WO2012175617A1

WO2012175617A1 - Diblock copolymer

Info

Publication number: WO2012175617A1
Application number: PCT/EP2012/061974
Authority: WO
Inventors: Garcia ARMANDO HERNÁNDEZ; Rindert Jacob DE VRIES; Frederik Anton De Wolf
Original assignee: Stichting Dutch Polymer Institute
Priority date: 2011-06-24
Filing date: 2012-06-21
Publication date: 2012-12-27

Abstract

The invention relates to a diblock copolymer comprising a collagen-like block consisting of a first amino acid sequence of m amino acids and a DNA binding block consisting of a second amino acid sequence of n amino acids, wherein m and n is an integer and wherein the ratio between m and n is at least 20.

Description

DIBLOCK COPOLYMER

The invention relates to a diblock copolymer, its use as a coating for a nucleic acid and a nucleic acid coated by such a copolymer and to the use of the coated nucleic acid in gene therapy, non-viral gene delivery, sequencing by transport through nanopores, nanoelectronics such as nanowire DNA templating artificial virus or DNA origami..

DNA is no longer only a carrier of genetic information, but rather a multi-purpose material that is being explored for both biological and non-biological applications.

Coating DNA with binder moieties is an effective way to modify physical parameters and modulate its interactions with the surrounding environment¹'²'³. For example, non-viral gene delivery relies on complexation of DNA with cationic lipids and/or polymers to promote passage of DNA across cell membranes. A controlled DNA coating would be advantageous to regulate doses of transfected DNA and also to direct vector shape. Also, a protein coating on ssDNA has been used to facilitate passage through nanopores. Binder moieties could in principle be used for modulating many properties as bending rigidity, thickness, stiffness and stretching. All of these are crucial to other nanotechnological processes involving DNA, such as nanoscale self assembly ("DNA origami"), templating of nanowires, and transport in nano- channels for the stretching, separation and sequencing of DNA⁴'⁵'⁶'⁷.

Control of the architecture of the coating using DNA binder moieties is an initial key element needed. Only a few studies so far have focused on understanding the requirements for coating single DNA molecules using rationally designed polymers. Typical for the use of polymeric DNA binders is the formation of random aggregates of uncontrolled size and irregular shape consisting of many nucleic acids molecules per aggregate ⁸'⁹' ¹⁰. This limits their applicability when a precisely controlled coating is desired. In this sense, an understanding of the polymer design requirements that control monomolecular DNA coating with regular shape and uniform molecular weight could benefit the entire field of DNA-based nanobiotechnology.

The simplest design for a DNA-coating macromolecule would be a diblock structure containing a DNA binding domain, connected to another domain that should confer colloidal stability. A number of such designs have been published in which the binding domain is a simple flexible polycation. For example, Kostiainen et at . have conjugated various globular proteins with polycationic blocks and showed that the conjugates coat dsDNA. A number of studies have also been performed on synthetic diblocks consisting of polycationic domain coupled to a large uncharged and hydrophilic flexible domain. A judicious analysis of these studies indicates clearly that the ratio between shielding and cationic block lengths (symmetry) it is important for the final architecture of the DNA molecule-polymer complex. In most of these cases the binding domains were still quite large (on the order of 20 positive charges or more) leading to the formation of multimolecular aggregates by "bridging" several DNA molecules¹¹'¹²'¹³ and/or to the folding of single DNA molecules (intramolecular aggregation) creating a diverse population of shapes and sizes¹⁴'¹⁰. Intermolecular aggregation can be quite slow in these cases, and is typically enhanced at high DNA and salt concentrations¹⁵.

A few of these studies have claimed monomolecular DNA coating. Radler and coauthors ¹⁶ reported controlled coating of single DNA using PEG-PEI (50 positive charges). Borner and coauthors¹¹, used PEO-PAMAM (20 positive charges) to form a mixed population of single- plasmid toroids with different folded degrees and supercoiled dsDNA. Kataoka and coauthors reported a PEG-poly-L-Lys (19 positive charges) polymer that multi-fold single plasmid DNA¹². Intramolecular folding/aggregation is presumably driven by polycations bridging neighboring pieces of dsDNA. This type of bridging does not distinguish between inter- or intramolecular bridging (or folding). If intramolecular bridging occurs, it is therefore very likely that over time, or at high DNA concentrations, intramolecular aggregation will also occur. Despite the fact that they used much longer binding blocks (at a roughly similar size of the neutral block), Radler et. al.¹⁶ report neither intermolecular aggregation nor intramolecular folding. This may be related to the fact that they have used DNA that was quite short DNA and concentrations that were quite low.

It is an object of the present invention to provide a diblock copolymer that can coat DNA with a reduced degree of aggregation even at a higher concentration of DNA.

Accordingly, the present invention provides a diblock copolymer comprising a collagen-like block consisting of a first amino acid sequence of m amino acids and a DNA binding block consisting of a second amino acid sequence of n amino acids, wherein m and n each independently stand for an integer and wherein the ratio between m and n is at least 20.

The present invention also provides a diblock copolymer comprising a collagen-like block consisting of a first amino acid sequence and a DNA binding block consisting of a second amino acid sequence, wherein the ratio between the molecular weight of the first amino acid sequence and the molecular weight of the second amino acid sequence is at least 15.

It was surprisingly found that a high asymmetry between shielding-to-binding block lengths in the diblock copolymer according to the present invention results in the prevention of formation of random aggregates of uncontrolled size and shape over time. Decreasing binding block length into a minimal and increasing enough the asymmetry (difference of the length ratio between block lengths) between shielding/binding blocks will lead into a controlled single DNA coated rod assembly; reducing significantly the presence of populations with different shapes and thus homogenizing the size and shape of the coated single DNA molecules. The present invention provides a technology for producing a highly asymmetric neutral-cationic diblock protein diblock copolymer that accurately coats a long single dsDNA, which shows no signs of intermolecular aggregation even at high DNA concentrations. Preferably, the ratio between m and n is at least 25, more preferably at least 30. Preferably, the ratio between m and n is at most 100, more preferably 80, 60 or 50.

