Animal‐cell culture media: History, characteristics, and current issues

2.1. Dawn of cultivation experiments (1882–1907)

In 1882, Sydney Ringer developed Ringer's solution, a balanced salt solution of a composition that is close to that of bodily fluids, and successfully kept frog hearts beating after dissection and removal from the body.4, 5 This is said to be the first instance of in vitro cultivation of animal tissue. Balanced salt solutions were developed one after another in the wake of Ringer's report, including Locke's solution,6 Tyrode's solution,7 the Krebs–Ringer bicarbonate solution,8 Gey's solution,9 Earle's solution,10 and Hanks' solution.11 The composition of these balanced salt solutions is simple and includes only inorganic salts, sometimes with glucose added as a nutrient. Nonetheless, their pH, osmotic pressure, and inorganic salt concentrations were calibrated to physiological conditions and these solutions can be used successfully to keep tissues and cells outside the body alive for short periods, generally up to a few days.

After the success of Ringer's solution, researchers began to pay attention to cells in culture devices and tried to maintain the cells. Nonetheless, the cells usually did not survive and rarely showed mitotic figures.12, 13, 14 In 1907, however, Ross G. Harrison successfully monitored an apparent outgrowth of nerve fibers of a frog for several weeks in lymph fluid that had been freshly drawn from the lymph sacs of an adult frog.15 This experiment is considered to be the beginning of animal cell cultivation.

2.2. Use of natural media (1907‐)

Alexis Carrel was a French surgeon and biologist who received the Nobel Prize in Physiology and Medicine in 1912 for his research on the vascular suture and transplantation of blood vessels and organs. He contributed greatly to tissue culture technology by devising a prototype of the cell culture flask that is used widely today and by establishing the aseptic manipulation technique.16

The first success of animal cell culture by Harrison inspired Carrel to send Montrose T. Burrows to work under Harrison's supervision in 1909. There, Burrows found that lymph is unsuitable for the cultivation of cells from warm‐blooded animals and used plasma instead. Thereafter, blood plasma had become a major culture medium for a variety of animal cells. He successfully cultivated chicken embryonic cells by using chicken blood plasma, which is readily available,17 and later successfully cultivated mammalian cells as well.18 In 1912, Carrel demonstrated that the long‐term cultivation of the cells that have been obtained from the connective tissues of chick fetuses is possible (for several months) with a periodic exchange of the medium.19 In 1913, he discovered that adding embryonic extract to blood plasma can dramatically increase cellular proliferation and extend the culture period of fibroblasts from the chick embryo heart.20, 21 Meanwhile, because the composition of the lymph, plasma, and embryonic extract was unknown, it became a new scientific inquiry regarding which of their components affected the survival and growth of animal tissues and cells. This situation led to a period when researchers attempted to identify the growth‐promoting substances within these ingredients of natural origin and to replace them with ingredients of definite composition.

Carrel is widely believed to be the first person in the world to successfully culture mammalian somatic cells, but the Biographical Memoirs of the National Academy of Sciences contends that the cultivation of guinea pig bone marrow specimens in 1908 by Margaret Reed, the future Mrs. Warren H. Lewis, precedes his success.22

2.3. Endeavors toward synthetic media (1911‐)

Margaret R. Lewis and Warren H. Lewis (1911) demonstrated that the Locke–Lewis solution—which is modified Locke's solution that additionally contains amino acids, bouillon, and glucose (or maltose)—is more effective for chick embryo cell cultivation than simple balanced salt solutions.23, 24 They reported that glucose is especially important: if the concentration of glucose is not sufficient in the medium, the chick embryo cells show vacuolar degeneration and die within a few days.25 In contrast, in their search for the active ingredients in embryonic extract, it was ascertained that the active substance is in the protein fraction26 and that the partially hydrolyzed proteins effectively promote the cell growth of chick embryo fibroblasts.27 Those researchers also confirmed the growth‐promoting activity of amino acids28 and glutathione for chick embryo fibroblasts.29 They hypothesized that glutathione is required for the control of the redox environment during cell cultivation. Carrel's medium was supplemented with several natural products, such as casein digests (thymus‐derived) nucleic acids, liver ash, and hemoglobin. Thus, their research did not lead to the molecular identification of any substance that is essential for cultured cells. Joilannes P. M. Vogelaar and Eleanor Erlichman, in contrast, confirmed the effectiveness of two hormones (insulin and thyroxine) by successfully cultivating human fibroblasts for >3 months in a medium that consisted of Ringer's solution with these hormones, along with glucose, cysteine, hemin, peptone, and blood plasma.30 Researchers continued the efforts to specify the composition of media thereafter, including the development of Baker's medium, which contains vitamin A, ascorbic acid, vitamin B₁, vitamin B₂, glutathione, and blood plasma.31 Nevertheless, this approach did not lead to culture media that did not include natural products, such as blood serum or plasma.

