Fluorescence and Failure

By Jack Sewell

Project aims

This post is about the malaria lab project that we've been working on full-time this semester. To avoid giving away too much information about what the lab is investigating, I won't mention the specific proteins under study. Essentially, the project has involved each of us studying the expression and localisation of a separate protein during the life cycle of Plasmodium, the malaria parasite. In addition, we aimed to study the morphology of the parasite when the protein was absent using a PTD construct. This removed expression of the protein from every stage in the life cycle, except for the blood stage where it is known to be essential. The point of studying the proteins was to identify a potential drug target for malaria treatment. The first thing we did was to transfect Plasmodium with a fusion gene, which coded for our protein of interest followed by the green fluorescence protein (GFP) tag. This would later allow us to see where our protein was in the cell. However, first we had to battle through the process of producing a Western blot, in order to show that the protein and GFP had been successfully fused. As Carrie has written about previously, this resulted in lots of failed attempts, with each one inevitably falling foul of a mistake that would leave us scratching our heads.

It's in there somewhere.

Life cycle imaging

Eventually though, we all had a band that we could confidently call our protein, and could then focus on the fluorescence imaging of the various Plasmodium life cycle stages. Because each attempt to transfect our PTDs into Plasmodium failed, this meant we could only study the protein localisation and not the phenotypes of the knockout. What I would consider to be the major difficulty in this part of the project was the time taken to culture the different life cycle stages. The process of growing Plasmodium in mouse blood takes around 3-4 days, after which time the mouse must be euthanised to prevent unnecessary suffering. The blood can then be used to image the parasite blood-stages (rings and trophozoites), or cultured for a further 24 hours to produce either schizonts or a mix of gametocytes, zygotes and ookinetes.

Plasmodium stages that can be imaged from mouse blood

Furthermore, studying the mosquito stages of the life cycle takes a total of 21 days following the collection of mouse blood. The mice must first be treated to eliminate all parasites except the gametocyte stages, then the mosquitoes are permitted to feed on them. At this point, the gametocytes in the mosquito gut become activated and leave the red blood cells. You then need to wait 10 days before some of the mosquitoes can be checked for oocysts lining the midgut, and mature oocysts can be imaged on day 14. After another week, on day 21, it should be possible to image sporozoites in the mosquito salivary glands. This completes the stages of the life cycle that can be studied, because the host's liver-stage parasite infections cannot be studied without complex dissection that would also render the mouse useless for study of the other stages. Unfortunately, the day 10 check of the mosquito infection for my protein showed no oocysts, meaning that it may not be possible to complete the oocyst and sporozoite imaging. This has therefore been a potential waste of 10 days, and would need the use of more mice and mosquitoes if a second attempt is to be made.

The Fluorescence Microscope

The actual process of taking parasite fluorescent images starts by mixing infected blood with DAPI, a DNA-specific stain that highlights the nucleus blue. In addition, the activated female, zygote and ookinete stages can also be tagged in red. This works by using a red fluorophore conjugated to an antibody the specifically binds to a protein expressed on the surface of these cell types. Once the fluorophores have been added, the blood can be searched for the desired stages under the fluorescence microscope. Any cells of interest must then be measured for their blue and red fluorescence, as well as the green fluorescence emitted by the GFP-tagged protein.

Activated female gametocyte with its membrane, nucleus and protein highlighted.

Getting a good image was always more difficult than you'd think, given that we had to make do with a single microscope shared between 4+ people, who often all had parasites to image at the exact same time. As well as this, we had to get the cell concentration exactly right, or they would be too sparse to find anything of interest or too dense to get the cell on its own. After many mice though, we were each able to create a complete collage to show the localisation of our protein at each of the ring-to-ookinete stages. All that remains now is to finish the study of the mosquito stages, and the project will be completed.

All in all there's been a lot of frustration and disappointment, but with some perseverance we got good results in the end. We also gained a lot from being able to adapt and learn from the things that went wrong, which in the end is a very important skill for a lab scientist to have. 

My (Our) project - Cell division in Plasmodium Berghei

What am I doing?

I am investigating the role of an Aurora-like protein kinase, PbARK2, in cell division in Plasmodium berghei, the rodent model for Plasmodium falciparum.

Why am I doing this?

Plasmodium undergoes cell division in a very atypical way – while most eukaryotes undergo open mitosis, in which the nuclear membrane breaks down before sister chromatids are pulled apart, Plasmodium cells maintain their nuclear envelope at all times. This is called closed mitosis. In addition to closed mitosis, their nuclear division is asynchronous – in a multinucleate cell, each nucleus divides independently of the ones around it. This suggests that there is no global control of mitosis within a cell, and a recent study highlighted that CDC20, a protein essential for mitotic regulation in all eukaryotes is in fact dispensable in the blood stages of parasite development.

Given that Plasmodium uses a drastically different system for cell division, many of the proteins involved may be very different to those in humans, making them good drug targets as off-target effects will be limited. By understanding the underlying biology of the system, we can better make predictions and investments into drug development.

