PTP1B is a deceptively simple target with an outsized influence on human metabolism. The enzyme works by removing phosphate groups from tyrosine residues on signaling proteins, and among its substrates are the insulin receptor and downstream mediators of the insulin signaling cascade. When PTP1B runs unchecked, it effectively short-circuits insulin signaling, forcing cells to become less responsive to the hormone, a hallmark of type 2 diabetes. Genetic studies in mice made the case decades ago: animals lacking functional PTP1B show enhanced insulin sensitivity and resistance to diet-induced obesity without obvious developmental defects. That biological profile has made the enzyme a magnet for pharmaceutical chemists, but developing selective inhibitors has proved difficult because PTP1B shares a highly conserved catalytic pocket with closely related phosphatases, and the charged active site makes it hard to design drug-like molecules that bind tightly and get into cells. Natural products, with their structural diversity and often favorable safety profiles, remain an important hunting ground for new chemotypes against this target.

The Korean and Vietnamese team turned their attention to hemp seed cake, the protein- and polyphenol-rich solid residue left after oil is pressed from seeds of Cannabis sativa. Hemp cultivation for food and feed is expanding worldwide, and the cake is already valued as a source of protein, minerals, and phenolic compounds for applications such as fortified bread. Previous studies had identified phenylpropionamides in hemp seed husks and shown that hemp hull fiber and its constituent compounds can shape the human gut microbiome, but the cake itself had not been systematically explored for enzyme inhibitors relevant to metabolic disease. Using chromatographic separation guided by spectroscopic analysis and comparison with published data, the researchers identified two compounds: N-trans-caffeoyltyramine and N-trans-feruloyltyramine. Both belong to a family of cinnamoyl phenethyl amides, molecules in which a hydroxycinnamic acid moiety, the same aromatic framework found in coffee’s chlorogenic acid chemistry, is joined by an amide bond to tyramine, a phenolic amine.

The enzyme assays delivered the headline result. N-trans-caffeoyltyramine inhibited PTP1B with a half-maximal inhibitory concentration, or IC50, of 62.0 micromolar, while its feruloyl cousin proved considerably more potent, achieving an IC50 of 14.0 micromolar. For a pair of small phenolic amides isolated from a food-industry residue, those numbers are respectable, particularly because the compounds were tested as purified molecules rather than as a crude extract whose activity might be inflated by synergistic effects. The difference in potency between the two is chemically informative: the feruloyl compound carries a methoxy group in place of one of the caffeoyl compound’s hydroxyl groups, and the extra methyl ether appears to help the molecule engage the enzyme, a structure-activity clue that could guide the design of more potent analogs.

To understand how the more potent compound suppresses the enzyme, the team performed kinetic analysis, the classical enzymology method that reveals where and how an inhibitor binds. The results showed that N-trans-feruloyltyramine acts as a mixed inhibitor with an inhibition constant, Ki, of 30.9 micromolar. Mixed inhibition means the compound can bind both the free enzyme and the enzyme-substrate complex, indicating that it does not sit exclusively in the catalytic pocket but also makes contact with adjacent or allosteric regions of the protein. From a drug-discovery standpoint, that behavior is potentially advantageous. Inhibitors that engage sites outside the highly conserved catalytic cleft are more likely to discriminate between PTP1B and its close relatives such as T-cell protein tyrosine phosphatase, which has long frustrated efforts to develop selective compounds. The mixed-mode profile also suggests the feruloyl amide exploits binding determinants beyond the phosphate-binding loop that all members of the phosphatase family share.

Molecular docking simulations provided a structural picture consistent with the kinetics. Docking places a small molecule into a three-dimensional model of the target protein and scores the predicted binding pose and interaction energy. The simulations showed that N-trans-feruloyltyramine can occupy the PTP1B active-site region and form specific interactions with residues lining the pocket, including the hydrogen-bonding contacts expected between its phenolic hydroxyls and the polar residues that normally recognize phosphorylated tyrosine substrates. The aromatic rings and amide linkage of the compound allow it to span a meaningful stretch of the binding site, anchoring itself through multiple simultaneous contacts. While docking is a computational hypothesis rather than proof, the agreement between the predicted binding mode, the measured IC50, and the kinetic signature of mixed inhibition gives the pharmacological case for the compound internal coherence that many natural-product screening hits lack.