2.4. Birth of established cell lines (1940‐)

It is rare for healthy somatic cells that are derived from animals to acquire unlimited proliferative capacity during cultivation. They typically stop growing after a certain number of divisions (ie, the Hayflick limit).32 As a result, researchers would conduct studies by using cells that had been sampled from animals in each experiment. In 1940, Wilton R. Earle et al. used carcinogens to successfully create immortal mouse fibroblasts (L cells),10 revealing that proliferation from a single cell is possible.33 In 1951, George O. Gey and his coworkers created an infinitely proliferating human cell line from a tissue of a patient with uterine cervical cancer (HeLa cells).34 Due to the emergence of these established cell lines, the sampling of cells from the tissues of animals in each experiment became unnecessary, enabling researchers worldwide to perform assays by using the same homogenous population of cells. This state of affairs made it easier to examine and to precisely quantify the subtle differences in the effects of culture media on cells. Thus, the development of culture media advanced rapidly as a result.

2.5. Establishment of basal media and research into protein‐free media (1946‐)

Baker's medium and the other media that had been developed up to this point contained naturally derived components of unknown composition, including plasma, serum, bouillon, peptone, and tissue extracts. In order to find the crucial components in those natural materials and to develop defined media that are comparably efficient in the cultivation of cells, relative to the media containing natural ingredients, two main strategies were undertaken. The first strategy was to use dialyzed serum for the support of cells at minimum levels and to add defined components to maximize the proliferation of cells. The second strategy was not dependent on serum, or even proteins at all, and involved the formulation of media exclusively from definitive components.

Fischer was a pioneer of the first strategy. He dialyzed blood plasma to remove the low‐molecular‐weight fraction. Culture media that were supplemented with dialyzed blood plasma could sustain cells only for a short period, indicating that the low‐molecular‐weight fraction was essential for the survival of cells. Then, he discovered that the amino acids are the key substance in the low‐molecular‐weight fraction.35, 36 This creation of dialyzed media, along with the established cell lines, made it possible to determine, in a systemic way, whether the cells require low‐molecular‐weight substances under culture conditions. For example, in 1955, Harry Eagle studied the minimum necessary amounts of low‐molecular‐weight components that are required by mouse L cells and HeLa cells by using a balanced salt solution with the addition of dialyzed serum, on the basis of Fischer's method.37, 38 He found that 13 amino acids and eight vitamins are necessary.39 Going on to study the amino acid requirements of a variety of cells, Eagle then developed the minimum essential medium (MEM), which is composed of the minimal essential components that he identified (glucose, six inorganic salts, 13 amino acids, eight water‐soluble vitamins, and dialyzed serum).40 Renato Dulbecco and Marguerite Vogt,41 Clifford P. Stanners et al.,42 and Norman N. Iscove and Fritz Melchers43 made improvements to MEM, according to the desired cell type and culture purpose (Table 2). Later, using dialyzed serum that had been prepared in accordance with Fischer's method, Thomas A. McCoy et al. confirmed that carcinosarcoma cells require pyruvate.44 At this time, the Roswell Park Memorial Institute (RPMI) modified the 5A medium that had been developed by McCoy,45 in terms of its calcium and magnesium concentrations. The result then was reported as the RPMI 1640 medium for use in lymphocyte cultivation.46, 47

