The invention relates to a method for mixing a first stream of droplets with at least one second stream of droplets to obtain a third stream of fused droplets, to a device for generating and mixing a first stream of droplets with at least one second stream of droplets to obtain a third stream of fused droplets, to a droplet, a plurality of droplets or to a two-dimensional or three-dimensional scaffold generated by this method, according to the independent claims.

In recent years, micro-scale mixing has become increasingly im ^¬ portant in various fields ranging from biomedical diagnostics and drug development to food and chemical industries. An attrac- tive application, in which the very precise and fast picoliter- range mixing of liquids is required, is the microencapsulation of cells in an ultra-high-throughput fashion. This technique al ^¬ lows the generation of combinatorial cellular microenvironments for the screening and discovery of two dimensional and three di- mensional cell culture substrates.

Notably, the application of such substrates has a high potential in stem cell research. Stem cells are drawing increasing attraction of the scientific community for having the potential to ex- pand and generate many types of tissue specific cells. However, such potential is strongly related to the microenvironment in which they reside in-vivo, the so-called niche. To recapitulate such a complex network of cues, stem cell biologists need great ^¬ er control over cell culture conditions. Furthermore, stem cells usually show high heterogeneity when kept in culture. In order to systematically study such populations, single cell manipula ^¬ tion becomes a crucial requirement. On the other hand, micro-scale mixing is an important aspect to be addressed in the fabrication of three dimensional artificial tissues. In this context, ink jet printers have been increasing ^¬ ly exploited to handle micro-volumes of biologically relevant solutions. Importantly, high cell compatibility of such tools has been achieved. Cells have been printed without negatively affecting their viability within a multitude of biomaterials . A typical droplet size of less than 100 ym is appropriate to con ^¬ tain a single mammalian cell (smaller than 25 ym) . This parame- ter can be controlled through modification of the dispensing parameters .

De Gans and Schubert have provided a comprehensive review on commercially available instrumentation for ink jet printing of polymer micro-arrays (Macromol. Rapid Commun. 2003, 24, 659- 666) . They explain that even with nowadays commercially availa ^¬ ble ink jet equipment, a number of problems remains unsolved, most notably the problem of how to conduct the mixing of compo ^¬ nents when creating micro-arrays of polymer blends. There are three main choices to address this issue: mixing before print ^¬ ing, mixing during printing or mixing on the substrate. In order to effect mixing during printing, the collision of two different droplets in air is described. However, according to de Gans and Schubert, this approach is far too complicated to be applicable in a routine practice.

The patent application US 2003/0175410 Al describes methods and apparatuses for selectively depositing so-called bio ink solu ^¬ tions in order to build up a three-dimensional biomimetic scaf- fold. In particular, a method is provided for preparing such scaffolds by co-depositing two or more of bio ink solutions. These structures are prepared using a solid free form fabrica ^¬ tion system (i.e. a 3D printer) comprising an apparatus employ- ing one or more focused micro dispensing devices, which permits the co-depositing of bio inks in controllable manner. During operation, a first micro dispensing device and a second micro dis ^¬ pensing device may selectively dispense a focused volume of a first bio ink and of a second bio ink at a plurality of dispens ^¬ ing locations on a surface. Hence, the bio inks are mixed up on deposition on the surface. This technology can be used to cause solidification or gelation of the bio ink in order to produce the desired three-dimensional structure of a biomimetic scaf- fold.

However, the described system has the drawback that the mixing of the bio inks occurs in a rather uncontrolled manner. Conse ^¬ quently, the solidification or gelation of the bio inks, for in- stance the formation of a hydrogel by chemical or enzymatic re ^¬ actions, proceeds relatively slow. This attenuates the resolu ^¬ tion of such a device. Furthermore, micro encapsulation of liv ^¬ ing cells cannot be achieved and patterning of the material is impossible .

It is a problem underlying the present invention to overcome these drawbacks in the state of the art.

