Not the aqua color but I've seen strong yellow and red/pink in a lysate pellets.  It was due to metal precipitates and complexed metals that were identified by classic qualitative inorganic analysis.  ICAP confirmed the content and concentrations.  Some were phosphate or hydroxide precipitates (high trace metals), some were organic iron complexes (insufficient bioavailable Fe).  Notice that there are striations where the color is appearing in at least 3 relatively thin layers and in different hues as though there are different densities of colored material; this is quite telling.  Might be interesting to know the nature of your recombinant protein.  The Brookhaven NL media already contains Mo, Cu, Co, Mn, and Fe chromogenic metals and significant phosphate levels so if these are not unusually high in concentration then complexation could exaggerate the intensity of the color and amount of putative metal.  Probably not any non-metal organic chromophore.  How well were the cell pellets washed before lysis and was there any trace of color in the cell pellets?  Was the lysis alkaline and does pellet washing with water reduce the color intensity?  

That is quite odd! I've gotten a lot of inclusion bodies out of various media including Studier's autoinduction formulations and I have NEVER seen such a pellet! Is the protein a known metal chealtor? Does it have a His tag, or expected metal center? Even Cu/Zn SOD looks plain grey in IBs, what were the lysis conditions? Could there be an order of magnitude error in metals added (these can happen, guess how I know ;))

 Cool! I imagine you have a lot of some metalloprotein there. Would love to know what it is.

Not the aqua color but I've seen strong yellow and red/pink in a lysate pellets.  It was due to metal precipitates and complexed metals that were identified by classic qualitative inorganic analysis.  ICAP confirmed the content and concentrations.  Some were phosphate or hydroxide precipitates (high trace metals), some were organic iron complexes (insufficient bioavailable Fe).  Notice that there are striations where the color is appearing in at least 3 relatively thin layers and in different hues as though there are different densities of colored material; this is quite telling.  Might be interesting to know the nature of your recombinant protein.  The Brookhaven NL media already contains Mo, Cu, Co, Mn, and Fe chromogenic metals and significant phosphate levels so if these are not unusually high in concentration then complexation could exaggerate the intensity of the color and amount of putative metal.  Probably not any non-metal organic chromophore.  How well were the cell pellets washed before lysis and was there any trace of color in the cell pellets?  Was the lysis alkaline and does pellet washing with water reduce the color intensity?  

Thanks for the responses everyone! 

The recombinant protein has an N-term Profinity eXact fusion-tag (BioRad), no His-tag. There is a possible metal-binding domain in my target protein based on sequence (no characterization has been done yet). I have worked with metalloproteins before, but have never seen anything like this. It is quite incredible! 

The pellets did have some blue spots in them when harvesting the cells (they were not throughout the pellets but in small clumps similar to darker areas I am used to seeing in large-scale pellets). I washed each 1L pellet with 500 mL of PBS one time. The color did not go away after the wash. 

Lysis buffer is at pH 7.4 

I work on fluorescent proteins. I have seen pellets like this with one of our blue proteins.