WO2015071883A1 - Process for growing and preparing cells in-vitro for metabolic profiling - Google Patents

Abstract

A method for pretreating a sample of cells for metabolite profiling comprises the steps of: immobilizing cells on a solid membrane, washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate for a time period of up to 30 seconds and contacting the immobilized cells with a quenching solution comprising dichloromethane (DCM) and ethanol (EtOH). The use of at least one ¹³C-labeled carbon source for pretreating a sample of cells for metabolite profiling is disclosed.

Description

Process for growing and preparing cells in-vitro for metabolic profiling

The present invention relates to the field of metabolomics. In particular, provided is a method for pretreating a sample of cells for metabolite profiling comprising the steps of immobilizing cells on a solid membrane, washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate for a time period of up to 30 seconds and contacting the immobilized cells with a quenching solution comprising dichloromethane (DCM) and ethanol (EtOH). The invention, furthermore, relates to the use of at least one ¹³C-labeled carbon source for pretreating a sample of cells for metabolite profiling.

Metabolomic analyses of isolated cells such as cells of a cell culture are a valuable tool for determining metabolic changes. However, metabolomics approaches are sensitive towards artificial changes of the metabolites, e.g., due to degradation reactions which may occur rather rapidly in the samples prior to the actual analysis.

The currently available analysis technology for metabolomics is essentially based on either liquid or gas chromatography-based mass spectrometry or liquid or gas chromatography-based NMR spectroscopy.

There are protocols for efficient metabolite extractions (see WO 201 1/003945) as well as for derivatization for gas chromatography in metabolomics. However, the step of sample pretreat- ment is still sensible with respect to a potential degradation of metabolites or other changes of the actual metabolic composition of samples (Bolten 2007, Anal Chem 79: 3843-3849).

Many approaches aim at quenching metabolite activity and preventing leakage of the cells (Meinert 2013, Anal Biochemistry 438: 47-52; Volmer 201 1 , Biotechnol Lett 33: 495-502).

Furthermore, meaningful metabolite profiling data can only be obtained if variations in the amount of starting material used for metabolite extraction are properly accounted for. This issue concerns any type of biological material, but it is especially eminent in the metabolic analysis of in vitro cultivated cells, where usually only small sample volumes are available. Various parameters can impact the amount of generated cell material in a sample, as for example cell line type, treatments with cytotoxic substances, cultivation conditions, or harvesting procedures.

So far, several normalization strategies have been described. Among others, cell number is a commonly employed parameter for the normalization of raw metabolite peak data across multiple samples (Dettmer 201 1 , Anal Bioanal Chem 399: 1 127-39). However, accurately determining the number of adherent cells cultured on a solid membrane, or cells recovered after filtration and washing of a suspension culture is at least difficult, if not impossible.

Alternatively, Cao et al. (Cao 201 1 , Anal Bioanal Chem 400: 2983-93) describe the identification of specific marker metabolites that correlate with protein content that can be used for spectral data adjustment. Moreover, internal normalization of spectral data is commonly applied in quantitative metabolomics by adjusting single peak areas to the sum of the peak areas of all ana- lytes. However, both peak area sum and cell count are only linear in a small concentration range of cells (less than one order of magnitude, Hutschenreuther 2012, Anal Methods 4: 1953- 63).

A further way of accounting for biological variability during an experiment with cell cultures is to normalize the spectral data against the total protein content of a cell sample as described in Balcke et al. (Balcke 201 1 , Toxicol Lett 203: 200-209). In this case, parallel control cell cultures have to be set up for all timepoints or experimental treatments, thereby significantly increasing the cost and complexity of a study.

Thus, methods for an efficient pretreatment and or normalization of samples for metabolomics analyses are not yet available but nevertheless highly desired.

Accordingly, the technical problem underlying the present invention might be seen as the provision of means and methods for allowing an efficient extraction and analysis of metabolites present in biological samples such as cell culture samples avoiding the drawbacks referred to before. This technical problem is solved by the embodiments characterized by the claims and herein below.

The present invention relates, thus, to a method for pretreating a sample of cells for metabolite profiling comprising the steps of:

a) immobilizing cells on a solid membrane;

b) washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate for a time period of up to 30 seconds; and

c) contacting the immobilized cells with a quenching solution comprising dichloromethane (DCM) and ethanol (EtOH).

The method may comprise the aforementioned steps, i.e. it may contain additional steps such as cell cultivating and/or further pretreatment steps, such as freezing or cooling steps, extraction of metabolites and/or the steps of the actual analysis of the sample for the presence, absence or quantity of metabolites comprised therein. Moreover, it is also envisaged that the method may essentially consist of the aforementioned steps. The method may essentially or in part be assisted by automation, e.g., by suitable robotic devices carrying out the cell culture, sample taking and pretreatment steps.

The phrase "pretreating a sample of cells for metabolite profiling" as used herein refers to treating the said sample of cells in such a way that the metabolites comprised in the said sample can be analyzed by metabolic analyzing devices, such as LC-MS and/or GC-MS devices or NMR devices, without further ado. However, it will be understood that the metabolites comprised in the sample shall after the pretreatment still reflect the metabolites which were present before the said pretreatment, i.e. the metabolites present in the sample of cells. Accordingly, it is envisaged that degradation or other changes of the metabolites originally present in the sample of cells shall be prevented or at least significantly reduced.

The term "sample of cells" as used herein refers to samples comprising suspension cells, adherent cells, dispersed cells and other types of samples comprising cells. The cells referred to in this context may be prokaryotic, such as bacterial cultures, or eukaryotic cells, such as yeast or fungal cell cultures, single-cellular organisms or eukaryotic higher cells, e.g., cells of animal or human cell culture or cancer cells. Preferably, the cells to be used in the method of the present invention are either adherent cell culture cells or suspension cell culture cells. If adherent cells are envisaged, the said cells are, preferably, cultivated directly on the solid membrane to be applied in the method of the present invention. In another preferred embodiment, the sample of cells to be used in this invention are either in a healthy or diseased condition. In a further preferred embodiment the sample of cells are treated with a drug or untreated before sampling. Preferably, the cells were not grown in the presence of an isotopically labeled substrate. More preferably, the cells were not grown in the presence of a ¹³C labeled substrate.

The term "metabolites" as used herein refers to small molecule compounds, such as substrates for enzymes of metabolic pathways, intermediates of such pathways or the products obtained by a metabolic pathway. Metabolic pathways are well known in the art and may vary between species. Preferably, said pathways include at least citric acid cycle, respiratory chain, photosynthesis, photorespiration, glycolysis, gluconeogenesis, hexose monophosphate pathway, oxidative pentose phosphate pathway, production and β-oxidation of fatty acids, urea cycle, amino acid biosynthesis pathways, protein degradation pathways such as proteasomal degradation, amino acid degrading pathways, biosynthesis or degradation of: lipids, polyketides (including e.g. flavonoids and isoflavonoids), isoprenoids (including e.g. terpenes, sterols, steroids, carot enoids, xanthophylls), carbohydrates, phenylpropanoids and derivatives, alcaloids, benzenoids, indoles, indole-sulfur compounds, porphyrines, anthocyans, hormones, vitamins, cofactors such as prosthetic groups or electron carriers, lignin, glucosinolates, purines, pyrimidines, nucleosides, nucleotides and related molecules such as tRNAs, microRNAs (miRNA) or mRNAs. Accordingly, small molecule compound metabolites are preferably composed of the following classes of compounds: alcohols, alkanes, alkenes, alkines, aromatic compounds, ketones, aldehydes, carboxylic acids, esters, amines, imines, amides, cyanides, amino acids, peptides, thiols, thioesters, phosphate esters, sulfate esters, thioethers, sulfoxides, ethers, or combinations or derivatives of the aforementioned compounds. The small molecules among the metabolites may be primary metabolites which are required for normal cellular function, organ function or animal growth, development or health as well as for plant growth. Moreover, small molecule metabolites further comprise secondary metabolites having essential ecological function, e.g. metabolites which allow an organism to adapt to its environment. Furthermore, metabolites are not limited to said primary and secondary metabolites and further encompass artificial small molecule compounds. Said artificial small molecule compounds are derived from exogenously provided small molecules which are administered or taken up by an organism but are not primary or secondary metabolites as defined above. For instance, artificial small molecule compounds may be metabolic products obtained from drugs by metabolic path-ways of the animal. Moreover, metabolites further include peptides, oligopeptides, polypeptides, oligonucleotides and polynucleotides, such as RNA or DNA. More preferably, a metabolite has a molecular weight of 50 Da (Dalton) to 30,000 Da, most preferably less than 30,000 Da, less than 20,000 Da, less than 15,000 Da, less than 10,000 Da, less than 8,000 Da, less than 7,000 Da, less than 6,000 Da, less than 5,000 Da, less than 4,000 Da, less than 3,000 Da, less than 2,000 Da, less than 1 ,000 Da, less than 500 Da, less than 300 Da, less than 200 Da, less than 100 Da. Preferably, a metabolite has, however, a molecular weight of at least 50 Da. Most preferably, a metabolite in accordance with the present invention has a molecular weight of 50 Da up to 1 ,500 Da.

