How DPP-4 Destroys Incretins and Why Inhibitors Block It
Incretins
<2 min half-life
Native GLP-1 has a circulating half-life of less than 2 minutes because DPP-4 cleaves it almost immediately after L-cell secretion, rendering 60-80% of total circulating GLP-1 inactive.
Deacon, Peptides, 2018
Deacon, Peptides, 2018
View as imageYour gut releases GLP-1 every time you eat. Within 120 seconds, most of it is already gone. The enzyme responsible is dipeptidyl peptidase-4 (DPP-4), a serine protease that clips two amino acids from the front of GLP-1 with remarkable speed and precision.[1] This rapid destruction creates a paradox at the center of diabetes pharmacology: the body produces a powerful blood sugar-lowering peptide and then immediately destroys it. Understanding exactly how DPP-4 works, where it acts, and why it matters explains two entire drug classes: DPP-4 inhibitors and GLP-1 receptor agonists. For the broader picture of what incretins are and why they matter, see our pillar article on GLP-1 and GIP: the two incretins.
Key Takeaways
- DPP-4 cleaves GLP-1 at the Ala8-Glu9 bond, producing inactive GLP-1(9-36) amide that constitutes 60-80% of total circulating GLP-1 (Deacon, 2018)
- Native GLP-1 has a half-life under 2 minutes; GIP survives approximately 7 minutes before DPP-4 degradation (Thornberry and Gallwitz, 2009)
- Endothelial DPP-4, not circulating soluble DPP-4, is the primary site of incretin degradation according to tissue-specific knockout studies (Mulvihill et al., 2017)
- DPP-4 inhibitors raise intact GLP-1 levels 2 to 3-fold but cannot replicate the pharmacological GLP-1 levels achieved by GLP-1 receptor agonists (Rhee et al., 2014)
- DPP-4 also cleaves over 30 other bioactive peptides, including GIP, NPY, PYY, SDF-1, and substance P (Juillerat-Jeanneret, 2014)
- DPP-4 inhibition increases intact GIP more than intact GLP-1 in human studies, suggesting GIP may contribute more to gliptin efficacy than previously thought (Rhee et al., 2014)
The enzyme: what DPP-4 actually is
DPP-4 (also called CD26) is a 766-amino-acid serine protease expressed on cell surfaces throughout the body. It exists in two forms: a membrane-anchored version embedded in the cell surface and a soluble form circulating in plasma.[2] Both forms are catalytically active.
The enzyme's specificity is narrow but devastating for incretins. DPP-4 cleaves dipeptides from the N-terminus of peptides that have a proline or alanine at position 2. GLP-1 has alanine at position 2. GIP has alanine at position 2. Both are perfect substrates.[1]
DPP-4 is expressed at high density on endothelial cells of capillaries, kidney proximal tubules, hepatocytes, and T-lymphocytes. This wide distribution means newly secreted GLP-1 encounters DPP-4 almost immediately after leaving the L-cells of the small intestine.[3]
The cut: how DPP-4 destroys GLP-1
GLP-1(7-36) amide is the bioactive form secreted by intestinal L-cells after a meal. DPP-4 cleaves the peptide bond between alanine at position 8 and glutamate at position 9, releasing a His-Ala dipeptide and leaving behind GLP-1(9-36) amide.[1]
GLP-1(9-36) amide is the predominant form in circulation. Deacon's 2018 review quantified this: 60-80% of total circulating GLP-1 is already the inactive (9-36) fragment.[1] The intact, bioactive GLP-1(7-36) amide that reaches pancreatic beta cells is a fraction of what was originally released.
The kinetics are extreme. Native GLP-1 has a circulating half-life under 2 minutes.[2] This means that after a meal stimulates L-cell GLP-1 secretion, the peptide has less than 2 minutes to activate GLP-1 receptors on beta cells, alpha cells, vagal afferent neurons, and other targets before DPP-4 renders it inactive.
GIP faces the same enzyme. DPP-4 cleaves GIP(1-42) at the Tyr1-Ala2 bond to produce inactive GIP(3-42). GIP's half-life is approximately 7 minutes in healthy individuals and even shorter (approximately 5 minutes) in people with type 2 diabetes.[2]
Where the cleavage happens
A longstanding assumption was that circulating soluble DPP-4 in plasma degrades incretins as they travel through the bloodstream. Mulvihill et al.'s 2017 study in Cell Metabolism overturned this model using tissue-specific DPP-4 knockout mice.[3]
Their findings: endothelial cell-derived DPP-4 is the primary site of incretin degradation, not circulating soluble DPP-4 and not bone marrow-derived DPP-4. When researchers selectively deleted DPP-4 from endothelial cells, intact GLP-1 levels rose and glucose tolerance improved. Deleting DPP-4 from bone marrow-derived cells had no effect on incretin metabolism or glucose homeostasis.[3]
This matters because GLP-1 must pass through the capillary bed immediately after secretion from L-cells. The endothelial DPP-4 lining these capillaries acts as a first-pass destruction filter. Much of the GLP-1 is cleaved before it even reaches the general circulation. This geographic arrangement explains the sub-2-minute half-life: the enzyme is positioned exactly where new GLP-1 must travel.
