AUTOMATED CULTURE OF STEM CELLS

The present invention relates to methods for large-scale automated production of stem cells, including embryonic stem cells. The invention also relates to methods for large-scale automated production of differentiated cells derived from stem cells in culture. Also provided are populations of stem cells or differentiated cells and apparatus adapted for large-scale automated culture of stem cells. The invention further relates to the use of stem cells and the differentiated progeny thereof for the production of proteins or viruses.

The establishment and maintenance of stem cell cultures in vitro, including cultures of pluripotent stem cells such as embryonic stem (ES) cells, is well known. For example, the culture of pluripotent stem cell cultures in the presence of medium containing serum and Leukaemia Inhibitory Factor (LIF) is described in Smith et al. (1988) Nature 336: 688-90. Stem cell cultures can also be induced to differentiate in vitro and hence provide a source of differentiated cell types, including progenitor and stem cells, which are otherwise difficult to obtain. For example, mouse ES cell lines can be expanded in vitro and have the ability to give rise to cells from all three germ layers (Smith, 2001).

The increasing availability of stem cell lines, has led to an increased awareness of the value and potential of such cell lines for research and for potential therapeutic applications. For example, one area of interest is the drug discovery industry, which is currently hindered by a lack of functional cell culture models that could be used as screening tools for drug discovery and development (Pouton and Haynes, 2005). Hence, the drug discovery industry often uses cells with low clinical relevance and costly animal models. Stem cells, in particular mouse embryonic stem cells, have been in use for a number of years in drug discovery to develop genetically modified mice for target validation, target selectivity, model development and toxicity evaluation. With advances in stem cell technology it is becoming realistic that stem cells, including human stem

cells, can offer the drug discovery community physiologically relevant screening options, available in quantities unconstrained by the demands of high throughput screens (HTS) and showing normal growth and genetic structure (Gorba and Allsopp, 2003). Thus, stem cell lines and specific differentiated cell types can offer the drug discovery community an in vitro model that is more physiologically relevant than immortalised and recombinant cell lines and that avoids the expense and time of in vivo experiments and the development of animal models, and the inconsistency of primary cultures.

However, if stem cells are to be widely adopted as research tools it will be necessary to develop culture processes for stem cell growth and differentiation that reliably and reproducibly yield the quantity and quality of cells required (Pouton and Haynes, 2005). Such processes have not been provided to date. In the case of high throughput screening assays, current HTS technologies can screen millions of compounds against a single target in a timescale of a few weeks; therefore the number of cells required is in the order of billions of cells per day. Providing this number of cells by manual culture would be expensive in terms of time and personnel. Current approaches used for in vitro culture of stem cells, including ES cells, operate at a manual bench scale. Scale-up by the generation of multiple parallel manual processes is unattractive because of high labour cost and potential variability of output. Moreover, it is essential that the culture conditions do not result in spontaneous differentiation of the stem cells or any loss of differentiation potential.

Human embryonic stem cells and other human stem cells are also likely candidates for many future therapies. The tissue specific cell types derived from human ES cells may have therapeutic potential for the treatment of Parkinson's disease, spinal cord injury, heart disease, diabetes and other degenerative conditions (Li et al, 2004). It is thought that large numbers of functioning cells are needed for cell-based therapies, although the exact number will be dependent on therapeutic area. Again, processing these cells is important for the faithful

maintenance of their phenotype and generating these high quality cell populations in sufficient quantity is a bottleneck in the capability to generate such therapies (Vacanti et al, 2006).

Thus, there is a requirement for both reproducible expansion of stem cell populations and reproducible control of the phenotype of the cells within the expanded cultures. Current multiple parallel manual culture processes fail to satisfy these criteria.

A number of automated platforms capable of cell culture have been developed including SelecT, manufactured by the Automation Partnership. Other suitable devices are manufactured by Tecan and RTS Life Science International. However, such platforms have not been used to date for the successful production of stem cells, e.g. ES cells, in numbers suitable for screening for drug discovery or potential therapeutic applications.

Terstegge et al (2007) describe the expansion of ES cells using the Cell ^host system, based on a Hamilton Microlab STAR pipetting robot. However this is a relatively small-scale system based on culturing cells in 6-well plates, which only provide about 9.6 cm ² of culture area per well. This scale of culture is not suitable for providing cells in the numbers required for screening for drug discovery or therapy.

Thus, there remains a need for methods for large-scale production of stem cells, e.g. for use in HTS or therapy. Accordingly, an object of the present invention is to provide methods for large-scale automated production of stem cells or of differentiated cells derived from stem cells. Another object of the invention is the provision of populations of stem cells or differentiated cells in numbers suitable for HTS or therapeutic applications. A further aim of the invention is to provide apparatus adapted for such large-scale automated culture of stem cells.

Accordingly, a first aspect of the invention provides a method for large-scale automated production of stem cells comprising:

(a) providing a population of stem cells in a first culture vessel; (b) introducing the first culture vessel into a robotic cell culture apparatus; and

(c) culturing the stem cells under conditions conducive to stem cell proliferation.

The methods of the invention may be used to culture any type of stem cell that is of interest including mouse stem cells and human stem cells. Other mammalian stem cells can also be cultured according to the methods of the invention, including rat, American mink, hamster, pig, sheep, cow and primate stem cells.

It is intended, for the purposes of the present invention, that the term stem cell embraces any cell having the capacity for self-renewal and the potential to differentiate into one or more other cell types. Thus, the term stem cell includes pluripotent, multipotent or unipotent stem cells and progenitor cells from any tissue or stage of development. For example, the desired stem cells may be pluripotent stem cells, optionally ES cells, EC cells, or EG cells. In other embodiments, the stem cells can be haematopoietic stem cells, epidermal stem cells, mesenchymal stem cells, adipose tissue-derived stem cells, muscle stem cells or neural stem cells. In some embodiments the stem cells are modified for production of a protein or proteins, or virus particles. In some embodiments, the stem cells comprise an expression vector, which can be episomally maintained in the stem cells or integrated into the stem cell genome. In other embodiments, the stem cells comprise a virus.

The methods described herein are suitable for use with media containing serum and with serum-free media. However, it is preferred that the culture medium is serum-free. The replacement of serum or other incompletely defined or

undefined medium components with defined medium components can also result in greater reproducibility of the methods of the invention.

A number of suitable culture media, including serum-free media, are commercially available for the culture of stem cells and are suitable for use in the methods of the invention. Typically the medium used to expand populations of stem cells according to the methods of the invention will be the same medium used for conventional manual culture in plates or tissue culture flasks. However, in some embodiments a different medium may be used in the methods of the invention or the medium may be supplemented with additional components that promote proliferation or survival of the stem cells and/or prevent differentiation. For example, the medium may, in some embodiments, comprise a combination of a MEK inhibitor, a GSK3 inhibitor and, optionally, an antagonist of an FGF receptor. This combination is the subject of a co-pending patent application, US 11/694,351 , filed on 30 March 2007, the contents of which are incorporated herein by reference. In other embodiments, the medium may comprise a mixture of insulin and progesterone, and optionally other factors such as putrescine or sodium putrescine, sodium selenite, and transferrin or apotransferrin. These combinations are the subject of a co-pending patent application, US 11/825,068, filed on 3 July 2007, the contents of which are incorporated herein by reference.

Optionally, the stem cells contain a selectable marker gene operatively linked to a promoter that is preferentially or differentially expressed in the stem cells and cells other than the stem cells being cultured. The application of selection (e.g. use of an antibiotic when expression of the selectable marker confers antibiotic resistance) can be used to preferentially culture or purify the stem cells. Suitable methods are described in WO94/24274, the contents of which are incorporated herein by reference.

In some embodiments of the invention, step (c) includes feeding the stem cells. This may be by means of periodically replacing all or a portion of the culture

medium. For example, all or a proportion of the culture medium can be removed from the culture vessel by pipetting or by pouring used medium to waste and fresh medium can then be added. If medium is to be removed by pipetting, the culture vessel can be positioned to assist removal of the medium, e.g. by tilting the culture vessel.

It is envisaged that in some embodiments feeding of the cells can be achieved without removal of medium from the culture vessel, for example by adding a volume of fresh culture medium or by adding one or more medium components to the culture. The proportion of medium volume replaced or added will vary between different embodiments of the invention and may be 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80% or 90% of the culture volume. The frequency of medium replacement will also vary, and may be, for example, every 12, 24, 36 or 48 hours. The precise proportions and frequencies chosen in different embodiments will depend on the type of cells being cultured, the culture medium, the type of culture vessel, and other culture parameters, and can be readily determined by the user.

In preferred embodiments of the invention, step (c) comprises periodically passaging the stem cells into fresh culture vessels. Typically, for adherent stem cells, the passaging comprises (i) dissociating the stem cells in the first culture vessel to form a suspension; (ii) transferring the stem cells to at least two further culture vessels; and (iii) culturing the stem cells in the at least two further culture vessels under conditions conducive to stem cell proliferation. In other embodiments, e.g. when it is desired to maintain a population of cells rather than expanding the population, passaging can involve transfer of a proportion of the cells from the first culture vessel to a single further culture vessel.

The suspension may comprise single cells or small clusters of stem cells. In some embodiments, the clusters of stem cells will each contain from about 10 to about 20 or to about 30 cells. In other embodiments, particularly if automated

cell counting is to be used, the clusters of stem cells will each contain fewer than about 10 cells, and preferably at least 3 cells. For some stem cell types, it is preferred that the dissociated stem cells are not in single cell suspension as this can lead to reduced survival of the stem cells in subsequent culture. For some types of stem cell, including human ES cells, it is found that dissociation into single cells tends to lead to cell death. However, survival in subsequent culture is significantly improved if clumps of stem cells are used to inoculate the cultures.

Step (i) can be omitted in embodiments in which the stem cells grow in suspension, although, for some types of stem cells, it may be desirable to dissociate clusters of cells that may be present in the culture.

In embodiments in which passaging requires dissociation of the stem cells, the dissociation is conveniently carried out by adding a cell dissociation reagent to the first culture vessel. The cell dissociation reagent may be an enzymic cell dissociation reagent, such as trypsin-EDTA, or Accutase (Chemicon), or a non- enzymic cell dissociation reagent, such as Cell Dissociation Buffer (Invitrogen), Cellstripper (Mediatech, Inc). However, when culturing sensitive stem cells, such as ES cells, it is often desirable to avoid prolonged contact between the stem cells and an enzymic cell dissociation reagent. In embodiments in which serum containing media are used, the addition of fresh medium to the cell suspension following dissociation results in quenching of residual enzyme activity by the serum. However, this is not possible when serum-free media are used, and in such embodiments non-enzymic dissociation reagents are preferably used.

In manual culture methods, enzymic and non-enzymic dissociation reagents are typically removed by centrifuging before the cells are transferred to fresh culture vessels. However, in the automated methods of the invention it is preferred that the passaging does not comprise a centrifugation step. This is, in part, due to the difficulty and considerable expense of integrating a centrifuge into an automated

cell culture system. An advantage is to avoid exposing the stem cells to the shear forces that result from centrifugation.

A consequence of the absence of a centrifugation step in preferred embodiments of the invention is that it is not possible to completely remove the cell dissociation reagent from the stem cells before they are transferred to fresh culture vessels. This might be expected to have negative consequences for the resultant stem cell cultures as, for example, the presence of trace amounts of enzyme could, for example, disrupt cell adhesion or cause undesired differentiation of the stem cells. Surprisingly, however, the present inventors have found that stem cells are amenable to automated culture methods lacking a centrifugation step and have observed no adverse effects resulting from the carry-over of cell dissociation reagents into the subsequent cultures. Nevertheless, it is an option to take steps to minimise the carry-over of cell dissociation reagents, for example by diluting the cell dissociation reagent (e.g. 1:3 with medium or balanced salt solution) or by adding the cell dissociation reagent, manipulating the culture vessel to ensure an even film of reagent across the cell sheet, and removing the excess reagent prior to incubation.

