We have constructed 26-amino acid transmembrane proteins that specifically transform cells but consist of only two different amino acids. Most proteins are long polymers of amino acids with 20 or more chemically distinct side-chains. The artificial transmembrane proteins reported here are the simplest known proteins with specific biological activity, consisting solely of an initiating methionine followed by specific sequences of leucines and isoleucines, two hydrophobic amino acids that differ only by the position of a methyl group. We designate these proteins containing leucine (L) and isoleucine (I) as LIL proteins. These proteins functionally interact with the transmembrane domain of the platelet-derived growth factor β-receptor and specifically activate the receptor to transform cells. Complete mutagenesis of these proteins identified individual amino acids required for activity, and a protein consisting solely of leucines, except for a single isoleucine at a particular position, transformed cells. These surprisingly simple proteins define the minimal chemical diversity sufficient to construct proteins with specific biological activity and change our view of what can constitute an active protein in a cellular context.T he chemical diversity of the 20 standard amino acid sidechains found in proteins supports myriad biochemical activities essential for life, and additional chemical diversity is generated by posttranslational amino acid modifications. The number of potential protein sequences is enormous: for proteins only 300 amino acids long, ∼10 400 possible sequences exist, a number larger by many orders-of-magnitude than the number of atoms in the known universe. This immense number of sequences and the chemical and conformational complexity of naturally occurring proteins hinder our ability to understand protein structure, folding, and function, and complicate protein engineering efforts. In addition, many different related amino acid sequences can fold into nearly identical structures, and even proteins with quite divergent sequences or protein folds can use similar chemistry to carry out related functions or display the same catalytic activity (e.g., refs. 1-3). These considerations further complicate attempts to clearly identify and understand the key structural features of proteins. Thus, the isolation of biologically active proteins with drastically reduced chemical complexity would represent a significant advance in protein science by facilitating the correlation of specific structural features with function.Up to 30% of all proteins contain ∼20-25-amino acid, primarily hydrophobic segments that span cell membranes (4). These transmembrane domains often play critical roles in cellular processes by engaging in highly specific protein-protein and protein-lipid interactions required for the proper folding, oligomerization, and function of transmembrane proteins (5-8). For example, receptor tyrosine kinases (RTKs) such as the PDGF-β receptor (PDGFβR) usually consist of an extracellular ligand binding do...