**11. The problem of Ig directed against nonspecific targets**

#### **11.1. The "MRZ pattern"**

In MS, the presence of TLO is associated with a more intense subpial demyelination (cortical lesions type III) and cell loss [169]. A strong topographical relationship is observed between TLO and subpial inflammation and an incidental observation showed that acute subpial lesions are associated with active local neurodegeneration (**Figure 5**) [180]. Owing to their small size and low number (about six follicles in positive blocks [168]), TLOs are probably largely underestimated by the sampling process of pathological examinations. Nevertheless, it was recently demonstrated that late post‐contrast FLAIR sequences on 3T‐MRI sequences might sample few leptomeningeal lesions (one lesion in 65% of patients, ≤6 in all of them) in a third of progressive patients [181], and this prevalence could be higher with 7T‐MRI. Pathological examination of one of them confirmed the congruence of the MRI lesion with TLO. Given the massive underestimation of TLO lesions, the practical consequences of this technique are currently being examined for their predictive clinical value. The longitudinal evolution of these structures and the effect of immunosuppressive therapies on them are unknown. Up to now, TLO could only be examined in the removed CNS structures including brain and leptomeninges but excluding skull and dura mater. Therefore, the recently discov‐ ered dura mater lymphatic network, which drains CSF to CLN, has never been examined in the context of MS. Since lymphatics are closely associated with TLO in all other tissues, confirmation of these structures might open up unexpected avenues in studies of inflammatory

Epitope spreading in EAE and MS is mainly documented for T‐cells where it occurs in association with TLO formation [176]. The contribution of B‐cells to epitope spreading, especially for their intrathecal counterpart, is less well known. The intimate mechanisms driving epitope spreading are speculative and it remains unclear to what extent SHM con‐

It was observed some years ago that the ITS of IgG becomes enriched during the early phase after MS onset [27, 42, 68–70, 74]. The frequency of the MRZ pattern increases, OCB‐negative patients become positive, and more OCBs occur in the few patients initially displaying a low number of OCB. Assessment of IgG synthesis with multiplex antigen arrays now allows epitope spreading against CNS antigens to be monitored over time. Serum studies have shown that both intramolecular and intermolecular spreading occurs early after the very first demyelinating index event in children but only in patients demonstrating a further clinical MS conversion [42]. By contrast, pediatric patients remaining monophasic fail to diversify their antigenic response. A failure to regulate the aberrant autoimmune response is tentatively thought to explain this observation. Studies on early harvested CSF would give rewarding clues about CNS B‐cell epitope spreading. A longitudinal description of the network of interactions in B‐ and T‐cell epitopes (functional immunomics) at both serum and CSF level, and their interaction with HLA polymorphisms, might shed light on the pathophysiology of

cell studies trafficking and maturation [182].

**10. Epitope spreading**

68 Trending Topics in Multiple Sclerosis

tributes to this process.

MS [183].

ITS may occur in supposedly healthy patients or in those suffering from nonimmune CNS disorders, but in such cases, synthesis is always directed against a former healed CNS infection such as Lyme disease, VZV, or HSV [22].

The long‐known "MRZ pattern" is an ITS of Ig directed against measles, rubella, and varicella zoster virus, the herpes virus sometimes being included ("MRZH"). Although the IgG fraction that belongs to this specific response in the CSF is estimated to represent less than 2% of the total amount of CSF IgG [78, 184, 185] and only a minor fraction of total OCB [77, 186, 187], the reaction is considered to be highly specific of MS. None of the MS patients in whom the MRZ pattern has been reported suffered from a history of clinical encephalitis involving these viruses. The proportion of MS patients with an elevated antibody index against ≥1 of these viruses increases over time and MS evolution [78]. This MRZ reaction is sometimes present in OCB‐negative MS patients and in CIS patients who will convert and it becomes more pro‐ nounced over time, as demonstrated by follow‐up LP [39, 188, 189].

ITS is polyspecific in a quarter of MS patients (against 2, 3, or 4 AI in 17, 4, and 2%), but remains monospecific (one elevated AI against M, R, Z, or H) in 22% [39]. The number of elevated AI strongly correlates with both age at spinal tap and disease duration and slightly changes over time [39, 190]. Moreover, there is a trend to a higher proportion of elevated AI in SP‐MS than in RR‐MS patients [191]. Immunosuppressive treatments including natalizumab are ineffective to prevent the persistence of the MRZ pattern [38, 103, 189]. Unlike in controls, no decline in AI levels occurs with age in MS patients considering either serum or CSF titers, and there is even a slight increase [192, 193].

