Yes and no. The Golgi apparatus (or Golgi to its friends) is named after Camillo Golgi, who first reported in 1888 a reticular structure in the cytoplasm of many cell types that he found by silver chromate staining. The text book story, which most people probably do know, emerged with the advent of electron microscopy (EM) more than half a century later, when the structure was revealed to be a set of flattened membrane-bound compartments, or cisternae, that are typically arranged in a stack (Figure ​(Figure1).1). Radiolabeling studies then led to the current dogma that the Golgi is the organelle through which newly made secretory and membrane proteins pass as they move from the endoplasmic reticulum (ER) to the plasma membrane, or other membrane-bound compartments of the cell; and it is now also part of the classic picture that the Golgi elaborates and edits the generic glycan structures that are attached to proteins in the ER.

Well, first of all, the Golgi is by no means always as described by its discoverer. The structure that Camillo Golgi first observed was a typical mammalian Golgi, with individual stacks that are linked to form long connected ribbons arranged around the microtubule organizing center near the nucleus. However, in some cell types - for example muscle - and in most invertebrates and plants, the Golgi is not a linked ribbon but rather exists as one to hundreds of individual stacks scattered through the cytosol. Worse still, some species lack altogether the classic 'stack of plates' arrangement of the cisternae. This led to the belief that some species - fungi in particular - had no Golgi apparatus, and suggestions that the organelle emerged only after the first eukaryotes. However, it transpired that in these fungi the Golgi cisternae are present but spend most, and perhaps all, of their time apart. In more extreme cases such as microsporidia the Golgi is no more than a cluster of tubes and vesicles. Figure ​Figure2,2, which shows an immunofluorescence image of the Golgi in a mammalian and a yeast cell, illustrates the different arrangement of the membranes. The current view is that all eukaryotic cells have a Golgi of some sort, and thus it was a feature of the last common eukaryotic ancestor.

Yes and no - again. Despite varying greatly in shape and size, the Golgi performs some roles that are almost certainly shared by all cells and species. All eukaryotes make their membrane and secreted proteins in the ER, and require some mechanism for routing them out of the ER to various destinations inside the cell, or out of it. One key role of the Golgi is as the main sorting point for all of this post-ER traffic (Figure ​(Figure3a).3a). Proteins for different destinations all exit the ER in vesicles coated in the specialized coat protein COPII, and these fuse to each other and to the first, or 'cis', cisterna of the Golgi. ER resident proteins that have escaped in the COP II vesicles are then recycled back to the ER in COPI vesicles, and the remaining proteins exit from the opposite, or 'trans', side of the Golgi. These proteins must be sorted into carriers to be transported to different destinations. Secreted and cell surface proteins are transported to the plasma membrane, whilst lysosomal proteins are initially sorted to endosomes that mature and subsequently fuse with lysosomes. All this departing traffic depends on specialized cargo receptors and traffic machinery that are recycled and returned to the Golgi, probably at the trans cisterna.

Not exactly the same, no. All cells require that the Golgi provides the traffic route from the ER to the rest of the cell, but the Golgi also performs functions that are more variable between cell types. First, some cells make specialized secretory structures, such as the insulin-containing granules of the pancreas, and these too form at the Golgi (Figure ​(Figure3a).3a). Second, the Golgi adds post-translational modifications to the cargo proteins traveling through from the ER. The most prominent of these activities is the trimming and extension of the glycan core structures that are attached in the ER (Figure ​(Figure3b).3b). The precise glycan structures attached vary between protein, cell type and species, and they are synthesized by resident enzymes that are often arranged in the stack in the order in which they act. The diversity of glycosylation is illustrated by human blood groups, which arise from different alleles of a particular Golgi glycosyltransferase. In some cells long polymers of glycans are attached, and the Golgi has a major role in the biosynthesis of proteoglycans in animals and pectins in plants. Other modifications include sulfation, phosphorylation and proteolysis, and all of the enzymes involved are integral membrane proteins. This is strikingly distinct from the ER, where many of the resident proteins are soluble within the ER lumen, and may account for one morphological specialization of the Golgi membranes: it may be that the flattened shape of the cisternae reflects the need to keep the soluble cargo proteins in the lumen close to the 'catalytic carpet' of enzymes that lines Golgi membranes and applies the appropriate modifications to passing cargoes.

Ah well. That is a rather embarrassing question. Although we know what arrives and departs, how proteins move from one side of the Golgi to the other has been debated for decades, and still is. The original model from electron micrographs was that cisternae formed on the cis side, and then 'progressed' or 'matured' through the stack until they broke up at the trans side or fused with a pre-existing trans cisterna. This was then challenged by the proposal that transport vesicles carry cargo forward through the stack. Although cisternal maturation is back in favor with most in the field, some recent papers have suggested that tubular connections form between cisternae to allow rapid forward movement of cargo. For a lucid account of the details of this long-running debate the interested reader is referred elsewhere (see below). It also remains to be seen if the nature of intra-Golgi transport has any implications for the way that the Golgi performs its fundamental roles for the cell.

It's not. Another is the mechanism that ensures that the Golgi resident enzymes remain in the stack rather than departing with exiting cargo. There is evidence that transmembrane domains can contribute to retention, but how different enzymes are targeted to different parts is not known, or even how the transmembrane domains act.

Much of the molecular machinery of the Golgi was identified by biochemical assays and yeast genetics. More recently, genome-wide RNA interference screens in metazoan cells have identified a few further components that were missed or were absent from yeast. More unexpectedly, an increasing number of rare genetic diseases are being found to be caused by null alleles of genes encoding Golgi proteins. It seems that the loss of some Golgi proteins that are ubiquitously expressed and well conserved in evolution results in defects that, although severe, are not cell lethal but instead specific for particular tissues or cargo proteins, or result in reduced levels of glycosylation. This suggests that some Golgi machinery is not required for basic traffic but for ensuring that the organelle functions at maximum efficiency, especially when large amounts of material are being secreted.

Emr S, Glick BS, Linstedt AD, Lippincott-Schwartz J, Luini A, Malhotra V, Marsh BJ, Nakano A, Pfeffer SR, Rabouille C, Rothman JE, Warren G, Wieland FT: Journeys through the Golgi--taking stock in a new era. J Cell Biol 2009, 187:449-453.