Massive stars are primary sources of chemical yields and mechanical & ionizing luminosities. Therefore, understanding massive stars' evolution is crucial for comprehensive studies of the chemical and ionization evolution of galaxies. In this talk, I will focus on the chemical evolution of galaxies, especially discussing the origin of Nitrogen at low metallicities. Most observational calibrations of metallicities (O/H) implicitly depend upon the intrinsic N/O ratio at low metallicities, and therefore, understanding the nucleosynthetic origin of N in metal-poor systems is essential if we are to complete our picture of the chemical evolution of galaxies at high redshift. However, the observed N/O ratio (log(N/O)∼ −1.5) is nearly independent of O/H, albeit with a very large scatter (~1.5 dex) at low metallicities (12+log(O/H) <= 7.5). This plateau and scatter in N/O at low metallicities puzzled astronomers for decades and the origin of them had remained unexplained to date. In this talk, I will describe how these several heretofore unexplained features of the N/O distribution at low O/H can be explained by the N seen in metal-poor galaxies being mostly primary nitrogen that is returned to the ISM via pre-supernova winds from rapidly rotating massive stars (M >= 10 MSun, v/vcrit >= 0.4). This mechanism naturally produces the observed N/O plateau at low O/H. I will further explain the origin of large scatter in N/O at low O/H that arises naturally from variations in star-formation efficiency. By contrast, models in which the N and O come primarily from supernovae provide a very poor fit to the observed abundance distribution. These peculiar abundance patterns observed at low O/H are a signature that dwarf galaxies retain little of their SN ejecta, leaving them with abundance patterns typical of winds.