SUPPORTING INFORMATION
Materials. Unless otherwise noted, all chemicals were obtained from the Aldrich Chemical
Company.
Synthesis of PiEVE Monomer. Potassium phthalimide (1400 g, 1eq) was suspended in the DMF
(5L). The resulting mixture was heated to 93 °C, and then tetrabutylammonium bromide (48.67
g, 0.02 eq) was added. The mixture was stirred for 0.5h. The reaction mixture was cooled to
60~70 °C, and 2-chloroethyl vinyl ether (1006.8 g, 1.25 eq) was added dropwise over 30 min.
The reaction mixture was heated to 90~100 °C for 1h. HPLC showed that ~8% potassium
phthalimide was remaining. The mixture was cooled to 50~60 °C, and then water (2.0 L) was
added dropwise over 1hr. The mixture was cooled to 20~30° C and aged for 30 mins. The
mixture was filtered through a Celite pad, the solid was washed with water (2*1 L). The above
solids were dissolved in CH2Cl2 (7 L). The solution was concentrated to about 3~4L, until solids
appeared. Cyclohexane (5 L) was added dropwise and the mixture was further concentrated to
2~3 L at 45~50 °C. The mixture was cooled to 10~20 °C and stirred for 30 mins. The mixture
was filtered through a Celite pad and the cake was washed with cyclohexane (2 L~3 L). The wet
cake was dried under vacuum at 40 °C for 12h. The desired product (PiEVE, 1.37 kg) was
obtained with 99.87% LCAP (0.13% of alcohol impurity) and 99.6% LCWP in 83.2% corrected
yield. 1H NMR (CDCl3, 500 MHz): δ 7.86 (dd, 2H, J = 5.3, 3.0 Hz), 7.72 (dd, 2H, J = 5.4, 3.0 Hz),
6.42 (dd, 1H, J = 14.4, 6.8 Hz), 4.20 (d, 1H, J = 14.4, 2.2 Hz), 4.01 (m, 3H), 3.94 (m, 2H). GC was
used to measure residual solvent. The following solvents were present in the following
quantities; MeOH <100 ppm, DCM 540 ppm, cyclohexane 320 ppm, 2-chloroethyl vinyl ether
<100 ppm and DMF <100 ppm.
Purification of Butyl Vinyl Ether. n-Butyl vinyl ether was purchased from Aldrich and contained
stabilizer, 0.01% KOH. The peroxide content of butyl vinyl ether was measured prior to
distillation and was measured to be <1 ppm (BLD) using QuantofixPeroxide 100 test strips. An
NH3-rinsed and oven-dried 2 L, 3-neck round-bottomed flask with magnetic stir bar and internal
thermocouple probe was set up in a heating mantle and was fitted with a 14/20 Vigreux column
(15 cm), distillation head with water-cooled condenser, and collection flask in a cold bath and
purged with nitrogen for 15 min. The flask was charged with butyl vinyl ether (1.50 L). The
nitrogen line was moved from the distillation pot to the collection flask and the pot was heated
to 100 ºC (internal temperature of 95 ºC) to begin distillation into a dry ice / acetone-cooled
collection flask. Fractions (150 – 200 mL, colorless liquid) were collected until <200 mL
remained in the pot. Fractions were checked by 1H NMR for purity (CDCl3). Fractions
containing large amounts of water, noted by ice crystals in the collection flask, were discarded.
Purification of Octadecyl Vinyl Ether. Octadecyl vinyl ether was purchased from TCI. The
distillation apparatus was set up behind a blast shield. An ammonia-rinsed 500 mL, 3-neck
round-bottomed flask with large magnetic stir bar and internal thermocouple probe was fitted
with a 24/40 Vigreux column (30 cm) with distillation head and air-cooled condenser. The flask
was set up in a heating mantle. The Coolant (water or air) in condenser should be warmed to
room temperature to prevent clogging of distillation equipment. Octadecyl vinyl ether (500 ml)
was first warmed to 45 oC to liquefy and this was then charged to the flask. The system was
evacuated and purged with nitrogen 3 times to remove air (oxygen). The flask was evacuated
(1 mmHg) and heated to 190-195 ºC (internal temperature of 185-190 ºC). The flask was
wrapped with glass wool with aluminum foil outside that to contain heat but the Vigreux
column was left open to the air to create a better temperature gradient. Fractions were
collected until <100 mL remained in the pot. Each fraction was warmed to 40 oC to melt prior
to sampling for GC analysis. By GC, the ratio of C18-vinyl ether:C16-vinyl ether >98 : 2 was
achieved after distillation.
