Global evolution of the tumor microenvironment associated with progression from preinvasive invasive to invasive human lung adenocarcinoma

Abstract

To investigate changes in the tumor microenvironment (TME) during lung cancer progression, we interrogate tumors from two chest computed tomography (CT)-defined groups. Pure non-solid (pNS) CT density nodules contain preinvasive/minimally invasive cancers, and solid density nodules contain invasive cancers. Profiling data reveal a dynamic interaction between the tumor and its TME throughout progression. Alterations in genes regulating the extracellular matrix and genes regulating fibroblasts are central at the preinvasive state. T cell-mediated immune suppression is initiated in preinvasive nodules and sustained with rising intensity through progression to invasive tumors. Reduced T cell infiltration of the cancer cell nests is more frequently associated with preinvasive cancers, possibly until tumor evolution leads to a durable, viable invasive phenotype accompanied by more varied and robust immune suppression. Upregulation of immune checkpoints occurs only in the invasive nodules. Throughout progression, an effector immune response is present but is effectively thwarted by the immune-suppressive elements.

Keywords: CP: Cancer; CT scan density; GGO; Tregs; extracellular matrix; lung ground glass lesions; preinvasive lung adenocarcinoma; tumor microenvironment.

Figure 1.. Transcriptome differences between normal lung and pure non-solid nodules

(A) Volcano plot of differentially expressed genes upregulated (red) and downregulated (green) in pure non-solid (pNS) nodules compared with normal lung with a ≥2-fold difference in expression and an FDR ≤0.05 (Benjamini-Hochberg correction). Comparison of 10 normal lung with 23 pNS nodules. (B) Some biological pathways enriched among the significantly upregulated genes in pNS nodules compared with normal lung (FDR <0.05). The David Gene Ontology pathway identifiers (GO numbers) are noted. (C) Significantly upregulated genes in pNS nodules relative to normal lung that are annotated to extracellular matrix (ECM) biological pathways. The FPKM of the individual genes are scaled to the average value in normal lung. Average values of the scaled fragments per kilobase of exon per million mapped fragments (FPKM) ±SEM for 10 normal lung samples and 23 pNS nodules. (D) Chemokines and cognate receptors significantly upregulated in the pNS nodules compared with normal lung. The target cells (those cells that highly express the receptors) are noted. (E) Some biological pathways enriched among the significantly downregulated genes in pNS nodules compared with normal lung (FDR <0.05). The David Gene Ontology pathway identifiers (GO numbers) are noted. (F) Chemokine genes significantly downregulated in pNS nodules compared with normal lung. Target cells for these chemokines are noted.

Figure 2.. Transcriptome differences between pNS and solid nodules

(A) Volcano plot of differentially expressed genes upregulated (red) and downregulated (green) in solid tumors compared with pNS tumors with a ≥2-fold difference in expression and an FDR <0.05 (Benjamini-Hochberg correction). Comparison of 23 pNS nodules with 21 solid nodules. (B) Some biological pathways enriched among the significantly upregulated genes in solid tumors compared with pNS tumors (FDR <0.05). Gene Ontology pathway identifiers (GO numbers) are noted. (C) ECM Gene Ontology pathway genes upregulated in solid tumors compared with pNS tumors. Average values of the scaled FPKM ±SEM for 23 pNS nodules and 21 solid nodules. (D) Type I interferon response genes upregulated in solid tumors compared with pNS tumors. Average values of the scaled FPKM ±SEM for 23 pNS nodules and 21 solid nodules.

Figure 3.. Genes progressively upregulated from normal lung to solid nodules

(A) Some biological pathways enriched among the 90 genes that are upregulated by ≥2-fold in pNS compared with normal lung and further upregulated by ≥2-fold in solid tumors relative to pNS tumors (FDR <0.05). (B) ECM gene set score for each sample group generated as the sum of the Z scores of the 36 ECM pathway genes differential expressed between normal lung and pNS nodules (Figure 1C). Sidak’s p value for multiple comparison are shown. 10 normal lung, 23 pNS nodules, and 21 solid nodules. (C) Gene set score for chemokine/chemokine receptor genes upregulated in pNS compared with normal lung generated as the sum of the Z scores of the individual genes (Figure 1D). Sidak’s p value for multiple comparison are shown. 10 normal lung, 23 pNS nodules, and 21 solid nodules. (D) Gene set score for chemokine/chemokine receptor genes downregulated in pNS compared with normal lung generated as the sum of the Z scores of the individual genes (Figure 1F). Sidak’s p value for multiple comparison are shown. 10 normal lung, 23 pNS nodules, and 21 solid nodules.

