Solid-phase synthesis, characterization and RNAi activity of branch and hyperbranch siRNAs

Article abstract of DOI:10.1016/j.bmcl.2013.08.033

Linear, branch and hyperbranch siRNAs were effectively prepared for down-regulating GRP78 expression and inducing cell death in HepG2 liver cancer cells. Branch and hyperbranch GRP78 siRNAs were synthesized by automated solid-phase synthesis in good yields (44-78%) and isolated in excellent purities (>99%) following HPLC purification. Moreover, siRNAs adopted stable intramolecular hybrids as discerned by native PAGE and thermal denaturation studies. These sequences also exhibited the pre-requisite A-type helical trajectory for triggering RNAi activity as determined by CD spectroscopy. Biological studies confirmed potent suppression of GRP78 expression (50-60%) while compromising cancer cell viability by ～20%. Thus, branch and hyperbranch siRNAs may serve as potent siRNA candidates in cancer gene therapy applications.

Full text of DOI:10.1016/j.bmcl.2013.08.033

Bioorganic & Medicinal Chemistry Letters 23 (2013) 5270–5274
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Solid-phase synthesis, characterization and RNAi activity of branch
and hyperbranch siRNAs
Anthony Maina a, Brittany A. Blackman b, Christopher J. Parronchi b, Eva Morozko a, Maria E. Bender a,
Allan D. Blake b, David Sabatino a,
⇑
a Department of Chemistry and Biochemistry, Seton Hall University, 400 South Orange Avenue, South Orange, NJ 07079, USA
b Department of Biological Sciences, Seton Hall University, 400 South Orange Avenue, South Orange, NJ 07079, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Linear, branch and hyperbranch siRNAs were effectively prepared for down-regulating GRP78 expression
and inducing cell death in HepG2 liver cancer cells. Branch and hyperbranch GRP78 siRNAs were synthe-
sized by automated solid-phase synthesis in good yields (44–78%) and isolated in excellent purities
(>99%) following HPLC purification. Moreover, siRNAs adopted stable intramolecular hybrids as discerned
by native PAGE and thermal denaturation studies. These sequences also exhibited the pre-requisite A-
type helical trajectory for triggering RNAi activity as determined by CD spectroscopy. Biological studies
confirmed potent suppression of GRP78 expression (50–60%) while compromising cancer cell viability
by ꢀ20%. Thus, branch and hyperbranch siRNAs may serve as potent siRNA candidates in cancer gene
therapy applications.
Received 11 June 2013
Revised 26 July 2013
Accepted 5 August 2013
Available online 12 August 2013
Keywords:
Branch siRNA
Hyperbranch siRNA
GRP78 silencing
RNAi
Published by Elsevier Ltd.
Oncogene therapy
Among the multiple structures and functions of ribonucleic
acids1,2 (RNA), double-stranded RNAs (dsRNAs) have been vastly
implicated in the regulation of gene expression.3 The latter pos-
sesses significance for therapeutic applications, specifically via
the mechanism of RNA interference (RNAi). In this case dsRNA, re-
ferred to as short-interfering RNA (siRNA) is capable of down-reg-
ulating gene expression of mRNA by recruiting the silencing
activity of the RNA-Induced Silencing Complex (RISC) in vivo.4 In
the fight against cancer, siRNAs have been used to down-regulate
oncogene expression of cancer cell proteins, leading to apoptosis
in melanoma, neuroblastoma, and breast cancer cells, among many
others.5–7 In spite of their promise, siRNAs are still plagued by poor
pharmacokinetic properties8 which limits their therapeutic poten-
tial and raises the need to develop more potent constructs.
Combination approaches are rapidly emerging as a leading
strategy to improve siRNA efficacy in gene therapy applications.
For example, a vector-based approach has been developed, in
which a small library of siRNAs have been transfected and ex-
pressed in mouse bone marrow macrophages to target up to six
genes belonging to the tumor-necrosis factor-associated factors
(TRAF) family of proteins.9 These multiple siRNA vectors proved
more potent in silencing gene expression relative to the single
siRNA expression cassette. However, these cloning methods are
not amenable to the incorporation of modifications that may
improve the ‘drug-like’ properties of siRNAs. Chemically derived
siRNAs facilitate the incorporation of modifications and the assem-
bly of higher-order structures for optimizing the pharmaceutical
potential of siRNAs. For example, a tripartite-interfering RNA
(tiRNA)10a,b and related supramolecular RNA nanostructures10c,d
were self-assembled in vivo to increase the silencing efficiency of
a single or multiple genes. These combination approaches pro-
duced a more potent and long-lasting gene silencing effect relative
to the linear siRNA controls, albeit with the pre-requisite formation
of a highly intricate and specific self-assembly for triggering RNAi.
