Renal monocyte chemoattractant protein-1: an emerging universal biomarker and therapeutic target for kidney diseases?

Monocyte chemoattractant protein-1 (MCP-1) (also known as CCL2) is a member of a large family of chemotactic cytokines (chemokines) known to be important soluble mediators of innate immunity and tissue inflammation. MCP-1 was one of the earliest cytokines found to have a selective role in controlling monocyte/macrophage migration (although it is now known to act also on T cells and basophils). This is in contrast to other inflammatory mediators, such as complements and platelet activity factor (PAF), that have broader chemoattractant activity. MCP-1 is synthesized by a range of cell types in the kidney, including intrinsic kidney cells and infiltrating leucocytes (Figure 1A). Glomerular mesangial cells release MCP-1 in vitro following stimulation with a variety of stimuli, including inflammatory cytokines [interleukin-1 (IL-1) and tumour necrosis factor alpha (TNF-α)], immune complexes, metabolic factors (glucose, glycation end products) and danger-associated molecular patterns (DAMPs), such as extracellular adenosine triphosphate [1]. Podocytes release MCP-1 following stimulation with glycation end products, TNF-α and transforming growth factor beta (TGF-β) in vitro [2–4]. Stimulation of podocytes [which express CC chemokine receptor type 2 (CCR2)] by MCP-1 has also been shown to downregulate nephrin, an alteration that could contribute to development of proteinuria in diabetic nephropathy and focal segmental glomerulosclerosis (FSGS) [5]. Renal tubular cells also produce MCP-1 in response to IL-1 and TNF-α [6], thrombin [7] and albumin [8]. The wide variety of cells and stimuli that are known to release and respond to MCP-1 highlight its importance but also raise questions as to whether it could ever be useful as a specific disease biomarker or selective therapeutic target (recently reviewed) [9]. The renal MCP-1–CCR2 axis. (A) Cytokines, inflammatory cells and the pathogenesis of kidney disease. MCP-1 can be synthesized by a range of cell types including macrophages, podocytes, mesangial cells, endothelial cells and tubular cells. (B) The interaction between MCP-1 and CCR2 is not exclusive, as MCP-1 can also bind to CCR1, CCR3 and CCR5. Conversely, CCR2 has been shown to bind multiple chemokines such as CCL7, CCL8, CCL11, CCL12, CL13, CCL24 and CCL26. Although an interaction between MCP-1 and CCR2 is the best-known ligand–receptor interaction, it is not an exclusive one. MCP-1 has been shown to interact with multiple receptors, e.g. CCR1, CCR2, CCR3 and CCR5. Conversely, CCR2 can bind to other chemokines apart from MCP-1, such as CCL7, CCL8, CCL11, CCL12, CCL13, CCL24 and CCL26 (Figure 1B). Therefore selective CCR2 blockade should not be considered the same as MCP-1 neutralization (with an anti-MCP-1 antibody). Nevertheless, the severity of kidney injury (macrophage infiltration, type I collagen expression and tubulointerstitial fibrosis) was reduced in Ccr2 knockout mice or following treatment with a CCR2 antagonist [10], suggesting that CCR2 could be the major receptor mediating the effect of increased MCP-1 in kidney disease. In this month’s issue of NDT, three papers highlight the role of MCP-1 in a range of kidney diseases as a clinical biomarker of disease activity or in disease pathogenesis. The first two studies report that the combination of urinary MCP-1 (uMCP-1) with other soluble mediators can improve its performance as a biomarker of disease activity. In the first study, Moran et al. [11] found that the combination of urinary soluble CD163 (sCD163) and uMCP-1 improved the diagnostic accuracy of renal flares in patients with anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis. In the second study, Wu et al. [12] reported a significant association between urinary epidermal growth factor (uEGF) and uMCP-1 with disease progression in diabetic nephropathy. Finally, Wilkening et al. [13] demonstrated an increased expression of the MCP-1 receptor CCR2 in kidney biopsies of patients with FSGS and found a positive effect of CCR2 deletion in a murine model of