Coculturing diverse clonal populations prevents the early-stage neoplastic progression that occurs in the separate clones

Abstract

Most human cancers are of monoclonal origin and display many genetic alterations. In an effort to determine whether clonal expansion itself could account for the large number of genetic alterations, we compared spontaneous transformation in cloned and uncloned populations of NIH 3T3 cells. We have reported that progressive transformation of these cells, which is driven by the stress of prolonged contact inhibition at confluence, occurs far more frequently in cultures of recent monoclonal origin than in their uncloned progenitors. In the present work we asked how coculturing six clones at early and late stages of progression would affect the dynamics of transformation in repeated rounds of confluence. When coculture started with clones in early stages of transformation, marked by light focus formation, there was a strong inhibition of the progression to the dense focus formation that occurred in separate cultures of the individual clones. In contrast, when coculture started after the individual clones had progressed to dense focus formation, there was selection of transformants from the clone producing the largest and densest foci. Mixing the cells of a single clone with a large excess of uncloned cells from a subline that was refractory to transformation markedly decreased the size of dense foci from clones in transit from light to dense focus formation, but had much less effect on foci from clones with an established capacity for dense focus formation. The major finding of protection against progression by coculturing clones in early stages of transformation suggests that the expansion of a rogue clone in vivo increasingly isolates many of its cells from genetically stabilizing interactions with surrounding clones. Such clonal isolation might account for the increase in mutation rates associated with the dysplasia in colorectal adenomas that signifies the transition between benign and malignant growth.

It generally is accepted that most human cancers are clonal in origin (1–3). Extremely large numbers of genetic changes have been found in these growths. For example, more than 25% of the heterozygous loci in sporadic cases of colorectal cancer exhibit loss of heterozygosity (LOH) (4). More than half of the cases of esophageal adenocarcinoma reveal extensive LOH of polymorphic microsatellite markers along the entire length of a single chromosome (5). A large increase in the frequency of mutations is found in the highly dysplastic lesions that mark the transition in colorectal adenomas to carcinomas (6). This finding suggests that the lesions have to reach a certain minimal size and degree of disorganization to drive the many genetic changes that underlie progression to malignancy. It had been proposed earlier that a critical mass of cells bearing the same or similar mutations was needed to produce the malignant state (7). Quantitation of the critical mass concept closely fit the age-related increase in deaths from mammary cancer in women, although it was acknowledged later that a requirement for multiple mutations also could account for this pattern (8). Experiments with C3H/10T1/2 mouse cells in culture indicated that monoclonal foci of initiated cells had to reach a minimal size against a confluent background of confluent cells before overt transformation was functionally and morphologically expressed (9, 10).

To explore this problem further, we initiated a study in NIH mouse 3T3 cells of the effect of clonal isolation on the dynamics of progressive neoplastic transformation brought about by prolonged incubation under the contact inhibition of confluence. We first found that recently isolated clones underwent progression from light to dense focus formation more quickly than their uncloned parental cultures (11). Six clones of the original transformation-sensitive line then were maintained in frequent low-density subculture, during which the only clone that produced dense foci in the first round of confluence lost that capacity for quick progression (12). This loss allowed us to undertake the present experiments in which we measure the effects of mixing and coculturing the six clones on progressive transformation in repeated rounds of confluence. We also examined the effects of coculturing clones that had already progressed to dense focus formation before mixing. The major result of these studies was that coculture of the clones that had been producing only light foci strongly inhibited their progression to dense focus formation. This partial reconstruction, therefore, complemented and reinforced the earlier finding that recently isolated clones were more readily transformed than their polyclonal progenitors. We discuss the possible role of clonal expansion and destabilization as the driving force for the multiple genetic changes that underlie progression of human tumors to malignancy.

Materials and Methods

Cell Culture and Clonal Mixing.

