In a recent Opinion article in Trends in Endocrinology and Metabolism [1], Day and Gorr review recent work in the chromogranin field, including our own work on chromogranin A (CgA) and our proposal that CgA is an ‘on/off’switch, acting at a post-transcriptional level, for the biogenesis of neuroendocrine large dense-core secretory granules [2]. They raise the question of whether CgA plays a role as a universal master gene,‘on/off’switch, or assembly factor in the control of dense-core secretory granule biogenesis in endocrine and neuroendocrine cells. They argue that CgA is unlikely to act as such a switch only at the transcriptional level. We agree: our data show that suppression of CgA synthesis in PC12 cells causes an abolition of large dense-core secretory granule proteins, but not the mRNAs that encode them, clearly pointing to a post-transcriptional component of the action of CgA in granule biogenesis. Day and Gorr also point out that secretory granules exist in cells that do not synthesize CgA. These exceptions are particularly interesting. As pointed out by us previously [2], mast cells use a nonprotein substance, heparin, to enable mast cell granule formation, and lactotrophs contain prolactin rather than CgA in their secretory granule. Our definition of an ‘on/off’switch for granulogenesis is a factor that regulates and is required for initiation of granule formation at the level of the trans-Golgi network (TGN). According to this definition, molecules other than CgA, such as heparin [3], could function as ‘on/off’switches for granule biogenesis, depending on cell type. Based on the fact that most endocrine and neuroendocrine cells [4–6], as well as some exocrine cells [7], synthesize CgA, and our data implicating CgA in large dense-core granule biogenesis in PC12 and 6T3 cells [2], CgA is likely to be the most widely used granulogenic factor in neuroendocrine cells.

What of experiments in which overexpression of CgA does not reconstitute large dense-core secretory granule formation [8], or in which the spontaneous loss of regulated secretion in a PC12 variant cell cannot be rescued by exogenous CgA expression [9]? Tellingly, forced expression of CgA completely reconstitutes granule formation and function in PC12 cells in which endogenous CgA suppression first leads to loss of granulogenesis and regulated secretion [2], and forced expression of CgA in non-neuroendocrine cells likewise results in the formation of significant numbers of CgA containing large dense-core granule-like vesicles, albeit less abundant than the granule complement of normal endocrine cells [2]. Thus, in a cell with all other components predisposing to granulogenesis (PC12 cells), CgA is the post-translational ‘switch’that turns it on, whilst in nonendocrine cells (eg fibroblasts), CgA might well act only as an assembly factor, because structures or organelles formed in the CgA-expressing fibroblasts (CV-1 cells) are most likely not bona fide secretory granules, in the sense that they contain prohormones and have prohormone processing capability, and so on. PC12 variant cells that have also permanently lost neurosecretion competence [9–12] cannot be expected to recover simply by overexpressing CgA [9], unless it can induce the expression of all the necessary genes, thus acting as a ‘master’gene. However, CgA has never been proposed as such a gene. By contrast, PC12 and 6T3 cells have not permanently lost their synthesis of regulated secretory pathway proteins upon downregulation of CgA synthesis and thus exhibit reversible competence for neurosecretion upon restoration of CgA synthesis. Thus, our data from PC12 and 6T3 cells support the hypothesis that CgA not …