Monday, December 10, 2012

Technologies That Will Change Everything

Every now and then, a technology comes along and changes things so radically that it disrupts whole economic structures and ways of life.  No doubt, just about anyone can name quite a few of these off the top of his head -- the internal combustion engine, the printing press, the computer, etc.

Here are three technologies which I think are on the cusp of doing just that.

3D Printing

This technology will revolutionize manufacturing.  Using a CAD program and one of these devices, practically any three-dimensional structure can be produced out of a plastic resin --

Other materials are also in the works. Right now this technology is still rather expensive, so it mostly only finds limited use in the production of relatively high-value, custom parts (dental implants, prototype parts, etc.) But the cost is rapidly falling, and already low-end consumer models are available, some of which are being used for purposes that the finger-waggers might not like.

Talk about decentralization. Oh, and goodbye cheap plastic Chinese crap. Whole economies are going to need new economic models.

A Breakthrough Treatment for Leukemia

About five years ago, I was talking to a chemistry professor after a seminar, and I told him that I thought that 'the' cure for cancer would come from a combined approach using genetics and immunology. I thought that once immunology had reached a certain point, it would become possible to genetically modify immune systems to better recognize cancers. These modified cells could be injected back into the patient (or possibly created there, for that matter), where they would seek out and destroy the cancer which they had been 'trained' to find.

This was not going much out on a limb -- the human immune system already has the capacity to destroy cancers that it can recognize; in fact it does so on a regular basis. This is one reason that AIDS sufferers and other immune compromised people so regularly contract cancers that almost nobody else ever gets. Normally, the immune system destroys these cancers, but if you don't have an immune system, well, tough luck for you.

Anyway, several researchers appear to have used just this approach against a type of leukemia, with amazing success. Again, I will not go much out on a limb, and call this 'the' cure for cancer.

(Sorry, could not embed the video, as it was somehow ruining my html...)

The really, really nice thing about this approach...well, there are too many really nice things about it. They appear to have solved a number of problems that have plagued this type of approach in the past.
  • First, though they don't tout this so much, is that the approach is modular.  They identified an antigenic marker for the targeted cells, and then 'trained' T-cells to identify and destroy cells with this marker by genetically modifying them.  This is a general strategy -- you just need to identify an antigen on the cells you would like to destroy.  If you can find the antigen, you've got a more or less packaged approach for producing an overwhelming immune response against it.  Most likely there will be complications in every particular case to overcome, yes, but it is still remarkably general.
  • They found a way to use a virus for 'gene therapy' (i.e., human genetic engineering) without using the virus directly on a human.  This has been a big obstacle for gene therapy.  Viruses are immensely useful genetic tools, but you've got to be very careful about engineering them and using them on people.  After all, a number of cancers and lots of really nasty diseases are known to be caused by viruses.  By using the virus on human tissue outside of the body, the tissue can be 'checked' before it is used, to confirm that there haven't been any oncogenic or other untoward side effects.  Assuming, of course, that you've got a good screen for this...
  • They seem to have solved the 'oomph' problem.  Many, many approaches to cancer therapy have made use of the basic strategy of 1) identifying a good antigen or other cell marker, 2) finding something to bind to the marker, then 3) attaching some really nasty poison or other toxic agent to the binding agent, so that the coupled system forms a sort of 'guided missile' to seek out and destroy the specified tissue.  The problem that most of these attempts have run into is that they only partially work.  They attack some of the target, but don't have quite enough 'oomph' to finish the job and eradicate the cancer, so that the patient isn't really cured.  This approach seems to have done very well in this regard -- mainly because the 'guided missile' is itself alive and capable of flourishing inside the body on its own.
  • The approach is highly personalized, using the patient's own living tissue and a target specific to the cancer being treated.  It avoids a great number of complications and difficulties in this regard, such as the general bodily malaise caused by chemotherapy, and the necessity of large doses and prolonged application of very expensive materials.  One patient was given a dose of cells more appropriate to a mouse.  Nevertheless, once inside his body, the cells multiplied and gave the cancer holy hell anyway.
To use this approach on other cancers will of course require a rather encyclopedic knowledge of cancer itself, in order to identify the requisite cellular markers specific to each case.  But a great deal of this knowledge has already been compiled, and better tools are being developed every day to add to it.  It will also require a tailoring of this approach for each type of cancer, so that a virus carrying the requisite marker will be produced for each type of cancer that the treatment is to be applied to.  Nevertheless, I expect that either this technology, or one very much like it, will rather soon begin seeing widespread application as a personalized and cost effective treatment for a wide range of cancers.    

