Getting Back on The Bike of Research: the Basics of Science Research

This week I moved back to the United States to work at my lab at the NIH for two months before my thesis defense in July. Not only was I in Oxford for 6 months, I was also concentrating on writing my thesis, and didn’t go into the lab except to meet with supervisors. So now I’m back in the lab on a daily basis and doing science. That said, I ‘m currently in the stage of simply accumulating supplies before I can do actually do any experiments.

Originally, the plan had been to simply write up, have my viva, and graduate, without returning back to the NIH. However, my examiners were busy, and couldn’t schedule by defense until July. This gives me some time now to return and try to generate more data. If the data is good, this means that I will have more work to support both my thesis defense and also to help make more coherent papers for publication. Since I’m only going to be back for 2 months, this really limits the scale of experiments that I can do, and certainly rules out completing any new transplant experiments since they take a minimum of 8 weeks to run, with the preparation and analysis times extra. However, I’ve come up with an experiment plan that closely matches what I’ve already done, but to take it a step further in some cases, or to simply do a more sophisticated study to glean better results.

I’m not concerned about coming back into the lab having been away for 6 months. With the exception of one major technique, all others are methods that I’ve done before. And I find that for planning and executing experiments, the logic and the process never change. It’s in line with belief that once you learn to ride a bike, you never really forget. I’ve been seriously studying science for 10 years now, 6 of which have been spent doing genuine research (instead of planned coursework experiments), and over time on learns the basics of experimental logic:


1)   How to choose an assay that will allow you to observe the specific factor you wish to

2)   How to minimize distraction by other factors by designing more simplistic experiments

3)   How to use proper controls to make sure that your experiment design is correct, even if the results you get are negligible or puzzling

4)   How to manage resources and time

5)   How to minimize human error (this involves knowing your personal limitations)


Over the next two months I’ll keep you informed about what I’m doing, the techniques and reasoning, and also my adventures into paper writing, along with updates on developments in tissue engineering and bone research.


Technique: Chemotaxis Revisited – the Boyden Chamber Assay

Chemotaxis, introduced in my post on January 26, is the process by which cells migrate along a gradient of growth factor. Researchers study chemotaxis for many different reasons. First, there’s the biological study of how these cells move. Then, there’s the study of why these cells move. In my post from September 28, I describe how random movement of cells in response to an even distribution of growth factor can be measured. However, the next step is to measure whether this response can be seen in a directional manner in response to a gradient.

The easiest method, perhaps, for studying chemotaxis is the Boyden Chamber assay. The concept of the Boyden Chamber assay is straightforward. Known as a transwell system, a smaller well is suspended over the main well of a tissue culture plate. However, the bottom of this insert is not solid plastic, put a permeable membrane. Cells are added to the top of this membrane, in solution, and their migration across the membrane to the other side in response to stimuli can be measured simply by counting the cells that have passed through. A great product for this are the FluoroblokTM inserts from BD Biosciences. Cells can be fluorescently labeled, and light from cells on the top side of the membrane won’t shine through to the bottom. The layout of a basic Boyden Chamber Assay is shown (blue dots are cells, green dots are stimuli):

boyden blog

The Boyden Chamber Assay may be simple, but requires a high level of finesse. The first main concern is the use of proper controls. Not only are negative and positive controls important, but also we need to measure the directionality of response to control for random movement. I.e., although, for example, growth factor might be added to the bottom well, and cells to the top, we need to show that the a gradient is actually maintained and that it is not simply a case of growth factor being evenly distributed on either side of the membrane (equilibrium) and cell movement increased in general.

Also, there are several steps in which human inaccuracy can lead to deviation between wells, ultimately resulting in lack of statistical significance because of too much variability. The assay is indeed incredibly sensitive to both the number of cells and the amount of stimulus, both of which tend to be very low, meaning that everything must be incredibly accurate. The first year of my PhD I conducted a gigantic Boyden Chamber assay, but got no significant results, just trends, because the error bars were too large.

