EP0958380A1 - Macromolecule-lipid complexes and methods for making and using - Google Patents

Abstract

The invention provides novel compositions involving macromolecule-lipid complexes and method for making them. These compositions and methods of the invention are significant improvements in the field of macromolecule-lipid complex processing, macromolecule targeting and delivery to various biological systems.

Description

MACROMOLECULE-LIPID COMPLEXES AND METHODS FOR MAKING AND USING

This application is claiming priority under 35 U.S.C. §119(e) of provisional application U.S. Serial No. 60/032,163, filed December 6, 1996.

This invention was made with Government support under NSF grants DMR-9624091 and DMR-9632716. The Government has certain rights in this invention.

Throughout this application various publications are referenced. The disclosures of these publications in their entireties are hereby incorporated by reference into this application in order to more fully describe the state of the art to which this invention pertains.

BACKGROUND OF THE INVENTION

Conventional macromolecule delivery and release technologies, which in the past have concentrated on improvements in mechanical devices such as implants or pumps to achieve more targeted and sustained releases of drugs, is now advancing on a microscopic and even molecular level. Recombinant technology has produced a variety of new potential therapeutics in the form of nucleic acids, proteins and peptides and these successes have driven the search for newer and more flexible macromolecule delivery and targeting methods and systems.

Microencapsulation of different molecules within biodegradable polymers and lipid complexes has achieved successes in improving the targeting and delivery of a variety of molecules including nucleic acids and various chemotherapeutic agents. For example, lipid complexes are currently used as delivery vehicles for a number of molecules where sustained release or target release to specific biological sites is desired. In the case of nucleic acids, charged nucleic acid-lipid complexes are utilized to enhance transfection efficiencies in somatic gene transfer by facilitating the attachment of nucleic acids to the targeted cells. Success in somatic gene therapy depends on the efficient transfer and expression of extracellular DNA to the nucleus of eucaryotic cells, with the aim of replacing a defective or adding a missing gene (1). Viral-based carriers of DNA are presently the most common method of gene delivery, but there has been a tremendous activity in developing synthetic nonviral vectors. In particular, cationic liposomes (CLs), in which the overall positive charge of the cationic liposome-DNA (CL-DNA) complex enhances transfection by attaching to anionic animal cells, have shown gene expression in vivo in targeted organs, and human clinical protocols are ongoing (2-4). Cationic liposome transfer vectors exhibit low toxicity, nonimmunogenicity, and ease of production, but their mechanism of action remains largely unknown with transfection efficiencies varying by up to a factor of 100 in different cell lines (2-6).

This unpredictability, which is ubiquitous in gene therapy (7) and in particular in synthetic systems, may be attributed to a lack of knowledge regarding the interactions between DNA and CLs and the resulting structures of CL-DNA complexes. DNA membrane interactions might also provide clues for the relevant molecular forces in the packing of DNA in chromosomes and viral capsids. Studies show regular DNA condensed morphologies induced by multivalent cations (8) and liquid-crystalline (LC) phases at high concentrations of DNA both in-vitro (9) and in- vivo in bacteria (10). More broadly, the nature of structures and interactions between membranes and polymers, either adsorbed (11) or tethered to the membranes (12), is currently an active area of research.

Feigner et al. (3) originally proposed a "bead-on-string" structure of the CL-DNA complexes picturing the DNA strand decorated with distinctly attached liposomes. Electron microscopy (EM) studies have reported on a variety of structures including string-like structures and indications of fusion of liposomes in metal-shadowing EM (13), oligolamellar structures in cryo-TEM (14), and tube-like images possibly depicting lipid bilayer-covered DNA observed in freeze-fracture EM (15).

A variety of modifications of the lipid membranes have been attempted with limited success, including polymerizing or crosslinking the molecules in the bilayer to enhance stability and reduce permeation rates, and incorporating polymers into the bilayer to reduce clearance by macrophages in the bloodstream. While these modifications have proved beneficial, without means to overcome the inherent unpredictability of these complexes by controlling crucial factors such as lipid membrane thickness and the intermolecular spacing of the encapsulated molecules, the use of these molecules is severely limited. The present invention is directed to overcoming this limitation.

SUMMARY OF THE INVENTION

The invention provides novel compositions involving macromolecule-lipid complexes and methods for making them. These compositions and methods of the invention are significant improvements in the field of macromolecule-lipid complex synthesis, macromolecule targeting and delivery to various biological systems.

The present invention provides methods for making macromolecule-lipid complexes and methods for controlling components of the macromolecule-lipid complexes such as the membrane thickness and intermolecular spacing of the complex constituents.

In one embodiment for making macromolecule-lipid complexes, the method comprises mixing a lipid combination (e.g., a neutral lipid and a charged lipid) in a sufficient amount with a macromolecule so as to form a complex with specific geometric and charge qualities. By varying the relative amounts of (1) the charged and neutral lipids, (2) the weight amount and/or the macromolecule and (3) the assembly solution, conditions distinct complexes can be generated having desired isoelectric point or charged states.

By utilizing this process for controlling both the exterior lipid structure and interior macromolecular ordering, an extremely versatile molecular targeting and delivery system can be developed for a variety of applications. The invention has applications in the numerous methods which utilize lipids and various macromolecules such as gene therapy, nucleic acid based vaccine development and peptide and protein delivery.

BRIEF DESCRIPTION OF THE FIGURES Figure 1(A) is a series of high resolution differential interference contrast microscopy images of cationic liposome-DNA complexes showing the formation of distinct condensed globules in mixtures of different lipid to DNA weight ratios. The scale bar is lOμm.

Figure 1(B) is a plot of the average size of the lipid-DNA complexes measured by dynamic light scattering.

Figure 2(A) is a series of small-angle x-ray scattering scans in water as a function of different lipid to DNA weight ratio (L/D). (Inset is under extreme dilute conditions).

Figure 2(B) is plot of the spacing d and doNA as a function of L D.

Figure 2(C) is a series of small-angle x-ray scattering scans of the lamellar Lα phase of DOPC/DOTAP water mixtures done at lower resolution (rotating anode x-ray generator).

Figure 3(A) is a schematic picture of the local arrangement in the interior of lipid-DNA complexes.

Figure 3(B) is a micrograph of the DNA-lipid condensates under bright light.

Figure 3(C) is a micrograph of DNA-lipid condensates under crossed polarizers.

Figure 4(A) is a series of small-angle x-ray scattering scans of CL-DNA complexes at approximately the isoelectric point.

Figure 4(B) is d_{D A} and d from figure 4(A) plotted as a function of L/D.

Figure 4(C) the average domain size of the ID lattice of DNA chains derived from the width of the DNA peaks shown in 4(B). Figure 5a is a schematic representation showing the macromolecule-lipid complex formation from the negatively charged DNA and positively charged liposomes. Schematics of lamellar and inverted hexagonal complex

Figure 5b is the powder X-ray diffraction patterns of two distinct ( and ) liquid- crystalline phases of CL-DNA complexes.

Figures 6a-d are video-microscopy images of CL-DNA complexes in H}₇ and t_a.

Figure 7 are two SAXS scans obtained following the transformation from £_a to H_π phase in the case when the macromolecule is DNA (Left) or a polynucleotide T (right).

Figure 8 shows the variation of structural parameters in l_a and H_; complexes with the three different types of polyelectrolytes and correlative schematic diagrams showing the structure of a unit cell in the three H„ complexes (with DNA, Poly-T, or PGA as the macromolecule).

Figure 9 is a schematic of DNA-lipid complex oriented in microchannels with applications in nanolithograph and separations (or in oriented multilayers).

DETAILED DESCRIPTION OF THE INVENTION

DEFINITIONS:

As used in this application, the following words or phrases have the meanings specified.

As used herein, the term "lipid" means any surfactant both biologically and non- biologically derived.

As used herein, the term "lipid combination" means any mixture of two or more lipids. s As used herein, the term "sufficient amount" means a concentration of a given component that is determined to be adequate to produce the desired effect or characteristic.

As used herein, the term "making" means constructing in a systematic manner.

As used herein, the term "complex" means a substance composed of two or more molecules, components, or parts.

As used herein, the term "isoelectric point state" means the set of conditions under which the electric charge of the complex is approximately zero.

As used herein, the term "negative state" means the set of conditions under which the electric charge of the complex has a net negative charge.

As used herein, the term "positive state" means the set of conditions under which the electric charge of the complex has a net positive charge.

As used herein, the term "charged state" means the set of conditions under which the electric charge of the complex has some net charge or zero charge.

As used herein, the term "the macromolecule interaxial distance (di i)" means the perpendicular distance between the cylinder axis of neighboring macromolecules or the average distance between macromolecules.

As used herein, the term "membrane thickness of the lipid combination (δ_m)" means the thickness of a bilayer of lipid molecule made up of a particular lipid combination.

As used herein, the term "macromolecule area (A_M)" means the cross section area of the macromolecule.

As used herein, "macromolecule density (P_M)" means the density of the macromolecule. As used herein "lipid density (P_L)" means the density of the lipid combination.

As used herein "inverted hexagonal complex phase" means the phase wherein the lipid combination forms a monolayer around the macromolecule (i.e., with lipid tails pointing outward); thereby creating a lipid monolayer macromolecule tube which then assembles into a hexagonal lattice.

As used herein "regular hexagonal complex phase" means the phase wherein the lipid combination assembles into a cylindrical rod (i.e. with lipid tails pointing inward) and macromolecule attached to the outer surface of the rod; thereby creating cylindrical rods with attached macromolecules which then assemble in a hexagonal lattice.

In order that the invention herein described may be more fully understood, the following description is set forth.

METHODS OF THE INVENTION

The invention provides methods for regulating the structure of a charged macromolecule- lipid complex having a selected characteristic or multiple characteristics. These characteristics include interaxial distance (d_M), membrane thickness of the lipid combination (δ_m), macromolecule area (A_M), macromolecule density (P_M), lipid density (P_L). and the ratio (L/D) between the weight of the lipid combination (L) and the weight of the macromolecule (D). The benefits of being able to precisely control the micromolecular structure of macromolecule-lipid complexes is that it will be possible to tailor make specific structures which have defined chemical and biological activities. For example specific structural attributes of cationic lipid-DNA structures are known to impact transfection efficiencies in different biological systems. By being able to manipulate these structural attributes, the chance of success in somatic gene therapy, which depends on the efficient transfer and expression of extracellular DNA to the nucleus of eucaryotic cells, will be greatly improved.

