I block randomise* the injections when multiple groups/treatments are present (e.g. Ctrl and KO or isotype control/depleting antibody). The goal is for each cage to have all of the cell lines/treatments.
Starting with Cage 1, I inject different cell lines assigned to each mouse within that cage, and then move to the next cage. This way, any bias with cells sitting on ice for extended periods of time is applied equally across all the cell lines used.
*Block randomisation prevents the situation where all mice within one cage receive the same cell line/treatment. The way I do it: A random number is assigned to each mouse in a cage (block), and the treatments assigned from lowest to highest number within each cage (e.g Ctrl1 to lowest number, Ctrl2 for the second lowest, etc.). Random number is generated using the ‘rand’ function in Excel, but copied and pasted as ‘values’ so they would not change every time.