Dear Zeeshan,

SLE is an immune complex disease that occurs by Type III hypersensitivity reaction. Apoptoctic or necrotic cells derived ds-DNA is one of main target of humoral immunity. Antigen-antibody complexes occur in the blood and these complexes trigger complement activation and the other inflammatory events.

The best research I've seen towards answering this question suggests that the antibodies only bind to cells undergoing apoptosis. Evidently many of the nuclear antigens are expressed on the cell surface during apoptosis. Thus, it is likely the antibodies aren't really causing SLE to occur, they are merely a response to the increased apoptosis.

That is something really picking my brains, so it means that this hypersensitivity reaction is produced in case of abuse only, as some genetic link is thought to be there but none of the genes involved have been identified. And I think in apoptosis the dsDNA is not exposed in the tissue spaces, thus I can understand the autoimmune reaction against the unusual phospholipids resulting from flip flop of membrane in the apoptosis. To my understanding the dsDNA is exposed only in the necrosis resulting from inflammation.

This is a good question not often enough asked, Zeeshan.

The first point is that lupus manifestations involve a wide range of effector mechanisms, each relating to a different type of autoantibody, of which there are often many in lupus. Anti-phospholipid syndrome is perhaps the easiest to understand. Antibodies to phospholipid disturb the coagulation cascade. Antibodies to cellular elements in blood are also probably important in aspects like thrombocytopenia, lymphopenia and neutropenia, although these may also arise from interactions with immune complexes.

Apart from these cases the focus is on immune complex-mediated disease. What is important to note here is that there are several quite different ways immune complexes can cause disease. In glomerulonephritis large complexes deposit in the basement membrane region of the glomerular filter. They probably deposit in vessel walls in many other tissues but the glomerulus being a specialised filter is likely to suffer most. It is highly likely that there is enough nuclear material being released into the circulation from mechanically damaged senescent white cells or from other sources to provide antigen. In normal people such complexes should be cleared by complement. Remember that within the circulation complement is anti-inflammatory - a clearance mechanism. It is only pro-inflammatory outside the circulation. Deposition of large complexes in lupus therefore probably has a lot to do with complement malfunction - which is very common in lupus. Not only are there complexes but they are not being cleared.

The second major role of immune complexes is in activating macrophages to produce cytokines in response to ligation of Fc receptor - perhaps mostly FcR3. This is likely to be the basis of arthritis and pericarditis. Exactly what the source of nuclear material would be is uncertain. However, note that years ago both rheumatoid and lupus were said to show 'fibrinoid necrosis'. The difference between the two diseases was that the 'fibrinoid' in RA was red (eosinophilic or proteinaceous) whereas the fibrinoid in lupus was blue-purple (haematoxyphilic or full of DNA). Free DNA is not unusual in sections of damaged tissue of all sorts. Most of it may get mopped up very quickly in the normal situation but I see no reason to think that there is not enough to form complexes if the antibodies are there. Remember that in normal people about 10gm of DNA is being chucked out during red and white cell turnover every day. Just lose a few micrograms and you can have complexes.

Then there are the antibodies to nuclear proteins. It is known that some of these can penetrate cells and many are associated with very characteristic signs of 'cell poisoning'. Cell penetration by autoantibodies has had too little attention although Alarcon Segovia showed clearly that it occurs many years ago and you can demonstrate it on living cells if you antibody treat them and then wash, permeabilise and stain with anti-Ig. Thus anti-topoisomerase in scleroderma is associated with progressive exhaustion of vascular cell replication and so to ischaemia. Anti-synthetase antibodies produce a syndrome that looks quite like ricin poisoning (myonecrosis and pulmonary oedema) which targets ribosomes. And so on.

Risk factor genes for lupus are well known. The complement genes are best understood with C1q deficiency most closely associated with lupus - and a more complex link with the other preMAC components, especially C4 alleles. The onset of autoantibody production in people without homozygous complement deficiencies is apparently stochastic, except in drug induced cases, but that would be consistent with a self generating feedback loop within B cell survival, as suggested in the paper with Cambridge and Abrahams on my RG page.

