Tiny but mighty: microproteins emerge as regulators of disease and future precision therapeutics

For decades, the human proteome was catalogued with an implicit assumption: real proteins are long, complex molecules produced from well-annotated genes. That assumption is now crumbling. A comprehensive review published in the journal 3 Biotech by Muhammad Azhar, Faiz-ul Hassan and Muhammad Asif Arain synthesizes a decade of discoveries showing that thousands of tiny proteins, known as microproteins, are continuously manufactured inside human cells from genomic regions once dismissed as genetic dark matter. These molecules, typically encoded by small open reading frames (sORFs) within what were labelled non-coding RNAs or overlooked stretches of canonical transcripts, are not molecular noise. According to the review, they participate in fundamental cellular processes including mitochondrial homeostasis, apoptosis, metabolic reprogramming and immune signaling, and their dysregulation is increasingly implicated in cancer, cardiovascular disease, metabolic disorders, neurodegeneration and immune-related conditions.

The technical foundation of this field rests on ribosome profiling, or Ribo-seq, a technique that captures snapshots of ribosome positions across the transcriptome at nucleotide resolution. When Ingolia and colleagues first applied genome-wide ribosome profiling in 2009, it revealed pervasive translation outside annotated protein-coding genes. Subsequent refinements, including improved library preparation protocols that expose cryptic translation events, showed that non-canonical open reading frames (ncORFs) are actively translated in mammalian cells. These ncORFs hide in 5′ and 3′ untranslated regions, within upstream open reading frames, on overlapping alternative frames of canonical mRNAs, and on long non-coding RNAs and circular RNAs. Some initiate at non-AUG start codons; the review highlights the example of POLGARF, a protein produced from an overlapping reading frame in the POLG mRNA that begins at an unusually efficient CUG codon, demonstrating that translation initiation in human cells is far more flexible than textbook models suggested.

Translation evidence alone, however, does not prove biological function, and the review is emphatic about this distinction. Proteogenomics, which combines mass spectrometry-based peptide detection with genome annotation, has provided direct biochemical confirmation of many sORF-encoded polypeptides in human cell lines and tissues. Early peptidomic studies identified dozens of such peptides, and comparative proteomic profiling has since catalogued hundreds of unannotated microproteins and alternative proteins. Yet the authors stress a persistent gap between computational prediction and experimental validation: algorithms predict vast numbers of candidate sORFs, but only a fraction have been shown to produce stable, functional peptides. Rigorous orthogonal validation, they argue, requires CRISPR-based genome editing to mutate specific sORFs without disturbing the host transcript, ribosome release assays, targeted mutational analysis, high-resolution mass spectrometry and functional studies. Discrepancies in microprotein annotation across databases remain a significant obstacle to reproducibility.

Some of the most compelling functional stories come from muscle and heart biology. The micropeptide DWORF, encoded by a transcript annotated as long non-coding RNA, enhances SERCA calcium pump activity in muscle, and in a mouse model of dilated cardiomyopathy it improved contractility and prevented heart failure. A family of SERCA-inhibiting micropeptides, including sarcolipin-related regulators, exerts widespread control over calcium signaling, while another-regulin integrates neuroendocrine signaling with SERCA2a activity to regulate cardiomyocyte calcium handling. Myoregulin and phospholemman tune pump performance in skeletal and cardiac muscle. These examples illustrate a recurring mechanistic theme: microproteins often act as fine-tuning subunits of large protein complexes, occupying interaction surfaces that classical regulatory proteins cannot access, and thereby adjusting physiological output with remarkable precision.

Mitochondria have emerged as a particularly rich source of microproteins. Humanin, the first mitochondrial-derived peptide discovered, was initially identified through its neuroprotective activity in Alzheimer’s disease models and has since been shown to influence apoptosis, insulin sensitivity and inflammatory markers in an age-dependent manner. MOTS-c, another mitochondrially encoded peptide, promotes metabolic homeostasis, reduces obesity and insulin resistance in animal models, and can translocate to the nucleus to regulate nuclear gene expression in response to metabolic stress. SHLP2 has been identified as a protective factor in Parkinson’s disease, and a naturally occurring variant of SHLP2 is associated with reduced disease risk, while the SHMOOSE microprotein, revealed through mitochondrial DNA variation studies, links mitochondrial genotype to Alzheimer’s disease biology. MOXI enhances fatty acid beta-oxidation, mitoregulin regulates fatty acid metabolism at the mitochondrial outer membrane, and LINC00116-encoded peptides link respiration to lipid metabolism. A mitochondrial microprotein has also been shown to regulate the endoplasmic reticulum stress response, underscoring how these tiny molecules coordinate communication between organelles.