Energy metabolites as referred to herein are, preferably, metabolites of the glycolysis or the pentose phosphate pathway. Preferably, said energy metabolites are phosphorylated metabolites. The said metabolites are well known in the art (Buescher 2010, Anal Chem 82: 4403-4412; Ritter 2008, Anal Biochemistry 373: 349-369) and include, e.g., glucose, glucose-6-phosphate, Fructose-1 ,6-bisphosphate, 1 ,3-bisphosphate glycerate, 3-phospho glycerate, 2-phospho glycerate, phosphoenol pyruvate, pyruvate, glucose-6-phosphate, 6-phosphogluconolactone, 6- phosphogluconate, ribulose-5-phosphate, ribose-5-phosphate, xylulose-5-phosphate, glycer- alaldehyde 3-phosphate, seduheptulose 7-phosphate, fructose 6-phosphate, erythrose 4- phosphate, xylulose 5-phosphate, glyceralaldehyde 3-phosphate, and fructose 6-phosphate.

The term "metabolite profiling" as used herein refers to the qualitative and/or quantitative analysis of the entire detectable metabolites comprised in the sample or a subgroup or class of these metabolites. Thus, the said metabolite profiling may be broad spectrum profiling aiming at the detection of the entire detectable metabolites in a sample of cells or may be profiling of a certain type of metabolites, such as, preferably, energy metabolites or lipids. Metabolic profiles may be generated by various analysis techniques.

Preferred techniques for analyzing the metabolites and for establishing metabolite profiles are mass spectrometry-based or NM R-based techniques. Preferably, mass spectrometry is used, in particular, gas chromatography mass spectrometry (GC-MS), liquid chromatography mass spectrometry (LC-MS), ion-pair-UPLC-MS/MS, direct infusion mass spectrometry or Fourier transform ion-cyclotrone-resonance mass spectrometry (FT-ICR-MS), capillary electrophoresis mass spectrometry (CE-MS), high-performance liquid chromatography coupled mass spectrometry (HPLC-MS), quadrupole mass spectrometry, any sequentially coupled mass spectrometry, such as ESI-MS-MS or MS-MS-MS, inductively coupled plasma mass spectrometry (ICP-MS), pyrolysis mass spectrometry (Py-MS), ion mobility mass spectrometry or time of flight mass spectrometry (TOF). Said techniques are disclosed in, e.g., Niessen, Journal of Chromatography A, 703, 1995: 37-57, US 4,540,884 or US 5,397,894, the disclosure content of which is hereby incorporated by reference. Most preferably, UPLC and ESI-MS/MS is used.

As an alternative or in addition to mass spectrometry techniques, the following techniques may be used for metabolite determination: nuclear magnetic resonance (NM R), magnetic resonance imaging (MRI), Fourier transform infrared analysis (FT-IR), ultra violet (UV) spectroscopy, refraction index (Rl), fluorescent detection, radiochemical detection, electro-chemical detection, light scattering (LS), dispersive Raman spectroscopy or flame ionisation detection (FID). These techniques are well known to the person skilled in the art and can be applied without further ado.

Moreover, the metabolites can also be analyzed by a specific chemical or biological assay. Said assay shall comprise means which allow specifically detecting a metabolite in the sample. Preferably, said means are capable of specifically recognizing the chemical structure of the metabolite or are capable of specifically identifying the metabolite based on its capability to react with other compounds or its capability to elicit a response in a biological read out system (e.g., induction of a reporter gene). Means which are capable of specifically recognizing the chemical structure of a metabolite are, preferably, antibodies or other proteins which specifically interact with chemical structures, such as receptors or enzymes. Specific antibodies, for instance, may be obtained using the metabolite as antigen by methods well known in the art. Antibodies as referred to herein include both polyclonal and monoclonal antibodies, as well as fragments thereof, such as Fv, Fab and F(ab)2 fragments that are capable of binding the antigen or hapten. The present invention also includes humanized hybrid antibodies wherein amino acid sequences of a non-human donor antibody exhibiting a desired antigen-specificity are combined with sequences of a human acceptor antibody. Moreover, encompassed are single chain antibodies. The donor sequences will usually include at least the antigen-binding amino acid residues of the donor but may comprise other structurally and/or functionally relevant amino acid residues of the donor antibody as well. Such hybrids can be prepared by several methods well known in the art. Suitable proteins which are capable of specifically recognizing the metabolite are, preferably, enzymes which are involved in the metabolic conversion of the said metabolite. Said enzymes may either use the metabolite as a substrate or may convert a substrate into the metabolite. Moreover, said antibodies may be used as a basis to generate oligopeptides which specifically recognize the metabolite. These oligopeptides shall, for example, comprise the enzyme^'s binding domains or pockets for the said metabolite. Suitable antibody and/or enzyme based assays may be RIA (radioimmunoassay), ELISA (enzyme-linked immuno-sorbent assay), sandwich enzyme immune tests, electrochemiluminescence sandwich immunoassays (ECLIA), dissociation-enhanced lanthanide fluoro immuno assay (DELFIA) or solid phase immune tests. Moreover, the metabolite may also be identified based on its capability to react with other compounds, i.e. by a specific chemical reaction. Further, the metabolite may be analyzed due to its capability to elicit a response in a biological read out system. The biological response shall be detected as read out indicating the presence and/or the amount of the metabolite comprised by the sample. The biological response may be, e.g., the induction of gene expression or a pheno- typic response of a cell or an organism

The term "immobilizing" as used herein refers to fixing the cells to the solid membrane. Said fixation may be either occur as a result of adherent growth of the cells on said solid membrane or it may result from a fixation of the cells to the solid membrane by either applying a physical force or by chemical linkage. Physical force can be applied, for example, in the case of suspension cells by filtering a culture medium comprising the said suspension cells throughout a solid membrane by applying a vacuum. A subsequent low temperature treatment, e.g., shock freezing using liquid nitrogen, will further fix the cells on the solid membrane. The term "solid membrane" as used herein refers to any solid support which can be transferred between different vials and which allows cells to adhere thereon either as consequence of growth on the said solid support or as a consequence of a filtration step. Suitable membranes are well known in the art and include track edge membranes and any kind of hydrophilic or hydrophobic support suitable for the cultivation of adhering cells or filtration of suspension cells. For example, the membrane to be used in accordance with the method of the present invention is a track edge membrane. Advantageously, said membranes are destroyed during the extraction process and, therefore, release the adherent cells or filtered cells without the need of scrapping, tedious digestion, or other procedures which may interfere with efficient quenching. Preferably, said solid membrane is a polytetrafluoroethylene (PTFE) or polyvinylidene fluoride (PDVF) filter membrane and, preferably, a LUMOX membrane (Sarstedt AG & Co., Nurmbrecht, Germany).

