I'm in the process of designing a simple flow panel for human PBMCs and am looking for one marker that would best identify neutrophils. CD66b seems like a relatively new and specific marker, but I'm open to suggestions!
Dear Meaghan, most of the neutrophils will actually not be in the PBMC (=Mononuclear Cells, i.e. monocytes + lymphocytes) fraction under normal conditions, but in the granulocyte/erythrocyte layer after Ficoll. If you want to enumerate neutrophils, it would be better to use whole blood, not PBMC. Granulocytes (neutros/eos) can be distinguished in the FSC/SSC as SSC high cells (for an example, see here: http://www.nature.com/jid/journal/v132/n10/fig_tab/jid2012282f2.html). Inside the granulocyte gate, neutrophils with be positive for CD16, eosinophils will be negative. CD66b is also expressed on eosinophils, so it is not specific for neutrophils.
Further to Christina's info above, several neutrophil subsets have now been identified (see Pillay J, JCI for example) some of which are thought to be CD16 low. In reality, there is not a single marker expressed on human neutrophils that will identify all of them.
I agree with Christine that most neutrophils will not be found in the interface layer after Ficoll density gradient centrifugation, however...
We found that there is increasing neutrophil contimation of PBMCs after performing a Ficoll IF there is a delay in processing the blood, especially after 24 hours. Heparin seems to cause more contamination than EDTA.
As for neutrophil phenotype, I agree with Charlotte, there isn't a single marker to identify them and a combination of markers (including negative ones to 'gate out' cells you don't want) is the best strategy. CD15 and CD66b are good starting points (and HLA-DR negative/low).