In particular, it is a problem underlying the present invention to provide methods and means to effect reliable and precise mi ^¬ cro-scale mixing of multiple components, even in routine appli ^¬ cations. The mixing should be compatible with physiological sys ^¬ tems, in particular with living cells. It should be cost- efficient and safe. Furthermore, a high degree of automation is desirable and the mixing technology should allow for high- throughput applications. It should also allow for the mixing of a wide variety of different liquids in nanoliter quantities. Moreover, it is a problem underlying the present invention to provide corresponding micro-scale droplets with improved proper ^¬ ties and enhanced two-dimensional or three-dimensional scaf ^¬ folds . These problems are solved by a method according to claim 1, a device according to claim 16 and products according to claims 19 and 20.

The present invention refers to a method for mixing a first stream of droplets with at least one second stream of droplets to obtain a third stream of fused droplets by collision and fu ^¬ sion of at least one droplet of the first stream with at least one droplet of the second stream. The method is characterized in that the first stream of droplets and the least one second stream of droplets are generated in a synchronized manner.

By directly colliding and fusing at least two droplets, effi ^¬ cient micro-scale mixing is achieved. To this end, the synchro ^¬ nized generation of the droplets is of critical importance in order to achieve controlled collision and fusion. In the present context, the term "synchronized generation" refers to a genera ^¬ tion of two or more streams of droplets in a manner that sub ^¬ stantially all droplets of the first stream undergo collision and fusion with a droplet of a second stream, and vice versa. The term "substantially all" refers to a ratio of at least 85%, preferably 90%, even more preferably 95% or 99%. As the droplets are generated and mixed streamwise, the method is suitable for high-throughput applications. At least one of the streams of droplets can be generated by a piezoelectric or thermal ink jet dispenser. The ink jet technol ^¬ ogy is a well-established method for the generation of droplets in a picoliter volume range and allows for the formation of droplets in a very high frequency and a narrow size distribu ^¬ tion.

The collision and the fusion of at least one droplet of the first stream with at least one droplet from a second stream can occur in a flying phase. In a flying phase collision and fusion, the incoming flying trajectories of at least two droplets com ^¬ bine to a final trajectory that is a combination of the original ones. This in-flight-mixing allows the union of two droplets without any interference of a support-vessel or - surface.

Upon collision and fusion, a reaction, in particular a cross- linking-reaction, can occur between a first component contained in the first stream of droplets and a second component contained in a second stream of droplets. Accordingly, the mixing- technique provides an attractive approach for conducting various chemical or biological reactions on a picoliter-scale .

A crosslinking-reaction can occur to form a stream of hydrogel droplets. This allows mixing two precursor-solutions of a hydro ^¬ gel in order to form a hydrogel droplet under well-defined con ^¬ ditions (e.g. volumes and mixing-time) . In this context, a hy ^¬ drogel is a highly swollen network of cross-linked polymer chains with the swelling agent water. Hydrogels can contain over 90% of water. They possess a degree of flexibility which is very similar to natural tissue. This allows very broad application of hydrogels as scaffolds in tissue engineering.

The hydrogel droplets can be collected in a liquid bath or de ^¬ posited on a substrate. This underlines the high flexibility of this method. The hydrogel droplets can be deposited on a substrate to gener ^¬ ate a two-dimensional or three-dimensional scaffold. Such a scaffold can have tissue-like characteristics. The method is therefore amenable for tissue engineering.

The first component can contain a naturally derived macromole- cule, preferably an alginate, and a second component can be a solution of a crosslinking ion, preferably a metal ion, even more preferably a calcium ion. The crosslinking-reaction of alginates with calcium ions proceeds with very short reaction times. The formed hydrogels exhibit high biocompatibility and allow for various applications, in particular in tissue engineering .

On the other hand, the first component can contain a synthetic macromolecule, in particular a PEG-based polymer, preferably modified with oligopeptides, and a second component can be a so ^¬ lution of a crosslinking agent, preferably an enzyme, even more preferably a transglutaminase.

