Differential Evanescence Nanometry

This is an extraordinarily confusing paper titled Differential Evanescence Nanometry: Live-Cell Fluorescence Measurements with 10-nm Axial Resolution on the Plasma Membrane in Biophys J. After looking over it a few times, I think I get the gist of it, but I'm hampered by the fact that I know almost nothing about "clathrin coated pits". But, I find even the descriptions of the theory and control experiments to be incredibly confusingly written. But, basically, it appears to be basically just another realization of VA-TIRFM.

Briefly, one of the chief problems with TIRFM is that, while it can be very useful for tracking objects and determining their positions, it only gives you a 2D projection of position, in the focal plane of the microscope objective. Getting information about the z-axis position of an object is very difficult. There are basically two standard ways to get at this information, and a large number of papers proposing different schemes for realizing them:

1) Some information about the z-axis position of objects in the field of view can be determined by the shape of the point spread function. As an object goes out of focus, the point spread function broadens, and by measuring the width of the PSF, you can learn about its z-axis position. See, for instance:

• Three-dimensional tracking of fluorescent nanoparticles with subnanometer precision by use of off-focus imaging

There's also a more clever version of this realized recently by Watanabe, et al, which uses a dual-view type splitting to image two focal planes at once, and thereby get differential information:

• Three-dimensional nanometry of vesicle transport in living cells using dual-focus imaging optics

Yet another twist on this approach is to utilize a long focal-length cylindrical lens in the imaging path to create elliptical PSFs. One axis of the PSF will focus to a different plane than the other, and the eccentricity of the PSF can be used to determine the axial position. This approach was utilized in one of the Zhuang lab's recent papers which I blogged about briefly:

• Three-Dimensional Super-Resolution Imaging by Stochastic Optical Reconstruction Microscopy
  

2) The other method is to exploit the fact that the TIRF field has exponentially decaying intensity, and to determine the axial position based on fluorescence intensity. This has a lot of potential problems due to fluctuations in intensity that are not related to axial position (for instance, due to dye fluctuations, blinking, or local environment). So, most of these approaches are differential, comparing intensities at different TIRF angles. Two good examples are:

• Observing secretory granules with a multiangle evanescent wave microscope.
• Variable-angle total internal reflection fluorescence microscopy (VA-TIRFM): realization and application of a compact illumination device.

The recent paper referenced at the beginning is basically an example of this, except that instead of varying the TIR angle, the switched between TIR and widefield, using widefield imaging as their "reference". They were also able to do imaging in two colors, in order to compare differential aggregation of two different proteins in to cytosolic pits.

While this isn't technically (or even mostly) single molecule work, novel approaches to fluorescence imaging is another one of my areas of interest, and I did a significant amount of work on 3D single molecule tracking in my doctoral work (as you may have already guessed.) My work also involved implementing a z-axis stabilizer for the microscope stage, since drift of the microscope stage is a serious issue for high resolution tracking, especially along the axis of the microscope focuser.



If you're really interested, you can get the relevant chapter of my thesis here, and a poster I presented on 3D single molecule particle tracking at BPS in 2006 here.

My mother, on blogging

"Blogs are a great mystery to me and those of my age group...and not a mystery we are very interested in solving, by the way." -Norma Gordon

W.E. Moerner wins Wolf Prize in Chemsitry

W.E. Moerner, the "father" of single molecule imaging, has been awarded the Wolf Prize in Chemistry. Hooray for single molecules!

Single molecule sequencing with nanopores

Lots of people have been trying for a long time to use nanopores to sequence DNA molecules (David Deamer at UCSC and Amit Meller at BU being two of the biggest proponents). The basic idea is to electrophoretically force a piece of DNA through a nanopore, and watch how the current changes in order to figure out which bases are moving through the pore. But, there have always been problems in getting this to work at single base resolution, mostly due to the DNA moving too fast through the pore, and other factors which cause the noise to overwhelm the tiny differences in signal when the different bases are in the pore. This has made it unusable for sequencing.

Now, a paper in JACS (along with a review by Vincent Croquette) demonstrates the first sequencing of a DNA using a nanopore. The technique relies on two improvements: first, a short ssDNA molecule is blocked on both ends, by a biotin/streptavidin blockade on one, and by double stranded region on the other, which cannot fit through the pore. This allows the molecule to be cycled back and forth several times through the pore. Second, by attaching a long PEG chain to the single stranded end, a large dielectric gradient is formed at the very end of the single stranded region. When this region is in the pore, the current becomes very sensitive to the position of the PEG/DNA junction in the pore. So, by adding polymerase and DNA bases one at a time, changes in the position of the PEG/DNA junction can be detected when one or more NTPs are incorporated. By flowing in NTPs sequentially, it can be determined what the next base or bases is by which base is present during the next elongation step. This is similar in spirit to the
Quake approach, but the latter used incorporation of fluorescent bases, and TIR imaging. (This approach became the basis of Helicos HeliScope "single molecule sequencer." I don't really know much about how this thing performs in the field, and how many of these they've sold.) The Croquette review also draws comparisons with Will Greenleaf's 2006 paper from our lab demonstrating single molecule sequencing using high resolution optical trapping of RNA polymerase.

