rexresearch.com
Marcus E. PETER, et
al.
siRNA vs Cancer
https://www.dailymail.co.uk/health/article-6330575/The-end-chemotherapy-Scientists-discover-cancerous-cells-KILL-CODE.html
The end of chemotherapy? Scientists
discover all cancerous cells have a KILL CODE that can be
triggered without the gruelling treatment
Researchers at Northwestern University
found that our cells can kill themselves
Currently, this is triggered by
disease itself or the artificial use of chemotherapy
Now, experts believe the 'kill codes'
could be synthetically duplicated for use
By Peter Lloyd
Every cell in the human body contains a 'kill code' which can be
triggered to cause its own self-destruction.
That's the discovery made by researchers at Northwestern
University, Illinois, who believe it could be utilized for the
future fight against cancer.
Specifically, they predict malignant cells could be encouraged to
'commit suicide' without toxic chemicals pumped into the body -
which currently turns them on.
And, in the process, it could mean an end to grueling rounds of
chemotherapy.
Currently, as soon as the cell's inner bodyguards sense it is
mutating into cancer, it self-activates the kill code to
extinguish itself.
Embedded in ribonucleic acids, scientists estimate these evolved
more than 800 million years ago in part to protect the body from
diseases.
However, while they are already somewhat successful, they can't
always compete with aggressive tumours. Hence they they are
artificially prompted with drugs.
But these could be even more powerful if synthetically duplicated
- not least because they would benefit a patient without the
side-effects of chemo.
''Now that we know the kill code, we can trigger the mechanism
without having to use chemotherapy and without messing with the
genome,' said lead author Marcus E. Peter, the Tomas D. Spies
Professor of Cancer Metabolism at Northwestern University Feinberg
School of Medicine.
'We can use these small RNAs directly, introduce them into cells
and trigger the kill switch.'
'My goal was not to come up with a new artificial toxic
substance,' he added. 'I wanted to follow nature's lead. I
want to utilize a mechanism that nature designed.'
He added: 'Based on what we have learned in these two studies, we
can now design artificial microRNAs that are much more powerful in
killing cancer cells than even the ones developed by nature.'
However, he stressed that a potential therapy is many years off.
The paper describing the kill code and identifying how the
cancer-fighting microRNAs use the code to kill tumor cells was
published in the Nature Communications, today.
https://news.northwestern.edu/stories/2017/october/suicide-molecules-kill-any-cancer-cell/
Suicide molecules kill any
cancer cell
Time traveling hundreds of millions of
years to unleash one of nature’s original kill switches
By Marla Paul
Small RNA molecules originally developed as a tool to study gene
function trigger a mechanism hidden in every cell that forces the
cell to commit suicide, reports a new Northwestern Medicine study,
the first to identify molecules to trigger a fail-safe mechanism
that may protect us from cancer.
The mechanism -- RNA suicide molecules -- can potentially be
developed into a novel form of cancer therapy, the study authors
said.
Cancer cells treated with the RNA molecules never become resistant
to them because they simultaneously eliminate multiple genes that
cancer cells need for survival.
“It’s like committing suicide by stabbing yourself, shooting
yourself and jumping off a building all at the same time,” said
Northwestern scientist and lead study author Marcus Peter. “You
cannot survive.”
The inability of cancer cells to develop resistance to the
molecules is a first, Peter said.
“This could be a major breakthrough,” noted Peter, the Tom D.
Spies Professor of Cancer Metabolism at Northwestern University
Feinberg School of Medicine and a member of the Robert H. Lurie
Comprehensive Cancer Center of Northwestern University.
Peter and his team discovered sequences in the human genome that
when converted into small double-stranded RNA molecules trigger
what they believe to be an ancient kill switch in cells to prevent
cancer. He has been searching for the phantom molecules with this
activity for eight years.
“We think this is how multicellular organisms eliminated cancer
before the development of the adaptive immune system, which is
about 500 million years old,” he said. “It could be a fail safe
that forces rogue cells to commit suicide. We believe it is active
in every cell protecting us from cancer.”
This study, which will be published Oct. 24 in eLife, and two
other new Northwestern studies in Oncotarget and Cell Cycle by the
Peter group, describe the discovery of the assassin molecules
present in multiple human genes and their powerful effect on
cancer in mice.
Looking back hundreds of millions of years
Why are these molecules so powerful?
“Ever since life became multicellular, which could be more than 2
billion years ago, it had to deal with preventing or fighting
cancer,” Peter said. “So nature must have developed a fail safe
mechanism to prevent cancer or fight it the moment it forms.
Otherwise, we wouldn’t still be here.”
Thus began his search for natural molecules coded in the genome
that kill cancer.
“We knew they would be very hard to find,” Peter said. “The kill
mechanism would only be active in a single cell the moment it
becomes cancerous. It was a needle in a haystack."
But he found them by testing a class of small RNAs, called small
interfering (si)RNAs, scientists use to suppress gene activity.
siRNAs are designed by taking short sequences of the gene to be
targeted and converting them into double- stranded RNA. These
siRNAs when introduced into cells suppress the expression of the
gene they are derived from.
Peter found that a large number of these small RNAs derived from
certain genes did not, as expected, only suppress the gene they
were designed against. They also killed all cancer cells. His team
discovered these special sequences are distributed throughout the
human genome, embedded in multiple genes as shown in the study in
Cell Cycle.
When converted to siRNAs, these sequences all act as highly
trained super assassins. They kill the cells by simultaneously
eliminating the genes required for cell survival. By taking out
these survivor genes, the assassin molecule activates multiple
death cell pathways in parallel.
The small RNA assassin molecules trigger a mechanism Peter calls
DISE, for Death Induced by Survival gene Elimination.
Activating DISE in organisms with cancer might allow cancer cells
to be eliminated. Peter’s group has evidence this form of cell
death preferentially affects cancer cells with little effect on
normal cells.
To test this in a treatment situation, Peter collaborated with Dr.
Shad Thaxton, associate professor of urology at Feinberg, to
deliver the assassin molecules via nanoparticles to mice bearing
human ovarian cancer. In the treated mice, the treatment strongly
reduced the tumor growth with no toxicity to the mice, reports the
study in Oncotarget. Importantly, the tumors did not develop
resistance to this form of cancer treatment. Peter and Thaxton are
now refining the treatment to increase its efficacy.
Peter has long been frustrated with the lack of progress in solid
cancer treatment.
“The problem is cancer cells are so diverse that even though the
drugs, designed to target single cancer driving genes, often
initially are effective, they eventually stop working and patients
succumb to the disease,” Peter said. He thinks a number of cancer
cell subsets are never really affected by most targeted anticancer
drugs currently used.
Most of the advanced solid cancers such as brain, lung, pancreatic
or ovarian cancer have not seen an improvement in survival, Peter
said.
“If you had an aggressive, metastasizing form of the disease 50
years ago, you were busted back then and you are still busted
today,” he said. “Improvements are often due to better detection
methods and not to better treatments.”
Cancer scientists need to listen to nature more, Peter said.
Immune therapy has been a success, he noted, because it is aimed
at activating an anticancer mechanism that evolution developed.
Unfortunately, few cancers respond to immune therapy and only a
few patients with these cancers benefit, he said.
"Our research may be tapping into one of nature’s original kill
switches, and we hope the impact will affect many cancers,” he
said. “Our findings could be disruptive.”
Northwestern co-authors include first authors William Putzbach,
Quan Q. Gao, and Monal Patel, and coauthors Ashley Haluck-Kangas,
Elizabeth T. Bartom, Kwang-Youn A. Kim, Denise M. Scholtens,
Jonathan C. Zhao and Andrea E. Murmann.
The research is funded by grants T32CA070085, T32CA009560,
R50CA211271 and R35CA197450 from the National Cancer Institute of
the National Institutes of Health.
https://www.newsweek.com/cancer-super-assassin-huntingtons-disease-802967
Cancer 'Super Assassin': Huntington's
Disease Molecule Can Target and Kill All Tumor Cells,
Scientists Find
by Katherine Hignett
Scientists have destroyed numerous types of human cancer cells
with a toxic molecule characteristic of fatal genetic illness
Huntington’s disease.
The researchers hailed the molecule—which has killed both human
and mouse ovarian, breast, prostate, liver, brain, lung, skin and
colon cancer cell lines in mice—as a “super assassin.”
Their results were published Monday in the journal EMBO Reports.
Huntington’s disease is a progressive illness caused by an excess
of a specific repeating RNA sequence in the huntington gene, which
is present in every cell.
The defect causes the death of brain cells, and gradually worsens
a person’s physical and mental abilities. The disease has no cure.
“We’ve never seen anything this powerful”
Researchers believe that the defect may be even more powerful
against cancer cells than nerve cells in the brain, and the team
hopes it can be harnessed to kill cancer cells without causing
Huntington’s symptoms.
“This molecule is a super assassin against all tumor cells,” said
senior author Marcus Peter, a professor of cancer metabolism at
Northwestern University Feinberg School of Medicine, in a press
statement. “We’ve never seen anything this powerful.”
Peter collaborated with Feinberg colleague Shad Thaxton, associate
professor of urology, to deliver the molecule in the form of
nanoparticles to mice with human ovarian cancer. The targeted
molecule decreased tumor growth with no toxicity to the mice.
"Kill switch"
First author Andrea Murmann, a research assistant professor who
discovered the cancer-killing mechanism, used the molecule to kill
numerous other human and mouse cancer cell lines. Building on
previous research into a cancer “kill switch,” Murmann looked to
diseases associated with low rates of cancer and a suspected RNA
link.
“I thought maybe there is a situation where this kill switch is
overactive in certain people, and where it could cause loss of
tissues,” Murmann said in the statement. “These patients would not
only have a disease with an RNA component, but they also had to
have less cancer.“
There is up to 80 percent less cancer in people with Huntington’s
disease than the general population.
Murmann recognized similarities between the kill switch and the
toxic Huntington’s disease RNA sequences.
Based on their results, the team believe the “super assassin”
molecule could be used to fight cancer in humans. “We believe a
short-term treatment cancer therapy for a few weeks might be
possible, where we could treat a patient to kill the cancer cells
without causing the neurological issues that Huntington’s patients
suffer from,” Peter said.
The scientists next aim to refine the molecule’s delivery method
to improve tumor targeting, and to stabilize the nanoparticles for
storage.
https://www.ncbi.nlm.nih.gov/pubmed/29063830
Elife. 2017 Oct 24;6.
pii: e29702.
doi: 10.7554/eLife.29702.
Many si/shRNAs can kill
cancer cells by targeting multiple survival genes through an
off-target mechanism.
Putzbach W, et al.
Abstract
Over 80% of multiple-tested siRNAs and shRNAs targeting CD95
or CD95 ligand (CD95L) induce a form of cell death characterized
by simultaneous activation of multiple cell death pathways
preferentially killing transformed and cancer stem cells. We now
show these si/shRNAs kill cancer cells through canonical RNAi by
targeting the 3'UTR of critical survival genes in a unique form of
off-target effect we call DISE (death induced by survival gene
elimination). Drosha and Dicer-deficient cells, devoid of most
miRNAs, are hypersensitive to DISE, suggesting cellular miRNAs
protect cells from this form of cell death. By testing 4666 shRNAs
derived from the CD95 and CD95L mRNA sequences and an unrelated
control gene, Venus, we have identified many toxic sequences -
most of them located in the open reading frame of CD95L. We
propose that specific toxic RNAi-active sequences present in the
genome can kill cancer cells.
https://www.nature.com/articles/s41467-018-06526-1#Abs1
Nature Communicationsvolume 9, Article number: 4504
(2018)
6mer seed toxicity in tumor
suppressive microRNAs
Quan Q. Gao, et al.
Abstract
Many small-interfering (si)RNAs are toxic to cancer cells
through a 6mer seed sequence (positions 2–7 of the guide strand).
Here we performed an siRNA screen with all 4096 6mer seeds
revealing a preference for guanine in positions 1 and 2 and a high
overall G or C content in the seed of the most toxic siRNAs for
four tested human and mouse cell lines. Toxicity of these siRNAs
stems from targeting survival genes with C-rich 3′UTRs. The master
tumor suppressor miRNA miR-34a-5p is toxic through such a G-rich
6mer seed and is upregulated in cells subjected to genotoxic
stress. An analysis of all mature miRNAs suggests that during
evolution most miRNAs evolved to avoid guanine at the 5′ end of
the 6mer seed sequence of the guide strand. In contrast, for
certain tumor-suppressive miRNAs the guide strand contains a
G-rich toxic 6mer seed, presumably to eliminate cancer cells.
Introduction
RNA interference (RNAi) is a form of post-transcriptional
regulation exerted by 19–21 nt long double-stranded RNAs that
negatively regulate gene expression at the mRNA level. RNAi-active
guide RNAs can come from endogenous siRNAs and micro(mi)RNAs. For
an miRNA, the RNAi pathway begins in the nucleus with
transcription of a primary miRNA precursor (pri-miRNA)1.
Pri-miRNAs are first processed by the Drosha/DGCR8 microprocessor
complex into pre-miRNAs2, which are then exported from the nucleus
to the cytoplasm by Exportin-53. Once in the cytoplasm, Dicer
processes them further4,5 and these mature dsRNA duplexes are then
loaded into Argonaute (Ago) proteins to form the RNA-induced
silencing complex (RISC)6. The sense/passenger strand is
ejected/degraded, while the guide strand remains associated with
the RISC7. Depending on the degree of complementarity between the
guide strand and its target, the outcome of RNAi can either be
target degradation—most often achieved by siRNAs with full
complementarity to their target mRNA8—or miRNA-like
cleavage-independent silencing, mediated by
deadenylation/degradation or translational repression9. The latter
mechanism can be initiated with as little as six nucleotide
base-pairing between a guide RNA’s so-called seed sequence
(positions 2–7) and fully complementary seed matches in the target
RNA10,11. This seed-based targeting most often occurs in the 3′UTR
of a target mRNA12,13.
A number of miRNAs function either as tumor suppressors or as
oncogenes14. Their cancer-specific activities are usually
explained by their identified targets, being oncogenes or tumor
suppressors, respectively14. Examples of targets of
tumor-suppressive miRNAs are the oncogenes Bcl-2 for miR-15/1615
and c-Myc for miR-34a16. While many miRNAs have been reported to
have both tumor suppressive and oncogenic activities depending on
the cancer context, examples for widely established
tumor-promoting miRNAs are miR-221/222, miR-21, miR-155, and
members of the miR-17~92 cluster, or its paralogues miR-106b~25
and miR-106a~36317,18. In contrast, two of the major
tumor-suppressive miRNA families are miR-15/16 and the p53
regulated miR-34a/c and miR-34b19.
We recently discovered that many si- and shRNAs can kill all
tested cancer cell lines through RNAi by targeting the 3′UTRs of
critical survival genes (SGs)20. We called this mechanism DISE
(for death induced by SG elimination). Cancer cells have
difficulty in developing resistance to this mechanism both in
vitro and when treated in vivo21. We reported that a 6mer seed
sequence in the toxic siRNAs is sufficient for effective
killing20. We have now performed a strand-specific siRNA screen
with a library of individual siRNAs representing all 4096 possible
6mer seed sequences in a neutral RNA duplex. This screen, while
based on siRNA biochemistry, was not designed to identify targets
that are degraded through siRNA-mediated slicing activity but to
identify toxicity caused by moderately targeting hundreds of genes
required for cell survival in a mechanism similar to miRNA-induced
silencing.
We report that the most toxic 6mer seeds are G-rich with a G
enrichment towards the 5′ end targeting SGs with a high C content
in their 3′UTR in a miRNA-like manner. Many tumor-suppressive
miRNAs such as miR-34a-5p but none of the established oncogenic
miRNAs contain G-rich 6mer seeds and most of miR-34a-5p's toxicity
comes from its 6mer seed sequence. Mature miRNAs from older and
more conserved miRNAs contain less toxic seeds. We demonstrate
that for most miRNAs the more abundant mature form corresponds to
the arm that contains the less toxic seed. In contrast, for major
tumor-suppressive miRNAs, the mature miRNA is derived from the arm
that harbors the more toxic seed. Our data allow us to conclude
that while most miRNAs have evolved to avoid targeting survival
and housekeeping genes, certain tumor-suppressive miRNAs function
to kill cancer cells through a toxic G-rich 6mer seed targeting
the 3′UTR of SGs...
https://www.medicalnewstoday.com/articles/323506.php
'Triggering the kill switch' without chemo
...In last year's paper, Prof. Peter and his team found a sequence
of six nucleotides contained in small RNAs that made these
molecules toxic to cancer cells. A nucleotide is "the basic
structural unit and building block for DNA" and RNA.
In the first recently published study, Prof. Peter found that
about 3 percent of all the large RNAs can be "cut" into small
pieces that then act as toxic microRNAs that can kill cancer.
In the second recent study, Prof. Peter's team tested almost 4,100
different possible combinations of nucleotide bases from those six
initial nucleotides in an attempt to find the deadliest, most
toxic combination.
"Based on what we have learned in these two studies, we can now
design artificial microRNAs that are much more powerful in killing
cancer cells than even the ones developed by nature," Prof. Peter
explains. "We absolutely need to turn this into a novel form of
therapy."