1 The complex comprises a macromolecule and lipid combination. Preferably, both the macromolecule and lipid combination are charged. Further, the charge of the lipid combination is typically opposite of the charge of the macromolecule. By varying the relative amount of the charged and neutral lipid, and the weight of the macromolecule, distinct complexes can be generated having selected isoelectric point or charged states. For example, the lipid combination and the macromolecule can be associated so as to form a complex in an isoelectric point state, Alternatively, the lipid combination and the macromolecule can be associated so as to form a complex in a positively charged state. Further alternatively, the lipid combination and the macromolecule can be associated so as to form a complex in a negatively charged state.

Additionally, in accordance with the practice of the invention, a lipid combination comprises a neutral lipid and a charged lipid. For example, the ratio of the neutral lipid component relative to the charged lipid component can be 70/30, 50/50, 0/100, or 10/90. It clear that in the embodiment, wherein the ratio of the neutral lipid component relative to the charged lipid component is 0/100, a lipid combination is not used but only a single lipid component is used.

Examples of suitable macromolecules include nucleic acid molecules, peptides, proteins, polysaccharides, combinations of a protein and carbohydrate moiety and a synthetic macromolecule of non-biological origin, e.g., doped polyacetylene macromolecules (J.G.S. Cowie "Polymers Chemistry and Physics of Modern Materials", Chapter 7, (Blackie Academic & Professional Press) (1993)).

Examples of suitable neutral lipids include but are not limited to: dioleoyl phosphatidyl cholin, 1 ,2-dioleoyl-sn-glycero-3-phosphoethanolamine, 1 ,2-dicaproyl-sn-glycero-3- phosphoethanolamine, 1 ,2-dioctanoyl-sn-glycero-3-phosphoethanolamine, 1 ,2-dicapryl- sn-glycero-3-phosphoethanolamine, l,2-dilauroyl-sn-glycero-3-phosphoethanolamine, 1 ,2-dimyristoyl-sn-glycero-3-phosphoethanolamine, 1 ,2-dipentadecanoyl-sn-glycero-3- phosphoethanolamine, l,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine, 1,2- diphytanoyl-sn-glycero-3-phosphoethanolamine, l,2-dipalmitoleoyl-sn-glycero-3- phosphoethanolamine, 1 ,2-distearoyl-sn-glycero-3 -phosphoethanolamine, 1 ,2- dipretrselinoyl-sn-glycero-3-phosphoethanolamine, l,2-dielaidoyl-sn-glycero-3- phosphoethanolamine, 1 ,2-dilauroyl-sn-glycero-3-ρhosphoethanolamine, 1 ,2-dilinoleoyl- sn-glycero-3-phosphoethanolamine, l,2-dilinolenoyl-sn-glycero-3-phosphoethanolamine, 1 ,2-diarachidoήoyl-sn-glycero-3-phosphoethanolamine, 1 ,2-docosahexaenoyl-sn-glycero- 3 -phosphoethanolamine, 1 ,2-myristoleoyl-sn-glycero-3-phosphocholine, 1 ,2- dimyristelaidoyl-sn-glycero-3-phosphocholine, 1 ,2-palmitoleoyl-sn-glycero-3- phosphocholine, 1 ,2-palmitelaidoyl-sn-glycero-3-phosphocholine, 1 ,2-petroselinoyl-sn- glycero-3-phosphocholine, 1 ,2-dioleoyl-sn-glycero-3-phosphocholine, 1 ,2-dielaidoyl-sn- glycero-3-phosphocholine, 1 ,2-dilinoleoyl-sn-glycero-3-phosphocholine, 1 ,2-linolenoyl- sn-glycero-3-phosphocholine, 1 ,2-eicosenoyl-sn-glycero-3-phosphocholine, 1 ,2- arachidonoyl-sn-glycero-3-phosphocholine, 1 ,2-erucoyl-sn-glycero-3-phosphocholine, 1 ,2-nervonoyl-sn-glycero-3-phosphocholine, 1 ,2-propionoyl-sn-glycero-3- phosphocholine, 1 ,2-butyroyl-sn-glycero-3-phosphocholine, 1 ,2-valeroyl-sn-glycero-3- phosphocholine, 1 ,2-caproyl-sn-glycero-3-phosphocholine, 1 ,2-heptanoyl-sn-glycero-3- phosphocholine, 1 ,2-capryloyl-sn-glycero-3-phosphocholine, 1 ,2-nonanoyl-sn-glycero-3- phosphocholine, l,2-capryl-sn-glycero-3-phosphocholine, l,2-undecanoyl-sn-glycero-3- phosphocholine, 1 ,2-lauroyl-sn-glycero-3-phosphocholine, 1 ,2-tridecanoyl-sn-glycero-3- phosphocholine, 1 ,2-myristoyl-sn-glycero-3-phosphocholine, 1 ,2-pentadecanoyl-sn- glycero-3-phosphocholine, 1 ,2-palmitoyl-sn-glycero-3-phosphocholine, 1 ,2-phytanoyl-sn- glycero- 3 -phosphocholine, 1 ,2-heptadecanoyl-sn-glycero-3-phosphocholine, 1 ,2-stearoyl- sn-glycero-3-phosphocholine, l,2-bromostearoyl-sn-glycero-3-phosphocholine, 1,2- nonadecanoyl-sn-glycero-3-phosphocholine, l,2-arachidoyl-sn-glycero-3-phosphocholine, 1 ,2-heneicosanoyl-sn-glycero-3-phosphocholine, 1 ,2-behenoyl-sn-glycero-3- phosphocholine, 1 ,2-tricosanoyl-sn-glycero-3-phosphocholine, 1 ,2-lignoceroyl-sn- glycero-3-phosphocholine.

Examples of suitable charged lipids include, but are not limited to, l,2-diacyl-3- trimethylammonium-propane, 1 ,2-dimyristoyl-3-trimethylammonium-propane, 1 ,2- dipalmitoyl-3-trimethylammonium-propane, l,2-distearoyl-3-trimethylammonium- propane, l,2-diacyl-3-dimethylammonium-propane, l,2-dimyristoyl-3- dimethylammonium-propane, l,2-dipalmitoyl-3-dimethylammonium-propane, 1,2- distearoyl-3-dimethylammonium-propane, and 1 ,2-dioleoyl-3-dimethylammonium- propane. _Q In accordance with the practice of the invention, the nucleic acid molecule can be single stranded, double stranded, triple stranded or quadruple stranded. Further, the nucleic acid molecule can be DNA or RNA. The DNA or RNA can be naturally occurring or recombinantly-made. Alternatively, it can be a synthetic polynucleotide. The polynucleotides include nucleic acid molecules having non-phosphate backbones which improve binding. The macromolecule may be linear, circular, nicked circular, or supercoiled.

In one embodiment of the invention, the method comprises selecting a selected characteristic or characteristics described above and modulating one or more of the non- selected characteristics from the group so as to regulate the structure of the macromolecule-lipid complex having the selected characteristic. Preferably, modulation is effected using the formula: dM = (L/D) (AMPM)/(δ_mpL)- Th^e relationship dM = (A /p /(d_m/p₁) (L/D) equates the cationic charge density (e.g., due to the cationic membrane) with the anionic charge density (e.g., due to the macromolecule). Here, PM = density of macromolecule (g/cc) and PL = densities of membrane, d_m the membrane thickness, and A the macromolecule area.

In another embodiment of the invention, the method comprises modulating any of the characteristics associated with the charged macromolecule-lipid complex as described above so as to regulate the structure of the macromolecule-lipid complex having the selected characteristic.

The method further comprises determining amounts of the macromolecule and the lipid combination so selected which would be sufficient to achieve the selected characteristic or characteristics thereby regulating the structure of the complex. In one embodiment this can be accomplished by selecting a selected characteristic or multiple characteristics to be achieved. These characteristics are macromolecule interaxial distance (d_M), membrane thickness of the lipid combination (δ_m), macromolecule area (A_M), macromolecule density (PM), lipid density (P_L), and the ratio (L/D) between the weight of t O uie n iu iuiiiυiiiαuυn \L,) ana tne weignt ot the macromolecule ). ι ucn mc characteristics not selected can be modulated so as to achieve the selected characteristic. After determining the proper amounts, the method provides mixing the macromolecule with the lipid combination in the amount so determined.

For example, when the selected characteristic is a specific value for the interaxial distance of adjacent macromolecules within the macromolecule-lipid complex, the method provides selecting a charged macromolecule and lipid combination, wherein the charge of the lipid combination is opposite of the charge of the macromolecule. The amounts of the macromolecule and lipid combination sufficient to regulate the structure of the complex is then determined using the formula d_M = (L/D) (A_MpM) (δ_mpL) • In one example the interaxial distance is in a range between 24.5 and 60 angstroms. In another example, the interaxial distance is about 60 angstroms. By regulating the interaxial distance of adjacent macromolecules in a complex, the distance between macromolecules within the complex or phase is necessarily regulated. Therefore, this invention also encompasses methods for regulating the distance between macromolecules.

Alternatively, when the selected characteristic is a specific value for the average density of macromolecules within a macromolecule-lipid complex, the amounts of the macromolecule and lipid combination sufficient to regulate the structure of the complex is determined using the formula, d_M = (L/D) (AMPM)/(5„,PL)-

Further, the macromolecule-lipid complex can be a multilamellar structure wherein the lipid combination forms alternating lipid bilayers and macromolecule monolayers. Alternatively, the macromolecule-lipid complex can form either an inverted hexagonal complex phase or a regular hexagonal complex phase. The complex, whether part of a multilamellar or hexagonal structure, comprises macromolecules associated with the lipid in an arrangement that can be regulated and controlled in accordance with the method of the invention.

