Editorial: Toll-like Receptor 7: More Than Skin Deep?

Authors


Skin rash is one of the most common manifestations of lupus and is of special significance for clinical investigations for 2 reasons. First, it is the only manifestation of lupus that is linked to an environmental trigger—sunlight (ultraviolet [UV] light) exposure. Second, UV light exposure is associated with both cutaneous and systemic disease flares. High doses of UV light from sunlight and artificial sources, such as psoralen ultraviolet A phototherapy and tanning booths, are reported to both precipitate and exacerbate systemic disease activity in vital internal organs, including the central nervous system and kidneys ([1]). Cutaneous inflammation, therefore, not only provides an important link between the skin and other manifestations of lupus but also provides clues to pathogenic mechanisms that are potentially responsible for systemic lupus.

Early research efforts suggested that UV light–induced death of keratinocytes and perhaps other cells is the initiating event that induces skin inflammation ([2]). However, progress in this area of research was hindered by a lack of understanding of the mechanisms of cell death and the pivotal role of the innate immune system in response to self antigens. Over the last decade, there have been major advances in understanding the role of intracellular sensors such as Toll-like receptors (TLRs), retinoic acid–inducible gene 1–like receptors, nucleotide-binding oligomerization domain–containing protein 1–like receptors, and AIM-like receptors that respond to nucleic acid ligands by stimulating production of inflammatory cytokines ([3]), as well as the nature of the immune response to dead and dying cells. In this issue of Arthritis & Rheumatology, Yokogawa et al ([4]) report a possible link between local skin inflammation (in this case stimulated by long-term repetitive activation of a single TLR, TLR-7) and systemic manifestations of lupus.

TLR-7 is an intracellular endolysosomal TLR that is most strongly expressed in B lymphocytes and plasmacytoid dendritic cells (PDCs) but is also expressed by myeloid cells. TLR-7 signaling is triggered by single-stranded RNA, which may be derived from viruses, or by certain bacteria. In addition, self RNA or RNP antigens complexed to autoantibodies to form immune complexes are potent agonists of TLR-7 following phagocytosis by PDCs ([5]). Activation of TLR-7 stimulates the production of type I interferon (IFN) and other proinflammatory cytokines, including interleukin-12 p40 subunit (IL-12p40), tumor necrosis factor (TNF), and IL-6. Genome-wide association studies as well as the observed correlation between anti-RNP antibodies and the IFN signature suggest that TLR-7 is involved in the pathogenesis of human systemic lupus erythematosus (SLE).

Functional and genetics studies also implicate TLR-7 in murine models of lupus. For example, MRL/lpr mice deficient in Tlr7 (but not Tlr9) are protected against disease, whereas a lupus-like disease develops in mice overexpressing TLR-7 (TLR-7–transgenic mice). Furthermore, an additional copy of TLR-7 was shown to be responsible for the accelerating effect of the Yaa mutation in BXSB mice. Of particular relevance to the study by Yokogawa et al is the observation that among the many strains of lupus-prone mice exposed to UVB light, only BXSB mice developed severe lupus ([6]). The finding that excess mortality occurred predominantly among male mice, which express an extra copy of Tlr7, suggests that the TLR-7 gene rather than other BXSB genes plays a role in promoting UV light–mediated systemic disease.

Based on the above-mentioned associations between TLR-7 and both human and murine lupus, Yokogawa et al examined the response of mice to administration of imiquimod. Imiquimod is a small molecule compound in the imidazoquinoline family that activates TLR-7 and TLR-8, resulting in potent antitumor and antiviral effects when applied to skin cancer– and viral infection–associated lesions in humans ([7]). Imiquimod treatment stimulates local production of cytokines such as type I IFN, TNF, IL-12, IL-1, and IL-6, recruits inflammatory cells (including PDCs), and culminates in activation of the adaptive immune response characterized by a dominance of Th1 cell and, to a lesser extent, Th17 cell responses ([8]). Yokogawa and colleagues first investigated the response of FVB/N mice to long-term topical application of imiquimod. Long-term exposure resulted in the development of a striking lupus-like disease characterized by the presence of autoantibodies, dermatitis, and nephritis. Furthermore, topical administration of imiquimod led to early mortality beginning at 8 weeks, and all imiquimod-treated mice died by 15 weeks after initiation of imiquimod exposure. The activation of immune cells in the spleen, especially the expansion of the myeloid cell population that has been previously associated with TLR-7 overexpression, demonstrates a potent systemic effect of topical application of imiquimod. This effect was not uniquely attributable to imiquimod (or the carrier vehicle), because similar effects were observed following topical application of a closely related compound, R848. In addition, the investigators demonstrated that TLR-7 was necessary for the development of disease, because features of lupus were not observed in TLR-7–deficient mice.

