Saturday, April 27, 2013

Qiagen Finder 45-pathway Cignal array: the good, the bad and the ugly

I am still not sure whether this is going to be a good review or a bad review. It may shake out one way or another as I write it.

We do a lot of signaling as we work on a intracellular signaling scaffold which functions in a lot of different pathways, so I always pay attention to new email from SABiosciences/Qiagen on their new PCR and luciferase arrays. Upon reading about it the Cignal Finder 45-pathway array seemed to provide an awesome way to apply very sensitive luciferase assays to test 45 pathways in a single shot.
It is a 96 well plate with 45 reporters in duplicates and 3 positive and negative controls that you can use to test you drug, siRNA, over expression plasmid on your cell line of choice. The protocol is simple and straightforward: the DNA is in the wells, you add the cells and transfection reagent and the next day you test for luciferase and renilla activity. At $1,600 for a 10 plate pack, plus dual-luciferase assay reagents, it can run around $200-300 per plate, which in triplicates for 2 conditions can be almost $2,000 per experiment. But screening 45-pathways at once could be totally worth it and SABiosciences always run some kind of deal on the plates (and you can use a fraction of the luciferase reagent recommended in the Promega protocol).

We know our protein of interest would really strongly activate NF-kappaB and we have done it with a different reporter, so that was our positive control. The first plate was amazing! There were very clear difference between reporters, the duplicates looked very consistent and a whole bunch of interesting pathways lit up in addition to NF-kappaB in our experimental condition. The dynamic range of response was incredible with some reporters with intensity in the 2,000,000 intensity units and some in the 2,000. And so we embarked in a huge study with over expression and siRNA...and ordered another 10-pack. Qiagen provided us with a handy Excel file for analysis where we just needed to plug the results of luciferase and renilla for each plate and everything would be calculated and plotted for us.

I was intrigued by some of the results and started paying a lot of attention to the intensity measurements, and that is when things started getting ugly. We had noticed that some of the duplicates were asymmetric in intensity when a low intensity reporter was next to a very high intensity one. I had had similar problems in the past with white plates in our luminometer, where we noticed spillover of fluorescence in neighboring wells. Some people say it's the fluorescence passing trough the white walls, others that the detector is far from the plate and the fluorescence spills over from above like a halo. Also instead of looking at the results in the Qiagen spreadsheet where they are in linear format, I analyzed everything in plate format, where I could see which well was above, below, left and right of any particular reporter. That is when half of our hits became uninterpretable because they could simply be due to spillover from the strong hit next door. Qiagen recommended to half the amount of lysate, but that didn't help much as the results were basically identical. So we were stuck with the option of spending $500 per troubleshooting attempt or find some other solution....

They said it must be our luminometer, which is a possibility, but that's the machine we have in the department and their product should work even for older machines. They gave us 2 plates for free for testing, which was nice, but still leaves for $500 per troubleshooting for each one of the other variables (machine, reagents, lysate amount). Right now I am faced with a conundrum: I have some data I believe from the array and some data I do not believe, which I think is due to spillover. I cannot publish the whole array because I would have to explain it is a problematic product and I have requested for a sample of the 6-8 reporters that are iffy, but I was told I had to buy them individually, which I am not inclined to do at this time, since I do not want to waste my grant money to troubleshoot their product for them.

At the same time, their reporters work beautifully and the data we have is very robust, though for us it was more a 30-pathway array than a 45-pathway because of the issues.....I am open to suggestions from other people who may have tried the product and who may have other ideas. Also, I am not entirely satisfied with the way they recommend to analyze the reporters and I'd love input on that also.

9 comments:

  1. Interesting observations!
    I just did the assay but disappointed due to poor transfection with the cell line that I used. The readings are really low for both firefly luciferase and renilla. In this context, can you share your inputs on raw data in terms of RLU. Which cell lines you have used and what is the cell seeding density. Can you specify in what ranges the values are? For how long the cells were transfected and any change in cell viability with transfecting agent. What was the transfection efficiency in terms of percentage? Also, in case you have treated transfected cells with any treatment, in what media, treatment was done.
    Sorry, many questions. Hope to receive response

