Why would ligation fail with different isoschizomer?

[anjama]

We're having a weird issue with trying to ligate Illumina adapters to DNA extracted from cattle egrets (a bird species). Doesn't matter if the DNA is extracted from blood or tissue. Initially, when we were learning how to do library preps a year ago, the DNA was digested with the EcoT22I enzyme, and it worked beautifully. However, we didn't have of the EcoT22I for the particular project involving cattle egrets, and the main source of it doesn't produce it anymore. So we resorted to purchasing different isoschizomers to try (NsiI and Mph1103I). They appeared to fragment the DNA, but the ligated DNA failed to amplify during the PCR stage. We went through a lot of troubleshooting to rule out different causes and couldn't find anything wrong. So we then tracked down a different source of EcoT22I, and it worked beautifully. So the question is, why might this be? Product pages for all three enzymes show the same methylation sensitivity, at least for the methylation sites listed. Is there something else about isoschizomers that could cause issues? Thanks

[nucacidhunter]

Possible causes:
1- Ligation was not successful which could be due to differences in RE buffer composition not being suitable for the ligase. This could happen if restriction reaction was not cleaned up and followed straight to ligation.
2- RE inactivation differences which could result in RE activity during ligation and cut the ligated adapters.