The role of Jagged1 as a dynamic switch of cancer cell plasticity in PDAC assembloids.

Background: Pancreatic ductal adenocarcinoma (PDAC), which commonly relapses due to chemotherapy resistance, has a poor 5-year survival rate (< 10%). The ability of PDAC to dynamically switch between cancer-initiating cell (CIC) and non-CIC states, which is influenced by both internal and external events, has been suggested as a reason for the low drug efficacy. However, cancer cell plasticity using patient-derived PDAC organoids remains poorly understood. Methods: First, we successfully differentiated CICs, which were the main components of PDAC organoids, toward epithelial ductal carcinomas. We further established PDAC assembloids of organoid-derived differentiated ductal cancer cells with endothelial cells (ECs) and autologous immune cells. To investigate the mechanism for PDAC plasticity, we performed single-cell RNA sequencing analysis after culturing the assembloids for 7 days. Results: In the PDAC assembloids, the ECs and immune cells acted as tumor-supporting cells and induced plasticity in the differentiated ductal carcinomas. We also observed that the transcriptome dynamics during PDAC re-programming were related to the WNT/beta-catenin pathway and apoptotic process. Interestingly, we found that WNT5B in the ECs was highly expressed by trans interaction with a JAG1. Furthermore, JAG1 was highly expressed on PDAC during differentiation, and NOTCH1/NOTCH2 were expressed on the ECs at the same time. The WNT5B expression level correlated positively with those of JAG1, NOTCH1, and NOTCH2, and high JAG1 expression correlated with poor survival. Additionally, we experimentally demonstrated that neutralizing JAG1 inhibited cancer cell plasticity. Conclusions: Our results indicate that JAG1 on PDAC plays a critical role in cancer cell plasticity and maintenance of tumor heterogeneity.

BRAK and APRIL as novel biomarkers for ovarian tumors.

Aims: To evaluate BRAK and APRIL in serum samples from healthy patients and an ovarian tumor group and analyze their effective value as biomarkers. Materials & methods: BRAK and APRIL were measured in 197 serum samples including 34 healthy controls, 48 patients with benign ovarian cysts and 115 patients with ovarian cancer, and the best statistical cut-off values were calculated. Then, the sensitivity, specificity, accuracy, positive predictive value and negative predictive value for selected cut-off points were assessed. Results: The healthy control group had statistically significant higher BRAK and lower APRIL than the ovarian tumor group. BRAK was excellent for differentiating healthy patients from patients with ovarian tumors, showing area under the receiver operating characteristic curve 0.983, 98.16% sensitivity and 100% specificity. When BRAK was combined with APRIL and CA-125, it also played a role in distinguishing benign cysts from malignancies with area under the curve 0.864, 81.74% sensitivity and 79.17% specificity. Conclusions: BRAK and APRIL are good candidates for ovarian tumor biomarkers.

KAI1(CD82) is a key molecule to control angiogenesis and switch angiogenic milieu to quiescent state.

BACKGROUND: Little is known about endogenous inhibitors of angiogenic growth factors. In this study, we identified a novel endogenous anti-angiogenic factor expressed in pericytes and clarified its underlying mechanism and clinical significance. METHODS: Herein, we found Kai1 knockout mice showed significantly enhanced angiogenesis. Then, we investigated the anti-angiogenic roll of Kai1 in vitro and in vivo. RESULTS: KAI1 was mainly expressed in pericytes rather than in endothelial cells. It localized at the membrane surface after palmitoylation by zDHHC4 enzyme and induced LIF through the Src/p53 pathway. LIF released from pericytes in turn suppressed angiogenic factors in endothelial cells as well as in pericytes themselves, leading to inhibition of angiogenesis. Interestingly, KAI1 had another mechanism to inhibit angiogenesis: It directly bound to VEGF and PDGF and inhibited activation of their receptors. In the two different in vivo cancer models, KAI1 supplementation significantly inhibited tumor angiogenesis and growth. A peptide derived from the large extracellular loop of KAI1 has been shown to have anti-angiogenic effects to block the progression of breast cancer and retinal neovascularization in vivo. CONCLUSIONS: KAI1 from PC is a novel molecular regulator that counterbalances the effect of angiogenic factors.

