Systemic lupus erythematosus (SLE) is an autoimmune disease of multifactorial aetiology, characterised by the formation of immune complexes (antigen–antibody) that are deposited in different organs, activation of the complement system, infiltration of inflammatory cells in target tissues, lymphocyte hyperreactivity and loss in clearance of immune complexes by the reticuloendothelial system.1 The change in homeostasis in the immune system, is the cause of tissue damage and the secondary systemic involvement. Renal impairment due to SLE occurs in about 50% of patients at some point of the disease and it is an indicator of morbidity and mortality.2 Lupus nephritis (LN) is characterised by a diversity of clinical presentation. Glomerular involvement causes haematuria, proteinuria and the progressive deterioration of renal function. The histological pattern is dynamic and changing, so its understanding is key for treatment and prognosis.

Despite advances in knowledge of the disease's pathophysiology and the currently available therapeutic arsenal, the response rates to therapy (complete remission+partial remission) in LN do not exceed 60%.2 Around 25% of patients with renal impairment due to SLE will develop glomerulosclerosis, which requires renal replacement therapy.

The recent knowledge of the TNF-related weak inducer of apoptosis (TWEAK/Fn14) signalling pathway has helped to establish its role in the pathophysiology of LN's. TWEAK is a type II membrane glycoprotein that belongs to the TNF superfamily, and it is produced endogenously in low concentrations by immune cells (monocytes/macrophages) and non-immune cells in different tissues, for example, renal cells, endothelial cells, synovial membrane, intestinal mucosa. Fn14 is a type I membrane protein from the TNF receptor superfamily, and its specific ligand is TWEAK.3 Experiments performed first in animals and then in humans have demonstrated the role of TWEAK/Fn14 in LN.4 Although the TWEAK/Fn14 pathway is not specific to this condition, its understanding will enable the development of future medical interventions that may modify the natural course of the disease.

This paper discusses the pathophysiological model of the TWEAK/Fn14 pathway in LN; it stresses the importance of urinary TWEAK (uTWEAK) as a biomarker of the disease; indicates the favourable results obtained by the inhibition of the TWEAK/Fn14 pathway as a therapeutic target in animals experiments; and defines the basis for future research.

As described above, in states of cellular rest, without stimuli associated with injury and inflammatory, the tissues’ expression of TWEAK and Fn14 is low.11 However, in inflammatory events such as SLE, there is an increase in Fn14 receptor expression in renal cells, a rise that is not proportional to mTWEAK/sTWEAK availability despite these proinflammatory conditions.4 In addition, it has been possible to establish, e.g. in patients with LN, an increase in Fn14 expression in 3 types of renal cells: podocytes, glomerular mesangial cells and the proximal tubular epithelium. This phenomenon may in part explain the mesangial proliferation, podocytopathy and tubular damage observed in the biopsies of patients with renal impairment due to SLE. The activation of the above-mentioned inflammatory pathway contributes to the glomerular and tubular infiltration of macrophages; the direct cytotoxicity mediated by these cells; and the mesangial and tubular apoptosis – all events that progressively deteriorate renal function. All these changes added to the activation of the complement system and the deposit of immune complexes accelerate kidney injury.

In current clinical practice, markers of disease activity have limitations, given their low sensitivity and specificity. From an immunological point of view, only 2 markers are used on a regular basis to track activity in LN: serum complement levels (C3–C4) and anti-dsDNA antibody titres.12,13 This is why research in recent years has focused on the search for new biomarkers able to predict in advance a relapse of the disease so, a more aggressive immunosuppressive regimen is applied and renal damage is prevented.14 Urine, as a biological sample from glomerular filtration, may directly reflect the inflammatory and cytotoxic events occurring during a relapse. Proteomic studies in urine samples in patients with SLE with renal impairment have made possible to recognise potential biomarkers that could serve as predictors of disease activity.15 By using a proteomic analysis of urine, Mosley et al. were able to identify 2 proteins as candidate biomarkers for distinguishing between active and inactive LN, and for differentiating between disease relapse and remission.16

Xuejing et al. found that after analysing urine samples from 46 SLE patients, with or without secondary renal impairment, uTWEAK levels were significantly higher in those with active LN vs. those without active LN.17 These uTWEAK levels correlated directly with the index of total activity and the index of glomerular and tubulointerstitial activity, but not with the index of chronicity. Moreover, there was a significant relationship between uMCP-1 levels and renal activity due to SLE.