This is an interesting technology, and a huge step forward for nanopore sequencing efforts. But, like the optical trapping approach, it has scalability problems, and isn't really ready for prime time.

I am in yer interwebz, bloggin yer science

Via Pharyngula.

Sit on this and rotate!

The Kinosita Lab has published another paper in Science on F1-ATPase. F-ATPases are used by mitochondria to transform proton gradients energy into ATP via rotation, like a miniature water wheel. But, Kinosita showed in 1997 that, supplied with ATP, the reaction was reversible, and F₁-F_o ATPase could act as a tiny rotary motor. By attaching fluorescently labeled actin filaments to the stalk, they were able to directly visualize rotation of the motor in vitro.


Further work showed that the motor takes discreet 120 degree steps, with 40 and 80 degree substeps. Their most recent paper studied truncated versions of the ATPasae, in which the "axel" that extends down into the "housing" was truncated. They showed that, even without the axel, the "wheel" will still rotate in the correct direction, demonstrating that the axel is not necessary for torque generation (though it makes it much more efficient.)

Happy Valentine's Day

More job hunters

The rumor mill says that Arne Gennerich from the Vale lab at UCSF is interviewing at UC Merced, where my buddy Tommy (previously mentioned) is also interviewing. Arne came down here visited last May, and he's a very nice guy. (He clued us in that the Anchor brewery does group tours with free tastings, and I've been meaning to get over there to get some free beer.) Ahmet from the Vale lab is also interviewng, as mentioned before. Hopefully not at the same places.

Perhaps a slight overstatement

With all due respect to my colleagues here in the Block lab, I don't think they really can claim to be the first ones ever to directly observe a 3-D molecule folding in real time. But, of course, they never claimed that. This is what happens when you let PR flacks get a hold of your science. You have been warned.

Was that a protocatech you ate?

Jody Puglisi's lab has published some data comparing the now common glucose oxidase/catalase deoxygenation system (which we used to call "gloxy" in grad school, but which my current lab now calls "scav") with the protocatechuic acid/protocatechuate-3,4-dioxygenase system. This is typically used to extend fluorescent dye lifetimes and for preventing photodamage from optical traps, by removing oxygen radicals from solution. We learned about this from some of our collaborators a few months ago, and it's good for systems which don't tolerate glucose well. The Puglisi lab reports that it has some advantages in terms of dye lifetime over the gloxy system, but I haven't looked at the details yet. Our experience, however, has been that protocatechuate is much more expensive than glucose oxidase for the same specific activity, and tends to be contaminated with RNAse. So, it has to be purified before it can be used in RNA systems. Your mileage may vary, as they say.

More Stupid AFM Tricks

On the subject of fun things to do with your AFM: A recent paper in Science on using an AFM for nano-assembly of DNA on surfaces. The authors used an AFM tip coated with DNA to pick up up complementary DNAs from one region of a surface and deliver them to another region patterned with yet more complementary DNAs. By tailoring the lengths of the complementary regions, this allowed them to essentially pick up and drop off DNA strands into arbitrary locations with nanometric precision. This presumably is another stepping stone on the way to making teeny-weeny machines of some sort, but I don't quite see how.

I was also reminded of a recent discussion we had on single-molecule NMR. It was suggested by somebody that this had never been done, but I actually presented an article on this for journal club some years ago in which this was done using a magnetic AFM tip. My paper journal-club file from grad school seems to have disappeared, but some dedicated Googling turned it up. Of course, it took them 13 hours to image a single spin in a chunk of glass, so we're not going to be using this to make biological measurements anytime soon.

Some highlights from BPS

Now that it's all over, some notes from BPS as brought back by the denizens of the Block Lab.

• This year, it was decided to hold all the single molecule talks and posters on one day, Tuesday, the penultimate day of the meeting. This was deemed a huge failure. It made it very difficult to see all the posters people were interested in, and it caused cross-scheduling of several talks with overlapping areas of interest. That was the biggest complaint that I heard.
• The Bustamante lab has apparently achieved the 'holy grail' of single molecule motility, by doing real-time measurements of ribosome motility in an optical trap. This has been a goal long sought-after by many labs, and it will be great to see the final published data showing how they did it. The ribosome is incredibly complex, requiring a whole host of proteins just to get it to do anything, so this was quite a feat.
• Lori Goldner's lab presented new data on their system for trapping water beads in a water/fluorocarbon emulsion, allowing them to do smFRET in what amounts to the world's smallest test tube. This stuff sounds really cool, and is proposed as a way to study things that are difficult to study when conjugated to a surface.
• A Japanese group, the Ando lab, apparently presented some stunning new data showing the use of high-speed AFM for visualizing myosin V motion. Their previous work demonstrated up to 30 ms time resolution for a 100x100 pixel region, and they showed motion of myosin V on a mica surface, but it wasn't motility, because there was no actin present. They apparently showed movies at the meeting of actual myosin V walking on actin, in real time, with enough resolution to resolve the stalk, the motor heads, and the actual diffusion of the motor head as it searched for the next binding site! It sounds like amazing work. Their web site has some of their older movies, but I can't wait to see this new work published and see these movies of myoV motility for myself.
• Markus Sauer's lab presented a poster on a new deoxygenation cocktail for SM fluorescescence, with they call "ROXS", for "reducing and oxidizing system". They claim that it is universally useful for all fluorophores, and shows better performance than the current glucose oxidase/catalase systems in use, but I didn't see their data, so I'll have to wait for the publication to see how it stacks up.

