The fact your protein is insoluble is the first issue. There might be enough soluble to purify, though the majority is insoluble. Lowering the temperature to 18oC before induction is the first thing to try. You can also add 10 mM (710uL/L) DMSO, which upregulates protein folding. There are also strains, such as Arctic express or those with GroEl/ES chaperones, that have enhanced protein folding which may help. 

Your cell lysis buffer is the correct pH for Ni-NTA - this deprotonates the side chain so it binds to the Ni2+ ions. If you go lower it does not bind and so low pH can be used as an elution step. HIgher pH may make your protein unstable. However DTT is not compatible with Ni-NTA as it reduces the Ni2+ to metallic nickel which will not bind the protein. You should use a little more imidazole - 30 mM is suitable for binding without allowing other proteins to bind.

I think you can use lower IPTG and slower shake speed . Or change a tag, like GST or MBP to increase your protein solubility ,which may change the PI also.

Change  a bacteria strain may also help

Here your protein goes into inclusion body it may be because of it not get properly fold during synthesis. You didn't tell how much conc. of IPTG you  are using?

You can optimize your expression with different IPTG Conc. with different Temperature with different induction time.

generally lower IPTG conc. at lower temp. ~24- 16 °C for 16-24 Hour incubation will help you to get your protein in soluble form.

If you still not get your protein you can also checked with different lysis buffer.

Best of luck

DTT at this concentration is not compatible with NiNTA purification. Most of the available resins get reduced wit as low concentration as 1mM. If you want to use nickel purification you have to change your reducing agent to beta mercaptoethanol. I would suggest not to use more than 15mM. b-ME is also more stable than DTT.

pH should be chosen based on your protein pI and stability range. It is said that buffer pH should be1 unit higher or lower than pI of your protein. It this case, when you are planning to use NiNTA purification you should be higher than 7 (below 7 Histag loses it's affinity towards resin because of the histidine protonation).

It is possible that your protein expresses in soluble form, but it precipitates. I would suggest using pH 7 and no imidazole in the lysis buffer. It is far enough from the pI of your protein and still compatible with NiNTA. Then you can wash it with 20mM and elute with 250mM imidazole.

If it is not going to help you will have to play with different fusion tags (e.g. MBP), temperature of induction, IPTG concentration, E.coli strains ect.

Is any of your protein in the soluble fraction? Does it contain any disulphide bonds? Can you say anything more about your protein?

@O'Neill.....Thanx for these tips..When I will add DMSO,, at the time of Inoculation or induction...??

@Dan.....I have no other expression host,....I will try with this low IPTG conc and very imp step is low shaking,,thanx for these tips....can you tell how much low conc of IPTG I can use..

@Ansari.......I'm using 500uM IPTG with 16"C,...ok I'm planing to decrease this conc to 100uM at 16"C...              Is it right that protein pI and Buffer pH will not matter for this precipitation...??

@Dawid Deneka....Is PMSF is ok at the time of sonication and in lysis buffer..

I'm trying with this single protein..it's not even coming to supernatant after centrifuge(13000rpm,30min,4"C)..after SDS-PAGE..every time I saw, it is in pellet..

What I discusses here is a single protein but, I made two recombinant variant N-ter and C-ter His tag, both of these protein is going to pellet.

I have another protein also that is working fine(in soluble fraction)..These all are bacteriophage proteins.

no it have no cys residue....also I checked for trans-membrane domain and signal peptide___It's dose not contain these things..

can you tell me what conc of NaCl, I can use for this lysis buffer... or what should be the lysis buffer composition??

PMSF is compatible and recommended (0.1-1mM). Be sure to add it just before lysis. It decays in water very quickly (half life~ 0.5-1h).

If you have at least some of the protein in the soluble fraction you can try to purify it. Sometimes it may not be even visible on gel. It is quite common that part is soluble and the rest goes into inclusion bodies.

If the change in the buffer and expression conditions doesn't help try MBP or GST as a fusion partner.

One more thing you can do is refolding. Isolation of inclusion bodies is easy and efficient. If you will not be able to acquire soluble expression it is one more option.

@Dawid Deneka......thanx

I suggest a 0.5,0.2,0.1mM gradient IPTG.

Here the problem is that ur protein is not soluble. Sometimes adding a fusion partners to your protein will help in the expression of your interested proteins. The best fusion partners are, Thioredoxin A, GST, MBP. If you have plasmids like pETM-20, pETM-Z and pETM-ZZ, you can try them. In most of the cases they help. 

There is a large list of things you may want to try or check, and some of them are listed below:

1. the presence of transmembrane/membrane-bound structural elements: this is not easy to tackle, but in some cases, such as in the overexpression of eukaryotic proteins, some of the structural elements of the protein can be bound or integrated into the cell membrane. In these cases, protein solubility is dramatically reduced. Your options in this scenario are to make a truncated protein without these regions (which can be identified with some confidence using servers available elsewhere) or to add detergents into your buffer.

2. the amount of IPTG and the overexpression time: sometimes, low temperature is just not enough. Bacterial cells do not have sophisticated degradation systems or chaperones for assisting folding as their eukaryotic counterparts, and therefore cannot deal with folding of complex proteins if the protein is being highly over-expressed. The fact that your protein does over-express but is insoluble suggest this is a plausible scenario, and therefore reducing the time of overexpression and the IPTG concentration can be important to improve the solubility of your protein, as Mohd already mentioned. Also, as Ellis suggested, using bacterial strains that include chaperones can be of great help.

3. the source of your protein: Might be important to check whether the protein gene has rare codons when compared with the bacterial codon usage, because this might be slowing down folding and promoting aggregation. Codon usage differences are obviously frequent when the protein gene comes from an eukaryotic source, and strategies such as using BL21 Rosetta strains are very common.

4. refolding as a strategy: if your protein does not have any (or many) disulfide bonds, you might be able to refold your protein from the inclusion bodies, as Dawid mentioned. Keep in mind that urea is mainly used in these cases because is cheaper than guanidine hydrochloride, and you will use a lot, depending on how much recovery you can get after refolding the protein. The presence of many disulfide bonds should be taken as a challenge that may take just too long to optimize with very little chances of success.

5. effect of the his-tag: have you ever tried to over-express your protein without the his-tag? although in my lab many of the proteins we study are over-expressed with his-tags, we (and many other researchers everywhere) have found that the presence of this affinity tags can decrease protein solubility and lead to aggregation or precipitation. As an example, a dimeric phosphofructokinase-2 that we used to study has been always purified through size exclusion and analog-affinity chromatography without his-tags. When we attempted to add a his-tag to make its purification easier, most of the over-expressed protein was insoluble, and the soluble protein yield was dramatically reduced. Moreover, we had to remove the his-tag as fast as we could after the nickel affinity chromatography due to severe aggregation in 24 h. The use of other tags that have been mentioned here can be of great help, but never step away from old-fashioned many-step purifications, because this can be one of those cases where it actually works best.

Good Science!