Cross Talk from Tubules to Glomeruli

Abstract

Tubular injury sensitizes glomeruli to injury. We review potential mechanisms of this tubuloglomerular cross talk. In the same nephron, tubular injury can cause stenosis of the glomerulotubular junction and finally result in atubular glomeruli. Tubular injury also affects glomerular filtration function through tubuloglomerular feedback. Progenitor cells, that is, parietal epithelial cells and renin positive cells, can be involved in repair of injured glomeruli and also may be modulated by tubular injury. Loss of nephrons induces additional workload and stress on remaining nephrons. Hypoxia and activation of the renin-angiotensin-aldosterone system induced by tubular injury also modulate tubuloglomerular cross talk. Therefore, effective therapies in chronic kidney disease may need to aim to interrupt this deleterious tubuloglomerular cross talk.

Chronic kidney disease (CKD) is characterized by glomerulosclerosis, tubular atrophy, and interstitial fibrosis (Zhong, Yang, and Fogo 2017). Previously, most studies focused on glomerulosclerosis with resulting ischemia to the downstream tubules, thus causing tubulointerstitial fibrosis. However, in human biopsies, interstitial fibrosis correlates better with progression of CKD than glomerulosclerosis, even when the initial injury is to the glomeruli (Nath 1992). Further, acute kidney injury (AKI), where the initial injury is directed primarily at the tubules, is now recognized as a major risk factor for subsequent CKD (Coca, Singanamala, and Parikh 2012; Chawla et al. 2014; Palant, Amdur, and Chawla 2017; Ferenbach and Bon-ventre 2015; Basile et al. 2016). While tubulointerstitial injury reflects a sampling of more nephrons than the limited number of glomeruli present in a biopsy and thus might be expected to provide better prediction of CKD, recent evidence indicates that tubulointerstitial injury may also lead to increased glomerular injury. Thus, tubulointerstitial fibrosis may contribute to adverse cross talk from tubular to glomerular injury, thus accelerating progressive kidney scarring. We propose that this adverse cross talk enhances CKD progression in all settings, whether there is initial isolated tubulointerstitial injury or combined glomerular/tubular injury. Previous glomerulocentric studies have delineated numerous mechanisms whereby glomerular injury causes tubulointerstitial injury, including ischemia, filtered proteins/cytokine elaboration, and so on. We will now focus on potential mechanisms of tubular injury sensitizing glomeruli to injury. We recently established a model of sequential tubular-glomeruli injury and found that even mild preexisting tubulointerstitial injury with clinical recovery sensitized glomeruli to subsequent podocyte-specific injury (Lim et al. 2017). Although the initial tubulointerstitial injury was not directly progressive in nature, it resulted in an exaggerated glomerular response to a subsequent second, glomerular-specific injury and thus reduced renal functional reserve (Johnson et al. 2002; Venkatachalam et al. 2010). We will review potential mechanisms, including those within the nephron, within the kidney and systemic mediators of this tubuloglomerular cross talk.

Intranephron Mechanisms: How Does Tubular Injury Affect Its Upstream Glomerulus?

Anatomically, tubulointerstitial injury could cause stenosis of the glomerulotubular junction and finally result in atubular glomeruli (ATG), that is, glomeruli without patent connection to the proximal tubule (PT). Tubular epithelial cell dysfunction, compression and obstruction of adjacent tubules by interstitial matrix, and transition of parietal epithelial cells (PECs) to fibroblast-like cells are potential mechanisms of ATG. Proximal tubular epithelial cells, especially at S1, are sensitive to injury, such as hypoxia, followed by mitochondrial dysfunction (Lan et al. 2016). In response to injury, tubular epithelial cells can regenerate and undergo complete repair or develop G2/M cell cycle arrest with impaired repair, resulting in tubular atrophy and interstitial fibrosis (Yang et al. 2010). In many disease conditions, both human and experimental models, ATG are present and are associated with decreased glomerular filtration rate (GFR) and disease progression (Galarreta et al. 2014; Forbes, Thornhill, and Chevalier 2011; White, Marshall, and Bilous 2008). Cystinosis is an inherited disorder resulting from a mutation in the CTNS gene, a lysosomal cystine carrier. Storage of lysosomal cystine results in progressive loss of mitochondria, flattened PT epithelial cells with thickening of the underlying tubular basement membrane, and finally leads to a so-called swan-neck deformity and ATG (Mahoney and Striker 2000; Galarreta et al. 2015). In unilateral ureteral obstruction (UUO) or polycystic kidney disease, downstream tubular obstruction results in continuing increased number and size of cysts and interstitial fibrosis, which aggravates anatomical narrowing of the glomerulotubular junction and subsequent formation of ATG (Forbes, Thornhill, and Chevalier 2011; Schulte et al. 2014; Galarreta et al. 2014). Diabetic nephropathy is also characterized by increased ATG with increasing severity of the disease (Najafian et al. 2006). In ATG, the PECs at the urinary pole of Bowman’s capsule may undergo transition to a mesenchymal phenotype and extend to seal the capsule (Schulte et al. 2014). The ATG maintain some perfusion of the glomerular tuft by modulating renin and renin-expressing cell distribution, but these glomeruli do not contribute measurably to filtration. Finally, increased ATG with resultant loss of filtration capacity will increase workload on the remaining connected glomeruli and thus promote prosclerotic mechanisms in these connected glomeruli (Forbes, Thornhill, and Chevalier 2011). In our sequential injury model, ATG were increased after tubular injury, even before induction of the second glomerular-specific injury induction. Thus, these ATG could contribute to sensitization of the kidney to a second hit directed to the glomeruli (Lim et al. 2017).