These studies also demonstrate that the RISC complex may
indiscriminately bind to multiple siRNA sequences and motifs for
triggering potent RNAi gene silencing effects.
The siRNA sequences described in this study (Table 1) are based
on the target nucleotides for down-regulating Glucose Regulated
Protein 78 (GRP78) expression in human cancer cells.11 In cancer,
GRP78 overexpression generates cell-surface GRP78 which signals
the un-folded protein response leading to rapid repair of cancer
cell proteins, cell proliferation and arrest of cancer cell
apoptosis.12,13
Thus, targeting and down-regulating cancer-cell GRP78 expres-
sion may function as an ideal method for inducing cancer cell
death. Towards this goal, we have developed a versatile yet simple
solid-phase synthesis procedure for linear, V-shape, Y-branched
and >-< hyperbranched GRP78 siRNAs. Specifically, V- and Y-
branch siRNAs (Table 1, 4 and 5, respectively) contain a single
branchpoint nucleotide which combines the sense and antisense
⇑
Corresponding author. Tel.: +1 973 313 6359; fax: +1 973 761 9772.
E-mail address: david.sabatino@shu.edu (D. Sabatino).
0960-894X/$ - see front matter Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.bmcl.2013.08.033

A. Maina et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5270–5274
5271
Table 1
Sequences synthesized in this study
Number
Sequencesa,b,c,d
1 a
b
2 a
b
3 a
b
4
50-AGUGUUGGAAGAUUCUGAU-30
50-AUCAGAAUCUUCCAACACU-30
50-GGAGCGCAUUGAUACUAGA-30
50-UCUAGUAUCAAUGCGCUCC-30
50-AGUUCAACGAGUAUCAGCA-30
50-UGCUGAUACUCGUUGAACU-30
2030 UCACAACCUUCUAAGACUA-50
50-rU
3050 AGUGUUGGAAGAUUCUGAU-30
2030 UCACAACCUUCUAAGACUA-50
50-UCACAACC-3050-rU
5
6
3050AGUGUUGGAAGAUUCUGAU-30
30-UAGUCUUAGAGGAUUGUGA-5020
rU-3050–UCA CAA CC–3050-rU
50-AUCAGAAUCUCCUAACACU-3050
2030 UCACAACCUUCUAAGACUA-50
3050-AGUGUUGGAAGAUUCUGAU-30
2030UCACAACCUUCUAAGACUA-50
3050 AGUGUUGGAAGAUUCUGAU-30
7
30-AGAUCAUAGUUACGCGAGG-5020
rU3050–UCA CAA CC–3050-rU
50-UCUAGUAUCAAUGCGCUCC-3
a
Sequences 1, 4, 5 and 6 derived from nucleotides 1236–1255 containing the initiation codon for GRP78 mRNA.11a
Sequence 2 derived from nucleotides 1887–1906 of GRP78 mRNA.11b
Sequence 3 is a non-specific control.
b
c
d
Sequence 7 contains sequences derived from a and b.
strands within a single molecular structure. Hyperbranch siRNAs
phosphate diester, 11, relative to the triester, 10, prior to acid
detritylation and chain extension reactions.14 This procedure
was completed by passing a solution of 2:3 v/v triethylamine:ace-
tonitrile through the column for 90 min followed by exhaustive
washing of the RNA-bound CPG with anhydrous organic solvents.
Detritylation with 3% dichloroacetic acid:dichloromethane
removed the 20-OMMT protecting group, followed by RNA synthe-
sis. An additional acetylation step caps the 30 or 50-ends of the ‘V-
shaped’ RNA molecule, 12, and prevents additional coupling
side-reactions by acetylating the terminal hydroxyl groups. A
manual delevulination procedure (0.5 M hydrazine hydrate in
3:2 v/v pyridine:acetic acid, 20 min) was used to liberate the
branchpoint 50-hydroxyl group15 followed by successive washing
are designed to contain two siRNA motifs embedded within the
same molecular structure by multiple branchpoint nucleotides.
The siRNAs target single or double sites of GRP78 oncogene expres-
sion (Table 1, 6 and 7, respectively). Furthermore, these types of
branch and hyperbranch siRNA motifs have not yet been described
and may offer novel insights into the substrate requirements for
triggering potent RNAi responses in cancer cells
The iterative semi-automated solid-phase synthesis protocol of
branch and hyperbranch siRNAs is next described (Scheme 1). The
first step entails linear RNA synthesis up to the branchpoint
nucleotide 9 (see Supplementary Scheme S1). Decyanoethylation
liberates the phosphate protecting groups affording a more stable
Scheme 1. Solid phase synthesis of branch siRNAs.

