Imprinted Polymers AND Carbon Nanotubes? Is That Even Allowed?

I ran across a paper a few days ago that caught my eye for two reasons.  The first reason is that the paper deals with applications of molecularly imprinted polymers, and applies molecular imprinting to a sensing scheme that I hadn't previously considered.  The second reason for my keen interest in this paper is that the analyte that the authors are trying to detect with their sensor is a cardiac protein called Troponin T (TnT), which is what much of my Masters research focused on - developing a fluorescence-based immunosensor for TnT.  The paper, titled Artificial antibodies for troponin T by its imprinting on the surface of multiwalled carbon nanotubes: Its use as sensory surfaces (subscription is likely required) is published in the October 2011 edition of Biosensors and Bioelectronics, which is one of my favorite journals.

I should probably briefly explain why TnT is important.  Cardiac TnT is part of a greater, multi-protein Troponin complex, and is found exclusively in cardiac muscle tissue.  When necrosis of cardiac tissue occurs, a hallmark of heart attack, the Troponin complex fragments into its three individual subunits (TnI, TnC, and TnT).  These subunits then become bloodborne and can be detected in the blood using standard laboratory testing.  Under normal, healthy conditions, the subunits are absent from the blood.  The current laboratory testing techniques for detection of TnT to diagnose heart attack are unfortunately somewhat lengthy, and there is a great deal of interest in developing a much more rapid diagnostic test that is both accurate and sensitive.  This was the goal of a large part of my Masters research.  And obviously, this is the goal of the authors of the paper described herein.

I'm not going to go into depth on the technical details of the sensing system that the authors developed, but it's important to have a general understanding of how the thing works.  The researchers started with multi-wall carbon nanotubes (MWCNTs) and chemically modified them, allowing them to bind covalently to TnT when it was introduced.  The MWCNTs with bound TnT were then coated in a functional polymer capable of forming a number of weak noncovalent bonds with the TnT.  Then, when the TnT was removed using a chemical extraction procedure, a molecularly imprinted binding site for TnT was left behind.  I've described this process previously, and it is diagrammed in the figure below, which is from the paper.

The MWCNTs coated in molecularly imprinted polymer were then suspended in PVC and coated onto a wire electrode.  The electrode was then exposed to a solution of TnT and voltage measurements were used to determine TnT binding.

Although I found this paper quite fascinating and an interesting use of molecularly imprinted materials, I have to say that the methods the authors used are somewhat flawed.  The most glaring problem that I noticed was the extraction procedure to remove the TnT template.  The authors claim that the TnT was removed by introducing the MWCNTs with the imprinted polymer to a solution of oxalic acid.  Now, I'm not all that familiar with oxalic acid, but I'm not entirely sure that this extraction procedure is robust enough to disrupt the covalent amide bond that was used to link the TnT to the surface of the MWCNT.  This would explain why the authors observed large peaks that are normally associated with proteins when x-ray spectroscopy measurements were taken.  Another fairly major problem is that the modified MWCNTs were embedded in PVC for the potential measurements.  Although TnT is not an abnormally large protein, it is a protein nonetheless, and is therefore a substantially sized molecule in the grand scheme of things.  Due to its size, it is unlikely that TnT would be able to infiltrate the PVC matrix.  That would mean only those MWCNTs that are exposed at the surface of the PVC layer would be able to effectively bind to TnT, rendering the rest of the MWCNTs - probably a majority of them - inaccessible to the analyte.


All in all, the idea behind this project is pretty cool.  It was poorly executed, though, and I have difficulty believing that they actually got the results that they claim to have gotten.  I can think of a number of ways to go about developing a carbon nanotube-based electrochemical sensor that uses molecular imprinting to impart selectivity, and I can honestly say that I wouldn't have decided on the method described here.

A Few Notes on Antibodies: Part 2

Now that I've covered the basic function of antibodies and how they're made, I think I'll turn my attention to immobilization techniques.  There are a number of different ways to stick antibodies onto a solid surface, and the strategy that is used is mostly dependent on the type of surface you're working with.  Gold, for example, is pretty easy.  This is because the cystein residues that are present all over the antibody structure will bind - albeit fairly weakly - to gold.  There is a natural attraction between the thiol group of the cystein residue and the gold.  So when you expose IgG antibodies to gold at a nice comfortable pH of 7.5-8.5, the antibodies will adsorb onto the gold surface.  Although this method is easy and relatively effective, it does not create a very stable bond.

To immobilize antibodies with a stronger binding scheme, the protein must be covalently attached to the surface.  This type of covalent linkage between a surface and a protein is often used when the surface is glass.  Glass is a material whose surface is surprisingly easy to modify using a class of compounds called silanes.  Silane molecules are most often based around a single silicon atom.  The silicon atom has three ethoxy or methoxy groups.  These groups will covalently bind to glass, creating extremely stable bonds that are also able to crosslink with other nearby silanes to further stabilize the silane layer.  The fourth valence electron is bound to an organic species - usually a functional group connected to the silicon via a short hydrocarbon linker.  One of the more common silanes used in immobilization techniques is mercaptopropyl(triethoxysilane), and its structure looks like this:

By looking at the structure, you can clearly see the three ethoxy groups (O-CH3) bound directly to the Si atom, and the one mercapto group (SH) connected to the Si atom by a three-carbons (propyl) bridge.

