3.2 Methods
3.2.8 Electrophoretic mobility shift assay
First, 1 % (w/v) of agarose gel was prepared. 1.5 g of agarose was dissolved in 150 mL of TBE buffer. The solution was heated up to 100 °C in the microwave until agarose was completely dissolved. When the solution cooled down to 50 °C, 150 μL of GelRed™ was added and the solution was poured into the electrophoresis unit (Sub-Cell GT Horizontal Electrophoresis System, 15 x 10 cm tray, BIORAD) and solidified at RT. The agarose gel was overlaid with a TBE-buffer (approximately 1 L), 5 μL of loading buffer were placed into the agarose gel pockets and subsequently 20 μL of polyplexses formed with 200 ng of pCMV-luc after 40 min of incubation. In the first pocket control free pCMV-luc was placed.
Electrophoresis was performed at 120 V for 70 min at RT. Picture of the electrophoresis assay was taken in the Dark Hood DH-40 using Canon EOS 2000 Kit.
37 3.2.9 Analytical methods
3.2.9.1 1H-NMR
1H-NMR spectra of all synthesized structures were recorded using an AVANCE III HD 500 (500 MHz) by Bruker with a 5 mm CPPBBO probe. Spectra were recorded without TMS as internal standard. Therefore, all signals were calibrated to the residual proton signal of the deuterium oxide (D2O) and chloroform-d, respectively. All chemical shifts were calibrated to the residual proton signal of the solvent (deuterium oxide; δ = 4.79 ppm and chloroform-d; δ = 7.26 ppm) and reported in ppm. The spectra were analyzed using the software MestreNova (Ver.11.0.4; Mestrelab Research). Integrals were set manually and normalized to the succinic acid peaks due to its high resolution.
3.2.9.2 ESI-MS
ESI-MS was carried out using Thermo Finnigan LTQ FT Ultra Fourier Transform ion cyclotron resonance mass spectrophotometer with resolution set to 100.000 at 400 m/z.
Depending on the sample, mass ranges from 50 to 2000 u were measured. The spray capillary voltage at the IonMax ESI head was 4 kV, the heater capillary temperature 250 ° C, the nitrogen sheath gas flow 20 and the sweep gas flow 5 units. Building blocks were dissolved in chloroform and analyzed in positive and negative mode.
3.2.9.3 MALDI-TOF MS
One μL of Super-DHB matrix was applied on a MTP AnchorChip (Bruker Daltonics, Bremen, Germany) and subsequently 1.5 μL of the sample solution (0.5 mg/mL in Millipore water) were added to the matrix spot under rapid pipetting. After the matrix dried and crystalized, samples were analyzed in positive or negative reflector mode using an Autoflex II mass spectrometer (Bruker Daltonics, Bremen, Germany).
38 4 RESULTS AND DISCUSSION
4.1 Synthesis of the Stp and Sph building blocks To improve the buffering capacity in
the lower endosomal pH, compaction of nucleic acid at a neutral pH, and subsequent transfection of nanocarriers, two building blocks were synthesized.
They were designed in a way that
enabled their use in SPAS using common Fmoc strategy and sequence-defined oligoaminoamides with precise topology and so that site-specific functionalization could be realized.(54) The blocks originate from PEI, which is very toxic, but if we break it into smaller segments and connect them in a more biodegradable and biocompatible manner, we can attain a very efficient transfection agent.(57) TEPA and PEHA are four and five sequences of a repeating diaminoethane motif, respectively and represent smaller, less toxic PEI sequences. Furthermore, coupling PEHA and TEPA with succinic anhydride resulted in reduced toxicity.(43) The purpose of succinic acid at one terminal primary amine, is to connect one oligoamine segment to the other and make the oligoaminoamide more biodegradable.(54) TEPA is a commercially available compound, whereas PEHA may be purchased, but is available only as a mixture of isomers. As recent studies reported that linear PEI are more efficient transfection agents, PEHA was purified, so that a linear isomer was obtained.(41) The linear PEHA isomer was distilled using high-vacuum distillation at a pressure of 1.5 mbar. Fraction with linear PEHA isomer collected at 190–195 °C was precipitated with concentrated HCl what resulted in a 24.1 % yield. In the past, PEHA was purified using Kugelrohr distillation method which is not as efficient as high-vacuum distillation. The latter provides more efficient rectification and therefore cleaner fractions can be obtained.(58) From here on, Stp and Sph building blocks were synthesized with the same steps, except for different amounts of reagents due to one additional amino group. The Sph building block was synthesized for first time according with the improved protocol, where the volumes of needed solvents for dissolving are smaller and some unnecessary extraction steps are skipped. This could be also the reason for lower yield at the end, still these changes should theoretically not affect drastically affect the final yield.