In another embodiment, the lipid combination and the macromolecule are associated so as to form a complex in an isoelectric point state and the complex has macromolecules exhibiting interaxial spacing of greater than 24.5 angstroms. The resulting complex can have a charge of about zero. In another embodiment, the lipid and the macromolecule is associated so as to form a complex in an isoelectric point state, wherein the amount of the neutral lipid component relative to the charged lipid component ranges from 2 to 95 percent. The resulting complex can have a charge of about zero. Further, the lipid and the macromolecule can associate so as to form a complex in a charged state, wherein the amount of the neutral lipid component relative to the charged lipid component ranges from 55 to 95 percent. The resulting complex can have a net charge.

Additionally, the lipid combination can form a bilayer membrane to which charged macromolecules are associated, and wherein the relative amounts of the lipid components generate the lipid bilayer membrane having a thickness of between 25 and 70 angstroms. Alternatively, the lipid combination can form a bilayer membrane to which charged macromolecules are associated and wherein the relative amounts of the lipid components generate the lipid bilayer membrane having a thickness of between 41 and 60 angstroms. Further, the lipid combination can form a bilayer membrane to which charged macromolecules are associated, and wherein the relative amounts of the lipid components generate the lipid bilayer membrane having a thickness of between 32 and 48 angstroms. Also, the lipid combination can form a monolayer membrane to which charged macromolecules are associated, and wherein the relative amounts of the lipid components generates the lipid monolayer membrane having a thickness of between 12 and 40 angstroms.

In addition to the bilayer membrane form (also referred to herein as lamellar or multilamellar), the resulting complex can form a monolayer (also referred to herein as being in a hexagonal phase, e.g. inverted hexagonal or regular hexagonal). For example, the lipid combination can form a monolayer membrane to which charged macromolecules are associated and wherein the relative amounts of the lipid components generate the lipid monolayer membrane having a thickness of between 15 and 35 angstroms. Alternatively, the lipid combination can form a monolayer membrane to which charged macromolecules are associated, wherein the relative amounts of the lipid components generate the lipid monolayer membrane having a thickness of between 16 and 30 angstroms. The invention further provides a macromolecule-lipid complex produced by the methods of the invention described above.

In one embodiment, the resulting macromolecule-lipid complex comprises a lipid combination having a charged lipid component and a neutral lipid component; and a charged macromolecule. The charge of the lipid combination being opposite of the charge of the macromolecule. The lipid combination macromolecule associate thereby forming a complex in an isoelectric point state. In this state the lipid combination forms a bilayer membrane to which the charged macromolecule is associated and the relative amounts of the neutral lipid component relative to the charged lipid component generates the lipid bilayer membrane having a thickness of between 25 and 75 angstcpms.

In another embodiment, in the resulting macromolecule-lipid complex, the lipids form a bilayer membrane to which the macromolecule is associated, wherein the relative amounts of the lipid components generate the lipid bilayer membrane having a thickness of between 25 and 75 angstroms; and the conformation of the complex has macromolecule exhibiting interaxial spacing of a range between 50 and 75 angstroms.

The invention further provides a process for generating formulations which form the basis for the processing of templates and for producing molecular sieves with precise control over pore size.

For example, the invention provides a process for creating a pattern on a surface using complexes having regulated structures made using the methods described above. The

»3 Additionally, the invention provides a process for creating a material having selected properties such as optical, mechanical, electronic, optoelectronic, or catalytic characteristics not previously realized from components of the material. This process comprises applying a macromolecule-lipid complex to a surface. The complex must have a regulated structure created by the methods of the invention. The process further provides applying molecules which make up the material onto the complex, wherein the molecules self-assemble based on its interactions with the complex. The complex is then removed from the surface thereby creating the material having a selected property. The complex can be in a multilamellar, regular hexagonal, or inverted hexagonal phase. The resulting material can function as a molecular sieve having precise pore size. The invention further provides a molecular sieve produced by the process above.

COMPOSITIONS OF THE INVENTION

The present invention provides nucleic acid-lipid complexes comprising a charged lipid combination and a charged nucleic acid molecule. In one embodiment of the invention, the charge of the lipid combination is opposite of the charge of the nucleic acid molecule. Further, the resulting complex has a desired isoelectric point state and nucleic acids exhibiting interaxial spacing of greater than 24.5 angstroms. In another embodiment, the interaxial spacing range is about between 24.5 and 60 angstroms. In yet another embodiment, the interaxial spacing is about 60 angstroms. In accordance with the practice of the invention, the conformation of the resulting complex can be a multilamellar structure with alternating lipid bilayers and nucleic acid monolayers.

Suitable examples of nucleic acid molecules include, but are not limited to, deoxyribonucleic acid (DNA), ribonucleic acid (RNA). The macromolecules may be linear, circular, nicked circular or supercoiled. The nucleic acid molecules can have phosphate backbones but not necessarily so. Alternatively, nucleic acid molecules having non-phosphate backbones which improve binding are also encompassed within this invention.

In one embodiment, the complex comprises a charged lipid combination; and a charged nucleic acid molecule. The charge of the lipid combination can be opposite of the charse of the nucleic acid molecule. Further, the lipid and the nucleic acid molecule are associated so as to form a complex in an isoelectric point state. In this state, the relative amounts of the lipid components generates the lipid bilayer membrane having a thickness of between 25 and 75 angstroms. Additionally, the conformation of the complex has nucleic acids exhibiting interaxial spacing of a range between 50 and 75 angstroms.

The present invention further provides macromolecule-lipid complexes comprising a charged lipid combination; and a charged macromolecule. Examples of suitable macromolecules include, but are not limited to, nucleic acid molecules such as single or double stranded deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) or hybrids thereof, or modified analogs thereof of varying lengths. In addition, the macromolecule can be a peptide, a protein (or modified analogs thereof). Further, the macromolecule may be a drug such as a chemotherapeutic agent or a modified analog thereof.

In one embodiment of the macromolecule-lipid complex the charge of the lipid combination is opposite of the charge of the nucleic acid molecule. Also, the lipid and the nucleic acid molecule are associated so as to form a complex in an isoelectric point state.

The lipid combination can have a charge lipid component and a neutral lipid component. The amount of the neutral lipid component relative to the charged lipid component can range from 2 to 95 percent.

Alternatively, in another embodiment of the macromolecule-lipid complex, the amount of the neutral lipid component relative to the charged lipid component ranges from 55 to 95 percent. Also, in accordance with the practice of the invention, the ratio of the neutral lipid component relative to the charged lipid component can be 70/30.

Suitable lipids include, but are not limited to, dioleoyl phophatidyl choline or dioleoyl phophatidyl ethanolamine and dioleoyl triethylammonium propane combination.

In a further embodiment of the macromolecule-lipid complex, the lipid combination can be a charged lipid combination and the macromolecule can be a charged macromolecule.

The lipids form a bilayer membrane in the complex to which the charged macromolecule i ∑- can be associated. In this embodiment, the charge of the lipid combination can be opposite of the charge of the nucleic acid molecule. Further, the lipid and the nucleic acid molecule are associated so as to form a complex in an isoelectric point state. Additionally, the relative amounts of the lipid components generates the lipid bilayer membrane having a thickness of between 25 and 75 angstroms.

In another embodiment of the macromolecule-lipid complex, the lipid and the nucleic acid molecule are associated so as to form a complex in a positively charged state, wherein the lipids form a bilayer membrane to which charged macromolecule is associated, and the relative amounts of the lipid components generates the lipid bilayer membrane having a thickness of between 41 and 75 angstroms.

Also, in another embodiment of the macromolecule-lipid complex, the lipid and the nucleic acid molecule are associated so as to form a complex in a negatively charged state, wherein the lipids form a bilayer membrane to which charged macromolecule is associated, and the relative amounts of the lipid components generates the lipid bilayer membrane having a thickness of between 32 and 75 angstroms.

In accordance with the practice of the invention, the lipid can be dioleoyl phophatidyl cholin or dioleoyl phophatidyl ethanolamine and dioleoyl triethylammonium propane. In this embodiment, the charge of the lipid combination in the complex can be opposite of the charge of the nucleic acid molecule. The dioleoyl phophatidyl cholin or dioleoyl phophatidyl ethanolamine and dioleoyl triethylammonium propane form a bilayer membrane to which the charged macromolecule is associated in an isoelectric point state, wherein the relative amounts of dioleoyl phophatidyl cholin or dioleoyl phophatidyl ethanolamine lipids relative to the dioleoyl triethylammonium propane generates the lipid bilayer membrane having a thickness of between 25 and 75 angstroms.

In accordance with the practice of this invention, in the macromolecule-lipid complex, the amount of the neutral lipid component relative to the charged lipid component ranges from 0 to 95 percent and whose charge is approximately zero. Alternatively, the amount of the neutral lipid component relative to the charged lipid component ranges from 55 to

95 percent and which has either a positive or negative charge.

/ fc There is a great flexibility in the structure of these complexes, which may vary greatly in their molecular ordering. These complexes may be relatively simple or may consist of a highly ordered structure. For example the conformation of such a complex can include a multilamellar structure with alternating lipid bilayers and nucleic acid monolayers.

The invention further provides formulations which form the basis for the processing of templates and for producing molecular sieves with precise control over pore size.

The following examples are presented to illustrate the present invention and to assist one of ordinary skill in making and using the same. These examples are not intended in any way to otherwise limit the scope of the invention.

EXAMPLE 1

Cationic liposomes complexed with DNA (CL-DNA) are promising synthetically based nonviral carriers of DNA vectors for gene therapy. The solution structure of CL-DNA complexes was probed on length scales from subnanometer to micrometer by synchrotron x-ray diffraction and optical microscopy. The addition of either linear λ-phage or plasmid DNA to CLs resulted in an unexpected topological transition from liposomes to optically birefringent liquid crystalline condensed globules. X-ray diffraction of the globules reveals a novel multilamellar structure with alternating lipid bilayer and DNA monolayers. We discovered that λ-DNA chains form a one-dimensional lattice with distinct interhelical packing states. Remarkably, in the isoelectric point state, the λ-DNA interaxial spacing expands between 24.5 and 60 angstroms upon lipid dilution and is indicative of a long-range electrostatic-induced repulsion possibly enhanced by chain undulations.