The original NIH 3T3 line of transformation-sensitive mouse embryo fibroblasts (13), which we designate SA′, was received in 1988, passaged once in DME containing 10% calf serum, and cryopreserved under liquid nitrogen in that medium with 7.5% DMSO (14). It was thawed in 1998 and cultured in MCDB 402 medium (15) containing 10% calf serum. The cells were passaged every 3 days for 5 weeks at low density to minimize spontaneous transformation. Twenty-nine clones were isolated from multiwell plates and expanded in 21-cm² plastic Petri dishes as described (11, 12). Clones were designated by column and row of the well they occupied on the multiwell plates. The low-density passages (LDPs) were resumed, and the clones as well as the uncloned parental culture were assayed for focus formation at confluence after varying numbers of LDPs in which we established that the clones were more easily transformed than their uncloned progenitors (11). Six of the clones were chosen for long-term studies of the clonal dynamics of transformation, which demonstrated the importance of contact inhibition for transformation (12). Once it was established that the clones were more sensitive to transformation than the uncloned parental culture, we wondered whether mixing and coculturing the clones would stabilize them against progressive stages of transformation. This test became possible after about 40 LDPs when the only clone that produced the more advanced dense foci in a single round of assay at confluence lost that capacity. We then mixed the six clones in equal proportions and assayed them for focus formation in three serial rounds of confluence. Another series of assays was initiated with clones that were mixed after the first assay that had endowed them with the capacity for dense focus formation in further assays.

Transformation Assays.

The serial assays for transformation were designated 1°, 2°, 3°, and 4°. Because three serial assay sets had been done previously with the six individual clones after earlier LDPs (12), the present assays were denoted by the prefixes 4 and 5. The transformation assays consisted of seeding 10⁵ cells 21-cm² dishes with MCDB 402 medium plus 2% calf serum (CS) and incubating for 14 days. The cultures then were fixed in Bouin's reagent and stained with 4% Giemsa in phosphate buffer, pH 7.0. Every assay was done on the mixture of the six clones and on each individual clone. Cells also were suspended by trypsinization of sisters to the stained cultures after each assay. They were counted electronically to determine the saturation density and seeded for further serial assays as described below. Foci were classified as light or dense depending on their staining characteristics (see Figs. 1 and 2). The light foci represent an early stage of transformation and consist of cells more densely packed than the surrounding cells but still mainly in monolayer (16). The dense foci, which represent progression to a later stage of transformation, are heavily multilayered and produce sarcomas within a few weeks of s.c. inoculation into nude mice (17). In the event that the seeding of 10⁵ cells might produce so many foci that they overlapped and could not be accurately counted, the cells from individual clones were diluted and 10⁴ or 10³ of the cells were seeded with 10⁵ cells of the transformation-refractory A′ subline. The latter formed a confluent, monolayered background for the display of the foci. This procedure allowed us to estimate the inhibitory effect of a confluent background of transformation-refractory cells on the proliferation of the transformed cells by the reduction in size of dense foci when compared with those seen in an earlier, direct assay of the unmixed transformation-sensitive clone. The purpose of these serial assays was (a) to induce progressive neoplastic transformation at confluence, (b) to select for the capacity of transformed cells to overgrow the confluent sheet of cells, and (c) to quantify the transformation at each stage by counting and photographing the light and dense foci as well as determining the saturation density of the cultures.

Figure 1.

Mixing clones before the 1° assay suppresses the progression from light to dense focus formation that occurs in separate clones in the 2° and 3° assays. The mixture was made at the 41st LDP (fifth series). (a) 5–1°. (b) 5–2°. (c) 5–3°. All assays shown were with 10⁵ cells except those of clones 1A and 3F in the 3° assay, which were assayed with 10⁴ cells plus 10⁵ transformation-refractory A′ cells.

Figure 2.

Mixing clones after the 1° assays allows selection of the clone that produces the largest dense foci. A 1° assay of each clone was done after the 37th LDP (fourth series) and was followed by mixing the clones. Each clone and the mixture were used in further serial assays. (a) 4–2°, 10⁵ cells. (b) 4–3°, 10⁴ cells. (c) 4–4°, 10³ cells. When 10⁴ or 10³ cells were assayed, they were mixed with 10⁵ transformation-refractory A′ cells.

Results

Mixing Clones Immediately Before the 1° Assay.