FSGS. Increased glomerular synthesis of CC chemokines, including MCP-1, has been shown to correlate with monocyte/macrophage infiltration in experimental models of antibody and immune complex–mediated glomerulonephritis [14, 15]. In clinical diseases, increased MCP-1 expression in human renal biopsies and excretion in urine has been detected in a number of proliferative glomerulonephritis associated with monocyte/macrophage infiltration, such as immunoglobulin A (IgA) nephropathy, lupus nephritis, renal vasculitis and Goodpasture’s disease [16, 17] (Table 1). Urinary and renal MCP-1 as a biomarker in studies of patients with kidney disease Urinary and renal MCP-1 as a biomarker in studies of patients with kidney disease In inflammatory diseases, the long-term effects of renal injury are determined by a balance between inflammation-induced injury and cellular repair or regeneration. EGF is produced by the thick ascending limb of the loop of Henle and distal renal tubule and has tubular-trophic effects. In a study of patients with IgA nephropathy, Torres et al. [19] found that the uEGF:uMCP-1 ratio provided a better prognostic assessment of renal outcome at 4 and 8 years after biopsy than the uEGF:creatinine or uMCP-1:creatinine (Cr) ratio [19]. The role of uMCP-1 as a disease marker has also been investigated in ANCA-associated renal vasculitis. In a seminal study, uMCP-1 was found to be elevated in patients with active renal vasculitis but not in patients with systemic vasculitis sparing the kidney. Most patients had a decrease in uMCP-1 following successful broad immunosuppressive treatment except, in one patient who had a further increase in uMCP-1 and progressed to renal failure despite immunosuppressive treatment [17]. These findings have since been replicated by several other groups [20–22]. Recently uMCP-1 has also been used as a pre-specified secondary endpoint in a randomized clinical trial of a complement C5a receptor inhibitor in patients with ANCA-associated renal vasculitis [23]. In the latest study by Moran et al. [11], the combination of creatinine-normalized urinary soluble CD163 (usCD163) and uMCP-1 was more sensitive than the individual tests in detecting subtle renal flares in ANCA-associated renal vasculitis (n = 88) in the longitudinal Vasculitis Clinical Research Consortium multicentre North American cohort. Mechanistically, this could reflect distinct independent aspects of macrophage activation (CD163) and recruitment (MCP-1), as there was poor direct correlation between them. The algorithm performed better when new-onset proteinuria was included as an independent variable. A major limitation of the study is the reliance on clinical assessment alone without histological confirmation, leading to potential misclassification. There was also a discrepancy with the uMCP-1 results as previously reported on a subset of patients with non-renal disease by a different group [21]. Finally, although the post-test probability of detecting a positive renal flare using this algorithm was >90% in patients with either moderate or high pre-test probability of having a renal flare, the equivalent post-test probability of a negative test excluding a flare was 22% (moderate) or 58% (high), respectively, indicating that the proposed algorithm will need much refinement and validation prior to clinical use. Of interest, MCP-1 may also play a role in the pathogenesis of chronic allograft failure. In a recent study, uMCP-1 was prognostic of renal allograft failure [35]. Although diabetic nephropathy is now the most common cause of kidney failure, the contribution of monocytes/macrophages in its pathogenesis has not been studied until recently. In renal biopsies of patients with diabetic nephropathy, increased expression of MCP-1 and macrophage infiltration in the glomeruli and tubulointerstitium has been reported. Similarly, increased numbers of glomerular macrophages have been detected in rodent models of diabetic nephropathy [24]. Of significance, proteinuria and glomerular macrophages were significantly reduced in Mcp1 knockout mice following the experimental induction of diabetes. Increased