Hydrogen sulfide, renin, and regulating the second messenger cAMP. Focus on “Hydrogen sulfide regulates cAMP homeostasis and renin degranulation in As4.1 and rat renin-rich kidney cell”

the nonessential amino acid cysteine is catabolized by desulfuration, thereby releasing sulfur in a reduced oxidation state and generating sulfane sulfur or hydrogen sulfide (H₂S). H₂S can be further oxidized to sulfate (10), a form of sulfur that can be excreted by the kidney. H₂S can be produced within many mammalian cell types, and consequently this gaseous metabolite has received increasing interest as a possible signaling molecule, especially because it has physical characteristics akin to those of the more widely studied receptor-independent gasotransmitters nitric oxide and carbon monoxide (10). H₂S has been reported to exist in as high as micromolar concentrations in various mammalian tissues, including the brain, kidney, liver, skin, lymphocytes, and vascular smooth muscle (10). Endogenous H₂S production is primarily the result of two enzymes: cystathione β-synthase and cystathione γ-lyase.

Different biological activities of this molecule, e.g., changes in membrane K⁺ conductance (which leads to voltage-sensitive channel opening in neurons), direct activation of nitric oxide synthase and synthesis of nitric oxide, vasodilation in the aorta and the portal vein, and induction of the cAMP-dependent protein kinase pathway in rat neurons and glial cells, have been described (10). To this end, the laboratory of Bian et al. (4, 5) has looked at possible mechanisms by which H₂S might act as a regulator of cardio-renal signaling. They report, contrary to the existing literature (10), that H₂S inhibits rather than stimulates β-adrenergic stimulation of cAMP production in cardiomyocytes and aorta and induces vasoconstriction (4). In this issue of Am J Physiol-Cell Physiol, Lu et al. (5) use renin-containing immortalized cultured tumor cells the As4.1 cell line (9) to extend the idea that H₂S might inhibit cAMP formation. Renin secretion from As4.1 cells has been shown to be cAMP dependent (3); similar to the established second messenger role of cAMP in the native juxtaglomerular (JG) cells (1). The new studies (5) provide convincing evidence that exogenous H₂S suppresses As4.1 cell adenylyl cyclase (AC) activity, stimulates phosphodiesterase (PDE) degradation of cAMP, and as predicted by those effects, attenuates the stimulated release of renin. Likewise, in a model of cystathione γ-lyase overexpression, excessive endogenous H₂S production also suppressed stimulated renin release from As4.1 cells. Bian and colleagues (4, 5) have now studied three different cell types in which H₂S seems to act as a relatively nonspecific inhibitor of cAMP accumulation, thus interrupting the downstream signaling cascade of this important second messenger molecule. Exactly what the mechanism(s) involved might be and the actual physiological relevance of such a pathway remains to be discovered.

As4.1 cells are a renin-expressing cell line originally isolated from the ascites fluid of transgenic mice harboring an intraparenchymal kidney tumor as the product of successful in vivo immortalization of renal renin-expressing cells, obtained after transgene-targeted oncogenesis to induce neoplasia in the cells (7, 9). As4.1 cells express high levels of processed renin mRNA from the endogenous Ren-1c locus, each cell containing up to 2,000 copies of renin mRNA, and constitutively secrete prorenin (2). Akin to native JG cells, As4.1 cells express the Ren-1c gene (7) and high levels of renin mRNA (2). Using As4.1 cells, Klar et al. (3) found that activation of AC by forskolin increases renin mRNA levels, heightens activity of the renin promoter, and increases prorenin secretion; effects attenuated by an inhibitor of protein kinase A (PKA). The authors concluded that cAMP stimulates renin gene expression in As4.1 cells by activating PKA and phosphorylation of the cAMP-responsive element (CRE) binding protein. However, Pan et al. (7) reported that inhibition of PKA significantly reduced Ren-1 c gene expression and that CRE was not induced by cAMP in these cells. They suggested there is constitutive activation of PKA in As4.1 cells independent of CRE but did not measure renin release. As4.1 cells have granules and contain pro-protein convertases that could activate prorenin; it has been reported that single As4.1 cells release 4 fg prorenin and 0.32 fg active renin·cell⁻¹·h⁻¹, such that inactive prorenin production is 13-fold higher than active renin (2). These cells have most commonly and appropriately been employed to study mouse renin gene molecular regulation. However, it is critical to note that these are not JG cells, and while they are useful as a vehicle for studying Ren1c gene regulation and expression, they cannot be considered a surrogate for studying the regulation of renin secretion from the JG cell because they do not share the regulatory phenotype of JG cells.