Optionally, passaging the cells is set up so that a predetermined number of cells are transferred to each of the further culture vessels. In some embodiments of the invention, the passaging includes counting the cells transferred from the first culture vessel, preferably using an automated cell counting device forming part of the robotic cell culture apparatus. Following counting, the predetermined number of cells is transferred to each of the further culture vessels.

In other embodiments, the actual number of cells transferred from the first culture vessel is not counted. Rather, the number of cells is estimated based on the size of the culture vessel and the growth characteristics of the stem cells under the particular culture regime being used. Thus, the passaging can comprise calculating the number of cells transferred from the first culture vessel based on

one or more of (i) the initial number of stem cells in the first culture vessel, (ii) the population doubling time of the stem cells, (iii) the culture area of the first culture vessel, and (iv) the culture volume. It will be appreciated that the culture area of the first culture vessel is particularly relevant when calculating the number of adherent stem cells obtained after a given period of culture, whereas the culture volume will be particularly relevant when calculating the number of stem cells growing in suspension. For automated culture of adherent stem cells, it is preferred that the number of cells transferred to each of the further culture vessels is from about 5 x 10 ³ to 5 x 10 ⁴ cells/cm ² culture area. For example, when culturing stem cells in T175 tissue culture flasks, 1.0 x 10 ⁶ , 2.0 x 10 ⁶ , 4.0 x 10 ⁶ , 6.0 x 10 ⁶ or 8.0 x 10 ⁶ cells can be transferred to each flask. Particularly good results have been obtained when from 6.0 x 10 ⁶ to 8.0 x 10 ⁶ mouse ES cells have been seeded into each T175 flask. Good results have also been obtained when from 2 x 10 ⁶ to 5 x 10 ⁶ mouse neural stem (NS) cells have been seeded into each T175 flask (e.g. 3 x 10 ⁶ cells per T175 flask when using the 46cNS cell line), and when from 1 x 10 ⁶ to 2 x 10 ⁶ human multipotent adipose- derived stem (MADS) cells have been seeded into each T175 flask.

Alternatively, it is an option to passage the cells so that the cells from the first culture vessel are divided between a predetermined number of further culture vessels. For example, in preferred embodiments of the invention, stem cells are passaged with a split ratio of from 1 :2 to 1 :10, preferably from 1 :2 to 1 :5. It has been found, for example, that good results are achieved by passaging mouse ES cell cultures every 48 hours with a split ratio of 1 :5; that is, the stem cells from a first culture vessel are divided between five further culture vessels for continued culture. In some embodiments, the cells from the first culture vessel are divided between culture vessels of different formats (e.g. a combination of flasks and multi-well plates). Optionally a proportion of the cells from the first culture vessel are discarded. For example, a split ratio of 1 :5 can be used to seed less than five culture vessels, each containing one fifth of the cells from the first culture vessel, with the excess cell being discarded.

Good results have also been obtained for automated culture of mouse NS cells by passaging every 3-4 days using a split ratio of from 1 :5 to 1 :10 and for automated culture of human MADS by passaging every 4 days using a split ratio of 1:5 or 1:6.

Regardless of whether passaging is carried out based on a predetermined cell number or a predetermined split ratio, the production of stem cells according to the invention is most effective when the stem cells transferred into the further culture vessels are of high quality. It is preferred that the stem cells are not confluent and are in log phase growth at the time of passaging. The precise frequency of passaging will depend on the type of stem cell and the doubling time obtained in the particular culture conditions used. Typically, for mouse ES cells, setting up a culture 24 hours in advance of passaging will result in non-confluent cells in log phase growth.

Thus, in further embodiments of the invention, passaging is carried out when the stem cells in the first culture vessel reach a predetermined percentage confluence (or, in the case of suspension cells, a predetermined cell density). Typically, the passaging is carried out when the stem cells in the first culture vessel are 50 to 90% confluent, preferably 60 to 80% confluent, more preferably 65 to 75% confluent. In the automated methods of the invention, it is desirable, for the sake of increased efficiency, that the percentage confluence is calculated rather than being determined prior to each passage by the operator. Thus, in preferred embodiments, the percentage confluence is calculated based on one or more of (i) the number of stem cells initially present in the culture vessel, (ii) the population doubling time of the stem cells, (iii) the culture area of the first culture vessel, and (iv) the culture volume. In other embodiments, confluence is recorded and/or estimated automatically, for example using a Genetix CloneSelecT imager. In the case of suspension cultures, similar calculations can be performed to determine the cell density in a culture, so that the passaging is

carried out when the cell density is, for example, 50 to 90%, preferably 60 to 80%, more preferably 65 to 75% of the maximum cell density at which log phase cell growth can be supported.

Preferably the passaging is repeated until either a predetermined number of culture vessels containing stem cells or a predetermined number of stem cells has been produced. In some embodiments, the point at which the predetermined number of stem cells has been produced will be estimated based on the growth characteristics of the stem cells and the previous process steps (e.g. the number of passages). It is also possible to calculate the number of stem cells obtained by calculating the yield per culture vessel and multiplying this value by the number of culture vessels containing stem cells that have been produced. By way of example, the methods of the invention can be used to produce in excess of about 10 ⁸ cells, in excess of about 10 ⁹ cells or even in excess of about 10 ¹⁰ cells. Thus, the methods of the invention provide automated production of stem cells on a significantly larger scale than has been possible previously.

Thus, the use of an automated cell culture system advantageously provides increased processing capacity as it can be set up to run 24 hours a day, 7 days a week if necessary. It not only allows greater production capacity, but also allows cells to be passaged and plated at the optimum time point, even if that falls outwith normal working hours. Based on normal operator hours and single shift facility operation, the experimental data presented in the specific examples indicate that it would be feasible to produce in excess of 2,000 96 well plates of mES cells per week. This would be sufficient to support the weekly demands of a typical high throughput screen (192,000 wells per week). In order to achieve the same output using manual culture techniques would require a minimum of eight staff working fulltime, even with the aid of multi-channel plating equipment. There would also be health and safety implications from staff carrying out such repetitive movements for the duration of an entire shift.

In addition to the increased capacity, one of the main advantages of automated culture is the improved reproducibility of culture conditions. An automated system is programmed to perform the same manoeuvres exactly the same way each time. This gives much improved reproducibility than a group of individuals who, although they follow the same standard operating procedure (SOP), will necessarily introduce variation in their technique. Automated culture according to the invention therefore provides better quality cultures, particularly if the automated platform is operated in conjunction with cell culture experts to ensure that the culture methodology developed produces cells of the highest possible quality.

When the passaging is repeated, the initial passaging and subsequent passaging may be carried out according to different embodiments as described above. For example, the stem cells can be counted during the first passaging step, and the number of cells obtained from the first culture vessel relative to the number of cells originally seeded into that vessel can be used to calculate the number of cells transferred to further culture vessels or the preferred split ratio used for subsequent passaging steps. Thus, in certain embodiments, the number of stem cells will be counted once and subsequent culture steps, including passaging protocols, will be carried out based on calculated values for, e.g. cell numbers, percentage confluence or cell density.

In some embodiments, the methods of the invention further comprise the step (d) harvesting the stem cells. The harvesting may comprise dissociating the stem cells, e.g. as described above in relation to passaging the stem cells. Thus, the dissociating may involve adding a cell dissociation reagent to the culture vessel or vessels, preferably a non-enzymic cell dissociation reagent.

Optionally, the harvesting further comprises dispensing the cells into one or more vessels according to the intended downstream use of the stem cells. The vessels may be any suitable vessels, including tissue culture flasks, spinner

flasks, bioreactors, culture dishes, multi-well plates, vials and tubes. For some downstream applications, it may be desirable to provide an extract of the stem cells, and hence some embodiments of the method include the further step of lysing the harvested stem cells.

It is generally preferred that, in the methods of the invention, the stem cells are cultured in the absence of feeder cells. This is particularly convenient in embodiments in which the actual cell count is recorded when the cells are passaged. However, the methods of the invention are suitable for production of feeder-dependent stem cells as well as feeder-independent/feeder-free cultures. This is of particular value for expanding feeder-dependent human ES cell lines, although other feeder-dependent stem cells can be cultured according to the invention. If feeder cells are to be used, the feeder cells are typically introduced into the culture vessels before the stem cells are introduced. Preferably, the feeder cells are permitted to proliferate until confluence prior to the introduction of the stem cells. It is also preferred that the feeder cells are inactivated prior to the introduction of stem cells, for example using known protocols for γ-irradiation or mitomycin c treatment.

It will be appreciated that the methods of the invention can be adapted for use with any type of culture vessel, including tissue culture flasks, dishes and multi- well plates. However, it is convenient to use flasks when producing large numbers of stem cells, as this advantageously reduces the number of processing steps required to obtain a given number of cells and thus reduces the potential for cell damage during handling. For example, if stem cells are carried out in 6- well plates, each well provides approximately 9.6 cm ² culture area, compared to 75 cm ² or 175 cm ² for T75 or T175 tissue culture flasks. Thus, large numbers of stem cells can be produced without excessive processing (e.g. large numbers of passages) using the methods of the invention. Stem cell production on such a scale has not previously been described and the methods of the invention permit production of stem cells in numbers far exceeding those previously available.

Optionally, the methods include coating the culture vessels with a substrate, e.g. gelatin, prior to the introduction of culture media or cells. Conveniently this step is automated.

A further advantage of automated production of stem cells using the methods of the invention is that all stages of the process can be tracked. For example, one or more culture vessels can be tracked, optionally using one or more bar code readers. The tracking can include recording details of each processing step carried out on a particular culture (including cell counts and percentage confluence values at one or more time points), the number of passages, and details of the media used including batch numbers of media and medium components. Thus, the methods of the invention can provide fully traceable populations of stem cells, with built in assurance of quality.

The methods of the invention can also be adapted to provide differentiated cells derived from stem cells. Thus, a second aspect of the invention provides a method for large-scale production of desired differentiated cells derived from stem cells comprising:

(a) providing a population of stem cells in a first culture vessel;

(b) introducing the first culture vessel into a robotic cell culture apparatus; and

(c) culturing the stem cells under conditions conducive to the differentiation of the stem cells into the desired differentiated cells.

The methods of the second aspect may be carried out in essentially the same way as those of the first aspect described above, other than that the media used in step (c) support the differentiation of stem cells into cells of a particular lineage or lineages. For example, the differentiated cells can be somatic stem cells, haematopoietic stem cells, epidermal stem cells, mesenchymal stem cells,

adipose tissue-derived stem cells, muscle stem cells or neural stem cells. Differentiated cells derived from somatic stem cells, haematopoietic stem cells, epidermal stem cells, mesenchymal stem cells, adipose tissue-derived stem cells, muscle stem cells or neural stem cells can also be obtained according to the invention. In some embodiments, the differentiated cells are a mixed population of cells belonging to one or more desired lineages. In preferred embodiments of the invention, the differentiated cells are neural cells. As a specific example, they are neurons. Good results have also been obtained for the production of adipocytes and osteoblasts by automated differentiation of human multipotent adipose-derived stem (MADS) cells.

