This is the first LRG1 study in the population of biopsy-proven LN so far. In this study, we found that in patients with LN, LRG1 was expressed in plasma, several white blood cells of peripheral blood, at proximal tubule and several inflammatory cells of kidney. And we also examined its potential effects on kidney.
In plasma, level of LRG1 might be a good indicator of renal function and renal disease activity of LN. pLRG1 was positively correlated with serum creatinine levels and stages of CKD. It was reported that high level of serum LRG1 pointed to a subclinical kidney insufficiency in classical galactosemic patients by proteomics methods [27]. Furthermore, fluctuations of pLRG1 levels were found to reflect renal function in LN patients, thus potentially serving as a helpful biomarker in the clinical follow-up. Meanwhile, pLRG1 was a biomarker of renal disease activity of LN. In patients with highest rSLEDAI score, level of pLRG1 was higher than that in lowest rSLEDAI score. LRG1 was reported to be involved in inflammatory and autoimmune diseases. In active LN, the inflammatory and autoimmune reactions were more serious. In previous study, serum LRG1 concentrations correlated with disease activity in SLE, rheumatoid arthritis, Crohn’s disease, ulcerative colitis and adult-onset Still’s disease [12, 20, 22, 24, 32].
It was reported that serum LRG1 was elevated in patients with SLE and correlated with disease activity [24]. Similarly, we found LN patients with higher SLEDAI had higher pLRG1 concentrations. Differently, in our study, we included biopsy-proven LN patients instead of SLE patients. We found pLRG1 was elevated in LN patients compared to HC and might be a good indicator of renal function and renal disease activity of LN.
One of the strengths of our study was the examination of pLRG1’s ability to predict renal pathology. Higher levels of pLRG1 existed in patients who had wireloop sign, heavy inflammatory cells infiltration and fibrosis in the kidneys. Wireloop sign characterized by a large number of immunocomplex deposited under the endothelium, and heavy infiltration of inflammatory cells indicated severe inflammatory and immune reactions in kidney, which was consistent with the association between pLRG1 and rSLEDAI above. Fibrosis in renal tubular interstitium reflected chronic injury and loss of function of kidney, which was also consistent with the association between pLRG1 and renal function above. Recently, Lee et al. revealed that urinary LRG1 was a potential biomarker for renal tubular injury in mouse albumin overload model [29]. Therefore, LRG1 might be associated with renal inflammation, immunity, fibrosis and injury.
In addition, we revealed that pLRG1 elevated in patients with hyperplasia of renal endothelial cells. And the stimulation of rhLRG1 induced proliferation of HUVEC cell line. This might result from the following two possible reasons. On the one hand, LRG1 might be involved in renal inflammation and deposition of immune complexes, causing injury and further hyperplasia of endothelial cells. On the other hand, LRG1 was reported mitogenic to endothelial cells by modulating endothelial TGF-β signaling [30]. Hypercellularity of endothelial cells was one of basic pathological regenerations after injury in the glomeruli, which would further cause hematuria, proteinuria and decrease in glomerular filtration rate.
In our study, level of pLRG1 was elevated in LN patients compared with that in HC. The possible causes were as follows. First, we found in neutrophils and NK cells of peripheral blood, the LRG1 expression increased in LN patients compared to that in HC. Second, as a secretory protein, LRG1 was mainly expressed at proximal tubule. IHC showed the expression of LRG1 increased in kidney tubules of LN patients. Moreover, several inflammatory cells including macrophages, T cells and B cells of kidney could also expressed LRG1. So the secretion from kidney of LN patients might be increased.
In kidney, the expression of LRG1 also increased in LN patients. And the expression of LRG1 was positive in proximal tubule, and negative in distal tubule or collecting tube. Besides, some inflammatory cells infiltrated in glomerulus and renal interstitium, including macrophage, T cell and B cell, could also express LRG1. LRG1 could be induced by plasma of LN patients in HK-2 cell line. In order to determine whether proinflammatory cytokines might promote the production of LRG1, we chose IL-1β, IL-6, TNF-α and INF-γ to directly stimulate HK-2 cell line because LRG1 was expressed mainly at proximal tubule. It was reported that these four cytokines increased in LN and were important pathogenic factors. We found LRG1 could be induced by IL-1β and IL-6, but not TNF-α and INF-γ.