Why evolution kept this system
A peptide hormone that gets destroyed within minutes of secretion seems like a design flaw. It is not. The rapid DPP-4 cleavage of incretins serves as a negative feedback mechanism that prevents excessive insulin release.[2]
GLP-1 stimulates insulin secretion in a glucose-dependent manner, meaning it only triggers insulin release when blood glucose is elevated. But GLP-1 also suppresses glucagon, slows gastric emptying, and reduces appetite through central nervous system signaling. Without rapid degradation, a single meal could trigger prolonged insulin secretion, suppressed glucagon, and delayed gastric emptying that persists far beyond the postprandial window. Hypoglycemia would be a constant risk.
DPP-4 ensures that incretin signaling is tightly coupled to the meal itself. High GLP-1 after eating, rapid decline after nutrients are absorbed. This pulsatile pattern of incretin action matches the body's need for tightly regulated glucose disposal.
The incretin defect in type 2 diabetes
In type 2 diabetes, the incretin system fails at multiple points. GLP-1 secretion from L-cells is reduced (though not absent), and the insulinotropic response to GIP is severely blunted.[4] DPP-4 activity itself does not appear dramatically different in diabetic versus healthy individuals, but with less GLP-1 being produced and the same rate of degradation, the net effect is profoundly reduced active incretin signaling.
Rhee et al.'s 2014 study in healthy volunteers demonstrated that DPP-4 inhibition with sitagliptin 200 mg increased intact GLP-1 by approximately 2-fold and intact GIP by approximately 3-fold after oral glucose ingestion.[5] The absolute increase in intact GIP was larger than the increase in intact GLP-1. This raises the possibility that the clinical benefits of DPP-4 inhibitors involve GIP preservation as much as GLP-1 preservation. For how the incretin effect creates more insulin from food than glucose alone, see our dedicated article.
DPP-4 inhibitors: blocking the scissors
Five DPP-4 inhibitors (sitagliptin, saxagliptin, linagliptin, alogliptin, and vildagliptin) are approved for type 2 diabetes treatment worldwide. They all work the same way: competitively or substrate-selectively inhibiting the DPP-4 catalytic site to prevent GLP-1 and GIP cleavage.[2]
Thornberry and Gallwitz described the mechanism in their 2009 Best Practice review: DPP-4 inhibitors prevent N-terminal degradation of endogenous incretins, resulting in increased plasma concentrations of intact, biologically active GIP and GLP-1. This amplifies the physiological insulin response to meals while maintaining glucose-dependent insulin secretion, which explains the low hypoglycemia risk.[2]
The magnitude of incretin preservation has a ceiling, however. DPP-4 inhibitors can raise intact GLP-1 levels 2 to 3-fold above baseline, but they cannot exceed the body's endogenous production capacity.[5] GLP-1 receptor agonists like semaglutide and liraglutide bypass this limitation entirely by providing pharmacological GLP-1 concentrations that are 5 to 10 times higher than physiological levels. This pharmacological difference explains the greater efficacy of GLP-1 receptor agonists over DPP-4 inhibitors for weight loss and glucose control.
Beyond incretins: other DPP-4 substrates
DPP-4 does not only cleave GLP-1 and GIP. Juillerat-Jeanneret's 2014 review catalogued over 30 bioactive peptides susceptible to DPP-4 degradation.[4] These include:
- Neuropeptide Y (NPY) and Peptide YY (PYY): appetite-regulating peptides involved in energy homeostasis
- Stromal cell-derived factor-1 (SDF-1/CXCL12): a chemokine involved in stem cell homing and tissue repair
- Substance P: a pain-transmitting neuropeptide
- Brain natriuretic peptide (BNP): a cardiac peptide involved in blood pressure regulation
- Growth hormone-releasing hormone (GHRH): the hypothalamic signal for GH secretion
This broad substrate profile means DPP-4 inhibition has effects beyond glucose metabolism. Some of these off-target effects may be beneficial (SDF-1 preservation may improve wound healing and cardiovascular repair). Others are less clear. The cardiovascular outcome trials of DPP-4 inhibitors have shown neutral-to-modestly-positive cardiac effects, which may partly reflect BNP and SDF-1 preservation rather than incretin effects alone.[4]
How GLP-1 receptor agonists sidestep DPP-4
GLP-1 receptor agonists represent a different pharmacological strategy. Rather than preventing degradation of endogenous GLP-1, they introduce modified GLP-1 analogs that DPP-4 cannot cleave.
