Methods for the detection of antibodies (Ab) and autoantibodies (aAb) to cytokines have included cytokine bioassays (e.g. antiviral neutralisation bioassays, ANB), immunometric assays, radioligand assays and various blotting techniques. Unfortunately, several problems hamper the specificity and accuracy of these assays. For example, (a)Ab may bind to denatured cytokines in ways that are entirely different from those operative in vivo, particularly under artificial conditions. Since Ab bind to both conformational and sequence restricted epitopes on proteins, immunoblotting techniques and immunometric assays using immobilised and a more or less denatured cytokine may hamper detection of (a)Ab restricted to the native conformation of the cytokine. Sensitive immunoblotting techniques may also reveal binding of (a)Ab even in cases where the affinity is low and with little or no clinical significance; this binding is often non-saturable and non-specific. Finally, interfering factors in the test sample such as soluble receptors, pre-existing cytokines (free and bound) and other factors may influence these assays in ways that are usually not adequately controlled.

Assay for binding antibodies (BAb)

In IIR, measurements of BAb to cytokines are usually carried out in direct binding assays using soluble 125I-labelled cytokines. This type of radioimmunoassay is usually preferable to the above mentioned techniques, because:
- the tracer can be validated and tested for preserved bioactivity.
- high specific activity of the radiolabelled cytokine (2-5 x 10^5 cpm/ng) and therefore high sensitivity is achieved (detection limit 10^7-10^8 molecules bound).
- the kinetics, avidity and capacity of the binding may be calculated.
- cross-binding to other factors in test materials and the binding specificity can be assessed/controlled.
- serum/BAb interference with target cell receptor binding and biological effects of the cytokines in the presence of BAb may be tested.
For example, cytokines such as interferon (IFN)-a and IFN-beta can be radioiodinated with preserved bioactivities to 2-5 x 10^5 cpm/ng, using Na125I, purified recombinant cytokines and chloramine-T. Tracers are purified on PD10 columns packed with Sephadex G-75 superfine and validated for specific activity, stability and immunoreactivity.
Samples are pre-incubated with 125I-IFN, and the amounts of free and IgG-bound tracer are determined by affinity chromatography using PD10 columns containing protein G Sepharose (the total capacity of serum to bind IFN may be detected as well by the use of a mol.size screening system). Steps are included to control for saturability of binding and non-(a)Ab factors interfering with IgG binding to 125I-IFN.

Assay for ex vivo neutralising antibodies (NAb) to IFN-alpha and IFN-b

ANB (see above) is the standard assay for measuring NAb against IFN. The test consists of a modified IFN bioassay, consisting of distinct steps, with controls incorporated for each step:

Notes

Antibody specificity
E.coli-derived recombinant cytokines are most often used for detection of anti-cytokine Ab. However, these cytokines are non-glycosylated, and naturally occurring cytokines are often glycosylated. If therapy-induced cytokine Ab are sought, the relevant structural form to use in the assay is the cytokine given to the patient. However, whenever binding to Ab has been detected with the recombinant cytokine, demonstration of cross-binding with the natural cytokine is relevant. This becomes mandatory, if one looks for cytokine aAb in both healthy and diseased individuals.

Detection
Some cytokine-Ab in complexes may escape detection because of a low exchange rate between Ab-bound, "cold" wild-type cytokine and the 125I-labelled cytokine during incubation. This may also occur if 125I-labelled cytokine and Ab form complexes with high dissociation rates allowing them to dissociate during the (brief) time it takes to separate bound from free tracer.

Bioassay for IFN NAb
It is important to notice the amount of IFN added vs. the amount of IFN that in the assay has to be neutralised in order to show a positive result. For example, using a high sensitivity assay (e.g. addition of 1.5 LU/ml of IFN) and testing a serum sample at 5% (v/v), 15 U/ml of IFN in full serum would be required to show 50% IFN-neutralisation. In contrast, using a low sensitivity assay (e.g. addition of 100 LU/ml of IFN), 1000 U/ml of IFN in full serum would be needed for 50% IFN-neutralisation. Hence, patients with sera which are positive only in our high sensitivity assay may not necessarily exhibit response failure to rec. IFN therapy. On the other hand, patients testing negative for NAb only in a low sensitivity assay may have significant and clinically important amounts of IFN NAb, and therefore at risk of developing response failure to continued IFN therapy!
It is also crucial to evaluate the level of endogenous IFN in a sample being tested for NAb, as this will interfere with the assay. Furthermore, controls for toxic molecules in the test sample must be carried out; if the sample by itself is toxic to the cells, the data obtained might be inconclusive.

Further information may be found in the following references.

KB 2003