Cancer-initiating cells in human pancreatic cancer organoids are maintained by interactions with endothelial cells.

Pancreatic ductal adenocarcinoma (PDAC) shows poor prognosis and high malignancy due to the presence of cancer-initiating cells (CICs) and characteristics of the tumor microenvironment (TME). Organoids are useful for studying PDAC, and establishing organoids is dependent on stem cell growth factors, including Wnt signaling. Herein, using a conventional organoid culture system, we demonstrated that CD44(+)CD24(+) and CD44(+)CD24(+)EpCAM(+) CICs were enriched >65% in a PDAC patient-derived organoid. CICs expressing CD44 formed lumen structures by gathering into circles. Additionally, organoid-derived CD44(-) cancer cells were capable of organoid re-formation and could be re-programed as CD44-expressing CICs in the organoid culture system. To mimic a TME absent artificial stem cell growth factors, a PDAC organoid with vascular niche was established. CICs in the PDAC tumor organoid were maintained by paracrine effects and direct interactions with endothelial cells. Interestingly, CD44(+) cells in PDAC tumor tissue were detected primarily in the vascular niche. Inhibiting both Wnt and Notch signaling in endothelial cells suppressed organoid formation and the maintenance of CD24(+)CD44(+) CICs. Collectively, our results suggest that PDAC patient-derived organoids maintain CICs by interacting with endothelial cells via Wnt and Notch pathways.

A Disintegrin and Metalloproteinase 8 as a Potential Blood Biomarker for Early Diagnosis of Gastric Cancer.

PURPOSE: We aimed to evaluate the clinical significance of a disintegrin and metalloproteinase 8 (ADAM 8) as a potential blood biomarker for gastric cancer (GC). MATERIALS AND METHODS: Blood ADAM 8 was measured by ELISA. Cytokines/chemokines [interleukin-23 (IL-23), stromal cell-derived factor 1α/CXC chemokine ligand 12 (SDF-1α/CXCL12), interleukin-8 (IL-8), and soluble CD40 ligand (sCD40L)] were measured by chemiluminescent immunoassay. They were compared among five groups; normal/gastritis, high-risk, early GC (EGC), advanced GC (AGC) without distant metastasis, and AGC with distant metastasis by one-way analysis of variance in both training (n=80) and validation dataset (n=241). Clinicopathological features of GC and GC-associated cytokines were evaluated for their correlations with blood ADAM 8. To evaluate the diagnostic accuracy to predict GC, receiver operating characteristic (ROC) curve and logistic regression were used. RESULTS: Blood ADAM 8 significantly increased along GC carcinogenesis in both training (ANOVA, p<0.001) and validation dataset (p<0.001). It was significantly higher in EGC compared to high-risk (post-hoc Bonferroni, p=0.041) and normal (p<0.001). It was also higher in AGC compared with high-risk (p<0.001) and normal (p<0.001) groups. However, no significant difference was found between cancer groups. Blood ADAM 8 was correlated with N-stage (Spearman's correlation, γs=0.320, p=0.011), but not with T-stage or M-stage. Pearson's correlations showed blood ADAM 8 was closely correlated with pre-inflammatory cytokines, IL-23 (p=0.036) and SDF-1α/CXCL12 (p=0.037); however, it was not correlated with pro-angiogenic cytokine IL-8 (p=0.313), and sCD40L (p=0.702). ROC curve and logistic regression demonstrated that blood ADAM 8 showed higher diagnostic accuracy (sensitivity, 73.7%; specificity, 86.2%) than CEA (sensitivity, 23.1%; specificity, 91.4%). Combination of ADAM 8 and CEA further increased the diagnostic accuracy to predict GC (sensitivity, 81.8%; specificity, 84.0%). CONCLUSION: Blood ADAM 8 is a promising biomarker for early detection of GC.

