Urinary Erythrocyte-Derived miRNAs: Emerging Role in IgA Nephropathy Open Access

Hematuria is one of the basic clinical manifestations of IgA nephropathy (IgAN). Either hematuria can be an isolated indicator unaccompanied by other renal abnormal indicators, or it can be accompanied by other renal abnormal indicators, such as proteinuria, hypertension, and elevated serum creatinine, and so on. Hematuria has been closely related to capillary proliferation and the formation of crescents in IgA nephropathy. However, the idea that hematuria may affect the progression of IgAN is still controversial. Most researchers believe that persistent microscopic hematuria is a poor prognostic factor of IgAN [1, 2]. However, other researchers believe that simple microscopic hematuria or episodes of gross hematuria are good prognostic factors. Presumably, one reason for the controversy is that some nephrologists ignore the erythrocyte contents themselves and oversimplify their role in IgAN.

The traditional view assumes that mature red blood cells are highly differentiated cells and in their maturation process, they gradually lose organelles and nucleic acids to prepare for the storage of hemoglobin. In the past, research had shown that erythrocytes did not have any nucleic acid substances, and they had no transcriptional activity [3]. So, according to the earlier point of view, erythrocytes were merely a filled hemoglobin pocket [3]. Subsequent studies indicate that they contain abundant long, non-coding RNA [4]，miRNA [4-7] and Y RNA [8]. Furthermore, the miRNA expression in the erythrocyte has been found to account for the majority of the miRNA expression in whole blood [5]. Moreover, 2289 types of proteins can be found in human red blood cells [9], including membrane skeletal proteins, metabolic enzymes, transporters and channel proteins, adhesion proteins, hemoglobin, cellular defense proteins, proteins of the ubiquitin-proteasome system, G proteins of the Ras family, kinases, chaperone proteins, proteases, and translation initiation factors. Therefore, with a systematic analysis of the effects of red blood cell contents (rather than the filled hemoglobin pocket) on IgAN, future studies could unearth new biomarkers and could provide new ideas regarding their involvement in IgAN progression.

Normal urine does not contain or only contains a very small quantity of red blood cells. Using 10 ml of fresh urine for centrifuging, smearing and observating with a microscope, for each high-power field of view, more than three erythrocytes can be considered to be hematuria. Hematuria can be divided into two types of hematuria: macroscopic hematuria and microscopic hematuria. The mechanisms of occurrence between the two are different. Research indicates that macroscopic hematuria generated in IgAN patients is closely related to CX3CR1-fractalkine axis activation [10]. Most studies confirm that macroscopic hematuria is negatively correlated with a long-term prognosis for IgAN [11]. And macroscopic hematuria is closely related to the incidence of crescents in IgAN patients [12]. Moreover, in patients with anticoagulation-associated gross hematuria for whom renal biopsy excluded glomerulonephritis, 66% of them do not recover baseline renal function [13]. This raises a note of caution about oral anticoagulation in patients with kidney disease.

However, whether isolated microscopic hematuria has an impact on the long-term prognosis of IgAN remains controversial. On one hand, some studies show that the long-term prognosis of patients with isolated microscopic hematuria is good. These patients have a low risk of developing hypertension or renal impairment [14]. On the other hand, a retrospective cohort study of 1.2 million by Israel showed that young patients with persistent isolated microscopic hematuria, compared with the general population, had a significantly increased risk of developing end-stage renal disease (ESRD). With an average of 22 years of follow-up, 0.7% of the patients with persistent isolated microscopic hematuria developed ESRD, which was significantly higher than 0.045% for the normal control group. The hazard ratio (HR) was 18.5 (95% confidence interval 12.4–27.6) [15]. We speculate that isolated microscopic hematuria is a low-risk factor for the long-term prognosis of chronic kidney disease (CKD), especially IgAN. However, when compared with the normal control group, it remains a certain risk, especially for a long follow-up time.

Most studies of IgAN find that microscopic hematuria combined with proteinuria (minimal or massive) is closely related with a deterioration of long-term renal outcomes [16-18]. One study revealed that during a median follow-up of 7 years, for patients in which IgA nephropathy presented with hematuria and minimal proteinuria (less than 0.4g/d), more than 40% of patients had evidence of further kidney injury, such as proteinuria of 1 g/d or more (33%), hypertension (26%), or renal dysfunction (7%) [17]. Moreover, patients with hematuria and minimal or no proteinuria may even have severe renal histological injury (Hass classification Grade II-IV) [18]. Study also showed that most patients with thin basement membrane nephropathy with persistent isolated microscopic hematuria could maintain renal function in their lifetime unless they had concurrent proteinuria [19]. The exact reasons are unclear, but we speculate that it may be related to a synergistic effect with other risk factors, such as abnormal glycosylated IgA1, microalbuminuria, and so on.

