The importance of water quality for media preparation

Transcription

1 The importance of water quality for media preparation K.E.Wiemer 1 ' 3, A.Anderson 1 and B.Stewart 2 institute for Assisted Reproduction, 200 Hawthorne Lane, Charlotte, NC 28233, and 2 Millipore Corporation, 80 Ashby Road, Bedford, MA 01730, USA 3 To whom correspondence should be addressed The variability in pregnancy rates achieved among in-vitro fertilization (IVF) clinics may be partially attributable to the disparate quality of the water used in the preparation of culture media. The removal of contaminants in the water is of paramount importance since water constitutes the predominant component in any media formulation. To assist in the selection, operation and maintenance of a water purification system, the level of contaminants must be carefully monitored. Conductivity and resistance are used to measure the purity of natural and ultrapure water respectively. Feed water is analysed by an assortment of direct chemical means to determine the necessary system filtration steps. In general, high quality water can be produced by combined reverse osmosis and electrodeionization of treated tap water. Processed water is supplied to an ultrapure water system to provide final polished water. A detailed water processing protocol is presented along with quality assurance guidelines to ensure the consistent production of high quality ultrapure water suitable for in-vitro human embryo culture. Key words: IVF/media preparation/water contaminants/water quality Introduction Pregnancy results following in-vitro fertilization (IVF) vary in success significantly from clinic to clinic. The tremendous variation seen today is likely due to a plethora of factors which include differences in infertile patient populations, stimulation protocols and laboratory conditions. It is generally recognized that laboratory conditions are of paramount importance in maintaining consistent success rates within a specific programme. An area of major concern is the negative impact that poor air and water quality have on embryonic development (Bavister et al, 1988; Rienhart et ai, 1988). Laboratory staff have long recognized that extraordinary steps are required to remove impurities from water. Removal of impurities is particularly important since the major component of any culture medium is water. This does not, however, reduce the importance of selecting pure reagents and materials which 166 European Society for Human Reproduction and Embryology Human Reproduction Volume 13 Supplement

2 The importance of water quality for media preparation comprise or contact culture media during handling and storage. Over the years a number of technologies have been used, in both single and combined stages, to purify water. Technologies that are currently in use include distillation, deionization, carbon absorption, microporous filtration, reverse osmosis, ultrafiltration and other unique methods including the collection and storage of rain water. This paper will examine the contaminants, variables and technology that are available to ensure a consistent production of ultrapure water for the development of human IVF culture media. Water sources and quality Water sources are generally categorized into two types: (i) ground water, which originates from wells; and (ii) surface water which originates from reservoirs, rivers and basins. Ground water often contains high concentrations of dissolved ionic solids due to its exposure to porous rock and minerals deep underground. For example, high amounts of dissolved iron or calcium in well water may be present. Ground water, however, tends to contain lower concentrations of colloidal, organic and particulate matter as a result of the natural filtration which occurs within soil and porous sedimentary rock. Surface waters quite often contain lower concentrations of dissolved solids, but higher levels of organic and particulate contamination. Moreover, a wide range of contaminants, e.g. fertilizers, pesticides, herbicides, detergents, industrial waste effluent, and waste solvents have been introduced into both surface and ground water through farming, mining and drilling, industry and waste disposal. Seasonal fluctuations in temperature and precipitation further affect snow melt runoff, levels of water tables, and inversions in lakes and rivers. Since water is an excellent 'universal' solvent and provides a medium for most biological and chemical reactions, it is also highly susceptible to contamination by more substances than any other common solvent. Contaminants found in water may be divided into the following main groups: (i) inorganics - comprising dissolved cationic and anionic species; (ii) organics; (iii) particles, both deformable and non-deformable; and (iv) micro-organisms, primarily bacteria, algae, and fungi. These contaminants may be naturally occurring or in many instances were introduced into the environment by man. Dissolved inorganics occur naturally in all waters, and organics, such as organic acids, originate from the decomposition of leaves, trees and animal matter. Chlorine and chloramines are added to water to reduce the level of micro-organisms to allow safe drinking. Polyionic substrates are used to remove contaminants in water treatment processes. Other components, such as ozone and fluorine, may be added for other safety and disinfection purposes (Meltzer, 1993). All such substances added for practical benefits must be removed from water to be used for critical cell culture media preparation. Measurement and quantification of contaminants is critical in the selection, operation and maintenance of a water purification system. The total dissolved 167

