SURFACTANT-STABILIZED FLUID INTERFACE AND DISPERSION COMPRISING ONE OR MORE DROPLETS HAVING A SURFACTANT STABILIZED FLUID INTERFACE

Abstract

The present invention relates to a surfactant-stabilized fluid interface, comprising a layer of surfactant and a first compound containing a hydrophobic part covalently linked to a molecular recognition site, wherein the surfactant-stabilized fluid interface has a first fluid on one side and a second fluid on the other side, wherein the surfactant stabilizes the fluid interface, wherein the hydrophobic part of the first compound is interacting with the layer of surfactant by secondary non-covalent interactions and the molecular recognition site of the first compound extends from the layer of surfactant, and wherein the surfactant is a block-copolymer having at least one hydrophilic block and at least one hydrophobic block. The present invention further relates to a dispersion comprising one or more droplets having a surfactant stabilized fluid interface.

Claims

1.-64. (canceled)

65. Surfactant-stabilized fluid interface, comprising a layer of surfactant and a first compound containing a hydrophobic part covalently linked to a molecular recognition site, wherein the surfactant-stabilized fluid interface has a first fluid on one side, a second fluid on the other side and a surfactant, wherein the surfactant stabilizes the fluid interface, wherein the hydrophobic part of the first compound is interacting with the layer of surfactant by secondary non-covalent interactions and the molecular recognition site of the first compound extends from the layer of surfactant, and wherein the surfactant is a block-copolymer having at least one hydrophilic block and at least one hydrophobic block.

66. The surfactant-stabilized fluid interface according to claim 65, wherein the surfactant stabilized fluid interface encapsulates an emulsion droplet having a center comprising the first fluid.

67. The surfactant-stabilized fluid interface according to claim 66, wherein the emulsion droplet is a microfluidic droplet.

68. The surfactant-stabilized fluid interface according to claim 66, wherein the emulsion droplet has a maximum dimension of 0.5 μm to 1,000 μm.

69. The surfactant-stabilized fluid interface according to claim 66, wherein the emulsion droplet is one of at least substantially ellipsoidal and at least substantially spherical.

70. The surfactant-stabilized fluid interface according to claim 66, wherein the emulsion droplet has an outer diameter of 0.5 to 1200 μm.

71. The surfactant-stabilized fluid interface according to claim 67, wherein the microfluidic droplet has a diameter selected in the range of 1 μm or greater to 100 μm or less.

72. The surfactant-stabilized fluid interface according to claim 65, wherein the first compound has a molecular recognition site which is hydrophilic.

73. The surfactant-stabilized fluid interface according to claim 65, wherein the hydrophilic part of the first compound is such that the first compound is water soluble.

74. The surfactant-stabilized fluid interface according to claim 65, wherein the hydrophobic part of the first compound is covalently bonded to the molecular recognition site.

75. The surfactant-stabilized fluid interface according to claim 65, wherein a spacer is provided between the hydrophobic part of the first compound and the molecular recognition site of the first compound.

76. The surfactant-stabilized fluid interface according claim 75, wherein the spacer is one of flexible and rigid.

77. The surfactant-stabilized fluid interface according to claim 65, wherein the first compound additionally contains a functional moiety.

78. The surfactant-stabilized fluid interface according to claim 75, wherein the spacer between the hydrophobic part of the first compound and the molecular recognition site of the first compound is a polymer.

79. The surfactant-stabilized fluid interface according to claim 65, wherein the molecular recognition site of the first compound is a polymer.

80. The surfactant-stabilized fluid interface according to claim 65, wherein the first compound consists of the hydrophobic part, the molecular recognition site and optionally a linker between the hydrophobic part and the molecular recognition site.

81. The surfactant-stabilized fluid interface according to claim 65, wherein the hydrophobic part of the first compound has at least one structural element based on one of cholesterol, tocopherol, ibuprofen, palmitate, stearate, trolox or porphyrine, hydrophobic amino acids or proteins, or derivatives of any of these compounds.

82. The surfactant-stabilized fluid interface according to claim 81, wherein the hydrophobic amino acid is at least one selected from the group consisting of methionine, phenylalanine, valine, leucine isoleucine, glycine, alanine and proline.

83. The surfactant-stabilized fluid interface according to claim 65, wherein the surfactant is made of a diblock copolymer comprising a hydrophobic perfluorinated polymer block and a hydrophilic polyether glycol block.

84. The surfactant-stabilized fluid interface according to claim 65, wherein the surfactant is made of a triblock copolymer comprising two hydrophobic perfluorinated polymer end blocks and there between a hydrophilic polyether glycol block.

85. The surfactant-stabilized fluid interface according to claim 65, wherein the surfactant comprises a statistic copolymer consisting of a combination of a diblock copolymer comprising a hydrophobic perfluorinated polymer block and a hydrophilic polyether glycol block and a triblock copolymer comprising two hydrophobic perfluorinated polymer end blocks and there between a hydrophilic polyether glycol block.

86. The surfactant-stabilized fluid interface according to claim 83, wherein the perfluorinated polymer block is a perfluorinated polyether block (PFPE).

87. The surfactant-stabilized fluid interface according to claim 65, wherein the surfactant comprises at least one of a polyether glycol (PEG) and a polyetheramine block.

88. The surfactant-stabilized fluid interface according to claim 65, wherein at least one of the hydrophobic blocks of the surfactant is fluorinated.

89. The surfactant-stabilized fluid interface according to claim 65, wherein the surfactant is a block copolymer comprising one hydrophilic block having a polyethylene glycol or polyetheramine segment and comprising another hydrophobic block which is fluorinated.

90. The surfactant-stabilized fluid interface according to claim 65, wherein the block copolymer is a diblock copolymer or a triblock copolymer.

91. The surfactant-stabilized fluid interface according to claim 65, wherein one or both of the first and second fluid are water-based.

92. The surfactant-stabilized fluid interface according to claim 65, wherein one or both of the first and second fluid are oil-based.

93. The surfactant-stabilized fluid interface according to claim 65, wherein the first fluid is water-based and the second fluid is oil-based.

94. The surfactant-stabilized fluid interface according to claim 65, wherein the first fluid is oil-based and the second fluid is water-based.

95. The surfactant-stabilized fluid interface according to claim 65, wherein the first or second fluid contains a second compound comprising a further molecular recognition site which is capable of selectively binding to the molecular recognition site of the first compound and optionally carries a covalently bound functional moiety.

96. The surfactant-stabilized fluid interface according to claim 65, wherein the molecular recognition site of the first compound is designed to form hydrogen bonds with the molecular recognition site of the second compound.

97. The surfactant-stabilized fluid interface according to claim 95, wherein the second compound is contained in that of the first and second fluid into which the molecular recognition site of the first compound extends from the layer of surfactant.

98. The surfactant-stabilized fluid interface according to claim 95, wherein the molecular recognition site of the first compound and of the second compound are based on nucleic acids.

99. The surfactant-stabilized fluid interface according to claim 95, wherein the molecular recognition sites of the first compound and of the second compound are based on DNA, RNA, XNA or antisense oligonucleotides.

100. The surfactant-stabilized fluid interface according to claim 95, wherein the molecular recognition sites of the first compound and of the second compound comprise 4 or more nucleobases.

101. The surfactant-stabilized fluid interface according to claim 95, wherein the molecular recognition site of the first compound comprises a sequence of nucleobases and the molecular recognition site of the seconds compound comprises a sequence of nucleobases, and wherein the sequence of the nucleobases of the molecular recognition site of the first compound is at least partially complementary to the sequence of the nucleobases of the molecular recognition site of the second compound.

102. The surfactant-stabilized fluid interface according to claim 95, wherein the molecular recognition site of the first compound is capable to form Watson-Crick base pairing with the molecular recognition site of the second compound.

103. The surfactant-stabilized fluid interface according to claim 65, wherein the molecular recognition site of the first compound comprises between 4 or more nucleobases and 100 or less nucleobases.

104. The surfactant-stabilized fluid interface according to claim 65, wherein the molecular recognition site of the first compound extends from the layer of surfactant into the first fluid.

105. The surfactant-stabilized fluid interface according to claim 65, wherein the molecular recognition site of the first compound extends from the layer of surfactant into the second fluid.

106. The surfactant-stabilized fluid interface according to claim 95, wherein the molecular recognition site of the second compound is bound to a functional moiety.

