Angiotensin II-independent upregulation of cyclooxygenase-2 by activation of the (Pro)renin receptor in rat renal inner medullary cells

Abstract

During renin-angiotensin system activation, cyclooxygenase-2 (COX-2)-derived prostaglandins attenuate the pressor and antinatriuretic effects of angiotensin II (AngII) in the renal medulla. The (pro)renin receptor (PRR) is abundantly expressed in the collecting ducts (CD) and its expression is augmented by AngII. PRR overexpression upregulates COX-2 via mitogen-activated kinases/extracellular regulated kinases 1/2 in renal tissues; however, it is not clear whether this effect occurs independently or in concert with AngII type 1 receptor (AT1R) activation. We hypothesized that PRR activation stimulates COX-2 expression independently of AT(1)R in primary cultures of rat renal inner medullary cells. The use of different cell-specific immunomarkers (aquaporin-2 for principal cells, anion exchanger type 1 for intercalated type-A cells, and tenascin C for interstitial cells) and costaining for AT(1)R, COX-2, and PRR revealed that PRR and COX-2 were colocalized in intercalated and interstitial cells whereas principal cells did not express PRR or COX-2. In normal rat kidney sections, PRR and COX-2 were colocalized in intercalated and interstitial cells. In rat renal inner medullary cultured cells, treatment with AngII (100 nmol/L) increased COX-2 expression via AT(1)R. In addition, AngII and rat recombinant prorenin (100 nmol/L) treatments increased extracellular regulated kinases 1/2 phosphorylation, independently. Importantly, rat recombinant prorenin upregulated COX-2 expression in the presence of AT(1)R blockade. Inhibition of mitogen-activated kinases/extracellular regulated kinases 1/2 suppressed COX-2 upregulation mediated by either AngII or rat recombinant prorenin. Furthermore, PRR knockdown using PRR-short hairpin RNA blunted the rat recombinant prorenin-mediated upregulation of COX-2. These results indicate that COX-2 expression is upregulated by activation of either PRR or AT(1)R via mitogen-activated kinases/extracellular regulated kinases 1/2 in rat renal inner medullary cells.

Figure 1

Characterization of long-term rat inner medullary cells (A-I). IM cells show specific immunoexpression of AQP-2 (red; A), AT1R (red; B), COX-2 (green; C), and PRR (green; D). AT1R (red) co-localizes with COX-2 (green) in the same type of cells (E), indicating that interstitial cells (arrows) strongly stained for COX-2 (green) also co-express AT1R (red) in the plasma membrane, as previously described (E). Cells expressing AQP-2 (red; F), the principal cells, do not co-express COX-2 (arrow, green; F). Tenascin C (arrow, green; G), a marker for interstitial cells, does not co-localize with AQP-2 (red; E). Likewise, anion exchanger type-1 (AE-1; red; H), a known immunomarker for intercalated type-A cells (arrowheads), does not co-localizes with AQP-2 (green; H). Evidence for the presence of PRR (red; I) co-localizing with COX-2 (arrow heads, green; I) is shown in I. Nuclei were counterstained with DAPI (4′,6-diamidino-2-phenylindole; blue). In addition, in normal rat kidney sections (3 μm) immunofluorescence studies demonstrate that PRR (apical green immunoreactivity) and AE-1 (basolateral red immunoreactivity) co-localize in tubular intercalated type-A cells (J). K shows specific immunoreactivity for COX-2 in the interstitial cells (arrows) and some tubular cells (arrowheads) in the rat inner renal medulla (brown chromogen, 3,3′-Diaminobenzidine DAB). In consecutive kidney sections (L and M), is evident that PRR (green; L) is immunoexpressed by negative AQP-2-expressing cells (red; M). Panel N displays examples of the co-localization of PRR (red; cell membrane) with COX-2 (green; intracellular localization) in the intercalated cells of the collecting ducts. In addition, this panel contains a high resolution microphotograph (left upper corner, 100X, oil immersion) for clear details. The lower panels O–Q show microphotographs (20X magnification) of a rat kidney section stained using dual immunofluorescence for COX-2 (red; O) and PRR (green; P) counterstained with DAPI (blue fluorochrome) demonstrate that COX-2 and PRR also co-localized in interstitial cells (merge, Q). Images were visualized using a Nikon Eclipse 50i immunofluorescence microscope and microphotographs were captured using a digital camera Nikon DS-U2/L2.

Figure 2

COX-2 mRNA (A) and protein levels (B) levels were augmented after 6 h of AngII treatment; candesartan (Cand; 1μM) blocks this effect. n=6–8; *P<0.05

Figure 3

COX-2 protein expression is upregulated by AngII and rat recombinant prorenin (rrPR) via ERK1/2 pathway. A. Representative Western blot showing phosphorylated ERK1/2 (p-ERK1/2) levels in response to AngII + rrPR and AngII + rrPR treatment at 0, 15, 30 and 60 min. *P<0.05 versus control (0 min, n=6). B. Representative Western blot showing that COX-2 protein levels are augmented by rrPR treatment in the presence of AT1R antagonist to avoid the possibility of AngII formation in IM cells. As shown before AngII also increase COX-2 protein levels. The ERK1/2 inhibitor UO126 (10 μM) blunted the effect of both treatments (*P=NS versus vehicle; n=6). No effect was observed on COX-1 protein levels.

Figure 4

PRR knockdown suppressed the PRR mediated upregulation of COX-2 in IM cells. COX-2 mRNA (C) and protein (D) levels were augmented by AngII (100 nmol/L) and this effect was suppressed by AT1R blockade with candesartan (Cand; 1 μM). Rat recombinant prorenin (rrPR; 100 nmol/L) plus Cand also upregulates COX-2, demonstrating an independent effect. PRR-mediated upregulation of COX-2 was completely blunted in IM cells previously transfected with PRR-shRNA. *P<0.05 versus vehicle; n=6.