Binding of the diblock copolymer to the DNA template is achieved by a short stretch of basic amino acids, i.e. the binding block having a positive charge. The binding block needs to be short enough to avoid bridging but at the same time it should be long enough to bind DNA strongly. The collagen block has to be large enough to confer colloidal stabilization, while at the same time not constraining the binding. Hence, m is preferably from 300 to 600, preferably from 325 to 550, preferably from 350 to 500, preferably from 375 to 450 or from 390 to 425. Preferably, n is from 10 to 25, preferably from 11 to 20 or from 12 to 15

Preferably, the ratio between the molecular weight of the first amino acid sequence and the molecular weight of the second amino acid sequence is 15-35, more preferably 18-30, more preferably 20-25.

Preferably, the molecular weight of the first amino acid sequence is 30-50 kDa, more preferably 32-40 kDa. Preferably, the molecular weight of the second amino acid sequence is 1 - 2 kDa, more preferably 1.2-1.8 kDa.

In order to bridge dsDNA, the binding block needs to be at least larger than two dsDNA molecules diameter (>~4.8 nm). The binding block is preferably shorter than around 4.8 nm. This is at least larger than two dsDNA molecules diameter, which allows bridging of the dsDNA.

The protein polymer design according to the present invention can serve as a scaffold for a step-wise addition of extra functional blocks for a tighter control of shape and functionality.

The diblock copolymer may comprise (biofunctional) domains that give specific

functionalities, beyond merely covering the DNA. This can be used as modulators of nanotechnological processes involving DNA. An attractive technology to produce

biocompatible polymers with a built-in possibility to incorporate biofunctional domains is the use of genetic engineering to produce designed protein-polymers.

Accordingly, the present invention provides a copolymer comprising the diblock copolymer according to the present invention and a functional block comprising a third amino acid sequence attached to the first amino acid sequence or the second amino acid sequence of the diblock copolymer according to the present invention.

The diblock copolymer according to the present invention may have a collagen-like block consisting of the first amino acid sequence represented by SEQ. NO. 1 or a sequence having at least 80%, 90%, 95%, 96%, 97%, 98%, or 99%identity with SEQ ID NO. 1.

The second amino sequence preferably consists of (i) positively charged amino acids selected from the group consisting of lysine, arginine and histidine and (ii) one non-polar amino acid at the end of the sequence. More preferably, the second amino sequence consists of positively charged amino acids selected from the group consisting of lysine and histidine and (ii) one non- polar amino acid at the end of the sequence. Said non-polar amino acid at the end of the sequence may be chosen from the group consisting of glycine, alanine, valine, leucine, isoleucine, methionine and proline. Preferably, said non-polar amino acid is glycine for its short length.

The second amino acid sequence is preferably attached to the first amino acid sequence with its end having the positively charged amino acid.

The second amino acid sequence may be represented by SEQ. NO. 2 or SEQ. NO. 3.

Ki2 (12 Lys; plus G corresponds to SEQ. N0.2) and (HK)₆ (six His-Lys dyads; plus G corresponds to SEQ. NO.3) show a different pH dependence of their charge density. (HK)₆ is more preferred of the two, since it coats DNA faster than K12. This is thought to be due to possible intercalation of the cationic aromatic imidazole rings of histidines between the purine nucleic acids

According to another aspect of the present invention, a process for making the diblock copolymer is provided, the process comprising the step of expressing the nucleic acid sequence encoding said copolymer in Pichia pastoris using recombinant expression techniques known per se.

Preferably, said step is performed at pH of at least 2.5 and at most 6.0. This range of pH is suitable for growing Pichia pastoris cells. The step is preferably performed at pH of at most 5.5, at most 5.0, at most 4.5, at most 4.0 or at most 3.5. A particularly preferred range is pH of 2.5-3.5. This range of pH of 2.5-3.5 gives a protective effect against degradation to the block copolymer of the present invention.

The present invention further provides a double strand nucleic acid coated with the diblock polypeptide according to the present invention. The double strand nucleic acid may be ribonucleotide, deoxy(ribo)nucleotide or a modified form of either type of nucleotide.

Preferably, the double strand nucleic acid is DNA, which may be in the form of a vector or a linear DNA.

The present invention further provides use of the coated nucleic acid according to the present invention in gene therapy, non- viral gene delivery, sequencing by transport through nanopores, nanoelectronics such as nanowire DNA templating artificial virus or DNA origami.

The present invention further provides an isolated amino acid sequence SEQ ID NO. 1 or a sequence having at least 80%, 90%, 95%, 96%, 97%, 98% or 99% 99% identity with SEQ ID NO. 1. The present invention further provides a nucleic acid sequence encoding the amino acid sequence SEQ ID NO.l or a sequence having at least 80%, 90%, 95%, 96%, 97%, 98% or 99%identity with SEQ ID NO. 1.

The present invention further provides a nucleic acid sequence encoding the diblock copolymer according to the present invention.

The skilled person will be aware of the fact that several different computer programs are available to determine the homology between two sequences. For instance, a comparison of sequences and determination of percent identity between two sequences can be accomplished using a mathematical algorithm. In a preferred embodiment, the percentage identity between two amino acid sequences is determined using the Needleman and Wunsch (J. Mol. Biol. (48): 444-453 (1970)) algorithm which has been incorporated into the GAP program in the Accelrys GCG software package (available at http://www.accelrys.com/products/qcq/), using either a Blossom 62 matrix or a PAM250 matrix, and a gap weight of 16, 14, 12, 10, 8, 6, or 4 and a length weight of 1, 2, 3, 4, 5, or 6. The skilled person will appreciate that all these different parameters will yield slightly different results but that the overall percentage identity of two sequences is not significantly altered when using different algorithms.

