Insulin

What is Insulin?

Insulin is a 51-amino-acid two-chain polypeptide hormone produced by the beta cells of the pancreatic islets of Langerhans. The mature molecule consists of a 21-residue A-chain and a 30-residue B-chain held together by two interchain disulfide bonds (A7-B7 and A20-B19) plus one intrachain disulfide (A6-A11). Insulin is synthesized as a single-chain 110-residue preproinsulin precursor that is processed to proinsulin (after signal peptide cleavage in the rough endoplasmic reticulum) and then to mature insulin plus the C-peptide by prohormone convertases PC1/3 and PC2 in beta-cell secretory granules — a precursor-product relationship discovered by Donald Steiner at the University of Chicago in 1967 (Science). Insulin signals through the insulin receptor (INSR), a transmembrane tyrosine kinase that on ligand binding undergoes conformational rearrangement, autophosphorylates, and recruits insulin receptor substrate (IRS) proteins to activate the downstream PI3K-Akt pathway (driving glucose uptake via GLUT4 translocation, glycogen synthesis, lipogenesis, and protein synthesis) and the Ras-MAPK pathway (driving growth and gene expression). The dominant physiological role is glucose homeostasis: postprandial insulin secretion suppresses hepatic glucose production, drives skeletal muscle and adipose tissue glucose uptake, and promotes glycogen synthesis — losing this regulation produces type 1 diabetes (autoimmune beta-cell destruction with absolute insulin deficiency), type 2 diabetes (combined insulin resistance and beta-cell dysfunction), MODY (monogenic diabetes), and other rarer forms. Insulin was discovered by Frederick Banting, Charles Best, John Macleod, and James Collip at the University of Toronto in 1921-22 and entered clinical use within months — the first life-saving peptide drug. Its primary structure was determined by Frederick Sanger at Cambridge in 1955 (the first protein ever sequenced), establishing that proteins have defined amino acid sequences. The clinical insulin landscape now spans more than a dozen approved formulations including the rapid-acting analogs (insulin lispro, aspart, glulisine), the long-acting analogs (insulin glargine, detemir, degludec), pre-mixed combinations, U-500 concentrated insulin, the inhaled formulation (Afrezza), biosimilar versions of human insulin and analog insulins, and the once-weekly basal insulin icodec approved in the EU and Canada in 2024. Recombinant human insulin replaced animal-source insulin (porcine, bovine) in the 1980s following Genentech's 1978 cloning of the human insulin gene and Eli Lilly's 1982 launch of Humulin — the first recombinant DNA-derived therapeutic.

What Insulin Is Investigated For

Insulin is one of the most consequential drugs in medical history — the first effective therapy for type 1 diabetes, the first protein to have its primary structure determined, and the first recombinant DNA-derived therapeutic. Its clinical roles are well-defined and well-validated. In type 1 diabetes, insulin replacement is mandatory and lifelong; the Diabetes Control and Complications Trial (DCCT, NEJM 1993) established that intensive insulin therapy reduces development and progression of retinopathy, nephropathy, and neuropathy by 35-76 percent compared with conventional therapy, and the long-term EDIC follow-up extended the benefit to cardiovascular outcomes. In type 2 diabetes, insulin is used when oral agents (metformin, SGLT2 inhibitors, DPP-4 inhibitors, sulfonylureas) and GLP-1 receptor agonists are insufficient — modern algorithms increasingly favor GLP-1/GIP agents (semaglutide, tirzepatide) over insulin escalation because of weight gain and hypoglycemia concerns with insulin. Insulin is the principal treatment for diabetic ketoacidosis (IV regular insulin), hyperosmolar hyperglycemic state, severe hyperkalemia (insulin plus dextrose to drive potassium intracellularly), and gestational diabetes inadequately controlled by lifestyle and metformin. The clinical insulin landscape is rich: rapid-acting analogs (lispro, aspart, glulisine) for prandial coverage, long-acting analogs (glargine, detemir, degludec, and the new once-weekly icodec) for basal coverage, premixed and concentrated formulations, and the inhaled insulin Afrezza. Biosimilar competition has reduced prices for some insulins but affordability remains a major U.S. policy issue. The honest framing is that insulin is essential, lifesaving, irreplaceable for type 1 diabetes, and one of the safest peptide therapeutics when properly dosed — but it is also a high-risk medication: hypoglycemia is the most common serious adverse event, dosing errors can be fatal, and modern type 2 diabetes care often delays or avoids insulin escalation by leveraging the newer incretin-based therapies. Self-prescribing or research-chemical-channel insulin is dangerous and not the same product as licensed pharmaceutical insulin.