In 1979, Birch and Fairley were the first to associate the shape of urinary erythrocytes with kidney disease [20]. Since then, this effective method of diagnosing kidney disease has been demonstrated by many studies [21-23]. The evaluation of red blood cell (RBC) morphology may prevent the patient from being exposed to inappropriate invasive diagnostic tests such as kidney biopsy or cystoscopy in the case of hematuria with isomorphic RBCs or dysmorphic RBCs, respectively. The proportion of dysmorphic erythrocytes found in the diagnosis of glomerular hematuria ranges from 10%–90% [21-23]. In addition, urinary acanthocytes have the highest specificity for the diagnosis of glomerular hematuria, greater than 1%, which is a hint of glomerular hematuria; greater than 5%, the specificity is 100% [24]. All methods for urine microscopy can be roughly divided into two kinds: automated urinary microscopy and manual urinary microscopy. There are also two types of machines for counting urine particles, using either flow cytometry with fluorescent dyes or software analysis of digitized microscopic images. The advantages of automated instruments are obvious; they can analyze large numbers of samples and afford much higher power fields in short periods. More fields can be surveyed to detect dysmorphic RBCs such as acanthocytes at very low percentages. Laser-based flow cytometry is very useful in screening hematuria [25, 26], but its produces scattergrams rather than images, and microscopy is still required to differentiate dysmorphic and isomorphic RBCs. A more relevant approach to urinary microscopy uses computerized analysis of digitized monochrome images of urinary particles [27]. One advantage of digitized instruments is that the actual images can be reviewed by the pathologist or nephrologist to inform their opinion on the patterns, and sending images to another hospital is also possible [27]. Manual urinary microscopy is inefficient and requires high expertise, but it is still the gold standard, inexpensive, and suitable for office practice. A phase-contrast microscope is the best tool for differentiating dysmorphic from isomorphic RBCs in urine microscopy.

The erythrocyte proteome is composed of 90% hemoglobin. The remaining 10% is largely unexplored. The mechanism for research in hematuria in IgAN mostly comes from hemoglobin. Glomerular hematuria, similar to proteinuria, may result in tubular injury. This evidence rests on insights from basic research on the molecular mechanisms of hemoglobinuria- and myoglobinuria-induced tubular injury as well as on the clinical evidence of macrohematuria-associated AKI in the context of IgA nephropathy [28]. Studies identify, from the casts (red blood cell casts and hemoglobin casts, etc.), a clogging effect progressing to the toxic effects of hemoglobin. The role of the cast blockage and oxidative stress caused by hemoglobin has been clearly identified. Studies on gross hematuria-related AKI in IgAN patients show that, when damaged red blood cells release hemoglobin, it is taken up by renal tubular epithelial cells, and causes the release of heme and eventually decomposes into free iron. Hemoglobin can be ingested by the proximal tubule epithelial cells [29], through the megalin-cubilin receptor. Hemoglobin can be decomposed into oxygen free radicals, lipid peroxides, heme, and free iron in renal tubular epithelial cells [30] and reduces NO (nitric oxide) production. Therefore, it will result in renal vasoconstriction and ischemia, and eventually lead to kidney damage [28]. Intracellular hemoglobin will also increase iron derived from hydroxyl radicals [30], apoptotic enzyme activation and apoptosis [31], and the release of pro-inflammatory cytokines [32]. Heme is a strong oxidizing agent; it can activate the pro-inflammatory and pro-fibrotic pathways, thereby inducing a chronic inflammation of the kidney [33, 34].

