**2. Immunological risks caused by silica particles**

As shown in **Table 1**, there are various immunological risks associated with exposure to silica particles. These findings were established by in vitro assays using peripheral blood mononu‐ clear cells (PBMC) derived from HD cultured with silica particles as well as freshly isolated immune cells derived from SIL. Additionally, various autoantibodies (aAbs) were detected from SIL [11–14]. All SIL comprised Japanese workers of a firebrick factory located at Bizen City, Okayama Prefecture, Japan, diagnosed with silicosis according to the International Labor Organization (ILO) 2000 guidelines for pneumoconiosis and monitored at Kusaka Hospital or Hinase Urakami Iin/Clinic at Bizen City. All SIL showed no symptoms related to autoimmune diseases or cancers.



**Table 1.** Immunological risks caused by silica particles.

#### **2.1. aAbs detected in SIL**

First, the risk of dysregulated autoimmunity assessed by the detection of particular aAbs will be discussed. Various aAbs have been detected in SIL, such as antinuclear antibody (ANA) [37–40], antismooth muscle aAb [41], antiglomerular basement membrane (GBM) aAb [41], antineutrophil cytoplasmic aAb (MPO-ANCA) [37, 42–48], rheumatoid factor (RF) [37–39, 49– 53], anti-Scl-70/topoisomerase I aAb [37, 54–60], anti-Fas/CD95 aAb [61], anticaspase 8 aAb [62], anticentromere/CENP-B (centromere protein B) aAb [63], antidesmoglein aAb [64], anti-PL 12 (aminoacyl tRNA synthetase) aAb [65], and anticollagen aAb [39], as found in publica‐ tions located via PubMed.

Of these aAbs, we investigated several Abs of interest, such as anti-Fas/CD95 Ab [61], anti‐ caspase 8 Ab [62], anti-Scl-70 Ab with respect to specific human leukocyte antigen (HLA) types [56–58], and anti-CENP-B Ab [63] and reported the case of antidesmoglein Ab-positive SIL [64].

We detected anti-Fas/CD95 aAb in approximately one-fourth of SIL [61]. Since T cells in SIL tend to be categorized into two classes, Fas/CD95-mediated apoptosis prone and resistant groups as described later in this chapter, it is important to determine whether the detected anti-Fas/CD95 aAb is functional in terms of the induction of Fas/CD95-mediated apoptosis. To examine this issue, we employed our established human sister myeloma cell lines, KMS-12PE and KMS-12BM. The former cell line was established from the pleural effusion of a myeloma patient, which showed high expression of Fas/CD95 on its surface as a result of apoptosis and growth inhibition caused by anti-Fas/CD95 agonistic antibody. The latter cell line was derived from bone marrow obtained from the same patients, who showed very low expression of Fas/ CD95 and no apoptosis caused by Fas/CD95 agonistic antibody [66]. Following cultivation of both cell lines with anti-Fas/CD95 aAb-positive serum from SIL, the growth of KMS-12PE was reduced by apoptosis, whereas the growth of KMS-12BM was unaffected [61]. These results indicated that anti-Fas/CD95 aAb is functional. Additionally, epitope mapping employing 12 amino acid polypeptides with the SPOT system of anti-Fas/CD95 aAb was analyzed. As a result, a minimum of four and a maximum of ten epitopes were found, and several amino acid residues involved in binding Fas ligand, such as C66, R87, L90 E93, and H126, were identified [61].

As in the case of anti-Fas/CD95 aAb, anticaspase 8 aAb was investigated in terms of the dysregulation of Fas/CD95-mediated apoptosis of lymphocytes in SIL [62]. The association of anticaspase 8 aAb with HLA types was examined. As a result, the frequencies of HLA-DRB1\*0406 were significantly higher in aAb-positive SIL (16.7%) compared with control individuals (3.0%, p<0.001). Additionally, HLA-DR4; DQB1\*0302 was found in one-fourth of positive SIL, and DPB1\*0601 was also higher in positive SIL (5.9%) compared with controls (0.6%, p<0.05), whereas DQB1\*0401 was lower in positive SIL (0%) compared with controls (13.3%, p<0.001). Furthermore, epitope mapping showed that a minimum of four and a maximum of thirteen polypeptides seemed to be involved. Among these, two important catalytic cysteine residues were found, cysteine Cys287 and Cys360, located in the unique pentapeptide motif QACQG [62].

Regarding the relationship between aAb and specific HLA type, we reported HLA types among anti-Scl70/topoisomerase I aAb-positive SIL [56–58]. Results indicated that the allelic frequency of HLA-DQB1\*0402 was significantly higher in aAb-positive SIL (28.6%) than in aAb-negative SIL (1.5%, p<0.001), as well as in controls (0.8%, p<0.001). Additionally, DQDB1\*0301, DQB1\*0601, and DPB1\*1801 were higher in aAb-positive SIL than in aAbnegative SIL, whereas no significant differences were found compared with controls [56–58].

