ADCY4 promotes brain metastasis in small cell lung cancer and is associated with energy metabolism

Abstract

Brain metastasis (BMs) in small cell lung cancer (SCLC) has a very poor prognosis. This study combined WGCNA with the mfuzz algorithm to identify potential biomarkers in the peripheral blood of patients with BMs. By comparing the significantly differentially expressed genes present in BMs samples, we identified ADCY4 as a target for further study. Expression of ADCY4 was used to cluster mfuzz expression pattern, and 28 hub genes for functional enrichment. PPI network analysis were obtained by comparing with differentially expressed genes in BMs. GABRE, NFE4 and LMOD2 are highly expressed in patients with BMs and have a good diagnostic effect. Immunoinfiltration analysis showed that SCLC patients with BMs may be associated with memory B cells, Tregs, NK cell activation, macrophage M0 and dendritic cell activation. prophytic was used to investigate the ADCY4-mediated anti-tumor drug response. In conclusion, ADCY4 can be used as a promising candidate biomarker for predicting BMs, molecular and immune features in SCLC. PCR showed that ADCY4 expression was increased in NCI-H209 and NCI-H526 SCLC cell lines. In vitro experiments confirmed that the expression of ADCY4 was significantly decreased after anti-PD1 antibody treatment, while the expression of energy metabolism factors were significantly different. This study reveals a potential mechanism by which ADCY4 mediates poor prognosis through energy metabolism -related pathways in SCLC.

Keywords: ADCY4; Anti-PD1; BMs; Mfuzz; SCLC; WGCNA.

Fig. 1

(A) Comparison of the original data of mRNA microarray detection of peripheral blood mononuclear cells in 6 patients with BMs and 6 patients without BMs in the GSE161968 dataset; (B) Comparative box plots of standardized and normalized gene expression profiles of samples; (C) Volcano maps of differentially expressed genes between patients with BMs and patients without BMs in the GSE161968 dataset, with marked differences in energy metabolist-related genes; (D) Significantly differentially expressed genes related to energy metabolism between patients with BMs and patients without BMs in the GSE161968 dataset.

Fig. 2

WGCNA screens Hub genes.(A) scale-free index analysis of each soft threshold power (β); (B) Average connectivity analysis for each soft threshold power; (C) The clustering tree of the sample, the tree of all differentially expressed genes is based on the measurement clustering of dissimilarity (1-TOM), and the ribbon shows the results of automatic single-block analysis; (D) Clustering correlation heat maps of samples; (E) Clustering of samples; (F) Heat maps of correlation of module characteristic genes in each sample.

Fig. 3

Screening of the most relevant modules of BMs. (A–B) Scatter plot of correlation between the weighted scores of genes in the red module and their significance respectively in samples without BMs and BMs; (C–D) Scatter plot of correlation between the weighted scores of genes in the violet module and their significance respectively in samples without BMs and BMs; (E–F) Scatter plot of correlation between the weighted scores of genes in the grey module and their significance respectively in samples without BMs and BMs. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Cluster analysis of mfuzz expression patterns was performed according to the expression level of ADCY4.

Fig. 5

Correlation analysis between ssGSEA expression pattern clustering and ADCY4 expression. (A) The significance of expression pattern clustering calculated by Ssgsea and the correlation between cluster scores and ADCY4 expression (sorted by number of genes); (B) The score comparison of expression pattern Cluster with Cluster 22 showed significant difference in the presence or absence of BMs, and the correlation was strongest; (C) Scatter plots of correlation between different clustering patterns and ADCY4 expression levels.

Fig. 6

Enrichment analysis. (A) Venn diagram of intersection between genes of ADCY4 cluster 22 and differential genes related to BMs; (B) Hub components of Cluster 22 intermolecular interaction networks among ensemble genes; intermolecular interaction networks. (C–D) Bar and bubble graphs for GO enrichment analysis of 28 ADCY4-related genes; (E–F) DO enrichment analysis of 28 ADCY4-related genes, bar and bubble graphs.

Fig. 7

Pathway enrichment map of KEGG enrichment analysis of 28 AdCy4-related genes. (A) Nicotine addiction pathway enrichment map. (B) Prolactin signaling pathway; (C) Enrichment map of Neuroactive ligand-receptor interaction pathway; (D) Concentration map of Bile secretion pathway. (E) Concentration map of GABAergic synapse pathway.

Fig. 8

Difference analysis and efficacy verification of 28 ADCY4-related genes on BMs; (A–B) GABRE was highly expressed in patients with BMs and had good diagnostic efficacy (AUC = 0.917). (C–D) NFE4 was highly expressed in patients with BMs and had good diagnostic efficacy (AUC = 0.889). (E-F)LMOD2 was highly expressed in patients with BMs and had good diagnostic efficacy (AUC = 0.889).

Fig. 9

Relationship between ADCY4 and immunity. (A) Cumulative bar chart of the degree of infiltration of immune cells in each sample in the GSE161968 dataset; (B) Differences in infiltration of immune cells in the BMs group in the GSE161968 dataset.

Fig. 10

Drug sensitivity analysis of ADCY4. Drug sensitivity difference between high and low ADCY4 expression groups. OSU-03012(A)、Tipifarnib(B)、Ruxolitinib(C)、Tivozanib(D)、Veliparib(E)、Imatinib(F)、BMs-509744(G)、Pyrimethamine(H)、GW 441756(I)、KIN001-135(J)、Roscovitine(K)、Salubrinal(L)、AMG-706(M)、Mitomycin C(N)、QS11(O)、PF-562271(P)、PHA-665752(Q)、IOX2(R)、UNC1215(S)、rTRAIL(T).

Fig. 11

Anti-PD-1 therapy inhibits progression of small cell lung cancer. (A) Expression of ADCY4 in BEAS-2B, NCI–H209 and NCI–H526 cell lines before and after anti-PD1 treatment. (B–C) Changes in cell viability in NCI–H209 and NCI–H526 cell lines before and after anti-PD1 treatment. (D–E) Changes in energy metabolism makers in NCI–H209 and NCI–H526 cell lines before and after anti-PD1 treatment. (F–G) Changes in apoptotic makers in NCI–H209 and NCI–H526 cell lines before and after anti-PD1 treatment. N = 3/group. *≤0.05, **≤0.01，***≤0.001，****≤0.0001. The results are presented as mean ± SEM.