History & Discovery

Insulin was discovered at the University of Toronto in the autumn of 1921 and through the spring of 1922 by Frederick Banting (an unemployed surgeon with a hypothesis), Charles Best (a medical student), John Macleod (the senior physiologist whose laboratory hosted the work), and James Collip (a biochemist who refined the extract for human use). Banting's central insight was that previous attempts to isolate a pancreatic anti-diabetic factor had failed because exocrine pancreatic enzymes degraded the active substance during extraction; he proposed ligating the pancreatic ducts of dogs to atrophy the exocrine tissue while preserving the endocrine islets, then extracting the islet hormone. Working through the summer of 1921 in Macleod's laboratory while Macleod was in Scotland, Banting and Best prepared 'isletin' from atrophied dog pancreas extracts and demonstrated that it lowered blood glucose in pancreatectomized diabetic dogs. Macleod, returning from Scotland and recognizing the significance of the work, redirected the laboratory's resources to the project. Collip, recruited for his expertise in protein chemistry, refined the extraction process to remove pyrogenic and other contaminants — producing material safe for human use. Leonard Thompson, a 14-year-old boy with type 1 diabetes weighing 65 pounds and near death at Toronto General Hospital, received the first clinically successful insulin injection on January 23, 1922. His blood glucose fell from 28.9 mmol/L to 6.7 mmol/L within 24 hours, and within weeks he was clinically improved and discharged. The world's first life-saving peptide drug had been discovered. Banting and Macleod received the Nobel Prize in Physiology or Medicine in 1923; Banting shared his prize money with Best, and Macleod shared his with Collip. Eli Lilly and Company industrialized insulin production in partnership with Toronto in 1922, using bovine and porcine pancreas as source tissue. The primary structure of insulin was determined by Frederick Sanger at Cambridge in a series of papers from 1951 to 1955 — the first protein ever sequenced. Sanger established the order of the 51 amino acids in the A-chain and B-chain and the disulfide-bond pattern, in the process inventing fundamental peptide-chemistry methodology (the Sanger reagent FDNB) and demonstrating that proteins have defined, reproducible sequences rather than indeterminate compositions. He received the 1958 Nobel Prize in Chemistry for the work and a second Nobel in 1980 for DNA sequencing methodology. Dorothy Hodgkin solved the three-dimensional crystal structure of insulin in 1969 — a 35-year project after she had received the 1964 Nobel Prize in Chemistry for solving penicillin and vitamin B12. The structure revealed the disulfide-stabilized insulin monomer and the zinc-coordinated insulin hexamer that constitutes the storage form in pancreatic beta cells. Donald Steiner at the University of Chicago in 1967 reported in Science (PMID 4291105) that insulin is biosynthesized as a single-chain precursor (proinsulin) cleaved to the mature two-chain hormone — establishing the precursor-product framework that became the standard model for peptide hormone biosynthesis. Steiner's work on prohormone convertases extended over four decades. Genentech cloned the human insulin gene in 1978 and produced recombinant human insulin in E. coli, and Eli Lilly launched Humulin in 1982 — the first recombinant DNA-derived therapeutic to reach the market. Recombinant human insulin rapidly displaced animal-source insulin over the 1980s. The first insulin analog (lispro, Humalog, Eli Lilly) was approved in 1996, followed by aspart (NovoLog, Novo Nordisk, 2000), glulisine (Apidra, Sanofi, 2004), and the long-acting glargine (Lantus, Sanofi, 2000), detemir (Levemir, Novo Nordisk, 2005), and degludec (Tresiba, Novo Nordisk, 2015). The once-weekly basal insulin icodec (Awiqli, Novo Nordisk) was approved in the EU and Canada in 2024 with U.S. FDA review pending. Biosimilar competition began in the late 2010s, and U.S. policy changes (Inflation Reduction Act Medicare insulin price caps, manufacturer voluntary price reductions in 2023) have improved affordability — though insulin pricing remains a contentious issue. Hybrid closed-loop systems combining CGM with rapid-acting insulin pumps (Tandem Control-IQ, Medtronic 780G, Omnipod 5) have transformed type 1 diabetes care over 2017-2025 and represent the leading edge of diabetes technology.

How It Works

Insulin is a hormone your pancreas makes after you eat. Its job is to tell the rest of your body that food is coming and that it's time to store energy. When insulin gets to your liver, muscles, and fat tissue, it tells them to take glucose out of the blood and either burn it or store it. Without insulin — as in type 1 diabetes — blood sugar rises uncontrollably, the body starts breaking down fat for fuel and producing dangerous ketones, and you can die within weeks. Insulin replacement, discovered in 1922 in Toronto, was one of the most important breakthroughs in the history of medicine. Modern diabetes care uses engineered versions of insulin (analog insulins) that are absorbed at different speeds — rapid-acting ones for meals, long-acting ones for the steady background level your body needs all the time.