A Promising Approach for Alzheimer's

OK, this one is a bit of a stretch.  But I really needed three to round out my list, as you can't have a respectable list with only two entries, and even so, it is a pretty good one.  I don't think these guys have hit upon the cure yet (for a number of reasons), nevertheless, the approach and the results are quite promising, and I wish more labs would use approaches like this --

What these guys have used is a sort of 'irrational' approach to problem solving. There are many labs using this kind of approach for all sorts of applications, mostly going under the heading of 'combinatorial.' The strategy is to throw up one's hands at trying to understand the problem and engineer a rational approach, admitting that the difficulty and complexity of the problem has defeated you, and simply start 'trying things,' usually on the basis of a 'lead.' You figure out some really cheap and effective way to produce loads and loads of 'candidate answers,' and start trying them one by one.

In this case, it would be too difficult and expensive to just start injecting elderly people suffering from dementia with experimental chemical compounds. Not to mention just plain mean. So, instead, they grew 'brain cells' in little dishes, treated them with their 'candidates,' and then subjected the cells to all sorts of abuse to mimic the conditions that prevail in the brain of a person suffering from Alzheimer's. If some of the 'candidates' were able to prevent the cells from dying -- effectively protecting them from the 'abuse,' they could be investigated further in animal models.

Yes, I'm vastly simplifying. Actually, you can tell from the rather awkward wording of their paper that they did a lot of improvisation and work-around, probably because their synthetic chemist appears to have had trouble making pure compounds, which is what you get our of a $40K/yr synthetic chemist. So, they actually wound up fishing out the useful compound from a complex mixture. I once had a Mongolian postdoc commend me on one of my reactions with "Ahhh! What a wonderful reaction! So many products!" Which passes for humor among such people (because, no, reactions that produce lots of different products are not good reactions...). Anyway, that appears to have been this guy's problem, but you get the basic idea.

As it turned out, they did indeed find a compound that 'cured' rats genetically engineered to have 'rat Alzheimer's.' But there has been no follow up on this paper that I can find. So, I suspect that it has died, probably because it has been found to have side effects, likely with regards to cancer, which is something that tends to happen when you start interfering with cell death.

Nevertheless, 1) it is a very promising approach, the surface of which is barely scratched in this paper, and 2) such compounds can often serve as useful 'leads' to help researchers identify other pathways involved in the disease, as well as pathways that might be used to alleviate things if the undesirable side effects can be minimized. With small molecules, it is often the case that the compound binds many different targets, most with little effect but sometimes a few will cause trouble. It is simply a problem of thermodynamics -- a small compound which binds one type of protein (or other target) is likely to bind many similar molecules, simply because it is too small to distinguish between them energetically. It only has so much surface area to work with. If the effects can be teased apart, however, more specific molecules can be designed.

On the other hand, if anyone would just develop my suggested technology, in a few generations we might not be worrying about a lot of this stuff at all...


  1. I totally agree the standpoint of upstairs. I often surfing on this forum when I m free and I find there are so much good information we can learn in this forum!
    we buy houses cash

  2. Why don't you do this kind of chemistry?

  3. Ummmm.... What kind of chemistry? You mean the Alzheimer's stuff? Yu did that kind of thing. I was never involved in it, but I did like the idea a lot.

    Now, of course, I'm not doing any chemistry at all, unless I'd like to take a 50-60% pay cut and become a post-doc. No thanks. I figure my life's better spent doing other things.

    Anyway, there's lot's of people doing this stuff. They just don't generally come out and say "We're not going to pretend; we don't really know what's going on, and we figure that since nobody's figured it out so far, it's probably really hard. What would really be useful are some new leads. So we're going to instead spend our brainpower on things we actually do know how to do well (create biological screening methods (which this group did very well), design combinatorial synthetic methodologies, etc.) instead of beating our collective head against a brick wall. Hopefully, something we find will shed some light on the problem."