The GFP Project: One Step Backward, Two Steps Forward

Oops.  After a week off for spring break, I returned to the GFP project and realized we had to do another transformation, because last time we had accidentally used up all our stocks of transformed bacteria.

So we did another transformation, and we also plated some of the leftover bacteria we had on a plate from the first transformation.  Hopefully we can use one of these sources to grow an overnight culture and extract GFP from.

We also had some interesting visitors!  We had Jonathan, an anthropologist in the science and technology field, who knew Mac Cowell from way back, and that we had some DIYBIO screens hidden away at BOSSLAB!  Time to make some T-Shirts!

We also had Kris Constable and Megan who are starting BioSpace, a which is exactly what it sounds like (a bio-hackerspace) up in Canada.  It was cool to meet international DIYBIO folks!

The other good news is that the stickers are finally in the envelopes and addressed!  They should go out this week.

GFP Project Week 2: Extract(ifying) GFP

GFP! GFP has an excitation peak in the blue light spectrum. Unfortunately, the blue LED is comparatively bright. Gotta find a green pass filter!

This week was Week 2 of the GFP project!  The extraction of the Green Fluorescent Protien (GFP) from the transformed host cells, using Hydrophobic Interaction Chromatography (HIC).  We used a kit from Carolina because I have not ever done this before, and I didn’t really know what to buy to do this completely DIY style.  Nonetheless, it was a success and a clear demonstration of the central dogma of biology (DNA–>RNA–>Protein), and a good way to add some value to our transformation.  Now we have not only altered the bacteria, but made something pseudo-useful, but definitely cool, with our alteration.

At a high level, the way HIC works is by selectively binding the protein to a “bead” (more like a fine powder) resin.  Once the cell has been lysed (the cell walls are broken up), the GFP-bearing lysate is mixed with a high salt binding solution.  This gently and reversibly denatures the GFP so that the hydrophobic (water fearing) inside is exposed to the solution.  This water fearing end acts like oil in water; it is attracted to other hydrophobic bits, so the bits all clump together.  Since the beads have hydrophobic sites, they bind to lots and lots of proteins that are hydrophobic.  At this stage, the liquid that contains non-binding proteins is removed from the matrix, so we remove it.  Then a medium-salt wash buffer is added, which removes some of the hydrophobic proteins that re-fold in lower salt, and are therefore less “sticky” than the gfp.  The extra liquid holding these proteins is then removed, and this is when the final low-salt TE buffer is added.  The low salt buffer is also known as the elution buffer, because it elutes (washes out with a solvent) the GFP from the matrix by allowing the GFP to re-nature (fold back into its original configuration), and hide its hydrophobic sites.  Then the liquid contains GFP!  The process is described in more detail below.

Overnight cultures fluorescing! You want to use only a few mls for these, to increase the amount of oxygen-per-volume in each culture.

The first step is to grow overnight cultures of the transformants.  To do this, aseptically pick off a colony from a plate of transformants that are (ideally) only a few days old, and use it to inoculate an LB+ampicillin liquid broth culture .  The antibiotic keeps the selection pressure on the transformed cells, and doing it a few days after the transformation helps ensure that you do not pick satellite colonies or have e. coli that have dropped the plasmid.  These are literally grown overnight, for about 12-16 hours in a shaker.  It is important that the cells are exposed to oxygen, because it is critical to the development of the GFP.  The cells should fluoresce brightly when blue light is shone on them.  In the picture above, the middle and right tube are fluorescing green, but the rightmost tube is not.

HIC resin and equilibration buffer

The next thing to do is to equilibrate the HIC resin with the equilibration buffer.  This is just some salt so that when you add your bound GFP (cell lysate in high-salt solution), the water that the hydrophobic beads are suspended in does not bring the salinity down.  To equilibrate, we added 300ul of HIC resin to 1mL of equilibration buffer, and centrifuged, removing the supernatant.  The supernatant is the liquid remaining on top of the pellet (the mass of solids at the bottom of the microcentrifuge tube) once it has been centrifuged.  Definitely remove the supernatant with a pipettor, because it is runny, so you need to keep the tube upright.

You can see the large green pellet at the bottom of the tube, after it has been centrifuged. The stuff on top is the supernatant.

The next steps you do twice.  Add 1mL of overnight culture to a microcentrifuge tube, spin down for ~2 min at max RPM, being sure to balance the centrifuge.  Then pour off the supernatant.  Repeat for another 1mL

The big green thing at the bottom is a huge pellet of cells!  Notice that the supernatant has been removed, and the cells are kind of stuck to one side of the tube.  That is the side of the tube the was facing outwards.

At this point, we have a great big lump of green cells.  Now we want to harvest the GFP, which is inside the cell membrane in the cytoplasm, with all the other cellular proteins.  To get it out, we add a lysis buffer.  I suspect that in this kit, it is SDS because it got bubbly/sudsy when we pipetted it.

