Figures and data in Mechanistic insight for T-cell exclusion by cancer-associated fibroblasts in human lung cancer

Figures
Tables
Additional files

7 figures, 4 tables and 1 additional file

Figures

Figure 1

Download asset Open asset

Summary of the article: Starting from experimental data on the structure and density of the different species on lung cancer, we first build a dynamical system, and then a spatiotemporal model.

We incorporate four species with specific interactions, that may lead to different outcomes.

Figure 2

Download asset Open asset

Major T-cell infiltration patterns observed in solid tumors.

(A) Lack of tumor antigen, inadequate priming, defects in antigen presentation and/or lack of presentation, and/or lack of T-cell-attracting chemokines result in the absence of T-cells in the tumor. (B) Presence of T-cells in invasive margins but absent in the tumor bed. Immune evasion may be due to stromal barriers, lack of chemokines, aberrant vasculature, or hypoxia. (C) High degree of T-cell infiltration forms a hot tumor.

Figure 3

Download asset Open asset

Structural organization in human Non Small Cell Lung Cancer (NSCLC).

Staining was performed by multiplex immuno-histochemistry (IHC). FFPE NSCLC sections were stained for keratin as a marker of cancer cells (gray), CD3 (yellow), and fibroblast markers αSMA (green), FAP (red). First row: CD3 excluded patient. (a) Scheme of the section showing CD3⁺ cell exclusion from tumor nests. The green arrow highlights border regions with contractile fibroblast barrier αSMA⁺ FAP⁺ and low CD3⁺ cells. (b) CD3⁺ cells are localized in the center of the stroma. (c) Dense αSMA staining at the tumor border are associated with a decrease of CD3⁺ cell abundance. (d) Recap of all the markers. Second row: CD3 infiltrated patient. (e) Scheme of the section showing CD3⁺ infiltration in the tumor islets. The green arrow shows αSMA⁺ staining on vessels. (f) CD3⁺ are localized in the tumor nest. (g) FAP⁺ staining is localized throughout the stroma. (h) Recap of all the markers.

Figure 4

Download asset Open asset

Evolution of the tumor microenvironment (TME) composition over time.

Different profiles are obtained from Equations 1–4 according to the set of parameters reported in Table 2 and consistent with the scenarios shown in Figure 2. In all the plots, $δ = 1$ and $α_{N A} = 10^{- 4}$ . (**a, b, c, d**) Immune-desert tumor. Cancer cells proliferate when the immune response is inefficient or when T-cells are not attracted to the tumor. e: Immune-excluded tumor. The cancer-associated fibroblast (CAF) barrier inhibits T-cell infiltration, and promotes tumor growth. (**f, g, h**) Immune-inflamed tumor. T-cell infiltration limits cancer cell growth. (i) Density plot of the equilibrium cancer cell fraction at fixed $δ$ and, $α_{N A}$ in function of the two parameters that control the growth: $α_{T C} δ_{C T} / (α_{N A, C} K δ_{T F})$ reflects the ability of T-cells to kill the tumor, and $α_{N A, C} + α_{T C}$ reflects the competition between species for space and resources.

Figure 5 with 1 supplement

Download asset Open asset

Small solid tumor growth.

We refer here to the different tumor phenotypes described in Figure 2. (A) Mass fraction of cancer cells at time $t = 0$ and profile of different mass fractions on a section of the tumor. (**B-C-D**) Mass fraction of cancer cells at time $t = 35 T_{0}$ and profile of mass fraction in immunodeficient tumor (B), when immune infiltration is inhibited by cancer-associated fibroblasts (CAFs) (C) and in immune-inflamed tumor (D) In D, the scale of the colorbar is 10⁻³ the values of (A), (B), (C), since this panel represents the case of efficient T-cells. (**E-F-G**) Development of different zones. (E) Surface fraction profile of the tumor nest for different scenarios, calculated as. $S / S_{t o t a l} = \int d V δ (C > 0.1) / S_{t o t a l}$ (F) Fibrotic surface fraction profile for different scenarios, calculated as. $S / S_{t o t a l} = \int d V δ (C A F > 0.1) / S_{t o t a l}$ (G) T-cell rich area fraction profile for different scenarios, calculated as $S = \int d V δ (T > 0.1) / S_{t o t a l}$ . (H) Cancer cell average mass fractions. (I) CAF average mass fractions. (J) T-cell average mass fractions. (K) Non-activated fibroblast (NAF) average mass fractions.

