Non-small cell lung cancer accounts for approximately 85% of lung cancer cases, and about 30% of these cases are LUSC. Although targeted therapy and immunotherapy have produced major advances in non-small cell lung cancer, LUSC remains difficult to treat with current traditional approaches. BMP2 has important functions in embryonic development and lung formation, but its role in cancer is context dependent. The source study notes that BMP2 can suppress cancer stem-cell expansion and induce osteogenic differentiation in some settings, while also functioning as an oncogene by regulating cancer-related genes. Earlier reports found elevated BMP2 in non-small cell lung cancer and an association between high serum BMP2 and poor survival, but its effects and regulatory mechanisms in LUSC required further investigation.

The investigators first combined five Gene Expression Omnibus datasets to search for genes associated with LUSC tumorigenesis or metastasis. Comparisons included tumor tissue against normal bronchial biopsies or tumor-free lung, as well as the KLN205 LUSC cell line and its metastatic subclone, KLN.LN4K1. Intersection of the differentially expressed genes identified BMP2, GAPDH and LIFR as common candidates. Analysis of The Cancer Genome Atlas through GEPIA2 associated high BMP2 expression with unfavorable survival. The researchers then examined protein expression in a tissue microarray, finding significantly higher BMP2 immunohistochemistry scores in LUSC tissue than in normal lung and in paired adjacent normal tissues.

Patients classified as having high BMP2 expression, defined by an immunohistochemistry score above the cohort mean, had shorter survival than those with low expression. High BMP2 was also associated with positive lymph-node status and higher N stage. In univariate and multivariable analyses that included tumor size, lymph-node status, histological grade and BMP2 expression, high BMP2 remained a risk factor for prognosis; the reported multivariable hazard ratio was 3.032, with a 95% confidence interval of 1.098 to 8.372. The tissue-microarray survival comparison included 42 patients per group. The association was not consistent in every public dataset: an analysis of 780 patients with LUSC in Kaplan–Meier Plotter found no significant relationship between BMP2 expression and survival, with a hazard ratio of 1.09 and p value of 0.39. Differences in cohort composition and platforms may have contributed to the discrepancy, and serum BMP2 was not measured in the study.

To test the effect of BMP2 directly, the team generated stable BMP2-overexpressing LUSC cell lines. In NCI-H2170 cells, BMP2 overexpression increased growth rates, the proportion of cells incorporating EdU during DNA synthesis, and migration and invasion in transwell assays. The same pattern appeared in NCI-H226 cells: elevated BMP2 was associated with faster growth, more EdU-positive cells and greater migratory and invasive capacity. Loss-of-function experiments produced the converse result. After the researchers depleted BMP2 in SK-MES-1 cells, those cells grew more slowly and showed reduced migration and invasion. Together, the complementary gain- and loss-of-function experiments support a functional role for endogenous as well as experimentally increased BMP2 in malignant cell behavior.

The effects were also examined in vivo. Subcutaneous implantation of BMP2-overexpressing NCI-H2170 cells into athymic nude mice produced xenograft tumors with significantly greater volume and weight than tumors formed by control cells, with eight animals per group. A separate tail-vein injection model provided preliminary evidence about metastasis. Bioluminescent signals appeared in the lung, liver and brain of three of ten mice receiving BMP2-overexpressing cells, compared with one of ten receiving controls; the control signal was confined to the lung. A brain lesion from a BMP2-expressing mouse was confirmed by hematoxylin and eosin staining. The difference in metastatic incidence did not reach statistical significance because of the small number of events, so the result suggests a possible increase in metastatic potential rather than establishing it.

The researchers next searched for transcription factors capable of controlling BMP2. Intersecting predictions from KnockTF, AnimalTFDB, CistromeDB and GTRD produced 10 candidates, including MYC and RELA. A separate comparison of GATA6-associated expression changes identified BMP2 as a potential target. In LUSC cells, forcing GATA6 expression reduced BMP2 messenger RNA and protein, whereas GATA6 depletion increased BMP2 messenger RNA. Promoter analyses predicted seven GATA6-binding regions within the 3,000-base-pair region upstream of the BMP2 gene. Reporter assays showed that GATA6 inhibited BMP2 promoter activity, and mutation of individual predicted sites identified regions at −793 to −784, −1286 to −1273 and −1717 to −1708 base pairs as functionally important.

