Both the Golgi and the endosomes have recently been proposed as the ma
in site of SM-synthase, the enzyme responsible for sphingomyelin (SM)
biosynthesis. To settle this confusion, we studied the subcellular dis
tribution of SM-synthase in human liver-derived HepG2 and baby hamster
kidney BHK-21 cells, To discriminate between Golgi and endosomes we m
ade use of 3,3-diaminobenzidine (DAB) cytochemistry. Cells were incuba
ted with a conjugate of transferrin (Tf) and horseradish peroxidase (H
RP), or with unconjugated HRP, to label the recycling pathway and the
complete endocytic pathway (including lysosomes) with peroxidase activ
ity, respectively. After cell homogenization, the peroxidase activity
was used to induce a local deposition of DAB-polymer, The total SM-syn
thase activity was not affected by this procedure, and, in contrast to
endosomes labeled with I-125-Tf, organelles containing SM-synthase di
d not increase in buoyant density as determined by Percoll density gra
dient fractionation, Thus, little, if any, SM-synthase localizes to th
e endocytic pathway of HepG2 and BHK-21 cells, In experiments performe
d at low temperature to inhibit vesicular transport, we found less tha
n 10% of newly synthesized short-chain SM at the cell surface, We conc
lude that most SM-synthase activity is present in the Golgi, and to a
small extent at the cell surface.