The other strategy—the total omission of serum and proteins—is thought to have begun in 1946, with full‐scale research by Philip R. White. He developed a chemically defined medium that was composed of glucose, inorganic salts, amino acids, iron, vitamins, and glutathione, with no protein at all. Using this medium, he successfully cultivated chick embryo‐derived fibroblasts and cardiac muscle cells for ~2 months.48 Other researchers were unable to repeat his findings and argued that White's Medium still needed 10%‐20% serum for the reported results.49 In 1950, Connaught Medical Research Laboratories (CMRL), led by Raymond C. Parker, developed Medium 199, which includes fat‐soluble vitamins, cholesterol, and nucleic acid precursors, in addition to the ingredients of White's Medium (Table 2). Instead of adding serum or proteins, Parker's team used the strategy of adding as many low‐molecular‐weight substances as possible, according to what was considered theoretically necessary for cell culture.50 Using this medium, they cultured chick embryo‐derived cells for 3‐4 weeks. Serum and embryonic extract were still necessary for the primary culture. Other researchers also reported the necessity of a small amount of serum in Medium 199 for cell proliferation.37 The team at CMRL thereafter developed a chemically defined medium, CMRL1066, using mouse L cells and their proliferation as an indicator of the effectiveness of the medium. This CMRL1066 medium, consisting of 58 components, was developed through several modifications, such as increasing the levels of reducing substances (cysteine, glutathione, and ascorbic acid), removing fat‐soluble vitamins, changing the nucleic acid precursors, and adding coenzymes (Table 2).51, 52 In contrast, during the same period, the National Cancer Institute's Tissue Culture Section (NCTC), led by Wilton R. Earle, developed a chemically defined medium that was composed of 68 ingredients, named NCTC109 (Table 2).53 These two media were difficult to prepare because of their complex composition. Thereafter, Charity Waymouth developed the MB 752/1 medium (Table 2), which is composed of 40 ingredients (glucose, inorganic salts, amino acids, vitamins, purine bases, hypoxanthine, and glutathione). This was the simplest‐possible chemically defined medium at the time.54 CMRL1066, NCTC109, and MB 752/1, however, all were designed for the best growth of mouse L cells. Therefore, they were not necessarily suitable for the serum‐free culture of other cell types. In addition, cell cloning was not possible in those protein‐free media. Waymouth hypothesized that the proteins that are produced by cells are necessary for proliferation under protein‐free conditions: the so‐called “auto/paracrine effect.” It was in this context that Richard G. Ham developed Ham's F‐10 medium in 1963 (Table 2). With two kinds of serum protein fractions (albumin and fetuin) instead of serum, he successfully made a single Chinese hamster ovary (CHO) cell form a colony under serum‐free conditions.53 The composition of this medium, however, was undefined because it contained serum protein fractions and the proliferative capacity it conferred fell short of that induced by serum‐containing media. There was also a challenge: serum supplementation remained necessary for the cultivation of cell types other than CHO cells. In order to solve these issues, Ham replaced the albumin and fetuin of biological origin with low‐molecular‐weight substances—linoleic acid and putrescine55, 56—and developed Ham's F‐12, a completely synthetic medium of definite composition. During this refinement of the medium, the concentration of each ingredient in Ham's F‐10 was reviewed and modified (Table 2).57 With the reduced amount of zinc, this medium supported colony formation by mouse L cells, but protein‐free cultivation of other cells remained difficult.58, 59 It became clear in later studies that trace elements contaminating the water or raw materials they used are necessary for the protein‐free cultivation of CHO cells in Ham's F‐12 medium, leading to the development of MCDB301, a medium that is supplemented with 20 trace elements (Table 2).60 Toshiko Takaoka and Hajim Katsuta successfully cultured various kinds of cells long‐term in a simple medium that did not contain proteins or lipids;61 however, only a few cell types can adapt to protein‐ and lipid‐free conditions and the fatty‐acid composition of the cells thus cultured differs greatly from the cells cultured in serum‐containing media.62

