Periostin in kidney diseases

Abstract

Chronic kidney disease is an incurable to date pathology, with renal replacement therapy through dialysis or transplantation being the only available option for end-stage patients. A deeper understanding of the molecular mechanisms governing the progression of kidney diseases will permit the identification of unknown mediators and potential novel markers or targets of therapy which promise more efficient diagnostic and therapeutic applications. Over the last years, periostin was established by several studies as a novel key player in the progression of renal disease. Periostin is de novo expressed focally by the injured kidney cells during the development of renal disease. In diverse cohorts of renal disease patients, the expression levels of periostin in the kidney and urine were highly correlated with the stage of the pathology and the decline of renal function. Subsequent studies in animal models demonstrated that periostin is centrally involved in mediating renal inflammation and fibrosis, contributing to the deterioration of renal structure and function. Genetic or pharmaco-genetic inhibition of periostin in animal models of renal disease was efficient in arresting the progression of the pathology. This review will summarize the recent advances on periostin in the field of kidney diseases and will discuss its utility of as a novel target of therapy for chronic kidney disease.

Introduction

Chronic kidney disease (CKD) is a major societal and economic burden for national health systems, with a continuously rising incidence that contributes to the increase of the number of deaths worldwide [1]. With currently no efficient prognostic or therapeutic options being available, the only possibility for the treatment of end-stage renal disease remains the renal replacement therapy through dialysis or transplantation. CKD may originate from various causes, such as diabetes, hypertension, immune or toxic stimuli attacking different kidney compartments; however, it is characterized by common pathologic pathways which involve chronic inflammation and excessive production of extracellular matrix (ECM) within the kidney, leading to gradual impairment of renal structure and function [2].

During the last years, a major focus in kidney research has been the deeper understanding of the pathophysiological mechanisms governing CKD progression, which will allow the development of early prognostic markers and targeted therapeutic treatments. This collective effort has led to the discovery of several novel mediators and promising targets of renal disease [3–7]. Drugs that were developed against some of these molecules, including well-known profibrotic and proinflammatory factors such as TGF-β1, CTGF, CCL2/CCR2, ET-1, TNF-α, have been efficient in preclinical studies and have entered early phase clinical trials on subgroups of CKD patients [8].

Periostin was identified by us and other investigators as a novel biomarker and/or mediator of CKD. Periostin is a 90 kDa matricellular protein with high expression in bone and dental tissues [9]. As is the case for other such proteins, the expression of periostin is high in development while restricted in adult tissues, but it is considerably upregulated in injury and wound-healing conditions [10]. The protein structure of periostin comprises distinct regions, including a tandem repeat of four fasciclin-I domains derived from the homologous insect domain which has been associated with neuronal adhesion. Periostin was shown to interact via its different domains with several proteins of the ECM, such as collagen-1, Notch-1, tenascin-C, BMP-1 and various cell-surface integrins [11–15]. Through these interactions periostin is capable to mediate signals both in the extracellular and the intracellular environment, thus modulating ECM assembly and cellular responses associated to adhesion, migration and differentiation.

Periostin has been shown to mediate inflammation and fibrosis the during disease of several organs, including the heart, lung, kidney, skin, liver, skeletal muscle and retina, while inhibiting periostin in animal models of these diseases was efficient in arresting the progression of the pathology [16–26]. The scope of this review is to summarize the latest research on the role(s) of periostin in kidney diseases and discuss the potential of periostin as a future target of therapy of CKD.

