A brief history of the cisternal progression-maturation model

This paper represents my view of the development of the concept of cisternal progression-maturation over the last few decades, to date. It is not meant to be particularly technical or exhaustive, and it is not addressed only to the specialist. For more detailed information, readers are referred to the key references. The goal is mainly to stimulate discussion, for which the cisternal maturation model appears to be a very apt subject. Indeed, perhaps more than other models in biology, cisternal maturation has been accompanied from the beginning by a number of controversies. The reason for this is not completely clear to me. It might be because the model addresses issues at the historic core of modern cell biology; or because it challenges long-established views; or perhaps it is because its relevance actually goes beyond the field of intra-Golgi trafficking, as it might also apply to other trafficking steps (for instance, from early to late endosomes)^1,2 and it gives rise to the concept of dynamic compartment identity,³ which is of broad relevance in cell biology. And last, but not least, the debate has suffered from uncertainties that are due to technical reasons: a few key questions cannot be addressed as directly as we would like, simply because today's microscopy imaging technologies do not yet possess the required resolution power (see below).

Cisternal Progression

The cisternal progression-maturation concept has a relatively old precursor, called the progression model, according to which the transport of cargo proteins through the Golgi complex occurs by the progression of cisternae from the cis face to the trans face of the Golgi stack.^4,5 This scheme envisions that the Golgi complex turns over constantly in a process that carries a steady flow of membranes from the endoplasmic reticulum to and through the Golgi complex, and thence to the plasma membrane (Fig. 1). The membrane flow begins with the exit from the endoplasmic reticulum of membranous carriers containing cargo proteins, and it continues with the movement of these membranes to the cis face of the Golgi complex, where they coalesce into a new cis cisterna. This process repeats itself, and at the same time, the trans-most Golgi cisternae disassemble into transport carriers that are directed towards the plasma membrane. As a result, each cisterna changes its position in the stack—repeatedly—until it reaches the trans face. Thus, the cisternae themselves act as carriers in this intra-Golgi trafficking segment. This model is simple and elegant, and it accommodated many morphological observations from the early decades of electron microscopy (EM). Among these, a key one was the presence of large objects, called ‘scales’, in the lumen of the cisternae of certain alga cells.⁶ As these scales are secreted, this finding was compelling evidence that they are transported by membrane flow inside the cisternae.

Figure 1.

Transport by cisternal progression. Cargo proteins leave the ER within dissociative carriers that reach the cis face of the Golgi stack, where these carriers coalesce to form a new cis-Golgi cisterna. At the same time, the trans-most Golgi cisterna (in this case, identifiable with the TGN) disassembles into transport carriers that are directed towards the plasma membrane. The process continues: new cis cisternae form, and old cisternae disassemble; as a result, each cisterna changes its position in the stack—repeatedly—until it reaches the trans face.

Anterograde Vesicular Transport

As we know, this simple state of affairs was destined to change. The main culprit, among several other findings, was a series of observations documenting the presence of specific compositional differences between adjacent cisternae within a given stack (reviewed in ref. 7). This showed that the Golgi cisternae are not similar to each other, as would be expected if the cisternae were simply progressing from cis to trans. For instance, it is now known that each cisterna has its own characteristic complement of glycosylating enzymes that differ from those of the other cisternae, and that those are arranged from cis to trans in the same order in which they are used in the sequence of glycosylation reactions that take place in the Golgi complex.⁸ The progression model could not explain how these enzymes remain in their cisternae while cargo proteins are swept forward (moreover, it had no role for the Golgi vesicles). This suggested a completely different organization of intra-Golgi transport. The cisternae must be stable compartments through which the cargo proteins are transported forward sequentially, by some dissociative carrier. Since the Golgi is surrounded by a number of spherical vesicles of regular size (60–70 nm in diameter), it was logical to propose that it is these vesicles that are the carriers that mediate trafficking from the proximal to the distal cisternae across the Golgi stack.⁷ This picture was not completely clear, though, because the transport of alga scales could not be mediated by these vesicles (as they are much larger than the Golgi vesicles); however, this was explained as an exception due to a ‘rare formula’ of transport present in a few evolutionarily distant organisms.⁷ Similarly, procollagen can be seen in the Golgi complex of fibroblasts in the form of aggregates that are much larger than vesicles;⁹ but again, this was set aside as an observation that would presumably be explained at later times. For instance, the large procollagen aggregates might not actually be transported; they might just be stationary bodies where a few procollagen molecules are occasionally trapped, while productive transport takes place in other parts of the cisternae (see ref. 10). Thus, despite these discrepancies, the anterograde vesicular model gained broad acceptance. And during the molecular era of trafficking, the Golgi vesicles were isolated and their components were identified (COP-I subunits, Arf, etc.), along with many other proteins that are involved in transport, and were shown to be essential for the structure and function of the Golgi complex.^11,12 These molecular findings strengthened the model of transport by anterograde vesicles that dominated the field for many years.