In the field of life sciences, the “reductive approach” and the “synthetic approach” are often used for elucidating a complicated biological process. The reductive approach is used to analyze and identify the essential parts in the complex biological process. The synthetic approach is used to reconstruct the biological process by putting the known parts together. Both approaches were instrumental in the development of synthetic culture media because these approaches had disadvantages relative to each other. First, the reductive approach was undertaken strenuously by Carrel and others because the synthetic approach was almost impossible without any information about the active substances for the cultivation of cells; accordingly, the substances that were important for cell culture in natural media were lined up. In contrast, the reductive approach could not be entirely successful at elucidating the complexities of the natural media because of the quantitative and qualitative limitations of the analytical method, complicated components of the natural media, and their variation from batch to batch. Accordingly, the candidate substances based on the knowledge of the reductive approach were tested by means of the synthetic approach by Eagle and others and the substances that are essential for cell culture were eventually identified. It seems to be important when and how these approaches are used for the successful development of culture media.

2.6. Identification of serum substitutes and the development of serum‐free media tailored to a cell type (1970‐)

Insulin was discovered earlier by Frederick Banting and Charles Best (1921), but full‐scale research into this peptide as a supplement for culture media began in the 1960s.63, 64 Initially, the effectiveness of insulin alone was found to be inferior to that of serum,65, 66 but the use of insulin in combination with low‐concentration serum yielded a higher level of efficacy of baby hamster kidney (BHK)21 cell growth. This finding led researchers to conclude that insulin acts in a coordinated manner with serum components. Growth factors were discovered one after another during this era: nerve growth factor,67 epidermal growth factor,68, 69 insulin‐like growth factor,70, 71, 72, 73 fibroblast growth factor (FGF),74, 75 platelet‐derived growth factor,76, 77, 78 and transforming growth factor (TGF).79, 80 The addition of these growth factors to a culture medium increased cellular proliferation. Nevertheless, their effect on cell proliferation, as with insulin, was found to be almost always inferior to the effect of serum.81, 82, 83, 84, 85

Under these circumstances, in 1976, three key reports were published that accelerated the development of serum‐free media. Ham's group discovered that a trace element of selenite is necessary for the serum‐free cultivation of human diploid cells86 and Larry J. Guilbert and Iscove showed that, besides selenite, a combination of transferrin and albumin is a good serum substitute.87 Izumi Hayashi and Gordon H. Sato discovered that a combination of several hormones and growth factors is an effective serum substitute.58 Prompted by their discoveries, attempts at serum‐free culture by using serum substitutes (eg, several hormones and growth factors, transferrin, and selenite; Table 3) grew in number and a variety of serum‐free media was developed, with each medium tailored to researchers' cell type of interest.

When developing serum‐free culture media, researchers typically test combinations of individual serum substitutes. As many cells require insulin, transferrin, and selenite,88 ITS, a supplement that contains a mixture of these three substances, has become commercially available. In addition, Hiroki Murakami et al. found that ethanolamine is essential for the serum‐free cultivation of hybridomas and developed the supplement, ITES, which consists of ITS plus ethanolamine.89 Various other supplements were designed for different cell types and cultivation purposes, with the aim of the addition of various trace elements to protein‐free culture media. These include Synthetic Serum Replacement, in which various trace elements are stabilized by ethylenediaminetetraacetic acid, citric acid, and aurintricarboxylic acid;90 the B‐27 supplement that was created for use with nerve cells (it contains progesterone, putrescine, triiodothyronine, fatty acids, vitamin E, bovine serum albumin [BSA], and glutathione, in addition to ITS);91, 92 and Knockout Serum Replacement (KSR), a formulation of various active ingredients of serum (eg, BSA, amino acids, antioxidants, ITS, and trace elements).93

