Macrophage Function in Calcium Oxalate Kidney Stone Formation: A Systematic Review of Literature

Background The global prevalence and recurrence rate of kidney stones is very high. Recent studies of Randall plaques and urinary components in vivo, and in vitro including gene manipulation, have attempted to reveal the pathogenesis of kidney stones. However, the evidence remains insufficient to facilitate the development of novel curative therapies. The involvement of renal and peripheral macrophages in inflammatory processes offers promise that might lead to the development of therapeutic targets. The present systematic literature review aimed to determine current consensus about the functions of macrophages in renal crystal development and suppression, and to synthesize evidence to provide a basis for future immunotherapy. Methods We systematically reviewed the literature during February 2021 according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Articles investigating the relationship between macrophages and urolithiasis, particularly calcium oxalate (CaOx) stones, were extracted from PubMed, MEDLINE, Embase, and Scopus. Study subjects, languages, and publication dates were unrestricted. Two authors searched and screened the publications. Results Although several studies have applied mixed modalities, we selected 10, 12, and seven (total, n = 29) of 380 articles that respectively described cultured cells, animal models, and human samples. The investigative trend has shifted to macrophage phenotypes and signaling pathways, including micro (m)-RNAs since the discovery of macrophage involvement in kidney stones in 1999. Earlier studies of mice-associated macrophages with the acceleration and suppression of renal crystal formation. Later studies found that pro-inflammatory M1- and anti-inflammatory M2-macrophages are involved. Studies of human-derived and other macrophages in vitro and ex vivo showed that M2-macrophages (stimulated by CSF-1, IL-4, and IL-13) can phagocytose CaOx crystals, which suppresses stone development. The signaling mechanisms that promote M2-like macrophage polarization toward CaOx nephrocalcinosis, include the ﻿NLRP3, PPARγ-miR-23-Irf1/Pknox1, miR-93-TLR4/IRF1, and miR-185-5p/CSF1 pathways. Proteomic findings have indicated that patients who form kidney stones mainly express M1-like macrophage-related proteins, which might be due to CaOx stimulation of the macrophage exosomal pathway. Conclusions This systematic review provides an update regarding the current status of macrophage involvement in CaOx nephrolithiasis. Targeting M2-like macrophage function might offer a therapeutic strategy with which to prevent stones via crystal phagocytosis.


INTRODUCTION
The prevalence of kidney stones has increased worldwide, and its high recurrence rate is also a factor that affects medical and economic resources (1)(2)(3)(4). Considerable research effort has been directed toward finding a cure, but the pathology of kidney stone formation is complex and awaits elucidation despite recent technological innovations. Kidney stones are recognized as a multifactorial disease similar to metabolic syndrome (MetS) (5,6). Renal function, mineral and lipid metabolism, inflammatory processes, oxidative stress, and insulin resistance can cause calcium oxalate (CaOx) crystals to develop (7).
Most kidney stones consist of calcium oxalate (CaOx) (8,9). The hypotheses presented to account for the pathogenesis of CaOx stones are free-and fixed-particle mechanisms (10); the latter is also known as Randall plaques (RP) that comprise apatite formed by calcium phosphate that grow in the interstitial space around the loop of Henle (11). In contrast to stones that develop within the tubular lumen or renal collecting system, crystal precursors of RP are surrounded by other molecular and cellular structures, which might be influenced by impaired homeostasis (12,13). Among such lithogenic environments, chemical and mineral component overload or other sources of inflammatory stimulation might act as first triggers that are followed by reactive oxygen species (ROS), which subsequently induce renal epithelial cell damage resulting in CaOx crystal deposition (14). The primary defense mechanism against such cellular impairment is autophagy involving endocytosis (15), and a secondary defense mechanism extends to peri-tubular cells in the interstitial space and immune cells from the circulation (16). The innate defense system that clears crystal deposits from renal tissues is key to finding a fundamental solution for developing novel treatments for kidney stones.
Understanding the role of macrophages (Mj) in renal crystal formation can help to identify a solution. Renal or peripheral Mj involvement in kidney stone disease was first reported by de Water et al. in 1999 (17). They discovered that Mjs migrate to crystal deposition sites and engulf the crystals. Given M1 proand M2 anti-inflammatory (18) polarization, the involvement of Mj in renal crystal formation is probably diversified in different ways. Because the ability to phagocytose crystals is greater for M2-than M1-like Mjs (19), regulating their polarization might have therapeutic value (20). Much about the clinical application of Mj to preventing kidney stone development has been reported over the last two decades (21).
This systematic review aimed to provide collective evidence of Mj function in renal crystal development and suppression, and to provide ideas for the future direction of Mj research into the clinical application of Mj immunotherapy.