In a previous study in which topical application of imiquimod in mice resulted in scaly skin lesions associated with epidermal proliferation and accumulation of neutrophils, the authors proposed imiquimod-induced dermatitis as a mouse model of psoriasis rather than lupus ([8]). This raises the question of what could explain the different outcome in the study by Yokogawa et al. First, most of the previously described psoriasis models involved use of short-term daily topical application rather than protracted topical application of the drug. Thrice-weekly application for 4 months allows long-term activation of the immune system through TLR-7/8. In addition, the FVB/N mouse strain has not previously been used to study lupus and is better known for its susceptibility to IgE-mediated disease, chemically induced squamous cell carcinomas, and retinal degeneration, suggesting interesting but as-yet-unidentified host susceptibility factors.

To address whether the lupus-like phenotype induced by imiquimod exposure was restricted to the FVB/N mice on a mixed genetic background, the authors used the same long-term topical imiquimod treatment strategy in BALB/c and C57BL/6 (B6) mice and demonstrated immune cell expansion (splenomegaly) in all 3 strains. Although B6 mice were not described in further detail, imiquimod did induce progressive glomerulonephritis, with increases in serum creatinine levels and proteinuria in both FVB/N and the BALB/c mice. BALB/c mice are also more “Th2-like” and are more susceptible to a lupus-like disease following administration of pristane, thus raising the interesting question of how Th cells are skewed in the imiquimod model of lupus described here.

TLR-7–overexpressing mice on a B6 genetic background produce antibodies to RNA, develop immune deposits in the kidney, but likely die of inflammatory liver disease or cytopenia ([9]). Similarly, Yokogawa et al observed that imiquimod-treated mice had a high frequency of cytoplasmic antinuclear antibodies (ANAs), which usually reflects anti-RNA activity. Despite the presence of anti-DNA antibodies, the rather weak IgG and C3 staining in the glomeruli raises the question of whether immune complex–mediated glomerulonephritis was the proximal cause of death in these imiquimod-exposed mice. In future studies, it will be informative to correlate hepatic inflammation, autoantibody titers, and autoantibody specificity with nephritis and mortality rates.

A major conclusion of the study by Yokogawa et al is that the skin is “the primary organ that allows TLR-7 agonists to initiate SLE.” This conclusion is based on their studies comparing the severity of lupus in mice receiving topical imiquimod and those receiving imiquimod by the intraperitoneal route. Mice receiving imiquimod intraperitoneally had lower ANA titers, no anti–double-stranded DNA (anti-dsDNA) antibodies, and less severe histologic changes in the kidney and liver compared with mice treated with topical imiquimod. Although the greater severity of disease in mice receiving topical versus systemic imiquimod is potentially an important finding, other interpretations are possible. First, different drug carriers were used in the topical formulations, and vehicle creams have been implicated in some of the effects of imiquimod ([10]). Second, and probably more important, topical imiquimod was administered 3 times weekly at a dose of 1.25 mg, whereas with intraperitoneal delivery, approximately one-tenth the dose (125 μg) was injected 3 times weekly. Systemic absorption has been observed after long-term administration of 5% imiquimod cream in humans ([7]), and splenomegaly has previously been observed in mice receiving topical therapy. For these reasons, quantification of serum drug levels after topical and systemic drug administration would be necessary to confirm that the skin is truly the preferred site for systemic disease activation following long-term stimulation of TLR-7.