The purpose of this experiment was to determine whether mixing clones in the earliest stages of transformation, when they produced only light foci in a 1° assay, would prevent the progression to dense focus formation seen in the separate clones by the 2° assay. Aliquots from the six clones (1A, 3F, 2B, 6H, 4B, and 3E) were mixed in equal cell numbers and used for 1°, 2°, and 3° assays in parallel assays with the separate clones. All of the clones except the nonproductive 4B clone produced light foci in the 1° assay as did the mixture of clones (Fig. 1a and Table 1). In the 2° assay, three of the clones (1A, 3F, and 2B) produced moderate numbers of dense foci as well as many light foci (Fig. 1b and Table 1). Clones 6H and 3E produced mainly light foci whereas 4B remained negative. The mixed culture produced only light foci although some 20 dense foci would be expected if progression had occurred to the same extent as in the individual clones (Table 1). In the 3° assay, there were so many dense focus formers in clones 1A and 3F that they had to be diluted and assayed with an excess of 10⁵ transformation-refractory A′ cells to form countable foci (Fig. 1c). Based on the numbers of dense foci in the assays of the clones, more than 1,500 of them were expected in the mixture but only one unequivocally dense focus and four moderately dense foci appeared, in addition to the many light foci (Fig. 1c and Table 1). It is evident that mixing the clones in the 1° assay in which five of them produced light foci almost completely suppressed the progression to dense focus formation seen in the separate clones.

Table 1.

Comparison of focus formers/10⁵ cells and saturation densities in serial assays of individual clones and of a mixture prepared from them before the 1° assay

	Focus formers/105 cells Serial assay			Saturation density (×10−5 cells) Serial assay
	1°	2°	3°	1°	2°	3°
1A	0*–41†	20–tntc	8,600–3,100	7.5	21.1	46.2
3F	1‡–15	75–tntc	570–400	4.2	18.5	27.5
2B	0–59	21–74	50–135	7.4	17.8	45.7
6H	0–46	1–tntc	0–tntc	8.7	9.1	14.2
4B	0–0	0–0	0–0	4.6	4.0	4.5
3E	0–2§	0–tntc	2–tntc	10.0	11.9	14.9
Average	0¶	20¶	1,537¶	7.1 ±2.3	13.7 ± 6.5	25.5 ± 17.4
Mixture	0–18	0–tntc	1–tntc	8.9	18.1	28.8

Clones were mixed in equal aliquots immediately before the 1° assay. Three serial assays for focus formation were done with the individual clones and with the mixture. Values for focus formation greater than 150 were normalized to that of 10⁵ cells based on seedings of 10⁴ or 10³ cells with an excess (10⁵) of transformation-refractory A′ cells. See Fig. 1 for photographs. tntc, too numerous to count. 

^*Dense foci. 

^†Light foci. 

^‡There was the small beginning of a dense focus in the 1° assay of clone 3F, but no such lesion appeared in the mixture, nor in the 1° assay of a later passage. 

^§Grainy. 

^¶The average refers only to dense foci. 

The saturation density of all but one of the clones shown in Fig. 1 increased in the sequential assays (Table 1). Clones 1A and 2B had the highest saturation densities in the 2° assay, and these were equaled by the clonal mixture. However, in the 3° assay, the saturation density of the mixture was far lower than that of the two highest-density clones. The mixture of clones therefore prevented the selective overgrowth expected had there been progression like that of clones 1A and 2B. This selective overgrowth does occur when the mixture is made after the individual clones have progressed to dense focus formation (see below). The measurement of saturation density supplemented the observation of foci in demonstrating the inhibitory effect of clonal mixing on progressive transformation. It also demonstrated, however, that the suppression was not complete because there was a gradual increase in saturation density of the clonal mixture in successive serial assays.

Mixing Clones After the 1° Assay.

This experiment was designed to see whether mixing clones would prevent the expression of dense focus formation by clones that had just acquired that capacity in separate assays. The clones were mixed after they had separately been through a 1° assay, in which five of them gained the capacity to produce dense foci in the 2° assay. The number of dense foci varied in the 2° assay from clone to clone, and the mixture produced them in numbers equal to the average of the separate clones (Fig. 2a and Table 2). The 3° assay had to be done with only 10⁴ cells from each sample mixed with 10⁵ A′ cells because there were so many focus formers in assays of 10⁵ cells that the foci overlapped in four of the clones and in the mixture. In fact, the mixture had more dense foci than the average of the clones (Fig. 2b and Table 2). The 4° assay was done with only 10³ cells of the samples (combined with 10⁵ A′ cells) because of the very large numbers of dense focus formers in four of the clones. The number of dense foci in the mixture was about half the average of the clones (Fig. 2c and Table 2). The morphology and number of the dense foci resembled those produced by clone 2B, which were the largest foci produced by any clone. This finding is consistent with selection of the most transformed cells at confluence in the serial assay.