We have carried out a combined in situ optical microscopy and x-ray diffraction (XRD) study of CL-DNA complexes (an embodiment of a macromolecule-lipid complex). On semi-macroscopic length scales, the addition of linear or circular plasmid DNA to binary mixtures of cationic liposomes induces a topological transition from liposomes into collapsed condensates in the form of optically birefringent LC globules with size on the order of 1 μm. The solution structure of the globules was revealed on the 1 to 100 nm length scale by high-resolution synchrotron XRD studies. Unexpectedly, the complexes consist of a higher ordered multilamellar structure with DNA sandwiched between cationic bilayers.

We have discovered distinct interhelical packing states for linear λ-phage DNA, above and below, and at the isoelectric point of the complex by varying the concentrations of DNA and the lipid components comprising the complex. Remarkably, in the isoelectric state of the CL-DNA complex the DNA interaxial distance doNA increases from 24.5 to 60 A as a function of lipid dilution and is quantitatively consistent with an expanding one- dimensional (ID) lattice of DNA chains. Thus, the DNA chains confined between bilayers form a novel 2D smectic phase.

DNA molecules can be readily labeled and imaged by fluorescence microscopy (16). Free λ-DNA in aqueous solution appears as a highly dynamic blob of = 1 μm in diameter, in agreement with a classical random coil configuration, while the contour length of λ-phage DNA is 16.5 μm. The CLs consisted of binary mixtures of lipids which contained either DOPC (dioleoyl phosphatidyl cholin) or DOPE (dioleoyl phosphatidyl ethanolamine) as the neutral co-lipid and DOTAP (dioleoyl trimethylammonium propane) as the cationic lipid. A mixture of DOPE/DOT AP (1:1, wt:wt) was prepared in a 20 mg/ml chloroform stock solution. 500 ml was dried under nitrogen in a narrow glass beaker and desiccated under vacuum for 6 hours. After addition of 2.5 ml Millipore water and 2 hx incubation at 40°C the vesicle suspension was sonicated by clarity for 10 minutes. The resulting solution of liposomes, 25 mg/ml was filtered through 0.2 μm Nucleopore filters. For optical measurements the concentration of SUV used was between 0.1 mg/ml and 0.5 mg/ml. All lipids were purchased from Avanti Polar Lipids, Inc. (Alabaster, Alabama).

The DOTAP/DOPC and DOTAP/DOPE CLs had a size distribution ranging between 0.02 to 0.1 μm in diameter, with a peak around 0.07 μm (18). We used highly purified linear λ-phage DNA (48,502 bp) in most of the experiments but some were carried out with

Escherichia coli DNA and pBR322 plasmid DNA (4361 bp); the latter, consisted of a mixture of nicked circular and supercoiled DNA. Purified λ-phage DNA and pBR322

I * plasmid were purchased from Biolabs, New England. Optical and x-ray data were taken with linear λ prepared in 2 ways: (1) used as delivered, and (2) by heating to 65°C and reacting with a surplus of a 12-base oligo complementary to the 3' COS end. Subsequently the DNA was ligated (T4 DNA ligase, Fischer). The methods gave the same result. For the optical experiments the DNA concentration used was between 0.01 mg/ml and 0.1 mg/ml. Condensation of CLs with λ-DNA was directly observed using differential interference microscopy (DIC) and fluorescence microscopy. A Nikon Diaphot 300 equipped for epifluorescence and high resolution DIC was used.

We show in Fig. 1A a series of DIC images 30 min after preparation in CL-DNA mixtures as a function of the total lipid to λ-DNA weight ratio L/D, where L = DOTAP + DOPE denotes the weight of lipid and D the weight of DNA. Figure 1A shows high-resolution DIC images of CL-DNA complexes forming distinct condensed globules in mixtures of different lipid to DNA weight ratio (L/D); scale bar is 10 μm.

Similar images were observed with λ-DNA replaced by the pBR322 plasmid DNA or DOPE replaced by DOPC. At low DNA concentrations (Fig. 1A, L/D = 50), in contrast to the pure liposome solution where no objects > 0.2 μm were found, 1 μm large globules are observed. The globules coexist with excess liposomes. As more DNA is added, the globular condensates form larger chain like structures (Fig. 1 A, L/D = 10). The Brownian motion of these globules suggests that they are linked by an invisible thread. At L/D = 5 the chain-like structures flocculate into large aggregates of distinct globules. For L/D < 5, the complex size was smaller and stable in time again (Fig. 1A, L/D = 2), and coexisted with excess DNA. Fluorescence-labeled DNA and lipid can be detected on each globule, indicating that the globules are DNA-lipid condensates. Sonicated DOPE-DOTAP (1:1) liposomes were prepared at 0.1 mg/ml with 0.2 mol % DHPE-Texas Red fluorescence label. DNA stained by YOYO (Molecular Probes) was added under gentle mixing at different lipid-to-DNA ratios (L/D). Polarized microscopy also shows that the distinct globules are birefringent indicative of their LC nature.

The size dependence of the complexes as a function of L/D (Fig. IB) was independently measured by dynamic light scattering (18). The large error bars represent the broad

/ 7 polydispersity of the system. The size dependence of the aggregates can be understood in terms of a charge-stabilized colloidal suspension. The charge of the complexes was measured by their electrophoretic mobility in an external electric field. For L/D > 5 (Fig. 1A; L/D = 50 or 10) the complexes are positively charged, while for L/D < 5 (Fig. 1A; L/D = 2) the complexes are negatively charged. The charge reversal is in good agreement with the stoichiometrically expected charge balance of the components DOTAP and DNA at L/D = 4.4 where L = DOTAP + DOPE in equal weights. Thus, the positively and negatively charged globules at L/D = 50 and L/D = 2 respectively, repel each other and remain separate, while as L/D approaches 5, the nearly neutral complexes collide and tend to stick due to van der Waals attraction. Remarkably, the size of the globules appears to be only weakly dependent on the length of the DNA in similar experiments carried out with Escherichia coli DNA or pBR322 plasmid (4361 bp).

Figure 2A shows a series of SAXS scans of CL-DNA complexes in excess water as a function of different lipid to DNA weight ration (L/D). The Bragg reflections at g₀oι =

0.096 A^"1 and ₀₀₂ ⁼ 0.1.92 A^"1 result from the multilamellar L_α structure with intercalated monolayer DNA (see Figure 3A). The intermediate peak at <?D_NA is due to the DNA- interaxial spacing _DNA as described in the text. Inset: SAXS scan of an extremely dilute (lipid + DNA = 0.014% volume in water) λ-DNA-DOPE/DOTAP (1 :1) complex at L/D = 10, which shows the same features as the more concentrated mixtures and confirms the multilamellar structure (with alternating lipid bilayer and DNA monolayers) of very dilute mixtures typically used in gene therapy applications.

The XRD experiments revealed unexpected structures for mixtures of CLs and DNA. Figure 2B shows the spacings d and D_NA as a function of L/D show that (i) d is nearly constant and (ii) two distinct states of DNA packing, one where the complexes are positive (L/D > 5, t o_NA approximately 46 A) and the other state where the complexes are negative (L/D > 5, C?_DNA approximately 35 A) Figure 2C shows SAXS scans of the lamellar L_α phase of DOPC/DOTAP (cationic)- water mixtures done at lower resolution (rotating-anode x-ray generator). A dilution series of 30% (d = 57.61 A), 50% (d = 79.49

A), and 70%) (d= 123.13 A) H₂0 by weight is shown. High resolution synchrotron x-ray scattering were performed at the Stanford Synchrotron Radiation Laboratory. Lower resolution XRD experiments were performed using a rotating anode source. Small angle x-ray scattering (SAXS) data of dilute (Φ_w = the volume fraction of water = 98.6% ± 0.3%) DOPC/DOTAP (1 :1) - λ-DNA mixtures as a function of L/D (L = DOPC + DOTAP) (Fig. 2A) are consistent with a complete topological rearrangement of liposomes and DNA into a multilayer structure with DNA intercalated between the bilayers (24) (Fig. 3A). The DNA-lipid condensates were prepared from a 25 mg/ml liposome suspension and a 5 mg/ml DNA solution. The solutions were filled in 2 mm diameter quartz capillaries with different ratios L/D respectively and mixed after flame sealing by gentle centrifugation up and down the capillary.

Figure 3 A shows a schematic picture of the local arrangement in the interior of lipid-DNA complexes (shown at two different concentrations in Figure 1A and in Figure 3B below. The semiflexible DNA molecules are represented by rods on this molecular scale. The neutral and cationic lipids comprising the membrane are expected to locally demix with the cationic lipids (red) more concentrated near the DNA. Micrographs of DNA-lipid condensates under (B) bright light and (C) crossed polarizers showing LC-like defects. Two sharp peaks at q = 0.0965 ± 0.003 and 0.193 ± 0.006 A^'1 correspond to the (001) peaks of a layered structure with an interlayer spacing d(= δ_m + δ_w) which is in the range 65.1 ± 2 A (Fig. 2B, open squares). The membrane thickness and water gap are denoted by δ_m and δ_w, respectively (Fig. 3A). The middle broad peak q_DNA arises from DNA-DNA correlations and gives dpNA - π/qoNA (Fig. 2B, solid circles). The multilamellar structure with intercalated DNA is also observed in CL-DNA complexes containing supercoiled DNA both in water, and also in Dulbecco's Modified Eagle Medium used in transfection experiments in gene therapy applications. This novel multilamellar structure of the CL- DNA complexes is observed to protect DNA from being cut by restriction enzymes. The intercalation of λ-DNA between membranes in CL-DNA complexes was found to protect it against a HincflH restriction enzyme which cuts naked λ-DNA at 7 sites (22).

In the absence of DNA, membranes comprised of mixtures of DOPC and the cationic lipid DOTAP (1 :1) exhibit strong long-range interlayer electrostatic repulsions that overwhelm the van der Waals attraction (27, 28). In this case, as the volume fraction Φ_w of water is increased, the L_α phase swells and d is given by the simple geometric relation d = δ_m/(l-Φ_w) (27). The SAXS scans in Fig. 2C shows this behavior with the (001) peaks moving to lower q as Φ_w increases. From d (= 2π/q(ooι>) at a given Φ_w we obtain δ_m = 39 ± 0.5 A for DOPC/DOTAP (1: 1). Liposomes made of DOPC/DOTAP (1 : 1) with Φ_w = 98.5%) do not exhibit Bragg diffraction in the small wave-vector range covered in Fig. 2A.

