Solid stress inhibits the growth of multicellular tumor spheroids

Introduction

The growth, morphogenesis, and homeostasis of tissue are tightly regulated by the stress field; examples can be found in the bone (external load) and blood vessel (systemic pressure). Vice versa, stress can be induced by cellular processes; examples are traction forces and residual stresses in tissues.

Researchers hypothesized that the stress may affect

• Tumor growth rate
• Growth pattern in vitro/in vivo
• Tumor physiology and blood flow - vessel wall collapse, chronic vascular/lymphatic occlusion
• Metastasis

However, the stress generated was difficult to quantify and we did not know how the stress affects proliferation/apoptosis/cellular density. The tumor growth in vivo also involves a gradual displacement of the surrounding matrix.

In this study, the authors addressed these concerns by seeding tumor cells in an agarose gels (the cells formed spheroids) - an in vitro study. The effect of the external stress was simulated by deformation of the gel. Also, when the spheroids grow, they generated stress, which can be calculated from the size of the spheroids and properties of the gel.

Techniques used:

• TUNEL (Terminal deoxynucleotidyl transferase dUTP Nick End Labeling) - apoptosis
• PCNA (Proliferating Cell Nuclear Antigen) marker - proliferation
• PI (Propidium iodide) staining - cell density

Results

Kinetics of tumor spheroid growth modulated by solid stress?

Increasing agarose concentration increases initial stiffness of the matrix.

In human colon adenocarcinoma (LS174T):

• Low concentrations (<0.8%): Growth rates and final sizes were shown to not change much when the concentration changed, but were significantly lower than free-suspension controls.
• High concentrations (>0.9%): Growth was inhibited, lower clonal efficiency (?) (the ability of the individual cells to form colonies in culture)
  • Threshold for significant growth inhibition at 0.9-1% conc. Sizes not decrease progressively.
• Clonal efficiencies are similar across all conditions until very high gel conc.

Same trends for other lines.

So, increase in initial matrix stiffness does accelerate response to growth-induced stress, as shown by spheroid growth kinetics and clonal efficiency.

(reversibly inhibit?)

Growth responded to local threshold level of stress (by gradual displacement)?

• Look at solid stress on the surface of the spheroids
• Levels of accumulated stress around plateau-phase spheroids are comparable, though the final sizes varied with initial gel conc.
• Variability due to
  • Assumption that there is no matrix relaxation
  • Error in data on mechanical prop. of gels and spheroid diameter measurements
  • (Small) variation in gel conc.
  • Size heterogeneity of spheroids
• Stress profile dropped back to pregrowth value at distance of the spheroid radius - assumption that spheroids not interact mechanically to each other

So, spheroids grow (in mechanically resist matrix) until a growth-inhibitory threshold level of stress is reached

Growth resumes when stress is allevated

• Digestion of agarose to release the spheroids > place as free-suspensions > resume growth
• Growth rates and sizes statistically not different from those in controls (free-suspensions)

So, effect of stress is reversible

Nonuniform stress affects shape of tumor aggregates

In this experiment, the authors grew cells in cylindrical glass capillary tubes, with the diameter of the tube about the size of final spheroids in suspension (control). The tube causes the radial stress to increase more rapidly that the axial stress during growth.

• Controls: nearly perfect spheroids (longitudinal axis vs. radial axis)
• 0.7% gels in tube: ellipsoid shapes (longer axis on the longitudinal direction)
• 0.7% gels outside tube: continue to grow, near-spheroid shapes afterwards

So, tumor cell aggregates preferentially grow in the direction of least stress and the growth pattern is modulated by stress field

Solid stress and proliferation, apoptosis and cell density

Hypothesis: stress-dependent control of macroscopic growth is sensed at microscopic level by stress-induced changes in cellular growth parameters

Growth kinetics of spheroids >> Gompertz law (empirical relationship for volume growth)

<math>\ln \left ( \ln \frac{V} {V_0} \right )= -\alpha t + \frac{V_{max}} {V_0}</math>

• V = spheroid size
• <math>V_0</math> = initial size
• <math>V_{max}</math> = final size
• <math>\alpha</math> = proliferation rate in simple, two-compartment model (in this case, prolif. vs. nonprolif.)

