Lung cancer is one of the most commonly diagnosed human cancers,1 and in the year 2018 alone, more than two million new lung cancer cases were diagnosed worldwide.2 It is regarded as one of the most common causes of cancer-related death, and it was estimated that over 150 000 lung cancer-related deaths were reported in the United States of America in the year 2015.1 Lung cancer can generally be divided into two major groups, namely, small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC).3 SCLC makes up about 15% of all diagnosed lung cancer cases while the remaining 85% are contributed by different subtypes of NSCLC like squamous cell carcinoma (SCC) (30%), adenocarcinoma (40%), and large cell carcinoma (15%).3
Single-cell RNA sequencing (scRNA-seq) is a powerful and emerging technology that allows the study of the transcriptome profile at a single-cell level.4 Compared to the conventional bulk RNA-seq, which measures the mean gene expression quantification values from a wide variety of cells that largely depends on the abundance of different specific cell types and certain transcriptional profiles, scRNA-seq allows a more direct and precise measurement of the transcriptome profile of a particular cell.5 As such, scRNA-seq is very useful in cancer research to study cancer microenvironment,6 tumour heterogeneity,7 trajectory analysis,8 alternative splicing detection,9 single-nucleotide variants (SNV),5 and copy number variants (CNV).5,10
As scRNA-seq is becoming a popular tool used in human cancer research, few review reports have been published to highlight and summarize the potential applications, advantages, and limitations of scRNA-seq in studying human cancers like breast cancer11 and colorectal cancer.12 Like other human cancers, numerous in vitro, in vivo, and clinical studies have reported the roles and applications of scRNA-seq in studying human lung cancer.13–15 However, there is a lack of a comprehensive review that could effectively summarize the broad use of scRNA-seq in studying lung cancer. Therefore, this review aimed to discuss and outline the various applications of scRNA-seq in human lung cancer research and explore the potential challenges and future directions of scRNA-seq in studying human lung cancer. The review would first briefly elaborate about the concept and principle of scRNA-seq, followed by discussing the applications of scRNA-seq in lung cancer research based on the findings from in vitro, in vivo, and clinical studies. Finally, the challenges faced in scRNA-seq related lung cancer studies would be discussed, and how would scRNA-seq help in realizing personalized lung cancer treatment would be explored.
2. SINGLE-CELL RNA SEQUENCING
Generally, the workflow of scRNA-seq can be divided into four main stages,11 namely, (1) single-cell isolation, (2) library preparation, (3) sequencing, and (4) data analyses (Fig. 1). The first and most crucial step in scRNA-seq is the effective isolation of a single cell from a pool of cells.18 There are several methods, which could be utilized to isolate single cells, and these include limiting dilution, capillary pipetting, laser-capture microdissection (LCM), microfluidics, fluorescence-activated cell sorting (FACS), magnetic-activated cell sorting (MACS).16
Limiting dilution is a manual technique, which utilizes a micropipette to isolate a single cell via a series of dilution until a single cell is obtained.19 This is a relatively common method to isolate single cell,19 but it could cause loss of cells during the process, and it is not an ideal method when the cell sample is limited.20 Capillary pipetting is a technique that makes use of a capillary pipette to isolate a single cell under a microscope, but the method could be time-consuming with low yield.19 LCM utilizes a computer-aided laser system to isolate single cell from solid samples19 or suspensions,21 and LCM could limit the disruption to the samples while minimizing contamination as it does not require the use of chemicals like in the case of FACS.21 However, LCM requires manual operation, which can be laborious and prone to human error.21 Compared to limiting dilution, capillary pipetting, and LCM, microfluidics is a method that employs controlled micro-scale structure to isolate single cell, and its advantage is that the automated platform allows automated reverse transcription immediately after single-cell isolation.11 Microfluidics efficiency can be reduced if the cells are sticky or nonspherical as these factors would limit the cell capturing capacity.22 FACS is a fluorescence-based cell sorting method that isolates single cell based on the specific marker present on the cell surface, and it is one of the most preferred ways to obtain highly purified single cell.19 The limitation of FACS is that it requires a high number of cells to begin the sorting (>10 000 cells) and it may require the use of monoclonal antibodies to recognize the specific cell surface proteins.19 In MACS, specific antibodies conjugated with magnetic beads are used to tag the cells of interest.23 An external magnetic field will be applied to isolate single cells from the pool of cells.23 Like FACS, MACS is a powerful cell isolation technique that can produce highly purified single cell (>90%), but its drawback is that it requires high running cost due to the needs to purchase separation magnet and magnetic beads with conjugated antibody.23
The amount of RNA extracted from a single cell is minimal, and the effectiveness of reverse transcription and cDNA amplification is essential in ensuring the accuracy of scRNA-seq.11 There are currently many different protocols (Table 1) that have been reported in scRNA-seq for reverse transcription process and cDNA amplification.11,17 These scRNA-seq protocols differ in terms of amplification technology, coverage of transcript, strand specificity, early multiplexing, and use of a unique molecular identifier (UMI).11,17 For reverse transcription, poly(A) tailing and template-switching are two commonly used methods.11 Poly(A) tailing involves the addition of poly(A) tail at the targeted ends of the RNA molecule by poly(A) polymerase before subjected to reverse transcription and cDNA synthesis.33 Template-switching approach utilizes the template-switching activity of the Moloney murine leukaemia virus reverse transcriptase to generate the first cDNA strand from the mRNA molecule, followed by second-strand synthesis.34 Tang method25 and Quartz-Seq26 are scRNA-seq protocols that use poly(A) tailing while examples of scRNA-seq protocols that utilize template-switching approach for reverse transcription include Smart-Seq,27 Smart-Seq2,28 STRT-Seq,29 and Drop-Seq.30 After the second strand synthesis by using either poly(A) tailing or template-switching approaches, polymerase chain reaction (PCR) can be performed for the cDNA amplification purpose.25–30 PCR after poly(A) tailing and template-switching are considered nonlinear amplification processes, and thus, the efficiency and effectiveness of the process are sequence-dependent.22 In vitro transcription (IVT), on the contrary, is a linear amplification protocol, which incorporates T7 promoter in the PCR primer to amplify cDNA following reverse transcription.11 Examples of scRNA-seq protocols that employ IVT amplification approach include CEL-Seq,31 MARS-Seq,32 and InDrop.24 CytoSeq is a scRNA-seq method that utilizes predefined or specific genes primers to run the PCR after reverse transcription, and this method is useful to study the specific genes expression profiles of thousands of single cells.17
Table 1 -
Comparison of different commonly reported scRNA-seq methods
||Full length: Tang Method, Quartz-Seq, Smart-Seq, Smart-Seq2
|5′-end only: STRT-Seq
|3′-end only: CEL-Seq, MARS-Seq, Drop-Seq, InDrop
|Specific predefined genes only: CytoSeq
|Strand specificity, UMI utility, and early multiplexing
||No: Tang Method, Quartz-Seq, Smart-Seq, Smart-Seq2
|Yes: STRT-Seq, CEL-Seq, MARS-Seq, Drop-Seq, InDrop, CytoSeq
||PCR post-polyA tailing: Tang method, Quartz-Seq,
|Template-switching-based PCR: Smart-Seq, Smart-Seq2, STRT-Seq, Drop-Seq
|In vitro transcription: CEL-Seq, MARS-Seq, InDrop
|Gene-specific primer-based PCR: CytoSeq
A total of 10 commonly reported scRNA-seq methods were compared based on several parameters like transcript coverage, strand specificity, UMI utility, early multiplexing and amplification approaches.
ScRNA-seq = single-cell RNA sequencing; PCR = polymerase chain reaction; UMI = unique molecular identifier.
Another key feature that differentiates various scRNA-seq protocols is the transcript coverage.11,17 Some protocols like Tang method,25 Quartz-Seq,26 Smart-Seq,27 and Smart-Seq228 cover the full-length transcript while some only cover the 5′-end of the transcript like STRT-Seq29 or 3′-end like CEL-Seq,31 MARS-Seq,32 Drop-Seq,30 and InDrop.24 Full transcript length coverage allows the users to study alternative splicing of the transcripts and RNA modification11,23 while targeted sequencing at the 3′ or 5′-end of the transcripts allows the users to incorporate UMI in their study.11 Inclusion of UMI in next-generation sequencing (NGS) has become a popular choice because these molecular barcodes could correct PCR amplification bias and improve the data accuracy.35 For scRNA-seq protocols that cover specific ends of the transcripts or genes,17,24,29–32 strand specificity is vital to provide information on the specific regions like overlapping regions of transcriptions between sense and antisense transcripts,36 and the strand specificity is often lost in the full-length scRNA-seq protocols during the PCR amplification process.17 Besides, for full-length scRNA-seq protocols,11,17 early multiplexing is always not performed, but this is usually carried out in scRNA-seq protocols that cover the specific ends of transcripts like STRT-Seq,29 CEL-Seq,31 MARS-Seq,32 Drop-Seq,30 InDrop,24 and CytoSeq.29 Multiplexing allows simultaneous sequencing of different samples simultaneously, which allows cost and time-saving.37
After establishment of the cDNA library, the next step will be the sequencing and data acquisition step.11 During this step, the sequenced data will first be presented as raw reads, which will then be further processed via trimming, deduplication, and alignment to produce the read count for data normalization.38 The reported scRNA-seq data could be affected by two types of variations: technical or biological variations.19,39 Technical variations may include batch effect when different batches of samples are used, data dropouts when the RNA amount is too low to be detected or amplification errors.39 Biological variation can be because of the stochastic expression of genes, suggesting that similar cell type may have different genes expression.19
The final cleaned and normalized scRNA-seq data can be analysed at the cellular, gene, or molecular level.11 At the cellular level, scRNA-seq allows the cells to be clustered based on the specific gene expression patterns, which is useful in determining cellular heterogeneity.4 Trajectory analyses are another form of advanced scRNA-seq data analysis that allows the users to study the dynamic alterations in a cell’s gene expression over different time points, which is particularly useful in understanding cellular differentiation and cell cycle progression.19 Compared to scRNA-seq, clustering and trajectory analyses are not possible for bulk RNA-seq, and this makes scRNA-seq more powerful than bulk RNA-seq when it comes to data analyses at the cellular level.11 At the gene level, scRNA-seq enables the study of the differential genes expression of thousands of single cells, and this could be useful in pathway analyses to study a specific disease pathway.40 At the molecular level, scRNA-seq provides powerful insights to study CNV41 and SNV.42 Both CNV and SNV are gaining popularity in the life science research because there are increasing shreds of evidence that suggest the changes in the nucleotide compositions of the human genome are related to a wide variety of diseases.43
3. APPLICATIONS OF scRNA-seq IN HUMAN LUNG CANCER RESEARCH
In human lung cancer research, scRNA-seq has been used widely to study tumour microenvironment (TME),44 tumour heterogeneity,45 tumour cells transcriptome profiles,46 and tumour response towards antineoplastic therapies (Fig. 2).40 In this section, the various applications of scRNA-seq in lung cancer research would be explored and discussed (Table 2).