Cisternal Progression-Maturation

In the mid-1990s, however, the transport models were re-examined critically in a few different laboratories. On the one hand, there was a feeling that the complexity of the morphological observations in different cell types could not be explained by the simple vesicular model, and that other carriers (e.g. tubules) might be involved.^13,14 On the other hand, and more importantly, there was the increasing recognition of the conservation of the basic cellular mechanisms across species, including those for transport. This made it difficult to continue dismissing the transport of scales as a mere ‘exception.’ Instead, it was realized that the old cisternal-progression model that could explain the transport of both small and large cargoes could be modified to also accommodate the very observations that had led to its demise; namely the fact that different Golgi cisternae have different, and apparently stable, enzymatic compositions.

To this end, it was necessary to introduce the idea of cisternal maturation. According to the maturation concept, cisternae change in composition, and hence, ‘identity’ through retrograde trafficking of the Golgi enzymes (and of other components) in lock-step with their progression (Fig. 2). This idea was discussed initially through personal exchanges between interested members of the scientific community, and then in a brief space of time, several groups published review articles in visible mainstream journals proposing the cisternal maturation idea.^14–17 Other reviews with more refined models came later.^18,19 It should be noted, however, that a version of the maturation model based on studies of the secretion of plant slimes had actually been presented earlier in a review in Protoplasma,²⁰ though it had gone mostly unnoticed.

Figure 2.

Transport by cisternal progression-maturation. Cargo proteins leave the ER within dissociative carriers that reach the cis face of the Golgi stack, where these carriers coalesce and receive cis-Golgi enzymes that recycle from the cis cisterna, forming a new cis cisterna in the process. Subsequently, this cis cisterna receives medial-Golgi components from the medial cisterna and loses its cis components to a new forming cis cisterna. This changes its composition and it matures into a medial cisterna, and it also progresses to a medial position in the Golgi stack. This maturation process repeats itself until the original cis cisterna, now matured into a trans-Golgi network element and located at the trans-Golgi face, recycles its resident enzymes back to the underlying trans cisterna and breaks down into cargo-laden carriers, which move to the plasma membrane.

The first study explicitly designed to test the cisternal progression-maturation model vis-a-vis the vesicular model was published in 1998. Bonfanti et al.¹⁰ used synchronized trafficking of procollagen in fibroblasts, accompanied by a series of stringent controls, to show that procollagen actually progresses in the form of aggregates from the cis to the trans Golgi in 10–15 minutes, without ever leaving the lumen of the cisternae. These observations could not be dismissed as incomplete or ascribed to an exceptional transport formula (see above) and, coupled with the constancy of the enzymatic compositions of cisternae, could only be interpreted as being due to the progression-maturation mechanism, although the maturation (compositional change) of the cisternae was not visualized in these experiments (see below).

