An overall pathological hallmark of MS is the loss of myelin and oligodendrocytes resulting in demyelination and lesion formation in the CNS. Although cases of MS patients suffering from GML were already described in the early 20th century, research has focused mainly on WML, since conventional histochemical staining and imaging techniques were not, or only to a limited extent, able to visualize GML. Since the discovery of new imaging techniques, it is nowadays possible to visualize GML in vivo as well, which has led to a strong increase in research of this type of lesion. WML can be categorized based on their pathological profile. The presence of infiltrated immune cells and activated glial cells are characteristic for active WML. These immune cells can still be found at the border of the lesions and not in the center of the lesion when the lesion becomes chronic active. Inactive lesions are characterized by a relative absence of infiltrated immune cells and activated microglia, while hypertrophic astrocytes fill the lesion and form a scar. Interestingly, WML and GML differ in their cellular profile. WML show BBB damage and influx of immune cells. Yet, BBB damage is not evident in GML and influx of immune cells is significantly less compared to WML. Therefore, it might be that, although demyelinating lesions occur in WM and GM of MS patients, the formation of WML and GML has a distinct underlying pathogenic mechanism and, consequently, development of novel therapies should take this into consideration. Currently available therapies for MS focus on the prevention of the infiltration of immune cells or dampening of the immune response in the CNS. However, these therapies do not stop the progressive phase of MS which involves mostly neurodegeneration in GML. Thus, it is of utmost importance to further characterize WML and GML and determine similarities and differences between these types of lesions. This may lead to a further understanding of the pathogenesis of WML and GML and lead to other or additional targets for therapy.
The process of infiltration of immune cells through the BBB involves many molecules and receptors, including glial-derived factors such as chemokines. The activation of microglia and astrocytes is a crucial and early event in the pathogenesis of both WML and GML formation. However, GML present with less glial cells compared to WML, which questions whether differences between WM and GM glial cells contribute to the observed pathological cellular difference between WML and GML.
In an attempt to further characterize WML and GML pathology, we focused on hippocampal pathology for two different reasons. First, the hippocampus harbours both WM as well as GM, which makes this brain structure suitable to study WML and GML differences. Secondly, the hippocampus plays a pivotal role in the processes of learning and memory. MS patients often suffer from cognitive problems, e.g. memory problems, and accumulating evidence suggests that hippocampal lesions are the underlying cause. Indeed, the hippocampus is often affected in MS patients and hippocampal lesions correlate with memory deficits in these patients. The major neurotransmitters involved in hippocampal learning and memory are glutamate, Ach and GABA. Disturbed glutamate signaling has already been described in MS. However, the status of cholinergic and GABAergic neurotransmission in MS patients is not yet known.

Hence, the two aims of the studies described in this thesis were:
1) to determine the glial cell responsiveness in GM and WM of MS patients.
2) to identify changes in cholinergic and GABAergic neurotransmitter systems in the hippocampus of MS patients.

In this final chapter, our main findings are summarized and discussed. In addition, directions for future research are given.

In chapter 2, we reviewed the literature to identify possible underlying mechanism(s) of pathological differences between WML and GML. First, the observed paucity of infiltrated immune cells in GML compared to WML might be, partly, due to the age of the lesions. Research to identify cells present in these types of lesions depends mostly on immunohistochemistry on post mortem tissue. Since analysis of biopsy material of MS patients did show infiltrated immune cells in GML, it might be that these immune cells are present during the very early stage of GML formation.
Alternatively, differences between WML and GML can be the result of the local environment. While the GM is mainly composed of neurons and GM astrocytes and microglia, the WM harbours myelin forming oligodendrocytes, WM astrocytes and WM microglia. Neurons express several immune dampening molecules, e.g. CD200, CD47, ICAM-5 and CX3CL1. In addition, neurotransmitters that are released by neurons also modulate the immune response. Although the effects of glutamate and acetylcholine release on the immune response are dual, GABA is known for its immunosuppressive effects. Since GABA levels have been described to be higher in healthy GM compared to WM, this could contribute to the immunosuppressive environment of the GM. In addition, during MS, immune cells are directed against myelin and myelin breakdown and oligodendrocyte apoptosis occurs. The myelin debris that is left induces an immune response. Therefore, the immune response in GML could be less compared to WML due to lower levels of myelin debris in GML compared to WML.
Another difference between WM and GM that might account for the observed paucity of activated and infiltrated immune cells in GML is that BBB damage has been observed in WML, leading to infiltrating leukocytes in the parenchyma. Although there is no evidence for a physiological or anatomical difference in BBB in WM compared to the BBB in the GM, BBB damage has not been shown in GML.
Several lines of evidence indicate different reactivity of microglia and astrocytes in WM compared to GM. Although it remains unclear whether microglia and astrocytes from the WM are different types of cells compared to those found in the GM, it has been shown that glial cells in the WM express different molecules and can have a different phenotype than their GM counterparts. In humans, WM microglia have been described to be more numerous compared to microglia in the GM. In addition, WM microglia seem to have a stronger tendency to get activated and to be pro-inflammatory upon disturbed homeostasis. Morphologically, WM astrocytes are described as with a fibrous-like morphology whereas GM astrocytes have a protoplasmic-like morphology. In addition, in mice the presence of certain glial-derived chemokines involved in the migration of immune cells through the BBB is higher in WM compared to GM. In this context, we analyzed the expression of the pro-inflammatory interleukin-1ß (IL-1ß) (chapter 3) and monocyte-chemotactic protein-1 (CCL2) (chapter 4) in WML and GML.

