How do you make a nanobot?

Here is an essay I wrote while reflecting on and researching a lecture I heard by Yale Goldman, Professor of Physiology at the Pennsylvania Muscle Institute, and associate director of the NBIC at Penn.

How do you make a nanobot?

Imagine you are playing with a set of Legos. This set is one of the newer

ones. It has all kinds of blocks, gears, wheels, pulleys, batteries, etc. Put different

combinations of them together and you can build a house or maybe if you’re feeling

clever, a machine. Suppose you make a small battery-powered car. Imagine shrinking

this car. As it shrinks, the number of atoms in each Lego block must become fewer.

Eventually, you may reach the limit where each block is one atom. What if we could

make machines like this?

Eric Drexler had this vision in his 1987 book Engines of Creation: The Com-

ing Era of Nanotechnology. He imagined if we could manipulate atoms like Legos,

we could build nanomachines from the ground up, atom by atom. One could create

different kinds of machines by simply using different assemblies of atoms. Drexler

hoped that these “nanobots” could be used to do all sorts of things, like deliver drugs

to specific sites within the body. Micheal Crichton in his 2002 novel Prey envisioned

these nanobots existing, but in a typical Crichton ending the bots take over and kill

us. While Drexler’s vision is intriguing (and Crichton’s vision is scary), these hopes

(fears) should be looked at skeptically. For starters, it’s simply not possible to build

nanomachines using Drexler’s scheme.

Why is this? Consider a Lego block floating in a cup of water. This block is

surrounded by water molecules. The water molecules are constantly jostling about

due to a phenomenon called Brownian motion. You won’t see the effect of this phe-

nomenon with the Lego. Change the Lego to a speck of dust. The dust is much

smaller than the Lego and so is more affected by the jostling of the water molecules.

(You can actually see the dust being randomly jiggled around if you do this experi-

ment.) Now, imagine changing the speck to a single sugar molecule, comparable in

size to the water molecules. What was gentle jiggling of the dust translates into vio-

lent jostling of the sugar molecule. In fact, the sugar molecule collides with the water

molecules about a trillion times per second! So we really can’t place atoms together

in an orderly fashion like we can with Legos: the wind is too strong. In brief, the

nanoscale is a much harsher environment than the macroscale.

Despite the seemingly impossible construction challenge of Brownian motion,

nanomachines are ubiquitous. They are around us and already inside of us! Fear not,

as they are supposed to be there. They are naturally occurring biological molecules

that keep us functioning. So in order to make artificial nanomachines, we might need

to steal ideas that nature is already using. Biological nanomachines make use of a

few tricks, that scientists one day hope to copy. First, their molecular makeup is pre-

programmed by a DNA sequence. Secondly, they undergo self-assembly to make up

their final shape. Lastly, they use the Brownian motion of the surrounding molecules

to their advantage rather than detriment.

One example of a natural nanomachine is myosin. There are several different

varieties of myosin. They are all motor proteins that “walk” along long actin fibers.

All myosins are actuators, machines that convert one form of energy to another. Like

the motor in a battery-powered car converts electrical energy into motion, myosin con-

verts the energy of a chemical reaction (ATP to ADP) into motion. Clearly myosin

is a great model nanomachine to study for future technology. One common variety

of myosin (II) is responsible for flexing our muscles by sliding the filaments against

each other. One variety (V), is responsible for moving large amounts of cargo along

actin fibers. Myosin V has two “hands”, and seems to move hand-over-hand along

actin, similar to a child on monkey bars [Corrie, J.E.T., et. al. Nature 422, 399

(2003)]. However there does seem to be a key difference. The child on monkey bars

releases one hand, then does a power stroke with the free hand to get to the next

rung. Myosin does the same motion, but its power stroke only takes it 2/3 of the

way. It then lets the Brownian motion of the surrounding water push the rest of the

way [Shiroguchi, K. et. al. Science 316, 1208 (2007)]. By taking advantage of this

Brownian motion the efficiency of the system is about 50%, very comparable to an

electric motor!