Preferably, the pretreatment as set forth in the method or the use of the present invention, thus, does not comprise the step of scraping the cells from membrane. Also, preferably, the pretreatment does not comprise a trypsinization of the sample (in particular of the cells).

The term "washing" as used herein refers to a treatment of the immobilized cells with a washing solution such that remaining cell culture media, impurities and/or cell debris are removed. The washing shall not affect the integrity of the cells and the metabolite composition in the cells. Accordingly, said washing solution is, preferably, an isotonic NaCI solution with respect to the osmolality of the culture media used for cultivating the cells. More preferably, it is a solution having the same or essentially the same osmolality as the culture media, in particular if the cells are suspension cells. Also preferably, it is a NaCI concentration of about 0.9 %, in particular if the cells are adherent cells. The washing solution according to the present invention shall lack phosphate ions. The presence of phosphate ions in the washing solution could negatively affect the subsequent analysis of the metabolites present in the sample, in particular if the samples are analyzed by mass spectrometry. E.g. the presence of phosphate ions may inhibit the ionization of the analytes. Thus, the commonly used phosphate buffered solutions are not suitable as washing solutions according to the present invention. Washing can be carried out at least once, i.e. the method of the present invention also encompasses more than one washing steps, preferably, two, three or more washing steps. More preferably, two washing steps are carried out. The washing shall be carried out for a time period no longer than 30 seconds. Preferably, washing is carried out for 25 to 30 seconds, most preferably for about 30 seconds. Washing for this time period allows for an effective removal of the remaining cell culture media, impurities and/or cell debris without affecting the cell integrity and without affecting the metabolites present in the cells, e.g., by degradation processes.

Moreover, the washing solution, preferably, may comprise at least one ¹³C-labeled carbon source (see, e.g., Kleijn 2010, J. Biol. Chem., 285:1587-1596. If such a washing solution is applied in the method of the present invention, the said washing the immobilized cells can be extended up to a time period of 60 seconds. It has been surprisingly found in accordance with the present invention that the presence of a ¹³C-labeled carbon source, such as entirely or partially labeled ¹³C glucose, inhibits significantly the enzymatic activity of various degrading enzymes present in the sample of cells. Preferably, the washing in the presence of ¹³C-labeled carbon source the preserves metabolites, in particular ¹²C metabolites, from enzymatic metabolization. As a consequence, the washing time can be prolonged such that the remaining cell culture media, impurities and/or cell debris can be even more efficiently removed. A suitable ¹³C carbon source according to the present invention, preferably, may be a ¹³C labeled metabolite of the glycolysis. More preferably, a suitable ¹³C carbon source is a ¹³C-labled glucose. The labeling may be complete, i.e. all ¹²C atoms may be replaced by ¹³C, or it may be partial, e.g., 1 ,3,6 ¹³C glucose or 1 ,2,3 ¹³C glucose. A partially labeled ¹³C carbon source allows for using an entirely labeled ¹³C carbon source in later steps as an internal standard and vice versa since it is possible to distinguish between the said carbon sources based on the mass differences.

It will be understood that the cell culture medium may, prior to washing, be collected and stored also for metabolite profiling. In such a case, it is preferred to freeze the cell culture medium supernatant in suitable vials to a temperature of between -20 to -80°C or in liquid nitrogen.

Freezing as used herein refers to bringing the sample or solution to a temperature of -20°C or less, preferably, a temperature between -20°C and -80°C or into liquid nitrogen, i.e. to a temperature of about -196°C. Particular preferred temperatures between -20°C and -80°C are within the range of -25 to -80°C, -30°C to -80°C, -35°C to -80°C, -40°C to -80°C, -45°C to -80°C, - 50°C to -80°C, -55°C to -80°C, -60°C to -80°C, -65°C to -80°C, -70°C to -80°C and, more preferably, the temperature is -70°C, -71 °C, -72°C, -73°C, -74°C, -75°C, -76°C, -77°C, -78°C, - 79°C, or -80°C.

The term "contacting" as used herein refers to bringing the quenching solution into physical contact to the cells comprised in the sample of cells such that the components of the quenching solution can exert its activity to the cells comprised in the sample. A quenching solution to be applied in the context of the method of the present invention shall comprise dichloromethane (DCM) and ethanol (EtOH). Preferably, said quenching solution comprises DCM and EtOH in a ratio of about 9/1 1 . Such a quenching solution can be, preferably, be applied if broad spectrum profiling is envisaged. The said quenching solution may, also preferably, comprise DCM and EtOH in a ratio of about 2/1. Such a quenching solution can be preferably applied if energy metabolite profiling is envisaged. However, the quenching solution may further comprise DCM and EtOH in a ratio which differs from the aforementioned ratios. Thus, the present invention is not limited to the aforementioned ratios. Suitable ratios can be determined by the skilled person without further ado. For example, the quenching solution may comprise DCM and EtOH in a ratio of 2/1 to ratio about 1/5, or to about 1/10. Also, the quenching solution may comprise 100% ethanol. Alternatively, the quenching solution may be a solution which comprises methanol. E.g. the quenching solution may be 80% methanol (i.e. methanokwater, 80:20, v/v), acetonitrile (e.g. 80%) or other denaturating solvents.

The term "about" as used herein refers to either the precise value indicated afterwards or to a value differing +/- 20%, +/- 10%, +/- 5%, +/- 2% or +/-1 % from the said precise value. Preferably, said immobilized cells are frozen in the quenching solution to a temperature of a temperature of between -20 to -80°C or in liquid nitrogen. The step of freezing in, preferably, liquid nitrogen, may be carried out prior or after contacting the cells with the quenching solution. In case adherent cells are applied in the method of the invention, freezing is envisaged after the contacting. In case of suspension cells it is, preferably, envisaged to first freeze the immobilized cells on the solid membrane and subsequently carry out the step of contacting.

Preferably, the method of the invention may also encompass the step of extracting metabolites for metabolite profiling from the frozen immobilized cells in quenching solution. To this end, various extraction techniques may be used. For example, depending on whether a broad spectrum profiling is envisaged or whether a specific type or subset of metabolites shall be determined, different extraction techniques either covering the entirety of metabolites or the specific type or subset of metabolites can be used. The entirety of metabolites may require extraction of polar and apolar metabolites by polar and apolar extraction solutions, respectively, while the extraction of lipids, for example, may require the use of an apolar extraction solution. Extraction may also require a physical destruction of the cells of the sample. Suitable techniques are well known to the skilled artisan. Extraction can be, preferably, done by as described in the accompanying Examples, below. Moreover, suitable extraction techniques are disclosed in WO 201 1/003945, the disclosure content of which is herewith incorporated by reference.

Analysis of the extracted metabolites may be carried out by various techniques including those referred to elsewhere herein in detail.

In particular, the method of the present invention may comprise the following steps in order to pretreat a sample of suspension cells for cells for broad spectrum metabolite profiling comprising the steps of:

a) immobilizing cells on a solid membrane;

b) washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate and comprising at least one ¹³C-labeled carbon source for a time period of up to 30 seconds;

c) freezing the cells on the solid membrane; and

d) contacting the frozen immobilized cells with a quenching solution comprising dichloro- methane (DCM) and ethanol (EtOH).

In particular, the method of the present invention may comprise the following steps in order to pretreat a sample of suspension cells for cells for energy metabolite profiling comprising the steps of:

a) immobilizing cells on a solid membrane;

b) washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate for a time period of up to 30 seconds;

c) freezing the cells on the solid membrane; and

d) contacting the frozen immobilized cells with a quenching solution comprising dichloro- methane (DCM) and ethanol (EtOH). In an embodiment, the isotonic washing solution lacking phosphate as used in step b) may comprise at least one ¹²C carbon source. Preferably, the concentration of the carbon source is the same or essentially the same as in the culture medium used for culturing the cells.