In the second recent study, Prof. Peter's team tested almost 4,100
different possible combinations of nucleotide bases from those six
initial nucleotides in an attempt to find the deadliest, most
toxic combination.
"Based on what we have learned in these two studies, we can now
design artificial microRNAs that are much more powerful in killing
cancer cells than even the ones developed by nature," Prof. Peter
explains. "We absolutely need to turn this into a novel form of
therapy."
"Now that we know the kill code, we can trigger
the mechanism without having to use chemotherapy and without
messing with the genome. We can use these small RNAs directly,
introduce them into cells and trigger the kill switch."
US2018251762
USE OF TRINUCLEOTIDE REPEAT RNAs TO TREAT CANCER
[ PDF ]
Inventor: PETER MARCUS, et al.
Applicant: UNIV NORTHWESTERN [US]
Disclosed are compositions and methods related to RNA interference
(RNAi) and the use of RNAi active sequence for treating diseases
and disorders. Particular disclosed are toxic RNAi active
sequences such as siRNA and shRNA for killing cancer cells. The
disclosed toxic RNAi active sequences typically include
trinucleotide repeats and preferentially target the expression of
multiple essential genes for cell survival and/or growth
BACKGROUND
[0004] The field of the invention relates to RNA interference
(RNAi) and the use of RNAi active sequences for treating diseases
and disorders. In particular, the field of the invention relates
to the use of toxic RNAi active sequences for killing cancer
cells.
[0005] Cancer therapy is only marginally effective and not
curative because tumor cells will often develop resistance and
metastasize. This resistance is driven by the enhanced mutagenesis
rate cancer cells experience and is particularly effective in
circumventing drugs designed to target a single molecule or
pathway.
[0006] We recently reported that siRNAs and shRNAs derived from
CD95, CD95L (17) and other genes in the human genome (47) kill
cancer cells through RNAi by targeting a network of critical
survival genes (15,17). DISE (“death induced by survival gene
elimination”) was found to involve simultaneous activation of
multiple cell death pathways and cancer cells have a hard time
developing resistance to this form of cell death (16). DISE was
found to preferentially affect transformed cells (16) and among
them cancer stem cells (44). We can artificially induce DISE in
cancer cells by introducing si/shRNAs that correspond to certain
gene transcript sequences and by using artificially designed
siRNAs that do not exist in nature optimized to induce DISE in
cancer cells (15). In fact, treating tumor bearing mice with
nanoparticle coupled DISE-inducing siRNAs resulted in reduced
tumor growth with no toxicity to the mice. We hypothesized that
DISE is part of a natural anticancer mechanism (15). We therefore
wondered whether this mechanism could be accidentally triggered
resulting in human pathology. Diseases that could be caused by
overactive DISE would have to result in loss of various tissues
and an RNA component would have to be involved in the disease
pathology.
[0007] Here, we report that certain repeat sequences (CAG/CUG
repeats) that are found to cause tissue loss in a number of
degenerative diseases (i.e., Huntington's disease), are highly
toxic to cancer cells by targeting multiple survival genes
containing complementary trinucleotide repeat sequences. The
significance of this discovery is that cancer cells would need to
develop multiple concurring mutations to become resistant to this
form of DISE because multiple death pathways are activated at once
(17). Indeed, we have not found a single compound or knockdown of
any gene that can rescue cancer cells from DISE and this DISE-like
cell death preferentially affects cancer cells (17,44). Here we
describe an entirely new way to kill cancer cells in vivo by
introducing trinucleotide repeat derived siRNAs that
preferentially target a large number of genes that are critical
for the survival of cancer cells.
SUMMARY
[0008] Disclosed are polynucleotides, compositions, and methods
related to RNA interference (RNAi). The disclosed polynucleotides,
compositions, and methods may be utilized for treating diseases
and disorders through RNAi.
[0009] Particularly disclosed are toxic RNAi active seed sequences
and methods of using toxic RNAi active seed sequences for killing
cancer cells. The disclosed toxic RNAi active seed sequences
typically include trinucleotide RNA nucleotide repeats and
preferentially target and inhibit the expression of multiple
essential genes for cell survival and/or growth. The disclosed
toxic RNAi active seed may be presented or administered in siRNAs,
shRNAs, and/or vectors that express siRNAs and/or shRNAs.
BRIEF DESCRIPTION OF THE FIGURES
[0010] FIGS. 1A-E. An siRNA duplex comprised of CAG and CUG
repeats is super toxic to various cancer cell lines of human and
mouse origin.
[0011] FIG. 1A. Sequence of the siCAG/CUG duplex. dA,
deoxyadenosine.
[0012] FIG. 1B and FIG. 1C. Confluences over time of human (B) and
mouse (C) cell lines transfected with 10 nM of either siNT, siL3,
or siCAG/CUG. M565, 3LL and B 16F10 cells were reverse
transfected. Values are mean −/+SEM. The 13 cell lines were tested
between 1 and 6 times in each case with between 3 and 6 technical
replicates.
[0013] FIG. 1D. Confluence over time of HCT116 cells transfected
with either siNT or siCAG/CUG at 0.1 or 0.01 nM. Values are mean
−/+SEM. n=2 biological replicates, 6 technical replicates each.
[0014] FIG. 1E. Viability (ATP content) of HeyA8 cells transfected
with different concentrations of siNT, siL3 or siCAG/CUG. Values
are mean −/+SD. n=3 biological replicates, 3 technical replicates
each.
[0015] FIGS. 2A-E. Identification of the most toxic TNR-based
siRNAs.
[0016] FIG. 2A. Confluence over time of HeyA8 (top) or A549
(bottom) cells transfected with 1 nM of either siNT or the four
trinucleotide repeat (TNR) based siRNAs, siCAG/CUG, siCGG/CCG,
siGAA/UUC and siCGA/UCG. Values are mean −/+SEM. n=3/2 biological
replicates (HeyA8/A549), 8 technical replicates.
[0017] FIG. 2B. Results of toxicity screens of 60 TNR-based siRNAs
in human HeyA8 (top) and mouse M565 (bottom) cells. Cells were
reverse transfected in triplicate in 384 well plates with 1 nM of
each siRNA. In all cases, the complementary (sense) strand was
inactivated by a 2′-OMe modification in positions 1 and 2.
Trinucleotide repeats were characterized as effecting: (i) no or
less than 50% loss in viability; (ii) >50% loss in viability;
or (iii) >75% loss in viability. Family 7 is the only family in
which all 6 duplexes were super toxic. The six members of Family 5
have the same GC content and the same nucleotide composition as
the members of family 7. Each family contains three TNRs in the
forward frame (for) and three TNRs that are the reverse complement
(rev) of these TNRs. Values are mean −/+SD. 3 technical
replicates.
[0018] FIG. 2C. Correlation between the average of two viability
screens performed in HeyA8 and two performed in M565 cells. The
data points for the 6 members of TNR family 5 and family 7 are
labeled. Pearson correlation and p-value is given.
[0019] FIG. 2D. Results of toxicity screen of 60 TNR-based 6 mer
seeds in a nontoxic backbone siRNA (see FIG. 10B) in HeyA8 cells.
Values are mean −/+SD. n=3 technical replicates.
[0020] FIG. 2E. Correlation between the average of two viability
screens performed in HeyA8 with the 60 TNR based siRNAs and the
screen performed with the 60 TNR based seeds. The data points for
the 6 members of TNR family 7 and family 10 are labeled. Each
member of family 7 effected < ̃20% viability and no member of
family 10 effected < ̃75% viability. Pearson correlation and
p-value is given.
[0021] FIGS. 3A-E. siCAG/CUG kills cancer cells through RNAi.
[0022] FIG. 3A. Western blot analysis of HeyA8 and A549 cells
treated with a control SmartPool (siCtr) or an AGO2 siRNA
SmartPool for 48 hrs.
[0023] FIG. 3B. Confluence over time of HeyA8 and A549 cells after
transfection with 1 nM siNT or siCAG/CUG. Cells were first
transfected with either 25 nM of a Ctr SmartPool or the AGO2
SmartPool and then after 24 hrs transfected with siNT or
siCAG/CUG. Values are mean −/+SEM. n=3 biological replicates, 3
technical replicates each.
[0024] FIG. 3C. Left: Sequences with positions of the
2′-O-methylation labeled in either the passenger/sense (S), the
guide/antisense (AS) or both (A/AS) strands of siCAG/CUG. Right:
Confluence over time of HeyA8 cells transfected with 10 nM of the
four duplexes depicted on the left and two similarly modified
duplexes derived from siNT. siCAG/CUG and siCAG (5-OMe) effected a
significant reduction in % confluence. Values are mean −/+SEM. n=2
biological replicates, 4-6 technical replicates each.
[0025] FIG. 3D. Venn diagrams showing the overlap between 1185
survival genes (genes identified as critical survival genes in two
genome-wide lethality screens [45, 46], blue) and genes
significantly downregulated (red) in cells transfected with either
siCAG/CUG (top) or siCGA/UCG (bottom) when compared to siNT.
[0026] FIG. 3E. Venn diagram comparing the 3466 genes
downregulated in the siCAG/CUG treated HeyA8 cells (>1.5 fold
adj. p <0.05) as determined by RNA-Seq (siCAG/CUG down) (see
FIG. 3D) and the 1236 genes expression of which inversely
correlate with the length of CAG repeats recently reported in HD
patients (CAGnome down) [24].
[0027] FIGS. 4A and 4B. The toxicity of the TNR-based siRNAs
correlates with the presence of long complementary repeats in the
ORFs of genes.
[0028] FIG. 4A. Correlation between the average of the two
viability screens performed in HeyA8 (top panels) or M565 (bottom
panels) cells with the percentage of TNRs that are part of 6 mer
or higher longer repeat sequences in either the ORF (left panels)
or 3′UTR (right panel) of the genes. The data points for the 6
members of TNR family 5 are labeled in blue and those of family 7
are labeled in red. Pearson correlations and p-values are given.
[0029] FIG. 4B. The amino acids that are coded by the targeted
trinucleotide repeats in families 5 and 7.
[0030] FIGS. 5A-G. Killing cancer cells using siCAG/CUG coupled to
TLP nanoparticles both in vitro and in vivo.
[0031] FIG. 5A. Left: Confluency over time of HeyA8 (Nuc-Red)
cells treated with 10 nM of either siNT-TLP siL3-TLP, or
siCAG/CUG-TLP. Right: Phase and red fluorescence image of HeyA8
(Nuc-Red) cells 90 hrs after transfection with 15 nM of TLPs.
Values are mean −/+SEM. n=3 technical replicates.
[0032] FIG. 5B. Confluency over time of different human and mouse
cancer cell lines treated with either 10 nM (OVCAR3, HepG2, M565)
or 20 nM (A549, T98G) of TLPs. Values are mean −/+SEM. n=2
biological replicates, 6 technical replicates each.
[0033] FIG. 5C. Percent cell death (Trypan blue counting) of
GIC-20 neurospheres derived from a patient with glioblastoma six
days after adding the TLPs (30 nM). Values are mean −/+SD. n=3
technical replicates. **p<0.01.
[0034] FIG. 5D. Treatment scheme.
[0035] FIG. 5E. Tumor growth over time based on small animal
imaging of 10<5 >HeyA8-Nuc-red-Luc-neo cells injected i.p.
into NSG mice treated with either siNT-TLPs or siCAG/CUG-TLPs.
Treatment group 1 received 18 injections over four weeks and
treatment group 2 10 injections over three weeks. The
bioluminescence signal of IVIS #4 and #5 for individual mice is
shown (right panel). The experiment represents one of two similar
experiments. Values are mean −/+SD. * p<0.05; **p<0.01;
***p<0.0001, NS, not significant.
[0036] FIG. 5F and FIG. 5G. Change in red object count (growth) of
tumor cells from 3 mice of the siNT-TLP and the siCAG/CUG-TLP
treatment group 1 either after transfection with 1 nM siNT or
siCAG/CUG (F) or after incubation with 7.5 nM siNT-TLP or
siCAG/CUG-TLP (G). 1000 cells per well were plated. Values are
mean −/+SEM. n=3-8 technical replicates.
[0037] FIG. 6A and 6B. Morphological changes and cell death in
cells transfected with siCAG/CUG.
[0038] FIG. 6A. Left: DNA fragmentation in HeyA8 and A549 cells
120/72 hours after transfection with indicated siRNAs. n=3-4
technical replicates. Right: Viability of human and mouse cancer
cell lines 96 hours after transfection with the indicated siRNAs.
Values are mean −/+SD. n=2 biological replicates, 3 technical
replicates each.
[0039] FIG. 6B. Human and mouse cell lines transfected with either
siNT or siCAG/CUG. Time points after transfection of picture
taking is given. Size bar=100 μm.
[0040] FIG. 7. Properties of toxic TNR based siRNAs. Left:
Sequences of the siCAG repeat in three different frames. Right:
Confluence over time of HeyA8 cells transfected with 1 nM of
either siNT or the three duplexes depicted on the left. Values are
mean −/+SEM. n=3 biological replicates, 4-8 technical replicates
each.
[0041] FIG. 8. The toxicity of siL3 is solely based on its guide
strand. Left: Scheme showing positions of the 2′O-methylation in
either the passenger/sense (S) or the guide/antisense (AS) strand
of siL3. The siL3 seed region is shown as a green box. Right:
Confluence over time of HeyA8 cells transfected with 10 nM of the
four duplexes depicted on the left and two similarly modified
duplexes derived from siNT. Values are mean −/+SEM. n=3 technical
replicates.
[0042] FIG. 9. Reproducibility of the TNR siRNA screens. Variation
between the two screens of the 60 TNR based siRNAs in HeyA8 (left)
and in M565 (right) cells. The data points for the six members of
TNR family 5 are labeled in blue those of family 7 are labeled in
red. Pearson correlations and p-values are given.
[0043] FIG. 10A and 10B. DISE inducing activity of siL3 is mostly
based on its seed sequence.
[0044] FIG. 10A. Left: Scheme showing the different mutant siNT
and siL3 duplexes used. The four light blue boxes indicate the
four positions that in the nontargeting siRNA (siNT) were
identical to the same positions in siL3. In siL3 the seed sequence
is shown as a green box, and the siNT seed sequence is shown as a
grey box. siNT/siL3 is a chimeric duplex comprised of the seed
sequence of siL3 and the rest of siNT. In the siL3 seed duplex,
the four siL3 positions in siNT were replaced with the
complementary nucleotides (i.e. an G:C was changed to a C:G).
Right: Confluence over time of HeyA8 cells transfected with 10 nM
of the five duplexes depicted on the left. Values are mean −/+SEM.
n=6 technical replicates.
[0045] FIG. 10B. Schematic showing the design of the seed siRNAs
tested in FIG. 2D. Y and Z indicate the Watson-Crick complementary
nucleotides of the 6 mer seed. The two red Xs indicate the
position of the 2′O-methylation in the passenger strand.
[0046] FIG. 11. Increased sensitivity of HCT116 Drosha<−/−>
cells compared to HCT116 wild-type cells. Confluency over time of
either HCT116 wt (left) or Drosha<−/−> (right) cells
transfected with 0.1 nM of either siNT or siCAG/CUG. Transfection
efficiency of the two cell lines was similar as assessed by uptake
of siGLO [1]. p-value according to polynomial distribution is
given. Values are mean −/+SEM. n=3 biological replicates, 3-4
technical replicates each.
[0047] FIG. 12A and 12B. siCAG/CUG induced growth reduction of
mouse embryonic stem cells requires Ago2.
[0048] FIG. 12A. Sensitivity of mouse embryonic stem cells to
siCAG/CUG lacking expression of all four AGO proteins with the
same cells in which we rendered the RISC functional by
re-expression of human AGO2. Western blot analysis of Ago1-4 k.o.
mouse embryonic stem cells expressing a Tet inducible AGO2
protein. Lanes are as follows: 1: Low Dox; 2: 4 days without Dox;
3: 3 days without Dox-1 day with high Dox; 4: 3 days without Dox-1
day with high Dox-1 day without Dox; 5: 3 days without Dox-1 day
with high Dox-2 days without Dox; 6: 3 days without Dox-1 day with
high Dox-3 days without Dox; 7: 3 days without Dox-1 day with high
Dox-4 days without Dox; 8: 3 days without Dox-1 day with high
Dox-7 days without Dox. Low Dox=0.1 μg/ml; high Dox=2.5 μg/ml. n=2
biological replicates.
[0049] FIG. 12B. Confluence over time of Ago1-4 k.o. cells with
induced AGO2 with high Dox (left) or without Dox (right) after
transfection with 5 nM siNT or siCAG/CUG. Two-way ANOVA is given.
Equal transfection efficiency was established by transfecting
cells with 5 nM of either siNT or 5′Cy5 labeled siNT followed by
FACS analysis (inserts). Values are mean −/+SEM. n=3 biological
replicates, 4 technical replicates each. AGO2 expressing Ago1-4
k.o. ESCs over-expressing human AGO2 showed a very low but
reproducible susceptibility to siCAG/CUG. In contrast, Ago1-4 k.o.
cells were completely resistant to this form of toxicity, despite
similar transfection efficiencies (see inserts) between Ago1-4
k.o. cells and those expressing AGO2. The low sensitivity of the
Ago1-4 k.o./AGO2 cells could either be due to normal cells being
less sensitive to this form of cell death or could point at a
functional role of Ago family members other than Ago2 in this
process.