In yet another embodiment, the percent identity between two nucleotide sequences is determined using the GAP program in the Accelrys GCG software package (available at http://www.accelrys.com/products/gcg/), using a NWSgapdna.CMP matrix and a gap weight of 40, 50, 60, 70, or 80 and a length weight of 1, 2, 3, 4, 5, or 6. In another embodiment, the percent identity of two amino acid or nucleotide sequence is determined using the algorithm of E. Meyers and W. Miller (CABIOS, 4: 1 1 -17 (1989) which has been incorporated into the ALIGN program (version 2.0) (available at the ALIGN Query using sequence data of the Genestream server IGH Montpellier France http://veqa.iqh.cnrs.fr/bin/aliqn-quess.cqi) using a PAM120 weight residue table, a gap length penalty of 12 and a gap penalty of 4.

Based on the present description and possibly in combination with e.g. , Sambrook and Russel, Cold Spring Harbor Laboratory Press, 2001 hereby incorporated by reference, the skilled person can produce the copolymer of the present invention using the recombinant expression techniques described herein.

Figure 1 shows design of diblock copolypeptides for nucleic acid binding, (a) Schematic modular design, green is the colloidal stability collagen-like block and red is the DNA binding block, (b) Model of protein diblock copolymer as a random coil, (c) Amino acid sequence of the colloidal stability block and of the two used binding blocks.

Figure 2 shows molecular mass characterization of protein diblock copolymers, (a) SDS- PAGE image: MM: molecular markers, lane 1 : C4K12 harvested medium at the end of the fermentation, 2: Purified C4K12 1 g/L, 3: Purified C4K12 0.1 g/L, 4: C4HK6 harvested medium at the end of the fermentation, 5: Purified C4HK6 sample 1 g/L 6: Purified C4HK6 sample 0.1 g/L. (b) MALDI-TOF of purified C4K12. The molecular weight for M+ is 38,586, M2+ 19,305 and 2M+ 76,967. (c) MALDI-TOF of purified C4HK6. The molecular weight for M+ is 38,432, M2+ 19,228.

Figure 3 shows electrophoretic mobility of DNA-Protein diblock copolymer complexes on 1% agarose gel. Top row shows N/P ratio. Lane 1 : DNA molecular markers, 2: pDNA alone (N/P = 0).

Figure 4(a). shows light scattering intensity for DNA-protein diblock copolymer complexes formed at different N/P. # C4K12 complexes, * C4HK6 complexes.

Figure 4(b) shows DLS intensity and hydrodynamic radius for complexes formed between diblock copolypeptides and pDNA at different N/P values or salt conditions, (a) C4K12-pDNA nanoplexes measured at Θ = 12.8°. (b) C4HK6-pDNA nanoplexes measured at Θ = 12.8°

Figure 5 shows disassembly of DNA-protein nanoplexes by salt addition, a) Light scattered intensity by size distribution of C4K12-pDNA nanoplexes (N/P = 1.1) at different salt concentrations. Appearance of free diblock dynamics of the nanoplexes was followed, b) Light scattered intensity of complexes at different salt concentrations.

Figure 6 shows typical atomic force microscopy (AFM) images and height profiles of rod nanoplexes formed between diblock copolypeptides and 3.7 kbp pDNA a) C4K12-pDNA at N/P = 2.2 b) C4HK6 -pDNA at N/P = 1.5. Scale is 2 μιη.

Figure 7 shows atomic Force Microscopy in air of deposited single pDNA molecules complexed with protein diblock copolymers at different N/P ratios. Top row C4K12-pDNA nanoplexes: (a) N/P = 0.95, (b) N/P = 2.2. (c) N/P = 8.3. Bottom row C4(HK)6-pDNA nanoplexes (d) N/P = 0.4, (e) N/P = 1.5, (f) N/P = 8.5. Bar scale 200nm.

Figure 8 shows coated Linear DNA with C4K12 protein diblock copolymer. Scale bar: 200 nm. L is 1258 nm. 0.45 nm height, 24nm width. In mica.

Figure 9 schematically illustrates side -chain stretching.

Figure 10 schematically illustrates coated supercoiled DNS. Experiments

The choices of the DNA binding block and the collagen-like block were made such that the DNA binding block was shorter (12 amino acids) than the shortest ones reported and the collagen-like block was longer (400 amino acids) than the longest ones reported (see Table 1). Thermodynamically stable single-DNA complexes were obtained. The single-DNA complexes were not only kinetically stable at low concentrations or short times, but were stable in conditions that ultimately will show either intermolecular bridging or intramolecular bridging (folding) if the copolymer according to the present invention is not used. Table 1

Binding Block Colloidal Block

[DNA] Reference

Type n₊ Type Mw

-50

Branched PEI PEG 20 kDa <0.33 ng^L 1

2 kDa 20

polyamidoamine PEO 2.7 kDa 2

2.6 kDa

Poly-L-Lys 17 PEG 12 kDa 33.3 ng/uL 3

12 Collagen¬

Lys 36 kDa 43 ng/uL This study

-1.5 kDa like

12 Collagen¬

His-Lys 36 kDa 43 ng/uL This study

-1.5 kDa like

1. DeRouchey, J.; Schmidt, C; Walker, G. F.; Koch, C; Plank, C; Wagner, E.; Radler, J. O. Biomacromolecules. 2008, 9, 724-732¹⁶.

2. Hartman, L.; Hafele, S.; Peschka-Suss, R.; Antonietti, M.; Borner, H. G. Chem. Eur. J. 2008, 14, 2025-2033^u.

3. Osada, K.; Oshima, H.; Kobayashi, D.; Doi, M.; Enoki, M.; Yamasaki, Y.; Kataoka, K. JACS. 2010, 132, 12343-12348¹⁴.

Biosynthesis of asymmetric cationic diblock copolypeptides. The two diblock protein copolymers were successfully produced using recombinant yeast strains of P. pastoris carrying the artificial genes coding for their amino acid sequences, prepared by molecular cloning. After simple and scalable steps of differential precipitation (see materials and methods) lyophilized dib locks were recovered with a yield up to lg/L of cell-free medium. The precipitation steps were chosen in such a way that they did not precipitate the low concentrations of extracellular P. pastoris proteins present in the medium, but only the diblock protein copolymers. Note that biosynthesis of cationic block polypeptides has already been reported in bacterial systems such as E. coli , but not in yeast ¹⁷'¹²'¹⁸. As far as we know, this is the first report of the production of cationic block copolypeptides in eukaryotic yeast cells with high yields. Biosynthesis of highly cationic polypeptides has been difficult since they are generally toxic because they bind to the dominantly negatively charged biomolecules inside the cell, disrupting biological functions.