Figure 12: Artificial aminoacid, Stp and Sph
39
Figure 13: Reaction scheme of Sph synthesis
The primary amines were firstly transiently protected with TFAEt as trifluoroacetamides to allow site-specific Boc- protection of the secondary amines by Boc anhydride in a one-pot-reaction. Furthermore, the Boc group served as protection due to unwanted side-reactions with secondary amines when using SPAS. TFAEt preferentially reacts with primary amines due to their stronger nucleophilicity in comparison to secondary amines. Only 2.1 equivalents were added dropwise into the cooled mixture to minimize the side-reaction with secondary amines. The intermediate products bis-tfa-TEPA(Boc3)/bis-tfa-PEHA(Boc4) were separated from the remaining TFAEt, Boc anhydride and other related structures through extraction with 5 % NaHCO3 solution and water. At the end, purification occurred via crystallization to remove some smaller 1,2-diaminoethane sequences that could still be present, due to an impure starting component. Structures were confirmed with 1H-NMR and ESI-MS. At this step, pentaboc- and tetraboc-adduct, respectively, could occur. However, the spectra showed this type of adduct appearing only in traces. The yield for bis-tfa-TEPA(Boc3) was 58.98 % and for bis-tfa-PEHA(Boc4) 54.62 %. The reason for a lower yield of Sph might lie in crystallization. In the next step, the trifluoroacetamide protection group was removed with alkaline hydrolysis using a 3 M NaOH solution, and the organic compound was extracted with DCM to remove sodium salt of TFA. Structures were confirmed with 1H-NMR and ESI-MS. This step resulted in a very high yield of 96.7 % and 92.39 % for TEPA(Boc3) and PEHA(Boc4), respectively. In the last step, primary amines were coupled with succinic anhydride and Fmoc-OSu and the final Fmoc-Stp(Boc3)-OH and Fmoc-Sph(Boc4)-OH were obtained. To selectively couple succinic acid only to one primary amine, the mixture was cooled down to -75 °C using dry ice and only 1.25 equivalents of
40
succinic anhydride was added dropwise. The reaction is preferentially directed to the one primary amine because after one primary amine is coupled, the second one becomes less nucleophilic, which leads the
reaction to new molecule with no reacted primary amines.
Subsequently, the last primary amine was coupled with a reactive ester, Fmoc-OSu, in the presence of DIPEA, which serves as proton scavenger, making the reaction very efficient. The Fmoc group is base labile group and is introduced to prevent random polymerization.