IL-1ß is an important pro-inflammatory cytokine, being the principal driver of immune responses in the CNS. IL-1ß has been described to be involved in many processes during WML formation, e.g. oligodendrocyte apoptosis and leukocyte migration. We questioned whether IL-1ß is also present in GML, in addition to WML. For this purpose, we analyzed the presence of IL-1ß and its naturally occurring anti-inflammatory counterpart, the IL-1ß receptor antagonist (IL-1ra) in the brain and spinal cord of cr-EAE rats. This experimental model mimics inflammatory pathology of MS and enabled us to analyze the presence of IL-1ß and IL-1ra throughout the whole brain during the early phases of demyelination (chapter 3). We demonstrated that IL-1ß and IL-1ra are present in GML, in addition to WML. This increase was strongest during the early phases of cr-EAE and suggests the involvement of IL-1ß in both WML and GML formation, which was not prevented by the presence of endogenous IL-1ra.

CCL2 has been described extensively with regard to its ability to attract leukocytes across the BBB. A discrepancy between WM and GM expression of CCL2 and its receptor CCR2 may, therefore, play an important role in the pathological cellular difference observed between WML and GML. This hypothesis is corroborated by the findings described in chapter 4. The expression of CCL2 and CCR2 was analyzed in WML and GML of post mortem hippocampal tissue of MS patients. First, CCL2 en CCR2 mRNA levels were determined in hippocampi of healthy control subjects and myelinated and demyelinated hippocampi of MS patients. A significant upregulation of CCL2 and CCR2 mRNA was observed in demyelinated hippocampi compared to control hippocampi and myelinated hippocampi of MS patients. Immunohistochemistry for CCL2 and CCR2 was used to identify the spatial and cellular expression of CCL2 and CCR2 in active and inactive hippocampal lesions. Interestingly, CCL2 was barely detected in hippocampal GM, but was slightly increased in inactive hippocampal WML, and significantly upregulated in active hippocampal WML. Similarly, CCR2 was significantly increased in active hippocampal WML. However, in contrast to our findings on CCL2, CCR2 was also significantly upregulated in active GM hippocampal lesions. However, the number of CCR2 expressing cells was significantly higher in hippocampal WM compared to GM. The higher expression of CCL2 and CCR2 in hippocampal WML and low expression of CCL2 in GML could contribute to the observed difference in the number of infiltrating immune cells between WML and GML.

More than 50% of MS patients suffer from cognitive deficits, among which problems with memory is among the most frequently reported.20 Hippocampal lesions have been linked to memory deficits in MS patients. The three major neurotransmitter systems involved in learning and memory are glutamate, Ach and GABA. Previous studies have already shown disturbed glutamatergic neurotransmission in MS. However, it is unknown whether cholinergic and GABA-ergic neurotransmission are affected in the hippocampus of MS patients.

The neurotransmitter Ach is of importance for learning and memory. The enzyme responsible for Ach synthesis is ChaT, while the enzyme AchE hydrolyzes Ach thereby reducing its concentration in the synaptic cleft. In chapter 5, ChaT and AchE activity and protein expression were examined in hippocampi of MS patients compared to hippocampi of patients with AD and control hippocampi. A severe decrease in ChaT and AchE activity in the hippocampus had already been established in AD, a neurodegenerative disease characterized by severe memory loss. We found that ChaT activity and protein expression was significantly reduced in hippocampi of MS patients compared to control hippocampi. This reduction was comparable to the reduction observed in hippocampi of AD patients. Interestingly, AchE activity and protein levels remained unchanged in MS patients compared to controls, while these were significantly reduced in AD patients. These results suggest that in hippocampi of MS patients there is a reduced production of Ach, while the clearance of Ach from the synaptic cleft remains unchanged resulting in lower levels of Ach compared to controls. Therefore, treatment of cognitive decline in MS patients could possibly benefit from pharmacological inhibition of AchE activity.

In chapter 6, we examined PV and GAD67 expression, which are involved in GABAergic neurotransmission. PV represents a subgroup of GABAergic interneurons and GAD67 is an enzyme that converts glutamate into GABA, and is considered a marker for GABA. GABAergic neurotransmission is essential for optimal memory function. The number of PV positive interneurons did not significantly differ between MS patients and controls subjects. Interestingly, we observed that the number of GAD67 positive interneurons significantly increased in and around active CA1 lesions of MS patients. In addition, GAD67 positive astrocytes were significantly more numerous in hippocampal WML than GML of MS patients and the level of GAD67 immunoreactivity in astrocytes is significantly increased in MS patients with active hippocampal lesions, and relates to self-reported cognitive impairment. Thus, this suggests that in MS patients GABAergic changes occur in neurons and astrocytes of which the latter possibly contributes to cognitive impairment.