SLE is a complex and heterogenous disorder. There is a lot more involved in the pathogenesis than just type III hypersensitivity. Antigen-antibody complex or immune complex mediated disease would explain mainly the renal involvement. Other features such as hemolytic anemia is better explained by direct antibody dependent cytotoxicity (ADDC), IgG-C3 mediated, usually. Thrombosis and pregnancy morbidity is explained by antiphospholipid antibodies. Lung and CNS involvement by complement activation, not to mention, increased expression of BLYS (BAFF), and alpha interferon, etc. There is evidence for accelerated apoptosis of T lympocytes, and this may or may not be behind the overly reactive B cells, characteristic of the disease. ANA production is likely a by-product, not directly behind tissue damage, except perhaps for DNA-anti-DNA deposits, as detected in the renal tissue of patients with lupus nephritis.

I agree with a lot of that Emilio, but why make ANA a byproduct? In rheumatoid arthritis a whole bunch of features that occur in lupus can disappear after rituximab: arthritis, serositis, lung disease, vasculitis, normochromic anaemia. These all fit with immune complex mediated macrophage activation. The kinetics mirrors autoantibody decline. Why shouldn't the same apply in lupus? The trial of rituximab in lupus failed but belimumab was effective and there are problems with B cell killing in lupus with rituximab that may reflect complement malfunction. There is also a difference in plasma cell half life for anti-DNA and anti-nucleoprotein antibodies that complicates things. We don't have any other plausible mechanism for these features as far as I know, since nobody has found any autoreactive T cells so I go for antibody mediation as the most plausible option.

Thank you, Jonathan; I don't disagree; may be "by-product" was not the right word to use in this context,

This is now rather the right amount of data, that I could found in the usual immunology books and would have needed a lot of digging of literature. Thank you all gentlemen.

Although Jon Edwards and Emilio González have given some sound answers, I would like to make some additional comments:

1. Of all the SLE-associated antinuclear antibodies (I will leave aside antiphospholipid and other autoantibodies, as they were not part of the original question) only anti-dsDNA and anti-Ro/SSA appear to have a clear pathogenetic role.

2. Anti-dsDNA lead to immune complex-mediated (type III) damage, particularly upon binding to DNA-nucleosome complexes bound to anionic (mainly sialic acid) sites in basal membranes. The wide range of damage seen in different patients probably relates to antibody isotypes or to genetic traits not entirely defined yet, but DNAse 1 polymorphisms could contribute.

4. Antigen substrates for anti-dsDNA probably raise from all types of dead (not only apopthotic) cells.

4. Anti-Ro/SSA autoantibodies are responsible for neonatal lupus after placental transfer from the mother by mechanisms yet to be completely understood.

5. No other SLE-related ANA have been shown convincingly to cause damage either inside or ouside cells. As nice as antibody penetration into living cells appeared when first published in 1977 by Alarcon-Segovia, data never stood and lack of reproducibility, together with spurious results obtained with the same original anti-RNP, have left that once nice hypothesis, as a thing of the past, very unlikely to bear any relevance. 

Dear José,

I agree that the pathogenic roles of other ANAs are not proven, but since other ANAs are associated with disease in other contexts, like scleroderma and myositis I tend to think it is simplest to think in terms of all these antibodies have pathogenic roles.

And I see no reason to think of Alarcon Segovia's idea as a thing of the past. We now know that a component of the Ro complex is TRIM21 which has recently been shown to have a specific function in binding and internalising antibody within cells!! Greg Towers has recently published on this from UCL.

And I have no doubt that antinuclear antibodies penetrate living cells as we have demonstrated this ourselves in the lab. This was not with an anti-RNP, but with a myositis related antibody. I am sure that penetration is an inconsistent feature of immunoglobulins but I find it very hard to dismiss, particularly as it would make so much more sense of all the other ANA related syndromes - anti-synthetase, anti-topoisomerase etc etc. Nobody has ever found any satisfactory T cell effector mechanism for these diseases so cell penetrating antibody is about all we have on offer.

We may have to agree to differ but a dogmatic statement that the A-S idea is a thing of the past seems to me unwise!!

Dear Jon,

I really appreciate your comments and it is true that A-S idea in its origin is not bad at all, but we need to learn a lot to give it a place in our understanding of cell biology. There is no known trafficking mechanism for intracellular transport of antibodies to specific cell sites, especially to the nucleous. I agree that antimyositis antibodies must do something inside the cell as many of them are directed at protein biosynthesis components. Unfortunately, we are still far from understanding how antibodies reach the cytosol and how it relates to myositis.

Thanks again for your comments. 

Maybe TRIM-21 is the key? - although it is unlikely to be that simple I think.