Protein TOLLIP puts the brakes on sugar metabolism that fuels psoriasis in skin cells
Mouse and cell studies suggest TOLLIP restrains PKM2-driven glycolysis in keratinocytes, and delivering it by AAV eased psoriasis-like disease in mice.
Summary
Psoriasis skin cells shift to a fast, sugar-burning metabolism called aerobic glycolysis, which feeds overgrowth and inflammation. This study identifies TOLLIP, an adaptor protein, as a natural brake on that process. TOLLIP levels are higher in human psoriatic lesions and track with disease markers. Even so, mice lacking Tollip developed worse imiquimod-induced psoriasis-like disease. Mechanistically, TOLLIP binds the glycolytic enzyme PKM2 through its C-terminal CUE domain (amino acids 179–274), promotes PKM2 ubiquitination, dampens its enzyme activity and reduces glycolysis. This limits keratinocyte proliferation and inflammatory signaling. Delivering full-length Tollip or only the CUE fragment by AAV vectors improved disease in mice. The work is preclinical, but it points to the TOLLIP–PKM2–glycolysis axis as a potential therapeutic target.
Detailed Summary
Psoriasis affects roughly 2–3% of people worldwide and is linked to cardiovascular disease, metabolic syndrome, fatty liver and psychiatric illness. Skin cells called keratinocytes drive a self-amplifying loop with immune cells, mediated by IL-17, IL-22 and TNF-α. Increasingly, researchers see a metabolic shift toward aerobic glycolysis (the Warburg effect) as a core part of this loop, and the glycolytic enzyme PKM2 is a known contributor. What controls PKM2 in psoriatic skin has been unclear.
The authors tested whether Toll-interacting protein (TOLLIP), a 274-amino-acid adaptor involved in inflammatory signaling, autophagy and trafficking, restrains this metabolic reprogramming. They first mined public human datasets. In two bulk skin transcriptome cohorts (GSE13355 and GSE30999), TOLLIP mRNA was higher in lesional skin than in non-lesional or healthy skin. Single-cell data from five patients showed the rise in keratinocytes. TOLLIP expression separated lesions from healthy skin with AUCs of 0.81 and 0.74. It correlated positively with KRT16, S100A9, IL17A and CXCL10, and it fell in patients treated with etanercept or brodalumab. The team also examined skin from five patients and five controls by immunohistochemistry. They followed TOLLIP in imiquimod (IMQ)-treated mouse skin and in HaCaT keratinocytes exposed to a four-cytokine cocktail (IL-17A, IL-22, TNF-α, OSM).
Functional experiments, as described in the paper's abstract and introduction, point the other way from the expression data. Tollip knockout mice had worse epidermal thickening, inflammation and disease severity in the IMQ model. Re-expressing Tollip in keratinocytes improved the exacerbated phenotype. In HaCaT cells, TOLLIP overexpression reduced cytokine-induced proliferation and inflammation, and knockdown had the opposite effect. Primary keratinocytes from knockout mice behaved consistently. The rise of TOLLIP in lesions therefore looks like a protective, compensatory response rather than a driver of disease.
On mechanism, the authors used published interactome data to show that TOLLIP binds PKM2 through its CUE domain (amino acids 179–274). This promotes PKM2 ubiquitination, reduces its enzymatic activity and lowers glycolytic flux. In mice, AAV-delivered full-length Tollip or the 179–274 fragment alone ameliorated psoriasis-like disease. This suggests that a small functional domain could be enough for therapeutic effect.
The work is entirely preclinical. The IMQ model captures only part of human psoriasis. The excerpt supplied here covers only the first results section in detail, so the metabolic assays, ubiquitination data and AAV outcomes are described here as the authors summarize them, not from the full figures. Safety, delivery to human skin, durability and efficacy against established disease have not been tested. Whether TOLLIP loss or PKM2 dysregulation occurs in patients is also unknown.
Key Findings
- TOLLIP mRNA is higher in psoriatic lesions than in non-lesional or healthy skin, with ROC AUCs of 0.81 and 0.74 in two cohorts.
- TOLLIP expression correlates with KRT16, S100A9, IL17A and CXCL10 and falls after etanercept or brodalumab treatment.
- Tollip-knockout mice develop more severe imiquimod-induced psoriasis-like disease; keratinocyte-specific Tollip re-expression rescues it.
- TOLLIP binds PKM2 via its CUE domain (aa 179–274), promoting ubiquitination, reducing PKM2 activity and curbing glycolysis.
- AAV delivery of Tollip or the 179–274 fragment improved psoriasis-like disease in mice.
Methodology
The authors combined analysis of public human transcriptomic datasets (bulk skin and single-cell RNA-seq) with immunohistochemistry on patient skin (n=5 per group). They then used the IMQ mouse psoriasis model with Tollip-knockout mice, keratinocyte-specific rescue and AAV delivery, plus HaCaT cell and primary keratinocyte gain- and loss-of-function experiments with a cytokine cocktail. Mechanistic work covered the TOLLIP–PKM2 interaction, PKM2 ubiquitination and enzyme activity, and glycolysis.
Study Limitations
Evidence comes from mouse models and cultured keratinocytes; the IMQ model does not fully reproduce human psoriasis, and human tissue validation was small (n=5 per group). The text provided was truncated after the first results section, so the glycolysis, ubiquitination and AAV findings are summarized from the abstract and introduction rather than detailed data. Gene-therapy safety, skin-targeted delivery and effects on established disease remain untested.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: