Method for providing genetically-modified cells

Abstract

This invention concerns a method for providing genetically modified cells having a defined phenotype comprising providing a microorganism or a cell of a multicellular organism, performing random mutagenesis on said microorganism or said cell such as to obtain a population of genetically modified cells, optionally, incubating the population of genetically modified cells in the absence of light between 300 nm and 500 nm for a period of resting time. In a subsequent step, culturing the cells in growth medium for a period of recovery time, encapsulating of at least a portion of the genetically modified cells in a plurality of droplets, detecting the presence of a phenotypic marker which is indicative of a defined phenotype of the genetically modified cell in the droplets, and selecting the droplets comprising the detected phenotypic marker, or comprising a quantity of the detected phenotypic marker which exceeds a predefined threshold. The invention also concerns the use of the so fabricated droplets containing the genetically-modified cells with the desired phenotype for the production of macroscopic engineered living materials or other industrial applications.

Description

Method for providing genetically-modified cellsTechnical domain

[0001] The present invention concerns a method for providing genetically modified cells of a desired phenotype. The invention also concerns the fabrication of an engineered living material using such genetically modified cells.Related art

[0002] Engineered living materials (ELMs), which are composed either entirely or partly of living cells, represent an enticing new class of materials with structure and properties governed by metabolically active biological cells. Such cells can be temporarily used to generate intricate structures not accessible via conventional manufacturing and are often kept alive in the host structure to create a material with living, adaptive functionalities.

[0003] Different from biological ELMs, which are entirely composed of living cells, find their application in production of organoids, hybrid ELMs, also called HLMs, combine living cells with structural material to produce composites with functional properties. HLMs are usually built in a top-down process. The cells incorporated in the HLM provide these materials with the ability to grow, self-heal, remodel and react to environmental stimuli.

[0004] The functionalities of the ELMs are encoded in the DNA of the biological cells contained in the material. ELMs are usually grown in water under ambient temperature and pressure and offer therefore an energetically efficient alternative to synthetic counterparts.

[0005] To provide the desired functionalities to ELMs, the cells comprised therein must display the desired phenotypic trait(s) to match the engineering needs. Living cells used in HLMs include for example self-regenerating mycelia and bio-cementing bacteria.

[0006] In order to tailor the ELM to achieve desired functionalities, specific traits may be introduced into the living cells contained in the ELM. The cells may be for example re-programmed using known synthetic biology tools and methods. This compelling strategy has been employed to create genetically engineered living materials for catalysis, energy conversion, sensing, electronic and biomedical applications.

[0007] Despite the demonstrated success of these strategies, a mismatch often exists between the biological function of wild-type species and the desired engineering functionalities. At the level of the whole microorganism, the envisioned functionalities might display trade-offs and also require the concerted action of multiple genes. In other instances, the genes encoding for a specific function are also unknown, thus preventing a clear connection between genotype and phenotype. Even if a functionencoding gene is known, the fitness of the whole microorganism to an engineering setting will likely involve more than one specific trait.

[0008] These challenges call for the development of other strategies for the selection of genetically programmable microorganisms for engineered living materials.

[0009] Directed evolution has been widely applied to improve the selectivity and activity of enzymes by exploring the huge design space available in the genome of microorganisms. Following the principles of natural selection, this approach accelerates the iterative process of diversification and selection of genes that encode for a specific desired phenotype. Gene diversification typically occurs by inducing targeted or random mutations in the genome of the microorganism that produces the enzyme of interest. This results in a library of distinct mutants that are subsequently screened based on the performance of the enzyme.

[0010] However, since the probability of finding a mutant with improved performance is very low, directed evolution of enzymes is often performed with an initial library containing between 104to 107mutants. The rare high-performance mutants may be selected from this vast pool using microfluidic approaches, which can reach very high throughputs if performed inside droplets.

[0011] Such a microfluidic selection method for cells having the desired phenotype from mixed cell populations have been disclosed previously, for example inUS20090068170 describing a microfluidic screening platform for screening and sorting fluidic droplets, which may comprise immortal or non-immortal cells, such as hybridoma cells or B-cells.

[0012] Even though this high-throughput selection approach has accelerated screening of large populations of cells, the problem of low yield of genetically-engineered cells with a desired phenotype following a random mutagenesis approach still persists. This low yield reduces the efficiency of generating the desired mutant strains. This in turn renders the development of new ELMs cumbersome and inefficient.

[0013] The efficacy of providing new ELM materials which exhibit desired engineering functionalities remains therefore still a significant challenge today.Short disclosure of the invention

[0014] An aim of the present invention is to improve the efficiency of providing genetically engineered single cells, in particular genetically engineered microorganisms such as bacteria, which have a desired phenotype.

[0015] It is another aim of this invention that the yield of genetically engineered single cells having a desired phenotype increased.

[0016] It is yet another aim to find an alternative method for providing genetically modified cells having a desired phenotype.

[0017] According to the invention, one or more of these aims are attained by the object of the attached claims, and especially by a method for providing genetically modified cells having a defined phenotype comprising the steps of providing a microorganism or a cell of an organism, performing random mutagenesis on said microorganism or said cell. Following the random mutagenesis, the population of cells is cultured for a period of recovery time in a recovery step prior to encapsulation.

[0018] Using a random mutagenesis approach to generate the genetically modified cells provides the advantage there is no need to have knowledge of all the genes or genetic modifications contributing to a desired phenotype. Instead, this method can beused to establish such a functional link between a genotype and a specific phenotype. It can also be used to identify further genetic markers, which may contribute to or enhance a particular phenotype.

[0019] During random mutagenesis the microorganism or cell is exposed to mutagens, for example to UV light or to mutagenic chemicals for a period of mutagenesis time.

[0020] It is a particular challenge in such whole-genome mutagenesis approaches that a large number of mutations are introduced at the same time, which potentially impairs the ability of the microorganism to efficiently grow.

[0021] By introducing the recovery step, which naturally pre-selects the mutated cells based on their growth capability in a given environment, the yield of mutated cells being metabolically active is significantly increased. The percentage of cells in the cell population, which present the desired phenotype is as a consequence also higher.

[0022] The liquid medium used in the recovery step can be configured to reproduce the environment in which the microorganism will later be used in the applications. For example, this liquid medium can contain carbon sources from agricultural waste, can contain ions and molecules extracted from the soils subjected to biomining processes, or can have the pH, temperature and ionic concentration tuned to reflect the environment of specific microbiomes.

[0023] In one embodiment culture conditions in recovery step are chosen to apply an adverse selective pressure on the cultured cells. The selective pressure conditions may advantageously be chosen to enhance the chance of surviving and reproducing of one or more phenotypes of the genetically modified cells, which are able to withstand the selective pressure. As a result, the cells having such a resilient phenotype recover better than and / or outgrow cells having alternative phenotypes. In this embodiment the recovery step provides a selection of genetically modified cells which are resilient to the applied selective pressure conditions. The recovery step of this embodiment therefore also serves as a preselection step to selectively amplify cells having a resilient or a desired phenotype. This further increases the efficiency of providing genetically modified cells of a defined phenotype.

[0024] The selective pressure conditions may be defined according to a desired phenotype of the genetically modified cells.

[0025] Selective pressure may for example be applied by culturing the genetically modified cells in the presence of a high concentration of metal ions, for example 10'000 to 100'000 ppb (parts per billion) of essential metal ions in the culture liquid. Essential metal ions are cations of the alkali metals, Mg, Ca, Sr, Fe, Co, Cu and Zn.

[0026] Selective pressure may for example be applied by culturing the genetically modified cells in the presence of a high concentration of bacteriotoxic metal ions in the culture liquid, for example 100 to 10'000 ppb of bacteriotoxic metal ions in the culture liquid. Bacteriotoxic metal ions are ions of metals Be, Ba, V, Cr, Ni, Zr, Nb, Mo, Tc, Ru, Rh, Pd, Ag, Cd, Hf, Ta, W, Re, Is, Ir, Pt, Au, Hg, Al, Ga, In, Tl, Ge, Sn, Pb, As, Sb, Bi, Se, Te, Po, or polyatomic ions thereof.

[0027] Selective pressure may for example be applied by culturing the genetically modified cells in the presence of a high concentration of rare earth element ions or actinide ions or polyatomic ions thereof, for example 1'400 to 32'000, or 14 to 800 ppb of rare earth element ions or actinide ions or polyatomic ions thereof in the culture liquid.

