Madin‐Darby canine kidney cell sialic acid receptor modulation induced by culture medium conditions: Implications for the isolation of influenza A virus

Abstract Background The influenza A virus (IAV) binds to α‐2,3‐ and α‐2,6‐linked sialic acid (SA) receptors expressed by Madin‐Darby canine kidney (MDCK) cells. The receptor distribution may therefore be important in regulating IAV propagation. Serum‐free medium (SFM) avoids variability in conventional culture medium containing fetal bovine serum (FBS), which can have variable composition and may contain endotoxins. However, little is known about the distribution of SA receptors on cells maintained in SFM. Objectives We assessed the influence of culture media on MDCK cell SA receptor distribution along with the effect of SA receptor distribution on IAV recovery. We hypothesized that SFM would increase the proportion of α‐2,6‐linked SA receptors present and alter isolate recovery. Methods Madin‐Darby canine kidney cells were cultured in medium containing FBS and two SFMs. Cell surface distribution of α‐2,6‐ and α‐2,3‐linked receptors was determined using flow cytometry. Recovery of swine‐ and avian‐lineage IAVs from MDCK cells maintained in each medium was quantified as TCID50. Results Madin‐Darby canine kidney cells cultured in UltraMDCK SFM expressed both SA receptors and supported the growth of both swine‐ and avian‐lineage IAVs. Cells maintained in other medium inconsistently expressed each receptor and the avian IAV grew to lower titers in cells cultured with FBS. Conclusions Medium conditions altered the distribution of SA receptors present on MDCK cells and affected IAV recovery. Culture in UltraMDCK SFM resulted in cells expressing both receptors and IAVs grew to higher titers than in the other culture condition, indicating that this medium may be useful for culturing IAV from multiple species.


| INTRODUC TI ON
Influenza A virus (IAV) is a common pathogen that infects numerous species and adversely affects both public and animal health.

Thousands of people become infected with seasonally circulating
IAVs each year resulting in substantial morbidity and mortality. 1 Outbreaks of influenza in commercial poultry and swine operations result in large economic losses and can facilitate reassortment events leading to zoonotic transmission with pandemic potential. 2,3 Continued surveillance and characterization of IAV from animal host species are needed to moderate the risk to public health.
Galactose at the receptor binding site of IAV hemagglutinin (HA) protein binds to α-2,3-linked or α-2,6-linked sialic acid (SA) on host cells to facilitate infection. 4 The type of SA receptor linkages found on host cells contributes to a transmission barrier for IAV strains between host species. 5 Avian-lineage IAVs generally have preferential binding to SA cellular receptors in the α-2,3-linked conformation, while swine-and human-lineage IAVs bind SA receptors in the α-2,6linked conformation. 6 Therefore, changes in the type of linkage of SA (ie, α-2,3-linked vs α-2,6-linked) and the level of expression of each type of SA on the cell surface could ultimately impact the efficiency of IAV infection and replication.
Several techniques have been developed to isolate IAVs including the use of embryonated chicken eggs and immortalized cell lines. [7][8][9][10][11][12][13] Madin-Darby canine kidney (MDCK) cell lines are widely available, easily amplify in culture, and are commonly used for isolating IAVs from a variety of species. 12,14 Previous studies indicated that MDCK cells cultured in the presence of fetal bovine serum (FBS) express both α-2,6-linked and α-2,3-linked SA receptor types and showed that some cells co-express both receptors. 9,14 In contrast, another study found that 98% of MDCK cells adapted to serum-free media (SFM) expressed only α-2,6linked SA receptors, with 2% of cells expressing both α-2,3-and α-2,6-linked SA receptors. 15 However, detailed characterization of SFM-adapted MDCK cells is needed because variations in SFM preparations may cause the cells to express different amounts of each SA receptor, altering IAV culture success. Indeed, MDCK SIAT1 cells, which overexpress α-2,6-linked SA receptors, improve isolation rates for human IAV versus standard MDCK cells. 16 Increasing the recovery of IAV from culture systems is important for the timely detection of seasonal and novel IAV to prevent further spread and permit appropriate medical treatment for symptomatic individuals. We hypothesized that culture in SFM would increase the proportion of MDCK cells expressing α-2,6-linked SAs and subsequently increase the recovery of mammalian origin IAV. We assessed MDCK cell SA receptor distributions over serial passages in commercially available SFM using flow cytometric analysis and quantified IAV recovery.

| Effect of culture media on sialic acid receptor expression
To determine whether alterations in SA receptor distribution are a consequence of depletion of medium nutrients due to cell growth, MDCK cell SA receptor distribution during culture in medium A and B was evaluated. MDCK cells were seeded at densities that ensured confluence at 1 day (high-8 × 10 6

| Medium B
When MDCK cells were maintained in medium B, a mean of 80% (range 52% to 100%) expressed both receptors regardless of time since passage ( Figure 2B), 16% (range 0.4% to 46%) only expressed α-2,3-linked, and 2% (range 0.02% to 6%) expressed only α-2,6-linked SA receptors. The percentage of cells expressing both receptors (mean 80%) was statistically different from the percentage of cells expressing both receptors cultured in medium A after three (mean 21%) or 4 days (mean 63%) in culture (P = .002, P ≤ .005, respectively). Cells maintained in medium B stained more intensely with the SNA stain than medium A and C cells, indicating the presence of more α-2,6-linked SA receptors per cell than cells maintained in medium A or C ( Figure 3A) even though most cells expressed both SAs.
Cells maintained in medium C stained more intensely with the MAA stain than medium A and B cells, indicating that these cells have more α-2,3-linked SA per cell ( Figure 3B).

| Effect of culture media on sialic acid receptor expression
Based on the results from the first experiment, we sought to de-  Interestingly, however, the proportion of cells co-expressing both SA receptors declined after day 2 post-seeding (at all densities). An average of 62% (range 52% to 74%) of cells expressed both receptors for the first 2 days after seeding, but this dropped to 21% (range 16% to 27%) on days three through seven.
The same method used above for medium A was applied to medium B. MDCK cells maintained in medium B expressed high proportions of both receptors regardless of initial seeding density ( Figure 4B). 44% (range 26% to 53%) of cells cultured in medium B expressed both receptors, 18% (range 10% to 41%) expressed only α-2,6-linked SAs, and 24% (9% to 33%) expressed only α-2,3-linked

| D ISCUSS I ON
The use of SFM is increasingly popular to avoid the variability and contaminants associated with FBS. 8 Transitioning cells in culture to SFM has the potential to alter the sensitivity of IAV recovery systems because changes in culture medium can alter the expression of SA receptors on cell membranes. 20 Since IAVs from different species typically prefer binding to either α-2,3-linked or α-2,6-linked SA receptor types, cells that express both could be useful for the  requires. FBS is commonly added because it contains many growth factors (eg, epidermal growth factor, fibroblast growth factor, nerve growth factor, endothelial cell growth factor, insulin-like growth factors, transforming growth factors). However, the use of FBS can be problematic because it is an animal-harvested product and each lot varies in composition and endotoxin contamination. 26 We suspect that the cycling of receptor distributions seen in experiment 1, with cells cultured in medium A, was due to variability in the lot of FBS used in that medium. One hypothesis is that the cells in this experiment were using nutrients in medium A in a way that caused the cells to change receptor expression from one passage to the next. SA expression is dependent on nutrients available and processing by the Golgi. 27 One group showed that cells cultured with modified SA derivatives will express modified SA after 48 hours. 20

CO N FLI C T O F I NTE R E S T S
The authors declare no conflicts of interest with respect to the research, authorship, and/or publication of this article.