[0050] FIG. 13. Efficient knockdown of CUG repeat containing genes
in cells transfected with siCAG. HeyA8 cells were transfected with
1 nM of either siNT, siCAG/CUG, siCAG (with the CUG containing
passenger strand modified by 2′O-methylation), or siCUG (with the
CAG containing passenger strand modified by 2′O-methylation). RNA
was quantified by real-time PCR. The genes are ranked according to
their highest fold downregulation in the RNA Seq experiment.
Values are mean −/+SD. n=2 biological replicates (for siNT and
siCAG/CUG), 3 technical replicates each.
[0051] FIG. 14A and 14B. Genes containing the 19 mer targeted by
siCAG are poorly conserved between human and mouse.
[0052] FIG. 14A. Venn diagram of human and mouse ORFs and 3′UTRs
containing the CAG (left) or CGA (right) trinucleotide repeats.
[0053] FIG. 14B. Venn diagram of human and mouse ORFs and 3′UTRs
containing the 19 mer sequences completely complementary to the
CAG (left) or CGA (right) based 19 mer.
[0054] FIGS. 15A-C. No adverse effects in mice treated with
siCAG/CUG-TLPs.
[0055] FIG. 15A. Weight of the ten mice in treatment group 1 (see
FIG. 5D) over the course of the treatment. Values are mean −/+SD.
NS, unpaired p-value not significant.
[0056] FIG. 15B. H&E stained liver sections of two of the mice
that were treated with either siNT-TLP or siCAG/CUG-TLP on day 27
in the experiment shown in FIG. 5E).
[0057] FIG. 15C. Serum analysis of the same two mice per treatment
group. 1=Sample assay value is less than the dynamic range. For
most assays, the dynamic range low limit is reported. 2=Sample was
diluted for testing. Assay value for sample was below dynamic
range, but results have been corrected for dilution. 3=Assay is a
calculated value. Either or both assay values used in the
calculation were below the dynamic range of the assay, therefore
no result is reported.
DETAILED DESCRIPTION
[0068] The disclosed technology relates to nucleic acid and the
use of nucleic acid for treated diseases and disorders. The terms
“nucleic acid” and “oligonucleotide,” as used herein, refer to
polydeoxyribonucleotides (containing 2-deoxy-ribose),
polyribonucleotides (containing ribose), and to any other type of
polynucleotide that is an N glycoside of a purine or pyrimidine
base. As used herein, the terms “A,” “T,” “C”, “G” and “U” refer
to adenine, thymine, cytosine, guanine, uracil as a nucleotide
base, respectively. There is no intended distinction in length
between the terms “nucleic acid,” “oligonucleotide,” and
“polynucleotide,” and these terms will be used interchangeably.
These terms refer only to the primary structure of the molecule.
Thus, these terms include double- and single-stranded DNA, as well
as double- and single-stranded RNA. For use in the present
invention, an oligonucleotide also can comprise nucleotide analogs
in which the base, sugar or phosphate backbone is modified as well
as non-purine or non-pyrimidine nucleotide analogs.
[0069] A “fragment” of a polynucleotide is a portion of a
polynucleotide sequence which is identical in sequence to but
shorter in length than a reference sequence. A fragment may
comprise up to the entire length of the reference sequence, minus
at least one nucleotide. For example, a fragment may comprise from
5 to 1000 contiguous nucleotides of a reference polynucleotide. In
some embodiments, a fragment may comprise at least 5, 10, 15, 20,
25, 30, 40, 50, 60, 70, 80, 90, 100, 150, 250, or 500 contiguous
nucleotides of a reference polynucleotide; in other embodiments a
fragment may comprise no more than 5, 10, 15, 20, 25, 30, 40, 50,
60, 70, 80, 90, 100, 150, 250, or 500 contiguous nucleotides of a
reference polynucleotide; in further embodiments a fragment may
comprise a range of contiguous nucleotides of a reference
polynucleotide bounded by any of the foregoing values (e.g. a
fragment comprising 20-50 contiguous nucleotides of a reference
polynucleotide). Fragments may be preferentially selected from
certain regions of a molecule. The term “at least a fragment”
encompasses the full length polynucleotide. A “variant,” “mutant,”
or “derivative” of a reference polynucleotide sequence may include
a fragment of the reference polynucleotide sequence.
[0070] Regarding polynucleotide sequences, percent identity may be
measured over the length of an entire defined polynucleotide
sequence, for example, as defined by a particular SEQ ID number,
or may be measured over a shorter length, for example, over the
length of a fragment taken from a larger, defined sequence, for
instance, a fragment of at least 20, at least 30, at least 40, at
least 50, at least 70, at least 100, or at least 200 contiguous
nucleotides. Such lengths are exemplary only, and it is understood
that any fragment length supported by the sequences shown herein,
in the tables, figures, or Sequence Listing, may be used to
describe a length over which percentage identity may be measured.
[0071] Regarding polynucleotide sequences, “variant,” “mutant,” or
“derivative” may be defined as a nucleic acid sequence having at
least 50% sequence identity to the particular nucleic acid
sequence over a certain length of one of the nucleic acid
sequences using blastn with the “BLAST 2 Sequences” tool available
at the National Center for Biotechnology Information's website.
(See Tatiana A. Tatusova, Thomas L. Madden (1999), “Blast 2
sequences—a new tool for comparing protein and nucleotide
sequences”, FEMS Microbiol Lett. 174:247-250). Such a pair of
nucleic acids may show, for example, at least 20%, at least 30%,
at least 40%, at least 50%, at least 60%, at least 70%, at least
80%, at least 85%, at least 90%, at least 91%, at least 92%, at
least 93%, at least 94%, at least 95%, at least 96%, at least 97%,
at least 98%, or at least 99% or greater sequence identity over a
certain defined length.
[0072] A “recombinant nucleic acid” is a sequence that is not
naturally occurring or has a sequence that is made by an
artificial combination of two or more otherwise separated segments
of sequence. This artificial combination is often accomplished by
chemical synthesis or, more commonly, by the artificial
manipulation of isolated segments of nucleic acids, e.g., by
genetic engineering techniques known in the art. The term
recombinant includes nucleic acids that have been altered solely
by addition, substitution, or deletion of a portion of the nucleic
acid. Frequently, a recombinant nucleic acid may include a nucleic
acid sequence operably linked to a promoter sequence. Such a
recombinant nucleic acid may be part of a vector that is used, for
example, to transform a cell.
[0073] The nucleic acids disclosed herein may be “substantially
isolated or purified.” The term “substantially isolated or
purified” refers to a nucleic acid that is removed from its
natural environment, and is at least 60% free, preferably at least
75% free, and more preferably at least 90% free, even more
preferably at least 95% free from other components with which it
is naturally associated.
[0074] Oligonucleotides can be prepared by any suitable method,
including direct chemical synthesis by a method such as the
phosphotriester method of Narang et al., 1979, Meth. Enzymol.
68:90-99; the phosphodiester method of Brown et al., 1979, Meth.
Enzymol. 68:109-151; the diethylphosphoramidite method of Beaucage
et al., 1981, Tetrahedron Letters 22:1859-1862; and the solid
support method of U.S. Pat. No. 4,458,066, each incorporated
herein by reference. A review of synthesis methods of conjugates
of oligonucleotides and modified nucleotides is provided in
Goodchild, 1990, Bioconjugate Chemistry 1(3): 165-187,
incorporated herein by reference.
[0075] The term “promoter” as used herein refers to a cis-acting
DNA sequence that directs RNA polymerase and other trans-acting
transcription factors to initiate RNA transcription from the DNA
template that includes the cis-acting DNA sequence.
[0076] As used herein, the term “complementary” in reference to a
first polynucleotide sequence and a second polynucleotide sequence
means that the first polynucleotide sequence will base-pair
exactly with the second polynucleotide sequence throughout a
stretch of nucleotides without mismatch. The term “cognate” may in
reference to a first polynucleotide sequence and a second
polynucleotide sequence means that the first polynucleotide
sequence will base-pair with the second polynucleotide sequence
throughout a stretch of nucleotides but may include one or more
mismatches within the stretch of nucleotides. As used herein, the
term “complementary” may refer to the ability of a first
polynucleotide to hybridize with a second polynucleotide due to
base-pair interactions between the nucleotide pairs of the first
polynucleotide and the second polynucleotide (e.g., A:T, A:U, C:G,
G:C, G:U, T:A, U:A, and U:G).
[0077] As used herein, the term “complementarity” may refers to a
sequence region on an anti-sense strand that is substantially
complementary to a target sequence but not fully complementary to
a target sequence. Where the anti-sense strand is not fully
complementary to the target sequence, mismatches may be optionally
present in the terminal regions of the anti-sense strand or
elsewhere in the anti-sense strand. If mismatches are present,
optionally the mismatches may be present in terminal region or
regions of the anti-sense strand (e.g., within 6, 5, 4, 3, or 2
nucleotides of the 5′ and/or 3′ terminus of the anti-sense
strand).
[0078] The term “hybridization,” as used herein, refers to the
formation of a duplex structure by two single-stranded nucleic
acids due to complementary base pairing. Hybridization can occur
between fully complementary nucleic acid strands or between
“substantially complementary” nucleic acid strands that contain
minor regions of mismatch. Conditions under which hybridization of
fully complementary nucleic acid strands is strongly preferred are
referred to as “stringent hybridization conditions” or
“sequence-specific hybridization conditions.” Stable duplexes of
substantially complementary sequences can be achieved under less
stringent hybridization conditions; the degree of mismatch
tolerated can be controlled by suitable adjustment of the
hybridization conditions. Those skilled in the art of nucleic acid
technology can determine duplex stability empirically considering
a number of variables including, for example, the length and base
pair composition of the oligonucleotides, ionic strength, and
incidence of mismatched base pairs, following the guidance
provided by the art (see, e.g., Sambrook et al., 1989, Molecular
Cloning—A Laboratory Manual, Cold Spring Harbor Laboratory, Cold
Spring Harbor, N.Y.; Wetmur, 1991, Critical Review in Biochem. and
Mol. Biol. 26(3/4):227-259; and Owczarzy et al., 2008,
Biochemistry, 47: 5336-5353, which are incorporated herein by
reference).
[0079] As used herein, the term “double-stranded RNA” (“dsRNA”)
refers to a complex of ribonucleic acid molecules having a duplex
structure comprising two anti-parallel and substantially
complementary nucleic acid strands.
[0080] As used herein, the term “nucleotide overhang” refers to an
unpaired nucleotide or nucleotides that extend from the 5′-end or
3′-end of a duplex structure of a dsRNA when a 5′-end of one
strand of the dsRNA extends beyond the 3′-end of the other strand,
or when a 3′-end of one strand of the dsRNA extends beyond the
5′-end of the other strand.
[0081] As used herein, the term “blunt” refers to a dsRNA in which
there are no unpaired nucleotides at the 5′-end and/or the 3′-end
of the dsRNA (i.e., no nucleotide overhang at the 5′-end or the
3′-end). A “blunt ended” dsRNA is a dsRNA that has no nucleotide
overhang at the 5′-end or the 3′-end of the dsRNA molecule.
[0082] As used herein, the term “anti-sense strand” refers to a
strand of a dsRNA which includes a region that is substantially
complementary to a target sequence (i.e., where the target
sequence has a sequence corresponding to the sense strand).
[0083] As used herein, the term “sense strand,” refers to the
strand of a dsRNA that includes a region that is substantially
complementary to a region of the anti-sense strand and that
includes a region that substantially corresponds to a region of
the target sequence.
[0084] As used herein, RNAi active sequences may include “siRNA”
and “shRNA” and dsRNA that is processed by nucleases to provide
siRNA and/or shRNA. The term “siRNA” refers to a “small
interfering RNA” and the term “shRNA” refers to “short hairpin
RNA.” RNA interference (RNAi) refers to the process of
sequence-specific post-transcriptional gene silencing in a cell or
an animal mediated by siRNA and/or shRNA.
[0085] As used herein, the term “siRNA targeted against mRNA”
refers to siRNA specifically promote degradation of the targeted
mRNA via sequence-specific complementary multiple base pairings
(e.g., at least 6 contiguous base-pairs between the siRNA and the
target mRNA at optionally at least 7, 8, 9, 10, 11, 12, 13, 14,
15, 16, 17, 18, 19, 20, 21, 22, or 23 contiguous base-pairs
between the siRNA and the target mRNA).
[0086] The terms “target, “target sequence”, “target region”, and
“target nucleic acid,” as used herein, are synonymous and refer to
a region or sequence of a nucleic acid which may be selected as a
sequence to which the anti-sense strand of siRNA or shRNA is
substantially complementary to and hybridizes to as discussed
herein. A target sequence may refer to a contiguous portion of a
nucleotide sequence of an mRNA molecule of a particular gene,
including but not limited to, genes that are essential for
survival and/or growth of cells and in particular cancer cells.
The target sequence of a siRNA refers to a mRNA sequence of a gene
that is targeted by the siRNA due to complementarity between the
anti-sense strand of the siRNA and the mRNA sequence and to which
the anti-sense strand of the siRNA hybridizes when brought into
contact with the mRNA sequence.
[0087] As used herein, the term “transfecting” means “introducing
into a cell” a molecule, which may include a polynucleotide
molecule such as dsRNA. When referring to a dsRNA, transfecting
means facilitating uptake or absorption into the cell, as is
understood by the skilled person. Absorption or uptake of dsRNA
can occur or may be facilitated through passive diffusive or
active cellular processes, or through the use of auxiliary agents
or devices. Transfection into a cell includes methods known in the
art such as electroporation and lipofection. However, the meaning
of the term “transfection” is not limited to introducing molecules
into cells in vitro. As contemplated herein, a dsRNA also may be
“introduced into a cell,” where the cell is part of a living
organism. For example, for in vivo delivery, a dsRNA may be
injected into a tissue site or may be administered systemically.
[0088] As used herein, the terms “silencing” and “inhibiting the
expression of” refer to at least partial suppression of the
expression of a target gene, for example, as manifested by a
reduction of mRNA associated with the target gene.
[0089] As used herein, the phrase “effective amount” shall mean
that drug dosage that provides the specific pharmacological
response for which the drug is administered in a significant
number of patients in need of such treatment. An effective amount
of a drug that is administered to a particular patient in a
particular instance will not always be effective in treating the
conditions/diseases described herein, even though such dosage is
deemed to be a therapeutically effective amount by those of skill
in the art.
[0090] As used herein, the term “pharmaceutical composition” may
include be defined as a composition that includes a
pharmacologically effective amount of a dsRNA and a
pharmaceutically acceptable carrier for delivering the dsRNA to
target cells or target tissue. As used herein, the term
“pharmaceutically acceptable carrier” refers to a carrier for
administration of a therapeutic agent which facilitates the
delivery of the therapeutic agent (e.g., dsRNA) to target cells or
target tissue. As used herein, the term “therapeutically effective
amount” refers to that amount of a therapeutic agent that provides
a therapeutic benefit in the treatment, prevention, or management
of a disease or disorder (e.g., a cell proliferation disease or
disorder such as cancer).
[0091] In one aspect, the present inventors disclose an isolated
double stranded short interfering ribonucleic acid (siRNA)
molecule or small hairpin ribonucleotide acid (shRNA) molecule
that silences expression of one or more mRNA' s of essential genes
that are required for survival and growth of cells such as cancer
cells. Preferably, the disclosed siRNA molecules or shRNA
molecules silence the expression of multiple mRNA's of essential
genes that are required for survival and growth of cells such as
cancer cells through a process similar to the process called
“death-induced by survival gene elimination” or “DISE.”
[0092] The mechanism of action of siRNA and shRNA is understood by
the skilled person. Interfering RNA (RNAi) generally refers to a
single-stranded RNA (ssRNA) or double-stranded RNA (dsRNA). The
dsRNA is capable of targeting specific messenger RNA (mRNA) and
silencing (i.e., inhibiting) the expression of a target gene.
During this process, dsRNA (which may include shRNA) is
enzymatically processed into short-interfering RNA (siRNA)
duplexes of ̃21-23 nucleotides in length. The anti-sense strand of
the siRNA duplex is then incorporated into a cytoplasmic complex
of proteins (RNA-induced silencing complex or RISC). The RISC
complex containing the anti-sense siRNA strand also binds mRNA
which has a sequence complementary to the anti-sense
strand-allowing complementary base-pairing between the anti-sense
siRNA strand and the sense mRNA molecule. The mRNA molecule is
then specifically cleaved by an enzyme (RNase) associated with
RISC resulting in specific gene silencing. For gene silencing or
knock down (i.e., mRNA cleavage) to occur, anti-sense RNA (e.g.,
siRNA) has to become incorporated into the RISC. This represents
an efficient process that occurs in nucleated cells during
regulation of gene expression.