MATERIALS AND METHODS
This systematic literature review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (22).

Search Strategy
We searched PubMed, MEDLINE, Embase, and Scopus databases using the MeSH key terms, "macrophages" OR "macrophage", OR "monocyte", AND "urolithiasis", OR "calcium oxalate", OR "kidney stone". Article type, publication date, language, or species were not restricted during the initial search.

Eligibility Criteria
Articles that focused on the relationship between Mj and monocyte function and the pathogenesis of urolithiasis were included. Articles describing findings of inflammatory cytokines/ chemokines secreted by Mjs and monocytes were excluded because direct connections with Mjs and/or monocytes were limited. Since we focused on experimental and translational findings, original articles were favored over reviews or commentary articles, which referred to and/or summarized findings published by others.

Data Collection and Description
Two authors (KT and RU) independently reviewed the titles and abstracts of the articles identified in the initial search during February 8 th , 2021. Data were extracted from articles that met the eligibility criteria and reassessed in full-text articles during primary and secondary screening, respectively. The most recent of duplicate published articles was included. Disagreements and discrepancies between the two screens were resolved through discussion and consensus with the other authors. The following data were extracted from all eligible full-text articles: first author, journal name, year of publication, type of study, methodology, type of experimental sources, and main findings concerning Mj/monocyte function.
The Mj phenotypes were reported in accordance with the latest nomenclature, experimental guidelines (18), and reviews regarding organ-specific findings (20,23,24).

RESULTS
We identified 380 articles that met the search strategy and criteria across all databases in the initial literature search. After the first screen of titles and abstracts, the second screen of full texts filtered them into 29 articles that were eligible for review. Figure 1 shows the PRISMA flowchart. Table 1 summarizes the findings of 29 articles describing various studies that were published in China, Germany, Japan, the Netherlands, Thailand, and the USA between 1999 and 2020. These articles described mostly translational research studies including rat and mouse models of nephrocalcinosis caused by hyperoxaluria in vivo, murine, canine, and human renal tubular epithelial cell lines in vitro, and human renal papillar, peripheral blood, and urine samples ex vivo.

In Vitro and Ex Vivo Human Monocyte Function Against Calcium Oxalate Crystals in Kidney Stone Formers
Thongboonkerd et al. were the first to report the proteomic interactions between human monocytes and CaOx monohydrate (COM) crystals. The researchers found that COM could drive oxidative stress, resulting in increased cellular apoptosis and expression of prohibitin, plasminogen activator inhibitor-2, Alix, lamin A/C, and moesin, as well as reduced cellular survival, protein synthesis and stability, mRNA stability, and lipid metabolism. In addition, levels of La protein, heterogeneous nuclear ribonucleoprotein H1, elongation factor-2, otubain-1, heat shock protein (HSP) 105, and acyl-CoA thioester hydrolase were reduced (25). The proteomics associated with the monocyte functions evoked by COM in renal tubular epithelial cells (RTECs) have also been investigated. The researchers identified six secreted proteins with significantly altered expression levels in Mardin-Darby canine kidney (MDCK) cells incubated with COM crystals. Furthermore, they reported that enolase-1, a COM crystalbinding protein, activated monocytic cell migration by binding to the surface of monocytic cells (26).
Williams et al. further investigated monocyte mitochondrial function in human serum specimens and reported significantly reduced monocyte mitochondrial maximal respiration, reserve      Figure 2). Internalized COM crystals become surrounded by phagosomes that fuse with lysosomes to promote dissolution and spontaneous elimination (29). Mj functions have also been investigated in RTEC. M-1 (murine collecting duct) and RAW264.7 cells incubated with murine Mjs developed into a pro-inflammatory state with increased COM crystal adhesion, especially when these cells were co-cultured with adipocytes. This paracrine mechanism is associated with increased mRNA and protein levels of OPN, MCP-1, and tumor necrosis factor (TNF)-a (30). A previous study on Mjs derived from human monocytes incubated with COM crystals in vitro identified an association between HSP90 and F-actin on the phagosome membranes engulfing COM crystals. Blocking HSP90 with an siRNA reduced phagocytic activity and Mj migration, suggesting that HSP90 and F-actin are involved in Mj function during the CaOx interaction (31). Exosomes secreted by Mjs are also affected by COM crystals. In particular, COM-treated Mj exosomes have increased membrane fragility and are able to enter the renal interstitium and trigger RTEC to release IL-8, which worsens tissue inflammation (32). Thus, COM-treated Mj exosomes enhance monocyte and T-cell migration, monocyte activation, and Mj phagocytosis. However, the suppression of vimentin with an siRNA abolishes these effects, suggesting that Mj exosomal vimentin plays an important role in the immune response to COM crystals (33).