5272
A. Maina et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5270–5274
Figure 1. (A) Denaturing and (B) native PAGE of linear, branch and hyperbranch siRNAs.
Figure 2. (A) Thermal denaturation and (B) CD Spectroscopy of linear, branch and hyperbranch siRNAs. Equimolar concentrations of complimentary single strands (0.4
and branch/hyperbranch siRNAs (0.1–0.2 M) were dissolved in physiological phosphate buffer (140 mM KCl, 1 mM MgCl2, 5 mM Na2HPO4 adjusted to pH 7.2). Samples were
denatured at 95 °C for 5 min and annealed (rt to 5 °C) overnight prior to thermal denaturation studies by UV–Vis spectroscopy and structural studies by CD spectroscopy.
lM)
l
of the column with anhydrous organic solvents and completion of
the branched siRNA 13 (Table 1, 5). This cycle was repeated for
the generation of hyperbranch siRNAs (Table 1, 6 and 7). Upon
completion of the synthesis cycle, CPG-bound RNA was cleaved
and deprotected using 1:1 v/v solution of ammonium
hydroxide/methylamine for 20 min at 65 °C. The crude
oligonucleotides were suspended in mixture of 1:1 v/v
dimethylsulfoxide/triethylamine trihydrofluoride for 2 h at 65 °C
to complete 20-desilylation. Following post-synthesis work-up
conditions siRNAs, (4–7) were obtained in isolated yields of
a

A. Maina et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5270–5274
5273
47–78% and purities >99% by RP-IP HPLC (see Supplementary
Table S1).
(Fig. 2B). Linear siRNAs demonstrated typical CD profiles for A-
form helices, displaying a minimum peak at 240 nm and a broad
maximum between 255 and 290 nm.18 In the case of branch and
hyperbranch sequences (4–7), the A-type broad maximum
and minimum bands were observed between 255–290 nm and
240 nm, respectively, albeit with a decrease in the amplitudes of
molar ellipticities at these characteristic wavelengths. In spite of
this change, branch and hyperbranch sequences maintained a CD
signature consistent with the A-like helical structures. In light of
their enhanced duplex stability and A-type helices, branch and
hyperbranch siRNAs were next evaluated for RNAi activity (Fig. 3).
Confocal microscopy confirmed the presence of GRP78 in
HepG2 liver carcinoma cells (Fig. 3A). HepG2 cells were plated on
4 well chamber glass slides and treated with anti-GRP78 primary
and a secondary antibody conjugated to fluorescein isothiocyanan-
te (FITC). Images were captured with a 40X oil objective lens, from
which FITC labeled GRP78 was detected in green and DAPI-stained
cell DNA in blue. The RNAi activity of hyperbranch siRNA 7
(40 pmol) was assessed in HepG2 cells at 37 °C for 48 h. After
transfection, cells were plated on glass slides and treated with an
anti-GRP78 primary and appropriate secondary antibodies to
examine GRP78 expression levels (Fig. 3B). Hyperbranch siRNA 7
treatment potently silenced GRP78 expression as shown by the
minimal FITC-signaling relative to the control untransfected cells
(Fig. 3A and B).
With purified sequences in hand, a native and denaturing PAGE
was conducted in order to determine the propensity for siRNA
hybridization (Fig. 1). Under native conditions, branch and hyper-
branched siRNAs (4–7) showed slower electrophoretic mobility
and the formation of stable higher-order complexes relative to lin-
ear controls. Alternatively, under denaturing conditions, most siR-
NAs migrated as their linear single-stranded oligonucleotides.
Exceptionally, branch and hyperbranch siRNAs (4–7) maintained
stable higher-order complexes under denaturing conditions, sug-
gesting that intramolecular base-pairing owns greater stability rel-
ative to a duplex forming between two complementary strands.