Figure 4.. Immunofluorescence immune profiling

(A) Immune-cell compositions of normal, pNS tumors, and solid tumors determined by xCell deconvolution. (B) Density of CD3⁺ cells determined as the percentage of total cells per ROI. (C) Coefficient of variations of CD3⁺ cell densities of the 6 ROIs per tumor of the pNS and solid groups. (D) Densities of CD4⁺ cells determined as percentage of CD3⁺ cells. (E) Treg densities (CD3⁺CD4⁺FoxP3⁺) determined as the percentage of CD3⁺CD4⁺ cells. (F) CTLA-4⁺ Treg densities (CD3⁺CD4⁺FoxP3⁺CTLA-4⁺) as the percentage of CD3⁺CD4⁺FoxP3⁺ cells. (G) Cytotoxic T cell densities (CD3⁺CD8⁺) determined as the percentage of CD3⁺ cells. (H) Ratios per tumor of Tregs (CD3⁺CD4⁺FoxP3⁺) to cytotoxic T cells (CD3⁺CD8⁺) in normal lung and pNS nodules. (I) Active cytotoxic T cells densities (CD3⁺CD8⁺GZB⁺Ki67⁺) determined as the percentage of Ki67 + CD3⁺CD8⁺GZB⁺ cells. GZB, granzyme B. (J) Correlation of CTLA-4⁺ Treg (CD3⁺CD4⁺FoxP3⁺CTLA-4⁺) and active cytotoxic T cell (CD3⁺CD8⁺GZB⁺) densities as percentages of total cells per ROI. The p value is for the difference in the slope of the correlation line from 0. (K) T cells (CD3⁺) within the cancer cell nests as a percentage of total T cells in the 3 subcategories of solid tumors segregated by panCK⁺ PD-L1⁺ expression. (L) PD-L1-expressing panCK cells in pNS- and solid-nodule groups. The percent of nodules per group with <1% or ≥1% panCK⁺PD-L1⁺ cells. Fisher’s exact test. In all panels, the symbols for normal tissue are data from a single ROI per sample, whereas in the pNS and solid groups, each symbol is the mean of 3–6 ROIs per tumor. Total cells were determined by 4′,6-diamidino-2-phenylindole (DAPI)-stained nuclei per ROI. The group means ± SEM are shown for 10 normal lung, 25 pNS nodules, and 27 solid nodules. p values. ANOVA followed by Kruskal-Wallis test of p values for multiple comparisons. All significant differences are noted.

Figure 5.. TME morphology: SMA pattern

(A) Fibroblasts cell (CD3⁻panCK⁻SMA⁺) density in pNS- and solid-nodules groups. (B) Representative pseudo-colored immunofluorescence and virtual H&E images of pNS and solid nodules with organized or disorganized αSMA barriers. panCK in red, αSMA in green, and DAPI in blue. Images acquired at 200× magnification. Scale bar, 200 μm. (C) Predominant tumor αSMA morphology by tumor group. p value, Fischer’s exact test. (D) Comparison of nodules, grouped by αSMA pattern, for an activated lung fibroblasts gene signature (Peyser et al., 2019) generated as sum Z score for the genes of the signature for each sample. Because the RNA-seq is from bulk tumor tissue, only those nodules that had either continuous or discontinuous morphologies, independent of CT density group, for each of the 3 ROIs were used for these analyses: n = 14 continuous and n = 14 discontinuous. Mann-Whitney nonparametric t test. (E) Comparison of nodules, grouped by αSMA pattern, for a lung CAF gene signature (Navab et al., 2011) generated as sum Z score for the genes of the signature for each sample. Because the RNA-seq is from bulk tumor tissue, only those nodules that had either continuous or discontinuous morphologies, independent of CT density group, for each of the 3 ROIs were used for these analyses: n = 14 continuous and n = 14 discontinuous. Mann-Whitney nonparametric t test. (F) Comparison of nodules, grouped by CT density, for an activated lung fibroblasts gene signature (Peyser et al., 2019) generated as sum Z score for the genes of the signature for each sample. Mann-Whitney nonparametric t test. Means ± SEM for 23 pNS and 21 solid nodules. (G) Comparison of nodules, grouped by CT density, for a lung CAF gene signature (Navab et al., 2011) generated as sum Z score for the genes of the signature for each sample. Mann-Whitney nonparametric t test. Means ± SEM for 23 pNS and 21 solid nodules. (H) Comparison of nodules, by αSMA pattern, for CD3⁺ cell infiltration into cancer cell nests as the percentage of T cells in the ROI (cancer cell nests plus stroma). Mann-Whitney nonparametric t test. Means ± SEM for 16 continuous and 19 discontinuous nodules with the same fibroblast morphology pattern in all three ROIs.