How am I doing this?

ARK2 has already been shown to be essential for blood stage development in P. berghei. The plan was therefore as follows: place ARK2 under the control of an alternative promoter that is active in the blood stage to escape the lethal phenotype, allowing phenotypic analysis at later stages of development. This was to be achieved using a PTD construct (promoter trap using double homologous recombination), generated using molecular cloning techniques in the lab, before transfecting it into wild-type parasites.

Schematic representation of the PTD construct for placing Ark2 under the control of ama1

To provide additional insight into the function of ARK2, a previously generated parasite line expressing an ARK2-GFP tag was to be used to visualise the cellular distribution of the protein and determine expression patterns throughout the life cycle. This would be complimented by qPCR data produced for each life cycle stage.

What’s the story so far?

Last semester started pretty well – there are four of us MSci students in the lab doing very similar projects with different proteins so we felt like we weren’t all on our own. The initial molecular cloning went well, with the direct supervision of Declan, our resident lab-God, knower-of-all-things, provider-of-reagents and all-round-science-boss (the lab technician). With the 3’ fragment successfully inserted into the vector, we were left to do the 5’ section ourselves – this is where things started to go wrong. I managed to get lucky and only need to do one step twice, but Jack had to repeat his bacterial transformation four or five times. However, after a slightly shaky semester we had our constructs completed, sequenced to check accuracy, and ready for transfection. Two weeks before the end of the semester we were poised to begin our transfections and get on with the meat of the project. December 5^th came around, everything went well, nothing could have gone wrong, we felt confident and excited until a week later, when nothing survived the drug selection step, meaning our transfections hadn't worked.

Merry Christmas!

Four weeks later we were back and ready to get back into it. We knew the protocol better this time, we had read more about it, we wanted it more this time. And then it didn’t work. Neither did the third attempt, but the positive control worked just fine. Slightly disheartened, we decided to move on with getting our Western blots to show our protein was correct in the GFP-tagged lines. Carrie has a fantastic post telling you just how well that went… While a few weeks went by desperately hunting for meaning in the Rorschach tests we were creating, we started to learn how to use the fluorescence microscope to visualise our proteins. Jack has a detailed post talking about some of the trials and tribulations we experienced with the microscope, particularly the bit where more than four people need to use one microscope to do time sensitive investigations. Somewhere along the line we went from “just taking practice images” to being expected to produce a full collage of every life cycle stage – I’m still not sure exactly when that was, but I found that this was something I could confidently do. In addition to regular life cycle stages, we discovered that my protein was only transiently expressed during ookinete development, meaning we needed to image timepoints in ookinete development. This is where it gets exciting, because I got my first actual set of data usable in my write-up!

Collage of ookinete developmental stages showing punctate distribution of ARK2 dependent on stage.

Having finally got a decent set of images, I hoped that shortly after everything else would click into place – my Western would come out beautifully, my transfection would be successful and I’d have a clear picture of everything-ARK. However, the next transfection also failed and I still, 3 months and 14 Westerns later, don’t have a band for my protein. We’ve got a few plans for how to fix this but time is running out and it’s getting a bit worrying.

That being said, having spoken to my friends who are doing PhDs, and the post-docs in the lab, in research you have to get used to failure. The trick to being a good scientist is to persevere despite things not going right and work out how to fix it. Use the scientific method and think logically about what you can change, and maybe, just maybe, you’ll get the result you want.

If you’ve had a project come right down to the wire, or want to share any of your science woes please let me know in the comments so I can feel like it’s not just me that’s had this experience.

Taming your Western

During our projects we’ve learnt a multitude of molecular techniques, but the most elusive so far has been mastering the Western blot. Before Christmas we’d completed our promoter trap constructions, using molecular cloning to transfer our genes of interest (Kinesin13, Ark2, Rap9 and Smc3) under the control of the AMA1 promoter to restrict expression. We're currently waiting for our malaria transfections to show any sign of success, and for the meantime working on Western blotting and fluorescence imaging of our GFP-tagged proteins. On Ching has previously posted a summary of her project, so to avoid repeating ourselves I'd like to share one of the challenges of our project: the Western blot.  

The humble Western blot has many uses throughout biomedical research, including disease diagnosis, antibiotic efficacy studies, and the FIFA 2014 World Cup doping tests. For our purposes, we'd like to demonstrate that the fluorescence shown on our images indicates our protein of interest. Jack will shortly be writing a post about our experiences with microscopy, and I'd like to talk through the many mistakes we've made with Western blotting.

Image courtesy of David Taylor

In striving for the good blot, our initial results all too often resembled the bad and the ugly, and frankly things didn’t improve until we altered several steps of our lab protocol. We hope sharing our experiences troubleshooting Western Blotting will save some frustration for other students, and at the very least provide a snapshot of numerous ways it can go wrong. For any newcomers to Western Blotting - or anyone who just feels like punishing themselves -  we’d recommend this starter paper as a background to protein blotting.