[0093] As such, siRNA-mediated RNA interference may be considered
to involve two-steps: (i) an initiation step, and (ii) an effector
step. In the first step, input siRNA is processed into small
fragments, such as ̃21-23-nucleotide ‘guide sequences.’ The guide
RNAs can be incorporated into the protein-RNA RISC complex which
is capable of degrading mRNA. As such, the RISC complex acts in
the second effector step to destroy mRNAs that are recognized by
the guide RNAs through base-pairing interactions. RNA interference
via use of siRNA may be considered to involve the introduction by
any means of double stranded RNA into a cell which triggers events
that cause the degradation of a target RNA, and as such siRNA may
be considered to be a form of post-transcriptional gene silencing.
The skilled person understands how to prepare and utilize siRNA
molecules. (See, e.g., Hammond et al., Nature Rev Gen 2: 110-119
(2001); and Sharp, Genes Dev 15: 485-490 (2001), the contents of
which are incorporate herein by reference in their entireties).
[0094] For purposes of this application, the anti-sense strand of
the disclosed RNA molecules (e.g., siRNA molecules) may comprise a
contiguous nucleotide sequence, where the base sequence of the
anti-sense strand has substantial or complete sequence
complementarity to the base sequence of a contiguous nucleotide
sequence of corresponding length contained in an mRNA sequence of
the targeted mRNA (e.g., in a non-coding 3′-end of an mRNA
sequence). Substantial complementary permits some nucleotide
mismatches (i.e., non-pairing nucleotides) and as such, the
anti-sense strand of the siRNA need not have full complementarity.
[0095] In some embodiments, at least a portion of an anti-sense
strand of the disclosed RNA molecules (e.g., siRNA molecules)
comprises or consists of a sequence that is 100% complementary to
a target sequence or a portion thereof. In another embodiment, at
least a portion of an anti-sense strand of an siRNA molecule
comprises or consists of a sequence that is at least about 90%,
95%, or 99% complementary to a target sequence or a portion
thereof. For purposes of this application, the anti-sense strand
of the disclosed RNA molecules (e.g., siRNA molecules) preferably
comprises or consists of a sequence that specifically hybridizes
to a target sequence or a portion thereof so as to inhibit
expression of the target mRNA.
[0096] In some embodiments, the disclosed RNAs, including siRNAs
administered in RNAi therapy, may include repeat sequences. For
example, in some embodiments, the disclosed RNAs may include
trinucleotide repeats such as any of: (AAA)n, (AAC)n, (AAG)n,
(AAU)n, (ACA)n, (ACC)n, (ACG)n, (ACU)n, (AGA)n, (AGC)n, (AGG)n,
(AGU)n, (AUA)n, (AUC)n, (AUG)n, (AUU)n, (CAA)n, (CAC)n, (CAG)n,
(CAU)n, (CCA)n, (CCC)n, (CCG)n, (CCU)n, (CGA)n, (CGC)n, (CGG)n,
(CGU)n, (CUA)n, (CUC)n, (CUG)n, (CUU)n, (GAA)n, (GAC)n, (GAG)n,
(GAU)n, (GCA)n, (GCC)n, (GCG)n, (GCU)n, (GGA)n, (GGC)n, (GGG)n,
(GGU)n, (GUA)n, (GUC)n, (GUG)n, (GUU)n, (UAA)n, (UAC)n, (UAG)n,
(UAU)n, (UCA)n, (UCC)n, (UCG)n, (UCU)n, (UGA)n, (UGC)n, (UGG)n,
(UGU)n, (UUA)n, (UUC)n, (UUG)n, and (UUU)n, where n is an integer,
typically selected from 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
15, to about 100 or higher. Preferably, the disclosed RNAs may
include trinucleotide repeats such as any of: (AGC)n, (CAG)n,
(CUG)n, (GCA)n, (GGU)n and (UGC)n, where n is an integer,
typically selected from 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14,
15, to about 100 or higher.
[0097] Methods for preparing and isolating siRNA also are known in
the art. (See, e.g., Sambrook et al., Molecular Cloning, A
Laboratory Manual (2.sup.nd Ed., 1989), the content of which is
incorporated herein by reference in its entirety). The disclosed
siRNA may be chemically synthesized, using any of a variety of
techniques known in the art. The disclosed siRNA may include
modifications, for example, modifications that stabilize the siRNA
and/or protect the siRNA from degradation via endonucleases and/or
exonucleases. In some embodiments, the disclosed siRNA may include
nucleic acid protecting and coupling groups, such as
dimethoxytrityl at the 5′-end and/or phosphoramidites at the
3′-end.
[0098] In one embodiment, the disclosed RNAs comprise a double
stranded region of about 15 to about 30 nucleotides in length.
Preferably, the disclosed RNAs are about 20-25 nucleotides in
length. The disclosed RNAs of the present invention are capable of
silencing the expression of a target sequence in vitro and in
vivo.
[0099] In one embodiment, the dsRNA disclosed herein comprises a
hairpin loop structure and may be referred to as shRNA which may
be processed to a siRNA. In another embodiment, the dsRNA or siRNA
has an overhang on its 3′ or 5′ ends relative to the target RNA
which is to be cleaved. The overhang may be 2-10 nucleotides long.
In one embodiment, the dsRNA or siRNA does not have an overhang
(i.e., the dsRNA or siRNA has blunt ends).
[0100] In another embodiment, the disclosed RNA molecules (e.g.,
siRNA molecules) may contain one or more modified nucleotides,
including one or more modified nucleotides at the 5′ and/or 3′
terminus of the RNA molecules. In yet another embodiment, the
disclosed RNA molecules may comprise one, two, three four or more
modified nucleotides in the double-stranded region. Exemplary
modified nucleotides may include but are not limited to, modified
nucleotides such as 2′-O-methyl (2′OMe) nucleotides,
2′-deoxy-2′-fluoro (2′F) nucleotides, 2-deoxy nucleotides,
2′-O-(2-methoxyethyl) (MOE) nucleotides, and the like. The
preparation of modified siRNA is known by one skilled in the art.
In some embodiments, the disclosed dsRNA molecules include one or
more modified nucleotides at the 5′-terminus of the passenger
strand of the dsRNA that prevent incorporation of the passenger
strand into RISC. (See, e.g., Walton et al., Minireview:
“Designing highly active siRNAs for therapeutic applications,” the
FEBS Journal, 277 (2010) 4806-4813).
[0101] In some embodiments, the disclosed RNA molecules are
capable of silencing one or more target mRNAs and may reduce
expression of the one or more target mRNAs by at least about 20%,
30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% relative to a control
RNA molecule (e.g., a molecule not exhibiting substantial
complementarity with the target mRNA). As such, in some
embodiments, the presently disclosed RNA molecules targeting the
mRNA of essential genes may be used to down-regulate or inhibit
the expression of essential genes (e.g., by at least about 20%,
30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% relative to a control
RNA molecule).
[0102] The disclosed RNA molecules may conveniently be delivered
to a target cell or a target tissue through a number of delivery
systems. For example, RNA may be delivered via electroporation,
lipofection, calcium phosphate precipitation, plasmids, viral
vectors that express the RNA, viral nucleic acids, phage nucleic
acids, phages, cosmids, nanoparticles, or via transfer of genetic
material in cells or carriers such as cationic liposomes. In one
embodiment, transfection of RNA may employ viral vectors, chemical
transfectants, or physico-mechanical methods such as
electroporation and direct diffusion of DNA.
[0103] Also disclosed herein are pharmaceutical compositions
(e.g., pharmaceutical compositions comprising therapeutic RNA) and
methods of administering pharmaceutical compositions for treating
diseases and disorders (e.g., cell proliferative diseases and
disorders such as cancer). The pharmaceutical composition may
comprise one or more RNAs as therapeutic agents for inhibiting the
gene activity of one or more essential genes and a pharmaceutical
acceptable carrier. Pharmaceutically acceptable carriers include,
but are not limited to, excipients such as inert diluents,
disintegrating agents, binding agents, lubricating agents,
sweetening agents, flavoring agents, coloring agents and
preservatives. Suitable inert diluents include sodium and calcium
carbonate, sodium and calcium phosphate, and lactose, while corn
starch and alginic acid are suitable disintegrating agents.
Pharmaceutical compositions containing RNA may be administered to
a mammal in vivo to treat cancer. In one embodiment, the
pharmaceutical formulation includes a dosage suitable for oral
administration. In another embodiment, the pharmaceutical
formulation is designed to suit various means for RNA
administration. Exemplary means include uptake of naked RNA,
liposome fusion, intramuscular injection via a gene gun,
endocytosis and the like.
[0104] Toxic RNAi Active Sequence for Killing Cancer Cells
[0105] Disclosed herein are polynucleotide sequences that may be
utilized in therapeutic methods of killing cancer cells in a
subject in need thereof. The disclosed polynucleotide sequences
may be referred to as “toxic RNAi active sequences.” Particularly
disclosed are toxic RNAi active sequences such as siRNA and shRNA
and methods of using toxic RNAi active sequence for killing cancer
cells. The disclosed toxic RNAi active sequences typically include
trinucleotide repeats and preferentially target and inhibit the
expression of multiple essential genes for cell survival and/or
growth through a process similar to the process called
“death-induced by survival gene elimination” or “DISE.”
[0106] In some embodiments, the disclosed polynucleotide sequences
include nucleotide trinucleotide repeats such as any of: (AAA)n,
(AAC)n, (AAG)n, (AAU)n, (ACA)n, (ACC)n, (ACG)n, (ACU)n, (AGA)n,
(AGC)n, (AGG)n, (AGU)n, (AUA)n, (AUC)n, (AUG)n, (AUU)n, (CAA)n,
(CAC)n, (CAG)n, (CAU)n, (CCA)n, (CCC)n, (CCG)n, (CCU)n, (CGA)n,
(CGC)n, (CGG)n, (CGU)n, (CUA)n, (CUC)n, (CUG)n, (CUU)n, (GAA)n,
(GAC)n, (GAG)n, (GAU)n, (GCA)n, (GCC)n, (GCG)n, (GCU)n, (GGA)n,
(GGC)n, (GGG)n, (GGU)n, (GUA)n, (GUC)n, (GUG)n, (GUU)n, (UAA)n,
(UAC)n, (UAG)n, (UAU)n, (UCA)n, (UCC)n, (UCG)n, (UCU)n, (UGA)n,
(UGC)n, (UGG)n, (UGU)n, (UUA)n, (UUC)n, (UUG)n, (UUU)n, where n is
an integer, typically selected from 3, 4, 5, 6, 7, 8, 9, 10, 11,
12, 13, 14, 15, to about 100 or higher. Preferably, the disclosed
toxic RNAi active sequences may include trinucleotide repeats such
as any of: (AGC)n, (CAG)n, (CUG)n, (GCA)n, (GGU)n, and (UGC)n,
where n is an integer, typically selected from 3, 4, 5, 6, 7, 8,
9, 10, 11, 12, 13, 14, 15, to about 100 or higher. For example,
the disclosed toxic RNAi active sequences may include the
nucleotide repeat (CAG)n, or (CUG)n, where n is 3, 4, 5, 6, 7, 8,
9, 10, 11, 12, 13, 14, 15, to about 100 or higher.
[0107] In some embodiments, the disclosed polynucleotide sequences
comprise a passenger strand and a guide strand, and may include
siRNAs and/or shRNAs. The disclosed polynucleotide sequences also
optionally and preferably comprise: (i) an RNA nucleotide
trinucleotide repeat sequence (X1X2X3)n, wherein X1, X2, and X3
independently are selected from any ribonucleotide A, C, G, and U,
and n is an integer from 3-10; and where optionally and preferably
(ii) one or more modified nucleotides at the 5′-terminus of the
passenger strand that prevents loading of the passenger strand
into the RNA-induced silencing complex (RISC). Preferably, the RNA
trinucleotide repeat sequence has a GC content of at least 66%.
[0108] In some embodiments, the trinucleotide repeat sequence of
the disclosed polynucleotide sequences is selected from the group
consisting of (ACC)n, (ACG)n, (AGC)n, (AGG)n, (CAC)n, (CAG)n,
(CCA)n, (CCC)n, (CCG)n, (CCU)n, (CGA)n, (CGC)n, (CGG)n, (CGU)n,
(CUC)n, (CUG)n, (GAC)n, (GAG)n, (GCA)n, (GCC)n, (GCG)n, (GCU)n,
(GGA)n, (GGC)n, (GGG)n, (GGU)n, (GUC)n, (GUG)n, (UCC)n, (UCG)n,
(UGC)n, and (UGG)n, Preferably, the trinucleotide repeat sequence
of the disclosed polynucleotide sequences is selected from the
group consisting of (AGC)n, (CAG)n, (CUG)n, (GCA)n, (GGU)n, and
(UGC)n, where n is an integer, typically selected from 3, 4, 5, 6,
7, 8, 9, 10, 11, 12, 13, 14, 15, to about 100 or higher.
Particularly, in some embodiments, the trinucleotide repeat
sequence of the disclosed polynucleotide sequences is CAG or CUG
and is present in the guide strand.
[0109] Typically, the disclosed polynucleotide sequences comprise
one or more modified nucleotides at the 5′-terminus of the
passenger strand that prevents loading of the passenger strand
into the RNA-induced silencing complex (RISC). In particular, the
passenger strand may comprise at least two modified nucleotides at
its 5′-terminus that prevents loading of the passenger strand into
the RNA-induced silencing complex (RISC). Suitable modified
nucleotides may include, but are not limited to 2′-O-methyl
(2′OMe) nucleotides, 2′-deoxy-2′-fluoro (2′F) nucleotides,
2′-deoxy nucleotides, and 2′-O-(2-methoxyethyl) (MOE) nucleotides.
[0110] The disclosed polynucleotide sequences may include a 3′
overhang of one or more nucleotides at the passenger strand, the
guide strand, or both strands of the double-stranded
polynucleotide. In some embodiments, the double-stranded
polynucleotide comprises a 3′ overhang of one or two
deoxyribonucleotide residues (A, C, G, or T) in the passenger
strand (optionally one or two thymidine residues) and/or the
double-stranded polynucleotide comprises a 3′ overhang of one or
two deoxyribonucleotide residues (A, C, G, or T) in the guide
strand (optionally one or two adenosine residues).
[0111] The identified polynucleotides that exhibit toxicity to
cancer cells may be formulated as pharmaceutical compositions, for
example, as pharmaceutical compositions for treating cell
proliferative diseases and disorders such as cancer. The disclosed
pharmaceutical compositions may be administered to a subject in
need thereof, for example, a subject having a cell proliferative
disease or disorder such as cancer.
Illustrative Embodiments
[0112] The following Embodiments are illustrative and should not
be interpreted to limit the scope of the claimed subject matter.
[0113] Embodiment 1. A double-stranded polynucleotide comprising a
passenger strand and a guide strand, the double-stranded
polynucleotide optionally comprising: (i) a trinucleotide repeat
sequence (X1X2X3)n, wherein X1, X2, and X3 independently are
selected from any ribonucleotide A, C, G, and U, and n is an
integer from 3-10; and optionally where (ii) one or more modified
nucleotides at the 5′-terminus of the passenger strand that
prevents loading of the passenger strand into the RNA-induced
silencing complex (RISC).
[0114] Embodiment 2. The double-stranded polynucleotide of
embodiment 1, wherein the RNA is an siRNA or an shRNA.
[0115] Embodiment 3. The double-stranded polynucleotide of
embodiment 1 or 2, wherein the trinucleotide repeat sequence has a
GC content of at least 33%.
[0116] Embodiment 4. The double-stranded polynucleotide of any of
the foregoing embodiments, wherein the trinucleotide repeat
sequence has a GC content of at least 66%.
[0117] Embodiment 5. The double-stranded polynucleotide of any of
the foregoing embodiments, wherein the trinucleotide repeat
sequence is selected from the group consisting of (ACC)n, (ACG)n,
(AGC)n, (AGG)n, (CAC)n, (CAG)n, (CCA)n, (CCC)n, (CCG)n, (CCU)n,
(CGA)n, (CGC)n, (CGG)n, (CGU)n, (CUC)n, (CUG)n, (GAC)n, (GAG)n,
(GCA)n, (GCC)n, (GCG)n, (GCU)n, (GGA)n, (GGC)n, (GGG)n, (GGU)n,
(GUC)n, (GUG)n, (UCC)n, (UCG)n, (UGC)n, and (UGG)n, preferably
wherein the trinucleotide is selected from the group consisting of
(AGC)n, (CAG)n, (CUG)n, (GCA)n, (GGU)n, and (UGC)n, where n is an
integer, typically selected from 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, to about 100 or higher.
[0118] Embodiment 6. The double-stranded polynucleotide of any of
the foregoing embodiments, wherein the trinucleotide repeat
sequence is CAG or CUG and the trinucleotide repeat sequence is
present in the guide strand.
[0119] Embodiment 7. The double-stranded polynucleotide of any of
the foregoing embodiments, wherein the passenger strand comprises
at least two modified nucleotides at its 5′-terminus.
[0120] Embodiment 8. The double-stranded polynucleotide of any of
the foregoing embodiments, wherein the one or more modified
nucleotides at the 5′-terminus of the passenger strand are
selected from the group consisting of 2′-O-methyl (2′OMe)
nucleotides, 2′-deoxy-2′-fluoro (2′F) nucleotides, 2′-deoxy
nucleotides, and 2′-O-(2-methoxyethyl) (MOE) nucleotides.