The third part of the assay which requires delicacy is the counting of the cells. As I mentioned, the cells are fluorescently stained. However, in order to calculate fluorescence, because the desired cells are on the bottom of the membrane, in order to use spectrophotometry, you’d have to have a bottom plate reader. Our spectrophotometer could not be reprogrammed to do this (although in theory it had the ability). This meant that the membrane had to be cut off and mounted upright before being imaged. The cutting of the membrane is tricky, because it becomes easy to crease, which can lead to problems when trying to mount it flat (it’s very difficult to image a crinkled sample due to focus issues).

The protocol for Boyden Chamber Assays:

  1. Trypsin cells, centrifuge, resuspend in desired medium at desired density
    1. Make sure this cell suspension is evenly mixed
    2. Place cells in the top of the transwell
    3. Place desired medium in the bottom chamber
      1. Make sure the volume of the bottom corresponds to the top volume so no fluid pressure is exerted across the membrane (manufacturer’s recommendations)
      2. Incubate for the desired amount of time (good idea to perform a time-course with controls as a preliminary study)
      3. Gently aspirate medium on top and bottom, wash 3x with PBS
      4. Incubate 15 minutes in 4% PFA in PBS in order to fix the cells
      5. Wash 3x with PBS
      6. Incubate with 1:10000 DAPI in TBS with 0.1% Tween for 5 minutes at 37°C.
      7. Wash 3x with PBS
      8. Carefully cut out membranes and place, bottom-up on slides
        1. Mount with fluorescence-stabilizing hard-set mounting medium
        2. Pressing down to eliminate any air bubbles and make sure membrane is flat
        3. Image at 4x  magnification
        4. Use ImageJ to increase brightness/contrast to max, and convert to binary for quantification

Ten Hundred Words of Science Challenge

I was recently told about the amazing Ten Hundred Words of Science challenge. The challenge is really simple. Try to explain what your scientific project entails, using only the ten hundred (thousand) most commonly used words in the English language.

That sounds easy enough, until you try to use the text editor, and realize that out of the paragraph above there are many many words that would not be allowed: recently, challenge, scientific, project, entails, thousand, commonly, English, language… and SCIENCE. Science is not one of the top 1000 words in English!

This is my project description I ended up writing:

My college paper is about a fix for when a human breaks a leg or other body part. Many people in the world are working on such things. The best way to do this is to come up with a thing, that turns into part of the leg when it is put into the person, instead of trying to make a leg outside of the person and then put it in. To do this we need to focus on having good blood roads quickly, and then good leg can form. We also need to give our pretend leg everything it needs so that it can become real leg over time.

Thank goodness leg and blood was allowed! Neither bone, nor transplant, nor vessels, were allowed, which are very key words in my thesis. In the current thesis draft I mention the word “bone” over 500 times.Above, “college paper” translates into PhD thesis, leg = bone, and blood roads = blood vessels. Oh, and “thing” and “pretend leg” translate to “transplant”.

Are you up for the challenge? Scientific or not, I dare you to attempt to explain what you do in only ten hundred words! The tool can be found here:

The Up-Goer Five Text Editor

And the website to view other people’s scientific projects is here:

Ten Hundred Words of Science

Technique: The Simplest Way to Make a Collagen Scaffold

Collagen scaffolds are very common within the field of tissue engineering. They are varied in composition as different types of collagen can be used, and virtually anything added s a supplement. In this post I’m going to discuss the simplest method of creating a collagen scaffold, a method that I employed for two out of the three materials I made for my PhD project. The first material was a plain fibril bovine collagen type I scaffold, and the second was the same except with an added hydroxyapatite component. I actually made the hydroxyapatite myself, through the standard precipitation method (topic for a future technique post) in order to eliminate any byproducts present in commercially available product.

Basically, all materials in the final scaffold are mixed together to form a homogenous aqueous solution, which is then centrifuged and vacuumed in order to remove any air bubbles. The solution is then placed in a mold (we used PVC) and frozen at the desired temperature (in my experiments I used -20 degrees Celsius, though in past experiments I have also used -80 degrees Celsius freezers). The samples are then lyophilized (a.k.a. freeze-dried) for a few hours, pushed out of their molds, and then lyophilized some more. Note: beware of removing the samples too early; they dry from the outside, so if moisture is still present in the center they can collapse later. Finally, before use, samples are trimmed and cut as desired. I found that some trimming was necessary as sem micrographs showed walls of collagen on the outer edges, hindering cell and nutrient penetration.