So once the surface of the glass is functionalized with a silane layer, it is much more reactive than the fairly inert native glass surface.  The next step would then be to connect the functional layer of the modified glass surface to one of the amino acids of the IgG antibody.  This is accomplished through the use of a crosslinker.  To give give an example, the crosslinker that I have the most experience with is GMBS (long chemical IUPAC name:  4-Maleimidobutyric acid N-hydroxysuccinimide ester).  And this is what it looks like:

GMBS is known as a heterobifunctional crosslinker because the two ends of the molecule are different and are reactive towards different type of functional groups.  In this case, the maleimide group on the left binds covalently to the mercapto group of the functionalized glass.  The succinimidyl ester on the right then binds to amine groups found on the peptide chains that make up the antibody protein.  Once this reaction successfully completes - which happens fairly quickly - you end up with a glass surface that is coated in IgG antibodies.  And therefore, the surface is now capable of selectively binding the antigen of interest.

One last note about antibody immobilization: the the steric position of the antibodies on the glass is important.  By simply crosslinking the protein directly to the glass surface, you have no way of controlling the position of the antibody.  For instance, the crosslinkage could occur at or near the antigen binding site of antibody.  This would mean that this particular antibody would end up immobilized 'upside down,' with the antigen binding sites so close to the glass surface that the antigen would be unable to bind.  To remediate this problem, you can first crosslink special proteins, such as Protein A or Protein G, to the surface.  Protein A and Protein G have a binding site that is specific to a highly conserved region near the 'bottom' of the antibody, on the opposite side of the IgG from the antigen binding sites.  After immobilization of Protein A or G, you can introduce the antibody, it will bind to the Protein A or G, and you end up with a surface in which all of the antibodies are pointed 'up' with the antigen binding sites exposed and available.

A Few Notes on Antibodies: Part 1

During my doctorate research, I shifted my research focus from more traditional biosensor technologies, such as immunosensors, to sensing applications of molecularly imprinted polymers.  However, I still dabble in immunosensors fairly often, and my expertise and previous experience in antibody-based sensors comes in handy, particularly for other researchers who come to me with questions and guidance in this field.  Here at LU, my research group is currently waiting to receive a fluorescent microscope that we recently ordered.  When the new scope arrives, I will be using it to conduct a study examining targeted antibody immobilization onto gold electrodes.  The plan is to block specific areas of our sensor substrate so that antibodies will only be able to bind to certain regions.  When fluorescent bacteria are introduced, they'll bind specifically to the antibodies, and this phenomenon should be clearly visible under the fluorescent microscope.  So in the spirit of beginning this work, I thought I'd discuss some of the basics of antibodies and antibody immobilization.  I'm going to be brief, so I may have to turn this into a series of posts.

First off, I need to clarify one thing:  when I say antibody, I am referring solely to immunoglobulin G (IgG) antibodies.  Next, I guess I should describe what antibodies are and what they do.  The IgG antibody is a relatively high molecular weight protein.  It's produced by our immune cells (B cells) in response to infection.  When the antibodies are produced during infection, they bind to the bacterium or virus or whatever, coating its surface and acting as a sort of signaling beacon.  Other immune cells are able to detect this beacon and attack and destroy the infection.  So in a very general sense, that's the natural function of antibodies.

This natural function of the IgG antibody is made possible by one of its most important properties; selectivity.  The IgG antibody is a large Y-shaped protein that looks something like this:

At the top ends of the two 'arms' of the antibody are binding sites that are capable of binding to one single antigen, which is the target bacteria or virus or whatever, and only that one antigen.  This is what is referred to as selectivity.

Researchers like me who are interested in using the properties of antibodies for biosensors and other applications must be able to readily produce or purchase these antibodies.  As it turns out, there is a fairly effective way of harvesting the antibodies that are selective for a particular antigen of interest.  Let's say, for instance, that you want an antibody that is selective for the flu virus.  To produce these antibodies, you would take a syringe that contained the flu virus and inject it into the lumen of the gut - or some other space where the antigen would not leak out into the bloodstream - of a mammal.  The mammal used is commonly a mouse, rabbit, or goat.  The B cells infiltrate the gut lumen of the animal and begin churning out antibodies that are selective for the flu virus.  Then you would go back and draw out the fluid from the lumen of the gut and purify the antibodies that were produced.  And there you have it - anti-influenza antibodies.

Having only touched the surface of this topic, there will definitely be more to come.

Learning Curve

Whoa...Blogger has made some changes to their interface, so hopefully my confusion won't result in any illegible posts.  I must say, though, that they've simplified things nicely.  Perhaps too simple, as it's been difficult for me to figure out how to do tasks that used to be almost second nature. 

For example, the old interface had a navigation bar at the top of the page, with one of the nav buttons displaying the word 'STATS.'  When clicked, you were taken to a page in which you could view various statistics and information on your readership (number of blog and individual post views, redirect URLs, etc.).  To get to that same stats page with the new interface is a little less intuitive.  There is a button near the top of the page with an icon that resembles a piece of paper with text on it.  Next to that is an upside down triangle.  When you click on the upside down triangle, a drop-down box opens that contains a list of link options.  One of the options is 'Stats,' and when you click on it, you are directed to the statistics page that I described previously.

Don't get me wrong, the interface is very sleek and I think I'll come to prefer it to the old one.  It's just going to take some getting used to.  I hope.