[0121] Embodiment 9. The double-stranded polynucleotide of any of
the foregoing embodiments, wherein the double-stranded
polynucleotide comprises a 3′ overhang of one or more nucleotides
at the passenger strand, the guide strand, or both strands of the
double-stranded polynucleotide.
[0122] Embodiment 10. The double-stranded polynucleotide of any of
the foregoing embodiments, comprising a 3′ overhang of one or two
deoxyribonucleotide residues (A, C, G, or T) in the passenger
strand, optionally wherein the one or two deoxyribonucleotide
residues are thymidine residues.
[0123] Embodiment 11. The double-stranded polynucleotide of any of
the foregoing embodiments, comprising a 3′ overhang of one or two
deoxyribonucleotide residues (A, C, G, or T) in the guide strand,
optionally wherein the one or two deoxyribonucleotide residues are
adenosine residues.
[0124] Embodiment 12. An expression vector that expresses the
polynucleotide of any of the foregoing embodiments or a
single-stranded portion thereof.
[0125] Embodiment 13. The expression vector of embodiment 12
comprising a eukaryotic promoter operably linked to DNA encoding
the polynucleotide or a single-stranded portion thereof.
[0126] Embodiment 14. The expression vector of embodiment 12 or
13, wherein the expression vector is a plasmid or a viral
expression vector.
[0127] Embodiment 15. A pharmaceutical composition comprising: (i)
the double-stranded polynucleotide of any of the foregoing
embodiments or a single-stranded portion thereof, or a vector for
expressing the double-stranded polynucleotide of any of the
foregoing embodiments or a single-stranded portion thereof; and
(ii) a pharmaceutically acceptable excipient.
[0128] Embodiment 16. A method for treating a disease or disorder
in a subject in need thereof, the method comprising administering
to the subject the pharmaceutical composition of embodiment 12.
[0129] Embodiment 17. The method of embodiment 13, wherein the
disease or disorder is a cell proliferative disease or disorder
such as cancer.
[0130] Embodiment 18. A method of inhibiting the growth of a cell
or killing a cell, the method comprising introducing into the cell
the double-stranded polynucleotide of any of embodiments 1-11 into
the cell or a single-stranded portion thereof, or introducing a
vector that expresses the double-stranded polynucleotide of any of
embodiments 1-11 or a single-stranded portion thereof.
[0131] Embodiment 19. A nanoparticle comprising the polynucleotide
of any of embodiments 1-11 or a single-stranded portion thereof.
[0132] Embodiment 20. The nanoparticle of embodiment 19, wherein
the nanoparticle is a nanoparticle formed from lipoproteins and/or
phospholipids (e.g., wherein the nanoparticle is a liposome or a
micelle).
[0133] Embodiment 21. The nanoparticle of embodiment 19 or 20,
wherein the polynucleotide is a siRNA and the siRNA is coupled to
a lipoprotein of the nanoparticle.
[0134] Embodiment 22. An expression vector that expresses the
polynucleotide of any of the foregoing embodiments or a
single-stranded portion thereof.
[0135] Embodiment 23. The expression vector of embodiment 22
comprising a eukaryotic promoter operably linked to DNA encoding
the polynucleotide or a single-stranded portion thereof.
[0136] Embodiment 24. The expression vector of embodiment 22 or
23, wherein the expression vector is a plasmid or a viral
expression vector.
EXAMPLES
[0137] The following Examples are illustrative and are
illustrative and should not be interpreted to limit the scope of
the claimed subject matter.
[0138] Title—Small interfering RNAs based on huntingtin
trinucleotide repeats are highly toxic to cancer cells
[0139] Reference is made to Murmann et al. “Small interfering RNAs
based on huntingtin trinucleotide repeats are highly toxic to
cancer cells,” (2018), the content of which is incorporate herein
by reference in its entirety.
[0140] Abstract
[0141] Trinucleotide repeat (TNR) expansions in the genome cause a
number of degenerative diseases. A prominent TNR expansion
involves the trinucleotide repeat CAG in the huntingtin (HTT) gene
responsible for Huntington's disease (HD). Pathology is caused by
protein and RNA generated from the TNR regions including small
siRNA-sized repeat fragments. An inverse correlation between the
length of the repeats in HTT and cancer incidence has been
reported for HD patients. We now show that siRNAs based on the CAG
TNR are toxic to cancer cells by targeting genes that contain long
reverse complimentary TNRs in their open reading frames. Of the 60
siRNAs based on the different TNRs, the 6 members in the CAG/CUG
family of related TNRs are the most toxic to both human and mouse
cancer cells. siCAG/CUG TNR-based siRNAs induce cell death in
vitro in all tested cancer cell lines and slow down tumor growth
in a preclinical mouse model of ovarian cancer with no signs of
toxicity to the mice. We propose to explore TNR-based siRNAs as a
novel form of anti-cancer reagents.
[0142] Introduction
[0143] Trinucleotide repeat (TNR) expansions are the cause of a
large number of degenerative disease syndromes characterized by
amplification of DNA triplet motifs [1]. They include
spinocerebellar ataxias (SCAs), spinobulbar muscular atrophy
(SBMA), myotonic dystrophy type 1 (DM1), and Huntington's disease
(HD) [1, 2]. HD is a dominantly inherited neurodegenerative
disorder caused by expansion of CAG repeats in the huntingtin
(HTT) gene. It has been shown that the resulting glutamine
expansions (polyQ) in HTT are toxic to cells [3, 4] and that the
length of the CAG amplifications determines severity and onset of
the disease [2, 4]. In addition to polyQ toxicity,
repeat-associated, non-ATG translation (RAN translation) was
discovered as another translation-level pathogenic mechanism of
CAG repeat-containing mRNAs [5]. More recent evidence however,
also points toward RNA playing a role in affecting cell viability
by poly-triplet repeats [1, 6]. Indeed, many of the repeats in
several TNR diseases are not located in open reading frames (ORFs)
but in introns or untranslated regions (UTRs) [4]. DM1 is the
best-characterized disease regarding RNA toxicity. The CUG repeats
are in the 3′UTR of the dystrophia myotonica protein kinase (DMPK)
gene, causing most of their toxicity by forming hairpin structures
[7]. These hairpins are believed to recruit a number of
RNA-binding proteins to nuclear RNA foci [8]. Another mechanism by
which CAG/CUG TNRs could be toxic at the RNA level is by
interfering with cellular splicing. This has been shown for CUG in
DM1 [9] and CAG in HD [10].
[0144] Mounting evidence suggests the CAG TNR expansions are toxic
at the RNA level. It was shown in Drosophila that the toxicity of
the CAG repeat disease gene spinocerebellar ataxia type 3 (SCA3)
protein ataxin-3, is in large part caused by the trinucleotide
repeat RNA and not the polyQ protein [11]. Replacing some of the
glutamine coding CAG repeats with the other codon coding for
glutamine, CAA, mitigated the toxicity despite similar polyQ
protein expression levels. Direct toxicity of mRNA with extended
CAG repeats was also demonstrated in mice [12]. Finally, there is
convincing evidence that CAG/CUG repeats can give rise to
RNAi-active small RNAs. In human neuronal cells, expression of the
CAG expanded exon 1 of HTT (above the threshold for complete
penetrance which is >40) [6] caused an increase in small CAG
repeat-derived RNAs (sCAG) of about 21 nt in length. Above a
certain length, CAG/CUG repeats were found to be cleaved by Dicer,
the enzyme that generates mature miRNAs from pre-miRNAs before
they are incorporated into the RNA induced silencing complex
(RISC) [13]. The CAG repeat derived fragments could bind to
complementary transcripts and downregulate their expression via an
RNAi-based mechanism. In a mouse model of HD treatment of the mice
with a locked nucleic acid-modified 20 mer antisense
oligonucleotide complementary to the CAG TNR (LNA-CTG) which
reduced the expression of sCAGs but not of HTT mRNA or protein
reversed motor deficits [14]. This study identified sCAG as a
disease causing agent. Since sCAGs, isolated from HD human brains,
when transfected reduced viability of neurons [6], these sequences
might affect cell viability through RNAi by targeting genes that
regulate cell survival.
[0145] We recently reported that siRNAs and shRNAs derived from
CD95, CD95L [15], and other genes in the human genome [16] kill
cancer cells through RNAi by targeting a network of critical
survival genes [15]. DISE (death induced by survival gene
elimination) was found to involve simultaneous activation of
multiple cell death pathways, and cancer cells have a hard time
developing resistance to this form of cell death [17]. DISE was
found to preferentially affect transformed cells [17]. Because the
length of the CAG repeats in different CAG repeat diseases has
been inversely correlated with cancer incidence in various organs
[18-21], we were wondering whether RNAi active CAG based TNRs
might be responsible for this phenomenon and whether they could be
used to kill cancer cells.
[0146] We have now identified an entire family of TNR-based
siRNAs—which contains the CAG repeat that causes HD—to be at least
10 times more toxic to cancer cells than any tested DISE-inducing
si/shRNA. Our data suggest this super toxicity is caused by
targeting multiple complementary TNR expansions present in the
open reading frames (ORFs) of multiple genes, rather than in their
3′UTRs. As a proof of concept, we demonstrate that siCAG/CUG can
be safely administered to mice to slow down growth of xenografted
ovarian cancer cells with no obvious toxicity to the animals. We
are proposing to develop super toxic TNR expansion-based siRNAs
for cancer treatment.
[0147] Results
[0148] siCAG/CUG kills all cancer cells in vitro. CAG repeats are
the defining factor in Huntington's disease, and their complement
CTG is amplified in myotonic dystrophy type 1 (DM1) [1]. We were
interested in determining whether a 19 mer duplex of CAG and CUG
repeats (siCAG/CUG) (FIG. 1A) would affect the growth of cancer
cells. When transfecting siCAG/CUG into various human (FIG. 1B)
and mouse (FIG. 1C) cancer cell lines at 10 nM, all cancer cells
stopped growing within hours of transfection and eventually most
of the cells died with no outgrowth of recovering cells (FIG. 6A).
All cancer cells transfected with siCAG/CUG showed morphological
changes similar to the ones we observed in cells undergoing DISE
(FIG. 6B, [15, 17]). We found that siCAG/CUG killed HCT116 cells
even when transfected at 10 pM (FIG. 1D). Compared to any other
si- or shRNA we have tested siCAG/CUG is ̃10-100 times more toxic
depending on the assay used. When monitoring cell viability (ATP
content), the IC50 for siL3, the most toxic DISE inducing siRNA we
have used, was determined to be 0.8 nM and for siCAG/CUG was 0.039
nM (FIG. 1E).
[0149] Identification of the most toxic TNR based siRNAs. The
siCAG/CUG repeat 19 mer in all three frames showed roughly the
same level of toxicity when transfected into HeyA8 cells (FIG. 7).
To test whether other TNR disease-derived sequences were toxic to
cancer cells when introduced as siRNAs, the repeats siGAA/UUC (GAA
is amplified in Friedreich's ataxia [22]), siCGG/CCG (CGG found in
fragile X tremor ataxia syndrome [FXTAS] and CCG found in Fragile
XE mental retardation [FRAXE] [1]) were transfected into HeyA8
(ovarian) and A549 (lung) cancer cells (FIG. 2A). In addition,
siCGA/UCG was tested because it has the same base composition as
the super toxic siCAG/CUG TNR. Interestingly, among the four
tested TNR siRNA duplexes two were super toxic to both cell lines,
and two showed no toxicity. Most remarkable was the observation
that siCGA/UCG was among the nontoxic repeats. This finding
pointed at a sequence specific mechanism behind this phenomenon
rather than a response of the cells to dsRNA of a specific base
composition.
[0150] To identify the most toxic TNR sequences in an unbiased
screen, we designed a library of 19 mer siRNAs based on the 60
possible TNRs (that contain more than one type of nucleotide). To
reduce passenger strand loading and determine the toxicity of each
repeat when loaded into the RISC as a guide strand, we replaced
positions 1 and 2 of the passenger strand with 2′-O-methylated
(OMe) nucleotides. To confirm the effect of the OMe modification,
we modified the toxic CD95L-derived siRNA siL3 in this way. While
the siL3 duplex modified on the intended passenger strand (S-OMe)
was slightly more toxic to cells than unmodified siL3, likely
reflecting a low level of passenger strand loading of siL3,
neither siL3 modified on the antisense strand (AS-OMe) nor on both
strands (S/AS-OMe) showed any toxicity (FIG. 8).
[0151] All 60 TNRs were now synthesized with the sense strand
carrying the OMe modification in positions 1 and 2, allowing us to
determine the toxicity of each of the 60 antisense sequences.
HeyA8 cells were transfected with 1 nM of each of the 60 TNRs and
viability was quantified 96 hrs after transfection. The 60 TNRs
can be grouped into 10 families [23]. Each family is comprised of
3 triplets shifted by one nucleotide plus its three complementary
triplets. In total 30 (50%) of the TNRs were not toxic, 11 (18%)
were moderately toxic (>50% loss of viability, shown in
yellow), and 19 (31.7%) were super toxic (>75% loss of
viability, shown in red) to HeyA8 cells (FIG. 2B, top panels).
Among the nontoxic TNRs were all 6 members of family 3 (0% GC
content) and all 6 members of family 10 (100% GC content). All
other TNR families contained nontoxic and toxic TNRs.
Interestingly, in some cases just shifting the TNR sequence in the
19 mer by one nucleotide resulted in opposite effects on viability
(i.e. AGG and GGA in family 8). In other cases, members of a
family showed toxicity of one strand but no toxicity of its
complement (i.e. families 1, 2, 4, 5, 6, 8 and 9). This finding
suggests a sequence-specific and in some cases frame specific
activity of the TNRs consistent with RNAi being involved. Due to
the different base composition of targeted RNAs, the comparison of
TNR families with the same GC content and base composition is most
meaningful. Two families contain a balanced GC content of 66.7%
and identical base composition: family 5 and 7. Remarkably, while
family 5 contained toxic and nontoxic members, all six TNRs in
family 7 were super toxic (boxed in red in FIG. 2B). Family 7
stands out as it contains all permutations of both the CAG and the
CUG repeats we identified as killing all cancer cells.
[0152] To determine how much of these activities were conserved
between human and mouse cancer cells, the screen was repeated with
the mouse liver cancer cell line M565 (FIG. 2B, bottom panels).
The results for the siRNAs in TNR families 1, 2, 4, 5, 8, and 9
were somewhat similar to the ones obtained with the human cell
line, but also showed clear differences. This could be due to
differences in tissue origin, cell line, or species between the
two cell lines. Three of the TNR families performed in an
identical fashion between the two cell lines. Similar to HeyA8
cells, none of the 12 TNR-derived siRNAs in families 3 or 10
showed any toxicity in M565 cells. Most strikingly however, was
the finding that again all 6 members of family 7, which contain
both the CAG and the CUG repeat, were super toxic to the mouse
cell line. Screens in both HeyA8 and M565 cells were repeated and
results showed a high degree of congruence, especially in the
results of family 7 (FIG. 9). When the average of the screen in
HeyA8 cells was plotted against the averages of the two screens in
M565 cells, a significant correlation between the screens was
found (FIG. 2C) and again the six TNRs in family 7 were most
consistently toxic. The data suggests the toxicity of this TNR
family is conserved and it is independent of tissue, cell line and
species.
[0153] We recently reported that the 6 mer seed sequence of siL3
was the main determinant of its toxicity [15]. We therefore
wondered how much of the toxicity of the super toxic TNRs was due
to complete complementarity of the siRNA and how much was
dependent on just the 6 mer seed sequence. The data on siL3 were
obtained by generating chimeric siRNA duplexes between a nontoxic
control siRNA (siNT) and siL3 by replacing siL3 sequences from
either end of the duplex with siNT sequences [15]. To generate an
artificial nontoxic siRNA backbone in which to test all 60 TNR 6
mer seed sequences, we first replaced 4 positions in the center of
siNT still identical to the same position in the siL3 sequence
with the complementary nucleotides, thereby removing any identity
between siNT and siL3 outside the seed, while maintaining GC
content (FIG. 10A). This siL3 seed siRNA (siL3 seed) was almost as
toxic to HeyA8 cells as siL3, confirming that the 6 mer seed
determined a substantial part of the toxicity of siL3. We
therefore used the modified siNT backbone to test all possible
TNR-derived 6 mer seed sequences (FIG. 2D, FIG. 10B). While some
TNR derived seeds were toxic to HeyA8 cells, there was only a
moderate level of congruence between the screen with the entire
TNR 19 mers and one just with the 6 mers in the modified siNT
backbone (FIG. 2E). Of the 6 super toxic TNRs in family 7 only one
was also toxic in the 6 mer screen (FIG. 2D and 2E).
Interestingly, most of the 6 mers in family 10 were toxic although
no toxicity was observed in the TNR screen (FIG. 2B). We interpret
this as the inability of these 6 TNRs with their 100% GC content
to properly enter the RISC. Together these data suggest 19 mer TNR
siRNAs are toxic to cancer cells by a mechanism distinct from the
process of DISE which relies on just the seed sequences targeting
the 3′UTRs of survival genes [15].
