Timeline of interferon, autoantibodies, and autoimmune disease development

[jnmaciuch]

Posting several links exploring temporal dynamics between interferon, autoantibodies, and autoimmune diseases (primarily lupus).

(Review) Emerging concepts of type I interferons in SLE pathogenesis and therapy
Note: SLE = systemic lupus erythematosus

The genetic susceptibility of SLE is complex, with >100 loci known at a genome-wide level that involve diverse immune pathways 20. Many newly identified loci are linked to activation of the interferon pathway 21. The importance of interferon regulatory factors such as IRF-5 was highlighted above, as the risk haplotype is associated with increased levels of serum IFNα and a risk of clinical progression to SLE in ANA-positive individuals 13. In particular, the IRF5 rs2004640-T allele enables expression of several distinct IRF-5 isoforms and conveys an important genetic risk for SLE 22. The IRF5 rs10488631 allele has an independent association with enhanced transcript and protein levels of IRF-5 in patients with SLE 23. Single-nucleotide polymorphisms (SNPs) in IRF7 loci also confer a strong genetic susceptibility to SLE 24. Additionally, genetic variations at the IRF7–PHRF1 locus are associated with high levels of serum IFNα and positivity for anti-Sm antibody in African American patients 25. Similarly, the IFIH1 rs1990760-T allele is associated with anti-dsDNA positivity as well as a high type I interferon gene signature (IFNGS) in peripheral blood mononuclear cells (PBMCs) of patients with SLE 26. SNPs in a STAT4 haplotype, such as the rs7574865 allele, are strongly linked to SLE, and homozygosity of this risk allele carries a more than doubled risk of developing the disease (compared with absence of the risk allele) 27. The STAT4 rs7574865-T allele is also linked to sensitivity to type I interferon signalling 28.

Notably, although there are many genetic associations involving components of the downstream signalling pathways of type I interferon, no polymorphisms within genes encoding most type I or II interferons are known to be linked to SLE susceptibility. An exception is the IFNK locus. As identified by a pooled genome-wide study, the rs1031154 SNP (which affects a site 25 kb upstream of the promoter for IFNK) is strongly associated with SLE susceptibility (rank number 51 of 116,204 SNPs) 29. SNPs at the IFNK locus are also associated with serum type I interferon activity as well as cutaneous lupus clinical phenotype. These results call into question whether IFNα has a particular importance in SLE pathogenesis compared with other type I interferon subtypes, such as IFNκ.

Genetic-transcriptomic studies can provide information on the tissues affected by pathogenic variants in SLE susceptibility genes. In a combined analysis of genotypic and transcriptomic data, susceptibility variants associated with SLE, including those in the type I interferon pathway, were shown to regulate gene expression across a range of blood and non-blood tissues 30.

Links to references are active in the quote. Being predisposed towards higher interferon levels confers causal risk for developing autoantibodies and lupus, though does not guarantee disease. Interestingly, this review also noted that mutations in most interferon or interferon receptor genes themselves are not associated with lupus--likely because individuals with those mutations simply develop severe "interferonopathies" first. I think it is an interesting concept that the degree and exact relationship to interferon regulation across various risk alleles can create different clinical outcomes in conjunction with environmental triggers.

Functionally impaired plasmacytoid dendritic cells and non-haematopoietic sources of type I interferon characterize human autoimmunity
Study followed SLE and primary Sjögren’s Syndrome (pSS), at-risk, and control populations. Individuals were classfied as "at risk" if they had positive anti-nuclear antibodies (ANAs) but did not meet clinical criteria for autoimmune disease. Higher interferon levels are seen in the blood, and much higher levels are seen in the skin, in both at-risk and autoimmune individuals compared to control. Although plasmacytoid dendritic cells (pDCs) are the main source of type I interferon in acute viral infection, they are not the source of high interferon levels seen in SLE and pSS. Interferon in the skin was determined to originate from keratinocytes (outermost layer skin cells).

