I wont try to amplify 40kbp from genomic DNA by any sense, any polymerase. Less complicated template, like phage, may, just may be able to PCR such big amplicon. In general, longer than 10kb PCR, you will get 50% chance meeting problem. 20 kb PCR? 1% chance it will yield correct amplicon.

Not Taq, you have to use a high fidelity polymerase. An example would be Q5 from NEB, or Kapa.

I wont try to amplify 40kbp from genomic DNA by any sense, any polymerase. Less complicated template, like phage, may, just may be able to PCR such big amplicon. In general, longer than 10kb PCR, you will get 50% chance meeting problem. 20 kb PCR? 1% chance it will yield correct amplicon.

That's a very large product.  It's worth the investment in a high fidelity enzyme.  I've had good luck with Pfusion by Thermo Fisher.  You might want to consider contacting your product sales representatives and asking for samples of their best long product enzymes to test.  There is a good chance they will give them to you for free if you are willing to share your results.

For large DNA amplification better to use Phusion high fidelity enzyme. 

I agree to Artu Burzynski and Ziguo Zhang, there is no chance to amplify 40 bp of DNA by PCR. The 2 best Polymerases for Long-Range-PCR are Q5 and Phusion. But they deal only with sequence up to 20 kp in best case.

Maybe you amplify the sequence in multiple steps in parts of 5 kp. Afterwards you can ligate the DNA-parts together with seamless Ligation methods like SLiCE or something else.

I have used Expand Long Template PCR System from Roche to amplify up to 12kb from bacterial DNA (attempted 18kb, but could not get it to work). As suggested above, I would not recommend amplifying anything above 10kb and if the template is not good (eg. high GC content), wouldn't go above 6kb.

Phusion Polymerase

As other colleagues mentioned before: PCR is probably not a good option. One must wonder why would you need to do that (making a mouse?). Without knowing more about your project or the species you working with I can suggest to order one or more BACs that contains the region of interest  and work from there. I explained how to do this as an answer to a prevous question that you might find useful.

https://www.researchgate.net/post/How_do_I_amplify_a_5kb_intronic_fragment

Hi,

I think all the comments are important to consider.

But thought this might help you as well. 

http://www.epibio.com/docs/default-source/forum-archive/forum-08-1---high-fidelity-pcr-amplification-of-dna-up-to-40-kb-using-the-masteramp-extra-long-pcr-kit.pdf?sfvrsn=6

thermos fusion

To amplify large fragment of DNA, high fidelity DNA pol is more preferred like phusion and Pfu. 

I agree with Juhi about the type of polymerases that can be used. but your fragment is to long to be amplified without problems even using such reagents. you should try to set up your PCR conditions carefully. 

You can try TAKARA primstar GXL polymerase,but still amplifying a 40kb fragment is not that easy.

i think any high-fidelity polymerase would be fine. Are you trying to amplify a gene cluster? if you think the gene is too big you can amplify its different parts, maybe 8*5kb and overlap them, or use Gibson assembly.

good luck

Try TAKARA  LA DNA polymerase. It is one of the best for long PCR up to 50kb.

Good luck

What is your end game? If it is sequencing, design internal primers  in smaller fragments and put the sequences together? 

Interestingly nobody asked... What is your purpose to amplify 40 kbps ? Maybe the way you chose is not the only way. Buy a BAC clone instead of amplification.

Why do you doubt that TAQ Polymerase may not work for large DNA PCR.

Using a special type of template PAC or BAC DNA preparation and purification procedure to quantitatively obtain undegraded strands we were able to consitently amplify 30-60 kb. The Eppendorf Triple-Master-Kit plus additional normal taq working polymerase did well. See our 2010 paper of Rocchi et al. in Human Gene Therapy 21: 1077-92.   If not familiar with preparing BAC/PAC DNA in agarose and pulsed field electrophoresis you may require 3-6 months or so to set up everything. Good luck

Might want to try 5-6 overlapping fragments amplification and finally full length product with Q5 or Takara LA. bests 

Hi again,

 I want to thank all of you for giving me your time and answering my question.

Thanks for shearing your experiences as well.

Very Cordially! 

Dear prof Artur Burzynski, 

For: Save yourself a lot of frustration and design new strategy. Do you suggest any other strategy instead of PCR to get a high copy number of my DNA, I have tried in vivo amplification and I had autoinduce my recombinant bacteria to increase the yield of fosmid DNA but the yield still not high and not enough for sequencing. So I am looking for in vitro strategy.  

Well, if I had to try (not something I'd recommend...) I'd choose q5, used for site directed mutagenesis of large plasmids it works okay... But 40kb of genomic DNA is not going to be easy. 

Dear Soledad Miranda-Rottmann

Thank you for answering

The aim of this part of my work is to do sequencing of high-throughput screened recombinant bacteria to find an untapped gene. and to do Hiseq I need an important quantity of DNA which I could not get after extraction, So I was thinking to amplify my fosmids with PCR.

 Dear Da Di 

Thanks for your answer, Is there any protocol for that ?

Dear Ozge Kuddar 

Thank you for answering

The aim of this part of my work is to do sequencing of high-throughput screened recombinant bacteria to find an untapped gene. and to do Hiseq I need an important quantity of DNA which I could not get after fosmid extraction from E.coli. I have tried the in vivo way to amplify the fosmids but the DNA yield was not that much So I was thinking to amplify my fosmids with in vitro way with PCR.

any suggestion!

Dear Sarah Leah Downs 

Thank you for answering, 

Yes I am going to sequence my DNA but I do not have any information about the sequence so I can not design primers, is there any option to design primers without knowing the DNA sequence? 

Dear prof Jai Ghosh 

Actually I found that there is high fidelity polymerase for large sequences and Taq in not one of them.