Transcription and Translation

With the sequencing of the human genome, the public at large has been told that biologists now have a full picture of how life works. This is far from the truth. In this series of posts I’ll try to outline what we don’t know – in other words gaps in our knowledge.

Today we’ll look at how proteins acquire sugar modifications, aka glycosylation.

Look at any eukaryotic cell and you’ll notice that one of the main differences between it and its prokaryotic cousins is its elaborate systems of membranes and its complicated secretion machinery. Unlike prokaryotes, which inject secreted proteins directly into the extracellular (or periplasmic) space, eukaryotes pump newly made secreted and membrane bound proteins into the endoplasmic reticulum (ER). These proteins are then extensively modified by that same elaborate systems of membranes that characterize eukaryotes. The modification that occurs in these organelles is called glycosylation. In other words, long sugars chains are added to certain asparagine, serine or threonine residues found in the newly synthesized polypeptide.

i-d93d0583e45e5f6379ea60dca9934d47-golgi.jpgSo getting back to our story … in the ER, the proteins are folded and the root of each sugar chain is added to certain asparagine amino acids. These proteins are then stuffed into vesicles that bud off of the ER and are transported to the first layer of the Golgi complex. This organelle is composed of a series of huge flat membrane compartments, each called a cisternae, that form a large stack that resemble a dish of 4-6 layered pancakes (see image to the right). The main function of the Golgi complex is to extensively modify all the secreted protein’s sugar chains so that they resemble large branched trees. As the protein moves from layer to layer, the branches are pruned and are grown. It used to be thought that the proteins might be transported from one cisternae to the next, however recent results from a few labs indicate that the stacks themselves are maturing.

What do I mean by this?

Take a the first layer, called the cis-Golgi cisternae. In this model, vesicles from the ER would fuse together and with vesicles that contain enzymes that are normally found in this Golgi layer to form a new cis-Golgi cisternae. Over time the enzymes are replaced – the cis-Golgi enzymes are packed into vesicles that bud off of this compartment and are replaced by medial-Golgi enzymes that are found in vesicles that fuse with the same layer. Our cis-Golgi cisternae slowly transforms into a medial-Golgi and a new cis-Golgi is assembled beneath it. Because each Golgi cisternae contains different sugar pruning and sugar growing enzymes the glycosylation patterns on each processed protein slowly changes as the cisternae matures. In some late Golgi compartments we even get some new glycosylation on serine and threonine residues. By the end of the process, different proteins acquire different sugar trees. Some proteins gain gigantic sugar trees while other proteins are barely touched. Eventually the proteins that were synthesized in the ER end up in the last (or most mature) cisternae, the trans-Golgi, which is really not a continuous layer, but instead a loose connection larger vessicles found after the last true Golgi cisternae. In the trans-Golgi, proteins are packed into vesicles and shipped to different parts of the cell such as the endosome, the plasma membrane etc.

So you might be thinking that the cell goes through a lot of trouble to transport its newly made secreted, membrane and organelllar proteins through this maze of membranes in order to properly glycosylate these substrates. What is it all for? The short answer is that the addition of all these sugars drastically alters the properties of any given protein. The next thought that you might have is what is the sugar, or glycosylation, code? To be honest we really don’t have a full picture. In fact we’re pretty far from that point.

Flip through any cell biology text book and the rudiments of the glycosylation process is known, but if you were given an amino acid sequence, besides pointing out where the sugar trees are added, you would not be able to predict the type of tree that the protein will acquire when it leaves the Golgi complex. In fact it is worse than that. We have yet to develop a technique that could accurately analyze the nature of these trees. Yes, we know very little. And many many many gradstudent and postdoc lives have been sacrificed in the name of glycosylation.

So what to do?

First off – we need a new method to analyze glycosylation patterns. One potential answer could be a modified form of tandem mass spec (glycoproteomics?). The problem here is that sugar trees are branched making it very hard to distinguish how the sugars are connected on a given protein at a particular residue. Perhaps the answer lies with a combination of mass spec and other spectroscopic methods. More likely some novel technique using some trick needs to be developed. (For a discussion of how new technologies come about, click here.)

This data would be the start of many new lines of investigation. For example:

Do glycosylation patterns change between cell types? between different conditions? between cells experiencing different stimuli?

Many secreted proteins act as signals between cells. Do differential glycosylation patterns aftect how cells communicate?

Many secreted proteins form the extracellular matrix (ECM) in which cells live. The ECM, in turn, affects the internal chemistry of the cells that come into contact with it. How is the activity of the ECM affected by differential glycosylation?

Once they exit the Golgi, secreted, membrane and organellar proteins must be sorted into distinct vesicles so that they can be sorted to distinct locations. For example some vesicles are destined to be fused with one portion of the plasma membrane while others must be transported to another part of this same surface. It is likely that glycosylation patterns may act as determinants for the proper sorting of these proteins.

Proteins synthesized in the ER also end up populating most of the other organelles. How do differential glycosylation patterns affect organellar function?

So many questions to be answered. All we need is that first breakthrough.


  1. #1 mycobacterium
    June 4, 2008

    I resent your implication I cannot glycosylate my proteins! Balderdash! ER, SmER I say!

  2. #2 Alex Palazzo
    June 4, 2008

    Not at all. Glycosylation is fundamental to all life. But prokaryotes do it without the baroque system of membrane organelles that euks have.

    I will give you this: both prokaryotic and eukaryotic research would benefit from that foirst critical breakthrough in glycoproteomics.

  3. #3 Dave Bridges
    June 4, 2008

    What about glucosylation? There was a paper a little bit ago about insulin stimulating glucosylation (single glucose residue) on some important proteins. This appears to be a post-translational modification of already processed proteins, rather than a trafficking signal. One thing about these modifications too, the normal way of looking at proteins (boiling in SDS before western blotting) usually clips off sugar modifications meaning a lot of these modifications might have been easily missed.

  4. #4 PhysioProf
    June 4, 2008

    Nice post! One minor quibble: You kind of imply that glycosylation is the only important thing that goes on the in the post-ER secretory pathway. This is, of course, not correct. There are all kinds of post-translational processing events that occur post-ER, such as lipid modifications, endoproteolytic processing of preproproteins, etc.

    One thing about these modifications too, the normal way of looking at proteins (boiling in SDS before western blotting) usually clips off sugar modifications meaning a lot of these modifications might have been easily missed.

    This is not correct. Ordinary SDS-PAGE methods do not usually remove sugar chains from proteins.

  5. #5 Alex Palazzo
    June 4, 2008


    I am going to write a separate post on all sorts of protein modifications such as acetylation, glycylation, glutamylation, ADP-ribosylation, acylation, lipidation, etc. There are MANY protein modifications that have yet to be explored, but glycosylation is a special case.

    And also PP is right, glycosylation withstands SDS denaturation, although many membrane bound proteins do get proteolyzed if you boil them too vigorously.

  6. #6 Sunil
    June 4, 2008

    I’m not going to contribute to this discussion in any meaningful way, but as a biochemist who dabbles in cell biology, I love reading these more extensive cell biology posts of yours, as they serve as great refreshers on the bigger picture questions in cell biology.

  7. #7 Arun Chandrashekar
    June 28, 2008

    I wonder whether an RNai system tagged to an inducible promoter in conjunction with a tagged target protein would help. this is assuming that proteins required for glycosylation are continually being made. Another presumption that could be made that it is the partailly folded protein that is recognized. Do we see glycosylation in the inside of a protein molecule or only on its surface ? So there must perhaps be a realtion between structure of a protein and glysoylation ?