Figure 5—figure supplement 1

Download asset Open asset

Free growth of a tumor nest.

(**A-G**) Cancer population density plots for different scenarios: Cancer cells (C-c) + non-activated fibroblasts (NAFs) + Inefficient T-cells, Cancer cells + NAFs,Cancer cells alone. (A) Mass fraction of cancer cells at time $t = 0$ and profile of different mass fractions on a section of the tumor. (B) Mass fraction of cancer cells at time $t = 35 T_{0}$ and mass fraction profile for a case of a stroma composed by C-c, NAFs, and inefficient T-cells. (C) Mass fraction of cancer cells at time $t = 35 T_{0}$ and profile of the different mass fractions on a cut of the tumor for a case of a stroma composed by C-c and NAFs. (D) Mass fraction of cancer cells at time $t = 35 T_{0}$ and mass fraction profile for a case with no stroma in the tumor microenvironment. (E) Surface fraction of the tumor nest in the different scenarios. (F) Average mass fraction of the cancer cells in the different scenarios. (G) Average mass fraction of the NAFs in the different scenarios. (H) Time evolution of the mass fraction profile of the cut of the simulation window in the case of a tumor in the presence of efficient T-cells and NAFs only, without T-cells.

Figure 6

Download asset Open asset

Growth process of two small solid tumors.

Density plots showing cancer cell population. We start with two tumor nests placed at the same distance from the center of the domain (30 μm), sharing the same mass fraction of cancer cells and we present different scenarios. (A) Nests coalesce in immune-desert tumor, i.e., in the presence of inefficient T-cells. (B) Nests interacting in immune-excluded tumor, with chemotactic T-cells and cancer-associated fibroblasts (CAFs). (C) Nests coalesce in immune-inflamed tumor.

Figure 7 with 1 supplement

Download asset Open asset

Anisotropy.

Following biological observation reported in Figure 3, we here show different cases for describing anisotropy. (A) Example of the tumor microenvironment in the presence of two blood vessels (in white). The tumor nest is in cyan ( $C > F_{A} + F_{N A} + T$ ), the cancerous stroma in red ( $F_{A} > F_{N A} + T$ ) and the healthy stroma in green ( $F_{N A} + T > F_{A}$ ). The tumor shape is dictated by the localization of the blood vessels from which T-cells and non-activated fibroblasts (NAFs) are issued. (**B-C**) Tumor behavior in the presence of normal (B) or orthonormal (C) fibers. The tumor fraction is indicated in colors from black to gray which we limit for $C > 0.25$ to better visualize of the contour of the nest. The flux $C v_{C}$ is indicated with blue arrows whose thickness is proportional to the magnitude of the flux. The nematic order is shown with red lines whose thickness is proportional to the determinant of the matrix $Q$ (see SI, Nematic order) and the fibroblast fraction is indicated with a pink color gradient.

Figure 7—figure supplement 1

Download asset Open asset

Different tumor phenotypes in non-spherical domains.

Deformed shapes result in anisotropic configurations by forcing non-activated fibroblasts (NAFs) and T-cells to enter the domain from only one side of the boundary. Different colors indicate separated regions where the nest (in light blue) is confined by the boundary of cancer-associated fibroblasts (CAFs) (in red) within a healthy stroma (in green). (A) Immune-inflamed case leads to disappearance cancer cells at $t = 5 T_{0}$ . (B) Elliptical area leads to slower dynamics.