#### **11.2. Broadening the "MRZ pattern" to all encountered infectious agents?**

A more comprehensive analysis of ITS with a larger range of infectious agents demonstrated that the MRZ pattern is an interesting concept that could be broadened. In fact, ITS against many other infectious agents has been confirmed in MS: rotavirus, HHV6 (20–30%), mumps, influenza, parainfluenza, adenovirus, respiratory syncytial virus, distemper virus, Coxsackie virus B4, vaccinia, JC virus (3%), CMV, poliovirus, toxoplasmosis (10%), *Borrelia burgdorferi* (26%), and mycoplasma (complete review in reference [194]). A previous silent CNS infection can be ruled out since the seroprevalence for each infection was the same in both MS and controls, whereas the latter had no specific reaction. Moreover, an ITS against tetanus and diphtheria toxoids was also demonstrated in vaccinated patients whose CNS was never exposed to native toxins [195, 196]. In a study of ITS against 17 infectious agents, 57% of MS patients had ITS against ≥5 infectious agents, with an increasing number over time [197], and virtually all patients demonstrated a reaction against at least one of nine viruses [198].

The pattern of reaction mirrors the individual's history of previous infections and the immu‐ nization level of the population: reaction against rubella is more observed in MS from Germany than in Cuban patients (lower incidence and immunization campaign in Cuban females) [199]. The rate of intrathecal reaction against a given infectious agent correlates with the rate of seroprevalence of this agent in the population.

Some major consequences can be drawn from these findings as follows:


Polyspecific ITS against all infectious agents is just a component of polyspecific ITS, which is thought to be characteristic of MS, and could throw light on the pathophysiology of MS.

#### **11.3. Nonspecific synthesis against infectious agents is lower in MS than in CNS infections**

The fraction of a specific antibody response against an infectious agent within the complete intrathecal IgG response is called the specific fraction (*Fs*) [200]. For example, an *Fs* value for measles of 2% means that 2% of the total intrathecal IgG response is directed against mea‐ sles. Neuroinfections are thought to be associated with very high *Fs* against viruses: *Fs* are 8– 45% for viral encephalitis (HSV, measles, VZV, subacute sclerosing panencephalitis). In oth‐ er words, about 55–92% of the intrathecal Ig in infectious pathologies are directed against non‐causative antigens [184, 200, 201], that is, a nonspecific ITS is very common even in neu‐ roinfections. In MS, each specificity in the MRZH reaction typically retains a very low me‐ dian *Fs* of 0.2–1.3% (ranging from 0.03 to 5.3) [200, 202, 203] and comparable results are obtained for *Fs* anti‐EBV [203]. These specific *Fs* results are about 40‐fold lower in MS pa‐ tients than those found in neuroinfections without overlap.

The amount of intrathecally secreted specific antibodies should be proportional to the number of intrathecal plasma cells. If one considers that circulating plasma cells are nonspecifically and randomly selected from blood to home to CNS TLO, the relative proportion of each specific IgG synthesis in CSF should grossly parallel to their proportion in blood [197]. The ranking of specific antibody concentrations in blood and CSF differs in 67% of MS patients, confirming that CSF IgG secretion does not simply mirror blood secretion. ITS of the specific antibodies occurs independently from each other [200]. We propose two non‐mutually exclusive explan‐ ations. A first hypothesis involves the differential intrathecal proliferation of specific B‐cells after being recruited from blood but before being committed to terminal differentiation into plasma cells, owing to a favorable intrathecal lymphoid environment. A second hypothesis posits a nonrandom brain homing of circulating plasmablasts. It seems unlikely that plasma‐ blast homing is driven by IgG specificity. Rather, the critical intensity of plasmablast/plasma cell recruitment to the brain owing to the intensity of the peripheral immune response to infection could be involved.

#### **11.4. The paradox of low intrathecal anti‐EBV reaction**

The rate of intrathecal reaction against a given infectious agent correlates with the rate of

**1.** ITS against an infectious agent in MS considered in isolation should not be considered as a clue for a chronic infection in MS, since the coexistence of reaction against many

**2.** The concept of "MRZ pattern" should be extended to a "polyspecific infectious pattern."

**3.** The hypothesis of the antigenic mimicry of a single infectious agent should always be considered in the light of a broad reaction against various infectious agents, whose

Polyspecific ITS against all infectious agents is just a component of polyspecific ITS, which is thought to be characteristic of MS, and could throw light on the pathophysiology of MS.

**11.3. Nonspecific synthesis against infectious agents is lower in MS than in CNS infections**

The fraction of a specific antibody response against an infectious agent within the complete intrathecal IgG response is called the specific fraction (*Fs*) [200]. For example, an *Fs* value for measles of 2% means that 2% of the total intrathecal IgG response is directed against mea‐ sles. Neuroinfections are thought to be associated with very high *Fs* against viruses: *Fs* are 8– 45% for viral encephalitis (HSV, measles, VZV, subacute sclerosing panencephalitis). In oth‐ er words, about 55–92% of the intrathecal Ig in infectious pathologies are directed against non‐causative antigens [184, 200, 201], that is, a nonspecific ITS is very common even in neu‐ roinfections. In MS, each specificity in the MRZH reaction typically retains a very low me‐ dian *Fs* of 0.2–1.3% (ranging from 0.03 to 5.3) [200, 202, 203] and comparable results are obtained for *Fs* anti‐EBV [203]. These specific *Fs* results are about 40‐fold lower in MS pa‐