While artificial nanomachines are still only a vision, it’s clear that natural

nanomachines are excellent prototypes. Unfortunately, our technology is not at the

stage where we can really imitate them. What scientists can do (and are doing now)

is try to understand more deeply how these fascinating natural nanomachines work,

and use this understanding to inform future technologies.

[1] Drexler, Eric. Engines of Creation: The Coming Era of Nanotechnology, 1987

[2] Crichton, Micheal. Prey, 2002

[3] Goldman, Yale. “Nature’s Nanotechnology: Biomolecules Explored One at a Time” Penn Science Café Lecture, 20th October 2010.

[4] Corrie, J.E.T., et. al. Nature 422, 399 (2003).

[5] Shiroguchi, K. et. al. Science 316, 1208 (2007).

Cartoon of still SEM images of myosin processing along actin from: Matthew L. Walker, Stan A. Burgess, James R. Sellers, Fei Wang, John A. Hammer III, John Trinick & Peter J. Knight. Two-headed Binding of a Processive Myosin to F-actin. Nature, 405, 804-807 (2000).

AWIS meeting report

Here is a report I did on the AWIS Philadelphia meeting. The topic was ethical considerations when pharmaceuticals and academic medicine mix. I am posting the full text here, or you can go to the AWIS-PHL website for a pdf version.


October 2010 Meeting Report by Kerstin Nordstrom

Kathryn Ross (MBE, DMH (c)), the Research Coordinator for Quality Research at the American Board of Internal Medicine, gave a presentation entitled “Pharma’s Ties with Academic Medicine:“Ethical Concerns?” on October 19th at The College of New Jersey. This was a joint meeting between the Central New Jersey and Philadelphia AWIS chapters. She opened by formally defining a conflict of interest, a key issue for any ethical concern. A conflict of interest is a set of circumstances that create a risk that actions regarding a primary influence (e.g. patient health, research validity) will be unduly influenced by a secondary influence (e.g. financial gain) [1].

Conflicts of interest (COIs) are present in every stage of life as a physician: in medical school, in residency, in practice. Ms. Ross presented results of a survey of residency programs conducted by the ABIM. Program directors and senior residents were asked, “What constitutes a conflict of interest?” There were several striking differences between the two groups. For example, 61.8% of senior residents classified excessive moonlighting as a COI, contrasted with 79.4% of program directors. Other notable differences were: personal use of sample medication (73.4% vs. 82.5%), accepting pharmaceutical gifts/meals/entertainment (75.8% vs. 89.7%), and accepting unrestricted educational grants (62.1% vs. 17.5%). The survey further went on to ask, “What is allowed at your institution?” The notable differences of opinion were: pharmaceutical detailing of residents (14.9% vs. 33%), distribution of free educational material such as textbooks (34% vs. 56.7%), industry support of continuing medical education (13% vs. 47.4%), unrestricted educational grants (6.8% vs. 80.4%), and faculty serving as paid consultants (26.3% vs. 61.9%). Clearly, the survey results indicate a discrepancy that points to a lack of clear COI guidelines. The survey further found about half of the programs had COI guidelines for trainees, about half had COI curriculum, but only a small percentage (<10%) had any COI competency evaluation for trainees.

Additionally, there was a difference in the opinion on unrestricted educational grants (UEGs) between university settings (22.4% classified UEGs as COIs) and community settings (12.7%). This may reflect that university settings are more likely to be pharm-free. One might come to the conclusion that academic settings are then doing “better” with regards to COI than non-academic settings. Perhaps, but the overall results of the survey also show that neither setting is doing especially well in establishing clear guidelines. Also, academic settings are in a better position to effect change from the ground up by actually instituting formal COI curriculum.

Ms. Ross went on to emphasize that the stigma attached to the term “conflict of interest” must be thrown out. Conflicts of interest technically only refer to situations where there is risk of undue influence. Classifying a situation as a conflict of interest does not mean any transgressions have occurred. In other words, COIs do not necessarily need to be abolished (though specific kinds may be); there is simply a need for clear COI policies.

