Create a column containing specific string expressions derived from another column

In a dataframe, I would like to create a column out of an existing one. The new column (name="Symbol") should contain only a subset of the expressions found on the existing column (name="Description) based on a matching pattern, which in my case is defined by the prefix GN=. For those cells missing the matching pattern from the existing column, "Not available" should be returned in the new column. From here:

View(df[,1:3])

Accession   Description                                                                               Sample1
A0FGR9      Extended synaptotagmin-3 OS=Homo sapiens GN=ESYT3 PE=1 SV=1 - [ESYT3_HUMAN]               117.920
A6NHJ4      Zinc finger protein 860 OS=Homo sapiens GN=ZNF860 PE=1 SV=3 - [ZN860_HUMAN]               30.218
A0A0C4DH68  Immunoglobulin kappa variable 2-24 OS=Homo sapiens GN=IGKV2-24 PE=3 SV=1 - [KV224_HUMAN]  524.706
P0DOX7      Immunoglobulin kappa light chain OS=Homo sapiens PE=1 SV=1 - [IGK_HUMAN]                  503.110

I would like to get here:

View(df[,1:4])

Accession   Description                                                                               Symbol          Sample1
A0FGR9      Extended synaptotagmin-3 OS=Homo sapiens GN=ESYT3 PE=1 SV=1 - [ESYT3_HUMAN]               ESYT3           117.920
A6NHJ4      Zinc finger protein 860 OS=Homo sapiens GN=ZNF860 PE=1 SV=3 - [ZN860_HUMAN]               ZNF860          30.218
A0A0C4DH68  Immunoglobulin kappa variable 2-24 OS=Homo sapiens GN=IGKV2-24 PE=3 SV=1 - [KV224_HUMAN]  IGKV2-24        524.706
P0DOX7      Immunoglobulin kappa light chain OS=Homo sapiens PE=1 SV=1 - [IGK_HUMAN]                  Not available   503.110

Thank you in advance for the fruitful suggestions.