Prediction of autoimmune connective tissue disease in an at-risk cohort: prognostic value of a novel two-score system for interferon status
Follow-up study of at-risk (ANA positive but not presenting with clinical disease) individuals. Interferon A and B scores (not to be confused with interferon-alpha and beta) were calculated using different subsets of genes stimulated by interferons. A contained genes stimulated by type I interferons, B contained genes stimulated by type I, II, and III. At-risk and SLE individuals both had higher interferon A scores than healthy controls. SLE individuals had higher interferon B scores compared to at-risk and HC (and at-risk/HC did not significantly differ from each other).

(Abstract/Poster) Prediction of response to rituximab in SLE using a validated two-score system for interferon status
Using same interferon-stimulated gene scores from above. High interferon A and B scores predicted non-responsiveness to rituximab (though only B score was p < 0.05).

Altered Type II Interferon Precedes Autoantibody Accrual and Elevated Type I Interferon Activity Prior to Systemic Lupus Erythematosus Classification
Longitudinal data for individuals who eventually developed lupus and age-and-sex-matched controls. Autoantibodies and elevated type II interferon could be seen several years (up to ~8) prior to clinical disease, and in healthy controls that never progressed to disease. Type I interferon levels peaked 1-2 years before clinical disease onset, but were not detected as elevated in the serum prior to detection of autoantibodies. This suggests that development of lupus is a combined effect of autoantibodies and elevated type I interferon over time. A caveat is that pre-autoantibody measurements were taken at random, and therefore cannot exclude the possibility of a precipitating event involving high interferon (e.g. viral infection) triggering autoantibody formation (which did not result in permanently elevated type I interferon prior to autoantibodies).

Induction of Dendritic Cell Differentiation by IFN-a in Systemic Lupus Erythematosus
Circulating monocytes from SLE participants were found to present antigen more than controls, and serum from SLE induced monocytes to differentiate into dendritic cells (which have higher expression of MHCI/II and induce higher T-cell responses). This differentiation effect was recapitulated only with interferon-alpha. The authors speculate that type I interferon priming may explain increased likelihood to develop autoantibodies.
https://pmc.ncbi.nlm.nih.gov/articles/PMC3694565/#S5
(Review) The innate immune system in SLE: type I interferons and dendritic cells

Through their direct effect on B cells, type I IFNs enhance primary antibody responses to soluble proteins and induce the production of all subclasses of IgG in mice.33 IFN-α upregulates CD38, a germinal centre B-cell and plasma cell marker, on B lymphocytes and BAFF (B cell–activating factor) on monocytes and mDCs. BAFF in turn contributes to the survival of autoreactive B lymphocytes.34 In addition, IFN-α promotes the differentiation of activated B lymphocytes into plasmablasts. pDCs activated with viruses secrete IFN-α and IL-6, which permits plasmablasts to become antibody-secreting plasma cells.35 The same effect is observed when pDCs are activated with SLE ICs containing nucleic acids that bind TLRs.36, 37 This could contribute to amplify the production of type I IFN and subsequently the differentiation of autoreactive plasma cells that would further secrete autoantibodies, thus, perpetuating this pathogenic loop.

Auto-antibody production induced by type I interferon has not been directly observed, though there is evidence that type I interferon may create optimal conditions for auto-antibody generation. Selected quote is only for effect of type I interferon on B-cells, there are other potentially salient effects on other cell types as well outlined in the linked review.

Inactive disease in patients with lupus is linked to autoantibodies to type I interferons that normalize blood IFNa and B cell subsets
Just an interesting tidbit--turns out that some lupus patients naturally develop auto-antibodies against type I interferon, and this ends up having a protective effect reducing disease severity.

Low-titre auto-antibodies predict autoimmune disease during interferon-α treatment of chronic hepatitis C
Patients with pre-existing anti-nuclear antibodies were more likely to develop autoimmune disease during interferon alpha therapy, although there were a few de-novo cases as well.