[0154] Super toxic TNR-based siRNAs kill cancer cells through RNAi
resulting in the loss of survival genes. To address the question
whether the super toxic TNR-based siRNAs killed cancer cells
through RNAi, we first compared the toxicity of siCAG/CUG in
HCT116 wild-type and HCT116 Drosha<−/−> cells. DISE inducing
si- and shRNAs kill Drosha<−/−> cells more efficiently than
wild-type cells [15]. We had interpreted this as the RISC being
more available in the absence of most cellular miRNAs, which rely
on Drosha for processing. While siCAG/CUG was highly toxic to both
cell lines at early time points, Drosha <−/−> cells were
more sensitive to growth reduction induced by siCAG than their
wild-type counterparts (FIG. 11, p=0.038, according to polynomial
fitting model). To directly test the requirement of AGO2 in the
siCAG/CUG induced toxicity we knocked down AGO2 in both HeyA8 and
A549 cells (FIG. 3A) and transfected the cells with either siNT or
siCAG/CUG (FIG. 3B). Removal of AGO2 from the cells almost
completely prevented the toxicity of siCAG/CUG confirming
dependence on the RISC. A dependence on Ago2 for siCAG/CUG
toxicity was confirmed in Ago1-4 knock-out mouse embryonic
fibroblasts with re-expressed AGO2 (FIG. 12). These data indicated
that siCAG/CUG was negatively affecting cells through canonical
RNAi involving the RISC complex. To confirm this, we modified the
siCAG siRNAs with the 2′-O-methylation to selectively block
loading of either the siCAG or the siCUG based strand into the
RISC (FIG. 3C). When the CAG-based guide strand was modified
(siCAG AS-OMe), the toxicity of the siCAG/CUG duplex was severely
reduced. It was not affected when the CUG repeat containing strand
was 2′-O-methylated (siCAG 5-OMe), confirming that most of the
toxicity of the siCAG/CUG repeat comes from the CAG repeat strand.
siCAG/CUG did not have any toxicity when both strands were
modified indicating most, if not all, of its toxicity requires
RISC loading confirming that RNAi was responsible for cell death.
[0155] We recently reported that DISE-inducing CD95L derived sh-
and siRNAs kill cancer cells by targeting the 3′UTR of critical
survival genes through canonical RNAi [15]. To test whether the
super toxic siCAG/CUG duplex also killed cancer cells through this
mechanism, we transfected HeyA8 cells with siNT, siCAG/CUG or the
nontoxic siCGA/UCG, and subjected the RNA 48 hours after
transfection to a RNA-Seq analysis. Interestingly, in the cells
transfected with siCAG/CUG, 3466 genes were down and 867 genes
were upregulated (>1.5 fold, adjusted p value<0.05) (data
not shown). A DAVID gene ontology analysis of the upregulated
genes did not reveal any evidence of an interferon response by the
cells induced by the transfected siRNA (data not shown). In cells
transfected with the nontoxic siCGA/UCG, only 194 genes were found
to be downregulated and 420 genes upregulated.
[0156] We performed a gene set enrichment analysis for a group of
̃1800 survival genes and ̃400 nonsurvival genes that were
identified in a genome-wide CRISPR lethality screen [45] after
transfecting cells with either siCAG/CUG or siCGA/UCG using a.
non-targeting siNT served as a control. Similar to cells
undergoing DISE, when transfected with siCAG/CUG, the ̃1800
critical survival genes but not the ̃400 nonsurvival control genes
[15] were significantly enriched in the downregulated genes in
cells transfected with siCAG/CUG but not in cells transfected with
siCGA/UCG (data not shown). In fact, we detected a ̃12-fold
increased percentage of survival genes compared to the
non-survival genes among the downregulated RNAs in the siCAG/CUG
treated cells (FIG. 3D)—a higher difference than seen in cells
treated with DISE-inducing sh- or siRNAs (data not shown).
[0157] We also performed a metascape analysis of 4 RNA-Seq data
sets of cells into which were introduced siL3, two CD95L derived
shRNAs (shL1 and shL3), a CD95 derived shRNA (shR6) ORF,
previously described [15], and the downregulated genes in cells
transfected with siCAG/CUG. The Metascape gene ontology analysis
comparing the downregulated genes in cells treated with either
CD95 or CD95L derived si- or shRNAs with the data from the
siCAG/CUG treated cells showed a strong overlap in the GO terms
including cell cycle, response to DNA damage, mitosis, and
chromatin organization suggesting that cells died through a
mechanism similar to DISE (data not shown). When the RNA-Seq data
of cells treated with siCGA/UCG was included in the analysis, not
a single GO cluster overlapped (data not shown).
[0158] Interestingly, a large genome-wide comparison of
lymphoblastoid cell lines from 107 HD patients reported an inverse
correlation between CAG repeat length and downregulated genes.
Biological pathways that were significantly affected were
ribosomal process, energy metabolism and cell death pathways [24]
all consistent with reduced cell viability. We compared the genes
that were reported to be negatively and significantly correlated
with the length of the CAG repeats in these patients (1236 genes,
according to Pearson correlation) with the 3466 genes
downregulated in the HeyA8 cells transfected with siCAG/CUG. Of
the 1236 genes downregulated in the patients 182 (14.7%) were also
downregulated in the siCAG/CUG treated HeyA8 cells (FIG. 3E). In a
DAVID gene ontology analysis with these 182 genes the two most
significantly enriched clusters were consistent with genes playing
a role in cell division and mitosis, consistent with a major
effect of siCAG/CUG on mitosis (data not shown). In summary, these
data suggest that the toxicity of the CAG repeat based siRNA may
involve loss of survival genes and that this form of cell death
could be related to the TNR activities seen in patients with
extended CAG repeats.
[0159] Super toxic TNR-derived siRNAs kill cells by targeting TNR
sequences present in the ORF of genes complementary to the toxic
siRNA guide strand. To determine which genes and what part of the
mRNAs could be targeted by toxic TNR-derived siRNAs, we subjected
ranked lists of downregulated genes of cells treated with either
siCAG/CUG or siCGA/UCG to a Sylamer analysis [25]. This method
detects enrichment of seed matches in mRNAs that are complementary
to the seed of the introduced siRNAs. In particular, we performed
a 6-nucleotide Sylamer analysis of the ORFs and the 3′UTRs of
genes deregulated in HeyA8 cells transfected with either siCAG/CUG
or siCGA/UCG and ordered the genes from most down-regulated to
most up-regulated.
[0160] When the seed length was set to 6 nts, we detected a minor
enrichment of the 6 mer TGCTGC in the 3′UTRs of the downregulated
genes in the cells treated with siCAG/CUG. TGCTGC is the expected
seed match (position 2-7) of the siCAG 19 mer guide strand (data
not shown). No significant seed match enrichment was found in
cells treated with siCGA/UCG or when the ORFs of these genes were
analyzed.
[0161] We also performed a 10-nucleotide Sylamer analysis of the
ORFs and 3′UTRs of genes deregulated in HeyA8 cells transfected
with either siCAG/CUG or siCGA/UCG and ordered the genes from most
down-regulated to most up-regulated. The three most highly
enriched sequences were as follows:
ORF 10-mers, siCAG/CUG:
CTGCTGCTGC, (SEQ ID NO: 9)
TCTGAGACCA, (SEQ ID NO: 10)
TGCTGCTGCT (SEQ ID NO: 11)
ORF 10-mers, siCGA/UCG:
GGGGGTGGGG, (SEQ ID NO: 12)
CCTCCCTCCC, (SEQ ID NO: 13)
CCCCGCCCCC (SEQ ID NO: 14)
3′UTR 10-mers, siCAG/CUG:
GGCCCTGGCC, (SEQ ID NO: 15
CACTCCCCAC, (SEQ ID NO: 16)
GGCAGGGGTG (SEQ ID NO: 17)
3′UTR 10-mers, siCGA/UCG:
GGGGGTGGGG, (SEQ ID NO: 18)
CCTCCCTCCC, (SEQ ID NO: 19)
CCCCGCCCCC (SEQ ID NO: 20)
[0162] When we analyzed the ORFs of cells treated with siCAG/CUG
when setting the seed length to the maximum of 10 nts, we found a
very profound enrichment of two 10 nt sequences (p-value
̃10<−50>) that corresponded to positions 1-10 and 2-11,
respectively of the targeting siCAG 19 mer, which were CTGCTGCTGC
(SEQ ID NO:9) and TGCTGCTGCT (SEQ ID NO:11). No such enrichment
was found when the 3′UTRs of the genes were used for the analysis.
These data suggest that in contrast to DISE-inducing si/shRNAs,
siCAG/CUG killed cancer cells by targeting long repeat sequences
located mainly in ORFs. Consistent with this conclusion, genes
containing either of the two targeted lOmers in their ORFs were
very strongly enriched among the downregulated genes in siCAG/CUG
treated cells, while only a weak enrichment was found when the
3′UTRs were analyzed.
[0163] Now knowing that the toxicity of the siCAG/CUG correlated
with the presence of targeted trinucleotide repeats found in the
ORFs of genes, we wondered whether the toxicity across all 60 TNR
derived siRNAs correlated with the presence of higher ordered
reverse complementary TNRs in ORFs that could be targeted by the
TNR siRNAs and whether this was conserved between human and mouse
cells. We analyzed the frequencies of all triplets combined in
mouse and human ORFs and 3′UTRs and counted the number of genes
containing 6 mers, 10 mers, or 19 mers targeted by the 60 TNRs. We
also performed a set enrichment analysis for genes containing a
GCTGCTGCTGCTGCTGCTG (SEQ ID NO:21) 19 mer (targeted by the siCAG
19 mer) in their ORFs and 3′UTRs in cells transfected with
siCAG/CUG when compared to cells transfected with siNT. When
counting all 60 single triplets in both the ORFs and 3′UTRs of all
human and mouse genes they were found to be slightly more abundant
in the ORF of genes. When separating into individual triplets of
the 10 families most triplets are found at similar frequencies in
both the ORFs and 3′UTRs in humans and in mice. This situation
changed when we focused on targeted (reverse complements of the
targeting TNRs) repeats. We plotted the results for 6 mers (as
this is the minimum sequence required for RNAi-based targeting),
10 mers (the maximum seed length allowed by Sylamer), and 19 mers
(the length of the siRNAs used). In all cases, with longer
sequences, the preference for certain triplets became clearer.
When analyzing the 19 mers, in both human and mouse, the most
abundant TNRs are members of the super toxic family 7 and barely
any 19 mers were found in the related family 5. Genes containing
the 19 mer TNR (or longer) in their ORF targeted by siCAG were
enriched in the most downregulated genes in cells transfected with
siCAG/CUG. We selected 5 of the 6 most highly expressed and
downregulated genes from the RNA-Seq analysis for validation (FIG.
13). HeyA8 cells were transfected with siRNAs at 1 nM, and the
mRNAs levels were quantified by real time PCR 10, 20 or 40 hrs
after transfection. When transfecting siCAG/CUG all 5 CUG
repeat-containing mRNAs were downregulated as early as 10 hrs with
maximal downregulation at 40 hrs. Specificity of the targeting was
established by transfecting the cells with either siCAG or siCUG
(in which the passenger strand was disabled by adding the
2′-O-methylation). Only the siCAG-based siRNAs were active in
silencing the CUG TNR containing genes. These data strongly
suggested that the toxicity of TNR-based siRNAs in general might
be explained by the presence of extended reverse complementary
repeat sequences present preferentially in the ORF of targeted
genes. This was confirmed by plotting the viability of cells
treated with any of the 60 TNR siRNAs and the number of targeted
TNR sequences of 6 nts or longer (FIG. 4A). The highest and most
significant correlation was found for both human and mouse ORFs.
Remarkably, while 80-90% of triplets targeted by the 6 members of
the CGA containing family 5 are present as singular events (blue
dots), between 50 and 70% of the triplets targeted by members of
the CAG containing family 7 are found to be part of 6 mers or
higher ordered TNRs (red dots). Interestingly, the TNRs targeted
by the 6 toxic members of family 7 code for 5 different amino
acids (FIG. 4B). As all the 6 members of family 7 are equally
toxic to cancer cells, this suggests that this involves targeting
long repeat elements (i.e. TNRs), rather than a requirement for
poly-homo amino acid coding stretches. This view is supported by
an analysis of species conservation (FIG. 14). Of the genes that
contain targeted 19 mers, only 7 of the 99 genes found in either
the mouse and the human genome overlapped and the genes that are
targeted do not have shared functions (FIG. 14B). In summary, our
data provide evidence that TNR-based siRNAs are toxic to cells by
targeting a number of genes that contain high order trinucleotide
repeats that are reverse complementary to the targeting TNR. The
resulting cell death has features of what we recently described as
DISE, with the main difference that DISE is the result of a
miRNA-like targeting of short seed matches in the 3′UTRs of
survival genes, whereas the TNR-induced cell death is an on-target
effect affecting a larger number of genes that contain targeted
sequences in their ORF. This now provides an explanation for why
the most toxic TNR-based siRNAs are much more toxic than
DISE-inducing siRNAs. Intriguingly, the CAG repeats found in HD
are part of the most toxic family of TNRs and their reverse
complementary 19 mers that can serve as targets are the most
abundant TNR sequences in the ORFs of both human and mouse
genomes.
[0164] Super toxic CAG/CUG TNR based siRNAs slow down tumor growth
in vivo with no toxicity to normal tissues. We were wondering
whether the super toxic TNR-based siRNAs could be used for cancer
therapy. We decided to deliver the siRNAs to cancer cells in vivo
using templated lipoprotein (TLP) nanoparticles [26]. Before using
the TLP particles loaded with the siCAG/CUG duplex (siCAG/CUG-TLP)
in vivo, we tested their effects on tumor cells in vitro. They
killed HeyA8 cells more efficiently than siL3-TLPs (FIG. 5A) and
also slowed down growth of the tested human or mouse cancer cell
lines (FIG. 5B and data not shown). They also killed neurospheres
derived from patients with glioblastoma (FIG. 5C). To test the
activity of the siCAG/CUG-TLPs in vivo we treated orthotopically
xenografted HeyA8 ovarian cancer cells in mice. Mice injected with
100,000 tumor cells were i.p. injected with nano particles 5 times
a week for two weeks (FIG. 5D). After the tenth treatment mice
were split into two groups, one group continued to receive
treatment in the third week and the second group did not. This was
done to determine whether large established tumors would still
respond to the treatment. The large tumors in treatment group 1
still benefited from the effect of siCAG/CUG in the third week of
treatment (FIG. 5E, left panel). In contrast, some tumors in the
mice in treatment group 2 grew out rapidly, while others showed
persisting growth reduction (FIG. 5E, right panel). These results
suggest that established tumors respond to the siCAG/CUG
treatment. This was confirmed in another experiment in which
10<6 >HeyA8 cells were injected and mice were first treated
3 times a week and then switched to daily treatment 19 days after
tumor cell injection (data not shown).
[0165] To determine whether siCAG/CUG was detrimental to mice,
mice in treatment group 1 were treated a few more times with the
siRNA and were analyzed just when the control treated mice were
moribund at around day 27. We did not see any signs of toxicity in
any of the mice. They were feeding well (not shown), did not lose
weight (FIG. 15A), had normal liver histology (FIG. 15B), and
showed no increase in liver enzymes in the serum (FIG. 15C). These
data demonstrated that super toxic CAG/CUG TNR-based siRNAs
delayed tumor growth in vivo 5-6 days with no gross toxicity to
normal cells and that they could be safely administered using TLP
nanoparticles.
[0166] To determine whether tumor cells acquired resistance to the
treatment, we tested tumors from three mice in treatment group 1
ex vivo. Three tumors of mice treated with siNT-TLP and three
tumors of mice treated with siCAG/CUG-TLP were transfected with
the same siRNAs in vitro a day after tumor isolation (FIG. 5F). In
parallel, these tumor cells were also treated with the
nanoparticles again (FIG. 5G). In all cases, the tumors from the
mice that had received the toxic siRNA were as sensitive to the
toxic effects of siCAG/CUG in vitro as the tumors from mice
treated with siNT suggesting that cancer cells cannot become
resistant to cell death induced by the toxic TNR-based siRNAs, at
least not in the classical sense observed after targeted therapy,
and that preferentially tumor cells were responding as there were
no signs of toxicity in the mice.
[0167] Discussion
[0168] The siRNA we used for cancer treatment was a duplex between
the basic TNR module found in HD (CAG) and the fully complementary
strand CUG found in Myotonic Dystrophy (DM1). In a screen of all
TNR-derived siRNAs both CAG and CUG were part of a family that
contains 6 members all of which were highly toxic to cancer cells
of both human and mouse origin. This hybrid duplex between the two
disease molecules was also recently tested in a well-established
Drosophila model of DM1 [27]. Expression of the two transcripts
led to the generation of Dicer-2 (dcr-2) and ago2-dependent 21-nt
TNR-derived siRNAs resulting in high toxicity to the cells. In a
separate study, it was shown that expression of these
complementary repeat RNAs leads to dcr-2-dependent
neurodegeneration [28]. These results suggest that co-expression
of CAG and CUG repeat-derived sequences may dramatically enhance
toxicity in human repeat expansion diseases in which anti-sense
transcription occurs. Antisense transcription was reported to
occur in SCA8 in two genes encompassing the repeats are expressed:
ATXN8 (CAG repeat), on the sense strand and ATXN8OS (CTG repeat)
on the antisense strand [29].
