Synaptic loss is a key feature of the secondary progressive phase of multiple sclerosis (MS) and is related to clinical and cognitive functioning. However, whether the excitatory or inhibitory synapses are more susceptible to MS pathology is insufficiently clarified to date.
To quantify GABAergic and glutamatergic synaptic densities in a sample of post-mortem MS brains and, hence, to investigate whether there is reason to suspect an imbalance in excitatory versus inhibitory neurotransmission.
Brains of 33 neuro-pathologically verified MS cases (21 women, mean age=63±12y) and 9 non-neurological controls (NC, 5 women, mean age=72±6y) were dissected shortly after death (mean post-mortem delay in MS: 5:37±1:29h; in NC: 9:19±2:85h). Sections of the superior frontal cortices were stained for myelin, parvalbumin- and calretinin-expressing interneurons and glutamatergic and GABAergic synapses. Subsequently, synaptic densities were quantified through confocal microscopy in sections of pre-determined regions of interest (ROIs) and image analyses. Data were analyzed using linear mixed-effects models.
Of the 71 defined ROIs in MS tissue, 24% were demyelinated, the remaining were normal-appearing grey matter (NAGM). No differences in densities of calretinin- and parvalbumin-expressing interneurons were observed between groups. For both excitatory and inhibitory synapse densities, there was a significant interaction between tissue type (NC NAGM, MS NAGM and MS demyelinated cortex) and cortical layer (P=.003; P=.001, respectively). Post-hoc testing revealed that the densities of both synapse types were reduced in cortical layer 6 (excitatory: P=.004; inhibitory: P=.002). NAGM in MS cortical layer 6 showed reductions of 12.5% (excitatory) and 14.9% (inhibitory) synaptic density as compared to NC values (P<.05). In demyelinated MS cortex a loss of 18.5% in excitatory synapses and 29.3% in inhibitory synapses was noted respective to NC values (P<.05).
In post-mortem MS tissue, we found a significant loss of excitatory and inhibitory synapses in layer 6 of superior frontal cortex. Interestingly, in NAGM the proportion of synaptic loss was similar for both synapse types, while in demyelinated cortex inhibitory synapses were affected more. The differential effects of fairly subtle differences in excitatory versus inhibitory synapse loss on functional measures may nonetheless have a substantial effect on cellular and network functioning. This effect is now being investigated in a corticothalamic mean-field model (results pending).