Degradation of GM1 and GM2 by mammalian sialidases

In mammalian tissues, the pathway known for the catabolism of G(M1) [Gal beta 3GalNAc beta4(Neu5Ac alpha3)Gal beta 4GlcCer; where Cer is ceramide] is the conversion of this ganglioside into G(M2) [GalNAc beta4(Neu5Ac alpha3)Gal beta 4Glc beta Cer] by beta -galactosidase followed by the conversion of G(M2) into G(M3) (Neu5Ac alpha 3Gal beta 4Glc Cer) by beta -N-acetylhexosaminidase A (Hex A). However, the question of whether or not G(M1) and G(M2) can also be respectively converted into asialo-G(M1) (Gal beta 3GalNAc beta 4Gal beta 4GlcCer; G(A1)) and asialo-G(M2) (GalNAc beta 4Gal beta 4Glc beta Cer, G(A2)) by mammalian sialidases has not been resolved. This is due to the fact that sialidases purified from mammalian tissues always contained detergents that interfered with the in vitro hydrolysis of G(M1) and G(M2) in the presence of an activator protein. The mouse model of human type B Tay-Sachs disease created by the disruption of the Hexa gene showed no neurological abnormalities, with milder clinical symptoms than the human counterpart, and the accumulation of G(M2) in the brains of affected mice was only limited to certain regions [Sango, Yamanaka, Hoffmann, Okuda, Grinberg, Westphal, McDonald, Crawley, Sandhoff, Suzuki and Proia (1995) Nat. Genet. 11, 170-176]. These results suggest the possible presence of an alternative catabolic pathway (the G(A2) pathway) in mouse to convert G(M2) into G(A2) by sialidase. To show the existence of this pathway, we have used recombinant mammalian cytosolic sialidase and membrane-associated sialidase to study the desialylation of G(M1) and G(M2). We found that the mouse membrane-bound sialidase was able to convert G(M1) and G(M2) into their respective asialo-derivatives in the presence of human or mouse G(M2) activator protein. The cytosolic sialidase did not exhibit this activity. Our results suggest that, in vivo, the stable NeuAc of G(M1) and G(M2) may be removed by the mammalian membrane-associated sialidase in the presence of G(M2) activator protein. They also support the presence of the G(A2) pathway for the catabolism of G(M2) in mouse.