Summary:

Genetic studies indicate that higher levels of type I interferons predispose the development of lupus, though congenitally high type I interferon levels do not guarantee autoimmunity and are not present in all cases of autoimmunity. Anti-nuclear antibodies themselves do not create disease either since they are observed in the absence of disease. The combination of auto-antibodies and induced high levels of type I interferon are both necessary to manifest tissue damage associated with auto-immune disease, and auto-antibodies may be drivers of later type I interferon production immediately preceding disease [edit: which may encourage clonal expansion to then create a feedback loop of autoantibodies and interferon].

One additional caveat is that serum levels are not reflective of tissue-specific levels, and none of these studies explore parenchymal interferon production beyond the skin. Therefore, it is possible that any disease manifestations dependent on type I interferon are actually dependent on tissue-specific presence of interferon, rather than just in the blood.

 

[Jonathan Edwards]

Genetic studies indicate that higher levels of type I interferons predispose the development of lupus, though congenitally high type I interferon levels do not guarantee autoimmunity and are not present in all cases of autoimmunity. Anti-nuclear antibodies themselves do not create disease either since they are observed in the absence of disease. The combination of auto-antibodies and induced high levels of type I interferon are both necessary to manifest tissue damage associated with auto-immune disease, and auto-antibodies may be drivers of later type I interferon production immediately preceding disease [edit: which may encourage clonal expansion to then create a feedback loop of autoantibodies and interferon].

One additional caveat is that serum levels are not reflective of tissue-specific levels, and none of these studies explore parenchymal interferon production beyond the skin. Therefore, it is possible that any disease manifestations dependent on type I interferon are actually dependent on tissue-specific presence of interferon, rather than just in the blood.

I am trying to catch up after 2 days offline. Some very interesting material here.

Type I interferons have important effects on B cell dynamics and there has been interest in lupus for a while in terms of predisposing to an autoantibody producing state. The story is obviously cleare for complement where homozygous deficiencies of several complement components produce lupus, or a lupus-like syndrome with nephritis. I interpret the complement effect as a double whammy of both impairing selection of B cell receptors on new B clones in marrow (and later in lymphoid tissue) and impairing clearance of complexes generated once autoantibodies appear. My guess is that interferons probably contribute more in terms of shifting thresholds rather than being directly involved in rogue clonal deletion.

I think ANAs do create disease in themselves. The argument from ANAs in normals is what led people astray in the 1980s, until work from the endocrine diseases (people like Botazzo) and fine separation of extractable nuclear antigens made it clear that just measuring autoantibody in an old -fashioned precipitation system (or equivalent) missed the fact that autoantibodies come in a huge range of functional types. Some are pathogenic, others are not. Everyone has some rheumatoid factor. Lots of people have low level ANA. But almost no healthy people have significant titre anti-DNA by specialised assays. One of the most beautifu demonstrations of the need to look at the functional subtype of an autoantibody population came when Jorg Schifferli questioned me in Basle as to why, if rheumatoid factor complexes really cause synovitis through ligating tissue macropahge FcR3a then why didn't people with Waldenstrom's hypergammaglobulinaemic purpura have arthritis? They have huge rheumatoid factor titres and circulating complexes. I read up the papers and the answer seemed to be that the rheumatoid factors in WHP are always IgG3. And IgG3 has poor access to normal tissues because of the weird long Fc/Fab connections.

The evidence that ANA's produce lesions is pretty good. We can see their complexes in membranous lupus nephritis without needing to invoke any cellular reaction. They destroy circulating platelets. The produce a small immune complex pattern like RA, of synovitis, serositis and alveolitis.

It would be interesting if high interferons were necessary for tissue targeting effector mechanisms of lupus antibodies but do we have any evidence for that? Neonates can have lupus, myasthenia or Sjogren's related heart block purely from maternal antibody crossing placenta.