Tubular injury affects glomerular filtration function through tubuloglomerular feedback. Tubuloglomerular feedback is a well-known physiologic cross talk mechanism between tubules and glomeruli, inversely regulating glomerular filtration rate according to intratubular salt concentration and flow (Araujo and Welch 2009; Singh and Thomson 2010). Proximal tubular injury with impaired absorptive capacity is a common and uniform feature in various forms of intrarenal AKI (Singh and Okusa 2011). In septic AKI, pro-inflammatory cytokines downregulate tubular chloride entry transport proteins, that is, CLCK-1, CLCK-2, and Barttin, and sodium transport proteins, that is, NHE3, Na+/K+-ATPase, ROMK, NKCC2, and NCC, which ultimately increases distal tubular delivery of sodium and chloride, activates tubuloglomerular feedback, and reduces GFR (Morrell et al. 2014; Schmidt et al. 2007a, 2007b). Tubuloglomerular feedback can also activate local paracrine mediators of glomerular disease. In the fawn-hooded hypertensive rat model of sclerosis, upregulation of nNOS and COX-2 in macula densa cells and renin expression in juxtaglomerular cells contribute to glomerulosclerosis (Wei-chert et al. 2001). Our previous data showed prominent increase in nNOS protein, which is mainly expressed in the macula densa within the kidney, after tubular injury. These data suggest tubular injury can activate the macula densa and result in abnormal tubuloglomerular feedback, with continuing arteriolar vasodilation, and consequently result in maladaptive increased glomerular pressure and severe podocyte injury (Lim et al. 2017).

Tubular injury changes progenitor cells involved in repair of injured glomeruli. Podocytes have limited or no replicative capacity and, when injured, must be replaced by progenitor cells. PECs and renin positive cells located at the juxtaglomerular apparatus are recognized as potential progenitor cells for podocytes (Lasagni and Romagnani 2010; Sagrinati et al. 2006; Shankland, Anders, and Romagnani 2013). PECs and podocytes originate from the metanephric blastema and share a common phenotype until the S-shaped body stage in development (Smeets and Moeller 2012). A large body of evidence indicates that PECs can replenish podocytes by migrating toward the vascular stalk of Bowman’s capsule or directly reach the glomerular tuft by forming new migratory tracks along adhesions from the tuft to Bowman’s capsule (Lasagni and Romagnani 2010). Particularly in male mice, some PECs located close to the PT have a columnar shape and express stem cell markers (CD133 and CD24; Berger et al. 2014; Smeets et al. 2013; Romagnani 2011). After UUO, a model of tubulointerstitial injury and fibrosis, apoptotic and necrotic columnar epithelium of the capsule, and PTs appears early at one week. At week two after UUO, the normal columnar PECs at the urinary pole become flattened (Forbes, Thornhill, and Chevalier 2011). In our own observations (unpublished data), normal male mice show increased columnar versus squamous-type PECs, and this ratio did not change after podocyte-specific injury. However, when podocyte injury was induced after preexisting tubular injury, the ratio decreased, suggesting that preexisting tubular injury perturbed PEC function and potential transformation. These findings suggest that the transition of progenitor PECs could be affected by tubular injury, which could affect glomerular repair potential.

Renin positive cells, located at the juxtaglomerular apparatus, can also migrate to Bowman’s capsule or into the glomerular tuft, replacing PECs, podocytes, and mesangial cells after podocyte injury (Pippin et al. 2013, 2015; Kaverina et al. 2017). This repair capacity is quite striking, since podocytes are normally derived from a different embryonic structure than the renin lineage cells, namely, the cap mesenchyme cells. The nature of the signal(s) that triggers juxtaglomerular cells to migrate and transdifferentiate into podocytes or mesangial cells has not yet been identified. Renin positive cells are surrounded by the afferent and efferent arterioles and the macula densa. They also express connexin 40 which assembles to mold a hemichannel among adjacent cells and allows the cytoplasm of cells to connect (Brunskill et al. 2011). These channels build a gap junction and thus permit the cells to share small molecules and respond to extracellular signals in a coordinated way. In connexin 40 knockout mice, renin positive cells were not present in a juxtaglomerular location. The recruitment phenomenon of these cells following severe sodium depletion also did not occur at its usual location, that is, in the wall of the afferent arteriole (Kurtz et al. 2007). Therefore, cell-to-cell communication may adjust renin-positive cell location and migration. Whether tubular injury affects this intercellular communication is unknown. Of note, the secretion of renin is regulated by nNOS or adenine produced by the macula densa, which could be regulated by tubuloglomerular feedback.