Diabetes-Induced Jagged1 Overexpression in Endothelial Cells Causes Retinal Capillary Regression in a Murine Model of Diabetes Mellitus: Insights Into Diabetic Retinopathy.

BACKGROUND: Several mechanisms have been proposed to account for diabetes-induced microvasculopathy (DMV). Although Notch signaling was reported to be affected by glucose metabolism in endothelial cells during developmental angiogenesis, it has not been investigated in vascular remodeling of adult capillaries in relation to diabetes mellitus. METHODS: We induced diabetes mellitus in 8-week-old adult mice by intravenously administering streptozotocin. After 6 weeks, we harvested organs, including retina, heart, and skeletal muscle, and evaluated the capillaries with immunofluorescence and confocal microscopy. We modulated endothelial Notch signaling using chemical inhibitors in wild-type mice or transgenic mice, inducing conditional knockout of Jagged1 or Mib1. RESULTS: DMV was characterized by capillary remodeling, regression, and decreased density. Notch ligand Jagged1, but not δ-like ligand 4, was markedly increased in endothelial cells of diabetic mice. Using endothelium-specific Jagged1 knockdown mice, we found that blocking Jagged1 prevented DMV even under diabetic conditions. Furthermore, in the inducible endothelium-specific Jagged1 knockdown mice, blocking Jagged1 even at 4 weeks after the establishment of DMV could reverse it, leading to normalization of retinal vasculature. A search for downstream signals revealed that diabetes mellitus decreased the nuclear localization of Notch1 intracellular domain and reduced the expression of VE-cadherin and N-cadherin in endothelial cells. Chemical Notch inhibition phenocopied DMV in normal mice. CONCLUSIONS: Our findings indicate that diabetes mellitus induces Jagged1 overexpression and suppresses Notch signaling in endothelial cells, leading to DMV in adult mice. We conclude that dysregulated intercellular Notch signaling may be a novel mechanism of DMV.

Priming mobilized peripheral blood mononuclear cells with the "activated platelet supernatant" enhances the efficacy of cell therapy for myocardial infarction of rats.

AIM: Various methods are used to augment the efficacy of cell therapy in myocardial infarction (MI). In this study, we used the "activated platelet supernatant (APS)" to prime autologous "granulocyte colony-stimulating factor-mobilized peripheral blood mononuclear cells ((mob) PBMCs)" and investigated the efficacy of cell-based therapy in MI. METHOD: Rat (mob) PBMCs were isolated after daily subcutaneous injections of G-CSF at 100 µg/kg for 3 days. APS was isolated separately after activating rat platelets with thrombin 0.5 U/mL for 2 hours. Priming was performed with APS for 6 hours. To check the paracrine effect of primed (mob) PBMCs, we used the 36-hour culture supernatant of the primed cells. A rat MI model was used for an in vivo model. RESULT: Cytokines such as IL-1ß, IL-10, and TGFß were 3.7±0.9-fold, 3.4±1.2-fold, and 1.2±0.1-fold higher in APS, respectively, compared with naïve platelet supernatant. By APS priming, (mob) PBMCs showed M2 polarization and upregulation of angiogenic molecules (i.e., TEK, IL-10, CXCL1, and CX3CR1). APS-primed (mob) PBMCs had a 2.3-fold increased adhesion ability, induced by upregulated integrins. Rat endothelial cells cultured in the 36-hour culture supernatant of APS-primed (mob) PBMCs showed a 1.6-fold augmented proliferation and capillary network formation. In vivo transplantation of APS-primed (mob) PBMCs into rat MI models showed a significant trend of reduction in fibrosis area (P=.001) and wall thinning (P=.030), which lead to improvement in cardiac function measured by echocardiography. CONCLUSION: Our data reveal that APS priming can enhance the wound-healing potential of (mob) PBMCs. APS priming may be a promising method for cell-based therapy of MI.

CD82/KAI1 Maintains the Dormancy of Long-Term Hematopoietic Stem Cells through Interaction with DARC-Expressing Macrophages.