[0169] One could argue that a cancer therapy based on delivering
siCAG/CUG could be detrimental to patients as TNR expansion
patients suffer from various pathologies. However, similar to many
other genes with amplified CAG repeats, HTT is ubiquitously
expressed throughout the body with somewhat higher expression in
the brain and in testis [30]. The disease is characterized by
neurodegeneration affecting the cerebral cortex and neuropathology
in the striatum, but it also affects other tissues [31]. So, if
sCAGs are produced in multiple tissues the effects on most normal
tissues seems to be moderate. Even in the brain, while detrimental
to HD patients long term, most patients do not have major symptoms
before the age of 40 [4, 31]. Short-term exposure to toxic sCAGs
for cancer therapy, as suggested by our mouse experiments, may not
have a dramatic effect on normal tissues but may be enough to kill
cancer cells. If a CAG based siRNA were to produce side effects
particularly in the brain it may be possible to protect the brain
through local administration of neuroprotecting LNA-CTGs as
described [14].
[0170] What could be the mechanism of the relative resistance of
normal versus tumor cells to the toxic siRNAs? Our recent data
suggested that miRNAs inhibit DISE and in fact may protect normal
cells from it [15]. Both Drosha and Dicer k.o. HCT116 cells were
found to be hypersensitive to a DISE inducing siRNA. This is
entirely consistent with reported activities of CAG repeats which
may also act though RNAi. Pathogenic Ataxin-3 with amplified CAG
repeats showed strongly enhanced toxicity in HeLa cells after
knockdown of Dicer [32]. In addition, it was previously shown in
Drosophila that impairing miRNA processing dramatically enhanced
neurodegeneration caused by the CAG repeat gene Ataxin-3. Two fly
mutants were tested, one with a deficiency in dcr-1, the
Drosophila Dicer ortholog that is required for miRNA biogenesis,
and another with a deficiency in R3D1, a gene required for dcr-1
to function [33]. The authors concluded that “miRNA pathways
normally play a protective role in polyQ-induced
neurodegeneration”. In the light of our data it is possible that
miRNAs might actually protect cells from the toxic effects of TNR
based siRNAs.
[0171] While HD is the best known disease caused by CAG repeats,
one of the two diseases first discovered to be caused by TNR
repeat expansions is the neurodegenerative disorder SBMA/Kennedy
disease [34] wherein the pathogenic CAG repeat is found in exon 1b
of the androgen receptor (AR). If patients with amplified CAG
repeats produce sCAGs, which according to our work may be
detrimental to cancer cells, one would have to expect that such
patients have a reduced cancer incidence due to toxic siRNA
expression in many tissues. Indeed reduced cancer incidence was
reported for HD and SBMA patient populations in Sweden [18], and
HD patients in France [19], Denmark [20], and England [21].
[0172] Possibly the clearest connection to cancer has been
reported for the CAG repeats in the AR gene and prostate cancer
(PCa). The CAG repeat length in the AR has been inversely linked
to PCa. While longer repeats (>20 CAGs) confer a protective
effect among the PCa patients 45 years or older [35], shorter CAG
repeats have been shown to result in a two-fold increased cancer
risk [36], a more aggressive disease, and a high risk of distant
metastases [37-39]. Shortening of CAG repeat length was found in
in situ lesions of PCa and its possible precursors [40],
suggesting that PCa avoids longer CAG repeats. This is consistent
with our finding of super toxicity of CAG based siRNAs.
[0173] There are two observations that suggest the targeted TNRs
present in the ORFs of certain genes are not there because these
proteins require stretches of the same amino acid for their
function which would presumably be conserved between human and
mouse: first, all six members of the TNR family 7 were super toxic
targeting 6 different reverse complementary TNRs that code for 5
different amino acids (FIG. 4B), and second, the genes with the
longest repeats with complete complementarity to the most toxic
siCAG/CUG of 19 nts showed little overlap between human and mouse
(Fig S 14B) and the different targeted genes do not share similar
functions based on a Metascape analysis (data not shown).
Interestingly, both the AR and the HTT genes contain some of the
longest CAG repeats in the human genome, however, those are not
found in the mouse orthologs at the same positions in the ORF.
[0174] So, if there was no pressure to maintain these TNRs in
specific genes but rather anywhere in the genome, could there be
an evolutionary link between TNRs and cancer? A hint may come from
the way the repeat expansion are generated. It is believed that
among other mechanisms DNA replication slippage and/or defective
base excision repair causes expansion of TNRs [41]. Therefore CAG
repeats could be part of a mechanism used during evolution to
maintain genome integrity and, in the context of multicellular
organisms, to prevent cancer formation by producing toxic siRNAs.
This would occur whenever too many mutations start accumulating in
cells, one property all cancers have in common.
[0175] While the treatment with siCAG/CUG requires optimization,
our data on the toxicity of CAG TNR based siRNAs for cancer cells
but not normal cells when administered in vivo and the reported
decreased incidence rate for different types of cancer in patients
with CAG expansions suggest that TNR-based siRNAs may be useful
for cancer therapy.
[0176] Materials and Methods
[0177] Cell lines and tissue culture. All cells were grown in an
atmosphere of 5% carbon dioxide (CO2) at 37° C. Unless indicated
otherwise base media were supplemented with 10% heat-inactivated
fetal bovine serum (FBS; Sigma-Aldrich) and 1%
penicillin/streptomycin and L-Glutamine (Mediatech Inc.). Cells
were dissociated with 0.25% (w/v) Trypsin-0.53 mM EDTA solution
(Mediatech Inc.). The following cell lines were cultured in
supplemented RPMI1640 Medium (Mediatech Inc.): Ovarian cancer cell
lines HeyA8 (RRID:CVCL_8878), OVCAR3 and OVCAR4 (both from Tumor
Biology Core, Northwestern University), and lung cancer cells A549
(ATCC CRM-CCL-185) and H460 (ATCC HTB-177). The GBM cell line T98G
(ATCC CRL1690) was cultured in Eagle's Minimum Essential Medium
(EMEM) (ATCC). Melanoma B16F10 cells (ATCC CRL-6475) and 293T
cells (RRID:CVCL0063) were cultured in DMEM (Cellgro). HepG2 (ATCC
HB-80645) was cultured in EMEM (ATCC). IDB, a mouse ovarian cancer
cell line, was cultured in DMEM supplemented with 4% FBS, and 10
mg/1 Insulin, 5.5 mg/1 Transferrin, 6.7 μg/ml Selenium (ITS,
Mediatech, Inc., 1:10 diluted). and 3LL Lewis lung cancer cells
(ATCC CRL-1642) were cultured in DMEM. FOSE2 cells are
spontaneously immortalized ovarian surface epithelial cells, and
M565 cells are from a spontaneously formed liver cancer in a
female and male mouse, respectively, both isolated from mice
carrying a floxed Fas allele [17]. Both were cultured in DMEM/F12
(Gibco #11330), 1% ITS (Insulin-Transferrin-Selenium Gibco
51300-044). M565 cells were dissociated with Accutase detachment
reagent (Fisher Sci.). HCT116 Drosha<−/−> were generated by
Narry Kim [42]. HCT116 parental (cat#HC19023, RRID:CVCL_0291) and
the Drosha<−/−> clone (clone #40, cat#HC19020) were
purchased from Korean Collection for Type Cultures (KCTC). All
HCT116 cells were cultured in McCoy's medium (ATCC, cat#30-2007).
Mouse Ago1-4 k.o. embryonic stem cells inducibly expressing human
FLAG-HA-AGO2 were described in [43]. CELLSTAR tissue culture
dishes (Greiner Bio-One, cat#664160, cat#639160) were coated with
0.1% Gelatin solution (Sigma, cat#ES-006-B) for 10 to 30 minutes
before use. Cells were cultured in DMEM media (Gibco,
cat#12430054) supplemented with 15% Fetal Bovine Serum (Sigma,
cat#F2442), 1% NEAA solution (HyClone, cat#SH3023801), 1% GlutaMAX
100× (Gibco, cat#35050061), 0.0007% 2-Mercaptoethanol (Fisher,
cat#BP176100), and 10<6 >units/L LIF (Sigma, #ESG1107). The
cell culture media was refreshed daily. FLAG-HA-AGO2 is under the
control of a TRE-Tight (TT) doxycycline (Dox)-inducible promoter.
100 ng/ml doxycycline (Sigma, cat#D9891) was added to the media in
order to induce moderate level of Ago2 expression to maintain
normal cell growth. To deplete Ago2 expression in cells,
doxycycline was withdrawn from media for 4 days. To induce wild
type level of hAgo2 expression, 2.5 μg/ml doxycycline was added to
the media. The human GBM derived neurosphere cell line GIC-20
(infected with pLV-Tomato-IRES-Luciferase) was obtained from Dr.
Alexander Stegh. Cells were grown as neurospheres in DMEM/F12
50:50 with L-glutamine (Corning), supplemented with 1% PenStrep,
B27 (Invitrogen), N2 (Invitrogen), human-Epidermal Growth Factor
(hEGF; Shenandoah Biotech), Fibroblast Growth Factor (FGF;
Shenandoah Biotech), Leukemia Inhibitory Factor (LIF; Shenandoah
Biotech), and GlutaMAX (Life Technologies). HeyA8 xenografted
tumors nodules were dissected from mice, cut, washed in sterile
PBS, and dissociated with dissociated with 0.25% (w/v)
Trypsin-0.53 mM EDTA solution for 20 minutes at 37° C. The
digestions was stopped by adding full RPMI-1640 medium. After
centrifugation, the trypsin solution mix was removed, and the
tumor cells were resuspended in fresh full medium, and strained
through 70 micron cell strainer. The tumor cell suspension was
plated over night on 10 cm tissue culture dishes. The following
day, cells were harvested, counted, and plated on 96-well plates
for further experiments. Cells were transfected with siRNAs after
cells had adhered or incubated with siRNA TLPs and then plated.
[0178] Western blot analysis. Primary antibodies for Western blot:
anti-β-actin antibody (Santa Cruz #sc-47778, RRID:AB_626632),
anti-human AGO2 (Abcam #AB186733, RRID:AB_2713978). Secondary
antibodies for Western blot: Goat anti-rabbit; IgG-HRP (Southern
Biotech #SB-4030-05, RRID:AB_2687483). Reagents used: propidium
iodide (Sigma-Aldrich #P4864), puromycin (Sigma-Aldrich #P9620)
and Lipofectamine RNAiMAX (ThermoFisher Scientific #13778150).
Western blot analysis was performed as recently described [15]
[0179] Transfection with short oligonucleotides. For transfection
of cancer cells with siRNAs, RNAiMAX was used at a concentration
optimized for each cell line, following the instructions of the
vendor. Cell lines were either transfected after cells had adhered
(forward transfection), or during plating (reverse transfection).
For an IncuCyte experiment cells were typically plated in 200 μl
antibiotic free medium, and 50 μl transfection mix with RNAiMAX
and siRNAs were added. During growth curve acquisitions the medium
was not exchanged to avoid perturbations. All individual siRNA
oligonucleotides were ordered from Integrated DNA Technologies
(IDT). Individual RNA oligos were ordered for the sense and
antisense oligo; the sense strand had 2 Ts added to the 3′ end;
antisense strand had 2 deoxy As at the 3′ end. When indicated the
first two positions at the 5′-end were 2′-O-methylated. Sense and
antisense oligos were mixed with nuclease free Duplex buffer (IDT,
Cat.No# 11-01-03-01; 100 mM Potassium Acetate, 30 mM HEPES, pH
7.5) to 20 μM (working solution), heated up for 2 minutes at 94°
C., and then the oligos were allowed to cool down to room
temperature for 30 minutes. siRNA solutions were aliquoted and
stored at −80° C. The cells were transfected with siRNAs at a
final concentration of 0.01 nM -10 nM. The following siRNA
sequences were used: siNT (siNT#2): UGGUUUACAUGUUGUGUGA (SEQ ID
NO:1) (non targeting in mammalian cells), siNT1:
UGGUUUACAUGUCGACUAA (SEQ ID NO:2) (non targeting in mammalian
cells), siL3: GCCCUUCAAUUACCCAUAU (SEQ ID NO:3) (human CD95L exon
1), siNT/siL3: UGGUUUACAUGUCCCAUAA (SEQ ID NO:4). siNT seed:
UGGUAAACUAGUUGUCUGA (SEQ ID NO:5), siL3 seed: UGGUAAACUAGUCCCAUAA
(SEQ ID NO:6). All TNR based 19 mer siRNAs were designed as
follows: The TNR based siRNA was named according to its
antisense/guide strand: 2 nt 3′ overhangs were added as described
above. All TNR based siRNAs were fully complementary 19 mers. For
instance, the siCAG/CUG sequences are: S: CAGCAGCAGCAGCAGCAGCdAdA
(SEQ ID NO:7), AS: GCUGCUGCUGCUGCUGCUGTT (SEQ ID NO:8). In all
siRNAs used in screens the sense/passenger strand was disabled by
2′-O-methylation in positions 1 and 2 of the sense strand.
[0180] For transfecting Ago1-4 k.o. mouse ESC, cells were cultured
without doxycycline for three days. Half of the cells were then
cultured for one more day without doxycycline before transfection.
The other half was cultured in media containing 2.5 μg/ml of
doxycycline for one day to induce WT level AGO2 expression before
transfection. For IncuCyte experiments, The ESCs were transfected
with either 5 nM siNT or siCAG/CUG using reverse transfection
method in a 96 well plate coated with 0.1% gelatin. 5000
cells/well and 0.2 μl RNAiMAX/well were used. One day after
transfection, 100 μl of media (with or without 2.5 μg/ml
doxycycline) was added to each well. After that, media was
refreshed every 2-3 days until the cells grew confluent. For flow
cytometry experiments, cells were transfected with either 5 nM
unlabeled siNT or siNT labeled with Cy5 on the 5′ end of the
antisense strand using reverse transfection method in a 12 well
plate coated with gelatin in triplicates. 300,000 cells/well and 1
μl RNAiMAX/well were used. Flow cytometry measurements were
conducted 24 hours after transfection. For AGO2 knockdown
experiment, 100,000 cells/well HeyA8 or 200,000 cells/well A549
cells were reverse transfected in 6-well plate with either
non-targeting (Dharmacon, cat#D-001810-10-05) or an AGO2 targeting
siRNA SMARTpool (Dharmacon, cat#L004639-00-005) at 25 nM. 1 μl
RNAiMAX per well was used for HeyA8 cells and 6 μl RNAiMAX per
well was used for A549 cells. 24 hours after transfection with the
SMARTpools, cells were reversed transfected in 96-well plate with
either siNT or siCAG/CUG at 1 nM and monitored in the IncuCyte.
0.1 μl/well RNAiMAX was used for HeyA8 cells and 0.6 μl/well
RNAiMAX was used for A549 cells.
[0181] Total RNA isolation and RNA-seq analysis. HeyA8 cells were
transfected in 6-wells with siNT or either siCAG/CUG or siCGA/UCG
oligonucleotides at 1 nM. The transfection mix was removed after 9
hours. Total RNA was isolated 48 hours after transfection using
the miRNeasy Mini Kit (Qiagen, Cat.No. 74004) following the
manufacturers instructions. An on-column digestion step using the
RNAse-free DNAse Set (Qiagen, Cat.No.: 79254) was included. NGS
RNA-SEQ library construction and sequencing was performed by the
University of Chicago Genomics Facility. The quality and quantity
of RNA samples was assessed using an Agilent bio-analyzer. RNA-SEQ
libraries were generated using Illumina Stranded TotalRNA TruSeq
kits according to the Illumina provided protocol and sequencing
was performed using the Illumina HiSEQ4000 according to Illumina
provided protocols and reagents. The resulting paired end reads
were aligned to the hg38 assembly of the human genome with
Tophat2. HTseq was used to associate the aligned reads with genes,
and EdgeR was used to identify genes significantly differentially
expressed between treatments, all as recently described [15]. The
accession number for the RNA-Seq and expression data reported in
this work are GSE104552.
[0182] Real-time PCR. Real-time PCR was performed a described
recently [15] using the following primers: GAPDH (Hs00266705_g1),
RPL14 (Hs03004339_g1), LRRC59 (Hs00372611_m1), CNPY3
(Hs01047697_m1), CTSA (Hs00264902_m1), and LRP8 (Hs00182998_m1).
[0183] Monitoring growth over time and quantification of cell
death. To monitor cell growth over time, cells were seeded between
125 and 4000 per well in a 96-well plate in triplicates. The plate
was then scanned using the IncuCyte ZOOM live cell imaging system
(Essen BioScience). Images were captured at regular intervals, at
the indicated time points, using a 10× objective. Cell confluence
was calculated using the IncuCyte ZOOM software (version 2015A).