The DNA that condenses on the CLs strongly screens the electrostatic interaction between lipid bilayers and leads to condensed multilayers. The average thickness of the water gap δ_w = d-δ_m = 65.1 A - 39 A = 26.1 A ± 2.5 A is, just sufficient to accommodate one monolayer of B-DNA (diameter = 20 A) including a hydration shell (29). We see in Fig. 2B that d is almost constant as expected, for a monolayer DNA intercalate (Fig. 3A). In contrast, as L/D decreases from 18 to 2, do_NA suddenly decreased from = 44 A in the positively charged state just above L/D = 5 (near the stoichiometric charge neutral point) to = 37 A for the negatively charged state (Fig. 2B). In these distinct states, lamellar condensates coexist with excess giant liposomes in the positive state, and with excess DNA in the negative state. The multilamellar structure of the complex (with λ-DNA) and the distinct DNA interhelical packing states was also found in SAXS data in binary mixtures of cationic lipids which contained DOPE [which has a high transfection efficiency (2)] as the neutral co-lipid. However, the complexes showed a phase- separation into two lamellar phases.

The driving force for higher order self-assembly is the release of counterions. DNA carries 20 phosphate groups per helical pitch of 34.1 A, and due to Manning condensation 76% of these anionic groups are permanently neutralized by their counterions, which leads to a distance between anionic groups = the Bjerrum length = 7.1 A (31). During condensation, the cationic lipid tends to fully neutralize the phosphate groups on the DNA in effect replacing and releasing the originally condensed counterions (both those bound to the ID DNA and to the 2D cationic membranes) in solution.

To improve on the signal-to-background intensity ratio the synchrotron XRD experiments were carried out at concentrations (lipid + DNA = 1.4 ± 0.3% volume in water), which, although dilute, were nevertheless greater than the concentrations used in the microscopy work. The DNA-lipid condensates were prepared from a 25 mg/ml liposome suspension

3Λ and a 5 mg/ml DNA solution. The solutions were filled in 2 mm diameter quartz capillaries with different ratios L/D respectively and mixed after flame sealing by gentle centrifugation up and down the capillary.

A typical SAXS scan in mixtures at the optical microscopy concentrations (Fig. 1A) is shown in Fig. 2A (inset) which exhibits the same features and confirms that the local multilayer and DNA structure (Fig. 3A) is unchanged between the two concentrations. The x-ray samples consisted of connected yet distinct globules (Fig. 3B). What is remarkable is the retention of the globule morphology consistent with what was observed at lower concentrations in DIC (Fig. 1 A). Under crossed polarizers (Fig. 3C) LC defects, both focal conies and spherulites (32), resulting from the smectic-A-like layered structure of the DNA-lipid globules are evident. The globules at the lower concentrations (Fig. 1A) show similar LC defects.

We further probed the nature of λ-DNA-packing within the lipid layers by conducting a lipid dilution experiment in the isoelectric point state of the complex. The total lipid (L = DOTAP + DOPC) was increased while the charge of the overall complex, given by the ratio of cationic DOTAP to DNA, was kept constant at DOTAP/DNA = 2.40 ± 0.1. The projected charge density of DNA (two anionic charges per 68 A ) is very nearly matched by two cationic head groups on DOTAP of = 70 A² each and thus permits near complete neutralization of the complex (Fig. 3 A).

Figure 4A shows a series of SAXS scans of CL-DNA complexes at DOTAP/DNA = 2.4 ± 0.1 (approximately the isoelectric point) which shows the DNA peak (arrow) moving toward smaller q as L/D increases (that is, increasing the DOPC to DOTAP ratio at a constant DOTAP/DNA; L = DOTAP + DOPC, D = DNA). Figure 4B shows d_OΗA and d from (A) plotted as a function of L/D (see Fig. 2A for notation). Circles are synchrotron data, and triangles are rotating anode. The solid line is the prediction of a packing calculation (with no adjustable parameters) where the DNA chains form a space-filling ID lattice. Figure 4C shows the average domain size of the ID lattice of DNA chains derived from the width of the DNA peaks shown in (B) [corrected for resolution and powder averaging broadening effects]. The SAXS scans in Fig. 4A, (arrow points to the

DNA peak) show that d_DNA = 2π/q_DNA increased, with lipid dilution from 24.54 A to 73.5

A 3 A as L/D increased with lipid dilution between 2.45 and 13.8 (Fig. 4B). The most compressed interaxial spacing of 24.55 A at L/D = 2.45 approaches the short-range repulsive hard-core interaction of the B-DNA rods containing a hydration layer (29).

The DNA interaxial spacing can be calculated rigorously from simple geometric considerations. If we assume that all of the DNA is adsorbed between the bilayers and that the orientationally ordered DNA chains separate to fill the increasing lipid area as L/D increases, while maintaining a ID lattice (Fig. 3A), then:

Here, po = 1.7 (g/cc) and p = 1.07 (g/cc) denote the densities of DNA and lipid respectively, δ_m the membrane thickness, and A_D the DNA area. A_D = Wt(λ)/(poL(λ)) = 186 A², Wt(λ) = weight of λ-DNA = 31.5 x 10⁶/(6.022 x 10²³) g and L(λ) = contour length of λ-DNA = 48502 x 3.4 A. The solid line in Fig. 4B is then obtained from Eq. 1 with no adjustable parameters and clearly shows a remarkable agreement with the data over the measured interaxial distance from 24.5 to 73.5 A. The observed deviation from linear behavior both in the data and the solid line arises from the slight increase in δ_m as L/D increases. The variation in the interlayer spacing d (= δ_w + δ_m) (Fig. 4B) arises from the increase in the membrane bilayer thickness δ_m as L/D increases (each DOPC molecule is = 4 A to 6 A longer than a DOTAP molecule). δ_m was obtained at each L/D by measuring d in the L_α phase multilayer membranes at the corresponding DOTAP to DOPC ration and using the relation δ_m — d ( 1 - Φ_w), Φ_w = water volume fraction. The measured δ_m and d, gave δ_w = 25 A ± 1.5 A close to the spacing for the DNA monolayer (see Fig. 3A).

The existence of a finite-sized ordered lattice is made unambiguous from the line widths of the DNA peaks (Fig. 4A) where we find that the ID lattice of DNA chains has a correlated domain size extending to near 10 unit cells (Fig. 4C). Thus, the DNA chains form a ID ordered array adsorbed between 2D membranes; that is, they form a novel finite-sized 2D smectic phase.

A The lattice expansion at the isoelectric point covering interaxial distances with negligible short-range hydration forces (29) (B-DNA diameter « 20 A) is indicative of a long-range repulsion. The distribution of the counterion lipid (DOTAP) concentration according to the Poisson-Boltzmarm equation along the top and bottom monolayer which bound the DNA molecules (Fig. 3A) will lead to a long-range electrostatic-induced interhelical interaction from the counterion lipid pressure (due to the expected local demixing of the cationic and neutral lipids) and the electric field. Preliminary salt dependent experiments which show shifts in the DNA peak indicate that long-range electrostatic induced interactions are present. Additionally, because of the semi-flexible nature of λ-DNA [consisting of between 170 and 340 persistence lengths (ξ_p) in dilute solution (ξ_p « between 500 and 1000 A)], we expect the long-range repulsions to be further enhanced by chain-undulation interactions. A similar enhancement has been observed in a hexagonal lattice of DNA (29, 36). This phase of ID DNA chains is the lower dimensional analog of 2D fluid membranes in that it may either be dominated by electrostatic-induced forces (27, 28) or the interplay between electrostatics and undulations (37-39).

Further experiments are needed to elucidate the precise nature of the intermolecular forces and the interplay between electrostatic and chain undulation interactions (40). Future studies may also reveal states with 3D correlations between the DNA chains from layer to layer in analogy to recent theoretical findings in highly condensed DNA phases (41). The observed quantitative control over the structural nature of the DNA packing in CL-DNA complexes may lead to a better understanding of the important structural parameters relevant to transfection efficiencies in gene therapy; in particular, they should be directly relevant to our understanding of the interactions of the complex with cellular lipids and the mechanism of DNA transfer across the nuclear membrane.

EXAMPLE 2

This example provides the hexagonal phase of a cationic lipid-polyelectrolyte complex (an embodiment of a macromolecule-lipid complex). This embodiment is a LC structure of the complex achieved by varying the lipid composition. It is a novel LC phase with DNA double-strands surrounded by lipid monolayers arranged on a regular hexagonal lattice. This embodiment interacts differently with giant negatively charged liposomes, compared to the lamellar phase, and represents the simplest model of outer cellular membranes. We demonstrate the generality of the lamellar-hexagonal transformation by observing it in complexes of cationic lipid with two other negatively charged biopolymers - polyglutamic acid (PGA), a model polypeptide and poly-thymine (polyT), a model single-stranded oligo-nucleotide. We identify the interactions leading to the transformations between the two complex phases for the three different biological polyelectrolytes. Aside from the significance for gene therapy, our findings suggest new pathways for controlling structural parameters of polyelectrolyte-surfactant complexes, which has been suggested as templates for the formation of new soft materials.

Example 1 shows that mixing linear DNA with liposomes of DOPC/DOTAP mixtures leads to a topological transition into CL-DNA complexes of lamellar structure L^c _a , where DNA monolayers are sandwiched between lipid bilayers (43). In this example, the existence of a completely different inverted hexagonal HJ, liquid-crystalline state in complexes of linear 1-DNA with liposomes of DOPE/DOTAP mixtures is unambiguously demonstrated for the first time using synchrotron small-angle x-ray diffraction and optical microscopy. We show how changing the ratio of cationic DOTAP to neutral DOPE lipid in the liposomes leads to CL-DNA complexes with lamellar or hexagonal structure (Figure 5 a).

