Live-cell imaging has a fundamental problem: imaging inevitably damages the sample. That’s because the contrast in images is generated by the interaction of the sample with subatomic particles such as electrons or photons. In light microscopy, researchers fire photons at their sample; in electron microscopy, it’s electrons. You can improve sensitivity by increasing intensity, but this means the cells endure an even heavier bombardment. Overall, it’s just a matter of time before your cells give up, roll over, and die.
“It’s almost surprising to me because quantum light has been around for three decades, a bit more, and biology has always been considered one of the most attractive applications of it,” says Warwick Bowen, an associate professor at the University of Queensland in Australia. “I think it’s basically because biology is a messy sort of science. Quantum physicists like clean experiments, and biology is generally not that.”
But now, a new breed of physicists are ready to get their hands dirty and begin developing new microscopes that reduce or eliminate the damage to biological samples while improving sensitivity down to the subnanometer scale.
Steady Stream of Photons
Ten years ago, while Bowen was a postdoc at the Australian National University, he first began studying quantum imaging. “We weren’t really at a point where we could apply this sort of technique to real applications,” he says. “It was quite unclear what the real benefits of these sorts of quantum imaging techniques would be in something like biology.”
Because small, nanosized objects do not scatter much light, you need a lot of light and a system such as optical tweezers, which uses a laser beam to trap and track nanoparticles. But biologists don’t have that luxury when working with living samples because the cells will die. So the question for Bowen became: How could you get as much information per photon? The answer was quantum light.
In traditional light microscopy, researchers fire photons at a sample, the photons are reflected by the sample, and a sensor detects those scattered photons to create an image. But the firing of photons is random, so in the end you can have multiple photons hitting the detector at the same time, which doesn’t produce a very clean signal overall because you don’t know when the next photon will arrive at the detector. So if you’re tracking a particular molecule in real time, this produces uncertainty in your measurements.
But by squeezing the light, researchers in quantum optics have shown that you can increase the precision of measurements by correlating photons in time, reducing the quantum noise. Instead of firing photons randomly, each photon would be fired at a precise time. This would produce a steady stream of evenly distributed photons rather than an unorganized, overlapping bombardment. Overall, this decreases the level of uncertainty in each measurement. As a result, you could decrease the intensity of the light used because it takes fewer photons to get the same statistical information on the location of your target molecule.
The idea was quite simple. But when Bowen’s group began building their microscope, there were several real world issues that had to be resolved. As a result, the project got vastly more complicated.
Quantum Light is Fragile
In traditional microscopy, if you lose a photon, you still have many other photons that will be reflected and detected, so there isn’t any harm overall. But in squeezed light, losing a single correlated photon will actually create a substantial amount of quantum noise that could compromise the detection of the other correlated photons, putting you in a worse position than before.
Any technical noise source would negate the advantages of working with quantum light in the first place. There were a lot of noise sources that Bowen’s team had to eliminate, including vibrations, electronic noise, and laser noise. And this problem was only exacerbated by the low frequencies of light that are used to image biological samples.
“Quantum physicists tend not to care about those noise sources because their proof-of-principle experiments operate at high frequencies, where aren’t these sorts of noise problems,” says Bowen. “But if you want to do real experiments, then you can’t get away with that any more, right?”
To overcome those noise issues, Michael Taylor, one of Bowen’s students, had an idea for a completely new optical tweezers setup, one that incorporates lock-in amplification. The lock-in amplifier allowed the group to isolate the low frequency signals from the quantum noise surrounding it by using a reference frequency and phase. “It was just a beautiful idea that made the whole thing possible,” says Bowen.
In a paper published in Nature Photonics in February 2013 (1), Bowen and colleagues at both the University of Queensland and the Australian National University reported the first microscope that uses a quantum light source to improve the information per photon obtained. In the paper, Bowen and his team used their microscope to track a nanoparticle in Saccharomyces cerevisiae yeast cells. Overall, the development of Bowen’s quantum microscope took five years and about $600,000. By comparison, most of his other projects lasted about two years and cost significantly less.
“They are basically using squeezed light to reduce the noise in measurements, it’s not quantum imaging really, it’s about using an optical light source, a kind of quantum squeezed state to make a measurement,” says Mehmet Fatih Yanik, a physicist at the Massachusetts Institute of Technology.
Using the technique, researchers can track a single nanoparticle in a cell over time with a sensitivity of 1 nanometer, which is beyond the 380 nanometer lower wavelength limit of visible light. Future experiments could track such a particle through a cell in 3-D, providing a more comprehensive look at the cellular environment. But Bowen’s group hasn’t gotten there just yet.
One biological question that Bowen plans to explore with the microscope is how polymer networks function in the cytoplasm. To do so, one could track an embedded lipid particle to see how the cytoplasm functions. “People have done this kind of measurement before, but using quantum light we could do it better,” says Bowen. “You build a better microscope, and you might see something interesting.”
Look but Don’t Touch
During his postdoctoral work at Stanford University, Yanik spent his coffee breaks thinking about the fundamental limitations of electron microscopy. “I thought about how big of an impact it would be if we could see the inside of living cells, for example the nucleus or the synapses, at the resolution of an electron microscope,” recalls Yanik.