[0028] Selective pressure may also be applied by providing a culture medium lacking a certain type of nutrient, or by limiting the amount of a certain type of nutrient to a suboptimal concentration in the medium.

[0029] In addition or alternatively, an unusual nutrient may be provided, or the concentration of a certain nutrient may significantly exceed its optimal concentration for a cell culture medium.

[0030] Selective pressure may be applied by providing high concentration of salt in the culture medium, for example a salt concentration of 10% or more, of 2% or more, or of 0.9% or more. Cells may for example be incubated in sea water, which naturally contains a high concentration of salt compared to fresh water, or in a liquid medium which is prepared with sea water instead of fresh water.

[0031] Selective pressure may be applied by letting the genetically modified cells recover at a specific temperature, at which the desired phenotype is expected to grow.

[0032] Selective pressure may be defined on the basis of the environmental conditions the genetically modified strain is presumed to encounter in its intended application. For example, the cells may be incubated in a liquid solution obtained by washing mining residues, soil samples, or any solid phase expected to be present in the environment in which the genetically modified cells are intended to be applied. The liquid solution may be directly obtained from the environment, for example from contaminated water reservoirs, or from mining ponds and heaps.

[0033] Generally, selective pressure may be applied using any other adverse conditions, which is expected to negatively impact on the metabolism and / or growth of cells which do not have the desired phenotype. One or more adverse conditions may be combined to define multiple selection criteria for a more targeted selection.

[0034] As an optional step, which should be performed if a UV mutagenesis approach is performed, the population of genetically modified cells is incubated in the dark for a period of resting time prior to the recovery step.

[0035] In one embodiment two or more cycles of the mutagenesis, the optional resting step and the recovery step are performed. Repeating this step further increases the yield of viable mutated microorganisms and cells.

[0036] In an UV mutagenesis approach the cells may be exposed to UV doses ranging from 0.5 mJ / cm2to 100 mJ / cm2. Other UV doses may be chosen, according to the cell type and the experimental conditions. The skilled person may follow conventional protocols for UV-induced mutagenesis, and, if required, adapt the protocols according to the provided species or cell.

[0037] Crucially, following the UV-induced mutagenesis, the cells are allowed to recover in a two-step resting / recovery process. First, the cells are incubated in the absence of light having a wavelength from 300 nm to 500 nm for a period of resting time. The cells may for example be incubated in the dark for a period of resting time. The resting time should be at least 30 minutes. The resting time may range from 30minutes to 3 h, the cells may for example rest for lh. Incubation in the absence of light of a wavelength of 300 nm to 500 nm prevents the photoreactivation repair mechanism to become active. DNA damage introduced during the mutagenesis will therefore be maintained in the cells and not repaired. Incubating the cells in the dark for the resting period therefore results in a higher percentage of mutated cells. Subsequently to the resting step, the cells are cultured in a rich culture medium for a period of recovery time to allow the viable mutated cells to amplify.

[0038] The term "cultured" as used herein means incubating the cells at appropriate culture conditions, including appropriate temperature and pressure. The appropriate conditions are specific to the type of cells which were provided for mutagenesis and are known to a skilled person. As a general guideline, microorganisms derived from soil or water may for example be incubated at temperatures of less than 30°C, for example ranging from 20°C to 30°C, at ambient pressure. Cells which were isolated from warmblooded bodies, or pathogenic microorganisms may be for example be incubated at temperatures from 30°C to 37°C, at ambient pressure.

[0039] In one embodiment the microorganism is a bacterial cell. The microorganism may however be any other type of microorganism, for example a yeast, a fungi or a microalgae.

[0040] It was observed as part of this invention, that the yield of genetically modified cells increased significantly when the microorganisms or cells were cultured to recover.

[0041] The recovery time may be 0.5 h to 96 h. The suitable recovery time may vary depending on the microorganism or the type of cell.

[0042] In one embodiment the recovery time corresponds to the period or time for culturing 1 to 5 generations of the cells. The times to culture a generation, also called the generation time, is specific for a microorganism or a cell type. It corresponds to the amount of time it takes for a cell population to double in number. It may be determined for a species of a cell type, or it may be based on generally known, or on estimated times.

[0043] For many common bacteria, the generation time is ranges between 20-60 minutes under optimal culture conditions. However, some bacteria may have significantly longer generation times.

[0044] Compared to UV-mutagenesis performed on microorganisms or cells from multicellular organisms performed without the additional resting and recovery steps, the population of genetically-modified cells obtained by this method comprises a significant increase in mutants with the desired phenotype.

[0045] The method further comprises the steps of encapsulating of at least a portion of the genetically modified cells in a plurality of droplets or capsules, detecting the presence of a phenotypic marker which is indicative of a defined phenotype of the genetically modified cell in said droplets or capsules, and selecting the droplets or capsules comprising the detected phenotypic marker, or comprising a quantity of the detected phenotypic marker which exceeds or falls below a defined threshold. Alternatively, if the desired phenotype is defined by the absence of a phenotypic marker, droplets which do not comprise detectable amounts of said marker may be selected.

[0046] The droplet may be a water in oil droplet, a hydrogel in oil droplet, which may be transferred to water, or double emulsion water in oil in water droplet. The term "droplet" as used hereinunder also refers to a capsule.

[0047] The phenotypic marker may for example be a marker molecule.

[0048] It is however also possible, that the phenotypic marker is another trait of the strain, for example an ability to withstand certain environmental conditions, or an altered growth behaviour.

[0049] The term "marker molecule" as used herein is not particularly limited to a specific type of molecule but is used in its broadest meaning. The marker molecule may be a biological or a biochemical analyte. The marker molecule may be an inorganic analyte. The marker molecule may be an organic molecule. The marker molecule may be a sugar. The marker molecule may be a peptide of a protein. Themarker molecule may be a monomer. The marker molecule may be a macromolecule, it may for example be a polymer, for example cellulose.

[0050] The marker molecule may be a product secreted by the cell.

[0051] The marker molecule may be soluble in aqueous solution. The marker molecule may be insoluble in aqueous solution. The marker molecule may be a water-immiscible compound. The marker molecule may also be particle or a compound, which is degraded by the microorganism or cell having the desired phenotype. In this case the decrease or the absence of the marker molecule is the criterion for selecting a cell, respectively a droplet.

[0052] In one embodiment the method comprises the additional step of culturing the genetically modified cells within the droplets prior to the detection step for a period of expression time. This expression time serves to permit the genetically-modified cells to be metabolically active and to synthesize and / or secrete the marker molecule which is detected in a subsequent step.

[0053] The expression time may for example range from 30 minutes to 1 week , or from 12 h to 24 h. Depending on the microorganism or the type of cell, the expression time may be shorter or longer.

[0054] In some embodiments the expression time can take up to a week, for example if microorganisms need to amplify in the droplet in order for the phenotypic marker to be measurable.

[0055] In some embodiments of this invention, assays to detect a phenotypic marker can be performed directly after the encapsulation, without requiring incubation of the cells to express mutant proteins.

[0056] In one embodiment the genetically modified cells are encapsulated in the plurality of droplets at an average ratio of no more than one cell per droplet. The majority of the droplets may contain either one genetically modified cell or no cell.

[0057] In one embodiment the encapsulation step comprises contacting a genetically modified cell of the population of cells with an aqueous phase comprising nutrients for the cell and necessary reagents for the assay under conditions sufficient to form a droplet comprising the cell. In another embodiment, cells are contacted with an aqueous hydrogel containing the nutrients for the cell and necessary reagents for the assay. These conditions may include contacting the aqueous suspension containing the modified cell with a fluorocarbon oil comprising a fluorosurfactant.

[0058] In one embodiment the detection of a marker molecule comprises contacting the marker molecule with a detection agent. The detection agent may comprise a fluorescent dye, it may for example be an agent with specific binding affinity for the marker molecule, said agent being coupled to a fluorescent dye. The specifically binding agent may be a ligand or an antibody adapted for binding the marker molecule. The detection agent may also be a fluorescent dye which binds or interjects with the marker molecule.

[0059] Insofar as the detection agent is or comprise a fluorescent dye an optical detection method is suitable for detecting the presence and / or the quantity of marker molecule in the droplet. The optical method includes exposing one or more droplets to light at the excitation wavelength of the fluorescent dye and to optically detecting light emitted by the fluorescent dye present in the droplet. The amount of fluorescent dye in the droplet being defined by the quantity of marker molecule to which the dye is directly or indirectly coupled.