In particular, the method of the present invention may comprise the following steps in order to pretreat a sample of adherent cells for cells for broad spectrum metabolite profiling comprising the steps of:

a) immobilizing cells on a solid membrane;

b1 ) washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate for a time period of up to 30 seconds;

b2) preferably, freezing the immobilized cells on the solid membrane;

c) contacting the immobilized cells with a quenching solution comprising dichloro-methane

(DCM) and ethanol (EtOH); and

c) freezing the cells on the solid membrane in the quenching solution.

In an embodiment, the isotonic washing solution lacking phosphate as used step b1 ) comprises at least one ¹³C-labeled carbon source.

In particular, the method of the present invention may comprise the following steps in order to pretreat a sample of adherent cells for cells for energy metabolite profiling comprising the steps of:

a) immobilizing cells on a solid membrane;

b1 ) washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate for a time period of up to 30 seconds;

b2) preferably, freezing the immobilized cells on the solid membrane;

c) contacting the immobilized cells with a quenching solution comprising dichloro-methane

(DCM) and ethanol (EtOH); and

c) freezing the cells on the solid membrane in the quenching solution.

In an embodiment, the isotonic washing solution lacking phosphate as used in step b1 ) may comprise at least one ¹²C carbon source. Preferably, the concentration of the carbon source is the same or essentially the same as in the culture medium used for culturing the cells.

Advantageously, it has been found in the studies underlying the present invention that a specific timing of sample pretreatment, such as washing, allows for an efficient conservation of the metabolites in a sample of cells. It is important to not exceed certain time frames in order to have a low level of metabolite degradation and an efficient removal of cell culture media, cellular debris and other contaminations. Moreover, it was found in the studies underlying the present invention that the degradation of metabolites and, in particular energy metabolites, can be further inhibited by applying a washing solution supplemented with a ¹³C carbon source. Thanks to the present invention, the reliability of metabolomics analysis of cell culture samples shall improve. The explanations and definitions of the terms made above apply mutatis mutandis for the following preferred embodiments and the other embodiments specified below.

It follows from the above that in a preferred embodiment of the method of the invention, said washing solution is an isotonic NaCI solution.

In a particular preferred embodiment of the method of the invention, said metabolite profiling is broad-spectrum profiling.

In yet a preferred embodiment of the method of the invention, said washing solution further comprises at least one ¹³C-labeled carbon source. In a more preferred embodiment of the said method, said washing the immobilized cells is extended up to a time period of 60 seconds.

In another preferred embodiment of the method of the invention, said quenching solution comprises DCM and EtOH in a ratio of 9/1 1 .

In a preferred embodiment of the method of the invention, said metabolite profiling is energy- metabolite profiling. In a more preferred embodiment of the said method, said quenching solution comprises DCM and EtOH in a ratio of 2/1 .

In a further preferred embodiment of the method of the invention, said cells are adherent cells cultivated on the solid membrane. In a more preferred embodiment of the said method, said solid membrane is a PTFE or PDVF filter membrane and, preferably, a LUMOX membrane.

In another preferred embodiment of the method of the invention, said cells are suspension cells.

In yet an embodiment of the method of the invention, said immobilized cells are frozen in the quenching solution to a temperature of -80°C or less.

In a furthermore preferred embodiment of the method of the present invention, said method further comprises the step of extracting metabolites for metabolite profiling from the frozen immobilized cells in quenching solution.

The frozen cells obtained by any of the aforementioned methods may be further treated for metabolite profiling (e.g., for broad spectrum profiling or energy metabolite profiling).

As discussed above, such a further treatment encompassed in accordance with the present invention may comprise extraction of the metabolites from the quenched, immobilized and/or frozen cells obtained by any of the methods of the invention described before.

For extraction of the metabolites, ammonium acetate buffer can be, preferably, added to the cells and metabolites can be extracted, e.g., using a bead milling process. In particular, for broad spectrum profiling, the obtained extract shall be filtered and subsequently mixed with di- chloromethane. Centrifugation of the sample is carried out in order to separate the polar and lipid phase of the sample. Aliquots are taken from both phases and subjected to metabolite profiling. In cases where energy metabolite profiling is envisaged, the cell homogenate is centri- fuged and the upper polar phase is, preferably, re-extracted in the bead mill. After a final cen- trifugation step, the upper phase shall be filtered, diluted with water, and stored for further investigation at -80°C. Details on a preferred extraction method are also to be found in

WO201 1/003945 as referred to herein before. Preferably, extraction is, thus, carried out by extracting a cell sample with an extraction buffer comprising a phase separator and a volatile neutral ammonium salt under conditions which allow for immediate disruption of cells comprised by the said sample. For further investigation, e.g., for broad spectrum or energy metabolite profiling, the extracted metabolites are separated by chromatography and analyzed by spectroscopy and, preferably, by mass spectroscopy techniques.

Moreover, the further treatment envisaged in accordance with the present invention may also encompass a normalization procedure. To this end, the total protein content in the sample aliquots and, preferably, the aliquots of the polar phase of a sample, may be determined.

In particular, depending on expected protein concentration, e.g., 5-50 μΙ_ and, more preferentially 20 μΙ_, of the aqueous polar phase from the cell extract will be aspirated and transferred to a microwell plate. For each sample, duplicate measurements are, preferably, performed on two separate plates. In addition, calibration samples containing, e.g., 10 μΙ_ of known amounts of bovine serum albumin (BSA) or other protein standards, ranging from, e.g., 0 to 250 μg/mL, shall be included in triplicate on the microplates for calculation of a calibration curve. Subsequently, the microplates will be sealed with adhesive foil and kept below freezing point of the samples (preferentially at -80 °C) until completely frozen. In order to remove excess ammonium acetate from the solutions that would compromise the protein determination reaction, the adhesive film is taken off, and the frozen samples are lyophilized overnight. Prior to the measurement, a second set of calibration samples is applied to the same microplate without freeze- drying. Protein quantification will be achieved with the bicinchoninic acid (BCA) assay, which relies on the formation and colorimetric detection of a colored complex between cuprous cations produced in the reaction of peptide bonds of proteins with alkaline Cu²⁺ (the biuret reaction), and BCA (Wiechelman 1988, Anal Biochem 175: 231 -237).

Thus, the further treatment envisaged in the accordance with the present invention, preferably, encompasses normalization of the investigated sample for total protein content.

More preferably, said normalization for total protein content comprises measuring the total protein content in an aliquot of the sample to be investigated by metabolic profiling, preferably, with the biuret reaction using BCA.

More preferably, protein standards, such as varying concentrations of BSA in solution, may be measured in parallel in order to establish a calibration curve. Preferably, such calibration curves are established for freeze-dried and non-freeze-dried protein standards. For total protein content normalization, the amount of a metabolite of interest represented by, e.g., the peak area of a corresponding peak in a mass spectrum, will be, preferably, related with the determined total protein content, e.g., by establishing the ratio of both values.

By following the above normalization procedure for total protein content, a simple and cost- effective way for normalization of raw metabolite profiling data across multiple cell culture samples is provided. Since the method uses leftover material from the extractions (i.e. aliquots of the sample to be investigated rather than control samples), it is not necessary to prepare expensive control cultures that would be otherwise needed for the estimation of cell counts. Moreover, laborious and time-consuming cell counting can be avoided. Another advantage is that normalization for total protein content based on aliquots from individual samples will account for biological variability in cell proliferation within a group treatment. Furthermore, this normalization method is also suitable for normalization of results between different group treatments.

