That False Negative Problem

So whenever you test someone for some disease, you can have a negative, positive, false negative, or false positive.  As far as the falses go, a false positive is probably not that bad.  In the case of our system, that would mean that an output was produced when the person does not actually have cancer. After the output was produced, another test would be performed to actually find the cancer with a scope and it would be discovered that there is no cancer.  Although it would cause the patient a lot of unnecessary grief over their diagnosis, a false positive eventually just ends with the patient being relieved.  A false negative, however, is much more frightening.
Recently, our group discovered a really probable false negative that our system could give when interacting with GI cancer.  Since our system only produced output when Beta-catenin was present without Wnt, the cell would be free to have free Beta-catenin and Wnt and not show that it is cancerous.  If the presence of Wnt and free Beta-catenin was just an occasional occurrence in cancer cells and they would produce the output at all other times, this would not be a problem. However, some cancer cells suffer from hyperactive Wnt (too much Wnt).  Therefore, there would be no output would be produced, but the cell would be cancerous.
We are in the amidst of figuring out how to deal with hyperactive Wnt.  Our current solution is to have two systems.  One system is the old system and the other involves miRNA degrading a second output.  We have also looked at possible ways to use Wnt presence as a repressor and promoter for the same output.  This one-system system would create fewer parts for the cell to produce in order for the system to work, but may be less accurate.  This is a dilemma into which we are still looking.
This issue has made me think of the part of the project on which I was just working. I had to find a way to amplify the output.  I think that we are going to use the modular amplifier that I found even though we have to produce a few more parts.  This reminds me of the two-system vs. one-system debate because it forces the engineer to decide between a larger system built with modular parts vs. a simpler system that is tailored specifically to the issue/goal at hand.  Maybe the second way is better, but I am very grateful that papers have been published and the Registry of Biological Parts has been created to help make the whole field more modular and a little easier to grasp by people just learning things like me.

New Chassis

After receiving feedback from the tech spec, the team now has new direction: to find a new and improved chassis that won't die of malnutrition in the ocean waters.

We identified several candidates, but after much deliberation it was decided that cyanobacteria would, after all, probably make a better chassis than Lee would. So we've been looking into several strains that look promising. Exactly which strain will fit our needs best remains to be determined, and further researching that is our next step. After that is decided, it seems like we're just going to transfer the old device-level system onto the new chassis by looking for new parts that are compatible with the new strain of cyanobacteria. I find it pretty amazing that you can just transfer this kind of system, and here again the versatility of bioengineering, at least in theory, strikes me.

I kind of wish we could actually try this in a lab. It would be so cool.

I Love Viruses! (and other things I never thought I'd say)

This past few weeks in 20.020, I've come to a startling realization:

I love viruses!

Note: that should not be read as "I love experiencing viruses." A clear distinction must be made here. Regardless, recently my job in Team Human Health/Cancer/Turtle has been to look into viral vectors and their human implications, and I am finding that I actually really enjoy it. I have been able to research many different kinds of viruses that are widely used in clinical applications and what makes them unique and effective. The virus we decided on is the EBV/PAC viral vector, which has been widely used in medicine and can hold an insane amount of genetic material. It was recently used to construct a human genomic library. In other words, it can pack a lot of information into a small space. Most people have already had EBV, so they are immune to the wild-type virus. This means that clinical trials have been pretty successful and low on side-effects.

Because of the fact that most people have already had EBV, one concern of ours was that the immune system would recognize and target the host virus and destroy it using existing antibodies. However, further research showed that the immune response to the virus is dependent on the presence of helper viruses found in the wild-type infection. Without these viruses (which are not present in the manufactured vector), EBV can successfully inundate its target cell.

The research on viruses is pretty much complete at this point, so I'll be moving on to helping the team with other aspects of the system; but the EBV vector will always have a special place in my heart (another thing I never thought I would say). I'm excited to see what other interesting things come up as we move on with the project!

We're Makin' It

In the past two weeks, I've seen our decomposition project grow from hand-wavey ideas and concepts to something concrete, with detailed parts specifications and a plan of action to bring it all together. It is very exciting, being a part of this process of creation and seeing how your personal contribution plays just as important a role as everyone else's.

I think it's hilarious how, despite our thoroughness in designing the system, we entirely overlooked the method by which the e. coli would receive their sustenance! As it currently stands, our chassis wouldn't have lasted much time at all, when you consider the amount of energy required to pump salt out and build decomposition enzymes, all while being starved in the middle of the salty ocean.

Wednesday's inventor/investor activity was highly illuminating in terms of just how much capitalism gets in the way of advancing the current states of science and medicine. I voluntarily hindered another team's progress just so I can sign an exclusive deal and rake in the dough, and while the stakes were significantly lower in our little social experiment, I can't help but to think how much these things happen in real life, on a large scale.