Hematopoiesis is regulated by crosstalk between long-term repopulating hematopoietic stem cells (LT-HSCs) and supporting niche cells in the bone marrow (BM). Here, we examine the role of CD82/KAI1 in niche-mediated LT-HSC maintenance. We found that CD82/KAI1 is expressed predominantly on LT-HSCs and rarely on other hematopoietic stem-progenitor cells (HSPCs). In Cd82(-/-) mice, LT-HSCs were selectively lost as they exited from quiescence and differentiated. Mechanistically, CD82-based TGF-ß1/Smad3 signaling leads to induction of CDK inhibitors and cell-cycle inhibition. The CD82 binding partner DARC/CD234 is expressed on macrophages and stabilizes CD82 on LT-HSCs, promoting their quiescence. When DARC(+) BM macrophages were ablated, the level of surface CD82 on LT-HSCs decreased, leading to cell-cycle entry, proliferation, and differentiation. A similar interaction appears to be relevant for human HSPCs. Thus, CD82 is a functional surface marker of LT-HSCs that maintains quiescence through interaction with DARC-expressing macrophages in the BM stem cell niche.

M-CSF from Cancer Cells Induces Fatty Acid Synthase and PPARß/δ Activation in Tumor Myeloid Cells, Leading to Tumor Progression.

We investigate crosstalk between cancer cells and stromal myeloid cells. We find that Lewis lung carcinoma cells significantly induce PPARß/δ activity in myeloid cells in vitro and in vivo. Myeloid cell-specific knockout of PPARß/δ results in impaired growth of implanted tumors, and this is restored by adoptive transfer of wild-type myeloid cells. We find that IL-10 is a downstream effector of PPARß/δ and facilitates tumor cell invasion and angiogenesis. This observation is supported by the finding that the CD11blowIL-10+ pro-tumoral myeloid cell is scarcely detected in tumors from myeloid-cell-specific PPARß/δ knockout mice, where vessel densities are also decreased. Fatty acid synthase (FASN) is shown to be an upstream regulator of PPARß/δ in myeloid cells and is induced by M-CSF secreted from tumor cells. Our study gives insight into how cancer cells influence myeloid stromal cells to get a pro-tumoral phenotype.

Erythropoietin priming improves the vasculogenic potential of G-CSF mobilized human peripheral blood mononuclear cells.

AIMS: From our previous clinical trials, intracoronary infusion of granulocyte-colony stimulating factor (G-CSF)-mobilized peripheral blood mononuclear cells ((mob)PBMCs) proved to be effective in improving myocardial contractility and reducing infarct volume in acute myocardial infarction. We tested the effect of priming (mob)PBMCs with erythropoietin (EPO) to augment its therapeutic efficacy. METHODS AND RESULTS: (mob)PBMCs were obtained from healthy volunteers after a 3-day subcutaneous injection of G-CSF (10 µg/kg). About 40% of (mob)PBMCs were EPO receptor (EPOR) (+) and responded to 6 h EPO-priming (10 IU/mL) by increasing the expression of vasculogenic factors (i.e. IL8, IL10, bFGF, PDGF, MMP9) and adhesion molecules (i.e. integrin αV, ß1, ß2, ß8) through the JAK2 and Akt pathway. These responses were also observed in PBMCs from elderly patients with coronary disease. The conditioned media from EPO-primed (mob)PBMCs contained various cytokines such as IL8, IL10, TNFα, and PDGF, which enhanced the migration and tube formation capability of endothelial cells. EPO-primed (mob)PBMCs also showed increased adhesion on endothelial cells or fibronectin. Augmented vasculogenic potential of EPO-primed (mob)PBMCs was confirmed in a Matrigel plug assay, ischaemic hindlimb, and myocardial infarction models of athymic nude mice. There were two action mechanisms: (i) cellular effects confirmed by direct incorporation of human (mob)PBSCs into mouse vasculature and (ii) indirect humoral effects confirmed by the therapeutic effect of the supernatant of EPO-primed (mob)PBMCs. CONCLUSION: Brief ex vivo EPO-priming is a novel method to augment the vasculogenic potential of human (mob)PBMCs, which would help to achieve better results after intracoronary infusion in myocardial infarction patients.