107. The surfactant-stabilized fluid interface according to claim 95, wherein the functional moiety bound to the second compound comprises one of reactive groups, amine-groups, carboxylic groups, thiol groups and the like, or a DNA nanostructures, or microspheres, or nanoparticles, or proteins, or actin filaments or living cells.

108. The surfactant-stabilized fluid interface according to claim 95, wherein the functional moiety bound to the second compound is one of either amine-groups, a DNA lattice, plain microspheres, actin filaments or living cells.

109. The surfactant-stabilized fluid interface according to claim 95, wherein the second compound consists of a molecular recognition site and a functional moiety.

110. The surfactant-stabilized fluid interface according to claim 95, wherein the molecular recognition site of the second compound is complementary to the molecular recognition site of the first compound.

111. The surfactant-stabilized fluid interface according to claim 65, wherein the surfactant stabilized fluid interface encapsulates an emulsion droplet having a center comprising the first fluid, wherein the emulsion droplet comprises a third compound having a molecular recognition site and a functional moiety, wherein the molecular recognition site of the third compound is complementary to the molecular recognition site of the first compound.

112. The surfactant-stabilized fluid interface according to claim 65, wherein the layer of surfactant is substantially free of lipids.

113. The surfactant-stabilized fluid interface according to claim 65, wherein the first fluid is identical to the second fluid.

114. The surfactant-stabilized fluid interface according to claim 65, wherein the first fluid is different from the second fluid.

115. Dispersion comprising one or more droplets having a surfactant stabilized fluid interface, the surfactant stabilized fluid interface comprising a layer of surfactant and a first compound containing a hydrophobic part covalently linked to a molecular recognition site, wherein the surfactant-stabilized fluid interface has a first fluid on one side, a second fluid on the other side and a surfactant, wherein the surfactant stabilizes the fluid interface, wherein the hydrophobic part of the first compound is interacting with the layer of surfactant by secondary non-covalent interactions and the molecular recognition site of the first compound extends from the layer of surfactant, wherein the surfactant is a block-copolymer having at least one hydrophilic block and at least one hydrophobic block, and wherein the surfactant stabilized fluid interface encapsulates the one or more droplets, wherein the droplets comprise the first fluid and wherein the second fluid is the continuous phase of the dispersion.

116. The dispersion according to claim 115, wherein the dispersion comprises the droplets in an amount of 1015 or less droplets per ml.

117. Separating process using a surfactant-stabilized fluid interface for separating a component A from a first fluid, the surfactant-stabilized fluid interface comprising a layer of surfactant and a first compound containing a hydrophobic part covalently linked to a molecular recognition site, wherein the surfactant-stabilized fluid interface has a first fluid on one side, a second fluid on the other side and a surfactant, wherein the surfactant stabilizes the fluid interface, wherein the hydrophobic part of the first compound is interacting with the layer of surfactant by secondary noncovalent interactions and the molecular recognition site of the first compound extends from the layer of surfactant, wherein the surfactant is a block-copolymer having at least one hydrophilic block and at least one hydrophobic block, wherein the method comprises the following steps: providing the first fluid in which the component A and a first compound are dissolved or dispersed, wherein the first compound comprises a hydrophobic part and a molecular recognition site, wherein the hydrophobic part is covalently linked to the molecular recognition site, and wherein the molecular recognition site is capable of binding component A; providing a second fluid in which a surfactant which is a block-copolymer having at least one hydrophilic block and at least one hydrophobic block is dissolved; and bringing the first and second fluid into contact with each other; wherein the second fluid is immiscible with the first fluid such that a phase separation between the first and second fluid occurs and an interface between the first and second fluid is formed when the first and second fluid contact each other, and wherein the interface between the first and second fluid is the surfactant-stabilized fluid interface.

118. The separating process according to claim 117, wherein the method further comprises an additional step of mixing the first and second fluid to form a dispersion comprising one or more droplets having the surfactant stabilized fluid interface, wherein the surfactant stabilized fluid interface encapsulates the one or more droplets, wherein the droplets comprise the first fluid and wherein the second fluid is the continuous phase of the dispersion.

119. The separating process according to claim 117, wherein the method further comprises an additional step of destabilizing the surfactant-stabilized fluid interface formed between the first and second fluid.

120. Separating process according to claim 119, wherein between mixing and adding of the destabilizing surfactant an incubation step is performed.

121. Separating process according to claim 117, wherein after forming the surfactant-stabilized fluid interface one or more external stimuli for releasing component A from the molecular recognition site of the first compound is applied.

Description

[0037] Now the invention will be described in more detail also by referring to Figures and Examples which, however, do not limit the present application.

[0038] FIG. 1 shows the interaction of cholesterol-tagged and cholesterol-free DNA with surfactant-stabilized droplets according to the invention. Scale bars: 10 μm.

[0039] FIG. 2 shows FRAP measurements of microfluidic water-in-oil droplets according to the invention functionalized with cholesterol-tagged At-to-488-labeled single-stranded DNA. The three phases of the experiment (i-iii) are indicated.

[0040] FIG. 3 shows how microfluidic water-in-oil droplets according to the invention can be functionalized with various components via complementary DNA-tags. Scale bars in the lower row of images are: 30 μm in a and b, 20 μm in c and e, 5 μm in d.

[0041] FIG. 4 shows the temperature-responsive reversibility of droplet functionalization.

[0042] FIG. 5 shows a schematic illustration of the DNA lattice design, adapted from Kurokawa et al..sup.3

[0043] FIG. 6 shows a schematic layout of microfluidic devices for the production of surfactant-stabilized droplets according to the invention.

[0044] FIG. 7 shows interaction of cholesterol-tagged (according to the invention) and cholesterol-free DNA with surfactant-stabilized droplets employing a fluorescence labeled aqueous phase. Scale bars: 10 μm.

[0045] FIG. 8 shows that in microdroplets according to the invention cholesterol-tagged DNA remains stable at the droplet interface. Scale bars: 50 μm.

[0046] FIG. 9 is a graph showing the diffusion coefficient of cholesterol-tagged At-to-488-labelled DNA encapsulated in surfactant-stabilized droplets as a function of surfactant concentration (in weight percent).

[0047] FIG. 10 shows confocal fluorescence imaging of droplets according to the invention containing cholesterol-tagged DNA and a complementary DNA strand. Scale bar: 50 μm.

[0048] FIG. 11 shows that cholesterol-tagged DNA binds polystyrene microspheres. Scale bar: 30 μm.

[0049] FIG. 12 shows the functionalization of water-in-oil droplets according to the invention with beads (diameter: 2 μm) via DNA-tags. Scale bar: 30 μm in a,b,c; 10 μm in d.

[0050] FIG. 13 shows the functionalization of water-in-oil droplets according to the invention with carboxylate-modified beads (diameter: 0.3 μm) via DNA-tags. Scale bar: 30 μm.

[0051] FIG. 14 shows the formation of cortex-like actin structures inside droplets according to the invention via DNA-tags. Scale bars: 50 μm in a) and 5 μm in the other images.

[0052] FIG. 15 shows that cholesterol-tagged DNA attaches to Jurkat cells. Scale bars: 3 μm in a, b, d and e; 10 μm in c and f.

[0053] FIG. 16 shows that Jurkat cells are located at the bottom of surfactant-stabilized droplets in the absence of DNA linkers. Scale bar: 30 μm.

[0054] FIG. 17 shows that Jurkat cells can form droplet spanning multicellular clusters in the presence of the DNA linkers in droplets according to the invention. Scale bar: 20 μm.

[0055] FIG. 18 schematically shows a separating process according to the invention.