Table 2 -
Various applications of scRNA-seq in lung cancer-related research (n = 30)
|Applications of scRNA-seq in lung cancer research
||Study designs (sample types, size, cell types)
||Key study findings
|Study of tumour cells genes expression profiles
||Clinical sampling (Stage IV NSCLC patients = 35, health subjects = 20), in vitro (HCC827, H661, A549, H1650, H1975, HCC827, and PCS-201)
||ScRNA-seq from both in vitro and in vivo data discovered that TERT and MET were highly expressed in the lung cancer cells, suggesting these two genes could serve as promising cancer biomarkers. In term of the EGFR mutation in different patients or lung cancer cells, it was observed that the patterns of EGFR mutations were quite heterogeneous between different patients or cell lines.
|In vitro (multiple cell lines like U2OS, H1299, HCT116, MDA-MB-231, hTERT, and REF52)
||ScRNA-seq analysis revealed that high level of E2F1 would promote apoptosis in multiple cancer cell lines, including lung cancer cell and low level of E2F1 would promote cellular proliferation.
|Clinical sampling (n = 1144 samples from Cancer Genome Atlas Network), in vitro (A549, CALU-1, NCl-H520, COR-L23, LCLC-103H, COLO-699, LXF-289, SKMES-1, and SKLU-1)
||ScRNA-seq data revealed that FBXO17 was co-localized with epithelial cell marker (EPCAM) and was dysregulated in lung cancer (5% of all lung cancer cases). Overexpression of FBXO17 would activate PI3K/Akt/mTOR pathway in lung cancer cells.
|Study of TME
||Clinical sampling (NSCLC tissue samples = 164), in vitro (A659 and NCI-H1299)
||Two types of B cells were infiltrated in the NSCLC tissue samples, namely, plasma-like and naïve-like B cells. The former was associated with enhanced tumour cells growth in advanced NSCLC, while the latter was related to good NSCLC prognosis. Four proteins (VNN2, IF130, SERPINA9, and PIK3AP1) released by naïve-like B cells were believed to suppress tumour cells growth.
|Study of TME and heterogeneity
||Clinical sampling (metastatic lung cancer patients = 44), survival analysis (TCGA data, LUAC = 494, and LUSC = 490)
||ScRNA-seq revealed that resident myeloid cells were replaced by monocyte-derived macrophages and dendritic cells in the TME and T cell exhaustion was shown. The changes in the immune and stromal cells dynamics eventually created a pro-cancerous and immunosuppressive TME.
|Study of TME and genes expression profiles
||Clinical sampling (NSCLC patients = 14), in vivo (C57BL/6 and Balb/c mice, n = unclear), survival analysis (data with only top 25% tumour mutation burden)
||T cell exhaustion occurs during prolonged antigen stimulation, and TOX was reported to be promoting CD8+ T cells exhaustion by upregulating immune checkpoints (IC) genes. Thus, inhibiting TOX protein can impede T cell exhaustion, potentially improving immune checkpoint inhibitor (ICI) effectiveness.
|Clinical sampling (lung cancer biopsy samples = 8, formalin-fixed paraffin-embedded (FFPE) tissues = 90), in vivo C57BL/6J and NOD/SCID (n = 5/group), in vitro (LLC)
||Tumour-associated macrophages (TAM) are essential in contributing to cancer progression. The use of scRNA-seq revealed that mincle/Syk/NF-κβ pathway is responsible for promoting the pro-tumoural activity of TAM in the development of lung cancer.
|Study of tumour cells heterogeneity and genes expression profiles
||Xenograft sampling (34 single cells of human LUAC obtained from immunocompromised female NOG mice)
||The authors identified a gene module named G64, which contains the cell-cycle-related genes predominantly. This G64 distinguished two intratumoural subgroups and E2F1 was discovered as the transcription factor that mostly mediates the G64 module expression in a single LUAC cell. Besides, intertumoural heterogeneity was also detected between the different lung cancer cell, and single lung cancer cells with upregulated G64 were associated with poor prognosis.
|Clinical sampling (human LUSC samples = 502, normal lung samples = 51 from TCGA), in vivo NIH Swiss mice (n = 2)
||Unique clonal mutation patterns were detected in the LUSC in which some clones carried mutant Trp53 genes while some carried mutant Igfbp7 gene. The expressions of genes involved in regulating immune checkpoints like PD-L1, LAG3, VISTA, and TIM3 were significantly dysregulated.
|Clinical sampling (lung cancer patients samples = 2), in vitro (LC2/ad-R and LC2/ad)
||Intra- and inter-tumoural heterogeneity were observed in the lung cancer cells. The downregulation of IFN-γ pathway genes was associated with the acquired resistance phenotype in the lung cancer cells. Besides, tumour related antigens like neoantigens and cancer-testis antigens were heterogeneously expressed in single cancer cells.
|Study of tumour cells heterogeneity and genes expression profiles
||Clinical sampling (Stage 1A LUAC patients = 2), in vivo (KrasLSL-G12D/WT, p53flox/flox, Rosa26LSL-eYFP mice, n ≥ 4)
||Using both human and mouse lung epithelial cells, overexpression of KRAS reduced the expression of differentiation and maturation markers (CD74, LYZ2, SFTPC, SFTPD, and NKX2-1) in the early-stage lung cancer AT2 cells. The KRAS overexpressed AT2 cells are transcriptionally distinct and underwent development transition than maturation state.
|Clinical sampling (SCLC patients = 21), in vitro (SCLC cell line), in vivo (Rb1fl/fl, Trp53fl/fl, Rbl2fl/fl mice, n = 6/group)
||ScRNA-seq revealed that SCLC consists of various molecular subtypes like NEUROD1+, YAP+, ASCL1+, and POU2F3+. It was shown that MYC would activate Notch to de-differentiate the SCLC cells, and subsequently, caused the SCLC to shift from ASCL+ to NEUROD1+ and finally YAP1+ subtype. All SCLC subtypes arise from the same neuroendocrine cell of origin and MYC is important in reprogramming the tumour cells.
|Clinical sampling (human LUAC patients = 5, healthy control = 3)
||Expression of CEACAM6 and SCGB3A2 was elevated in the CTCs found in the CSF of lung cancer patients with brain metastases, and these two candidate genes could serve as biomarkers for lung cancer brain metastases. ScRNA-seq analysis discovered tumour heterogeneity in the CSF-CTCs between different lung cancer patients and different individual cells from the same patients.
|Study of tumour cells heterogeneity, genes expression profiles and treatment response
||Clinical sampling (lung cancer patients = 2), in vivo (female NOG NOD/Shi- SCID/IL-2Rγ-null mice, n = unclear)
||Patient-derived xenograft lung adenocarcinoma cells (PDX) were sequenced using scRNA-seq, and 50 tumour-related single-nucleotide variations like KRASG12D were detected. Depending on KRASG12D and risk score status, PDX cells with high KRASG12D expression and the high-risk score would be drug-resistant. The use of multiple anti-neoplastic agents with different modes of action would reduce the KRASG12D expression in the PDX cells and sensitize the cells towards chemotherapy.
|In vitro (LC2/ad, PC-9, VMRC-LCD)
||ScRNA-seq revealed that the treatment of multi tyrosine kinase inhibitor vandetanib would reduce the relative expression of ribosomal and housekeeping genes in the lung cancer cells, but, the expressions of genes targeted by these therapies were not changed much by these therapies and remain constant. The diversity in the expressions of cancer-related genes sometimes remained latent in an untreated state, and the expressions of these genes could become more diverse following drug treatment.
|Study of TME, heterogeneity and genes expression profiles
||Clinical sampling (NSCLC treatment-naïve patients = 14), survival analysis (TCGA data = 473 lung cancer patients)
||ScRNA-seq revealed two groups of CD8+ T cells, one at “pre-exhausted state” while another group was at “exhausted state”. High ratio of “pre-exhausted” to “exhausted” T cells were associated with good NSCLC prognosis. For tumour regulatory T cells (Tregs), a bimodal distribution of TNFRSF9 was observed, suggesting heterogeneity among the Tregs. Activated Tregs with high expression of IL1R2 was associated with poor prognosis in NSCLC.