Synthesis of Polymer 1. Synthesis of 18 kDa 15:4:1 [PiEVE:BE:ODVE] by Flow Polymerization.
The monomer solution was prepared by dissolving octadecyl vinyl ether (0.734 g, 2.5 mmol, 1
eq), n-butyl vinyl ether (0.99 g, 9.9 mmol, 4 eq) and N-(2-Vinyloxy-ethyl)phthalimide (8.03 g, 37
mmol, 15 eq) dichloromethane (150 mL) with a water content of 100 ppm. The catalyst
solution was prepared by dissolving boron trifluoroetherate (0.107 g, 1.5 mol% vs monomers)
dichloromethane (5.23 mL) with a water content of 100 ppm. The quench solution was
prepared by dissolving 2M ammonia in methanol (1.5 mL, 4 eq versus boron triflouride diethyl
etherate) dichloromethane (103 mL). Stream 1 is pumped at 1.429 mL/min through 1/16" PTFE
and 316 stainless steel tubing introduced to a controlled bath set at -30 °C. Stream 2 is pumped
at 0.0714 mL/min through 1/16" PTFE and 316 stainless steel tubing introduced to the
controlled bath. Streams 1 and 2 mix in a 1 mm ID 316 stainless steel tee before entering a 30
mL coil of 1/8" 304 stainless steel tubing. Stream 3 is pumped at 1.429 mL/min through 1/16"
PTFE and 316 stainless steel tubing introduced to the controlled bath before mixing with the
resulting stream from the 30 mL coil (mixture of Streams 1 and 2) in a 1 mm ID 316 stainless
steel tee. The resulting stream exits the controlled bath to a collection vessel. The collected
polymer was isolated by removal of dichloromethane under reduced pressure to afford the
product Polymer 1 with a molecular weight of Mw = 18.1 kDa and polydispersity index of 1.7.
1H NMR (500 mHz, CDCl ) = δ 7.8-7.6 (om), 3.9-3.0 (om), 1.9-1.1 (om), 1.0-0.7 (om).
3
Synthesis of Polymer 2. Synthesis of 18 kDa 15:4:1 [PiEVE:BE:ODVE] by Batch Polymerization.
In a dry 3-neck flask fitted with an overhead stirrer, nitrogen inlet, and temperature probe was
charged dichloromethane (112 ml, <10 ppm H2O). The flask was cooled to -40 °C (cryocool
bath) then charged BF3-Et2O (1.19 ml, 9.44 mmol). In a separate round bottom flask dissolved
n-butyl vinyl ether (48.9 ml, 378 mmol), octadecyl vinyl ether (28 g, 94.0 mmol), and N-(2Vinyloxy-ethyl)phthalimide (308 g, 1416 mmol) in dichloromethane (2100 ml, <10 ppm H2O).
The solution of vinyl ether monomers were then charged to the reaction vessel over 1 hr at a
constant rate using a Knauer pump. The solution was then let stir for an additional 1 hr and
then quenched with NH3 in MeOH (2.0 M, 47 ml) and removed from the cooling bath. Once at
room temperature, the reaction mixture was transferred to a recovery flask concentrated in
vacuo to yield product Polymer 2 (376 g, 598 mmol, 101% yield) as a foamy white solid with a
molecular weight of Mw = 18.3 kDa and polydispersity index of 1.8. 1H NMR (500 mHz, CDCl3) =
δ 7.8-7.6 (om), 3.9-3.0 (om), 1.9-1.1 (om), 1.0-0.7 (om).