The use of cationic lipids can be extended to deliver other negatively charged biopolymers into cells, in particular polypetide-based drugs and single-stranded oligonucleotides for antisense therapy (23, 24). We show that these polyelectrolytes also form complexes with cationic lipids of lamellar and hexagonal structure, similar to the CL-DNA complexes. Comparison of the three types of complexes allows to gain an insight on how the polyelectrolyte charge density and diameter tune the interactions between lipids and polymer, shifting the phase boundaries between L^c _a and HJ, complexes.

Figure 5a shows the formation pathway of a complex from the free DNA and liposomes. 1-DNA in solution has a random-coil configuration of ~1 μm diameter. The Cls consisting of binary DOPE/DOTAP mixture have an average size of 0.06μm. In order to reduce the electrostatic free energy, both DNA and lipid charges are partially neutralized by their respective counterions. During the CL-DNA complex formation cationic lipids replace DNA counterions, releasing the [Na⁺] and [Cf] ions into solution with a very large entropic free energy gain (of order k_BT per released counterion). The result is a close association between DNA and lipid in a compact complex of ~0.2μm size. The overall charge of the complex is determined by the weight ratio r of cationic lipid and DNA. The complexes are positive for r>2.2 and negative for r<2.2, indicating that charge reversal occurs when complexes are stoichiometrically neutral with one positive lipid per each negatively charged nucleotide base.

Surprisingly, the internal structure of the complex changes completely with DOPE/DOTAP ratio. Defining the volume fraction of DOPE as φπε as the fraction of neutral DOPE in the lipid mixture, the complex is lamellar L^c _a for Φ_PE<0.41 and has inverted hexagonal H°_u structure for Φ_PE >0.7. In complexes with 0.41<φp_£ <0.7 the two structures coexist. Small-angle x-ray scattering (SAXS) data of complexes with φp_E =0.41 and 0.75 (Figure 5b) clearly shows the presence of two completely different structures. The two sharp peaks at q=0.099A^_I and 0.198A^"1 correspond to (001) and (002) peaks of a lamellar structure with interlayer spacing d=63.4A. Since DOPE/DOTAP bilayer has thickness δ_m=40A at Φ_PE =0.41 , the water gap between bilayers d_w=d-d_m=23.4A is just large enough to accommodate a monolayer of DNA with a hydration shell of water. This structure is analogous to the one previously reported in DOPC/DOTAP-DNA complexes (Example 1). The middle broad peak at q_DNA arises from regular 2D-smectic arrangement of DNA, giving the spacing between the DNA strands

For Φ_PE >0.7 the peaks of the SAXS scan index perfectly on a hexagonal lattice with a repeat spacing of a = 4π/ q_\o = 67.8A. We were able to observe Bragg peaks up to 7th order, indicating a high degree of regularity of the structure. Schematic of the new HJJ phase is shown in Figure 5a. Each of the DNA molecules is surrounded by a monolayer of lipid and the unit cells of DNA/lipid inverted cylindrical micelles are arranged in a hexagonal lattice. The structure resembles that of the inverted hexagonal ? ( HJJ) phase of pure DOPE in excess water (30), with the water space inside the lipid micelle filled by DNA. The higher electron density of DNA with respect to water leads to the relative suppression of (23) and Bragg peak intensities compared with that in pure lipid HJJ phase. Assuming again an average bilayer thickness of 4θA, the diameter of micellar void in the H_j phase is ~28A, again sufficient for a DNA molecule with approximately two hydration shells.

To improve the signal/background ratio, samples for synchrotron SAXS experiments were prepared at lipid and DNA concentrations about 100 times greater then typically used in optical microscopy and transfection studies. SAXS scans of mixtures at typical transfection concentrations, also shown in Figure 5b, have Bragg peaks at exactly the same positions as in corresponding more concentrated samples. This confirms that the internal L^c _a and H_u structures of the complexes and the phase boundaries between them are independent of the overall DNA and lipid concentrations.

In either of the condensed phases the complexes appear as highly dynamic birefringent aggregates when viewed with video-enhanced optical microscopy (Figure 6a,b). Each complex consists of several connected blobs close to charge neutrality, with the aggregates becoming smaller and eventually dissociating into individual blobs with the increasing complex charge. Interestingly, the shape of aggregates is different in the two complex phases: the L^c _a phase forms linear structures, while in the H), phase the aggregates are predominantly branched. Microscopy of DNA and lipids with appropriate fluorescent labels allows us to image their respective distributions in the complex. This observations show that the complex is indeed a compact object, with a close association of lipid and DNA, since in both phases the complexes exhibit fluorescence in DNA and lipid modes. The complexes coexist with excess DNA for r<2.2 and with excess lipid when r>2.2. However, we never observe presence of macroscopic lipid aggregates. proving that the only condensed liquid crystalline structures in the CL-DNA mixtures are complexes. On a larger length-scale and at higher lipid and DNA concentrations, bigger LC aggregates are observed (Figure 6a), with very different defect structures in the two phases. H_n° phase never exhibits the spherrulites characteristic of the L^c _a phase. The spherulites are an unmistaking signature of lamellar liquid-crystalline structure (33), and are not present in hexagonal phases.

The membrane of giant anionic liposome is a good model of the outer cell membrane - the first barrier to the complex on its way to DNA delivery. There is a striking difference in the way H₇ and L complexes interact with model anionic lipid membranes. We show in Figure 6c,d typical micrographs of slightly positively charged (r=4) complexes attached to the fluid membranes of giant liposomes. The L°_a complexes attached to anionic membrane remain stable for many hours. The compact complex morphology can be seen in DIC as well as in DNA and lipid fluorescence. Clearly there is no fusion between the complex and the giant liposome. A strikingly different behavior is observed with H„ complexes. They lose their compact structure immediately upon attaching to the liposome, spreading and fusing with it. Since the amount of lipid in the complex is comparable with that in liposome, and since the fusion occurs very quickly, it results in formation of a local multilamellar structure on the giant liposome surface. The loss of the compact complex structure and the subsequent spreading of the DNA fluorescence are clear indications of fusion and the first observed example of the effect of complex structure on its interaction with a membrane. This finding unambiguously demonstrates the importance of complex internal structure for the efficiency of CL-DNA vectors.

The presence of H), and L^c _a phases is universal in complexes of DOPE/DOTAP mixtures with other anionic polyelectrolytes. Figure 7 shows SAXS scans of complexes with DNA and oligonucleotide polyT (100 bases long) as a function of Φ_PE. As Φ_PE increases, the complexes undergo a first order phase transition from lamellar to hexagonal structure with a broad range of φ_PE were the two phases coexist. The same structures are also observed in complexes of DOPE/DOTAP with anionic polypeptide PGA (M =81,000). The only difference in the structure of complexes between DNA and the shorter polyelectrolytes is the absence of polymer-polymer correlation peak in the L^c _a phase. We attribute this difference to the difference in length and rigidity between very long and stiff DNA and shorter, more flexible polyT and PGA. We compare the phase diagrams of CL-polyelectrolyte complexes for the three different polymers in Figure 8, which also shows the variation of repeat distances of complex structure as a function of Φ_PE. TO understand the phase sequence in complexes it is useful to consider structure of DOPE/DOTAP mixtures without the polyelectrolytes. These phase boundaries are indicated on top of Figure 8. Pure lipids also form lamellar L_a and inverted hexagonal H_π structures, although the phase boundaries are very different from CL-polymer systems and the H_π phase is only present in coexistence with L_a structure. Therefore the phase sequence in the CL-polyelectrolyte mixtures mimics the ones preferred by the pure lipids, with stabilization of the pure inverted hexagonal phase.

DOPE forms stable H_π phases, whereas DOTAP has stable lamellar structures. Once the complex is formed and lipid and polymer counterions are released, the internal structure of the complex will be affected by several comparable free energy contributions. Since DOPE monolayers have negative spontaneous curvature and bending energy of only a few k_BT *", increasing φrm will allow the lipid layers to curve around the polyelectrolites, forming the Hj_j structure. Additionally, the lipid head-group area and correspondingly chain length will adjust itself so as to further minimize the free energy of the system, since the stretching energy of the lipid chain is only slightly greater then the bending energy of the monolayers. The three polyelectrolites which we have studied have different diameters (2θA DNA, 13A PGA which has a-helix conformation inside the complex, ~lθA poly-T) and different linear charge densities (l=2e73.4A DNA, le^'/1.5A PGA, ~le73.4A poly-T). This changes the relative magnitude of electrostatic interaction in the complex, as well as the required amount of lipid monolayer bending in the j_j phase, thus shifting the phase boundaries and structure of a unit cell in the complex.

Further insight into the relative phase boundaries and structures in the three CL-polymer complexes may be gained if one considers that the charge densities of polyelectrolyte and lipid monolayers have to match within the H_u unit cell, I = were A is the lipid head-group area and D is the radius of lipid monolayer, which may be larger then polyelectrolyte diameter. Let us assume first that the lipid layer thickness remains fixed at

2 0 d_m=4θA in the H°_u complex. Then in CL-DNA complex D=24A and A=65A (normal value), giving Φ_PE =0.5, close to experimentally observed lower boundary of the HJ_; phase. This implies closely matched diameters of DNA and lipid monolayers in the complex unit cell (Figure 8). In CL-pT complex D=25A and A=65A, giving at φpε =0.75, again close to the experimentally observed value. This corresponds to a loosely bound unit cell, as shown in Figure 8. Higher Hj_j phase boundary and greater difference between polymer and monolayer diameters arise because of the weaker electrostatic interaction and larger monolayer bending in CL-pT complex compared with CL-DNA. In CL-PGA Hj, phase, a reasonable phase boundary may be only achieved if the head-group area is substantially smaller, resulting in stretching of the lipid chains and increase in lipid layer spacing. With A=4θA and D=2θA one obtains Φ_PE =0.6, in reasonable agreement with experiment. Here stronger electrostatic interaction and small polymer diameter result in crowding of lipid heads. The additional free energy of stretching the chains may be the cause of the very narrow region of stability of pure Hj phase in CL-PGA system.