Subsequently, the organic phase was washed five times with a trisodium citrate buffer (0.1 M; pH 5.5) to remove unreacted primary
amines which are protonated and therefore hydrophilic. Still, other impurities are present, such as “double Fmoc” (Fmoc on both primary amines, Fmoc-PEHA(Boc)4-Fmoc), or not reacted Fmoc, or succinic acid on both primary amines, or not reacted TEPA(Boc3) and PEHA(Boc4), respectively. To remove these byproducts, Stp and Sph were isolated using DCVC (See Figure 8: Dry column vacuum chromatography apparatus). Silica gel was used as a stationary phase. A solvent gradient of n-heptane-ethyl acetate was used to remove Fmoc byproducts, occurring around the 30th fraction, and ethyl acetate-methanol to elute our final product. To determine, if Fmoc byproducts were separated from our product, each fraction was analyzed by TLC using silica gel aluminum plates for fluorescence quenching at 254 nm and solvent mixture of 7:3 CHCl3/MeOH. Fmoc byproducts are more lipophilic and have a Rf > 0.9 whereas the final product has a lower Rf of 0.6. The Stp building block separated from Fmoc byproducts very efficiently, whereas Sph did not. At the fraction, where the strongest fluorescence quenching of our product was detected, a strong spot of Fmoc byproducts also appeared. Consequently, these fractions were collected separately.
Isolated Stp and Sph were analyzed by 1H-NMR and ESI-MS. Spectra of Stp and Sph that had no traces of Fmoc byproducts on TLC plate were as expected, whereas a 1H-NMR Figure 14: Example of DCVC purification TLC. At the bottom, individual fractions are presented.
Fluorescence quenching spots with Rf values around 0.9 are caused by Fmoc by-products and product Fmoc-Stp(Boc3)-OH has an Rf of 0.6.
41
spectrum of another Sph batch had shown an additional spot on the TLC plate at a Rf around 0.9 where Fmoc byproducts can be observed. The spot displayed an additional Fmoc group, meaning that not all molecules reacted with the succinic anhydride, instead coupling into an additional Fmoc group, or have possibly couped to the succinic anhydride. However, the beside and Sph peak, the ESI-MS spectrum had, an additional intensive pick at 1099.55 g/mol, which corresponds to the molar mass of a “double Fmoc” adduct, and there was no sign of an adduct that had an additional Fmoc group coupled to succinic acid. To check if this byproduct interferes with further steps of our synthesis, where it should theoretically not, the [G-Sph-G-Sph-G-Sph-G] sequence was synthesized for both batches of Sph using an automatic synthesizer. The sequence mass was measured by MALDI-TOF MS, and the results showed that the mass of the clean batch was as expected 1186.93 g/mol, whereas the batch containing Fmoc byproducts was 893.57 g/mol. According to their mass, Fmoc-PEHA(Boc4)-Fmoc (“double Fmoc”) could interfere with our synthesis. Double Fmoc cannot couple with free amines, meaning that the Sph block could be omitted from the sequence, and that the mass of the dummy structure comprising the impure Sph, is lower for exactly one Sph building block. The reason why double Fmoc appeared later and not during the first composition of solvents, might be due to encapsulation into the Sph building blocks, which eluted with the new mixture of solvents, consequently releasing “double Fmoc”, and producing a batch with impurities. Nevertheless, the yield was still high enough to synthesize all the structures. The final Stp and Sph step resulted in a 49.6 % and 46.2 % yield, respectively. If we combine all reactions the overall yield, was 28.29 % and 5.48 % for Stp and Sph, respectively. The peak of Sph can be seen at 955.54 g/mol, whereas the intensive peak appears at 1099.55 g/mol.
O
Figure 15: Stp and Sph building block for SPAS
42
4.2 Four-arm oligoaminoamides design and synthesis
Four-arm oligoaminoamides were synthesized by SPAS using a common Fmoc strategy.
This method uses linkers attached to small porous beads, resins, for building peptide chains with consecutive coupling, washing, and deprotection steps. Amino acids are protected with protection groups that enable selective deprotection. The Fmoc base labile protection group is used as protection of a reactive amine that will react in the next coupling step, whereas other reactive groups such as thiol, hydroxy or additional amino groups are protected with other acid labile protecting groups. Standard steps in SPAS encompass resin loading, coupling, washing, deprotection, washing and subsequent cycle repeats. After every washing step, Kaiser’s test is performed to determine if the coupling, deprotection or washing steps were executed successfully.
Figure 16: Illustration of repeating coupling-washing-deprotection-washing steps in SPAS.