[0060] The method of this invention may further comprise a step of further comprising the step of separating the selected droplets from remainder of the plurality of droplets.

[0061] In one embodiment this separation is performed dielectrophoretically using a microfluidic system. In this approach, electrical pulses may be triggered based on optical detection of the presence of a phenotypic marker, or based on optical signal surpassing a predefined threshold.

[0062] This approach to separation allows for high throughput screening with subsequent sorting of the genetically modified cells displaying the desired phenotype.Combining the optical detection method with the microfluidic separation method is particularly time efficient, as it allows for the screening and sorting to be automated.

[0063] The separated droplets may be used to fabricate engineered living material. The droplets comprising the genetically-modified cells of the desired phenotype may for example be provided together with structural material to produce macroscopic ELM, for example in a casting process, in a molding process, or in a 3D printing process.

[0064] Compared to previously known methods this invention is therefore adapted to provide genetically-modified cells with a desired phenotype at an increased yield. The method of this invention is therefore suited for providing living cells for ELMs and other industrial application.

[0065] Genetically-modified cells produced according to this method may find their application in any type of biotechnological process, such as the production of biofuels, fermented food, plant-based food, polymers, minerals, as well as the degradation of natural polymers (e.g. lignocellulose, lignin, cellulose, chitin, keratin, etc) and synthetic polymers (e.g. polyolefins, polyesters, polycarbonates, etc). Further applications also include biomining processes for the extraction of metals and rare earth elements from mineral ores or from recycled electronics and energy storage devices (e.g. batteries). Genetically-modified microorganisms produced according to this method may also be used as biofertilizers and for biocementation of building materials.

[0066] The method of this invention also facilitates the discovery of links between genetic mutations and physical traits, i.e. links between the genotype of a cell and its phenotype.

[0067] So that the invention may be more readily understood, certain terms are defined hereinunder.

[0068] The term "microorganism" as used herein refers to procaryotic or eukaryotic organisms of microscopic size, which may be a single cell, a cell cluster or a colony of cells.

[0069] The term "cell" as used herein refers to a single cell, which may be a eukaryotic cell or a prokaryotic cell, it may be a microorganism, or it may be a single cell which is isolated from a multicellular organism.

[0070] The term "genotype" refers to the genetic constitution of an individual organism.

[0071] The term "phenotype" refers to the observable traits of an individual organism or cell. It is determined by the genotype of the organism or the cell, as well as by environmental factors.

[0072] The term "genetically modified" microorganism or cell is a microorganism or cell whose genome has been engineered in the laboratory in order to favour the expression of desired physiological traits.

[0073] The term "macroscopic" means dimensioned to be seen without the use of a microscope.

[0074] The singular "a", "an" and "the" include the plural equivalents unless the context distinctly indicates otherwise. Likewise, "or" is intended to include "and" unless the context distinctly indicates otherwise.

[0075] The terms "comprises" or "contains" means "includes" in a non-limiting sense.

[0076] The terms "for example" and "such as" are used herein to indicate a nonlimiting example.Short description of the drawings

[0077] Exemplar embodiments of the invention are disclosed in the description and illustrated by the drawings in which:Figure la is a flow-chart depicting the steps of the method according to this invention;Figure lb is a flow-chart depicting essential and optional screening steps of a method according to this invention;Figure lc is a diagram showing the survival rates of cell suspensions exposed to different UV-C doses compared to a non-exposed sample; the culture exposed to 10 mJ / cm2(indicated by an arrow) was selected for the directed evolution process;Figure Id is a diagram depicting the impact of a cell resting and recovery process in the directed evolution process on the cell concentration after mutagenesis; error bars represent the standard deviation;Figure 2a illustrates the change in quantity of fluorescently-labeled cellulose over time, presenting the cellulose-correlated fluorescence in individual droplets as determined by confocal microscopy for about 1'200 sample droplets, with error bars corresponding to standard error of the mean;Figure 2b shows a histogram of the cellulose fluorescence signal in droplets containing single cells after mutagenesis (10 mJ / cm2) with 1 h or 50 h of recovery (A~0.1 CFU / droplet, 28'000 droplet events); the screening was performed after incubation of the droplets at 28°C for 24h;Figures 3a to 3f show single-cell laden droplets at different incubation times, respectively quantified cellulose production in the droplets at different incubation times, withFigures 3a, 3c and 3e showing representative confocal microscopy images taken after Oh, 6h and 24h incubation time, as indicated, with fluorescently labelled cellulose in cyan (excitation: 405 nm; emission: 432-460 nm) shown in fair grey in the dark grey droplets, the scale bar corresponding to 100 pm; and withFigures 3b, 3d, and 3f showing the corresponding histograms of cellulose fluorescence per droplet, quantified by image analysis (n ~1200 droplets) after Oh, 6h and 24h incubation time, as indicated;Figure 4a shows a histograms of cellulose fluorescence signal in droplets containing either the native K. sucrofermentans strain JCM 9730 (ATCC 700178), the control sample which was not irradiated, or the evolved K. sucrofermentans strain Ev5 (A ~0.1 CFU / droplet, 10'000 droplet events); the major fluorescence peaks were fitted with a Gaussian distribution;Figure 4b is a schematic presentation of a portion of the droplet sorting apparatus;Figure 4c depicts a histogram of the cellulose fluorescence signal in droplets containing the genetically-engineered cells;Figure 4d shows images of washed and freeze-dried bacterial cellulose pellicles having a diameter of 9.65 cm, produced in 100 ml cultures of the native parental K. sucrofermentans strain and by the evolved K. sucrofermentans Ev5 strain, with the scale bar corresponding to 1 cm;Figure 4e weight measurements of washed and air-dried bacterial cellulose pellicles grown for 12 days from single colonies having a diameter of 3 cm, grown in 5 ml cultures at 28°C under static conditions; weight was measured with both a scale and via thermogravimetric analysis (TGA) for each sample; error bars correspond to the error propagation; evolved strains (Ev 2, Ev3, Ev4, and Ev5) showed significantly increased cellulose production (54-70%) compared to the native strain (*p < 0.05, n = 3);Figure 5 is a shear stress diagram showing the rheological properties of the 3D-printing ink without bacteria (white circles / triangles) or with the native K. sucrofermentans strain (black circles / triangles);Figure 6a is an image of a centimeter-scale engineered living object obtained by 3D printing the bacteria-laden ink, the scale bar corresponding to 1 cm;Figure 6b is a representative stitched confocal image of fluorescently labelled bacterial cellulose (excitation: 405 nm; emission: 432-460) in the 3D-printed disks containing the native parental K. sucrofermentans strain after one day of incubation; the scale bar corresponding to 3mm;Figure 6c is a representative stitched confocal image of fluorescently labeled bacterial cellulose (excitation: 405 nm; emission: 432-460) in the 3D-printed disks containing the evolved K. sucrofermentans strain Ev5 after one day of incubation; the scale bar corresponding to 3mm;Examples of embodiments of the present invention

[0078] The individual steps of a method according to this invention, including optional steps indicated by a dashed arrow, are depicted in the flow-chart shown in Figures la and lb.

[0079] As an initial step SI, parental cells intended for mutagenesis are provided. The cells are subjected to random mutagenesis by exposing the cells to UV radiation, shown in step S2. The UV mutagenesis is not particularly limited to any specific protocol. Duration of UV irradiation and dosage may vary according to known protocols. The UV irradiation may be applied continuously for a period of time. The UV irradiation may also be pulsed. The cells may be exposed to UV pulses of a certain duration and energy over a period of time.

[0080] The UV irradiation causes a series of mutations in the genome of the cell. The number of mutations generally increases with longer exposure times, respectively a higher radiation energies applied to the cells. Due to the accumulating mutations, longer exposure, respectively high radiation energy levels reduce the viability of the cells. The radiation protocol should therefore be adapted such as to maintain a reasonable percentage of viable cells.

[0081] The population of viable mutated cells may subsequently be exposed to a defined selective pressure for directed evolution. The selective pressure is chosen suchas to select for a specific trait in the cell. The selective pressure may for example be provided by the presence and / or the absence of a certain compound or nutrient in the culture medium.