They therefore represented a turning point in the field, and initiated a period of rethinking of the transport mechanisms. Moreover, they carried important implications, quickly grasped by a few investigators, that impinged on the question as to whether cellular compartments may have dynamic (rather than stable) identities.²¹ They were followed initially by a series of counter-proposals that tended to support the vesicular model. In one, it was proposed that large cargoes, such as procollagen aggregates, can cross the Golgi stack within ‘megavesicles.’ This was based on experimental evidence that an artificial large polymeric cargo can be visualized in large peri-Golgi containers that are apparently separated from the Golgi cisternae in thin EM sections.²² However, a large number of 3D reconstructions of procollagen aggregates in the Golgi showed that these are never seen separated from the lumen of the cisternae; moreover, megavesicles were absent even under conditions (block of the fusion machinery) where vesicles can form but cannot fuse, i.e., under conditions where megavesicles should accumulate, if they existed.²³ Another proposal was that the progression- maturation mechanism might represent a slow trafficking mode that is specialized for a few oversized cargo proteins, such as procollagen, while most of the other cargo proteins would follow a more common and fast route that would be mediated by the COPI vesicles.²⁴ However, shortly after this proposal, it was shown that VSVG, a transmembrane protein that has been widely used in many laboratories as a common traffic marker, was transported through the Golgi complex in a fashion that was indistinguishable from that of procollagen.²³ Thus, VSVG was also most likely moving by cisternal maturation, suggesting that this transport model applies broadly. Together with this, two reports came out: one that indicated that COP-I vesicles mediate the retrograde transport of the Golgi enzymes that are required for cisternal maturation²⁵ (although this aspect remains controversial; see refs. ²⁶ and ²⁷), and the other that proposed a mechanism for the sorting of the Golgi enzymes into these vesicles.²⁸ These results provided strong support for the maturation model.

However, the most critical test came a few years later, with the direct demonstration by two independent laboratories (one of which had been among the early proponents of the model; see ref. 17) that the Golgi cisternae indeed change composition in a fashion and at a rate that is consistent with cisternal maturation being the transport mechanism.^29,30 These experiments were based on video microscopy in yeast, with the dynamic visualization of individual cis, medial and trans-Golgi cisternae, each tagged with fluorescent cisternal markers. Importantly, the use of yeast as the model organism was necessary because yeast cisternae are not arranged in stacks; rather, they move apparently freely in the cytosol, and can therefore be monitored individually in vivo by video-microscopy at the resolution that was currently available (stacked cisterna could not be resolved by the technology of the day). The use of yeast, however, also raised questions as to whether these conclusions that were drawn for yeast are actually applicable universally, to all eukaryotic cells. Given the degree of conservation of the fundamental cellular mechanisms found across species, including yeast and mammals, this is indeed likely to be the case. Nevertheless, experimental confirmation needs to be obtained. Another question left open by these experiments was due to the technical difficulty of visualizing fluorescent cargo proteins together with the cisternal markers. Because the cargo could not be visualized, an objection was that the maturing cisternae being analyzed might not be involved in cargo transport. Again, this appears very unlikely; however, experimental confirmation is desirable. In any case, and in spite of the difficulties, these direct observations of cisternal maturation provided essential and direct support for this model. In fact, together with the results on procollagen transport in fibroblasts, these data represent the experimental foundation of the cisternal progression-maturation concept.

Post-Golgi Compartment Maturation Models

At almost the same time, other researchers showed that the concept of compartment maturation is applicable also to other transport steps, and specifically to early-to-late endosomal trafficking in mammalian cells. Here, Zerial and colleagues showed that over time, early endosomes lose their characteristic marker Rab 5 and acquire instead Rab 7, the main late-endosome marker, i.e., these early endosomes mature into late endosomes (ref. ¹; however, see also ref. 31). The same study also proposed that the mechanism of Rab conversion is the central event in compartment maturation. This was a key discovery in the context of the issue of compartment maturation. Moreover, an important extension of this work was published recently, whereby the molecular mechanisms of this Rab conversion were further unravelled in Caenorhabditis elegans.² Also recently, other compartments (in this case, those involved in exit from the Golgi complex) were shown to undergo maturation (in yeast), and the underlying molecular mechanisms related to the Rab conversion were further clarified.³² Finally, it has been known for some years that phagosomes mature into phagolysosomes (albeit extremely slowly).³³ Thus, evidence for compartment maturation and its role in trafficking, and for the importance of this Rab conversion in the control of compartment identity, is accumulating.