Another normalization procedure envisaged as a further treatment according to the present invention is the median normalization. Median normalization as referred to in accordance with the present invention means that each determined metabolite value of a sample is normalized to the median of all metabolite values determined for that sample. The rationale behind that approach without being bound by theory is that the integrated peak areas of all considered metabolites should correlate with the quantity of the biological material used for the measurement.

The median normalization produces a new set of values X™^ed according to the following expression: X₁i]^ied= io(¹°g(^xij)-^medianOog(Xi))) _{> W}jth X;. = (Χϋ, χ_ί2, ... , X_im), representing the values from the i^th sample. Here, the index i = 1 ,2 n denotes the samples and j = 1 ,2 m denotes the metabolites, so that represents the pool normalized ratio of metabolite j from the sample i.

In the presence of missing values, median normalization is more robust if the median values of each individual metabolite across samples are close to each other. This can be achieved by either first normalizing each metabolite to its median across samples or by normalizing each metabolite to the values measured in pooled samples prior to the median normalization of the samples.

Median normalization could be used, e.g., when cell number or protein data are not available. In cases where either treatment or inherent biological characteristics of different cell lines influence cell number, median normalization still could facilitate reliable and sound statistical comparisons, since it might adjust inter-group contrast to a more realistic level. However, lack of sample pre-dilution would still lead to undesirable matrix-suppression effects and linearity issues, especially when treatment and/or cell line effect differences are relatively high. Nevertheless, such effects cannot be adjusted for by median normalization.

In the case that large differences in medians between different groups of samples are observed (e.g. very different cell lines or very different cell culture medium) due to different correlations between protein content/cell number and median, and the preserving of this offset is desired, the adjusted median normalization can be performed. To this end, an adjustment parameter is calculated and introduced into median normalization to re-include the group difference. In cases where data on cell number/protein is available it could be used, e.g., for either pre-dilution of samples or alternatively for protein/cell number normalization after measurement. Pre-dilution steps would reduce potential linearity issues. Though, the time and effort for pre-dilution can be quite high for large sample numbers.

Often protein content/cell numbers are only known for parallel cultures. Then the normaliza- tion/pre-dilution would only apply one value for each group of samples and thus, correct for the inter-group differences but not for intra-group variability. In this situation, median normalization would be a more suitable approach to reduce intra-group variability.

Thus, the further treatment envisaged in the accordance with the present invention, preferably, encompasses median normalization of the investigated sample. Preferably, said median normalization comprises the steps of:

a) determining the amount of the metabolites comprised in the sample;

b) identifying the median value among the amounts of the metabolites determined in step a); and

c) relating the amount of a metabolite of interest (i.e. the metabolite to be normalized) to the median value, preferably, by establishing a ratio of both values, whereby the normalization of the metabolite of interest is achieved.

Preferably, the median normalization produces X-J^iedas described above.

Moreover, the present invention, in general, provides for a method of metabolite normalization comprising the steps of:

a) determining the amount of the metabolites comprised in a sample;

b) identifying the median value among the amounts of the metabolites determined in step a); and

c) relating the amount of a metabolite of interest (i.e. the metabolite to be normalized) to the median value, preferably, by establishing a ratio of both values, whereby the normalization of the metabolite of interest is achieved.

Moreover, the present invention, in general, provides for a method of metabolite profiling comprising the steps of:

a) extracting metabolites from a sample

b) determining the amount of the metabolites comprised in a sample;

c) identifying the median value among the amounts of the metabolites determined in step b); and

d) relating the amount of a metabolite of interest (i.e. the metabolite to be normalized) to the median value, preferably, by establishing a ratio of both values, whereby the normalization of the metabolite of interest is achieved,

e) calculating the normalized amount(s) of the metabolite(s) of interest. The sample in the context of the aforementioned method of median normalization and the aforementioned method of metabolite profiling may be any biological or environmental sample containing metabolites. Preferably, the sample is a biological sample such as a sample derived from or comprising organisms, such as animals, plants, microorganisms or cultured cells as specified elsewhere herein (e.g. cells immobilized on a solid membrane). In an embodiment, the sample is derived from a body fluid. In particular, the body fluid is blood, serum, plasma, lymph, saliva, cerebrospinal liquid, sudor, sperm, vaginal fluid, tears, faeces or urine. Preferably, the body fluid is derived from a vertebrate, more preferably, from a mammal and, most preferably, from a human. If the sample is derived from a plant, the sample can be any plant tissue or extract including root, stem, leaf, or seed tissue.

Preferably, the median normalization produces X-J^iedas described above.

In connection with the method of median normalization and with the method of metabolite profiling it is not necessary to determine the amount of all metabolites comprised by the sample and to determine the median value among all determined amounts. Preferably, the median value to be identified is the median value among the amounts of at least 50 %, more preferably, of at least 60 %, and most preferably, of at least 70 % of the determined metabolites. This applies, in particular if the amount of a large number of metabolites is determined (in particular for broad spectrum profiling). If the amount of a limited number of metabolites is determined (e.g. less than 500 or 300, e.g. for energy metabolite profiling), it is envisaged that the median value to be identified is the median value among the amounts of at least 80 %, more preferably, of at least 90 %, and most preferably, of at least 95 % of the determined metabolites. However, the percentage may be lower (e.g. 70 %), if the amount of each metabolite would be normalized to the amount determined in pooled samples prior to median normalization (as described above).

The invention further encompasses, in general, the use of at least one ¹³C-labeled carbon source for pretreating a sample of cells for metabolite profiling. Preferably, said pretreating encompasses the preserving metabolites, in particular ¹²C metabolites, from enzymatic metaboli- zation. Preferably, said cells are immobilized on a solid membrane.

Moreover, the present invention relates to the use of a ¹³C-labeled carbon source for prolonging the washing time of a sample of cells for metabolic profiling. Preferably, said cells are immobilized on a solid membrane.

All references referred to above are herewith incorporated by reference with respect to their entire disclosure content as well as their specific disclosure content explicitly referred to in the above description.

FIGURES Fig. 1 shows a time course experiment detecting the changes of a ¹²C analyte in the presence or absence of ¹³C glucose in the washing solution. It is evident that up to 60 seconds no changes are observed in the presence of ¹³C glucose in the washing solution.

Fig. 2 shows a schematic sampling protocol for a suspension cell culture sample.

Fig. 3 shows a schematic representation of the two succeeding washing steps of the excised LU MOX membrane in isotonic NaCI solution in a 6-well plate.

Fig. 4 shows the influence of different sampling protocols on the abundance of selected energy metabolites observed in eukaryotic cells grown in suspension cultures. Fast filtration was conducted as described in the text. For excess solvent quenching 5 ml. of the cell culture were mixed with 20 ml. of ice-cold quenching solution containing 60% (v/v) methanol and NaCI. The cells were pelleted by centrifugation at 1000xg for 1 min and washed with an isotonic NaCI solution containing glucose. Harvesting of cells using the fast filtration method results in significantly higher levels of most energy metabolites compared to the traditional excess solvent quenching protocol. This advantage is due to the fast and gentle separation of cells from the culture medium by fast filtration, preserving cell integrity and metabolic state.

Fig. 5 shows the results of a time course experiment investigating the metabolic response of eukaryotic cells, grown in suspension cultures, to different sampling speeds. The levels of the measured energy metabolites remain stable for about 30 seconds, extending the filtration time to 60 or even 300 seconds leads to a significant decrease in metabolite abundance due to rapid intracellular turnover.

Fig. 6 depicts intracellular levels of fructose-1 ,6-diphosphate in prokaryotic cells depending on the applied sampling method. Due to quick and efficient sampling, the highest levels of this early glycolysis metabolite were obtained with fast filtration.

Fig. 7 depicts intracellular levels of glutamine in prokaryotic cells depending on the applied sampling method. This metabolite is also a media component, which is efficiently depleted by fast filtration, whereas high levels are observed using methods without a washing step due to contaminating glutamine from the adherent medium.