[0082] Following the random mutagenesis of the cells, it is important that the population of mutated cells are incubated in the absence of nutrients and light of a wavelength from 300 nm to 500 nm, for a period of time, for example for lh. This incubation in the absence of light of 300 nm to 500 nm, or in the dark is a resting period S3.1.

[0083] In a subsequent recovery step S3.2, following the resting period S3.1, the cells are cultured in rich media for a period of recovery time. Culturing the cells in a suitable growth medium allows the viable cells to become metabolically active and to amplify. Optionally, in order to favour the survival of cells with the desired phenotype, a selective pressure may be exerted on the population of mutated cells during this period.

[0084] It was found that a resting period followed by a longer recovery period, i.e. 50h as compared to lh, resulted in a higher number of cells having the desired phenotype, in this case cellulose overproducers.

[0085] Light has been demonstrated to activate DNA repair enzymes that repair damage caused by exposure to UV-light as part of a process known as photoreactivation. The photolyase enzymes of this process require exposure to radiation at 300 nm to 500 nm. Culturing the cells in the absence of light of this wavelength therefore allows the viable cells to grow and to maintain the mutations introduced by UV-mutagenesis. As only the viable cells amplify, the yield of viable, mutated cells in the population provided for the subsequent screening steps is higher following the recovery step.

[0086] The recovery time should not exceed 15 generation times, or 10 generation times, or 5 generation times, as cells comprising mutations with less impact on their metabolism tend to outgrow other mutants, thereby shifting their proportional representation in the mutant cell population in their favour.

[0087] Once the cells have recovered, the population of mutated cells can be screened to select the desired phenotype in step S4 of the procedure.

[0088] Figure lb provides more detail on this screening step S4. To allow for rapid identification of the desired phenotype, the cells of the mutant population are encapsulated in step S4.1 in a plurality of droplets at an average ratio of no more than one cell per droplet. The droplets may be liquid droplets, or they may be provided as a gel bead.

[0089] In one embodiment, the plurality of droplets are incubated at appropriate culturing conditions of the cell in a step S4.2, in order to allow for expression of relevant enzymes or material, for example cellulose fibers, which are causative the desired phenotype. The droplets may for example be cultured for 12h to 24h.

[0090] Following the encapsulations step S4.1 or the incubation step S4.2, the droplets are sorted based on phenotypic markers displayed by the cells encapsulated in the droplets in step S4.3.

[0091] In one embodiment the sorting is performed using a microfluidic approach. The presence of a desired phenotype may be detected optically, for example by using dyes to stain cells exhibiting the desired traits and detecting such dyes optically.

[0092] Once the presence of a phenotypic marker in the droplets has been determined, the cells may be separated.

[0093] In one embodiment the droplets are separated dielectrophoretically by applying an electrical pulse, which is triggered by the presence or by a certain quantity of a detected phenotypic marker. The electrical pulse is arranged to direct the droplets comprising a cell of the desired phenotype in one microfluidic channel, while the droplets which tested negative flow in another microfluidic channel.

[0094] It is however also possible to separate the droplets using other sorting techniques, such as commercially available Fluorescence-activated cell sorting (FACS) machines, or using another probe than fluorescence, for example absorbance, spectroscopy, or mechanical properties, either actively or passively.

[0095] Optical detection and separation may be automated or semi-automated.

[0096] In one embodiment, the selected droplets comprising cells of the desired phenotype are then used for the fabrication of an engineered living material, in particular a hybrid ELM (HLM). The droplets may for example be provided in an ink for 3D printing of macroscopic objects.

[0097] In one embodiment the droplets including the genetically-modified cells or the genetically modified cells selected according to this invention are used in engineered living material, as a biofertilizer, or for self-healing of concrete.

[0098] The method of this invention was performed in a study using Komagataeibacter sucrofermentans as a test strain ATCC 700178. The aim of this study was to identify genetic modifications which result in increased cellulose production of this naturally cellulose-producing strain.

[0099] K. sucrofermentans is a Gram-negative aerobic bacterium that produces long cellulose nanofibers found in food products, wound dressing materials and high-end acoustic membranes.

[0100] K. sucrofermentans metabolizes sugars to UDP-glucose, which is then catalyzed by the transmembrane cellulose synthase complex into P( 1-4) glucan chains that self-assemble into cellulose fibers while exported through the cell wall. The cellulose synthase complex in K. sucrofermentans is encoded by many copies of the bcsA, bcsB, bcsC, and bcsD genes, some of which are gathered in operons. Since the function of some of those genes are not yet fully understood and that other genes are expected to be involved in the regulation of the cellulose formation process, a wholegenome directed evolution approach was chosen for the generation of mutant strains with the desired traits, i.e. an increase in cellulose production.

[0101] The directed evolution process described herein aimed to boost bacterial cellulose production within a 24-hour fermentation time frame. The aim was to generate strains comprising one or more genetic mutations resulting in increased cellulose production of the mutated strain compared to its parental strain. To enhance the efficacy of the mutation and selection process, the direct evolution process wascombined with a microfluidic approach. The approach mutagenic approach is further detailed in the methods section below.

[0102] The approach taken was to enhance the bacterial production of cellulose by exploring the entire genome of the microorganism rather than introducing mutations only in the genes encoding a specific target protein. To this end, a random mutagenesis approach using UV light was chosen and a library of mutants was created. This approach allowed to tap into the broad diversity of possible genetic mutants without prior assumptions about the genes controlling cellulose bio-synthesis. From this initial mutant library, mutants with the desired phenotype were selected by performing direct evolution cycles in a droplet microfluidic platform as described in the material and methods section. Encapsulating a single bacterium in the individual droplet provides the advantage that a direct link between genotype of the single bacterium and its phenotype is directly established.

[0103] The microfluidic platform used herein consisted of three parts: a droplet generator for cell encapsulation, an incubation container for cell growth, and a highspeed droplet sorter.

[0104] To enable directed evolution of individual cells, one microorganism per droplet was encapsulated. This was achieved by adjusting the initial cell concentration in the culture medium accordingly. After encapsulation, cell-laden droplets were incubated in glass containers for 1-3 days at optimum growth conditions in the presence of a cellulose-staining fluorescent dye.

[0105] Droplets containing mutants that overproduce cellulose were dielectrophoretically separated from the remainder of the population by applying an electrical pulse. The electrical pulse was triggered when the measured fluorescence of the droplet surpassed a defined threshold level.

[0106] Sorted droplets were compared to droplets comprising the native parental strain as well as to a control sample droplet, which were treated like the population of mutated cells but which had not been subjected to UV mutagenesis.

[0107] UV-C irradiation is used in the mutagenesis approach and is known to severely impact the viability of bacteria due to excessive DNA damage. In other words, if the DNA damage resulting from UV-C irradiation is too severe the viability of cells and, as a consequence, the efficacy of the direct evolution process is compromised.

[0108] In order to create a library of bacteria with a large fraction of viable mutants, the effect of UV-C irradiation dose on the survival rate of cells was investigated in greater detail.

[0109] To test this, a liquid salt suspension of K. sucrofermentans was irradiated with UV-C light (wavelength 254nm) at doses up to 100 mJ / cm2. After UV exposure, the irradiated samples were put in the dark for 1 h, and then resuspended in rich culture medium to allow the cells to recover from the stress of UV exposure. The bacteria were incubated for lh to 50h at 28°C, which is the conventional culture temperature of K. sucrofermentans. The incubation in the dark served to inhibit the onset of DNA repair mechanisms, which counteracts the UV-induced mutagenesis. To quantify the cell survival rate, bacterial culture samples exposed to different light doses were directly frozen after lh or recovery for subsequent colony forming unit (CFU) counting analysis.

[0110] The survival rate of cell suspensions exposed to different UV-C doses compared to a non-exposed cell suspension is shown in Figure lc. The survival rate was calculated relative to this control sample, which underwent the same steps but was not irradiated. The cell counting analysis revealed that the survival rate of the irradiated bacteria decreased from 98% to 17% upon an increase in UV dose from 0.5 mJ / cm2to 10 mJ / cm2. Exposure to the highest dose of 100 mJ / cm2led to severe DNA damage and complete cell death. Based on these results, a dose of 10 mJ / cm2, which is indicated by an arrow in Figure lc, was found to be most appropriate to ensure a high fraction of mutants in the population while keeping a reasonable percentage of bacteria alive. As shown in Figure lc, the survival rate under these radiation conditions was about 17%.

