Removal of agarose gel and suspension culture

Spheroids in agarose gels were treated with agarase (10Uml⁻¹, Sigma) for 2 days. Spheroids were separated from the degraded agarose gel by centrifugation (1000r.p.m., 3min), and resuspended with culture medium.

Morphological study of spheroids under solid stress

Spheroids cultured in agarose gel for 10 days were fixed with 4% paraformaldehyde for 3h, washed in PBS and then fixed in osmium (1%) for 2h. The spheroids were then dehydrated in increasing concentrations of ethanol and embedded in Polybed 812. Sections with a thickness of 2μm were stained with toluidine blue.

Labelling of hyaluronan with the hyaluronan binding protein

AT-2.1 and AT-3.1cells were grown in 1% agarose or in free suspension. After 3 weeks, spheroids were fixed in 4% paraformaldehyde. To facilitate handling, free suspension spheroids were concentrated by centrifugation and embedded in agarose. Sections 5μm thick were prepared from paraffin blocks. The paraffin was removed, and the slides were washed with PBS between each step. The slides were incubated with the biotinylated hyaluronan binding protein (Calbiochem, CA, USA), then with streptavidin-coupled horseradish peroxidase), and finally with diaminobenzidine. Negative and positive controls ensured that the staining was specific for hyaluronan (data not shown). The slides were counterstained with haematoxylin and Scott's water, and then dehydrated and coverslipped.

Solid stress inhibits spheroid growth

To determine the effect of increasing solid stress on growth kinetics, we cultured both cell lines in gels of various concentration and measured the spheroid size as a function of time. We also fit the growth curves with the Gompertz equation. The model yields two parameters: the specific growth rate, a (day⁻¹) and the maximum (asymptotic) spheroid diameter, D_max (μm). As shown in Figure 2, increasing the gel concentration from 0.5 to 1.5% agarose decreases D_max from ~150 to ~65μm. In 1.8% agarose gels the growth is completely arrested. We suggest that this inhibition in growth results from increasing stress as trapped spheroids grow within denser gels, consistent with our previous data on five different cell lines: LS174T (human colon adenocarcinoma), MCaIV (murine mammary carcinoma), and the clones (A, B, C) of a rat rhabdomyosarcoma, BA-HAN-1 (Helmlinger et al, 1997a).

Unlike our previous study, we found that the specific growth rate, a, decreased with increasing gel concentration for AT2.1 (P=0.023), but not for AT3.1 (Figure 2C). Without knowing the cellular proliferation rates (e.g. PCNA staining), apoptosis rates (e.g. TUNEL staining) and cell densities (e.g. PI staining), it is difficult to determine the mechanisms underlying this difference. In parallel shorter-term experiments, cells were seeded in agarose gels at both higher (up to 20×10³cellsml⁻¹) and lower (2.5×10³cellsml⁻¹) seeding densities, and corresponding growth curves were also fit with the Gompertz equation. Although the limiting size D_max could not be determined accurately for these experiments due to their shorter duration, spheroids grown in gels with low seeding densities reached sizes of 300μm (AT2.1) and 250μm (AT3.1) by day 30, exceeding maximum measurements at other seeding densities, even at longer times. Accurate estimation of the growth rate a, which is determined largely by the rapid growth phase and not the long-term data, was possible and showed no significant variation with seeding density at any gel concentration. Thus, we note that seeding density can affect asymptotic spheroid size but does not appear to affect growth rate (data not shown). We speculate that this result is due to solid stress: stress increases more rapidly in gels containing a high density of (growing) spheroids which thus limits their growth.

Relieving stress causes loss of spheroid integrity

When stress around plateau–phase spheroids was relieved by treatment of the gel with agarase, the two cell lines responded differently. Similar to our previous findings (Helmlinger et al, 1997a), the AT2.1 spheroids retained their morphology and resumed their growth as in free suspension (Figure 3). The AT3.1 cells also retained their morphology for the first several days, but lost their integrity by days 6–8 and disintegrated into a loose collection of cells. The latter were reminiscent of cellular aggregates in free suspension. However, the spheroids did not lose their integrity as rapidly upon treatment with agarase as would be expected for unassociated cells held in proximity only by gel. Evidence for this comes from the fact that pelleting by centrifugation did not contribute to aggregation, and the pelleted cells were readily separated (data not shown). These results again support the hypothesis that solid stress facilitates spheroid formation by inducing some form of adhesion between cells.

Open in a separate window

Figure 3

Enzymatic digestion of agarose gel cultures after 30 days of growth. AT2.1 and AT3.1 are shown in the left and right columns, respectively. Images shown are from 3, 6, and 8 days after release of spheroids from gel. Note that the AT3.1 cells return to a dis-aggregated state, but the AT2.1 maintain their spheroidal morphology at 8 days after release. Scale bar=100μm.

Morphological differences may result from the stress response

To gain insight into the differential behaviour of AT2.1 and AT3.1 cell lines with respect to spheroid formation, we prepared histological sections of spheroids. As seen in Figure 4, the cells of AT3.1 spheroids in agarose gels have polygonal shapes and are separated by uniform spaces with a regular width (0.45–0.8μm) occupied by the extracellular matrix (Figure 4A). On the other hand, there appears to be (at least at the light microscopic level) no discernible spacing between most of the cells in AT2.1 spheroids in agarose (Figure 4B), and AT2.1 and AT3.1 cells in free suspension (Figure 4C). Large extracellular spaces are observed occasionally in AT2.1 spheroids and in free suspensions of 2.1 spheroids and 3.1 aggregates (Figure 4B,C). Spheroid formation by AT3.1 cells growing in agarose could result solely from the stiffness (or the limited compliance) of the agarose gel which would compact the cells and the matrix. Alternatively, cell–matrix adhesion may also contribute to spheroid formation in AT3.1. This is supported by the observations that (a) AT3.1 cells are closely apposed in free suspension aggregates but separated by matrix in agarose gels (Figure 4A,C), and (b) the adhesion holding together the large spheroids is not immediately reversed upon removal from the gel. Although formation and compaction of spheroids can be mediated by E-cadherin (Kantak and Kramer, 1998; St. Croix et al, 1998) it has been shown that AT2.1 and 3.1 cell lines do not express E-cadherin (Bussemakers et al, 1992). Hyaluronan may also be important in spheroid formation, as hyaluronan participates in cell–cell adhesion and cellular aggregation of transformed cells (Underhill and Toole, 1981), and it has been shown to influence both the formation and compactness of spheroids by tumour cell lines (St. Croix et al, 1996, 1998). The addition of hyaluronidase to compact (tight) spheroids induced spheroids with a loose morphology (St. Croix et al, 1996). Also, treatment of tumour cells with hyaluronidase can prevent the formation of cellular aggregates and spheroids (St. Croix et al, 1998; Mueller et al, 2000).

Open in a separate window

Figure 4

Histological sections of spheroids in agarose gels. AT3.1 cells in agarose, are loosely-packed with a uniform matrix space observed between the cells (A). On the other hand, in spheroids of AT2.1 (B) and aggregates of AT3.1 (C) there is no visible matrix between most of the cells. Scale bar=20μm.

This report was supported by an Outstanding Investigator Grant (R35-CA-56591) and a Bioengineering Research Partnership Grant (R24-CA-85146) to RK Jain. TD McKee is supported by a grant from the National Institutes of Health (5 T32 GM08334) Interdepartmental Biotechnology Program, Biotechnology Process Engineering Center at the Massachusetts Institute of Technology.