In the current study, we quantified specific populations of EVs in the urine of CSFs compared to NSFs. Results indicate that the number of EVs carrying immune/inflammatory cell markers including those of macrophages, leukocytes, and neutrophils were lower in CSFs compared to controls. In addition, the number of EVs bearing markers of proteins important in calcium and phosphorus regulation including FGF23, PiT1, PiT2, and Klotho were also lower in the CSFs compared to NSFs. In general, the number of EVs did not differ between CSFs with high versus low amounts of RP. These results indicate that specific populations of urinary EVs may reflect ongoing pathological events in the kidney of the CSFs, but perhaps those pathways are independent or differ in some way from RP formation.
Under normal conditions, nanocrystals can form and grow in tubular fluid but pass out as crystalluria [23]. Generally speaking, the literature suggests stone formers on average excrete a greater number of crystals of larger size [23, 24]. Observations made in culture cells in vitro, experimental animals in vivo, and kidney tissue from CSFs suggest that CaOx crystals can adhere to tubular epithelial cells, become transcytosed to the renal interstitium, and undergo dissolution within cells [8, 11, 12, 25–27]. The kidney harbors a variety of resident immune cells including macrophages and lymphocytes [28]. CaOx crystal deposition can activate renal immune cells to increase release of chemokines and pro-inflammatory cytokines, which in turn can recruit additional inflammatory cells including monocytes and neutrophils to the site [11, 12, 28, 29]. EVs secreted by these immune cells can serve as a biomarker of their presence and activation, and may also serve signaling functions in vivo, including antigen presentation, immune suppression, and tissue remodeling [30]. EVs secreted by innate immune cells such as macrophages appear to impact innate immune regulation primarily as pro-inflammatory and paracrine mediators [31]. In contrast, some subsets of immune cells and their signaling molecules can suppress an immune response [13, 32]. For example, neutrophils secrete EVs that have anti-inflammatory and immunosuppressive effects, mainly on dendritic cells and macrophages [31].
Although it is assumed that urinary EVs are mainly derived from kidney cells, evidence suggests that circulating exosomes can also enter the urine via trans tubular release [33]. Interestingly, in the current study the number of urinary EVs bearing immune cell markers CD68, CD45, and CD15 were reduced in CSFs compared with NSFs. Previously we had demonstrated that urinary EVs carrying the inflammatory mediator, monocyte chemoattractant protein-1 was also lower in CSFs compared to NSFs [16]. Although matching kidney tissue was not available from CSFs in the current study in order to quantitate immune/inflammatory cell populations and correlate with urinary EV populations, previously published studies do suggest that the number of immune cells within the kidney bearing CD68 may differ in CSFs compared to NSFs [27]. In the current study the number of inflammatory cell-derived EVs did not differ between high RP and low RP CSFs. This finding is consistent with previous reports that RP is not associated with inflammation [8]. Thus, in this study the populations of EVs that differed between CSFs and NSFs likely reflect events involved in stone formation but that are independent of RP formation, and instead may relate to processing of those crystals that are retained in the kidney.
Singhto et al., [13, 29]. reported that exposure of macrophages to calcium oxalate monohydrate (COM) crystals altered expression of 26 exosome proteins involved in immune signaling. They also demonstrated that exposure of macrophages with exosomes derived from COM-treated macrophages enhanced their COM binding capacity and increased crystal migration through the extracellular matrix [29]. These macrophages manifest increased fragility due to actin cytoskeleton alterations [29]. To some extent, these findings may partially explain why, in the current study, the number of urinary EVs derived from immune cells was lower in CSFs than NSFs; however, further studies are needed to elucidate this mechanism.
In bone and cartilage the transmembrane proteins PiT1 and PiT2 transport inorganic phosphate (Pi) into matrix vesicles, promoting nucleation and crystallization of Ca2+−PO4 [34]. Fibroblast growth factor 23 (FGF23) is produced primarily by osteocytes [35]. FGF23 acts on the proximal tubule to decrease phosphorus reabsorption and reduce serum levels of 1,25-dihydroxyvitamin D3 [1,25(OH)2 Vitamin D3] [35–37]. The proximal tubule is responsible for reclaiming the majority of phosphorus filtered from the blood [38]. Klotho is a co-receptor that increases the binding affinity of FGF23 to its FGF receptors [39]. In this study, urinary excretion of EVs carrying all four of these calcium and phosphorus related proteins (FGF23, Klotho, PiT1, and PiT2) were significantly lower in CSFs compared to NSFs. In addition, urinary excretion of FGF23- and PiT1- carrying EVs were significantly lower in high RP CSF compared to NSFs individuals. Although the exact mechanisms are not clear, these results suggest that alterations and phosphorus transport in the proximal tubule may influence susceptibility to RP formation.
Urinary EVs are a mixture of exosomes and microvesicles [14]. Exosomes are formed within the endosomal network including early endosomes, late endosomes (multivesicular bodies, MVBs), and recycling endosomes [14]. Clathrin has been found in early endosomes, which form from clathrin-coated buds [40]. HIP1 recruits clathrin to endosomes through its central helical domain, which binds directly to highly conserved clathrin light chains (CLCs) [41]. HIP1 binding to CLC is necessary for HIP1 targeting to clathrin-coated pits and clathrin-coated vesicles [41]. Biogenesis of microvesicles occurs via outward budding and fission of the plasma membrane [14]. Anoctamin4 (ANO4), a Ca2+-dependent phospholipid scramblase, not only takes part in exposing phosphatidylserine from the inner leaflet to the outer leaflet, [42, 43] but also alters membrane curvature and facilitates EV release [43]. Thus, HIP1 and ANO4 play essential roles in membrane budding and EV formation and secretion [43]. PiT1 and PiT2 are present in matrix vesicles as noted above. We found that the numbers of EVs carrying plasma membrane EV- biogenesis markers (ANO4) were significantly lower in CSFs compared to NSFs, while the number of EVs carrying HIP1 also trended lower. It is possible that the reduced number of these matrix vesicles relates to their ongoing rupture and calcification within the interstitium of the CSFs.
Differences in the urinary biochemical profile of the CSF versus NSF group in the current study were not surprising, including lower urine pH and greater calcium and phosphorous excretion and uric acid SS (Table 1) in CSF [4, 16, 27, 44]. Urine calcium and phosphorus excretion in high RP participants were markedly higher compared with NSFs, however those of low RP were not significantly different. Thus, metabolic factors may differ between high versus low RP CSFs, as we have previously reported [20]. We also found that EVs expressing PiT1 and PiT2 were negatively correlated with urinary phosphorus excretion, and urine excretion of EVs expressing FGF23 negatively correlated with urinary calcium and phosphate excretion. Thus, EVs containing FGF23, PiT1, and PiT2 may reflect underlying metabolic features that favor RP formation. However, further study will need to be completed to determine if quantification of EVs expressing PiT1, PiT2 and FGF23 will add additional clinically useful information to traditional urinary super saturation profiles, perhaps reflecting risk for RP and serving as a “liquid biopsy” for USD [5].
Our investigation has some limitations. The sample size is relatively small because we were limited to available biobanked urine samples from surgically mapped CSFs at Mayo Clinic, Rochester, MN. Thus findings need to be verified in larger cohorts of CSFs. However, the results suggest that urinary EVs differ between CSFs and controls may influence directly or indirectly calcium stone pathogenesis.