[0111] The viability of cells exposed to 10 mJ / cm2was confirmed by growing the bacteria for 50 hours in rich culture. In this time interval, the concentration of mutant bacteria was found to increase by more than 8-fold between 1 h of incubationand 50 h of incubation. As shown in Figure Id, the determined cell concentration was 4.7 M CFU / ml, when recovering for 1 h, and 38.7 M CFU / ml, when the cells were allowed to recover for 50h. Error bars represent the standard deviations.

[0112] The relatively concentrated cell suspension obtained after a 50 h recovery period and enriched in fast-growing bacteria was used as feedstock for the microfluidic encapsulation process.

[0113] Mutant bacteria were encapsulated in droplets using a microfluidic device via a step emulsification approach. The diameter of the droplets was about 49 pm. By operating at a high throughput rate in the order of thousands of droplets / second and reaching a droplet polydispersity index (PDI) of only 10-3, the microfluidic platform provided a large library of mutants which was ideal for directed evolution experiments.

[0114] In the emulsification process, droplets of the recovered, diluted bacteria suspension were emulsified in a fluorocarbon oil with the help of a biocompatible surfactant, in this case a fluorosurfactant. The number of cells encapsulated in a single microfluidic droplet is known to follow Poisson statistics and to depend on the concentration of bacteria in the feedstock suspension.

[0115] The concentration of the bacterial suspension was adjusted (A. = 0.1 CFU / droplet) to obtain a droplet population with about 90.5% of empty droplets, about 9.0% of droplets containing a single encapsulated cell, and about 0.5% of droplets with more than 1 cell.

[0116] An important task in droplet-based directed evolution processes is the appropriate design of an assay for the fast, on-chip quantification of the performance of encapsulated mutants. Given the ease and availability of fluorescent-based detection tools, we used the Fluorescent Brightener 28 to quantify the production of cellulose by the encapsulated bacteria. This commercially available dye is known to selectively bind to the (1-4)|3 bonds present in cellulose. To detect cellulose formation inside the droplets during incubation, we added the dye directly into the bacterial suspension used in the encapsulation process. The cellulose-forming capabilities of the encapsulated bacteria was measured after incubation of the monodisperse droplets at28°C for up to 4 days. Measurements were performed in a confocal microscope using a 405 nm laser and a 432-460 nm detector.

[0117] The results of these measurements are shown in Figure 2a. It was observed that the majority of the cellulose was produced within the first day of droplet incubation, with a maximum fluorescence reached after two days of incubation (left axis). Error bars correspond to the standard error of the mean. About 7% of droplets contained a cellulose-producing bacteria, which is close to the 9.5% value expected from Poisson statistics for a cell concentration A. of 0.1 CFU / droplet (right axis).

[0118] The positive effect of the resting and recovery steps on the yield of genetically-modified cells of the desired phenotype becomes evident from Figure 2a. This Figure shows histograms of the cellulose fluorescence signal in droplets containing single cells after mutagenesis, which were cultured to recover in rich medium for 1 h or for 50 h. Mutants recovered for 50 h showed a significant increase in cellulose production compared to mutants which recovered for 1 h only, as evidenced by cellulose fluorescence in the droplets. Using a library of mutated cells which recovered for a longer period of time, for example for 30 to 50 hours is therefore expected to drastically increase the chances of finding a mutant strain having the desired phenotype.

[0119] Confocal microscopy images taken during incubation show that 6 hours were already sufficient for the bacteria to start producing cellulose inside the droplets. This indicates that the confinement in droplets does not impair the ability of the microorganism to synthesize cellulose. At this early stage, the cellulose was observed to form primarily around the encapsulated bacteria, as seen in Figures 3a to 3d. The results indicate that their confinement in droplets does not impair the ability of the microorganism to synthesize cellulose. At this early stage, the cellulose was observed to form primarily around the encapsulated bacteria, as seen in Figure 3b.

[0120] When extending the incubation period to 24 hours, the bacteria-derived cellulose is no longer limited to the bacteria but occupies most of the volume of the droplets, as seen in Figure 3c. Figures 3a to 3f show fluorescence distribution data obtained by image analysis, wherein Figures 3a, 3c and 3e display representative confocal microscopy images of single-cell-laden droplets over the incubation timeindicated in the respective figure. The droplets were exposed to an excitation wavelength of 405nm, and an emission range of 432nm to 460nm was detected. The scale bar represents 100pm. These Figures clearly evidence the emergence of a population of cellulose-containing droplets after 6 and 24 hours of incubation. These results validate the fluorescence-based approach to quantify the cellulose-producing capabilities of the mutant bacteria encapsulated in droplets. Figures 3b, 3d, and 3f show the corresponding histograms of cellulose fluorescence per droplet as quantified by image analysis with a sample size of about 1'200 droplets.

[0121] Due to the high emulsification rate of 2000 droplets / second, the microfluidic platform enabled the generation of a library of 1.2 million droplets in 10 minutes. This corresponds to approximately 100'000 single-cell encapsulated mutants. A fraction of this large pool of mutants (430'000 droplets, ~40'000 mutants) was further analyzed in a microfluidic droplet sorter to select for possible cellulose overproducers after 24 hours of droplet incubation off-chip. Selection was performed by fluorescence-activated droplet sorting (FADS) in a high-speed customized microfluidic device (800 Hz). The fluorescence emitted by the produced cellulose was measured as the droplet moved along the main microfluidic channel. A photomultiplier tube (PMT) was used to quantify the droplet fluorescence at 488 nm using a 405 nm laser for excitation. The PMT voltage was utilized as a proxy for the fluorescence emitted by the droplet. Experimental results confirmed the importance of a long recovery time, for example 50 hours, after UV-C exposure to obtain a strong fluorescent signal after 1 day of incubation as shown in Figure 2b.

[0122] A bifurcation at the end of the main channel was used to separate cellulose overproducers from the pool of mutants. The few droplets with fluorescence above a pre-defined threshold are dielectrophoretically pulled into an outlet channel of sorted bacteria, while the vast majority was discarded in a waste outlet. The PMT voltage threshold was set to 2.75 V to separate the 1.25% most fluorescent variants from the pool.

[0123] Sorted droplets containing evolved bacteria were collected and spread on solid media to enable the growth of individual mutants into colonies. Five colonies of evolved bacteria (Evi to Ev5) were chosen for further analysis. The evolved bacteria were compared to native and control strains by measuring their cellulose-formingcapabilities inside droplets or in bulk pellicles. For the analysis in droplets, the bacteria were re-encapsulated (A ~0.1 CFU / droplet), incubated for 24 h and screened using the same microfluidic platform used for the initial evolution experiment. This time no sorting was performed. Instead, the analysis of the fluorescence intensity distribution obtained for the evolved, native and control bacteria was performed directly.

[0124] The evolved mutants clearly outperformed the native and control bacteria in terms of cellulose production inside droplets. Fluorescence intensity histograms obtained for these three variants show a Gaussian-like distribution that is strongly skewed towards high fluorescence value, as observed in Figure 4a, showing histograms of cellulose fluorescence signal in droplets containing native, control and evolved (Ev5) bacterial cells, with A ~0.1 CFU / droplet and measurements for approximately 10'000 droplet events.

[0125] To quantify the performance of the distinct bacteria, we fitted the distributions with a Gaussian function and compared the values of peak fluorescence intensity. The evolved and control bacteria respectively displayed a mean cellulose fluorescence 52% and 36% higher than that measured for the native strain, respectively. This enhanced cellulose-producing capability is also evidenced by the stronger maximum fluorescence signal measured for droplets loaded with the evolved bacteria.

[0126] A schematic illustration of a portion of a droplet sorter as used in the present study is shown in Figure 4b. The fluorescence of the droplets passing a microfluidic channel is determined using a laser beam 1 at excitation wavelength of the fluorescence dye. Electrodes 2 positioned just above the bifurcation of a main microfluidic channel provide the electrical pulses for directing selected droplets 3 into a first side channel, while droplets which tested negative 4 flow into a second side channel.

[0127] Figure 4c shows a histogram of the cellulose fluorescence signal in droplets containing the mutated cells which had been exposed of UV-C radiation of 10 mJ / cm2 and were subsequently incubated in the dak for 1 h and then allowed to recover in rich media for 50h. The droplets were incubated at 28°C for 24 hours (A ~0.1 CFU / droplet, 430'000 droplet events, ~40'000 variants screened). All droplets with acellulose fluorescence signal corresponding to a PMT voltage above 2.75, i.e. the droplets whose results are depicted right of the dashed line, were sorted, collected, and spread onto solid media. The sorted droplets correspond to 0.12% of all droplets and ~1.25% of all droplets comprising mutant bacterial cells.

