Archives
br Colabeling evidence for ACh
Colabeling evidence for ACh and GABA cotransmission
Though functional demonstrations of ACh/GABA cotransmission remain largely limited to the retina and our recent analyses of cortex (Lee et al., 2010, Saunders et al., 2015a, Saunders et al., 2015b), evidence suggestive of ACh/GABA cotransmission has long existed in the literature. The earliest evidence comes from the chick ciliary ganglion. Neurons in this ganglion receive input through a large and reliable calyx-type synapse that depolarizes the ganglion neurons via electrical synapses and by activating nicotinic 490 receptors following ACh release in the calyx (McEachern et al., 1985). However, numerous studies have shown that these neurons are also sensitive to exogenous GABA, through the opening of GABAA receptors. With no local interneurons and the only input coming from the cholinergic neurons of the accessory oculomotor nucleus (AON), this suggests that the presynaptic neurons of the AON may cotransmit ACh and GABA (McEachern et al., 1985, Engisch and Fischbach, 1990, Engisch and Fischbach, 1992). However, the dual cholinergic/GABAergic identity of these neurons has not been confirmed and, to our knowledge, coexpression of ACh and GABAergic synthetic and packaging enzymes has not been examined in the chick AON.
Immunolabeling of neurotransmitters or proteins necessary for neurotransmitter synthesis and packaging had also hinted that cholinergic neurons of the rodent basal forebrain may be GABAergic. Kosaka and colleagues examined consecutive 40 micron fixed tissue sections that transected individual neurons, staining one section with choline acetyltransferase (ChAT), the enzyme that synthesizes ACh, and the other with glutamate decarboxylase (GAD), the enzyme that synthesizes GABA, or GABA itself. They reported individual neurons immunopositive for both sets of markers in the basal forebrain, retina (most likely corresponding to the SACs), as well as the spinal cord (Kosaka et al., 1988b). Another study confirmed this finding of overlapping expression of ChAT and GAD in consecutive tissue sections of individual basal forebrain neurons retrogradely labeled with cortically-injected wheat germ agglutinin bound to horse-radish peroxidase (Fisher and Levine, 1989). However, not every study has observed colabeling of GAD and ChAT proteins, instead finding intermingled, non-overlapping populations of GAD- and ChAT-positive neurons (Gritti et al., 1993). One probable explanation for this discrepancy is the differences in subcellular localization of the two major GAD isoforms, GAD65 and GAD67. The GAD67 isoform, which is commonly used as a marker for GABAergic identity, is found throughout the entire cell, incuding the soma. GAD65, in contrast, is localized in synaptic terminals and largely not detected at the soma (Soghomonian and Martin, 1998). If cholinergic neurons only express GAD65, then somatic immunostaining may miss GAD expression. Indeed, genetic labeling of GABAergic neurons shows overlapping expression of ChAT with neurons that express VGAT (slc32a1) and gad2, the gene that encodes GAD65, but not gad1, the gene for GAD67 (Saunders et al., 2015a). Notably, overlap of ChAT with GABAergic markers was limited to the forebrain and did not extend to the midbrain pedunculopontine nucleus, which also contains population of ChAT-expressing neurons. Because GAD protein may not always localize in the cell body, a more reliable way to assess if a neuron expresses GAD is to examine mRNA directly. One study using single-cell RT-PCR did show that individual neurons of the basal forebrain (BF) and globus pallidus (GP) do express transcripts of proteins indicative of both GABA release, such as gad and slc32a1, and ACh release, like ChAT (chat) and VAChT (slc18a3) (Tkatch et al., 1998).
To directly demonstrate expression of GABA synthetic enzymes in cholinergic neurons, we used fluorescent in situ hybridization to visualize expressed mRNA transcripts and thus avoid the confounds of missing protein expression due to non-somatic localization. We designed two probes with the same fluorophore against mRNA for both isoforms of GAD (gad1 and gad2), as well as a probe for chat, and performed in situ hybridization with tissue sections from adult, wild-type mice. We observed that cell bodies of the globus pallidus and nucleus basalis of the basal forebrain fluorescently labeled with probes against ChAT also showed GAD expression, though GAD expression appeared weaker in ChAT+ neurons than in neighboring ChAT− neurons (Fig. 1). This demonstrates that cholinergic neurons do express the necessary machinery to synthesize GABA for release in adulthood.