Travels Through the Golgi

A brief introduction to one of the more intriguing organelles of eukaryotes.

One of the star organelles of eukaryotes is the Golgi apparatus- great name, elegant structure, and mysterious function. Sharp-eyed readers might have spotted one in last week's post about coronavirus replication, though that virus mostly circumvents the Golgi apparatus in its trip through the secretory system to infect more people (going directly from post-endoplasmic reticulum vesicles to the exterior). This week, using a recent theoretical paper, we will delve into the nature and rationale of the Golgi apparatus.

What we do know is that the Golgi organelle is where a lot of protein processing happens. It is the major way-station from the endoplasmic reticulum (ER)- where proteins destined for the membranes, vesicle organelles, and exterior of the cell are synthesized- to their final states and destinations. So it both directs traffic, and also directs post-translational modifications like glycosylation, sialylation, and phosphorylation, by way of short sequences, or addresses, on the individual proteins. And all this is done via a series of vesicle budding and fusion events, since it is vesicles that carry proteins from the ER to the cis face of the Golgi, and onwards from the ends of each Golgi stack to the next (medial and trans) stacks of the Golgi, and thence off to destinations like the plasma membrane and lysosome. For example, insulin is a secreted protein, translated into the ER. Its key disulfide bonds form in the ER, but its cleavage into two interconnected peptides happens in the Golgi, after which it is stabilized and stored in secretory vesicles, ready for release when glucose signals arrive at the pancreas. These processing steps often have to be done sequentially. For example, sialylation can only happen after the core glycosylation has happened. This gives some direction and rationale to the gradual and stepwise nature of the Golgi transit / processing system.

Electron micrograph (right) and schematic (left) of a Golgi apparatus, with ER-originating vesicles entering from top, and secretory vesicles leaving from the bottom, towards the plasma membrane and exterior of the cell. The story is one of vesicles, both for inputs and outputs, but also as the carriers of traffic between the various internal stages, or cisternae, of the apparatus.

We also know that the key enzymes of the Golgi, which carry out the protein processing and regulate the apparatus's own stability, are often membrane proteins, have their own addressing system, and engage in retrograde (i.e. backward) vesicular transport, which keeps them localized to the stack where they are supposed to stay, against the general flow of proteins going forward through the apparatus. There is also a special set of switch-like GTPase proteins, called Rab proteins, that control some aspects of Golgi form and function, and additional proteins, GRASPS and golgins, which all affect Golgi structure, by mutational studies. There are also cytoskeletal interactions, as the Rab proteins appear to regulate the activity of specialized myosins that are motors on actin, and dynein/microtubule activity as well. These at very least help to orient the Golgi with respect to the source (ER) and destination (exterior) locations, but may have more intimate roles in the shape of the Golgi and the paths that its voluminous vesicle traffic takes, within its compartments, and externally.

So there are many clues and ideas, but as yet we do not know fully how this structure forms and maintains itself- why do stacks form at all? Are the stacks stable, or do their pancakes progress and mature, like a slow conveyor belt from cis to trans, with new ones forming behind? A recent paper tries to synthesize past theories and evidence to come up with a unified model of self-organization for the Golgi. Unfortunately, it struggles to even represent the golgi, let alone explain it, so I will summarize briefly.

They set up a few relations- the preference vesicles generally have to fuse with similar membranes, the maturation of each membrane micro-domain and locations near it through time, as processing of the nearby proteins (both on the membrane surface and within the vesicle) takes place, and the formation of vesicles out of such matured subdomains. From these more or less empirical premises, they develop a model that they can tune on several parameters- the budding rate, the fusion rate, and the rate of purification of nearby domains, thus the purity of budding vesicles, with regard to stages of secretory processing. Vesicles of mixed purity can fuse in either direction, retrograde or anterograde (forward with the next cistern of the Golgi apparatus). But the more pure the segregation of components is at vesicle budding sites, (and indeed within the budded vesicles, while they are underway), the more forward-biased the whole process becomes. This leads, with certain parameters, to a reasonably realistic, if highly abstract, model of the Golgi.

Author's general model of the substructure of the Golgi apparatus, and some alternate models of traffic flow. Here the direction is turned around, with ER proteins arriving from below. Models that propose wholesale maturation and movement of Golgi sub-compartments are unlikely to be true, in light of genetic experiments that dissociate and tie down portions of the Golgi to mitochondria, and find that their character remains stable. PM = plasma membrane; TGN = trans-Golgi network; ERGIC = endoplasmic reticulum- Golgi intermediate compartment.

All this simply follows from their assumptions and modeling. For instance, it is entertaining to see that raising the budding vs fusion rates can cause vesicle sizes to decline, leading to a flurry of tiny vesicles. Whether this modeling helps account for the features of known actors and mutants, of which there are many that affect Golgi structure and function, is claimed, but not easy to evaluate. In the end, it would probably be much better to address the actual molecular function of these proteins, and place them in a biochemical network of interactions and regulation, to build an explicit model of what is going on, rather than one that treats the Golgi as an abstraction. Reconstitution of Golgi activity from pure components, which would be the ideal method, is likely to be extremely difficult, given its great complexity. Genetics will doubtless remain the main tool.

• Fear and the conservative mind.
• Ditto.