The stem cells or the differentiated cells can be modified for production of a protein or proteins, or virus particles. In some embodiments, the stem cells or the differentiated cells comprise an expression vector, which can be episomally maintained or integrated into the genome of the stem cells or differentiated cells. In other embodiments, the stem cells or the differentiated cells comprise a virus. In general, the stem cells will be modified, and subsequently differentiated using the methods described herein. However, in some embodiments, the stem cells are differentiated and the differentiated cells are then modified and, optionally, further expanded of differentiated, for example using the methods of the invention.

A number of media and supplements suitable for supporting and/or inducing differentiation of stem cells down one or more chosen lineages are available to the skilled person and appropriate media can be selected without undue burden. Optionally, the stem cells can contain a selectable marker gene which is preferentially or differentially expressed in cells of the chosen lineage or lineages and cells of other, undesired, lineages, and selection can be used to preferentially culture, purify or isolate cells of the chosen lineage or lineages. Suitable methods for achieving such selection are described in WO94/24247 and WO99/53022, the contents of which are incorporated herein by reference.

In some embodiments, step (c) comprises (i) culturing the stem cells under conditions conducive to the proliferation of stem cells, and (ii) culturing the stem cells under conditions conducive to the differentiation of the stem cells into the desired differentiated cells. In other embodiments, step (c) comprises (i) culturing the stem cells under conditions conducive to the differentiation of the stem cells into the desired differentiated cells, and (ii) culturing the resultant differentiated cells under conditions conducive to the proliferation of the differentiated cells.

It will be appreciated that methods of the invention can also be adapted to produce proteins or viruses in stem cells or the differentiated progeny of stem cells. For example, stem cells and their differentiated progeny can be used to produce proteins of therapeutic interest or vaccine components. The use of stem cells and their progeny as a base line for protein or vaccine production is particularly attractive due to their capacity for physiologically appropriate protein folding, assembly and post-translational modification. For example, stem cells can be used to produce human proteins of therapeutic value, which are correctly formed, assembled and modified. The cells can be used to assemble multi- subunit protein complexes or the subunits can be produced separately and subsequently combined. In addition, such cells are generally stable over long term culture, are genotypically normal and capable of differentiation into biologically relevant cell types at scale. The normal genotype of stem cells and their progeny provides an advantage over immortalised cell lines, which may exhibit abnormal physiology as a result of mutations acquired in culture and, in some cases, do not retain introduced genetic constructs.

The production of proteins and viruses in stem cells or their progeny is also useful for drug discovery applications. For example, a recombinant protein of interest, preferably a human protein such as a G protein-coupled receptor, can be expressed (stably or transiently) in stem cells or their progeny and the cells

can subsequently be used in an assay. According to the assay, the protein can be retained within the cells, presented on the cell surface or secreted.

Accordingly, a further aspect of the invention provides a method for producing a protein in stem cells, comprising:

a) introducing a DNA molecule encoding the protein into a stem cell to obtain a modified stem cell; b) culturing said modified stem cell under conditions conducive to the proliferation of stem cells, thereby obtaining a population of modified stem cells; c) further expanding the population of modified stem cells, preferably using an automated process; and d) extracting or isolating the protein from the cells or the culture medium.

The DNA will comprise all or part of one or more genes that can encode any protein or proteins of interest, including proteins of therapeutic interest. The DNA also suitably comprises regulatory elements (e.g. promoters), such as are well known in the field of recombinant DNA technology, to enable the introduced gene or genes to be expressed. If more than one gene is to be expressed, it may be desirable to include an internal ribosome entry site (IRES). For example, it may be desirable to use a construct containing an IRES to express a protein of interest and a marker gene. Suitable technology, methods and constructs are described in WO94/24301 , the contents of which are incorporated herein by reference.

The stem cells can be expanded using any suitable method, including the methods of the invention. Step b) can also be carried out using an automated process, e.g. by using the methods of the invention.

In a related aspect, the invention provides a method for producing a protein in differentiated cells derived from stem cells, comprising:

a) introducing a DNA molecule encoding the protein into a stem cell to obtain a modified stem cell; b) culturing said modified stem cell under conditions conducive to the proliferation of stem cells, thereby obtaining a population of modified stem cells; c) inducing differentiation of the population of modified stem cells, preferably using an automated process, to obtain a population of desired differentiated cells; and d) extracting or isolating the protein from the cells or the culture medium.

It will be appreciated that the differentiation of the modified stem cells can be induced using, for example, the methods for large-scale production of differentiated cells described herein. In addition, step b) can optionally be carried out using an automated process, e.g. by using the methods of the invention.

The DNA molecule can be introduced into the stem cells or differentiated cells using any one of the numerous techniques known to the skilled person, including electroporation, lipofection, nucleofection, calcium phosphate/calcium chloride and viral transduction. For example, neural stem (NS) cells can conveniently be transfected using nucleofection technology (Amaxa). In some embodiments, the

DNA will be introduced in the form of an expression vector (e.g. a plasmid or viral vector) and hence the stem cells or the differentiated cells comprise an expression vector.

The transfection can be transient, whereby the introduced DNA may be lost when the cells undergo mitosis, or it may be stable, so that the introduced DNA remains in the transfected cell and its daughter cells. Selection, e.g. using an antibiotic or a marker allowing physical separation of cells containing the

introduced DNA (e.g. by FACS), may be used to ensure selective maintenance of cells containing the introduced DNA. Many well-published strategies for achieving such selection are available to the skilled person and suitable for use according to the invention.

The introduced DNA may be maintained episomally or may be integrated into the genome of the stem cells or differentiated cells. For episomal expression, it is preferred in some embodiments that the cell contains a first vector which provides a replication factor, and a second vector for expression of the DNA of interest the replication of the second vector being dependent on the presence of the replication factor. Suitable vectors and methods are described in WO98/32868, the contents of which are incorporated herein by reference.

In embodiments in which the introduced DNA is integrated into the host cell genome, the integration may be random or may be targeted using homologous recombination. Suitable methods are well known to the skilled person. For example, if the DNA is to be introduced into mouse stem cells, the construct can be targeted to integrate in the rosa26 locus, which is constitutively active in all cell types. Alternatively, sequence elements can be included in the construct to prevent silencing of the integrated DNA.

Following introduction of the DNA into the stem cells, the transfected cells are cultured, optionally under selection to isolate cells containing the DNA of interest. If e.g. antibiotic selection is used, colonies of positive cells can be picked manually or using an automated clone picking device and optionally expanded using conventional manual tissue culture methods prior to expansion or differentiation using automated methods, e.g. the methods of the invention.

In preferred embodiments, the protein is secreted into the culture medium, thus allowing convenient harvesting of the protein from the culture medium. Typically, the introduced DNA will encode a signal peptide, which directs translocation of

the expressed protein to the cell membrane and secretion into the medium. Suitable sequences are well known to the skilled person. The protein of interest can be harvested from the medium during the culture process or taken from the culture vessel following completion of the culture process. In some embodiments, no further purification is required, for example if the expressed protein is a growth factor to be used as conditioned media. Alternatively, the protein can be purified using standard techniques appropriate to the intended use. In other embodiments, the protein is not secreted and is harvested from the cells, e.g. as a cell lysate. In embodiments in which the protein is intended for therapeutic use, the cells are preferably cultured in serum-free media, more preferably media free of all animal components.

In some embodiments in which the protein is to be produced in differentiated cells, the population of modified stem cells of step (b) is divided into two or more sub-populations of stem cells, and steps (c) and (d) are carried out in parallel on each sub-population. For example, each sub-population may be induced to differentiate into cells of a different lineage or type. In some such embodiments, the proteins produced by the different cell types will differ due to e.g. alternative splicing, different post-translational modifications or incorporation into complexes comprising different components. Thus, stem cells and their differentiated progeny provide a useful resource for obtaining different, physiologically relevant, cell-type specific protein variants.

The invention also provides a protein obtainable by the methods of the invention. In some embodiments, the protein is assembled into a multi-subunit complex, either in the stem cells or differentiated cells or subsequently, following harvesting of the protein. In other embodiments, the protein is a single subunit of a protein complex. If the protein is to be used for therapeutic purposes, it is optionally combined with one or more excipients or carriers.

A further aspect of the invention provides a method for producing a virus in stem cells, comprising:

a) introducing a DNA molecule encoding the virus into a stem cell or infecting a stem cell with a virus to obtain a modified stem cell; b) culturing said stem cell under conditions conducive to the proliferation of stem cells, thereby obtaining a population of modified stem cells; c) further expanding the population of modified stem cells, preferably using an automated process; and d) extracting or isolating the virus from the cells or the culture medium.

A related aspect provides a method for producing a virus in differentiated cells derived from stem cells, comprising:

a) introducing a DNA molecule encoding the virus into a stem cell or infecting the stem cell with a virus to obtain a modified stem cell; b) culturing said stem cell under conditions conducive to the proliferation of stem cells, thereby obtaining a population of modified stem cells; c) inducing differentiation of the population of modified stem cells, preferably using an automated process, to obtain a population of desired differentiated cells; and d) extracting or isolating the virus from the cells or the culture medium.

The expansion or differentiation of the modified stem cells can be carried out using the methods described herein, for example the culture methods used for protein production in stem cells and their differentiated progeny.

The methods of the invention are particularly suitable for the production of viruses that infect specific cell types, e.g. neurotropic viruses. For example, modified stem cells can provide a source of virus, which when differentiated into neural cells permits amplification of the virus by means of infection and further

amplification of the virus in other neural cells present in the culture. The viruses may be recombinant viruses or modified viruses, which have been modified to provide improved operator safety (e.g. for human viruses) or to provide alternative and/or improved vaccine components.

The viruses obtained using the methods of the invention can be combined with one or more excipients or carriers, for use e.g. as vaccines, vaccine components or candidate vaccines.

The proteins or viruses produced according to the methods of the invention can be harvested in a variety of ways. For example, a batch process can be used in which the medium and/or cell population is harvested at the end of the culture process. It will be appreciated that in some such embodiments, the cells are passaged and/or fed by adding additional medium or other components to the culture vessel or vessels during the process. Alternatively, media can be continually (or periodically) added and media and/or cells can be continually (or periodically) harvested throughout the culture process. For example, when the invention is used to produce a secreted protein or virus particles, all or a proportion of the culture medium can be harvested and replaced with fresh medium without transferring the cells to fresh culture vessels. In other embodiments, cells or media can be harvested when the cells are passaged. For example, a proportion of the cells and/or medium is used to seed further cultures and a proportion of the cells and/or medium is harvested. If required, the protein or virus can then be extracted and/or purified from the cells and/or medium using a number of well-known protocols.

In a further aspect, the invention provides use of a robotic cell culture apparatus for large-scale culture of stem cells in the absence of feeder cells. Preferably, the apparatus comprises automated means for one or more of: inoculating cells into a culture; changing or adding culture medium to a culture; adding further components to a culture; and harvesting cells from a culture. It is envisaged that

the aforementioned means will include at least a robotic arm (or equivalent means, such as a track system) for manipulation of culture vessels and a pipetting station for dispensing cell suspensions, media and/or medium components.

Preferably the apparatus further comprises means for providing conditions conducive to stem cell proliferation and/or differentiation. For example, the apparatus may include incubators for any culture vessel format described herein, typically including at least one of an incubator for flasks and an incubator for tissue culture plates. In use, the apparatus will typically provide control of one or more of the temperature, the CO2 level, the O ₂ level and the relative humidity at which the stem cells are cultured. For example an automated flask incubator will typically provide precise control of the incubation temperature and CO ₂ level and an automated plate incubator will additionally provide relative humidity control so as to minimise medium loss via evaporation. Typically, the temperature will be maintained at about 37°C ± 0.5 ⁰ C, the CO2 level will be maintained at about 5- 7% and the relative humidity will be maintained at about 95% or above. For some cell types it may be desirable to maintain the O ₂ level below ambient. The apparatus will also provide aseptic conditions to prevent contamination of cultures and ensure operator safety, suitably using a negative pressure laminar airflow hood. Optionally, the environment in such a laminar flow hood is controlled for one or more of the temperature, the CO ₂ level, the O ₂ level and the relative humidity.

