λ-Carrageenan It is well known that cGMP levels are reduced
It is well known that cGMP levels are reduced in cerebellar slices of hyperammonemic rats; that this is due to enhanced activation of NMDA receptors and that cGMP levels are normalized by blocking NMDA receptors with MK-801 . We therefore tested whether blocking NMDA receptors also normalizes cAMP levels. MK-801 did not affect cAMP levels in control rats, but reduced cAMP in hyperammonemic rats (p<0.001) to 51±6% of basal (Fig. 4B), the same levels reached in the presence of 8-Br-cGMP. This supports that increased activation of NMDA receptors is responsible for increased cAMP levels and PKA activity in hyperammonemic rats.
We then tested whether blocking NMDA receptors with MK-801 modulates phosphorylation and membrane λ-Carrageenan of GluA2. MK-801 completely normalized phosphorylation of GluA2 at Ser880 in hyperammonemic rats, reducing it to 111±24% (p<0.05) reaching similar levels to control rats (Fig. 5A). This was associated with increased membrane expression of GluA2, which reached 117±19% (p<0.05) (Fig. 5B).
These results suggest therefore that in hyperammonemic rats enhanced activation of NMDA receptors reduces cGMP levels and, subsequently, PDE2 activity, resulting in increased cAMP levels and PKA activity, which activates PLC and PKC, increasing phosphorylation of GluA2 at Ser880 which results in reduced membrane expression (Fig. 6).
As Ser831 of GluA1 may be also phosphorylated by PKC, we then assessed if the same pathway is increasing phosphorylation and membrane expression of GluA1 in hyperammonemic rats. Phosphorylation of Ser831 of GluA1 is increased (p<0.05) in hyperammonemic rats to 183±34% of controls (Fig. 1E) and is reduced by blocking NMDA receptors with MK-801 (to 77±16%, p<0.01; Fig. 7A), by inhibiting PKA (to 89±11%, p<0.001; Fig. 7B), by inhibiting PKC (to 40±10%, p<0.001; Fig. 7C) or by inhibiting PLC (to 57±9%, p<0.001; Fig. 7D). This supports that the pathway depicted in Fig. 6 is also responsible for increased phosphorylation of GluA1 at Ser831 in hyperammonemic rats.
In contrast, phosphorylation of Ser845 of GluA1 is not modulated by this pathway. It is reduced in hyperammonemic rats to 79±9% (p<0.01) of control rats (Fig. 1G) and was not normalized by MK-801 (90±7%, Fig. 7E), by inhibition of PKA (71±18%, Fig. 7F), by inhibition of PKC (72±14%, Fig. 7G) or by inhibiting PLC (to 73±9% of controls; Fig. 7H).
For membrane expression of GluA1 the results obtained were similar to those for its phosphorylation at Ser831. Membrane expression was increased (p<0.01) to 184±32% of controls in hyperammonemic rats and was reduced by blocking NMDA receptors with MK-801 (to 87±6%, p<0.001; Fig. 8A), by inhibiting PKA (to 63±17%, p<0.001; Fig. 8B), by inhibiting PKC (to 109±13%, p<0.01; Fig. 8C) or by inhibiting PLC (to 94±18%, p<0.001; Fig. 8D). This supports that the NMDAR-PKA-PLC-PKC pathway shown in Fig. 6 is also responsible for altered membrane expression of GluA1 in hyperammonemic rats.
Discussion The results show that hyperammonemia alters membranes expression of the GluA1 and GluA2 subunits of AMPA receptors in opposite ways, increases membrane expression of GluA1 and reduces that of GluA2. This effect is the opposite observed by Liu and Cull-Candy for repetitive synaptic activity, which triggers the loss of synaptic GluA2-lacking (Ca-permeable) receptors and drives a delivery of GluA2-containing (Ca-impermeable, Na+-permeable) receptors to the synapse . This results in different post-synaptic responses due to activation of different signal transduction pathways. This would result in altered glutamatergic neurotransmission. In other situations membrane expression of the GluA1 and GluA2 subunits may be modulated independently. For example, surface expression of GluA1 in hippocampus is reduced during contextual fear conditioning in aged rats while GluA2 is not affected , indicating that GluA1 and GluA2 are not mutually exclusive. Their proportion in the AMPA receptors in the membrane surface depends on different mechanisms (see ref. 11 for a review). We show here that hyperammonemia alters one of these mechanisms resulting in increased membrane expression of GluA1 and reduced membrane expression of GluA2. The net result is an increase of the abundance of heterotetramers containing GluA1 and a decrease in those containing GluA2. According to Liu and Cull-Candy  this would be due to removal of tetramers containing GluA2 from the surface and insertion of tetramers containing GluA1, and not to exchange of subunits in the surface tetramers .