Chromatin immunoprecipitation followed by polymerase chain reaction in NCI-H2170 cells showed enrichment of the BMP2 promoter around all three functional sites in anti-GATA6 immunoprecipitates relative to immunoglobulin G controls, with no enrichment at a negative control region. These results provide direct evidence that GATA6 occupies the BMP2 promoter in the native chromatin of the cancer cells. The mechanism is not universal: the source article cites earlier work in which GATA6 increased, rather than reduced, BMP2 promoter activity in an adrenocortical cancer line. The researchers suggest that differences in co-regulator recruitment or promoter epigenetics may account for this tissue-dependent behavior, a question that would require chromatin profiling and comparisons of GATA6-associated proteins across cell types.

Functional experiments connected this transcriptional control to tumor behavior. The authors had previously identified GATA6 as a tumor suppressor in LUSC. When GATA6 was overexpressed by itself, LUSC cells grew less rapidly and showed reduced migration and invasion, but simultaneous BMP2 overexpression counteracted those suppressive effects. RNA sequencing of BMP2-overexpressing NCI-H2170 cells identified 156 upregulated and 111 downregulated genes at a fold-change threshold greater than two. Pathway analysis pointed to the non-canonical MAPK route, and immunoblotting showed increased phosphorylation of ERK1/2 without a significant rise in phosphorylated SMAD1/5/8/9. BMP2 also increased expression of MMP9, a protein implicated in extracellular-matrix remodeling, while genes associated with extracellular-matrix organization were prominent among the affected pathways.

Treatment of BMP2-overexpressing NCI-H2170 cells with the MEK inhibitor U0126 reduced the elevated ERK1/2 phosphorylation and abolished the increased proliferation and migration, bringing both phenotypes to levels comparable with controls. The inhibitor experiments support the necessity of MEK/ERK signaling for the BMP2-driven phenotypes in this cellular model, rather than merely showing an association between pathway activity and growth. Overall, the study proposes a sequence in which reduced GATA6 transcriptional repression permits BMP2 elevation, BMP2 activates non-canonical ERK1/2 signaling and alters extracellular-matrix-related programs, and these changes support proliferation and invasion. The evidence comes mainly from established cell lines and immunodeficient mouse models; orthotopic tumors and genetically engineered models would better reproduce the human lung environment. The work therefore identifies a mechanistic target for further study, not an established treatment or a clinically validated prognostic biomarker for LUSC.

Subject of Research: Molecular regulation of lung squamous cell carcinoma by the GATA6–BMP2–ERK1/2 pathway

Article Title: BMP2, transcriptionally regulated by GATA6, accelerates the proliferation and invasion of lung squamous carcinoma cells

Article References: Yang, H., Huang, C., Wang, H., & Chen, L. (2026). BMP2, transcriptionally regulated by GATA6, accelerates the proliferation and invasion of lung squamous carcinoma cells. iScience, 29(11), Article 117882. https://doi.org/10.1016/j.isci.2026.117882

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117882

Keywords: lung squamous cell carcinoma, BMP2, GATA6, ERK1/2 signaling, MAPK pathway, cancer-cell proliferation, tumor invasion, mouse xenografts, transcriptional regulation, transcriptionally, regulated, accelerates

The embryology behind the condition is elegantly precise. The pancreas forms from two outgrowths of the primitive foregut: a ventral bud and a dorsal bud. The ventral bud gives rise to the head of the pancreas and its uncinate process, whereas the dorsal bud builds the majority of the organ, including the neck, body, and tail, along with the accessory duct of Santorini. When the dorsal bud never appears, the distal pancreas simply does not exist. The clinical consequences can range from complete silence, discovered only incidentally, to abdominal pain, acute or recurrent pancreatitis, exocrine pancreatic insufficiency, and diabetes mellitus. The diabetes is thought to follow directly from the missing tissue: a substantial fraction of the insulin-producing beta cells reside in the body and tail, so their congenital absence reduces endocrine reserve from the very start of life.

In the weeks before his emergency presentation, the young man had been losing weight, roughly four kilograms over eight weeks, and had developed the classic warning signs of rising blood sugar: excessive urination and unquenchable thirst. He had also noticed intermittent discomfort after meals. On the day of admission, the picture crystallized. His blood glucose measured an alarming 412 milligrams per deciliter, his urine tested strongly positive for ketones, and his venous pH of 7.31 with a bicarbonate of 17.2 milliequivalents per liter indicated mild metabolic acidosis, treated clinically as mild diabetic ketoacidosis. At the same time, his serum lipase of 486 units per liter and amylase of 312 units per liter both exceeded their upper reference limits more than threefold, and combined with his epigastric pain and imaging evidence of peripancreatic inflammation, this satisfied the standard diagnostic criteria for acute pancreatitis.