Synthesis of Polymer 3. Synthesis of 27.1 kDa 15:4:1 [PiEVE:BE:ODVE] by Flow
Polymerization. The monomer solution was prepared by dissolving octadecyl vinyl ether (6.31
g, 21.27 mmol, 1 eq), n-butyl vinyl ether (8.52 mL, 85.07 mmol, 4 eq) and N-(2-Vinyloxyethyl)phthalimide (69.30 g, 319.02 mmol, 15 eq) in dichloromethane (900 mL) with a water
content of 50 ppm. The catalyst solution was prepared by dissolving boron trifluoroetherate
(0.92 g, 6.48 mmol, 1.5 mol% vs monomers) and dichloromethane (45 mL) with a water content
of 50 ppm. The quench solution was prepared by combining 2M ammonia in methanol (12.96
mL, 25.91 mmol, 4 eq versus boron triflouride diethyl etherate) was dissolved in
dichloromethane (887 mL). To generate polymer, stream 1 (monomer solution) was pumped at
1.429 mL/min through 1/16" PTFE and 316 stainless steel tubing introduced into the controlled
temperature bath set at -30 °C. Stream 2 (catalyst solution) was pumped at 0.0714 mL/min
through 1/16" PTFE and 316 stainless steel tubing introduced into the controlled temperature
bath. Streams 1 and 2 were mixed in a 1 mm ID 316 stainless steel tee before entering a 30 mL
coil of 1/8" 304 stainless steel tubing. Stream 3 (quench solution) was pumped at 1.429 mL/min
through 1/16" PTFE and 316 stainless steel tubing introduced to the controlled bath prior to
mixing with the resulting stream from the 30 mL coil (mixture of Streams 1 and 2) in a 1 mm ID
316 stainless steel tee. The resulting stream exited the controlled bath into a collection vessel.
The collected polymer was isolated by removal of solvent under reduced pressure to afford the
product Polymer 3a with a molecular weight of Mw = 27.1 kDa and polydispersity index of 2.2.
1H NMR (500 mHz, CDCl ) = δ 7.8-7.6 (om), 3.9-3.0 (om), 1.9-1.1 (br), 1.0-0.7 (om).
3
The same procedure was utilized to synthesize two subsequent batches of Polymer 3.
Polymer 3b: Mw = 29.4 kDa and polydispersity index of 2.0. 1H NMR (500 mHz, CDCl3) = δ 7.87.6 (om), 3.9-3.0 (om), 1.9-1.1 (om), 1.0-0.7 (om).
Polymer 3c: Mw = 28.0 kDa and polydispersity index of 2.8. 1H NMR (500 mHz, CDCl3) = δ 7.87.6 (om), 3.9-3.0 (om), 1.9-1.1 (om), 1.0-0.7 (om).
Polymer Deprotection and Purification. In a 3-neck flask fitted with an overhead stirrer, reflux
condenser, and nitrogen inlet was slurried Polymer 2 (50.0 g, 79 mmol) in 2-Propanol (1000
mL). Hydrazine (25% wt in H2O) (499 ml, 3889 mmol) was charged to the reaction vessel and
the reaction vessel was heated (65 °C). After 16 hrs, the reaction mixture was cooled to room
temperature. A constant volume distillation was performed to remove 2-propanol while adding
0.1 M NaOH to maintain a volume of 1500 mL of total reaction volume. The distillation was
continued until the amount of 2-propanol remaining in the reaction mixture was below 1
percent of the total volume as monitored by GC. The aqueous polymer solution was then
subjected to TFF (tangential flow filtration) purification (PALL centremate membrane, 1K MW
cutoff, OS001C12) with NaOH (0.25 N) until the HPLC of the retentate polymer solution
indicated complete removal of phthalhydrazide. The TFF process was continued using water
until the pH of waste stream (permeate) became neutral (pH 7-8). The aqueous solution was
then freeze-dried to obtain the product Polymer 2-deprotected (20.3 g) as a sticky oil. The
water content of the isolated polymer was determined by Thermogravimetric Analysis (TGA).
The sodium content of the isolated polymer was determined by ICP-MS. The weight percent of
the isolated polymer was determined by subtracting the weight of water and sodium hydroxide
in the polymer solid from the total weight of polymer solid. 1H NMR (500 mHz, D2O) = δ 3.83.0 (om), 2.8-2.6 (om), 2.0-1.0 (om), 0.9-0.7 (om). All other polymers were deprotected and
purified using the same general procedure as that described above.