[0056] FIG. 1 shows the interaction of cholesterol-tagged and cholesterol-free DNA with surfactant-stabilized droplets according to the invention. Cholesterol-tagged DNA is an example for a first compound containing a hydrophobic part covalently linked to a molecular recognition site, wherein the cholesterol tag is the hydrophobic part and a DNA fragment is the molecular recognition site. In image a) of FIG. 1 a schematic illustration of a water-in-oil droplet is provided, wherein the part left of the red dotted line indicates a droplet with cholesterol-tagged DNA and the right side of the dotted line indicates a droplet with cholesterol-free DNA. Water is for example the first fluid and oil is for example the second fluid. Images b and c in FIG. 1 show confocal fluorescence images of a water-in-oil droplet with cholesterol-tagged Cy3-labeled DNA (with SEQ ID No. 1; =550 nm) (according to the invention) (b) or with cholesterol-free Atto488-labeled DNA (with SEQ ID No. 2; =488 nm) added to the internal aqueous phase (c) (not according to the invention). In image d) of FIG. 1 a schematic illustration of an oil-in-water droplet is provided, wherein the part left of the red dotted line indicates a droplet with cholesterol-tagged DNA and the right side of the dotted line indicates a droplet with cholesterol-free DNA. A confocal fluorescence image of an oil-in-water droplet with cholesterol-tagged Cy3-labeled DNA (with SEQ ID No. 1) added to the external aqueous phase is shown in image e of FIG. 1 and a corresponding image of a droplet with cholesterol-free Atto488-labeled DNA (with SEQ ID No. 2) added to the external aqueous phase is shown in image f of FIG. 1. The images a to f of FIG. 1 indicate that cholesterol-tagged DNA self-assembles at the droplet periphery, whereas cholesterol-free DNA remains homogeneously distributed in the aqueous phase images b, c, e and f were produced using 10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8. Scale bars in these images are 10 μm.

[0057] In part g of FIG. 1 an enlarged view of a section of a microfluidic droplet 10 according to the invention is schematically shown. The microfluidic droplet 10 comprises a center 12 encapsulated by a layer of surfactant 14. The microfluidic droplet 10 further comprises a first compound 16 containing a hydrophobic part 18 covalently linked to a molecular recognition site 20. The center 12 of the microfluidic droplet 10 contains a first fluid 22, e.g. a polar liquid such as an aqueous solution or water. The microfluidic droplet 10 is surrounded by a second fluid 24 which is different from the first fluid 22, e.g. a hydrophobic liquid such as oil. The layer of surfactant 14 stabilizes the droplet 10 in the second fluid 24. The layer of surfactant is of a block-copolymer 28 having at least one hydrophilic block 30 and at least one hydrophobic block 32. Hydrophilic blocks 30 and hydrophobic blocks 32 phase separate from each other so that the hydrophilic blocks are located on the site of the layer of surfactant 14 which faces that of the first and second fluid 22, 24 which is more hydrophilic. In the situation schematically illustrated in part g of FIG. 1, the first fluid 22 is more hydrophilic than the second fluid 24 so that the molecular recognition site 20 of the first compound 16 extends away from the layer of surfactant 14—in which the hydrophobic part 18 is anchored—into the center 12 of the microfluidic droplet 10.

[0058] As illustrated in part g of FIG. 1, the microfluidic droplet can contain a second compound 34. The second compound 34 comprises a further molecular recognition site 36 which is capable of selectively binding to the molecular recognition site 20 of the first compound 16. The second compound 34 illustrated in part g of FIG. 1 further comprises a functional moiety 38 which is linked to the molecular recognition site 36. At an interface 40 between the first compound 16 and the second compound 34 interactions occur which attract the first compound 16 and the second compound 34 to one another. The interaction at the interface 40 is not based on covalent bonds, so that the agglomerate of the first compound 16 and of the second compound 34 can disintegrate so that the second compound 34 is released from the layer of surfactant 14. The interaction at the interface 40 can be based for example on hydrogen bonds between nucleobases 42, 44 provided at the molecular recognition sites 20, 36 of the first and second compounds 16, 34.

[0059] FIG. 2 shows FRAP measurements of microfluidic water-in-oil droplets functionalized with cholesterol-tagged Atto488-labeled single-stranded DNA. a) Confocal fluorescence images of a droplet (bottom plane) i: before bleaching, ii: directly after bleaching (circular bleaching area clearly visible, λ ex=488 nm), and iii: after recovery. The bleached area is highlighted with a red dashed circle. b) Representative recovery curve. Mean normalized intensity values within the bleaching area are plotted as a function of time. The solid line represents an exponential fit, which was used to determine the diffusion coefficient, here: D=0.412 μm.sup.2s.sup.−1. The three phases of the experiment (i-iii) are indicated.

[0060] FIG. 3 shows how microfluidic water-in-oil droplets according to the invention can be functionalized with various components via complementary DNA-tags. Schematic illustrations in the top row show a close-up of a single cholesterol-tagged DNA handle (correspond to the above mentioned first compound 16) attached to the droplet periphery interacting with the added component. The bottom row shows confocal fluorescence images of entire water-in-oil droplets, in groups or as a close up of a single droplet. Added components were as follows: a) complementary amine-tagged DNA (as above mentioned second compound 34 having SEQ ID No. 3 Atto488-labeled, compound #3); b) an Atto488-labeled hexagonal DNA-lattice formed from compounds #4i to #4iii having SEQ ID No. 4 to 6; c) plain multifluorescent polystyrene microspheres grafted with complementary FITC-labeled DNA (compound #8 with SEQ ID No. 10; =488 nm) via hydrophobic interactions; d) filamentous actin (containing 10 mol % biotinylated actin and 1 mol % Atto488-labeled actin). The actin filaments were functionalized with complementary biotinylated DNA via biotin-streptavidin linkage; e) T-lymphocyte (Jurkat) incubated with complementary cholesterol-tagged DNA. The cholesterol-tagged DNA inserts into the lipid membrane of the cells and through two complementary DNA strands binds the cholesterol-tagged DNA on the periphery of the droplets (see FIG. 15). Scale bars in the lower row of images are: 30 μm in a and b, 20 μm in c and e, 5 μm in d. For DNA sequences see Supporting Information, Table 1 and Sequence listing.

[0061] FIG. 4 shows the temperature-responsive reversibility of droplet functionalization. a) Confocal fluorescence images of cholesterol-tagged DNA (top row, labeled with Atto488, λ.sub.ex=488 nm) and complementary DNA (bottom row, labeled with ROX, λ.sub.ex=580 nm) in the same surfactant-stabilized droplets for two consecutive temperature cycles (heating from 20° C. to 60° C., cooling to 20° C., then heating to 60° C.). Scale bars: 30 μm. b) Average fluorescence intensity within the droplets (periphery excluded) during the temperature cycles for the cholesterol-tagged DNA (orange) and its complementary strand (green) as a function of time. The temperature is plotted below. Error bars correspond to the standard deviation of 17 measurements. The time points when the four images shown in a) were taken are indicated. The complementary strand binds to the cholesterol-tagged DNA handle at the periphery at lower temperatures. Above the melting temperature of the duplex, the complementary cholesterol-free DNA was homogeneously distributed within the aqueous phase of the same droplet (10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8). Furthermore, the cholesterol-tagged DNA anchor remained at the droplet periphery, showing the high entropical benefit of cholesterol-insertion into the surfactant layer. For DNA sequences, see Supporting Information, Table 1.

[0062] In FIG. 5 a schematic illustration of the DNA lattice design, adapted from Kurokawa et al..sup.3 is provided. The lattice is composed of three DNA sequences (see Table 1, compounds #4i, ii, and iii containing SEQ ID No. 4 to 6) with sticky ends for polymerization into hexagonal lattices. One of the DNA strands has a 3′ cholesterol tag (red, oval shape), another is tagged with Atto488 (green circle). The lattice was assembled by mixing the three strands at equimolar concentrations of 2 μM in a buffer containing 10 mM Tris, 1 mM EDTA and 10 mM MgCl.sub.2. The mixture was annealed in a thermocycler (Biorad) by heating to 60° C. for 10 minutes and subsequently encapsulated into microfluidic droplets via the aqueous phase.

