I was once purifying inclusion bodies of protein containing a lot of cysteine residues and they were red, probably due to iron binding. My friend's protein was black in inclusion bodies because of it's zinc binding properties. Normally it should be white.

Full bacterial pellet can change its color in response to expressed protein's toxicity. It is easily observed during expression of different fab fragments. Normal pellets are gray/yellow/greenish and these ones have high yield of protein. When it changes to slightly orange/purple the yield is dropping. But Fabs are expressed to periplasm so I cannot be sure if it works for proteins expressed to cytoplasm.

Now, I could be wrong, but I don't believe it has to do with toxicity, solubility or yield of a protein.  Let me share my experience with a colored pellet:

 I was purifying a protein that I found out later to bind to iron sulfur clusters.  I started to notice that the pellets had a brown hue and sometimes I could see black color in streaks throughout it.  When I purify other proteins that do not bind metals, the pellets were white.

My guess is that it may have something to do with a property of the proteins you are purifying.  When you purify the proteins themselves, are they colored as well?  

from personal experience and I don't claim that it is always correct:

normally E. coli pellets are brown.

white pellet is clear sign of  of inclusion bodies

if the protein is toxic I'd expect smaller, fluffy pellets and a lot of foam

never seen anything black yet..

- lysate appearance is much more informative. 

Hope this helps

I also think this may have something to do with inclusion bodies, after sonication I used to also see blackish pellets when trying to work with a very insoluble protein.

@Valeria J Carabetta.....I agree with your experience...Metal binding proteins do tend to give colour...But my question is what about when protein has no metal binding properties....Thanks for your input

@Misha Soskine & Gary Steven McDowell....Thanks for sharing your valuable experience...

I would like add some more details observation of clone which tends to give me white pellet....Just after freshly cloning and transformation...I used to get normal brown pellet....but very little pellet....I assumed little pellet was due to toxic nature of my protein...As protein yield was sufficient for my experiments so i dint bother much...But in subsequent purification....I started noticing pellet color tends to become more whitish and protein yield decreased to some extent..Also bacterial pellet volume increased too....I'm very much curious to know can i learn to correlate bacterial color pellet with toxicity, solubility or yield of expressed protein

I agree with the white pellet being a sign of inclusion bodies (which are made of insoluble protein). Now, you can have any shade between brownish and white depending on the amount of inclusion bodies. You should compare with a control with your empty expression vector, to know what the "normal" color should be. Finally, if you get irreproducible blackish pellets, think of everything that can contaminate your sample (like damaged O-rings in a French press cell or in centrifuge bottles, which can generate a kind of black dust, or corroded metal form the probe tip if you sonicate)

I was once purifying inclusion bodies of protein containing a lot of cysteine residues and they were red, probably due to iron binding. My friend's protein was black in inclusion bodies because of it's zinc binding properties. Normally it should be white.

Full bacterial pellet can change its color in response to expressed protein's toxicity. It is easily observed during expression of different fab fragments. Normal pellets are gray/yellow/greenish and these ones have high yield of protein. When it changes to slightly orange/purple the yield is dropping. But Fabs are expressed to periplasm so I cannot be sure if it works for proteins expressed to cytoplasm.