CLN025 is the strongest answer to “what is the smallest protein?” It is a designed chain of only 10 amino acids with experimentally determined crystal and solution structures, reversible cooperative folding, and a protein-like energy landscape. Its sequence is YYDPETGTWY.[1][2][3]
There is no official universal record, however, because protein and peptide do not have a hard length boundary. Under other definitions, TAL/pri is a natural 11-amino-acid gene product, MOTS-c is a reviewed 16-amino-acid human peptide, and RPL41 encodes a conventional 25-amino-acid human ribosomal protein.[5][6][8]
| Topic | Answer | Size | Description |
|---|---|---|---|
| Smallest folded protein | CLN025 | 10 aa | Designed rather than natural |
| Shortest natural translated product | TAL/pri | 11 aa | Usually described as a peptide |
| Short reviewed human peptide | MOTS-c | 16 aa | Mitochondrial-derived peptide |
| Shortest conventional human protein | RPL41/eS32 | 25 aa | Functions as part of the ribosome |
| Smallest unit of protein | An amino acid | — | Called a residue once incorporated into a chain |
| Minimum protein length | No universal number | — | A 40–50-aa boundary is only a convention |
| Smallest enzyme subunit | 4-OT | 62 aa mature | Bacterial and active as a homohexamer |
Why is there no single smallest protein?
A shortest-protein record changes with the definition. A researcher may count a designed molecule or require a natural gene product; count the translated precursor or the processed mature chain; require an autonomous fold or accept a subunit that is structured only inside a larger complex; and distinguish a predicted short open reading frame from a product supported by protein-level experiments.
The familiar peptide–protein boundary does not settle the issue. Chains below roughly 40 amino acids are often called peptides, while UniProt commonly uses “peptide” for active products no longer than about 40–50 residues that are processed from precursors. These are naming conventions, not physical laws.[15][17] CLN025 is important precisely because a 10-residue chain displays several properties normally associated with proteins.[1]
What is the smallest protein?
CLN025 is the smallest linear polypeptide with a particularly strong experimental case for behaving as an independently folded protein. It contains 10 amino acids, forms a compact beta-hairpin, undergoes reversible cooperative thermal unfolding, and occupies a funnel-like folding energy landscape.[1]
| CLN025 property | Value |
|---|---|
| Sequence | YYDPETGTWY |
| Length | 10 amino acids |
| Molecular mass in the PDB entries | 1.29 kDa |
| Main fold | Compact beta-hairpin |
| Solution structure | PDB 2RVD, solution NMR |
| Crystal structure | PDB 5AWL, 1.11 Å X-ray structure |
| Reported melting temperature | 343 K, about 70°C |
The molecule was designed from chignolin by replacing its two terminal glycines with tyrosines. Those substitutions make CLN025 substantially more stable while preserving the same 10-residue length.[1][4]
Is 1UAO the smallest protein?
PDB 1UAO is chignolin, not CLN025. Chignolin is another designed 10-amino-acid molecule with the sequence GYDPETGTWG. Entry 1UAO contains its solution NMR ensemble and reports a molecular mass of about 1.08 kDa.[4]
CLN025 is the stabilized chignolin variant YYDPETGTWY. Its corresponding experimental structures are PDB 2RVD for solution NMR and 5AWL for X-ray crystallography.[2][3] Both molecules are 10 residues long, but CLN025 has the stronger claim because its stability and folding behavior were characterized more extensively.
What is the smallest natural protein?
There is no universally accepted smallest natural protein. One of the shortest compelling natural gene products is the 11-amino-acid peptide encoded by the Drosophila tarsal-less/polished rice, or TAL/pri, gene.
TAL/pri contains 33-nucleotide open reading frames that directly encode 11-residue peptides. Experiments showed that these translated products regulate gene expression and animal development, and that one of the tiny peptides can provide the gene's biological activity.[5]
TAL/pri is therefore a strong answer to “shortest natural translated protein product,” but most authors call it a peptide or micropeptide rather than a conventional folded protein. It does not replace CLN025 as the best-supported record for the shortest autonomous fold.
What is the smallest protein in the human body?
