Augmented Lipocalin-2 Is Associated with Chronic Obstructive Pulmonary Disease and Counteracts Lung Adenocarcinoma Development

Abstract

Rationale: Early pathogenesis of lung adenocarcinoma (LUAD) remains largely unknown. We found that, relative to wild-type littermates, the innate immunomodulator Lcn2 (lipocalin-2) was increased in normal airways from mice with knockout of the airway lineage gene Gprc5a (Gprc5a^-/-) and that are prone to developing inflammation and LUAD. Yet, the role of LCN2 in lung inflammation and LUAD is poorly understood.Objectives: Delineate the role of Lcn2 induction in LUAD pathogenesis.Methods: Normal airway brushings, uninvolved lung tissues, and tumors from Gprc5a^-/- mice before and after tobacco carcinogen exposure were analyzed by RNA sequencing. LCN2 mRNA was analyzed in public and in-house data sets of LUAD, lung squamous cancer (LUSC), chronic obstructive pulmonary disease (COPD), and LUAD/LUSC with COPD. LCN2 protein was immunohistochemically analyzed in a tissue microarray of 510 tumors. Temporal lung tumor development, gene expression programs, and host immune responses were compared between Gprc5a^-/- and Gprc5a^-/-/Lcn2^-/- littermates.Measurements and Main Results:Lcn2 was progressively elevated during LUAD development and positively correlated with proinflammatory cytokines and inflammation gene sets. LCN2 was distinctively elevated in human LUADs, but not in LUSCs, relative to normal lungs and was associated with COPD among smokers and patients with LUAD. Relative to Gprc5a^-/- mice, Gprc5a^-/-/Lcn2^-/- littermates exhibited significantly increased lung tumor development concomitant with reduced T-cell abundance (CD4⁺) and richness, attenuated antitumor immune gene programs, and increased immune cell expression of protumor inflammatory cytokines.Conclusions: Augmented LCN2 expression is a molecular feature of COPD-associated LUAD and counteracts LUAD development in vivo by maintaining antitumor immunity.

Keywords: chronic obstructive pulmonary disease; immunity; lipocalin-2; lung neoplasms.

Figure 1.

Increased Lcn2 during lung adenocarcinoma (LUAD) development in Gprc5a^−/− mice. (A) Heat map showing differentially expressed transcripts between normal airways of wild-type (WT) and Gprc5a^−/− mice (n = 6 per genotype) by RNA sequencing (RNA-seq). Columns indicate individual mouse airways and rows indicate transcripts (red, upregulated; blue, downregulated). (B) Comparison of Lcn2 mRNA expression in normal airways between WT (blue) and Gprc5a^−/− (red) mice. (C) Temporal assessment of Lcn2 mRNA levels by RNA-seq in normal airways of Gprc5a^−/− mice (n = 6 per time point) before treatment (baseline) and at various time points after NNK exposure (end of NNK and 2, 4, and 6 mo after NNK). (D) Analysis of Lcn2 mRNA expression by RNA-seq in normal lung tissues (blue, n = 9–10 per time point) with matched lesions (hyperplasias, adenomas, LUADs [red], n = 4–23 per time point) of Gprc5a^−/− mice at end of NNK and 3 and 7 months after NNK treatment. (E) Analysis of LCN2 protein in BAL fluid of Gprc5a^−/− mice at the different time points after NNK exposure (n = 7–10 per time point) by ELISA. P values denoting comparisons between two groups were obtained using unpaired Student’s t test, and values for analysis of Lcn2 expression progressively with time were obtained using ANOVA. (F) Representative histopathologic (H&E) and LCN2 immunoreactivity from NNK-exposed Gprc5a^−/− mice with hyperplasia (left), adenoma (middle), and lung adenocarcinoma (right). Scale bars: top row, 100 μm; middle and bottom rows, 50 μm. *P < 0.05, **P < 0.01, and ***P < 0.001. H&E = hematoxylin and eosin; NNK = nicotine-specific nitrosamine ketone.

Figure 2.

Elevated LCN2 expression in human lung adenocarcinoma (LUAD). (A) LCN2 mRNA levels were examined between normal lung tissues (blue, n = 59) and LUADs (red, n = 514) as well as between normal lung tissues (blue, n = 51) and lung squamous cancers (LUSCs) (red, n = 502) in the The Cancer Genoma Atlas (TCGA) data sets (20, 21). (B) LCN2 mRNA levels were also directly examined between LUSCs (blue, n = 502) and LUADs (red, n = 514) in the TCGA data sets. (C) LCN2 mRNA levels were statistically compared between LUSCs (blue, n = 57) and LUADs (red, n = 152) in the MD Anderson dataset, PROSPECT (Profiling of Resistance Patterns and Oncogenic Signaling Pathways in Evaluation of Cancers of the Thorax) (22). (D) LCN2 mRNA levels were also statistically compared between LUADs that are wild type (WT) for KRAS and KRAS-mutant LUADs (KM-LUADs) in the following cohorts: Okayama and colleagues (blue, WT, n = 206; red, KM, n = 20); Selamat and colleagues (blue, WT, n = 36; red, KM, n = 22); Beer and colleagues (blue, WT, n = 46; red, KM, n = 39) (–25). (E) Representative photomicrographs of histopathologic (H&E) and LCN2 immunohistochemical analyses of LUSC (left) and LUAD (right) specimens from the assessed tissue microarray (see Methods). (F) LCN2 protein expression was statistically compared between LUSCs (blue, n = 84) and LUADs (red, n = 187) using tumor with positive LCN2 H-scores. (G) LCN2 protein levels were also statistically compared between WT- and KM-LUADs in a tissue microarray using tumors with positive LCN2 H-scores (blue, WT, n = 36; red, KM, n = 50) and in TCGA reverse-phase protein array data set (blue, WT, n = 41; red, KM, n = 20). Solid horizontal lines represent median LCN2 log₂ expression values, H-scores, or protein z-scores. Differences in LCN2 levels between two groups were statistically assessed using the unpaired Student’s t test. Scale bar, 100 μm. *P < 0.05, **P < 0.01, and ***P < 0.001. H&E = hematoxylin and eosin; ns = not significant.

