Now you can leave the IDs as concatinated strings, but that breaks database design and requires more work. What you would have to do is use a query to merge the table1 and table2 data, split the values out of the CourseID values, and execute additional queries. I can't see anyone designing a database this way unless they had no idea what they were doing. Is there a reason why you want it to be this way?
P.S. Use code tags so that your data looks nice. It's hard to look at your data when it doesn't line up.