There is no single accepted smallest human protein. The most defensible answer depends on whether the question means a mature signaling peptide, a reviewed short gene product, a conventional protein-coding gene, or a newly detected non-canonical open reading frame.
| Human category | Example | Length | Gene or precursor | Evidence and qualification |
|---|---|---|---|---|
| Reviewed mitochondrial-derived peptide | MOTS-c | 16 aa | MT-RNR1 | Reviewed UniProt entry with protein-level evidence; translation mechanism remains unusual |
| Reviewed mitochondrial-derived peptide | Humanin | 24 aa | MT-RNR2 | Protein-level evidence, but the precise source and production mechanism remain debated |
| Conventional ribosomal protein | RPL41/eS32 | 25 aa | RPL41 | Reviewed, protein-level evidence, and present in experimental human ribosome structures |
| Research-stage short products | 221 sORF candidates | Up to 15 aa | Multiple non-canonical ORFs | Potentially translated candidates, not one settled smallest functional protein |
| Mature processed hormone | TRH | 3 residues | 242-aa pro-TRH precursor | A cleaved and chemically modified peptide, not a standalone protein |
| Familiar small hormone | Insulin | 51 aa mature | 110-aa preproinsulin | Small, but not the smallest human protein |
MOTS-c: a reviewed 16-amino-acid human peptide
UniProt lists MOTS-c as a reviewed 16-amino-acid human peptide with the sequence MRWQEMGYIFYPRKLR, a mass of 2,175 Da, and evidence at the protein level. The entry reports detection in plasma and skeletal muscle.[6]
MOTS-c is an excellent short answer when “protein” is being used broadly enough to include independently annotated bioactive peptides. UniProt itself calls it a mitochondrial-derived peptide and notes that its translation mechanism is not fully resolved, so it should not be presented as an undisputed universal record.
Humanin: a reviewed 24-amino-acid human peptide
UniProt displays Humanin as a reviewed 24-amino-acid human peptide with protein-level evidence. It also cautions that mitochondrial translation would produce a 21-residue sequence and that the physiological source of Humanin remains unresolved.[7] Humanin is therefore another strong short human example, but not a simple universal record.
RPL41: a 25-amino-acid conventional human protein
For a conventional, reviewed human protein-coding gene with protein-level evidence, RPL41 is a stronger answer than insulin or thymosin beta-4. UniProt lists its product as a complete 25-amino-acid, 3.46-kDa ribosomal protein. Its sequence is MRAKWRKKRMRRLKRKRRKMRQRSK.[8]
A comparative study reported that the human RPL41 coding sequence is only 78 base pairs, the shortest among the ribosomal-protein genes it analyzed.[9] Together with UniProt's 25-residue sequence, that figure is consistent with 75 nucleotides of sense codons plus a three-nucleotide stop codon.
The protein was historically called 60S ribosomal protein L41. Current UniProt nomenclature calls it small ribosomal subunit protein eS32 and notes that structural evidence places it at the interface between the ribosomal subunits.[8]
RPL41 is annotated as disordered rather than as an autonomous globular fold.[8] It functions as a highly basic component of the much larger ribosome, which is why CLN025 remains the better answer to the separate question of the shortest independently folded protein.
Are there human proteins shorter than MOTS-c or RPL41?
Possibly, but the record is not settled. A 2023 study identified 221 previously missed human short open reading frames potentially translated into products no longer than 15 amino acids. The authors reported endogenous peptide-level evidence for 38 of them, while emphasizing the technical difficulty of detecting products this short. The study did not establish one universally accepted smallest human protein.[10]
The TransCODE Consortium showed in 2026 why the answer remains fluid. Across 95,520 proteomics experiments, about 25% of 7,264 analyzed non-canonical open reading frames produced detectable peptides. The consortium introduced “peptidein” for translated protein molecules whose status as functional proteins remains indeterminate.[11]
Very short sequences are also intrinsically difficult to certify by standard proteomics rules. The TransCODE study notes that reference-level evidence normally requires two distinct unique peptides of at least nine residues covering at least 18 residues—criteria that a protein shorter than 18 residues cannot satisfy in the usual way.[11]
Is the smallest human protein 11 amino acids long?
No widely accepted 11-amino-acid human protein record exists. The well-known 11-residue example is TAL/pri from Drosophila, not humans.[5] Human products of 15 residues or fewer have been reported as short-ORF candidates, but they should not yet be turned into a single definitive gene-and-protein record.[10][11]
Is insulin the smallest human protein?
