Ligases

Join a 5’ phosphate to a 3’ hydroxyl
Requires energy; the reaction is not favorable on its own
Not glue. It is a list of requirements, and every one of them is a way to fail

What a Ligase Requires

T4 DNA Ligase

No homolog in E. coli, the phage brought its own
Requires ATP: one per bond. Tired buffer is a leading cause of dead ligations
The workhorse for joining ends made by restriction enzymes

5’-cttcgCTGCA              Ggctggcaga…-3’
3’-gaagcG        +     ACGTCcgaccgtct…-5’

↓ T4 DNA Ligase: the same TGCA overhang on both ends, so they anneal

5’-cttcgCTGCAGgctggcaga…-3’
3’-gaagcGACGTCcgaccgtct…-5’

The PstI site is back. Cut it again and the insert comes out

A nick, not a gap. And held in register.

Blunting: the other direction

Same plasmid, same EcoRI. But this time, instead of chewing the overhangs off, fill them in with Klenow and dNTPs, then ligate. What is the product?

5’-…gatttctgGAATTCgcggcc…-3’
3’-…ctaaagacCTTAAGcgccgg…-5’

↓ EcoRI

5’-…gatttctgG-3’        5’-AATTCgcggcc…-3’
3’-…ctaaagacCTTAA-5’        3’-Gcgccgg…-5’

↓ Klenow + dNTPs, fills in the recessed 3’ ends

5’-…gatttctgGAATT-3’        5’-AATTCgcggcc…-3’
3’-…ctaaagacCTTAA-5’        3’-TTAAGcgccgg…-5’

↓ T4 DNA Ligase, blunt to blunt

5’-…gatttctgGAATTAATTCgcggcc…-3’
3’-…ctaaagacCTTAATTAAGcgccgg…-5’

Four base pairs gained: where mung bean lost four. The site is dead either way

The other two ligases

E. coli and Taq DNA ligase, NAD⁺, not ATP
Both seal nicks only. Not blunt ends, and not 4-base sticky ends
The native job: repair, and sealing Okazaki fragments
It is the ligase in Gibson and in ligase chain assembly. Use the thermostable one
Run hot, Taq ligase also refuses a mismatched junction, sharpest at the 3’-OH