S‐adenosylmethionine and methylthioadenosine boost cellular productivities of antibody forming Chinese hamster ovary cells

The improvement of cell specific productivities for the formation of therapeutic proteins is an important step towards intensified production processes. Among others, the induction of the desired production phenotype via proper media additives is a feasible solution provided that said compounds adequately trigger metabolic and regulatory programs inside the cells. In this study, S‐(5′‐adenosyl)‐ l‐methionine (SAM) and 5′‐deoxy‐5′‐(methylthio)adenosine (MTA) were found to stimulate cell specific productivities up to approx. 50% while keeping viable cell densities transiently high and partially arresting the cell cycle in an anti‐IL‐8‐producing CHO‐DP12 cell line. Noteworthy, MTA turned out to be the chemical degradation product of the methyl group donor SAM and is consumed by the cells.


| INTRODUCTION
Therapeutic proteins such as monoclonal antibodies (mAB) dominate the global market for biopharmaceuticals and are mostly produced in Chinese hamster ovary (CHO) cells (Walsh, 2018). The last decades of biomanufacturing witnessed a steady rise of product titers in conventional fed-batch modes reaching 5-8 g/L in 12-14 days process time (Reinhart, Damjanovic, Kaisermayer, & Kunert, 2015;Schaub et al., 2010;Wurm, 2004). However, those volumetric productivity improvements predominately mirror elevated viable cell densities rather than the equal rise of cell specific productivities (CSPs). The latter however, is of outstanding importance when the next generation of intensified bioprocesses should be realized, in particular when using perfusion processes (Becker, Junghans, Teleki, Bechmann, & Takors, 2019;Templeton, Smith, et al., 2017;Zhang et al., 2013).
Consequently, novel approaches are needed that boost CSPs for pharmaceutical proteins. As a prerequisite, detailed understanding of the roles of interacting partners in the complex regulatory patterns of CHO metabolism is necessary. As a result of this, novel switches to enhance CSPs may be derived. Promising studies focused on metabolism (Handlogten et al., 2018;Junghans et al., 2019;Templeton, Xu, Roush, & Chen, 2017), transcription (Pfizenmaier, Junghans, Teleki, & Takors, 2016;Sha, Bhatia, & Yoon, 2018), and on epigenetics. Epigenetic adaption mechanisms that is histone modification via acetylation and DNA methylation are basic features of CHO cell adaption. Nematpour et al. (2017) showed that stabilized acetylation by inhibition of histone deacetylases with medium additives enhanced mAB productivity in CHO cells. Furthermore, DNA methylation which links genetics with transcriptomics (Wippermann, Rupp, reason for particular downregulation and upregulation patterns found in high-producer cell lines (Feichtinger et al., 2016;Harreither et al., 2015). Recent studies of Dhiman et al. (2019) demonstrated that more than half of the analyzed nuclear genes are susceptible to methylation. Consequently, the addition of molecules interacting with epigenetic regulation mechanisms offers the intrinsic potential to improve CSPs.
In this context, the cosubstrate S-(5′-adenosyl)-L-methionine (SAM) was chosen as a medium additive whose structure was discovered in the early 1950s (Cantoni, 1951). SAM may be involved in aminopropylation (polyamine synthesis), trans-methylation and -sulfuration underlining its crucial role for cellular growth and regulation. Accordingly, SAM limitation may affect gene expression, membrane fluidity, and glutathione availability reflecting its roles in DNA methylation, methylation of phospholipids, and trans-sulfuration, respectively (Finkelstein, 1990;Mato, Alvarez, Ortiz, & Pajares, 1997). Metabolically, the methyl group donor SAM is created by methionine adenosyltransferase (MAT) using adenosine triphosphate (ATP) and L-methionine as substrates. In CHO, the latter needs to be supplemented in the medium. In methylation reactions, SAM is de-methylated to S-(5′-adenosyl)-L-homocysteine (SAH) further de-adenylated to homocysteine, re-methylated to methionine, and finally re-adenosylated to SAM again (SAM cycle; Finkelstein, 1990;Mato et al., 1997). Accordingly, the cycle not only regenerates the key methyl donor SAM, it also provides important precursors: Homocysteine serves as a precursor for glutathione synthesis (Lu, 1998) which is an important antioxidant in cells. By analogy, SAM enters the polyamine synthesis (Finkelstein, 1990). Investigations on liver disease demonstrated that extracellular SAM influences the cellular glutathione level. Likely, SAM is taken up by the cells (Mato et al., 1997) which may explain why SAM-treated liver cells showed decreased growth rate, prevented the development of liver cancer (Cai, Mao, Hwang, & Lu, 1998;Pascale et al., 1993) and disclosed antiapoptotic mechanisms (Ansorena, 2002).
Summarizing, SAM is a crucial cosubstrate that serves as a key methyl donor thereby linking metabolism with cellular regulation and epigenetics. Consequently, it offers the intrinsic potential to function as an additive that boosts CSPs in CHO cells.