1) Protein Handling – Lysis, Denaturation and Loading

The ugly: poor lysis, degradation
and overloading. 

Shown is one of our first gels, and to this day I still remember the look on our PIs face. The underlying nature of a Western relies on the quality of the protein that’ll be used. Our projects have involved isolating parasite proteins from infected mice blood, using CF11 columns for leukocyte depletion and then lysing cells to gain access to our proteins. During our first few westerns, we were having difficulty with protein degradation during cell lysis, which was compounded by chronic overloading of protein onto the gel. After several smeared and blotchy attempts, we rewrote the lab protocols for the lysis steps, lowering temperatures by lysing on ice and centrifuging at 4^oC, incorporating protease inhibitors and also restricting repeated freeze-thaw cycles.

Throughout this process we’ve also learnt to select detergents depending on our protein’s hypothesised cellular location – NP40 may be fine for cytoplasmic Kinesin 13, but for the nuclear Ark2 Sarkosyl would have been a better choice. A general rule is that the more compartments between you and your protein, the harsher the detergent needed. 

2) Supervising your Western

The overheated: a cautionary tale 

Borrowing the words of the Upturned Microscope: ”Your western blot hates you. Never leave it unsupervised”. After setting gels running, it’s usual to notice rising bubbles alongside fluctuating voltage and amp values. We’ve also found it’s a good idea to keep half an eye on our westerns, checking at least halfway through the cycle and not leaving it unsupervised for too long. During one of our earlier gels, there was an incident with overused electrophoresis buffer that resulted in the gel just melting within the chamber. A cursory glance over the tank will quickly reveal if everything is running as normal: checking the dye front is symmetrical, the tank feels cool to the touch and your protein hasn’t run off the gel edge will save a headache later.

The disturbed: temperature and
electrolyte fluctuations

If running for an extended period of time, it’s also best not to disturb the tank. For one of our experiments we were attempting to resolve a high molecular weight protein which refused to leave its well. We were running two gels, the first for 1hr 30min and the second for 2hrs, to reach better resolution. We made a rookie error: running the two in the same tank. Removing the first gel at the end of it's run meant turning off electrophoresis, removing the gel and refilling the electrolyte solution - all resulting in salt redistribution and temperature changes. The difference between the ladder on our two gels can clearly be seen – one resembles bands and the other blotchy smears.

3) Transfer

We’ve found that thoroughly removing any trapped air from the sandwich is an essential step once it’s been assembled onto the transfer system. Murphy ’s Law states that any bubbles occurring will appear just at the right height to obscure your protein band, and if you’re really unlucky these bubbles can span several lanes. Remembering to equilibrate sandwich components in transfer buffer for 2 – 10 minutes before assembling and then rolling out any air will help. This equilibration period also gives the gel time to cool which avoids transfer problems, and most buffers contain methanol to reduce air retention within your sandwich.

4) The Antibodies

Have you checked your primary antibody species matches your secondary? If you’ve ever had a western with no signal, this could be the culprit. Whilst this may seem like an obtuse mistake, it’s particularly easy to make when using Western Reagent kits. Our lab uses WesternBreeze by ThermoFisherScientific, which contains two delightfully similar bottles of mouse and rabbit antibodies, and once produced a confusingly blank blot. It is also possible to strip membranes and begin again from the beginning, but as this isn't something we've tried yet we'd love to hear our viewers comments on the process.

5) Background

The obscured: high background.

After our initial problems with high background on our westerns, we increased the frequency of our washing steps.If there’s too much background on the western, we’veincreasing washing steps will help to resolve the issue. However, if the background is disguising smearing or low signal to background levels, you may wish to examine how you’re handling the membrane and check your antibody dilutions. If assembling your sandwich involves the membrane sliding over the gel repeatedly (or when developing has frequent repositioning of your film within the cassette) it’s also possible to create smears and blotches.

If you’re having similar troubles with your western, we’d love to hear about it! You can share your worst westerns in the comments. For any students looking for a more comprehensive troubleshooting gallery, we’ve found the SDS-PAGE Hall of Shame curated by Rice University to be particularly inspiring. Join us again next fortnight when Jack shares his experiences with fluorescence microscopy in our next post: “Fluorescence and Failure”. 

References

1.  Baume N, Jan N, Emery C, et al (2015) 
   Antidoping programme and biological monitoring before and during the 2014 FIFA World Cup Brazil. British Journal of Sports Medicine. 49: 614-622
   
   
2. Kurien, B. T., & Hal Scofield, R. (2015). Western blotting. Methods. 38(4): 283–293 https://doi.org/10.1007/978-1-4939-2694-7
   

   
   
3.  Johnson, M. (2013). Detergents: Triton X-100, Tween-20, and More. Materials and Methods, [online] 3 (163). Available at: www.labome.com/method/Detergents-Triton-X-100-Tween-20-and-More.html.