Penetration of Anti-DNA Antibodies into Live Cells
The ability of ANA to enter the nucleus of live cells was initially reported by Alarcón-Segovia et al. in 1978
[1]. Using a direct immunostaining method without a second antibody, they documented the internalization of anti-RNP antibodies obtained from a patient with mixed connective tissue disease into normal peripheral blood mononuclear cells (PBMCs). Soon after, they reported similar findings with anti-DNA antibodies as well
[2]. Initially, these findings met with skepticism, but gradually, many studies confirmed this phenomenon
[3][4][5]. The mechanisms responsible for internalization are multifarious. Some anti-DNA antibodies enter cells via Fc-receptor-mediated endocytosis, but there are examples showing that recombinant single-chain fragments of the variable chains (scFv) lacking the Fc region can still enter cells
[6][7]. Some anti-DNA antibodies enter the nucleus and bind to chromatin DNA, while others remain in the cytoplasm: the factors that determine how much movement remains unidentified.
Anti-DNA antibodies form immune complexes with DNA in vivo in the plasma or in vitro in a culture medium. Although these immune complexes would be trimmed using DNase, when researchers use pure antibodies, they must be washed thoroughly in a high salt buffer and/or alkaline buffer
[8][9]. Because the ratio of absorbance at 260 nm and 280 nm changes only slightly but significantly before and after washing, it is speculated that, for example, without sufficiently thorough washing, short oligonucleotides could remain attached to the antigen-binding cleft of the antibodies purified using the protein G column. Even after the preparation of ultra-purified antibodies, however, they still bind again to DNA in the medium or on the cell surface when added to cell cultures. Therefore, even very highly purified anti-DNA antibodies should be considered immune complexes in most studies, even without the addition of exogenous DNA.
In parallel with the discovery of various intracellular nucleic acid sensors, it has been suggested that the DNA that enters the cells accompanying anti-DNA antibodies stimulates Toll-like receptors (TLRs) or other nucleic acid sensors expressed in the endosome or in the cytosol, leading to the production of cytokines relevant to lupus pathogenesis
[10][11]. In line with this, it was shown that the mouse monoclonal antibody 2C10, which specifically recognizes dsDNA and does not bind to single-stranded (ss)DNA, enters the nucleus of PBMCs from healthy subjects and induces the expression of cytokines commonly implicated in lupus, including IFN-α, IFN-β, TNF-α, IL-1β, and MCP-1
[12]. The internalization of 2C10 is significantly inhibited by the macropinocytosis inhibitor cytochalasin D but not by an Fcγ-receptor blocker. Cytokine expression was suppressed by cytochalasin D and the TLR-9 inhibitor chloroquine. In addition, the NLRP3 inhibitor shikonin suppressed the secretion of certain cytokines, including IL-1β. These results suggest that 2C10 was endocytosed mainly by monocytes via macropinocytosis, and the accompanying DNA ligated TLR-9 in the endosome, and after leaking into the cytosol, stimulated AIM-2.
Clinical phenotypes of NPSLE are diverse and are classified into neurological syndromes (including headache, seizure disorders, and cerebrovascular disease) and diffuse psychiatric or neuropsychological syndromes (including cognitive impairment, mood disorder, anxiety disorder, and psychosis)
[13]. At least some neurological symptoms are ascribed to the pathological effects of antiphospholipid antibodies (aPL) on the vascular system. Although it is hypothesized that some autoantibodies are involved, the pathogenetic role of autoantibodies in diffuse psychiatric or neuropsychological syndromes remains undefined
[14]. In addition to the blood–brain barrier, however, several other interfaces may serve as sites of antibody transfer into the central nervous system (CNS), such as the meningeal barrier, the glymphatic pathway, and the blood–cerebrospinal fluid barrier; the permeability of these barriers is considered to increase under pathological conditions
[13]. It is noteworthy that Stamou et al.
[15] documented the internalization of IgG-anti–IgG immune complexes by newborn rat hippocampal cells via Fcγ receptors. Based on these findings, recently, whether 2C10 enters cells of the CNS was tested and found that it does enter the nucleus of rat astrocytes, but not neurons, in in vitro cultures . The effects of 2C10 internalization on the function of astrocytes have not yet been determined, but given the pivotal role of astrocytes in regulating brain activity , they might be relevant to the pathogenesis of diffuse psychiatric or neuropsychological syndromes in neuropsychiatric SLE.
Another monoclonal anti-DNA antibody, WB-6, which is cross-reactive with dsDNA, ssDNA, and cardiolipin-β
2GPI, was observed to enter normal monocytes and induce tissue factor expression
[9][16]. Since internalization was diminished by the pretreatment of cells with DNase 1, WB-6 was suggested to enter cells by binding to the cell surface’s DNA. As a result, WB-6 stimulated not the TLR-4 axis, which has been suggested to be the major route in previous studies
[17], but the TLR-9 pathway, leading to tissue factor expression and a prothrombotic state in a mouse model
[18].
Using the well-studied mouse monoclonal anti-DNA antibodies 3D8 and 3E10, molecular mechanisms of cell penetration have been explored and reviewed in detail
[19]. Briefly, following the binding to the cell surface, the heparan sulfate proteoglycan, 3D8, is engulfed into early endosomes and then dissociates from the heparan sulfate, changes its conformation, and escapes into the cytosol. By contrast, 3E10 is proposed to enter the cells via a mechanism not involving endocytosis but in a manner dependent on equilibrative nucleoside transporter 2 (ENT2). ENT2 is an integral membrane protein widely expressed in most cell types, playing a role in transporting nucleosides. The knockdown of ENT2 or adding the ENT2 inhibitor dipyridamole reduces the penetration of 3E10 into the cell. Further, 3E10 traffics to the nucleus via an uncertain mechanism. It is noteworthy that in a mouse model, a dimeric scFv structural 3E10 variant (designated DX1) was suggested to be transcytosed through the endothelial cells of the brain and, thus, cross the blood–brain barrier
[20]. Dipyridamole reduced the transfer of DX1. Such a study aimed to develop antibody-based immunotherapy for brain tumors could also be relevant to the pathological mechanism of NPSLE. Antibody transcytosis across the brain’s endothelial cells is a hot topic
[21], and it would be intriguing to explore the molecular mechanisms of how DX1 interacts with ENT2 and enters and exits the brain’s endothelial cells.