Antimalarial agents have a long history of efficacy in systemic lupus erythematosus (SLE). Initially understood to treat skin disease, they are known to also relieve arthritis and pleurisy. In addition to these therapeutic benefits, they likely have a role in prophylaxis of disease activation. Should all SLE patients be taking antimalarials? The answer to this requires more information regarding its prophylactic efficacy and is complicated by continuing concerns about retinal toxicity. The clinical use of antimalarial drugs might be significantly aided by better understanding of how they work. We will discuss here advances in several areas that provide a basis for understanding how antimalarials work in SLE and how they might be better utilized in the future.
Over the last several years, a paradigm shift in understanding the importance of the innate immune system in SLE has been driven by the recognition that a class of pattern recognition receptors (PRRs), collectively known as Toll-like receptors (TLRs), likely contributes to pathogenesis. TLRs are best known for their ability to discriminate microbial macromolecules from host tissue and thereby rapidly activate the innate immune system. However, it has become increasingly apparent in recent years that select host molecules, especially nucleic acids, can serve as endogenous TLR ligands (for review, see refs.1 and2), perhaps promoting responses to damaged tissue (3). Notably, these receptors are harbored in intracellular compartments and therefore not normally accessible to circulating nucleic acids. However, receptors expressed by B cells (surface IgM), dendritic cells (DCs) (Fcγ receptor [FcγR]), or other antigen-presenting cells can act as key intermediates, trafficking nucleic acids down a pathway that eventually intersects the relevant TLR-containing compartment.
The first indication that PRRs might be involved in SLE pathogenesis came from a series of elegant studies carried out by Ronnblom and colleagues. Ronnblom and Alm demonstrated that immune complexes (ICs) containing DNA or RNA, present in SLE sera, could trigger the production of high levels of type I interferon (IFNα) (4). The IFN-producing cells were subsequently found to be a specific class of DC, now referred to as plasmacytoid DCs, and the relevant receptor was identified as FcγRIIa (5). Since other kinds of ICs did not elicit this kind of IFNα response, the results of these studies indicated a unique function of nucleic acids in activation of the innate immune system. Although CpG motifs found in microbial DNA are particularly active in stimulating TLR-9 responses, mammalian DNA and RNA in ICs can, in the proper context, also efficiently engage TLR-9 and TLR-7, respectively. Importantly, in the past 2 years it has become apparent that the ability of ICs to activate plasmacytoid DCs depends on FcγR-mediated delivery to a cellular compartment containing TLR-9 or TLR-7 (6–9). Relevant ICs incorporate autoantibodies that are directly reactive with DNA or RNA, or autoantibodies specific for proteins complexed with DNA or RNA, such as SmD or Ro.
If ICs play such an integral role in SLE pathogenesis, then key to understanding the disease process is understanding the factors that contribute to the production of the relevant autoantibodies. Here again it appears that TLRs might have an important role. B cells that express an immunoglobulin transgene receptor specific for IgG2a can be activated by ICs in which IgG2a autoantibodies target DNA, RNA, or a DNA or RNA binding protein. This response is also dependent on TLR-9 and/or TLR-7 and blocked by inhibitors of the TLR-7/9 signaling cascades (10, 11). This same B cell receptor/TLR coengagement paradigm applies to the activation of DNA-reactive B cells (12) and is reflected, in vivo, by the failure of TLR-9–deficient autoimmune-prone mice to produce antibodies reactive with double-stranded DNA (13). Such B cell autoantibody responses may be triggered by the availability of excess cell debris that has not been adequately cleared by the scavenger system. Of note, B cell responses to RNA or RNA-associated autoantigens are markedly enhanced by IFNα, through up-regulation of TLR-7 (11, 14). Thus, one can envision a model in which IFNα produced in response to a viral infection promotes autoantibody production if autoantigen is available; autoantibodies induced during the infection might then form ICs that activate plasmacytoid DCs, leading to additional production of IFNα and autoantibody.
