Oiling the wheels of autoimmunity

Oily substances in the skin have now been shown to contain structures that activate a population of skin-homing, self-reactive T cells. The responses of these immune cells may contribute to local defences, but also to autoimmune disease.

Immunology students are taught that the immune system responds to foreign entities while remaining tolerant to 'self' structures. This is not strictly true, however, because there are specialized populations of immune cells that are self-reactive. Such cells have the potential to initiate undesirable autoimmune reactions, so their existence raises several questions. What are the origin and structure of the self-antigens to which these cells respond, and how is this potentially dangerous self-recognition regulated? Reporting in Nature Immunology, de Jong et al.1 identify hydrophobic self-antigens in the skin that are recognized in an unusual manner by a specialized subset of skin-resident immune cells.

B and T cells are the white blood cells responsible for immune recognition in the adaptive immune system. Populations of self-reactive T cells reside in or near the epithelial surfaces of the skin and intestine2,3, where there are rich concentrations of microorganisms. Although it may seem paradoxical that self-reactivity is prevalent where microbes are abundant, it is possible that self-reactivity at these surfaces involves 'sentinel' immune cells that can rapidly respond to general signs of cellular stress or barrier disruption, without the need for specific recognition of microbes.

The antigens recognized by most T cells are peptides that are displayed on the surface of other cells, bound in the groove of antigen-presenting proteins of the major histocompatibility complex (MHC) family. Lipid antigens, by contrast, are bound by the hydrophobic grooves of CD1 antigen-presenting proteins, which are related to the MHC proteins. Humans have four CD1 proteins4: CD1a, CD1b, CD1c and CD1d. Some self and microbial lipid antigens that bind to CD1 proteins have been identified, but research on this antigen-presentation system has mostly been restricted to CD1d molecules.

De Jong et al. concentrated on T cells that recognize antigens bound to CD1a, which are more prevalent in human blood than T cells recognizing other CD1 proteins5,6. CD1a-reactive T cells are also found in the skin; when stimulated, these cells produce IL-22 (ref. 5), a cytokine protein involved in microbial defence and in inducing the proliferation of skin cells called keratinocytes. Moreover, Langerhans cells, which are antigen-presenting cells that reside in the skin's epidermal layer, express particularly high amounts of CD1a.

The authors show that a CD1a molecule purified from a human cell line activates CD1a self-reactive T cells by binding to their antigen receptor. The antigen-binding grooves of MHC and CD1 proteins are always filled, but the proteins do not discriminate self from non-self — this is the job of the antigen receptors that react to them. Therefore, the authors sought to uncover the CD1a-bound antigens that triggered the self-reactive T cells. Using mass-spectrometry analysis, they found more than 100 molecules corresponding in mass to lipids (glycolipids and phospholipids) with different chain lengths and degrees of unsaturation. They then used several strategies to detect antigenic activity in this melange, including tests of synthetic molecules that had the same molecular masses as the material eluted from CD1a, and of partially purified lipids from various cell types.

The authors found that epidermal lipids bound to CD1a were more stimulatory for the CD1a-reactive T cells than lipids from other cell types. More specifically, they showed that CD1a-reactive T cells recognized highly hydrophobic compounds such as squalene and triacylglyceride, which are found naturally in the skin, when these were bound to CD1a. This reactivity was selective, and other hydrophobic molecules such as cholesterol were not recognized by the cells.

Although CD1 molecules have a hydrophobic antigen-binding groove, the completely hydrophobic character of the antigens presented by CD1a is surprising. T-cell antigen receptors typically recognize a composite structure of the antigen-presenting molecule and the antigen, with the exposed portion of the antigen participating in engaging the T-cell receptor (Fig. 1a). The exposed portion of lipid antigens is normally hydrophilic, containing a sugar or phosphate group, with the hydrophobic chains buried in the CD1 groove7. But the CD1a-binding self-reactive T cells did not obey this rule, because they did not require an exposed hydrophilic portion of the stimulatory lipid-containing antigen.

