Many of the tips for doing RNAi screens can be gleaned from the 'Overview of the screening process' page and also the FAQ pages. Below are some specific points to consider when doing each type of screen.
Tips not specific to each screen are listed first then specificities about each one are in the following paragraphs. Simple things to remember are make sure reagents and enzyme reaction temperatures are constant, the reagents are made in batches and each batch is validated before use. We generally buy a single batch of serum and use that throughout the screen and all the development and validation processes.
The cells can change from batch to batch, so make sure you have a healthy 3 or more flasks growing before you begin on the screens. You'll need the cells to be at the same passage and you should have frozen cells down before you start with each batch. You'll need at least 3 T75 flasks before you being screening anyway and these are usually amplified to 6-9 T75 flasks before seeding.
Luciferase screens, even though only giving a read-out of a single variable can provide enough data to dissect pathways or processes. However, luciferase screens can provide misleading results and consideration must be taken when comparing 'hits'.
For example if you have not used a general read-out for cell viability, ie, renilla being driven by a constitutive or' housekeeping' promoter, then care must be taken as your 'hit' maybe causing changes to cell viability, growth etc. Generally we suggest for screeners to use a dual luciferase reagent, ie your reporter driving firefly and then a cell viability reporter of actin driving renilla. Once both sets of data have been acquired from the screen, we also plot the ratios for firefly and renilla Z-scores, as we do this for each one individually. This gives us some insight into what the dsRNAs are doing, as ratios can sometimes be misleading.
GFP reporter assays can be recorded in more than one way. For example the fluorometry aspect of our plate reader can be used, the high content microscope can be used to determine general fluorescence, more commonly called in cell Western, and of course the microscope can be used in a high content manner to determine location as well as abundance of the tag. The user needs to establish which method would be appropriate for their screen.
Remember go for a fast and simple method over higher quality data if all you need to know is if a cell is green or not. The other thing you need to consider is that a whole genome replicate can be done in a single day on the plate reader or within 24 hours if in cell Westerns are being used. However, using a 40x objective with 8 acquired fields will slow the process and give you a huge amount of imaging data that you will have to analyse and store. The analysis is usually the headache part of the process, so you can see the benefits of the fast methods. Also you have to bear in mind that screens will identify proteins that are involved in GFP production or turn-over.
High content screens (HCS) provide the most information from of all the screens available. However, care should be taken when considering fluorophore combinations and secondary stains are always recommended as a point of reference for each screen. Most users consider that they will be capable of doing live cell imaging when they first come to the facility, but once one considers acquiring 20000 wells with at least 4 fields acquired per well, the user's mind is changed towards fixing the cells and storing them before acquisition.
Generally most of our users use a Hoechst nuclear DNA stain in combination with the GFP, but there are many other possibilities. The most time that users spend on the HCS assay is usually at the development stage of the algorithm writing, either on predefined canned algorithms or by writing journals. Some help will be given at each of these stages. We suggest to booking times with the ImageXpress expert in advance and having a good idea of the question you wish to answer.
Remember we have many users using the SRSF and so a short wait may be expected, as we're busy.