The researchers then widened the lens with a network pharmacology analysis, a systems-level approach that has become a staple of natural-product research. The method begins by identifying putative protein targets of a compound, often through databases that link chemical structures to known or predicted targets, and then maps those targets onto disease-associated genes and signaling networks. For N-trans-feruloyltyramine, the analysis pointed toward molecular mechanisms relevant to the treatment of type 2 diabetes mellitus, connecting the compound’s PTP1B inhibition to broader insulin-signaling and metabolic pathways. Network pharmacology cannot confirm efficacy, but it generates testable hypotheses about how a food-derived molecule might exert pleiotropic effects, and it helps prioritize which cellular assays and animal models should come next. In this case, the network view supports the idea that a PTP1B inhibitor from hemp could influence glucose homeostasis through the same axis that pharmaceutical PTP1B inhibitors are designed to modulate.

Context matters for any compound isolated from a food matrix, and here the feruloyl amide has a head start. A 2025 safety assessment published in Human and Experimental Toxicology evaluated N-trans-feruloyltyramine using in vitro genotoxicity studies and a 90-day toxicity study in rats, work that provides a preliminary toxicological foundation for a molecule now being flagged as an enzyme inhibitor. The compound also occurs naturally in hemp hull fiber, where it has been shown to influence the composition of the human gut microbiome in vitro, raising the possibility that it reaches the intestinal environment in a bioactive form when hemp-derived foods are consumed. Whether oral consumption of hemp seed cake delivers blood concentrations sufficient to inhibit PTP1B in human tissues is a question the current study cannot answer, and the authors are careful to frame their findings as a basis for further investigation rather than a dietary recommendation.

The broader significance of the study lies in what it says about agricultural side streams. Hemp seed cake is produced in quantity wherever hemp oil is pressed, and much of it is undervalued relative to its chemical content. Earlier work has catalogued its proteins, minerals, polyphenols, and glycoproteins, and extrusion processing has been shown to modify its polyphenol profile and biological activities. The new results add a specific, mechanistically characterized bioactivity to that inventory: PTP1B inhibition by phenylpropionamides, with potency data, kinetic mode, docking evidence, and a disease-relevant network context. That combination is exactly the kind of package that turns a phytochemistry observation into a credible lead for functional-food or nutraceutical development.

Considerable work remains before any of this reaches a clinic or a supermarket shelf. The IC50 values, while promising, are in the low-to-mid micromolar range, well above the nanomolar potencies typical of clinical drugs, so potency optimization through structural modification or synergistic formulations would be needed. Selectivity against related phosphatases must be demonstrated experimentally, bioavailability and metabolism in vivo remain uncharacterized, and the network pharmacology predictions require validation in cellular and animal models of insulin resistance. Still, the study delivers what early-stage natural-product science is supposed to deliver: a verified chemical lead from an unexpected source, a mechanistic hypothesis grounded in kinetics and structure, and a clear map of what to test next. In the effort to find new tools against type 2 diabetes, the humble residue of hemp oil pressing has just earned a place on the list of materials worth watching.

Subject of Research: PTP1B inhibitory activity of phenylpropionamides isolated from hemp seed cake as potential antidiabetic compounds

Article Title: Inhibitory effects of phenylpropionamides from the hemp seed cake on protein tyrosine phosphatase 1B: in vitro, molecular docking, and network pharmacology approach

Article References: Cao, T. Q., Yeo, C. E., Yu, S., Chi, H. N., Tran, T. Y. N., Tuong, L.-T., Kim, J.-S., & Hahn, D. (2026). Inhibitory effects of phenylpropionamides from the hemp seed cake on protein tyrosine phosphatase 1B: in vitro, molecular docking, and network pharmacology approach. Food Science and Biotechnology. https://doi.org/10.1007/s10068-026-02293-8

Image Credits: AI Generated

DOI: 10.1007/s10068-026-02293-8

Keywords: hemp seed cake, PTP1B, phenylpropionamides, N-trans-feruloyltyramine, N-trans-caffeoyltyramine, type 2 diabetes, molecular docking, network pharmacology, enzyme inhibition, natural products, Cannabis sativa, insulin signaling

The numbers behind the review are sobering. Between 1990 and 2021, the global burden of chronic kidney disease driven by type 2 diabetes nearly doubled, reflecting an increase of roughly 80 to 90 percent. In the United States, annual all-cause healthcare costs for an insured patient with diabetic kidney disease climb from about 30,000 dollars in early stages to approximately 119,000 dollars as the disease advances. These figures capture more than economic strain; they reflect a disease whose cellular and molecular underpinnings remain only partially mapped, and whose progression is rarely reversible once structural injury takes hold. The authors of the review, Grace LeBleu and Fawaz G. Haj, contend that aberrant tyrosine phosphorylation—a reversible chemical switch that controls how cells respond to insulin, growth factors, and inflammatory cues—is a hallmark of diabetic kidney injury, and that the enzymes which remove these phosphate marks deserve far greater therapeutic attention.

