Characterization of the substrate specificity of a human 5-hydroxymethyluracil glycosylase activity

David Baker, Pingfang Liu, Artur Burdzy, Lawrence C. Sowers

Research output: Contribution to journalArticlepeer-review

40 Scopus citations


The oxidation of pyrimidine 5-methyl groups, derived from either thymine or 5-methylcytosine, can generate 5-hydroxymethyluracil (HmU) in DNA. An activity from HeLa cells that removes 5-hydroxymethyluracil (HmU) from DNA has been partially purified and characterized using a battery of oligonucleotides containing modified bases. This partially purified activity preferentially removes HmU mispaired with guanine. The HmU repair activity also acts on uracil and fluorouracil but not 5-substituted uracil derivatives with halogens larger than fluorine. However, neither mispaired thymine nor ethenocytosine are substrates. HmU is readily removed when paired with guanine, hypoxanthine (deoxyinosine), and purine (deoxynebularine), but not from single-stranded substrates. Upon the basis of these substrate preferences, we conclude that (1) the mispaired HmU repair activity is distinct from previously reported glycosylases including UDG, TDG, MUG, and SMUG1 activities, (2) the binding pocket is highly selective for the 5-hydroxymethyl group, and (3) the preference for mispaired HmU derives from reduced thermal stability of the mispair, as opposed to selective recognition of the mispaired guanine residue in the opposing DNA strand.

Original languageEnglish (US)
Pages (from-to)33-39
Number of pages7
JournalChemical Research in Toxicology
Issue number1
StatePublished - 2002
Externally publishedYes

ASJC Scopus subject areas

  • Toxicology


Dive into the research topics of 'Characterization of the substrate specificity of a human 5-hydroxymethyluracil glycosylase activity'. Together they form a unique fingerprint.

Cite this