|Clinical sampling (lung cancer patients = 5), survival analysis (TCGA data and other ublication dataset in which LUAC = 1027; LUSC = 545)
||Lung cancer cells are embedded in a TME consisting of a complex population of stromal cells. The use of scRNA-seq revealed that at least 52 stromal cell subtypes were present in the lung cancer samples. Further analysis discovered that tumour endothelial cells suppressed immune attraction pathways, and B cells are highly enriched in the tumour. Besides, lung cancer houses five different types of fibroblasts, and high expression of stromal markers was associated with decreased survival in LUSC patients.
|Clinical sampling (early stage NSCLC patients = 4)
||ScRNA-seq revealed that myeloid cells populations exhibited dynamic changes between normal and lung cancer cells, and identified a pathway that suggested differentiation of CD14+ monocytes to M2 macrophages. Genes involved in regulating oxidative phosphorylation and p53 like TREM2, PPARG, MSR1/CD204, and MRC1/CD206) were upregulated. In contrast, genes involved in regulating TNF-α and NF-κβ pathways like CXCL2 and ILIB were downregulated.
|Study of TME, heterogeneity and genesexpression profiles
||Clinical sampling (lung cancer samples = 7)
||Distribution of nine pro-oncogenes was analysed using scRNA-seq. BUB1B, BUB1, and TTK were expressed on T cells, NK cells, and myeloid cells. CDC45 was mostly expressed on NK and myeloid cells. CDC6, CCNB1, and CCNB2 were expressed on B and plasma cells, T cells, NK cells, and myeloid cells. CHECK1 was expressed on all cell types while RRM2 was not expressed on any cell types. High TTK protein expression level was associated with advanced disease stage and poor prognosis.
|Clinical sampling (stage I–III NSCLC patients samples = 41, paired normal tissues = 41), in vitro (A659 and NCI-H1299), survival analysis (n = 888 from both Southampton University Hospital cohort and TCGA database)
||The authors classified the tumour-associated macrophages into M1-like and M2-like, and the former has anti-tumour properties while the latter has pro-tumour properties. Therefore, the presence of M1-like subtypes is associated with better prognosis as this macrophages subtype could activate tissue-resident memory T cells to eradicate the tumour cells. M1-associated markers include CXCL9, CXCL10, CXCL11, CXCL12, STAT1, and FAM26F.
|Clinical sampling (Stage I–III NSCLC patients samples = 100); survival analysis (TCGA cohort = 6853)
||The relationships between tumour mutation burden (TMB) and T cell development were studied in the study. It was found that TMB was correlated tightly to T cell differentiation skewing, which led to the loss of TCF7 expressing CD4 T cells and increased the number of dysfunctional CD8 and CD4 T cells. The T cells shifting from functional to dysfunctional states were associated with poor survival among lung cancer patients.
|Study of TME, heterogeneity and genes expression profiles
||Clinical sampling (Stage I or II LUAC samples = 7, adjacent normal tissues = 5), in vitro (A549 and NCI-H1975)
||Myeloid and T cells were the most commonly found stromal cells in the lung tumour and adjacent normal tissues. There was also an increase in the number of CD1C+ dendritic cells and macrophages in the tumour tissues, and these cells showed pro-tumoral activities. Tumour-infiltrating T cells showed regulatory T-cells and exhausted features. The LUAC cells were classified into several subtypes based on the unique expression signatures in various pathways like cell cycle, cell metabolism, translation initiation, antigen presentation, hypoxia, and glycolysis.
|Clinical sampling (NSCLC treatment-naïve patients = 14)
||This scRNA-seq-based study was aimed to identify specific transcript isoforms that are present in different T cells, and isotype analysis showed that particular isotypes of 456 genes were found in CD4+ T cells while specific isotypes for 245 genes were present in CD8+ T cells. The isoform switching could help understand the regulatory mechanism of tumour-infiltrating T cells, especially during the T-cell activation or differentiation stages.
|Clinical sampling (lung adenocarcinoma with GGN = 5, lung SADC = 5)
||Using scRNA-seq to compare cellular and molecular characteristics of lung adenocarcinoma with GGN and SADC, it was noted that in the former lung adenocarcinoma subtype, angiogenesis signaling pathway was downregulated, more immune cells were activated, and the fibroblasts expressed lower collagens levels. This could explain why lung adenocarcinoma with GGN has a better prognosis than SADC.
|Study of TME, heterogeneity and genes expression profiles
||Clinical sampling (four pairs of NSCLC and normal tissues from Gene Expression Omnibus [GEO])
||A total of 14 lung tumour-related cells were clustered, and the top five cell types include CD4+ T cells, cancer stem cell (CSC), NK cell, B cell, and plasma cell. The overly expressed genes in the lung tumour cells were those involved in regulating immune-related pathways like TNF, toll-like receptor, cytokine-cytokine receptor interaction, and IL-17 signaling pathway. Besides, helper T cells, memory CD4+ T cells, naïve B cell, and M1 macrophages were found to have a higher proportion in the lung tumour tissues than normal tissues.
|Clinical sampling (NSCLC patients = 7), in vivo C57BL/6 mice study (n ≥ 2), survival analysis (n = 720 LUAC patients)
||Tumour-infiltrating myeloid cells (TIM) consist of neutrophils, monocytes, macrophages, and dendritic cells. ScRNA-seq revealed that TIM populations were conserved across different lung cancer patients and different species (human and mice). Several genes of the TIM like ISG15 and PI3 were correlated negatively to the patients’ survival while CD207 was associated with good prognosis.
|Study of TME, heterogeneity, genes expression profiles, and treatment response
||Clinical sampling (lung cancer biopsy samples = 27 in which responders = 7, treatment-naïve = 8, nonresponders = 12), ex vivo (n ≥ 6)
||Presence of tissue-resident memory T cells (TRM) was associated with good prognosis in lung cancer patients, and TRM that co-expressed both PD-1 and TIM3 was shown to improve treatment response towards PD-1 inhibitors. Besides, TRM cells were also enriched in transcripts that are responsible for inducing TRM proliferation and cytotoxicity.
|Study of TME, heterogeneity, genes expression profiles, and treatment response
||Clinical sampling (lung SCC samples = 4, LUAC = 3), in vitro (Lewis lung carcinoma cell [LLC], 293T, primary human umbilical vein endothelial cells [HUVECs], primary human tumour and peritumoural endothelial cells), in vivo (C57BL6/J mice, n = 6/group)
||ScRNA-seq of the endothelial cells from the human and murine lung cancer showed different, distinct subpopulations of endothelial cells that expressed specific signatures related to immune cells regulation, breakdown of the basement membrane and antigen presentations. The transcriptome signatures of the tip endothelial cells correlated to the patients’ survival and these populations of cells were most sensitive to VEGF blockade.
|Clinical sampling (metastatic lung cancer patients = 30), in vitro (PC9 & H3122), survival analysis (MSK-impact data =141, TCGA = unclear)
||Individual cells that showed molecular diversity and therapy-induced plasticity were observed in both tumour and microenvironment cells. Cancer cells at the progressive disease state showed upregulation in the kynurenine, plasminogen, and gap-junction pathways while cancer cells at the residual disease state showed alveolar-regenerative signatures that indicated therapy-induced primitive cell transition state.
|Clinical sampling (stage IV NSCLC patients = 4)
||PD-1 receptor blockage is a type of immunotherapy used in treating NSCLC, and the authors reported dynamics changes in individual T cells clones during anti-PD-1 treatment. This included the drastic decrease in PD-1+ T cells following PD-1 receptor blockade and increased cytotoxic activity in the tumour-related CD4+ T cells clones compared to the CD8+ T clonal populations. A total of 25 genes were either up- or downregulated following treatment progressions, and these include genes like CSCR4, DUSP2, and ZFP36. Overall, the PD-1 blockage resulted in the decreased activity in the cytokine-cytokine receptor interaction pathway in the T cells, which may help kill the tumour cells due to the reduced level of dysfunctional T cells.
ScRNA-seq was widely used to study the lung cancer TME, tumour heterogeneity, gene expression profiles, and treatment response. This review altogether summarized the findings from 30 in vitro, in vivo, and clinical studies on the various applications of scRNA-seq in human lung cancer research.
CSF = cerebrospinal fluid; CTCs = circulating tumour cells; GGN = ground glass nodule; LUSC = lung squamous cell carcinoma; NK = natural killer; NSCLC = non-small cell lung cancer; SADC = solid adenocarcinoma; ScRNA-seq = single-cell RNA sequencing; TME = tumour microenvironment; VEGF = vascular endothelial growth factor.
3.1. Dissecting TME
Cancer does not involve only an isolated mass of malignant cells, but it also involves many other surroundings, supporting cells.69 TME is a term used to describe the complex interaction between the cancer cells and surrounding noncancerous cells.70 The noncancerous cells that could be present in the TME include immune cells, stromal cells-like fibroblasts, adipocytes, vascular cells, and neuronal cells.70 The immune cells that could be found in the TME include natural killer (NK) cells, T lymphocytes, B lymphocytes, plasma cells, monocytes, macrophages, and neutrophils.69,70 TME is a highly heterogeneous and dynamic environment, and the presence of various substances like cytokines, chemokines, growth factors, enzymes, and extracellular matrix (ECM) in the TME is responsible in sustaining tumour growth.69,70 However, in some other study,71 it was proposed that TME could act as a double-edged sword in promoting or inhibiting tumour growth, and this is dependent on the phases of tumour progression. In the early stage of tumour growth, it is believed that TME has tumour-inhibitory effects, and as the tumour continues to grow, the effect becomes tumour-promoting.71 Out of all the different stromal cells that could be present in the lung cancer TME, myeloid and T cells were reported to be the most commonly found stromal cells according to a scRNA study finding65 whereas, in another study,44 it was found that T cells, B cells, plasma cells, and NK cells were the several cell types, which were commonly found in lung cancer TME. The myeloid cell is recognized as one of the most abundant stromal cells found in cancer TME.72 Still, in specific cancer types like breast cancer, fibroblasts were the most abundant stromal cells in the cancer TME.73 Therefore, the type of commonly found TME cells could be varied in different studies or different studied cancer type.
