[Arabidopsis] Can overly expressed GFP by 35S be a secretory protein?

[Hee-Jin Kim]

Dear Arab-gen and colleagues 

How are you doing?

Does anybody know if overly expressed GFP with 35S can be found extracellular fraction of cells in addition to the typical localizations (cytoplasm and nucleoplasm) of GFP ?  One of my soluble protein (no  signal peptide), abundantly over-expressed mGFP5 fusion with 35S  was found in the cell wall fractions by Western blot although microscopic picture showed most portion of the fusion protein is in cytoplasm.  

I found a paper that 31% of overly expressed mGFP5-ER with 35S in BY2 was secreted into medium according to Western blot analyses (Liu et al., 2001, J Biotechnology 87: 1-16).  However,  I could not find so far if overly expressed GFP alone in plant cells can be secreted.   
Any comment how the levels of GFP expressions can affect localization would be appreciated.  

Hee-Jin