Insulin is encoded by the INS gene on human chromosome 11p15.5 and synthesized as a 110-amino-acid preproinsulin precursor in pancreatic beta cells. The 24-residue signal peptide is removed cotranslationally in the rough endoplasmic reticulum, producing 86-residue proinsulin, which folds and forms the three disulfide bonds that stabilize the mature hormone. Proinsulin is then trafficked to the Golgi and packaged into immature secretory granules, where the prohormone convertases PC1/3 and PC2 cleave it at two paired-basic-residue sites — releasing the C-peptide (a 31-residue connecting peptide) and producing mature insulin (a 21-residue A-chain plus a 30-residue B-chain held together by disulfide bonds). The precursor-product relationship was established by Donald Steiner at the University of Chicago in 1967 (Science, PMID 4291105), changing the conceptual model for how peptide hormones are biosynthesized. Mature insulin is stored as zinc-containing hexamers in dense-core secretory granules and released by glucose-stimulated exocytosis through ATP-sensitive K+ channel closure, membrane depolarization, voltage-gated Ca2+ influx, and granule fusion with the plasma membrane. Insulin signals through the insulin receptor (INSR), a 320 kDa transmembrane glycoprotein composed of two extracellular alpha-subunits and two membrane-spanning beta-subunits with intracellular tyrosine-kinase domains, joined by disulfide bonds in an alpha2-beta2 quaternary structure. Insulin binding to the alpha-subunits induces a conformational rearrangement that activates the beta-subunit tyrosine kinase, leading to autophosphorylation and tyrosine phosphorylation of insulin receptor substrate (IRS) proteins (IRS-1 through IRS-6) and the Shc adaptor protein. Phosphorylated IRS proteins recruit phosphoinositide 3-kinase (PI3K), generating PIP3 at the plasma membrane and activating Akt (PKB) — the central effector kinase of insulin's metabolic actions. Akt phosphorylates and inactivates GSK3 (driving glycogen synthase activation and glycogen synthesis), phosphorylates AS160/TBC1D4 (releasing GLUT4 storage vesicles for translocation to the plasma membrane and increasing glucose uptake in muscle and adipose tissue), and phosphorylates FoxO transcription factors (suppressing hepatic gluconeogenic gene expression). A parallel Ras-MAPK pathway is activated through Shc and drives mitogenic and growth-related effects of insulin signaling. The insulin-receptor-related insulin-like growth factor 1 receptor (IGF1R) shares ~50% sequence identity and substantial signaling overlap, complicating selective pharmacology. Functionally, insulin is the master anabolic hormone. In the liver it suppresses gluconeogenesis and glycogenolysis (lowering hepatic glucose output), promotes glycogen synthesis, and drives lipogenesis through SREBP-1c activation. In skeletal muscle (the principal site of postprandial glucose disposal) it drives GLUT4-mediated glucose uptake, glycogen synthesis, and protein synthesis through mTORC1 activation. In adipose tissue it drives glucose uptake, lipogenesis, and inhibits lipolysis through inhibition of hormone-sensitive lipase. In the brain it has appetite-suppressing and cognitive effects through receptors in the hypothalamus and elsewhere. The integrated effect is rapid postprandial glucose disposal, suppression of hepatic glucose production, and inhibition of catabolic pathways — losing this regulation in type 1 diabetes (autoimmune beta-cell destruction) or type 2 diabetes (combined insulin resistance and beta-cell dysfunction) produces the cascade of acute and chronic complications that define diabetes mellitus. Clinical insulin formulations are engineered for specific pharmacokinetic profiles. Regular human insulin (recombinant E. coli- or yeast-produced human insulin, sequence-identical to native) self-associates into hexamers at therapeutic concentrations, slowing absorption from subcutaneous tissue (onset 30-60 min, peak 2-4 hr, duration 6-8 hr). Rapid-acting analogs (lispro: B28 lysine, B29 proline; aspart: B28 aspartate; glulisine: B3 lysine, B29 glutamate) are engineered to disrupt hexamer formation and absorb as monomers (onset 5-15 min, peak 30-90 min, duration 3-5 hr). Long-acting analogs use different strategies: insulin glargine (A21 glycine, B-chain extended with two arginines) precipitates at neutral subcutaneous pH after acidic injection, slowly redissolving (peakless, ~24 hr duration); insulin detemir (B29 myristoyl fatty acid, B30 deletion) reversibly binds plasma albumin, slowing distribution; insulin degludec (B30 deletion plus B29 hexadecanedioic acid) forms multihexamer chains in subcutaneous tissue that slowly dissociate (~42 hr duration); insulin icodec (multiple modifications plus C20 fatty acid for albumin binding) extends to once-weekly dosing with ~196 hr half-life. Concentrated U-500 regular insulin is reserved for severely insulin-resistant patients. The inhaled formulation Afrezza uses Technosphere microparticles for pulmonary alveolar absorption.