We have provided a first demonstration for the existence of distinctly different lamellar and hexagonal LC structures of CL-DNA complexes. These structures are formed at different lipid compositions and interact differently with model anionic membranes. The two LC phases also form in other Cl-biopolyelectrolyte complexes used for intra-cellular delivery. Comparison between the complexes in three different systems also improves the understanding of interactions shaping complex structure. This will be important for controlled design of the new class of surfactant-polyelectrolyte materials (48), of which our complexes are examples.

Figure 5a shows the schematic of the complex formation from the negatively charged

DNA and positively charged liposomes. Complete topological rearrangement of lipids and DNA in this process is driven by release of partially-bound counterions from the diffuse screening layers into bulk solution, which lowers the electrostatic free energy of the system. However, once the counterions are released and the lipids are bound to DNA, the liquid-crystalline structure of the complex will depend on the interplay of various comparable contributions to the complex free energy. These vary with the lipid

3 ) composition of the complex, resulting in two different observed structures: the lamellar complex L^c _a when the volume fraction of neutral DOPE lipid (Φ_PE) is φ_P£ <0.41 and the inverted hexagonal complex Hj, for φp_E >0.7. The two structures coexist for intermediate

Figure 5b provides the powder X-ray diffraction patterns of the two distinct liquid- crystalline phases of CL-DNA complexes. Scan of the Hj_; complex at Φ_PE =0.75 (open circles, top) shows the first three order Bragg peaks of the hexagonal DNA/lipid lattice at qιo=0.107A^"1, and q₂o=0.214A^_1. Scan of the lamellar L°_a complex at φp_E =0.41 (filled circles, bottom) shows the peaks at q₀oι=0.099A^"' and qo^O.^δA^"1 resulting from the lamellar periodic structure with DNA intercalated between lipid bilayers and a peak at q_DNA ⁼0.172A^*] due to the smectic structure of the intercalated DNA. In both cases the samples were prepared by mixing concentrated deionized water solutions of DNA (5mg/ml) and lipid (25mg/ml) directly in a 1.5mm diameter quartz x- ray capillary with r=3. Because these concentrations are higher then typically used in preparation of CL-DNA complexes for cell transfection, we have also recorded SAXS patterns of complexes made from dilute DNA (O.Olmg/ml) and lipid (O.lmg/ml) solutions (solid lines). The peak positions are the same for experiments done with concentrated and dilute complexes, indicating that the complex phases remain the same at lipid and DNA concentrations typically used for cell transfection.

Figures 6a-b provides video-microscopy images of CL-DNA complexes in (a) HJ, and (b) L^c _a phases. In all cases complexes were viewed in DIC (left), lipid fluorescence (middle) and DNA fluorescence (right). For fluorescence experiments cationic lipids were labeled with 0.2 mol% of DΗPE-TexasRed and DNA was labeled with Yo Yo- 1 iodide at 1 dye molecule/15bP ratio. The complex morphology is different in the two phases: branched in the H_tI and linear in the L^c _a phase. In both phases the lipid is closely associated with DNA, as evidenced by the exactly same morphology of complexes in the two fluorescence modes. Complexes were prepared by gently mixing DNA (O.Olmg/ml) and lipid (0.1 mg/ml) stock solutions with φ_PE =0.73 (a) and φ_Pε =0.3 (b) to yield the r=3 weight ratio (slightly positively charged complexes). The complexes were further diluted with deionized water for observation. Scale bar is 2μm in DIC and 4μm in fluorescence images.

Figures 6c-d provides video microscopy of positively charged Hj_j (c) and L^c _a (d) complexes that interact differently with the negatively charged giant liposomes. The lamellar complexes simply stick to the liposomes and remain stable for many hours, retaining their blob-like morphology. The blobs are localized in DIC as well as lipid and DNA fluorescence modes. The hexagonal complexes break-up and spread immediately after attaching to giant liposomes, indicating a fusion process between the complex and the liposome lipid bilayer. Spreading of the complex is evident in both lipid and DNA fluorescence modes. Giant unilamellar liposomes were prepared from the mixtures of 90%) DOPC (neutral) and 10% DOPG (negatively charged) lipids. CL-DNA complexes were prepared as described above with r=4. Scale bar is lOμm in both DIC and fluorescence images.

Figure 7 provides SAXS scans following the transformation from L^c to Hj_j phase with increasing amount of DOPE for complexes with DNA (i) and poly-Thymine (ii). The dashed line indicates L^c _a phase peaks. At very high DOPE content (Φ_PE >0.85) the H_u complexes coexist with the excess Hπ phase of pure DOPE (peaks marked with arrows). In both (i) and (ii) r=3, slightly above charge-neutrality.

Figure 8 shows variation of structural parameters in L^c _a and Hj_j complexes with the three different types of polyelectrolites (i) 1-DNA, (ii) poly-Thymine (polyT), (iii) polyglutamic acid (PGA). In all cases a « -J3/2 d, were a is the repeat distance of pure Hj, and d is the membrane repeat distance in pure L^c _a complex. Thus U_a and Hj, phases are always epitaxially matched, but this condition is not satisfied for the regions of phase coexistence. The arrows on top of the figure indicate the phase boundaries in the mixtures of DOPE and DOTAP lipids, indicating that the presence of polyelectrolites stabilizes the pure lamellar and hexagonal phases. Schematic representations show the structure of a unit cell in the three Hj, complexes, demonstrating that the thickness of water layer and the stretching of the lipid chains should be different in the three polyelectrolyte-lipid complexes. - -> EXAMPLE 3

Recently we have found that cationic liposomes (CL) complexed with DNA (CL- DNA) form a novel self-assembled structure consisting of a higher ordered multilamellar structure with DNA sandwiched between cationic lipid bilayers shown schematically in Fig. 5. These series of x-ray diffraction experiments lead to the observation of a variation in the DNA interaxial distance as a function of the lipid to DNA (L/D) weight ratio in multilayers which unambiguously showed that the x-ray technique was directly probing the DNA structure in multilayer assemblies. It was found that the linear DNA confined between bilayers forms an expanding one- dimensional lattice of chains with the center to center distance between DNA varying in a controlled manner in the nanometer range 25 A < dp^_A ^< 60 A.

Microstructures with submicron linewidths as substrates for confining and orienting this multilamellar CL-DNA structure is shown schematically in Fig. 9. The oriented multilamellar structure would have many important technological applications. For example, in developing nano-scale masks in lithography and molecular sieves with nanometer scale cylindrical pores (Fig. 9).

REFERENCES:

1. R. G. Crystal, Science 270, 404 (1995); R. C. Mulligan, Science 260, 926 (1993).

2. P. L. Feigner, G. Rhodes, Nature 349, 351 (1991). 3. P. L. Feigner, et al., Proc. Natl. Acad. Sci. USA 84, 7413 ( 1987).

4. N. Zhu, D. Liggitt, Y. Liu, R. Debs, Science 261, 209 (1993).

5. G. J. Nabel, et al., Proc. Natl. Acad. Sci. USA 90, 11307 (1993); N. M. Caplen, et al., Nature Medicine 1, 39 (1995).

6. D. Lasic, N. S. Templeton, Advanced Drug Delivery Review, (in press). 7. E. Marshall, Science 269, 1050 (1995); E. Marshall, Science 270, 1751 (1995).

8. V. A. Bloomfield, Biopolymers 31, 1471 (1991).

9. F. Livolant, A. M. Levelut, J. Doucet, J. P. Benoit, Nature 339, 724 (1989).

10. Z. Reich, E. J. Wachtel, A. Minsky, Science 264, 1460 (1994).

11. E. Sackmann, Science 211, 43 (1996); C. Ligoure, G. Bouglet, G. Porte, Physical Review Letters 71, 3600 (1993).

12. H. E. Warriner, S. H. J. Idziak, N. L. Slack, P. Davidson, C. R. Safinya, Science 271, 969 (1996); A. K. Kemworthy, K. Hristova, D. Needham, T. J. Mclntosh, Biophysical J. 68, 1921 (1995).

13. H. Gershon, R. Ghirlando, G.S.B., A. Minsky, Biochemistry 32, 7143 (1993). 14. J. Gustafsson, G. Arvidson, G. Karlsson, M. Almgren, BBA 1235, 305 (1995).

15. B. Sternberg, F. L. Sorgi, L. Huang, FEBS letters 356, 361 (1994).

16. S. B. Smith, L. Finzi, C. Bustamante, Science 258, 1122 (1992); T. T. Perkins, E. S. Douglas, S. Chu, Science 264, 819 (1994).

17. J. P. Behr, Bioconjugate Chemistry 5, 382 (1994). 18. The liposome and complex sizes were measured by dynamic light scattering

(Microtrac UPA 150, Leeds and Northrup).

19. A. Singhal, L. Huang, Gene Therapeutics: Methods and Applications of Direct Gene Transfer. J. A. Wolff, Ed., (Birkhauser, Boston 1994).

20. Feigner, J., et al. J.Biol.Chem. 269, 2550-2561 (1994). 21. Remy, J.-S., Siriin, C, Vierling, P. & Behr, J.-P. Bioconjugate Chem. 5, 647-654

(1994).

22. Raedler, J. O. et al. Science 275, 810-8 (1997)

3^ 23. Farhood, H., Serbina, N. & Huang, L. Biochim.Biophys.Acta 1235, 289-295 (1995).

24. Hui, S.W., et al. Biophys.J. 71, 590-599 (1996).

25. Raedler, J. O. et al. Science 275, 810-8 (1997) 26. Chiang, M.-Y., et al. J.Biol.Chem. 266, 18162-18171 (1991).

27. D. Roux, C. R. Safmya, J. Physique France 46, 307 ( 1988).

28. C. R. Safmya, in Phase Transitions in Soft Condensed Matter R. Tormod, D. Sherringion, Eds. (Plenum, New York, 1989) pp. 249-270.

29. R. Podgornik, D. C. Rau, V. A. Parsegian, Macromolecules 22, 1780 (1989). 30. Lappalainen, K., et al. Biochim.Biophys.Acta 1196, 201-208 (1994).