In Vivo Macrophage Involvement in Calcium Oxalate Nephrocalcinosis
While there are limitations in adapting the findings to human patients owing to differences between human kidney stone formation and nephrocalcinosis in rodents, CaOx nephrocalcinosis models in rats and mice by hyperoxaluria are widely utilized for research, on the basis of similarities in intratubular crystal formation and retention (34). Mjs and multinucleated giant cells encapsulating interstitial CaOx crystals in both kidneys in a hyperoxaluric rat model and in patients with oxalosis were described in 1999 (17). The authors further confirmed that ectodermal dysplasia 1 (ED1)-positive Mjs were predominantly increased around renal crystal deposition sites compared with CD45 and major histocompatibility complex (MHC) class II positive and other mononuclear cells during the time course of hyperoxaluria in rats (35). These early studies showed that Mj dissolve CaOx crystals, which might be related to renal defense against stone development. Okada et al. found spontaneously eliminated renal CaOx crystal deposition in a mouse model of nephrocalcinosis induced by glyoxylate (36), and that subsequent transcriptome studies associated the activation of monocytes/macrophages with this phenomenon involved  M(CSF-1) derived from human peripheral blood mononuclear cells demonstrate greater ability to phagocytose both CaOx crystals and natural kidney stones than that does M (GM-CSF). Inhibitor assays have demonstrated that kidney stone clearance is mediated through clathrin-dependent phagocytosis and endocytosis (42). Dominguez-Gutierrez et al. reported that CaOx, but not potassium or ZnOx, induced the M1-like Mj differentiation of human monocyte cell lines and primary human monocytes expressing CD86 and CD68, and secrete cytokines and chemokines. Furthermore, supernatants of CaOx-treated monocytes can enhance M2Mj CaOx crystal phagocytosis (43).

In Vivo Polarized Macrophage Function in Renal CaOx Crystal Development
Several studies have assessed Mj polarization in renal stone formation. In accordance with the mouse model of nephrocalcinosis with hyperoxaluria, more renal and urinary crystal deposition develops in CSF-1-deficient mice with fewer M2-like Mjs (identified as CD11b + F4/80 + CD163 + CD206 hi cells), than wild-type mice (39). In contrast, a surge in M1-like Mjs (identified as CD11b + F4/80 + CD11c int Ly6C hi cells) has been identified in MetS model mice with a leptin deficiency and CaOx crystal deposition in the kidney under treatment with ethylene glycol (EG), and fed with a high-fat diet (44). Further studies of the roles of M1 and M2Mj roles in stone development have revealed that the induction/transfusion of M1-and M2-like Mjs respectively accelerated and attenuated renal crystal development in C57BL/6J wild-type mice administered daily with intra-abdominal glyoxylate (19). Anders et al. examined the role of the nucleotidebinding oligomerization domain, leucine-rich repeat, and pyrin domain-containing 3 (NLRP3), a central molecular mediator of inflammation in crystallopathies, in CaOx nephrocalcinosis formation in Nlrp3-deficient mice on a high-oxalate diet. They found that NLRP3 inhibition induces a shift in infiltrating renal Mjs from the M1-like (CD45 + F4/80 + CD11b + CX3CR1 + CD206 -) to the M2-like (CD45 + F4/80 + CD11b + CX3CR1 + CD206 + TGFb -) phenotype and attenuation of renal fibrosis. Therefore NLRP3 appears to promote nephrocalcinosis-related fibrotic kidney disease by promoting a shift from anti-inflammatory M1, to proinflammatory and profibrotic M2-like Mjs (45).

