Multiple Stellar Populations observed in Globular Clusters (GCs) are one of the major unsolved problems in the local Universe. Old GCs, which were previously believed to host Simple Stellar Populations, unambiguously show abundance variations and anti-correlations that are found in no other environment. Supermassive stars (SMS) with a mass of 10^3-4M⊙ formed within the proto-GCs at their early stages are proposed to be the source of the observed chemical enrichments in GC stars. However, such massive objects have not been observed in the Universe before.

Based on the predicted spectrophotometric properties of the SMS by Martins et al. (2020), I developed observational strategies and applied them to the massive star clusters in two nearby galaxies (NGC628 and M83). The atmospheric models for bright, cool, and inflated SMS predict a Balmer break (BB) in their spectra due to non-LTE effects. I found that color-color plots using the filters across and before/after the break have the potential to identify such inflated SMS hosting clusters. The spectra of such SMS hosting clusters will have a unique nature, as they exhibit both BB and nebular emission due to the presence of gas and young massive stars within the clusters. A detailed analysis performed using HST photometry and MUSE spectroscopy revealed that, although a small fraction of clusters satisfied the criteria, the clusters in our sample were not massive enough to host such an inflated SMS.