Fig. 8 shows (A) a correlation of sample medians from cell lysates of 15 different cell lines vs. the respective protein amount, (B) Schematic description of the impact of median normalization on intra- and inter-group variability in cases where treatment and cell line effects on metabolic profiles can be expected, and (C) the calculated effect of median normalization on intra-group variability for the same 15 cell lines as used for Fig. 8A.

Fig. 9 shows a schematic illustration of the LUMOX procedure and two alternative cell sampling methods that were applied to two different breast cancer cell lines. Shown are the duration to perform a single step of the protocol (time in grey) as well as required total times (in black) to complete the entire sampling protocol. Fig. 10 shows an overview of metabolic changes the of main energy-generating pathways including glycolysis, TCA cycle, and potential urea cycle precursors/intermediates including indicators of connection points with amino acid metabolism and lipid synthesis/oxidation. Dashed lines implicate enzymatic reactions which were omitted for the sake of clarity. Grey nodes and labels indicate intracellular metabolites which were not evaluated. Significantly different metabolite levels between samples harvested with TRYPSIN method versus LUMOX method (for both cell lines) are marked with labels, i.e. down means lower metabolite levels and up means higher levels in trypsin-harvested samples. Metabolites without labels were not significantly different between both methods. The labels are colored according to the p-value with black for p<0.01 and grey for p<0.05.

EXAMPLES

The following Examples shall merely illustrate the invention. They shall, whatsoever, not be construed to limit the scope in any respect.

Example 1 : Sampling of cells and supernatant from adherent cell cultures for metabolome analysis a) Sampling for MxP^e Broad profiling

At least 1 Mio eukaryotic cells were required per profiling sample. The adherent cells were cultivated on a LUMOX plate with a filter membrane at the bottom.

Before sampling, the quenching solution (DCM/EtOH 9:1 1 ) and the polypropylene vials were pre - cooled on dry ice or liquid nitrogen. 6-well plates are prepared and 5 mL isotonic NaCI solution (0,9%, pre-warmed to 37°C) is added into each well.

Before sampling of the adherent cells, 1 mL of supernatant was sampled into a polypropylene vial and floating cells or cell debris removed by a quick centrifugation step (e.g. 900 rpm at 4°C). The supernatant was transferred into a new polypropylene vial and frozen in liquid nitrogen.

Two LUMOX plates with remaining supernatant, covering the membrane, were set each onto the first well of a row of the 6-well plate. The membrane of the LUMOX plate was cut, so that the membrane and the remaining supernatant were falling into the washing solution. Following, the washing of the membrane was conducted by dipping the membrane into the 2 washing solutions of one row (Fig. 3). Subsequently, the membrane cutting was transferred into a pre-cooled polypropylene vial, 600 μί of pre-cooled quenching solution added and the polypropylene vial frozen in liquid nitrogen.

The cutting of the membrane until freezing should not take more than 30s. b) Sampling for MxP^® Energy

At least 1 Mio eukaryotic cells were required per profiling sample. The adherent cells were cultivated on a LUMOX plate with a filter membrane at the bottom.

Before sampling, the quenching solution (DCM/EtOH 2:1 ) and the polypropylene vials were pre - cooled on dry ice or liquid nitrogen. 6-well plates are prepared and 5 ml. isotonic NaCI solution (0,9%) containing 4,5 g/l Glucose (pre-warmed to 37°C) is added into each well.

Two LUMOX plates with remaining supernatant, covering the membrane, were set each onto the first well of a row of the 6-well plate. The membrane of the LUMOX plate was cut, so that the membrane and the remaining supernatant were falling into the washing solution. Following, the washing of the membrane was conducted by dipping the membrane into the 2 washing solutions of one row (Fig. 3). Subsequently, the membrane cutting was transferred into a pre-cooled polypropylene vial, 900 μί of pre-cooled quenching solution added and the polypropylene vial frozen in liquid nitrogen. The cutting of the membrane until freezing should not take more than 30s.

Example 2: Sampling of cells and supernatant from suspension cell cultures for metabolome Analysis

At least 1 Mio eukaryotic cells or 1 billion prokaryotic cells were required per profiling sample.

Before sampling, the quenching solution (DCM/EtOH 9:1 1 for MxP^® Broad Profiling and

DCM/EtOH 2:1 for MxP^® Energy) and the corresponding polypropylene vials were pre - cooled on dry ice or liquid nitrogen.

The filter funnel with the activated filter (presoaked in EtOH) was placed on the vacuum manifold and the necessary sample culture volume was added to the filter funnel (not less than 1 mL). The broth was filtered through the filter by opening the 3 port stopcock and by applying a vacuum force of 35 mbar. The supernatant (filtrate) was collected in a polypropylene vial and immediately frozen in liquid nitrogen. After the supernatant has passed the filter, the 3-port stopcock was switched to the wash container and washing solution in equal volume of the filtered culture broth was added to the filter funnel. The washing solution was discarded.

In case of MxP^® Broad Profiling the washing solution consisted of uniformly labeled [U-¹³C6- Glucose] and NaCI.

In case of MxP^® Energy the washing solution consisted of ¹²C-Glucose and NaCI.

In both cases, the volume of the washing solution was equal to the overall sampled volume of all individual samples. Likewise, the ionic strength (i.e. concentration of NaCI) of the washing solution was identical to the ionic strength of the culture media at the sampling time point. If the culture media contained glucose as carbon source with a concentration of 4,5 g/l at the time point of sampling, the amount of glucose in the washing solution was adjusted to 4,5 g/l.

After the washing solution has passed the filter, the filter funnel was disconnected from the vacuum manifold and liquid nitrogen was added to the filter membrane. When the liquid nitrogen was evaporated, the filter was transferred to a new pre-cooled polypropylene vial and pre- chilled quenching solution was added to the vial (600 μΙ_ DCM/EtOH in a ratio 9:1 1 for MxP^® Broad Profiling or in case of MxP^® Energy 900 μΙ_ DCM/EtOH in a ratio of 2:1 ). The polypropylene vial was then frozen in liquid nitrogen.

In both cases (MxP^® Broad Profiling, MxP^® Energy) the steps of filtering and washing of the filter until the first freezing step with liquid nitrogen should not take more than 30s.

In order to demonstrate the superior efficiency of the fast filtration method with respect to stopping metabolism and keeping cells intact during their separation from the medium, this sampling procedure was compared to the traditional excess solvent quenching protocol. Eukaryotic cells grown in suspension culture were harvested using either fast filtration or excess solvent quenching and the abundance of intracellular metabolites was quantified.

The amount of most energy metabolites, in particular AMP, ADP, and ATP, recovered in the cells was significantly increased when fast filtration was used for sampling compared to excess solvent quenching (Fig. 4). This is due to the fact that the use of methanol during excess solvent quenching leads to leakage and loss of intracellular metabolites into the culture medium. Viability measurements of cells treated with the excess solvent quenching solution show a reduction by about 50%, clearly supporting this observation by demonstrating severe cell damage. Moreover, the cellular energy charge, as determined from the concentrations of AM P, ADP, and ATP according to Atkinson and Walton (1967), was calculated as 0.81 for cells sampled with fast filtration (in the published range of most biological cells) compared to 0.69 for excess solvent quenching. Another positive aspect of fast filtration is improved quality of the obtained data set due to an approximately twofold reduction in biological variability as compared to excess solvent quenching.

Fig. 5 shows the results of a time course experiment using fast filtration demonstrating a decrease in the abundance of most energy metabolites after 30 seconds of sampling. These results underline the need for a fast and efficient way of cell sampling; otherwise some energy metabolites can be lost due to rapid intracellular turnover.