levels of uMCP-1 as well as other cytokines and profibrotic growth factors [TGF-β, connective tissue growth factor (CTGF)] have been detected in patients with diabetic nephropathy. In a longitudinal study over 6 years, the uMCP-1:Cr ratio was found to be predictive of the rate of estimated glomerular filtration rate (eGFR) decline [25]. These results have been validated in several other studies even when adjusted for baseline covariates, such as diabetic control, blood pressure, age and sex [26]. The idea of measuring the uEGF:uMCP-1 ratio was first described in a study of IgA nephropathy [19]. Two later studies (including one in this current issue) have now demonstrated the prognostic value of the uEGF:uMCP-1 ratio in diabetic nephropathy [12, 27]. The first study from the Joslin Diabetic Centre reported a negative correlation between uEGF:Cr and uEGF:uMCP-1 and the occurrence of diabetic kidney disease (DKD) [27]. In a longitudinal follow-up, the uEGF:uMCP-1 ratio correlated more closely to the rate of loss of eGFR than uEGF:Cr or uMCP-1:Cr. When a composite endpoint (end-stage kidney disease or 30% reduction in baseline eGFR) was studied, uEGF:Cr and uEGF:uMCP-1 were negative predictors, whereas uMCP-1:Cr was an independent positive predictor of disease progression [27]. In this issue, Wu et al. [12] confirm that the uEGF:uMCP-1 ratio in the urine is a better prognostic biomarker than the uEGF:Cr or uMCP-1:Cr ratio [12]. However, there are several major limitations to the current study. First, the conclusions were based on the cross-sectional analysis of two very different Chinese population-based cohorts with early [Incident, Development and Prognosis of DKD (INDEED), n = 1811] and advanced [Chinese Cohort Study of Chronic Kidney Disease (C-STRIDE), n = 208] DKD. Second, >90% of patients had type 2 diabetes mellitus and therefore the conclusions may not apply to type 1 patients. Third, although statistically significant, the use of uEGF:Cr or uEGF:uMCP-1 only marginally added to the discriminatory value of the urine albumin:creatinine ratio in distinguishing DKD from those without DKD in the INDEED cohort. Similarly, uEGF:Cr, uMCP-1:Cr and uEGF:uMCP-1 showed limited clinical value (overlapping confidence intervals) in predicting composite kidney outcomes (end-stage renal failure and 30% decline in eGFR) in the C-STRIDE cohort. Although promising, these results will clearly need to be validated in a larger multi-ethnic cohort with longer-term follow-up. The clinical relevance of the MCP-1/CCR2 pathway in DKD has been tested in a trial of CCX140B, a small-molecule inhibitor of CCR2, in type 2 diabetic patients with well-controlled blood pressure (BP) (average age 62.5 years, women 15%, ethnic origin 97% Caucasian, average BP 137/78 mmHg, average eGFR 62 mL/min/1.73 m2) receiving the maximal tolerable dose of angiotensin-converting enzyme inhibitor or angiotensin receptor inhibitor. Low-dose, but surprisingly not high-dose treatment with CCX140B resulted in a significant reduction in albuminuria compared with placebo [36]. However, the clinical development of CCX140B requires further evaluation because of the lack of efficacy of high-dose treatment. In a Phase 2A clinical trial in type 2 diabetic patients, a different CCR2 antagonist, emapticap pegol (NOX-E36), was studied. Although the drug was safe and well tolerated, the reduction in albuminuria (29% treated versus 16% placebo) observed did not reach statistical significance compared with the placebo treatment [37]. Increased uMCP-1 had been previously reported in patients with FSGS [28] and increased renal tubular expression of MCP-1 was detected in FSGS renal biopsies [6]. Wilkening et al. [13] show that there is increased gene expression for MCP-1 and CCR2 in microdissected FSGS human glomeruli. This is an interesting finding since FSGS is not a typical proliferative glomerulonephritis. However, these results were based on microarray analysis without validation by other techniques. The authors did not determine the intrinsic glomerular cell types expressing MCP-1 and CCR2 in diseased glomeruli. However, they did show that soluble