Like any other polymer scaffold, the size and shape of the pores can be easily modified for different desired pore sizes etc. Given a set amount of collagen in an area, pore size is directly controllable by modifying the freezing temperature. At lower temperatures, ice crystals nucleate faster, and therefore the crystals formed will be more numerous and smaller. Since final pore size is directly related to aqueous crystal size, this means that the lower the freezing temperature, the smaller the pores. If the pores are more numerous, this means that the number of walls between pores also rises; but, since the total amount of collagen remains the same, this means each individual wall is thinner. One of the advantages to collagen and this fabrication method is the variation of the scaffold in its micro-porosity under 100 micrometers. It has actually been shown that variation on this scale is more conducive to biological activity, a long as the larger pore structures average the same as an otherwise similar scaffold.

Oxford – Bring on the Thesis

I’ve been back in Oxford for two weeks now.  Finally got over my jet lag. Maybe I shouldn’t have slept during the daytime the entire flight over from the States. But it took me longer than usual to get over the jetlag – in part due to a newfound love of afternoon coffee and partly due to what is possibly the best duvet in the world.


So how’s the thesis writing going, you ask? Well, I must say that the Radcliffe Camera in Oxford is an amazing place to work. Opened in 1749, it’s a place full of history, and the sound of students working. There are also many coffee places I go to when in need for more ambient noise, and caffeine.


But the writing, I hear you ask, how is the writing going? Well, the first thing to do is to figure out the data that you have. To this end I need to finish analyzing some data, which is what I’m currently focusing on. Also, for the writing part, I’m working on the words that go in the results section. The Results section is always trickier than it might appear because all the figures and relevant graphs need to appear there in order to present the data, but in terms of prose it’s important to provide context for the figures, while at the same time not including that which should instead belong in either the Methods or Discussion sections.  Often, so far, I’ve found myself writing and then thinking that a paragraph really belongs in another location in the thesis.


So here are the steps now:

1)     Finish analyzing data

2)     Write Results section

3)     Make figures and insert into Results section

4)     Reformat Methods and Intro to fit around the Results section

Stuck in a Hurricane – Don’t Know When I’ll Be Back Again!

Hurricane Sandy will go down in history for what it did to New York and New Jersey. But it will also be a story I will tell my grandchildren.


It’s stressful enough to move out of an apartment.  You have to sell or give away all of your furniture. You have to throw out all the detritus of a private life that you would keep using if you were still there – but which everyone else already seems to possess and doesn’t want more of: stainless steel pots anyone?  You have to clean the apartment thoroughly if you expect any of your deposit back. And you have to pack your entire life into 2 check-in bags (I paid extra and brought 3), one computer bag, and one “personal item”.


It’s also stressful to finish your PhD experiments (see my previous post: “Experiments – Done!!”). To simply circle a date on a calendar and say, “I’m going to finish this day and what I’ve got is what I’ve got. It is, what it is, and I’m going to write up and be done.” Particularly stressful when it gets to 3 days before you leave and you’re starting an 18 hour experiment and still have immunohistochemistry to work out, and to make sure you have all your files off all the various computers in the department.


And then: Hurricane Sandy. Oh, how I love you Hurricane Sandy. I’ve seen stronger wind and rain growing up in Albuquerque; and yet, for some reason, Washington D.C. completely shut down. I had just gotten back from my birthday/going away party on Sunday night, when they announced that the metro was going to be shut all day Monday. I therefore had a choice: I could either be stuck in the apartment with no way to get into work, or I could be stuck at work, with no way to get back to my apartment.  The choice was simple, I had an experiment that had to run overnight Monday to Tuesday. Either that experiment was going to be in my thesis or it was not. Also, my immunohistochemistry wasn’t working yet. I had to go into lab. So, I grabbed a set of pajamas, and hoofed it into lab.