The fast filtration protocol was also compared with other sampling methods using prokaryotic suspension cells. The fast filtration method was applied to 5 ml. of an E. coli culture, with subsequent washing of the retained cells with 5 ml. washing solution containing 4.5 g/L U-13C- glucose and 0.9% NaCI.

For excess solvent quenching 5 ml. of culture were mixed with 20 ml. of a 0.9% NaCI solution and centrifuged at 4000 rpm for 1 min at -5°C. Harvesting of the cells using centrifugation was achieved by spinning 5 mL of culture for 1 min at 4000 rpm with or without subsequent washing of the cell pellet with 5 mL of a 0.9% NaCI solution.

Excess solvent quenching leads to leakage and loss of intracellular metabolites. Centrifugation and washing with NaCI only, results in fast depletion of early glycolysis metabolites like Fruc- tose-1 ,6-bisphosphate, which has a rapid intracellular turnover rate (Fig. 6). Without washing, media components like glutamine contribute to intracellular levels due to the adherent medium. Excess solvent quenching results in contribution of media components to intracellular metabolite levels due to the loss of cellular membrane (cell wall) integrity (Fig. 7).

Example 3: Metabolite analysis a) MxP® Broad Profiling

The metabolites were extracted from supernatant or cellular material fixed to the membrane by the addition of 10 μί Ammonium acetate buffer (10,0 g Ammonium acetate / 25 mL H2O) 350μί h O, 100 iL internal standard lipid and 100 μί internal standard polar and steel beads to the polypropylene vial containing the supernatant or the membrane. Cell rupture, protein denaturing, and metabolite extraction was achieved in one step within 5 min via a bead milling process. The extract was transferred to a spin filter and the filtrate collected via spin filtration for 5 min at 12000 rpm. 100 μΐ of DCM was directly added to the filtrate, agitated for 5 min at 1400rpm. The phase separation was accomplished via centrifugation for 5 min at 12000 rpm. The upper and lower phases were subjected to MxP® Broad profiling.

For MxP® Broad profiling all the metabolites of the extract were measured by gas chromatog- raphy-mass spectrometry and liquid chromatography-mass spectrometry. Samples collected as described in the previous chapter were prepared and subjected to metabolite profiling by gas chromatography-mass spectrometry (GC-MS) and liquid chromatography-mass spectrometry (LC-MS/MS) analysis. Metabolites from the supernatant and the membrane with fixed cells from suspension cell culture or adherent cell culture were extracted with polar (water) and non-polar (ethanol/dichloromethane/acetonitrile) solvents. The extract was fractioned into an aqueous, polar phase (Polar fraction) and an organic, lipophilic phase (lipid fraction). For GC-MS analyses, the lipid fraction was treated with methanol under acidic conditions to yield the fatty acid methyl esters derived from both free fatty acids and hydrolyzed complex lipids. Further, components of the lipid backbone (i.e. glycerol) were quantified in the non-polar phase (carrying the term "lipid fraction" following the metabolite name). For example, "glycerol, lipid fraction" represents glycerol liberated from complex lipids - in contrast, "glycerol, polar fraction" represents glycerol present originally in the polar phase of the biological sample. The polar and lipid fractions were further derivatized with O-methyl-hydroxylamine hydrochloride (20 mg/mL in pyridine, 50 iL) to convert oxo-groups to O-methyloximes and subsequently with a silylating agent (MSTFA, 50 μί) before GC-MS. For LC-MS/MS analyses, both fractions were reconstituted in appropriate solvent mixtures. High performance liquid chromatography (H PLC) was performed by gradient elution using methanol/water/formic acid on reversed phase separation columns. Mass spectrometric detection technology was applied as described in the patent US 7196323, which allows targeted and high sensitivity "Multiple Reaction Monitoring" profiling in parallel to a full screen analysis.

The samples were analyzed in semi-randomized analytical sequence design with pooled samples (= "pool") generated from extra samples provided for this purpose. The raw peak data were normalized to the median of pool per analytical sequence to account for process variability (so called "ratios versus pool"). b) MxP^® Energy

The metabolites were extracted from supernatant or cellular material fixed to the membrane by the addition of an extraction buffer (100 μΙ_ 1 ,5 M ammonium acetate buffer, 4 °C) an isotopical- ly labeled cell extract (¹³C- carbon source, 4 °C) and steel beads to the polypropylene vial containing the supernatant or the membrane. Cell rupture, protein denaturing, and metabolite extraction was achieved in one step within 30 seconds via a bead milling process under cryogenic conditions (using a FastPrep24 device, MP biomedicals Inc.). Phase separation was accomplished via centrifugation for 2 minutes at 14000 rpm and 4°C. The upper phase was optionally transferred to a new polypropylene vial, 150 μΙ_ of 1 .5 M ammonium acetate added and the bead milling step as described above repeated for 30 seconds. Phase separation was again accomplished via centrifugation for 2 minutes at 14000 rpm and 4°C. The upper phase was transferred to a spin filtration vial and the filtrate was collected via spin filtration for 5 minutes at 14000 rpm and 4°C. An aliquot of the filtrate was diluted with water, frozen at -80 °C, and lyoph- ilized subsequently.

For the chromatographic separation of phosphorylated and/or carboxylated polar metabolites ultra-high pressure ion pairing liquid chromatography IP-UPLC was applied. A chromatographic gradient between a solvent A (deionized water) and a solvent B (50% acetonitrile, 50% water (v/v)) was processed, whereas a constant column flow of 0.4 mL/min and a column oven temperature of 45 °C was maintained. As a volatile additive thbutylamine was added to both eluents and the pH value was adjusted to a pH value of 6.2 using glacial acetic acid to form the ion pairing tributylammonium cation. The lyophilized sample was dissolved in a low volume of eluent A before injection of 1 to 20 μΙ_ of the extract.

Negative mode electrospray tandem mass spectrometry (-ESI-MS/MS) was used to assess the polar metabolites separated by U PLC. The tandem MS/MS was operated in the so-called scheduled or selected multiple reaction monitoring mode (sMRM) whereas unique mass adjustments with unit resolution were defined. Isotopically labeled and non-labeled forms of individual metabolites were distinguished by different mass traces. The response of each metabolite present in the sample as ¹²C was normalised to the response of the equivalent ¹³C metabolite from the isotopically labeled cell extract. Example 4: Median normalization

The sample based medians from 15 different cell lines obtained from broad profiling were correlated versus the measured protein amount for each cell line (Fig. 8A). As apparent, a good correlation between the median and the protein amount is obtained (R² = 0.75). This suggests that the median is an adequate parameter to correct for differences in cell number/protein content within a biological group (intra-group variability due to slight differences of cell number/protein content between biological replicates). Furthermore, differences in cell number/protein content between biological groups can also be adjusted for to give more realistic inter-group contrasts (resulting from either treatment effects or differences between cell lines) (Fig. 8B). In fact, reduced intra-group variability following median normalization can be determined in the same da- taset for 1 1 out of 15 cell lines (Fig. 8C).

Median normalization was produced by calculating X™^ed according to the following sion: X₁i]^ied= io(¹°g(^xij)-^medianOog(Xi))) _{> wit}h χ.. = ... , X_im), representing the values from the i^th sample. Here, the index i = 1 ,2 n denotes the samples and j = 1 ,2 m denotes the metabolites, so that represents the pool normalized ratio of metabolite j from the sample i.

Applicability of median normalization in different situations:

Median normalization will be used when cell number or protein data is not available. In cases where either treatment or inherent biological characteristics of different cell lines influence cell number, median normalization still could facilitate reliable and sound statistical comparisons, since it might adjust inter-group contrast to a more realistic level. However, lack of sample pre- dilution would still lead to undesirable matrix-suppression effects and linearity issues, especially when treatment and/or cell line effect differences are relatively high. Nevertheless, such effects cannot be adjusted for by median normalization.