Evidence Snapshot

Research Gaps & Open Questions

What the current literature has not yet settled about Insulin:

• 01Whether closed-loop and bi-hormonal (insulin plus glucagon) artificial-pancreas systems will fully obviate the need for patient decision-making in type 1 diabetes — current hybrid closed-loop systems still require meal announcements and are imperfect during exercise and illness.
• 02Whether immunomodulatory or beta-cell regenerative therapies (teplizumab for T1D delay; emerging stem-cell-derived islet cell therapies like VX-880; verapamil for residual beta-cell function) will eventually reduce or eliminate insulin replacement requirements in subsets of T1D.
• 03The optimal positioning of insulin in modern type 2 diabetes care given the strong cardiovascular and weight-loss profiles of GLP-1 receptor agonists (semaglutide), GLP-1/GIP dual agonists (tirzepatide), and emerging triple agonists (retatrutide) — guideline algorithms continue to evolve.
• 04Whether oral, smart, or glucose-responsive insulin formulations currently in development will reach clinical use and reshape patient experience.
• 05The long-term durability and clinical outcomes of once-weekly basal insulin (icodec) compared with daily long-acting analogs, particularly with respect to hypoglycemia risk during illness or exercise.
• 06The contribution of insulin therapy itself (versus the underlying hyperinsulinemic insulin-resistant state) to cancer risk, cardiovascular outcomes, and weight gain — a long-debated set of questions without clean answers.
• 07Whether the insulin-affordability landscape will continue to improve with biosimilar competition, manufacturer price cuts, and policy changes, or whether structural barriers will reassert themselves.

Forms & Administration

Insulin is administered by subcutaneous injection (the dominant route in chronic outpatient management — typically abdomen, thighs, upper arms, or buttocks), continuous subcutaneous infusion via insulin pump (CSII, using rapid-acting analog insulin), intravenous infusion (in DKA, HHS, surgery, and ICU glycemic control protocols), or pulmonary inhalation (the Afrezza Technosphere formulation, available for prandial coverage in select patients). Approved formulations span: rapid-acting analogs (lispro/Humalog, aspart/NovoLog, glulisine/Apidra, plus the ultra-rapid faster aspart/Fiasp and lispro-aabc/Lyumjev); regular human insulin (Humulin R, Novolin R, plus U-500 concentrated regular); intermediate-acting NPH (Humulin N, Novolin N); long-acting analogs (glargine/Lantus, Basaglar, Toujeo U-300; detemir/Levemir; degludec/Tresiba U-100 and U-200); the once-weekly basal insulin icodec (Awiqli, EU/Canada-approved 2024); and pre-mixed combinations (70/30, 75/25, 50/50) of intermediate and rapid-acting components. Biosimilar versions of glargine, lispro, and aspart are available in many markets. Pump and continuous glucose monitor (CGM) integration via hybrid closed-loop systems (Tandem Control-IQ, Medtronic 780G, Omnipod 5) automates basal insulin delivery and most correction boluses based on real-time CGM data. Self-administered insulin requires patient education on dosing, injection technique, hypoglycemia recognition, sick-day management, and travel adjustments.

Common Questions

Who Insulin Is NOT For

Drug & Supplement Interactions

Insulin's drug-interaction profile is dominated by additive effects on glucose homeostasis. Drugs that lower blood glucose — sulfonylureas, meglitinides, GLP-1 receptor agonists, GIP/GLP-1 dual agonists (tirzepatide), GLP-1/GIP/glucagon triple agonists (retatrutide), SGLT2 inhibitors, DPP-4 inhibitors, metformin, pramlintide, and alcohol — increase hypoglycemia risk when combined with insulin and typically require insulin dose reduction. Drugs that raise blood glucose — glucocorticoids (oral, IV, intra-articular, high-dose inhaled), atypical antipsychotics (olanzapine, quetiapine, clozapine), HIV protease inhibitors, beta-blockers (which also mask hypoglycemia awareness), thiazide diuretics (high-dose), and oral contraceptives (modest effect) — reduce insulin sensitivity and may require insulin dose increases. Beta-blockers blunt the autonomic warning symptoms of hypoglycemia (sweating, tremor, tachycardia), creating a particular risk; cardioselective agents are generally preferred. Hypokalemia-inducing diuretics combined with insulin (which drives potassium intracellularly) can produce significant total-body potassium derangement during DKA treatment. ACE inhibitors and ARBs may modestly potentiate insulin effects through improved insulin sensitivity, occasionally requiring dose reduction. Thiazolidinediones may increase risk of fluid retention and heart failure when combined with insulin.

Safety Profile

Legal Status

Regulatory status changes over time. Verify current local rules with a qualified professional.

Myths & Misconceptions