In fact, many conflicts of interest have positive aspects that ensure they are around to stay. An example is the industry financing of medical research. All three parties potentially benefit: industry may develop a new product, academia may discover an effective treatment, and ultimately the patient is the beneficiary. These benefits must be weighed and discussed openly against the main risk of the COI: the patient’s health may be influenced by the profit motive of industry. However, as long as this COI is disclosed and discussed openly, action can be taken to minimize risk.

Therefore, in order to effectively deal with COIs, the first step is disclosure. However, evidence suggests even this basic first step is not always undertaken. In another study [2] a sample of 41 paid medical device company consultants was taken. From this number, 31 had published an article on their research. Of the 31 papers, 25 were randomly selected. None of the 25 articles published COI disclosure. This points to a need for a standard practice for disclosure. A standard practice would not only increase the rate of disclosure, it would also reduce the burden for disclosing. An indirect benefit is that this would create a larger data set on which to base policies.

Once disclosure practices are in place, the next step needed to address a COI is to assess the likelihood and severity of harm. As medical COIs usually involve patient health, this step is critical. Finally, one can determine if a COI needs to be prohibited outright or may be managed.

The timing of the meeting was serendipitously on point. Just two hours earlier, NPR and ProPublica began broadcasting the results of their “Dollars for Docs” investigation [3]. Under the recent Health Care and Education Reconciliation Act was included a law called “Physicians Payment Sunshine Provisions” where all pharmaceutical companies must disclose all financial transactions with physicians starting in 2012. Some companies have started doing this already, and the investigation tracked the payments. The investigators looked at 384 of the highest paid consultants and found that many had no board certification and many had allegations or convictions of professional misconduct. Additionally, several of the companies had recently settled massive lawsuits regarding the promotion of off-label drug use. While these examples may simply be “bad apples”, clearly this does nothing to placate public notions that industry may be “buying” or influencing doctors in unsavory ways. With better COI procedures in place, doctors and industry may hope to win back the trust of the patient.

Interestingly, the nature of industry influence is evolving as well. Historically pharmaceutical companies have been able to get prescription data from pharmacies. They can then target their marketing to specific doctors with data mining. However, some states have recently restricted this practice. Anyone with a television has seen the result of these measures: an exponential increase in direct-to-consumer marketing because of this restricted access to physicians. The patients’ involvement in the marketing only compounds the complexity of COIs, underscoring the need to address them.

While some previous examples paint an unflattering picture of some companies, it is important not to demonize the entire pharmaceutical industry. Many companies have their own sets of ethics guidelines and may follow them well. Companies are also in business, so it is unfair to judge them for desiring to increase sales or to develop a new drug. Further, they are the experts on the drugs/devices, and so are the informational authorities. Certainly many COIs in academic medicine could be avoided if all research funding was pharm-free and all medical education funding was pharm-free. But neither option is realistic in our society. Therefore, it is important to establish clear guidelines regarding COIs and to begin to establish formal COI training.

Ms. Ross ended the presentation by summarizing a recent meeting report from the Institute of Medicine [1]. The report summarizes recommendations to help medicine address COIs. Recommendations for academic medical centers (AMCs) are organized into three main points: 1) At AMCs, Prohibit gifts, ghostwriting, speakers bureaus, and limit drug samples, consulting, sales representatives. 2) Provide education on relationships with industry and COIs. 3) Develop a new system of funding accredited continuing medical education that is free of industry. Hopefully, AMCs can adopt these policies and standardized practices will result in the future.

While some might dismiss professional ethics as a personal rather than professional skill, the bottom line is over 90% of physicians get some form of industry support, and practically all medical research is industry-funded. While this is not necessarily a bad thing, many of these relationships necessarily entail conflicts of interest. It is of utmost importance to ensure these COIs are well-managed and prohibited when necessary. By managing them effectively, patients may be assured that their doctors have their best interest in mind: their health.

The presentation was followed by a fruitful discussion with the audience as well as a tour of our host institution’s science facilities.

[1] Conflict of Interest in Medical Research, Education and Practice, IOM, April 21, 2009
[2] Chimonas, Susan, Frosch, Zachary, Rothman, David J. “From Disclosure to Transparency: The Use of Company Payment Data” J Arch Intern Med (2010)