    ReplyDelete
  2. Hi, I'll tell you what I can.
    - We are using HEK cells and we had used all our reagents (plasmids, siRNA) before so we know they transfect at very high efficiency >70%
    - We follow the protocol exactly using the number of cells recommended and the amounts of DNA/siRNA recommended by the Cignal protocol.
    - We transfect with Lipofectamine 2000 from Invitrogen as recommended from the manufacturer and again following the Cignal protocol exactly as far as media (I think they require a certain % of serum, which is not part of our normal protocol)
    - For the overexpression we let the cells go for 24 hours because our overexpression has a very strong effect. For the siRNA we wait 48 hours to make sure the knockdown is complete. Again all these conditions were worked out before as we have been using these reagents for a long time.
    - The values for the overexpression range between 2,000 and 2,000,000 RLU within each experiment with renilla values between 2,000,000 and 4,000,000 depending on the experiment. In each experiment, the renilla values are very consistent.
    - Because of the bleeding between wells, we actually only take 10ul of the lysate and add it to 30ul of luciferase reagent and then 30ul of Stop and Glo. We use the Promega Dual Luciferase kit...
    - We are attempting to do some drug treatments right now, but are still working our conditions before we use the Cignal. We were just planning to add them to the regular medium.
    Hope this helps!

    ReplyDelete
  3. Thank you so much for all the details.

    I have used 80000cells/well and used Attractene transfection reagent as recommended by the company. I standardized in my lab with the GFP plasmid construct and I could achieve 40 - 50% efficiency with OVACAR-3. I did not get same efficiency when I used their assay plate. I am trying to troubleshoot the problems. As per your experience do you think, with approx 50% transfection efficiency with their assay plate, will the assay work?

    Secondly, the reason I asked about raw data values was low RLU I got. I used Dual Glo luciferase assay kit which do not have an extra lysis step as you have when using Dual luciferase reporter assay kit. As per the online, they mentioned that values will be low unlike with dual reporter kit. As you have used dual reporter kit, you mentioned that you transferred 10ul of lysate to fresh assay plate. In this connection, how much lysis buffer you have added per well and how long you incubated the cells with lysis buffer? What is the time gap between addition of first substrate and reading. Which instrument you have used for measurement and did you use injectors for addition of reagents? If possible, can you specify the settings especially integrations in milliseconds and pattern of reading.

    With regard to use of serum during treatment, company recommends 0.5% FBS in optiMEM for drug treatment. With my cells, at that low serum conc. cells die with the concentration of the drug that I intend to test. I may need to test at lower conc. to avoid the cell death but hopefully should be able to see cell signaling changes. However, untreated cells can survive low FBS and are healthy atleast for 24-48hrs. In case you have done your expt with drug treatment, please share your experience. As per the brochure, high serum induces many signaling pathways, cross talk and high background, hence false results. Please share your experience when treated in normal media.

    Once again thank you so much for responding thoroughly to my concerns. Please respond at your convenience,
    Regards

    ReplyDelete
  4. I think 50% efficiency would work fine with their assays, since the intensities we get are really high. I had never seen anything in the million RLUs before using these reporters and most of the issues we had with this kit were due to the very high intensity. We add 25ul of PLB to each well and have to use only 10 because the luciferase activity is too much.
    I actually freeze the plate in PLB at -80 to get better lysis. I found that freezing the cells in PLB increases lysis a lot and it's one of the recommend steps on the Dual Luciferase protocol. We don't have a luminometer in the lab and that also helps if we are running experiments in the weekend and need to store the lysates until Monday.
    When we do the reading we bring all the reagents to the machine (which is a very old Wallac Victor plate reader) and add the LAR with a multichannel immediately before reading. Reading conditions are 1 s shaking and 1 s reading.
    No drug treatment yet, because we have yet to optimize conditions....

    ReplyDelete
  5. Thank you so much!

    ReplyDelete
  6. Try using black plates, it may help cut down on the crosstalk between cells.

    ReplyDelete
    Replies
    1. ... oops, I meant wells not cells.

      Delete
  7. I don't know how they put the plasmid in the plate but I guess you can put few ul water in those wells you want to repeat the experiment (should be a new plate never been used) and then use 1-2ul to transfect e.coli competent cell. Pick few clones and you will get either reporter plasmid or renilla. Then you can do your individual assay without considering the array price.

    ReplyDelete
  8. We thought about that...One thing I'm not sure is whether the plasmids can be replicated in coli. If Qiagen was really sneaky they would just make then replication incapable. We never needed one of the clones so much to waste a plate to try, but we may at some point and I would love to know if someone was ever successful in retrieving one of these clones.

    ReplyDelete