This document discusses epigenetic modulation through inhibition of histone demethylases like LSD1. It summarizes that:
1) Polyamino(bis)guanidines and polyaminobiguanides can inhibit the histone demethylase LSD1 in vitro and in human colon cancer cells.
2) These inhibitors are non-competitive inhibitors of LSD1 and promote increased histone H3 lysine 4 dimethylation.
3) One inhibitor, verlindamycin (compound 2d), re-expresses tumor suppressor genes silenced in cancer cells and reduces tumor growth in mouse models of human colon cancer, especially in combination with 5-azacytidine.
Turning Genes On and Off with Epigenetics: Can You Control Gene Expression
1. Flipping the Switch: Can You
Turn Genes "On" and "Off?”
Craig J. Kutz, 1,2 Steven L. Holshouser,1 Robert A. Casero, Jr.,3 Donald R.
Menick2 and Patrick M. Woster1
1Department of Drug Discovery and Biomedical Sciences, Medical University
of South Carolina, 70 President St., Charleston, SC 29425
2Department of Medicine, Medical University of South Carolina, 141 Ashley
Ave., Charleston, SC 29425
3Sidney Kimmel Comprehensive Cancer Institute, Johns Hopkins School of
Medicine, 1650 Orleans St., Baltimore, MD 21231
2. Lou Hawthorne, the Missyplicity Project and
Genetic Savings and Clone
Lou Hawthorne and Missy Rainbow Copycat
Allie and Copycat Rainbow and Copycat
4. DNA Methylation and Histone Modifications:
Compartmentalization of the genome into domains of
different transcriptional potentials (important for
development and differentiation).
hypoacetylated histones
Dense DNA methylation
Histone H3-K9 methylation
Histone H3-K27 methylation
hyperacetylated histones
Low DNA methylation
Histone H3-K4 methylation
From P. Vertino
5. Epigenetics
• Heritable traits that do not involve changes to
the underlying DNA sequence.
• Epigenetic changes can lead to gene
silencing or gene activation, depending on
the chromatin mark involved.
• Regulated by changes in DNA CpG
methylation and histone protein modification.
6. Control of Gene Expression Through Post-Translational
Modification of Histones
Histone PTMs are coordinated with methylation of
CPG Islands at DNA promoter sites.
LYS
or
ARG
NH2
LYS
or
ARG
HN
Histone
Tail
Histone
Tail
Histone
Tail
Histone
Tail
histone acetyltransferase
histone lysine methyltransferase
protein arginine methyltransferase
histone deacetylase
histone demethylase
Altered
Gene
Expression
Post-translational
Modification
epigenetic writers
epigenetic erasers
epigenetic
readers
7. Components of Cellular Epigenetic Modulation
Via Histone Post Translational Modifications
8. Epigenetic Writers, Erasers and Readers
Epigenetic Writers
Histone acetyltransferases (HATs)
GNAT family (Gcn5, PCAF and ELP3)
p300/CBP family (p300 and cyclic AMP-responsive element
binding protein)
MYST family (Tip60 and MYST 1-4)
Histone lysine methyltransferases
Protein lysine methyltransferases (KMTs) – SET1, SET2,
SUV39, EZH1,EZH2, PRDM, other SET, non-SET
All but non-SET contain SU(VAR)3–9, enhancer-of-Zeste, Trihorax)
Protein arginine methyltransferases (PRMTs)
PRMT Type I, PRMT Type II
Histone lysine phosphorylases (H3 Thr3, Ser10, Thr11, and Ser28)
11. Transcriptional control via histone lysine methylation
Methylation of specific lysine residues on histone tails can lead to either transcriptional activation or
repression
17 lysine residues and 7 arginine residues have been shown to undergo methylation/demethylation
10 lysine methyltransferases and nine arginine methyltransferases are known
Lysine demethylases:
lysine-specific demethylase 1 (LSD1) bound to CoREST complex – specific for
H3K4me1 and H3K4me2 (activating chromatin mark)
lysine-specific demethylase 2 (LSD2) - specific for H3K4me1 and H3K4me2 but not bound
to CoREST or another protein complex
LSD1 bound to androgen receptor – specific for H3K9me1 and H3K9me2 (a deactivating
chromatin mark)
Jumonji (JmjC)-domain containing demethylases
JHDM1A – specific for H3K6me1 and H3K6me2
JHDM2A – H3K9me1 and H3K9me2
Other Jumonji demethylases specific for trimethylated lysines
12. Epigenetics and Cancer
• DNA methylation and histone modifications contribute to aberrant gene
silencing.
• A functional link of aberrant epigenetic gene silencing to the pathophysiology
of cancer has been established.
• Tumor-suppressor genes are frequently inactivated in association with
promoter CpG island methylation.
