Remarks on the geometric quantization of the Kepler problem