Intrarenal Mechanisms: How Does Tubular Injury Affect Glomeruli within the Same Kidney?

The loss of nephrons regardless of cause induces more workload and potentially more oxidative and other stress in remaining nephrons and thus can accelerate progression. Communication among nephrons in the same kidney also involves the interstitium and vasculature. Infiltrating inflammatory cells, activated myofibroblasts, peritubular capillary endothelial cells, cytokines, and growth factors can mediate this process. One of these factors is hypoxia-inducible factor (HIF). The kidney has a relatively narrow range of local oxygen levels and is susceptible to hypoxic injury (Lϼbbers and Bäumgartl 1997; Brezis and Rosen 1995). Hypoxic damage to cultured tubular epithelial cells can induce cellular apoptosis or convert cells to a myofibroblast phenotype (Manotham et al. 2004; Tanaka et al. 2003). When tubular injury is severe, interstitial fibrosis is induced, which in turn aggravates hypoxia because the fibrosis increases the distance between capillaries and tubules, leading to reduced oxygen diffusion efficiency (Mimura and Nangaku 2010). HIFs are oxygen-sensitive transcription factors, which regulate oxygen delivery and cellular adaptation to oxygen deprivation. When cells experience hypoxia, increased HIFs activate genes involved in energy metabolism, angiogenesis, erythropoiesis, and other biological processes that counteract adverse effects of hypoxia (Haase 2006, 2009; Kapitsinou et al. 2014; Majmundar, Wong, and Simon 2010). Pharmacologic inhibition of HIF degradation, such as dimethyloxalylglycine, ameliorated oxidative stress and reduced tubulointerstitial injury in experimental models (Nordquist et al. 2015). Whether it will blunt the adverse cross talk of tubule to glomeruli is the subject of ongoing studies.

The kidney contains all elements of the renin-angiotensin-aldosterone system (RAAS), and the local RAAS appears to act in a paracrine/autocrine manner to modulate renal function by regulating glomerular hemodynamics and tubule sodium transport and activating inflammatory and fibrotic pathways (Siragy and Carey 2010). Selective knockout of the type 1 angiotensin receptor (AT1aR), the dominant receptor transducing hypertensive, profibrotic angiotensin II effects, in PT cells lowered blood pressure to the same extent as that seen in systemic AT1aR knockout mice (Gurley et al. 2011; Crowley et al. 2011). These data suggest the tubular actions of angiotensin II are essential for maintenance of angiotensin II-dependent blood pressure. In a model of ischemia/reperfusion, intrarenal and urinary angiotensin II levels increased with no significant change in plasma levels (Kontogiannis and Burns 1998; Allred et al. 2000). Intrarenal RAAS is upregulated in patients with acute tubular necrosis and its level is associated with the severity of AKI (Cao et al. 2016). A study of patients with AKI found elevated blood pressure at 180 days after AKI, suggesting intrarenal RAAS may be involved in the AKI to CKD transition (Hsu et al. 2016). In the rat model of ischemia/reperfusion AKI, preadministration of losartan induced early restoration of renal blood flow, increased HIF1, and reduced glomerulosclerosis at the chronic stage (Rodriguez-Romo et al. 2016). Similarly, patients with AKI associated with cardiac surgery who received RAAS inhibitor showed lower rate of ensuing CKD and longer median CKD-free survival time (Chou et al. 2017). These studies suggest that intrarenal RAAS could be a regulator not just of classic tubuloglomerular feedback but also modulate tubuloglomerular cross talk.

Currently, few studies have investigated whether targeting tubuloglomerular cross talk could protect against glomerular injury. One study showed that pioglitazone attenuated the glomerular hyperfiltration and hyperfiltration-associated glomerular injury in diabetic nephropathy by restoration of altered macula densa signaling (Asakura et al. 2012). More interesting are the effects of sodium-glucose cotransporter 2 (SGLT2) inhibitors in diabetic nephropathy. SGLT2 is expressed at the apical border of the PT. Inhibition of glucose transport in the tubule increases NaCl delivery to the macula densa and results in more adenosine release. Treatment with SGLT2 inhibitor increased adenosine receptor CD73 and reduced another adenosine receptor Adora2b, followed by less glomerular injury, suggesting the benefit could be in part through normalizing tubuloglomerular feedback (Wang et al. 2017). However, none of these studies could exclude direct glomerular protection from these drugs.

Conclusion

Tubulointerstitial injury and even mild fibrosis initiates and perpetuates tubuloglomerular cross talk, which enhances susceptibility to a second hit on the glomerulus. We propose that the ensuing glomerulosclerosis then further promotes increased tubulointerstitial fibrosis, ultimately promoting CKD progression. Effective therapies in CKD may therefore need to aim to interrupt this deleterious tubuloglomerular cross talk and thereby protect against tubular atrophy and interstitial fibrosis and promote its regression, in addition to podocyte/glomerulosclerosis protection strategies.