Figure 1: Skin-antigen recognition by self-reactive T cells.

a, T cells recognize peptide antigens bound to MHC molecules on the surface of antigen-presenting cells, or lipid antigens that are presented by CD1 proteins. In both cases, the antigen protrudes from the groove of the antigen-presenting molecule to engage the T-cell receptor. De Jong et al.1 show that some T cells that react to CD1a molecules (a subset of the CD1 family) recognize oily substances found in sebum — a hydrophobic layer secreted onto the outermost layer of the skin. These self-antigens nestle deep within the CD1a groove, such that there might be direct contact only between CD1a and the T-cell receptor. b, The authors propose that skin-barrier disruption, through trauma or infection, allows Langerhans cells, which express CD1a, to acquire oily antigens from sebum and move from the skin's epidermis to the dermis. There, they make contact with self-reactive T cells and activate them to produce cell-signalling molecules, such as IL-22, that promote an immune response to barrier disruption without requiring specific recognition of invading microbes.

In fact, binding of lipids bearing hydrophilic groups to CD1a inhibited the response of these T cells, presumably by competing with more strongly hydrophobic antigens for binding into the CD1a groove. Therefore, it seems that the self-reactive T-cell antigen receptor requires a view of CD1a that is unimpeded by exposed hydrophilic groups; the bound lipid may simply permit or stabilize CD1a into the correct conformation. As a consequence, rather than recognizing single compounds with high specificity, these T cells can be stimulated by a range of highly hydrophobic substances that fit in the CD1a groove.

The skin forms a barrier to microbes through the generation of sebum — a highly hydrophobic substance synthesized in the sebaceous glands and secreted onto the outermost layer of the skin. Using microdissected sebaceous glands, de Jong et al. demonstrated that sebum is highly stimulatory for CD1a-dependent self-reactive T cells, and that it is rich in antigenic compounds, such as squalene. However, sebum is not typically in contact with the underlying dermal and epidermal layers that contain T cells and Langerhans cells. In normal skin, this physical separation may prevent CD1a-binding self-reactive T cells from being constantly exposed to their antigens. But disruption of the skin barrier by injury, infection or inflammation might allow sebum contents to permeate the epidermis and bind to CD1a-expressing Langerhans cells, thereby stimulating T-cell responses (Fig. 1b). Although this may aid general immune defences, in cases of prolonged barrier disruption, constant exposure of immune cells to sebum could contribute to autoimmune skin diseases such as psoriasis and atopic dermatitis.

Interestingly, squalene is currently used as an immune booster (adjuvant) to enhance the efficacy of vaccines and immunotherapies, and as a carrier for topical delivery to hair follicles of drugs for treating hair loss8. An autoimmune syndrome has also been described that is induced by adjuvants, including squalene9. It is possible that activation of CD1a-binding self-reactive T cells contributes to this compound's immune-stimulating effects. Certainly, further investigation of the regulation of these T-cell responses to skin oils is warranted, both for understanding immunity and autoimmunity and in light of the increasing therapeutic use of such agents.


  1. 1

    de Jong, A. et al. Nature Immunol. 15, 177–185 (2014).

  2. 2

    Cheroutre, H., Lambolez, F. & Mucida, D. Nature Rev. Immunol. 11, 445–456 (2011).

  3. 3

    Witherden, D. A. & Havran, W. L. J. Leuk. Biol. 94, 69–76 (2013).

  4. 4

    Salio, M., Silk, J. D. & Cerundolo, V. Curr. Opin. Immunol. 22, 81–88 (2010).

  5. 5

    de Jong, A. et al. Nature Immunol. 11, 1102–1109 (2010).

  6. 6

    de Lalla, C. et al. Eur. J. Immunol. 41, 602–610 (2011).

  7. 7

    Zajonc, D. M. & Kronenberg, M. Curr. Opin. Struct. Biol. 17, 521–529 (2007).

  8. 8

    Aljuffali, I. A., Sung, C. T., Shen, F.-M., Huang, C.-T. & Fang, J.-Y. AAPS J. 16, 140–150 (2014).

  9. 9

    Vera-Lastra, O., Medina, G., Cruz-Dominguez, M. Del P., Jara, L. J. & Shoenfeld, Y. Expert Rev. Clin. Immunol. 9, 361–373 (2013).

Download references

Author information

Correspondence to Mitchell Kronenberg or Wendy L. Havran.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Kronenberg, M., Havran, W. Oiling the wheels of autoimmunity. Nature 506, 42–43 (2014).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.