Molecular characterization of diblock copolypeptides. Purity and molecular weight of C₄- K 2 and C4(HK)₆ after purification were characterized using SDS-PAGE and MALDI-TOF (figure 2). SDS-PAGE showed well defined single bands for both purified diblocks that corresponded with the dominant bands in the lane of the harvested medium. The molecular weight of the purified polypeptides as determined by SDS-PAGE (98kDa) does not correspond to the expected value (38 kDa). The reduced mobility is caused by the very low content of hydrophobic amino acids in C4 block that leads to poor binding of SDS. For C4(HK)₆ a second upper band is observed. This may corresponded to a multimer, since the same band is not present in the lane of the SDS-PAGE corresponding to the harvested medium sample.

Accurate experimental molecular weights for the diblocks protein polymers have been determined using MALDI-TOF. For C₄K₁₂ three signals are found (19,306, 38,587 and 76,967 Da) that correspond respectively to M2+ M+ 2M+. For C₄(HK)₆ , two signals are found (19,228 and 38,433 Da) corresponding to M2+ and M+, respectively. The MALDI-TOF sizes are in very good agreement with the expected theoretical sizes of -38,407 and 38,461 Da for C4K42 and C₄(HK)₆, respectively. Small deviations are possible, since Glu-Ala repeats can be left when pre -pro signal is removed from the diblock sequence before being secreted. The possible presence of these extra amino acids has no effect on the DNA binding properties, since they are located on the opposite side of the proteins.

Electrophoretic mobility of DNA-polypeptide complexes. Agarose gel mobility retardation was used as an initial assay for the interaction between the two protein diblock copolymers and plasmid DNA. Figure 3 shows that electrophoretic mobility of DNA is reduced when diblock copolypeptides are added. The effect increases at higher binding block to DNA phosphate ratio (N/P), and saturates at about N/P =5. This behavior is similar to that of some cationic polypeptides also carrying a short binding block, where a total retardation was not observed either. In contrast, less asymmetric neutral-cationic diblock polymers completely neutralize DNA charge or even reverse the mobility ^u'¹⁶. As a control, we have also verified that there is no mobility shift when DNA is mixed with C4 block alone. Although the net charge of the resulting complexes was almost neutral at high N/P it never became positive. Presumably, this is because for the highly asymmetric diblock, the coating is not dense enough to neutralize all the charge even at high N/P. Finally, the outside C₄ blocks are weakly negatively charged, and this may also play a role in determining the observed mobility of the complexes. (The predicted isoelectric point of the C₄ block is pi = 4.9).

Light Scattering studies. Light Scattering was used to further study the interaction between DNA and the diblock copolypeptides, since it is a convenient technique for obtaining solution sizes and molecular weight estimates for the complexes that does not require any labeling. Figure 4(a) shows the absolute intensity of scattered light (or Rayleigh ratio, see materials and methods) Rg of the complexes as a function of the N/P ratio. For both copolypeptides, the scattered intensity increased rapidly at N/P < 1 , and then saturates at N/P » 1. The scattered intensity of the fully covered DNA was about a factor 100 larger than that of the naked DNA. For a rough estimate of the degree of coverage, it may be assumed that the scattering at N/P»l is dominated by the complexes and that the contribution of excess free polymer to the scattering can be neglected. Assuming full charge neutralization, the predicted molar mass M of a single coated DNA molecule is

M = (1 + «)M DNA

where M_DNA and M_p are respectively, the DNA and polymer molar mass, n_p is the number of basepairs covered by a single bound C₄B₁₂ polymer, and M₀= 600 Da is molar mass of a single basepair. Assuming n_p = 6bp (one C₄B₁₂ polymer per 12 DNA phosphate charges or 6 basepairs), gives a = 10.7. In the limit of zero scattering angle and zero concentration, the predicted absolute scattering intensity R_g for the complexes is

R_g = K_RCM

where C = (1 + a)C_DNA is the weight concentration of complexes. Literature values for the refractive index increments for dsDNA and the C₄B₁₂ polymers are within the same range of dn/dc = 0.16...0.18 ml/g (see materials and methods). Using a single average value of dn/dc = 0.17 gives a value of the Rayleigh constant of K_R = 2.105-10 ⁵ mol m² kg^"2 (see materials and methods). Roughly speaking, both C and M increase by a factor of about 10 by the polymer coating, such that the light scattering increases by a factor of about 100, as observed. The estimated absolute scattering intensity of = 0.319 m^"1 for charge neutral complexes is indicated as a dashed line in Fig 4a, and lies somewhat above the limiting values for both types of complexes. Consistent with the results of the mobility shift assay, this indicates that at saturation, for both complexes, the majority, but not all of the DNA phosphate charges have been neutralized by complexation with binding domains.

