Biocurious is a weblog about biology, quantified.

The perils of research

by PhilipJ on 20 October 2006

For the past year, I’ve been spending the majority of my time starting a collaborative project between my lab and a group of microbiologists at the school across town. This past week, unable to really get things off the ground, I threw in the towel.

In past research-related posts, I was rather coy about what I was doing, and when I went to Japan, I didn’t fill everyone in on the project deemed interesting enough to fly me across the pacific. Unfortunately, basically since then, I’ve made no significant progress. To anyone wondering what I’ve been doing, and what I’ve started doing instead, here goes.

If you want to get DNA inside of cells (for, say, growing new proteins, or making a strain of cells that is resistant to certain antibiotics), there are only a couple of options. For E. coli, one commonly incubates cells with your DNA of interest, and either zap them with electricity, or heat them up quite rapidly in warm water. Both have the effect of making the cell membrane significantly more permeable to, well, everything, and DNA somehow passes through and is then able to get expressed by the cell’s transcription and translation machinery. As you can imagine, the cells don’t like this, and many die. Those that don’t, however, now have new DNA inside of them.

There are other cells, however, which have special complexes which span the cell membrane that can actively bind and internalise DNA, for reasons which are still contested. Two of the main ideas are food and gene transfer. As a nutrient source, DNA can be broken down into individual nucleotides, which could then be used for any number of dogmatic processes. As a gene transfer mechanism, uptaken DNA could be incorporated into the genome, or simply expressed as is, as a way to gain new genes. The common bacteria Bacillus subtilis is one example of a bug which is naturally able to internalise DNA.

In the bacterium Haemophilus influenzae, the story is even more interesting. Haemophilus is a gram-negative (having two membranes) bacteria, so any DNA uptake machinery is going to have to span two separate membranes (B. subtilis’ need only span one). Complicating things even further, to get efficient DNA uptake, a 9 basepair Uptake Signaling Sequence (pPIU should now make sense to you!) is required, which is thought to be perhaps a binding site on a receptor protein on the cell surface. It isn’t entirely clear, but without this sequence, uptake basically doesn’t occur at all.

Given our laboratory’s ability to tug on single DNA molecules, a group across town got excited and contacted us about starting the collaboration to look at DNA uptake with live cells at the single molecule level. It didn’t sound hard to begin with; make some DNA that has both a USS and will be suitable for uptake measurements in the optical trap, and find some way to immobilise cells to a surface. To anyone who has never started a brand new project before, please take note: everything is hard to begin with.

Though I have no formal training as a molecular or microbiologist, I didn’t see any reason why I couldn’t do this kind of work, even if there might be a learning curve to start out. Surprisingly, the DNA cloning steps took only a month or two, but immobilising the H. influenzae cells to a surface has been much trickier than we hoped.

We wanted to stick cells to the surface of beads, either by non-specific interactions, or via antibodies to surface proteins found on the cells, as cells don’t usually live long in the focus of the optical trap, and our micropipette tips are already finicky enough without clogging them with cells. By immobilising the cells onto beads, we could save ourselves a lot of hassle. Our lab has all kinds of beads with different coatings (carboxy groups, protein G, anti-digoxigenin, streptavidin), but the cells didn’t seem to bind naturally to any of them. They also don’t seem to immobilise on glass coverslips.

Up next was contacting another lab for antibodies to common outer membrane proteins. I was sent three different kinds of sera, made to target the unimaginatively named P2, P4, and P5, all residing on the outer membrane, with P2 accounting for as much as 20% of all membrane proteins! This sounded promising.

Through a number of tests (dot blots), I can coat beads with these antibodies, and the cells seem to bind the antibodies as well, but at no point have I seen cells bind to antibody-coated beads. Never. Zero times! I’m not able to explain why.

Finally, we decided I should attempt grabbing cells in the pipette tip to test for uptake events; and should I see any with reasonable frequency, to brainstorm new ways to attempt to immobilise our bacteria. After testing a number of cells (and clogging a few pipette tips along the way) without observing any kind of uptake events, it was decided to give things a break.

I’m not going to lie, I’ve had a really hard time trying to stay positive these past couple of months as experiments continued to fail. It is no secret that science is hard, or that some (most?) experiments end in failure or with entirely ambiguous results, but it is so easy to forget that when they aren’t your own experiments.

So, what now? I’m in the physics department, so they unfortunately won’t really care that I’ve put a lot of work into something that we’re more or less putting on the back burner for a while, since that something happened to be primarily biology. I’m also a little too proud to write a masters thesis without feeling like I’ve accomplished something worthwhile, so I’ve changed to a new project involving single-stranded DNA which can form neat secondary structures, looking at the unfolding of the structure under different conditions. A bit less flashy, and taking so long for a masters degree doesn’t feel great, but with a little luck things will work out.

  1. Jean-Claude Bradley    3776 days ago    #

    Philip – I applaud you for taking the time to describe your unsuccessful project. As you say, you put in a lot of work that has no way of being shared by traditional publication mechanisms and that is very frustrating. I would encourage you to share as much experimental and conceptual details as you can – you never know which components will turn out to be useful for someone else, even if your objectives were not met. Putting data in a blog is a pretty good way of doing this since Google tends to index the information quickly to make it findable for people looking.

  2. Alex    3776 days ago    #

    Sorry to hear your first project didn’t work. Best of luck on your next one !

    To echo what was said above, preserving the stuff that didn’t work is definitely useful. I’d go a bit further than what you’ve posted abve, and consider putting it on OpenWetWare; while there currently isn’t a category for “Stuff that didn’t work”, I’m sure the folks that run OWW would be open to one [and I can put you in touch with them if you’re interested].

    I swallowed my pride a while ago, and expect my Master’s thesis to just barely drag me across the finish line. Then again, I think I have a slightly different set of pressures [ie a baby] than you :-)

  3. Bill Hooker    3776 days ago    #

    Rosie Redfield works on the same problem, DNA uptake by H. influenzae. I strongly agree with Jean-Claude and Alex, that you should share as many details as you can, and I would add that sending Prof Redfield a heads-up would be a good idea as well. You never know what use someone might make of your work; it might even resuscitate your project.

  4. PhilipJ    3776 days ago    #

    Jean-Claude – Thanks! I’m sure a Journal of Negative Results would be by far the most popular place to publish, and I’d go so far as to bet the impact factor wouldn’t be too low either. The project isn’t being given up on entirely, but my involvement is basically ending.

    Alex – I was going to comment over at Malletrivia about that; I’m sure little Zander will find a way to thank you for your decision later. Indeed, our reasons are quite different. I’m scrambling around trying to get this project off the ground now too, so once I catch my breath in a little while perhaps I’ll get in touch with the OWW crowd.

    Bill – That’s who I’ve been working with. :) There will be a more formal account written up for our two labs, and hopefully something will come of it further down the road.

  5. Jean-Claude Bradley    3775 days ago    #

    Philip – I don’t think we need a third party to share our research results. Just make your lab notebook public. I’m leading the session on open notebook science at the Science Blogging conference in NC in January. Your participation would be most welcome.

  Textile help