5. Library Packaging and Transduction of Target Cells

To package your lentiviral sgRNA, shRNA, or barcode library as VSV-G pseudotyped lentiviral particles, follow the general lentiviral packaging protocol. However, when packaging a library for screening, it is critical to package sufficient amount of the library to ensure representation of all the library constructs. The chart below indicates the number of 15-cm plates to use depending on the library complexity (i.e. number of constructs in the library). We do not recommend scaling down the lentiviral packaging protocol due to risk of compromising the representation of the library.

Pooled library screens require quantification of changes in the fraction of cells bearing each sgRNA/shRNA sequence in selected vs. control cells (or starting library). A “hit” is identified when selected cells have significantly more or fewer cells bearing a particular sgRNA/shRNA sequence. Whether one is looking at enrichment of sgRNA/shRNA sequences in the selected cell population vs. control (positive selection) or depletion of sgRNA/shRNA sequenced in selected cell population vs. control (negative selection), it is critical that the screens begin with a sufficient number of cells expressing each sgRNA/shRNA to ensure that measured changes in the fraction of cells bearing any given sgRNA/shRNA sequence are statistically significant. This means that if there are very low numbers of cells bearing specific sgRNA/shRNA sequences at the start of the screen, random changes in a drifting population may be difficult to differentiate from significant trends.

Simply put, a loss of 2 cells is a 20% change if there are only 10 initially vs. 2% if there are 100. For this reason, at least a few hundred cells on average need to be infected with each sgRNA to initiate a good screening. This is demonstrated in the figure below which shows the effect on reproducibility of different infection and splitting ratios. A smaller population of just 50 cells per shRNA (third bar) leads to significantly more variation than starting with a population of 200 cells on average for each shRNA (first bar). For this reason, for example, at least 25 million cells (500 × library complexity) are needed to start a screen with libraries of 50,000 sgRNAs.

Reproducibility is adversely affected if starting infected cell populations have fewer than 200 cells per shRNA (third bar). Allowing representation to drop below 1000X representation at any time is even more detrimental to reproducibility, as shown in the middle bar, when splitting causes the representation to drop to 200 cells per shRNA.

For experiments with pooled lentiviral libraries of shRNA, sgRNA, or barcodes, it is important to have more cells than infecting viral particles. For sgRNA/shRNA screens, typically, 2-3 times more cells than infecting viral particles is needed (i.e., a multiplicity of infection [MOI] of 0.3-0.5) to ensure that the majority of cells are only infected with one sgRNA/shRNA-carrying virus, so you need to have 2-3 times more cells than the number targeted for infection. For barcode labeling of cells, lower MOIs may be desirable to ensure almost all transduced cells only pick up a single barcode.

The lower the MOI, the more cells you need to start the screen so it is tempting to use a high MOI. However, you should consider that a higher MOI produces a higher percentage of infected cells bearing two or more different sgRNA/shRNA constructs.

For most loss-of-function screens, we recommend optimizing conditions and performing genetic screen transductions at no more than 0.5 MOI (a 40% transduction efficiency), which balances these two considerations. Please note that to accurately calculate the MOI, it is critical to determine the library titer directly in your target cells prior to beginning your experiment. Once conditions are established to achieve ~40% transduction efficiency in the titering assay, scale up all conditions proportionately to accommodate the larger number of transduced cells needed for the genetic screen.

To ensure a comprehensive screen, it is not simply sufficient to start with the right amount of cells. During the screening process, incorrect propagation of the cells can completely undercut the representation set up at the initiation of the screen. This is especially true for a negative selection screen, such as a viability screen, where one is interested in identifying sgRNAs or shRNAs that kill or inhibit proliferation of cells, and, therefore, drop out of the population. It is critical to maintain the full library representation that was initially used at the start of the screen.

If a portion of propagated cells is removed during propagation (e.g., cells are split), the representation of the library can be skewed in the sample, which introduces significant random noise. This effect is readily seen in the middle bar in the above figure where the effect of starting with sufficient cells (i.e., 200-fold library complexity) is completely undercut by splitting cells during propagation so that the final count of cells after 10 days is the same as the initial number of transduced cells (i.e., 200-fold library complexity). The correlation between triplicates falls dramatically when the cells are split to this degree.

Once you have calculated the titer of the packaged sgRNA or shRNA library and you know the MOI and starting number of cells you need to maintain representation, transduce your target cells (Cas9- or dCas9-positive cells if performing a CRISPR screen) using the appropriate number and size of plates and the correct amount of virus.