The limitations of distributed systems to satisfy the combination of consistency, availability and network partitioning tolerance requirements are well-documented and formally proven. There is however a limited amount of works discussing the impact of these limitations on designing applications native to the Cloud. This work addresses this particular need by proposing an approach for considering these requirements while designing an application. Our contributions are therefore a methodology for Cloud-native application design which considers consistency, availability and network partitioning tolerance, and a framework instantiating this methodology by using design patterns and their realization solutions on the Cloud. We also show how the proposed methodology can be used in practice to design an application solution with desired properties.