There are other data, as well, that support the notion of an important role for IFNα in SLE. Elevated serum IFNα levels have been observed in patients with SLE (15). Type I IFN therapy for nonautoimmune disorders can cause the production of antinuclear antibodies and, in rare cases, more overt symptoms of SLE (16, 17). More recently, gene expression profiling has revealed a strong type I IFN signature in a high percentage of SLE patients (18–20). Further, IFN and IFNα markers identified in gene expression studies have recently been shown to correlate with clinical disease activity (21, 22). Type I IFNs may mediate several features of SLE pathogenesis. In addition to stimulating B cell maturation, type I IFNs promote activation of monocytes into cells with features of DCs thought to amplify the autoimmune process (23).
The report by Costedoat-Chalumeau et al in this issue of Arthritis & Rheumatism, showing that low blood levels of hydroxychloroquine (HCQ) are associated with more frequent SLE flares (24), is of particular interest because antimalarial agents inhibit activation of intracellular TLRs (TLR-3, -7, and -9). As indicated above, unlike many of the TLRs, the nucleic acid–binding TLRs do not reach the cell surface. Instead, after translocating from the endoplasmic reticulum, these TLRs bind nucleic acid ligands in the late endosome/early lysosome (25). The acidic lysosomal environment is favorable to binding of nucleic acids to intracellular TLRs (26). Antimalarial agents target microsomes, stabilize the microsomal membrane, and disrupt proper endosomal maturation and acidic pH, blocking TLR interaction with nucleic acid ligands (27). Although it is impossible to directly extrapolate in vitro activities of antimalarials to in vivo effects, it is notable that the in vivo concentrations that were associated with decreased frequency of subsequent flare in the study by Costedoat-Chalumeau and colleagues (>1,000 ng/ml) are in the same range as those shown to block intracellular TLRs in vitro (10). In view of the increasing data implicating type I IFNs in SLE pathogenesis, these observations strongly support the notion that the key activity of antimalarial drugs in SLE is inhibition of TLR activation.
This is not the first mechanism suggested for antimalarial agents, which have multiple biologic activities in vitro (27). Antimalarials have long been known to interfere with antigen presentation by affecting lysosomal acidification. Apparently unrelated to their effects on lysozymes, antimalarials have other profound effects on macrophages, inhibiting production of inflammation mediators such as interleukin-1 and interleukin-6. These or other in vitro activities might also be important in SLE or perhaps more important in rheumatoid arthritis, in which intracellular TLRs are not likely to play such a prominent role.
These observations lead to several easily testable speculations regarding the mechanism of action of antimalarial agents and how they might be used more effectively in SLE: Do antimalarials at currently accepted dosages inhibit IFN-regulated genes? Does altered metabolism of the antimalarial agent HCQ lead to drug resistance? Do decreased serum HCQ levels correlate with incomplete inhibition of IFN-regulated genes? Is retinal toxicity related more closely to dosage or to serum levels, and can dosages be safely adjusted based on serum concentrations or other parameters?
An exciting aspect of these questions is the potential to link IFN-regulated gene expression to clinical management. Microarray gene expression analysis, with its remarkable informational content and rapid decline in price, promises to become a clinically available tool in the near future. Serum IFN is often undetectable in SLE patients despite striking elevations in levels of IFN-regulated genes, likely making study of the latter a more reliable means of detecting IFN activity. If blocking of TLR activation, and thereby IFN, accounts for the efficacy of HCQ, then the clinician might titrate the dosage of this or other medications based on successful inhibition of IFN-regulated gene expression, rather than serum levels.
Clarifying whether antimalarial agents work by TLR inhibition in SLE might not only increase our sophistication in dosing of these drugs, but would provide proof of principle for the next generation of TLR and IFNα modulators likely to arrive soon in clinical trials. The goals with these agents will be to provide more effective and safer ways to inhibit TLRs, but before they become available, we may already have effective agents. Can we show that effective TLR inhibition with antimalarial agents blocks SLE disease activity and inhibits future flares? Such data would provide exciting new links between the scientific advances in innate immunity and therapy for SLE.