When we plot <math>\ln \left ( \ln \frac{V} {V_0} \right )</math> vs. time, and approximate by linear fit, we can get the proliferation rate

• For LS174T cells, the rates are identical in 0.3-1% gels and controls and do not change after release of spheroids from the gels

So, the solid stress does not affect proliferation rate, although it inhibits the macroscopic growth

Looking at the proliferation and apoptosis at plateau phase, the percentage of both proliferating and apoptotic cells decreased, while cell density increased with gel conc. (stiffer gel > proliferate less/die less and more dense cell population)(?)

Apoptosis was detected primarily in the core, while proliferation was on the surface of the tumor mass. Free-suspension spheroids often have necrosis in the core, forming voids - not found in gel-cultured spheroids.

The percentage of apoptotic cells increased with size in free-suspension groups, but not those in gels.

So, Solid stress does not affect proliferation rate, but decrease apoptotic rate

Discussion

• Gel is a 3D model mimicking the microenvironment of tumor in vivo
• Isotropic solid stress inhibits spheroid growth, regardless of tumor cell lines, tissues of origin, and differentiation states
• The level of solid stress imposed by a semisolid matrix that controls the tumor growth - not the limitations in nutrient supply or waste removal
• Free-suspension spheroids have less favorable nutrient environment (diffuse over longer distances - one-sided medium supply and intercalated cell-free agarose gel layer) - but grow faster than gel-embedded spheroids (medium supplied from upper and lower compartments)
• Diffusion coeff do not change significantly in high gel conc.
• fluids in both compartments equilibrate rapidly > high hydraulic conductivity.
• Growth inhibition may be from accumulation of agarose-bound toxic substance (at high gel conc - high gel strength) -> cannot explain sharp decrease in growth between low and intermediate stiffness gels and high stiffness gels.
• Cells growing in collagen gel also show similar inhibition
  • Tumor cells actively interact with collagen - not accurately model the environment
• The authors showed that stress reduced spheriod growth
  • Plateau-phase spheroids have a decrease in apoptotic rate, negligible necrosis, and similar proliferation rate compared to controls
  • The decrease in apoptotic rate and reduction in macroscopic spheroid growth can be explained by increase in cellular packing density
• This leads to hypothesis that stress-induced inhibition of macroscopic tumor growth may give a survival advantage to the tumor
  • Cell-cell interactions - presumably strengthened at higher cell density - inhibit apoptosis
  • Increased cell density (compaction) (induced by external solid stress or found in inner cell layers of the spheroid) > multicelllular-dependent mechanism of increased radiation resistance and drug resistance (proven by Rak and Kerbel)
• Tumor dormancy in vivo may be from the balance between proliferation and apoptosis of cells
  • Aggressive tumor growth may not be due to proliferation, but rather decreased apoptotic rate
• Solid stress may decrease apoptotic rate' by inducing the increase in the rate of accumulation of viable cells in the quiescent state
  • The proliferation rate is found to not change significantly after the release of spheroids from the gel - cells are in quiescent states
  • The macroscopic growth resume, presumably by the recruitment of those cells back to the proliferation state - corroborated by other findings

- with lag period increased

• (Vaage) dormant tumors were completely surrounded by a highly fibrous stroma and disruption of the capsule resumed the tumor growth - reflects the change in solid stress
• Tumors in vivo develop higher interstitial fluid pressure (IFP) than normal tissues
  • Solid stress: mechanical work by the cells required to deform the matrix
  • IFP (Fluid stress): rapid transmembrane stress equilibrium > the effect needs further investigation

Can proliferating tumor cells collapse blood or lymphatic vessels locally by growth-generated solid stress?

• 45-200mm Hg of stress accumulation may be the upper bound of stress that tumor spheroids can exert

> Sufficient to induce local collapse (average tumor microvascular pressures = 6-17mm Hg)

> Chronic vascular collapse and spatial heterogeneity in blood flow

IFP is known to equilibrate with tumor microvascular pressure

> Too small pressure differences to induce vascular collapse

Solid stress (100mm Hg) may severely collapse lymphatic vessels

Further questions

• Nature of stress transduction mechanism
• Stress within tumor vs. external stress
• Which elements of ECM are important and relevant to solid stress generation?
• Can stress be modified by matrix manipulation?