No. Mature human insulin contains 51 amino acids, divided between a 30-residue B chain and a 21-residue A chain. It is translated first as a 110-amino-acid preproinsulin precursor and then processed.[13]
Insulin is a famous small protein and hormone, but MOTS-c, Humanin, and RPL41 are all shorter under their respective definitions. See the full processing and residue count in How many amino acids are in insulin?.
Is a three-amino-acid hormone the smallest human protein?
Thyrotropin-releasing hormone, or TRH, is a three-residue mature signaling peptide. However, the human TRH gene encodes a 242-amino-acid precursor containing repeated copies that are cleaved and chemically modified into the active tripeptide.[12] It is therefore an example of an extremely short mature peptide, not a standalone three-amino-acid protein.
Is glutathione the smallest protein?
No. Glutathione contains three amino acids, but it is synthesized through two ATP-dependent enzymatic reactions rather than translated as a gene-encoded polypeptide on a ribosome.[14] It is a tripeptide metabolite, not a protein.
What is the smallest unit of protein?
The smallest unit of a protein is called an amino acid. Once an amino acid is incorporated into a peptide or protein chain, it is usually called an amino acid residue.
Amino acid residues are linked by peptide bonds, and their order forms the protein's primary structure.[15] A linear chain with n residues has n − 1 peptide bonds, so CLN025 has 10 residues and 9 peptide bonds.
The word monomer is also correct when discussing amino acids as the repeating building blocks of a protein polymer, but “amino acid” is the clearest answer to “the smallest unit of protein is called what?”
What is the smallest amino acid?
Glycine is the smallest standard amino acid. Its side chain is a single hydrogen atom, making it the simplest and lowest-mass protein-forming amino acid.[16] Glycine is a building block of proteins; it is not itself the smallest protein.
What is the smallest peptide?
A dipeptide is the smallest peptide in the chemical sense: two amino acid residues joined by one peptide bond. A single amino acid is not a peptide, and a dipeptide is not normally called a protein.[15] Human TRH and glutathione are both tripeptides, but they reach that size through very different biological routes.
What is the minimum number of amino acids in a protein?
There is no universally accepted minimum number of amino acids required for a protein. Two amino acids can form a dipeptide, but adding a particular next residue does not suddenly convert a peptide into a protein.
A boundary around 40–50 amino acids is often used informally, especially when distinguishing processed bioactive peptides from larger proteins.[15][17] The boundary has many exceptions and does not describe folding, stability, function, evolutionary origin, or experimental evidence. CLN025 demonstrates that a 10-residue sequence can display protein-like folding properties.[1]
The best answer is therefore:
There is no fixed minimum protein length. Ten amino acids is the shortest strong example of an independently folded protein, not a universal cutoff for all proteins.
How small are proteins?
At the experimentally demonstrated lower end, a folded protein-like molecule can be 10 amino acids long and about 1.29 kDa, as shown by CLN025.[1][2] A conventional human ribosomal protein can be 25 residues and about 3.46 kDa, as shown by RPL41.[8]
Most proteins are much larger. Typical proteins contain hundreds of residues, so these record candidates sit at an extreme tail of the size distribution. See average protein size for the broader comparison.
What counts as a small protein?
A common operational definition treats a small protein as a translated product shorter than 100 amino acids, as used by the SmProt database.[25] The cutoff is not universal: some fields use 50 residues, while a 2026 review defines microproteins as polypeptides of 100–150 amino acids or fewer.[26]
“Small protein,” “microprotein,” “micropeptide,” and “short-ORF-encoded peptide” can therefore overlap without being exact synonyms. Length alone also does not show whether a sequence folds independently, is processed from a precursor, or has a demonstrated function.
