Requirements for the identification of dense-core granules Jacopo Meldolesi, Evelina Chieregatti and Maria Luisa Malosio Vita-Salute San Raffaele University and Scientific Institute San Raffaele, Department of Neuroscience and Immunology and Diabetes Unit, Centre of Excellence in Physiopathology of Cell Differentiation, via Olgettina 58, 20132 Milan, Italy Dense-core granules (DCGs), cytoplasmic organelles competent for regulated exocytosis, show considerable heterogeneity depending upon the specificity of their expressing cells – primarily neurons and neuro- secretory cells. DCGs have been mainly identified by detecting their cargo molecules, often members of the granin family, and using conventional electron microscopy and immunocytochemistry. However, by a critical analysis of the various stages of DCG ‘life’ within neurosecretory cells, we have highlighted several specific molecular and functional properties that are common to all these organelles. We propose that these properties be considered as strict requirements for the identification of DCGs. Dense-core granules (DCGs), also referred to as large dense-core vesicles, are typical yet heterogeneous orga- nelles competent for regulated exocytosis and expressed by neurons and neurosecretory cells. In contrast to synaptic vesicles (SVs) and synaptic-like microvesicles (SLMVs), which contain low molecular-weight neurotransmitters, DCGs contain proteins of the granin family, such as chromogranins (Cgs) and secretogranins (Sgs), as well as a large spectrum of active molecules including peptide hormones and transmitters, ATP and biogenic amines. Most of the information about DCGs has been obtained by studying a few types of neurosecretory cells, primarily bovine chromaffin cells and their rat tumor counterpart, PC12 cells. In most other cells, including neurons, recognition has been often based on only a few criteria; these include detection by conventional electron microscopy of dense, often round, cores of variable sizes (60 – 300 nm in diameter) surrounded by a single mem- brane; a distribution in the proximity, or in direct contact with, the plasma membrane; and the detection of granins and/or peptides by immunocytochemistry. The question that we intend to raise is whether these properties are sufficient or additional specific and more comprehensive criteria are required for the identification of DCGs. In our opinion, this question is relevant not only to cell and neurobiology but also to pathology because DCGs are expressed by a large number of tumors. Moreover, granin- rich organelles appear in non-neurosecretory cells, such as fibroblasts, after the transfection of cDNAs of the DCG cargo. Therefore, without adequate criteria, there is a risk that the organelles that have little (if anything) in common with bona fide DCGs are identified as such, with ensuing confusion in an important area of cell science. In the following four sections, the four stages of the intracellular ‘life’ of DCGs – biogenesis and maturation, storage, exocytosis and surface recycling – are critically analyzed. To highlight specific aspects, additional gran- ules that share some properties with DCGs, such as the insulin granules of pancreatic b cells, are considered. To distinguish between the secretory processes, the last section is devoted to constitutive secretion, a process distinct from secretion of DCGs, which is regulated by Ca 2þ . Our aim is to define a minimum set of criteria, based on common molecular and functional properties, that are sufficient to distinguish DCGs from other organelles; these criteria should be useful as a general guideline for cell biological and physiological studies of regulated secretion in cells that have a complex structure and organization such as neurons and neurosecretory cells. Biogenesis and maturation of DCGs The trans-Golgi network (TGN) is the sorting station of trafficking proteins addressed to not only DCGs but also other compartments, such as endosomes, lysosomes and vesicles of constitutive secretion. At least two mechanisms play major roles in sorting the cargo proteins in the DCG lumen. The first mechanism depends upon ionic conditions and occurs in discrete TGN domains (i.e. sorting by retention); the second process involves direct binding of cargo proteins to subpopulations of secretory and mem- brane proteins that work as ‘sorting receptors’ (i.e. sorting for entry). Here, we briefly discuss only a few aspects that are relevant to this Opinion article; detailed discussion of these mechanisms is found elsewhere [1–3]. Not only the expression but also the aggregation of lumenal proteins, which in part accounts for the consider- ably high concentration of these proteins at the TGN – granule boundary (Figures 1,2), varies in a cell-dependent manner; thus, DCG cargos can be heterogeneous with respect to their granin content. In some cells, such as AtT-20 and pancreatic b cells, granins and hormones are sorted to the same granules, whereas in others, such as pituitary somatomammotrophs and GH4 cells, they are sorted to different granules [3–5]. CgA, which is the major protein in the DCGs of neurons and chromaffin cells, is a minor component of many PC12 clones [6], is absent from GH4 cells [5] and from many neuroendocrine tumors. In Corresponding author: Jacopo Meldolesi (meldolesi.jacopo@hsr.it). Opinion TRENDS in Cell Biology Vol.14 No.1 January 2004 13 http://ticb.trends.com 0962-8924/$ - see front matter q 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.tcb.2003.11.006