These results are not entirely surprising, considering RNA hairpins
possess thermodynamically more stable folded structures relative
to their linear double-stranded sequences.16
In order to confirm the enhanced duplex stabilities of branch
and hyperbranch siRNAs relative to their linear controls, thermal
denaturation experiments were conducted. siRNAs (0.1–0.4 lM)
were annealed in physiological phosphate buffer (140 mM KCl,
1 mM MgCl2, 5 mM Na2HPO4, pH 7.2) and analyzed for bio-physi-
cal studies by thermal melt experiments17 (Fig. 2A). Consistent
with the gel data, branch and hyperbranch siRNAs displayed great-
er thermal stabilities (Tm: 68–89 °C) and enhanced hyperchromic-
ities (%H: 5–35%) relative to their linear controls (Tm: 67–72 °C and
%H: 5–25%). Moreover, branch and hyperbranch siRNAs, 4–7
exhibited broad thermal denaturation curves that are indicative
of multiple phase transitions with increasing temperatures, as is
characteristic of higher-order oligonucleotide structures.19 In order
to assess whether siRNAs adopted the pre-requisite A-type helical
trajectory for activity, CD spectroscopy18 was used to determine
their secondary structures relative to single strand controls
Western blotting experiments were conducted to explore the
extent of GRP78 silencing in HepG2 cells with linear, branch and
hyperbranch siRNAs. After transfection, cell lysates were prepared
(RIPA lysis buffer, Invitrogen) and resolved by SDS PAGE. GRP78
expression levels were detected with an anti-GRP78 primary anti-
body and a secondary antibody coupled to horseradish peroxidase
(HRP). GRP78 expression levels correspond to the redox activity of
Figure 3. (A and B) Confocal microscopy images of HepG2 cells without and with hyperbranch GRP78 siRNA 7, respectively, (C) Western blot of HepG2 cells0 GRP78
expression following transfection with linear (2) and (1+2) siRNA controls and with V-shape (4), branch (5) and hyperbranch siRNAs (6,7), (D) GRP78 expression levels were
detected by autoradiography and quantified using NIH ImageJ, (E) cell viability data determined by Trypan blue exclusion. The data shown are the mean SEM for 3 or more
experiments, each performed in duplicate and analyzed with GraphPad Prism 4.0 and statistical significance was determined by ANOVA (p <0.05).

5274
A. Maina et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5270–5274
the 3,30,5,50-tetramethylbenzidine (TMB) substrate.21 The gel
(Fig. 3C) was analyzed by autoradiography to determine GRP78
expression levels. The data shows effective silencing (40–60%) of
GRP78 expression by linear (ꢀ40–60%), V-shape (ꢀ55%), Y-branch
(ꢀ50%) and hyperbranch (ꢀ50–60%) siRNAs (Fig. 3D).
References and notes
1. Joyce, G. F. Nature 2002, 418, 214.
2. Wallace, J. C.; Edmonds, M. Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 950.
3. (a) Hammond, S. M.; Caudy, A. A.; Hannon, G. J. Nat. Rev. Genet. 2001, 2, 110; (b)
Mello, C. C.; Conte, D. Nature 2004, 431, 338.
4. Fire, A.; Xu, S.; Montgomery, M. K.; Kostas, S. A.; Driver, S. E.; Mello, C. C. Nature
1998, 391, 806.
5. (a) Heidenreich, O. Curr. Pharm Biotechnol. 2004, 5, 349; (b) Scherr, M.; Battmer,
K.; Schultheis, B.; Ganser, A.; Eder, M. Gene Ther. 2005, 12, 12; (c) Tai, J.; Wang,
G.; Liu, T.; Wang, L.; Lin, C.; Li, F. J. Biochem. Mol. Toxicol. 2012, 26, 499.
6. Gartel, A. L.; Kandel, E. S. Biomed. Eng. 2006, 23, 17.
7. Pai, S. I.; Lin, Y. Y.; Macaes, B.; Meneshian, A.; Hung, C. F.; Wu, T. C. Gene Ther.
2006, 13, 464.
Cell viability data was used to establish the extent of HepG2 cell
death following GRP78 knockdown. A Trypan blue cell staining
assay22 was used to track the number of dead cells. Approximately
10–20% cell death was observed with the branch and hyperbranch
siRNAs (4–7). The observed cell death percentages for linear siRNA
controls were consistent with previous studies,20 approximately
5% (1), compared to about 17% with V-shape siRNA (4), about
11% with Y-branch siRNA (5) and 15% with hyperbranch siRNAs
(7) (Fig. 3E). Interestingly, hyperbranch siRNA 7, targeting double
sites of GRP78 mRNA was found to elicit more potent GRP78
knockdown (60% vs 40%) and cell death (15% vs 5%) relative to
the linear controls (1 and 2), highlighting its synergistic effects in
gene therapy applications.
8. Bumcrot, D.; Manoharan, M.; Koteliansky, V.; Sah, D. W. Nat. Chem. Biol. 2006, 2,
711.