In preferred embodiments the apparatus also comprises means for automated cell counting to provide consistent and accurate cell densities when seeding new culture vessels. Means for automated determination and/or estimation of percentage confluence (e.g. the Genetix CloneSelecT imager) can also be included.

The apparatus can also comprise imaging equipment or other detection means. Such means can, for example, be used to detect the expression of fluorescent reporter genes (e.g. GFP) in the cells being cultured. For example, the cells may express a reporter gene, optionally provided by means of a construct comprising an internal ribosome entry site (IRES) as described in WO94/24301 , the contents of which are incorporated herein by reference. The percentage of reporter- positive cells can be used to determine when to passage or induce differentiation of the stem cells in a culture. Imaging equipment can also be used to assess when to harvest cells. In the case of differentiated cells, the readiness of cells for harvest can be assessed by reference to predetermined criteria, e.g. the number of processes in neurons or the number of fat globules in adipocytes.

The apparatus can also comprise means for tracking culture vessels and/or media, thus providing traceability and audit of cultures and media. Suitably, the means for tracking culture vessels and/or media comprise one or more bar code readers. These readers will typically be connected to a computer system providing an interface with the user. Thus, the user can access the full audit trail.

In preferred embodiments, the apparatus comprises means for handling two or more culture vessel formats, which are, for example, selected from tissue culture flasks, multi-well plates, culture dishes, vials, and tubes. This advantageously allows the efficient expansion of stem cells to large numbers, e.g. in flasks, and subsequent dispensing of the stem cells into alternative vessel formats for downstream use. It may be desirable, for example, to use the apparatus to obtain large numbers of cells, suitably using the methods of the first or second aspects of the invention, and then dispense the cells into multi-well plates ready for use in drug screening assays. In other embodiments, it may be desirable to obtain large numbers of stem cells and then expose sub-populations to different differentiation regimes, for example using the methods of the second aspect of the invention.

The use of this aspect of the invention is suitable for the culture of any stem cells described herein, including culturing stem cells so as to obtain differentiated progeny according to the second aspect of the invention. Thus the stem cells can be pluripotent stem cells, suitably ES cells, haematopoietic stem cells, epidermal stem cells, mesenchymal stem cells, adipose tissue-derived stem cells, muscle stem cells or neural stem cells. The stem cells can be modified for production of a protein or proteins, or virus particles. In some embodiments of this aspect of the invention, the stem cells comprise an expression vector, which can be episomally maintained or integrated into the genome of the stem cells. In other embodiments, the stem cells comprise a virus.

It will be appreciated that the present invention also extends to populations of cells obtained using the methods of the invention. Accordingly, a further aspect of the invention provides a population of stem cells or the differentiated progeny thereof obtainable by any of the methods described herein. Thus, the population may, in some embodiments, comprise pluripotent stem cells, optionally ES cells; haematopoietic stem cells; epidermal stem cells; mesenchymal stem cells; adipose tissue-derived stem cells; muscle stem cells or neural stem cells. The population may also comprise differentiated cells derived from such stem cells. In one preferred embodiment, the differentiated progeny are neural cells.

The populations of cells provided by the invention are not limited in size, but preferably comprise cells in sufficient numbers to be used e.g. in HTS assays or in therapy. Thus, it is preferred that the population comprises in excess of about

8 9 10 cells, more preferably in excess of about 10 cells, and most preferably in

10 excess of about 10 cells.

It will be apparent from the foregoing description that the invention provides methods for automated production of stem cells in hitherto unobtainable numbers. Thus, further aspects of the invention provide a population of stem

8 cells comprising in excess of about 10 cells; a population of stem cells

g comprising in excess of about 10 cells; and a population of stem cells

10 comprising in excess of about 10 cells. Such populations of stem cells may comprise any type of stem cell, including those described herein. Thus, in preferred embodiments of the invention, the population comprises pluripotent stem cells, optionally ES cells; haematopoietic stem cells; epidermal stem cells; mesenchymal stem cells; adipose tissue-derived stem cells; muscle stem cells or neural stem cells. The populations of cells can comprise stem cells or the differentiated cells which are modified for production of a protein or proteins, or virus particles. In some embodiments, the stem cells or the differentiated cells comprise an expression vector, which can be episomally maintained or integrated into the genome of the stem cells or differentiated cells. In other embodiments, the stem cells or the differentiated cells comprise a virus.

In a further aspect, the invention provides apparatus adapted or arranged for carrying out the methods of the invention. Thus, the invention provides an apparatus for large-scale automated production of stem cells comprising: a) robotic means for handling culture vessels; b) means for inoculating stem cells into a culture; c) means for changing or adding medium to a culture; and d) programmable control means; wherein the apparatus is adapted to passage the stem cells when they reach a predetermined percentage confluence.

As has been discussed above, the invention seeks to provide stem cells of the highest possible quality, and it is therefore important that the stem cells are in log phase growth when they are passaged. That is, the stem cells are passaged before they reach confluence or, in the case of suspension cultures, the maximum population density at which log phase growth can be maintained.

Thus, in preferred embodiments, the apparatus is adapted to passage the stem cells when they are 50 to 90% confluent, more preferably when they are 60 to

80% confluent, and most preferably when they are 65 to 75% confluent.

The percentage confluence may be estimated or calculated as described above, thus avoiding the necessity for direct operator monitoring of every passaging operation. Thus, in preferred embodiments, the apparatus is programmed to calculate the percentage confluence based on one or more of (i) the number of stem cells initially present in the culture vessel, (ii) the population doubling time of the stem cells, (iii) the culture area of the culture vessel, and (iv) the culture volume.

The apparatus can optionally be programmed to repeat the passaging until a predetermined number of culture vessels containing stem cells has been produced. Alternatively, the apparatus can be programmed to repeat the passaging until a predetermined number of stem cells has been produced. In preferred embodiments the predetermined number of stem cells is in excess of

8 9 about 10 cells, more preferably in excess of about 10 cells, and most preferably

10 in excess of about 10 cells.

As discussed above, the cessation of further passaging may be triggered by the attainment or exceeding of a calculated value for the number of stem cells produced, rather than by physically counting the cells, for example based on the calculated yield of stem cells per culture vessel and the number of culture vessels containing stem cells that have been produced. Thus, the apparatus can be programmed to calculate the number of stem cells obtained based on one or more of (i) the number of stem cells initially present in the first culture vessel, (ii) the population doubling time of the stem cells, (iii) the culture area of the culture vessels, (iv) the culture volume in each culture vessel, and (v) the number of passages.

In other embodiments, the apparatus can be programmed to cease passaging when a predetermined number of passages has been carried out.

In a further aspect, the invention provides an apparatus for large-scale automated production of desired differentiated cells derived from stem cells comprising: a) robotic means for handling culture vessels; b) means for inoculating cells into a culture; c) means for changing or adding medium to a culture; and d) programmable control means; wherein the apparatus is adapted to passage the stem cells or the differentiated progeny thereof when they reach a predetermined percentage confluence or a predetermined number of cells.

Preferably, the apparatus is adapted to passage the stem cells or the differentiated progeny thereof when they are 50 to 90% confluent. In more preferred embodiments, the apparatus is adapted to passage the stem cells or the differentiated progeny thereof when they are 60 to 80% confluent or when they are 65 to 75% confluent.

Again, the apparatus can be programmed to calculate the percentage confluence based on one or more of (i) the number of stem cells initially present in the culture vessel, (ii) the population doubling time of the stem cells and/or the differentiated progeny thereof, (iii) the culture area of the culture vessel, and (iv) the culture volume.

In a similar way to the apparatus for production of stem cells, the apparatus for production of differentiated cells can be programmed to repeat the passaging until a predetermined number of culture vessels containing the desired differentiated cells has been produced or, alternatively, until a predetermined number of desired differentiated cells has been produced. Preferably, the

8 predetermined number of desired differentiated cells is in excess of about 10 g cells, more preferably in excess of about 10 cells, and most preferably in excess

10 of about 10 cells.

Once again, the number of cells obtained can be determined by calculation as well as by physical counting of the cells. Thus, in a preferred embodiment, the apparatus is programmed to calculate the number of desired differentiated cells obtained based on one or more of (i) the number of stem cells initially present in the first culture vessel, (ii) the population doubling time of the stem cells and/or the differentiated progeny thereof, (iii) the culture area of the culture vessels, (iv) the culture volume in each culture vessel, and (v) the number of passages.

The apparatus of the invention can be adapted to culture stem cells or the differentiated progeny thereof in serum-free media as well as in serum containing media, suitably by using the methods described herein. Similarly, the apparatus can be adapted to culture stem cells or the differentiated progeny thereof either in the presence or in the absence of feeder cells. Preferably, the apparatus is adapted to culture stem cells or the differentiated progeny thereof in the absence of feeder cells.

The apparatus may be selected from any of a number of automated platforms for cell culture that are currently available and adapted for large-scale production of stem cells or differentiated cells derived from stem cells. The present inventors have obtained good results using the CompacT SelecT platform, manufactured by the Automation Partnership, but it will be understood that other systems can be adapted to provide apparatus according to the invention, which can be used to carry out the methods of the invention.

Generally, the apparatus will include robotic means for handling and/or manipulating culture vessels. This can conveniently be provided by a robotic arm having the capacity to import and export culture vessels, to move or alter the orientation of culture vessels within the apparatus to facilitate the carrying out of necessary processing steps, and/or to open and close culture vessels. The apparatus will also include means for inoculating cells into a culture and means

for changing or adding medium to a culture. Such means are conveniently provided using an automated pipetting station, preferably using disposable pipettes, and, optionally, additional liquid pumps, thus permitting programmable medium selection and additive dispensing of different media and or reagents without risk of cross-contamination. Thus, the apparatus may further comprise means for adding further components to a culture. In some embodiments, separate systems will be provided for adding or removing media, reagents and/or cells to and from culture vessels of different types. For example, the apparatus may comprise separate dispensing stations for tissue culture flasks and multi-well plates. Additional means may also be supplied, e.g. for adding growth factors or cell dissociation reagents.

The apparatus will also comprise programmable control means, typically in the form of a computer system which controls the operation of the apparatus. The control means provide an interface with the user, which permits adaptation of culture protocols to the cells being cultured and scheduling of tasks to be carried out. Thus, it is envisaged that the user will be able to set the number of cells, or the volume of a cell suspension to be added to each culture vessel, the type of culture vessel to be used, the type and volume of medium to be added (including the type and volume of any additional supplements), the incubation time between passages, the volume of cell dissociation reagent (if any) to be used when passaging, the time for which the cells are incubated in the presence of the dissociation reagent, the desired split ratio when passaging, and/or the criteria for terminating a particular culture process (e.g. reaching a predetermined cell count or number of culture vessels).

Preferably the apparatus will also comprise means for harvesting cells or medium (e.g. medium comprising secreted proteins or viruses produced in stem cells or their differentiated progeny) from a culture. This may be achieved using elements previously described, such as a robotic arm and pipetting and/or medium dispensing stations. In some embodiments, the apparatus may

comprise additional elements, for example for dispensing cells into vessels suitable for their intended downstream use. In some embodiments, the apparatus is adapted to dispense the harvested cells into multi-well plates ready for use in drug screening assays.

