Here, we describe a new method that enables the anaerobic long-term cultivation of established cell lines. The maximum survival time that was tested is 17 days. This method is suitable for the testing of cytotoxic agents and exploring the physiology of anoxically replicating cells.
Most mucosal surfaces along with the midpoints in tumors and stem cell niches are geographic areas of the body that are anoxic. Previous studies show that the incubation in normoxic (5% CO2 in air) or hypoxic (low oxygen levels) conditions followed by an anoxic incubation (an absence of free oxygen) results in limited viability (2–3 days). A novel methodology was developed that enables an anoxic cell cultivation (for at least 17 days; the maximum time tested). The most critical aspect of this methodology is to ensure that no oxygen is introduced into the system. This is obtained by the degassing of media, and by flushing tubes, dishes, flasks, and pipettes with an anaerobic gas mixture (H2, CO2, N2) followed by permitting the materials to equilibrate to the anoxic (non-oxygen) environment prior to usage. Additional care must be exercised when acquiring photomicrographs to ensure that the micrographs obtained do not contain artifacts. In the absence of oxygen, cell morphology is significantly altered. Two distinct morphotypes are noted for all anaerobically-grown cells. The ability to grow and maintain mammalian cells in the absence of oxygen can be applied to the analysis of cell physiology, polymicrobial interactions, and the characterization of biosynthetic pathways for novel cancer drug development.
Cells from solid tumors, stem cell niches, and those lining the mucosal surfaces exist in environments that experience reduced oxygen levels, including anoxia1,2,3,4. In normal physiologic states, oxygenation varies beyond that of hypoxia to anoxia (the complete absence of oxygen)5,6.The realization that atmospheric oxygen adversely affects mammalian cell replication and that in vitro cell growth can be optimized under depleted oxygen conditions was reported in the early 1970s. Richter et al.7 showed that 1–3% of oxygen levels (hypoxia) enhanced plating efficiency as compared to atmospheric oxygen (20%). The human diploid cell lifespan is also extended in the hypoxic culture conditions8.
In vivo, hypoxic conditions occur when oxygen stores are depleted (e.g., during intense exercise), wherein the ATP production is switched in the skeletal muscle from aerobic respiration to fermentation (anaerobic respiration) with the end-product of lactic acid9. Pathologically, in cancerous tumors, the interior of the tumor mass is hypoxic to anoxic due to poor vascularization10. The effect of the limited perfusion on tumor interiors is independently validated by tumor interiors colonized by obligate anaerobes1. Mechanistically, tumor cell survival in hypoxia is thought to be solely dependent on the expression of the hypoxia-inducible factor 1-alpha gene (HIF1-alpha), which is the initial spontaneous response to hypoxia4,11,12. HIF1-alpha is induced under hypoxic conditions by heat shock proteins that bind the HIF1-alpha promoter and upregulate the gene transcription12. These heat shock proteins are believed to aid in the induction of the various phenotypes seen in the tumor hypoxic microenvironment. These phenotypes exhibit an increased expression of the cell membrane glucose transporters and the rate of glycolysis (the Warburg effect)13. The result is a switch from mitochondrial oxidative phosphorylation to lactate fermentation.
Anoxic survival can also utilize alternatives to glucose to support the survival phenomenon14,15. The best studied mammalian example is the mole rat, which can survive for nearly 20 min without oxygen through a fructose-driven glycolytic fermentation pathway14. An alternative adaptation occurs in certain fish (e.g., carp [Carassius sp.], which can survive for significantly longer time periods using glycolysis with ethanol as the terminal by-product)15. In both cases, fermentation drives the metabolism enabling the survival in the absence of oxygen. The current hypothesis for anoxic survival is that so long as HIF1-alpha is activated during hypoxia, mitochondrial respiration, without the need for oxygen, occurs under anaerobic conditions16. Furthermore, it is postulated that the use of a fermentative pathway for hypoxic/anoxic survival enhances tumor survival since the cells avoid oxidative stress which could prove to be detrimental to the cell survival17. This postulate is supported by a recent study which shows that in cardiomyocytes, hypoxia reduces the oxidative stress placed on the tumor cell17.
