You're gonna need to make it in a eukaryote through the secretory pathway and it probably needs a suite of modifying enzymes to cleave and glycosylate it properly, but I don't know if those enzymes are widespread in the eukaryotes or animal-specific.
If the yeast doesn't produce the enzyme on it's own why not simply add the gene for it? If the enzyme messes up the yeast too much you can always switch from yeast towards some other unicellular organism. It might cost a bit to do the bioengineering but when it comes to commercial usage of the protein I don't think that's an issue.
Okay, I actually went and looked at the papers and followed a chain of citations back until I found where they are actually getting the protein.
Looks like my assessment was unnecessarily pessimistic. Those that I could find the source of their proteins who studied this protein in isolation seem to have bought a recombinant product from Peprotech, a biotech company that sells huge numbers of proteins. Their website (http://www.peprotech.com/en-US/Pages/Product/Recombinant_Human_GDF-11/120-11) seems to indicate that it is being produced in E. coli. I susp...
In experiments performed on mice, blood transfusions from young mice reversed age-related markers in older mice. The protein involved is identical in humans.
http://mic.com/articles/88851/harvard-scientists-may-have-just-unlocked-the-secret-to-staying-young-forever