Other studies, such as the one by Schwartz et al. in 2006 in which 2 cohorts of SLE patients were studied; the Ohio SLE Study (OSS); and the Albert Einstein College of Medicine (AECOM) study, showed a direct correlation between uTWEAK levels and rSLEDAI, a rating system used to assess the degree of renal activity in patients with SLE.18 Likewise, and as expected, there was a correlation directly proportional between uTWEAK levels and anti-dsDNA titres, and inversely proportional to serum complement (C3–C4) levels. Interestingly, the type of immunosuppressive therapy used did not have a significant influence in the change in uTWEAK levels; this suggest the therapeutic arsenal currently available does not block the TWEAK/Fn14 signalling pathway. This same group19 showed that there was a statistically significant difference between uTWEAK levels, not only between patients with active and inactive LN, but also with respect to healthy individuals and patients with rheumatoid arthritis (RA) and osteoarthritis. Thus, despite the fact that there is an inflammatory state in these other clinical scenarios, uTWEAK may be a more specific biomarker of renal involvement in SLE.

Not only has uTWEAK been identified as a potential biomarker in LN, but also, in 2011, El-Shehaby et al. showed how levels of uMCP-1, urine osteoprotegerin (uOPG) and urine IL-8 were significantly higher in patients with active LN than in those without LN.20 There was also a direct correlation between uTWEAK, uMCP-1, urine osteoprotegerin levels and the rSLEDAI score.

Hutcheson et al.21 evaluated the role of adipokines, proteins derived from adipose tissue, in both blood and urine, and their usefulness as biomarkers for predicting renal impairment due to SLE in 38 LN patients. The authors found that patients with LN had higher serum levels of adiponectin, leptin and resistin as compared with healthy control groups. However, urine resistin was the only one that showed predictive capacity. Of all the adipokines evaluated, only serum resistin was directly correlated with indicators of kidney injury (creatinine, BUN and proteinuria/creatininuria index).

Much research has been done to identify the usefulness of these biomarkers for predicting not only the clinical course of the disease but also the type of histology of renal involvement in SLE,22–24 which plays a fundamental part in the diagnosis, treatment and prognosis of the disease. Studies such as the one by Brunner et al.,25 for example, identified how the combination of uMCP-1, urinary alpha-1 acid glycoprotein (uAAG) and urine ceruloplasmin were excellent biomarkers for predicting active proliferative LN when added to the use of the proteinuria/creatininuria index as a clinical marker of the disease. They also found that urinary neutrophil gelatinase-associated lipocalin (uNGAL) and uMCP-1 in combination with serum creatinine clearance as a clinical marker were predictors of LN chronicity, and, as with uAAG, C4 serum levels together with creatinine clearance were excellent predictors of membranous LN. Despite these findings, to date there is no biomarker capable of replacing renal biopsy as the gold standard for the diagnosis of the type of LN in progress.

The usefulness of other urine biomarkers in SLE and other glomerulopathies is beyond the scope of this article, but they have been reviewed recently.26,27

Based on the full characterisation of the TWEAK/Fn14 signalling pathway, it has been possible to investigate in clinical settings other than the model of renal impairment due to SLE. Without question, one of the fields in which the most progress has been made is that of atherosclerotic cardiovascular disease. Like diabetes mellitus, chronic kidney disease (CKD) is considered a cardiovascular risk factor. In a cohort of 1058 patients with stages 3-5D CKD and no known history of cardiovascular disease, Fernández-Laso et al.28 found that serum levels of sTWEAK decreased as the stage of CKD progressed. They also found that those individuals with a drop in sTWEAK levels had a higher degree of thickening of the carotid intima-media, a trend that was maintained with respect to the presence and severity of atherosclerotic plaques – calcified or not – on this level. Once the variables were adjusted for the Cox logistic regression model, it was found in this cohort of patients that individuals with lower serum levels of sTWEAK had a higher rate of fatal and non-fatal cardiovascular events. This drop in serum concentrations of sTWEAK, relative to the severity of the endothelial injury, seems to be explained by the increased binding of sTWEAK to Fn14, expressed by the injured endothelial cells, which significantly decreases the serum levels of sTWEAK.

In another study, Akdoğan et al.29 evaluated the serum levels of sTWEAK and MCP-1 in a cohort of 97 patients with stages 2–3 CKD, who underwent heart catheterisation for invasive stratification of coronary disease using the Gensini score (the maximum score of 32 points indicates 100% coronary occlusion). It is striking that they found a directly proportional relationship between the levels of sTWEAK and MCP-1 and the Gensini score obtained.

Jasiewicz et al.30 evaluated the relationship of TWEAK in other cardiovascular scenarios. This group performed a comparative study of 26 patients with pulmonary hypertension, confirmed by right heart catheterisation and 24 healthy volunteers. Their results showed that individuals with pulmonary hypertension had a higher expression of the CD163 marker (macrophage scavenger receptor) and lower serum levels of sTWEAK as compared with healthy volunteers. The explanation for these results is not entirely clear, but as previously mentioned, the fact that sTWEAK binds to Fn14 might explain the reduction of serum concentration. Also, the increased expression of the CD163 receptor might indicate an activation of the negative regulation mechanisms of TWEAK/Fn14. These are, however, unconfirmed hypotheses.