The amount of intrathecally secreted specific antibodies should be proportional to the number of intrathecal plasma cells. If one considers that circulating plasma cells are nonspecifically and randomly selected from blood to home to CNS TLO, the relative proportion of each specific IgG synthesis in CSF should grossly parallel to their proportion in blood [197]. The ranking of specific antibody concentrations in blood and CSF differs in 67% of MS patients, confirming that CSF IgG secretion does not simply mirror blood secretion. ITS of the specific antibodies occurs independently from each other [200]. We propose two non‐mutually exclusive explan‐ ations. A first hypothesis involves the differential intrathecal proliferation of specific B‐cells after being recruited from blood but before being committed to terminal differentiation into plasma cells, owing to a favorable intrathecal lymphoid environment. A second hypothesis posits a nonrandom brain homing of circulating plasmablasts. It seems unlikely that plasma‐ blast homing is driven by IgG specificity. Rather, the critical intensity of plasmablast/plasma cell recruitment to the brain owing to the intensity of the peripheral immune response to

Some major consequences can be drawn from these findings as follows:

presence depends mostly on epidemiological variations.

tients than those found in neuroinfections without overlap.

infection could be involved.

seroprevalence of this agent in the population.

70 Trending Topics in Multiple Sclerosis

infectious agents is the rule in MS.

Seroprevalence for EBV is virtually complete in MS patients, so the probability of ITS should be at least equal to or higher than other nonspecific viruses. However, anti‐EBV intrathecal values have shown that an unexpectedly small proportion of patients have ITS (in the range of <20%) [203–211]. Comparing CSF OCB against HHV6 and EBV, OCB occurred twice more frequently against HHV6 than against EBV [211]. Moreover, AI against EBV is sometimes twofold lower than AI against each MRZ component [207]. A difference in seroprevalence can be ruled out. Measles and rubella seroprevalence resulting from either natural infection or vaccination exceed 90% in most populations and vaccination campaigns started decades before [39]. Furthermore, varicella seroprevalence exceeds 90% in all European populations and EBV seroprevalence is virtually complete in MS patients.

The lower‐than‐expected intrathecal response against EBV is not consistent with the strong correlation linking high serum anti‐EBV levels and MS activity. In the light of this finding, such an extreme discrepancy can be interpreted as a strong clue for EBV infection preceding MS clinical onset [203]. Possible pathophysiological explanations have been developed elsewhere in the light of TLO [194]. The peculiar relationship between EBV and MS pathology is reinforced by the demonstration of a high intrathecal EBV‐specific CD8+ cytotoxic activity only early in MS patients, without recruitment of CD8+ cells against different targets (CMV‐ specific CD8+ cells) [212] and the clearance of ITS in a case receiving autologous CD8+ T‐cells activated against EBV [109].

#### **11.5. Nonspecific synthesis may be a simple property of meningeal tertiary lymphoid organs**

Chronic nonspecific synthesis is constant in MS, but may not strictly indicate a bystanding reaction since they are almost absent in other mostly acute CNS inflammations [194]. These unspecific reactions rather indicate a dedicated specific function suggestive of the persistence of TLO in CNS. During immune activation in the periphery, naïve B‐cells undergo hypermu‐ tation of the Ig genes driven by germinal centers in the spleen and ganglia, and surviving cells are committed to plasmablasts released in the blood for a few days on their way to survival niches, where they differentiate into long‐lived plasma cells. Most of the niches are situated in the bone marrow but they are also to be found in secondary and tertiary lymphoid organs, where they display the ability to retain the newly formed plasma cells. Cultures of synovial fragments (containing TLO) from rheumatoid arthritis retain the ability to secrete non‐disease‐ specific antibodies, such as anti‐tetanus toxin IgG [213]. Three weeks after, lupus‐prone mice were vaccinated against OVA, and the same number of antibody‐secreting cells against OVA was recovered in inflamed kidneys as in the bone marrow [214]. Therefore, nonspecific homing of plasma cells may be a common feature of inflammatory tissues and lead to local nonspecific synthesis.

The prolonged survival of retained plasma cells necessitates an anti‐apoptotic environment provided by cell niches. For example, CXCL12, which is a major determinant of PC retention in niches, is elevated in MS CSF, expressed by astrocytes and in the vicinity of lymphoid infiltrates [215]. Therefore, we previously suggested that nonspecific Ig synthesis in MS might simply be a common function of TLO harvesting Ig‐secreting cells, as is observed in most inflammatory disorders [194, 213]. An experimental demonstration was given by a series of MS patients who developed CSF IgG against TT after tetanus toxoid vaccination [196]. According to this assumption, the recruitment of intrathecal antibody‐secreting cells should be progressive and the complete MRZ pattern might take years or even decades to achieve. This was observed in clinical studies where the mean age at spinal tap and disease duration correlated with the number of elevated antibody indices [39]. Half of the patients having fewer than 5 years of disease duration had no elevated AI (MRZ), whereas 70% of those with more than 10 years duration had AI ≥2 [39]. A similar trend was observed in a study based on 17 antigens [197].