urokinase-type plasminogen activator receptor and recombinant TNF-α stimulated cultured podocytes to produce MCP-1. Importantly, their study demonstrated that genetic deletion of CCR2 in the adriamycin-induced murine FSGS model reduced macrophage numbers in the glomeruli and tubulointerstitium and simultaneously reduced the severity of glomerular and tubulointerstitial fibrosis [13]. Increased renal tubular expression of MCP-1 has been detected in the renal biopsies of patients with a variety of glomerulonephritis (IgA nephropathy, membranous nephropathy, membranoproliferative glomerulonephritis and FSGS) in close proximity to infiltrating monocytes/macrophages [38]. Furthermore, inflammatory cytokines such as IL-1 and TNF-α are capable of stimulating cultured renal tubular epithelial cells to release MCP-1 [6]. In another study of patients with a range of kidney diseases (including thin glomerular basement disease, IgA nephropathy, ischaemic/hypertensive nephropathy, FSGS, membranous nephropathy, diabetic nephropathy, minimal change nephropathy, primary amyloidosis, light-chain nephropathy), uMCP-1 was shown to correlate with the number of interstitial macrophages [28] (Table 1). Thus diseases of diverse aetiology may share a final common pathway for disease progression characterized by monocyte/macrophage infiltration and chronic tubulointerstitial injury. Of interest, the uMCP-1:Cr ratio was found to be elevated in patients with congenital obstructive nephropathy and surprisingly improved following surgical treatment [29]. Unexpectedly, very high levels of uMCP-1 were also detected in patients with Fanconi syndrome due to Dent’s disease, Lowe’s syndrome and autosomal dominant idiopathic Fanconi syndrome [30]. It is unclear whether these high levels relate to increased local tubular synthesis or decreased tubular reabsorption of filtered MCP-1. Autosomal dominant polycystic kidney disease (ADPKD) is the third or fourth most common cause of kidney failure worldwide [39]. The major focus in ADPKD research has been on understanding how increased tubular cell proliferation and fluid secretion lead to cyst growth. Nonetheless, there is now increasing evidence that other factors (including metabolism, autophagy, inflammation and innate immunity, oxidative stress, epigenetic modification) could contribute to the severity of disease progression in ADPKD (recently reviewed by Chang and Ong [40]). Although ADPKD is not an obvious disease associated with monocyte/macrophage infiltration, early data from experimental models indicated a modifying role for MCP-1 in disease pathogenesis. Increased renal expression of MCP-1 was detected and localized to the cyst-lining epithelium in a non-orthologous rat PKD model (Han: SPRD) [41] and in murine Pkd1 models [42]. Significantly tubule-specific deletion of Mcp1 or administration of a CCR2 inhibitor to Pkd1 mutant mice reduced the number of pericystic macrophages, cyst formation and tubular injury and improved kidney function [43]. MCP-1 has also been detected in renal cyst fluid, cyst lining of end-stage ADPKD kidneys and urine (but not serum) samples of ADPKD patients, indicating de novo tubular synthesis [31]. Importantly, uMCP-1 levels were positively correlated with the rate of loss of eGFR in another study [34]. In the pivotal Tolvaptan Efficacy and Safety in Management of Autosomal Dominant Polycystic Kidney Disease and Its Outcomes 3:4 trial in patients with ADPKD, there was an 11–24% reduction in uMCP-1 over 36 months in the tolvaptan-treated group [33], thus raising the possibility that MCP-1 could be a sensitive biomarker of therapeutic response in ADPKD trials. In the CCX140B trial in type 2 diabetic patients with nephropathy, blocking CCR2 resulted in a dose-dependent increase in plasma MCP-1 compared with the placebo group [36]. Only the lower dose of the CCR2 blocker reduced albuminuria significantly. The increase in plasma MCP-1 observed at the higher dose could have led to an increase in systemic inflammation counterbalancing the beneficial effects on renal inflammation observed at the lower dose. Conversely, the systemic delivery of MCP-1 