3 K.E.Wiemer and B.Stewart solids (TDS) concentration found in water is measured by determining the ability of a water sample to conduct an electrical current and is represented by the conductivity. Conductivity is commonly expressed in microsiemens per centimeter ( as/cm), and is used to measure the purity of natural, tap or surface waters. It is highly recommended that either a simple hand-held or built-in sensor type conductivity measurement device should be available to observe conductivity values. Resistance, the inverse of conductivity, is used to measure the ionic purity of high purity water and is expressed in megohm-cm (Mfl-cm). Since the conductivity or resistance of a water sample is a function of the specific ions dissolved in solution, one may calculate the conductivity of an ideal water sample by using the specific conductance of H + and OH", the constituents of pure water. The resulting conductivity value of pure water is (is/cm or (the inverse) 18.2 MQ-cm. Water, in this ultrapure state, becomes a rather aggressive solvent that will try rapidly to establish an equilibrium with the environment. In fact, atmospheric carbon dioxide rapidly reacts with water to form aqueous carbonic acid. The dissociation of carbonic acid yields a hydrogen ion concentration, at an equilibrium of ~2X 10~ 6, equal to a ph of 5.7. This explains the slightly acidic ph detected when analysing newly stored ultrapure water (Nebergall et al, 1980). Organic levels in water are often expressed as total organic carbon (TOC); this represents the total mass of organic material in water excluding inorganic carbon (e.g. CO 2, bicarbonate). Organic measurement is done by the controlled UV oxidation of neutral organics to produce a net change in aqueous conductivity, which can then be quantified. Bacterial number is expressed in colony-forming units per millilitre (cfu/ml). Water-borne bacteria from nutrient deprived environments require special conditions for culture. A proper total count or R2A media formula must be used to support the bacteria from pure water. Also, incubation from h at a maximum of 30 C is required to develop visible colonies for enumeration. Pyrogenic endotoxins are detected using the Limulus amebocyte lysate (LAL) test in which a unique protein-endotoxin reaction occurs resulting in the formation of a gel. Kinetic detection methods relate endotoxin concentration to the rate of gel formation, while end-point methods yield a positive or negative solid gel which correlates with a known standard quantity of LAL substrate. The test is very sensitive; the limit of detection the end-point gel clot test ranges from 0.1 to 0.3 EU/ml (endotoxin units per millilitre) (Pearson, 1985). Analysis of the feed water source is crucial to determine the proper system filtration steps required. A variety of direct chemical methods are used to measure chlorine, hardness, silica, ph, iron, bacteria, alkalinity, CO 2 and other substances. The particle concentration of the feed water source is measured by passing water through a 47 mm, 0.45 (im filter membrane at 30 p.s.i.g. (constant) and measuring the flow decay rate of the product water stream. The flow decay value is expressed as the silt density index. Feed water temperature and ph should also be measured on-site to prevent natural sample changes which occur from storage and time. 168

4 The importance of water quality for media preparation Water filtration Over the years, we have developed the following water processing protocol and quality assurance programme in order to maintain consistent high quality water production in our geographical region. These general criteria can most likely be applied throughout the world, however, specific adaptations and changes would be necessary to meet certain regional water requirements. Particulate removal from tap water was achieved using a 12 inch 5 im nominal cut-off depth polypropylene filter. Chlorine and coarse organic removal was achieved using a cartridge containing activated carbon combined with a polyethylene fibre matrix. Both cartridges were installed in a series of housings that were mounted prior to the pure water system. Total chlorine in feed and product water was measured and monitored every 21 days to ensure no detectable chlorine (Hach DPD Chlorine test reagent kit; HACH Chemical Co, Loveland, CO, USA). Feed water, ionic and silica levels were monitored to note fluctuations coming from seasonal, environmental and maintenance sources. Treated water was processed via an Elix pure water system (Millipore Corporation, Bedford, MA, USA). Elix systems combine reverse osmosis and electrodeionization (EDI) to yield high quality surface water. Pre-filtered water first passed through reverse osmosis which removed 95-99% of a broad spectrum of organics, ions, bacteria, pyrogens and particles. Water processed through reverse osmosis was directed to an EDI module which used a combination of ion-exchange (IE) membranes, IE resin and direct current to remove ions from water. Elix product water was sent to a temporary storage reservoir in which water was recirculated for 2 h/day through a 185 nm UV oxidation system which oxidized organics and destroyed bacteria in the stored water to reduce any inherent contamination. This is an important aspect since we have removed the chlorine which had previously reduced bacterial counts. In our laboratory, reverse osmosis performance is monitored daily by recording the percentage of ionic rejection, feed and product conductivity as well as temperature. Acceptable rejection rates range from 95-99%. EDI purification provided 10-15MQ-cm water with <30.parts per billion (ppb) TOC. The storage reservoir supplied Elix-processed water to a Milli-Q PF Plus ultrapure water system (Millipore). Within the Milli-Q PF system, Elix water was further processed through a purification pack which contained nuclear-grade, high-purity, IE resins, and a special organic-scavenging resin. This process produced 18.2 M 2-cm water with organic levels <8 ppb TOC. Ultrapure water was further purified through a 5000 Da hollow-fibre ultrafilter. All proteins, particles, organics and endotoxins >5000 Da were removed as the water was processed from the exterior into the interior lumen of the fibres. The final polished water was then passed through a 0.22 iim filter to scavenge any trace particles from the water and also to prevent reverse bacterial contamination from incoming air from the environment. Quality control was maintained by daily recording of water system performance for above parameters. Because silica diffuses slowly and ionizes weakly in pure 169