Tyrosine phosphorylation is governed by a tug of war between two enzyme families: protein tyrosine kinases, which add phosphate groups to proteins, and protein tyrosine phosphatases, which remove them. While kinases have long been celebrated drug targets, phosphatases were for years dismissed as undruggable, largely because their catalytic pockets are highly conserved and positively charged, making selective inhibition difficult. The human genome encodes more than 100 phosphatase genes, of which 38 are classical phosphatases that specifically target phosphotyrosine residues—21 of the receptor-type and 17 of the non-receptor-type. The review systematically catalogs the evidence linking each of these 38 enzymes to diabetic kidney disease, ranging from directly established contributors to those implicated only indirectly or not at all.

Among the most firmly established is PTP1B, encoded by the PTPN1 gene. Discovered in the late 1980s and early 1990s as a negative regulator of insulin signaling, PTP1B dephosphorylates the insulin receptor and dampens downstream cascades. In podocytes—the specialized cells whose foot processes form the kidney’s filtration barrier—PTP1B knockdown stabilizes the cytoskeleton and preserves insulin signaling under high-glucose conditions. Recent work identified the cytoskeletal protein alpha-actinin-4 as a PTP1B substrate in insulin-stimulated podocytes, linking the enzyme directly to podocyte motility. In mouse models, global PTP1B deletion markedly attenuates glomerular injury and reduces albuminuria without major changes in blood glucose or pressure, while podocyte-specific deletion or pharmacological inhibition preserves podocyte architecture and blunts glomerular basement membrane thickening. Notably, PTP1B expression has been reported to be elevated in glomerular biopsies from patients with diabetic kidney disease, lending human relevance to the animal findings.

SHP2, the product of PTPN11, has emerged as another central player. In podocytes exposed to high glucose, SHP2 expression rises alongside increased endoplasmic reticulum stress, activation of the inflammatory transcription factor NF-κB, and heightened cell motility; silencing the enzyme prevents these changes. In diabetic mice, elevated SHP2 phosphorylation appears in macrophages, glomeruli, and podocytes, and podocyte-specific genetic deletion or pharmacological inhibition preserves renal function, lowering albuminuria and blood urea nitrogen. Immunohistochemical analysis of renal biopsies from patients with diabetic kidney disease reveals elevated phosphorylated SHP2 at a key activating site, and small-molecule inhibitors such as SHP099 and PHPS1 have shown renoprotective effects in rodent models of both type 1 and type 2 diabetes. SHP1, encoded by PTPN6, tells a similar story: hyperglycemia raises its expression in podocytes, disrupting insulin signaling and promoting apoptosis, while podocyte-specific deletion of SHP1 reduces albuminuria and reverses features of disease progression, apparently by restoring SUMOylation of the slit diaphragm protein podocin.

The review also spotlights endothelial and podocyte receptor-type phosphatases with distinctive renal roles. VEPTP, encoded by PTPRB, is largely restricted to endothelial cells in the adult kidney, where it dephosphorylates the angiopoietin receptor TIE2. Hyperglycemia, osmotic stress, and hypoxia all increase VEPTP expression, and inducible genetic deletion in diabetic mice improves glomerular filtration rate, reduces matrix accumulation, and preserves podocyte integrity—although an antibody-based inhibition approach alone failed to improve albuminuria, hinting that endothelial protection must be paired with therapies addressing the filtration barrier itself. CD148, encoded by PTPRJ, dephosphorylates growth factor receptors including VEGFR2 and EGFR, and an agonistic anti-CD148 antibody that boosts the enzyme’s activity attenuated albuminuria and mesangial expansion in diabetic mice. GLEPP1, encoded by PTPRO, sits on the apical surface of podocytes, and its deletion produces structurally abnormal foot processes, hypertension, and reduced filtration capacity, while its expression is diminished in sclerosed glomeruli from patients with diabetic kidney disease.