Contrast-enhanced computed tomography delivered the decisive finding. The scan showed complete absence of the pancreatic neck, body, and tail distal to the splenic vein, together with an enlarged pancreatic head measuring 4.1 centimeters in its anteroposterior diameter, minimal peripancreatic fat stranding, and no signs of necrosis, pseudocyst, or fluid collections. A subtle but diagnostically valuable clue, the dependent stomach sign, was also present: with the distal pancreas missing, the stomach settles into the space it would normally occupy. Magnetic resonance cholangiopancreatography, performed eight weeks after discharge, added ductal confirmation by demonstrating a preserved ventral pancreatic duct, the duct of Wirsung, and a completely absent dorsal pancreatic duct, the duct of Santorini. Together, the morphological and ductal evidence established the congenital nature of the defect beyond reasonable doubt.

Reaching that conclusion required excluding a long list of impostors. Pancreatic lipomatosis, in which functioning tissue is replaced by fat, can create the illusion of a missing gland. Chronic pancreatitis can cause distal atrophy and ductal changes, but this patient had no history of chronic inflammation. Prior trauma, necrotizing pancreatitis, or surgery were ruled out by his history. A pancreatic tumor was considered unlikely given his age and the absence of any discrete mass or obstructive lesion on imaging. The etiological workup for pancreatitis also came back clean: triglycerides of 184 milligrams per deciliter and calcium of 9.1 milligrams per deciliter excluded the two major metabolic causes, liver enzymes showed no evidence of biliary obstruction, no gallstones were seen, and the patient denied alcohol, tobacco, drugs, and relevant medications.

The diabetes workup proved equally revealing. Twelve weeks after admission, his fasting C-peptide level was just 0.31 nanograms per milliliter, confirming profoundly low insulin production, while tests for the GAD65, IA-2A, and ZnT8 autoantibodies, the immunological fingerprints of classic type 1 diabetes, were all negative. This combination, severe insulin deficiency without autoimmune markers, in the presence of a major structural pancreatic defect, points toward a structural rather than autoimmune explanation. The authors were careful, however, not to over-classify. Type 3c diabetes, or pancreatogenic diabetes, formally requires evidence of exocrine pancreatic insufficiency in addition to structural disease and absent autoimmunity. In this patient, fecal elastase measured 210 micrograms per gram, within acceptable range, and there was no steatorrhea or malabsorption, so the team preferred the more cautious description of diabetes associated with dorsal pancreatic agenesis and severe insulin deficiency.

Treatment followed established pathways for both conditions simultaneously. The patient received intravenous fluid resuscitation, insulin infusion, analgesia, electrolyte monitoring, and a gradual reintroduction of oral feeding, and his abdominal symptoms resolved without any surgical intervention. Because his admission potassium of 4.3 milliequivalents per liter was normal, no supplementation was needed before starting insulin. He was discharged on a basal-bolus regimen combining once-nightly glargine insulin with pre-meal lispro. Pancreatic enzyme replacement was deliberately withheld given his preserved exocrine function. By week twelve, his fasting glucose ranged from 110 to 140 milligrams per deciliter and postprandial values from 140 to 180, his HbA1c had fallen from 9.73 percent to 7.1 percent, and he had experienced no hypoglycemic episodes, no recurrence of abdominal pain, and no further attacks of pancreatitis.

The case also carries an instructive lesson about the limits of diagnostic imaging. An initial ultrasound described the pancreatic body and tail as normal, but retrospective review showed this reflected inadequate visualization rather than genuine confirmation, a discrepancy fully explained by the well-known difficulty of imaging the distal pancreas through the stomach. The authors were similarly transparent about their own limitations: the only archived CT images were labeled as plain scans, preventing independent verification of the contrast phase, β-hydroxybutyrate levels were never measured, the paired glucose reference for the C-peptide test was not recorded, genetic testing was not performed, and follow-up lasted only twelve weeks. These caveats matter, because they remind readers that even striking case reports rest on imperfect data, and that the enlarged pancreatic head seen during active pancreatitis could reflect inflammatory edema rather than true compensatory hypertrophy.