Firstly, the C-terminal amino acid is loaded on the resin and repeating cycle of coupling, washing, deprotection and washing steps is preformed to elongate the peptide chain. After coupling with all amino acids, obtained peptide is cleaved off the resin with TFA and the remaining protection groups are removed.
Its ability to remove excess reagents makes SPAS synthesis so convenient. It allows beads to be washed through a filter in the syringe reactor, preventing the loss of product due to immobilization of peptide on the resin beads that cannot permeate through the filter pores.
Using this technique, peptide structures can be synthesized very quickly, and with improving technology SPAS can also be automated. In addition to commercially available amino acids, other reagents are used in SPAS synthesis. PyBOP/HBTU and HOBt are both used as
43
activation reagents, as well as suppressors of racemization. A non-nucleophilic base like DIPEA is used as a proton scavenger. DIPEA ensures both the deprotonation of the amine, as well as the carboxylic function, facilitating easier active ester formation with the activation reagent. As organic solvents, DCM and DMF are mainly used to create nonaqueous environment for successful coupling, while facilitating swelling of the resin during the reaction at the same time.(59) The final product was cleaved off the resin using a cleavage cocktail that includes TFA – a very strong acid that can also remove other remaining protection groups. During this process highly reactive cationic species are generated from the protecting groups and resin-linkers. These species can react and modify electron-rich amino acid functional groups, such as tyrosine, tryptophan, methionine, and cysteine. The appearance of cationic species called for adding various nucleophilic reagents, known as scavengers, to the cleavage cocktail: water is a moderately effective scavenger for t-butyl cations and the products of the cleavage of arylsulphonyl-based protecting groups;
EDT is the best scavenger for t-butyl cations and protects unprotected tryptophan against sulphonation, while also facilitating the removal of the trityl protecting group from cysteine:
TIS is very efficient at quenching highly stabilized cations liberated after cleaving Trt. (60) TFA was mainly used at 95 % to achieve successful deprotection of all protection groups and to cleave the oligoaminoamide off the resin. For example, the Trt protection group on histidine needs TFA of at least 50 %, and a concentration of up to 90 % must be used to remove protection groups like Boc, OtBu, tBu, Trt on cysteine.(61) Groups such as t-butyl-based protecting groups, Pmc and Pbf from arginine, and Trt groups from asparagine, glutamine, histidine, and cysteine often call for TFA at 95 % to obtain a fully side deprotected peptide.(60)
In the first part of synthesis, the branching cores were synthesized. Ala-Wang resin was used instead of 2-chlorotritylchloride resin due to thermal hydrolysis of the trityl ester bond at a higher temperature, that we used for coupling amino acids with an automated synthesizer.(62) Only 0.25 eq. of Fmoc-L-Lys(Fomc)-OH relative to the free
resin-bound amines were coupled at the first Figure 17: Graph of dibenzofluene-piperidine adduct absorption peak at 301 nm
44
step to reduce the density of four-arm oligoaminomaides which are prone to aggregation at a high density.(63) The remaining free resin-bound amines were acetylated to prevent reacting with further reagents. Lysin was introduced two times as a symmetrical branching point, owing to two primary amines, to branch the structure into four-arm structures. Histidine was coupled after every branching point with the intent of fine-tuning the proton sponge activity.(40) After the branching cores were synthesized, the load was determined spectrophotometrically. By the Fmoc deprotection, Fmoc group reacts to dibenzofluene-piperidine adduct absorbance of which can be measured at the absorption peak of 301 nm and can give us a quantitative amount of the loaded resin beads.(56)
resin
Figure 18: Reaction scheme of dibenzofluene-piperidine adduct formation