To date, the essential nature of a fermentative pathway for anoxic mammalian cell survival has been ingrained in the literature, due, in large part, to an inability to culture mammalian cells in the complete absence of oxygen for more than 3 days. However, an alternative to glycolysis for the anaerobic survival occurs in bacteria. In certain bacteria, nitrogen or sulfate (among other compounds) can serve as terminal electron acceptors for the cytochrome oxidase system in the absence of oxygen18. Although the bacterial and eukaryotic evolution occurred in parallel since diverging from the last universal common ancestor, it is estimated that mitochondria were providing energy to cells for 1.542 million years before the oxygenation of the oceans19. Since researchers have shown that the isolated mitochondria can produce ATP in the absence of oxygen, with nitrite as the terminal electron acceptor, it is reasonable to assume that cells could function for periods of time longer than 3 days in the absence of oxygen20,21,22,23. The methodology described in this study has a utility for the anaerobic mammalian cell growth of numerous cell lines.
1. Prepare Media for Anaerobic Culture of Various Mammalian Cell Lines
2. Anoxic Cultivation of Various Mammalian Cell Lines
3. Assessment of Phenotypic Cell Differentiation by Microscopy of Anaerobically Cultured Cells
The strength of this protocol lies in its support of the longevity and the growth of multiple cell lines and in the recognition that there are a profound alteration and divergence in the cell morphology25. The most critical element of this study is the transfer and maintenance of the cells under strict anaerobiosis. This requires an anaerobic chamber organized to maximize the protocol (Figure 1) and the assurance that the cells removed for either microscopy or other testing are kept in a sealed environment. To date, the range of anoxic survival was 15–17 days for immortal cancer cell lines and up to 17 days for transformed cell lines (nine cell lines screened for anoxic growth). The levels of viable cells, once these cells were moved to anoxic conditions, declined at least 10%, to a maximum of 90%, before they stabilized. All cell lines reverted to the dual phenotypes (tethered and runagate) with the proportion of cells in the suspension increasing over time. Even after an extended incubation of several weeks, there remained some tethered cells, but the majority of the cells were in the rounded runagate morphology (Figure 2). As reported, these runagate cells are viable and proceeded through their cell cycle. It should be noted that the variables that may need to be adjusted include the glucose levels and the nitrite concentration. That said, from the over 60 different media formulations that were tested for their ability to support both normoxic and anoxic incubation, the PS-74656 medium was able to support the survival and replication of nearly all the cell lines tested.
An essential observation was the differentiation of anoxic cells, regardless of the cell tested, into two distinct phenotypes, tethered and runagate (Figure 2). Analogous to the normoxic monolayer cells, the tethered cells were attached to the tissue culture plastic. However, the attachment was weaker, easier to trypsinize or dislodge by scraping, than the attachment that occurred under the normoxic conditions. In addition, the cell morphology was smaller with more spindle (dendritic) extensions26. The second cellular morphotype observed were cells in suspension (i.e., runagate cells). These cells spontaneously converted into free-floating single cell suspensions in the presence of tissue culture-treated plastic, which were rounded with a smaller cell size and less cytoplasm. They differed from the classical suspension culture cells, which typically require an enzymatic adaptation, grow in aggregates, require non-tissue culture-treated containers, and are larger than their normoxic counterparts27. Over time, the anoxic cell populations could be observed to shift from a predominantly tethered monolayer to that of the runagate cells in suspension. These cells in suspension were viable and replicating; thus, care had to be exercised should the supernatant be removed. Also, of note is that the medium did not require changing over the entire anaerobic incubation period. The medium pH indicator under the anoxic incubation, in contrast to its normoxic control, was also unchanged (i.e., remained red). Should the investigator wish to harvest the runagate cells, or change the medium, the cells can be obtained using anaerobic gas-flushed micro-centrifuge tubes that are closed tightly within the anaerobic chamber. For the cell washing of tethered cells, the procedure must also be performed in the chamber to eliminate any oxygen exposure.