In cases where data on cell number/protein is available it will be used for either pre-dilution of samples or alternatively for protein/cell number normalization after measurement. Pre-dilution steps would reduce potential linearity issues. Though, the time and effort for pre-dilution can be quite high for large sample numbers. Often protein content/cell numbers are only known for parallel cultures. Then the normalization/pre-dilution would only apply one value for each group of samples and thus, correct for the inter-group differences but not for intra-group variability. In this situation, median normalization would be a more suitable approach to reduce intra-group variability.

The presented results show that the median of all metabolites per sample can be a useful parameter for data normalization. However, although sample based medians correlate with cell number/protein content, it has to be considered that the median does not only contain information on cell number/protein content, but could also depend on treatment effects as well as on inherent biological characteristics of different cell lines. Example 5: Comparison of the LU MOX sampling method with trypsinization and mechanical scraping of adherent cells

In order to evaluate the impact of the sampling method on the metabolome of adherent cells, the LU MOX procedure was compared to the classical trypsinization and mechanical scraping methods using two different breast cancer cell lines, respectively (Fig. 9).

For M BA-M B-231 250.000 cells and for MCF7 300.000 cells were seeded in 2 mL RPM I-1640 medium (10% FBS (fetal bovine serum), 2 mM glutamine, 100 μg/mL penicillin/streptomycin) and incubated for 48h under standard growth conditions (37°C, 5% C02) reaching 80% confluence. Cells were grown on LU MOX dishes (sampling method LU MOX) or standard 6 well plates (sampling method TRYPSI N and SCRAPI NG).

TRYPSIN method for MxP® Broad Profiling and MxP® Energy: Supernatant was removed, cells were washed once with PBS and incubated for 5 min with 0.2 mL 0.05% Trypsin / 0.02% EDTA solution at 37°C. Two volumes of fresh medium (0.4 mL) were added. Detached cells were re- suspended, counted, and centrifuged (900 rpm, 4°C, 3 min). These extra supernatants were collected for investigation by metabolomics. Cell pellets were washed twice with PBS (4°C) and finally snap-frozen in liquid nitrogen.

SCRAPI NG method for MxP® Broad Profiling and MxP® Energy: Supernatant was removed, cells were washed twice in ice-cold 5% maltose water (H PLC grade) and quenched with 1 mL 100% methanol (H PLC grade). The cells were detached with a cell scraper and the methanol cell solution was transferred into a tube and snap frozen in liquid nitrogen.

LU MOX method for MxP® Broad Profiling: Supernatant was removed, the membrane was cut out, dip-washed twice in isotonic NaCI (room temperature), pushed into a tube with 600 μί quenching solution (9: 1 1 (v/v) dichlormethane (DCM) ethanol, precooled on dry ice), and snap- frozen in liquid nitrogen.

LU MOX method for MxP® Energy: Supernatant was removed, the membrane was cut out, dip- washed twice in isotonic NaCI containing 4.5 g/L glucose (room temperature), pushed into a tube with 900 iL quenching solution (2: 1 (v/v) dichloromethane (DCM) ethanol, precooled on dry ice), and snap-frozen in liquid nitrogen.

All samples were vented for 10 min on dry ice to remove excess liquid nitrogen and stored at - 80°C. Per cell line, cell culture sampling method and metabolomics platform six biological replicates were collected and investigated.

As depicted in Fig. 10, the considerably longer handling time of the TRYPSI N method leads to an energy-depleted state of the cells. Many glycolysis and TCA cycle intermediates exhibit lower levels, simultaneously ATP levels were lower while ADP levels increased, when samples were collected using the TRYPSI N method.

Metabolic analysis of the culture extra supernatants (collected after fresh medium was added to cells to inhibit activity of trypsin) revealed that cell membrane intactness was strongly compromised due to the trypsinization treatment. Several metabolites known to be preferably intracellu- lar were found to be enriched in the extra supernatant, i.e. glucose-6-phosphate, NAD, IPP, fumarate while metabolites typically excreted as waste products were taken up, i.e. uric acid.

Furthermore, the mechanical scraping of adherent cells in solvent, wash buffer or quenching solution is connected with several disadvantages. First, the washing of plates, typically by repeated aspiration and addition of washing solutions, is more time-consuming than the here proposed dip washing. Depending on the number of washing steps and washing solutions, perturbations of intracellular metabolism will occur. Washing without carbohydrate sources results in a depletion of the early steps of glycolysis and washing with carbohydrates sources other than glucose will lead to a shift in early metabolic steps of the selected carbohydrate. Additional remaining residual amounts of the washing solution are higher on plates than after dip washing of foils and can further perturb subsequent extraction or measurements (e.g. ion suppression). Second, the often necessary removal or evaporation of scraping solvents such as Methanol to allow efficient extraction leads to a degradation of labile metabolites, e.g. polyunsaturated fatty acids or phosphates. This results for example in a substantial decrease of nucleotide triphosphates and in an increase in degradation products such as nucleotide mono- or diphosphates. Third, many adherent cells are so strongly attached that especially cell membrane lipids on the side of attachment are not fully detached. Additionally, depending on the scraping solvent and the dish/scraper material lipids can adsorb to the dish/scraper leading to a substantial loss of lipids. Additionally, differences in efficiency to scrape and transfer cells into the collection tube will increase variability of metabolomics data when "external" normalization factors such as cell numbers or protein concentration of a corresponding replicate cell dish will be used.

In conclusion, the TRYPSIN and SCRAPING methods affect the intracellular metabolome to such great extent that it results in misleading metabolite levels and subsequent misleading interpretation of activity of different central metabolic pathways relevant in numerous eukaryotic cells.

Claims

Claims

1 . A method for pretreating a sample of cells for metabolite profiling comprising the steps of: a) immobilizing cells on a solid membrane;

b) washing the immobilized cells on the solid membrane with an isotonic washing solution lacking phosphate for a time period of up to 30 seconds; and

c) contacting the immobilized cells with a quenching solution comprising dichloro- methane (DCM) and ethanol (EtOH).

2. The method of claim 1 , wherein said washing solution is an isotonic NaCI solution.

3. The method of claim 1 or 2, wherein said metabolite profiling is broad-spectrum profiling.

4. The method of claim 3, wherein said washing solution further comprises at least one relabeled carbon source.

5. The method of claim 4, wherein said washing the immobilized cells is extended up to a time period of 60 seconds.

6. The method of any one of claims 3 to 5, wherein said quenching solution comprises DCM and EtOH in a ratio of 9/1 1 .

7. The method of claim 1 or 2, wherein said metabolite profiling is energy-metabolite profiling.

8. The method of claim 7, wherein said quenching solution comprises DCM and EtOH in a ratio of 2/1 .

9. The method of any one of claims 1 to 8, wherein said cells are adherent cells cultivated on the solid membrane.

10. The method of claim 9, wherein said solid membrane is a PTFE or PDVF filter membrane and, preferably, a LUMOX membrane.

1 1 . The method of claims 1 to 8, wherein said cells are suspension cells.

12. The method of any one of claims 1 to 1 1 , wherein said immobilized cells are frozen in the quenching solution to a temperature of -80°C or less.

13. The method of any one of claims 1 to 12, wherein said method further comprises the step of extracting metabolites for metabolite profiling from the frozen immobilized cells in quenching solution.

14. Use of a ¹³C-labeled carbon source for prolonging the washing time of a sample of cells for metabolic profiling.

15. Use of at least one ¹³C-labeled carbon source for pretreating a sample of cells for metabolite profiling.

16. The use of claim 15, wherein said pretreating encompasses the preserving metabolites from enzymatic metabolization.

17. The use of any one of claims 14 to 16, wherein the cells are immobilized on a solid membrane.

18. The use of any one of claims 14 to 17, wherein the cells are washed in an isotonic washing solution, in particular an isotonic NaCI solution.

19. The use of any one of claims 18, wherein the washing solution lacks phosphate.