Figure 4(b) shows the effective hydrodynamic radii (R_h) determined using DLS, as a function of N/P. For C4HK6, the size rapidly increases from 75 nm in the absence of polymer, to 150 nm at N/P = 1.5. For C4K12 + DNA, the final size is somewhat smaller, Rh = 135 nm at N/P = 1.5

These values are close to values for similar bulky cationic polypeptides¹⁸. Finally, for the individual diblocks, in the absence of DNA, the scattered intensity was very low. The hydrodynamic radius for the diblocks was 5.4 nm for C4K 2 and 6.1 nm for C4(HK)₆. It is important to note that the reported effective R_h of the complexes is calculated from a measured diffusion D constant assuming a spherical particle size,

where k_BT is the thermal energy, and η is the solvent viscosity. A 3.7kb supercoiled plasmid, both with and without adhering diblock polymers, presumably is closer to a rod-like than to a spherical particle. Assuming the diffusion of both naked and coated supercoiled plasmid can be approximated by that of a cylindrical particle with length L and diameter d, the predicted diffusion constant is:

3^L

A= ln(L/ d) + 0.312 + 0.565 /(L/ d) - 0.1/(L/ d)²

This approximation was used by deRouchey et al.¹⁶ for short (rod-like) linear DNA (<lkb) coated with cationic -neutral diblock polymers, to extract an estimated thickness d of the coated DNA, using the known length L. For a very rough estimate of the effective thickness of the (coated) plasmid "cylinder", we use L = 650 nm, which is half the contour length of 3.7kb DNA. Then, for uncoated DNA we find d = 12nm, which is not unreasonable for the diameter of a plectonemic supercoil. For the fully coated supercoils we find i =87 nm (assuming R_h = 135nm for C₄K₁₂) and d=\ 12nm (assuming R_h = 150 nm for C₄(HK)₆).

Charge Screening by salt addition. To confirm that complexes are formed due to electrostatic interactions, N/P =1.1 complexes were exposed to increasing concentration of salt (NaCl) and their disassembly was followed using DLS (figure 5). These experiments were carried out at a larger scattering angle of 173° , corresponding to a larger value of the wavevector q (see materials and methods) to allow for a more sensitive detection of the diffusion of the much smaller free polymers. Figure 5. a shows that the scattering intensity decreases rapidly for salt concentrations in between 0.1M and 0.25 M. Figure 5.b shows the size distributions as determined using DLS. The contribution due to diffusion of free diblocks (R_H = 6 nm) starts to become visible at 0.25 M, implying that at this concentration, most diblocks are no longer bound to the DNA. Above 0.4M the scattering intensity is constant, implying that no polymers are bound anymore. These experiments demonstrate that the complexation is clearly electrostatic in nature. This also means that further tuning of the binding strength may be achieved via changes of pH that affect the charge on the binding blocks, and that makes the design can be tailored to specific pH dependency requirements.

Atomic Force Microscopy. AFM studies in air were carried out to determine the morphology of the complexes and to confirm that complexes consist of individual DNA molecules coated by diblock copolymer. We have found that the collagen-like C₄ blocks rapidly adsorb on the mica and silica surfaces that we have used as a substrate for AFM. Whereas naked DNA does not adhere on on bare mica or silica, all of the complexes did without requiring divalent cations or poly-Lysine. This is advantageous because it this excludes any possible influence of these additives on the morphology of the adhering DNA complexes.

The figure 6 shows typical images of C4-K12 complexes at N/P=1.5 and C4HK6-DNA complexes at N/P = 2.2. For both cases the complexes are well dispersed. They appear as semiflexible single rods of regular shape. The average contour length of the nanoplexes at N/P = 2.2 is 649.4 ± 15 nm for C4K12 and 624.7 ± 29.3 for C4HK6 (N = 20). The height for both complexes is very uniform with an average around 1 nm (1.07 ± 0.1 and 0.99 ± 0.2 (N = 20) for C4K12 and C4HK6, respectively).

A series of images of individual complexes at increasing values of N/P (for both polymers) is shown in Fig. 7. Clearly, for both polymers, complexes become thicker, stiffer and more homogeneously coated at higher N/P. At low N/P, complexes appear thin and less rigid, and an occasional loop of the supercoiled DNA molecules can still be seen. C4HK6-DNA complexes appear thicker at lower N/P as compared to C4K12 complexes: N/P = 0.4 complexes of C4HK6 are similar to N/P = 1 complexes of C4K12. Also, both types of complexes continue to increase in thickness for N/P > 1. Both findings are in agreement with the light scattering and electrophoresis results.

A good test for the absence of intramolecular folding is to compare with complexes of linearized pDNA: this should give a contour length exactly equal to the contour length of the 3.7kb pDNA (L_DNA = 1258 nm). A representative image of a C4K12- linearized pDNA complex at N/P = 1 is shown in Figure 7. For this case the contour length of the complexes indeed equals that of the naked DNA, indicating that individual DNA molecules get coated, and that there is no intramolecular folding due to bridging of the binding blocks.

In conclusion, the AFM measurements are completely consistent with the light scattering measurements and indicate that the complexes consist of individual DNA molecules, coated with diblock copolymers: contour lengths for the pDNA complexes are in a very good agreement with the theoretical half contour length for 3.7 kbp DNA (-629 nm, 0.34 nm/bp), and for the complexes with linearized pDNA the contour length equals the full contour length of the naked DNA, as expected. Furthermore, the observed heights are very uniform, which would not have been the case if there had been a spread in the number of dsDNA chains per complex.

Numerical Estimates. DNA has two charges every 0.34 nm (one bp), such that the bare linear charge density v of dsDNA (elementary charges e per unit length) is v~ 5.9 nm^"1.

Assuming a single B₁₂ block neutralizes 12 DNA phosphate charges, it covers an estimated length of dsDNA of /z_min = 2 nm. DNA-bound diblock copolypeptides start feeling each other when they are bound at a spacing h along the DNA contour smaller than the coil size of the C4 block. This occurs at a critical spacing h* and critical degree of neutralization J*

h^* « 2R ¾ 12nm

f ^ h^ /h^ O.17

where R ~ 6 nm is the coil radius of the C4 block, as determined using DLS. For N/P > f, electrostatic binding of the diblocks is opposed by steric repulsion of neighbouring C4 blocks. As is clear from the experimental data, binding continues to increase well beyond N/P =f, and this will lead to stretching of the C₄ blocks. When the free energy penalty due to side-chain stretching exceeds the binding free energy of the B12 block, no more additional polymers are bound, even if not all of the DNA charges have been neutralized. Both the electrophoresis and the static light scattering suggest that this is the case for our diblocks at N/P » 1.