In electron microscopy, the biological sample is bombarded with high levels of radiation, killing most cells instantaneously. It’s roughly equivalent to the amount of radiation you would be exposed to if a 10-megaton hydrogen-bomb exploded about 30 yards away. Obviously, that’s not a very hospitable place for the living.
What Yanik needed was a way of looking without touching. Previously, researchers had developed a measurement technique in quantum optics called interaction-free measurement, a way to detect an object without light interacting with it (2). To do so, the researchers used an optical cavity. When the optical cavity was empty, a photon could pass through it, so you could infer that there was nothing there based on the path the photon travelled. But if an object was in the optical cavity, the photon would bounce off the surface of that cavity, so you could infer that something was there.
“What if we could take those concepts and convert them in such a way that we could do it with electron beams, and we could then build an electron microscope that could take measurements without touching electrons ever touching the imaged sample,” says Yanik. “This could allow us to sample this with high resolution... theoretically.”
But getting an interaction-free measurement method with electrons is trickier because, unlike photons, electrons like to interact. Electrons are so strongly attracted to materials that they usually get absorbed almost immediately.
After moving to the MIT in 2006, where he now heads the High-Throughput Neurotechnology Group, Yanik published a paper in 2009 in Physical Review A that describes his concept of noninvasive electron microscopy (3). In the quantum world, electrons don’t just travel in a straight line from point A to point B; they travel like waves using multiple pathways from A to B. And every pathway is affected by what is happening in all the other pathways. So the idea is to sense the imaged object by placing it into one of these pathways and looking at what is happening in the other pathways. If the electron pathways in the microscope’s sensor are disturbed, then you know that there’s some matter there. In principle, this could be done in such a way to dramatically reduce the actual interaction of the electrons with the object.
The following year, Yanik was invited by Japanese electronics company Hitachi, Ltd. and the Japanese government to present his quantum electron microscope concept at a major meeting on the next generation of electron microscopes. “I was kind of scared about speaking there, coming from another completely different field of expertise and making such a bold claim,’” Yanik recalls. His talk sparked a debate; while some were enthusiastic, others were doubtful that such a microscope could ever be built.
But proving that it could be built would require a significant investment. “For doing quantum measurements with photons, I can buy something online, set it up in my basement, and do it,” says Yanik. “But for electrons, even in my lab with million dollar annual budget, I couldn’t do it.” So Yanik’s idea was shelved for two years while he pursued more mainstream projects.
Not Completely Crazy
One morning, the phone in Yanik’s MIT office rang. It was Gary Greenburg, a program officer from the Gordon and Betty Moore Foundation. Knoblock began asking Yanik questions about his quantum electron microscope paper and whether he believed that it was feasible. And if so, the foundation wanted to fund its development.
The Moore Foundation does not accept unsolicited proposals. If they are interested in your work, they’ll find you. Their internal management team invests in big initiatives with the potential to make a big impact. And they believed that Yanik’s concept fit that bill, but they wanted to make sure the microscope could actually be built.
The next thing Yanik knew, he was in California at a closed-door meeting with the Moore Foundation and a handful of his electron microscopy colleagues to determine if his idea was feasible. At the end of the meeting, Yanik and his colleagues convinced the Moore Foundation that it was worth a shot.
As a result, the Moore Foundation announced in February 2013 that they would invest $4 million over the next 3½ years to get some proof-of-principle measurements. The idea is that those measurements will convince both the Moore Foundation and other funding agencies to make larger investments in the technology. The investment will fund the development of multiple instruments not only at MIT but also at Stanford University, the Max Planck Institute of Quantum Optics in Germany, and the Delft University of Technology in the Netherlands. Each group is taking a different approach because no one is sure which will work.
“This kind of stuff is still theoretical, yet promising a resolution far beyond that of super-resolution optical microscopy which has only tens of nanometers of resolution. Quantum electron microscopy might allow us image live biological processes at the nanometer scale” says Yanik. In the end, researchers might be able to see the first live molecular-resolution movies of chromatin dynamics, cellular differentiation, synpases, and other unknowns without staining or disrupting the sample.
Overall, quantum electron microscopy should be fairly compatible with scanning electron microscopy (SEM), says Bowen. Interaction-free optical measurements didn’t seem practical to him because the photon only provides one bit of information, but a SEM scans the sample with an electron beam. So it should work, in theory. “I think it’s a cute idea, and I’d love for it to be successful,” says Bowen. “Although I’m not sure I’d argue that it’s quantum actually. I would argue that it’s a purely classical phenomenon that you could explain using Maxwell’s equations without resorting to quantum mechanics, but quantum-inspired...”
Whether quantum or quantum-inspired, what’s clear is that quantum physics is beginning to play an important role in the development of biological imaging technology. “I hope that our results this year have shown that these sorts of things aren’t completely crazy, you can combine biology with quantum light,” says Bowen.
1. Taylor, M. A., J. Janousek, V. Daria, J. Knittel, B. Hage, A. BachorHans, and W. P. Bowen. 2013. Biological measurement beyond the quantum limit. Nat Photon. 7(3):229-233.
2. Kwiat, P., H. Weinfurter, T. Herzog, A. Zeilinger, and M. A. Kasevich. 1995. Interaction-Free measurement.Physical Review Letters. 74(24):4763-4766.
3. Putnam, W. P., and M. F. Yanik. 2009. Noninvasive electron microscopy with interaction-free quantum measurements. Physical Review A .80(4):040902+.