Billo, A. | Search bioRxiv

bioRxiv2026

IGF1 modulates lesional skin inflammation in checkpoint inhibitor-induced lichen planus

Immune checkpoint inhibitor-induced lichen planus (ICI-LP) is a cutaneous immune related adverse event (irAE) that shares key clinicopathologic features with spontaneous lichen planus (LP) but differs histologically and in the sex distribution of its incidence, and may therefore reflect a distinct tissue inflammatory state. To define the cellular programs that distinguish ICI-LP from LP, we profiled lesional skin by single cell and spatial transcriptomic approaches. We found few differences in the T cell and keratinocyte compartments between ICI-LP and LP, which shared similar inflammatory signatures. Rather, the dominant transcriptional features differentiating these two eruptions occurred within the fibroblast and myeloid cell compartments. Fibroblasts in ICI-LP were enriched for IGF1, FGF7, and androgen-response-associated programs, whereas myeloid cells exhibited amplified JAK-STAT and interferon-responsive states spanning both type I and type II interferon signatures. The potential role of androgen response in shaping lichenoid inflammation was supported by a striking loss of androgen receptor expression in lesional keratinocytes by immunohistochemistry. Furthermore, using spatial RNA and transcriptomic approaches, we identified anatomically segregated IFNG, IL17A, and IL13 niches within lesional skin, suggesting that regional immune compartmentalization with differences in local immunoregulation may explain the mixed inflammatory features reported in both ICI-LP and LP. Collectively, these data indicate that ICI-LP is not simply a more inflamed form of LP, but a distinct form of the disease with more prominent inflammatory perturbations within stromal and innate immune cell populations.

immunology↗

bioRxiv2026

High Diversity Gene Libraries Facilitate Machine Learning Guided Exploration of Fluorescent Protein Sequence Space

While protein language models (PLMs) have shown great promise for protein design, their performance is fundamentally constrained by the diversity and completeness of available training data. In particular, PLMs often struggle to extrapolate to sequences that fall outside the distribution spanned by their training sets, limiting their ability to discover proteins in sparsely sampled regions of sequence space. Here we test the hypothesis that experimentally expanding training diversity can convert extrapolation into interpolation and thereby enable discovery of functional sequences beyond natural protein manifolds. Using large-scale gene synthesis and DNA shuffling, we generate libraries that span a broad region of fluorescent protein sequence space and create chimeric variants that bridge between distant homologs. Functional screening for blue fluorescence yields thousands of active variants distributed across diverse sequence lineages. Fine-tuning ProtGPT2 on this expanded dataset enables generation of diverse fluorescent proteins, including designs that extend beyond the regions occupied by known natural sequences while retaining function. This work illustrates how synthetic approaches can help address key limitations in machine learning-guided protein design, especially for small or sparsely populated protein families, by actively creating novel sequences across unexplored but functional regions of sequence space.

synthetic biology↗