Preferably, the apparatus will comprise means for providing conditions conducive to cell proliferation and/or differentiation. Typically, this will be provided in the form of one or more incubators connected to or integrated with the apparatus, and in which the cells are maintained between processing steps. For example, the apparatus may include incubators for any culture vessel format described herein, typically including at least one of an incubator for flasks and an incubator for tissue culture plates. The apparatus will typically provide control of one or more of the temperature, the CO ₂ level, the O ₂ level and the relative humidity at which the stem cells are cultured. For example an automated flask incubator will typically provide precise control of the incubation temperature and CO ₂ level and an automated plate incubator will additionally provide relative humidity control so as to minimise medium loss via evaporation.

In use, the incubators will typically be maintained at a temperature of 37 ⁰ C ± 0.5 ⁰ C, the CO ₂ level will be maintained at about 5-7%, for example by supplying the incubators with air enriched with CO ₂ , and the relative humidity will be maintained at or above about 95%. In some embodiments, it may be desirable to maintain the O ₂ level below ambient. However, the operator will be able to determine whether alternative culture conditions are more suitable for the production of a particular stem cell or differentiated cell type.

The apparatus will also provide aseptic conditions to prevent contamination of cultures and ensure operator safety, suitably using a negative pressure laminar airflow hood, preferably fitted with a HEPA filter. Optionally the environment within such a laminar flow hood is controlled for one or more of the temperature, the CO ₂ level, the O ₂ level and the relative humidity.

In preferred embodiments the apparatus also comprises means for automated cell counting to provide consistent and accurate cell densities when seeding new culture vessels. The cell counting means may additionally have the capacity to assess cell viability, for example by using Trypan blue exclusion. It is envisaged that the cell counting apparatus will supply information to the control means to enable calculation of any of the parameters discussed herein (e.g. the percentage confluence after a given incubation time, the number of cells to be transferred to each further culture vessel when passaging, and the total number of cells produced). Means for automated determination and/or estimation of percentage confluence (e.g. the Genetix CloneSelecT imager) can also be included.

The apparatus can also comprise means for tracking culture vessels and/or media, thus providing traceability and audit of cultures and media. Suitably, the means for tracking culture vessels and/or media comprise one or more bar code readers. These readers will typically be connected to a computer system providing an interface with the user. Preferably the tracking means will record details of all processing steps carried out on a particular culture vessel, including the history of the cells therein, thus providing the user with access to the full audit trail.

In preferred embodiments, the apparatus comprises means for handling two or more culture vessel formats, which are, for example, are selected from tissue culture flasks, multi-well plates, culture dishes, vials, and tubes. This advantageously allows the efficient expansion of stem cells to large numbers, e.g. in flasks, and subsequent dispensing of the stem cells into alternative vessel formats for downstream use. Suitable vessel formats for downstream use include spinner flasks and bioreactors. It may be desirable, for example, to use the apparatus to obtain large numbers of cells, suitably using the methods of the

first or second aspects of the invention, and then dispense the cells into multi- well plates ready for use in HTS assays.

It will be appreciated that the apparatus of the invention can be used to for large- scale production of any stem cells or differentiated cells obtained from stem cells, including the specific cell types described herein. Thus, the apparatus can be used for the automated culture of pluripotent stem cells, optionally embryonic stem (ES) cells and of haematopoietic stem cells, epidermal stem cells, mesenchymal stem cells, adipose tissue-derived stem cells, muscle stem cells or neural stem cells. When the apparatus is used to obtain differentiated progeny of stem cells, the differentiated cells can be haematopoietic stem cells; epidermal stem cells; mesenchymal stem cells; adipose tissue-derived stem cells; muscle stem cells or neural stem cells. The differentiated cells can also be the progeny of such stem cells. In preferred embodiments, the differentiated progeny are neural cells.

It will also be appreciated that the apparatus can be used to produce stem cells or differentiated cells that are modified for production of a protein or proteins, or virus particles, as described herein. In some embodiments, the stem cells or the differentiated cells comprise an expression vector, which can be episomally maintained or integrated into the genome of the stem cells or differentiated cells. In other embodiments, the stem cells or the differentiated cells comprise a virus.

Further aspects of the invention relate to automated assays, e.g. for drug screening. Ideally, automated production of stem cells or differentiated cells according to the present invention is combined with the automated assay procedure. However, in some embodiments the expansion of stem cells or differentiated stem cells and the subsequent carrying out of the assay may be physically or temporally separated.

Accordingly, another aspect of the invention provides an automated assay

comprising:

(a) providing a population of stem cells or differentiated cells derived from stem cells;

(b) determining the effect of one or more test compounds on an activity of the cells.

Typically the initial population of stem cells or differentiated cells will be produced using the automated methods and/or the apparatus described herein. This is particularly advantageous when multiple test compounds are to be tested, as the automated processes of the invention are able to provide large numbers of cells of consistently high quality. In such embodiments, step (a) will typically comprise dividing the population of cells into a plurality of sub-populations. This is conveniently achieved by dispensing each sub-population into a separate well of one or more multi-well plates, although other culture vessels could also be used. Each of the plurality of sub-populations can then be contacted with a different one or more test compounds. Preferably, at least one of the sub-populations is contacted with a control or reference compound. The nature of the control or reference compound will depend on the assay, but may include a positive control compound of known activity or a compound that facilitates quantitation of the assay results.

The test compounds will typically be candidate drug compounds with potential to alter, e.g. increase or decrease, an activity or function of the cells. The test compounds may, for example, be small molecules, siRNA (small interfering RNA), shRNA (short hairpin RNA), proteins, peptides, or antibodies, including antigen binding fragments such as Fab fragments, F(ab')2 fragments. In some embodiments, the test compounds are candidate activators or inhibitors of a functional protein expressed by the cells, e.g. an enzyme.

In some embodiments, the assay is carried out on differentiated cells obtained from the initial population of cells. Accordingly, step (a) can further comprise

inducing differentiation of the plurality of sub-populations to provide differentiated cells. This is of particular application when the assay is to be carried out on terminally differentiated cells such as neurons as, for example, neural stem cells can readily be dispensed into the requisite number of culture vessels or wells of multi-well plates and differentiation can then be induced, providing almost unlimited numbers of subpopulations of neurons for use in the assay. Differentiation of the sub-populations may also be appropriate if the assay is to be carried out on an adherent cell type. In such circumstances, higher quality cells may be obtained for the assay by dispensing stem cells or precursor cells into e.g. multi-well plates before inducing differentiation to the adherent cell type rather than harvesting and dispensing the adherent cells themselves.

A further aspect of the invention provides an automated assay, comprising:

(a) providing an initial population of cells, comprising desired stem cells or cells from which desired stem cells can be derived;

(b) obtaining a reservoir population of desired stem cells from the initial population;

(c) separating desired stem cells from the reservoir population to carry out an assay.

Typically, the reservoir population will be maintained to provide a source of desired stem cells that can be used to provide cell samples for use in the assay. By maintaining such a reservoir population, e.g. using the automated culture methods described herein, the assay can be carried out whenever required by obtaining one or more cell samples from the reservoir population.

Thus, the assay can further comprise:

(d) maintaining and/or expanding the reservoir population; and

(e) repeating step (c) to carry out a further assay.

A further aspect of the invention provides an automated assay, comprising

providing a reservoir population of desired stem cells and repeatedly separating desired stem cells from the reservoir population to carry out an assay, whilst maintaining and/or expanding the reservoir population.

In aspects of the invention involving maintenance of a reservoir population of cells, the separated desired stem cells are typically contacted with one or more test compounds in order to carry out the assay. In some embodiments, the separated desired stem cells are induced to differentiate and the differentiated cells thus obtained are contacted with one or more test compounds. A discussed previously, this is of particular application when the assay is to be carried out on terminally differentiated cells, but may also be used e.g. when the assay is to be carried out on an adherent cell type.

In preferred embodiments of the assays of the invention the stem cells or desired stem cells are neural stem cells. However, the assays can also be carried out using other stem cell types, e.g. the stem cell types described herein, including pluripotent stem cells, haematopoietic stem cells, epidermal stem cells, mesenchymal stem cells, adipose tissue-derived stem cells and muscle stem cells.

In other preferred embodiments, the assay is carried out on differentiated cells.

For example, the differentiated cells are neurons.

When the assay is carried out using neural stem cells or neurons, the assay may be to identify a compound with potential to treat a neurodegenerative disease. For example, the assay may be to identify a compound with potential to treat Alzheimer's disease, prion diseases such as Creutzfeldt-Jakob disease, and Parkinson's disease. Preferably, the neurodegenerative disease is Alzheimer's disease.

In general, the assay can comprise determining the activity of an enzyme

expressed by the stem cells or differentiated cells. In some embodiments the stem cells or differentiated cells express a gene encoding the substrate for the enzyme. Conveniently, the enzyme promotes formation of a product from a substrate and the activity of the enzyme is determined by detecting the product in the media in which the cells are cultured. Various suitable detection methods are available to the skilled person, depending on the particular substrate and enzyme used in the assay. A non-limiting example of a suitable method is detection of the product by ELISA. In preferred embodiments, the detection of the product is automated.

In a particularly preferred embodiment, the assay is to determine the effect of a test compound on the enzyme β-secretase. β-secretase promotes the cleavage of the extracellular part of the transmembrane protein amyloid precursor protein (APP). Further cleavage of APP by γ-secretase results in the release of a 39 to 42 amino acid peptide, beta-amyloid, which is implicated in the progression of Alzheimer's disease. Thus, modulation of β-secretase activity, e.g. by inhibiting β-secretase activity, is a potential route to treating Alzheimer's disease and the assays of the invention can conveniently provide automated, high throughput screening of potential β-secretase modulators or inhibitors. The effect of potential inhibitors can be detected, as in the presence of β-secretase inhibition, reduced quantities of beta-amyloid will be released into the medium. Such assays are most conveniently carried out on neurons, because β-secretase expression is upregulated when neural stem cells differentiate to form neurons. Thus, in preferred assays the substrate is (APP) and/or the product is beta- amyloid protein.

In some embodiments of the assays of the invention, the stem cells or differentiated cells express a heterologous protein. For example, the stem cells may be genetically modified to express a heterologous protein, e.g. as described herein in relation to the production of proteins in stem cells. The heterologous protein can be a substrate for an enzyme, the activity of which is to be assayed.

In some embodiments, the heterologous protein is a mutant or modified form of an endogenous protein expressed by the unmodified stem cells or differentiated cells. The heterologous protein may, for example, provide a substrate for an enzyme on which the enzyme acts with increased activity compared to the wild type substrate.

In a preferred aspect, the heterologous protein is a modified or mutant form of amyloid precursor protein. Various mutant forms of APP have been described including forms which provide an improved substrate on which β-secretase acts with increased activity. Thus, in the absence of a β-secretase inhibitor, differentiation of neural stem cells to neurons results in an increase in β- secretase expression and a subsequent sudden and large increase in beta- amyloid present in the culture medium. Reduction in beta-amyloid levels in the presence of a test compound would indicate that the test compound has potential as an agent for treatment of Alzheimer's disease.

Yet another aspect of the invention provides an automated assay comprising:

(a) (i) providing a population of neural stem cells;

(ii) optionally, dividing the population of neural stem cells into a plurality of sub-populations;

(iii) inducing differentiation of the neural stem cells to form neurons; and

(b) determining the effect of one or more test compounds on the activity of an enzyme expressed in the neurons..

A related aspect of the invention provides an automated assay, comprising:

(a) providing an initial population of cells, comprising neural stem cells or cells from which neural stem cells can be derived;

(b) obtaining a reservoir population of neural stem cells from the initial population;

(c) separating neural stem cells from the reservoir population to carry out

an assay.

