Getting Small August 25, 2000
 
Most universities attract great people. Of the many great people I have met at CalTech, I hold the highest regard for Richard Feynman. He demonstrated to me that a scientist did not have to fit into a specific mold. That you could be interested in a multitude of things. That the best scientists always looked at the world in a 'kooky' way. Because only 'kooky' ideas really helped move the world forward. (I was quite pleased to see Apple include him in several of their 'Think Different' ads. Just a picture on the back of a magazine or on a wall. Without any other indication of who he is.).

Richard Feynman was an amazing man. I am of the firm belief that he is one of the giants of the last half-century. Much like Einstein, his scientific work stands on its own but it is his influence on others that makes him great. He had a tremendous ability to see things before any one else did and to detail just what to do experimentally. His quote on this week's Discovery Researcher (I think the problem is not to find the best or most efficient method to proceed to a discovery, but to find any method at all.) is typical. His Lectures on Physics are incredible things to read, even if you do not understand the science because they make you think along new and different lines.

During the committee hearings on the Challenger explosion he cut through the mind-numbing, and rear-end covering, dialog that was going on by performing an experiment that was devastating in its simplicity. Following a demonstration of how the O-rings worked in the shuttle, Feynman clamped a piece of the O-ring and put it in his ice water for a few seconds. After removing it, he could easily see that it was slow in returning to its original shape. A 30 second experiment overshadowed all the talk and begged the question "Why had no one done a similar check?" They had no Feynman.(But then it does appear he had some help, also.)

In December 1959, Feynman gave a talk at the American Physical Society that, while many years ahead of its time, was a very influential speech. The title was 'There's Plenty of Room at the Bottom.' In it he brought up for the first time many ideas that we take for granted today, such as using photolithography to create ultrasmall gadgets that can act as motors or to store data. He described the incredible storage capability of the head of a pin; the creation of small levers and tools to perform microscale experiments; the use of miniature surgical machines that are injected into the body and find their way to the site of damage, repairing it when needed. This speech invented and foreshadowed the field of nanotechnology. We are on the cusp of seeing just what this can do.

Atomic force microscopy (AFM) is an amazing example of such technology. An incredibly tiny needle point is dragged across a sample. Cantilevers, lasers and computerized electronics are used to measure the landscape of the surface the tip is dragged across. Much like a phonograph stylus, the needle will move up and down as it interacts with the surface. So you get a readout of the molecular topography as you drag the needle up and down and across the surface.

Now, the original AFMs had pretty coarse needles. Incredible technology but not too useful. Newer models have much finer tips, allowing the AFM to reveal much greater detail of the surface. By adhering biologically important molecules to a surface, an AFM can now reveal highly detailed information regarding the structure of the molecules. Or even the structure of a single molecule!

A recent paper in Nature Structural Biology used AFM to examine the protein-protein interactions between single molecules of the chaperonins, GroEL and GroES. They developed new cantilevers that are much more delicate than previous ones, able to detect forces that are 30 times smaller. They dragged these probes over a mica surface which had GroEL protein bound to it. They could detect the molecules as differences in height, from 0 nm to 15 nm. The average diameter of a molecule was 14.6 ± 2.2 nm, closely matching X-ray crystallography measurements. They then added GroES and ADP (both are required for binding to GroEL) and detected a 3 nm increase in height, very close to that seen in crystallography. They could repeatedly scan the same section and still see the same molecules bound together.

Now the cool part. Because of the nature of the electronics used, it is much easier to scan the array in one dimension than in the other. They could scan the same y-axis every 100 ms. So, if they simply have the probe repeatedly scan back and forth across the same axis, they get a 'tube' of data, highlighting the temporal nature of the GroEL-GroES interaction for a row of single molecules.

This figure shows the results. The scan allows you to see directly the binding characteristics of single protein molecules to each other. Now we do this every time we do a binding assay but there is a huge difference. The binding assays average the kinetic constants over millions of molecules. This apparatus examines the binding kinetics of a single molecule. We can see one molecule of GroES bind to one molecule of GroEL, for how long and the amount of time before another rebinds.

It is just mind-boggling to me that we have the ability to visualize a single protein molecule and to examine its ability to interact with another. Speeding up the electronics and developing smaller probe tips will allow us to examine even smaller molecules, with much greater resolution and shorter time periods.

One dream of AFM is the direct sequencing of DNA. Most constructed cantilevers are much too large to be able to do this. However, extremely narrow carbon nanotubes have been constructed and can be used as part of the cantilever. The use of these single-walled carbon nanotubes (SWNT) in AFM was applied recently to DNA haplotyping.

Single nucleotide polymorphisms (SNP) hold tremendous promise for examining human genetic disease. They will allow us to quickly identify and map disease loci. However, in many cases one SNP will not be sufficient to accurate detect the disease gene and multiple SNPs, all closely linked, will be needed. A problem arises due to the diploid nature of the genome. It is often difficult to determine if 2 closely linked SNPs reside on the same chromosome, or are on homologous regions of the 2 separate, diploid chromosomes. In one case, the gene may be expressed correctly, while in the other there may be incorrect expression.

So, Woolley at al. use AFM to examine DNA molecules. The use of SWNT gives them a resolution of about 10 bases. What they do is anneal oligonucleotides to specific sequences of the DNA. These primers have large molecules attached to one end, say a streptavidin. These show up really well using AFM and allow them to localize exactly where on the DNA molecule these oligos anneal. By using different sized labels they can examine two or more oligonucleotide primers simultaneously.

Using these approaches they could correctly localize a SNP to within 10 bases on a DNA molecule over 10,000 bases long. They demonstrated that they could do the correct haplotyping of some authentic polymorphic sites. They believe that they can make some simple improvements to extend the resolution to 10 bases amongst 100 kb of DNA sequence. Of real interest is the multiplexing abilities of this approach. A single SWNT probe could examine 200 samples a day, with a redundancy of 10 images per sample. However, they can create a 32 x 32 grid of AFM cantilevers and haplotype over 200,000 samples in one day!! We may not be too far from the day when you can quickly determine what a person's SNP pattern really is. Talk about the data generation from such a project as this. And the possible ethical problems that could result.

Even more incredibly, SWNT can be made even smaller. Ones with radii of 0.25 nm have been made. Since these are smaller than the spacing between DNA base pairs, they hold the possibility of directly determining the sequence of a DNA molecule. We could then sequence any type of DNA, without the need for cloning, overcoming some of the problems still inherent in current technology.

Many of the early, and great, molecular biologists of the 1950's were trained as physicists. They brought their theoretical knowledge, and their tool building approaches, with them, applying these to solve biological problems. It appears that they are again bringing many of their tools, developed for other reasons, and derived from a speech given by a physicist over 40 years ago. It seems that as we continue to get smaller, the stature of Richard Feynman grows larger.