Occurrence and control of sporadic proliferation in growth arrested Swiss 3T3 feeder cells

Abstract

Growth arrested Swiss mouse embryonic 3T3 cells are used as feeders to support the growth of epidermal keratinocytes and several other target cells. The 3T3 cells have been extensively subcultured owing to their popularity and wide distribution in the world and, as a consequence selective inclusion of variants is a strong possibility in them. Inadvertently selected variants expressing innate resistance to mitomycin C may continue to proliferate even after treatment with such growth arresting agents. The failure of growth arrest can lead to a serious risk of proliferative feeder contamination in target cell cultures. In this study, we passaged Swiss 3T3 cells (CCL-92, ATCC) by different seeding densities and incubation periods. We tested the resultant cultures for differences in anchorage-independent growth, resumption of proliferation after mitomycin C treatment and occurrence of proliferative feeder contaminants in an epidermal keratinocyte co-culture system. The study revealed subculture dependent differential responses. The cultures of a particular subculture procedure displayed unique cell size distribution and disintegrated completely in 6 weeks following mitomycin C treatment, but their repeated subculture resulted in feeder regrowth as late as 11 weeks after the growth arrest. In contrast, mitomycin C failed to inhibit cell proliferation in cultures of the other subculture schemes and also in a clone that was established from a transformation focus of super-confluent culture. The resultant proliferative feeder cells contaminated the keratinocyte cultures. The anchorage-independent growth appeared in late passages as compared with the expression of mitomycin C resistance in earlier passages. The feeder regrowth was prevented by identifying a safe subculture protocol that discouraged the inclusion of resistant variants. We advocate routine anchorage-independent growth assay and absolute confirmation of feeder disintegration to qualify feeder batches and caution on the use of fetal bovine serum.

Fig 1. P6 and P7 cell yields and responsiveness to Mitomycin C.

Column chart representing the influence of subculture schemes on Swiss 3T3 cell outputs in T75 flasks at 7^th passage (P7) and responsiveness to Mitomycin C (MC). The uniformly set up 6^th passage (P6) cultures from a working bank were subcultured to P7 by schemes of 3K3D, 3K4D and4K3D (* P<0.05). Shaded columns represent exhibition of resistance to MC and each column represents a mean value from 3 flasks with the standard deviation.

Fig 2. Feeder regrowth in 3K4D culture after mitomycin C treatment.

The newly formed bipolar cells (F) appeared amidst the attached MC treated large cells (f) and the floating cellular debris (arrowhead, A) after four weeks of post-treatment period. Five weeks later, the same field showed a distinct large collection of multipolar cells with high nucleus to cytoplasm ratio (B). Scale bar: 100 μm.

Fig 3. Influence of subculture schemes on responsiveness to Mitomycin C.

The 3K3D feeders plated either alone (3T3) or co-cultured with human epidermal keratinocytes (3T3 + Kc) showed degeneration of feeders (arrowhead) and a large well circumscribed keratinocyte (K) colony surrounded by inactivated feeders. The 4K3D feeders plated alone exhibited newly formed compact proliferative feeder cells with high nucleus to cytoplasm ratio (arrows) over a background of enlarged (f) and degenerating cells (white arrow head). The co-culture with human epidermal keratinocytes showed the proliferative foci (arrows). The 3T3 alone and the co-cultures were 2 and 1 weeks old, respectively. Scale bar: 100 μm.

Fig 4. Proliferative feeder cell contamination.

Keratinocyte cultures plated alone without the feeders showing cytological features of attached keratinocytes (K) and the contaminating non-proliferative (f) and proliferative feeders (F). The 3K3D feeder cells that came along with the keratinocytes as contaminants (Left panel) presented no dividing cells and were broad with large nuclei in phase contrast and showed coarse chromatin aggregates (arrows) that stained bright in Hoechst (inset of Hoechst image). The keratinocytes comprised of a mix of small polygonal or broad terminally differentiated cells which presented the dull small nuclei. The 4K3D contaminating feeder cells (Right panel) consisted of well circumscribed keratinocyte colonies (K) enveloped by numerous newly formed narrow-bodied feeder cells (F) with several cell divisions (inset of Hoechst image) in addition to a few broad non-proliferating feeder cells (f) showing vesicular nuclei. Scale bar: 100 μm; Left inset: 200 μm; Right inset: 450 μm.

Fig 5. Anchorage-independent growth assay.

Swiss 3T3 fibroblasts from different passages during serial subculture by the 3K3D and 4K3D schemes were plated in methylcellulose. Discrete cells from 7^th Passage (P7) cultures of 3K3D (A) and 4K3D (D) subculture schemes showed smooth outline. Cells with angular outlines having a shallow contrast were observed at P15 of 3K3D (B) and P11 of 4K3D (E) with occasional linear aggregates. Conspicuous 3-dimensional spheres appeared in the subsequent passages of P16 (C) and P12 (F). Scale bar: 100 μm.

Fig 6. Cell yields during serial subculture.

Scattered plot showing the influence of serial subculture of Swiss 3T3 fibroblasts by schemes of 3K3D and 4K3D on cell outputs and responsiveness to Mitomycin C treatment at each passage. The 3K3D cultures were serially subcultured either by the same scheme or the 4K3D scheme. Shaded markers denote passages at which MC treated cells showed proliferative foci. The columns in the inset represent the number of cells seeded (Input) and yield (Output). Each column represents average cell number of all the passages from each scheme. The shaded column represents consistent resistance to Mitomycin C at all passages.

Fig 7. Establishment of spontaneously transformed clone.

The culture of 7^th passage Swiss 3T3 fibroblasts initiated with 3000 cells per cm²after 3 days showed cells with uniform morphology (A) which reached confluence in five days (B) and formed few transformation foci in ten days (C). After two weeks, the cells from such foci were scraped, trypsinized briefly and the single cells were re-plated in a fresh flask. Several discrete clusters of small cells with compact cell bodies and high nucleus to cytoplasm ratio appeared in two days (D) which grew steadily forming conspicuous colonies in four days (E) and enlarged further in ten days (F) over a background of normal looking broad cells. Two weeks later, the trypsinized cells from this culture formed discrete spheres in methyl cellulose (G) which were subjected to the single sphere cloning in 24-well plates (H). The spheres readily anchored overnight (I), spread out to form small colonies in three days (J) and established as large dense colonies in ten days (K). One such colony further expanded in flasks presented narrow bipolar and tripolar cells (L). Scale bar: 100 μm.

Fig 8. Failure of growth arrest in spontaneously transformed clone.

The 2^nd passage culture of the clone exposed to Mitomycin C and replated alone (A) exhibited the proliferative focus consisting of the newly formed compact cells (small arrows) surrounded by the broad cells (arrowhead) after 14 days of incubation. The co-culture of the growth-arrested clone cells with the human epidermal keratinocytes (B) presented the well-spread out degenerating cells (arrowhead) and the newly formed cells (small arrows) at the periphery of a large colony of keratinocytes (K) after 7 days. Scale bar: 100 μm.

Fig 9. Growth curves.

Growth curves of Swiss 3T3 cells of P5, spontaneously transformed clone and the P7 cells produced by subculture schemes of 3K3D and 4K3D.

Fig 10. Graphical representation of cell size distribution.

Area plot showing Swiss 3T3 cells of P5, spontaneously transformed clone and the P7 cells produced by subculture schemes of 3K3D and 4K3D.