Renal Nerve Ablation in Nephritis and Beyond

Sympathetic nerves innervate the kidney and help regulate glomerular perfusion and filtration, and after renal denervation, altered regulation may affect the course of kidney disease. However, our understanding of denervation’s effects on kidney function in humans is incomplete. The autonomous nervous system has well established hemodynamic effects, but it also influences immune functions in the kidney and other organs. The effects on inflammatory kidney diseases have not been studied as extensively as the hemodynamic effects. In this issue of JASN, Böhner et al. examine the effects of renal denervation on antibody-induced nephritis and pyelonephritis in mice, illuminating a potential role of renal nerves in inflammatory kidney disease. Their study also points to gaps in our understanding of the effects of renal innervation and denervation.¹

Studies exploring inflammation, immune activity, and hypertensive damage to the kidney in experimental models of hypertension have demonstrated that catecholamines promote inflammation and sclerosis in the kidney and an accumulation of monocytes/macrophages.² Such studies have found that T cells and cytokines derived from T cells play a role in the forms of experimental hypertension that feature pathologically increased central sympathetic outflow. Excessive levels of catecholamines or hypertension-inducing agents such as angiotensin II or deoxycorticosterone acetate-salt favor increased numbers of T cells with an effector phenotype that infiltrate the kidney and perivascular regions of arteries and arterioles. These T cells release cytokines, including IL-17, IFN-γ, TNF-α, and IL-6, that can induce renal and vascular dysfunction, with progressive development of organ damage. Renal effects of cytokines may include increased angiotensinogen production, reabsorption of sodium, and other effects that have yet to be defined.

For years, it has been known that kidneys exert a neurogenic influence on each other in health and disease, an effect often referred to as the reno-renal reflex.³ This neurogenic crosstalk between left and right kidney (which Böhner et al. also mentions) is best understood in the context of afferent renal nerves: afferent renal nerve traffic influences sympathetic activity of the contralateral kidney. These reno-renal reflexes are impaired in arterial hypertension. Afferent nerves have an additional, puzzling feature: antidromic depolarization from the first neuron of afferent peripheral nerve fibers in the dorsal root ganglia increases the release of proinflammatory neurogenic peptides—namely substance P and calcitonin gene-related peptide (CGRP)—in peripheral target areas.⁴ In normotensive mesangioproliferative nephritis, decreased afferent renal nerve activity is likely accompanied by increased release of neuropeptides.^5,6 Hence, the control of sympathetic nerve activity via afferent nerve fibers might be significantly altered during pathologically increased intrarenal levels of proinflammatory neurogenic substances.

Renal innervation is most densely expressed in the renal pelvis. The idea that innervation at this site might be somewhat independent from the other intrarenal neuronal pathways is plausible if one recalls that the renal pelvis develops out of the Wolffian duct, whereas the kidney arises from the metanephrogenic mesenchyme. It is not known to if these differences in developmental anatomy point to different functions within the renal sympathetic system. However, the dense innervation of the renal pelvis fits well with the data on pyelonephritis presented by Böhner et al. They found that unilateral renal denervation ameliorates inflammation in the denervated kidney, and that after denervation, neither increased sympathetic activity to the renal pelvis nor increased release of proinflammatory neurogenic peptides by renal afferent nerve endings were able to exert their deleterious influences.

With intrarenal innervation of the cortex, the situation is somewhat more complex. Böhner et al. report that after renal denervation, mice were more susceptible to LPS- and antibody-induced renal injury, which they suggest was a consequence of increased glomerular filtration in denervated kidneys in these models. However, persistent hyperfiltration after renal denervation is not consistently observed in animal models or humans. In other research in rats, investigators found that complete renal denervation (afferent and sympathetic) ameliorated mesangioproliferative glomerulonephritis, reduced the influx of inflammatory cells, markedly tempered the increase in cytokines, and had a protective effect against renal sclerosis.⁷ In nondenervated nephritic animals, ED1-positive macrophages were found in close proximity to both sympathetic and peptidergic afferent nerve fibers. Dendritic cells positive for CD11c, OX62, or OX6 (MHC class II) were also found close to efferent and afferent renal nerves.

Other investigators found that specific inhibition of substance P in mesangioproliferative glomerulonephritis ameliorated the disease in rats during complete denervation.⁵ The inflammatory response was influenced by substance P, specifically via the recruitment of dendritic cells. Another study found that blocking the effects of CGRP (which is proinflammatory but acts mainly as a vasodilator) under these circumstances lowered urinary protein excretion and reduced the number of interstitial and glomerular macrophages as well as the proliferation of glomerular cells.⁶ Renal denervation will likely combine these effects in addition to the beneficial consequences that the loss of sympathetic activity might contribute to overall disease activity.² Interestingly, some researchers have reported that putatively increased intrarenal release of neuropeptides (substance P, CGRP) in nephritis is accompanied by decreased afferent nerve activity and loss of control of efferent sympathetic activity.^5,6

In a rat model of renovascular hypertension, animals exhibited a decrease in renal inflammatory response (besides a slight drop in blood pressure) 2 weeks after selective afferent denervation from a kidney with stenosis.⁸ On the other hand, data from a study in deoxycorticosterone acetate (DOCA) salt-hypertensive rats suggested that afferent renal nerve activity may mediate the hypertensive response to DOCA-salt, but inflammation depended primarily on efferent renal sympathetic nerve activity.⁹ There is evidence that renal nerves might not contribute to the maintenance of renal inflammation in DOCA-salt hypertensive rats, as researchers demonstrated that neither specific afferent denervation nor a total renal denervation—cutting all afferent and efferent sympathetic pathways—appear to have an effect on ongoing inflammatory damage.¹⁰ These data in hypertensive animal models suggest that the role of afferent and efferent renal innervation must be always carefully analyzed with respect to the inflammatory situation and immunologic mechanisms involved in the specific model used.

An important insight from the study by Böhner et al. is that persistent glomerular hyperfiltration might impair possible beneficial effects of renal denervation, at least in normotensive nephritis. It is worth noting that the extent to which renal nerve ablation has utility as a therapeutic intervention is a different question, not directly addressed by these studies. Nevertheless, this work suggests that better understanding of the effects of sympathetic nerves is important; perhaps the development of drug therapies addressing these pathways may prove more useful than the permanent destruction of these complicated multifaceted control mechanisms (Figure 1).

Figure 1.: Presented are the sympathetic renal pathway (broken line with solid arrowhead to the kidney icon) and the direction of antidromic depolarization of afferent renal nerve fibers (enlarged open arrow heads on solid line) toward the kidney. Sympathetic fibers will aggravate renal inflammation via catecholamines whereas antidromic depolarization will induce enhanced retrograde release of the proinflammatory neuropeptides substance P (SP) and CGRP intrarenally and in the pelvis. The renal pelvis (B) is more densely innervated than the kidney itself (A). These proinflammatory mechanisms could be ameliorated by renal denervation (X). Beneficial effects of denervation on renal inflammation may be reduced by renal hyperfiltration or immunologic mechanisms independent of neurogenic influences (in nephritis and hypertension). Influence of afferent renal nerve activity (solid line with arrowhead to central nervous system icon) on sympathetic outflow (C) must be discussed separately. For details see text.

Disclosures

T. Ditting reports other interests/relationships with Deutsche Hochdruckliga. R. Veelken reports other interests/relationships with Deutsche Hochdruckliga and Deutsche Forschungsgemeinschaft.

Funding

R. Veelken was supported by a grant-in-aid from the Deutsche Forschungsgemeinschaft (VE 104/4-1) and Interdisciplinary Center for Clinical Research of the University Erlangen.