Activated platelet supernatant can augment the angiogenic potential of human peripheral blood stem cells mobilized from bone marrow by G-CSF.

Platelets not only play a role in hemostasis, but they also promote angiogenesis and tissue recovery by releasing various cytokines and making an angiogenic milieu. Here, we examined autologous 'activated platelet supernatant (APS)' as a priming agent for stem cells; thereby enhance their pro-angiogenic potential and efficacy of stem cell-based therapy for ischemic diseases. The mobilized peripheral blood stem cells ((mob)PBSCs) were isolated from healthy volunteers after subcutaneous injection of granulocyte-colony stimulating factor. APS was collected separately from the platelet rich plasma after activation by thrombin. (mob)PBSCs were primed for 6h before analysis. Compared to naive platelet supernatants, APS had a higher level of various cytokines, such as IL8, IL17, PDGF and VEGF. APS-priming for 6h induced (mob)PBSCs to express key angiogenic factors, surface markers (i.e. CD34, CD31, and CXCR4) and integrins (integrins α5, ß1 and ß2). Also (mob)PBSCs were polarized toward CD14(++)/CD16(+) pro-angiogenic monocytes. The priming effect was reproduced by an in vitro reconstruction of APS. Through this phenotype, APS-priming increased cell-cell adhesion and cell-extracellular matrix adhesion. The culture supernatant of APS-primed (mob)PBSCs contained high levels of IL8, IL10, IL17 and TNFα, and augmented proliferation and capillary network formation of human umbilical vein endothelial cells. In vivo transplantation of APS-primed (mob)PBSCs into athymic mice ischemic hindlimbs and Matrigel plugs elicited vessel differentiation and tissue repair. In safety analysis, platelet activity increased after mixing with (mob)PBSCs regardless of priming, which was normalized by aspirin treatment. Collectively, our data identify that APS-priming can enhance the angiogenic potential of (mob)PBSCs, which can be used as an adjunctive strategy to improve the efficacy of cell therapy for ischemic diseases.

Human podoplanin-positive monocytes and platelets enhance lymphangiogenesis through the activation of the podoplanin/CLEC-2 axis.

Emerging studies suggested that murine podoplanin-positive monocytes (PPMs) are involved in lymphangiogenesis. The goal of this study was to demonstrate the therapeutic lymphangiogenesis of human PPMs by the interaction with platelets. Aggregation culture of human peripheral blood mononuclear cells (PBMCs) resulted in cellular aggregates termed hematospheres. During 5-day culture, PPMs expanded exponentially and expressed several lymphatic endothelial cell-specific markers including vascular endothelial growth factor receptor (VEGFR)-3 and well-established lymphangiogenic transcription factors. Next, we investigated the potential interaction of PPMs with platelets that had C-type lectin-like receptor-2 (CLEC-2), a receptor of podoplanin. In vitro coculture of PPMs and platelets stimulated PPMs to strongly express lymphatic endothelial markers and upregulate lymphangiogenic cytokines. Recombinant human CLEC-2 also stimulated PPMs through Akt and Erk signaling. Likewise, platelets in coculture with PPMs augmented secretion of a lymphangiogenic cytokine, interleukin (IL)-1-ß, via podoplanin/CLEC-2 axis. The supernatant obtained from coculture was able to enhance the migration, viability, and proliferation of lymphatic endothelial cell. Local injection of hematospheres with platelets significantly increased lymphatic neovascularization and facilitated wound healing in the full-thickness skin wounds of nude mice. Cotreatment with PPMs and platelets augments lymphangiogenesis through podoplanin/CLEC-2 axis, which thus would be a promising novel strategy of cell therapy to treat human lymphatic vessel disease.

High glucose-induced jagged 1 in endothelial cells disturbs notch signaling for angiogenesis: a novel mechanism of diabetic vasculopathy.