However, here we need to go back to the specific case of intra-Golgi transport and examine the current status of the cisternal-maturation model, with a focus, of course, on its difficulties. A general problem is our lack of knowledge about several aspects of the Golgi-maturation process, and especially the molecular ones. For instance: we do not know if Rab conversion is required for intra-Golgi trafficking, and if so, which Rabs might be involved; we do not agree on the means by which the Golgi enzymes are recycled in the cis direction, i.e., via COP-I vesicles or via tubules or tubular continuities; we do not know how cisternal membranes are stacked at the cis-Golgi face, and unstacked and disassembled at the trans face; and we do not know whether and how the Golgi-matrix proteins are involved in this process. These uncertainties certainly do not strengthen the model.

Modified Cisternal Progression-Maturation

However, at this stage, it is more important for us to ask whether there are major difficulties that might lead us to discard this model altogether, or to alter it radically. There are two main potential cases of this kind, in my view. One is related to the kinetics of exit of cargo from the Golgi complex. While the progression-maturation scheme predicts a linear exit rate, the kinetics that have been experimentally observed turn out to be mono-exponential.³⁴ This is apparently incompatible with cargo progression-maturation, and has led some researchers to reject the model and to propose a radically different scheme, called the mixing-partitioning scheme. In my view, and in the view of others, the partitioning hypothesis fails to explain major and well-established observations in the field.³⁵ It has the merit, however, of posing the kinetic problem forcefully, and stimulating further reflection. Indeed, a conceptually simple explanation that might reconcile the maturation scheme with the kinetic data can be envisaged: cargo proteins might pass through the Golgi stack by progression-maturation and then accumulate in a post-Golgi compartment (which might well be the trans-Golgi network, for instance), from which the cargo proteins would then be released mono-exponentially towards the plasma membrane (Fig. 3). This is an important modification of the maturation model that needs to be experimentally tested. If confirmed, it would explain all of the observed data in a logical way. Another potentially serious problem would arise if some cargoes were found to traverse the Golgi complex at different rates than those predicted by the progression-maturation mechanisms. Here, again, we would have a case where the progression-maturation model would have to be either rejected or enriched/modified, for instance by a combination with other transport principles (importantly, the main transport principles are not mutually exclusive). It is therefore important to study the trafficking patterns of more, and indeed all, of the cargo classes.

Figure 3.

Transport by a modified cisternal progression-maturation mechanism. This mechanism is identical to cisternal maturation-progression, except that the maturation ends at the trans-Golgi level, rather than at the trans-Golgi network (TGN). Then, the trans-Golgi element recycles its resident enzymes back to the underlying cisterna and breaks down into cargo-laden carriers, which move to and accumulate in, the TGN. This modification to the cisternal maturation-progression model accommodates data indicating a mono-exponential rate of export from the TGN.

Perspectives

One of the results of the introduction of the maturation model and of all of the interesting discussion that has followed has been to free the minds of those of us in the field, and to attract more speculation. In this context, it is interesting to look at the recently proposed cisternal-progenitor model. This is motivated, I believe, simply by the difficulty of envisioning that cisternal stacking is reversible, as required by cisternal maturation. This model proposes that even large supramolecular cargo, like procollagen aggregates, can move across cisternae located at different levels in neighboring stacks through large transient continuities that are established by a Rab-conversion-related mechanism.³⁶ In my view, this model does not explain the kinetic data (see above) and the maturation of cisternae (as observed in yeast), and it may not apply to transport in individual stacks (e.g., in algae), but it is a clever exercise that enlarges our repertoire of ideas and helps us to prepare our minds for the next 2–5 years, when new technologies will give us the tools to address many of the key questions more directly.

So, the game is still on… and we are anxiously awaiting comments, concerns, and ‘attacks’ from the specialist and the curious alike, here at the CellLogBlog.

Comment on this article:

www.landesbioscience.com/journals/cellularlogistics/blog/2

Abbreviations

EM

electron microscopy

TGN

trans-Golgi network