Post Tech Spec Thoughts

The feedback we received after the tech spec was very useful: to either switch our chassis or our target location, and to simplify our system. Looking back on it, it seems like it should have been obvious that we should use an ocean organism to solve an ocean problem, but I think we were drowning in our many devices to the point where we forgot to even think of what our system would look like as we deployed it. Our E. coli wouldn’t have had a carbon source or even a native salt water survival system and wouldn’t have lasted very long in the sea. Right now, we have agreed to switch to a cyanobacterial chassis but are still deciding what strain to use- a common strain that floats, or a less-well-researched species that can adhere to plastics and other hydrophobic objects.

As for the device list, we are narrowing down our focus from four targets (PET, polyethylene, polystyrene, and phthalates) to just one, polystyrene. (We will add in others later if we have the time.) We’ll also have to remove my favorite part, the OmpW salt pump, as cyanobacteria naturally thrive in the salty ocean water. We are not sure now whether to even include a similar cell death mechanism, because how much damage could these cyanobacteria really do if they got loose?

The cell death question is now one of a large number of open questions on this project, and it’s kind of difficult to know how to proceed in answering them. I’m almost wishing for another Tech Spec—another chance to regroup, reassess, and move us forward in the best direction possible. This project is also making me realize just how challenging the different aspects of an open-ended project are: narrowing down research so the problem does not seem overwhelming, planning out a timeline for discussing certain aspects of the problem, and balancing higher-level thinking with detail specification. They are all super hard! Hopefully by the end of the semester I’ll have gotten at least a little better at these skills, even if we don’t have a design that would really work.

After reviewing our feedback and planning out a new direction of research, our group is going strong.  Go Team 99 Bottles!

Coloring With Cancer

I really enjoy brainstorming. The process of developing ideas under the assumption that everything is perfect and nothing can go wrong is fun. Actually fleshing out an idea is a much harder process involving some serious reality checking, and that is exactly what I discovered this week. Our project is to develop a biological mechanism to tag GI cancer. Ideally this would replace the current, highly-invasive methods implemented in hospitals today. Most widely used are endoscopies, which involve sticking a long tube with a camera down the patient's throat-- a less-than-pleasant sounding procedure. Hopefully our project would allow patients to physically see something in their excrement that would alert them to the presence of cancer cells. 

This sounds like a simple enough concept. I thought so too, until we actually began looking into how cancer environments actually work. Many of the surface proteins that we had intended to tag are found on normal healthy cells, which increases the possibility of a false-positive. In addendum, the high turgidity and low pH of tumor microenvironments make them nearly indelible to drug access. Our method will need to be able to detect cells on a very nominal level, before that microenvironment establishes. However, this creates a problem, as then the surface proteins will be in similar concentrations to the surrounding cells. 

Despite these obstacles, the team managed to conduct some stellar research and find surface tags that are specific to GI cancer cells. We're still settling on which receptor we are going to use, but at least now we have options that can latch onto our target cells. Furthermore, we found a high-capacity adenoviral vector into which we could transform our desired gene sequence. Our plan involves a large amount of recombinant DNA, so the high storage capacity of the virus is a major draw. It also is highly effective at accessing target cells, so breaking into the tumor microenvironment should be much easier. While the term "adenovirus" may raise some red flags, all of the clinical trials involving its use in other cancer studies saw minimal side effects. 

As I said, brainstorming with no consequences is fun, but I am finding that I definitely prefer the actual design process of a real idea. Sure, it's harder and there are plenty of obstacles, but our project is beginning to come to life, and I'm really excited to see what challenges we overcome next and how our design continues to grow from here.

The Devil's in the Details

      The last two weeks (aside from the much need interlude that was spring break) have largely been spent investigating at ever increasing levels of specificity the projects we have chosen.  It's very easy to see why curing cancer is the subject of such a large amount of research; there are seemingly infinite possible mutations that can lead to the disease making it difficult (if not impossible, at least for the moment) for any one detection or treatment method to work on all possible cancerous phenotypes.  As such we've decided to narrow our primary goal to detecting gastrointestinal (GI) cancers, as that is an area of the body that is fairly accessible.  You can ingest a pill, put something in a piece of food or just straight up use a scope to get things into the GI tract; much easier than something like the brain which has that messy blood-brain barrier to contend with.
      Our decision to narrow the scope of our project has made our job of detection (at least a little) easier.  Since GI cancers are in the same areas and involve many of the same kinds of cells, they tend to have more in common than any two generic cancer cells from other forms of cancer, making it easier to avoid false positives on our test.  The job of finding a reliable marker was still in no way trivial.  We investigated a variety of surface proteins that are common on cancer cells but ruled them out because most were also upregulated in several other non-cancerous diseases.  In the end we decided to focus in measuring an activity within the cells that we are screening for cancer.  According to research we found, one of the most common mutations in GI cancers is loss of function in the APC gene.  We'll go into more detail about how we plan to check for that mutation in our tech spec review this week but I'm pretty excited about what we've come up with thus far.  Overall, I think we have a pretty solid system planned for checking for APC mutation and while I hope we can find a more generalized solution (perhaps something involving the common loss of DNA repair mechanisms), things are looking pretty good.