Such a PEG-hydrogel provides an attractive alternative to an al- ginate-based hydrogel. Even though the reaction times for cross- linking are usually longer, such hydrogels show an increased stability. Furthermore, these hydrogels can be readily tethered with biologically active signals.

The first component can also contain a reactive synthetic macro- molecule, preferably a PEG-based polymer, comprising maleimidie or vinylsulfonate functional groups, in particular at terminal positions. In such an application, a second component can con ^¬ tain a counter-reactive hydrogel precursor, preferably a syn ^¬ thetic macromolecule, in particular a PEG-based polymer, com ^¬ prising thiol groups, in particular at terminal positions. How- ever, a second component can also contain a cysteine containing peptide instead, in particular an oligopeptide or a polypeptide. These hydrogel systems have beneficial properties. Their cross- linking reactions proceed with sufficiently short reaction times for the formation of three-dimensional structures, in particular in a layer-by-layer fashion. Furthermore, they form peptidic substrates for proteases, in particular for metalloproteases .

Importantly, the crosslinking kinetics of such systems can be increased by blending PEG-based hydrogels with alginate, which, if desired, can be readily removed after PEG crosslinking by ly ^¬ ase treatment. This "hybrid hydrogel" strategy is advantageous because both hydrogel systems share calcium (Ca2+) as the common entity enabling crosslinking. That is, the alginate component can be rapidly cross-linked to form a framework within which the PEG component may cross- link via a reaction which proceeds more slowly. The alginate framework ensures that the PEG hydrogel re ^¬ tains a three-dimensional structure while the slower PEG cross- linking reaction proceeds. The alginate may then be removed to leave only the PEG hydrogel.

It should be noted that the above-mentioned hybrid hydrogel net ^¬ work concept can be implemented with other slowly crosslinkable hydrogel systems. Accordingly, the second component can further comprise a macromolecule, preferably a naturally-derived macro- molecule, in particular fibrinogen or hyaluronic acid. As such, a wide range of hydrogel systems can be tailored to be useful for in situ droplet mixing. At least one of the first stream of droplets and the second stream of droplets can contain a, preferably mammalian, living cell that is encapsulated into the hydrogel droplet. The in-vivo encapsulation of cells is particularly attractive in stem cell research to provide a well-defined microenvironment . Further ^¬ more, the encapsulated cells can also be deposited in a two- dimensional or three-dimensional scaffold to generate a tissue like artificial living structure.

On the other hand, the first stream of droplets can contain at least one, preferably mammalian, living cell of a first type and a second stream of droplets can contain at least one, preferably mammalian, living cell, in particular of a second type, the cells are then co-encapsulated into a hydrogel droplet. The co- encapsulation of living cells into the same hydrogel droplet is an attractive approach for studying cell-cell interactions (whether of cells of the same type, or cells of different types) .

Another aspect of the present invention refers to a device for generating and mixing a first stream of droplets with at least one second stream of droplets to obtain a third stream of fused droplets. The mixing preferably proceeds by an above described method. The device comprises

- a first dispenser for generating the first stream of droplets,

- at least one second dispenser for generating at least one second stream of droplets,

- holding means for holding the first and at least one second dispenser .

The first and/or at least one second dispenser preferably is/are a piezoelectric or thermal ink jet dispenser (s) .

The device is characterized in that it further comprises syn ^¬ chronizing means for synchronizing the operation of the first and at least one second dispenser to effect collision and fusion of at least one droplet of the first stream with at least one droplet of the second stream to obtain the third stream. Such a configuration allows for the efficient and reliable mix ^¬ ing of droplets. In particular, collision and fusion of substantially all droplets of the first stream with a droplet of a sec ^¬ ond stream, and vice versa, can be achieved. Synchronization of the operation of the first and at least one second dispenser can be achieved by a reiterative process com ^¬ prising the steps of:

- mechanically adjusting the nozzle positions inside the hous- ings

- establishing continuous flows from the nozzles

- visually assessing crossing of the flows

- repeating the procedure until proper crossing is achieved

- once proper crossing is achieved, setting the ejection speed at each dispensing unit to the same value, by acting on dispensing parameters, such as voltage and pulse length.