[0063] FIG. 6 provides a schematic drawing of the layout of microfluidic devices for the production of surfactant-stabilized droplets. a) Device with one oil-inlet and one water-inlet used for the encapsulation of cholesterol-tagged DNA as well as the joint encapsulation of cholesterol-tagged and complementary DNA (amine-tagged, cholesterol-tagged, fluorescently-labeled DNA and the DNA lattice); b) Device with one oil-inlet and two water-inlets used for actin and bead encapsulation with DNA. The cholesterol-tagged DNA was supplied via one inlet and the respective complementary DNA strands via the second one to avoid aggregation prior to droplet formation; c) Device with a coiled water inlet used for Jurkat cell encapsulation. The coil provides better separation of the cells. The microfluidic PDMS devices (Sylgard184, Dow Corning, USA) were fabricated according to a previously published protocol..sup.43 For confocal imaging, the droplets were collected from the outlet and sealed in a simple observation chamber as described previously..sup.44

[0064] FIG. 7 shows the interaction of cholesterol-tagged and cholesterol-free DNA with surfactant-stabilized droplets employing a fluorescence labeled aqueous phase. a) Schematic illustration of a water-in-oil droplet (left of dotted red line: with cholesterol-tagged DNA, right side: with cholesterol-free DNA); b,d) Confocal fluorescence images of a water-in-oil droplet with cholesterol-tagged Cy3-labeled DNA (λ.sub.ex=550 nm) (b) or cholesterol-free Atto488-labeled DNA (λ.sub.ex=488 nm) (d) in the internal aqueous phase; c,e) Aqueous phase labeled via dissolving an Alexa405 dye (λ.sub.ex=405 nm) for droplets containing cholesterol-tagged DNA (c) or cholesterol-free DNA (e) in the internal aqueous phase; f) Schematic illustration of an oil-in-water droplet (left of dotted red line: with cholesterol-tagged DNA, right side: with cholesterol-free DNA), g,i) Confocal fluorescence images of an oil-in water droplet with cholesterol-tagged Cy3-labeled DNA (g) and cholesterol-free Atto488-labeled DNA (i) in the external aqueous phase; h,j) aqueous phase labeled via an Alexa405 dye for cholesterol-tagged DNA (g) and cholesterol-free DNA (i) in the external aqueous phase. Whereas cholesterol-tagged DNA self-assembles at the droplet periphery, cholesterol-free DNA remains homogenously distributed in the aqueous phase, which consisted of 10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, 1 per mille Alexa405 dye (Thermo Fisher Scientific), pH 8. Scale bars: 10 μm.

[0065] FIG. 8 shows images demonstrating that cholesterol-tagged DNA remains stable at the droplet interface. a,b,c,d,e) Confocal fluorescence images of ATT0488-labeled cholesterol-free DNA encapsulated within water-in-oil droplets at 0 h, 24 h, 48 h, 96 h and 208 h; f,g,h,i,j) Confocal fluorescence images of cholesterol-tagged Cy3-labeled DNA encapsulated within water-in-oil droplets at 0 h, 24 h, 48 h, 96 h and 208 h. The cholesterol-free DNA remains homogeneously distributed in the droplet's lumen throughout several days, indicating both that there is no unspecific binding to the periphery and that the water-in-oil droplet remains stable. In contrast, the cholesterol-tagged DNA is only distributed along the droplets periphery showcasing the strong hydrophobic surfactant-cholesterol interaction. The aqueous phase contained 10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8 and 4 μM DNA. Scale bars: 50 μm.

[0066] FIG. 9 shows the diffusion coefficient of cholesterol-tagged Atto488-labelled DNA encapsulated in surfactant-stabilized droplets as a function of surfactant concentration (in weight percent). The error bars correspond to the standard deviation of at least 9 independent measurements. If the droplet moved during the acquisition interval, the recorded data was not taken into consideration. The data points were fitted with an exponential decay function (solid line).

[0067] FIG. 10 provides a confocal fluorescence imaging of droplets containing cholesterol-tagged DNA and a complementary DNA strand. a) Cholesterol-tagged DNA labeled with Cy3 at the droplet periphery; b) Complementary DNA labeled with Cy5 without a cholesterol-tag in the same water-in-oil droplets (containing 10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8). The complementary DNA co-localizes with the cholesterol-tagged DNA, indicating successful DNA duplex formation. Scale bar: 50 μm.

[0068] FIG. 11 shows that cholesterol-tagged DNA binds polystyrene microspheres. a) Fluorescent (YG) microspheres in an aqueous solution (10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8); b) Cy5-labeled DNA attached to the same beads via cholesterol-tagged DNA; c,d) Same as (a,b) but at a higher bead concentration, resulting in increased bead-bead interactions and clustering. Scale bar: 30 μm.

[0069] FIG. 12 shows images of the functionalization of water-in-oil droplets with beads (diameter: 2 μm) via DNA-tags. a,b) Fluorescent (YG) microspheres (green) in water-in-oil droplets containing an aqueous solution (10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8) without cholesterol-tagged DNA (control). For clarity, brightfield- and fluorescence images were overlaid. The beads were distributed randomly, neither favoring the droplet lumen nor its periphery. c,d) Fluorescent (YG) microspheres (green) in an aqueous solution containing cholesterol-tagged as well as a complementary Cy5-labeled DNA (red). The beads are localized at the droplet periphery. Scale bar: 30 μm in a,b,c; 10 μm in d.

[0070] FIG. 13 shows the functionalization of water-in-oil droplets with carboxylate-modified beads (diameter: 0.3 μm) via DNA-tags. a) Fluorescent (YG) carboxylate-modified microspheres (green) in water-in-oil droplets. The aqueous solution consists of 10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8 with added DNA, namely cholesterol-tagged DNA (compound #7 in table 1) and its complementary Cy5-labeled DNA (red, compound #7 in table 1) but without the interconnecting strand (compound #8 from table 1). For clarity, the fluorescence images were overlaid. The beads were distributed randomly, neither favoring the droplet lumen nor its periphery. b) Fluorescent (YG) microspheres (green) in an aqueous solution containing cholesterol-tagged DNA and its complementary Cy5-labeled DNA (red) as well as the interconnecting strand (compound #8 from table 1). The beads were localized at the droplet periphery. This opens up opportunities for diverse means of functionalizing droplets via modified beads. Scale bar: 30 μm.

[0071] FIG. 14 shows the formation of cortex-like actin structures inside droplets via DNA-tags. a) Filamentous (F-) actin inside surfactant-stabilized droplets without DNA. Actin filaments were distributed homogenously within droplets; b,c) F-actin inside surfactant-stabilized droplets with 10% biotinylated actin binding a strand of biotinylated DNA via streptavidin. The filaments are tethered to the droplet periphery via a complementary cholesterol-tagged DNA handle (like the one shown in FIG. 3d). F-actin is found at the droplet interface. Scale bars: 50 μm in a) and 5 μm in the other images. For the actin polymerization protocol see the below Material and Methods section.

[0072] FIG. 15 shows that cholesterol-tagged DNA attaches to Jurkat cells. a) Cy3-labeled cholesterol-tagged DNA mixed with the cell culture medium attached to the cell membrane; b) ROX-labeled DNA without a cholesterol-tag, added with the cell culture medium was homogeneously distributed in the extracellular space; c) The ROX-labeled strand attached to the cell periphery in the presence of its complementary cholesterol-tagged DNA strand; d-f) Corresponding brightfield images. Scale bars: 3 μm in a, b, d and e; 10 μm in c and f.

[0073] FIG. 16 shows that Jurkat cells are located at the bottom of surfactant-stabilized droplets in the absence of DNA linkers. a,b,c) Jurkat cells were encapsulated in droplets with RPMI cell culture medium without the addition of DNA. The cells sank to the bottom of the droplets indicating a lack of directed interaction with the droplet interface. This is the case both in single cells (a) and at higher cell concentrations (b,c). Scale bar: 30 μm.

[0074] FIG. 17 shows that Jurkat cells can form droplet spanning multicellular clusters in the presence of the DNA linkers. a,b) A high concentration of Jurkat cells (10.sup.7 per milliliter) encapsulated in droplets with RPMI cell culture medium together with the DNA linkers. A representative composite brightfield and fluorescence image (DNA is FITC-labeled) is shown. Higher cell and DNA concentrations result in an increase of cell-cell interactions, resulting in droplet-spanning multicellular clusters. Scale bar: 20 μm.

[0075] FIG. 18 shows a schematic illustration of the separating process according to the present invention. In a first step A an aqueous phase (first fluid 22) and an oil phase (second fluid 24) are provided such that these phases are in contact with each other and an interface is formed there between. The first fluid 22 comprises component A 42 and a further component B 44 and the second fluid 24 comprises a surfactant 28. A first compound 16 containing a hydrophobic part covalently linked to a molecular recognition site which can selectively bind the component A 42 that is supposed to be removed from the mix is contained in the first fluid 22. The first compound 16 attracts component A 42 to the fluid interface and the supernatant—containing only the remaining components (component B 44, single or multiple components)—can be recovered. The separation efficiency depends on the interface area between the first fluid 22 and the second fluid 24. This area can be increased drastically by forming emulsion droplets. Vigorous shaking or vortexing induces the formation of a dispersion 46 containing emulsion droplets—drastically increasing the surface area for attachment of the first compound 16 and the component A. The same can be achieved with microfluidics. After an incubation period (normally minutes are sufficient), the droplets are broken up by addition of a destabilizing surfactant (e.g. perfluoro-1-octanol (PFO)). As before, the remaining components can be recovered. It is also possible to recover component A 42 by subsequently denaturing the binding site (e.g. by heating, enzyme digestion, not shown in FIG. 18).