5 K.E.Wiemer and B.Stewart Table I. Water maintenance overview Action Replace carbon and depth filters Monitor chlorine Chlorine sanitization of Elix Dummy (mock) sanitization of Elix Replace RO pretreatment pack UV recirculation of storage tank NaOH sanitize storage tank NaOH sanitization Milli-Q PF Plus Replace resin purification pack Replace 0.22 Jim final filter unit Monitor silica and TOC at final product Interval every 21 days bi every 3 months quarterly bi quarterly daily RO = reverse osmosis; TOC = total organic carbon. water, its retention and removal via IE resins is reduced. As a result, early silica breakthrough may indicate the relative capacity of the IE system, and this was monitored in feed and product waters to alert of any potential drop in ionic quality. An online total organic carbon meter (Anatel Corporation, Boulder, CO, USA) performed a controlled oxidation of the water sample in order to measure TOC content of the water sample. Acceptable levels for ultrapure product water range from 1-7 ppb. Endotoxin levels were verified using LAL testing and only accepted if levels were <0.03 EU/ml. Table I gives a brief outline of our current water system sanitization procedures. Guidelines for water quality Personnel in IVF laboratories generally regard the maintenance of a water system as an extremely burdensome task. However, the establishment of a daily monitoring schedule and timely sanitization can reduce this somewhat arduous task to a minimum. The obvious alternative to one manufacturing their own water for media production is to purchase commercially prepared media. This concept is certainly a 'leap of faith' since the embryology staff totally absolves itself of any real control over quality assurance. If one is willing to purchase media, be certain that this product is made by individuals who understand the consequences of poor quality culture media. It is our opinion that the specific formulations of culture media are extremely important, they are nonetheless not as crucial to outcomes as the quality of the product in question. In our laboratory, we chose to make media from scratch, using tissue culture-tested chemicals and water produced under conditions described above. For laboratories committed to manufacturing their own water for embryo culture, reverse osmosis rejection rates should continuously range between 95-99%. Feed water conductivity directly reflects the quality of the product feeding the final polishing system. In our laboratory normal feed water conductivity values range from p,s/cm. Moreover, product water conductivity values 170

