 2007.

REFERENCES

Top