Involvement of periostin in renal disease

Insights from patient biopsies

Periostin was first implicated in the progression of renal disease by a study that identified high expression of the protein in cyst-lining cells and cystic fluid from autosomal dominant polycystic kidney disease patients [27]. Subsequently, given the known involvement of matricellular proteins in fibrotic diseases, Sen et al. performed a transcriptomic analysis in isolated glomeruli specimens from patients with proteinuric diseases, including focal segmental glomerulosclerosis, IgA nephropathy, lupus nephritis and mesangial nephropathy, to discover differentially expressed matricellular transcripts [28]. This study identified periostin as the most highly expressed protein among the targets and demonstrated a broad association of its expression with the progression of human nephropathies. In immunohistochemical analyses in biopsies of these patients, periostin was localized in areas of mesangial expansion, the tubulointerstitium and sites of fibrosis, while the degree of periostin expression was negatively correlated with renal function. Interestingly, the profibrotic factor TGF-β1 was shown as a potent inducer of periostin in cultured mesangial cells, while treatment of the cells with exogenous periostin was associated with increased proliferation and reduced apoptosis.

Subsequent studies in patient biopsies and urine samples confirmed the elevated expression of periostin in diverse renal diseases and highlighted the potential utility of periostin as a tissue or urine biomarker of the disease progression. In lupus nephritis patients, periostin was highly expressed in periglomerular areas, fibrous vessels and areas with interstitial fibrosis, while its expression levels were correlated with the chronicity of the disease and the decline of renal function [29]. Another study revealed high expression of periostin in sclerotic glomeruli and tubular epithelium of type 2 diabetes patients. Urinary periostin levels of these patients were significantly upregulated compared with healthy controls, whereas significant variations of the urine periostin values were also observed between normo-, micro- or macroalbuminuric diabetes patients [30]. In a large cohort of IgA nephropathy patients, high urinary periostin concentration was associated with an advanced stage of the disease and could predict worsening renal outcomes [31]. Accordingly, in a group of transplant recipients, urine periostin levels could accurately distinguish chronic allograft nephropathy patients from transplant subjects with normal renal function [32]. Interestingly, another study showed that urine periostin/creatinine ratio is significantly higher in patients with both proteinuric and non-proteinuric renal diseases, uncoupling urine periostin excretion from the dysfunction of the glomerular filtration barrier [33].

Insights from animal models of CKD

Models of vascular injury

To identify factors crucially implicated in the progression of renal disease, we administered losartan, an angiotensin-II receptor blocker, to l-NAME-treated hypertensive rats at the non-return point of the disease. This method resulted in either reversal or progression of renal disease after losartan treatment, which was followed by comparative analysis of the kidney transcriptome of the two experimental groups, aiming to reveal novel putative targets of therapy. Periostin was identified in this analysis as the target with the highest expression in the progression compared to the reversal group [34]. Moreover, periostin was predominantly localized in the injured fibrotic regions of the diseased kidneys, while its expression levels showed a close correlation with classical parameters of renal function such as plasma creatinine, proteinuria and renal blood flow. This was the first study to denote periostin as a potential diagnostic target for the evaluation of the progression or response to therapy of kidney diseases.

Our results where more recently confirmed in the two-kidney, one-clip (2K1C) rat model of hypertensive nephrosclerosis, where proteomic analysis of the renal cortex identified periostin as one of the most highly upregulated proteins, with a principal expression around fibrotic vessels [35]. Independent in vitro studies showed that angiotensin-II, a major factor involved in hypertensive nephropathy, is a potent inducer of periostin in vascular smooth muscle cells and cardiac fibroblasts through complex pathways mediated by PI3 kinase and Ras/ERK/TGFβ1 signaling, respectively [36, 37].

Models of tubular aggression

A first study investigating the expression and function of periostin in the models of 5/6 nephrectomy and unilateral ureteral obstruction (UUO), identified high expression of periostin in distal tubular epithelium and in tubulointerstitial areas [33]. Periostin overexpression in cultured mouse distal tubular cells downregulated the expression of E-cadherin and induced expression of mesenchymal markers like fibroblast specific protein-1 (FSP-1) and matrix metalloproteinase-9 (MMP-9), associating periostin function with the induction of a mesenchymal phenotype in tubular cells.

Subsequent studies by our group revealed a critical role for periostin in the progression of renal disease. Mice with genetic deletion of periostin displayed reduced interstitial fibrosis and inflammation associated with preservation of the renal epithelial phenotype in the UUO model, compared with their wild-type littermates [38]. Expression of periostin by renal epithelial cells was attributed to TGF-β1, while treatment of the cells with periostin could upregulate the expression of collagen I and stimulate mitogen-activated protein kinase (MAPK) pathways.