• Aberrant DNA methylation and histone modifications have been shown to
have potential in risk-assessment, early detection, disease classification
and prognosis prediction in a variety of cancers.
• DNA-methyltransferase inhibitors reactivate functional expression of tumor-
suppressor genes silenced in cancer.
Baylin et al. Nature Reviews Cancer 6, 107–116 ,
2006
13. Klose and Zhang, Nat Rev. Molec. Cell Biol, 2007 8, 307-318
a | The LSD1 reaction mechanism detailing the removal
of a mono-methyl group. LSD1 is proposed to mediate
demethylation of mono- and di-methylated lysine residues
through an amine oxidation reaction using FAD as a
cofactor. Loss of the methyl group from mono-methyl lysine
occurs through an imine intermediate (1), which is
hydrolysed to form formaldehyde by a non-enzymatic
process (2). b | A polypeptide backbone cartoon structure
of LSD1 bound to Co-REST and the cofactor FAD. The
two-lobed amine oxidase (AO) domain is shown in orange
and yellow. The Tower domain is in green and the SWIRM
domain in blue. The Co-REST linker region (pink)
associates with the LSD1 Tower domain and the SANT
domain (red) situated at the top of the Tower domain.
c | Depiction of the potential association of LSD1–Co-REST
with nucleosomal DNA. The bottom half shows a
nucleosome with the core histone octamer in the centre
and the associated DNA double helix in blue. The LSD1–
Co-REST complex modelled onto a nucleosome indicates
that the SANT domain of Co-REST (red) could interact with
nucleosomal DNA, whereas LSD1 targets the histone H3
tail where it protrudes from the DNA gyres (shown by the
arrow). d | LSD1 as part of the Co-REST complexes
contributes to repression of neuronal genes in non-neuronal
cells. LSD1 contributes to repression by removing H3K4
methylation. e | When bound to the androgen receptor (AR),
LSD1 is converted from a transcriptional repressor to an
activator by changing the substrate specificity of LSD1 so
that it catalyses the removal of H3K9 methylation.
16. Polyamino(bis)guanidines and Polyaminobiguanides Inhibit Purified LSD1
0
20000
40000
60000
80000
100000
1a
1c
1f
1d
1e
2a
2e
2f
2c
2b
2d
1b
1g
Untreated
(pmol/mgprotein/min)
rLSD1activity
“You can observe a lot by just watching.”
-Yogi Berra
18. Polyamino(bis)guanidines and Polyaminobiguanides are Non-Competitive Inhibitors of LSD1
-100
-50
0
50
100
150
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
1/H3K4me2 (mM)
1/V
0 mM
0.25 mM
0.5 mM
1 mM
2.5 mM
-100
-50
0
50
100
150
-0.3 -0.2 -0.1 0 0.1 0.2 0.3
1/H3K4me2 (mM)
1/V
0 mM
0.25 mM
0.5 mM
1 mM
2.5 mM
Compound 1c
Compound 2d
21. The Polyaminobiguanide Verlindamycin is a Potent Epigenetic Modulator
-Acts as a non-competitive inhibitor of recombinant LSD1/CoREST (KI = 6.7 mM)
-Promotes a 6.5-fold increase in global H3K4me2 in HCT116 cells in vitro; no change in
methylation levels at H3K9 or H3K27
-Causes significant re-expression of aberrantly silenced tumor suppressor proteins
SFRP1, 4 and 5 and GATA 5.
Huang, Y. et al.: Proc. Nat. Acad. Sci. USA 2007, 104,
8023-8028.
Huang, Y. et al.: Clin. Cancer Res. 2009, 15,
7217-7228
22. In vivo effects of compound 2d in the presence and absence of 5-azacytidine
Verlindamycin (2d) Is Effective In Vivo in Combination with 5-Azacytidine
23. Epigenetics in the Heart
Epigenetics refers to alterations in gene expression independent of
the genetic code.
HDACs have been extensively studied in cardiovascular disease
HDAC inhibitors shown to be cardioprotective in both ischemia
reperfusion injury and heart failure
Recent evidence implies a crosstalk, or even an interdependency,
of HDACs with histone demethylases
Chandrasekaran, S. et al.: Histone deacetylases facilitate sodium/calcium exchanger
up-regulation in adult cardiomyocytes. FASEB J. 2009, 23(11), 3851-3864.
