POVRay Bug
Biophysical Imaging & Spectroscopy

(Back to research)

Eukaryotic motility and holography

This page is split into two sections:

The fanciest movies are right at the bottom, along with links to the .mp4 (H.264 encoded) versions of all movies - give those a go if you have trouble playing the movies on this page. Also note that these movies can be quite large (the largest is 36 MB) so you may have to be patient when downloading them. If you'd like to have a go at some holographic microscopy, we have software available for download (follow the link on the left), though you'll need the LabVIEW software package to use it in its current form. If you're more inclined towards code written using IDL, David Grier at NYU maintains an excellent resource for holography software (see links page) - some of our routines are based on his code.

Eukaryotic motility

In this area of our research, we focus on understanding how eukaryotic flagella work (eukaryotes are cells with a nucleus). These whip-like structures are found in plants, animals, fungi and single-celled protozoa; they are therefore thought to have evolved very early on. These organelles play a sensory role, but they also let cells move around or pump fluid, and this is the aspect we're interested in. They come in several varieties, though they have unifying structural motifs. At the heart of the flagellum is a structure called an axoneme. This is a bundle of microtubule doublets often, but not always, found with a pair of microtubule singlets in the center (with the singlets, this is known as a '9+2' axoneme). To the right is a simple cartoon of the structure. In motile axonemes, the outer doublets are connected to each other by dynein molecules that shift the tubules lengthwise relative to each other, causing the structure to beat. These flagella are closely related to the cilia that are found in the lungs and the brain (among other places). Some people claim that there's a clear difference between cilia and flagella (presence/absence of central pairs, overall length), but I've found the terms to be used almost interchangeably in the literature.

On the right is an image of a Chlamydomonas algae cell swimming (I confess that I didn't know that algae swum until recently!) You can just about make out the flagella, which are two hair-like structures that look like antennae, above the cell body (dark circle at the bottom). This microorganism swims with a breast-stroke motion - the image is a link to a movie of this. The movie is around 14 MB in size with XviD compression (see bottom of this page for mp4 versions) so you may have to save the file to disk before viewing. The movie was acquired at 1000 frames per second, and is played back at 25 frames per second, so is 40 times slower than real life. For scale, the cell body has a diameter of around 10 μm. The flagella are sometimes a little hard to see because the beating plane rotates around the swimming direction as the cell swims. The flagella are most visible at the start and end of the movie.

Malaria is caused by a eukaryote parasite from the genus Plasmodium. We work with P. berghei which causes the disease in rodents, but not in humans; although it shares a lot of features with the human version, it's a lot safer to work with. Malaria has a complicated life cycle, with some parts occurring in a mosquito, and others in a vertebrate host. The particular stage we're interested in is the sexual reproduction stage that occurs inside a mosquito. When a mosquito takes a blood meal, it triggers a response in certain types malaria cell that live inside red blood cells. The 'male' version of these cells (there are male, female and neutral types) quickly synthesize flagella in about 20 minutes; these then burst out of their confining red blood cell and go swimming off looking for their female counterparts. Below are links to two videos. On the left, flagella bursting out of the remnants of a blood cell (3 MB, played at real speed). The video on the right shows a high-speed movie of free swimming flagellum, slowed down by a factor of 20 (5 MB). The circular objects are mouse red blood cells, and the magnification is such that the movies measure 60 μm on a side.

It turns out that these swimming flagella are excellent 'model' systems for trying to figure out how flagella work. To the best of our knowledge, it's the only time that the flagellum naturally exists on its own. In the past, other model systems like algae or sperm cells (in particular sea urchin sperm) have been studied, but in these cases, there's a large cell body, or head attached to the flagellum. The hope is that these Plasmodium flagella offer a clear-cut opportunity to find out how the 'native' axoneme works.

Holographic microscopy

One of the trickier aspects of studying how flagella work is that their beat pattern is often three-dimensional. Previous studies have looked at regular video microscopy images to look at the two-dimensional projection of these shapes, but if we use holography we can capture the full thing. Holograms are different from photographs in the way they record light. Without going into too much detail, photographs record the amplitude of light waves squared, whereas holograms record the amplitude and phase. The upshot of this is that a hologram encodes all the information you need in order to completely reconstruct a light wave; a photograph does not. This accounts for the name 'holography', which compounds the Greek words for 'whole' and 'drawing'. Holography has been around for some time (Dennis Gabor won the Nobel prize in 1971 for inventing it), and there are many different variations, but we use a fairly simple implementation in our experiments.

The full details are in our 2012 Optics Express paper (free to access) but the basic idea for the process is summarized in the cartoon at the top of this section. We illuminate a sample from above using a plane (flat) wave, which we make using an LED, a pinhole and a lens. Objects in the sample scatter part of this light, which interferes with the unscattered light at the focal plane. The 'floor' in the cartoon is a real holographic image (located where the focal plane would be), which we capture using a standard camera - the trick is all in the illumination. The spherical particles in our sample cause 'ripples' in the light field, and these ripples encode information about how far the particle lies from the focal plane. If we know the size of the ripples, and the position of their center, we can locate particles in 3D. That's great for transparent spheres, but what about the malaria flagellum?

To reconstruct the position of more complicated objects, we use Huygens' principle, from classical optics. This is a simple, but powerful idea saying that every point on a wave acts like a source of waves with the same frequency and phase. We can use this idea to refocus a holographic image computationally. We say that each pixel is a point source, emitting with the same frequency as our original LED, and ask what the total field is at a plane some distance from the original. We can use this idea to generate a series of images (a 'stack'), that's the equivalent of refocusing a single image at different distances. The picture to the right is a link to a movie (160 kB), where we've refocused the starting image at a series of different depths. The center of the flagellum is dark in the first frame, turning light as it passes through focus. We use this dark-light shift to determine when parts of the object are 'perfectly' in focus, and hence where they are in relation to the focal plane.

Each image we capture contains information about the whole sample volume; this means that our volume-scanning rate is limited by the camera frame rate. The camera we use for holography (a Mikrotron MC1362) is quite fast, allowing up to 500 frames per second at a resolution of 1280x1024 pixels. We can get faster frame rates if we take smaller movies, but we find that 500 frames per second is as fast as we need to go to capture the motion of the malaria flagella. Below are a couple of movies of reconstructed holographic data, rendered using the POV-Ray raytracing package, available here. The top movie (31 MB) is a standard reconstruction, and the one below that is a 3D anaglyph movie (36 MB), which needs to be viewed with red/blue glasses to be seen in its full glory. The silver object on the left is the raw data, and the bronze object on the right is the best-fit flagellar shape. The tiles on the ground are 2 μm on a side, the fitted contour is 200 nm in diameter (the actual size of the flagellum), and the mouse red blood cells in the background are rendered to scale (6-8 μm across). The raw data was acquired at 500 frames per second, and is played back at 50 frames per second, i.e. 1/10th the true speed.