[0076] The separating process is applicable for any fluid purification, as long as it is possible to couple a molecular recognition site for the component of interest to a hydrophobic part. Thereby, the size and the molecular weight of the components can range from atoms to cells: Applications could include water purification (e.g. removal of contaminants like heavy metals), filtration of body fluids (e.g. removal of pathogenic components), cell sorting (e.g. removal of bacteria, sorting of mixed cell populations according to their surface markers) and the like.

[0077] Here, we present a broadly applicable method for functionalizing microfluidic droplets utilizing the hydrophobic interaction of cholesterol-tagged DNA with the droplet-stabilizing surfactant. We show that DNA handles can serve as reversible anchoring points for various components including reactive groups, DNA nanostructures, beads, proteins or even cells. The use of off-the-shelf available DNA holds considerable advantages compared to standard methods for droplet functionalization, including: the broad scope of options for site-directed chemical functionalization, the addressability and programmability due to specific base pairing as well as the reversible stimuli-responsive binding properties of DNA. Therefore, cholesterol-tagged DNA handles bear great potential for various applications in droplet-based microfluidics. An overview of DNA handles described below is provided in Table 1.

TABLE-US-00001 TABLE 1 SEQ Com- ID pound No. # DNA sequence Fig.  1  1 5′Cy3 (or Atto488)/TTCCTTCTATGCATCA/3′CholTEG 1b, e; 2; 3d  2  2 TCTATGCATCA/3′Atto488 1c, f  3  3 5′Atto488/TGATGCATAGAAGGAA/3′Amine 3a  4  4i 5′Atto488/GCTCGAGCCAGTGAGGACGGAAGTTTGTCGTAGCATCGCACC 3b  5  4ii GCTCGAGCCAACCACGCCTGTCCATTACTTCCGTCCTCACTG/3′CholTEG 3b  6  4iii GCTCGAGCGGTGCGATGCTACGACTTTGGACAGGCGTGGTTG 3b  7  5 ACCAGACAATACCACACAATTTT/3′CholTEG 3b  8  6 5′Atto488/TTCTCTTCTCGTTTGCTCTTCTCTTGTGTGGTATTGTCTAAGAG 3d AAGAGTT/3′BioTEG/  9  7 CTACTATGGCGGGTGATAAAAAACGGGAAGAGCATGCCCATCCAA/3CholTEG 3d 10  8 5′FITC/GGATGGGCATGCTCTTCCCGTTTTTTATCACCCGCCATAGTAGGAG 3c, e GTAAGTTATGACAGGTCCA 11  9 5′Cy5/GGATGGGCATGCTCTTCCCGTTTTTTATCACCCGCCATAGTAGAGGAC 3c, e CTGTCATAACTTACCTG 12 10 5′Atto488/TTCCTTCTATGCATCA/3′CholTEG 4 13 11 TCTATGCATCA/3′ROX 4

Results and Discussion

[0078] Surface-functionalization of surfactant-stabilized droplets with cholesterol-tagged DNA

[0079] Droplet-based microfluidics relies on amphiphilic polymer-based agents (surfactants) for the formation of stable water-in-oil or oil-in-water emulsion droplets. We designed a short 3′ cholesterol-tagged DNA oligomer to investigate its self-assembly into the surfactant-layer of microfluidic droplets. The 16 base-long DNA (for DNA sequences see Table 1 and sequence listing) additionally contained a covalently linked fluorophore (Cy3) for visualization purposes. Having tested that the fluorescent dye does not influence the binding properties of the DNA to the droplet periphery, an Atto488-tagged DNA oligomer without cholesterol served as a control.

[0080] FIG. 1a shows a sketch of a DNA-containing water-in-oil droplet. The DNA is supplied via the aqueous inlet (2 μM DNA in 10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8) of a microfluidic droplet-production device (flow focusing PDMS device, see Supporting Information, FIG. 6). The confocal image in FIG. 1b shows the self-assembled cholesterol-tagged DNA at the water-facing periphery of the surfactant-stabilized droplet. This process happens within milliseconds after droplet formation (see Supporting Information). The cholesterol-tagged DNA remains stably attached to the droplet interface for days (FIG. 8). This is in accordance with interfacial tension (IFT) measurements, which showed no dramatic change of IFT in droplets containing cholesterol-tagged DNA compared to droplets containing cholesterol-free DNA (see Supporting Information, Table 2). In contrast to cholesterol-tagged DNA, cholesterol-free single-stranded DNA is homogeneously distributed in the aqueous phase inside the droplet (FIG. 1c, for clarity the aqueous phase was labeled with a fluorescent dye in FIG. 7). This proves that the self-assembly of the DNA at the droplet periphery is due to hydrophobic interactions between the DNA-linked cholesterol moiety and the surfactant molecules—rather than electrostatic interactions between DNA and surfactants. To broaden the range of possible applications of DNA-functionalized droplets, we demonstrate that our system also works for oil-in-water droplets (see Materials and Methods). FIG. 1e shows cholesterol-tagged DNA attached to the aqueous exterior of the droplet, while the cholesterol-free DNA remains homogeneously distributed in the aqueous phase surrounding it (FIG. 1f).

[0081] To gain more insights into the interaction of the cholesterol-tagged DNA and the droplet periphery, we set out to probe the mobility of the DNA in the surfactant layer. To this end, we performed fluorescence recovery after photobleaching (FRAP) measurements with the water-in-oil droplets containing cholesterol-tagged DNA. A confocal plane at the bottom of the droplet was selected and a circular bleaching area (5 μm diameter) was defined. The measurement procedure is illustrated in FIG. 2a. It consists of a pre-bleaching time span (i), followed by bleaching of the circular area (ii), and a recovery period (iii). The mean intensities acquired from the images before bleaching were used to calculate normalized intensity values. From an exponential fit of the normalized recovery intensities, the diffusion coefficient was calculated (for more details please see Supporting Information below).

[0082] A representative FRAP intensity curve as a function of time is shown in FIG. 2b. First, it should be noted that the cholesterol-tagged DNA is indeed diffusive when incorporated into the droplet periphery (FIG. 2a). We obtained a diffusion coefficient of D=(0.41±0.01) μm.sup.2s.sup.−1 based on 17 independent experiments. This value is comparable to the diffusion coefficient of block copolymer surfactants,.sup.6 but two orders of magnitude smaller than typical diffusion coefficients of highcholesterol content lipid membranes..sup.28 We further demonstrated that the diffusion coefficient increases with decreasing surfactant concentrations (Supporting Information, and FIG. 9).

Droplet Functionalization Via DNA Handles

[0083] In order to verify that the single-stranded DNA handles at the droplet periphery are accessible for duplex formation, a complementary Atto488-labeled DNA sequence (without any cholesterol-tag) was supplied via the aqueous phase during droplet production. In the presence of the DNA handles (which were not fluorescently labeled, and thus are not visible in the confocal images), the cholesterol-free DNA strand was found at the droplet periphery, indicating successful duplex formation (Supporting Information, FIG. 10). Without the DNA handles, the cholesterol-free DNA was homogeneously distributed within the droplet as shown in FIG. 1c.