 

[jnmaciuch]

I think ANAs do create disease in themselves.

Some are pathogenic, others are not. Everyone has some rheumatoid factor. Lots of people have low level ANA. But almost no healthy people have significant titre anti-DNA by specialised assays.

Which specific antibodies by specialized assays are you referring to here?

I agree that very high titres of anti-dsDNA (if that's what you meant) are not seen in healthy people, but I think the Munroe et al. longitudinal study and the Bell et al. interferon-a therapy study both paint a picture of a two-phase interaction between those ANAs, interferon, and ultimate disease state.

We have:
1) the precipitating event that creates those ANAs initially, which may never progress past low titre levels
and
2) an escalation that ultimately manifests in clinical disease, characterized by both high interferon levels and high titres.

The interferon therapy paper seems to suggest that high dose interferon alpha therapy (in conjunction with some sort of genetic susceptibility, no doubt) will facilitate that escalation for those who already have low titres, and can also create ANAs anew.

A feed-forward loop between antibodies and interferon seems like the obvious candidate for that later escalation given that longitudinal data. But there still must be something that sets everything into motion.

It would be interesting if high interferons were necessary for tissue targeting effector mechanisms of lupus antibodies but do we have any evidence for that?

There is some evidence for high interferons being necessary in cutaneous lupus (in mice, take it or leave it)
The early local and systemic Type I interferon responses to ultraviolet B light exposure are cGAS dependent

Additionally, interferon kappa levels in the skin are highly correlated with cutaneous lupus in humans and mutations in interferon kappa are very predictive for CLE (per the first review in my original post).

Additionally, for psoriasis, there is evidence of type I signaling being necessary for the formation of lesions. Keratinocyte IFN-b production and pDC activation precedes T-cell infiltration and blockade of IFNAR prevents formation of lesions.
Plasmacytoid predendritic cells initiate psoriasis through interferon-α production

And interferon alpha therapy is sufficient for triggering psoriasis in some cases.
Induction and exacerbation of psoriasis with Interferon-alpha therapy for hepatitis C: A review and analysis of 36 cases

All in all, my sense is that an exacerbated type I interferon response is necessary to actually call in the cavalry from the circulation to see effector-mediated tissue damage. Though high cutaneous interferon levels can also be observed in at-risk individuals prior to development of disease--so there must be some additional step that turns over on the adaptive side. I'd put good money on viral infection triggering that "turnover event" in a good proportion of cases--a sudden spike in system interferon response triggering clonal expansion, perhaps?

 

[Jonathan Edwards]

A feed-forward loop between antibodies and interferon seems like the obvious candidate for that later escalation given that longitudinal data. But there still must be something that sets everything into motion.

Yes the feed forward loop was central to our general theory for autoimmune rheumatic disease in 1999. The generation of the antibody is expected to increase the permissive factor/s that allow its clones to expand. That doesn't mean the permissive factror is involved in tissue targeting.

All you need to set it in motion is the generation of the right antibody by random Ig gene rearrangement maybe with some further mutation. An inherently unstable system dynamic that allows for the production of positive cycles affecting randomly generated protein species requires nothing more to start that off than the random generation of the species. The story will be more complex because you need clonal co-operation and recognition of more than one epitope in most of the options known. But there is no need for any other trigger. Which of course fits with the epidemiology of these diseases mostly fittin genes + stochastic rather than genes plus environment, as Stastny pointed out when describing the DR4 link to RA.

I wouldn't be surprised if interferon was needed for cutaneous lupus, which isn't really a big problem. In something like skin it might well be necessary for expression of rash but is there any evidence in humans that you get more cutaneous lupus if skin is already overdosed with IFN in some way?

I still don't really get what story you are trying to put together. Genetic predisposition to IFN production or sensitivity would make sense but that doesn't give a 'trigger' for the onset of acquired disease. IFN may be involved down the line in some lesions but they are going to be triggered by antibody from time to time.

