Download Supplementary Information - Molecular Cancer Therapeutics

6α-Acetoxyanopterine: A Novel Structure Class of Mitotic Inhibitor Disrupting
Microtubule Dynamics in Prostate Cancer Cells
Claire Levrier1,2, Martin C. Sadowski1, Anja Rockstroh1, Brian Gabrielli3, Maria Kavallaris4,5,
Melanie Lehman1,6, Rohan A. Davis2, and Colleen C. Nelson*,1
1
Australian Prostate Cancer Research Centre−Queensland, School of Biomedical Sciences,
Institute of Health and Biomedical Innovation, Queensland University of Technology, Princess
Alexandra Hospital, Translational Research Institute, Brisbane, QLD 4102, Australia
2
Eskitis Institute for Drug Discovery, Griffith University, Brisbane, QLD 4111, Australia
3
The University of Queensland Diamantina Institute, Translational Research Institute;
Brisbane, QLD 4102, Australia
4
Tumour Biology and Targeting Program, Children’s Cancer Institute, Lowy Cancer Research
Centre, UNSW Australia, NSW, Australia
5
ARC Centre of Excellence in Convergent Bio-Nano Science and Technology and Australian
Centre for NanoMedicine, UNSW Australia, NSW, 2052
6
Vancouver Prostate Centre, Department of Urologic Sciences, University of British Columbia,
Vancouver, Canada
Supplementary Information
Supplementary Figures S1−S9 and figures legends
Supplementary Tables S1−S6
Supplementary Videos S1−S5 and legends
1
Supplementary Figure S1. Quantitative image analysis of PHH3-positive cells and αtubulin intensity using CellProfiler software.
Example (LNCaP cells treated with 2.5 nM of 6-AA for 24 h) of the automated image
segmentation and object quantification performed with CellProfiler software. First, nuclei were
identified based on DAPI staining with a mean fluorescence intensity threshold that captured
all nuclei (relaxed and condensed chromatin, top left panel). Next, PHH3-positive nuclei were
identified based on PHH3 staining (top right panel). To quantify PHH3-positive cells, all nuclei
were scored for PHH3-positive (red) and -negative staining (blue, bottom left panel). Mitotic
cells displayed strong nuclear PHH3 staining, condensed chromatin and a round cell
morphology. Cell outlines were identified based on α-tubulin staining (bottom left panel) and
the α-tubulin staining mean intensity per cell was correlated with α-tubulin polymer mass per
cell.
2
Supplementary Figure S2. 6-AA induces a time- and concentration-dependent growth
arrest and inhibits cell migration.
A, following treatment of HeLa cells with various doses of 6-AA, proliferation was analyzed
for 72 h using real-time cell imaging (IncuCyte), showing that 6-AA caused a concentrationdependent growth inhibition. DMSO (vehicle) and vinblastine (Vinb) were used as controls
(n=3, mean±SD).
3
B, 6-AA inhibited PC-3 cell migration. Wound closure was measured by live-cell imaging
(IncuCyte) after 16h of treatment with vehicle control (DMSO), 6-AA, vinblastine (Vinb) or
the positive control cytochalasin D (Cytoch) (left panel, n=2, mean±SD). Cytochalasin D is an
actin inhibitor commonly used as positive control in migration assays. Representative images
of treatments with DMSO, 6-AA and cytochalasin D are shown (right panel).
C, phase contrast and GFP-H2B images acquired on an Olympus IX81 microscope showed
that 6-AA- and vinblastine-treated HeLa GFP-H2B cells presented the classical features of
mitotic arrest (round cells with condensed chromatin) and apoptosis (membrane blebbing,
DNA ultra-condensation and fragmentation). Scale bar=50 µm.
D, FACS analysis revealed that 6-AA (5 nM) arrests LNCaP cells in the G2-M phase of the cell
cycle and induces cell death in a time-dependent manner. DMSO and vinblastine (Vinb) were
used as controls (n=2, mean±SD). Two-ways ANOVA with Sidak’s multiple comparisons test
was used (ns=non-significant, * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001).
4
Supplementary Figure S3. DNA microarray analysis of 6-AA-treated LNCaP cells.