[0084] With this system in place, we made use of the large toolbox of chemical modifications available for DNA and functionalized the droplet periphery with a variety of components. A broadly applicable modification is the attachment of commonly used functional groups. One example is a commercially available amine-terminated DNA oligo (see Materials and Methods). We used confocal fluorescence imaging to confirm the successful attachment of the amine-modified DNA (labeled with Atto488) to the unlabeled DNA handles at the droplet periphery (see FIG. 3a). As a next step, we demonstrated that the DNA handles can also serve as anchoring points for DNA nanostructures. For this purpose, we assembled a hexagonal Atto488-labeled DNA lattice (see Supporting Information, FIG. 5 and Table 1). Its adhesion to the compartment periphery is visible in the confocal image in FIG. 3b. Beyond functional groups and DNA nanostructures, we employed the cholesterol-tagged DNA handles as attachment points for polystyrene microspheres with a diameter of 2 μm (see Materials and Methods). For this purpose, we designed a longer DNA-based linker, which provides more space between the microspheres and the droplet periphery. The linker is made up of two cholesterol-tagged DNA duplexes that each have a single-stranded DNA overhang. These overhangs can interconnect, creating a strong linker, as shown in the illustration in FIG. 3c (for sequences see Table 1 and sequence listing). This linking method requires just one rather than two different cholesterol-tagged DNA sequences, making it more cost-effective than the direct link. The microspheres were first incubated with one of the cholesterol-tagged duplexes. As shown in FIG. 11, the DNA adheres to the surface of the microsphere. The second cholesterol-tagged DNA duplex served as an attachment point at the droplet periphery, binding DNA-coated microspheres via the single-stranded DNA overhang. As visible in the confocal image in FIG. 3c, the microspheres attached to the droplet periphery, whereas when DNA was missing, they were distributed randomly (see FIG. 12). Note that it is also possible to use modified microspheres of different sizes carrying functional groups (see FIG. 13). Next, we went one step further by attempting to assemble an actin cortex inside the microfluidic droplet. Again, our approach relied on the DNA-handle-system, which we modified to link actin filaments to the droplet periphery. We first polymerized a mixture of G-actin monomers that were modified with either Alexa488 or biotinylation or that were left unmodified into up to 5 μm long filaments. This mixture was then added to a solution containing biotinylated DNA and streptavidin and encapsulated into the droplets. The confocal image in FIG. 3d shows that we were successful in binding actin filaments to the droplet periphery, building up a well-defined actin cortex. The filaments are arranged in a spherical manner at the droplet periphery, whereas excess labeled monomers can be found in the droplet lumen. In a control without the DNA handles, the actin filamets remained homogeneously distributed within the droplet (FIG. 14). In living cells, the binding of actin to the periphery involves a multitude of proteins. Remarkably, we have achieved an artificial imitation using much fewer components—demonstrating the potential of DNA-based linkers in bottom-up synthetic biology.

[0085] Increasing the level of complexity one step further, we went from the attachment of purified proteins to living cells. Droplets have previously been used as microreactors for cells to perform high-throughput single cell analysis. Nonetheless such applications often require controlled interactions between encapsulated cells and the droplet periphery. We therefore set out to test if our DNA-handle-system is capable of anchoring living cells at the droplet periphery. We first incubated Jurkat cells with a cholesterol-tagged DNA duplex as used previously for the microspheres. The linkage via four rather than two DNA strands ensures a larger distance between the cell and the droplet periphery, leaving the cells in their native environment and avoiding unwanted interactions with the surfactant molecules. The confocal images in FIG. 15 show the DNA inserting into the lipid bilayer membrane of the cells. As with the microspheres, we could bind the cells to the DNA-functionalized droplets periphery via the single-stranded overhang on the DNA duplex. FIG. 3e shows an encapsulated Jurkat cell, which was attached to the droplet periphery via the described DNA linkers. Increasing the concentration of cells, we also observed DNA-mediated cell-cell interactions, building up droplet-spanning multicellular clusters (see FIG. 17). The fluorescent DNA clusters in the droplet lumen, visible in FIG. 3e, likely result from the use of cell culture medium containing proteins and hydrophobic molecules with which the cholesterol can interact. All these experiments exemplify the ability of DNA to act as a linker between the droplet periphery and various components making it a universally applicable tool for the functionalization of surfactant-stabilized droplets.

Temperature-Responsive Reversibility

[0086] For many applications, stimuli-responsive droplet functionalization is highly desirable. This requires mechanisms to turn interactions with the droplet periphery on and off upon demand. With the DNA-handle-system, we can make use of the temperature-dependency of Watson-Crick base pairing between strands of DNA to achieve stimuli-responsive reversibility of the binding.

[0087] We carried out temperature cycling experiments during which the DNA-containing droplets were placed on a temperature-controlled stage during confocal imaging. These experiments can reveal whether a) the cholesterol-tagged DNA is stable at the periphery even at elevated temperatures and b) the complementary DNA (which can carry diverse functional groups, as shown in FIG. 3) can be reversibly detached from the periphery. FIG. 4a shows confocal fluorescence images of the cholesterol-tagged DNA (top row, Atto488-labeled) and the complementary DNA (bottom row, ROX-labeled) encapsulated together in a droplet, during temperature cycling experiments. While the cholesterol-tagged DNA remains bound to the periphery at temperature between 20° C. and 60° C., the complementary DNA detaches at elevated temperatures. The melting temperature of the DNA duplex (i.e., the temperature when half of the complementary DNA should be unbound) formed by compounds #10 and #11 was calculated to be approximately 28.3° C. In FIG. 4b, we plotted the average fluorescence intensity within droplets (periphery excluded) and the applied temperature as a function of time. When the fluorophore-tagged DNA was bound to the periphery this value was low but it increased as more DNA detached. The intensity was measured for cholesterol-tagged Atto488-labeled DNA (orange) as well as for the complementary ROX-labeled DNA (green). The cholesterol-tagged DNA is found at the periphery at all times and temperatures—hence the fluorescence intensity in the droplet lumen remains low. This shows that the hydrophobic interaction between surfactant and cholesterol is stronger than the sum of the hydrogen bonds between all DNA bases. The insertion of the cholesterol-tagged DNA is entropically beneficial and stable. In contrast, the plot clearly shows the periodic oscillations of the complementary DNA with each temperature cycle. This means that the complementary DNA is located mainly at the periphery at 20° C. and found within the droplet lumen at higher temperatures. This is in agreement with the calculated melting temperature of the employed DNA of 28.3° C. Because of the low melting temperature, even at 20° C., not all of the complementary DNA may be bound. The binding can be enhanced by using longer DNA sequences with higher melting temperatures as we did in FIG. 3. In addition, hysteresis effects can be observed by comparing the intensity maxima with the temperature, according to which the maximum is reached when the temperature drops back to 50° C. This is conclusive considering that the system needs some time to equilibrate, which becomes more clear, when looking at the temperature of the heating plate as depicted in FIG. 4. From these experiments, we can conclude that the functionalization of the droplet periphery is reversible.

Conclusion

[0088] In the present study we demonstrated a versatile method for the functionalization of microfluidic droplets using DNA-tags. We showed that cholesterol-tagged DNA self-assembles into the periphery of surfactant-stabilized water-in-oil and oil-in-water droplets due to the hydrophobic interactions between the surfactant molecules and the cholesterol-tag at the interface between oil and water. Furthermore, by means of FRAP measurements we found that the DNA anchored in the surfactant layer is diffusive with D=(0.41±0.01) μm.sup.2s.sup.−1 and that the diffusivity depends on the surfactant concentration. Importantly, we proved that the DNA strand attached to the periphery remains fully addressable. It can act as a sequence-specific and programmable anchor for a variety of components attached to a complementary DNA strand. By creating a link between the droplet periphery and one of either amine-groups, a DNA lattice, plain microspheres, actin filaments or living cells, we showcased the versatility of our DNA handles. In principle, the system can be extended to any other functional group and component that can be linked to DNA, e.g. thiol-groups, adenylation, phosphorylation, various DNA nanostructures, gold nanoparticles, lipid vesicles or other types of proteins and cells. Conveniently, various functional groups covalently linked to DNA are commercially available. Lastly, we showed that droplet functionalization as a result of DNA duplex formation, is temperature dependent and fully reversible. The temperature at which unbinding happens is fully controllable by choosing a DNA sequence with the desirable melting point. In addition, other stimuli-responsive DNA motifs, like pH- and light-responsive elements, could be incorporated to achieve reversibility of droplet functionalization. All in all, this study shows the potential of DNA handles to specifically and reversibly functionalize surfactant-stabilized droplets for diverse applications in droplet-based microfluidics. It is to be understood that the effects demonstrated for DNA based recognition sites anchored to the periphery of microdroplets can be expected to also appear with other non-DNA based recognition sites based on nucleic acid fragments, including RNA and artificial nucleic acids, such as for example peptide nucleic acids or other molecular recognition sites, based for example on antibody-antigene recognition.