2.7. Improvements to basal media (1970‐)

In addition to being a source of hormones, growth factors, carrier proteins, and lipids, serum increases the levels of various low‐molecular‐weight compounds in basal media. As a result, traditional basal media from which serum is excluded were sometimes unable to adequately support cell growth.94 For example, Barnes et al. and Murakami et al. reported that the performance of DMEM, MEM, and Ham's F‐12 as basal media in serum‐free cultures is inadequate, but that the DMEM/F‐12 medium, in which Ham's F‐12 and DMEM are combined in a 1:1 ratio, shows better performance occasionally when used for certain types of cells.95, 96 The reason seems to be the large number of constituents in Ham's F‐12 and the high concentration of several nutritional constituents in DMEM: mixing the two allows each to complement the weaknesses of the other. Moreover, Murakami et al. reported that the RDF medium—a 2:1:1 mixture of RPMI 1640, DMEM, and Ham's F‐12—yields more effective cell growth of hybridomas than does the DMEM/F12 medium.97 Mixed media, however, do not always show a level of performance that is better than that of a single medium. For example, the ferrous sulfate that is contained in Ham's F‐12 is toxic to nerve cells and nerve cells proliferate more readily at a reduced osmolarity. Thus, DMEM alone is more effective than DMEM/F‐12 in this case.91 Naturally, the composition of a basal medium that is used for serum‐free culture should be optimized for each cell type. In addition, it seems that the optimization also depends on the scale of the culture and its method.98

2.8. Medical and industrial applications of animal‐cell culture technology (1978‐)

3.1. Selection of a basal medium: the roles of the medium components

4.1. Setting up a serum‐free culture

There are two points to note when working with a serum‐free culture system. First, unintended clones of cells in a culture vessel can be selected during subculturing because serum‐free media can promote the proliferation of a particular cell clone or subtype of cell more often than can serum‐containing media. Second, impurities in a serum‐free medium, both avoidable and inevitable, might have stronger effects on the cultured cells than can serum‐containing media because of a lack of the toxin‐neutralizing activity that serum contains. In addition, there are other precautions for serum‐free cultures: that is, the minimization of the concentration of trypsin and the careful selection of adhesion factors.57

4.2. pH changes

The pH of the culture medium that is used for animal‐cell cultures generally is maintained by the equilibrium relationship between sodium bicarbonate (NaHCO₃) in the culture medium and CO₂ in the incubator (Fig. 1). Many laboratories set the CO₂ concentration in the incubator to 5%‐6% by convention, but technically, it should be adjusted according to the concentration of the NaHCO₃ in the culture medium. For example, 5% CO₂ is appropriate for a medium with 26 mmol L⁻¹ NaHCO₃ added (eg, MEM medium), 2% CO₂ for a medium with 14 mmol L⁻¹ added (eg, Ham's F‐12 medium), and 10% CO₂ for a medium with 44 mmol L⁻¹ NaHCO₃ added (eg, DMEM medium). These numbers are theoretical values that are calculated by using the Henderson–Hasselbalch equation (Fig. 1): in practice, checking the pH of a culture medium after it reaches equilibrium and then making minute adjustments to the CO₂ concentration are preferable.

The 4‐(2‐hydroxyethyl)‐1‐piperazineethanesulfonic acid (HEPES), with its strong buffering capacity, might have buffering effects in a medium. Especially when cultured cells are handled outside the CO₂ incubator, the pH of the medium with HEPES is more stable than with only bicarbonate. It also might be beneficial for cultures of high cell density; for instance, when the pH can drop rapidly due to an accumulation of metabolites, such as lactate. With serum‐free media, which lack the buffering capacity of serum, the pH‐buffering role of HEPES could be more relevant. Nevertheless, HEPES might have negative effects on certain types of cells, such as chick embryo epiphyseal chondrocytes and ES cells.107, 132 It also has been demonstrated that HEPES‐containing media produce an increased level of cytotoxic products, mostly hydrogen peroxide, during exposure of the medium to visible light.133 Researchers should be aware of such drawbacks of HEPES and confirm that the addition of HEPES is harmless to the cells that they use.

Phenol red, added to a medium as a pH indicator, sometimes affects the proliferation of cells and shows estrogenic activity.134, 135 Therefore, investigators should not forget to test whether such characteristics of phenol red affect their cells and empirical results, especially when estrogen‐sensitive cells, like mammary gland cells, are involved.