[0128] The over-production of cellulose by the evolved bacteria is not limited to droplets but can also be translated into the bulk manufacturing of thicker cellulose pellicles. To illustrate this, the evolved, native and control bacterial cells from single colonies were cultivated in a liquid culture medium which allowed for the growth of a BC pellicle at the air-water interface in static conditions. A long incubation time of 12 days was chosen to ensure cells reached the stationary growth phase and to maximize cellulose production. This also minimized the possible effect of slight differences in the initial inoculation amounts. Pellicles were then washed and dried to quantify the amount of cellulose produced by the different variants. Freeze-drying was performed to obtain macroscopic samples, whereas thermogravimetric analysis and weight measurements in a lab scale were carried out to quantify the amount of cellulose formed by the different strains.

[0129] Visual inspection of the freeze-dried films clearly showed the production of thicker pellicles by the evolved bacteria compared to the native species, as the images taken of the washed and freeze-dried pellicles of Figure 4d demonstrate. Both scale measurements and thermogravimetric analysis of air-dried pellicles heated up to 650°C revealed that 4 of the 5 evolved strains (Evi, Ev2, Ev3, Ev4, and Ev5) chosen for analysis produced 54wt% -70wt% more cellulose compared to the native bacteria, as depicted in Figure 4e, which shows weight measurements of washed and air-dried bacterial cellulose pellicles grown for 12 days from single colonies.

[0130] The long fibers obtained from these strains show the typical interwoven microstructure of bacterial cellulose and were similar to the fibres produced by the parental strain ATCC 700178.

[0131] To evaluate the stability of the enhanced phenotype, passage experiments in which bacterial cellulose pellicles were grown from previous pellicles over multiple generations were performed. The results demonstrate that the higherproduction of cellulose by the selected evolved strain (Ev5) was stable even after 5 passages.

[0132] K. sucrofermentans strain Ev5 was named K. sucrofermentans JML 2321 and deposited with the Culture Collection of Switzerland (CCOS) under the accession number CCOS 2107.

[0133] In addition to this stability experiment, an additional directed evolution cycle of the Ev5 strain to explore the possibility of further phenotype improvement was performed. To this end, the evolved strain was mutated, encapsulated, incubated, and sorted with two different thresholds: (i) the same condition as in the first round (2.75 V) and (ii) more stringent selection pressure (3.60 V). The selected strains of this second round of directed evolution did not show a further increase in cellulose production, but maintained a similar production as in the first evolution round.

[0134] In addition to producing cellulose in the form of thick bulk pellicles or in the form of pellets or granules, the ability of the evolved bacteria to overproduce cellulose also opens new opportunities for the manufacturing of more complex engineered living materials.

[0135] To further explore this possibility, a 3D printable gel that can be loaded with the evolved cellulose-producing bacteria was designed. Networking-forming silica particles and viscosity-modifiers hyaluronic acid and K-carrageenan were used to tune the rheological properties of the gel according to a known protocol (Schaffner M et al, 2017, "3D Printing of Bacteria into Functional Complex Materials", Sci Adv. 3, DOI: 10.1126 / sciadv.aao6804). Oscillatory rheology experiments confirmed that gels with or without cells display elastic response at low shear stresses and become a viscous fluid above a well-defined yield stress of 200 Pa, as shown in Figure 5a. Figure 5a depicts the rheological properties of the ink with no bacteria (white) or with the native K. sucrofermentans strain (black). The crossover between the storage (G') and loss (G') moduli occurs at a yield stress of 200 Pa for both inks. It could therefore be demonstrated that the properties of the ink laden with bacteria fulfill the rheological requirements for 3D printing via the extrusion-based direct ink writing (DIW) technique.

[0136] Cell-laden gels were successfully printed into a three-dimensional object with complex geometry at the centimeter scale, an example of which is shown in Figure 6a. In addition to the rheology modifiers, the ink also contained the nutrients required for the proliferation and growth of the embedded bacteria. This allowed for the in-situ production of cellulose fibers within the printed gel during an incubation period of 1 day.

[0137] The evolved bacteria may therefore be used to generate cellulose-based macroscopic objects using state-of-the-art manufacturing technologies, such as casting, molding and 3D printing.

[0138] To gain insights into the effect of the bacterial strain on the formation of cellulose within the gel, we printed discs containing the cellulose-binding dye and loaded with the evolved or native strains, and imaged them in a confocal microscope. The fluorescence of the cellulose-binding dye was taken as a proxy for the local concentration of cellulose produced by the bacteria. To restrict oxygen supply to the edges of the sample, the 12-mm printed discs remained sandwiched between a Petri dish and a cover slip during incubation and imaging.

[0139] Confocal images of the printed discs revealed that both strains produce cellulose predominantly along the edge of the sample next to the air-water interface (Figure 6c). However, the evolved strain produces more cellulose than the native strain at the edge, as indicated by the increased fluorescence, but in a narrower ring (Figure 6b). Radially integrated fluorescence measurements of the samples further provided clear evidence of the distinct cellulose patterns created by the two bacterial strains. This example illustrates how the architecture of engineered living materials can be tuned at different length scales by combining the top-down manufacturing capabilities of 3D printing with the bottom-up self-assembly processes controlled by microorganisms.

[0140] To establish the origin of the enhanced cellulose production in the evolved strains, the genome of the evolved bacteria was compared with that of the native strain. For this, a high-quality reference genome was obtained, using both shotgun and long read methods to sequence the native strain. The assembled genome contained a single 2.95 Mbp genome and 4 plasmids, as schematically depicted inFigure 7. The genomes of the evolved and control strains were sequenced with the Illumina platform and aligned to the reference genome. The breseq algorithm known in the art was used to predict mutations from annotated sequences.

[0141] The genomic analysis revealed no mutation in the control strain but a consistent and unique mutation in four of the evolved strains found to be overproducers of cellulose (Ev2, Ev3, Ev4, and Ev5): a 12-base pair deletion within the reading frame of the dpA gene (SEQ ID NO 1). The missing nucleotides code for a sequence of four amino acids of the N-terminal domain of the CIpA protein, i.e. amino acids QRVI at positions 84 to 87 in the CIpA polypeptide (SEQ ID NO 2).

[0142] Surprisingly, this highly significant deletion was found in a gene encoding a polypeptide of the CIpAPS protease complex, and not in any of bcs genes encoding the multiple protein domains forming the cellulose synthase complex.

[0143] The CIpA protein is part of the CIpAPS complex, which is responsible for the hydrolysis of misfolded and degraded proteins inside cells. The function of CIpA is to bind, unfold and lead proteins to CIpP, where the hydrolysis process takes place. Under normal conditions, the proteins to be degraded are fixed through the specific interaction between the N-terminal domain of CIpA with the adaptor protein CIpS, as previously demonstrated for the Escherichia coli CIpAPS complex.

[0144] In fact, it could be established that the deleted 12 base pairs in the mutant strains code for four amino acids of the ClpS-binding domain of CIpA. In particular, the Arginine residue (R85) in K. sucrofermentans was shown to be highly conserved across species and involved in the CIpS bonding to CIpA (Zeth K et al, (2002), Nature Structural Biology 9, 906 - 911, doi:10.1038 / nsb869).

[0145] The results therefore provide evidence for a previously unknown function of the CIpAPS complex, respectively adverse effect of CIpA / CIpS binding, in bacterial cellulose production.

[0146] The directed evolution of the microorganism K. sucrofermentans in microfluidic droplets allowed for the generation of a novel bacterial strain for the efficient production of cellulose in engineered living materials.

[0147] A commonly used fluorescent dye was found to be an effective probe for the high-speed quantification of cellulose produced by single microorganisms inside the microfluidic droplets.

[0148] Starting with a library of 40'000 randomly mutated variants, 4 microorganisms were identified as cellulose overproducers at the end of the directed evolution process.

[0149] Using hydrogel inks as host for the selected bacteria, it could be demonstrated that the evolved microorganisms can be 3D printed into cellulose-based complex objects with structural features at macroscopic scale.