The side -chain stretching has been discussed extensively in the literature on so-called bottle- brush polymers, consisting of a main chain covered with regularly spaced side chains¹⁹.

Assuming good solvent conditions, the scaling estimate for the degree of stretching of the grafted side chains (each consisting of N segments of length /) is:

ϋ oc r^1/4

R

where R_s is the mean-square end-to-end distance of the stretched chain and R is the mean- square end-to-end distance of the unstretched chain. Note that in the absence of a numerical prefactor for this scaling estimate, it can only be used to get order-of-magnitude estimates. Assuming / = 0.5 nm and N=400, for the fully charge -neutral complex with h=h_min =2nm, we find Γ^{1 4} « 3. This implies that there is only moderate stretching. From dynamic light scattering, we inferred diffusion diameters the coated supercoils of D « 100 nm. Assuming the diffusion diameter of the coated supercoils is about 4 times the radius R_s of the stretched side- chains (see sketch), this gives R_j, « 25 nm, or a stretching factor R_j, / R « 4 , which is of the expected order of magnitude.

A further issue that is highly debated in the literature on bottle-brush polymers, is the stiffening of the main chain, induced by the side chains. Current simulations suggest that at least for relatively short side -chains (N < 100) the contribution l_PtS of the side chains to the total persistence length of dense bottle-brush polymer is typically on the order of its thickness.

According to numerical self-consistent field calculations, very strong stiffening only occurs for densely grafted, extremely long side chains ( N» 1000): l_p,s * // - ^ N²

This is mainly due to a very small numerical prefactor, μ « 0.02 . For our case (again considering the limiting case of assuming full charge neutralization, h = 2 nm), the latter equation would imply 1 _s « 100 nm for a single dsDNA. This order-of-magnitude estimate seems to be consistent with the images of the fully coated 3.7kb supercoiled plasmid molecules, that are close to rod-like. Considering the entire supercoil as a semiflexible object, its estimated persistence length would increase to about 300nm after coating, about half the total supercoil contour length of about 650nm, such that is should indeed appear as close to rod-like.

Conclusions

Light scattering confirmed that the DNA was successfully coated with the diblock copolymer of the present invention.

Charge screening confirmed that complexes are formed due to electrostatic interactions.

AFM studies confirmed that complexes consist of individual DNA molecules coated by diblock copolymer rather than aggregates.

Materials and Methods

Materials, circular supercoiled plasmid DNA (3.7 kbp) used for binding studies was prepared using a Megaprep kit (Qiagen, Germany) from a recombinant E. coli broth culture grown overnight. The purified DNA was dissolved in sterile MQ-water for a final concentration of 232 ng^L (Nanodrop ND-1000 Spectrophotometer; Thermo Scientific, Waltham, MA). Portions were diluted till a final concentration of 116 ng^L in 10 mM acetate buffer at pH 5 and filtrated (0.2μιη) for DLS analysis. Restriction enzymes used for molecular cloning of the cationic protein polymers were purchased from New England Biolabs (Ipswich, MA) or Fermentas (ON, Canada). Acetate buffer was prepared from acetic acid and sodium acetate of analytical quality purchased from Sigma (St. Louis, MO).

Cloning of cationic protein polymers. C4 block and binding blocks were built separately and later put together. Two different cationic polypeptides C4-K12 and C4-HK6 were produced. "C4" refers a 4-fold repeat of a collagen- like block of approximately 100 amino acids (SEQ ID NO.l) and "B12" a binding block of twelve basic amino acids length either Lys ( 12) (SEQ ID NO. 2) or the dyad His-Lys [(HK)₆] (SEQ ID NO.3). The Collagen block (C4) was built it from pMTL23-C2 plasmid by directional ligation. The plasmid pMTL23-C2 was doubly digested with Van91I + Dralll restriction enzymes and the C2 fragment was isolated. Then, it was ligated into a previously Van91I digested/dephosphorylated pMTL23-C2 plasmid obtaining pMTL23-C4 where the Dralll restriction site is lost and only Van9 II kept in both sites of the duplicated segment. The binding blocks were build using two complementary oligonucleotides (Eurogentec, Belgium) codifying for the particular sequence of each block. They both carried out sticky ends for Van911 and EcoRI restriction enzymes. They were annealed by heating at 90 °C for 5-10 min and left at room temperature for 2-3 h. Upon this, they were ligated into the plasmid pMTL23-C4 previously digested with EcoRI + Van9 II, obtaining the plasmids pMTL23-C4-K12 and pMTL23-C4-HK6. These plasmids were propagated into E. coli after transformation using Z-competentTM E. coli transformation kit (Zymo Research Corporation). Plasmid sequencing and PCR-colony using Ml 3 primers were carried out to verify presence of intact blocks genes in the correct reading frame.

Fragments codifying for the corresponding C4-K12 (SEQ ID NO.l and SEQ ID NO.2) and C4HK6 (SEQ ID NO.l and SEQ ID NO.3) cationic protein polymers were obtained from pMTL23-C4-K12 and pMTL23-C4-HK6 through digestion with Xhol + EcoRI and ligated into pPIC9 plasmid previously digested with the same restriction enzymes. The resulting plasmids pPIC9-C4-K12 and pPIC9-C4-HK6 were linearized with Sail and electroporated into Pichia pastoris strain GS115 (Mut⁺). The plasmid integrates into the genome through homologous recombination at the his 4 locus providing high genetic stability of the resulting recombinant strains and leaves the cationic protein polymers genes expression under the control of AOX1 promoter enabling growth on methanol. Transformed yeasts were poured into Minimal Dextrose agar medium plates (1.34% Yeast Nitrogen Base, 2% dextrose, 4xl0^"5% biotin and 2% agar) and grew for 2-3 days at 30 °C. Some recombinant colonies were re-streaked and grown on plates with fresh medium. The presence of the genes was verified by PCR using specific primers for the AOX1 flanking sequences (5-AOX1 : GACTGGTTCCAATTGACAAGC (SEQ ID NO. 4) and 3-AOX1 : GCAAATGGCATTCTGACATCC (SEQ ID NO. 5)). Positive recombinant colonies were grown in YPD medium (1% yeast extract, 2% peptone plus 2% dextrose) until reaching an optical density (λ = 600 nm ) between 5-10 to be dissolved into 2/3 of 80% glycerol and stored at -80oC in vial ready-to-use for fermentation.