6 The importance of water quality for media preparation range from jis/cm. Higher conductivity values reflect a higher degree of ionic contamination. Environmental factors specific to certain geographic regions may significantly affect these values. Reverse osmosis product water at Mii-cm was purified to >15.0 M 2-cm via the EDI module. It is important to supply the polishing system with as high quality water as possible to facilitate the ability consistently to produce high quality 18 MQ-cm ultrapure water. It is also important to have low TOC levels since there is no clear relationhsip between TOC and ion levels. In other words, it is possible to have ion-free water with high TOC levels. In our laboratory, TOC levels were monitored daily and water was not used unless readings of <8 ppb could be achieved. The impact of high TOC levels on human embryonic development is unclear. Proper filtration systems in combination with IE resins are very effective at removing pyrogens, proteins and particles (Huang et al, 1995). The monitoring of silicates is an important issue since silicates are the first indicators for a contaminant breakthrough of the IE resins. Regardless of the seasonal changes, silicate levels in our region were stable at 5-7 parts per million (ppm) in tap water. However, high purity water from the point-of-use filter was consistently 0.0 ppm. Perhaps the contaminants most well known by laboratory personnel are endotoxins. High endotoxin levels are believed to have an adverse effect on embryonic development and subsequent pregnancy rates in human IVF (Dumoulin et al, 1991). It is believed that the presence of endotoxins causes slow rate of embryonic development in conjunction with abnormally high rates of embryonic fragmentation. Culture media quality and its role in clinical results Frequently when results are lower than expected in an IVF laboratory, the tendency is to blame the culture medium that is currently being used. The authors' personal opinion is that, to date, optimum culture formulations have yet to be elucidated. This is true because many successful programmes continue to use a wide array of culture media. One aspect that successful IVF programmes have in common is the extreme attention paid to chemical constituents and water quality regardless of the media formulation. However, before placing the blame for less than expected results on the culture media currently used, IVF programmes should be self critical of their overall laboratory environment, stimulation protocols and, above all, embryo replacement techniques. This is especially true if the embryonic development is optimum for the specific day of transfer. Recently, a decrease in implantation rates has been attributed to air contaminants (Cohen et al, 1997). During the suspected period of air contamination in the laboratory neither a decrease in fertilization rates nor inadequate embryonic morphology or development were noted. Determination of the biological suitability of culture media using one- or two-cell mouse embryo development to blastocyst stage, sperm viability and development of supernumerary human embryos to the blastocyst stage have all been performed with some success (Quinn et al, 1985). The ability to predict 171

7 K.E.Wiemer and B.Stewart media performance with the highest level of accuracy prior to actual use would have its obvious benefits. As previously mentioned, however, the impact of filtration materials, storage vessels, and preparation techniques themselves can adversely affect otherwise perfectly good culture media (Purdy, 1982). However, the scope of this discussion is to bring attention to the underlying importance that water quality has in the manufacturing of human IVF culture media irrespective of its chemical formulation. By recognizing the potentially adverse variables that can impact on water quality, IVF laboratory personnel can develop protocols consistently to to produce high quality water. Many of the parameters that have been discussed beforehand can be universally applied throughout most different initial water characteristics and different water systems. It is imperative that laboratory personnel become familiar with the subtle traits of their initial water source as well as the capabilities of the water systems on hand. At the very least, the measurement of the many variables discussed throughout this paper may ensure the production of high quality water for human embryo culture in an IVF laboratory. References Bavister, B.D. and Andrews, J.C. (1988) A rapid sperm motility bioassay procedure for qualitycontrol testing of water and culture media. J. In Vitro Fertil. Embryo Transfer, 5, Cohen, J., Gilligan, A., Esposito, W. et al. (1997) Ambient air and its potential effects on conception in vitro. Hum. Reprod., 12, Dumoulin, J.C, Menheere, P.P., Evers, J.L. et al. (1991) The effects of endotoxins on gametes and preimplantation embryos cultured in vitro. Hum. Reprod., 6, Huang, Y., Leblanc, P., Apostolou, V. et al. (1995) Comparison of Milli-Q PF Plus water to DEPC-treated water in the preparation and analysis of RNA. Biotechniques, 19, Meltzer, T.H. (1993) High-Purity Water Preparation for the Semiconductor, Pharmaceutical and Power Industries. Tall Oaks Publishing Inc., Littleton, Colorado, USA. Nebergall, W.H., Holtzclaw, H.F. and Robins, W.R. (1980) College Chemistry with Qualitative Analysis. 6th edn. D.C.Heath and Company, Lexington, MA, USA, pp Pearson, F.C. (1985) Pyrogens: Endotoxins, LAL Testing and Depyrogenation. Marcel Dekker Inc, New York, NY, USA, pp , Purdy, J.M. (1982) Methods for fertilization and embryo culture. In Edwards, R.G. and Purdy, J.M. (eds), Human Conception In Vitro, Academic Press, London, UK, pp Quinn, P., Warnes, G.M., Kerin, J.F. and Kirby, C. (1985) Culture factors affecting the success rate of in vitro fertilization and embryo transfer. Ann. N.Y. Acad. Sci., 412, Rienhart, J.S., Bavister, B.D. and Gerrity, M. (1988) Quality control in the in vitro fertilization laboratory: comparison of bioassay systems for water quality. J. In vitro Fertil. Embryo Transfer, 5,