Black hole-main sequence star (BH-MS) binaries are one of the targets of the future data releases of the astrometric satellite {it Gaia}. They are supposed to be formed in two main sites: a galactic field and star clusters. However, previous work has never predicted the number of BH-MS binaries originating in the latter site. In this paper, we estimate the number of BH-MS binaries formed in open clusters and detectable with {it Gaia} based on the results of {it N}-body simulations. By considering interstellar extinction in the Milky Way (MW) and observational constraints, we predict $sim 10$ BH-MS binaries are observable. We also find that chemical abundance patterns of companion MSs will help us to identify the origin of the binaries as star clusters. Such MSs are not polluted by outflows of the BH progenitors, such as stellar winds and supernova ejecta. Chemical anomalies might be a good test to confirm the origin of binaries with relatively less massive MSs ($lesssim 5M_{odot}$), orbital periods ($sim 1.5;$year) and higher eccentricities ($e gtrsim 0.1$).