What makes this report resonate beyond its rarity is the diagnostic challenge it poses to routine thinking. Diabetes in an 18-year-old is usually presumed to be autoimmune type 1 disease, and most of the time that presumption is correct. But this case demonstrates why severe new-onset diabetes deserves a structural look at the pancreas itself. Dorsal pancreatic agenesis has been linked to mutations in developmental genes such as GATA6, a recognized cause of human pancreatic agenesis, and recessive regulatory mutations in PTF1A have been described in isolated pancreatic agenesis, although no molecular mechanism could be attributed in this patient without genetic testing. Compared with previously published cases, including a systematic review of 53 patients and individual reports of DPA with diabetes or ketoacidosis, this patient stands out for the simultaneous occurrence of complete dorsal agenesis, acute pancreatitis, and antibody-negative severe insulin deficiency in one young man. The message for clinicians is clear: when the pancreas is missing half of itself, the diagnosis of diabetes should rest on imaging, C-peptide levels, and antibody studies together, not on age or assumption alone.

Subject of Research: Dorsal pancreatic agenesis presenting with new-onset diabetes and acute pancreatitis

Article Title: Dorsal Pancreatic Agenesis Associated With New‐Onset Diabetes Mellitus and Acute Pancreatitis in an 18‐Year‐Old Man: A Case Report

Article References: Rab, M., Javaid, M. H., Numani, R. N., Khanam, F., Khalid, M., & Segawa, F. (2026). Dorsal Pancreatic Agenesis Associated With New‐Onset Diabetes Mellitus and Acute Pancreatitis in an 18‐Year‐Old Man: A Case Report. Clinical Case Reports, 14(10), Article e73606. https://doi.org/10.1002/ccr3.73606

Image Credits: AI Generated

DOI: 10.1002/ccr3.73606

Keywords: dorsal pancreatic agenesis, pancreatitis, diabetes mellitus, diabetic ketoacidosis, C-peptide, MRCP, congenital anomaly, pancreatic imaging, type 3c diabetes, GATA6, insulin deficiency, case report

The research, led by Dong Cui of Henan Provincial Chest Hospital in Zhengzhou together with colleagues at Shanghai Pulmonary Hospital and Tongji University School of Medicine, centers on CYP3A5, an enzyme belonging to the cytochrome P450 family. Cytochrome P450 enzymes are best known for metabolizing drugs in the liver, but in recent years they have emerged as regulators of endothelial cell behavior within tumors. What remained unexplored until now was whether endothelial CYP3A5 plays a role in shaping the immune landscape of lung adenocarcinoma, and whether it could serve as a druggable target for therapies designed to reawaken the immune system against the cancer.

To answer that question, the team deployed an unusually broad methodological arsenal. They began with bioinformatic screening across large public datasets, including The Cancer Genome Atlas and the Genotype-Tissue Expression project, to pinpoint candidate endothelial targets in lung adenocarcinoma. They then layered on single-cell RNA sequencing and spatial transcriptomics, technologies that allow researchers to see which genes are active in individual cells and where those cells sit within the tumor architecture. This dual lens revealed that CYP3A5 is predominantly expressed in the endothelial cells of minimally invasive lung adenocarcinoma, an early stage of the disease, positioning the enzyme at a critical juncture in tumor evolution.

Having identified CYP3A5 as a candidate, the researchers asked what controls it. Through molecular docking, chromatin immunoprecipitation followed by quantitative PCR, DNA pulldown assays, and luciferase reporter experiments, they demonstrated that the transcription factor GATA6 sits upstream of CYP3A5, regulating its expression in endothelial cells. This established what the authors describe as a novel GATA6/CYP3A5/VEGF axis in the tumor-associated endothelium, a signaling cascade that links a nuclear regulator to an enzyme and ultimately to vascular endothelial growth factor, one of the most potent drivers of angiogenesis and immune suppression in cancer.

The single-cell and spatial analyses added another crucial piece to the puzzle: an enrichment of VEGFA–FLT1 signaling between the epithelial cancer cells and the endothelial compartment. This ligand-receptor conversation between tumor and vessel is a hallmark of the tumor microenvironment’s immunosuppressive wiring. Vascular endothelial growth factor is well known to inhibit T cell function and to promote the infiltration of regulatory T cells, the immune system’s own brakes. By mapping this axis with cellular resolution, the study clarified exactly where a therapeutic intervention could intervene to break the circuit.