Small protein examples
The table below compares short proteins, peptides, and protein-based tags that answer different versions of the question. Their residue counts are directly comparable; their biological categories are not.
| Example | Length | Type | Why it matters |
|---|---|---|---|
| CLN025 | 10 aa | Designed miniprotein | Strongest case for the shortest autonomous protein-like fold |
| Chignolin | 10 aa | Designed peptide/miniprotein | Original sequence represented by PDB 1UAO |
| TAL/pri | 11 aa | Natural translated peptide | Directly encoded natural developmental regulator |
| MOTS-c | 16 aa | Human mitochondrial-derived peptide | Reviewed entry with protein-level evidence |
| Trp-cage TC5b | 20 aa | Designed miniprotein | Compact, more than 95% folded in water at physiological pH |
| Humanin | 24 aa | Human mitochondrial-derived peptide | Reviewed human product with protein-level evidence |
| RPL41/eS32 | 25 aa | Human ribosomal protein | Conventional 25-residue human protein-coding gene product |
| Thymosin beta-4 | 44 aa | Human actin-binding protein | Reviewed small human protein; canonical database sequence is 44 aa |
| Insulin | 51 aa mature | Human hormone | Famous small protein, but not the human record |
| 4-OT | 62 aa mature subunit | Bacterial enzyme subunit | One of the shortest established natural enzyme subunits |
| NanoFAST | 98 aa | Fluorogen-activating tag | Very short genetically encoded fluorescence tag |
| miRFP670nano | 147 aa | Near-infrared fluorescent protein | Compact monomeric NIR tag |
Trp-cage is a useful intermediate example: its 20-residue sequence forms a compact structure and is more than 95% folded in water at physiological pH.[18] Thymosin beta-4 is a 44-residue reviewed human sequence that binds actin, but it is longer than RPL41 and the reviewed mitochondrial-derived peptides.[19]
The experimental PDB structures distinguish true folds from sequence length alone. Custom short sequences can also be built for inspection with PeptideBuilder.
The chart is a length comparison, not a single leaderboard. CLN025, TAL/pri, human peptides, a ribosomal protein, a multimeric enzyme subunit, and fluorescent tags satisfy different definitions and evidence standards.[1][5][6][7][8][18][19][20][21][23][24]
What is the smallest enzyme?
4-Oxalocrotonate tautomerase, usually abbreviated 4-OT, is one of the smallest well-established natural enzyme systems by subunit length. Its gene encodes 63 amino acids, and removal of the initiator methionine produces a 62-residue mature subunit.[20][21]
The active enzyme is not a lone 62-residue chain. Six subunits assemble into a homohexamer, so 62 amino acids is the subunit record being quoted rather than the total residue count of the catalytic assembly.[21] 4-OT is bacterial, not human.
A report sometimes summarized as a “one-amino-acid enzyme” used phenylalanine molecules assembled with zinc into a catalytic supramolecular material. The authors explicitly described it as non-proteinaceous, so it is not a one-residue natural protein enzyme.[22]
What is the smallest enzyme in the human body?
There is no standard, broadly accepted record for the smallest human enzyme. 4-OT should not be used as the answer because it is bacterial, and its 62-residue chain is one subunit of a six-subunit enzyme. Any proposed human record must specify whether it is counting a catalytic chain, a processed mature chain, or a complete functional complex.
What is the smallest fluorescent protein?
NanoFAST is a 98-amino-acid fluorogen-activating protein designed as a very small genetically encoded tag. Its 2021 paper described it as the shortest tag among known fluorescent and fluorogen-activating proteins at publication.[23] NanoFAST requires an added small-molecule fluorogen, so it is not directly equivalent to GFP-like proteins that form their own chromophore.
For near-infrared imaging, miRFP670nano is a 147-amino-acid, approximately 17-kDa monomer that binds biliverdin. Its authors described it as the smallest monomeric near-infrared fluorescent protein at publication.[24]
What is the simplest protein?
There is no objective “simplest protein” because simplicity can mean shortest sequence, fewest structural elements, fewest amino acid types, smallest mass, or simplest biological function.