In cancer biology, microproteins play roles on both sides of the oncogenic balance. Several act as tumor suppressors: the HOXB-AS3 peptide suppresses colon cancer growth by blocking oncogenic transcriptional elongation, a peptide encoded by circular LINC-PINT represses elongation in glioblastoma, CIP2A-BP inhibits triple-negative breast cancer progression, and ASRPS inhibits angiogenesis in the same tumor type. Others are oncogenic. The microprotein HDSP promotes gastric cancer progression by activating the MECOM-SPINK1-EGFR signaling axis; a microprotein derived from HCP5 drives gastric cancer through ferroptosis regulation; SMIM30 promotes hepatocellular carcinoma development by anchoring SRC and YES1 kinases to membranes and activating MAPK signaling; RASON controls oncogenic RAS signaling in KRAS-mutant cancers; and a LINC00263-encoded microprotein promotes breast cancer osteolytic bone metastasis by inducing osteoclastogenesis while inhibiting osteoclast ferroptosis. More recently described examples include MUCP1, which reprograms colorectal cancer metabolism by regulating mitochondrial succinate transport, and L3EMP, which triggers lung adenocarcinoma progression by catalysing deubiquitination of SIRT1. An oncopeptide that alters m6A RNA modification recognition by the reader protein IGF2BP1 further demonstrates that microproteins can hijack epitranscriptomic machinery.

Beyond cancer, the review documents microprotein involvement in immunity and development. A micropeptide encoded by the MIR155HG locus suppresses autoimmune inflammation by modulating antigen presentation, pointing toward therapeutic relevance in lupus and related conditions. The PC3-secreted microprotein functions as a high-affinity ligand for the CC chemokine receptor 2 and modifies the immune microenvironment in prostate cancer, and elevated PSMP levels have been linked to chronic antibody-mediated rejection in kidney transplantation. In developmental biology, the PLUM microprotein, encoded by an upstream open reading frame within the Lin28b mRNA, acts as a cytoplasmic determinant of pluripotency and embryonic development, while ELABELA, a hormone essential for heart development that signals through the apelin receptor, sustains embryonic stem cell self-renewal via the PI3K/AKT pathway. These findings suggest that microproteins are woven into the most basic circuits of life, from the earliest embryonic decisions to the maintenance of tissue homeostasis in adulthood.

Technological innovation is accelerating the pace of discovery. The review highlights artificial intelligence-assisted ORF prediction, exemplified by machine learning frameworks such as ShortStop that are trained to distinguish functional microproteins from translational noise. Single-cell translatomics now allows ribosome profiling at the resolution of individual cells, revealing tissue- and cell-type-specific microprotein expression that bulk measurements average away. Spatial proteomics approaches, including proximity-labeling methods such as MicroID, map the subcellular localizations of unannotated microproteins, providing crucial clues about function. Long-read techniques such as Ribo-STAMP simultaneously measure transcription and translation with isoform resolution, resolving which transcript variants actually produce which peptides. Integrated multi-omics pipelines that combine genomics, transcriptomics, translatomics and proteomics are reshaping how candidate microproteins are prioritized for functional study, and standardized annotation efforts are working to bring consistency to a fragmented nomenclature landscape.

The translational horizon is what makes the field so compelling for precision medicine. As naturally occurring, evolutionarily conserved molecules with defined disease associations, microproteins offer both biomarker and therapeutic opportunities. Circulating MOTS-c levels have been measured in breast cancer patients treated with metformin, mitochondrial-derived peptides correlate with age-related disease states, and microprotein signatures could eventually support disease subtyping and tailored treatment decisions. As therapeutics, microproteins or their mimetics could restore lost regulatory functions, as DWORF-like SERCA enhancers might do in heart failure, or block pathological interactions, as inhibitors of oncogenic microproteins might do in cancer. But the review is candid about the barriers. Peptide stability in circulation, unfavorable pharmacokinetics, the challenge of tissue-specific delivery, and immunogenicity all stand between laboratory discovery and clinical application, and rigorous preclinical and clinical validation remains essential. Safety assessment, including immunogenicity evaluation familiar from the broader peptide therapeutics field, will be a prerequisite for regulatory approval.

What emerges from this synthesis is a picture of biology that is richer and stranger than the standard genome annotation suggested. The boundary between coding and non-coding RNA is porous; thousands of small proteins are translated from the shadows of the transcriptome; and these molecules regulate calcium handling, mitochondrial respiration, metabolic flux, immune signaling and embryonic development. The authors frame their review as both an update and a critique, celebrating the field’s momentum while insisting on methodological rigor: computational predictions must be confirmed biochemically, functions must be established through perturbation experiments, and annotation standards must be standardized across databases. If those standards are met, the tiny proteins once overlooked in the interstices of the genome may become some of the most precise tools yet available for diagnosing disease and intervening in it, molecule by molecule, pathway by pathway.

Subject of Research: Microproteins encoded by small open reading frames as cellular regulators and precision medicine therapeutics

Article Title: The mighty microproteins: from versatile cellular regulators to precision medicine therapeutics

Article References: Azhar, M., Hassan, F.-U., & Arain, M. A. (2026). The mighty microproteins: from versatile cellular regulators to precision medicine therapeutics. 3 Biotech, 16(10), Article 408. https://doi.org/10.1007/s13205-026-05054-w

Image Credits: AI Generated

DOI: 10.1007/s13205-026-05054-w

Keywords: microproteins, small open reading frames, non-coding RNAs, ribosome profiling, proteogenomics, mitochondrial-derived peptides, cancer, precision medicine, biomarkers, peptide therapeutics, multi-omics, CRISPR validation