One class of endogenous non-coding RNA is microRNA, which is 20–25 nucleotides in length. In combination with the principle of complementary base pairing of target mRNA 3’ untranslated region, the nucleotides affect the level of translation of gene expression, resulting in the target mRNA appearing to be inhibited or degraded [35]. Rathjen et al. confirmed that mature red blood cells have microRNA during their research of plasmodium miRNA in 2006 [7]. Then, they systematically analyzed miRNA within mature red blood cells and reticulocytes. The miRNA chips showed that the mature red blood cell miRNA accounts for the vast majority of the entire miRNA expression in the blood [5]. Recently, with advances in high-throughput sequencing technologies, their study identified 287 known and 72 putative novel microRNAs that included many microRNAs that are either uncharacterized, or were not previously associated with erythrocytes [4]. Moreover, in addition to microRNAs, mature erythrocytes also contain long transcripts. Although smaller in number, erythrocyte long transcripts are highly enriched in annotated exons and largely encode proteins critical to erythrocyte differentiation and functions [4].

A large number of studies [36-38] comfirm that miRNAs can be biomarkers of IgA nephropathy, such as miR-148b, let-7b, miR-3613-3p, miR-200a, miR-200b, and so on. Due to being non-invasive and easily obtained, urinary sediment miRNAs have become a new direction for searching for non-invasive biomarkers of IgAN. We also carry out some useful attempts [37, 39], and find hundreds of candidate miRNA biomarkers of IgAN. Red blood cells are one of the main cell types existing in the urinary sediment of IgAN. Thus, can erythrocyte-derived miRNAs be non-invasive biomarkers of IgAN? Our research found that [39], compared with the normal control group and the disease control group, miR-25-3p, miR-144-3p, and miR-486-5p of urine sediment were significantly increased in the IgA nephropathy group. These three miRNAs had good specificity and sensitivity for the diagnosis of IgAN by receiver-operating characteristic curve analysis, in which the area-under-the-curve (AUC) value of miR-486-5p was the largest at 0.935. Based on CD235a (the markers of human mature RBC) magnetic bead separation of urinary sediment, we confirmed that these miRNA mainly came from red blood cells [39]. In addition, mononuclear cells, human primary renal tubular epithelial cells, and red blood cells are the main cell types in urinary sediment. Our research also compared the baseline expression levels of miR-25-3p, miR-144-3p, and miR-486-5p in these three cell types. We found the baseline expression levels of these miRNAs in red blood cells are the highest [39]. So erythrocyte-derived miRNAs could be specific biomarkers of IgAN. However, if hematuria as a indicator for the diagnosis of IgAN, the AUC value was only 0.787. Aside from these three miRNAs that we found, are there other erythrocyte-derived miRNAs that can be used as candidates for noninvasive biomarkers of IgA nephropathy? To that end, we perform a combined analysis of the top 50 expressions of erythrocyte-derived miRNAs [4] and IgAN urinary sediment miRNA arrays [39], trying to find more erythrocyte-derived miRNA biomarkers of IgA nephropathy. The results show that, among the top 50 expressions of erythrocyte-derived miRNAs, there are 33 (66%) increased expression levels in the IgAN urinary sediment miRNA chips (Table 1). MiR-144 and miR-451 (miR-144/451) are processed from a single gene locus. http://106.39.55.6: 8080/medlib/s/com/nature/www/G.http/onc/journal/v34/n25/full/onc2014259a.html - bib7 And miR-451 is one of the top 50 expression miRNAs. However, miR-451 was no statistical differences in our IgAN urinary sediment miRNA arrays. The P value between the IgAN group and control group was 0.0608, but the fold change was 182.67. The reason for it may be a small sample of miRNA arrays. Erythrocyte-derived miRNAs are an important part of IgAN urinary sediment miRNA. Future research on urinary erythrocyte-derived miRNAs could offer a new and promising direction in the search for IgAN biomarkers.