The final cell viability for a long-term cultivation was, to a great extent, dependent on the cell being used and the confluence to which the wells or the flask were initially seeded (Figures 3 and 4). That said, a good starting point is 80%, which had the greatest chance of optimal growth. In addition, a cell line-dependent loss of 10% to 60% of viability can occur in the initial 2–4 days. However, during the performance of this experiment, the remaining cells cycled through the cell cycle and replicated28. The seeding confluence effect on the longevity of the cell cultivation can be seen in Figures 3 and 4. Both HeLa 229 and Vero cells were seeded in 24-well plates to 40%, 60%, and 80% of initial confluence (100% confluence = 2.8 x 105 cells/well) and then incubated in the absence of oxygen for 17 days. The HeLa 229 cell anoxic growth was optimized at an initial seeding confluence of 80%. In contrast, the Vero cells optimal seeding confluence was 60%. This demonstrates the need to initially optimize the seeding density dependent on the cell line used.
Figure 1: Model of anaerobic chamber set-up for culture of mammalian cells. The minimal materials required include tissue culture plates, sterile transfer pipets, biohazard bags, de-gassed media, a tube for the discarding of media or other liquids, and resealable bags for the transfer of the plates to the microscope while maintaining anaerobic conditions.
Figure 2: Anoxic HeLa cells morphologic differentiation. (A) This panel shows the phase contrast microscopy of the normoxic HeLa cell controls 4 days post-inoculation, with a 20X magnification. (B) This panel shows the anoxic HeLa cells 17 days post-inoculation (with the unchanged from day 1 PS-74656 medium, 80% seeded confluence, and a 20X magnification). (C) This panel shows microscopy artifacts of water droplets on the interior of a sealed plastic bag containing a 24-well anoxic plate. Please click here to view a larger version of this figure.
Figure 3: Ratio of anoxic to normoxic HeLa 229 viable cells post-17 days incubation. This panel shows the effect of the seeding density on the HeLa 229 cell viability in the PS-74656 medium as compared to the normoxic control. The results are shown as the mean ± SEM.
Figure 4: Ratio of anoxic to normoxic Vero viable cells 17 days post-incubation. This panel shows the effect of the seeding density on the Vero cell viability in the PS-74656 medium as compared to the normoxic control. The results are shown as the mean ± SEM.
This method represents first-time mammalian cells that were cultured for extended periods of time in the absence of oxygen. Based on current observations, the anoxic growth capability via a non-fermentative pathway appears to be universal amongst mammalian cells lines28, where the anaerobic growth resulted in phenotype divergence. This was observed for all cell lines tested. With the anaerobic cultivation, increasing proportions of the cells became rounded, developed a suspension-like population, and were able to replicate. The secondary cell type remained attached to the substrate; however, the cell morphology reverted to a spindle (dendritic) shape. These findings point out a common error in analyses and observations with regard to anoxic tissue culture cultivation. It is generally accepted that when mammalian cells grown under normoxic conditions (5% CO2) round up and detach, they are dying. The contrary appears to be correct for cells grown in the absence of oxygen, in which case, cells that have rounded up and detached, are simply a morphologically divergent cell subpopulation.