Tables

Table 1

Composition of the tumor microenvironment.

This table presents the literature about the fraction of different species in the lung TME. The values found in the present studies are written in the two last rows (raw data is provided as Supplementary Material). Different methods have been used, in different subtypes of lung cancer. We introduce the following abbreviations. (C-c): Cancer cells. (MΦ): Macrophages. (T-c): T-cells. (Fb): Fibroblasts. (scRNA-seq): single-cell RNA seq,(NSCLC): Non Small Cell Lung Cancer, (LUAD): Adenocarcinoma, (SSN): Sub-Solid Nodule. (S): stroma region. (T): tumor nest region. (ST): Stroma + tumor nest region. (TSR): tumor stroma ratio. Data for Ref (Ireland et al., 2020; Kim et al., 2020; Laughney et al., 2020; Maynard et al., 2020; Qian et al., 2020) were extracted from Curated Cancer Cell Atlas, 2023.

Method	Sample	C-c (%)	T-c (%)	Fb (%)	MΦ(%)	TSR
Staining Sieren et al., 2010	LUAD	>25		>20		0.3
Staining Sieren et al., 2011	LUAD	67		>6.5		2.6
scRNA-seq Lambrechts et al., 2018	NSCLC		55 (S)	4 (S)	15 (S)
scRNA-seq Ireland et al., 2020	SCLC	76
scRNA-seq Kim et al., 2020	LUAD	33	19.5	∼6	20.4
scRNA-seq Laughney et al., 2020	LUAD	<12	42.6	∼1	12.2
scRNA-seq Maynard et al., 2020	NSCLC	<22.7	16	8	23
scRNA-seq Qian et al., 2020	mixed	19.9	28.1	∼2	27.5
scRNA-seq Xing et al., 2021	LUAD	16.4	30	∼2	18.4
scRNA-seq Altorki et al., 2022	LUAD	12	7.5
Staining Mathieson et al., 2022	NSCLC			75 (S), 0 (T)
Staining (present article)	LUAD	47±13		31±17 (S), 0 (T)		1.07±0.68
Staining (present article)	LUSC	54±10		35±12 (S), 0(T)		1.30±0.53

Table 1—source data 1 Raw data for the two last lines of the table.: https://cdn.elifesciences.org/articles/101885/elife-101885-table1-data1-v1.xlsx
Download elife-101885-table1-data1-v1.xlsx

Table 2

Scaling variation and estimation of the coefficients according to different scenarios.

The scenarios are described in Section Model Parameters and Fixed Point Analysis and shown in Figure 4. The coefficients above are those introduced in Equations 1–4. The left column summarizes the different roles that T-cells can play in a cell mixture and the values of the coefficients of the mixture are listed in the following horizontal lines of the table. The scaling of $δ$ is always: 1 , and $α_{N A} \sim δ ϵ^{4}$ .

	$δ_{C T}$	$α_{T C}$	$δ_{T F}$	$α_{N A, C}$	$K_{A}$
Control of:	Killing C-c by T-c	Attraction of T-c by C-c	Inhibition of T-c by F-c	Attraction of F-c to C-c	Activation of F-c
Efficient T-cells (T-c) but no attraction by (C-c)	$δ ϵ^{- 1}$ $\sim 10$	$ϵ^{2}$ $\sim 0.01$	No role	No role	No role
Efficient T-cells but inhibited by (F-c)	$δ ϵ^{- 1}$ $\sim 10$	$ϵ$ $\sim 0.1$	$δ ϵ^{- 1}$ $\sim 10$	1∼1	$δ$ $\sim 1$
Inefficient T-cellsno need of fibroblasts	$δ$ $\leq 1$	$ϵ$ $\sim 0.1$	No role	No role	No role
Efficient T-cells not inhibited by fibroblasts	$δ ϵ^{- 1}$ $\sim 10$	$ϵ$ $\sim 0.1$	$δ$ $\sim 1$	1∼1	$δ$ $\sim 1$
Efficient T-cells and fibroblasts not attracted	$δ ϵ^{- 1}$ $\sim 10$	$ϵ$ $\sim 0.1$	$δ ϵ^{- 1}$ $\sim 10$	$ϵ$ $\sim 0.1$	$δ$ $\sim 1$
Efficient T-cells fibroblasts not activated	$δ ϵ^{- 1}$ $\sim 10$	$ϵ$ $\sim 0.1$	$δ ϵ^{- 1}$ $\sim 10$	$ϵ$ $\sim 1$	$δ ϵ$ $\sim 0.1$

