New Microscope Watches Cells in 3D

Jamie found a story about a new 3D Microscope which creates 3D videos of cells in action. Traditionally scientists have had to choose between high resolution and animation, so no doubt this device will cure the common cold.

Comments

Move over YouTube...

[Delusion_]

...here comes FluTube!

Bad summary...

[ed.mps]

...and bad jokes. A little excerpt for those who didn't RTFA:

It can, for example, capture chromosomes spooling during cell division or a cervical cancer cell shriveling up when treated with acetic acid.

Death by light

[G4from128k]

Although the article does not say so, I'd bet that creatures don't live for very long. All of high-resolution imaging systems that I'm familiar with concentrate so much light on the subject matter that the creature dies within minutes.

Just think of the physics. Most digital sensors need about 10,000 to 100,000 photon to register a full response (i.e, "white") and to see 30 frames per second, that's 300k to 3 million photons per second per pixel. At high resolution a single cell might be 100 pixels by 100 pixels. That means that the poor creature is being hit by 3 to 30 billion photons per second. Even if there's no UV and all heat is removed from the subject, visible light photons in a high enough flux rate will induce various photochemical reactions that damage DNA, denature proteins, and photo-oxidize cellular chemicals. Or to put in another way. consider the amount of light needed to image the average landscape and then concentrate it on a single cell. Even with high-gain amplifiers (= grainy, low-light pictures), the shear concentration of light means the creature doesn't last long.

Re:Death by light

[kebes]

Photo-damage to cells is indeed a concern, but the described technique actually has the advantage that this can minimized as much as physically possible. Many visualization techniques involve either (1) having the cell absorb light, so that you can differentiate different regions based on absorption (may require staining with something sufficiently absorptive), or (2) having something fluoresce, which requires that species to absorb and then re-emit light (typically requires staining or genetic engineering so a target protein is fluorescent). Obviously both (1) and (2) require the sample to actively absorb photos, which means that some amount of photo-heating is unavoidable. Moreover fluorescent molecules often lead to undesired side-reactions and degrade over time (so-called "photo-bleaching"). With fluorescence imaging, you can select an excitation wavelength outside of the absorption bands of everything in solution (especially water!), and thereby minimize photo-heating and photo-damage.

The article says that they are actually imaging the refracted light. Since this technique doesn't require any amount of sample absorption at all, they can use a minimally absorbing wavelength, thereby keeping sample damage to an absolute minimum. In fact since they are measuring refracted light, the technique works best at wavelengths where absorption is as low as possible (but refractive index contrast is as high as possible).

From the description, it doesn't sound like the illumination would be much more intense than what a normal microscope generates. Most cells don't experience significant photo-damage under such illumination conditions.

Some current imaging systems use a raster-scanned focused-laser spot to generate the images. By using high-quality detectors the light-levels can be kept low enough that cell damage is prevented. Thus the technique from the article probably induces less cell damage than currently used techniques. Not to mention that the fact that you don't have to stain or modify the cells eliminates the toxicity (or perturbing effect) or those staining agents.

Press Release Science

[LightPhoenix7]

While interesting, the article had several fallacies in it.

For one, cells can be viewed while alive - fixative isn't always necessary. Motility studies, for exmaple, don't actually kill the cells (or sperm). For another, dyes aren't the only technique to view cells - plasmid insertion into bacteria with a fluorescent marker not only allows cells to be seen, but doesn't harm the cell.

Secondly, I find it decidedly inconvenient that this can only view small images. My current research is in bacterial biofilms - living and dead. I haven't had any trouble viewing living biofilms under a fluorescent or confocal microscope. What if you want to study the chemotaxis of groups of cells? Most cells, eukaryote or prokaryote, talk to each other and can respond differentially to external signals.

Thirdly, even if you can view these cells, only in very specific instances will it give clues about functionality. Sure, that's better than nothing, but it's not the miraculous panacea that the article describes. The mechanics of drug interaction are much more complex than can be determined by simply looking at a cell.

Finally, from a research standpoint, I have to ask how much this costs. Is the cost-benefit ratio really that good that spending large amounts of money to get this is worth it? Especially considering how in reality it has such a limited usage? I would tend to assume no. There may be some very useful things you can do with this, but it just seems like much more of a toy than anything.