I am trying to purify a bacterial cytosolic protein from E.coli as heterologous host. It has a C-terminal 6xhis-tag. I have previously purified reasonable yields on NiNTA or Cobalt columns, but observed a large fraction (>95%) remains insoluble under regular purification ocnditions in a variety of buffers. I figured these might be inclusion bodies, so tried solubilizing in 8M urea. However, protein still appears in the pellet. Now going to try 6M Gua-HCl, but am wondering about this behaviour. Any suggestions?