Klenow Assembly Method: Seamless cloning
The Klenow Assembly Method (KAM) for the seamless cloning of overlapping double stranded DNA fragments: Very cheap alternative to the Gibson Assembly
Unpublished: V1, 1-2 David MD Bailey* and Mohamed YH Mohamed. Chemical Engineering and Biotechnology. University of Cambridge.
For seamless assemblies.
1 Create double stranded fragments with overlaps of 15-30 bps
2 Clean fragments away from any plasmid template, primer–dimers and most importantly dNTPs
3 In a 5ul reaction comprising 1/2ul NEB Klenow M0210s + 1/2ul NEB2 buffer, place 30ng of each fragment- regardless of size: suggest < 6 fragments.
4 Incubate at 37C for 20-60mins. 30 mins is usually fine.
5 Transform 2ul or freeze then transform.
Gel purify any fragments showing additional spurious bands especially primer-dimers.
For assemblies containing more than three PCR generated inserts, always gel purify.
Overlaps of less than 15bp reduce effectiveness, overlaps of >40bps depress efficiency: we have had success where one end of a fragment overlap was only 8bps for a two fragment assembly. I recommend overlaps of 20-25bps as then you have the option of fusion PCR of fragments prior to assembly. With multi fragments assemblies, sometimes it is helpful to reduce the number of bits by fusion PCR-see below.
The fragments can be a mixture of restriction digested and PCR fragments or entirely PCR products. Restriction digested vectors do not need phosphatasing as no ligation events occur in vitro. Also any inserts evicted by restriction digestion do not need removing. Inactivation of the restriction enzyme is only necessary if the sites will exist in your recreated construct. I usually heat inactivate the enzyme out of habit.
The exo activity of Klenow is exploited so dNTPS must be removed.
Remove primer dimers as these can compete in the reaction, but their removal is not strictly necessary as their inclusion just depresses efficiency, i.e. you will need to screen by PCR to find the correct clones.
Aim for 5ul final, however it does not matter if the volume exceeds this, i.e. you can increase the volume up to 10ul total, without compensating the buffer or enzyme, as only a small drop in efficiency occurs.
I have assembled a four fragment assembly using a 60bp,10kb,1.5kb and a 6kb fragment, the same ammount of each, i.e. 30ng each. If one of your fragments is very dilute, put 30ng of the others in then put as much of the dilute fragment up to 10ul total. You wont get many colonies- but how many do you need –I always pick two or three. One of these to work up for sequencing, the others as frozen cell pellets, ready to process if the first example fails sequencing.
We have used KAM for assemblies of up to 8 fragments. I recommend you start with a maximum of 4-5 fragments. I reckon that for every additional fragment you introduce, the number of colonies reduces by factor of 5, so if a 2 fragment assembly gives you 20K CFU, an 8 fragment will only give you 1: you only need super competent cells for 6-8 fragment assemblies.
If you produce vectors by PCR, I recommend you break the vector up into several overlapping fragments of a size that allows efficient gel purification, i.e. 3kb or less, alternatively I digest the crude PCR products with DpnI for at least an hour to destroy residual template vector (1ul of DpnI into 50ul PCR reaction) and even better digest overnight. If you do produce the vector as several fragments by PCR, design the overlaps to hit key areas of the plasmid, i.e. the Ori and or the selection cassette. The logic is that only functional overlaps will be selected for, i.e. any oligo derived mutations contained within the vector backbone will be selected against; and so there’s no need to sequence these junctions as they have been functionally tested. I suggest you sequence any parts of the vector backbone that confer other properties such as elements of shuttle vectors, as very occassionally polymerase induced errors occur. So the message about amplifying vector backbones is do not be complacent. Restriction digested plasmids that have not undergone gel purification do not gain U.V. or polymerase induced mutations.
Use Q5 pol or similar proof reading pol (hot start is best) for PCR.
Sequence, at least, across all oligo overlap junctions (unless the overlap region is a non functional hinge), or you have a suitable functional assay. In my experience Q5 does not introduce any mutations, but the oligos are the source of potential mutation.
It is possible to produce a room temperature stable dried down formulation, where you simply add 5-10ul of DNA and incubate at 37C. The formulation below is stable for at least 6 months.
KAM Master mix for drying
58ul dd water
20ul NEB 2 restriction buffer
100ul 30% trehalose (T9531 Sigma-Aldrich)
2ul Dilute coloring see below
20ul Klenow (NEB M0210s)
Aliquot 5ul of master mix into PCR tubes or multi well plates , leave with the lids open and incubate for 24-48 hrs at 37C. Close lids and store in the dark before use. When adding DNA to the dried down formulation, re-suspend the pellet by pumping up and down- The Klenow mixture does not seem to mind.
The colour allows you to see where the dried pellet of reagents are, so is very convenient but not strictly necessary. I have used Sainsburys red food colouring successfully, their green colouring kills the reaction- be warned- test your colorants on a small scale.
Test colorants and assembly mixtures using gel electrophoresis by analysing the test mixtures on two small overlapping fragments of differing sizes, run the same fragment combinations untreated as a gel electrophoresis size control lane. Look for hetero-dimer formation in the treated combinations which indicates a successful reaction. Remember to heat kill the assemblies before this analysis, heat kill for 20 mins at 65C. Then leave at romm temperature fpr five minutes to allow the re-assembly of the heat dissociated fragments, before gel loading.