Living cells exhibit many types of membranes which participate in most biological precesses, one way or another. Imaging membranes is usually acheived by two types of reagents: chemical dyes or fluorescent proteins that are targeted to the membrane itself or inside an organelle.
The chemical dyes are usually targeted to an organelle based on a specific chemical property of that organelle.
Rhodamine 123, tetramethylrosamine, and Mitotracker are dyes that preferentially target mitochondria, due to its membrane potential. Mitotracker has thiol groups that allow it to bind to matrix proteins, thus making it more resistant to disruption of the membrane potential (e.g. by fixation).
Lysotracker are lypophilic, mildly basic dyes, which accumulate in the acidic lysosomes.
ER-tracker is a BODIPY (boron-dipyrromethene; a group of relatively pH insensitive dyes that are almost all water insoluble) based dyes which are linked to glibenclamide – a sulfonylurease – which binds to ATP sensitive Potassium channels exclusively resident in the ER membrane.
Long chain carbocyanines like DiL, DiO and DiD are lipophylic fluorescent molecules, which are weakly fluorescent in water, but highly fluorescent when incorporetaed into membranes, particularly the plasma membrane.
FM lipophylic styryl dyes bind the plasma membranes in a reversible manner and are also incorporated into internal vesicles.
On the other hand, fluorescent proteins (FP) are targeted to membranes or organelles by fusing them to either whole proteins that localize to a specific organelle, or to short peptides that carry a localization signal. Thus, a nuclear localization signal (NLS) targets the to the nucleus, mitochondrial targeting signal (MTS) to the mitochondria and a palmitoylation signal to the plasma membrane and endocytic vesicle.
There are advantages and disadvantages to each system, relating to ease of use, specificity, photostability etc… I do not want to go into that.
Here, I would like to mention two new methods to image the plasma membrane.
Tracking surface glycans in live cells
Many (most? all?) proteins are modified by linking to glycans – short or long sugar chains. Tracking surface proteins can enable studying of membrane structure and dynamics. Ben Ovryn’s lab developed a combined metabolic and chemical methods to fluorescently label specific surface glycans (i.e. specific glycans linked to surface proteins) using click chemistry. They then used TIRF microscopy (to allow imaging only the membrane surface without backgound from the cell body) with very short laser pulses (just a few milliseconds). This allowed imaging of single molecule surface glycans without blurring. Bleaching the majority of the fluorophores also reduced the signal to allow single-molecule resolution.
Imaging of these surface glycans, and tracking their trajectories, they were able to calculate diffusion of global surface proteins, discovering areas of sub-diffusion (transient immobilization) as well as highly mobile glycans. This method also allowed studying a particulalry fine membrane structure called membrane nanotube (NT, a.k.a. tunneling nanotubes). These structurs are difficult to image (i know that from personal experience). The surface glycans reveal the outline of the NT and the dynamyc movements of proteins on the surface (as has been suggested for Sonic Hedghog protein). Here’s a video.
mCLING for imaging plasma membrane and endocytic vesicles
As I mentioned above, there are many dyes to label the plasma membrane. Yet, most probes used for live cell imaging are either lost during fixation of the cells (e.g. for immunostaining) or even transfer to another organelle. Unlike organic dyes, membrane targeted fluorescent proteins remain on the membrane also after fixation (though many may have diminished brightness). However, it requires also requires cloning and transfection of the cells with the FP. This is not always that straight forward, and is more difficult if you wish to study primary cells or even whole organs.
The Group of Silvio Rizzoli from University Medical Center Gottingen developed a new type of membrane probe that is much more stable both in live and fixed cells, so it can also be used for super-resolution. The probe is an octapeptide, composed of one cystein (couples to a fluorophore) and 7 lysines, one of which is bound to a palmitoyl tail. This tail anchors the peptide to the membrane.
Their paper shows very nice images of both live and fixed cells with their plasma membrane and endocytic vesicles labeled with mCLING, much better than FM dyes mentioned above. They could show vesicles involved in ligand-receptor recycling or trafficking, study the composition of synaptic vesicles or membrane recycling.
This seemed like a very usefull probe and I wanted to try it myself.
Unfortunately, it did not go well. I compared mCLING-Atto647N to a membrane-targeted TagRFP-T.
Here’s the TagRFP-T image:
You can see clearly the membrane, its protrusions, and some vesicles or organelles inside.
Now the mCLING:
It looks like a negative image, right? What I think happened here is that it also attached to the glass bottom (which is coated with fibronectin – essential for my purposes) and so it is very bright compared to the cell. The “black” area is where the cell “protected” the glass from mCLING. Taking movies, I did see instances where the dark “shadow” remained were a cell used to be.
The vesicles inside are clearly visible. Most do not overlap with the TagRFP-T, indicating these are differnt vesicles.
mCLING was clearly much brighter and more photostable than the TagRFP-T: it barely bleached for the entire imaging session, whereas TagRFP-T bleached fast. (Imaging: 14 z sections, 10 sec intervals, 50ms and 25ms exposure TagRFP-T & mCLING).
I did try to label the cells with mCLING prior to plating them. It was better in a sense that I didn’t get any “shadows”. But by the time I got to image the cells (it takes at least 30min for the cells to attach), all of the dye was already in endocytic vesicles.
Actually, in this case, the TagRFP-T and mCLING labeled vesicles co-localize. Maybe it’s just a matter of time that the mCLING labels all endocytic vesicles here but not when I label the cells after they attach to the glass.
I asked Prof. Rizzoli concerning these issues. Here’s the reply:
We are aware of the speed of endocytosis – it unfortunately depends on the biological system you are trying to look for. Labeing cells in solution, followed by plating, is difficult, and takes indeed too long. The problem with mCLING on the coverslip comes from having a negatively charged glass surface, to which mCLING binds. I suggest using some blocking medium before applying the mCLING, such as tryptone/peptone. Alternatively, one could use another coverslip coating, such as BSA.
Prof. Rizzoli also suggested going to confocal instead of widefield imaging. However, as much as I wanted to, all of these suggestions are unsuited for my needs. Too bad.
Jiang H, English BP, Hazan RB, Wu P, & Ovryn B (2015). Tracking surface glycans on live cancer cells with single-molecule sensitivity. Angewandte Chemie (International ed. in English), 54 (6), 1765-9 PMID: 25515330
Revelo NH, Kamin D, Truckenbrodt S, Wong AB, Reuter-Jessen K, Reisinger E, Moser T, & Rizzoli SO (2014). A new probe for super-resolution imaging of membranes elucidates trafficking pathways. The Journal of cell biology, 205 (4), 591-606 PMID: 24862576