Cationic Protein Polymers Biosynthesis. Fed-batch fermentations using minimal basal salts medium were performed in 2.5-L Bioflo 3000 fermenters (New Brunswick Scientific, Edison, NJ). The methanol fed-batch phase for protein production lasted two to three days. A homemade semiconductor gas sensor-controller was used to monitor the methanol level in the off-gas and to maintain a constant level of 0.2%> (w/v) methanol in the broth. The pH was maintained at 3.0 throughout the fermentation by addition of ammonium hydroxide. At the end of the fermentation, the cells were separated from the broth by centrifugation for 15 min at 10000 x g (room temperature or 4 °C) in an SLA-3000 rotor (Thermo Scientific, Waltham, MA), and the supernatant was microfiltered (Pall Corporation, Port Washington, NY) to avoid possible proteolysis by remaining cells and immediately stored at 4 °C for subsequent purification.

Cationic Protein Polymers Purification. All centrifugations were done for 30 min at 20000 x g at 4 °C, interchangeably in a Sorvall SLA-1500 or SLA-3000 rotor. First, medium salts impurities were removed from the cell-free broth solution by precipitation by adjusting the pH to 8.0 with NaOH 1M followed by centrifugation. After this the diblock copolypeptides were selectively precipitated from the solution meanwhile P. pastoris proteins stayed in solution by adding ammonium sulphate till a saturation of 45% m/v and incubating over night at 4 °C. The pellet was resuspended in an equal volume (relative to the cell-free broth) of Milli-Q water and precipitation was repeated once at 4°C for 1-2 hrs. The pellet was resuspended in 0.2 volumes (relative to the cell-free broth) of Milli-Q water and sodium chloride (50 mM) and acetone (40% v/v) were added. After centrifugation the pellet was ruled out and acetone concentration was raised to 80% (v/v) and solution was centrifuged to separate the pellet containing the pure diblock copolypeptide. The o/n air-dried pellet was resuspended in Milli-Q water, extensively desalted by dialysis against Milli-Q water and lyophilized.

SDS-PAGE and Mass Spectrometry Analysis. SDS-PAGE was carried out using the NuPAGE No vex system (Invitrogen, Carlsbad, CA) with 10% Bis-Tris gels, MES SDS as running buffer and SeeBlue Plus2 prestained molecular mass markers. Gels were stained with Coomassie SimplyBlue SafeStain (Invitrogen). Matrix-assisted laser desorption ionization time- of-flight mass spectrometry (MALDI-TOF MS) was carried out in an Ultraflex mass spectrometer (Bruker, Billerica, MA). Proteins samples (lmg/mL) were prepared by the dried droplet method. The matrix was made up of 5 mg/mL 2.5-dihydroxyacetophenone, 1.5 mg/mL diammonium hydrogen citrate, 25% (v/v) ethanol, and 1% (v/v) trifluoroacetic acid on a 600 μιη AnchorChip target (Bruker). Measurements were made in the positive, linear mode. An external mass calibration was done based on Protein Calibration Standard II (Bruker).

Agarose Gel Mobility Retardation Assay. 46 ng of pDNA dissolved in 10 mM acetate buffer pH 5 were mixed with different volumes of C4-B12 lg/L solution dissolved in the same buffer for a final volume of 15 μΕ. After 30-60 min at room temperature they were poured in a 1% agarose gel and run at 90v for 60 min using TAE IX buffer. Bands were visualized using ethidium bromide.

Dynamic Light Scattering (DLS). Light scattering measurements were performed with a Zetasizer NanoZS apparatus (Malvern Instruments, UK) equipped with a 4mW He-Ne ion laser working at a wavelength of 633nm. DNA-cationic protein complexes were formed at room temperature mixing 5.6 μΕ of DNA (116ng^L) dissolved in buffer together with certain volume of desalted and purified C4-B12 (1-10 mg/mL) dissolved in the same buffer necessary to obtain the particular N/P ratio to be analyzed. DNA-diblock copolypeptide premixed solutions were filled up till 15 μΕ with 10 mM acetate buffer pH 5 and mixed. DNA, protein and buffer solutions were previously filtrated using 0.2μιη cut-off filters for all the complexing experiments. The resulting solutions were left to equilibrate for 1 -2 hrs at room temperature before DLS measurements. The intensity of scattered light by the DNA-cationic protein polymers complexes previously formed was determined from an average of 5 correlation measurements followed by 60 seconds at angle of 12.8 degrees and/or 173 degrees at 25°C. Effective hydrodynamic radius was automatically gotten from the apparent diffusion coefficient calculated by the apparatus. Atomic Force Microscopy. DNA-diblock copolypeptide complex solutions (prepared the same way as for DLS experiments) were diluted 10 or 20 times with filtrated Milli-Q water. Immediately 3-5 μΐ. of the resulting solution was added onto a clean hydrophilic lxl cm silicon wafer and left for 1 -2 min. Then it was washed with 500 μΐ. of filtrated Milli-Q water to remove salts and not absorbed particles followed by tissue removal of water excess and a slow drying by nitrogen stream. Samples were analyzed using a Digital Instruments NanoScope V equipped with a silicon nitride probe with a spring constant of 0.32 N/m in ScannAsyst mode (Veeco, NY, USA). Images were recorded between 0.488-0.965 Hz and 384-1024 samples/line. Image processing was done with NanoScope Analysis software. Contour length and long axis length measurements were performed with ImageJ software.