4.3. Oxidative stress

Culture experiments in general and especially when performed under a higher oxygen concentration, such as organ culture experiments, can expose cells to oxidative stress, which negatively affects the cells and tissues in culture. Supplementation with substances with an antioxidant activity (eg, vitamins [C, E], glutathione, selenite, β‐mercaptoethanol, dithiothreitol, or lipoic acid) is effective, especially when serum (it contains antioxidants) is not added to the medium. In addition, iron and copper ions in a free state promote the production of reactive oxygen species. It is therefore best for these ions to be complexed with appropriate carriers (eg, transferrin, albumin, or chelating agents) and to be supplied to the cells in this state, while they are isolated from reactive systems. Caution is also necessary with respect to free iron ions, which are readily hydroxylated and precipitate in an aqueous solution.

4.4. Nutrient requirements

Sufficient amounts of nutrients in the medium are a prerequisite for cells to behave properly. Some cell types require higher levels of nutrients than others do, depending on their metabolic activity and proliferation rate. Such characteristics of cells should be taken into account in the selection of a medium. DMEM, for example, was designed originally to contain glucose at 5.6 mmol L⁻¹. Now, a modified version of DMEM with a high glucose concentration, 25 mmol L⁻¹, which can be used for cells requiring greater amounts of glucose, is available from various suppliers. One caveat when using this high‐glucose medium for actively proliferating cells is the accumulation of metabolites, like lactate, and a plunge in the pH. Researchers are advised to replace the medium at proper intervals or to use HEPES to confer a stronger pH‐buffering capacity onto the medium.

Glutamine, one of the essential amino acids, is an energy source for mammalian cells in culture, in addition to being a biosynthetic material for nucleic acids and proteins. The glutamine requirements for cells in culture are ~3‐40‐fold greater than those of other amino acids.40 During the cultivation of cells with high nutrient requirements, the addition of glutamine could be helpful. In contrast, glutamine readily decomposes in culture media and generates cytotoxic ammonia. Consequently, researchers may consider adding glutamine to the medium immediately before use or to consider using glutamine derivatives, such as L‐alanyl‐L‐glutamine or glycyl‐L‐glutamine, which are resistant to degradation.

Non‐essential amino acids (NEAAs) can be biosynthesized in the cell. Thus, NEAAs are not included in some basal media, such as BME and MEM. Nevertheless, not every cell type can produce sufficient amounts of NEAAs in the cultured state. The addition of NEAAs to the medium, therefore, could ensure more favorable culture conditions. Even cells that can biosynthesize NEAAs at a necessary rate might benefit from the addition of NEAAs to the medium because this action alleviates the biosynthetic load of the cells.

Branched‐chain amino acids (BCAAs), including valine, leucine, and isoleucine, require special attention. They all belong to the group of essential amino acids and several cell types, including human fibroblasts and mouse myeloma cells, require larger amounts of BCAAs. For those cells, increased concentrations of BCAAs or replacing the medium at proper intervals can be considered in order to obtain the best results of an experiment.