In parallel, other investigators identified a role for periostin in the progression of polycystic kidney diseases (PKD). Periostin was found to be highly expressed by cyst-lining epithelial cells and accumulate in the tubulointerstitial matrix adjacent to cysts of PKD patients. Deletion of the periostin gene in a genetic mouse model of PKD resulted in the reduced number of cysts, decreased proliferation and autophagy in the cystic kidneys, which were accompanied by reduction of interstitial fibrosis and preservation of renal function [39].

Models of glomerular injury

In a recent study, we investigated the mechanism of induction and the role of periostin in a severe model of nephrotoxic serum (NTS)-induced glomerulonephritis. As with the aforementioned models, deletion of the periostin gene resulted in a marked reduction of proinflammatory and profibrotic mediators in the diseased kidney accompanied by preservation of renal structure and function [40]. In addition, we used a pharmacogenetic approach via administration of antisense oligonucleotides to block the expression of periostin after the onset of the disease in the NTS model. This strategy revealed that inhibition of periostin at a later stage of the disease after establishment of proteinuria is efficient in arresting the progression of the pathology and preserving renal function, demonstrating the utility of periostin as a future potential therapeutic target of CKD. Moreover, using a combination of bioinformatics analysis, in vitro reporter assays and in vivo chromatin immunoprecipitation analyses, we showed that NFκΒ and other proinflammatory transcription factors are the major inducers of periostin in the NTS model [40]. This is in accordance with other studies showing that periostin is highly induced by the interleukins IL-4 and IL-13 in inflammatory conditions such as bronchial asthma [18, 41]. Moreover, periostin was shown to mediate immune responses and promote production of proinflammatory chemokines in allergic inflammation and inflammatory lung diseases [19, 42]. The function of periostin in the NTS model was associated to activation of the integrin αvβ3 signaling cascade in glomerular podocytes and parietal epithelial cells, inducing migration, survival and proliferation pathways and amplifying the inflammatory and fibrotic responses [40]. Periostin has been previously shown to mediate adhesion and migration of different cell types, such as cancer cells, vascular smooth muscle cells and macrophages, though interactions with cell-surface integrins [15, 43, 44].

Another study examined the role of periostin in a genetic mouse model of lupus nephritis [45]. Lupus mice displayed upregulated periostin expression in glomerular mesangial cells accompanied by increased proliferation, ECM production and expression of platelet-derived growth factor-B (PDGF-B). The expression of periostin was induced by PDGF-B in cultured mouse mesangial cells and was shown to mediate increased proliferation and secretion of fibronectin downstream PDGF-B and the activation of PI3 kinase/Akt pathways. The described mechanisms of periostin induction and function in renal disease are summarized in Fig. 1 and Table 1.

Fig. 1.

Schematic illustration of how periostin mediates progression of renal disease. After an initial aggression, different growth factors, cytokines and signaling pathways upregulate periostin expression. Periostin interacts with its integrin receptors on the cell surface, inducing inflammation and matrix assembly and promoting cell adhesion, migration and proliferation. These pathways amplify the inflammatory and fibrotic responses inducing a vicious circle of renal damage

Table 1.

