1. Look like your target gene is a little bit long (and with GC-rich template), so:

You should use a (1) long-ranged (for longer product) and (2) high-fidelity DNA polymerase (for accuracy), and (3) with a buffer formulated for GC-rich DNA template (for GC-rich template). Commercial products including these properties are available.

2. Alternatively, you can add some additives such as 2% DMSO in your PCR reaction with GC-rich template. Except DMSO, some people also use BSA to increase the specificity and efficiency of PCR performance. An recent article published by BioTechniques shows that another novel additive, Bovine Thrombin (BT), is another effective additives. See attached article for more info.  

Try using additives like DMSO to tackle your GC rich problem, then for the amplification part I would suggest either USE Pfu Pol from Stratagene or Q5 from NEB..they both work great for amplicons upto 11kb..

remember that the DMSO conc. should not cross a maximum of 5% of the total reaction volume, ideally use 1-2%.

Addition of MgCl2 extra boosts up your amplification too.

good luck!!

study halophilic and alkaiphic bacteria

I agree with Sudheesh, although in our hands Phusion polymerase works better than Q5. 

If DMSO fails you: Our standard approach for high G+C actinomycetes is Phusion polymerase with 2.5M (yes, the concentration is correct) betain (+ 3% DMSO if that fails). Works well for with G+C contents up to 74%. The downside is a steep increase in the error rate, so you might start with 0.5M betain...

I suggest Q5 would be best  because your Amplicon though is of medium size but its increase GC content could be a problem. So to sort that u need Q5 HF GC Enhancer which have a very less error rate and very high proof reading capacity..And I agree for DMSO concentration with Sudheesh, as the increased DMSO concentration will give variations in primer annealing temperatures not suitable for any PCR assay..

Nowadays, a lot of commercial available DNA polymerases are powerful enough to easily handle amplifying a 3-kb DNA product. There is no need to go through such as Gibson Assembly PCR. 

Your question can be very simply answered by just saying that use appropriate primers. It may not be readily available. There are then 2 options only: either you make them yourself or ask somebody like Bangalore Genei to make them and supply it to you (in which case it may cost a bit more ). 

I have easily amplified a 3 kb fragments using a two-steps amplification PCR. You have to design your own primers that must be GC rich at the 3' end, and high Tm and choose an amplification program without the annealing step,  A programm containing denaturation and élongation steps. In this case you can set up the denaturation step at 95C for 30s and annealing/élongation step at 72 C for 90s to 120 s  for 35 - 40 cycles. This is particularly interesting if you have a non specific bands less than 3kb with others methods.

You can try this using the simple AmpliTaq. that is less expensive.

Alternatively, you can use Phusion (as previously recommanded by other colleagues) and add DMSO,

Another way is to use a mix of 2 Taq polymerases, we have amplified a fragments more than 15 Kb in lenght. or you can use a mix of two polymerase like Advantage GC 2 Polymerase Mix & PCR Kit (Clontech).