Monocytes and macrophages are critical immune cells found in the TME.74 For macrophages, it has been reported that there are two different phenotypes of macrophages in the TME, namely, activated (M1) and alternatively activated (M2) phenotypes.75 M1 macrophage is a kind of pro-inflammatory macrophage, which secretes several cytokines and chemokines, while M2 macrophage produces anti-inflammatory substances like IL-10.75 Thus, M2 phenotype is the usual macrophage found in the TME as they help maintain a pro-tumoural environment by creating an immunosuppressive environment for the tumour cells to grow.75 In lung cancer, a similar phenomenon was observed in which M1-like macrophages were found to have anti-tumoural properties while M2-like macrophages were said to have pro-tumoural properties.63 The presence of M1-like macrophages in lung cancer, especially NSCLC, could be associated with good prognosis and M1-associated markers, including CXCL9, CXCL10, CXCL12, and FAM26F.63 Like CXCL9 and CXCL10, some of these markers have been reported to be the M1-associated markers in human glioblastoma.76 In another lung cancer scRNA-seq study, M2 phenotype was also found to be the predominant macrophage that was present in NSCLC and LLC, and Mincle/Syk/NF-kB signaling pathway was shown to be responsible in driving the macrophage phenotype switching from M1 to M2 in the lung cancer TME.14 NF-kB is a homo- or hetero-dimer of the transcription factor Rel family.77 It has been reported that NF-kB plays a crucial role in determining the macrophage polarization in cancer, and this is mainly dependent on the cancer type and tumour growth stage.77 On the contrary, it was found that CD14+ monocytes could transform to become M2 macrophage, and this differentiation pathway was detected using trajectory analyses.41 This differentiation pathway was demonstrated in all four NSCLC patients, and several key transcription factors like JUN and NF-kB1A were found to be involved in regulating this monocyte-M2 differentiation.41 Both JUN and NF-kB1 (p50) have been proven to play essential parts in regulating the transformation of myeloid cells like M1 to M2 phenotypes by interacting with different targets like NF-kB.78,79 Besides, specific ligand-receptor interactions like SFTPA1-TLR2 and ICAM1-ITGAM were involved in regulating the monocyte-M2 differentiation, and this suggested that epithelial cells could also play roles in coordinating this cellular differentiation.41 In cancer-like prostate cancer, it was shown that prostate epithelial cells could also play a part in inducing the M2 polarization80 and this further supported that epithelial cells are also important in regulating the dynamic of TME.41,80
In TME, several T lymphocytes or T cells could be present, and these include CD8+ T cell, CD4+ T cell, and memory T cell.13 ScRNA-seq can be used to study alternate splicing,9 and this is important to understand T cell activation and differentiation.48 It was reported that different specific transcript isoforms could be present in different T cells, and based on a study finding,48 it was shown that specific isotypes of 245 genes were present in CD8+ T cells while specific isotypes of 456 genes were present in CD4+ T cells. The use of scRNA-seq to study alternate splicing in T cells enabled the understanding of isoform switching, and this is important to study T cell heterogeneity, T cell regulatory mechanism, T cell activation and differentiation.48 T cell exhaustion is a poorly understood cellular phenomenon that could affect immunotherapy efficacy.13 It could be because of chronic antigen exposure.81 Using the scRNA-seq approach, it was shown that tumour-infiltrating T cells showed exhaustion features, evidenced by the elevation of the exhaustion markers like TIGIT and CTLA4.65 TIGIT is a co-inhibitory receptor, which is expressed on T cell surface,82 while CTLA4 is a T cell co-receptor, which regulates cell motility,83 and both TIGIT and CTLA-4 are essential in mediating T cells exhaustion.82,83 To understand the mechanism that leads to T cell exhaustion, in a scRNA-seq study, it was reported that TOX could promote CD8+ T cells exhaustion in NSCLC by upregulating immune checkpoint genes like PD-1, TIM-3, CTLA-4, and TIGIT.13 This helped reveal the molecular mechanism that caused CD8+ T cells exhaustion in NSCLC and inhibiting TOX could slow down T cells exhaustion.13 Other than exhausted CD8+ T cell, another group of CD8+ T cells is known as pre-exhausted T cell in the lung cancer TME.61 Compared to exhausted T cells, it was shown that the presence of a higher proportion of pre-exhausted T cells in the TME could be linked to better prognosis among the NSCLC patients.61 To explain this, this could be because pre-exhausted T cells could respond better to immune checkpoint inhibitor therapy as they express a higher number of inflammatory-regulatory proteins like cytokines, TCF-1, and CXCR-5.84 Therefore, strategy to expand the tumour-infiltrating pre-exhausted T cells population could be applied to improve the lung cancer patients’ response towards the immune checkpoint inhibitor therapy.84 TCF7 is a transcription factor essential for T cells differentiation and in cancer in which the T cells are expressing TCF7, the cancer patients would respond better to immunotherapy and have better survival.85 In a lung cancer study,64 which combined the findings from scRNA-seq, bulk analysis, and flow cytometry, it was shown that patients with high tumour mutation burden would show T cell differentiation skewing, which was characterized by the loss of TCF7-expressing progenitor-like CD4 T cells and increased the number of dysfunctional CD8 and CD4 T cells. The T cells shifting from progenitor-like functional state to the dysfunctional state was associated with poor survival among lung cancer patients.64 Other than CD4+ T cells and CD8+ T cells, another type of T cells called tissue-resident memory T cells (TRM) was also present in the lung cancer TME, and the presence of this cells are said to correlate to good prognosis in lung cancer patients.49 TRM cells that co-expressed both PD-1 and TIM-3 could improve treatment response towards immunotherapy49 as these two targets have been recognized to play vital roles to reverse T cells exhaustion and to restore the anti-tumour immune response.86
B lymphocyte or B cell is one of the essential cellular components of the TME, and it plays a vital role in regulating adaptive immunity processes like antibody production and antigen presentation.87 In a scRNA-seq study, which involved the clinical sampling of NSCLC tissues, it was found that there are two types of B cells, which were present in the NSCLC samples and these included plasma-like and naïve-like B cells.53 Plasma-like B cells were said to suppress tumour growth in the early NSCLC stage but promote tumour growth in advanced NSCLC as the IgG extracted from early NSCLC would inhibit lung cancer cells growth in vitro but IgG isolated from advanced NSCLC would accelerate lung cancer growth in vitro.53 On the contrary, naïve-like B cells were found to be associated with good prognosis in NSCLC as subsequent analyses showed that naïve-like B cells could secrete four types of proteins to suppress NSCLC growth, and these proteins are VNN2, IF130, SERPINA9, and PIK3AP1.53 In a scRNA-seq study88 which focused on human melanoma, it was shown that the frequency of naïve-like B cells was significantly higher in patients who were responding well to immune checkpoint inhibitor therapy, and this further supported that naïve-like B cells could help improving prognosis in human cancer. The same study also demonstrated that plasmablast-like B cells could have both pro- and anti-inflammatory roles in human melanoma,88 and these suggested that plasmablast-like B cells could affect treatment response and patients survival by modulating the inflammatory process in the TME.
