Azurophil granule - has_part relationships too liberal?

[dosumis]

According to GO, azurophil granules are only found in neutrophils, but CL has them as part of many cell types that are not classified as neutrophil. Who is wrong?

CC @addiehl

[addiehl]

GO is too limited in their definition of azurophil granule. Per wikipedia: "An azurophilic granule is a cellular object readily stainable with a Romanowsky stain. In white blood cells and hyperchromatin, staining imparts a burgundy or merlot coloration. Neutrophils in particular are known for containing azurophils loaded with a wide variety of anti-microbial defensins that fuse with phagocytic vacuoles."

Per Quora (not necessarily a perfect source): "Azurophilic (nonspecific) granules are lysosomes found in the cytoplasm of all five kinds of white blood cells. They’re named for their property of staining with blue or violet dyes (azur = blue; philic = loving)."

PMID:5914694 describes in great detail the staining of developmental and mature stages of granulocytes and describes azurophil granules in progranulocytes, myelocytes, metamyelocytes, band cells, and mature PMN (neutrophils, basophils, and eosinophils).

Separately, NK cells are described as having azurophilic granules, and some T cells as well (PMID:28717070 talks about normal NK and T cells having such granules, which are found also in many leukemias).

Basically, GO should redefine their definition of azurophil granule to be more general. Perhaps, something derived from Wikipedia's description:
"An azurophilic granule is a primary lysosome readily stainable with a Romanowsky stain."
They would probably hate that.

I think CL is fine.

[addiehl]

GO has now implemented my suggestion to broaden their definition of azurophil granule, via geneontology/go-ontology#27271 (comment)
I will close this issue now.

[dosumis]

Thanks, good that incorrect GO axiom removed.