Physiopathological functions of periostin in animal models of CKD

Animal model	Periostin expression sites	Function	References
l-NAME-induced hypertensive nephropathy	Renal vessels, vascular smooth muscle cells, injured tubules	Periostin levels are correlated with the decline of renal function. Treatment of animals with periostin antisense protects from glomerulosclerosis and perivascular/interstitial fibrosis	[34, 38]
2-Kidney, 1-clip (2K1C) rat model of hypertensive nephrosclerosis	Fibrotic vessels	Relative abundance of periostin is sufficient to differentiate 2K1C animals from controls.	[35]
5/6 Nephrectomy (5/6 Nx)	Distal tubular cells, tubular interstitium	Periostin induces a mesenchymal phenotype in tubular cells by decreasing E-cadherin and upregulating MMP-9 and FSP1	[33]
Unilateral ureteral obstruction (UUO)	Tubular cells, tubular interstitium	Periostin treatment of cultured tubular cells stimulates expression of collagen I and MAPK pathways. Periostin null mice are protected from renal inflammation and interstitial fibrosis	[33, 38]
Polycystic kidney disease (pcy/pcy mice)	Cyst-lining tubular epithelial cells	Periostin activates the mTOR pathway to mediate cyst proliferation, autophagy and interstitial fibrosis	[27, 39]
Glomerulonephritis induced by nephrotoxic serum (NTS-induced GN)	Glomerular podocytes, parietal epithelial cells, vascular smooth muscle cells	Periostin is induced by proinflammatory transcription factors to activate integrin αvβ3 signaling pathway. Null mice display reduced inflammation and fibrosis and preservation of renal function	[40]
Lupus nephritis (MRL/lpr mice)	Glomerular mesangial cells	Periostin is induced by PDGF-B to promote mesangial cell proliferation and matrix production	[45]

Periostin as a target of therapy in CKD

During the last decades, a main focus of renal research has been the identification of novel biomarkers and/or targets of therapy for kidney diseases, which promise a more effective diagnosis and treatment of CKD patients. Although a lot of progress has been made towards the discovery of unknown mediators of renal disease, many among them were proven inefficient in clinical practice, while the considerable time and cost required for the generation and testing of potential drugs further impedes the clinical application of the new candidate targets.

Periostin fulfils several criteria for representing a promising biomarker of CKD. Although its expression is minimal in healthy tissue, periostin is highly upregulated in the kidney in diverse renal pathologies, both in animal models and human renal disease, while its expression levels were found to correlate with the decline of renal function [28–30, 34] (Table 1). Given the nature of periostin as a secreted molecule, it can be easily detectable in plasma or urine samples. Accordingly, periostin levels were found elevated in the urine of several types of CKD patients, including diabetes, IgA nephropathy, chronic allograft nephropathy, lupus nephritis, focal segmental glomerulonephritis and polycystic kidney disease [30–33]. Further studies are necessary to investigate whether measurement of periostin levels in biological fluids can be applied in the clinic as an early and sensitive marker of kidney damage.

The localization of periostin in the ECM renders the protein an easily accessible target of potential drugs. Moreover, the expression of periostin is focal in the damaged tissue, associated to the sites of injury, e.g. periostin is expressed by renal vessels in hypertensive nephropathy [34, 35], by renal tubules in UUO, 5/6 nephrectomy and polycystic kidney disease [33, 38, 39] and by glomerular cells in several glomerulopathies [28, 40]. Inhibition of periostin using knock-out mice or antisense oligonucleotides preserved renal structure and function in several animal models of CKD, demonstrating that periostin has the potential of being a target of therapy for kidney diseases [34, 38–40] (Table 1). However, targeting periostin in human renal disease requires the generation and validation of targeted drugs, for example specific neutralizing antibodies, blocking peptides or antisense with increased stability suitable for use in humans, which may necessitate further elaboration on the specific interplay between periostin and its interaction partners as well as collaboration of different fields of research.

Concluding remarks and perspectives

Understanding the molecular basis of kidney diseases will advance our knowledge on the mechanisms participating in the development of CKD and will provide us with novel tools to counteract the progression of end-stage renal disease. Periostin has been identified as a novel mediator of renal disease centrally involved in the processes of renal inflammation and fibrosis, with potency to constitute a future biomarker or therapeutic target of CKD. Periostin is de novo expressed in the kidney after renal damage while its inhibition efficiently protects from the progression of renal disease in animal models, which renders periostin an appealing biomarker or target of therapy. Targeting periostin in clinical practice will require the development of validated drugs for use in humans and the follow-up of large cohorts of patients, which is subject of further investigation.