If “simplest” means the shortest experimentally characterized autonomous fold, CLN025 is the best answer. It has only 10 residues and one compact beta-hairpin, yet still shows a defined three-dimensional structure and cooperative folding.[1] If the question means the simplest protein building block, the answer is glycine, the smallest amino acid—not a protein.[16]
Sources▼
- Crystal Structure of a Ten-Amino Acid Protein Journal of the American Chemical Society · 2008. https://pubs.acs.org/doi/10.1021/ja8030533
- 2RVD: NMR structure of a mutant of chignolin, CLN025 RCSB Protein Data Bank · August 20, 2026. https://www.rcsb.org/structure/2RVD
- 5AWL: Crystal structure of a mutant of chignolin, CLN025 RCSB Protein Data Bank · August 20, 2026. https://www.rcsb.org/structure/5AWL
- 1UAO: NMR structure of designed protein, chignolin, consisting of only ten amino acids RCSB Protein Data Bank · August 20, 2026. https://www.rcsb.org/structure/1UAO
- Peptides encoded by short ORFs control development and define a new eukaryotic gene family PLOS Biology · 2007. https://pmc.ncbi.nlm.nih.gov/articles/PMC1852585/
- Mitochondrial-derived peptide MOTS-c, UniProtKB A0A0C5B5G6 UniProt · August 20, 2026. https://rest.uniprot.org/uniprotkb/A0A0C5B5G6.txt
- Humanin, UniProtKB Q8IVG9 UniProt · August 20, 2026. https://www.uniprot.org/uniprotkb/Q8IVG9/entry
- Small ribosomal subunit protein eS32, UniProtKB P62945 UniProt · August 20, 2026. https://www.uniprot.org/uniprotkb/P62945/entry
- Characteristics and clustering of human ribosomal protein genes BMC Genomics · 2006. https://link.springer.com/article/10.1186/1471-2164-7-37
- Evolutionary origins and interactomes of human, young microproteins and small peptides translated from short open reading frames Molecular Cell · 2023. https://pmc.ncbi.nlm.nih.gov/articles/PMC10032668/
- Expanding the human proteome with microproteins and peptideins Nature · 2026. https://www.nature.com/articles/s41586-026-10459-x
- Pro-thyrotropin-releasing hormone, UniProtKB P20396 UniProt · August 20, 2026. https://www.uniprot.org/uniprotkb/P20396/entry
- Insulin, UniProtKB P01308 UniProt · August 20, 2026. https://rest.uniprot.org/uniprotkb/P01308.txt
- Glutathione: Overview of its protective roles, measurement, and biosynthesis Molecular Aspects of Medicine · 2009. https://pmc.ncbi.nlm.nih.gov/articles/PMC2696075/
- Biochemistry, Primary Protein Structure StatPearls, NCBI Bookshelf · August 20, 2026. https://www.ncbi.nlm.nih.gov/books/NBK564343/
- Glycine, ChEBI 15428 ChEBI · August 20, 2026. https://www.ebi.ac.uk/chebi/CHEBI:15428
- Peptide annotation definition UniProt · August 20, 2026. https://www.uniprot.org/help/peptide
- 1L2Y: NMR structure of Trp-cage miniprotein construct TC5b RCSB Protein Data Bank · August 20, 2026. https://www.rcsb.org/structure/1L2Y
- Thymosin beta-4, UniProtKB P62328 UniProt · August 20, 2026. https://www.uniprot.org/uniprotkb/P62328/entry
- 4-Oxalocrotonate tautomerase, an enzyme composed of 62 amino acid residues per monomer Journal of Biological Chemistry · 1992. https://pubmed.ncbi.nlm.nih.gov/1339435/
- 2-hydroxymuconate tautomerase, UniProtKB Q01468 UniProt · August 20, 2026. https://www.uniprot.org/uniprotkb/Q01468/entry
- Non-proteinaceous hydrolase comprised of a phenylalanine metallo-supramolecular amyloid-like structure Nature Catalysis · 2019. https://pmc.ncbi.nlm.nih.gov/articles/PMC6861134/
- NanoFAST: structure-based design of a small fluorogen-activating protein with only 98 amino acids Chemical Science · 2021. https://pubs.rsc.org/en/content/articlehtml/2021/sc/d1sc01454d
- Smallest near-infrared fluorescent protein evolved from cyanobacteriochrome as versatile tag for spectral multiplexing Nature Communications · 2019. https://www.nature.com/articles/s41467-018-08050-8
- SmProt: A Reliable Repository with Comprehensive Annotation of Small Proteins Identified from Ribosome Profiling Genomics, Proteomics & Bioinformatics · 2021. https://pmc.ncbi.nlm.nih.gov/articles/PMC9039559/
- Microproteins in Human Physiology and Pathology Biochemistry · 2026. https://pmc.ncbi.nlm.nih.gov/articles/PMC13143143/
Founder and computational chemist, ProteinIQ
Dr. Matic Broz is the founder of ProteinIQ and a computational chemist. He completed a PhD focused on protein structure, molecular dynamics, and neural networks, and writes about structural biology and scientific software.