heracles's posterous

John (www.intravitalimaging.com) and I came up with the idea of imaging EMT in cancer cells in vivo - but we had a couple of requirements:  1) To image EMT, the system needed to be inducible.  2) It needed to be in vivo (I'm being repetitive, but accepting this challenge is more difficult than one realizes).  3) Our imaging platform had to prepared to image cancer cells for days, even a week, because no one had any clue as to how long it takes a cell to revert from one morphology to another.  

To address 1), we chose to use Tom Wandless's Shield-DD induction system.  Its great because one can achieve immediate induction, without the transcription lag associated with Tet-ON/OFF systems.  Post-translational induction was the way to go.  2) We used the avian embryo model because the mets would be easily accessible and negate any surgery, which 3) allows us to image up to 3 days straight!  The entire work is summarized in PlosOne (http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0030177).

One of my favorite videos is the one with a MDA-MB-231LN micrometastatic colony (there's about 4-5 cells here clustered), which are in red.  At the outset, they are orange/red because they express tdTomato cytopllasmic protein.  Within the cluster, there is punctate zsGreen signal representing the E-cadherin-zsG-DD protein that we are about to induce with Shield-1, a small molecule designed to activate our "tunable E-cadherin" (E-cadherin-zsG-DD).  So in the absence of Shield-1, all tunable E-cadherin protein is being immediately degraded by the cell's proteasome.  However, when we IV inject Shield-1 we see the formation of junctions and then the individual cells change from a spindle shape into this round beach ball!

"Workers in cancer must make every effort to organize their work with
goals in view not just because they are 'interesting' but because they
will help in the solution of the cancer problem"

http://cancerres.aacrjournals.org/content/28/4/807.full.pdf

An amazing man measured by his scientific latitude and public education efforts.

...are not like cancer cells grown in culture on plastic.  I know it sounds obvious, and that the in vivo environment is the best place to understand cell morphology and cellular biophysics, but sometimes this is still a startling epiphany whenever I do an in vivo experiment.  

"what does it mean?" is a question that is (surprisingly) rarely asked among scientists. I do, however, get it all the time from Ann. Most often, I don't have a clue.
 
This video shows an individual HEp3-GFP cell within the intravascular space of the CAM. You'll have to imagine the stromal cells that the cancer cell is wrapped or contorting around, but I'm presenting this video to illustrate the subcellular structure of this cell and the dynamics of these structures in real time. Besides the nucleus (that kidney-shaped lobe with signal voids for nucleoli), you can actually see single granules or vacuoles moving within the cell. Now its pretty obvious that this movement is due to the hemodynamic stress on the right part of the cell. But whats also interesting is that most of those granular structures are concentrated in that "sock" or elongated structure.

Playing "what if?" - what if that sock were to become released from the rest of the cell (containing the nucleus)? What would happen to the integrity or the ability of that cell to proliferate or even migrate? What is in that "sock"?

Still, the "sock" needs to be released in order for us to even begin speculating the cell's fate. Stay tuned to see if the "sock" does get released.

This video was acquired using a 63X obj and acquisition every 75 msec (I'm only showing every other frame) at a single plane. This is a single HEp3-GFP cell within the intravascular space and since I'm using confocal microscopy (spinning disk), you can actually see subcellular structures within the cancer cell. These cells are very metastatic and therefore great for experimental metastasis expts.

Whats really neat is that you can actually watch the nucleus (fluorescent kidney shaped structure with an elongated signal void which represents nucleoli) quiver in the tiny space that it has. It's "quivering" because of the hemodynamic shear stress imposed upon it by the circulating blood cells. The shaking motion is also evident in the rest of the cell and especially the top part of the cell which has actually extended to the point where it is contact with the part of the cell containing the nucleus.

You might be wondering why there are all these circular voids around the cancer cells. Well, remember that there are stromal cells throughout this CAM plexus, so you'll have to use your imagination (I
didn't want to use Lectin-Rhodamine to label the endothelium here, but I usually do) and imagine that the spaces you see are where there are columns or "posts" of endothelial/stromal cells that support the CAM plexus lumen.

At the intracellular level, I wonder if the shaking motion affects the nucleus's ability to undergo cell division and other protein-processing/production processes: which may explain why we don't see cell proliferation during cancer cell dissemination. Just a thought.

Its observations like these that make me hate the way I think.  Before this, I always thought that microparticle (MP) formation was irreversible...that MP formation was a means for the cell to concentrate all PS and release it into the circulation to prevent apoptosis.  I always thought that a released MP was destined for the great black hole (or endoreticular system). 

It all started when I was trying to recapitulate cancer microparticle formation like the previous post (which by the way is alot easier now), and I saw this one microparticle forming...as I was rolling the camera, the MP never disconnected from the cancer cell; and as I waited longer (5 minutes now), the cell seemed to accept the fact that this microparticle was going nowhere, and slowly it re-incorporated the MP BACK into the cell!

I'm calling this process cancer microparticle retraction (as opposed to cancer microparticle formation/release in the previous post), and it really makes me re-think what cancer cells are doing once they are intravascular.  These cells are highly capable in this fast moving, highly populated capillary bed.  The still image is a 3D render of the same cell but with the CAM plexus walls surrounding this "amazing" cell, giving a perspective of where the cell is in the intravascular space.