RNAase presents in all types of body fluid and free miRNAs will quickly be degraded. Due to the special transportation of miRNA, such as microvesicles, exosomes and apoptotic vesicles (apoptotic bodies), and ribonucleoprotein complexes, miRNAs are very stable in blood circulation or urine. Apoptotic bodies are 50–5, 000 nm diameter and released by cells undergoing programmed cell death [40]. Microvesicles are membranous vesicles (100–1, 000 nm diameter) that are produced by budding from the plasma membrane [40]. Exosomes are 30–100 nm diameter vesicles considered to be of endocytic origin [40]. The possible transport types of erythrocyte-derived miRNAs are microvesicles and apoptotic vesicles. Erythrocyte-derived microvesicles are shed from the membrane surface, which expresses transmembrane proteins band 3, glycophorins, and blood group antigens but is depleted of cytoskeletal proteins, such as spectrin and ankyrin [41]. In normal circumstances, erythrocyte-derived microvesicles account for approximately 7.3% of the whole blood microvesicles, and 38.5% accounts for platelet-derived microvesicles. The endothelial-derived microvesicles have the largest number, accounting for 43.5% [42]. However, under the conditions of hemolysis, such as sickle cell anemia [43], thalassemia [44, 45], or malaria [46], the numbers of RBC microvesicles in circulation increase and can predominate. Moreover, malaria research finds that hemoglobin overflow stimulates oxidative stress, causing erythrocytes to increase their release of microvesicles, which can be inhibited by N-acetylcysteine [46]. So, the production of erythrocyte-derived microvesicles may be related to hemolysis and oxidative stress. This is similar to the hemolysis and oxidative stress mechanism of urinary erythrocytes when they are compressed while exiting out of an impaired nephron (Fig. 1). Dissolution of urinary red blood cells can also produce large amounts of hemoglobin and stimulate oxidative stress. Oxidative stress can not only make red blood cells release more microvesicles but also make red blood cells undergo apoptosis and secrete apoptotic vesicles (Fig. 1) [47].

Recent research shows that, similar to nucleated cells, erythrocytes may also undergo erythrocyte apoptosis (eryptosis), and its characteristics are cell shrinkage, cell membrane blebbing, and cell membrane phospholipid scrambling leading to the production of apoptotic vesicles [48]. The basic condition of inducing erythrocyte apoptosis is the opening of the unspecific cation channels, which increases the concentration of Ca2+ in the cytoplasm. Cytosolic Ca2+ further activates calpain and degrades cytoskeletal proteins, which causes cell membrane blebbing and apoptotic body formation. More importantly, Ca2+ can stimulate irregular cell membranes, accompanied by the asymmetric breakdown of membrane phospholipids, resulting in exposure of the phosphatidylserine of the cell membrane surface. Erythrocyte apoptosis can be stimulated by the change of osmotic stress, oxidative stress, energy consumption, the change in pH value, high fever, and many xenobiotics and endobiotic substances [49]. In addition, uremic toxins (such as indoxyl sulfate), prostaglandin E2, azathioprine, and ciclosporin can also promote erythrocyte apoptosis [49]. In addition, EPO (erythropoietin), chloride, nitric oxide, catecholamine, uric acid, and activated Adenosine 5’-monophosphate (AMP)-activated kinase (AMPK) can inhibit erythrocyte apoptosis [49]. Recent studies show that red blood cells can also occur in necroptosis [50], which shares the same key molecular mechanisms with nucleated cells, and is also accompanied by the formation of apoptotic bodies. Ceramides, the formation of NADPH oxidase and iron related to reactive oxygen species (ROS), and advanced glycation end products (AGEs) are involved in RBC necroptosis, and inhibiting these pathways can reduce the degree of RBC necroptosis. IgA nephropathy patients have oxidative stress (caused by the carbonyl stress product, hemoglobin, heme, and iron ions) [51], a change in osmotic pressure of different renal tubular segments, and a change in pH value and other factors. When urinary red blood cells are compressed while exiting out of an impaired nephron, under the stimulation of oxidative stress, the change in osmotic pressure and pH, local prostaglandin E2, and so on, red blood cells undergo erythrocyte apoptosis and necroptosis, cell shrinkage, cell membrane blebbing, and cell membrane phospholipid scrambling leading to the production of apoptotic vesicles (Fig. 1). Our study has also indirectly confirmed this view. We find that the level of urinary vesicles miR-486-5p in the IgAN group is abundant, and is significantly higher than that in the control group. The urinary sediment miR-486-5p in the IgAN group mainly comes from urinary erythrocytes.

A large number of studies demonstrate that circulating miRNA in microvesicles or exosomes can carry genetic information from one cell to another, and thus play an important role in intercellular communication [52-54]. After absorbing miRNA-rich microvesicles and exosomes, the transcriptome of recipient cells will change [55]. Thus, circulating miRNA is increasingly becoming an active area of research in nephrology for exploring different diseases, such as kidney transplant [56, 57], acute kidney injury [58] and chronic kidney disease [59], occurrence, progression, prognosis of biomarkers, and kidney injury mechanism.

