Abstract

Commentary

Autoantibodies to cytokines

K. Bendtzen, Rigshospitalet National University Hospital, Copenhagen, Denmark

Cytokines constitute a large group of intercellular signal peptides with multiple effects on a diversity of physiological and pathophysiological processes. They are centrally involved in normal myelopoiesis and in host responses to physical trauma, infectious agents and growth of certain tumours. Many cytokines exhibit pivotal pathogenic roles in human diseases, from life-threatening complications to surgery to chronic inflammatory diseases and AIDS.

Surprisingly little interest has been devoted to the fact that high-avidity (Kav's 10 -10 to 10 -12 M) antibodies to some, but not other even closely related cytokines are present in healthy individuals and in patients suffering from various immunoinflammatory disorders (1, 2). Natural IgG antibodies have now been found to interleukin (IL)-1a, IL-6, IL-10, interferon (IFN)a, IFN-beta, and granulocyte-macrophage colony-stimulating factor (GM-CSF), but not to for example IL-1b, IL-1 receptor antagonist, IL-2, IL-4, IL-8, IFN-gamma, and granulocyte colony-stimulating factor G-CSF; reviewed in (1). These autoantibodies specifically neutralise their respective cytokines in vitro. It is generally obscure, however, whether the in vitro neutralising autoantibodies always neutralise their respective cytokines in vivo, or whether they may exhibit carrier functions, or cytokine-protective or -stabilising functions (3).

It should be emphasised that the detection of cytokine antibodies is fraught with difficulties; in fact, this may be a major reason why some investigators still doubt the significance or, indeed, the existence of these major cytokine-modulating molecules. For example, false positive results may be obtained from the nonspecific and low-affinity binding that often occurs between IgG and recombinant human cytokines attached to plastic or nitrocellulose membranes, a procedure often used in immunometric or immunoblotting assays for cytokines (1). False positive results may also arise from the use of recombinant cytokines produced for example in yeast cells, because the glycosylated molecules contain a-galactose residues which react with natural antibody to a-galactose present in humans and higher primates (4). Demonstration of binding to the Fab fragments of the immunoglobulins, combined with saturation binding analysis and demonstration of cross-binding to the native cytokine are therefore essential to verify the presence in biological fluids of specific anti-cytokine antibodies (5).

Why and how high-avidity autoantibodies are induced to some cytokines and not to others is unknown. The antibodies characterised so far are usually polyclonally derived, but do not appear to show avidity maturation over time. One possibility is that activation of cross-reactive B- and T-lymphocytes is initiated by cytokine-related proteins coded for example by viruses; in some cases these infections may even obviate the need for co-stimulatory T-cells in that they also induce polyclonal activation of B-cells; e.g. Epstein-Barr virus (6). The immunological tolerance may also be broken if a native cytokine as a hapten binds to a carrier produced for example by an infectious agent; e.g. pox- and herpes viruses, who are known to produce a plethora of cytokine receptor-like molecules (7). Antigenic load as well as timing and site of immunisation may contribute.

In this issue, Elkarim et al. demonstrate the presence of cytokine antibodies not normally found in healthy individuals in the blood and/or cerebrospinal fluid of patients with multiple sclerosis, aseptic meningitis and stroke (8). Perhaps of particular interest is the finding in cerebrospinal fluid and plasma of multiple sclerosis patients of IgG exhibiting Fab-binding to the inflammatory cytokines, IFN-gamma and tumour necrosis factor (TNF)a. These cytokines are thought to play a pathogenic role in multiple sclerosis, and the development of specific antibodies to these cytokines during the course of the disease may therefore constitute an attempt by the immune system to down-regulate the disease-promoting effects of IFN-gamma and TNFa. Interestingly, the same group has unpublished findings that recovery from Guillan-Barré syndrome is associated with the induction of neutralising antibodies to IFN-gamma.

Although the paper by Elkarim et al. supports a role of induced anti-cytokine antibodies during immunoinflammatory neurological diseases, the interest in natural anti-cytokine antibodies in general is bound to increase when more researchers and clinicians realise that: 1) these molecules interfere with the measurement of certain cytokines in biological fluids (1); 2) pre-existing autoantibodies to cytokines almost invariably invalidate therapy with the respective cytokines (2, 9); 3) circulating levels of autoantibodies to selective cytokines can predict the outcome of certain chronic diseases (10); and 4) pharmaceutically prepared, normal human immunoglobulin (IgG) contain high-avidity antibodies to certain cytokines (1).