Beyond these established players, the review traces a broader web of phosphatases implicated in diabetes and renal pathology. TCPTP dephosphorylates the insulin receptor non-redundantly with PTP1B; its overexpression via gene therapy ameliorated albuminuria and renal morphology in diabetic mice, and its levels are reduced in patients with early-stage diabetic kidney disease. PEZ, or PTPN14, modulates the Hippo pathway through YAP and drives inflammatory and fibrotic signaling via TRIP6, with elevated expression detected in diabetic glomeruli and patient kidneys. RPTPζ serves as a receptor for interleukin-34, recruiting macrophages to injured tubules, while RPTPγ functions as a carbon dioxide and bicarbonate sensor in proximal tubular cells and rises in the serum as diabetic kidney disease progresses. Immune-associated phosphatases such as CD45, PTPN22, and HePTP connect renal inflammation and fibrosis to leukocyte signaling, and a cluster of enzymes including LAR, RPTPε, RPTPσ, and PTP-MEG2 modulate insulin signaling in muscle, liver, and adipose tissue, with several genetic variants linked to type 2 diabetes risk.

What makes the moment ripe for phosphatase-directed therapy is a confluence of pharmacological progress. Current standard care—RAAS blockade with ACE inhibitors or angiotensin receptor blockers, SGLT2 inhibitors validated by the DAPA-CKD, EMPA-KIDNEY, and CREDENCE trials, GLP-1 receptor agonists demonstrated in the FLOW trial, and the nonsteroidal mineralocorticoid receptor antagonist finerenone—slows disease but rarely reverses injury. Combination regimens are increasingly viewed as the path forward, and phosphatase modulators could slot into this framework as complementary agents. SHP2 inhibitors have already advanced through phase I, II, and III oncology trials, while the allosteric inhibitor SHP099 reduced albuminuria and blood urea nitrogen in diabetic mouse models. PTP1B inhibitors including ertiprotafib and trodusquemine reached clinical trials for type 2 diabetes and obesity, and an antisense oligonucleotide, IONIS-PTP-1BRx, lowered HbA1c and body weight in overweight patients with type 2 diabetes. VEPTP inhibition with AKB-9778 progressed to phase II trials for diabetic retinopathy, where patients showed a modest reduction in urinary albumin-to-creatinine ratio.

Novel modalities are expanding the toolbox further. Allosteric inhibitors that bind sites outside the conserved catalytic pocket promise improved selectivity, bivalent ligands engage both active and adjacent sites, and targeted protein degradation via PROTAC technology—exemplified by the SHP2 degrader SHP2-D26—offers a route to eliminate rather than merely inhibit the enzyme. Artificial intelligence and machine learning workflows have already been applied to PTP1B and SHP2 inhibitor discovery, helping overcome the historical barrier of conserved catalytic pockets by identifying selective binding modes and optimized physicochemical properties. The authors emphasize that pharmacological modulation need not mean inhibition in every case: for CD148, whose activity appears protective, an activation strategy is the goal, illustrating the nuanced directionality of phosphatase drug development.

The review’s bottom line is a call to translational action. Diabetic kidney disease remains the leading cause of end-stage renal failure, and existing therapies, however effective at slowing decline, rarely restore lost function. Classical protein tyrosine phosphatases sit upstream of the insulin, inflammatory, oxidative stress, and cytoskeletal cascades that converge to destroy the kidney in diabetes, and several are now pharmacologically tractable. The challenge ahead lies in bridging preclinical promise to clinical reality—defining cell-type-specific roles, establishing context-dependent effects in human disease, and testing phosphatase-targeted agents in combination with established therapies. If that bridge can be built, enzymes once written off as undruggable may finally deliver the precision treatments that patients with diabetic kidney disease have been waiting for.

Subject of Research: The roles of classical protein tyrosine phosphatases in the pathogenesis and treatment of diabetic kidney disease

Article Title: The role of classical protein tyrosine phosphatases in diabetic kidney disease and therapeutic implications

Article References: LeBleu, G., & Haj, F. G. (2026). The role of classical protein tyrosine phosphatases in diabetic kidney disease and therapeutic implications. Journal of Biomedical Science, 33(1), Article 89. https://doi.org/10.1186/s12929-026-01282-7

Image Credits: AI Generated

DOI: 10.1186/s12929-026-01282-7

Keywords: diabetic kidney disease, protein tyrosine phosphatases, PTP1B, SHP2, podocytes, insulin signaling, renal fibrosis, SGLT2 inhibitors, phosphatase inhibitors, drug discovery, chronic kidney disease, inflammation