Exenatide, the first approved GLP-1 agonist, is based on exendin-4, a peptide from Gila monster venom that activates the GLP-1 receptor but has a glycine at position 2 instead of alanine, making it resistant to DPP-4 cleavage.[1] Liraglutide attaches a fatty acid chain to human GLP-1, which causes it to bind albumin and resist DPP-4 access. Semaglutide goes further with additional amino acid modifications and a larger fatty acid chain, extending its half-life to approximately one week.
The practical difference: DPP-4 inhibitors raise GLP-1 within the physiological range (2 to 3-fold). GLP-1 receptor agonists provide pharmacological activation at concentrations the gut would never produce naturally. This distinction drives the clinical gap between the two drug classes, with GLP-1 agonists producing 3 to 5 times greater weight loss and HbA1c reduction than DPP-4 inhibitors. For more on this comparison, see our article on GLP-1 side effects.
The speed of destruction in context
To appreciate DPP-4's efficiency, consider the full timeline of a single meal's incretin response:
0-10 minutes: L-cells in the duodenum and jejunum detect nutrients and begin secreting GLP-1(7-36) amide into the portal circulation.
Within seconds: GLP-1 encounters DPP-4 on endothelial cells lining the intestinal capillaries. An estimated 25-50% of secreted GLP-1 is cleaved before reaching the portal vein.[3]
Portal circulation: Additional DPP-4 on hepatic endothelium degrades more GLP-1 during its first pass through the liver.
Systemic circulation: The small fraction of intact GLP-1 that reaches the general circulation has a half-life under 2 minutes. Kidney clearance removes both intact and degraded fragments.
Net result: 60-80% of measurable circulating GLP-1 is already the inactive (9-36) fragment. The biologically active peptide is a minority species at all times.[1]
This implies that a substantial portion of GLP-1's physiological action occurs locally, through vagal afferent nerve endings in the gut wall, before the peptide even enters the bloodstream. How these gut signals reach the brain through neural rather than hormonal routes is covered in our dedicated article.
DPP-4 expression and regulation
DPP-4 expression is not static. Several conditions alter DPP-4 activity:
Obesity: Circulating DPP-4 activity is elevated in obesity, though this does not appear to cause proportionally greater incretin degradation in obese individuals.[4]
Type 2 diabetes: Some studies report modestly increased DPP-4 activity in type 2 diabetes, but the increase is small relative to the already-overwhelming speed of GLP-1 degradation.
Inflammation: DPP-4 expression on T-cells (where it is known as CD26) increases during immune activation. This has implications for autoimmune conditions and has generated interest in DPP-4 as an immunological target beyond metabolism.[4]
Exercise: Acute exercise transiently increases circulating DPP-4 levels, though the clinical significance for incretin metabolism during exercise is unclear.
Clinical implications of the DPP-4 mechanism
Understanding DPP-4's mechanism explains several clinical observations:
Low hypoglycemia risk with DPP-4 inhibitors: Because they only preserve endogenous incretin levels (which themselves are glucose-dependent), DPP-4 inhibitors rarely cause hypoglycemia as monotherapy.[2]
Weight neutrality: DPP-4 inhibitors produce modest GLP-1 increases within the physiological range, insufficient to trigger the appetite suppression and gastric slowing seen with pharmacological GLP-1 agonist doses.
Fixed efficacy ceiling: No matter how completely you inhibit DPP-4, you cannot raise incretin levels beyond what the gut produces. Patients with severely impaired L-cell function may derive less benefit.
Why combination with metformin works: Metformin increases GLP-1 secretion from L-cells. Combining metformin (more GLP-1 produced) with a DPP-4 inhibitor (less GLP-1 destroyed) amplifies the incretin axis from both ends.
The Bottom Line
DPP-4 destroys GLP-1 with extraordinary efficiency, cleaving the active peptide within seconds of secretion and leaving 60-80% of circulating GLP-1 in its inactive form. Endothelial DPP-4 in intestinal and hepatic capillaries, not circulating plasma enzyme, is the primary site of this first-pass destruction. DPP-4 inhibitors partially restore incretin signaling by blocking this cleavage, raising intact GLP-1 2 to 3-fold. GLP-1 receptor agonists bypass the enzyme entirely by using modified peptides DPP-4 cannot cleave, achieving pharmacological GLP-1 levels that DPP-4 inhibitors cannot reach. Both drug classes trace their pharmacological rationale directly to this single enzyme's rapid, precise destruction of a 2-minute peptide hormone.