Roles and Polarization of Macrophage in Human Tissue
Microarray analysis of human kidneys revealed decreased expression of the M2-like Mj-related genes, peroxisome proliferator-activated receptor gamma (PPARg), CD163, and CD206, and increased expression of M1-like Mj-related genes, including nitric oxide synthase 2, CSF2, IL10, and C-C chemokine receptor type 2, in renal papillary tissues from patients with CaOx stones compared with those who do not form stones (19). The findings of causal network analyses associated the differentially expressed genes in RP papillary tissues with significantly higher immune cell activity, including Mjs and plasma cells, which are linked to IL11, PG-endoperoxide synthase 1, glutathione peroxidase 3, and monocyte-to-Mj differentiation in RP, than in normal papillary tissues (46). Urinary Mj-related cytokines in stoneforming adolescents have also been investigated. Levels of urinary IL-13/creatinine and Mj inflammatory protein-1b (MIP-1b)/ creatinine could serve as useful biomarkers for stones based on sensitivity with 50% and 58%, respectively, and 93% specificity for both in a study of a small sample (47). Urinary multiplex comparisons among individuals who did not form stones, and those who formed CaOx stones for the first time and those who formed recurrent CaOx stones, identified IL-1a, IL-1b, IL-4, IL-10, and GM-CSF as potential biomarkers affecting Mj and neutrophil function in stone development (48). Furthermore, M1-like Mj polarization increases pro-inflammatory cytokines such as TNFa, IL-1b, and IL-1, as well as M1/M2-like monocyte ratios in blood samples from patients with CaOx stones compared with those who do not form stones (49).

Therapeutic Target Altering Macrophage Phenotype Against Kidney Stone Disease
Sirtuin 3 (SIRT3), an NAD + -dependent deacetylase in the mitochondrial matrix, suppresses renal CaOx crystal deposition in vitro and in vivo by promoting M2-like Mjs through deacetylating forkhead box O1 (FOXO1) (49). The prevalence of kidney stones is influenced by sex hormones (47). Suppressing androgen receptor (AR) expression in RTEC in vitro increases Mj recruitment and causes an M2 polarization shift, which increases the phagocytosis of intrarenal CaOx crystals (47). Administering renal tubule-specific AR knockout mice and hydroxy-L-proline treated rats with an AR degradation enhancer in vivo has revealed that AR signaling suppresses CSF-1 expression via the upregulation of miR-185-5p, which results in decreased M2-like Mjs and accelerated CaOx crystal development (50). The potential of pioglitazone, a PPARg agonist, to treat kidney stones has been recognized (51,52). Pioglitazone increases M2-like Mj polarization and decreases renal CaOx crystal deposition and inflammatory damage and in murine bone marrow-derived Mj in vitro and in CaOx nephrocalcinosis mouse models in vivo. These results indicated that PPARg upregulates miR-23 expression and subsequently attenuates the expression of interferon regulatory factor 1 (Irf1) and Pknox1, which shifts the Mj phenotype from M1 to M2 (53). The effects of nuclear factor erythroid 2-related factor 2 (Nrf2) on PPARg and the anti-inflammatory process in individuals with CaOx stone have been investigated. The findings showed that Nrf2 attenuates the M1-like Mj polarization shift by suppressing tolllike receptor 4 (TLR4) and IRF in vitro. Moreover, sulforaphane, an activator of Nrf2, plays a protective role against CaOx crystal formation and renal injury via the Nrf2-miR-93-TLR4/IRF1 axis, which promotes M2-like Mj polarization and inhibits RTEC inflammation in vivo in mouse models of CaOx nephrocalcinosis (54). The involvement of the aryl hydrocarbon receptor (AhR) as a regulator of the phenotypic balance between M1-and M2-like Mjs in renal CaOx stone development has been investigated (55). Transcriptomic and proteomic analyses of bone marrow-derived Mjs and mouse models of CaOx nephrocalcinosis have revealed that stimulating the AhR-miR-142a-IRF1/hypoxia-inducible factor (HIF)-1a axis diminishes M1-like Mjs and promotes M2-like Mjs, leading to the suppression of renal CaOx crystal deposition and stone-related renal damage (55).

LIMITATIONS
This systematic review has several limitations. Firstly, most articles we found utilized the M1/M2Mj definition to be consistent with the prior literature in the nephrology and urology fields; therefore, in some cases it was difficult to summarize them according to the latest nomenclature. Secondly, findings from nephrocalcinosis mouse models might not be applicable to kidney stone patients owing to pathological differences between intratubular crystal deposition and CaOx stone formation. Lastly, only a few studies have assessed the direct relationship between CaOx stones/crystals and Mjs in human tissues; thus, additional live tissue studies are required.
Research focus has shifted from identifying Mj expression to analyzing Mj function and finding triggers that alter the Mj phenotypes to create novel therapeutic targets. However, evidence regarding Mjs in urinary sediments is scant, and the direct manipulation of Mj phenotypes for clinical use needs to be determined (56). Future investigations should strive to establish a urinary biomarker for liquid biopsies (57), and an Mj-specific target therapy using antibodies, vectors, and nanoparticles (58,59).