Given the structural homology between TNF and TWEAK, TNF's leading role has been extrapolated to the pathophysiology of other diseases and the relation to the TWEAK signalling pathway. Conditions like inflammatory bowel disease (IBD), psoriasis and RA are characterised by high levels of TNF alpha in affected tissues. For example, the expression of TWEAK messenger RNA in the mucosa of patients with IBD has been described with an exponential and direct increase according to the disease's activity.31 Furthermore, Bilgiç et al.32 studied the differences in the expression of TWEAK and other cytokines (IL-6, IL-23, TNF alpha) in serum in 45 patients with chronic psoriasis and 43 healthy volunteers. Their results showed that serum sTWEAK, IL-6, IL-23 and TNF alpha were significantly higher in individuals with chronic psoriasis than in healthy volunteers, but these findings did not correlate with disease severity as estimated using the Psoriasis Area and Severity Index (PASI). In RA, an increase in sTWEAK concentrations in the synovial fluid of patients with active disease has been found, in addition to an increase in RANKL expression by sTWEAK-induced synoviocytes33 and osteoblasts.34 Compared with psoriatic arthritis, patients with active RA have higher levels of sTWEAK, which in part explains the potential joint damage this disease can cause.35

Finally, sTWEAK has been used as a biomarker in terms of recovery in endothelial function following kidney transplantation. In 2013, Yilmaz et al.36 published a study in which they followed a cohort of 175 patients with kidney transplantation for 6 months. They evaluated endothelial function by measuring flow-mediated dilation of the brachial artery with Doppler ultrasound on day 0 and 180 days post-transplantation. They likewise evaluated the change in serum concentrations of sTWEAK over the same time period. Their results showed that improvement in Doppler blood flow correlated directly with normalisation in serum concentrations of sTWEAK, possibly in relation to its lower degree of binding to the Fn14 expressed by endothelial cells.

Several studies in experimental models in animals and also human studies have evaluated the usefulness of blocking the TWEAK/Fn14 signalling pathway37 (Table 1). Studies performed on animals with acute kidney injury following folic acid overdose have shown that blocking TWEAK/Fn14 is associated with improved renal function, reduced interstitial inflammation, decreased proliferation and tubular death.38 These findings were also reproduced in another animal experiment with acute kidney injury through the induction of kidney damage caused by ischaemia-reperfusion injury.39 In this work, for example, a decrease was also shown in residual renal interstitial fibrosis when TWEAK/Fn14 was blocked.

Xia et al. reproduced in a model of nephrotoxic nephritis, the immunological basis that started from the deposit of immune complexes at the glomerular level, thus simulating the conditions of LN, to demonstrate the clinical, paraclinical and histopathological course in a murine model by blocking the TWEAK/Fn14 pathway.40 Fn14+/+ WT (wild type) and Fn14−/− KO (knockout) mice were analysed comparatively. The Fn14−/− KO mice had a lower degree of proteinuria, less tubular disease and less glomerular crescent formation as compared with the Fn14-expressing strains. The use of anti-TWEAK monoclonal antibody injected in mice that did not die from uraemia after being inoculated with nephrotoxic serum showed a lower intensity in the Periodic acid–Schiff staining, a lower proliferation index (Ki 67+) and less glomerular infiltration by macrophages in each of the biopsies analysed.

Despite all of these potentially beneficial effects of blocking the TWEAK/Fn14 pathway in different diseases, not all studies have shown favourable results. In a murine model, Mustafa et al. explored whether the genetic or pharmacological ablation of TWEAK/Fn14 modified the histological changes of the substantia nigra and striated in Parkinson's disease.41 Their results failed to demonstrate a reduction in acute neurotoxicity mediated by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine after the genetic ablation of TWEAK and Fn14.

There are phase I and phase II studies of humans in progress that assess the role of TWEAK/Fn14 as a therapeutic target in RA,42 advanced solid tumours43 and LN.44 Their results will provide a better picture of the usefulness of blocking this pathway in these clinical scenarios.

Significant progress has been made in casting light on the TWEAK/Fn14 pathway in the pathophysiological model of several types of autoimmune and inflammatory diseases. For the time being, it is clear that TWEAK/Fn14 is involved in the processes of inflammation, proliferation and apoptosis of podocytes, glomerular mesangial cells and tubular epithelial cells of patients with renal impairment due to SLE. Based on the current evidence, we might consider the TWEAK/Fn14 pathway to be a necessary but insufficient mechanism to fully explain the potentially harmful events caused by this type of disease. The clinical usefulness of blocking the TWEAK/Fn14 pathway being evaluated in studies in progress will determine the real importance of this molecule as the leading actor in LN and other clinical scenarios.

The authors declare that they have no conflicts of interest.