HDAC activity causes
histone methylation
24. Assessment of Drug Effects in the Langendorff Heart Model
Normal rabbit heart Heart after ischemia reperfusion injury
25. Left Ventricular
HCT 116 human colorectal tumor xenograft in Balb/c mice
0 10 20 30 40 50 60 70 80 90 100
0
50
100
150
200
Time(mins)
mmHg Developed Pressure
Vehicle (n=6)
2d (n=5)
No IR (n=1)
ReperfusionNo-Flow
Ischemia
**** *** **
****p<0.0001; ***p<0.001; **p<0.01; *p<0.05
26. In vivo effects of compound 2d in the presence and absence of 5-azacytidine
HCT 116 human colorectal tumor xenograft in Balb/c mice
0 10 20 30 40 50 60 70 80 90 100
0
20
40
60
80
Time(mins)
mmHg End Diastolic Pressure
No IR (n=1)ReperfusionNo-Flow
Ischemia **** ***
****p<0.0001; ***p<0.001; **p<0.01; *p<0.05
Vehicle (n=6)
2d (n=5)
Left Ventricular
30. OH
F
CN
Cl O Cl
CN
O Cl
NH2
S
SN
NC
O
Cl
H
NN
S
NC
NH2H2N
EtOH
microwave
90oC, 10min
ether
microwave,
40oC, 5 min
DMSO, K2CO3
microwave, 190oC
6 min
+
LiAlH4
ether, 0oC, 24 h
CH3
CH3
H3C
O
Cl
H
NH
N
N
N
H2N
X
H
NN
N
NH
H2N
General Structure
Scheme 1
21 22 23
24
266
25
R1
R2
R3
R4
Cl
33. Cellular Effects of C1 and C15 in the Calu6 Lung Adenocarcinoma Cell Line
Vehicle
30µM TCP
1µM 6 10µM 7
1µM 6 10µM 7
DAPI F-Actin H3K4me2
Vehicle
30µM TCP
1µM 6 10µM 6
1µM 7 10µM 7
DAPI F-Actin H3K4me2
34. A B
!
C D
!
E
!
Figure S3. Comparison of the cytotoxicity of compounds 6 and 7 to known agents verlindamycin 2 and TCP in 5 cell lines in vitro using a standard MTS
reduction assay. Panel A: CA46 Burlitt’s Lymphoma cell line; Panel B: PC3 human prostate cancer cell line; Panel C: PANC-1 human pancreatic cancer cell
line; Panel D: MDA-MB-231 estrogen receptor negative breast cancer cell line; Panel E: MCF-10A human breast epithelial cell line. In Panels B and C,
verlindamycin 2 was run at 8 mM as a positive control, while in Panels A, D and E a dose-response curve was generated for 2. Each data point is the average of 3
determinations + standard error.
35. 0 10 20 30 40 50 60 70 80 90 100
0
50
100
150
200
Time(mins)
DevelopedPressure(mmHg) Left Ventricular Developed Pressure
(1-hr pretreatment)
Vehicle (n=6)ReperfusionNo-Flow
Ischemia C1 (n=3)
Verlindamycin (n=3)
No IR (n=2)
36. 0 10 20 30 40 50 60 70 80 90 100
0
20
40
60
80
Time(mins)
EndDiastolicPressure(mmHG) Left Ventricular End Diastolic Pressure
(1-hr pretreatment)
No IR (n=2)
ReperfusionNo-Flow
Ischemia C1 (n=3)
Verlindamycin (n=3)
Vehicle (n=6)
38. Figure 1. LSD1/HDAC/CoREST corepressor complex. LSD1 inhibitors or HDAC
inhibitors (HDACi) can independently re-express silenced promoters through post-
translational histone modifications.
Is the Major Effect of Inhibitors of Chromatin Remodeling Enzymes Mediated
at the Epigenetic Complex?
39. Primary feline cardiomyocytes were treated for 3 h with 5 mM verlindamycin (V), 1 mM C1
or 2 mM C15. The co-repressor HDAC:CoREST:LSD1 complex was initially pulled down with an antibody
for HDAC1 and a Western blot for CoREST was performed. The figure shows that verlindamycin and C1
disrupted the interaction between HDAC1 and CoREST, indicating that LSD1 inhibition may cause
disruption of the entire co-repressor complex.
Additional experiments were performed with a pull-down with LSD1 antibody, showing that C1 and
verlindamycin disrupted LSD1:HDAC1 interaction as well.
Pull-down Experiment for HDAC1/CoREST/LSD1 Complex
Wb: CoREST!
NC ! Veh! 2d! C1! C15!
IP: HDAC1!
IgG Veh V C1 C15
CoREST
Non-specific
40. Acknowledgements
MUSC Johns Hopkins University University of Pretoria
Dr. Donald R. Menick Robert A. Casero Lyn-Marie Birkholtz
Isuru Kumarasinghe Tracey Murray-Stewart Bianca Verlinden
Sun Choi Shannon Nowatarski Jandeli Niemand
Steven Holshouser Valentina Battaglia
Melissa Sokolosky Christina Destefano-Shields Jawarhal Nehru University
Craig Kutz Christin Hanigan Rentala Madhubala
Youxuan Li
Hereward “Cliff” Wimborne
Benefactors
NIH grant RO1 CA149095-01