PDCs are strongly implicated in the pathogenesis of cutaneous lupus erythematosus and SLE and are abundant in the skin biopsy specimens of patients with cutaneous lupus erythematosus ([11]). Although previous studies have shown that short-term cutaneous exposure to imiquimod induces a mixed cellular inflammatory response that includes PDCs ([8]), Yokogawa et al demonstrated that PDCs were also recruited to the dermis following long-term exposure to imiquimod. To evaluate the potential functional role of PDCs in the pathogenesis of imiquimod-induced lupus, the authors quantified the messenger RNA expression of Ifna and Mx1 (PDCs are the main producers of IFNα, and Mx1 is an IFN-stimulated gene) and observed that expression of these genes was increased in the spleen and the kidneys of treated mice, which again emphasizes the systemic effects of topical administration of the drug.

To further evaluate the requirement for PDCs in imiquimod-mediated lupus, PDCs were depleted by weekly administration of an antibody to plasmacytoid dendritic cell antigen 1 (PDCA-1) during topical imiquimod treatment. Depletion of PDCs led to a significant reduction in serum anti-dsDNA antibody levels. Whether the incomplete suppression of anti-dsDNA antibody generation indicates that anti-DNA production is not completely dependent on PDCs or whether transient reconstitution of PDCs between depletion treatments was sufficient to drive autoantibody formation is not completely clear. An important caveat of these studies is that anti–PDCA-1 is known to deplete other myeloid cells in the setting of inflammation; therefore, disease amelioration cannot be ascribed to PDCs alone. In addition, a detailed evaluation of organ-specific involvement of the kidneys, spleen, thymus, and skin would have been interesting to determine, because PDC depletion is considered to be a potential therapy for SLE and other autoimmune disorders associated with an IFN signature. Whether PDCA-1 depletion therapy leads to alterations in other key cytokines such as IFNγ and IL-17 would also be useful to know.

Remarkably, skin at sites distant from that exposed to topical application of imiquimod demonstrated IgG deposition at the dermal–epidermal junction, resembling the “lupus band” that is seen in cutaneous lupus in both clinically involved and uninvolved skin. Of further relevance to human lupus, when mice that had been sensitized by imiquimod on one ear were exposed to UV light, increased inflammation was also observed in the untreated ear. This finding reinforces the systemic nature of the imiquimod-mediated effects and also the notion that one inflammatory stimulus may prime for a second insult.

A pivotal unanswered question raised by the Yokogawa study is whether the pathology observed at skin sites distal from the site of topical application of imiquimod as well as the systemic manifestations described in this study are attributable to absorption of the drug into the circulation with consequent systemic exposure, or are attributable to local activation of immune cells in the skin followed by migration to distal sites. It has been known for some time that activation of immune cells in one part of the gut leads to trafficking to sites distal from the initial site—the “common mucosal immune system.” A similar phenomenon was observed in an experimental arthritis model in which arthritis that was initiated in one joint spread to the contralateral joint. If, as the authors propose, systemic trafficking is readily initiated following TLR activation in the skin, much can be learned about how systemic lupus can be exacerbated by UV light exposure and possibly other skin provocations.

Another important question is whether the phenomenon reported by Yokogawa et al is unique to TLR-7. The role of TLRs in cutaneous lupus has not been systematically studied, but one report suggests that, at least for the induction of inflammatory cytokines such as TNF, TLR-3 is selectively activated by UV light–mediated crosslinking of U1 RNA ([12]). Despite the apparently protective role of TLR-9 in some mouse models of lupus, it remains unknown whether local release of DNA from keratinocytes undergoing apoptosis or neutrophils undergoing NETosis stimulates inflammation through TLR-9 or other DNA sensors. The studies by Yokogawa et al will no doubt prompt many experimental approaches to probe beneath the skin.

AUTHOR CONTRIBUTIONS

Drs. Elkon and Sontheimer drafted the article, revised it critically for important intellectual content, and approved the final version to be published.

Ancillary