Material and Methods

Design and Assembly of DNA-Tags

[0089] A set of random fixed-length DNA sequences was generated in MATLAB (MathWorks, Inc.) using the randseq command. The sequences were analyzed in NUPACK 39 and chosen such as to provide stable base-pairing with complementary sequences at room temperature and an overall low tendency to form secondary structures and self-dimers. The selected DNA oligos are DNA SEQ ID No. 1 to 13, listed also in Table 1, were purchased from Integrated DNA Technologies, Inc. or biomers.net GmbH. HPLC purification was performed for DNA oligos carrying modifications (Cy3, Cy5, FITC, ROX, Atto488, cholesterol-TEG, biotin or amine). The DNA oligos were diluted in Milli-Q water at a stock concentration of 100 μM, aliquoted and stored at −20° C. until use. Before the experiment, cholesterol-tagged DNA oligos were heated to 60° C. for 5 minutes to reduce aggregation. A concentration of 2 μM of each oligonucleotide was used in the aqueous solution, if not stated otherwise.

Microfluidic Formation of DNA-Functionalized Surfactant-Stabilized Water-in-Oil Droplets

[0090] Microfluidic PDMS-based (Sylgard 184, Dow Corning, USA) devices for the formation of water-in-oil droplets were produced and assembled as described previously..sup.4 FIG. 6 shows a map of the device, which features a water and an oil channel that meet at a flow-focusing T-junction for droplet formation. The aqueous phase is made up of Milli-Q water containing 5 mM MgCl.sub.2, 1×+Tris-EDTA buffer, pH 8 (Buffer 1) and 2-4 μM DNA unless otherwise specified. The oil-phase contains either 2 wt % of Perflouro-polyether-polyethylene glycol (PFPE-PEG) block-copolymer fluorosurfactants (PEG-based fluorosurfactants from Ran Biotechnologies, Inc.) or 2.5 mM of a custom synthesized triblock-copolymer PFPE-PEG-PFPE triblock-copolymer surfactant.sup.1,2 dissolved in HFE-7500 oil (from DuPont). The custom synthesized triblock-copolymer PFPE-PEG-PFPE triblock-copolymer surfactant was synthesized as described by Platzman et al..sup.1 and by Janiesch et al..sup.2 The flow rates were generally set to 900 μlh.sup.−1 for the oil phase and 300 μlh.sup.−1 for the water phase using syringe pumps PUMP 11 ELITE (Harvard apparatus, USA). The fluids were injected with 1 ml syringes (Omnifix, B. Braun, Germany) connected by a cannula (Sterican R 0.4×20 mm, BL/LB, B. Braun, Germany) and PTFE-tubing (0.4×0.9 mm, Bola, Germany). To observe the production process, an Axio Vert.A1 (Carl Zeiss AG, Germany) inverse microscope was used. With these settings, homogeneous droplets with a diameter of approximately 30 μm were produced at a rate of 1 kHz.

Formation of Surfactant-Stabilized Oil-in-Water Droplets

[0091] Oil-droplets with a surrounding aqueous phase were produced manually by shaking. For this purpose, 10 μl of the surfactant-containing oil phase were added to 700 μl of the DNA-containing aqueous solution. The composition of the oil and the water phase were chosen as before. The probe was manually shaken until oil-in-water droplets formed (visible as a milky emulsion layer).

High-Speed Fluorescence Imaging

[0092] For imaging the production of water-in-oil droplets containing Cy3-labeled cholesterol-tagged or cholesterol-free DNA (10 μM), a fluorescence illuminator (HBO 100, Carl Zeiss AG, Germany), FS43HE Filter (Carl Zeiss AG, Germany) and a 40× objective (LD Plan-Neofluar 40×/0.6 Korr, Carl Zeiss AG, Germany) were used. For imaging purposes, lower flow rates (30 μlh.sup.−1 for aqueous phase and 120 μlh.sup.−1 for oil phase) were chosen.

Confocal Fluorescence Microscopy

[0093] A confocal laser scanning microscope LSM 800 (Carl Zeiss AG) was used for confocal imaging. The pinhole aperture was set to one Airy Unit and experiments were performed at room temperature. The images were acquired using a 20× (Objective Plan-Apochromat 20×/0.8 M27, Carl Zeiss AG) and a 63× immersion oil objective (Plan-Apochromat 63×/1.40 Oil DIC, Car Zeiss AG). Images were analyzed with ImageJ (NIH, brightness and contrast adjusted).

FRAP Measurements

[0094] FRAP measurements were performed using a Leica SP5 confocal microscope (Leica Microsystems, Germany), equipped with an argon laser and a 63× oil-immersion objective (HCX PL APO 63×/1.40-0.60; Leica Microsystems GmbH). Surfactant-stabilized water-in-oil droplets functionalized with compound #1 in Table 1 containing single-stranded DNA with SEQ ID No. 1 with an Atto488-tag on the 5′ end and a cholesterol-tag on the 3′ end were sealed in an observation chamber. Subsequently, a bleaching spot with a radius of 2.5 μm was defined at the confocal plane at the bottom of the droplet. Using the FRAP-WIZZARD, five images were recorded before bleaching, followed by 6.5 s bleaching and the acquisition of 35 images after bleaching. Representative images are shown in FIG. 2. The frame rate was set to 0.65 s. The diffusion coefficient, which was averaged from 17 independent measurements, was derived from the recorded images using a custom-written MATLAB (MathWorks, Inc.) code as detailed in the Supporting Information below.

Attachment of Amine Groups, DNA Nanostructures, Beads, Actin and Cells to the DNA Handles

[0095] For all attachment experiments, confocal imaging was carried out as described before.

[0096] Amine groups: A DNA oligo with an amine-group at the 3′ end and an Atto488-label at the 5′ end form compound #3 in Table 1, containing DNA having SEQ ID No. 3 was purchased from Biomers and corresponds to the second compound 34 mentioned above. It was mixed with complementary 3′-cholesterol-tagged DNA having SEQ ID No. 1 (corresponding to the first compound 16 mentioned above) at an equimolar concentration of 2 μM in the standard buffer (Buffer 1) and encapsulated in microfluidic droplets as described.

[0097] DNA nanostructures: To demonstrate the attachment of DNA nanostructures to the droplet periphery, we chose a hexagonal DNA lattice composed of 3 unique DNA sequences..sup.29 We modified one with a 3′ cholesterol-tag to give compound #4ii having DNA with SEQ ID No. 5 (Compound #4ii, Table 1, corresponds to first compound 16 mentioned above), a second one with a 5′ Atto488-tag having DNA with SEQ ID No. 4 (compound #4i, corresponds to second compound 34 mentioned above) and left a third one without modification having DNA with SEQ ID No. 6 (corresponds to a further second compound 34) (compound #4iii, #4i to #4iii all from Biomers) (see FIG. 5). The three strands were mixed in Buffer 1 at a concentration of 2 μM and heated to 60° C. for 10 minutes using a thermocycler (BioRad) before microfluidic encapsulation.

[0098] Polystyrene beads: Plain polychromatic polystyrene microspheres with a diameter of 2 μm were purchased from Fluoresbrite™, Polysciences, Inc. The beads were mixed with 4 μM of compound #7 and 2 μM of compound #8 and compound #9 (Table 1) at a concentration of approximately 10.sup.6 μl.sup.−1 in Buffer 1 before droplet production. Filter-free devices were used to prevent the beads from blocking the channel.

[0099] Actin: Actin (Cytoskeleton, Inc.) was stored in a buffer containing 2 mM TRIS/HCl, pH 8, 0.2 mM CaCl.sub.2, 0.2 mM ATP, 0.005% NaN.sub.3 and 0.2 mM DTT, at −80° C. The actin monomers were mixed with 10 mol % biotinylated-actin (cytoskeleton) and 1 mol % of Alexa488-labeled actin (LifeTechnologies). Before encapsulation, the monomers were polymerized into 5-10 μm sized filaments using an actin polymerization buffer (2.0 mM TRIS/HCl pH 8, 20 mM MgCl.sub.2, 0.2 mM CaCl.sub.2, 0.5 mM ATP, 0.005% NaN.sub.3 and 0.2 mM DTT) to give F-actin. After one hour of incubation (at room temperature), streptavidin (for biotin-binding) was added in 5-fold excess (10 μM) to the F-actin to enable the binding of biotinylated DNA (compound #6, Table 1). After incubation, the solution was centrifuged at 10.sup.6 g for 1 h and the F-actin pellet was afterwards resuspended in HEPES pH 8.2 buffer to a final actin concentration of 80 μM. The complementary biotinylated and the cholesterol-tagged DNA were supplied via a second aqueous inlet in the microfluidic device at a concentration of 8 μM, yielding a one-to-one ratio of streptavidin primed biotinylated F-actin and DNA strands. To prevent blocking, encapsulation experiments were performed using filter-free microfluidic devices.