In terms of anti-CENP-B/centromere aAb, the titer index (Log10) of anti-CENP-B autoantibody in SIL was higher than that of HV, and patients with systemic sclerosis (SSc) was higher than those of HV and SIL. This titer index was positively correlated with an assumed immune status for HV as 1, SIL as 2, and SSc as 3. Moreover, although the titer index of anti-CENP-B auto‐ antibody formed the same factor with anti-Scl-70 autoantibody, the Ig G value, and age of SIL, the property of other factors extracted indicated that anti-Scl-70 antibody was positively related with the Ig A value, while the converse was true for anti-CENP-B from the results of factor analysis. Those results indicated that the titer index of anti-CENP-B autoantibody may be employed as a biomarker in identifying dysregulation in SIL cases.

Taken together, various aAbs found in SIL have indicated that dysregulation of autoimmunity was caused by chronic and recurrent exposure to silica particles that remained in lung and related lymph nodes of various human cells, especially B cells. Some of these aAbs may be related to Fas/CD95-mediated apoptosis of lymphocytes and cause further dysregulation of autoimmunity such as in the case of long-surviving self-antigen recognizing clones in T cells [11–14].

Furthermore, examination of HLA types seemed to be important in revealing several aAbs in SIL. Although it can be mentioned that repeated and continuous screening of aAbs as well as the initial screening of HLA types seems to be necessary among workers in contact with silicarelated substances for the detection of dysregulation of autoimmunity, the use of genotyping, such as determining HLA types, is not permitted during employee selection procedures. However, a consideration of particular occupational health risks together with individual sensitivities is required in an effort to prevent occupational health hazards and associated future hardships.

#### **2.2. Fas/CD95-mediated apoptosis–related molecules in SIL**

Fas/CD95-related molecules analyzed in SIL are shown in **Table 1** [11–14, 67]. Regarding molecules that inhibit Fas/CD95-mediated apoptosis, the level of soluble Fas/CD95 was higher in the serum of SIL compared with HD, and similar to the level in systemic lupus erythema‐ tosus (SLE) [68], while higher mRNA expression, determined as the ratio of soluble to wildtype Fas/CD95, was present in SIL compared with HD in PBMC [69]. Additionally, higher amounts of various alternatively spliced variant messages of the Fas/CD95 gene were detected in PBMC from SIL compared with HD [70]. All of these variant messages, including soluble Fas/CD95, possess a Fas ligand-binding domain but lack a membrane-binding domain. Hence all of these translation products are secreted into the extracellular space and bind with Fas ligand, thereby protecting cells against membrane Fas-mediated apoptosis [70]. Furthermore, the expression of the protective molecule decoy receptor 3 (DcR3), which acts against the Trail molecule and similarly induces apoptosis via a Trail receptor and the same intracellular signaling molecules for apoptosis, such as caspase 8 and 10 [71, 72], was higher in SIL PBMC compared with HD [73]. These findings indicated that some types of T cells in PBMC from SIL provide protection against Fas/CD95- and Trail-induced apoptosis, which leads to long survival of these T cells and self-antigen recognizing clones [67].

However, several findings that showed accelerated Fas/CD95- and Trail-mediated apoptosis in PBMC of SIL were investigated. Messenger RNA expression in PBMC of several genes which act as physiological inhibitors of Fas/CD95- and Trail-mediated apoptosis, such as I-Flice (inhibitor of FADD-like interleukin-1β–converting enzyme), surviving, sentrin, and inhibitor of caspase-activated DNase (ICAD) was lower in SIL compared with HD [67, 74]. In addition to the aforementioned detection of functional anti-Fas/CD95 autoantibody, some types of T cells in PBMC from SIL possess enhanced Fas/CD95-mediated apoptosis [61]. Further studies revealed that this fraction may include Treg cells [13, 14]. Thus, a decrease in the number of Treg cells by apoptosis and an increase in the number of responder T cells caused by silica exposure may be the cellular biological mechanisms at work in SIL, which consequently impart susceptibility to autoimmune diseases in SIL.

We found higher expression of Fas/CD95 in Treg (CD4+, CD25+, and forkhead box P3 (FoxP3) +) [75, 76] and sensitivity to Fas-agonistic antibody–induced apoptosis in Treg cells from SIL [77]. Furthermore, when PBMC from HD were cultured with silica particles in vitro, Treg cell numbers were selectively reduced by apoptosis and the population of responder T cells was enhanced [77]. Thus, the aforementioned T cell population prone to Fas/CD95-mediated apoptosis seems to comprise Treg cells, and the imbalance that occurs as a result of a decreased Treg and surviving responder T cell population in SIL induces dysregulation of autoimmunity [13, 14, 77].

Moreover, there is evidence showing chronic activation of responder T cells. For example, CD69, an early activating marker of T cells, was gradually expressed in T cells when PBMC from HD were cultured in vitro with silica particles [78]. Expression of the program death protein 1 (PD-1) gene, another activation marker of T cells, in CD4+ CD25+ as well as in CD4+ CD25− T cell populations was higher in SIL compared with HD, which showed negligible expression [78]. Expression of serum soluble interleukin (IL)-2 receptor (sIL-2R) was also higher in SIL compared with HD [79].

Taken together, SIL possess a risk of developing dysregulation of autoimmunity. This risk can be detected using various markers mentioned above, such as serum soluble Fas, sIL-2R, and serum DcR3 (recently, the enzyme-linked immunosorbent assay (ELISA) kit is available for laboratory use), in SIL during their early clinical phases.
