I want to do crystallization for the neutral protease but the problem is protein solubility very less so the concentration of my protein in the solution is only 0.5 mg/ml.
It depends on the solubility of your protein. Some are more soluble than others. If you don't know how soluble your protein is try the so-called pre-crystallization screen (PCS). You may get the solutions e.g. from Hampton Research. If you get precipitation using these solutions this would mean that the concentration you have can be suitable for further exploring using crystallisation screens. If not, it is probably soo low. You may also try to make your own solutions using for example a series of PEG or ammonium sulfate (or both) concentrations and 3-4 different pH values, set up a screen of your own and check if you get any precipitation (or crystals if you are lucky!).
It is difficult to get crystals from low concentration of the protein. As it is suggested, it is good to optimize expression conditions, solution conditions for enhancing protein solubility and stability, and finally conditions for crystallization of your target protein. I personally recommend a high-throughput screening platforms to identify good expression, solution and crystallization conditions. It is good to apply a factorial design to check the effect of various parameters at different levels simultaneously.
I would try different buffers, additives and salt concentrations to increase the solubility of your protein and then concentrate it and try crystallization.
I myself made no good experience in trying to crystallize proteins at this low concentration.
in case of b)
you can try to improve expression levels or switch to a different expression system (E.col, Yeast, Sf9, CHO, HEC cells) depending what your lab can provide. But even then you should try to improve solbility by different buffers, pHs and so on.
I had one case where my protein had a low solubility in my buffer. I could not concentrate it to more than 5 mg/mL, otherwise I got precipitation.
I then used microscale thermophoresis (I used a device from NanoTemper) to determine melting temperature of my protein in different buffer conditions. I realised that in buffer conditions where my protein is more thermostable it was also more soluble. That might be also heloful in particular because you don't need much sample and not so high concentration of your protein using this device.
The initial information you provided is not enough to give you a complete and correct answer. Neutral proteases work in wide range of pH 5.5-9.0, so, you have more freedom in choosing the correct pH. The first step would be to calculate theoretical pI of your protein (http://web.expasy.org/protparam/ will do it for you).
The next step would be to move from that value lower or higher, and try different buffers in those spaces (acidic or basic).
Salt in most cases may increase solubility
Substrate or inhibitor like compounds may increase solubility.
You can build your own screen (pH, salt, additives) and run a thermal shift assay, to identify the best conditions for your protein.
Finally, we had crystals from protein solutions with concentration as low as 0.3 mg/ml.
During crystallization, a protein concentration is important for nucleations. The size of the crystal depends on the overall amount of the protein in the drop (a big drop - a big crystal). If you can induce nucleation with low the protein concentration, you can get big crystals in larger drops. Or you can use seeding, transferring seeds from small drops to bigger drops.