The ability to generate 3-dimensional models of the developing human kidney via the directed differentiation of pluripotent stem cells—termed kidney organoids—has been hailed as a major advance in experimental nephrology. Although these provide an opportunity to interrogate human development, model-specific kidney diseases facilitate drug screening and even deliver bioengineered tissue; most of these prophetic end points remain to be realized. Indeed, at present we are still finding out what we can learn and what we cannot learn from this approach. In this review, we will summarize the approaches available to generate models of the human kidney from stem cells, the existing successful applications of kidney organoids, their limitations, and remaining challenges.

通过定向分化多能干细胞（称为肾脏类器官）生成发育中的人类肾脏的 3 维模型的能力被誉为实验肾脏学的重大进步。尽管这些提供了审视人类发展的机会，但模型特异性肾脏疾病有助于药物筛选甚至提供生物工程组织；大多数这些预言性的终点仍有待实现。事实上，目前我们仍在寻找我们可以从这种方法中学到什么以及我们不能学到什么。在这篇综述中，我们将总结可用于从干细胞生成人类肾脏模型的方法、肾脏类器官的现有成功应用、它们的局限性和仍然存在的挑战。