149 Synaptic loss in the MS spinal cord Figure 3. Synaptic change comparison between control, nonlesional, and lesional gray matter (GM). (A, B) The difference in synaptophysin intensity between control, nonlesional, and lesional anterior horns is shown in A, with the comparison between the 2 multiple sclerosis (MS) groups shown at higher magnification in B. (C, D) Synaptic bouton area differences between control and MS anterior horns are shown in C, whereas the high-magnification graph in D shows only data corresponding to the lesional and nonlesional anterior horns. (E) The absolute and percentage differences between groups. (F) The difference in the number of neurons between control and MS sections. (G, H) The relationship between synaptic bouton area and anterior horn area is shown in G, whereas the relationship between synaptic bouton area and number of neurons is shown in H separately for sections from control cases (black circles), and MS sections with nonlesional (red circles) and lesional anterior horns (blue circles). (I, J) The correlations show the relationship between synaptic bouton area and (1) GM (I) and (2) cross-sectional area (CSA; J). Two separate standardized regression coefficient (src) and p values are provided and correspond to control (black circles) and MS (green circles) sections, respectively. All p values are from linear mixed models allowing for multiple slices per patient and adjusting for side. Error bars represent standard deviation. **p <0.05, ***p <0.01, ****p <0.001. GML = lesional GM; NLGM = nonlesional GM. differences in disease duration and/or severity between studies. Although the magnitude of neuronal loss detected is substantial, it is proportionally lower than the synaptic loss, certainly when synaptophysin is used as a marker. The difference in the magnitude of change in synaptophysin and synapsin immunostaining is likely explained by their differential expression and function20. Synapsin and synaptophysin are the most abundant synaptic vesicle proteins, with distinct functional roles. They are the most commonly used markers for presynaptic terminals, with synapsin being the most specific marker for presynaptic terminals21. Synapsin is the major peripheral membrane protein, accounting for 6% of the total synaptic vesicle protein22. It regulates the reserve pool of synaptic vesicles available for exocytosis23, and maintains the organization and abundance of vesicles at presynaptic terminals24. Synaptophysin is the major integral membrane protein, accounting for 6 to 10% of total synaptic vesicle protein25-27, regulating the kinetics of synaptic vesicle endocytosis28, and synaptic vesicle retrieval through its interaction with synaptobrevin29. Synapsin is found at all glutamatergic and γ-aminobutyric acidergic (GABAergic) synaptic boutons, whereas synaptophysin levels are highest at glutamatergic and very low at GABAergic terminals21,30. Importantly, the development of both synapsin and synaptophysin expression is required for the maturation of presynaptic function and stabilization of presynaptic boutons31. Furthermore, the balance between these proteins will affect vesicle cycling and likely impact the probability of transmitter release, especially after strong or sustained stimulation. The mechanisms and timing involved in synaptic injury and loss remain to be established. Studies using material of pwMS at earlier disease stages may allow establishing the sequence of events leading up to the extensive synaptic loss, which may have been partially confounded by lower limb inactivity. In animal models of MS, synaptic loss has been shown to be associated with changes in the balance of inhibitory and excitatory neurotransmission as a consequence of inflammation32. However, analysis of the relationship between inflammation and synaptic damage was not possible due to the scarcity of inflammation in our tissue samples10. A cellular infiltrate suggestive of inflammatory activity was only detected in 3 of 154 blocks of our sample. Evidence suggests the main signature of inflammation in the spinal cord may be located in the meninges33, which were not examined in this study. Although inflammation may contribute to synaptic damage/loss in the course of MS, it may not necessarily be involved in the initiation of the processes leading to connectivity breakdown in the spinal cord.
RkJQdWJsaXNoZXIy MTk4NDMw