Other than immune cells in the TME, scRNA-seq was also being utilized to study the functional and phenotypic roles of tumour endothelial cells in the lung cancer TME.67 Endothelial cells are key cellular components in the TME, which initiate and promote angiogenesis, and this cellular process is vital to sustaining cancer growth.89 Through scRNA-seq, more than 30 different phenotypes of lung cancer endothelial cells were discovered, and examples of these endothelial cells phenotypes include migratory tip and basement membrane remodeling phenotypes.67 Advanced analyses showed that the tip endothelial cells shared conserved markers between humans and mice.67 The tips endothelial cells that expressed molecular signatures that promoted endothelial cells growth, migration, and basement membrane remodeling were associated with poor disease prognosis.67 On the contrary, the endothelial cells at the tip or breach regions were most sensitive to anti-vascular endothelial growth factor (anti-VEGF) treatment,67 and these suggested that anti-VEGF treatment should be targeted at these cells. Using scRNA-seq to study cancer-associated endothelial cells is not new in cancer research, and it has also been previously used to study endothelial cells in other human cancers like breast cancer,90 melanoma,91 and bladder cancer.92 Therefore, by utilizing scRNA-seq, the various genotypes and phenotypes of the cancer-related endothelial cells can be studied at greater details, and this is very useful in planning an appropriate anti-angiogenesis treatment plan for the cancer patients.67
Some lung cancer scRNA-seq studies only focused on specific TME cell types like B cells53 and T cells,13 while some other scRNA-seq studies concentrated on different immune and stromal cells in the TME.54,63 In a clinical study54 that involved metastatic lung cancer samples from 44 patients, it was shown that monocyte-derived macrophages and dendritic cells replaced the resident myeloid cells in the TME, and T cell exhaustion was demonstrated. Myeloid cells were believed to play vital roles in regulating cellular homeostasis and maintaining an immunosuppressive microenvironment.93 The replacement of these cells by macrophages and dendritic cells suggested that TME remodeling promoted tumour growth.54 T cell exhaustion has been reported to link to reduced response to immunotherapy.13 The occurrence of T cell exhaustion further suggested that there were dynamic changes in the TME in metastatic lung cancer to promote carcinogenesis.54 On the other side, the use of scRNA-seq revealed that at least 52 stromal cell subtypes were present in the lung cancer samples.47 These include fibroblasts types that expressed different collagen sets and tumours endothelial cells that suppressed the immune attraction pathways.47 The detection of different fibroblasts types in the study suggested that different fibroblast could have different functional roles in the TME,47 which meant that different fibroblasts are specialized in their unique ways in maintaining the ECM.94 The discovery of tumour endothelial cells, which suppressed the immune homing and activation pathway suggested that they could contribute to tumour immunotolerance.47 Immunotolerance or immunological tolerance refers to the state in which the host immune system is unresponsive to the specific antigen, which is essential in preventing autoimmune disease.95 The tumour endothelial cells’ remodeling, which led to immunotolerance, would eventually cause cancer treatment resistance, particularly cancer immunotherapy.96 Other critical information reported in the same lung cancer scRNA-seq study was the detection of strongly enriched B cells and high stromal markers in lung cancer.47 It has been summarized in a previous review that the presence or infiltration of B cells in the lung cancer would promote tumour progression in either premalignancy or malignancy state.97 High stromal markers were found to correlate tightly to decreased survival among lung cancer patients, particularly the lung squamous cell carcinoma (LUSC) patients.47 It has also been reported that high stromal markers were associated with invasive colorectal cancer phenotypes and poor prognosis.98
3.2. Understanding tumour heterogeneity
Human cancer is a highly complex and heterogeneous disease, and the heterogeneities are reflected at the genetic, epigenetic, and phenotypic levels.42,99 Intratumoural heterogeneity means the different tumour cells within the same tumour may exhibit different genetic or phenotypic characteristics, and this can be observed in patients having a similar type of cancer.99 On the other hand, intertumoural heterogeneity suggests that heterogeneity is observed in different tumour types.42 To assess and understand tumour heterogeneity, scRNA-seq is more superior than bulk RNA-seq, which measures the transcriptome profile from RNA pool gathered from various cancer cells and microenvironment cells.99
In a study45 that employed scRNA-seq to study intratumoural heterogeneity in LUAC and lung cancer cell lines, it was demonstrated that genes that are involved in regulating cell cycle and antigen presentation were independently co-expressed among the single-sequenced cells, while genes that are involved in mediating IFN-γ pathway were heterogeneously expressed within the sequenced cancer cells. Also, potential targets for cancer vaccination like CTA and neoantigens were heterogeneously expressed in single cells.45 The intratumoural heterogeneity could potentially affect the clinical prognosis as different cells from same cancer may respond differently to the treatment given.59 For example, it has been established that the downregulation of the genes involved in the IFN-γ signaling pathway could be associated with immunotherapy resistance,100 Thus, a combination of multiple therapeutic strategies should be used in ensuring the complete eradication of lung cancer without tumour escape.45
SCLC is a poorly differentiated lung neuroendocrine tumour with poor prognosis.101 ScRNA-seq has been employed to study the intratumoural heterogeneity of SCLC, and it was found that SCLC has several distinctive molecular subtypes, which express YAP1, ASCL1, NEUROD1, and POU2F3.15 MYC, an oncogene linked to the development of multiple human cancers,102 was found to activate another gene called Notch to de-differentiate SCLC cells.15 Subsequent analysis showed that MYC could drive the SCLC’s temporal evolution from ASCL1+ subtype to NEUROD1+ subtype, and finally to YAP1+ subtype.15 All these different SCLC molecular subtypes arose from the same neuroendocrine cell of origin.15 This study finding suggested that scRNA-seq has essential value in studying both intratumoural heterogeneity and tumour evolution,15 and this technique has also been used in other cancers like breast11 and colorectal cancer12 to understand the tumour evolution. LUAC can be divided into solid adenocarcinoma (SADC) and adenocarcinoma with ground glass nodule (GGN).103 In a lung scRNA-seq study, which compared the cellular and molecular characteristics between SADC and GGN, it was found that the LUAC with GGN showed downregulation in the angiogenesis pathways, activation of the immune system, and reduced expression of collagens by the fibroblasts.46 All these molecular features like angiogenesis suppression have been recognized as crucial factors that improve the survival among the lung cancer patients.104 This finding provided useful insights to explain why LUAC patients with GGN features showed better prognosis than LUAC patients with SADC features.46
Apart from being useful in studying lung cancer intratumoural heterogeneity, scRNA-seq is also being used widely in understanding intercellular heterogeneity in lung cancer.62 For instance, different proto-oncogenes could be over-expressed in different cells of the TME, and for examples, BUB1B, BUB1, and TTK were highly expressed in T cells, NK cells, and myeloid cells, while CDC6, CCNB1, and CCNB2 were over-expressed in B cells and plasma cells.62 The unique genetic expression in different cell types enabled them to be clustered based on the highly expressed genes, and this is particularly useful in studying intercellular heterogeneity.62 All these proto-oncogenes have been proven to play crucial roles in driving tumourigenesis in different human cancers like gastric and breast cancers.105–107 Therefore, the expression patterns of these proto-oncogenes in different TME cells could be potentially used as biomarkers in lung cancer patients.62 In another study, which was aimed to study the heterogeneity between different types of T cells,48 it was shown that different isoforms of different genes were present in CD4+ T cells and CD8+ T cells, respectively. This meant the cellular heterogeneity between different T cells could be because of the alternate splicing of specific genes during T cells development.48
The ability of scRNA-seq to profile thousands of single cells has also enabled it to be a superior tool to study the cellular and molecular characteristics of different cell types across different species.108 In a scRNA-seq study, which focused on human and mice lung cancers, it was shown that both species shared common and unique neutrophil subsets, and conserved, distinct monocyte, and dendritic cell subsets were found in both species.66 Besides, the study also discovered macrophage subsets, which were species-specific.66 This study finding has helped dissect the cellular heterogeneity and the discovery of TME cell populations that were conserved between different species.66 This could help facilitate future translational research that involves both human and mice model.66 A similar study approach91 has also been applied to study tumour-associated myeloid cells in both human and mice melanoma tumours other than lung cancer. It was found that there were at least three distinct stromal populations, which were conserved between the human and mice melanoma tumours.91
3.3. Understanding genes expression profiles of tumour or tumour-related cells
Other than being useful in studying TME and tumour heterogeneity, scRNA-seq is also widely used in investigating the genes expression profiles of the tumour cells and the TME cells.65 FBXO17 is a negative regulator of GSK-3β, and it is vital in upregulating the oncogenic Wnt/β-catenin signaling pathway.109 The finding from a scRNA-seq study revealed that FBXO17 is dysregulated in human lung cancer cells.52 The overexpression of FBXO17 is linked to the upregulation of several targets responsible for driving cellular proliferation like Akt, ERK1/2, and mTOR.52 This finding suggested that FBXO17 could potentially act like a tumour-promoting gene in lung cancer.52 FBXO17 has also been proven to play an oncogenic role in other cancer like liver cancer.109
One of the distinctive features of lung cancer cells is aberrant differentiation.110 In a study that upregulated oncogene KRAS in both human and mouse lung epithelial cells,57 it was shown that the overexpression of this oncogene would reduce the expression of differentiation and maturation markers in the early-stage lung cancer cells. Examples of the differentiation and maturation markers investigated in the mentioned study included CD74, LYZ2, SFTPC, SFTPD, and NKX2-1.57 The distinctive transcriptional differences between the lung epithelial cells with KRAS overexpression and lung epithelial cells without KRAS overexpression suggested that the upregulation of this oncogene would make the lung epithelial cells to lose its lineage identity.57 This finding also meant that scRNA-seq is a useful tool to study the clonal evolution of lung cancer cells by considering its unique, dysregulated genetic expression profiles.56
In another in vitro study,51 scRNA-seq revealed that high expression level of the transcription factor, E2F1, was associated with apoptosis in multiple cancer cell lines, including lung cancer cells. In the past years, it has been widely established that E2F1 is a key player regulating cell cycle progression, apoptosis, and DNA damage response.111 Thus, upregulating E2F1 could be potentially employed as a therapeutic strategy to induce lung cancer cells apoptosis.51 However, in a different study,55 E2F1 was shown to be an upstream regulator of a gene module called G64 consisting mainly of cell-cycle regulating genes, and the LUAC cells with upregulated G64 were associated with poor prognosis. The contradicting findings from these two studies51,55 suggested that more study should be conducted to investigate whether upregulation of E2F1 would promote or suppress lung cancer progression.
G80 module is another set of cancer genes, which were found to be dysregulated in human and mice LUSC samples by using scRNA-seq analyses, and these genes are mainly involved in mediating six different cellular pathways that include focal adhesion, cell cycle, PI3K/Akt, p53, ErbB, and ECM-receptor interaction pathways.56 In other words, scRNA-seq could be used to effectively classify LUSC tumours based on the alteration in the expressions of the genes included in this G80 module.56 Besides, this G80 module could also be utilized as a predictive biomarker to assess patient survival as the dysregulation of the cellular pathways included in this G80 module like PI3K/Akt112 and p53 pathways113 have been long proven to associate tightly with enhanced tumourigenesis and poor prognosis.