Also, I'm curious to know who won the Young Fluorescence Investigator award this year, but I can't find it anywhere online. In the past, the award has kind of been a University of Illinois/Enrico Gratton/Laboratory for Fluorescence Dynamics love fest, but since Enrico moved the LFD to Irvine, I think it's probably been up for grabs.

Wired does BPS

Wired magazine did a photo shoot of some of the equipment vendors at BPS. Here's their photo of an automated AFM force-spectroscopy platform, the ForceRobot from JPK Instruments.

Gene aid

This doesn't have anything to do with single molecule biophysics. But I've been doing a lot of PCR lately, and this is either so retarded it's funny, or so funny it's retarded. I can't figure out which.

Gordon's Law of Maximal Uselessness

Today in lab, I reconfirmed what I like to call Gordon's Law of Maximal Uselessness. It is, to some extent, a corollary of Sturgeon's Law ("90% of everything is crap.") Gordon's Law of Maximal Uselessness states that, in the lab (as in life), there are two ways to do something: the Right Way(tm), and the Half-Assed Way(tm), and it is almost always a waste of time to do it the Half-Assed Way. The proof of Gordon's Law is as follows: If you do something the Half-Assed Way, and it doesn't work, you have learned nothing, because you don't know if it failed because of an intrinsic problem, or because you did it half-assedly. But (and here's the kicker), 99% of the time, nothing works anyway! Therefore, 99% of the time, if you do it the Half-Assed Way, you'll learn nothing, and you'll have to go back and do it the Right Way anyway. You would have saved yourself the time if you had just done it the Right Way to begin with, and been able to learn something from your failure.

The thing about Gordon's Law of Maximal Uselessness is that people only sometimes set out to do something the Half-Assed Way. What happens more frequently is that, halfway through doing something, they screw it up. At this point, one faces a branch decision: do I keep going, and see if it works anyway (the Half-Assed Way), or do I start over from the beginning (the Right Way.) Unfortunately, mostly as a result of the fallacy of sunk cost, it is common to just finish it off and see what happens. Gordon's Law of Maximal Uselessness tells you to never do this. It's a waste of time. And so ends my catechism.

Ghost town

It's pretty dead around here, with (almost) everybody at BPS. Just me and my gels. I'm holding down the fort, but please feel free to send field reports from BPS! I'll post them here if I get anything interesting.

All publicity is good publicity...

...although this makes me wonder...

Single molecule bomb

Happy Rotebruary! Rotebruary is the month during which rotation students in our lab (aka rotons) present journal club. If there are many rotation students, it sometimes extends into Rotarch, or, god forbid, Rotapril. (At one point, the name "Rapril" was suggested, but then was hastily withdrawn, for obvious reasons.)

Today one of our rotons presented the paper on single protein unfolding using optical traps discussed previously. One subject that came up during the discussion was the amount of energy which can be obtained by hydrolysis of ATP. The number is usually quoted as about 20 kT, but it is dependent on a number of parameters. For instance, the dissociation constant for the reaction:
ATP → ADP + P_i
is
K_d = [ADP][P_i]/[ATP]

The energy change on hydrolysis is


In the absence of any free phosphate, K_d = 0_. Hence, if we hydrolyze a single ATP molecule in the complete absence of any free phosphate, that single ATP molecule will release an infinite amount of energy. This led us to speculate about the value to DARPA of a single-molecule bomb, which can destroy an entire city (an entire universe, even) with a single molecule of ATP. (It is left as an exercise to discuss in the comments why this does not work.)

Waving of PALMs

Erdal Toprak forwarded me a bunch of brand new, hot off the presses articles from the Betzig lab on PALM, hot on the heels of the Zhuang lab STORM article:

• Rainbow PALMs, in PNAS: Dual-color superresolution imaging of genetically expressed probes within individual adhesion complexes.
• Tracking PALM, in Nature Methods: High-density mapping of single-molecule trajectories with photoactivated localization microscopy.
• And also in Nature Methods, a way of increasing pulse amplitude and frequency using interferometry (obviously also useful for PALM): High-speed, low-photodamage nonlinear imaging using passive pulse splitters.

As Erdal says, it's looking like a photoactivation year!