| Extracellular analysis
Samples were taken at least once a day. Viable cell density, viability, and average cell diameter were determined using trypan blue staining using a Cedex XS cell counter (Innovatis AG, Bielefeld, Germany) for read out.
The extracellular concentrations of D-glucose and L-lactate were determined using a Labotrace automatic analyzer (Trace Analytics GmbH, Braunschweig, Germany). The product concentration of immunoglobulin G (IgG) was determined applying enzyme-linked immunosorbent assay (ELISA; Pfizenmaier, Matuszczyk, & Takors, 2015). Each sampling and measurement procedure was performed in replicates.

| SAM and MTA determination
The amount of SAM and MTA was quantified on an Agilent 1200 Series Data were analyzed using MassHunter B.06.00 Analysis software.

| SAM addition boosted CSP while repressing growth
After 48 hr cultivation time, SAM was added reaching a concentration of 250 µM. Notably, REF received the equivalent volume of water to prevent any bias due to medium dilution. The SAM-treated cultures reduced maximum viable cell density (VCD) and growth rate (between 48 and 120 hr) by 10% and 18% compared with REF, respectively ( Figure 1). However, high cell viabilities remained.
Although, SAM supplemented cells showed lower VCD than REF they reached more than 50% higher antibody titer (SAM: 192.73 ± 18.24 mg/L and REF: 123.39 ± 12.63 mg/L). The increased titer (>48 hr) transformed to increased CSP between 0 and 120 hr: SAM supplemented cells produced 9.81 ± 0.89 pg/day whereas the reference only showed 6.24 ± 0.51 pg/day. This corresponds to a 57% increase in CSP ( Figure 2) after SAM addition.

| SAM degradation was not affected by the presence of cells
To investigate the effects of SAM addition and uptake extracellular SAM concentrations were monitored in the medium. The anticipated SAM "uptake" coincided with equally fast degradation rates observed in cell-free cultivation medium under similar conditions (medium + SAM + cells, 11.54 ± 0.32 µM/hr; medium + SAM, 10.95 ± 0.30 µM/hr).
Consequently, the direct cellular uptake of SAM seemed unlikely. To better understand SAM degradation, MTA as literature described degradation product was measured in the medium (Hoffman, 1986;Morana et al., 2002). Indeed, MTA was detectable in medium with and without cells but not in the reference (water addition). MTA concentration increased over time in medium without cells. In contrast, cell culture studies only showed high MTA levels at the beginning which decreased during the cultivation (Figure 3).

| SAM and its degradation products MTA and HSL
To check whether MTA is the inducer of the growth and IL-8 pro-

| SAM and MTA interact with the cell cycle
Propidium iodide staining revealed differences in the cell cycle phase distribution ( Figure 6). All cells started from a common preculture that was split after 48 hr cultivation before adding

