Cardiovascular disease is the leading cause of global mortality and morbidity. Cardiac dysrhythmias contribute significantly to this disease burden. Atrial fibrillation (AF) is the most common chronic dysrhythmia. Human induced pluripotent stem cell-derived atrial cardiomyocytes (hiPSC-AMs) present an exciting new model for AF but currently fail to reach maturity and so are limited in translational potential currently. We report a new approach using a combination of Gremlin 2 and retinoic acid treatment of human iPSCs for generating cardiomyocytes resembling atrial cells. More than 40% of myocytes generated by this approach showed rod-shaped morphology, expression of cardiomyocyte proteins (including RyR2 receptors, a-actinin-2, F-actin) and typically a striated appearance, all of which were broadly similar to the characteristics of adult atrial myocytes. Isolated myocytes were electrically quiescent until stimulated to fire action potentials with an atrial myocyte profile and an amplitude of approximately 100 mV, arising from a resting potential of approximately -70 mV. Single-cell RNA sequence (scRNASeq) analysis showed a high level of expression of several atrial specific transcripts including NPPA, MYL7, HOXA3, SLN, KCNJ4, KCNJ5 and KCNA5. Amplitudes of calcium transients recorded from spontaneously beating cultures were increased by the stimulation of α-adrenoceptors (activated by phenylephrine and blocked by prazosin) or β-adrenoceptors (activated by isoproterenol and blocked by CGP20712A). Thus, our new method provides an efficient approach for differentiating human atrial myocytes with mature characteristics from hiPSCs. This preparation will be very useful for studying signalling pathways in human atrial myocytes, and provides a valuable model for investigating atrial fibrillation and drug discovery.

Atrial fibrillation (AF) is the most common chronic arrhythmia presenting a heavy disease burden. We report a new approach for generating cardiomyocytes (CMs) resembling atrial cells from human-induced pluripotent stem cells (hiPSCs) using a combination of Gremlin 2 and retinoic acid treatment. More than 40% of myocytes showed rod-shaped morphology, expression of CM proteins (including ryanodine receptor 2, α -actinin-2 and F-actin) and striated appearance, all of which were broadly similar to the characteristics of adult atrial myocytes (AMs). Isolated myocytes were electrically quiescent until stimulated to fire action potentials with an AM profile and an amplitude of approximately 100 mV, arising from a resting potential of approximately −70 mV. Single-cell RNA sequence analysis showed a high level of expression of several atrial-specific transcripts including NPPA , MYL7 , HOXA3 , SLN , KCNJ4 , KCNJ5 and KCNA5 . Amplitudes of calcium transients recorded from spontaneously beating cultures were increased by the stimulation of α -adrenoceptors (activated by phenylephrine and blocked by prazosin) or β -adrenoceptors (activated by isoproterenol and blocked by CGP20712A). Our new approach provides human AMs with mature characteristics from hiPSCs which will facilitate drug discovery by enabling the study of human atrial cell signalling pathways and AF. This article is part of the theme issue ‘The heartbeat: its molecular basis and physiological mechanisms’.

Engineering human tissue with diverse cell types and desired cellular architectures and functions is a considerable challenge. The cerebral cortex, which has a layered cellular architecture composed of layer-specific neurons organised into vertical columns, delivers higher cognition through intricately wired neural circuits. However, current tissue engineering approaches cannot produce such structures. Here, we use a droplet printing technique to fabricate tissues comprising simplified cerebral cortical columns. Human induced pluripotent stem cells (hiPSCs) were differentiated into upper- and deep-layer neural progenitors, which were then printed to form cerebral cortical tissues with a two-layer organization. The tissues showed layer-specific biomarker expression and developed an integrated network of processes. Implantation of the printed cortical tissues into mouse brain explants resulted in substantial implant-host integration across the tissue boundaries as demonstrated by the projection of processes, the migration of neurons and the appearance of correlated Ca2+ signals. The approach we have developed might be used for the evaluation of drugs and nutrients that promote tissue integration. Importantly, our approach might be applied in personalised implantation treatments that restore the cellular structure and function of a damaged brain by using 3D tissues derived from a patient's own iPSCs.

Background: Bulk cancer and minimal residual disease (MRD) are characterised by different molecular drivers and therefore necessitate different therapeutic strategies. However, there are currently no 3D models that can faithfully recapitulate MRD ex vivo for therapy development. Methods: A microfluidic technique was implemented to construct 3D microtumours, in which tumour cells, either by themselves or with fibroblasts, were encapsulated in viscous hydrogels. The 3D microtumours were analysed for their response to first-line chemotherapeutics and characterised through RNA-Seq, by comparing them to both 2D cultures and clinical samples. Results: Our microfluidic platform guarantees the fabrication of 3D microtumours of tailorable size and cell content, which recreate key features of tumours such as hypoxia, characteristic organization of the cytoskeleton and a dose-response to chemotherapeutics close to the physiological range. The 3D microtumours were also used to examine non-genetic heterogeneity in ovarian cancer and could fully reflect the recently described 'Oxford Classic' five molecular signatures. The gene expression profile of 3D microtumours following chemotherapy treatment closely resembled that of MRD in ovarian cancer patients, showing the upregulation of genes involved in fatty acid metabolism. We demonstrate that these 3D microtumours are ideal for drug development by showing how they support the identification of a promising inhibitor of fatty acid oxidation, perhexiline, which specifically targets chemotherapy-resistant MRD ovarian cancer cells and not bulk cancer cells. Conclusion: We have obtained the first 3D model of ovarian cancer MRD by using microtumours generated through microfluidics. This system is ideal for high-throughput drug screening and, given its versatility, it can be readily extended to additional types of cancer, as well as accommodate multiple cell types to generate complex tumour microenvironments.