Metastases is an important cancer hallmark.114 A metastatic biomarker can help predict the probability of cancer occurrence and help in planning a personalized cancer therapy.115 In a study that focused on the circulating tumour cells (CTCs) isolated from the cerebrospinal fluid (CSF) of lung cancer patients, it was shown that two markers, namely, CEACAM6 and SCGB3A2, were highly expressed.58 CEACAM6 is a cell surface adhesion receptor, which modulates the ECM interactome, and its dysregulation has been reported to associate with cancers like pancreatic cancer.116 SCGB3A2 is a protein expressed in the alveolar epithelial cell, and it has been proven to be a novel biomarker for lung cancer in both human and mice models.117 The scRNA-seq finding from the CSF CTCs study suggested that both CEACAM6 and SCGB3A2 could be used as metastatic markers to identify lung cancer brain metastases and predict the overall disease prognosis.58 A different combined clinical and in vitro study demonstrated that TERT and MET were both highly expressed in the lung cancer cells and CTCs,50 suggesting that these two genes could be potentially used as lung cancer biomarkers. TERT activation has been proven to drive cellular immortalization and carcinogenesis.118 In contrast, MET is a tumour-promoting gene, which has been demonstrated to increase cellular growth, epithelial-to-mesenchymal transition and metastases.119
Apart from being applied in studying the specific genes expression profiles of lung tumour cells, scRNA-seq has also been widely used to investigate the genes expression profiles of TME cells.41 For example, TTK was found to be over-expressed in TME cells like NK cells and T cells, and the expression of this gene was associated with advanced disease stage and poor prognosis.62 TTK has been reported to promote tumour invasion and treatment resistance in solid cancer like breast cancer.105 This explained why high TTK expression in lung cancer TME cells would associate with poor prognosis.62 On the other hand, scRNA-seq was also being utilized to study the specific gene expression profiles associated with cellular differentiation.41 In a study that was aimed to investigate the particular genes which were responsible in driving the differentiation of CD14+ monocytes to M2 macrophages, it was found genes that are involved in regulating oxidative phosphorylation and p53 pathways like TREM2 and PPARG were upregulated, while genes that are involved in regulating TNF-α and NF-κβ pathways were downregulated.41 These findings were consistent to previous studies’ conclusions in which both TREM2 and p53 are responsible in modulating M2 polarization,120,121 while downregulation of TNF pathway is essential for the emergence of M2 macrophages.122 Other examples involving the use of scRNA-seq to identify specific markers related to TME cells include discovering M1-associated markers in a study.64 These markers include CXCL9, CXCL10, CXCL11, CSCL12, STAT1, and FAM26F.64
3.4. Study of tumour response towards lung cancer therapies
Therapy resistance is a major challenge in cancer therapy.123 To better understand the detailed cellular mechanisms that lead to therapy resistance, it is important to understand the specific molecular signatures that promote carcinogenesis.11,59 With the capabilities to dissect TME, study tumour-related transcriptome profiles, and tumour heterogeneity, scRNA-seq is a suitable tool to investigate the specific molecular signatures responsible for driving lung cancer treatment resistance.65
IFN-γ pathway has a tumour suppressive role in human cancer by promoting tumour ischaemia, tumour clearance, and tumour immune surveillance.124 In lung cancer, IFN-γ pathway has been demonstrated to have tumour suppressive role in the cancer progression, and the scRNA-seq finding from a study has found that the downregulation of the genes involved in the IFN-γ pathway could be correlated to the immunotherapy resistance phenotype in the lung cancer cells.45 In solid cancers like breast and colorectal cancer, IFN-γ was found to upregulate the expression of cell cycle inhibitory proteins like p16 and p21, as part of its efforts in suppressing the tumour proliferation.124
Oncogene KRAS is a key player in lung cancer development and progression.57 In a scRNA-seq study aimed to investigate the single-nucleotide variation of KRAS in lung cancer using human and mice models, it was noted that KRASG12D expression was associated with chemotherapy-resistant in the patient-derived xenograft lung adenocarcinoma cells (PDX).59 In a recently published study,125 KRASG12D has been reported to cause myeloproliferation. This further supported that KRASG12D could be playing essential roles in promoting both tumour growth and treatment resistance in human cancer.59 Therefore, it is recommended that combined therapies that comprises drugs with different modes of actions like cytotoxic agent or drug that targets specific molecular targets should be employed to maximize therapy success.59
PD-1 receptor blockage is a type of immunotherapy used to treat lung cancer like NSCLC.54 ScRNA-seq has been applied to study the dynamic changes in the T cells clones during anti-PD-1 treatment, and it was found that the level of PD-1+ T cells reduced drastically following anti-PD-1 therapy while the cytotoxicity of CD4+ T cells increased.40 Further analyses showed that 25 genes were either up- or down-regulated following anti-PD-1 treatment and the overall effects of anti-PD-1 treatment were decreased cytokine-cytokine receptor interaction, which could reduce T cells activity.40 In another study,68 scRNA-seq was used to identify the cellular pathways that were dysregulated in lung cancer cells resistant to targeted therapy. It was shown that kynurenine, plasminogen, and gap-junction pathways were upregulated in cancer cells at progressive disease state.68 The dysregulation of these cellular activities could promote tumourigenesis because, for example, kynurenine could serve as oncometabolite126 while plasminogen system is vital to regulate angiogenesis, invasion, and metastases.127 The use of scRNA-seq to identify resistant cancer cells populations or associated transcriptome signatures linked to treatment resistance was also being applied in other cancer types like breast cancer.128 In breast cancer, tumour cells populations that could contribute to targeted therapy-resistant in breast cancer were identified, and subsequently, this allowed the planning of a combinatorial therapy to eradicate better the breast cancer.128
Vandetanib is a multi-targeted receptor tyrosine kinase that could inhibit EGFR and VEGF.129 It has been studied in both preclinical and clinical phases for treating solid cancer.129 In a scRNA-seq study that involved the use of seven types of human lung cancer cell lines,60 it was demonstrated that the use of vandetanib would not alter the relative expression of the molecular targets, which are targeted by vandetanib, even though vandetanib could reduce the relative expression of some ribosomal and housekeeping genes. The reason why the expressions of the molecular targets of vandetanib remain relatively unchanged was not explored further in the reported study.60 Still, it could be because of reasons like resistance to inhibition or targets mutations.130 Besides, it was noted that the diversity in the expressions of cancer-related genes like EGFR sometimes would remain latent in the untreated state, but the expressions of these genes would become more diversified following vandetanib treatment.60 It is not new to discover that the expressions of some cancer-related genes would become more diversified after specific cancer therapy. A recently published study showed that silencing some cancer-related genes like mTOR would eventually result in more genetic evolution and variation, ultimately contributing to cancer treatment resistance.131
4. CHALLENGES AND FUTURE DIRECTIONS: PATHWAY TOWARDS PERSONALIZED THERAPY
Many studies have reported the use and application of scRNA-seq in lung cancer research, but, there are still several challenges to be overcome. Some of the reported scRNA-seq studies focused only on secondary lung cancer cell lines.51,60 Whether or not these findings can be translated into clinical findings are still unknown. The lung is an organ, which consists of left and right lung, and it can be further divided into different lobes.132 For the clinical study, to obtain a sufficient number of single cells from different lung zones for analyses, the lung tissues need to undergo mechanical or chemical dissociation.66 It may be hard to obtain healthy and sterile single cancer cells from the tissues for scRNA-seq analyses.12 Besides, for scRNA-seq study, which involves the use of CTCs, the problem is that CTCs could be present in low amount and there is a risk of contamination in the blood.133 A validated single-cell isolation method should be used to minimize the risk of contamination and enhance the single cells capturing.133 Other problems, which could limit the use of scRNA-seq in lung cancer research, include the high costs of the experiments and the needs of stringent quality control steps.134 Apart from the experimental challenges, bioinformatics analyses of the scRNA-seq data are another technical challenge to be overcome.12 Commonly faced problems in the scRNA-seq data analyses include background noise, batch effects, and technical errors.135 Compared to the bulk transcriptome analyses, scRNA-seq bioinformatics analyses are more complex and contain more noise.4 As time passes, it is believed that more robust bioinformatics analyses tools will be made available to analyse scRNA-seq data.12
Lung cancer is a highly heterogeneous and mutated disease,45 and different patients may carry tumours, which respond differently to various therapies.45,59 The use of scRNA-seq in lung cancer research has provided another option for the scientists to better understand the detailed cellular and molecular mechanisms that lead to lung cancer progression at single-cell resolution (Fig. 3).62,68 Besides, it also helps to identify tumour-promoting TME and specific biomarkers that can be used to predict the prognosis and treatment response of lung cancer patients.45,46,67 Another important application of scRNA-seq in lung cancer research is that it allows identifying unique treatment-resistant lung cancer cell clones and the associated transcriptome signatures that contribute to therapy resistance.40,59,68 This helps the clinicians formulate suitable personalized or combinatory therapies to better eradicate lung cancer in the clinical settings.45,52,59 Therefore, it is hoped that more clinical and translational research could be conducted in the future to integrate the use of scRNA-seq in the individualized human lung cancer management. Besides, it is also looking forward that there will be more advancement and breakthroughs in the single-cell isolation techniques, scRNA-seq methods, and scRNA-seq bioinformatics analyses pipelines so that this tool can be more widely used in studying human diseases.
In conclusion, scRNA-seq is a powerful tool in lung cancer research, and it is useful to understand the cellular and molecular mechanisms, which promote lung cancer progression. Besides, scRNA-seq can be applied to identify suitable biomarkers for lung cancer and delineate the mechanism that drives treatment resistance in lung cancer. Even though there are still some experimental and bioinformatics challenges, which could limit the use of scRNA-seq in studying human lung cancer, it is believed that as time passes, more advancements in this technique will be discovered, and more translational study can then be conducted to study human lung cancer further using scRNA-seq.
This study was supported by FRGS grant by Ministry of Higher Education, Malaysia, grant number FRGS/1/2018/STG05/UNIM/02/1 and FRGS/1/2014/SG05/UNIM/02/1.
1. de Groot PM, Wu CC, Carter BW, Munden RF. The epidemiology of lung cancer. Transl Lung Cancer Res. 2018; 7:220–33.