Following sequences of four-arm structures were synthesized automatically; [Stp-H-R]3, [Stp-H]3R3, [Stp-H-W]3, [Sph-H]3Y3, [Sph-H]3W3, [Sph-H]3R3, [Sph-H-R]3, [Sph-H-K]3, [Sph-H-W]3, and [SphH]3 twice. For more efficient coupling, amino acids were coupled at 75 °C and additional coupling was executed. The Stp and Sph building blocks were introduced to improve buffering capacity, compaction of nucleic acid and subsequent transfection. The Sph should be superior because of one additional amino group.(44) Arginine and lysine were introduced to improve binding ability. Arginine showed to be superior in regard to DNA compaction in comparison to lysine.(64) Tryptophan and tyrosine were introduced to improve hydrophobic interaction and to stabilize polyplexes.(49–51) To these sequences, cysteines and lysines with azido moiety were coupled manually, due to the tendency of cysteines to crosslink to each other at higher temperatures, and the possibility
45
of azido lysine causing an explosion in the automatic synthesizer’s microwave, because of the released gasses. The purpose of cysteine is to stabilize polyplexes through covalent disulfide cross-linkage that can be reduced in the reductive cytosol and facilitates the release of the payload into the cytosol.(40)
Table 8: Illustration of the synthesized four-arm structures and their ID numbers
Structure and topology
1494 1518
1495 1516
1514
1513 1519 1493
1515 606
1517
46
Azido lysine was introduced at the ends of the four arms, for further attachment of shielding agents or ligands using the DBCO click-chemistry strategy. None of the synthesized oligoaminoamides have been synthesized before, except for 606 that has served as a “gold standard” due to its so far obtained best transfection results. The 606 oligoaminoamide was thus used for comparison with the novel four-arm sequences.(40) For the synthesis of four-arm oligoaminoamides, 1 % of Triton™ X-100 (v/v) was added to the DMF and DCM to reduce the surface tension, as well as to make reactive groups more accessible, and to wash the resin beads properly after every coupling and deprotection step.(63) Due to bad deprotection results of the Fmoc group in the past experiments by using only 20 % of piperidine, additional 2 % of DBU was added to the deprotection solution which is a much stronger base in comparison to piperidine, and 2 % is sufficient for successful deprotection.(65,66) All four arms were purified through SEC, which is a good purification method of choice due to its capability of changing the TFA salt into the HCl salt. The latter is more viable for in vivo and in vitro experiments, because the TFA salt might induce toxicity connected issues.(67) For this reason, both dialysis as well as HPLC are not an applicable method duo to the polycationic nature of OAA. The structures of purified oligoaminoamides were confirmed by 1H-NMR (See Appendix). Polyplexes were formed and evaluated regarding nucleic acid binding efficiency, size, and zeta potential.
4.3 Formation and characterization of polyplexes
Polyplexes were formed with plasmid pCMV-luc that encodes the gene for Photinus pyralis firefly luciferase under the control of a cytomegalovirus promotor and enhancer, using water as medium.(68) pDNA was chosen due to difficult polyplex formation, owing to large molecule size. In addition, efficiency of nucleic acid encapsulation can be better determined.
Every four-arm structure was formed with pDNA at three N/P ratios; 6, 12 and 20 to determine the quantity of oligomers for successful encapsulation. The N/P ratio is a proportion of the polymer’s positively chargeable amine groups and negatively charged phosphate groups of nucleic acid. By increasing the N/P ratio, the size had reduced, whereas zeta potential increased. PDI increased on account of agglomeration due to too a high concentration of four-arm structures. Assembly of polyplexes was accomplished with rapid pipetting, 15-times, and subsequent incubation for 40 min at RT.
47
Table 9: Four-arm structure ID, main motif of four-arm structure, molecular weight and amount of total and protonatable amines
# Motif MW [g/mol] Amines in total Protonable
Since the size of polyplexes is a critical parameter for successful gene delivery and bigger particles are often prone to lower stability, the polyplex size was measured using Zetasizer Nano ZS. For more reliable results, polyplexes were formed two times with size and zeta
Since the size of polyplexes is a critical parameter for successful gene delivery and bigger particles are often prone to lower stability, the polyplex size was measured using Zetasizer Nano ZS. For more reliable results, polyplexes were formed two times with size and zeta