A, heatmap depicting the fold changes of all 228 reference genes differentially expressed in
LNCaP cells treated for 24 h with either 6-AA (10 nM) or vinblastine (Vinb, 3.25 nM), showing
that the directionality of change was the same for both compounds. Red indicates up-regulation
and blue down-regulation. The darker the shade of color, the higher the fold-change of
expression.
B, GeneOntology analysis (GOrilla) revealed “microtubule binding” as the top enriched
molecular function within the set of differentially expressed genes after 6-AA and vinblastine
treatments.
5
6
7
Supplementary Figure S4. 6-AA is a reversible inhibitor of mitosis.
A, immunofluorescence microscopy coupled with automated image analysis (CellProfiler) was
used to quantify PHH3-positive (mitotic) LNCaP cells (~3,000 cells/treatment) after the
indicated treatment conditions (n=2, mean±SD). 6-AA (1.25-10 nM) and vinblastine (10 nM)
induced a significant increase in PHH3-positive cells when treated for after 8 h (orange bars).
Longer treatment (24 h, purple bars) further increased the proportion of PHH3-positive cells.
Removal of 6-AA and vinblastine after 8 h of treatment followed by 16 h of recovery decreased
the number of PHH3-positive cells to levels seen in vehicle control (DMSO) irrespective of the
inhibitor dose. Two-ways ANOVA with Sidak’s multiple comparisons test was used (ns=nonsignificant, ** P < 0.01, **** P < 0.0001; orange label=statistical comparison to DMSO 8 h).
B, proliferation of LNCaP cells after different treatment modalities (continuous, 8 h and 24 h
before inhibitor removal) with the indicated concentrations of 6-AA (left panels) and
vinblastine (right panels) was monitored by real-time imaging (IncuCyte) for 72 h (n=2,
mean±SD). Removal of 6-AA and vinblastine after 8 h and 24 h of treatment enabled LNCaP
cells to visibly increase cell confluence at all inhibitor concentrations tested. Arrow=washout.
C, LNCaP cells were subjected to the same treatment modalities as described in B, and cell
viability was measured after 72 h (alamarBlue, n=2, mean±SD). Intermittent treatment with 6AA (8 h and 24 h) significantly less reduced cell viability compared to continuous treatment
(72 h). A shorter intermitted treatment period resulted in higher cell viability (8h vs 24h),
suggesting that the reversibility of the effect of 6-AA has a time-dependent component. Twoways ANOVA with Sidak’s multiple comparisons test was used (ns=non-significant,
** P < 0.01, *** P < 0.001, **** P < 0.0001).
D, cell cycle analysis (FACS) of LNCaP cells treated with DMSO, 6-AA (2.5 nM) continuously
for 24 h or with 6-AA (2.5 nM) for 8 h plus 16h recovery demonstrated that the 6-AA-induced
accumulation of cells in G2-M is reversible and that removal of 6-AA restored a cell cycle
distribution which was similar to control (DMSO). Representative DNA histograms of two
independent experiments are shown.
E, immunofluorescence microscopy coupled with automated image analysis (CellProfiler) was
used to quantify PHH3-positive (mitotic) PC-3 cells (~3,000 cells/treatment) after the indicated
modalities of transient and continuous treatments (n=2, mean±SD). Intermittent 6-AA
treatment for 8 h resulted in significant lower levels of PHH3-positive PC-3 cells when
compared to continuous treatment for 24 h at all concentrations tested. A similar, reversible
8
effect on the accumulation of PHH3-positive cells was observed with vinblastine (20 nM).
Two-ways ANOVA with Sidak’s multiple comparisons test was used (ns=non-significant,
* P < 0.05, *** P < 0.001).
F, real-time imaging of cell confluence (IncuCyte) revealed that washout of 6-AA and
vinblastine after 8 h allowed HeLa cells to resume proliferation, while transient treatment with
6-AA for 24 h visibly reduced cell proliferation (n=2, mean±SD). Arrow=washout.
Supplementary Figure S5. 6-AA induces mitotic spindle defects in PC-3 cells.
PC-3 cells treated with 6-AA, vinblastine, nocodazole (Noc), or vehicle control (DMSO) for
24 h were subjected to immunofluorescence microscopy of α-tubulin (green) and PHH3 (red)
with a DeltaVision Elite microscope (60×). Representative images of cells in metaphase with
misaligned chromosomes and abnormal spindle pole organization (6-AA, vinblastine and
nocodazole treatments) are shown. Scale bar=10 µm.