Further assays according to the invention include phenotypic screens. For example, test compounds are added to neural stem cells and the cultures are then screened for one or more features of differentiation such as neurite outgrowth or synapse formation. These features have been implicated in the development of diverse neurological and psychiatric conditions and hence can be used to screen potential therapeutic agents e.g. for treating depression, Alzheimer's disease or schizophrenia.

It will be appreciated that the assays of the invention can be used for high throughput screening of a library of compounds.

The invention also relates to apparatus adapted or programmed to carry out the assays of the invention. Accordingly, a further aspect of the invention provides an apparatus comprising: a) robotic means for handling culture vessels; b) means for inoculating stem cells into a culture; c) means for changing or adding medium to a culture; and d) programmable control means programmed to carry out an assay of the invention.

The apparatus can be substantially as described in relation to other aspects of the invention and preferably comprises means for adding test compounds to culture. In preferred embodiments, the apparatus further comprises means automated detection of a product in the culture medium.

The invention is now described in more detail in specific examples, illustrated by drawings in which:-

Fig.1 provides an overview of the experimental design. Each batch of cells is passaged twice and then plated into 96 well plates for further analysis;

Fig. 2 shows mouse ES cells plated using the CompacT SelecT system adapted/programmed according to the invention at different seeding densities. The cells were plated at five different densities in serum containing media and photographed 1 day after plating at a) 1 x 10 ⁶ , b) 2 x 10 ⁶ , c) 4 x 10 ⁶ , d) 6 x 10 ⁶ , and e) 8 x 10 ⁶ cells per T175 flask.

Fig. 3 shows mouse ES cells plated in serum containing and serum free media using either CompacT SelecT adapted/programmed according to the invention or manual methods post-passage, a) serum present, automated culture; b) serum present, manual culture; c) serum free, automated culture; d) serum free, manual culture.

Fig. 4 shows viability data obtained using Trypan blue exclusion methods for cells cultured using either CompacT SelecT adapted/programmed according to the invention or manual methods. Viable cell % ± SD in each condition over three passages (cells plated on third passage). Results are an average of a minimum of three batches.

Fig. 5 shows the levels of Oct-4 driven GFP expression in mouse ES cell cultures using either CompacT SelecT or manual methods. The GFP positive cell population ± SD when compared to a negative control is plotted for each condition. Results are an average of a minimum of three batches.

Fig. 6 shows the results of a clonal assay. The panels show alkaline phosphatase positive colonies formed from cells cultured in a) serum free CompacT SelecT culture, b) serum free manual culture, c) serum containing CompacT SelecT culture and d) serum containing manual culture.

Fig. 7 shows the presence of tubulin positive cells in monolayer neural differentiation cultures after 8 days in cultures seeded with mouse ES cells obtained using CompacT SelecT and manual methods.

Example 1 - Automated Culture of Murine Embryonic Stem Cell Lines

We demonstrated the successful automated culture of murine embryonic stem cell lines on a newly developed entry level cell culture platform. This platform allows the automation of seeding, feeding and other cell culture processes in order to maintain cell lines in standard T175 cell culture flasks. Embryonic stem cells have the capacity to generate all cell types in the body and their culture conditions must be strictly controlled. Cells cultured using automation are compared with parallel manual cultures in two relevant media formulations to ensure that viability, pluripotency and potential to differentiate into neuronal subtypes is not impaired. The automation of the culture of embryonic stem cells is a key step in providing a viable supply of such cells and their differentiated progeny for screening or therapeutic purposes.

Introduction

Whilst there are a number of automated platforms capable of cell culture, it could be argued that SelecT, manufactured by the Automation Partnership, has become the pharmaceutical industry standard for cell culture (Miret et al, 2005; Rush, 2004, Jenkins, 2004, Cacace, 2003). The SelecT system, and its newer model CompacT SelecT, consists of a robotic manipulator that can access a flask and plate incubator, a cell counter and plating equipment. Processing of the cells takes place within a HEPA filtered cabinet to ensure sterility. This technology has never been applied to the culture of ES cells and the existing machines are not set up for such cultures. We therefore modified existing manual methodologies for one of our mouse ES cell lines to a specially adapted

CompacT SelecT and compared the quality of the resulting cells against those from manual culture methods.

The pre-determined criteria for demonstrating equivalence of Oct4GFP mES cells resulting from manual and robotic culture were: i) equivalence of cell viability at the time of harvest ii) demonstration of equivalent cell morphology iii) demonstration of Oct-4 expression at the conclusion of the culture period examined iv) demonstration of clonal growth at the conclusion of the culture period examined and v) demonstration of neural differentiation. The method development leading up to the establishment of a working automated culture regime for Oct4GFP mES cells and the subsequent comparison of the five acceptance criteria outlined herein are now described in this study.

Materials and Methods

Experimental Design

Oct4GFP mES cells were used for the experiment. This cell line has EGFP- IRES-Puromycin under control of the Oct4 promoter, which is a marker for mouse ES cell pluripotency, allowing for convenient reporting of the ES cell phenotype by means of GFP expression (Ying et al, 2002).

Initial tests were carried out in order to determine appropriate parameters for enzyme incubation time (required to dissociate cells from the mother flask) and seeding density for daughter flasks. Once established, this incubation time and seeding density were used for the remainder of the experiment as described in the Results section.

Mother flasks (T175, BD Biosciences, Oxford, UK) were prepared manually using conventional cell culture techniques. Mother flasks were then passaged into a minimum of two daughter flasks according to the passage protocol below. Two

daughter flasks from the first passage were expanded for 48 hours and each was subsequently passaged into a further two daughter flasks. These flasks were expanded for 48 hours and the cells harvested and plated into 96 well plates using the plating protocol described below. The plated cells were promptly analysed for GFP expression using either the Guava EasyCyte (Guava Technologies, Hayward, CA) or FACS analysis (FACSCalibur, BD Biosciences). Further biological assays were carried out in order to verify the cells' ability to form colonies from single cells and that the cells retained their capacity to differentiate into neuronal subtypes as described below.

See Figure 1 for diagram of experimental design.

Medium

Serum containing experiments were conducted with GMEM (Sigma, St Louis, MO) supplemented with 10% (v/v) foetal calf serum (Biosera, East Sussex, UK), non-essential amino acids (Invitrogen, Glasgow, UK), β-mercaptoethanol (Invitrogen), glutamine (Invitrogen), sodium pyruvate (Invitrogen) and leukaemia inhibitory factor (LIF).

Serum-free experiments were conducted with ESGRO complete (Chemicon, Hampshire, UK), a proprietary formulation developed by Stem Cell Sciences (UK) Ltd. (Ying et al, 2003).

Clonal Assay

Cells were plated in duplicate dishes at 1000 cells per 10cm ² cell culture dish (Iwaki, Barloworld Scientific, UK) in 10 ml of either serum-containing or serum- free medium. Dishes were pre-coated with 0.1% (v/v) gelatin solution prior to cell seeding. The medium was changed on day 3. On day 5 the colonies were

stained for alkaline phosphatase (Sigma) and counted manually. Visualisation of cell culture was by light microscopy.

Neural Differentiation Assay

Cells were differentiated using a modified monolayer differentiation protocol (Ying et al, 2003). Cells were centrifuged at 200 rcf for 3 minutes at room temperature. The supernatant was removed and the cells resuspended in neural differentiation RHB-A media (Stem Cell Sciences (UK) Ltd.). Cells were plated onto gelatin coated 6-well plates (Iwaki) at 1 x 10 ⁵ cells/well in 2 ml RHB-A medium. The medium was changed every 2 days. After 8 days, the wells were fixed with 4% (w/v) paraformaldehyde and immunostained for neural markers Nestin (Chemicon) and Tubulin (Covance). Visualisation of the cell culture was by fluorescence microscopy.

Serum Containing Media - Determination of enzyme treatment time and seeding density

Three replicate 80% confluent T175 flasks of Oct4GFP mouse embryonic stem cells were passaged, each using a different enzyme incubation time. Accutase (Chemicon, Hampshire, UK) was used as the dissociation enzyme (2ml per T175 flask). The enzyme incubation times selected were 4 minutes, 7 minutes and 10 minutes at 37 ⁰ C. The flasks were passaged as per the passage protocol described below.

For cells produced by automated CompacT SelecT culture, cell counts and viability were determined by Cedex (Innovartis), which uses automated digital image recognition to establish cell viability by means of Trypan blue exclusion. Cells produced by manual culture were counted by haemocytometer and Trypan blue exclusion.

The cells from each flask were counted and split into 5 daughter flasks at different seeding densities in order to determine an appropriate number of cells for passage into new flasks. In CompacT SelecT, flasks can either be seeded according to a predetermined cell count, or by split ratio whereby a proportion of the total cells recovered from the mother flask is seeded into each daughter flask. In the first instance, it was decided to passage by cell count. The cell densities chosen were: 1.0 x 10 ⁶ , 2.0 x 10 ⁶ , 4.0 x 10 ⁶ , 6.0 x 10 ⁶ and 8.0 x 10 ⁶ cells per flask. The number of cells seeded was based on an average yield of 3 x 10 ⁷ cells from a T175 flask under standard cell culture conditions and a split ratio of 1 :5 for passage every 48 hours.

The most successful enzyme incubation time and seeding density were used for passaging the cells in the remainder of the experiment.

Serum-Free Media - Determination of enzyme treatment time and seeding density

In order to avoid the requirement for a centrifugation step, cell dissociation buffer (Invitrogen) was used instead of a dissociation enzyme. Failure to remove residual enzyme from serum free cultures can result in problems with cell adherence, therefore an enzyme free buffer was chosen. The incubation time was determined by comparison to the manual control culture. In manual culture, the spent media was removed from the flask and the cell sheet rinsed twice with 30 ml of PBS at room temperature. 3 ml of cell dissociation buffer was then added and the flask gently tilted in all directions to ensure an even coverage of the cell sheet. The flask was then placed into an incubator maintained at 37 ⁰ C, 7% (v/v) CO ₂ for 10 minutes after which time the cells were inspected under the light microscope. If the cells had not detached, the flask was replaced gently into the incubator for another 10 minutes. This process was repeated until the cells had obviously detached from the tissue culture plastic. The determined

incubation time was then used in CompacT SelecT to ensure that a single cell suspension was produced.

Cell counts and viability were determined by Cedex (Innovartis) in CompacT SelecT or by haemocytometer and Trypan blue exclusion in manual culture.

The cells were counted and split into 5 daughter flasks at different seeding densities in order to determine the appropriate number of cells for passage into new flasks. In CompacT SelecT, flasks can either be seeded according to a predetermined cell count, or by split ratio whereby a proportion of the total cells recovered from the mother flask is seeded into each daughter flask. In the first instance, it was decided to passage by cell count.

The cell densities chosen were: 1.0 x 10 ⁶ , 2.0 x 10 ⁶ , 4.0 x 10 ⁶ , 6.0 x 10 ⁶ and 8.0 x 10 ⁶ CeIIs per flask.

The determined cell dissociation time and most successful seeding density were then used for passaging the cells in the remainder of the experiment.

Serum Containing Media - Passage protocol

New tissue culture flasks (BD Biosciences) were pre-coated with 0.1% (v/v) gelatin solution prior to seeding with cell suspension.

Spent media was removed from the source cell flask and poured to waste. The cells were then washed with 30 ml of phosphate buffered saline (PBS) in order to remove any residual serum containing media. Care was taken not to pipette/ pump PBS directly onto the cell sheet. Accutase solution (2 ml per T175 flask) was added to the flask and the flask gently tilted in all directions to ensure an even coverage of enzyme across the cell sheet. The flask was placed into the flask incubator at 37 ⁰ C, 7% (v/v) CO ₂ for the predetermined incubation time.

