Hi James,

I asked a very similar question and started a discussion somewhere else I think. I am actually running end repair experiments to test T4 polymerase versus Klenow for the fill/blunt step and then running PNK separately at 37C. T4 polymerase is recommended by NEB to be used at 12 C for 15 min and with only 1U per ug. For Klenow, they recommend 1U per ug but at 25 C for 15 min. I am getting like 4-5% ligation efficiency on my fragments after Covaris shearing and using picogreen to measure mass and KAPA QPCR to measure "true library" in the mass of DNA. I have done experiments with libraries that have poor adapter containing fragments. I went back and sequentially "remade" the library from the same output of library construction. I found that ligation or adenylation and ligation alone did not improve the adapter containing fragments much. But when I end repaired again, my adapter containing fragments was much improved, but still not fantastic. I have done these "redo" libraries for over 400 samples. I strongly feel that Illumina designed end repair as a compromise to phosphorylate and fill simultaneously. Unfortunately, at these enzyme concentrations, blunting is not favorable no matter what! I calculated that in some of my preps, T4 polymerase was 25 fold excess over what is recommended by NEB...and it is done at 20C which is 8 degrees higher than what they recommend. I would bet Klenow alone (1U per ug DNA) at 25C with PNK for 25 minutes will work better at getting blunted ends than with T4 polymerase. If you go back to the older Illumina protocols, they use cocktail with all 3 enzymes together. I noticed with the NEBNext modules which I am now using, they leave Klenow out of the mix (KAPA does too for their end repair mix). I am glad to share with you guys my data next week!! Very few people are doing any library QC I would presume because they add PCR cycles after ligation. But you can imagine how this could severely impact dup rate, especially for target capture sequencing...which is what I do a lot of.

I think its only for related to size of fragments and nothing else

@james Gill

Hi James.....

First of all all polymerases polymerize only in 5' to 3' direction and not in 3'-5' as mentioned by you above for Klenow large fragments....The Klenow and T4 DNA polymerase are used for treating the 'imperfect blunt ends', if any, that are generally generated during 'non-enzymatic shearing' approaches. Like restriction enzymes the mechanical shearing doenot generates perfect ends that can be ligated...Shearing can generate blunt ends but the phosphodiester bonds can be broken either towards the 5' end or 3' end and some of them are not perfect for Ligations with the adapter...as you'll be doing in NGS protocol......

so the Kelnow and T4 polymerase are first utilised as 'exonucleases' to 'chew' the imperfect ends and then they work as 'polymerases' to make them 'perfectly blunt' ....(I am sorry to write this explanations in a sequential manner but these things must be happening very rapidly and simulateneously at different molecules).....Both these enzymes differ in their 'polymerizing' and 'exonucleolytic' capabilities and are used as a 'blend' to achiev the perfect blunt ends that serve as substrate for the next reaction...

You may be right in asking 'why Klenow is needed' and it may be possible in optimizing this reaction by using only T4 polymerase...but if they are working fine in a 'TEAM'...and doing the job perfectly...why worry about it....

see the link for some more info: http://lucigen.com/store/docs/literature/eLucidations/Means_End.pdf

Thanks Ajay,

I mispoke. Klenow polymerase does NOT catalyze DNA in the 3'-5 but the 5'-3' direction like you said. Even though Klenow is an E. coli product it has lost it's 5'-3' exonuclease activity but retains the 3'-5' chewing. Basically it's identical to the T4 polymerase and may be redundant or like you said contribute some unspecified benefit.

Hi James...

The two enzymes in the 'blend' doesnot contribute to an 'unspecified benefit'.....many things about the two are well know in the literature....They have different capabilites...for example....T4 DNA polymerase is more efficient in 3'-5' exonuclease activity (~200 time more than Klenow as claimed in the literature)....so controling a highly efficient enzyme will be difficult...and for 'Fill in' reactions Klenow I is known to be better that T4....In addition Klenow can do strand replacement while T4 cannot.....So keeping only one enzyme may not be good as in a sheared DNA there can be different types of 'imperfect termini' of the DNA molecules... and strand replacement activity of Klenow can make a difference ....depending upon the Phosphorylation status of your adapter....So 'blend' of the two would be ideal

Good to know. I didn't get any of that from your original link though. It seems Klenow has been around for a long, long time. All the articles I find are from the '80s.

Hi James,

I asked a very similar question and started a discussion somewhere else I think. I am actually running end repair experiments to test T4 polymerase versus Klenow for the fill/blunt step and then running PNK separately at 37C. T4 polymerase is recommended by NEB to be used at 12 C for 15 min and with only 1U per ug. For Klenow, they recommend 1U per ug but at 25 C for 15 min. I am getting like 4-5% ligation efficiency on my fragments after Covaris shearing and using picogreen to measure mass and KAPA QPCR to measure "true library" in the mass of DNA. I have done experiments with libraries that have poor adapter containing fragments. I went back and sequentially "remade" the library from the same output of library construction. I found that ligation or adenylation and ligation alone did not improve the adapter containing fragments much. But when I end repaired again, my adapter containing fragments was much improved, but still not fantastic. I have done these "redo" libraries for over 400 samples. I strongly feel that Illumina designed end repair as a compromise to phosphorylate and fill simultaneously. Unfortunately, at these enzyme concentrations, blunting is not favorable no matter what! I calculated that in some of my preps, T4 polymerase was 25 fold excess over what is recommended by NEB...and it is done at 20C which is 8 degrees higher than what they recommend. I would bet Klenow alone (1U per ug DNA) at 25C with PNK for 25 minutes will work better at getting blunted ends than with T4 polymerase. If you go back to the older Illumina protocols, they use cocktail with all 3 enzymes together. I noticed with the NEBNext modules which I am now using, they leave Klenow out of the mix (KAPA does too for their end repair mix). I am glad to share with you guys my data next week!! Very few people are doing any library QC I would presume because they add PCR cycles after ligation. But you can imagine how this could severely impact dup rate, especially for target capture sequencing...which is what I do a lot of.

Hi James,

Here is a small experiment I did looking at library quality. I just think ER will never be perfect. I did not end up comparing ER with and without Klenow, but 2 commercially available ER modules leave it out completely now. I am guessing there is probably good reason for this, although I don't know why yet.

https://www.researchgate.net/post/For_preparation_of_next-generation_sequencing_libraries_why_is_end_repair_performed_at_20_degrees?ev=tp_feed_post_xview

Thanks for the link Kenneth!

I was wondering if it would make sense to use the enzyme mix and then cycle between the three temperatures: one min at 37°C, then one min at 25°C, and one minute at 12°C, then back to 37 etc. Question would be how many cycles.

(PNK at 37C, T4 polymerase at 12 C, and Klenow at 25 C)

Hi Natascha, this would be interesting to compare to standard ER. Of course the best way would probably be each enzymatic step separately, but none of us have this kind of time...even if its automated!

I might try the Fast DNA End Repair kit from Thermo Scientific. They recommend an incubation for 5 min at 20°C (not to exceed 20 min). I could do a parallel reaction cycling between the three temperatures (maybe 1 min at each temp, for a total of 5 cycles = 15 min total) and see if it will work better.