9
Supplementary Figure S6. All anopterine analogs share the same mechanism of action.
A, chemical structures of tested C20 diterpenoid alkaloids, which included 6αacetoxyanopterine (6-AA, 1), 4'-hydroxy-6α-acetoxyanopterine (2), 4'-hydroxyanopterine (3),
10
and
11α-benzoylanopterine
(4),
anopterine
(5),
7β-hydroxyanopterine
(6),
7β,4'-
dihydroxyanopterine (7), and 7β-hydroxy-11α-benzoylanopterine (8).
B, LNCaP cells treated with 6-AA (1, 5 nM), 2 (30 nM), 3 (1000 nM), 4 (400 nM), 5 (100 nM),
6 (300 nM), 7 (400 nM), 8 (10 nM), vinblastine (Vinb, 10 nM), or vehicle control (DMSO) for
24 h were subjected to immunofluorescence microscopy of α-tubulin (green) and DNA (blue)
with an INCell 2200 automated imaging system (40×). Representative images of cells in
metaphase with misaligned chromosomes and abnormal spindle pole organization are shown.
Scale bar=10 µm.
C, anopterine (5), like vinblastine (Vinb), inhibited tubulin polymerization in a cell-free
system, while paclitaxel (Paclit) stimulated polymerization. A representative experiment (n=2)
is shown.
11
Supplementary Figure S7. 6-AA does not affect polo-like kinase-1 and Aurora A kinase
activities.
A, LNCaP cells treated with 6-AA, the Aurora A kinase inhibitor MNL8237, the polo-like
kinase 1 inhibitor BI2536, or vehicle control (DMSO) for 24 h were subjected to
immunofluorescence microscopy of Aurora A (green) and DNA (blue) with a DeltaVision Elite
microscope (60×). MNL8237 inhibited localization of Aurora A kinase at the spindle poles,
which was not affected by 6-AA or BI2536. Representative images are shown. Scale
bar=10 µm.
B, LNCaP cells treated with 6-AA, the polo-like kinase 1 inhibitor BI2536, or vehicle control
(DMSO) for 24 h were subjected to immunofluorescence microscopy of α-tubulin (green) and
PHH3 (red) with a DeltaVision Elite microscope (60×). BI2536 inhibited spindle pole
organization, uniformly generating cells with monopolar spindles and expanded chromosome
12
configurations, while 6-AA generated heterogeneous spindle pole abnormalities (monopolar,
bipolar and multipolar) with visibly contracted chromosome alignments. Representative
images are shown. Scale bar=10 µm.
Supplementary Figure S8. The microtubule-depolymerizing action of 6-AA in LNCaP
cells is reversible.
LNCaP cells were treated with high doses of 6-AA, vinblastine or colchicine for 2 h, then
washed and left to recover for 4 h. After washout of 6-AA and vinblastine, microtubules repolymerized, forming microtubule networks. Washout of the irreversible inhibitor colchicine
failed to restore microtubule networks. Microtubules were visualized by immunofluorescence
microscopy of α-tubulin with an INCell 2200 automated imaging system (40×). Scale
bar=5 µm.
13
Supplementary Figure S9. 6-AA reduce the length of the polymerizing portion of
microtubules in HeLa-EB1-GFP cells.
HeLa-EB1-GFP cells were treated with 6-AA or vinblastine for 2 h, and EB1-GFP was imaged
by spinning disk microscopy. EB1 comets were tracked and analyzed using Imaris software.
The average displacement length of EB1 tracks is represented (10 cells/treatment),
demonstrating that 6-AA caused a concentration-dependent decrease of the length of the
polymerizing portion of microtubules.