Structure of Amino Low Hex 9 (LH9):
5’-amil-iB-CUAGCUGGACACGUCGAUATsT-iB-3’ (SEQ ID NO.:1)
3’-UsUGAUCGACCUGUGCAGCUAU-5’ (SEQ ID NO.:2)
amil = amino linker
iB = Inverted deoxy abasic
CU = 2’-fluoro (F)
AGT = 2’-deoxy
UGA = 2’-methoxy (OMe)
AU= ribose
s= phosphorothioate linkage
Structure of Amino Zimmerman ApoB (ZimmApoB):
5’-amil-GGAAUCUUAUAUUUGAUCCAsA-3’ (SEQ ID NO.:3)
3’-UsCsCCUUAGAAUAUAAACUAGGUU-5’ (SEQ ID NO.:4)
amil = amino linker
UC = 2’-methoxy (OMe)
AUG= ribose
s= phosphorothioate linkage
Sata Modification of Amino Zimmerman ApoB Oligonucleotide. Amino Zimmerman ApoB
Oligonucleotide (1g, 0.0714 mmol) was dissolved in 0.1M sodium bicarbonate buffer (20 ml, 50
mg/mL) in a vial with magnetic stir bar and cooled to 0-5 °C in an ice water bath. In a separate
vial SATA (83 mg, 0.357 mmol, 5 equivalents; Thermo Scientific part 26102) was dissolved in
0.78 ml DMSO. The SATA solution was added over 1min and the clear, colorless reaction
mixture stirred at 0-5 °C for 2h. After 2h, the reaction mixture was sampled and analyzed by
HPLC for consumption of the starting oligonucleotide. If the amount of unmodified
oligonucleotide was greater than 5 %, then an additional charge of SATA in DMSO (2.0
equivalents) was added and the reaction aged at 0-5 °C for completion of the SATA conjugation
(confirmation by HPLC). The reaction mixture was purified by dialysis (MW cutoff and
manufacturer information) using endonuclease free water until HPLC indicated the removal of
N-hydroxysuccinimide, and N-succinimidyl-S-acetylthioacetate. The recovered solution was
lyophilized to afford the product sata-ZimmApoB as a white fluffy solid.
Polymer Conjugate Synthesis. Polymer Polymer 2-deprotected (1.2 g) was placed in a 40 mL
vial and was dissolved in 100 mM sterile TRIS buffer at pH 9 (120 mL, 10 mg/mL) and added to a
1 L sterile plastic bottle. To this solution was added SMPT (Thermo Scientific, ) as 1 mg/mL
solution in dmso (18 mg, 1800 uL) corresponding to 1.5 wt% with respect to the polymer
weight. The solution was stirred for 1 hr at rt to generate activated polymer. The activated
polymer solution was further diluted using 100 mM sterile TRIS buffer at pH 9 (496 mL),
followed by the addition of sata-modified siRNA sata-ZimmApoB as a solution in water (250
mg, 32.4 mg/mL, 7716 uL). This solution was aged for 4 hours at room temperature to generate
the siRNA-polymer conjugate. In a separate 1 L sterile plastic bottle, solid CDM-NAG (5.5 g) and
CDM-PEG (2.85 g) was added. The siRNA-polymer conjugate solution was transferred by
pouring into the plastic bottle containing the CDM-NAG and CDM-PEG solids. The mixture was
stirred for 2 minutes to dissolve all solids and then transferred by pouring into the original
plastic bottle which contained the siRNA-polymer conjugate. The reaction was stirred for 1
hour to generate the product masked siRNA-polymer conjugate. The pH of the final solution
was monitored to ensure the pH was 8-9 throughout the conjugation process. Purification of
the masked polyconjugate was performed using a tangential flow filtration (TFF) purification
process (get information from Sergey’s group). The amount of RNA covalently attached to
poly(vinyl ether) polymers (conjugation efficiency) was determined using strong anion
exchange chromatography (SAX). The conjugation efficiency of the product 2a’ was measured
as 89% by SAX (see Figure 10 and conjugation efficiency assay details for further information).