Figure 3.

Upregulation of LCN2 in human chronic obstructive pulmonary disease (COPD) and COPD-associated lung adenocarcinoma (LUAD). (A) LCN2 mRNA was statistically analyzed using unpaired Student’s t test in normal-appearing airway brushings (n = 238) from lung cancer–free patients in the data set by Steiling and colleagues based on COPD status (without COPD, blue, n = 151; with COPD, red, n = 87; 28). LCN2 mRNA levels in airway brushings were also statistically assessed based on the Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria for COPD staging (middle panel; blue, no COPD, n = 133; red, GOLD 1–4, n = 19–61 per stage) using unpaired Student’s t test. LCN2 mRNA levels were also correlated with FEV₁ percentage using Pearson’s correlation coefficient (right panel). (B) LCN2 mRNA levels were statistically compared between patients without and with antiinflammatory treatment using all patients with COPD (GOLD 1–4; left panel; red, no treatment, n = 64; blue, treatment, n = 23) and patients with moderate to severe COPD (GOLD 2–4; right panel; red, no treatment, n = 58; blue, treatment, n = 22) (28) using the Student’s t test. (C) LCN2 expression was statistically compared using unpaired Student’s t test between normal lung tissues adjacent to LUADs without (blue, n = 19) and with (red, n = 6) COPD as well as between normal lungs adjacent to lung squamous cancers (LUSCs) without (blue, n = 9) and with (red, n = 9) COPD in the data set by Bhattacharya and colleagues (29). (D) LCN2 expression was statistically compared using the unpaired Student’s t test between LUADs without (blue, n = 58) and with (red, n = 27) COPD as well as between LUSCs without (blue, n = 12) and with (red, n = 15) COPD in the MD Anderson cohort, PROSPECT (Profiling of Resistance Patterns and Oncogenic Signaling Pathways in Evaluation of Cancers of the Thorax) (22). (E) LCN2 H-scores were statistically compared using the unpaired Student’s t test between LUADs from the MD Anderson tissue microarray cohort without (blue, n = 98) and with (red, n = 43) COPD and between LUSCs from the same tissue microarray without (blue, n = 34) and with (red, n = 25) COPD. *P < 0.05 and ***P < 0.001. ns = not significant.

Figure 4.

Loss of Lcn2 augments lung adenocarcinoma (LUAD) development in tobacco carcinogen–exposed Gprc5a^−/− mice. (A) Schematic timeline depicting NNK i.p. injection of 8-week-old Gprc5a^−/− and Gprc5a^−/−/Lcn2^−/− mice divided into groups of 8–10 mice (per genotype and time point) and studied at baseline, end of NNK, and 3 and 7 months after NNK treatment. (B) LCN2 was quantified by ELISA in BAL fluid collected from Gprc5a^−/− and Gprc5a^−/−/Lcn2^−/− mice at the indicated time points (n = 7–9 mice per group). (C) LUAD incidence (number of tumors per mouse) was statistically compared using the unpaired Student’s t test between Gprc5a^−/− and Gprc5a^−/−/Lcn2^−/− mice (n = 8–10 mice per group) at 3 and 7 months after NNK. (D) Schematic illustrating lung orthotopic transplantation of Gprc5a^−/− LUAD MDA-F471 cells (2.5 × 10⁶ cells) in 8-week-old wild-type (WT), Gprc5a^−/−, and Gprc5a^−/−/Lcn2^−/− mice (n = 7–9 mice per genotype) and analysis at 1 month after implantation. (E) LCN2 ELISA of BAL fluid obtained from the same WT, Gprc5a^−/−, and Gprc5a^−/−/Lcn2^−/− mice (n = 7–9 mice per group) at 1 month after MDA-F471 cell implantation. (F) Analysis of tumor volume and tumor burden in orthotopically transplanted lungs of WT, Gprc5a^−/−, and Gprc5a^−/−/Lcn2^−/− recipient mice (n = 7–9 mice per group). Differences between two groups were statistically examined using the unpaired Student’s t test. *P < 0.05, **P < 0.01, and ***P < 0.001. i.p. = intraperitoneal; NNK = nicotine-specific nitrosamine ketone.