14
Supplementary Table S1. Primer sequences of primers used quantitative Real Time-PCR
(qRT-PCR)
Primer
Forward sequence
Reverse sequence
RPL32
5’-GCACCAGTCAGACCGATATG-3’
5’-ACTGGGCAGCATGTGCTTTG-3’
CCNB1
5’-AGAGCCATCCTAATTGACTG-3’
5’-CAACCAGCTGCAGCATCTTC-3’
CDC25B
5’-AGCGAAACCCCAAAGAGTCAGGT-3’
5’-AGCAGCCTCACCGGCATAGACT-3’
PLK1
5’-TGCCCTACCTACGGACCTGG-3’
5’-AAGTTGATCTGCACGCTGCC-3’
HMMR
5’-CTACCTTGCCTGCTTCAGCT-3’
5’-AGCTGAAGCAGGCAAGGTAG-3’
PTTG1
5’-GAAAAGCTGTTTCAGCTGGGC-3’
5’-CAGAATGCTTGAAGGAGACTGC-3’
CDKN3
5’-GGAAGAGCTTACAACCTGCC-3’
5’-ACAGGTATAGTAGGAGACAAGC-3’
Supplementary Table S2: Cytotoxic activities of 6-AA against a human malignant cell
lines.
Compound
IC50 ± SD (nM)
LNCaP
HeLa
PC3
6-AA
3.1 ± 1.8
3.2 ± 1.1
5.1 ± 0.2
Vinb
3.2 ± 0.1
11.6 ± 4.2
2.5 ± 0.4
IC50 values of 6-AA and vinblastine (Vinb) in the indicated cell lines were calculated based on
cell viability (alamarBlue) after 72 h of treatment (n=3, mean±SD).
15
Supplementary Table S3: List of selected mitosis-related genes upregulated by 6-AA (10
nM) as identified in the microarray.
Gene symbole
CDC20
AURKA
CENPA
CCNB1
HMMR
KIF18A
CKS2
KIF20A
CCNB2
NUF2
TPX2
CCNA2
NEK2
CDC25B
BUB1B
CDCA8
KIF4A
PLK1
BIRC5
BUB1
NDC80
CDKN3
KIF2C
SPC25
CDK1
KIF23
AURKB
KIFC1
PLK4
BORA
NCAPH
SMC4
CDK5RAP2
INCENP
CENPK
CNTRL
CKS1B
PCNT
PTTG1
Gene description
cell division cycle 20
aurora kinase A
centromere protein A
cyclin B1
hyaluronan-mediated motility receptor
kinesin family member 18A
CDC28 protein kinase regulatory subunit 2
kinesin family member 20A
cyclin B2
NDC80 kinetochore complex component
TPX2, microtubule-associated
cyclin A2
NIMA (never in mitosis gene a)-related kinase 2
cell division cycle 25
budding uninhibited by benzimidazoles 1 homolog beta
cell division cycle associated 8
kinesin family member 4A
polo-like kinase 1
baculoviral IAP repeat containing 5 (survivin)
budding uninhibited by benzimidazoles 1
NDC80 kinetochore complex component
cyclin-dependent kinase inhibitor 3
kinesin family member 2C
NDC80 kinetochore complex component
cyclin-dependent kinase 1
kinesin family member 23
aurora kinase B
kinesin family member C1
polo-like kinase 4
bora, aurora kinase A activator
non-SMC condensin I complex, subunit H
structural maintenance of chromosomes 4
CDK5 regulatory subunit associated protein 2
inner centromere protein
centromere protein K
Centriolin
CDC28 protein kinase regulatory subunit 1B
Pericentrin
pituitary tumor-transforming 1
Fold change
2.80
2.43
2.42
2.40
2.40
2.37
2.31
2.23
2.21
2.21
2.21
2.20
2.14
2.05
2.03
2.03
2.03
2.03
2.01
2.01
2.01
1.95
1.94
1.91
1.80
1.79
1.77
1.77
1.75
1.72
1.71
1.69
1.66
1.66
1.60
1.60
1.57
1.51
1.48
For genes with multiple probes, the ‘differentially expressed reference probe with the highest
fold change’ was used to represent the gene for functional analysis.
16
Supplementary Table S4: Cytotoxicity of 6-AA, 2-methoxyoestradiol, vincristine and
doxorubicin against resistant cell lines.