The percentage of amines in a poly(vinyl ether) polymer that are covalently modified with
disubstituted maleimides CDM-NAG and CDM-PEG (molar basis) was determined by HPLC
(masking efficiency). The masking efficiency of the product 2a’ was measured as 50% (see
masking efficiency assay for further details). The RNA concentration in the product 2a’ was
measured using ICP and this concentration was used in determining dilutions/dosage volumes
for in vivo studies.
Masked Polymer Conjugate Purification Process. Tangential flow filtration (TFF) process was
used to purify masked polymer conjugate formulations (i.e. 2a’) of un-incorporated
components and to exchange buffer to pharmaceutically acceptable formulation vehicle. The
TFF filter material was made of either modified polyethersulfone (PES) or regenerated cellulose.
The selection of molecular weight cutoff for these membranes was done with efficiency of
purification and retention of polymer conjugate in mind. The processing parameters, including
but not limited to feed pressure, retentate pressure, crossflow rate and filtrate flux were set to
allow reproducibility from batch to batch and linear scaling of the process. Using the difiltration
mode of TFF, the reaction impurities were filtered out into the permeate while the retained
polymer conjugate underwent a buffer exchange. After TFF, the final product was concentrated
to 0.4-2.0 mg/ml of siRNA and sterile filtered using a 0.2μm PES syringe filter and stored at -20
°C until use.
Proton NMR. The 1H and spectra were recorded on a Bruker AV or DPX series NMR
spectrometer at a frequency of 400 MHz or 500 MHz as noted, and are internally referenced to
residual HOD at 4.80?ppm, CHD2Cl2 at 5.32ppm or CHCl3 at 7.27 ppm. Data for 1H NMR are
reported as follows: chemical shift (δ), multiplicity (s = singlet, d = doublet, t = triplet, q =
quartet, m = multiplet, o= overlapped, ) integration and coupling constant (Hz). 1H NMR
spectra were in full accordance with the expected structures.
Determination of Monomer Ratio for Deprotected Polymer:
H2
C
OY
H2
C
H2
C
OX
CH
CH
CH
O
O
O
CH2
Y = any monomer
H2C
H2C
(a)
CH2
CH2
H2C
CH2
ND2
H3C
X = H, CH3
CH2
(H2C)14
(c)
(b)
H3C
5a + 3b + 3c = (CH/CH2, broad integral region from ~ 4.0-2.4 ppm)
a + 3b + 17c = (1/2CH2, broad integral region from ~ 2.0-1.0 ppm)
b + c = (1/3CH3, broad integral region from ~1.0-0.2 ppm)
Molecular Weight Determination Using SEC (Size Exclusion Chromatography). Polymer
molecular weight determination for protected polymers was performed on an Agilent 1100
high performance liquid chromatograph (HPLC) coupled with a Wyatt miniDAWN™ TREOS (3
angle multi angle light scattering (MALS) system) and a Wyatt Optilab® T-rEX (refractive index
detector). Chromatography was performed using two size exclusion chromatographic columns
in tandem, Waters Styragel HR3 Column, 5 um, 7.8 x 300 mm (THF) and Waters Styragel HR4E
Column, 5 um, 7.8 x 300 mm (THF) with 100% THF as mobile phase at a flow rate of 1.0 mL/min.
The temperature of the column was set at 25 °C and the UV detection wavelength was 260 nm.
The polymer sample was dissolved in THF at 1-10 mg/mL and 0.5 mg material was injected.
Instrument normalization and calibration was performed using 2-40K polystyrene standards
with a PDI of less than 1.1 (purchased from Polymer Laboratories). No calibration standards
were used in determination of molecular weights or polydispersities. The dn/dc values were
obtained for each injection assuming 100% mass elution from the columns. These values were
also independently verified by measuring the dn/dc independently using a Wyatt Optilab® T-rEX
refractometer. The data was collected and processed using Wyatt Astra software.
Deprotected polymer molecular weight was not measured directly. Deprotected
molecular weight values were calculated by adjusting the corresponding protected molecular
weight based on the mass loss due to removal of the protecting group.
