Water is a major component of all cell culture media and is therefore needed to prepare media, buffers and additives, as well as to serve many ancillary functions, such as heating, cooling, cleaning and rinsing. Thus, water quality plays an important role in the outcome of cell culture experiments. Contaminants in water used for cell cultures can occur in many forms, such as bacteria, yeasts or moulds. These contaminants are usually visible to the eye or by optical microscopy. However, contamination from chemicals or other biological agents may also affect the growth, morphology or behaviour of cultured cells, yet be undetectable to the unaided eye. Water used in cell cultures must therefore be free of microorganisms and, in particular, of endotoxins, inorganic ions (heavy metals such as lead or zinc, for example), and organic compounds (humic acids, tannins and pesticides). For more detailed information, please refer to the reference literature [1, 2]. Examples of typical impurities in mains water (tap water) and target values for cell culture work are shown in Table 1.
|Parameter||Mains water||Water for cell culture||% reduction|
|Conductivity (µS/cm)||50 to 900||0.2||99.95|
|Calcium (mg/l)||20 to 150||< 0.01||> 99.99|
|Sodium (mg/l)||20 to 150||< 0.01||> 99.99|
|Iron (mg/l)||0.01 to 0.1||< 0.001||> 98|
|Bicarbonate (mg/l)||30 to 300||< 0.01||> 99.99|
|Chloride (mg/l)||10 to 150||< 0.01||> 99.99|
|Sulphate (mg/l)||1 to 100||< 0.01||> 99.98|
|TOC (mg/l)||0.2 to 5||0.1||96|
|Free chlorine (mg/l)||0.1 to 0.5||< 0.01||> 97|
|Bacteria (CFU/100 ml)||100 to 1000||< 10||> 98|
|Endotoxin (IU/ml)||1 to 10||< 0.1||> 98|
|Turbidity||0.1 to 2||< 0.01||> 99|
The objective of the present test series was to evaluate whether pure water produced by Arium® pro UF can be readily used for cell culture applications without entailing any problems. In this study, PER.C6 EpCAM cells were cultivated in ready-made CDM4PERMab (Hyclone) media employed as controls, as well as in CDM4PERMab (Hyclone) powder media prepared using ultrapure water obtained with Arium® pro UF and RO water, respectively, for test purposes. The results of each culture were then used to assess whether water from Arium® pro UF is suitable for use in the cultivation of PER.C6 EpCAM cells. The PER.C6 cell line derived from human retinoblast cells described and employed in our test series is also used today for the expression of recombinant proteins and monoclonal antibodies and for the manufacture of therapeutic proteins and monoclonal antibodies.
Ultrapure Water System
The Arium® pro UF system (see Figure 2) is designed to produce ultrapure water from pre-treated water sources by removing trace levels of residual contaminants. Production of ultrapure water requires continuous recirculation and constant flow. In Arium® pro UF, this is carried out by a pump system that controls the pressure. The system measures the conductivity of the water at both the water feed inlet and the water outlet (downstream product).
The actual purification process depends on the Arium® type and technology used. The Arium® pro UF system works with two different cartridge kits. These cartridges are filled with a special active carbon adsorber and a special mixed-bed exchange resin designed to deliver high-purity water with low extractables. A final microfilter at the outlet is normally installed to remove any particles or bacteria from the ultrapure water as it is dispensed. The general process described for water purification with the Arium® pro UF system is depicted in Figure 3. For the tests, the feed water for the Arium® pro UF device (predecessor model with the same technical design as that of the current Arium® pro UF device shown in Figure 2) was pre-filtered by an Arium® RO reverse osmosis system (predecessor model of the Arium®advance). This configuration is in accordance with Whitehead  for water purification treatment for small scale cell cultivation in laboratories.
Materials and methods
PER.C6 EpCAM cells (see Figure 1) were cultivated in T-75 flasks, with vented caps (Nunc) for gas exchange, in duplicates (12ml media in each flask) for 10 passages and in 125ml spinner flasks (Wheaton, VWR) with 50ml of media in duplicates. The PER.C6 EpCAM cell line was cultivated in CDM4PERMab ready-made media (Hyclone) and in CDM4PERMab powder media (Hyclone). The CDM4 PERMab powder media was reconstituted with either Arium® pro UF water or with RO water, along with 4mM L-glutamine (Lonza), sodium bicarbonate (3.2g/L, Merck) and pluronic acid F-68 (0.5g/L, Sigma), and filtered through a final 0.2µm sterilising-grade filter using 1000ml disposable filtration units (Sartolab, Sartorius) under aseptic conditions. The cells were seeded at a seeding density of 0.3x106 cells/ml in T-75 flasks and 0.7x106 cells/ml in spinner flasks. The T-flasks and spinner flasks were incubated in a CO2 incubator (Forma direct heat CO2 incubator, Model 3-11 Thermo Scientific) at 37°C, 5% CO2 and 85% humidity. In the CO2 incubator, the spinner flasks were incubated on a magnetic stirrer (VWR) at 80 rpm with the side arm of each spinner flask loosely capped to facilitate gas exchange inside. Samples were taken every day except for weekends (days 4 and 5) from the spinner flasks and every third day from the T-flasks to determine the viable cell density. The viable cell density was measured according to the Trypan Blue exclusion method using a heamocytometer (Vasa Scientific). Comprehensive information about the basic techniques of cell cultivation is given in the reference literature .
|Controls (ready-to-use media)||Media reconstituted with Arium® water||*Media reconstituted with RO water|
|Cell count (x106 cells/ml)||Cell count (x106 cells/ml)||Cell count (x106 cells/ml)|
* The data for RO water were obtained using the predecessor model (AriumR RO) of the current AriumR advance system.