Washboard technique!

The secret to lysing pelleted cells quickly is using the washboard technique.  It works better (for me) than vortexing, and way better than flicking or pipetting up and down.  What you do is grab the top of the eppendorf, and run it along the top of a tube rack, like you would run a stick against a washboard for musical effect.  This subjects the tube to a lot of sudden shocks, and really blasts the pellet off the side of the tube.  When this is done, the pellet and lysis buffer should not be clear; it should have a milky green consistency.

Lysate on ice. you can see the pale green lysate in the tubes.

Another factor in support of the lysis buffer being SDS and not lysosome enzymes is that the ice would slow the enzymatic reactions, while icing cells without any kind or cryoprotectant (like glycerol) is reputed to break up the cell membrane by forming ice crystals, which would further our goals in this stage.

Pellet again! This time the supernatant lysate is green, and contains the GFP we are trying to extract,

The lysate is taken off ice and centrifuged after 15 minutes, and 250 ul of supernatant is pipetted into a clean eppendorf.  In this tube, we add the binding buffer, which gently denatures the GFP.  It is called binding buffer because it allows the GFP to bind to the HIC resin, not because it binds to the GFP.  It gets cloudy in this step, but still retains some greenness.

supernatant lysate in a clean tube.  Note bubbles from (probably) SDS

After the binding buffer has been added

The next step is to add the GFP bearing lysate in the binding buffer to the HIC resin.  You will want to mix this thoroughly to ensure that the GFP binds to the resin.

GFP in binding buffer sitting on the HIC resin

Once the beads and the lysate are mixed, you put the tubes into the centrifuge and pellet out the HIC resin, which is now bound to the GFP and many other proteins.  As you can see in the picture below, the supernatant is not fluorescent, so it does not contain any significant amount of GFP.

The resin glows now, but the supernatant does not!

Now the supernatant is pipetted off, and the wash buffer is added.  The wash buffer is a less salty buffer than the binding buffer, so some of the proteins that are less sensitive to salt change configuration and “hide” their hydrophobic sites from the HIC resin, making them fall back into solution.  This is centrifuged and the supernatant removed and discarded much like the binding buffer was.

The final step is to elute (to wash out with a solvent) the GFP from the resin with the TE (low salt) buffer.  This makes the GFP fall off of the resin and into the solution.  Then we pipetted the supernatant (now containing GFP) into a clean eppendorf.  Here are some shots of the results:

left tube is the extract, right tube is the crude cell lysate

with blue light from my arduino!

Now next weekend the GFP project will be on hiatus, because I will be away for spring break.  However, the sunday after that it will be back in full effect, probably with a plasmid DNA extraction adventure!




The GFP Project Week 1: A Lesson in Patience and Ingenuity

Arduino: the alternative lighting platform for molecular biology

It had been two days since the transformation yesterday, and I had not seen any growth on my plates.  I was becoming concerned.  It can be tricky to orchestrate the teaching and the actual transformation procedure, and I have had less than optimal results with this media, strain, and plasmid before.  I was not sure what I would do on Sunday if there were no transformants!  It would be a disaster, and very demoralizing.

I had a meeting to go to that would result on me being on the red line, so on the way back I stopped by and took a look at the cultures with my blue LED.  Nothing was growing, except for what appeared to be some e. coli growing on the ampicillin plate, which I took to be an ill omen.  I left my blue LED and some batteries there, because I figured that if there were transformants, they would be more likely to grow on the multitude of plates at sprout instead of the two that I had.

I wasn’t sure what to do there, so I grabbed a few items for doing another transformation and headed home, convinced that I would have to do another transformation.  Upon my return home, I obsessively checked again.

Perfect little colonies!

When I returned from dinner, I decided to take one last look.  I was greeted with two plates full of transformants, in perfect little green colonies.  Having left my sole blue LED at sprout, I initially thought that I would have no way of testing the fluorescence, until I spied my arduino nano on my desk.  I remembered that it had a blue LED on it, so I plugged it into my laptop and used it to light the plates, which indeed fluoresced.


Lessons learned:  Be patient.  Synch your life to the organism you are studying, not the other way around.  Also arduinos are good for many things.

cell growth phase chart, found via the google

The final question you may have is “why did they take so long to grow?”.  The answer there is something I should have recognized!  I totally forgot that there is often a “lag phase” of growth in bacteria.  This is the phase where bacteria are generating the needed metabolites and substrates to adjust to their surroundings.  The colonies are also growing (in this case) from a single bacteria!  So it makes sense that there was a pretty big lag from when they were plated, until  I could see them.

The GFP Project Week 1: Hands-on Transformation!

The shirt says it all.