Conjugation Efficiency (siRNA-polymer conjugation)
The strong anion exchange assay (SAX assay) is used to determine the
conjugation efficiency by analyzing the final masked polyconjugate with and without
dithiothreitol (DTT) treatment.
Chromatographic Conditions
Column:
Proteomix SAX-NP3 Column (Sepax
Technologies), 3.0 m, 100 mm x 4.6 mm
Column Temperature:
RT
Flow Rate:
0.4 mL/min
Detection: CAD and UV
UV = 260 nm, 234 nm and 343 nm (optional)
Injection Volume:
10 L
Run Time:
35 min
Mobile Phase:
A: 100 mM Tris, 10% Acetonitrile, pH8.0
B: 100 mM Tris, 10% Acetonitrile, 2 M LiCl,
pH8.0
Mobile Phase Program:
Time (min)
A%
B%
0
5
30
35
100%
100%
40%
0%
0%
0%
60%
100%
Equilibrate column for 5 min
Preparation of Solutions
Poly Conjugate: Inject as-is at 10 ul
DTT Treatment: equal volume of conjugate with 1.0M DTT, inject at 20 ul,
DTT solution: 1.0 M in DI water
Free RNA duplex as well as free RNA duplex-dimer was visualized using SAX
chromatography. Total RNA (both free and bound) was determined by using Inductively
Coupled Plasma (ICP) spectroscopy. Since the RNA is the only phosphorus containing species in
the formulations, determining the total phosphorus content can be used to directly determine
the total RNA concentration. Once the free RNA (duplex and duplex-dimer) and total RNA is
determined, the amount of RNA conjugated to the polymer can be calculated (i.e. conjugation
efficiency). Total RNA (bound and unbound RNA and RNA dimer) can also be determined and
visualized by pre-treatment of the polyconjugate with DTT prior to SAX chromatography.
Conjugation efficiencies are reported for all poly(vinyl ether) polymer conjugates in the text of
the paper.
Masking Efficiency
Total concentrations of CDM-NAG and CDM-PEG were determined using reversephase HPLC (UV at a wavelength of 260 nm) with mobile phases of 0.1% TFA in water and 0.1%
TFA in 70/30 methanol:acetonitrile. Rapid demasking of the polymer after injection onto the
column allows quantitation of CDMs with the polymer removed using a C18 guard column to
prevent chromatographic interference. Free (i.e. unbound) CDM-NAG and CDM-PEG is analyzed
by first filtering through a 10K centrifuge filter followed by analysis of the permeate using the
same reverse-phase HPLC method. Masking efficiency can be calculated by first calculating the
bound RNA, CDM-NAG and CDM-PEG. The polymer molecular weight in combination with the
total amines available for conjugation is then used with the bound ligands to calculate masking
efficiency. Masking efficiencies are reported for all poly(vinyl ether) polymer conjugates in
table 3.
In Vivo Evaluation of Efficacy in Mice
CD1 mice were tail vein injected with the siRNA containing polymer conjugates
at a specified dose (mg/kg) in a volume of 0.2 mL, 100mM TRIS/9% glucose, pH9, vehicle. Fortyeight hours post dose, mice were sacrificed and liver tissue samples were immediately
preserved in RNALater (Ambion). Preserved liver tissue was homogenized and total RNA
isolated using a Qiagen bead mill and the Qiagen miRNA-Easy RNA isolation kit following the
manufacturer's instructions. Liver ApoB mRNA levels were determined by quantitative RT-PCR.
Message was amplified from purified RNA utilizing primers against the mouse ApoB mRNA
(Applied Biosystems Cat. No. Mm01545156_m1). The PCR reaction was run on an ABI 7500
instrument with a 96-well Fast Block. The ApoB mRNA level is normalized to the housekeeping
PPIB mRNA and GAPDH. PPIB and GAPDH mRNA levels were determined by RT-PCR using a
commercial probe set (Applied Biosytems Cat. No. Mm00478295_m1 and Mm4352339E_m1).
Results are expressed as a ratio of ApoB mRNA/ PPIB / GAPDH mRNA. All mRNA data is
expressed relative to the vehicle control.