Hi Vicky -
Including a small linker (e.g., GGSG) does not usually cause any problems and can prevent them. That said, the degron is itself a fairly unstructured, flexible peptide so it may not be necessary to add a linker between your protein and the degron - it’s more likely to help between your protein and GFP.
As with epitope tags, the optimal insertion point for the degron should be determined by educated guessing and experimentation. Different people use different rules of thumb. My own approach is to quickly determine which part(s) of the protein are most highly conserved by downloading the Caenorhabditis orthologs at http://download.caenorhabditis.org/v1/json/CGP_orthology.html and generating a multiple sequence alignment using MAFFT (you can also download a pre-calculated MSA at that site). All things being equal, I would usually choose to tag the less conserved end (unless, for example, it’s only expressed in a subset of isoforms). If both ends are strongly conserved or predicted to be important, an internal insertion into a flexible loop may be worth testing.
Also, we recommend editing directly in strains expressing the TIR1 construct you plan to use - it’s simple and allows you to test the efficacy of your degron construct without crossing.
Good luck! If you have any other questions about this, I’m happy to try to address them.