I spent two nights in the lab.

The experiments got done.

The files were obtained.

The immunohistochemistry finally worked.

I ate the cheese, salami, and cracker platter leftover from the Friday birthday party from the lab fridge. I ate leftover pie.

Did I mention I spent my birthday in the lab with a hurricane outside?


Then I went home.

I packed everything I could.

I threw out everything I couldn’t.

I scrubbed every appliance and floor in the apartment.

I slept one night on an air mattress (a step up from the chairs in the lab).

Then I went to the airport and flew to England.

Experiments – Done!! a.k.a. I Never Want to Do Immunohistochemistry Again

Those who have been following me the past month on Twitter have witnessed my countdown to leaving the NIH. For those reading my blog, apologies for not posting, but time was limited and there was a lot of work to be done. So I now post a recap:


I needed to make sure I had all the information necessary to create a coherent project. To this end, the main focus of the last month was making sure I had all the staining I needed from my transplants in order to correlate angiogenesis and good bone formation, so that I could justify my thesis.


To this end, a preliminary survey of all of the “good transplants” – ones that I had transplanted since my return to the states in 2010 was made. This meant that 180 slides were observed in H&E. It was found that a number of transplants were in fact mammary glands or plain brown fat. The explanation for this is that sometimes transplants are resorbed – one of the reasons, after all, why I focused on collagen-based scaffolds is because they are resorbed over time. However, the ideal would be for the scaffold to be resorbed but simultaneously replaced by bone. Since this does not always happen, I was occasionally hunting around in the mouse for anything that resembled a transplant. After the preliminary survey, samples that were not transplants, or which were transplants but for reasons unknown were very fragmented and deemed “impossible to section” were abandoned. This left about 130 samples.


Some of the remaining samples were re-stained, and all were scanned using an Aperio slide scanner. This is an amazing machine, which is the saving grace for someone like me – a graduate student who vehemently despises microscopy. Slides are scanned at 40x resolution and then the entire image can be viewed on a computer, zoomed, and images captured of any section desired – without having to constantly refer back to the actual slide. Also very handy for me because it meant all I had to bring to Oxford was my computer. Welcome to the digital age of science.


Finally, all of the 130 slides to be analyzed were graded for bone formation and hematopoiesis (marrow formation) on a scale of 1-4 by 3 “blinded” researchers. I use blinded in quotation marks here, since the identities of what the samples were clearly written on the slide, but I have a habit of using acronyms for anything, so no one knew what they were. As one of the “blinded” researchers myself, I simply didn’t look at the slides based on identities, but by their numbered position in the box – and, as they were not arranged in the most logical order, my gradings can also be seen as unbiased.


Some of the slides appeared to be either pre-osteoid (i.e. in the process of forming bone) or cartilage, so I stained with toluidine blue. However, none of these samples turned out to be cartilage. For grading purposes these samples were deemed unsuccessful.


Then, I came to what I had been dreading for months. Immunohistochemistry.


Often shortened to IHC, immunohistochemistry is simply the staining of sections of tissue for specific expression of markers through the use of antibodies. Sounds simple, but not all antibodies are created equal and it took me about 10 different antibodies to find one that was specific for blood vessels and worked on my sample. Adding to the complexity of this process is the need for heat retrieval of some markers – for which every company and scientist will tell you a different protocol. Yet another variable exists in how to show where you have specifically selected with your antibody (i.e. choice of a secondary antibody). Since most antibodies aren’t themselves fluorescent or colorful (chromogenic) this means that you have to create another layer of antibody by selecting for the first one with your secondary which is either fluorescent or colorful. Thus, I think it’s easy to see that what, in theory, could be a very simple process, is indeed a very complex one.


But, I got it to work. I finally got really good immunohistochemistry of blood vessels on 16 representative samples from one experiment. The rest of the samples I am going to have to attempt to count the blood vessels by eye based off the H&E staining scans that I have.


And then there was the hurricane. But that’s for my next post!