Angiogenesis is a multistep process which is orchestrated by intercellular signaling. We developed an in vitro model of human angiogenesis to identify a pathologic angiogenesis and intercellular signaling in high glucose condition. We co-cultivated human endothelial cells (ECs) and smooth muscle cells (SMCs) in a spheroid on an SMC monolayer for 7 days either in high glucose or in control condition. We analyzed vascular growth and expression of notch or its ligands with confocal microscopy. Abnormal angiogenesis by high glucose condition was characterized by (1) increased sprouting and branching (high glucose vs. normal: number of sprouts 20.3±1.5 vs. 13.7±2.9, p=0.024; number of branching points 7.6±2.5 vs. 2.3±2.1, p=0.047), (2) decreased vascular diameter (diameter of the tubes 13.4±6. 1µm vs. 19.1±8.8 µm, p=0.012) and (3) destabilization of the tubes. We identified that high glucose induced jagged 1 and suppressed notch1 in ECs whereas it did not affect Dll4. Constitutive jagged 1 overexpression or inhibition of notch1 in ECs induced abnormal angiogenesis as the high glucose condition did. Endothelial-specific shRNA targeting jagged 1 rescued the aberrant angiogenesis in high glucose condition. High glucose condition induced an abnormal endothelial intercellular signaling leading to aberrant angiogenesis. It is a novel mechanism of diabetic microvasculopathy which can be a therapeutic target beyond glucose control.

Highly angiogenic CXCR4(+)CD31(+) monocyte subset derived from 3D culture of human peripheral blood.

Ex vivo expansion of human circulating angiogenic cells is a major challenge in autologous cell therapy for ischemic diseases. Here, we demonstrate that hematosphere-derived CXCR4(+)CD31(+) myeloid cells using peripheral blood possess robust proangiogenic capacity such as formation of vessel-like structures and tip cell-like morphology in Matrigel. We also found that CD31 positive myeloid cells are principal cellular component of hematospheres by magnetic cell sorting. Flow cytometry analysis showed that fresh peripheral blood contained 40.3 ± 15.2% of CXCR4(+)CD31(+) myeloid cells, but at day 5 of hematosphere culture, most of myeloid cells were CXCR4(+)CD31(+) by 86.9 ± 5.4%. Hematosphere culture significantly increased the production of angiogenic niche-supporting cytokines. Moreover, CD31-homophilic interaction and VEGF-VEGF receptor loop signaling were essential for sphere formation and acquisition of angiogenic capacity in hematospheres. Matrigel plug and ischemic hindlimb model provide in vivo evidence that hematosphere-derived myeloid cells have highly vasculogenic capacities, participate in new and mature vessel formation, and exert therapeutic effects on ischemic hindlimb. In conclusion, our strategy for ex vivo expansion of human CXCR4(+)CD31(+) angiogenic cells using hematospheres provides an autologous therapeutic cell source for ischemic diseases and a new model for investigating the microenvironment of angiogenesis.

New method to differentiate human peripheral blood monocytes into insulin producing cells: Human hematosphere culture.

Strategy to differentiate stem cells into insulin producing cells (IPCs) in vitro has been a promising one to get cell source of ß-cell replacement therapy for diabetes. It has been suggested that islets and neurons share features and nestin-positive cells could differentiate into IPCs. We have recently developed a three-dimensional culture system using human peripheral blood cells named as blood-born hematosphere (BBHS). Here we showed that most of BBHS were composed of nestin-positive cells. Under the four-stage differentiation protocol for IPCs, we plated nestin-positive BBHS onto fibronectin-coated dish. These cells form islet-like clusters and most of them expressed insulin. Pancreatic specific genes were turned on, such as transcription factors (Pdx-1, Ngn3 and Nkx6.1), genes related to endocrine function (Glut-2 and PC2) or ß cell function (Kir6.2, SUR1). Furthermore islet differentiation was confirmed by dithizone (DTZ) staining to detect zinc ion which binds insulin protein within the cells. Finally, IPCs derived from BBHS showed capability to secrete insulin in response to glucose stimulation. Taken together, our novel protocol successfully induced islet-like human insulin producing cells out of BBHS. This strategy of ex vivo expansion of IPCs using BBHS provides an autologous therapeutic cell source for the treatment of diabetes.