**8. Maturating B‐cells undergo bidirectional exchanges across the BBB**

Authors are often reluctant to localize the SHM process inside the CNS and prefer models that posit peripheral antigenic stimulation followed by CNS migration, although evidence is accumulating in favor of a local maturation process. High‐throughput sequencing techniques allow the analysis of clonal B‐cell populations on both sides of the BBB. A common finding is the demonstration of IgG cell lineages either restricted to one compartment (CNS, blood, and lymph node) or overlapping multiple compartments. In a study pairing CNS and cervical lymph node (CLN) B‐cells, about 6–15% of the IgG B‐cell sequences in plaques were recovered

**Figure 4. Schematic B‐cell lineage tree and putative migration routes**. Lineage founding B‐cell is postulated in the CLN (green), which drains antigens and B‐cells coming from the CNS. Immunoglobulins are subjected to somatic hy‐ permutation (SHM) and proliferation in cervical lymph nodes (CLN). Migrating B‐cell clones are sometimes encoun‐ tered in blood on their way to the CNS. B‐cell proliferation occurs both in CLN and in CNS. Evidence is growing that SHM may also occur inside the intrathecal compartment, possibly in meningeal tertiary lymphoid organs (TLOs). In such a model, rounds of B‐cell maturation and proliferation may occur in a recursive way both in CNS and in CLN [14, 149]. Long‐lived plasma cell (PC) differentiation occurs in the different compartments.

in the CLN [149]. Cell lineage analysis demonstrated that most of the founder cells originated in the CLN [149]. A tentative interpretation is that antigen‐driven affinity maturation of B‐cells takes place in the CLN, which drains CNS antigens [164], then B‐cells migrate to colonies populating the CNS and continue to traffic between the CNS and the periphery, notwith‐ standing the possibility that a bidirectional traffic occurs in association with clonal expansion in both compartments (**Figure 4**) [149]. Clonal populations of CSF IgM‐ and IgG‐secreting B‐ cells do not overlap and seem to have matured independently from each other [158]. This has been interpreted as a failure of IgM switching to IgG despite aberrant AID activity. However, only a partial overlapping of IgG and IgM populations has been demonstrated in various studies but it now needs to be replicated.

Naïve B‐cells are probably not randomly recruited in the inflammatory CNS since the VH family is biased [159]. Although the fate of naïve B‐cells emigrating to the CNS is still not clear, local maturation in the TLO structures and/or emigration to CLN are both probable (see below).

Most of the Ig peptides recovered from OCB by mass spectrometry match Ig‐secreting CSF B‐ cells [14]. Moreover, clusters of related B‐cells present in the CSF are also sometimes recovered from blood, the largest bi‐compartmental proportion being observed in association with a recent relapse [14]. This observation may explain the existence of OCB "mirror" patterns 3 and 4 with overlapping B‐cell lineage on both sides of the BBB.

#### **8.1. Local synthesis also occurs in the eye**

Inflammatory lesions are observed in the eye in association with MS and perivenous lympho‐ cyte cuffings (i.e., periphlebitis) are commonly observed, reminiscent of lymphoid aggregates [165]. Therefore, besides the classical intrathecal Ig synthesis, local intraocular synthesis may also affect the eyes and glands. The MRZ reaction occurs with single specificity in up to 76% of cases and with ≥2 specificities in 82% [166, 167], but the ratio of *Fs* between Fuchs hetero‐ chromic uveitis syndrome (FHUS) and MS is about 40‐fold, which is in the same range of ITS ratios as observed between MS and viral encephalitis. OCBs are always found in the CSF and aqueous humor of MS patients. Importantly, both OCB and MRZ patterns mismatch in most cases.