[0150] By enabling the evolution of entire microorganisms towards specific functionalities, the high-throughput directed evolution method of this invention holds great potential as a discovery tool for novel microbial strains, as well as for genotypephenotype correlations in engineered living materials. Moreover, the directed evolution of whole microorganisms might also be an effective strategy to enhance the efficiency of existing and prospective biotechnological processes.Material and MethodsCell culture for bacterial cellulose

[0151] Komagataeibacter sucrofermentans JCM 9730 (ATCC 700178) is the cell u- lose-producing strain used in this study. Their growth media was composed of 25 g / L D-mannitol (Thermo Fisher Scientific), 5 g / L yeast extract (Sigma-Aldrich), 3 g / L peptone (Sigma-Aldrich), and optionally 15 g / L agar (Sigma-Aldrich) for solid medium. Frozen stocks of K. sucrofermentans (-80°C) were streaked on media plates to isolate single colonies, which were then inoculated in 5 ml growth media in 50 ml Falcon tubes (TPP). The lid was replaced with a foam plug to allow optimal oxygen availability. The cultures were incubated in static conditions at 28°C for 8-12 days to form bacterial cellulose (BC) pellicles at the air-media interface. For passage experiments (phenotype retention), a sterile loop was used to rub the BC pellicles and streaked on solid medium, from which single colonies were picked for subsequent liquid cultures.Cell concentration

[0152] All experiments were started from frozen stocks of bacteria. For each stock, serial dilutions from 101to 10-6were prepared and plated on solid media in drops of 5 pl (n = 6). Colony-forming units (CFU) were counted at an appropriate dilution, averaged, and the stock concentration was back-calculated.Absorbance measurements

[0153] Absorbance measurements were taken with the Varioscan LUX (Thermo Fisher Scientific) at 600 nm, on 200 pl samples in 96-well plates (flat bottom, TPP) unless stated otherwise. Samples were always measured in triplicates, averaged, and blanks were subtracted. Blanks corresponded to media, with 2 vol% cellulase (Tricho- derma reesei ATCC 26921, Sigma-Aldrich) and 35 pg / pl chloramphenicol (Sigma-Aldrich) only when also present in the samples.UV-C mutagenesis

[0154] K. sucrofermentans from frozen stocks were inoculated in 150 ml of liquid media for 6 days at 28°C at 200 rpm. 2% cellulase was added to the culture to digest any cellulose produced. After measuring the absorbance, cells were spun down at 3275 ref for 10 min (Z306 Hermle) and resuspended in 0.9% NaCI (VWR) to reach a theoretical absorbance of 1. 10 ml of the cell suspension was added to each Petri dish (010 cm, TPP) to be exposed to different UV-C doses without the lid: 0, 0.5, 1, 2, 3, 5, 10, and 100 mJ / cm2(254 nm, UVP Crosslinker CL-3000, AnalytikJena). The plates were then left in the dark for 1 h, spun down as previously, resuspended in enriched media (2X concentration) with 2% cellulase, and incubated at 28°C and 200 rpm for either 1 h or 50 h. To store the cultures at -80°C for further analysis, aliquots with 20% glycerol (Fisher BioReagents) were prepared. To quantify the cell survival, the 1 h cultures were serial diluted and plated on solid media (5 pl dots, n = 6). Colony-forming units (CFU) were counted at the IO-4dilution and compared to the Control which was not exposed to UV-C (0 mJ / cm2). The 50 h aliquots were used for the directed evolution process, assuming more recovery time was beneficial to increase our chances of finding a strain which grows sufficiently well in our culture media.Fabrication of microfluidic devices

[0155] SU-8 (3000 series, MicroChem) was patterned on a Silicon wafer using standard photolithography methods. Those were used as masters for 1:10 PDMS (Syl- gard™ 184, Dow Corning) soft lithography. After curing, the devices were peeled off, inlet and outlet holes were punched, and the devices were air-plasma bonded to glass slides (Fisherbrand™ Superfrost™). To hydrophobize the devices, a 2 vol% solution of lH,lH,2H,2H-perfluorooctyltrichlorosilane (Fluorochem) in HFE-7500 (3M) was flushed through the inlet and then air dried. For the sorting devices, electrodes were created by inserting a low melting point solder (51ln / 32.5Bi / 16.5Sn; Indium Corporation, New York, USA) at one end of the electrode channel and a wire at the other end. The entire chip was subsequently placed on a 150°C hotplate for 3 minutes to melt and reflow the solder across the channel.-cell encapsulation and incubation

[0156] Cell loading: The desired cell loading was evaluated with the well-known Poisson distribution representing the probability of having k cells in a droplet knowing the average number of cells per droplet volume : p(k, ) = - . Knowing the drop- kl let size and frozen stock concentration, we aimed at a of 0.1 CFU / droplet, corresponding to a ~9.5% droplet occupancy. This low number was chosen to minimize coencapsulation events (<0.5%). For each encapsulation experiment, cells were thawed from frozen stocks, and their concentration was adjusted to a of 0.1 CFU / droplet (~1.5M CFU / ml). The occupancy was validated by image analysis, estimating the percentage of droplets in which cellulose was produced (see Droplet image analysis section).

[0157] Cell encapsulation: Cells were encapsulated in ~50 urn droplets using a step emulsification PDMS microfluidic device. The inner aqueous phase was composed of media, cells at low concentration ( = 0.1 cells / d roplet), 218 uM Fluorescent Brightener 28 (FB, Sigma-Aldrich) to stain the BC, and 26 uM fluorescein (Sigma-Aldrich) to get a background signal in the droplets. The outer phase comprised 2 wt% 008-Fluoro- Surfactant in HFE-7500 (RAN Biotechnologies). Both phases were flown at 500 uL / h for 10 minutes, corresponding to more than a million droplets formed.

[0158] Droplet incubation: Droplets were incubated in hydrophobized glass vials, on top of 200 pL HFE-7500 (3M), at 28°C in static conditions for 24h before the sorting process. The vials were horizontally placed to guarantee a monolayer of droplets and equal access to oxygen.analysis

[0159] Prior to screening in microfluidic devices, droplet sizes and cellulose content were analyzed using z-stacks of confocal images (30 slices, 3.58 urn each). A 405 nm laser was used to excite the cellulose-specific dye (Fluorescent Brightener 28, Sigma-Aldrich), which was detected between 432-460 nm (HyD detector). An additional PMT transmission detector was active to image the droplets. Using ImageJ,42the FB images were summed, and one of the droplet images was chosen for edge detection. Droplet diameters were estimated using the Hough Transform plugin (UCB Vision Sciences) after thresholding the image. For fluorescence quantification in the droplets, the FB sum stack and chosen droplet images were imported to Cell Profiler 4.1.3, a user-friendly software previously reported for droplet image analysis. Briefly, droplets were detected as objects in a desirable size range and filtered out if on the edge of the image, and then the object intensity of each object was measured. Those results were exported as an Excel file, and all were analyzed and plotted with MATLAB (R2023a, Math Works). Droplets were considered occupied when their measured FB fluorescence was exceeding the maximum fluorescence detected on Day 0.

[0160] Droplets were then transferred to a second PDMS microfluidic device, in which they were spaced with oil (HFE-7500, 3M). After passing under a 405 nm laser, the fluorescently labeled cellulose intensity was estimated with a 488 nm PMT detector, which we expect to be proportional to the voltage detected. Based on the fluorescein background in the droplets, the baseline was adjusted to 0.05 V. 0.43M droplets were detected in less than 10 min, of which 505 were sorted with a threshold of 2.75 V. Those droplets were collected in a sterile Eppendorf tube. After adding media and demulsifying, the droplets were plated onto solid media. Five evolved single colonies were then picked and grown in liquid media with 2% cellulase, and aliquots with 20% glycerol (Fisher BioReagents) were frozen for further analysis (Evl-Ev5).Cellulose dry weight

[0161] To quantify the amount of BC produced by the evolved strains compared to the native and control (0 mJ / cm2) strains, triplicates of BC pellicles were grown as previously explained. After 12 days, they were washed 3 times with 0.1 M NaOH (Fisher Scientific, UK) at 60°C in a water bath over a period of 24h and then brought back to neutral pH washing 3 times with MilliQ water (NANOpure Diamond, Barnstead). They were then dried in a 60°C oven for 48h on Teflon films (McMater-Carr, OH) to avoid sticking, and stored under vacuum until analyzed. Dried bacterial cellulose pellicles were weighed with a precision balance (UMT2 Microbalance, Mettler Toledo). Ther- mogravimetric analysis (Discovery TGA 5500, TA Instruments) was as well performed on the dried cellulose pellicles by increasing the temperature of 10°C / min until 650°C, with an isotherm of 15 min at 120°C to remove all the humidity contained in the samples. Weight loss was then calculated between the end of the 120°C isotherm and the end of the program at 650°C.Genome sequencing, assembly and analysis

[0162] Genome sequencing was performed by MicrobesNG (United Kingdom). The K. sucrofermentans DSM 15973 reference genome was sequenced with an Illumina NovaSeq 6000 (Illumina, San Diego, USA) using a 250 bp paired end protocol, as well as with GridlON (Oxford Nanopore Technologies, UK) to obtain long reads. Strains selected from the directed evolution process were sequenced with Illumina NovaSeq 6000 (Illumina, San Diego, USA) only.

