Use SDS-PAGE on the whole cells, even before sonication, to see if there is a new band corresponding to your protein's mass when the cells are induced with IPTG, compared with uninduced cells. If you can see such a band, then you can be confident that your protein was expressed.

Test that sonication was effective in breaking the cells by examining them under a microscope after sonication. If sonication was effective, you should see lots of debris but very few intact cells. If you have access to a French press or microfluidizer, use that instead. It's more effective than sonication and doesn't heat up your sample.

Following cell breakage, do a low speed centrifugation to pellet large debris and unbroken cells. Then do an ultracentrifugation to pellet all the membranes and insoluble protein. Use SDS-PAGE to compare the amount of your protein present in each of the pellets and supernatants. If your protein is in one of the pellets, then your problem is that the protein is insoluble. If your protein is in the high-speed supernatant, then you should proceed to purification.

If your protein doesn't bind to the column, it may be because the His-tag was lost due to proteolysis, or because the protein is aggregated. Use EDTA-free protease inhibitor cocktail during the cell breakage to reduce proteolysis and keep the sample cold at all times.

If the protein is in the form of soluble aggregates, you may be able to reduce aggregation by increasing (or decreasing) the salt concentration during extraction, including a non-ionic detergent, or lowering the temperature of the culture during induction.

I agree with Adam, you can also try a batch under denaturing conditions (8M Urea or 6 M GdnHCl) to check this. The protein should bind to the Ni column under denaturing conditions. If you get no binding under denaturing conditions it may be that the his tag is cleaved off during sample preparation, than use protein inhibitors. For column purification: Equilibration Buffer (pH 7.4) 50 mM sodium phosphate, 6 M guanidine­HCl

300 mM NaCl Imidazole. Elution Buffer (pH 7.4), 45 mM sodium phosphate 5 M guanidine­HCl,  270 mM NaCl, 150 mM imidazole.

Do not use guanidine if you want to run sdspage of your fractions, use urea instead. 

I consider that you have optimized the IPTG concentration and temperature for maximum protein expression levels. Incomplete lysis leads to the loss of protein in the pellet fraction. You can perform lysozyme lysis followed by sonication for maximum protein recovery. 

Hi Dominique,

I think you want to higher your protein yield. So I assume you have your protein of interest but its was low.

I may give you some ideas.

First of all, if your protein is a human protein, you should consider the host bacteria strain whether it contains human amino acid codons or not? Bacteria and human are different in the amount of some codons for protein synthesis. Some of the eukaryotic codons are not common in bacteria.  But genetically modified E.coli BL21 strain rosetta could be used for this problem.

Second, if your protein of interest is a long protein such as more than 300-400 aa or larger. You have to let bacteria to complete the translation of the protein. To do so, I suggest you incubate your bacteria at less than 25 degrees for 18 hours. If you grow the bacteria at 37 they just skip translation of your protein, so your protein will be short. But if you slow the growth rate of the bacteria by decreasing the temperature, you can give them chance to complete your protein. 

I generally check my induction with dot blot by anti-his antibody. You can check more colonies in this way. After IPTG induction I just add SDS loading buffer onto the pellet of the bacteria (50 ul 5xSDS buffer + 200 ul ddH2O). Then I take 20-30 ul to apply onto methanol activated PVDF membrane and incubate with anti-his antibody for 1 hour. It is very easy and effective. 

I hope this helps

Apart from the advice given by others (especially Adam's about using a French press), try adding some DNase to the cell lysate to reduce viscosity and interference.

It is also a good idea to keep track of your purification by noting protein concentration (mg/ml), total protein (concentration times volume), activity of your protein (U/ml, this can be enzymatic activity, antibody binding or any other test you happen to have), total activity and specific activity (U/mg) after each purification step. This tells you where you loose activity.

Adam Shapiro's suggestions above makes a great process check list - let me throw more details:

      1.  100 mM [IPTG] stock solution - I'm sure you are using concentrations the others in the lab are using who are having good success

      2.  To facilitate larger number of samples (a matrix of host/growth/induction conditions) your cell pellets may be snap frozen and stored @ -80. 

      3.  Proteinase Inhibitor cocktails are available from Sigma (P8849)  in general a solution in DMSO (I used to make them myself - 10 mg PMSF/mL, Final concentration of DMSO ~1% or lower in cell suspension - final PMSF ~10 ug/mL). 

     4.  Most industrial processes will use either a French press or a micro fluidizer.

     5.  Confirms the expression of protein of interest.  If you have the antibody specific for target, a Western blot would confirm that expression.

     6.  I have preformed the cell lysis in a cold room.  I have tricked my lysate from a French press into a T75 culture flask with the bottom submersed in a salt ice - water bath (similar slurry as you'd use for making Ice cream). 

      7.  Then I have centrifuged creating an S15 supernatant to run over my column.

Good Luck

Looking forward to hearing from you

Best Regards - good luck