31. G. S. Manning, Journal of Chemical Physics 51, 924 (1969).

32. P. Boltenhagen, O. D. Lavrentotovich, M. Kleman, Phys. Rev. A 46, 1743 (1992).

33. Seddon, J.M. Biochim.Biophys.Acta 1031, 1-69 (1989).

34. Boltenhagen, P., Lavrentovich, O.D. & Kleman, M. Phys. Rev. A 46, 1743-1746 (1992).

36. J. V. Selinger, R. F. Bruinsma, Physical Review A 43, 2922 (1991).

37. W. Helfrich, Z. Naturforsch A 33, 305 (1978).

38. C. R. Safmya, et al., Physical Review Letters 57, 2718 (1986).

39. E. A. Evans, V. A. Parsegian, Proceedings of the National Academy of Sciences U.S.A. 83, 7132 (1986).

40. N. Dan, BiophysicalJournal (in press).

41. Kamien, D. R. Nelson, Phys. Rev. E 53, 650 (1996).

42. (start example 2) Behr, J.-P. Bioconjugate Chem. 5, 382-389 (1994).

43. Raedler, J.O., Koltover, I., Salditt, T., Safmya, C. R. Science 275, 810-814 (1997). 44. Feigner, P.L., et al. Proc.Natl.Acad.Sci. USA 84, 7413 (1987).

45. Remy, J.-S., Kichler, A., Mordinov, V., Schuber, F. & Behr, J.-P. Proc.Natl.Acad.Sci. USA 92, 1744-1748 (1995).

46. Zhu, N., Liggitt, D., Yong, L. & Debs, R. Science 261, 209-211 (1993).

47. Gruner, S.M. J.Phys.Chem. 93, 7562-7570 (1989). 48. Antonietti, M., Comad, J. & Thunemann, A. Macromolecules 27, 6007-6011 (1994).

3

Claims

What is claimed is:

1. A method for regulating the structure of a charged macromolecule-lipid complex having a selected characteristic comprising modulating any of the characteristics associated with the charged macromolecule-lipid complex so as to regulate the structure of the macromolecule-lipid complex having the selected characteristic, wherein the characteristics associated with the charged macromolecule-lipid complex are the macromolecule interaxial distance (d_M), membrane thickness of the lipid combination (δ_m), macromolecule area (A ), macromolecule density (P_M), lipid density (Γ_L), and the ratio (L/D) between the weight of the lipid combination (L) and the weight of the macromolecule (D) and wherein the complex includes a charged macromolecule and a lipid combination.

2. A method for regulating the structure of a selected charged macromolecule-lipid complex comprising: a. obtaining a charged macromolecule; b. obtaining a lipid combination; the charge of the lipid combination being opposite of the charge of the macromolecule; c. determining an amount of the macromolecule of step (a) and the lipid combination of step (b) sufficient to regulate the structure of the complex by: i. selecting a desired characteristic or multiple characteristics of the complex from a group of characteristics consisting of macromolecule interaxial distance (d_M), membrane thickness of the lipid combination (δ_m), macromolecule area (A_M), macromolecule density (P_M), lipid density (P_L), and the ratio (L/D) between the weight of the lipid combination (L) and the weight of the macromolecule (D); and ii. modulating any of the characteristics not selected in (i) so as to achieve the desired characteristic thereby determining the amount of the macromolecule and lipid combination sufficient to regulate the structure of the complex; and

31 d. mixing the macromolecule with the lipid combination in the amount so determined so as to regulate the structure of the desired charged macromolecule-lipid complex.

3. A method for regulating the interaxial distance of adjacent macromolecules within a macromolecule-lipid complex comprising: a. selecting a charged macromolecule; b. selecting a lipid combination; the charge of the lipid combination being opposite of the charge of the macromolecule; c. determining an amount of the macromolecule of (a) and the lipid combination of (b) sufficient to regulate the structure of the complex by: i. selecting a desired macromolecule interaxial distance (d ); and ii. modulating any of membrane thickness of the lipid combination (5_M), macromolecule area (AM), macromolecule density (P_M), lipid density (p_L), and the ratio (L/D) between the weight of the lipid combination

(L) and the weight of the macromolecule (D)so as to achieve the desired macromolecule interaxial distance thereby determining the amount of the macromolecule and lipid combination sufficient to regulate the structure of the complex; and d. mixing the macromolecule with the lipid combination in the amount so determined so as to regulate the structure of the desired charged macromolecule-lipid complex.

4. A method for regulating the density of macromolecules within a macromolecule- lipid complex comprising: a. selecting a charged macromolecule; b. selecting a lipid combination; the charge of the lipid combination being opposite of the charge of the macromolecule; c. determining an amount of the macromolecule of (a) and the lipid combination of (b) sufficient to regulate the structure of the complex by: i. selecting a desired macromolecule density; and ii. modulating any of membrane thickness of the lipid combination δu), macromolecule area (AM), macromolecule interaxial distance, lipid density (P_L), and the ratio (L/D) between the weight of the lipid combination (L) and the weight of the macromolecule (D) so as to achieve the desired macromolecule density thereby determining the amount of the macromolecule and lipid combination sufficient to regulate the structure of the complex, d. mixing the macromolecule with the lipid combination in the amount so determined so as to regulate the structure of the desired charged macromolecule-lipid complex.

5. The method of claim 1 or 2, wherein the characteristic so selected from group is macromolecule interaxial distance or macromolecule density.

6. The method of claim 1 or 2, wherein the characteristic so selected from the group is macromolecule interaxial distance and macromolecule density.

7. The method of claim 1, 2, 3, or 4, wherein modulating is effected using the formula: d_M = (L/D) (A_MpM)/(δmpL).

8. The method of claim 1, 2, 3, or 4, wherein the macromolecule is a charged macromolecule and the charge of the lipid combination is opposite of the charge of the macromolecule.

9. The method of claim 1, 2, 3, or 4, wherein the macromolecule is a nucleic acid molecule.

10. The method of claim 1, 2, 3, or 4, wherein the macromolecule may be linear, circular, nicked circular or supercoiled.

11. The method of claim 10, wherein the nucleic acid molecule is a DNA or RNA.

12. The method of claim 1, 2, 3, or 4, wherein the macromolecule is a peptide, protein, polysaccharide, a combination of a protein and carbohydrate moiety.

3?

13. The method of claim 1, 2, 3, or 4, wherein the lipid combination comprises a neutral lipid and a charged lipid.

14. The method of claim 1, 2, 3, or 4, wherein the lipid combination and the macromolecule are associated so as to form a complex in an isoelectric point state.

15. The method of claim 1, 2, 3, or 4, wherein the lipid combination and the macromolecule are associated so as to form a complex in a positively charged state.

16. The method of claim 1, 2, 3, or 4, wherein the lipid combination and the macromolecule are associated so as to form a complex in a negatively charged state.

17. The method of claim 13, wherein the neutral lipid is dioleoyl phosphatidyl cholin (DOPC) or l,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE).

18. The method of claim 13, wherein the charged lipid is l,2-diacyl-3- trimethylammonium-propane (DOTAP).

19. The method of claim 1, 2, 3, or 4, wherein the macromolecule-lipid complex is a multilamellar structure wherein the lipid combination forms alternating lipid bilayers and macromolecule monolayers.

20. The method of claim 1, 2, 3, or 4, wherein the macromolecule-lipid complex forms either an inverted hexagonal complex phase or a regular hexagonal complex phase.

21. A macromolecule-lipid complex produced by the method of claim 1, 2, 3, or 4.

22. The macromolecule-lipid complex of claim 21, wherein the macromolecule comprises: i. a lipid combination having a charged lipid component and a neutral lipid component; and ii. a charged macromolecule; the charge of the lipid combination being opposite of the charge of the macromolecule; the lipid and the macromolecule being associated so as to form a complex in an isoelectric point state, wherein lipid combination forms a bilayer membrane to which the charged macromolecules are associated in an isoelectric point state, wherein the relative amounts of the neutral lipid component relative to the charged lipid component generates the lipid bilayer membrane having a thickness of between 25 and 75 angstroms.

23. A macromolecule-lipid complex of claim 21 , wherein the complex comprises : i. a charged lipid combination; and ii. a charged macromolecule; the charge of the lipid combination being opposite of the charge of the nucleic acid molecule; the lipid and the macromolecule being associated so as to form a complex in an isoelectric point state, wherein: a. the lipids form a bilayer membrane to which the macromolecule is associated, wherein the relative amounts of the lipid components generate the lipid bilayer membrane having a thickness of between 25 and 75 angstroms; and b. the conformation of the complex has macromolecule exhibiting interaxial spacing of a range between 50 and 75 angstroms.

24. A method for creating a pattern on a surface comprising: a. selecting a charged macromolecule; b. selecting a lipid combination; c. deterrnining an amount of the macromolecule of (a) and the lipid combination of (b) sufficient to regulate the structure of the complex by: i. selecting a desired macromolecule density or interaxial distance; and ii. modulating any of membrane thickness of the lipid combination (δ_m), macromolecule area (AM), lipid density (p_L), and the ratio (L/D) between the weight of the lipid combination (L) and the weight of the macromolecule (D) so as to achieve the desired macromolecule density or interaxial distance thereby determining the amount of the macromolecule and lipid combination sufficient to regulate the structure of the complex;

HI d. applying the lipid combination on the surface in amount so determined; and e. applying the macromolecule over the lipid combination of (a) in the amount so determined, wherein the macromolecule self assembles onto the lipid combination thereby forming a complex and creating a pattern on the surface.

25. The method of claim 24, wherein the pattern is used to create a mask.

26. A method for creating a material having desired properties comprising: a. applying a macromolecule-lipid complex to a surface by the method of claim 24; b. applying molecules which make up the material onto the complex of (a), wherein the molecules self-assemble based on its interactions with the complex; and c. removing the complex from the surface thereby creating the material having the regulated structure.

27. The method of claim 26, wherein the complex is in a multilamellar, regular hexagonal, or inverted hexagonal phase.

28. The method of claim 26, wherein the material so created is a molecular sieve.

29. A molecular sieve produced by the method of claim 26.

30. The method of claim 24, wherein modulating is effected using the formula: d_M = (L/D) (A_MPM)/(δ_mp_L).