Enter diosmetin, a natural flavonoid found in citrus fruits and certain medicinal herbs, which has previously shown anti-tumor properties but whose mechanism of action remained murky. The team’s bioinformatic screen flagged CYP3A5 as a functional target associated with diosmetin response. When the researchers tested this experimentally, they found something subtle and important: diosmetin inhibited the enzymatic activity of CYP3A5 without significantly altering the amount of the enzyme the cells produced. In other words, the flavonoid appears to act on the function of the enzyme rather than its abundance, suggesting that CYP3A5 may serve as a functional mediator of diosmetin’s activity in lung adenocarcinoma. The authors are careful to note that further biochemical studies will be required to confirm a direct physical interaction between diosmetin and CYP3A5, a caveat that reflects the rigor of the work.

The functional consequences of this intervention were striking. When endothelial cells were treated with diosmetin and then co-cultured with lung adenocarcinoma cells, the cancer cells lost much of their capacity to proliferate, migrate, and invade, the three behaviors that make tumors deadly. Enzyme-linked immunosorbent assays and Western blotting revealed that the treated endothelial cells also secreted fewer immunosuppressive cytokines, the chemical signals that tell patrolling immune cells to stand down. The tumor-vessel dialogue, in essence, was being rewritten from a conversation that protects the cancer into one that exposes it.

The story held up in living systems. In a xenograft model, diosmetin suppressed tumor growth and angiogenesis, and the treated tumors showed a measurable shift in their immune microenvironment. Because vascular endothelial growth factor is known to disable T cells and recruit regulatory T cells, the authors argue that suppressing endothelial CYP3A5 activity and the VEGF production it drives effectively disrupts an immune-evasive circuit. The dual action is what makes the finding conceptually exciting: a single intervention aimed at an endothelial enzyme simultaneously starves the tumor of new blood vessels and strips away one of its principal shields against immune attack. The researchers do acknowledge an important limitation, however, noting that further studies are needed to validate the immune-related effects in immune-competent models, since xenografts cannot fully recapitulate a functioning immune system.

The translational implications are considerable. Anti-VEGF therapies such as bevacizumab are already part of the standard arsenal against lung cancer, but they come with vascular toxicities and resistance mechanisms. Targeting CYP3A5 enzymatic activity upstream of VEGF offers a different point of entry into the same pathway, one that could theoretically complement existing angiogenesis inhibitors and immune checkpoint blockade. The fact that CYP3A5 expression was most prominent in minimally invasive adenocarcinoma also raises the intriguing possibility of early intervention, a window in which remodeling the endothelial microenvironment might prevent progression to fully invasive disease. And because diosmetin is a dietary flavonoid with a long history of human exposure, it provides a chemical starting point that is already amenable to medicinal chemistry optimization.

None of this means a diosmetin-based drug will reach patients tomorrow. The study is an early, mechanistic proof of concept, built on cell culture, omics data, and xenografts, and the authors themselves flag the need for direct binding studies and immune-competent validation. But the work does something scientifically valuable: it elevates an overlooked endothelial enzyme into a candidate vulnerability, connects it to a master transcriptional regulator, and traces a complete path from GATA6 through CYP3A5 to VEGF and the immunosuppressive microenvironment it creates. In doing so, it reframes the tumor vasculature not merely as a target for blood supply blockade but as an active immune gatekeeper whose enzymatic machinery can, in principle, be turned against the cancer it serves.

Subject of Research: Targeting endothelial CYP3A5 to remodel the immunosuppressive tumor microenvironment in lung adenocarcinoma

Article Title: Targeting endothelial CYP3A5 in lung adenocarcinoma remodels the immunosuppressive microenvironment via VEGF inhibition: Mechanistic insights from Diosmetin

Article References: Cui, D., Yang, Y., Chen, M., Feng, Y., Zhang, W., Zhou, X., & Qiu, R. (2026). Targeting endothelial CYP3A5 in lung adenocarcinoma remodels the immunosuppressive microenvironment via VEGF inhibition: Mechanistic insights from Diosmetin. Cancer Immunology, Immunotherapy. https://doi.org/10.1007/s00262-026-04538-1

Image Credits: AI Generated

DOI: 10.1007/s00262-026-04538-1

Keywords: CYP3A5, lung adenocarcinoma, diosmetin, tumor endothelium, VEGF, immunosuppressive microenvironment, GATA6, angiogenesis, cytochrome P450, tumor microenvironment, cancer immunotherapy, flavonoid