Table 3

Values of the parameters in the spatial model.

Parameter	Value	Parameter	Value
$D_{0}$	$10^{6} k g . μ m^{- 1} . T_{0}^{- 2}$	$δ$	$1.18 T_{0}^{- 1}$
$D_{C}$	$10^{5} k g . μ m^{- 1} . T_{0}^{- 2}$	$τ_{T}, τ_{N A}$	$2.5 \times 10^{- 3} μ m . T_{0}^{- 1}$
$D_{T}, D_{N A F}, D_{C A F}$	$2.5 \times 10^{5} k g . μ m^{- 1} . T_{0}^{- 2}$	$δ_{C T}$	$5 T_{0}^{- 1}$
$λ_{C C}$	$2.5 \times 10^{5} k g . μ m^{- 1} . T_{0}^{- 2}$	$ξ_{0}$	$3 \times 10^{2} k g . μ m^{- 3} . T_{0}^{- 1}$
$λ_{C T}$	$7.5 \times 10^{5} k g . μ m^{- 1} . T_{0}^{- 2}$	$ξ_{1}$	$3 \times 10^{5} k g . μ m^{- 3} . T_{0}^{- 1}$
$λ_{C F}$	$0 k g . μ m^{- 1} . T_{0}^{- 2}$	$K$	$10 T_{0}^{- 1} . μ m^{2}$
$κ$	$3.6 \times 10^{7} k g . μ m . T_{0}^{- 2}$	$λ_{c C}$	$30 μ m$

Appendix 1—table 1

Parameter values varying depending on the scenario in Figure 5—figure supplement 1.

Parameter	$τ_{T} (μ m . d^{- 1})$	$τ_{F} (μ m . d^{- 1})$
C-c+NAFs+inefficient T-cells	2.5×10⁻³	2.5×10⁻³
C-c+NAFs	2.5×10⁻⁶	2.5×10⁻³
C-c	2.5×10⁻⁶	2.5×10⁻³

Additional files

MDAR checklist: https://cdn.elifesciences.org/articles/101885/elife-101885-mdarchecklist1-v1.docx
Download elife-101885-mdarchecklist1-v1.docx

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Mendeley

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

Joseph Ackermann
Chiara Bernard
Philemon Sirven
Helene Salmon
Massimiliano Fraldi
Martine D Ben Amar

(2025)

Mechanistic insight for T-cell exclusion by cancer-associated fibroblasts in human lung cancer

eLife 13:RP101885.

https://doi.org/10.7554/eLife.101885.3

Share this article

Cite this article

Summary of the article: Starting from experimental data on the structure and density of the different species on lung cancer, we first build a dynamical system, and then a spatiotemporal model.

Major T-cell infiltration patterns observed in solid tumors.

Structural organization in human Non Small Cell Lung Cancer (NSCLC).

Evolution of the tumor microenvironment (TME) composition over time.

Small solid tumor growth.

Free growth of a tumor nest.

Growth process of two small solid tumors.

Anisotropy.

Different tumor phenotypes in non-spherical domains.

Composition of the tumor microenvironment.

Table 1—source data 1

Scaling variation and estimation of the coefficients according to different scenarios.

Values of the parameters in the spatial model.

Parameter values varying depending on the scenario in Figure 5—figure supplement 1.

MDAR checklist

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)