Date of Award

Winter 12-15-2014

Author's School

Graduate School of Arts and Sciences

Author's Department


Degree Name

Doctor of Philosophy (PhD)

Degree Type



The research presented herein focuses on the kinetics and inhibition of the KDM4 subfamily of Jumonji C (JmjC) domain-containing histone demethylases (HDMs). Belonging to the larger class of alpha-ketoglutarate (alpha-KG)-dependent, non-heme iron monooxygenases, the JmjC-HDMs remove methyl groups from mono-, di-, and tri-methylated histone lysine residues through an Fe(IV)-oxo-catalyzed hydroxylation reaction. JmjC-HDMs have been found to play integral roles in the maintenance of genomic integrity as well as in the regulation of transcription. Three KDM4 members were studied: the mixed H3K9/H3K36 demethylases KDM4A and KDM4C, and the pure H3K9 demethylase KDM4E. KDM4C is a hypoxia-inducible factor 1 (HIF-1) target gene and a putative oncogene, while KDM4A is found to be overexpressed at the protein level in multiple cancers. KDM4E is closely related to KDM4D, which has been found to regulate the tumor suppressor p53.

Three in vitro enzyme activity assays, including a novel continuous O2 consumption assay, were employed to determine the kinetics of the three HDMs. The kinetic parameters of the three substrates, with an H3K9me3 peptide in place of the full histone substrate, were determined. All three HDMs were found to have low apparent affinities for O2, suggesting that these enzymes may act as O2 sensors in vivo. Alpha-KG was found to inhibit KDM4C competitively with respect to O2, with KDM4C displaying optimal activity in vitro at alpha-KG concentrations similar to those found in cancer cells. Additionally, a 2.1 Å structure of KDM4A in complex with Ni(II) and alpha-KG was solved.

Various avenues were explored in the study of JmjC-HDM inhibition. First, a small library of simple primary- and secondary- substrate analogs was synthesized and screened against KDM4E. From this screen, two small molecules were identified as promising candidates for modification in future KDM4 inhibitor design studies. The analogs contained carbon-carbon triple bonds or 1,2,3-triazole groups. We find the triazole-containing compounds to be stronger inhibitors of KDM4E. Several of the alkyne-containing compounds were tested in KDM4-templated Huisgen 1,3-dipolar cycloaddition studies, with no evidence of enzyme-driven triazole formation.

Using a peptidomimetic approach, two novel, modified histone H3(7-11) peptides were synthesized and tested for inhibition against KDM4A and KDM4C. Lys residues were modified to incorporate two propargyl groups on the terminal nitrogen, with one of the peptides having an additional methyl group on the terminal nitrogen to make it quaternary. We find enhanced inhibition of KDM4A and KDM4C using the positively charged peptide vs. the neutral peptide. KDM4 members show no hydroxylase or demethylase activity toward the methylated peptide, suggesting that modification of this peptide could lead to enhanced specific inhibition of JmjC-HDMs. No evidence of HDM-driven triazole formation was found for either peptide in studies with KDM4A- and KDM4C- azide adducts.

The inhibition of KDM4A with respect to O2 for the pan-selective JmjC-HDM inhibitor JIB-04 was explored. KDM4A inhibition by JIB-04 increases with decreasing O2 concentration, suggesting that JIB-04 may inhibit KDM4A in part by disrupting the binding of O2. JIB-04 isomers were modeled into the KDM4A active site, revealing the predicted basis for JIB-04 isomer-specific inhibition of JmjC-HDMs. Finally, a 3.1 Å structure of KDM4A in complex with JIB-04 was solved, revealing only a fragment of the inhibitor in the enzyme active site.


English (en)

Chair and Committee

Liviu M Mirica

Committee Members

Michael Gross, Joseph Jez, John-Stephen Taylor, Timothy Wencewicz


Permanent URL:

Included in

Chemistry Commons