[0100] Jurkat cells: Jurkat, Clone E6-1 (ATCC® TIB-152 TM) were cultured in RPMI-1640 (ThermoFisher Scientific) containing 10% fetal bovine serum and 1% penicillin-streptomycin in a humidified incubator at 37° C. and 5% CO.sub.2. The medium was changed every two days and cells were maintained at a cell density of 5×10.sup.6 cells/ml. Prior to the experiment, cells were harvested, centrifuged and resuspended in culture media at a final concentration of 10.sup.6 cells/ml. 4 μM cholesterol-tagged DNA was added to the cell suspension before encapsulation. Because of typical cell sizes of 10-15 μm a filter-free device with 80 μm wide (instead of 30 μm) channels was used, resulting in droplets with a diameter of 80 μm.

Thermal Cycling Experiments

[0101] DNA containing compounds #10 and #11 were encapsulated in microfluidic droplets containing with Buffer 1. ROX was chosen as a fluorophore due to its known temperature stability..sup.41 Droplets were sealed in an observation chamber and imaged using a Leica SP5 microscope with a 60× oil immersion objective. The temperature was increased with a temperature-controlled microscope stage (Tokai Hit ThermoPlate TP-110) from 20° C. to 60° C. and subsequently decreased in steps of 10° C., each lasting 3 minutes. This cycle was then repeated 2.5 times. The region of interest (ROI) was changed for every temperature step to avoid bleaching. Fluorescence intensities within 20 droplets were analyzed using ImageJ.

Supporting Information

Interfacial Tension Measurements

[0102] A drop shape analyzer DSA25 (Kruss GmbH, Germany) tensiometer with CCDcamera and the pendant drop method was used to measure the surface tension at the interface of surfactants and aqueous phase in presence and absence of cholesterol-tagged DNA. The Laplace-Young equation was selected as a fittingmethod..sup.45 Oil and aqueous phase densities were set to 1.6 gcm.sup.−3 and 0.99 gm.sup.−3, respectively. For the measurements 1 mL of the aqueous phase (10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8) was used in a plastic cuvette. To study to which extent cholesterol-tagged DNA alters the IFT, three solutions were investigated: one DNA-free, one with 2 μM 3′ cholesterol-tagged DNA (SEQ ID No. 1) and one with 2 μM cholesterol-tagged and complementary DNA (SEQ ID No. 3)—all in the standard buffer (Buffer 1). The oil phase contained either commercial surfactant (2 wt % of Perflouro-polyether-polyethylene glycol (PFPE-PEG) block-copolymer fluorosurfactants from Ran Biotechnologies, Inc.) or custom-synthesized surfactant (2.5 mM of a triblock-copolymer PFPE-PEG-PFPE triblock-copolymer surfactant.sup.1,2) diluted in HFE-7500 (DuPont). Prior to the measurement, samples were drawn into 1 mL syringes equipped with a cannula (0.8 mm×22 mm blunt/dull) and lowered into the aqueous phase. Drops were created using the software controlled dosing unit. Droplet interfacial tension (IFT [mN m.sup.−1]) was analyzed over time. Once the IFT did not show any change, the measurement was stopped and the IFT was collected. For each surfactant/buffer experiment, three measurements were performed and the mean value as well as the standard deviation were calculated. All results are presented in Table 2.

[0103] Table 2: Interfacial tension between aqueous and oil/surfactant phase for DNA-free aqueous solution (10 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl.sub.2, pH 8), or solutions where either 2 μM cholesterol-tagged DNA or 2 μM cholesterol-tagged DNA and 2 μM of its complementary strand have been added. The measurements have been carried out for both surfactants used in this study. The results indicate no dramatic change of the IFT between aqueous and oil phase. Nonetheless, a trend to lower tension can be deduced in the presence of cholesterol-tagged DNA.

TABLE-US-00002 Cholesterol- Cholesterol-tagged and DNA-free tagged DNA complementary DNA Surfactant [mN m.sup.−1] [mN m.sup.−1] [mN m.sup.−1] Commercial 3.89 ± 0.07 3.78 ± 0.12 3.21 ± 0.10 Synthesized 4.72 ± 0.06 3.16 ± 0.27 2.09 ± 0.53

FRAP Analysis

[0104] For the FRAP analysis, normalized intensity values were calculated as follows:

[00001]$\begin{array}{cc}{I}_{\mathrm{normalized}}=\frac{I_{\mathrm{measured}}}{I_{\mathrm{pre}}},& \left(1\right)\end{array}$

[0105] I.sub.pre was calculated by taking the average of the five measured intensity values before bleaching. Due to the low number of acquired images, a correction for the bleaching of the dye during this pre-bleaching phase has not been taken into account. A non-linear least-square fit was applied to the normalized intensities from the recovery phase. The fit-function was of the form:

f(t)=A(1−exp(−λl))+x.sub.0,  (2)

where A and λ, are fit parameters and x.sub.0 is the time point after bleaching i.e. the start of the recovery phase. With this, following the protocol of Axelrod.sup.5 and Soumpasis,.sup.6 the diffusion coefficient can be calculated via:

[00002]$\begin{array}{cc}D=0.32\frac{r^2}{{\tau }_{1/2}}=-0.32\frac{\lambda r^2}{0.5},& \left(3\right)\end{array}$

where τ is the half-recovery time and r the radius of the circular bleaching spot (2.5 μm). In order to efficiently evaluate the measured data, a custom-written MATLAB script performing the above mentioned operations was used.

Correlation Between Diffusion Coefficient and Surfactant Concentration

[0106] In order to study the behavior of cholesterol-tagged DNA within the surfactant layer, we performed FRAP measurements for a variety of surfactant concentrations. The experiments were carried out with a commercial surfactant (008-FluoroSurfactant, RAN Biotechnologies). The results can be seen in FIG. 9, which shows the diffusion coefficient as a function of surfactant concentration. The droplets were around 30 μm in diameter.

[0107] The diffusion coefficient decreases with an increasing surfactant concentration. A nonlinear-least square fit has been applied using the formula:

D(c)=A exp(−λc),  (4)

where A and λ are fitting coefficients and c the surfactant concentration. A possible interpretation of this result is that the packing of the surfactants at the droplet periphery changes with the surfactant concentration. An increasing concentration results in a denser packing at the droplet periphery and therefore a decreased diffusion coefficient of the cholesterol-tagged DNA. It also makes sense that the packing of surfactants at the periphery is limited. As the data indicate, the diffusion coefficient remains constant above a surfactant concentration of 2 wt %. This could be due to the saturation of densely packed surfactant at the droplet periphery. It is possible that the insertion of cholesterol-tagged DNA into the surfactant layer provides an experimental handle to determine the diffusive properties of surfactants.

REFERENCES

[0108] References to the following non-patent literature have been made above, indicated by corresponding superscripted numbers: [0109] (1) Platzman, I.; Janiesch, J.-W.; Spatz, J. P. Journal of the American Chemical Society 2013, 135, 3339-3342. [0110] (2) Janiesch, J.-W.; Weiss, M.; Kannenberg, G.; Hannabuss, J.; Surrey, T.; Platzman, I.; Spatz, J. P. Analytical Chemistry 2015, 87, 2063-2067. [0111] (3) Kurokawa, C.; Fujiwara, K.; Morita, M.; Kawamata, I.; Kawagishi, Y.; Sakai, A.; Murayama, Y.; ichiro M. Nomura, S.; Murata, S.; Takinoue, M.; Yanagisawa, M. Proceedings of the National Academy of Sciences 2017, 114, 7228-7233. [0112] (4) Weiss, M. et al. Nature Materials 2017, 17, 89-96. [0113] (5) Axelrod, D.; Koppel, D.; Schlessinger, J.; Elson, E.; Webb, W. Biophysical Journal 1976, 16, 1055-1069. [0114] (6) Soumpasis, D. Biophysical Journal 1983, 41, 95-97.

REFERENCE SKINS

[0115] 10 microfluidic droplet [0116] 12 center [0117] 14 layer of surfactant [0118] 16 first compound [0119] 18 hydrophobic part [0120] 20 molecular recognition site [0121] 22 first fluid [0122] 24 second fluid [0123] 28 block-copolymer [0124] 30 hydrophilic block [0125] 32 hydrophobic blocks [0126] 34 second compound [0127] 36 molecular recognition site [0128] 38 functional moiety [0129] 40 interface [0130] 42 component A [0131] 44 component B [0132] 46 dispersion