By combined with Figure 2 and 3, we found in the fibrotic renal tissue, the expression of LRG1 was reduced. Because LRG1 was mainly and extensively expressed at the proximal tubule and proximal tubule was destroyed in the fibrotic tissue. However, several inflammatory cells, including macrophages, T cells and B cells, could also express LRG1. But, the total amount of LRG1 expression was reduced. Figure 1F showed that plasmic LRG1 level was higher in LN patients with renal fibrosis. The opposite conclusion might be explained by the following reasons: Firstly, Figure 1F was the result of LRG1 level in plasma, but not kidney; Secondly, in the process of renal fibrosis, the destroyed proximal tubules might release LRG1 into the blood. However, the exact reason needs further studies.
Furthermore, we revealed that LRG1 might be associated with inflammation, proliferation and apoptosis of the endothelial cell. It could induce chemokine (C-C motif) ligand 5 (CCL5), but reduce CCL20, CXCL2, IL-4, IL-6 and IL-23 in HUVEC cell line. pLRG1 was elevated in patients with high renal disease activity and LRG1 could be induced by some inflammatory cytokines. Meanwhile, the stimulation of LRG1 could increase CCL5, but reduce some other cytokines. LRG1 might have bidirectional regulation of inflammation.
The reduced proinflammatory cytokines IL-6, IL-23, CCL20 and CXCL12 played antiinflammatory role in LN. IL-6 is a lymphokine produced by activated T cells and fibroblasts, which can make B cell precursors become antibody producing cells. In collaboration with colony stimulating factor, IL-6 can promote the growth and differentiation of original bone marrow-derived cells and enhance the lysis function of natural killer cells [33]. IL-23 is released from antigen presenting cells and induces expansion of Th17 cells and is necessary for their maintenance, thus forming the IL-23/IL-17 axis [34]. CCL20 is a chemokine of CC subfamily, which directly participates in the directional migration of dendritic cells and T cells through its ligand CCR6. CXCL12 is a classic inflammatory chemokine and can be secreted by a variety of cells, including lymphocytes, monocytes, and endothelial cells, to bind to its receptor, CXCR4. In glomerulonephritis, CXCR4 is overexpressed in parietal epithelial cells of the kidney, triggering their migration into the glomerular tuft where they form hyperplastic lesions [35]. LRG1 might participate in the inflammatory response by reducing these proinflammatory factors.
In HUVEC cell line, we found the stimulation of rhLRG1 not only promoted cell proliferation but also inhibited apoptosis, mainly late apoptosis. It was well known that inadequate apoptosis and excessive apoptosis were both harmful to LN, such as inadequate apoptosis of lymphocytes targeting autoantigens and excessive apoptosis of renal tubular cell. In fact, for endothelial cells, apoptosis and hyperplasia were both its reactions to injury. LRG1 inducing proliferation and reducing apoptosis of endothelial cells might together lead to proliferation of endothelial cells, which influenced renal structural and functional integrity. In this respect, LRG1 might be might be involved in pathological changes of kidney to injury.
However, pLRG1 did not correlate well with proteinuria, which is another important marker of LN severity. The possible reasons are as follows: Firstly, proteinuria is a relative early manifestation of LN, and renal insufficiency is a more serious manifestation of the disease progression. Proteinuria and renal function are not necessarily parallel, and their risk factors are not identical. Secondly, in our study, we found LRG1 had more effect on proliferation of endothelial cells, which might influence renal function. Proteinuria was effected more on damage to podocytes and charge barrier. However, the exact mechanism needs our further research on lupus and LRG1.
Our study has several limitations. First, the sample size of LN biopsies was relatively modest. Second, due to the cross-sectional design, the serial change in pLRG level according to the change in the progression or relief of LN and the association between pLRG1 and prognosis of LN could not be defined. Further large-sized, longitudinal studies were warranted.