Compound
CEM
CEM/2ME2-14.4R
CEM/2ME2-28.8-R
IC50 (nM)
IC50 (nM)
RF
IC50 (nM)
RF
2ME2
765.9 ± 177.4
14,310.0 ± 4,851.4
18.7
23,410.0 ± 13,392.3
30.6
6-AA
2.6 ± 1.0
5.9 ± 0.7
2.3
4.5 ± 0.4
1.7
CEM
Vinc
CEM/VCR-R
IC50 (nM)
IC50 (nM)
RF
0.5 ± 0.3
6,168.0 ± 1,359.2
12,336.0
6-AA
2.6 ± 1.0
117.7 ± 50.7
45.3
Dox
99.4 ± 34.9
24,75.3 ± 222.0
24.9
Cytotoxicity (IC50) of 6-AA, vincristine (Vinc), 2-methoxyestradiol (2ME2), and doxorubicin
(Dox) was measured (72 h, alamarBlue) in indicated cell lines (upper panel, n=3, mean±SD).
The resistance factor (RF) expresses the quotient IC50 (drug-resistant cell line)/IC50 (CEM).
Supplementary Table S5: Cytotoxicity of 6-AA, vincristine, and doxorubicin in
combination with verapamil against the multi-drug resistant CEM/VCR-R cell line.
CEM
CEM/VCR-R
Compound
IC50 (nM)
IC50 (nM)
RF
Verap
>10,000
>10,000
-
Vinc + verap
1.3 ± 0.4
82.7 ± 42.2
63.6
6-AA + verap
1.6 ± 0.3
3.4 ± 0.4
2.1
Dox + verap
146.8 ± 54.7
163.9 ± 63.0
1.1
Cytotoxicity (IC50 ± SD) of indicated compounds in combination with verapamil (Verap) was
measured in indicated cell lines (72 h, alamarBlue, left panel, n=3). The resistance factor (RF)
expresses the quotient IC50 (CEM/VCR-R + verapamil)/IC50 (CEM + verapamil).
Supplementary Table S6: Statistical data for Figure 1D
Compound
Sub G0-G1
G0-G1
S
G2-M
6-AA (1.25 nM)
ns
ns
ns
ns
6-AA (2.5 nM)
**
**
ns
****
6-AA (5 nM)
**
***
ns
****
Vinb (10 nM)
ns
***
ns
****
For each phase of the cell cycle, one-way ANOVA with Dunnet's multiple comparisons test
was used (comparison to DMSO, ns=non-significant, ** P < 0.01, *** P < 0.001,
**** P < 0.0001).
17
Supplementary Videos
Supplementary Videos S1. Time-lapse microscopy of HeLa-H2B-GFP cells treated with
DMSO, vinblastine (20 nM) or 6-AA (0.62 nM). HeLa-H2B-GFP cells were imaged using
an Olympus IX81 (1 image/5 min for 24 h). Top panels: brightfield, bottom panels: H2B-GFP.
DMSO-treated cells normally progressed through mitosis, which was delayed with vinblastine
(20 nM) and low dose of 6-AA (0.62 nM). Both compounds induced asymmetric division. In
addition, vinblastine caused cell death. Scale bar=20 µm.
Supplementary Videos S2. Time-lapse microscopy of HeLa-H2B-GFP cells treated with
6-AA (2.5, 5 and 10 nM). HeLa-H2B-GFP cells were imaged using an Olympus IX81 (1
image/5 min for 24 h). Top panels: brightfield, bottom panels: GFP channel. 6-AA increased
in a concentration-dependent manner mitotic duration, leading to cell death or asymmetric
division. Scale bar=20 µm.
Supplementary Videos S3. Interphase microtubule dynamics in HeLa-EB1-GFP treated
with DMSO. EB1-GFP was imaged by spinning disk microscopy (1 image/s for 1 min). EB1
comets were clearly discernible and followed regular trajectories. Scale bar=10 µm.
Supplementary Videos S4. Interphase microtubule dynamics in HeLa-EB1-GFP treated
with 6-AA (2.5 nM). EB1-GFP was imaged by spinning disk microscopy (1 image/s for
1 min). EB1 comets followed irregular and short trajectories. Scale bar=10 µm.
Supplementary Videos S5. Interphase microtubule dynamics in HeLa-EB1-GFP treated
with vinblastine (20 nM). EB1-GFP was imaged by spinning disk microscopy (2 images/s for
2 min). EB1 comets appeared faint and followed irregular and short trajectories. Scale
bar=10 µm.
18