Upon removal of the flask from the incubator, the flask was shaken from side to side in order to dislodge the cells from the tissue culture plastic. A known volume of media was then added to the flask to dilute out the enzyme/cell suspension mixture. A 1 ml sample of the cell suspension was taken for cell count and viability analysis by Cedex / Trypan blue exclusion. A calculated volume of cell suspension was then added to 50ml of fresh cell culture media in a new flask to give the required seeding density. Flasks were incubated at 37 ⁰ C, 7% (v/v) CO ₂ for 48 hours, at which time the passage protocol was repeated.

This process was replicated as closely as possible by manual culture using standard tissue culture equipment, without inclusion of a centrifugation step.

Serum Free Media - Passage Protocol

New tissue culture flasks (BD Biosciences) were pre-coated with 0.1% (v/v) gelatin solution prior to seeding with cell suspension.

Spent media was removed from the source cell flask and poured to waste. The cells were then washed twice with 30 ml of phosphate buffered saline (PBS) in order to remove any residual media. Care was taken not to pipette PBS directly onto the cell sheet. 3 ml of cell dissociation buffer (Invitrogen) was added to the flask and the flask gently tilted in all directions to ensure an even coverage of cell dissociation buffer across the cell sheet. The flask was placed into the flask incubator at 37 ⁰ C, 7% CO ₂ for the predetermined incubation time. Upon removal of the flask from the incubator, the flask was shaken from side to side in order to dislodge the cells from the tissue culture plastic. A known volume of media was then added to the flask to dilute out the dissociation buffer/cell suspension mixture. A 1 ml sample of the cell suspension was taken for cell count and viability analysis by Cedex / Trypan blue exclusion. A calculated volume of cell suspension was then added to 50ml of fresh cell culture media in a

new flask to give the required seeding density. Flasks were incubated at 37 ⁰ C, 7% CO ₂ for 48 hours, at which time the passage protocol was repeated.

This process was replicated as closely as possible by manual culture using standard tissue culture equipment, without a centrifugation step.

Plating Protocol

A single cell suspension was formed from the source flask as per the passage protocol for either serum containing or serum free culture. Cells were counted using Cedex / Trypan blue exclusion and diluted with fresh media to 2.5 x 10 ⁵ cells/ml.

For CompacT SelecT culture, the diluted cell suspension was transferred into the cell pot and the cells kept in suspension by means of a magnetic stirrer bar. The cell suspension was then pumped to the Multidrop (Thermo Electron Corp.) plating equipment and 100 μl of cell suspension dispensed into each well of a 96 well plate. Cells were plated into three types of 96 well plates. Clear and black walled tissue culture treated (Greiner) 96 well plates were pre-gelatinised and used for microscopy. Round bottom non-tissue culture treated 96 well plates (Griener) were used to collect cell suspension for the clonal and neural differentiation assays and for analysis of GFP expression. All results are from a minimum of three processed batches of cells.

For manual culture, the diluted cell suspension was plated using a Multidrop dispenser situated within the biological safety cabinet. The same type of 96 well plates was used as in CompacT SelecT culture and the cells were collected for GFP analysis and assay in an identical way.

Cells plated into TC treated 96 well plates were incubated at 37 ⁰ C, 7% (v/v) CO ₂ for microscopy after 24 to 48 hours.

GFP Analysis

Cells which had been passaged twice and then plated into 96 well format were analysed for GFP expression using either conventional flow cytometry (FACSCalibur, BD Biosciences) or Guava Easycyte instrument (Guava Technologies).

The Oct4GFP cell line used in the experiment is engineered with GFP under control of the Oct-4 promoter. As previously stated, Oct-4 is often used as a marker for mES cell pluripotency and thus the level of GFP expression can be used to determine if the cell population remains pluripotent. The cells used as a negative control were 46C cells (Ying et al, 2003), which have GFP under control of Sox-1 , an early marker for neural fate.

Plated cell suspension samples were analysed for GFP expression using the Guava Easycyte Express Plus software module. The percentage GFP positive population was determined by overlay of the negative control population. Samples from CompacT SelecT and control cultures were analysed at the same time point as far as was reasonably practicable.

Due to unforeseen circumstances, some of the plated cell suspension samples were analysed for GFP expression using FACSCalibur CellQuest software. The percentage GFP positive population was determined by overlay of the negative control population. Samples from CompacT SelecT and control cultures were analysed at the same time point as far as was reasonably practicable.

Data acquired from the Guava Easycyte and exported into generic flow cytometry software for analysis gave the same results as data analysed in the Express Plus software module.

Results

Determination of enzyme treatment time and seeding density

In serum containing conditions, both the 7 and 10 minute Accutase treatment times formed single cell suspensions. In order to minimise any adverse effect on the cells and maximise the efficiency of the passage protocol, the shorter 7 minute incubation time was selected for use in the experiment.

In serum free conditions, a thirty minute incubation with cell dissociation buffer was sufficient to produce a single cell suspension and was used for the duration of the experiment (data not shown).

The viability data (Figure 4) indicated that neither of these incubation times had an immediate effect on cell survival. The mean viability over three passages for 7 minute Accutase incubation was 95.87% (CompacT) and 98.41% (Manual). The mean viability over three passages for thirty minute cell dissociation buffer incubation was 90.96% (CompacT) and 89.88% (Manual).

Images of mES cells plated out at different seeding densities (Figure 2) in serum containing conditions by CompacT SelecT (day 1 post plating) showed the expected variation in seeding density achieved by CompacT SelecT. The seeding densities used were a) 1 x 10 ⁶ , b) 2 x 10 ⁶ , c) 4 x 10 ⁶ , d) 6 x 10 ⁶ and e) 8 x 10 ⁶ cells per flask.

The mES cells adhered well to the gelatin coated tissue culture plastic and formed small colonies, even at the lowest plating density. At the higher plating densities of 6 x 10 ⁶ and 8 x 10 ⁶ CeIIs, the cells formed larger colonies which then merged together and gave the best yield of cells after 48 hours in culture. A seeding density in the range of 6-8 x 10 ⁶ cells was therefore considered appropriate for the duration of the experiment.

Passage Protocol

Cells passaged in serum and serum free conditions by CompacT SelecT and manual culture (Figure 3) showed normal mES cell morphology. Images were taken day 1 post passage and each flask was seeded with 6-8 x 10 ⁶ cells. The cells appeared healthy post-passage and there was little evidence of cell death, particularly in the serum containing conditions (Figures 3a, 3b). In the serum free conditions (Figures 3c, 3d) there were a slightly higher number of dead cells by visual assessment in the tissue culture flasks. These results were consistent with those from manual culture. A small proportion of differentiated cells were evident in both CompacT SelecT and manual cultures.

Viability

The viability of cells was determined by Trypan blue exclusion both by Cedex instrument in CompacT SelecT and manually (Figure 4). There was no apparent difference in cell viability between CompacT SelecT and manual cultures and viability of the cells was maintained throughout the course of the experiment. The mean viabilities were 95.8%±1.7% and 98.4%±0.5% for robotic and manual culture in serum respectively, and 90.9%±4.9% and 89.8%±7.6% for robotic and manual culture in serum free conditions respectively. There was no statistical significance (p>0.05) in viability between initial and final cell passages in serum containing media, however in serum-free flasks there was a significant increase (p<0.05) in both robotic and manual cultures. The viable cell population of the serum free cultures increased from 81.3%±2.3% to 95.3%±0.6% in manual culture and 85.3%±1.0% to 94.6%±0.3% over the following two passages.

GFP Expression

All cultures maintained GFP expression at >85% (Figure 5) giving a good indication that the cells are still expressing Oct-4. The mean GFP positive

populations at the final passage were 85.4%± 6.0% (CompacT) and 85.9% ± 5.0% (manual), in the serum containing cultures and 91.4%±0.9% (CompacT) and 91.26%±1.1% (manual) in the serum free cultures. There was no significant difference (p>0.05) between CompacT SelecT and manual control cultures.

Clonal Assay

All cultures formed alkaline phosphatase positive colonies when plated at clonal density (Figure 6). In serum containing conditions the number of colonies formed per 1000 cells plated was higher than in serum free conditions (data not shown). There was no significant difference in the number of colonies formed from cells cultured in CompacT SelecT or manually in serum containing conditions. In serum-free conditions, the cells cultured in CompacT SelecT gave rise to a larger number of colonies, but the difference was not statistically significant (p>0.05) (data not shown). In serum containing media, colonies had flatter morphology and contained some differentiated cells. In serum free media colonies in general were smaller and the cells tightly packed. Both of these observations are consistent with expectations.

Neural Differentiation

Cells from both CompacT SelecT and manual culture formed neural cells by undergoing the monolayer differentiation protocol. After 8 days the cultures contained a mixture of neuronal subtypes including cells positive for nestin (data not shown) and tubulin (Figure 7). There was no apparent difference by visual inspection in the number or type of neural cells formed from CompacT SelecT or manual cultures.

Discussion

We have developed automated cell culture protocols suitable for passage and plating of mouse ES cells using the CompacT SelecT system from The Automation Partnership.

Automation of passage and plating of mES cells

In general the mES cells adapted well to automated cell culture passage and plating, with no significant difference observed between the automated and parallel manual process. Cell characteristics including Oct4 expression, colony formation from single cells, viability and capacity for neural differentiation appeared unchanged.

An enzyme incubation time of 7 minutes was deemed acceptable for cell dissociation in serum containing cultures, although this can be further optimised for other cell lines and enzyme treatments. The cells readily re-attached to the gelatinised tissue culture flasks and expanded as normal.

With CompacT SelecT, the user can specify to passage cells based on either a pre-specified cell number or by a split ratio. In the first instance it was decided to passage based on replating a set number of cells into each daughter flask, however, as this must be decided prior to knowing the yield of cells from the flask, later passages were carried out using the split ratio method. Figure 2 gives a visual indication that CompacT SelecT can accurately plate out cells at different seeding densities, and this was used to determine the most appropriate cell number to use for seeding daughter flasks. It was decided to use either 6 x 10 ⁶ or 8 x 10 ⁶ cells per T175 flask.

Automation of passage and plating in serum free conditions

One of the main differences in protocol when automating mES cell culture is the absence of a centrifuge and wash step to remove any excess enzyme in the harvested cell suspension before replating. This step is not possible with CompacT SelecT. In serum containing conditions this is not normally an issue as the serum quenches any residual enzyme activity. However, in serum-free conditions, the cells can be very sensitive to even the smallest trace of residual enzyme leading to problems with cell adhesion end differentiation. For this reason, cell dissociation buffer was used in the serum-free experiments, rather than an enzyme based reagent. The incubation time must be sufficient to remove the cells from the tissue culture plastic and also dissociate the cell sheet into a single cell suspension. This enables accurate cell counting and reproducible passaging.

However, in order to form a single cell suspension with the cell dissociation buffer, the incubation time must be increased to 30 minutes. This is not ideal from a process viewpoint as flasks must be interleaved to increase efficiency, and the effect on the cells themselves is unknown. Viability data (by Trypan blue exclusion) showed that there was no obvious cell death caused by this longer incubation time. Cells re-attached to the gelatinised tissue culture flask and expanded as normal.

The only flasks which had a viability of less than 90% during the experiment were the serum free mother flasks, resulting in a significant difference in viability between first and final passages. This was caused by the cells becoming over confluent one day prior to the start of the experiment resulting in a larger than normal number of dead cells being carried over. It is encouraging to note that the viability of the serum free cell populations increased to greater than 90% over the subsequent two passages.