5.1. Design of experiments for the optimization of the medium components

Most of the culture media that generally are used today consist of a basal medium and several supplements. Researchers pay no attention to the possible interactions among the components of the basal medium and these supplements. The fact is that each component of a culture medium does not always act in isolation. The components can interact and their effects on the cells should be expected to be influenced by those interactions. Additionally, the concentration of a medium component and its effect on the cells might not follow a linear dependence. Cells' response to each component could be curvilinear, as shown in Figure 2. Therefore, the best way to find the optimal concentration and interactions of each component in a culture medium might be a statistical experimental strategy called a “design of experiments” (DoE), rather than the classical one‐factor‐at‐a‐time experiments. The latter approach might miss the optimal point of concentration because there are some unexamined areas in the range of experimental settings (Fig. 3A). By contrast, the DoE can allocate design points, namely the concentrations of the components to be examined, evenly throughout the area of settings. Then, the obtained empirical data will be applied to the appropriate mathematical models, such as the general linear model or the logistic regression model. In general, when the data corresponding to the response variable (also called the “objective variable”) are continuous, such as the production volume of a biopharmaceutical, the general linear model is used. With binominal data, such as the survival rate of cells, the logistic regression model is preferable. By means of these models, an optimal concentration of each component (which is expected to induce the best response of the cultured cells) will be estimated (Fig. 3B). The DoE mainly consists of screening designs and response surface designs. A screening design is used first when one is trying to extract the most important components from many, say more than five, candidates of the initial components. The strength of the screening designs, such as fractional factorial or the Plackett–Burman design, is that they can reduce the number of experimental runs significantly, in comparison with the number that otherwise has to be done with the full factorial design, which involves every possible combination of components. For example, when trying to find the key factors in Medium 199, which consists of 60 components, only 64 experiments are needed with the Plackett–Burman design, whereas 260 experiments are required with the full factorial design. Once the key factors are identified, preferably eight or fewer, a response surface design, such as the central composite design or Box–Behnken design, will be used to estimate a non‐linear response of the cells and to identify the optimized concentrations and interactions of each component. These approaches have become a powerful tool for the improvement of productivity in the field of cell culture.136 Moreover, a new screening design, called a “definitive screening design,” recently was reported, with the claim that it has the advantages of both a screening design and a response surface design.137 This design enables the estimation of not only the main effects (the effect of a certain component alone) but also the interactions of the components and the factors with non‐linear effects, at a minimal number of experimental runs. In the case of Medium 199, only 121 experiments are needed. This approach could be especially useful for studies where the sample size is limited, such as a primary culture, organ culture, and embryo culture, although there are some restrictions in this design. The establishment of effective development systems for culture media that combine in vitro and in silico approaches and the optimization of the combinations of all the components in the culture medium are crucial for further improvements in cell culture performance.

5.2. Undefined medium supplements

Supplements of biogenic origin, like serum, can be a cause of variation in the experimental results from batch to batch. They also carry a risk of microbial contamination of the culture medium. Thus, the replacement of those supplements with defined ones has been pursued in the history of culture media, as described above.43, 87 At the dawn of the technology, a human embryo was cultured in a medium containing serum. In the mid‐1980s, the serum could be replaced by serum albumin.138 Albumin is the most abundant protein in serum and is multifunctional. It binds to various water‐insoluble substances like lipids. The lipids that are carried by albumin become an energy source and biosynthetic substances for an embryo.139, 140, 141 In addition, serum albumin serves as an antioxidant, osmotic regulator, and neutralizer of toxins. These functions are key benefits that serum usually provides to the media. Nonetheless, the use of serum albumin in place of serum has not contributed much to the development of chemically defined media. First, most, if not all, commercial serum albumin versions contain >100 serum proteins, although these admixtures have very low concentrations.142, 143 Second, albumin can bind to potential toxins, like phthalates144 (common plasticizers) or endotoxins. Thus, those toxins can be present in the commercial versions of serum albumin. These impurities in serum albumin products even were reported to vary in concentration from batch to batch, thus affecting the results of the embryo culture.145, 146, 147, 148, 149, 150 Therefore, the development of the culture media without undefined supplements is desirable, especially for human embryos. Recently, highly purified recombinant human albumin became commercially available. It is worth trying it as a substitute for serum‐derived albumin. Summarized in Table 3 are the points of caution when researchers use undefined supplements.

5.3. Contamination of a medium

Foreign substances from unidentified sources can contaminate a culture medium, thus possibly affecting the empirical results. Typically, those contaminants include viruses, bacteria, mycoplasma, and endotoxins. There are, however, other types of contaminants, like plasticizers that might be eluted from plastic instruments144 or trace elements, even in water. These substances also can affect the cells in culture.60 It also was reported that some toxic substances are eluted from the microfilters that are used for sterilization.151 Some of this contamination seems to be even inevitable, but care must be taken to minimize it in order to make culture experiments reliable and highly reproducible. Thus, researchers may consider practicing the sterile technique strictly and selecting culture instruments carefully. Washing the instruments with the culture medium immediately before use is recommended in some special cases.151, 152, 153