So, can you patent naturally occurring substances, or not? That's a rather complicated question, and some recent Supreme Court decisions have recomplicated it in US patent law. Mayo v. Prometheus and Assoc. Mol. Pathology v. Myriad Genetics. The latter, especially, has sent the PTO (and the IP lawyers) back to staring out their respective windows, thinking about what to do next.
The Patent Office has now issued new guidelines for its examiners in light of these rulings, though, and things may be changing. Previous standards for patenting naturally occurring compounds have been tightened up - if I'm reading this correctly, no longer is the process of isolation and purification itself seen as enough of a modification to make a case for patentability. The four "judicial exception" categories, to be used in patentability decisions, are (1) abstract ideas, (2) laws of nature, (3) natural phenomena, and (4) natural products. And examiners are specifically asked to determine if a patent application's claims recite something "significantly different" than these.
Here's the blog of an IP firm that thinks that the USPTO has gone too far:
Now we learn that grant of these and similar patents were mistakes, that 100 years of consistent practice in the field of patents was wrong, that what was invented was no more than products of nature without significant structural difference from the naturally-occurring materials, and that the USPTO will endeavour to avoid such mistakes in future. . .
. . .Whatever workable rule of law is derivable from Prometheus, it is apparent from the opinion of Justice Breyer that it was not the Court’s intention to bring about a radical change in pharmaceutical practice. The opinion gives a warning against undue breadth:
“The Court has recognized, however, that too broad an interpretation of this exclusionary principle could eviscerate patent law. For all inventions at some level embody, use, reflect, rest upon, or apply laws of nature, natural phenomena, or abstract ideas.”
The problem (and it's the usual problem with fresh patent law) is that we really don't know what the phrases in the decisions or guidance mean, in practice, until there's been some practice. This is going to be thrashed out application by application, lawsuit by lawsuit, until some new equilibrium is reached. Right now, though, if you're trying to patent something that could be considered an isolated natural product, your life has become much more complicated and uncertain. Here's another IP law firm:
What is the "significantly different" standard? With respect to natural products, the Guidance offers that what is claimed should be "non-naturally occurring and markedly different in structure from the naturally occurring products". Again, it is unclear at this point how different "markedly different" will be. How different it needs to be will be worked out on a case-by-case basis, beginning at the level of the patent examiner at the USPTO.
So how can you protect your IP if it involves subject matter that could be considered a "product of nature" by a US examiner? Since we don't yet really know how different "markedly different" is, one prudent strategy would be to include multiple claims having varying degrees of modifications relative to the naturally occurring thing, to the extent these makes sense commercially and scientifically. The more different your claimed product is from the naturally occurring thing, the more likely it is to be considered patent eligible by the USPTO.