Evaluation of Antibiotic-Free Cell Cultures in Manual and Automated Cell Handling

Antibiotics are frequently added to cell culture media; the greatest benefit to this is that it suppresses contamination.1 However, routine use of antibiotics is associated with the development of resistances, the tendency to forgo aseptic working procedures, and the repression and carryover of cryptogenic contamination.2,3 The most common antibiotic for cell cultivation is a combination of penicillin and streptomycin. Penicillin belongs to the group of β-lactames, which impair wall synthesis of the cells.4 The aminoglycoside streptomycin interferes with the protein biosynthesis and causes cell death in the case of impaired translation of undamaged mRNA–tRNA complexes, resulting in an increase in inoperative proteins.5 Furthermore, common concentrations of penicillin and streptomycin affect protein expression.6

In scientific research and regenerative medicine, there has been increased interest in automated cell cultivation in combination with high-throughput screening.7 Changing the media, splitting the cells, and seeding a defined number of cells are central to automated cell cultivation.8 Automated cell cultivation increases the stability of cell culture because of the identical reiterations of processes. It is safe and efficient, and reduces human error and contamination from microorganisms or other parallel cultivated cell lines, thus reducing costs. A key benefit is the ability to set up scheduling to eliminate errors before cell culture processes begin.7,9

Because of the negative effects of using antibiotics in cell cultivation, the authors investigated whether it was necessary to use them in automated systems. The automated and manual cell culture processes of three test groups—manual/sterile, manual/unsterile, and automated/sterile—were compared in order to evaluate the proliferation rate using a proliferation and cytotoxicity assay. The growth curves were determined after cyclophosphamide treatment of the cells and detection of the proliferation rates.


Automated cell cultures

A principal component of the proprietary automated cell culture system (developed at the Centre for Life Science Automation, University of Rostock, Germany) is the Biomek Span-8 liquid handler (Beckman Coulter, Brea, CA) with steel cannulas for pipetting solutions and a gripper for the automated transport of cell culture flasks. To support sterile working conditions, the Biomek was equipped with a HEPA filter (Camfil, Stockholm, Sweden) and UV lights (Vilber, Eberhardzell, Germany). Installed on the liquid handler deck are two 3-D tilt racks for the simulation of manual cell flask handling, two shaker incubators, and a lift that connects the deck with a Cytomat™ incubator (37 °C and 5% CO2) (Thermo Fisher Scientific, Waltham, MA). A heatable automated labware positioner (ALP) enables heating of solutions to 37 °C. A Velocity 11™ VSpin centrifuge (Velocity 11, Palo Alto, CA) is also connected to the liquid handler deck to separate cells from supernatant. Cell count and viability are determined by the integrated Vi-CELL cell counter (Beckman Coulter). A port selection valve enables switching between six different media. A cool box chills the media and solutions (Figure 1).

Figure 1 – Cell cultivation system.

High-throughput bioscreening

The high-throughput screening system consists of the following components. The Biomek NX and FX liquid handlers and the control workstation are outside the housing. Inside the housing are a barcode reader; centrifuge; lid station; plate sealer; peeler cover; and PHERAstar, NOVOstar, and FLUOstar Galaxy readers (BMG Labtech, Ortenburg, Germany) for the detection of absorbance, luminescence, and fluorescence. The rail-mounted Motoman robot (Yaskawa, Fukuoka, Japan) connects all stations and transports the well plates. A Cytomat hotel and Cytomat incubator (37 °C, 5% CO2) are also integrated into the system.

Cell culture tests

Cervix carcinoma cells were cultivated in Dulbecco’s Modified Eagle Medium (DMEM) with 10% fetal bovine serum (FBS) with and without penicillin/streptomycin (P/S) under sterile or unsterile conditions. Three different test groups were investigated: manual/sterile, manual/unsterile, and automated/sterile. Work under unsterile conditions was carried out in a conventional laboratory for bioscreening research without a sterile bench; under manual and sterile conditions, the investigation was done in a cleanroom with a sterile bench, while under automated and sterile conditions the HEPA filter and UV lights were used. The cells were cultivated in media with or without antibiotics after thawing under the specified conditions.

The HeLa cells for manual handling were expanded in T75 cell culture flasks for three or four weeks. The cells for automated handling were seeded in an AutoFlask™ cell culture flask (Grenier Bio-One, Frickenhausen, Germany) and disseminated with the proprietary automated cell culture system. At a confluence of up to 90% in the cell culture flask, cells were rinsed with phosphate buffered saline (PBS) and harvested using trypsin. Medium was then added to stop the enzymatic reaction after an incubation period of 3 min. The cell suspension was transferred to a modular reservoir and the cell count was evaluated using the Vi-CELL.

For bioscreening, cells were seeded in well plates with different cell counts and incubated overnight (37 °C, 5% CO2). The manual steps were carried out with Eppendorf pipets (Hauppauge, NY). Automated cell handling was enabled using the Biomek NX liquid handler with steel cannulas. The automated steps were programmed and controlled with SAMI® software (Beckman Coulter) combined with the liquid handler software.