All in all, my sense is that an exacerbated type I interferon response is necessary to actually call in the cavalry from the circulation to see effector-mediated tissue damage.

But calling in the cavlry is often rather pathetic in lupus as opposed to e.g. RA. Immune complexes can cause a lot of trouble in glomeruli without a cellular response. In synovium there is rarely any major lymphoid infiltration as there is in RA (unless the person is also RF+) and quite often there is arthralgia with a rather low cell count fluid.

 

[jnmaciuch]

I still don't really get what story you are trying to put together. Genetic predisposition to IFN production or sensitivity would make sense but that doesn't give a 'trigger' for the onset of acquired disease.

I'm hypothesizing that those genetic predispositions not only increase baseline interferon, but also massively exacerbate interferon responses when triggered by e.g. viral infection or other situations where you already have a systemic interferon response. [Edit: was not arguing that genetic predisposition was itself a triggering event, the opposite actually. That genetic predisposition to high interferon] might not be necessary in all cases, just like an HLA mutation is not necessary but does increase likelihood of a stochastic event. But an [edit: initial] interferon response, whether genetically mediated or not, does change those thresholds and make things much more likely to occur and escalate, [edit: and is a necessary part of the feed-forward loop during escalation]

IFN may be involved down the line in some lesions but they are going to be triggered by antibody from time to time.

But calling in the cavlry is often rather pathetic in lupus as opposed to e.g. RA

But that is the opposite of what the evidence shows, at least in psoriasis. The lesions don't occur at all without interferon, and there is no presence of lymphocytes without interferon. And it's primarily neutrophils in CLE that first have the exaggerated IFN response to UV-induced cell free DNA preceding flares. The presence of autoantibodies seems to lead to an extreme cell-mediated response once the lymphocytes arrive in the tissue en masse, but if they never get called to the site by an exacerbated interferon response, you don't get tissue damage.

This might not be a requirement for SLE and nephritis where the circulating lymphocyte population will already be in contact with cfDNA. But for other tissue damage, it seems to be necessary.

In something like skin it might well be necessary for expression of rash but is there any evidence in humans that you get more cutaneous lupus if skin is already overdosed with IFN in some way?

Interferon itself doesn't create the rash in any case.

 

[Jonathan Edwards]

But that is the opposite of what the evidence shows, at least in psoriasis. The lesions don't occur at all without interferon, and there is no presence of lymphocytes without interferon. And it's primarily neutrophils in CLE that first have the exaggerated IFN response to UV-induced cell free DNA preceding flares. The presence of autoantibodies seems to lead to an extreme cell-mediated response once the lymphocytes arrive in the tissue en masse, but if they never get called to the site by an exacerbated interferon response, you don't get tissue damage.

Hang on, psoriasis is not an autoimmune disease and nothing to do with antibodies as far as I know.
Interesting if interferons are a necessary permissive factor in generating skin inflammation. But in ME/CFS the puzzle is that there is no inflammation anywhere.

 

[jnmaciuch]

Hang on, psoriasis is not an autoimmune disease and nothing to do with antibodies as far as I know.
Interesting if interferons are a necessary permissive factor in generating skin inflammation. But in ME/CFS the puzzle is that there is no inflammation anywhere.

I’m well aware, I was pulling extra evidence for interferons preceding adaptive immune mediated damage as my search did not reveal comparable research about timeline of interferon and lymphocyte infiltration in cutaneous lupus. I apologize for lack of clarification on that point

[Edit: and, to that last point, interferon signaling doesn’t necessarily mean inflammation either. It is necessary to initiate inflammation when there is also an adaptive shift towards targetting host tissue. But it can also cause symptoms on its own without inflammation, as we see in interferon therapies. Here I think the tissue localization matters quite a bit for how specific symptoms manifest]