Figure 5.

Loss of Lcn2 in tobacco carcinogen–exposed Gprc5a^−/− mice promotes gene expression programs associated with reduced antitumor immunity. (A) Differentially expressed genes between Gprc5a^−/− lungs at 7 months after nicotine-specific nitrosamine ketone (NNK) compared with baseline (top left, n = 1,116 gene features) and between Gprc5a^−/−/Lcn2^−/− lungs at the same time points (top right, n = 775 gene features; n = 9–10 mice per group) were identified by RNA-sequencing analysis as described in the online supplement section. Columns denote lung samples and rows represent gene features (red, upregulated; blue, downregulated). Pathways (bottom left) and gene sets (bottom right) enriched among Gprc5a^−/− and Gprc5a^−/−/Lcn2^−/− lung tissues with time were identified using Ingenuity Pathways Analysis (IPA) and then cross-compared and plotted. Activation of the pathway or gene set between 7 months after NNK and baseline for Gprc5a^−/− (white bars) and Gprc5a^−/−/Lcn2^−/− (black bars) is indicated by the z-scores. (B) Topological gene–gene network analysis of genes associated with stimulation of leukocytes in Gprc5a^−/− (left) and Gprc5a^−/−/Lcn2^−/− (right) lungs at 7 months after NNK compared with baseline (red, upregulated) were derived using IPA. Predicted activation of leukocyte stimulation based on the gene set is indicated by the orange color (in Gprc5a^−/− mice). (C) The following T-cell signatures indicative of different phenotypes were propagated and computed for each sample as described in the online supplement: CTL = cytotoxic T lymphocyte; CYT = T cell, cytolytic activity; TIS = tumor inflammation signature; IFN-γ; and T-cell expanded and T-cell exhaustion signatures. The T-cell signatures were then statistically compared between Gprc5a^−/− and Gprc5a^−/−/Lcn2^−/− lungs at baseline and at 7 months after NNK. P values were calculated between Gprc5a^−/− lung tissues (blue) at baseline and 7 months after NNK as well as between Gprc5a^−/−/Lcn2^−/− lungs (red) at the same time points using the unpaired Student’s t test. *P < 0.05, **P < 0.01, and ***P < 0.001. Activ. = activation; apop. = apoptosis; iCOS = inducible T-cell costimulator; Inhib. = inhibition; Leuk. = leukocytes; PKC = protein kinase C; resp. = response; Stimul. = stimulation; Th1 = T-helper cell type 1.

Figure 6.

Loss of Lcn2 decreases T-cell abundance and elevates proinflammatory cytokines during lung adenocarcinoma (LUAD) development. (A) Flow cytometry analysis of total (left) and IL-17A–expressing CD4⁺ T cells (right) in lungs of Gprc5a^−/− (blue) and Gprc5a^−/−/Lcn2^−/− (red) mice at baseline, end of NNK exposure, and at 3 and 7 months after NNK (n = 8–10 mice per group). (B) Flow cytometry analysis of PD-1 in CD4⁺ (left) and CD8⁺ (middle) T cells as well as in B cells (right) from lungs of Gprc5a^−/− (blue) and Gprc5a^−/−/Lcn2^−/− (red) mice at baseline and at 7 months after NNK (n = 4–5 per group). (C) Flow cytometry analysis of pro-IL-1β in dendritic cells from Gprc5a^−/− (blue) and Gprc5a^−/−/Lcn2^−/− (red) lungs before and after NNK exposure (n = 8–10 mice per group). Differences were statistically assessed between the two genotypes within time points using unpaired Student’s t test. (D) T-cell receptor β (TCRβ) sequencing was performed as described in the online supplement. Time-dependent changes in productive TCR rearrangements were statistically evaluated for Gprc5a^−/− (blue) and Gprc5a^−/−/Lcn2^−/− (red) lungs (n = 5–6 mice per group) using ANOVA. *P < 0.05, **P < 0.01, and ***P < 0.001. (E) Correlation between CD4⁺ T-cell density and LCN2 H-scores in LUADs with chronic obstructive pulmonary disease (COPD) (n = 43) was statistically interrogated using Pearson’s correlation coefficient. (F) Schematic summarizing the study’s main findings. Lcn2 expression in both normal-appearing lung tissues and tumors was augmented during tobacco carcinogen–associated LUAD development in Gprc5a^−/− mice and correlated with markers of inflammation. Human LCN2 was upregulated in LUAD but not in lung squamous cancer when compared with normal lung, and its expression in normal tissues or tumors was associated with COPD. Genetic deletion of Lcn2 in Gprc5a^−/− mice increased tobacco carcinogen–associated lung tumor development and protumor immune phenotypes exemplified by decreased antitumor immune gene signatures, increased proinflammatory cytokine production by CD4⁺ T cells and myeloid cells, and reduced T-cell richness—overall suggestive of a protective role for LCN2 induction in LUAD pathogenesis. MFI = mean fluorescence intensity; NNK = nicotine-specific nitrosamine ketone.