9. Wang, S.; Shi, Z.; Liu, W.; Jules, J.; Feng, X. BMC Biotechnol. 2006, 6, 50.
10. (a) Chang, C. I.; Lee, T. Y.; Yoo, J. W.; Shin, D.; Kim, M.; Kim, S.; Lee, D. K. Nucleic
Acids Ther. 2012, 22, 30; (b) Chang, C. I.; Lee, T. Y.; Kim, S.; Sun, X.; Hong, S. W.;
Yoo, J. W.; Dua, P.; Kang, H. S.; Kim, S.; Li, C. J.; Lee, D. K. J. Gene Med. 2012, 14,
138; (c) Nakashima, Y.; Abe, H.; Abe, N.; Aikawa, K.; Ito, Y. Chem. Commun.
(Camb.) 2011, 47, 8367; (d) Aviñó, A.; Ocampo, S. M.; Perales, J. C.; Eritja, R. J.
Nucleic Acids 2011, 586935, 1.
11. (a) Suzuki, T.; Lu, J.; Zahed, M.; Kita, K.; Suzuki, N. Arch. Biochem. Biophys. 2007,
468, 1; (b) Tsutsumi, S.; Namba, T.; Tanaka, K. I.; Arai, Y.; Ishihara, T.; Aburaya,
M.; Mima, S.; Hoshino, T.; Mizushima, T. Oncogene 2006, 25, 1018.
12. Pfaffenbach, K. T.; Lee, A. S. Curr. Opin. Cell Biol. 2011, 23, 150.
13. (a) Arap, M. A.; Lahdenranta, J.; Mintz, P. J.; Hajitou, A.; Sarkis, A. S.; Arap, W.;
Pasqualini, R. Cancer Cell 2004, 6, 275; (b) Kim, Y.; Lillo, A. M.; Steiniger, S. C.;
Liu, Y.; Ballatore, C.; Anichini, A.; Mortarini, R.; Kaufmann, G. F.; Zhou, B.;
Felding-Habermann, B.; Janda, K. D. Biochemistry 2006, 45, 9434; (c) Li, J.; Ni,
M.; Lee, B.; Barron, E.; Hinton, D. R.; Lee, A. S. Cell Death Differ. 2008, 15, 1460.
14. Damha, M. J.; Braich, R. S. Tetrahedron 1998, 39, 3907.
15. (a) Lackey, J. G.; Sabatino, D.; Damha, M. J. Org. Lett. 2007, 9, 789; (b)
Meldgaard, M.; Nielsen, N. K.; Bremmer, M.; Pesersen, O. S.; Olsen, C. E.;
Wengel, J. J. Chem. Soc., Perkin Trans. 1 1997, 1951; (c) Iwai, S.; Ohtsuka, E.
Tetrahedron Lett. 1998, 29, 5383.
16. Khvorova, A.; Reynolds, A.; Jayasena, S. D. Cell 2003, 115, 209.
17. Mergny, J. L.; Larcroix, L. Oligonucleotides 2003, 13, 515.
18. Gray, D. M.; Hung, S. H.; Johnson, K. H. Methods Enzymol. 1995, 246, 19.
19. Ballin, J. D.; Bharill, S.; Fialcowitz-White, E. J.; Gryczynski, I.; Gryczynski, Z.;
Wilson, G. M. Biochemistry 2007, 46, 13948.
In conclusion, we have demonstrated the synthesis, character-
ization and anti-cancer activity of a novel class of siRNA molecules.
In this study, branch and hyperbranch siRNAs were effectively syn-
thesized by semi-automated procedures. These siRNAs conferred
stable single molecular hybrid structures which maintained the
pre-requisite A-type helical structure for siRNA application. Of
interest, the branch and hyperbranch siRNAs were found to be po-
tent silencers of GRP78 expression (50–60%) while triggering sig-
nificant cancer cell death (ꢀ20%) relative to their linear controls.
Therefore, branch and hyperbranch siRNAs effectively extend the
repertoire of bio-active siRNAs that may be useful in the develop-
ment of more potent oncogene therapeutics.
Acknowledgments
20. Alhoot, M. A.; Wang, S. M.; Sekaran, S. D. PLoS One 2012, 7, e34060.
21. Badway, A. C.; West, F. M.; Tente, S. M.; Blake, A. D. Biochem. Biophys. Res.
Commun. 2004, 319, 1222.
The authors would like to thank Seton Hall University for con-
tinued support. The authors also extend their gratitude to Dr. Mark
Hail at Novatia LLC for ESI MS analyses of the siRNA sequences.
22. Lees, G. J. Histochemistry 1989, 91, 357.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2013.08.
033.