Erythrocyte apoptosis is the coordination of programmed cell death, accompanied by the formation of apoptotic bodies, and results in defective cells without cell membrane breakage and the release of the cell contents [48]. Erythrocyte apoptosis and the accompanying apoptotic body formation is a protective mechanism of red blood cells, to prevent the lysis of injured erythrocytes and release their contents (hemoglobin), thereby reducing the release of hemoglobin causing renal tubular injury and the incidence of AKI. Apoptotic bodies that are produced by erythrocyte necroptosis also contain Fas ligand (FasL) [60]. Erythrocyte apoptosis and hemolysis can produce miRNA-containing vesicles (including microvesicles and apoptotic vesicles). Studies carried out on animal models suggest that kidney tissue can uptake microvesicles in the blood [61]. Moreover, renal parenchymal cells, such as renal tubular epithelial cells, can phagocytose red blood cells [29, 30]. When kidney diseases occur with gross hematuria or microscopic hematuria, renal biopsy confirms the swallowed red blood cells present in the cytoplasm of the proximal tubular epithelial cells [62]. Therefore erythrocytes themselves or erythrocyte-derived microvesicles can be swallowed by renal parenchymal cells. Animal models of hematuria and human biopsy tissues also successfully capture the process of penetration in the Bowman capsule of white blood cells, red blood cells, and fibrinogen, and find that red blood cells exist in cellular crescents [63, 64]. The miR-486-5p is the most abundant microRNA in red blood cells [4]. Our research confirms that that urinary sediment miR-486-5p mainly comes from red blood cells. The expression levels of miR-486-5p in microvesicles extracted from urine supernatant are higher in the IgA nephropathy group than in the control group [39]. A review of the literature about miRNA array results for IgAN renal tissue [65] finds that miR-486 expression of IgAN renal tissue is 6.99 times higher than the normal control group, and miR-486 is the second highest of all IgAN-elevated miRNAs. Therefore, erythrocyte-derived vesicles (microvesicles and apoptotic vesicles) may be taken up by renal parenchymal cells and exert biological actions affecting the occurrence or development of IgAN. In the future, we will make a further study of this hypothesis, hoping to find more direct evidence of the pathophysiological role of urinary erythrocyte-derived vesicles in IgAN.

From the above series of studies, we can speculate on the general process of erythrocyte-derived vesicle intervention in renal parenchymal cells (Fig. 2). In the presence of IgA nephropathy, peripheral blood erythrocytes will be subjected to mechanical injury by extruding defective glomerular filtration membranes, or subjected to damage from the change in osmotic pressure and pH in the different renal tubular segments, or high ion concentration in the renal tubules. Therefore, erythrocytes will occur in hemolysis, erythrocyte apoptosis or even necroptosis, and erythrocyte-derived microvesicles, erythrocyte-derived apoptotic vesicles, and hemoglobin overflow greatly increases. When free hemoglobin is taken up by tubular epithelial cells, they can break down hemoglobin into heme and thus, stimulate oxidative stress, making red blood cells secrete more vesicles. These vesicles can be taken up by renal parenchymal cells and directly impact the occurrence and development of IgA nephropathy. In the future, a systematic analysis of erythrocyte-derived vesicles will enhance our knowledge of the role of hematuria in kidney disease, especially our comprehensive understanding of the impact of hematuria on IgAN prognosis.

Except for casts and the toxic effects of hemoglobin, there is a very important and long-neglected aspect of the impact of hematuria on IgAN: erythrocyte-derived miRNAs. This review, through an extensive examination of the literature that included the latest research progress on blood erythrocytes, combined with our research results of urinary erythrocytes in IgA nephropathy, sheds some light on the role of urinary erythrocyte-derived miRNAs on IgAN and its possible mechanism. Moreover, erythrocyte-derived miRNAs could be non-invasive biomarkers of IgAN. Urinary erythrocyte-derived miRNAs, through the release of vesicles (microvesicles and apoptotic vesicles), play a pathophysiological role by changing miRNA levels of the nephron and renal parenchymal cells.

All authors have no conflict of interest of any type.

This study was supported by the National Science and Technology Support Program (2015BAI12B06), the 863 program (2012AA02A512), the Natural Science Foundation of China (NSFC) (81171645, 81600548), the Technological Innovation Nursery Fund of Chinese PLA General Hospital (16KMM32), Science and Technology Project of Beijing China (Z161100000516225, D171100002817002) and the 973 program (2013CB530800).