inhibitor or a neutralization antibody could increase the concentration gradient between circulating plasma and tissue levels, paradoxically leading to increased leucocyte recruitment. In addition, circulating antibody–MCP-1 immune complexes may still retain biological activity. These potential issues have been highlighted by the results of a Phase 2 clinical trial of a human monoclonal antibody (mAb) to MCP-1 (ABN912) in rheumatoid arthritis [44]. After two infusions of ABN912, no clinical benefits were observed in the treated patients. However, there were very high serum levels of ABN912–MCP-1 immune complexes and synovial biopsies showed that the group treated with the highest dose of anti-MCP-1 mAb showed a paradoxical increase in the number of CD68+ macrophages in the synovium lining [44]. uMCP-1 has unexpectedly emerged as a sensitive biomarker of disease activity in a range of kidney diseases, not just in those characterized by acute inflammation, but also in others of a more chronic nature. The latter observation thus supports a disease-modifying role for chronic inflammation due to monocytes/macrophages in a common pathway of renal disease progression regardless of the initial aetiology. Definitive proof of this hypothesis is awaited since, to date, clinical trials attempting to block MCP-1 action (through CCR2 blockade) in diabetic nephropathy have been equivocal, possibly due to the redundancy of ligand–receptor interaction and/or unanticipated effects on systemic MCP-1 levels. Nonetheless, based on our emerging knowledge of disease pathogenesis, it seems likely that a combinatorial therapeutic approach targeting several key cytokines apart from MCP-1 in chronic kidney disease will be required. Other chemokines with powerful chemoattractant effects on monocytes/macrophages [such as Regulated on Activation, Normal T Cell Expressed and Secreted, macrophage inflammatory protein (MIP)-1α and MIP-1β) as well as profibrotic cytokines (TGF-β, CTGF) have been shown to be upregulated in preclinical and clinical studies. The simultaneous analysis of multiple urinary chemokines, including MCP-1, using multiplex assays is now achievable and may add to the utility of uMCP-1 as a prognostic, activity or therapeutic biomarker. These three articles add to the emerging body of evidence that uMCP-1, probably in combination with other soluble mediators (such as CD163, EGF and others), could be a clinically useful biomarker for disease onset or progression in acute and chronic kidney disease from multiple aetiologies. Future studies in larger multi-ethnic longitudinal patient cohorts should clarify whether it is a sufficiently sensitive biomarker, especially for chronic kidney disease. Whether inhibiting MCP-1 synthesis or action is beneficial in any form of kidney disease in humans remains to be shown. F.W.K.T. is supported by the Ken and Mary Minton Chair of Renal Medicine. This work was supported by the National Institute for Health Research Biomedical Research Centre, Imperial College Healthcare NHS Trust and Imperial College London. A.C.M.O. acknowledges funding from Kidney Research UK and the European Union. F.W.K.T. and A.C.M.O. contributed equally to this paper. F.W.K.T. has received research project grants from AstraZeneca, Baxter Biosciences, Boehringer Ingelheim, MedImmune and Rigel Pharmaceuticals and has consultancy agreements with Rigel Pharmaceuticals, Novartis and Baxter Biosciences. A.C.M.O. has received research support from ONO and served on advisory boards for Otsuka, Mironid and Sanofi-Genzyme. (See related articles by Moran et al. Urinary soluble CD163 and monocyte chemoattractant protein-1 in the identification of subtle renal flare in anti-neutrophil cytoplasmic antibody-associated vasculitis. Nephrol Dial Transplant 2020; 35: 283--291; Wu et al. Associations of urinary epidermal growth factor and monocyte chemotactic protein-1 with kidney involvement in patients with diabetic kidney disease. Nephrol Dial Transplant 2020; 35: 291--297; and Wilkening et al. C–C chemokine receptor type 2 mediates glomerular injury and interstitial fibrosis in focal segmental glomerulosclerosis.