REFERENCES

1. Hegner M, Smith S. B., Bustamante, C. PNAS, 1999.

2. Hall AR, van Dorp S, Lemay SG, Dekker C. Nanoletters, 2009.

3. Kostiainen MA, Szilvay GR, Lehtinen J, Smith DK, Linder MB, Urtti A, Ikkala O. ACS nano, 2007, 1.

4. Douglas S. M.; Dietz, H.; Liedl, T.; Hogberg. B.; Graf. F.; Shih, W. Nature. 2009, 459.

5. Maubach G, Born D, Csaki A, Fritzsche W. Small, 2005, 1.

6. Keren K, Berman RS, Buchstab E, Sivan U, Braun E. Science, 2003.

7. Smeets RMM (Smeets, R. M. M.)l, Kowalczyk SW (Kowalczyk, S. W.)l, Hall AR

(Hall, A. R.)l, Dekker NH (Dekker, N. H.)l, Dekker C (Dekker, C.)l nanoletters, 2009, 9.

8. Walker, G. F.; Fella, C; Pelisek, J.; Fahrmeir, J.; Boeckle, S.; Ogris, M.; Wagner, E. molecular therapy, 2004, 11.

9. Martin, E. M.; Rice, K. G. Peptide-guided gene delivery. AAPS journal, 2007, 9.

10. J.K.W. Lama, Y. Mab, S.P. Armesb, A.L. Lewisc, T. Baldwina and S. Stolnika, J. control release, 2004.

11. Hartmann, L.; Hafele, S.; Peschka-Suss, R.; Antonietti, M.; Borner, H. G. Chem. Eur.

J. 2008, 14, 2025-2033.

12. Numata, K.; Subramanian, B.; Currie, H. A.; Kaplan, D. L. Biomaterials. 2009. 30,

5775-5784.

13. Grayson, A. C; Doody, A. M.; Putnam, D. Pharm Res. 2006, 23, 1868-1876.

14. Osada, K.; Oshima, H.; Kobayashi, D.; Doi, M.; Enoki, M.; Yamasaki, Y.; Kataoka, K. JACS. 2010, 132, 12343-12348.

15. Mann, A.; Richa, R.; Ganguli, M. J Control Release. 2008, 125, 252-262.

16. DeRouchey, Walker, G. F.; Wagner, E.; Radler, J. O. J. phys. Chem B. 2006, 110, 4548-4554. 17. Canine, B. F.; Wang, Y.; Hatefi, A. J Control Release. 2008, 129, 117-123.

18. Zeng, J.; Wang. S. Biomaterials. 2005, 26, 679-686.

19. Hsu, H.; Wolfgang, P.; Kurt, B. Macromolecular theory and simulations, 2007.

SEQ ID NO. 1.

GPPGEPGNPGSPGNQGQPGNKGSPGNPGQPGNEGQPGQPGQNGQPGEPGSNGPQGSQG NPGKNGQPGSPGSQGSPGNQGSPGQPGNPGQPGEQGKPGNQGPAGEPGNPGSPGNQG QPGNKGSPGNPGPGNEGQPGQPGQNGQPGEPGSNGPQGSQGNPGKNGQPGSPGSQGSP GNQGSPGQPGNPGQPGEQGKPGNQGPAGEPGNPGSPGNQGQPGNKGSPGNPGQPGNE GQPGQPGQNGQPGEPGSNGPQGSQNPGKNGQPGSPGSQGSPGNQGSPGQPGNPGQPGE QGKPGNQGPAGEPGNPGSPGNQGQPGNKGSPGNPGQPGNEGQPGQPGQNGQPGEPGS NGPQGSQGNPGKNGQPGSPGSQGSPGNQGSPGQPGNGQPGEQGKPGNQGPAGG SEQ. ID NO. 2

KKKKKKKKKKKKG

SEQ. NO. 3

HKHKHKHKHKHKG.

SEQ.NO. 4

GACTGGTTCCAATTGACAAGC SEQ.NO.5

GCAAATGGCATTCTGACATCC

Claims

1. A diblock copolymer comprising a collagen- like block consisting of a first amino acid sequence of m amino acids and a DNA binding block consisting of a second amino acid sequence of n amino acids, wherein m and n each independently stand for an integer and wherein the ratio between m and n is at least 20.

2. The copolymer according to claim 1, wherein m is from 300 to 600, preferably from 325 to 550, preferably from 350 to 500, preferably from 375 to 450 or from 390 to 425.

3. The copolymer according to claim 1 or 2, wherein n is from 10 to 25, preferably from

11 to 20 or from 12 to 15.

4. The copolymer according to any one of the preceding claims, wherein the first amino acid sequence is represented by SEQ ID NO. 1 or a sequence having at least 80%, 90%, 95%, 96%, 97%, 98% or 99% identity with SEQ ID NO. 1.

5. The copolymer according to any one of the preceding claims, wherein the second amino acid sequence is represented by SEQ ID NO. 2 or SEQ ID NO. 3, or a sequence having at least 80%, 90%, 95%, 96%, 97%, 98%, 99% identity with SEQ ID NO.2 or SEQ ID NO. 3.

6. A process for making the diblock copolymer according to any one of the preceding claims, the process comprising the step of expressing the nucleic acid sequence encoding said copolymer in Pichia pastoris using recombinant expression techniques known per se.

7. The process according to claim 6, wherein said step is performed at pH of at least 2.5 and at most 6.0, preferably at most 5.0, at most 4.5, at most 4.0 or at most 3.5.

8. A copolymer comprising the diblock copolymer according to any one of the preceding claims and a functional block comprising a third amino acid sequence attached to the first amino acid sequence or the second amino acid sequence of the diblock copolymer according to the present invention.

9. A double strand nucleic acid coated with the diblock polypeptide according to any one of claims 1-5 or with the copolymer according to claim 8.

10. Use of the coated nucleic acid of claim 9 in gene therapy, non-viral gene delivery, sequencing by transport through nanopores, nanoelectronics such as nanowire DNA templating artificial virus or DNA origami.

11. A nucleic acid sequence encoding the diblock copolymer according to any one of claims 1 -5 or the copolymer of claim 8.