Pooled human IgG preparations are used for treatment of an increasing number of infectious and immunoinflammatory diseases. These preparations usually contain IgG from several thousand individuals and hence a broad range of binding specificities and idiotypes. The mechanisms by which pooled IgG influence immunoinflammatory reactions are incompletely understood but may include antigen neutralisation, Fc-receptor blockade, attenuation of complement activation, anti-idiotypic interactions and binding to other molecules involved in immunoinflammatory processes. Since, as mentioned above, autoantibodies to cytokines are found in apparently healthy individuals, it is not surprising that they are found in pooled human IgG preparations as well, in some cases at levels which interfere with selective cytokines during high-dose IgG therapy. Could it be that injection of highly specific and neutralising antibodies to inflammatory cytokines contributes to the mode of action of high-dose IgG in various immunoinflammatory and haematological disorders? If so, and as insight into the pathogenic roles of selective cytokines in infectious, autoimmune and haematological diseases increase, therapy with enriched or purified preparations of natural human IgG to specific cytokines might be worthwhile. On the other hand, IgG therapy of patients at risk of contracting infectious diseases, e.g. patients with hypo- or agammaglobulinemia, might benefit from the use of pools selectively depleted of neutralising antibodies to IFN-alpha, IFN-beta and perhaps other cytokines.

Several approaches to selective anti-cytokine therapy have been investigated in the past decade, including blockade of cytokine-receptor interactions by specific heterologous or humanised monoclonal antibodies to cytokines or their receptors, or by recombinant soluble receptors and naturally occurring or constructed receptor antagonists. A drawback in all these approaches, and particularly troublesome if there is a need for repeated therapy, is the formation of antibodies against the foreign proteins, including those generated by recombinant human DNA techniques. In this regard, an interesting perspective is vaccination with specific human cytokines or genetically engineered human cytokines, which do not retain biological activity (11).

References

1. Bendtzen K, Hansen MB, Ross C, Svenson M. Cytokine autoantibodies. In: Peter JB, Shoenfeld Y, editors. Autoantibodies. Amsterdam: Elsevier Science B.V.; 1996. p. 209-16.

2. Antonelli G. In vivo development of antibody to interferons: An update to 1996. J Interferon Cytokine Res 1997;17, Suppl. 1:S39-S46.

3. Bendtzen K, Svenson M, Jønsson V, Hippe E. Autoantibodies to cytokines - friends or foes? Immunol Today 1990;11:167-9.

4. Galili U, LaTemple DC. Natural anti-Gal antibody as a universal augmenter of autologous tumor vaccine immunogenicity. Immunol Today 1997;18:281-5.

5. Svenson M, Herbrink P. Measurement of cytokine antibodies. Test development. Biotherapy 1997;10:87-92.

6. Stuart AD, Stewart JP, Arrand JR, Mackett M. The Epstein-Barr virus encoded cytokine viral interleukin-10 enhances transformation of human B lymphocytes. Oncogene 1995;11:1711-9.

7. Henderson B, Poole S, Wilson M. Microbial/host interactions in health and disease: Who controls the cytokine network? Immunopharmacol 1996;35:1-21.

8. Elkarim RA, Mustafa M, Kivisakk P, Link H, Bakhiet M. Cytokine autoantibodies in multiple sclerosis, aseptic meningitis and stroke. Eur J Clin Invest 1998;28:295-9.

9. Special issue. Biotherapy 1997;10.

10. Jouvenne P, Fossiez F, Banchereau J, Miossec P. High levels of neutralising autoantibodies against IL-1a are associated with a better prognosis in chronic polyarthritis: a follow-up study. Scand J Immunol 1997;46:413-8.

11. Ciapponi L, Maione D, Scoumanne A, Costa P, Hansen MB, Svenson M, Bendtzen K, Alonzi T, Paonessa G, Cortese R, Ciliberto G, Savino R. Induction of interleukin-6 (IL-6) autoantibodies through vaccination with an engineered IL-6 receptor antagonist. Nature Biotechnol 1997;15:997-1001.