2. World Cancer Research Fund. Lung Cancer Statistics. 2020. Available at. https://www.wcrf.org/dietandcancer/cancer-trends/lung-cancer-statistics
. Accessed December 7, 2020.
3. Hassan Lemjabbar-Alaouia OH, Yanga Y-W, Buchanana P. Lung cancer: biology and treatment options. Biochim Biophys Acta. 2015; 1856:189–210.
4. Chen G, Ning B, Shi T. Single-Cell RNA-Seq technologies and related computational data analysis. Front Genet. 2019; 10:317.
5. Fan J, Slowikowski K, Zhang F. Single-cell transcriptomics in cancer: computational challenges and opportunities. Exp Mol Med. 2020; 52:1452–65.
6. de Vries NL, Mahfouz A, Koning F, de Miranda NFCC. Unraveling the complexity of the cancer microenvironment with multidimensional genomic and cytometric technologies. Front Oncol. 2020; 10:1254.
7. Hoffman JA, Papas BN, Trotter KW, Archer TK. Single-cell RNA sequencing
reveals a heterogeneous response to Glucocorticoids in breast cancer cells. Commun Biol. 2020; 3:126.
8. Gojo J, Englinger B, Jiang L, Hübner JM, Shaw ML, Hack OA, et al. Single-cell RNA-seq reveals cellular hierarchies and impaired developmental trajectories in pediatric ependymoma. Cancer Cell. 2020; 38:44–59.e9.
9. Wu S, Zhang H, Fouladdel S, Li H, Keller E, Wicha MS, et al. Cellular, transcriptomic and isoform heterogeneity of breast cancer cell line revealed by full-length single-cell RNA sequencing
. Comput Struct Biotechnol J. 2020; 18:676–85.
10. Serin Harmanci A, Harmanci AO, Zhou X. CaSpER identifies and visualizes CNV events by integrative analysis of single-cell or bulk RNA-sequencing data. Nat Commun. 2020; 11:89.
11. Ding S, Chen X, Shen K. Single-cell RNA sequencing
in breast cancer: understanding tumor heterogeneity and paving roads to individualized therapy. Cancer Commun (Lond). 2020; 40:329–44.
12. Tieng FYF, Baharudin R, Abu N, Mohd Yunos RI, Lee LH, Ab Mutalib NS. Single cell transcriptome in colorectal cancer-current updates on its application in metastasis, chemoresistance and the roles of circulating tumor cells. Front Pharmacol. 2020; 11:135.
13. Kim K, Park S, Park SY, Kim G, Park SM, Cho JW, et al. Single-cell transcriptome analysis reveals TOX as a promoting factor for T cell exhaustion and a predictor for anti-PD-1 responses in human cancer. Genome Med. 2020; 12:22.
14. Li C, Xue VW, Wang QM, Lian GY, Huang XR, Lee TL, et al. The mincle/Syk/NF-κB signaling circuit is essential for maintaining the protumoral activities of tumor-associated macrophages. Cancer Immunol Res. 2020; 8:1004–17.
15. Ireland AS, Micinski AM, Kastner DW, Guo B, Wait SJ, Spainhower KB, et al. MYC drives temporal evolution of small cell lung cancer subtypes by reprogramming neuroendocrine fate. Cancer Cell. 2020; 38:60–78.e12.
16. Gupta RK, Kuznicki J. Biological and medical importance of cellular heterogeneity deciphered by single-cell. Cells. 2020; 9:1751.
17. Picelli S. Single-cell RNA-sequencing: the future of genome biology is now. RNA Biol. 2017; 14:637–50.
18. Haque A, Engel J, Teichmann SA, Lönnberg T. A practical guide to single-cell RNA-sequencing for biomedical research and clinical applications. Genome Med. 2017; 9:75.
19. Hwang B, Lee JH, Bang D. Single-cell RNA sequencing
technologies and bioinformatics pipelines. Exp Mol Med. 2018; 50:1–14.
20. Zhang M, Zou Y, Xu X, Zhang X, Gao M, Song J, et al. Highly parallel and efficient single cell mRNA sequencing with paired picoliter chambers. Nat Commun. 2020; 11:2118.
21. Brasko C, Smith K, Molnar C, Farago N, Hegedus L, Balind A, et al. Intelligent image-based in situ single-cell isolation. Nat Commun. 2018; 9:226.
22. Kolodziejczyk AA, Kim JK, Svensson V, Marioni JC, Teichmann SA. The technology and biology of single-cell RNA sequencing
. Mol Cell. 2015; 58:610–20.
23. Hu P, Zhang W, Xin H, Deng G. Single cell isolation and analysis. Front Cell Dev Biol. 2016; 4:116.
24. Klein AM, Mazutis L, Akartuna I, Tallapragada N, Veres A, Li V, et al. Droplet barcoding for single cell transcriptomics applied to embryonic stem cells. Cell. 2016; 161:1187–201.
25. Tang F, Barbacioru C, Wang Y, Nordman E, Lee C, Xu N, et al. mRNA-Seq whole-transcriptome analysis of a single cell. Nat Methods. 2009; 6:377–82.
26. Sasagawa Y, Nikaido I, Hayashi T, Danno H, Uno KD, Imai T, et al. Quartz-Seq: a highly reproducible and sensitive single-cell RNA sequencing
method, reveals non-genetic gene-expression heterogeneity. Genome Biol. 2013; 14:R31.
27. Ramsköld D, Luo S, Wang YC, Li R, Deng Q, Faridani OR, et al. Full-length mRNA-Seq from single-cell levels of RNA and individual circulating tumor cells. Nat Biotechnol. 2012; 30:777–82.
28. Picelli S, Faridani OR, Björklund AK, Winberg G, Sagasser S, Sandberg R. Full-length RNA-seq from single cells using Smart-seq2. Nat Protoc. 2014; 9:171–81.
29. Islam S, Zeisel A, Joost S, La Manno G, Zajac P, Kasper M, et al. Quantitative single-cell RNA-seq with unique molecular identifiers. Nat Methods. 2014; 11:163–6.
30. Macosko EZ, Basu A, Satija R, Nemesh J, Shekhar K, Goldman M, et al. Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets. Cell. 2015; 161:1202–14.
31. Hashimshony T, Wagner F, Sher N, Yanai I. CEL-Seq: single-cell RNA-Seq by multiplexed linear amplification. Cell Rep. 2012; 2:666–73.
32. Jaitin DA, Kenigsberg E, Keren-Shaul H, Elefant N, Paul F, Zaretsky I, et al. Massively parallel single-cell RNA-seq for marker-free decomposition of tissues into cell types. Science. 2014; 343:776–9.
33. Botero LM, D’Imperio S, Burr M, McDermott TR, Young M, Hassett DJ. Poly(A) polymerase modification and reverse transcriptase PCR amplification of environmental RNA. Appl Environ Microbiol. 2005; 71:1267–75.
34. Zhu YY, Machleder EM, Chenchik A, Li R, Siebert PD. Reverse transcriptase template switching: a SMART approach for full-length cDNA library construction. Biotechniques. 2001; 30:892–7.
35. Sena JA, Galotto G, Devitt NP, Connick MC, Jacobi JL, Umale PE, et al. Unique Molecular Identifiers reveal a novel sequencing artefact with implications for RNA-Seq based gene expression analysis. Sci Rep. 2018; 8:13121.
36. Mills JD, Kawahara Y, Janitz M. Strand-specific RNA-Seq provides greater resolution of transcriptome profiling. Curr Genomics. 2013; 14:173–81.
37. Shin D, Lee W, Lee JH, Bang D. Multiplexed single-cell RNA-seq via transient barcoding for simultaneous expression profiling of various drug perturbations. Sci Adv. 2019; 5:eaav2249.
38. Klepikova AV, Kasianov AS, Chesnokov MS, Lazarevich NL, Penin AA, Logacheva M. Effect of method of deduplication on estimation of differential gene expression using RNA-seq. PeerJ. 2017; 5:e3091.
39. Hicks SC, Townes FW, Teng M, Irizarry RA. Missing data and technical variability in single-cell RNA-sequencing experiments. Biostatistics. 2018; 19:562–78.
40. Zhang F, Bai H, Gao R, Fei K, Duan J, Zhang Z, et al. Dynamics of peripheral T cell clones during PD-1 blockade in non-small cell lung cancer. Cancer Immunol Immunother. 2020; 69:2599–611.
41. Song Q, Hawkins GA, Wudel L, Chou PC, Forbes E, Pullikuth AK, et al. Dissecting intratumoral myeloid cell plasticity by single cell RNA-seq. Cancer Med. 2019; 8:3072–85.
42. Gay L, Baker A-M, Graham TA. Tumour cell heterogeneity. F1000 Res. 2016; 5:238.
43. Liu J, Zhou Y, Liu S, Song X, Yang XZ, Fan Y, et al.; DISCO (Deciphering disorders Involving Scoliosis and COmorbidities) Study. The coexistence of copy number variations (CNVs) and single nucleotide polymorphisms (SNPs) at a locus can result in distorted calculations of the significance in associating SNPs to disease. Hum Genet. 2018; 137:553–67.
44. Zhong R, Chen D, Cao S, Li J, Han B, Zhong H. Immune cell infiltration features and related marker genes in lung cancer based on single-cell RNA-seq. Clin Transl Oncol. 2021; 23:405–17.
45. Ma KY, Schonnesen AA, Brock A, Van Den Berg C, Eckhardt SG, Liu Z, et al. Single-cell RNA sequencing
of lung adenocarcinoma reveals heterogeneity of immune response-related genes. JCI insight. 2019; 4:e121387.