Results and discussion
The average cell density yield in the control T-flasks (cells cultivated in ready-made media as controls) was 1.52x106 cells/ml (see summary in Table 2) and the average viability for these controls was 95.23% (see Figure 4). For Arium® T-flasks (cells cultivated in media reconstituted with ultra pure water from the Arium®pro UF system) the average cell density was 1.73x106 cells/ml, and an average viability of 95.7% was obtained. In comparison with these results, an average cell density yield of 1.68x106 cells/ml and a viability of 95.59% (Table 2 and Figure 4) were obtained in the RO T- flasks (cells cultivated in media reconstituted in RO water). In another experiment, the PER.C6 EpCAM cell lines were cultivated in spinner flasks with ready-made media used as controls, with media reconstituted using water purified by Arium® pro UF and with media reconstituted using RO water (RO) . see Figure 5. The maximum cell density obtained in control spinner flasks was 5.42x106 cells/ml with an 88.47% viability, 6.24x106 cells/ml with an 88.55% viability in Arium water spinner flasks and 4.60x106 cells/ml with an 89.85% viability in RO water spinner flasks on day 6 of cultivation (Figure 5). Microscopic examination of the cells in the RO spinner flasks indicated that cells grown in media reconstituted with RO water look unhealthy compared with cells grown in the readymade media and in media reconstituted with water purified by the AriumR pro UF system. The viability of the cells grown in spinner flasks with media reconstituted in RO water decreased more rapidly compared with those cultivated in the ready-made media and media reconstituted with ultrapure water produced by the AriumR pro UF system. This decrease was not observed when the cells were cultivated in T-flasks. The rapid decline in viability inside the spinner flasks can be attributed to the presence of endotoxins and inorganic salts present in RO water which affect the growth and viability of the cells. However, these adverse effects of endotoxins or inorganic salts were not observed in static cultures (small-scale cultivation), such as in the T-flasks, because in the latter case, growth is limited by the O2 concentration in the medium and not by the endotoxin or inorganic salt concentration (no typical growth curve can be observed in the T-flasks compared with the cultures in the spinner flasks). The actual effects can be observed only in spinner flask cultivation where higher cell densities occur and O2 is not a limiting factor. In spinner flasks in which high cell densities can be achieved, the effect of higher concentrations of endotoxins and inorganic salts in media reconstituted with RO water result in a reduced growth rate (lower cell density and lower viability) compared with values obtained for the controls (ready-to-use media) or for the samples in medium reconstituted with Arium® ultrapure water. These results are confirmed by the antibody production experiments. Antibody production in spinner flasks (Figure 6) of cells cultivated in media reconstituted with ultrapure water obtained from the Arium® pro UF system was 0.84 mg/ml (example day 8) and is thus higher than the values obtained for antibody production with the manufacturerfs ready-made media as controls (0.71 mg/ml) or in the samples reconstituted with RO water (0.42 mg/ml). The productivity of cells, i.e Mab production in T-flasks, was not measured because the amounts of antibodies were too low and a reliable comparison of such low values was not statistically relevant.
The results demonstrate that dehydrated media (CDM4PERMab media) that are reconstituted with Arium® pro UF ultrapure water are suitable for use in cultivation of PER.C6 EpCAM cell lines instead of commercially available ready-to-use media. The growth characteristics of the PER.C6 EpCAM cell lines cultivated in media reconstituted with water purified by Arium® pro UF are similar to those of the PER.C6EpCAM cell lines cultivated in ready-touse CDM4PERMab media used as controls. Moreover, enhanced growth was observed in cell line samples cultivated in media reconstituted with Arium ultrapure water compared with the samples cultured in media reconstituted with RO water when the experiments were carried out in spinner flasks. In this case higher cell densities normally occur and O2 is not a limiting factor. It was therefore concluded that the higher concentration of endotoxins or inorganic salts in RO water caused this decrease in growth. These results were confirmed and reflected by the antibody production in the PER.C6 EpCAM cell line cultivated in spinner flasks. Antibody expression (Mab production) from PER.C6 cell lines showed the highest values in the samples reconstituted with in Arium ultrapure water, followed by those of the controls (ready-to-use media). These values were unlike those in the RO water samples, where antibody production decreased. Hence, it was concluded that water from the Arium® pro UF system is well suited for PER.C6 EpCAM cell cultivation because this water system minimises the content of impurities, such as inorganic ions, organic compounds and in particular reduces endotoxins to exceptionally low levels, as was recently confirmed in newer experiments .
 ASTM Standard Guide for Bio-Applications Grade Water D 5196-06.  Whitehead P. (2007): Water Purity and Regulations, in Medicines from Animal Cell Culture (eds. Stacey, G. and Davis, J.), John Wiley & Sons, Ltd.  Freshney, Ian, R. (2010): Culture of Animal Cells . A Manual of Basic Technique and Specialized Applications- 6th edition, John Wiley & Sons, Inc., USA.  Schmidt K. und Herbig E. (2012): gWeniger ist mehr . Quantitative Endotoxinbestimmung von Reinstwasserg, Laborpraxis, 5, 36. Jhg., 2012