Table 2.

Focus-forming cells and saturation densities in serial assays when clones were mixed after the 1° assay in preparation for the 2° assay

	Focus formers/105 cells Serial assay				Saturation density ×10−5 cells Serial assay
	1°	2°	3°	4°	1°	2°	3°	4°
1A	0*–12†	4–tntc	20–180	150–500	8.0	20.9	17.7	22.1
3F	0–13	20–43	660–tntc	800–1,400	7.3	19.3	31.7	24.2
2B	0–3	25–17	320–140	800–2,600	7.9	13.0	58.7	45.8
6H	0–5	0–40	0–tntc	400–8,000	6.1	9.4	13.1	13.1
4B	0–0	0–0	0–0	0–12	6.8	3.1	2.9	3.2
3E	1–0	21–14	1,160–tntc	7,700–1,600	16.2	21.1	35.0	25.1
Average	0.17‡	11.7‡	360‡	1,640‡	8.7 ± 3.1	14.5 ± 17.3	26.5 ± 19.8	22.3 ± 14.2
Mixture	—	10–21	780–tntc	800–2,300	—	22.6	57.2	51.3

Clones were mixed after the 1° assay, and three further assays were done with the individual and the mixed clones. Values for focus formation greater than 150 were normalized to that of 10⁵ cells based on seedings of 10⁴ or 10³ cells with an excess (10⁵) of transformation-refractory A′ cells. See Fig. 2 for photographs. tntc, too numerous to count. 

^*Dense foci. 

^†Light foci. 

^‡The average refers only to dense foci. 

The saturation density of all of the clones, except 4B shown in Fig. 2, increased in sequential assays, but clone 2B was clearly higher than the others in the 3° and 4° assays (Table 2). The clonal mixture had the saturation density of the highest two clones in the 2° assay and of clone 2B in the 3° and 4° assays. This finding indicates that the dense foci fully expressed themselves in the mixture, with clone 2B becoming dominant, as indicated by the type of foci visible in the 4° assay (Fig. 2c).

Variable Suppression of Dense Focus Formation in the Separate Clones by an Excess of Transformation-Refractory A′ Cells.

Although the dense focus formers were able to grow to the maximum extent in the clonal mixture, there remained a question whether they would be able to do so in the 2° assay if each individual clone was seeded with an excess (10⁵) of the transformation-refractory A′ cells. In fact, this combination resulted in an almost complete suppression of both dense and light focus formers in the 2° assay of even the most heavily transformed clones. Close inspection of the cultures revealed a few pinpoint foci in some of the clones (not shown), but nothing like either the dense or light foci in some clones of Fig. 2a. Hence, the excess of transformation-refractory A′ cells was more suppressive of focus formation at this early stage of dense focus formation than was coculturing the mixture of the six SA′ clones. This result is in contrast to the 3° and 4° assays of Fig. 2 b and c in which the clones and mixture apparently had undergone further progression because they did produce foci when diluted and seeded with an excess of the A′ cells. However, the foci on a confluent background of A′ cells were smaller than those seen in Fig. 2a in which the clones and the mixture were each seeded without the addition of the A′ cells. Therefore the refractory A′ cells still exerted some inhibitory effect on the proliferation of the transformed cells that had progressed beyond the intermediate, more suppressible stage at the borderline between the light foci of the 1° assay and the dense foci of the 2° assay. These results with the A′ background reveal different degrees of progression that are not evident by visual examination of the foci surrounded by nontransformed sister cells of the same clone.

Discussion

The most novel result presented here is that the progressive neoplastic transformation seen in separate clones under the contact inhibition of confluence is strongly inhibited if the clones are mixed when they are still in an early stage of transformation. This finding is complementary to our earlier observation that recently isolated clones are more sensitive to the transforming effect of prolonged confluence than their polyclonal parental populations (11). The latter observation was more striking in a transformation-refractory subline of NIH 3T3 cells, where the uncloned parental subline exhibited no sign of transformation in four serial assays at confluence, than it was in the transformation-sensitive line used in the present mixing experiments, probably because one of the clones from the latter had already produced dense foci in the 1° assay (11). Those transformed cells would tend to dominate the culture in a 2° assay, thereby obscuring the inhibition of progression by the less transformed clones. However, the transformation-refractory subline was not suitable for the clonal mixing experiment because it is difficult to discern discrete progressive changes in the relatively small and irregular foci produced by these cells. It became possible to do the mixing experiment in the transformation-sensitive line when the 1A clone, which produced the dense foci in the 1° assay, lost the capacity to do so in the course of LDPs (12).