The GFP project rolls forward!  Powered (funded) by the people doing the work, I managed to order all the materials that we would need for a transformation this weekend (by that, I mean today).  There was an awesome turnout; some new people came out, a lot of supporters came out, and a good time was had by all.

From left to right: LB broth, LB agar, 50 mM CaCl solution, alcohol burner, tape, streaked plate, arduino, pipettes in a fancy box, ziplock-o-wires

To start out, we had to sterilize some media and CaCl2 solution.  We made 200 ml of LB agar, 80 ml of LB broth, and 500 ml of 50mM CaCl2 salt solution.  The LB agar was used to make LB+ ampicillin plates.  These were prepared by adding 2 ml of 1% ampicillin to the still-molten agar once it had cooled to “warm enough to hold”, and then pouring it into sterile petri dishes.

The Calcium Chloride solution and the LB broth were needed for the transformation.  You only need 250 ul of each per-transformation.  We prepared the LB broth in bulk because it is useful to have on hand, and we prepared way to much CaCl2 because our scale was not sensitive enough to measure out a smaller quantity that would make a 50mM solution.  We could have made a 1M solution and diluted it, but it seemed simpler to just make 500ml.

Totally legit ice-bucket of science

The other necessities for the transformation that had to be prepared were the ice bath, and the 42 C water bath.  The ice bath was created using the insulated shipping crate that the plasmid and ampicillin came in by throwing some ice in the larger (bottom) container.  The water bath I have successfully made before by nuking (microwaving) the water and then  guesstimating that it was hot enough (more than body temperature, but cool enough to hold, between 37-50C).  Today I used what seemed to be a much more reliable method, which was a hot plate and a submersible temperature probe.

(Hot) Water bath setup. Sensor in falcon tube taped to beaker, arduino connected to sensor and computer

The temperature probe was improvised with my arduino mini, a DS1820 digital temperature sensor, some long wires, and a 15 ml falcon tube.  I submerged the falcon tube in the water bath, and stuck the temperature sensor (which had been soldered to the long wires) into the tube, allowing it to (hopefully) measure the temperature of the bath.  This was reporting the temperature back to my computer with some script I got to work a while ago.  A better way to set this up physiclly would be to cram the sensor in to a thin-wall reaction tube (PCR tube) and fill it with milliQ water (non conductive, very DI RO water), stick the sensor in the water and waterproof all that with sugru or epoxy.  This way there would be no air gap, and the sensor would be safe in DI water.

The game plan for the transformation in flowchart doodle format

During all of that setup we were also preparing the bacteria for heat shock, teaching people how to pipette and streak bacteria, and talking about biology in general.  Our transformation protocol, officially was this:

Gather all the things! Everything didn’t fit in this picture, but this is a lot of stuff!

1.  Gather reagents (LB broth, 50mM CaCl2, LB+Amp plates, e. coli plate, eppendorf tubes), prepare ice bath and water bath, and gather your tools (flame, inoculation loop, pipettes and tips)

Use your sweet pipetting skills to add salt solution to your eppendorf. keep on ice.

2.  add 250 ul of 50 mM CaCl2 to an eppendorf tube, and chill on ice

Scrape e. coli off your source plate. Be careful not to pick up any agar!

3.  scrape off some e. coli (enough to see on the loop) with a flame sterilized loop, and swish the loop around in the iced CaCl2 in the eppendorf.  Make sure they fall off into the CaCl2.

flick! flick! done!

4.  flick the tube until the bacteria are no longer clumped together.  The solution should be cloudy now.  Chill tube on ice for 1-5 min.

Adding tiny amounts of DNA means tiny pipette tips!

5. Add 20ul of plasmid DNA at .005ug/ul to the eppendorf.  This is .1ug of DNA.  Return tube to ice, incubate for 5 min

Heat shock!

6. heat shock at 42C for 90 seconds

7. add 250ul of LB broth

Spread the bacteria around with the sterile loop

8. plate 100ul on LB+Amp plates.  Pipette 100ul on to the plate, then flame loop and cool in agar somewhere that you did not pipette onto.  Use loop to spread pipetted transformant mixture.

With all of that done, we cleaned up and headed off to wherever.  Some people had to leave early, but most everyone seemed like they would be back!  There should be results on if this worked in just a few days.

Transformation 3.0: Today I learned about Ampicillin

Today I learned that you should immediately plate bacteria that have ampicillin, and that it is SOP to do so in our lab here at Olin.  It doesn’t matter that you add them immediately, because they have time to inactivate the amp before they divide, which is when the amp kills them.  Maybe this explains all the satellite colonies I saw before!  I was waiting a good 15-30 minutes, which allows the bacteria to express beta-lactamase, which breaks down ampicillin.  This can destroy nearby ampicillin on the surface of plates, destroying any antibiotic effect, creating satellite colonies.  Oops.  Time to revisit my transformation protocol.