EZ4U proliferation test

For regular proliferation tests without agent, cells were seeded with a cell count of 1000 up to 8000 cells per well. The total volume of media and cell suspension was therefore 200 μL/well. After one night of incubation at 37 °C and 5% CO2, the proliferation rate was detected using the EZ4U Kit (Biomedica Medizinprodukte GmbH & Co KG, Wien, Austria) and the high-throughput screening system. The EZ4U is a nonradioactive proliferation and toxicity screening assay based on the conversion of tetrazolium salt to formazan, which results in a color change. The liquid handler added 20 μL of the reagent to each well. After 3 hr of incubation, absorbance (OD 450 nm) was detected by the PHERAstar reader.

Growth curve under treatment with cyclophosphamide + EC50

The chemosensitivity of cervix carcinoma cells was investigated using the EZ4U proliferation test after cyclophosphamide treatment. To evaluate the growth rate, cells were seeded with a cell count of 10,000 cells per well and a ring of water to protect against evaporation. Then, cells were incubated at 37 °C and 5% CO2. The next day, cytostatic agent in a mother plate was diluted to a concentration of 35 mmol/L down to 0.27 mmol/L using the Biomek NX liquid handler. Afterward, the Biomek FX transferred the agent into the daughter plate containing the cervix carcinoma cells. Dimethyl sulfoxide (DMSO) was used as negative control and cell culture medium as positive control. After two days of incubation, the proliferation rates were detected using the EZ4U proliferation assay.

Results and discussion

Proliferation rate

The proliferation rate of the three different test groups was compared. These groups were defined as automated/sterile, manual/sterile, and manual/unsterile—half with antibiotic and the other half without antibiotic. The automated/sterile and manual/sterile test groups showed a higher proliferation rate in media without antibiotic up to 6000 cells/well. Automated and sterile treated cells without antibiotic showed a higher proliferation rate for all cell counts than cells with antibiotic. The differences in the proliferation rates for media with and without antibiotics increased from 5% at 1000 cells per well to 9% at 8000 cells per well under automated/sterile conditions. For cells cultivated under manual/sterile conditions, the measured differences were highest at 1000 cells per well (26%) and decreased to 6% at 6000 cells per well.

The highest percentage and most significant differences were detected in the manual/unsterile test group. In this group, cells treated without antibiotic showed a higher proliferation rate at all initial cell counts than cells with antibiotic. The major difference between the media conditioned with and without antibiotic was seen in the manual/unsterile test group. Variations of proliferation rates ranging from 39% at 2000 cells/well decreased to 11% at 8000 cells/well.

Figure 2 – Proliferation rate of HeLa cells (automated/sterile and manual/sterile) with and without P/S for cell counts of 1000 cells/well to 8000 cells/well + standard deviation (n = 6). Significance by t-test: *p<0.05, **p<0. 01, ***<0.001).

 Figure 3 – Proliferation rate of HeLa cells (manual/unsterile) with and without P/S for cell counts of 1000 cells/well to 8000 cells/well + standard deviation (n = 6). Significance by t-test: *p<0.05, **p<0. 01, ***p<0.001).

The proliferation rates of automated/sterile disseminated cells were slightly lower for all cell counts. In a comparison of media with antibiotic and media without antibiotic, the sterile test groups showed a higher proliferation rate without antibiotics. Contrary was the result of the manual and unsterile test group, in which a higher proliferation rate was detected for all cell counts. The variability between the sterile and antibiotic-free test groups decreased from 26% at 1000 cells/well to 1% at 8000 cells/well. In contrast, the authors noted increases from 4% at 1000 cells to 20% at 6000 cells/well for cells with antibiotic (Figures 2 and 3).

In general, antibiotics for cell cultures such as penicillin and streptomycin are routinely used because they are low-dose, nontoxic, and stable under different temperature conditions and pH values.10 Furthermore, Duewelkhenke et al.11 could not substantiate any effect of penicillin and streptomycin on proliferation, metabolic activity, and production of lactate in HeLa cells. Instead, Cohen et al.2 published the negative proliferation (mRNA expression) and differentiation effect of P/S compared to cells without antibiotic.

Cyclophosphamide and EC50 values

The growth rates of the three test groups were investigated, and it was found that the curve shapes of all groups were similar (Figures 4 and 5). Growth rates can also be used to evaluate the half-maximal effective dose (EC50). For cervix carcinoma cells EC50 values were detected in the range from 17.5 mmol/l to 8.75 mmol/L (Figure 4). Generally, the EC50 concentrations of cells with P/S are higher than of cells without P/S. The half-maximal concentrations of cells without antibiotic were identified in the range of 15–16 mmol/L and of cells with antibiotic from 12 to 15 mmol/L (Table 1).