The results from both the original experiments with the uncloned parental line (11) and the present clonal mixing experiment indicate that the interactions between diverse clonal populations inhibit the initiation of the genetic changes that underlie progression rather than merely inhibiting the development of foci by cells after they had transformed. It has been shown that contact inhibition of cell proliferation induces progressive transformation (12, 18) probably through nonspecific damage to cells (19). A strong precedent for such mutagenic effect of stressful conditions is the 10⁴-fold increase in excision of insertion sequences from DNA of bacteria subjected to starvation (20). The transforming role of contact in cultured cells and its mitigation in polyclonal cultures suggest that junctional communication among cells of diverse clones relieves some of the stress of contact inhibition perhaps through the sharing of small molecules across gap junctions (21, 22). Because cells in aging animals accumulate mutations at random (23) and these are likely to differ from clone to clone, the metabolic products from one clone would complement those lost by mutation in another clone and vice versa. The heterogeneous mixture of clones therefore would be more stable than its individual isolated components.

Although we are unaware of other reports of polyclonal interactions protecting against progressive transformation, a related phenomenon has been described in B16 mouse melanoma cells. Polyclonal mixtures of these cells stably maintained the different levels of metastatic capacity of the individual clones through many passages in vivo or in vitro (24). Separate serial passage of the individual clones in vivo or in vitro, however, resulted in great spread of metastatic capacity within each clone. By contrast, if six of the clones were cocultured, the characteristic metastatic capacity of each clone was maintained through many passages. Hence, polyclonal stabilization has been demonstrated for progressive transformation in NIH 3T3 cells and for metastatic capacity in B16 cells.

Most solid epithelial cancers progress from benign precursor states (25). For example, colorectal carcinoma arises from adenomas (26), mammary carcinomas develop from hyperplastic nodules or plaques (25), and hepatocarcinomas appear in persistent hepatocyte nodules (27). The benign growths result from clonal expansion (2, 28). The expanding size of the clone that gives rise to the benign tumor would be expected to shield many of its cells from contact interaction with surrounding clones. Our results suggest that such clonal isolation would increase the frequency of mutational events and thereby promote progression to the malignant state. It has been reported that colorectal adenomas exhibit a significant increase in mutations when they become highly dysplastic (6). The dysplasia, which includes multicellular disorganization, would tend to exacerbate the clonal isolation of the tumor cells (21, 29) and thereby could account for the observed increase in mutation frequency.

Destabilization by clonal expansion and isolation also may contribute to the exponential increase of cancer with age in humans (30). There is a decrease in functional units in tissues with age, e.g., the number of alveoli in human lungs may decrease from 300 × 10⁶ in young adults to 60 × 10⁶ in old age, and the number of nephrons in kidneys decreases from 2 × 10⁶ to 1 × 10⁶ in the same period (31). This decrease would allow for compensatory hyperplasia in the remaining functional units. The mammary epithelium of most young female Fischer rats has a small number of cells with a spontaneous oncogenic mutation of the Ha-ras-1 gene (32). The number of these cells selectively increases about 25-fold between 50 and 570 days of age, whereas the total number of mammary epithelial cells increases only 5-fold (32). The mutated cells remain normal unless the rat is treated with N-nitroso-N-methylurea (NMU), when they develop into mammary tumors (33). NMU, usually considered a mutagen, apparently selects for further growth of the spontaneously mutated H-ras-1 cells, which then undergo further mutations and progress to malignancy. This process is consistent with the idea that clonal expansion per se is mutagenic and hence carcinogenic. Other examples where this explanation might be considered for a role in age-related carcinogenesis may be found in the atrophy with compensatory hyperplasia of the prostate (34), stomach (35), and other organs in aging humans (36).

Abbreviation

LDP

low-density passage