| DISCUSSION
We added SAM to the medium of CHO cells to interact with the cellular metabolism and with SAM mediated regulation (Finkelstein, 1990;Mato et al., 1997). Clearly, SAM addition decreased the growth rate ( Figure 1) which is in agreement with previous findings in liver cells (Cai et al., 1998;Pascale et al., 1993). It may be anticipated that changes in the DNA methylation pattern caused the growth phenotype | 3243 (Feichtinger et al., 2016). Unfortunately, the identification of methylation differences with and without SAM addition was not manageable within the scope of the study.
Significantly increased CSP was found in CHO-DP12 cells after SAM addition (Figure 2), although growth rates were reduced.
Metabolically, this phenomenon may be understood by assuming that SAM enters the cell boosting the precursor supply of the SAM cycle and/or alleviating the ATP dependent recycling. Mato et al. (1997) anticipated that liver cells take up extracellular SAM. The hypothesis was deduced from rising intracellular glutathione levels. However, Figure 3 reveals that decreasing extracellular SAM levels were observed irrespective of whether cells were present or not. Consequently, the cellular uptake of SAM is unlikely.
A computational study demonstrated that SAM degrades to MTA and HSL in water (Lankau, Kuo, & Yu, 2017). Consequently, we investigated the addition of these degradation products to the media. Whereas HSL did not create any particular phenotype, MTA caused effects very similar to the SAM addition ( Figure 4). Noteworthy, MTA natively occurs in mammalian cells as part of the polyamine biosynthesis (Pegg, 1988;Williams-Ashman, Seidenfeld, & Galletti, 1982). There, SAM is decarboxylated to MTA to access spermidine and spermine. Because high MTA levels inhibit polyamine synthesis MTA is rapidly degraded by 5′-methylthioadenosine phosphorylase. MTA was shown to permeate the cellular membrane.
The reduction of polyamine pathway intermediates such as spermidine after the addition of metabolic inhibitors induced arrest in the cell cycle's S-phase of CHO cells (Alm & Oredsson, 2000).
In agreement with this, we observed the accumulation of cells in S-phase after MTA addition concomitantly with a diminishing number in G1-phase compared with the REF (Figure 6). We hypothesize that  (Subhi et al., 2003).
Both, SAM and MTA additions increased CSPs of CHO-DP12 cells ( Figure 5). One explanation could be that cell cycle arrest particularly boosted cellular production rates (Fox, Patel, Yap, & Wang, 2004;Hendrick et al., 2001;Kaufmann, Mazur, Fussenegger, & Bailey, 1999;Pfizenmaier et al., 2015). However, those improvements are neither necessarily nor exclusively linked to a single cell cycle phase. Instead, they may benefit from multiple phases (Aggeler, Kapp, Tseng, & Werb, 1982;Al-Rubeai & Emery, 1990;Kim, Kim, & Lee, 2014;Kubbies & Stockinger, 1990;Lloyd, Holmes, Jackson, Emery, & Al-Rubeai, 2000). Accordingly, one may argue whether the improvement of CSP mirrors the benefits of arresting the cell cycle alone. Lloyd et al. (2000) observed CSPs proportional to the cell volume investigating four different CHO cell lines. As cell volume and cell cycle status are linked too, unequivocal conclusions are hard to get. As a result, they concluded that rather cells in S-and G2/M-phase show improved cellular productivity than those in G1-phase. However, this statement somewhat contradicts the findings of other groups (Fox et al., 2004;Hendrick et al., 2001;Kaufmann et al., 1999;Pfizenmaier et al., 2015). In our studies, cell diameters rose after MTA and SAM addition ( Figure 6) which agrees with the findings of Lloyd et al. (2000). Strictly speaking, the observation does not exclude potential beneficial effects resulting from cell cycle arrest. In general, cell cycle arrest allows cells to use more energy for recombinant antibody production because biomass formation is hampered (Fussenegger, Mazur, & Bailey, 1997).
Besides the consequences of MTA and SAM on cell cycle and diameter, MTA may interact with a lot of other cell mechanisms (Ansorena, 2002;Lee & Cho, 1998;Maher, 1993;Riscoe et al., 1984).
For instance, MTA inactivates SAH hydrolase (Williams-Ashman et al., 1982) leading to increased SAH levels finally resulting in reduced DNA methylation (Iraburu et al., 2002). As methylation is a crucial part of epigenetic regulation (Feichtinger et al., 2016) affecting transcription too (Wippermann et al., 2015) further transcriptional regulation programs are likely to be activated as well.
Consequently, the substrate MTA will be further investigated, and its balanced addition is expected to intensify its boosting capability and avoid potential negative effects.

| CONCLUSION
The study was motivated by the idea to identify novel, nonconventional media components as promising additives to boost CSPs showcasing IgG-1 formation with CHO-DP12 cells. Testing the methyl group donor SAM, a multilevel effector was selected that may interact with different hierarchy levels of cellular regulation concomitantly. SAM addition improved CSPs by approx. 50% arresting the cell cycle and reducing cellular growth at the same time. A similar phenotype occurred when MTA was added which turned out to be the real effector resulting from SAM degradation and can transfer the cellular membrane.