IC50 values for siL3 and siCAG/CUG were determined using GraphPad
Prism 6 software (by logarithm normalized sigmoidal dose curve
fitting). Quantification of DNA fragmentation (subG1 DNA) was done
as previously described [15].
[0184] siRNA screens and cell viability assay. HeyA8 or M565 cells
were expanded and frozen down at the same passage. One week before
transfection, cells were thawed and cultured in RPMI1640 medium,
10% FBS and 1% pen/strep. Cells were split three times during the
week and each time seeded at 4×10<6 >cells total in one T75
flask. On the day of the transfection, RNA duplexes were first
diluted with Opti-MEM to make 30 μl solution of 10 nM (for the
duplexes with the 6 mer seeds) or 1 nM (for the TNR-based
duplexes) as final concentration in a 384-well plate by Multidrop
Combi. Lipofectamine RNAiMAX (Invitrogen) was diluted in Opti-MEM
(6 μl lipid+994 μl of Opti-MEM for HeyA8 and 15.2 μl lipid+984.8
μl of Opti-MEM for M565 cells). After incubating at room
temperature for 5 to 10 minutes, 30 μl of the diluted lipid was
dispensed into each well of the plate that contains RNA duplexes.
The mixture was pipetted up and down three times by PerkinElmer
EP3, incubated at room temperature for at least 20 minutes, and
then the mixture was mixed again by PerkinElmer EP3. 15 μl of the
mixture was then transferred into wells of three new plates
(triplicates) using the PerkinElmer EP3. 50 μl with 320 HeyA8 or
820 M565 cells was then added to each well containing the duplex
and lipid mix, which results in a final volume of 65 μl. Plates
were left at room temperature for 30 minutes then moved to a 37°
C. incubator. 96 hours post transfection, cell viability was
assayed using CellTiter-Glo (Promega) quantifying cellular ATP
content. 35 μl medium was removed from each well and 30 μl
CellTiter-Glo cell viability reagent was added. The plates were
shaken for 5 minutes and incubated at room temperature for 15
minutes. Luminescence was then read on the BioTek Synergy NEO2.
[0185] Treatment of xenografted ovarian cancer cells in vivo with
templated lipoprotein particles (TLP) loaded with siRNAs.
Synthesis of TLPs and production of siRNA-TLPs was done exactly as
recently described [44]. 10<5 >HeyA8 cells (infected with a
luciferase lentivirus and a NucRed lentivirus (Essen Bioscience))
were injected i.p. into 6-week-old female NSG mice [44] following
the Northwestern University Institutional Animal Care and Use
Committee (IACUC)-approved protocol. The growth of tumor cells in
the mice over time was monitored non-invasively using the IVIS®
Spectrum in vivo imaging system as recently described [44]. Each
mouse of a treatment group was injected with 150 μl of either
siNT-TLP or siCAG/CUG-TLP (1 μM stock).
[0186] Data analyses. To determine the number of triplets, 6 mer,
10 mer or 19 mer repeat sequences in the ORFs or 3′UTR of human or
mouse genes, all ORF and 3′UTRs were extracted from the Homo
sapiens (GRCh38.p7) or Mus musculus (GRCm38.p5) gene dataset of
the Ensembl database using the Ensembl Biomart data mining tool.
For each gene, only the longest deposited ORF or 3′UTR was
considered. Custom perl scripts were used to identify whether each
3′UTR or ORF contained an identical match to a particular triplet,
6 mer, 10 mer or 19 mer.
[0187] GSEA was performed using the GSEA v2.2.4 software from the
Broad Institute (www.http://software.broadinstitute.org/gsea);
1000 permutations were used. Two list of 1846 survival and 418
nonsurvival genes were used as recently described [15, 45]. They
were set as custom gene sets to determine enrichment of survival
genes versus the nonsurvival control genes in downregulated genes
from the RNA-seq data. Log(Fold Change) was used as the ranking
metric. p-values below 0.05 were considered significantly
enriched. The GO enrichment analysis shown was performed using all
genes that after alignment and normalization were found to be at
least 1.5 fold downregulated with an adjusted p values of<0.05,
using the software available on www.Metascape.org and default
running parameters. The data sets of HeyA8 cells with introduced
siL3, shL1, shL3 or shR6 were recently described [15].
[0188] Sylamer analyses [25] were performed using the RNA-seq
datasets from the HeyA8 cells transfected with siNT, siCAG/CUG or
siCGA/UCG as recently described [15]. The analyses were performed
using default settings. Enriched 6 or lOmer motifs were analyzed
using either ORFs or the 3′UTRs sequences. Sylamer (version
12-342) was run with the Markov correction parameter set to 4.
DAVID gene ontology analysis was performed using the tool at
https://david.ncifcrf.gov/home.jsp and default settings.
[0189] Statistical analyses. Two-way analysis of variances (ANOVA)
were performed using the Stata 14 software to compare growth
curves. One-tail student t-test was performed in the software
package R to compare tumor load between treatment groups. Wilcoxon
Rank Sum test was performed in R to compare IVIS signal between
treatment groups. The effects of treatment on wild-type versus
Drosha<−/−> cells were statistically assessed by fitting
regression models that included linear and quadratic terms for
value over time, main effects for treatment and cell type, and
two- and three-way interactions for treatment, cell-type and time.
The three-way interaction on the polynomial terms with treatment
and cell type was evaluated for statistical significance since
this represents the difference in treatment effects over the
course of the experiment for the varying cell types. All
statistical analyses were conducted in Stata 14 (RRID:SCR_012763)
or R 3.3.1 in Rstudio (RRID:SCR_000432) except for Pearson
correlation analyses, which were performed using StatPlus 6.2.2.
REFERENCES
[0190] 1. Nalavade R, Griesche N, Ryan D P, Hildebrand S, Krauss S
(2013) Mechanisms of RNA-induced toxicity in CAG repeat disorders.
Cell Death Dis 4: e752
[0191] 2. Gatchel J R, Zoghbi H Y (2005) Diseases of unstable
repeat expansion: mechanisms and common principles. Nat Rev Genet
6: 743-55
[0192] 3. Ross C A (2002) Polyglutamine pathogenesis: emergence of
unifying mechanisms for Huntington's disease and related
disorders. Neuron 35: 819-22
[0193] 4. Orr H T, Zoghbi H Y (2007) Trinucleotide repeat
disorders. Annu Rev Neurosci 30: 575-621
[0194] 5. Cleary J D, Ranum L P (2017) New developments in RAN
translation: insights from multiple diseases. Curr Opin Genet Dev
44: 125-134
[0195] 6. Banez-Coronel M, Porta S, Kagerbauer B, Mateu-Huertas E,
Pantano L, Ferrer I, Guzman M, Estivill X, Marti E (2012) A
pathogenic mechanism in Huntington's disease involves small
CAG-repeated RNAs with neurotoxic activity. PLoS Genet 8: e1002481
[0196] 7. Napierala M, Krzyzosiak W J (1997) CUG repeats present
in myotonin kinase RNA form metastable “slippery” hairpins. J Biol
Chem 272: 31079-85
[0197] 8. Wojciechowska M, Krzyzosiak WJ (2011) Cellular toxicity
of expanded RNA repeats: focus on RNA foci. Hum Mol Genet 20:
3811-21
[0198] 9. Ho T H, Savkur R S, Poulos M G, Mancini M A, Swanson M
S, Cooper T A (2005) Colocalization of muscleblind with RNA foci
is separable from mis-regulation of alternative splicing in
myotonic dystrophy. J Cell Sci 118: 2923-33
[0199] 10. Lin L, Park J W, Ramachandran S, Zhang Y, Tseng Y T,
Shen S, Waldvogel H J, Curtis M A, Faull R L, Troncoso J C, et al.
(2016) Transcriptome sequencing reveals aberrant alternative
splicing in Huntington's disease. Hum Mol Genet 25: 3454-3466
[0200] 11. Li LB, Yu Z, Teng X, Bonini N M (2008) RNA toxicity is
a component of ataxin-3 degeneration in Drosophila. Nature 453:
1107-11
[0201] 12. Hsu R J, Hsiao K M, Lin M J, Li C Y, Wang LC, Chen L K,
Pan H (2011) Long tract of untranslated CAG repeats is deleterious
in transgenic mice. PLoS One 6: e16417
[0202] 13. Krol J, Fiszer A, Mykowska A, Sobczak K, de Mezer M,
Krzyzosiak W J (2007) Ribonuclease dicer cleaves triplet repeat
hairpins into shorter repeats that silence specific targets. Mol
Cell 25: 575-86
[0203] 14. Rue L, Banez-Coronel M, Creus-Muncunill J, Giralt A,
Alcala-Vida R, Mentxaka G, Kagerbauer B, Zomeno-Abellan M T,
Aranda Z, Venturi V, et al. (2016)
[0204] Targeting CAG repeat RNAs reduces Huntington's disease
phenotype independently of huntingtin levels. J Clin Invest 126:
4319-4330
[0205] 15. Putzbach W, Gao QQ, Patel M, van Dongen S,
Haluck-Kangas A, Sarshad A A, Bartom E, Kim K Y, Scholtens D M,
Hafner M, et al. (2017) Many si/shRNAs can kill cancer cells by
targeting multiple survival genes through an off-target mechanism.
eLife 6: e29702
[0206] 16. Patel M, Peter M E (2017) Identification of
DISE-inducing shRNAs by monitoring cellular responses. Cell Cycle:
doi: 10.1080/15384101.2017.1383576
[0207] 17. Hadji A, Ceppi P, Murmann A E, Brockway S, Pattanayak
A, Bhinder B, Hau A, De Chant S, Parimi V, Kolesza P, et al.
(2014) Death induced by CD95 or CD95 ligand elimination. Cell
Reports 10: 208-222
[0208] 18. Ji J, Sundquist K, Sundquist J (2012) Cancer incidence
in patients with polyglutamine diseases: a population-based study
in Sweden. Lancet Oncol 13: 642-8
[0209] 19. Coarelli G, Diallo A, Thion M S, Rinaldi D, Calvas F,
Boukbiza O L, Tataru A, Charles P, Tranchant C, Marelli C, et al.
(2017) Low cancer prevalence in polyglutamine expansion diseases.
Neurology 88: 1114-1119
[0210] 20. Sorensen S A, Fenger K, Olsen J H (1999) Significantly
lower incidence of cancer among patients with Huntington disease:
An apoptotic effect of an expanded polyglutamine tract? Cancer 86:
1342-6
[0211] 21. Turner M R, Goldacre R, Goldacre M J (2013) Reduced
cancer incidence in Huntington's disease: record linkage study
clue to an evolutionary trade-off? Clin Genet 83: 588-90
[0212] 22. Clark R M, Dalgliesh G L, Endres D, Gomez M, Taylor J,
Bidichandani S I (2004) Expansion of GAA triplet repeats in the
human genome: unique origin of the FRDA mutation at the center of
an Alu. Genomics 83: 373-83
[0213] 23. Kozlowski P, de Mezer M, Krzyzosiak W J (2010)
Trinucleotide repeats in human genome and exome. Nucleic Acids Res
38: 4027-39
[0214] 24. Lee J M, Galkina E I, Levantovsky R M, Fossale E, Anne
Anderson M, Gillis T, Srinidhi Mysore J, Coser K R, Shioda T,
Zhang B, et al. (2013) Dominant effects of the Huntington's
disease HTT CAG repeat length are captured in gene-expression data
sets by a continuous analysis mathematical modeling strategy. Hum
Mol Genet 22: 3227-38
[0215] 25. van Dongen S, Abreu-Goodger C, Enright A J (2008)
Detecting microRNA binding and siRNA off-target effects from
expression data. Nat Methods 5: 1023-5
[0216] 26. McMahon K M, Plebanek M P, Thaxton C S (2016)
Properties of native high-density lipoproteins inspire synthesis
of actively targeted in vivo siRNA delivery vehicles. Adv Funct
Mater 26: 7824-7835
[0217] 27. Yu Z, Teng X, Bonini N M (2011) Triplet repeat-derived
siRNAs enhance RNA-mediated toxicity in a Drosophila model for
myotonic dystrophy. PLoS Genet 7: e1001340
[0218] 28. Lawlor K T, O'Keefe L V, Samaraweera S E, van Eyk C L,
McLeod C J, Maloney C A, Dang T H, Suter C M, Richards R I (2011)
Double-stranded RNA is pathogenic in Drosophila models of expanded
repeat neurodegenerative diseases. Hum Mol Genet 20: 3757-68
[0219] 29. Moseley M L, Zu T, Ikeda Y, Gao W, Mosemiller A K,
Daughters R S, Chen G, Weatherspoon M R, Clark H B, Ebner T J, et
al. (2006) Bidirectional expression of CUG and CAG expansion
transcripts and intranuclear polyglutamine inclusions in
spinocerebellar ataxia type 8. Nat Genet 38: 758-69
[0220] 30. Sharp A H, Loev S J, Schilling G, Li S H, Li X J, Bao
J, Wagster M V, Kotzuk J A, Steiner J P, Lo A, et al. (1995)
Widespread expression of Huntington's disease gene (IT15) protein
product. Neuron 14: 1065-74
[0221] 31. Vonsattel J P, DiFiglia M (1998) Huntington disease. J
Neuropathol Exp Neurol 57: 369-84
[0222] 32. Bilen J, Liu N, Burnett B G, Pittman R N, Bonini N M
(2006) MicroRNA pathways modulate polyglutamine-induced
neurodegeneration. Mol Cell 24: 157-63
[0223] 33. Jiang F, Ye X, Liu X, Fincher L, McKearin D, Liu Q
(2005) Dicer-1 and R3D1-L catalyze microRNA maturation in
Drosophila. Genes Dev 19: 1674-9
[0224] 34. La Spada A R, Wilson E M, Lubahn D B, Harding A E,
Fischbeck K H (1991) Androgen receptor gene mutations in X-linked
spinal and bulbar muscular atrophy. Nature 352: 77-9
[0225] 35. Gu M, Dong X, Zhang X, Niu W (2012) The CAG repeat
polymorphism of androgen receptor gene and prostate cancer: a
meta-analysis. Molecular biology reports 39: 2615-24
[0226] 36. Ingles S A, Ross R K, Yu M C, Irvine R A, La Pera G,
Haile R W, Coetzee G A (1997) Association of prostate cancer risk
with genetic polymorphisms in vitamin D receptor and androgen
receptor. J Natl Cancer Inst 89: 166-70
[0227] 37. Giovannucci E, Stampfer M J, Krithivas K, Brown M, Dahl
D, Brufsky A, Talcott J, Hennekens C H, Kantoff P W (1997) The CAG
repeat within the androgen receptor gene and its relationship to
prostate cancer. Proc Natl Acad Sci U S A 94: 3320-3
[0228] 38. Irvine R A, Yu M C, Ross R K, Coetzee G A (1995) The
CAG and GGC microsatellites of the androgen receptor gene are in
linkage disequilibrium in men with prostate cancer. Cancer Res 55:
1937-40
[0229] 39. Hakimi J M, Schoenberg M P, Rondinelli R H, Piantadosi
S, Barrack E R (1997) Androgen receptor variants with short
glutamine or glycine repeats may identify unique subpopulations of
men with prostate cancer. Clin Cancer Res 3: 1599-608
[0230] 40. Tsujimoto Y, Takakuwa T, Takayama H, Nishimura K,
Okuyama A, Aozasa K, Nonomura N (2004) In situ shortening of CAG
repeat length within the androgen receptor gene in prostatic
cancer and its possible precursors. Prostate 58: 283-90
[0231] 41. Liu Y, Wilson S H (2012) DNA base excision repair: a
mechanism of trinucleotide repeat expansion. Trends Biochem Sci
37: 162-72
[0232] 42. Kim Y K, Kim B, Kim V N (2016) Re-evaluation of the
roles of DROSHA, Export in 5, and DICER in microRNA biogenesis.
Proc Natl Acad Sci U S A 113: E1881-9
[0233] 43. Zamudio J R, Kelly T J, Sharp P A (2014)
Argonaute-bound small RNAs from promoter-proximal RNA polymerase
II. Cell 156: 920-34
[0234] 44. Murmann A E, McMahon K M, Halluck-Kangas A, Ravindran
N, Patel M, Law C, Brockway S, Wei J J, Thaxton C S, Peter M E
(2017) Induction of DISE in ovarian cancer cells in vivo.
Oncotarget 8: 84643-84658
[0235] 45. Wang T, Birsoy K, Hughes N W, Krupczak K M, Post Y, Wei
J J, Lander E S, Sabatini DM (2015) Identification and
characterization of essential genes in the human genome. Science
350: 1096-101
[0236] 46. Blomen V A, Majek P, Jae L T, Bigenzahn J W,
Nieuwenhuis J, Staring J, Sacco R, van Diemen F R, Olk N, Stukalov
A, et al. (2015) Gene essentiality and synthetic lethality in
haploid human cells. Science 350: 1092-6
[0237] 47. Ceppi P, Hadji A, Kohlhapp F, Pattanayak A, Hau A, Xia
L, Liu H, Murmann A E, Peter M E. (2014). CD95 and CD95L promote
and protect cancer stem cells. Nature Commun. 5:5238...