As4.1 cells produce copious amounts of inactive renin and some active renin, which may or may not respond to characteristic stimuli stimuli of renin secretion. The cells release small amounts of active renin through a cAMP-dependent mechanism but do not demonstrate critical regulatory characteristics of the JG cell and in particular the calcium paradox (1). In JG cells, unlike almost all secretory cells, the secretion of renin is inversely related to the extracellular and intracellular calcium concentrations (1). This has been shown to be the result of selective effects by increased intracellular calcium on the calcium-inhibited isoforms of AC, AC5, and AC6 and the calcium-stimulated phosphodiesterase PDE-1C (6). When we first tested these characteristics in the As4.1 cells, we found that low media calcium (nominally zero calcium plus 10⁻³ M EGTA) did not stimulate renin release, but high calcium media (5.4 mM) resulted in a 2.45-fold increase in renin release (Fig. 1). This result is opposite of what we found in studies with primary cultures of isolated murine JG cells, for which low calcium media stimulated a 3.4-fold increase in renin release and high calcium suppressed renin release (Fig. 1), similar to what is observed in studies with different in vitro models of native JG cells (1). Additionally, cAMP-mediated stimulation of renin secretion in As4.1 cells produces a complete degranulation and evacuation of stored renin, in contrast to release of only 2–6% of the stored active renin in native JG cells. In the study of Lu et al. (5), forskolin-stimulated cAMP production is accompanied by complete degranulation of their As4.1 cells. Thus, while one can stimulate renin release from this cell line through cAMP-mediated pathways, many laboratories have discounted this model because it does not appear to reflect the phenotype of the JG cell and thus would seem to be inappropriate for drawing conclusions regarding native JG cells. To their credit, the Lu et al. (5) do not claim that the As4.1 cell is a model of the JG cell, but instead they use these cells as a model to assess H₂S-mediated secretion.

Fig. 1.

Renin release from native mouse juxtaglomerular (JG) cells in primary culture (top) and from As4.1 cells in culture (bottom) incubated in normal, high or nominally zero calcium media. The JG cells demonstrate the classic calcium paradox, in which renin release is inversely proportional to the calcium concentration, while in contrast the As4.1 cells release renin in response to high media calcium.

However, the authors also provide limited additional data (5) using a primary culture of isolated mouse JG cells, in which they find that, while the introduction of H₂S has no effect on basal renin values, it does attenuate isoproterenol-stimulated renin release, which occurs via a cAMP-dependent pathway (1). While the authors do not measure cAMP, the result suggests the H₂S is acting on a cAMP-mediated stimulation of renin release in this model as well as in As4.1 cells.

So, if these cell types are so different, how do we reconcile these similar results in response to cAMP stimulation? Renin release from both cell types is characterized as cAMP dependent, but the difference may depend on the AC isoform that mediates the response. In JG cells the release of renin is completely dependent on the calcium-inhibitable isoforms (1), whereas the As4.1 cells appear to use calcium as a positive cofactor in release, presumably linked to calcium-activable AC isoform(s). Thus it suggests that the inhibitory effects of H₂S may not be AC isoform specific but probably through its ability to inhibit multiple AC isoforms. The same may be true for the H₂S effect on PDE (5). Overall, the current results of Lu et al. (5) provide an integrated general interaction with cAMP production and degradation that seems to be the prime target of this gaseous signaling molecule. There is certainly a precedent for some interaction or cross talk between AC and PDE regulation in the JG cell (6). If this is the case, and endogenous production of H₂S does reach levels that can influence the accumulation of cAMP, such results could make this gasotransmitter a target for further investigations in other cell types demonstrating cAMP-mediated events. The precise mechanism(s) by which H₂S might interact with these two enzymatic pathways controlling cAMP accumulation remain(s) to be determined.

GRANTS

This paper is funded by the National Institutes of Health Program Project Grant PPG 5PO1HL-090550-02.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: W.H.B. prepared figure; W.H.B. drafted manuscript; W.H.B. edited and revised manuscript; W.H.B. approved final version of manuscript.