It really depends on how much the molecule you are interested to is expressed on your cells. Generally, using an homogeneous population, 200,000-300,000 cells are sufficient. However, if it is poorly expressed, you might need to increase the number; on the other side, if it is widely expressed, you can even reduce to 100,000 (in any case, I would recommend to not go below 50,000 cells)
but, it also depends on cell size you are using like, in case of very small cells, number has to greater and in case of larger cells, number can be reduced.
And also depends on the flow rate of instrument (low or medium or high).
Genarally 0.5-1 * 106 cells /mL concentration has to be maintained.
But, you can decrease or increase this by your experience after 1 or 2 analysis.
I would suggest to analyse 1 or 2 sample just to standardized the cell number and the function you wish to assess first and then you will get the idea about it.
you can also have the reference of following publication but no need to necessarily follow as such:
I mean try with smaller number and see the results.