Due to the extensive PBS wash step within the protocol, most of the dead cells in the culture are removed prior to the cell dissociation step. This accounts for the larger number of dead cells visible microscopically (Figure 3) in the serum free cultures not being reflected in the viability data. The viability data does however indicate that the cells are in reasonable health following the incubation period.

mES retain Oct4 expression after automated passage and plating

The GFP data indicates that there is no significant difference in Oct-4 expression between automated and manual cultures after three passages. Unfortunately due to logistical difficulties it was not possible to analyse all the samples on the same instrument, however all pairs of samples (automated vs. manual culture) were analysed on the same instrument and positive populations were determined by comparison to a negative control analysed on the same instrument.

These data correlate well with the results of the biological assays, which show that the cells were still able to form colonies from single cells and differentiate into neural lineages. As the focus of this work was to develop cells which could go on to form neural cells in culture, this was the only lineage specification investigated. Future experiments will include differentiation into all three germ layers in order to ensure pluripotency is maintained.

Conclusions

This work has demonstrated that murine embryonic stem cell lines can be routinely propagated and plated into multiwell format using an automated cell culture platform. The cells maintain expression of Oct4, a marker of ES cell pluripotency, and subsequently are able to form colonies from single cells and differentiate into neuronal subtypes, comparably to cells grown using traditional manual methods. This provides robust and reproducible ES cell production suitable for e.g. screening purposes.

The protocols set out herein have also been successfully adapted for the culture of mouse neural stem (NS) cells and human multipotent adipose-derived stem (MADS) cells. For mouse NS cells, a seeding density of 2 x 10 ⁶ to 5 x 10 ⁶ cells per T175 flask (dependant on the NS cell line used, for example 3 x 10 ⁶ cells per T175 flask for 46cNS cells) was used. The cells were passaged every 3-4 days using a split ratio of 1 :5 to 1 :10, depending on the NS line (a ratio of 1 :10 was used for 46cNS cells). The cells were dissociated using cell dissociation buffer (diluted 1 in 3 with PBS) and an incubation time of 8 minutes. For human MADS cells a seeding density of 1 x 10 ⁶ to 2 x 10 ⁶ cells per T175 flask was used. The cells were passaged every 4 days using a split ratio of 1 :5 or 1 :6. The cells were dissociated using Accutase, described in the protocols for serum-containing media, and an incubation time of 5 minutes.

Example 2 - Production of a Protein by Stem Cells in Culture.

Stem cells can be used as a base cell line for protein or vaccine production. For protein production in stem cells, DNA encoding the gene of interest is transferred into the host cell (in this case a stem cell) by any suitable method, including electroporation, lipofection, nucleofection, calcium phosphate/calcium chloride and viral transduction. Neural stem (NS) cells can conveniently be transfected using the nucleofection technology (Amaxa). There are many well published strategies for transient and stable transfection and selection for cells containing the introduced DNA, commonly including the use of markers conferring antibiotic resistance in transfected cells, or incorporation of a marker allowing physical cell sorting of the transfected cells.

Following transient transfection, cells are typically plated out directly for use/assay.

In the case of a stable transfection cells containing the gene of interest are

selected e.g. by use of antibiotic selection, and colonies of transfected cells picked for expansion and checking of expression levels of the gene of interest. Colonies of cells are picked manually or using an automated clone picking device.

Cells are then be expanded from 24 well plates to 6 well plates to T flasks (T25, T75, T175), until sufficient cells are obtained for seeding large-scale stem cell cultures using e.g. the protocols provided in Example 1.

Commonly the protein of interest is secreted into the media, and used either as conditioned media or further purified depending on the usage requirements. Other methods include lysing the cells to obtain the protein of interest.

A protocol for transfection of stem cells and production of a secreted protein is provided below:

A. Prepare sufficient cells (e.g. a large T flask) to be available in log phase growth at the time of transfection.

B. Prepare DNA depending on transfection protocol of choice (plasmid or viral vector)

C. Transfect host cells using preferred transfection method (for example using the method set out below for electroporation of ES cells).

D. Use a selection methodology to isolate successfully transfected cells.

E. Pick colonies of cells and confirm expression of protein of interest. F. Expand clones of interest (typically highly expressing clones are chosen for further expansion).

G. Harvest media containing the secreted protein and further purify the protein, if desired.

The bioprocess is a batch process (whereby the medium and/or cell population is harvested at the end of the culture), which may include the steps of passaging

and/or feeding the cells by adding additional medium or other components to the culture vessel or vessels during the process. Alternatively, media are continually (or periodically) added and media and/or cells are continually (or periodically) harvested throughout the culture process.

Electroporation of ES cells - Protocol for gene targeting and stable transgene expression

DNA preparation 1. Linearize 10-50 μg of DNA by restriction digest.

2. Precipitate DNA by adding ethanol to the top of the Eppendorff tube.

3. Spin DNA down.

4. In a tissue culture hood, aspirate ethanol, leave the tube for a few minutes in the hood to allow residual ethanol to evaporate. 5. Resuspend DNA in 100 μl of PBS. Make sure DNA is dissolved before trypsinising ES cells. This can be done the night before electroporation. Store DNA in the fridge (4 ⁰ C).

ES Cells a. Passage ES cells into 1-2 medium flasks two days before electroporation. b. On the day of electroporation, Gelatin-coat an appropriate number of TC grade 100mm dishes (cells are plated at a density of 2-5 x 10 ⁶ /dish depending on the drug selection to be used, see note 1). c. Wash ES cells 2X in PBS and trypsinise as normal. d. Resuspend ES cells in 10 ml of PBS and count the number of cells. e. Plate 2 x 10 ⁶ cells in one dish as selection control (it is not necessary to electroporate these cells). f. Spin a minimum of 2 x 10 ⁷ cells. g. While cells are being centrifuged, measure the right amount of medium to be used for plating and transfer to a fresh container. Usually 60 ml (i.e. enough for 6 dishes) is enough.

h. Resuspend cells in 700μl PBS and transfer to electroporation cuvette, i. Add linearised plasmid DNA to the cuvette. The final volume of cuvette is thus 800μl. j. Electroporate cells. Parameters for normal integration: 3.0μF 800V (electroporation, the time parameter should read 0.1) k. Immediately add cells to GMEM/10% FCS (without selection). Plate 10 ml of cells to each dish (see note 2)

I. 24 hours later, add appropriate drug to each dish (see note 3). m. Change medium every 2 days until the cells on the control plate die and colonies are big enough to pick. This normally will take around 10 days, n. Pick colonies and expand into 24 well plates.

Notes:

1) Plating density is in the range of 2-5 x 10 ⁶ /dish depending on selection efficiency of drugs. Generally 5 x 10 ⁶ /dish for Puromycin, 3-4 x 10 ⁶ /dish for G418 and 2 x 10 ⁶ /dish for Hygromycin.

2) It has been shown that the electroporation efficiency is significantly reduced if the number of cells is fewer than 2 x 10 ⁷ and DNA less than 10μg. It is therefore preferred to do electroporation with more DNA and cells and subsequently plate a fraction of them.

3) Hygromycin: 100 μg/ml, G418: 200 μg/ml, Puromycin: 2 μg/ml.

4) For targeting experiments with unknown targeting efficiency, use minimum of 50 μg of DNA and 5 x10 ⁷ cells. Normal practice is to screen 200 clones (this means to start with 250 colonies). If high efficiency targeting has previously been achieved, further electroporation is done at a smaller scale (20 μg DNA, 2 x 10 ⁷ cells and pick 48- 96 colonies).

The invention thus provides large scale automated production of stem cells, including ES cells.

Example 3 - Candidate drug screening assay

Candidate agents for treatment of Alzheimer's disease are screened using an automated assay. The assay is based on the upregulation of β-secretase expression on differentiation of neural stem (NS) cells to neurons, resulting in cleavage of its substrate, the transmembrane protein amyloid precursor protein (APP), resulting in the release of beta-amyloid into the medium. The levels of beta-amyloid in the culture medium are detected by ELISA.

A protocol for carrying out the assay is provided below:

NS cells are transfected with a construct comprising DNA encoding a modified APP using the procedures of Example 2. The modified APP provides an improved substrate for β-secretase and hence a more sensitive assay.

NS cells containing the construct are expanded using the protocols described in Example 1 for NS cell culture to provide a reservoir population of from 1 to 50 flasks of NS cells.

Aliquots of NS cells from the reservoir population are dispensed to individual wells of multi-well plates coated with poly-D-lysine and laminin or poly-ornithine and laminin.

Candidate β-secretase inhibitors or control compounds are added to each well. Alternatively, the candidate β-secretase inhibitors or control compounds are added to each well prior to dispensing the NS cells.

Retinoic acid is added to each well to induce differentiation of the NS cells to neurons.

Samples of culture medium are removed from each well and subjected to ELISA analysis for the presence of beta-amyloid. Wells in which beta-amyloid is absent or present at levels lower than in control wells confirm β-secretase inhibition by the candidate compound added to those wells.

In a modified version of the assay, the NS cells are dispensed to individual wells of multi-well plates and induced to differentiate to neurons prior to the addition of the candidate β-secretase inhibitors or control compounds.

In other versions of the assay, differentiation of the NS cells to neurons is induced by removal of growth factors used to support NS cells from the medium (e.g. for mouse NS cells) or the addition of growth factors such as BDNF (brain- derived neurotrophic factor) or FGF (e.g. for human NS cells).

References:

Pouton CW ₁ Haynes JM: Pharmaceutical applications of embryonic stem cells. Adv Drug Deliv Rev. 2005; 57:1918-34.

Smith AG: Embryo-derived stem cells: of mice and men. Annu Rev Cell Dev Biol. 2001; 17:435-62.

Gorba T, Allsopp TE: Pharmacological potential of embryonic stem cells. Pharmacol Res. 2003; 47:269-78.

Li L, Baroja ML, Majumdar A, Chadwick K, Rouleau A, Gallacher L, Ferber I ₁ Lebkowski J, Martin T, Madrenas J, Bhatia M: Human embryonic stem cells possess immune-privileged properties. Stem Cells. 2004; 22:448-56.

Vacanti CA: History of tissue engineering and a glimpse into its future. Tissue Engineering. 2006; 12:1137-1142.

Miret J, Zhang J, Min H, Lewis K, Roth M, Charlton M, Bauer P: Multiplexed G- Protein-Coupled Receptor Ca ²⁺ Flux Assays for High-Throughput Screening. Journal of Biomedical Screening. 2005; 10:780-787.

Rush A: Strategies for optimizing development of robust stable cell lines for automated screening. Lecture presented at Society for Biomolecular Screening, Orlando FL 2004.

Jenkins O: High throughput protein expression for drug discovery; pitfalls and prospects. . Lecture presented at Society for Biomolecular Screening, Orlando FL 2004.

Cacace A: Industrialisation of cell culture for discovery. . Lecture presented at Society for Biomolecular Screening, Orlando FL 2004.

Ying Q, Nichols J, Evans E, Smith A: Changing potency by spontaneous fusion. Nature 2002; 416:545-547.

Ying QL, Nichols J, Chambers I, Smith A: BMP induction of Id proteins suppresses differentiation and sustains embryonic stem cell self-renewal in collaboration with STAT3. Cell 2003; 115:281-292.

Ying QL, Stavridis M, Griffiths D, Li M, Smith A: Conversion of embryonic stem cells into neuroectodermal precursors in adherent monoculture. Nature Biotechnology 2003; 21 :183-186.

Terstegge S, Laufenberg I, Pochert J, Schenk S, Iskovitz-Eldor J, Endl E, Brustle O: Automated maintenance of embryonic stem cell cultures. Biotechnology and Bioengineering 2007: 96:195-201.