[0163] Reference genome assembly combined both long and short reads using Unicycler version 0.4.0, and annotation was performed with Prokka 1.13. To quantify the quality of the assembly, coverage statistics were calculated using the short read data using samtools 1.3.1Mutations were detected using breseq 0.35.1, aligning the Illumina paired end reads to the reference genome. Bioinformatic analysis of specific genes and operons was performed with BioPython.3D printing

[0164] Ink Preparation: The 4.5% ink is composed of 1.5 wt% sodium hyaluronate (BulkSupplements), 1.5 wt% K-carrageenan (Acros Organics), and 1.5 wt% fumed silica (WDK V15, Wacker Chemie) in the standard media used for K. sucrofermentans described above. The ink was prepared following the previously published protocol (Schaffner, M et al. (2017): 3D Printing of Bacteria into Functional Complex Materials, Sc / . Adv. 3, eaao6804), UV-C sterilizing all powders before mixing and adding 218 uM of Fluorescent Brightener (Sigma-Aldrich) to stain the cellulose produced in the ink. For each experiment, a new batch of ink was prepared and separated in three equal amounts, in which 0.36M CFU / g of bacteria were added from frozen stocks. For controls, the same volume of sterile media was added (900 uL). The final inks were loaded into 10 ml syringes and kept at 4°C until 3D printing on the same day.

[0165] Ink Rheology: Rheological measurements of the inks were performed with a sandblasted parallel plate geometry (PP25-S, Anton Paar) at 25°C (MCR 302 compact rheometer, Anton Paar). The storage and loss moduli (G' and G") were analyzed with constant oscillatory measurements (1 s1) with shear strain amplitude increasing logarithmically from 0.01 to 100%. Flow curves were then recorded with rotational strain- controlled measurements with a shear rate ramping logarithmically from 0.01 to 100 s-1and then from 100 to 0.01 s-1.

[0166] 3D printing: All structures were 3D printed using a 10 ml syringe, a 0.84 mm needle, and a layer height of 0.7 mm. 1-layer disks of 12 mm diameter were 3D printed directly on Petri dishes. Immediately after the print was done, the disks were covered with coverslips, so oxygen was only available from the sides. Each Petri dish was then sealed with parafilm and imaged as is with a confocal microscope (TCS SP8, Leica) after a day of incubation at 28°C and 85% RH in static conditions. For each full disk, a 15-slice stack of 750 urn total was imaged (excitation: 405 nm; emission: 432- 460 nm) and stitched together using the Leica software.

[0167] Image analysis: All the image analysis was performed using ImageJ. Slices of the stacks were averaged, and the radial profile of each disk was plotted. Data was then smoothed and plotted with MATLAB (R2023a, Math Works).Growth curves

[0168] Each tested strain was thawed, spun down at 3000 ref for 10 min (5417R, Eppendorf), resuspended in 1 ml of media, and their absorbance was adjusted to 0.005. 2% of cellulase was added to each culture to digest the produced cellulose. 1 ml of each culture was then transferred to 24-well plates (flat bottom, TPP) in triplicates, and the well plates were covered with a breathable film (BREATHseal™, Greiner Bio- One) to avoid contamination between wells. The well plates were incubated at 28°C and 85% RH in shaking conditions (200 rpm). Every day, the breathable film was removed for the measurement, the absorbance at 600 nm of each well was measured, triplicates were averaged, and blanks composed of media and 2% cellulase were sub- tracted from the measurement. The breathable film was changed, and the plates were incubated until the next measurement.Statistical analysis

[0169] The significance of the cellulose increased production was evaluated with a one-way ANOVA (MATLAB R2023a, Math Works), followed by a pairwise comparison if the results showed a statistically significant difference between the groups (p < 0.05).A Bonferroni correction was applied to compensate for the effects of multiple comparisons.

Claims

Claims1. Method for providing genetically modified cells having a defined phenotype comprising(i) providing a microorganism or a cell of a multicellular organism,(ii) performing random mutagenesis on the microorganism or on the cell of an organism causing random mutations in said microorganism or in said cell such as to obtain a population of genetically modified cells,(iii) optionally, if the random mutagenesis by exposure to UV light is performed, incubating the population of genetically modified cells in the absence of light between wavelengths of 300nm and 500nm for a period of resting time,(iv) culturing the population of genetically modified cells in suitable growth medium for a period of recovery time,(v) encapsulating of at least a portion of the genetically modified cells in a plurality of droplets or capsules,(vi) in the droplets or capsules, detecting the presence of a phenotypic marker which is indicative of a defined phenotype of the genetically modified cell, and(vii) selecting the droplets or capsules, which comprise the detected phenotypic marker, which comprise a quantity of the detected phenotypic marker exceeding a defined threshold, which comprise a quantity of the detected phenotypic marker falling short of a defined threshold, or which do not comprise detectable amounts of the phenotypic marker.

2. The method of claim 1, wherein an adverse selective pressure is applied to the genetically modified cells during the culturing step (iv).

3. The method of claim 1 or 2, wherein the microorganism or cell is exposed to UV-light resulting in random mutations in said microorganism or in said cell such as to obtain a population of genetically modified cells.

4. The method of any of claims 1 to 3, further comprising a step of culturing the genetically modified cells within the droplets or capsules prior to detection for a period of expression time, for example for 30 min to 1 week, or for 12h to 24h.

5. The method of any of claims 1 to 4, wherein subsequent to the UV-induced mutagenesis step, the population of genetically modified cells is incubated in the dark for the period of resting time.

6. The method of any of claims 1 to 5, wherein the resting time ranges from 30 min to 3 h.

7. The method of any of claims 1 to 6, wherein the period of recovery time corresponds to 1 to 15 generations times, or 5 to 10 generation times.

8. The method of any of claims 1 to 7, wherein the genetically modified cells are encapsulated in the plurality of droplets or capsules at an average ratio of no more than one cell per droplet or per capsule.

9. The method of any of claims 1 to 8, wherein encapsulating comprises contacting a genetically modified cell of the population of cells with an aqueous suspension, or with a hydrogel, and emulsifying droplets in a fluorocarbon oil with the help of a biocompatible surfactant to form a droplet comprising the cell.

10. The method of any of claims 1 to 9, wherein the detection of a phenotypic marker, which is a marker molecule comprises contacting the marker molecule with a detection agent, for example a fluorescent dye or a marker specific ligand coupled to a fluorescent dye, which adheres to the marker molecule upon such contact.

11. The method of claim 10, wherein the detection agent is or comprise a fluorescent dye and wherein the detection method is an optical detection method comprising irradiating one or more droplets with a light having an excitationwavelength of the fluorescent dye and optically detecting light emitted in the droplet at an emission wavelength of the fluorescent dye.

12. The method of any of claims 1 to 11, further comprising the step of separating the droplets selected in step (iv) from the remainder of the plurality of droplets.

13. The method of claim 12, wherein the selected droplets are separated from the plurality of the droplets dielectrophoretically by applying an electrical pulse, which is triggered by the presence of a phenotypic marker, or by a defined quantity of a phenotypic marker.

14. The method of any of claims 1 to 13, wherein the provided microorganism is a bacterial cell.

15. Use of a genetically-modified cell produced and selected according to any of claims 1 to 14, or a selected droplet or capsule containing a genetically-modified cell according to any of claims 1 to 14 in an engineered living material.

16. Engineered living material produced using a genetically-modified cell produced and selected according to any of claims 1 to 14, or a selected droplet or capsule containing a genetically-modified cell according to any of claims 1 to 14.

Images