The holding means can be suitable to adjust the relative posi ^¬ tion and orientation of the first dispenser and the at least one second dispenser. This way, proper mixing can be attained by tuning the geometry of droplet collision. The angle of impact can be adapted to the energy of the incoming droplets, in order to promote fusion and to avoid re-separation of the droplets due to their inertial energy. The angle between the alignment axis of the first dispenser and the alignment axis of the second dis ^¬ penser can be adjustable in a range from 30° to 90°. Moreover, the possibility to tune the droplet-generation by acting on voltage, pulse-length and frequency allows properly synchroniz- ing the incoming jets of droplets and modifying droplet charac ^¬ teristics (volume and speed) in order to obtain a precise mixing ratio upon the impact to finally control the mixing parameters. The invention also relates to a droplet or a plurality of drop ^¬ lets generated by an above-described method. Such droplets have superior characteristics, as they exhibit a mixing-homogeneity that cannot be achieved by conventional methods. Furthermore, the invention also relates to a two-dimensional or three-dimensional scaffold generated by an above-described meth ^¬ od. Such scaffolds are valuable structures for tissue- engineering .

Further advantages and features of the invention are apparent from the following description of several embodiments and in the figures .

It is shown

Figure 1 : Schematic representation of the in-flight- mixing technique with several application examples;

Figure 2 : Formation of hydrogel droplets by a cross- linking reaction and collection of the droplets in a liquid bath;

Figures 3 to 5 : Series of microscopic images representing a droplet fusion and mixing process according to the present invention. Figure 1 provides a schematic overview of an in-flight-mixing technique for two streams of droplets according to the present invention. The first stream of droplets 1 is generated by the first dispensing unit 5 and the second stream of droplets (2) is generated by the second dispensing unit 6. The streams of drop ^¬ lets are collided to form a fused droplet 4 in which a cross- linking reaction occurs to form a stream of hydrogel droplets 3. The hydrogel droplets 3 can be used for several applications, such as 3D-patterning a, high throughput screening b or combina- torial studies c. In high throughput screening, it is possible by this technique to create a multitude of replicates of the same microenvironment and screen at series of different extrin ^¬ sic factors. Combinatorial studies represent an upgrade of the high throughput screening application, in which several variants of compositions are generated and screened according to specific requirements. Finally, 3D-patterning can be achieved by control ^¬ ling the deposition of the in-situ formed bio ink according to precise blueprints. 3D-patterning can also be combined with com ^¬ binatorial microenvironment formulation in order to create 3D- scaffolds with heterogeneous composition and function.

Figure 2 shows a further embodiment of the present invention in which a living cell 9 is encapsulated into a hydrogel droplet 3. To this end, a first stream of droplets 1 containing single liv- ing cells is collided with a second stream of droplets 2 to gen ^¬ erate a fused droplet 4 in which a crosslinking reaction of the hydrogel precursors occurs to form a stream of hydrogel droplets 3. These hydrogel droplets are collected in a liquid bath 7. The enlarged representation of a hydrogel droplet 3' shows that it contains several niches 8. One of those niches is occupied with the living cell 9. Figures 3 to 5 show a series of microscopic photographs repre ^¬ senting the droplet collision and fusion process according to the present invention. In figure 3, the impact of a first drop ^¬ let of the first stream 1 with a second droplet of a second stream 2 is shown. Figure 4 shows the first phase of the droplet fusion process. It can be seen that at this stage, the fused droplet 4 does not have a spherical shape yet. Figure 5 shows the second phase of the droplet fusion process in which the fused droplet 4' already has a spherical shape. The mixing of the droplets immediately commences upon the impact (Fig 3) and finishes at the late stage of phase two (Fig. 5) .