The optimization of anoxic growth conditions is cell line dependent. Medium PS-74656 overall supported the growth of all cell lines under normoxic and anoxic growth conditions, with the exception of MDA-MB-231, a breast cancer cell line which preferred growth in the absence of nitrite. An encouraging observation is that even with a 98% die-off of cells per well, the remainder replicates, resulting in a highly adapted population. The critical action is to maintain cultures even when only a few cells are directly observed by microscopy. Thus, care must be exercised to ensure that the detached cells are dead, either by trypan blue dye exclusion or by an alternative cell viability detection method. It should be noted that with such a large decrease in cell numbers, sufficient wells must be seeded to have adequate total cell numbers for subsequent experiments. The expected loss of viable cells is both cell line and seeding confluence dependent; thus, optimal cell numbers, over time, must be determined experimentally and cannot be extrapolated from findings with other cell lines. That said, a good starting point is an 80% confluence, which has the greatest chance of optimal growth. It must be noted that any exposure to oxygen during the anaerobic steps in the procedure, no matter how brief, will compromise the findings due to the reported spontaneous cellular HIF-1-alpha response11,12.
For this procedure, the most critical step is that of the de-gassing of the medium, tubes, and all other apparatuses and materials used. The introduction of even very low levels of oxygen could stop the transition from the fermentation metabolism to the anaerobic respiration. Alternatively, it could prolong the fermentative metabolism so that any experiments result in the measurement of hypoxia, not anoxia. The use of this novel procedure to measure cytotoxicity and chemotherapeutic activity could help to explain drug failure in the treatment of solid tumors.
The authors have nothing to disclose.
The authors thank the Midwestern University Office of Research and Sponsored Programs for their support.
Whitney A35 anaerobic chamber | Don Whitley Scientific | Microbiology International | The ability to remove the front of the chamber makes it easy to put instruments inside without concern about getting them through the portals. |
CO2 incubator | Fisher Scientific | 3531 | |
Tissue Culture Hood | Labconco | DO55735 | |
pipet aid | Drummond | 4-000-100 | |
sterile transfer pipets | Santa Cruz | sc-358867 | |
50cc sterile conical centrifuge tubes | DOT | 451-PG | |
vaccum jar | Nalgene 111410862889 | BTA-Mall 5311-0250 | |
DMEM high glucose (4.5 g/L) | CellGro | 10-017-CM | |
DMEM low glucose (1 g/L) | CellGro | 10-014-CV | |
FBS | VWR | 1500-500 | |
HBSS | VWR | 20-021-CV | |
trypsin | VWR | 25-052-CI | |
gentamicin | Gibco | 15750-060 | |
sterile pipets | CellTreat | 229210B, 229205B | |
Tissue culture flask (T75 or T150) | Santa Cruz | sc-200263, sc-200264 | |
24 well tissue culture treated dishes | DOT | 667124 | |
glad ziplock sandwich bags | Ziploc | Costco | |
inverted phase microscope (10, 20 40x objectives with camera mont) | Nikon Eclipse | TS100 | |
trypan blue | Invitrogen | T10282 | |
hemocytometer | Invitrogen | C10283 | |
Countess Automated Cell Counter | Life Technologies | AMQAF1000 | |
Rainin microtiter pipets | Mettler Toledo | Rainin Classic Pipette PR-100 | |
microtiter tips | Santa Cruz Biotechnology | sc-213233 & sc-201717 | |
test tube rack (50cc tubes) | The Lab Depot | HS29050A | |
sodium nitrite | Sigma | https://www.sigmaaldrich.com/catalog/product/sigald/237213?lang=en®ion=US | |
clear boxes with lids | Rosti Mepal | Rubber maid | |
paper towels | In House | ||
cell scraper | CellTreat | 229310 | |
PBS | In house prepared | ||
70% Ethanol | Fisher Scientific | 64-17-5 | |
microcentrifuge tubes sterile | Santa Cruz | sc-200273 | |
biohazard bags with holder (desktop) | Heathrow Scientific | HS10320 | |
Nitrile Gloves | VWR | 89428 | |
oxygen electrode | eDAQ | EPO354 | |
pH meter | Jenway | 3510 | |
pH paper/ pHydrion | Sigma Aldich | Z111813 |