 Figure 4 – Growth curve of HeLa cells (automated/sterile and manual/sterile) with and without P/S after treatment with cyclophosphamide + standard deviation (n = 6).
 Figure 5 – Growth curve of HeLa cells (manual/unsterile) with and without P/S after treatment with cyclophosphamide + standard deviation (n = 6).
Table 1 – Half-maximal effective dose (EC50) of cyclophosphamide at HeLa cells in three different test groups (n = 6)

Obviously higher concentrations of cytostatic agent had to be used to kill carcinoma cells in media with antibiotic. Normally the cytotoxicity and half-maximal concentrations are regularly quantified using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) tests in media with antibiotic. For example, EC50 values of different cancer cell types were 28 μmol/L for 9L glioblastoma cells, 22.5 μg/mL for A431 skin cancer cells, and 28 μg/mL MCF 7 for breast cancer cells.12,13


Cervix carcinoma cell cultures in three different test groups (manual/sterile, manual/unsterile, and automated/sterile) were investigated. The cells were disseminated manually and automatically with and without antibiotics under sterile and unsterile conditions. Growth curves were recorded to find EC50 values for cells treated with the chemosensitivity agent cyclophosphamide. Similar performance of automated and manual dissemination was indicated by proliferation rates, growth curves, and half-maximal effective dose (EC50).

The EZ4U proliferation test was used to evaluate the proliferation and toxicity of cells in automated bioscreening. The authors conclude that the antibiotic-free cultivation method is optimal, especially since the dissemination of cells without antibiotics using the automated cell cultivation system is stable. Chemosensitivity of agents should be detected in antibiotic-free media to prevent any influence of the antibiotics on the cells.


  1. Lincoln, C.K.; Gabridge, M.G. Cell culture contamination: sources, consequences, prevention, and elimination. Meth. Cell Biol.  1998, 57, 49–65.
  2. Cohen, S.; Samadikuchaksaraei, A. et al. Antibiotics reduce the growth rate and differentiation of embryonic stem cell cultures. Tissue Eng.  2006, 12(7), 2025–30.
  3. Kuhlmann, I. The prophylactic use of antibiotics in cell culture. Cytotechnology  1995, 19(2), 95–105.
  4. Tomasz, A. From penicillin-binding proteins to the lysis and death of bacteria: a 1979 view. Rev. Infect. Dis. 1979, 1(3), 434–67.
  5. Francois, B. Crystal structures of complexes between aminoglycosides and decoding A site oligonucleotides: role of the number of rings and positive charges in the specific binding leading to miscoding. Nucl. Acids Res. 2005, 33(17), 5677–90.
  6. Mathieson, W.; Kirkland, S. et al. Antimicrobials and in vitro systems: antibioticsand antimycotics alter the proteome of MCF-7 cells in culture. J. Cell. Biochem.  2011, 112(8), 2170–78.
  7. Kato, R.; Iejima, D. et al. A compact, automated cell culture system for clinical scale cell expansion from primary tissues. Tissue Eng. Part C: Methods  2010, 16(5), 947–56.
  8. Kino-oka, M.; Chowdhury, S.R. et al. Automating the expansion process of human skeletal muscle myoblasts with suppression of myotube formation. Tissue Eng. Part C: Methods  2009, 15(4), S. 717–28.
  9. Narkilahti, S.; Rajala, K. et al. Monitoring and analysis of dynamic growth of human embryonic stem cells: comparison of automated instrumentation and conventional culturing methods. BioMed. Eng. OnLine 2007, 6(1), 11.
  10. Bohannon, L.K.; Owens, S.D. et al. The effects of therapeutic concentrations of gentamicin, amikacin and hyaluronic acid on cultured bone marrow-derived equine mesenchymal stem cells. Equine Vet J.  2013.
  11. Duewelhenke, N.; Krut, O. et al. Influence on mitochondria and cytotoxicity of different antibiotics administered in high concentrations on primary human osteoblasts and cell lines. Antimicrob. Agents Chemother.  2006, 51(1), 54–63.
  12. Awasare, S.; Bhujbal, S. et al. In vitro cytotoxic activity of novel oleanane type of triterpenoid saponin from stem bark of Manilkara zapota Linn. Asian J. Pharm. Clin. Res.  2012, 5(Suppl. 4), 183–8.
  13. Jounaidi, Y. Enhanced antitumor activity of P450 prodrug-based gene therapy using the low Km cyclophosphamide 4-hydroxylase P450 2B11. Molec. Cancer Therapeut. 2006, 5(3), 541–55.

The authors are with the Centre for Life Science Automation, Central Scientific Facility of the University of Rostock, Rostock 18119, Germany; e-mail: ricarda.lehmann@celisca.de. The authors wish to thank the Federal Ministry of Education and Research (BMBF Germany) for the financial support (FKZ: 03Z1KN11), and Ms. Grit Koch (University of Rostock) for the manual cell cultivation and screenings.