This was quite amazing.  I've always wanted to capture this in vivo; in my previous reincarnation, I spent alot of time dreaming about platelet microparticles.  When I came to Ann and John's lab, I wanted to image cancer microparticle formation, but there was never a reason to do so. 

There still isn't a real reason why, but here is the first video of cancer cell microparticle formation and release in all its glory.  As you can see. there aren't many membrane surface blebs, as this cell just arrested and may be undergoing pre-apoptosis (maybe).  Acquisition was similar to the previous post, except I'm acquiring 5 images/second (I love EMCCD's!!)

You can see the top head slowly invaginate into a perfect sphere, then pinch off, and then teasingly remain tethered to the main cell body until it finally loses the "connection" and flies off into the circulation and probably into the arms of some macrophage (there is still quite a debate about this too).  The beauty of spinning disk confocal and CAM.  If not for the confocal, there would be no way to attain this kind of morphological detail.  You can even see the nucleus! 

Prior, this cell had already released one cancer microparticle and looked like it was going to pop another one out.  I'm glad I stuck around (only had to wait 1 minute).  The first video of cancer microparticle formation! (I may have to run this past Ian MacDonald though)

I always do my best imaging at 3AM!!

This movie nicely illustrates the membrane and endocytic dynamics of a single cancer cell trapped within the lumen (middle) of a vessel.  The animal being used is the avian embryo, and the vessel bed is the chorioallantoic membrane (the embryonic lung of the animal).  I have other pictures of the vessel bed in my earlier posts.

What's incredible is the membrane blebbing that takes place when its arrested within the plexus of the capillary bed.  I know its not "laser-toxicity" because I use a very low laser power and the cells look like this prior to acquisition.

I used a 63X oil imm obj with the spinning disk confocal, using the 491nm laser diode and acquiring two images every second.  What you're viewing is a single plane of the cell.  You can see the nucleus and its nucleolus (dark void), and some dark vacuoles being trafficked from the to of the cell into the nucleus.  Weird. 

The sheer dynamics of the cell makes me wonder if these in vivo membrane blebs resemble membrane ruffles when the cell is grown on plastic (in vitro).  Anyways, the cell is doing SOMETHING, maybe this is what a cell looks like when it resists anoikis.

I realize that the only way to really prove this in one video is to show a confocal 4D video of this phenomenon.  But I've seen it enough times to know that its actually happening and not a cell lying underneath the much larger, spread out cell.  

The cells in this video are MDA-MB-231LN cells that express the tdTomato cytoplasmic marker.  They also express another protein called zsGreen-DD whose signal will emerge later on in the video.  You'll first notice that the cells in the upper left are round and then become spread and clam-like...cancer cells typically round up in a ball just before dividing.  The bizarre thing is that one such "round body" actually gets engulfed by the very large and spread-out cancer cell.  While another "rounded" cell doesn't get engulfed and spreads out, the engulfed cells never demonstrate any kind of spreading effect...it just appears to be taken into the cytoplasm of the much larger cell!  The cool thing is that the green signal later emerges, meaning that the zsGreen-DD protein is working (because there's Shield-1 ligand in the system..more on this another day) and you'll also notice other "rounded bodies" stuck within the larger cell.  

What does this all mean?  Well, there's been a proposed theory describing the transformation of a cancer cell into a more malignant, metastatic version when it "fuses" with host immune cells such as macrophages and maybe even monocytes.  The main supporter of this hypothesis, Dr. John Pawelek believes it may be a key step in how a subpopulation of cancer cells within a tumour may morph into its deadly metastatic version when it assumes some of the cellular characteristics as bone marrow cells and white blood cells.

What do I think?  I think we need to image it real-time for me to really say.  Oh yes, and not in vitro- it would have to be done in vivo.  

I first did these videos in the fall of 2005 before I started the fibrin clot formation videomicroscopy (which is much more complicated to do); and it was a real coup for me at the time. The fact that I was able to setup a system to image clot lysis before any of the big wigs had (Weisel J, the Furies) and producing the most amazing videos essentially meant that everything was up for grabs. But it was also after this flurry of imaging that I realized that I had NO big question to ask; nonetheless the videos from that special time are here for the world to see.

The how: I used recombinant tPA to induce fibrinolysis and formed the clot with human thrombin and platelet poor plasma traced with Alexa488-fibrinogen (Molecular Probes). I was taking images every 15 seconds. The cool thing was that every time I messed up, it ended up revealing a new technique or perspective of the fibrinolysis process. I had clots that were in the shape of circular disks, muscle fibers, mesh curtains etc.. Truly a special time.

When I presented this work (it was a small part of what I was doing with PAI-1, vitronectin and tPA) at the iCAPTURE Centre "Research In Progress (R.I.P.) talks", I distinctly recall Bruce McManus exclaiming, "wow, those videos look like something out of a detergent commercial!!"

I think where the field of thrombolysis has to go is to resolve the fact that neutrophils and macrophages modify fibrin biochemistry (thanks to Jeff Weitz's PNAS paper on CD11b). I first stumbled upon this possibility when I investigated and analyzed aspirated clots from thrombolysis-resistant patients at Harefield Hospital (UK). The first thing that got me was the abundance of neutrophil nuclear nets everywhere in the thrombus, and then the fact that you just couldn't lyse these patient clots with fresh tPA and plasma. Anyways, perhaps more on this another day. Though I am truly happy to find these movies a permanent home on the 'net.