46. Lu T, Yang X, Shi Y, Zhao M, Bi G, Liang J, et al. Single-cell transcriptome atlas of lung adenocarcinoma featured with ground glass nodules. Cell Discov. 2020; 6:69.
47. Lambrechts D, Wauters E, Boeckx B, Aibar S, Nittner D, Burton O, et al. Phenotype molding of stromal cells in the lung tumor microenvironment. Nat Med. 2018; 24:1277–89.
48. Li J, Comeau HY, Zhang Z, Ren X. Landscape of transcript isoforms in single T cells infiltrating in non-small-cell lung cancer. J Genet Genomics. 2020; 47:373–88.
49. Clarke J, Panwar B, Madrigal A, Singh D, Gujar R, Wood O, et al. Single-cell transcriptomic analysis of tissue-resident memory T cells in human lung cancer. J Exp Med. 2019; 216:2128–49.
50. Park SM, Wong DJ, Ooi CC, Kurtz DM, Vermesh O, Aalipour A, et al. Molecular profiling of single circulating tumor cells from lung cancer patients. Proc Natl Acad Sci U S A. 2016; 113:E8379–86.
51. Shats I, Deng M, Davidovich A, Zhang C, Kwon JS, Manandhar D, et al. Expression level is a key determinant of E2F1-mediated cell fate. Cell Death Differ. 2017; 24:626–37.
52. Suber TL, Nikolli I, O’Brien ME, Londino J, Zhao J, Chen K, et al. FBXO17 promotes cell proliferation through activation of Akt in lung adenocarcinoma cells. Respir Res. 2018; 19:206.
53. Chen J, Tan Y, Sun F, Hou L, Zhang C, Ge T, et al. Single-cell transcriptome and antigen-immunoglobin analysis reveals the diversity of B cells in non-small cell lung cancer. Genome Biol. 2020; 21:152.
54. Kim N, Kim HK, Lee K, Hong Y, Cho JH, Choi JW, et al. Single-cell RNA sequencing
demonstrates the molecular and cellular reprogramming of metastatic lung adenocarcinoma. Nat Commun. 2020; 11:2285.
55. Min JW, Kim WJ, Han JA, Jung YJ, Kim KT, Park WY, et al. Identification of distinct tumor subpopulations in lung adenocarcinoma via single-cell RNA-seq. PLoS One. 2015; 10:e0135817.
56. Xiong D, Pan J, Yin Y, Jiang H, Szabo E, Lubet RA, et al. Novel mutational landscapes and expression signatures of lung squamous cell carcinoma. Oncotarget. 2018; 9:7424–41.
57. Dost AFM, Moye AL, Vedaie M, Tran LM, Fung E, Heinze D, et al. Organoids model transcriptional hallmarks of oncogenic KRAS activation in lung epithelial progenitor cells. Cell Stem Cell. 2020; 27:663–78.e8.
58. Ruan H, Zhou Y, Shen J, Zhai Y, Xu Y, Pi L, et al. Circulating tumor cell characterization of lung cancer brain metastasis in the cerebrospinal fluid through single-cell transcriptome analysis. Clin Transl Med. 2020; 10:e246.
59. Kim KT, Lee HW, Lee HO, Kim SC, Seo YJ, Chung W, et al. Single-cell mRNA sequencing identifies subclonal heterogeneity in anti-cancer drug responses of lung adenocarcinoma cells. Genome Biol. 2015; 16:127.
60. Suzuki A, Matsushima K, Makinoshima H, Sugano S, Kohno T, Tsuchihara K, et al. Single-cell analysis of lung adenocarcinoma cell lines reveals diverse expression patterns of individual cells invoked by a molecular target drug treatment. Genome Biol. 2015; 16:66.
61. Guo X, Zhang Y, Zheng L, Zheng C, Song J, Zhang Q, et al. Global characterization of T cells in non-small-cell lung cancer by single-cell sequencing. Nat Med. 2018; 24:978–85.
62. Chen J, Wu R, Xuan Y, Jiang M, Zeng Y. Bioinformatics analysis and experimental validation of TTK as a biomarker for prognosis in non-small cell lung cancer. Biosci Rep. 2020; 40:BSR20202711.
63. Garrido-Martin EM, Mellows TWP, Clarke J, Ganesan A-P, Wood O, Cazaly A, et al. M1 hot tumor-associated macrophages boost tissue-resident memory T cells infiltration and survival in human lung cancer. J Immunother Cancer. 2020; 8:e000778.
64. Ghorani E, Reading JL, Henry JY, de Massy MR, Rosenthal R, Turati V, et al. The T cell differentiation landscape is shaped by tumour mutations in lung cancer. Nat Cancer. 2020; 1:546–61.
65. He D, Wang D, Lu P, Yang N, Xue Z, Zhu X, et al. Single-cell RNA sequencing
reveals heterogeneous tumor and immune cell populations in early-stage lung adenocarcinomas harboring EGFR mutations. Oncogene. 2021; 40:355–68.
66. Zilionis R, Engblom C, Pfirschke C, Savova V, Zemmour D, Saatcioglu HD, et al. Single cell transcriptomics of human and mouse lung cancers reveals conserved myeloid populations across individuals and species. Immunity. 2020; 50:1317–34.
67. Goveia J, Rohlenova K, Taverna F, Treps L, Conradi LC, Pircher A, et al. An integrated gene expression landscape profiling approach to identify lung tumor endothelial cell heterogeneity and angiogenic candidates. Cancer Cell. 2020; 37:421.
68. Maynard A, McCoach CE, Rotow JK, Harris L, Haderk F, Kerr DL, et al. Therapy-induced evolution of human lung cancer revealed by single-cell RNA sequencing
. Cell. 2020; 182:1232–51.e22.
69. Balkwill FR, Capasso M, Hagemann T. The tumor microenvironment at a glance. J Cell Sci. 2012; 125(Pt 23):5591–6.
70. Jin MZ, Jin WL. The updated landscape of tumor microenvironment and drug repurposing. Signal Transduct Target Ther. 2020; 5:166.
71. Klein-Goldberg A, Maman S, Witz IP. The role played by the microenvironment in site-specific metastasis. Cancer Lett. 2014; 352:54–8.
72. Engblom C, Pfirschke C, Pittet MJ. The role of myeloid cells in cancer therapies. Nat Rev Cancer. 2016; 16:447–62.
73. Mao Y, Keller ET, Garfield DH, Shen K, Wang J. Stromal cells in tumor microenvironment and breast cancer. Cancer Metastasis Rev. 2013; 32:303–15.
74. Ham S, Lima LG, Lek E, Möller A. The impact of the cancer microenvironment on macrophage phenotypes. Front Immunol. 2020; 11:1308.
75. Hao NB, Lü MH, Fan YH, Cao YL, Zhang ZR, Yang SM. Macrophages in tumor microenvironments and the progression of tumors. Clin Dev Immunol. 2012; 2012:948098.
76. Gjorgjevski M, Hannen R, Carl B, Li Y, Landmann E, Buchholz M, et al. Molecular profiling of the tumor microenvironment in glioblastoma patients: correlation of microglia/macrophage polarization state with metalloprotease expression profiles and survival. Biosci Rep. 2019; 39:BSR20182361.
77. Biswas SK, Lewis CE. NF-κB as a central regulator of macrophage function in tumors. J Leukoc Biol. 2010; 88:877–84.
78. Porta C, Rimoldi M, Raes G, Brys L, Ghezzi P, Di Liberto D, et al. Tolerance and M2 (alternative) macrophage polarization are related processes orchestrated by p50 nuclear factor kappaB. Proc Natl Acad Sci U S A. 2009; 106:14978–83.
79. Yang Y, Qin J, Lan L, Li N, Wang C, He P, et al. M-CSF cooperating with NFκB induces macrophage transformation from M1 to M2 by upregulating c-Jun. Cancer Biol Ther. 2014; 15:99–107.
80. Han IH, Song HO, Ryu JS. IL-6 produced by prostate epithelial cells stimulated with Trichomonas vaginalis promotes proliferation of prostate cancer cells by inducing M2 polarization of THP-1-derived macrophages. Plos Negl Trop Dis. 2020; 14:e0008126.
81. Pauken KE, Wherry EJ. Overcoming T cell exhaustion in infection and cancer. Trends Immunol. 2015; 36:265–76.
82. Zhang Q, Bi J, Zheng X, Chen Y, Wang H, Wu W, et al. Blockade of the checkpoint receptor TIGIT prevents NK cell exhaustion and elicits potent anti-tumor immunity. Nat Immunol. 2018; 19:723–32.
83. Brunner-Weinzierl MC, Rudd CE. CTLA-4 and PD-1 control of T-cell motility and migration: implications for tumor immunotherapy. Front Immunol. 2018; 9:2737.
84. Ando M, Ito M, Srirat T, Kondo T, Yoshimura A. Memory T cell, exhaustion, and tumor immunity. Immunol Med. 2020; 43:1–9.
85. Crispin JC, Tsokos GC. Cancer immunosurveillance by CD8 T cells. F1000Research. 2020; 9:F1000 Faculty Rev-80.
86. Sakuishi K, Apetoh L, Sullivan JM, Blazar BR, Kuchroo VK, Anderson AC. Targeting Tim-3 and PD-1 pathways to reverse T cell exhaustion and restore anti-tumor immunity. J Exp Med. 2010; 207:2187–94.
87. Hu X, Zhang J, Wang J, Fu J, Li T, Zheng X, et al. Landscape of B cell immunity and related immune evasion in human cancers. Nat Genet. 2019; 51:560–7.
88. Griss J, Bauer W, Wagner C, Simon M, Chen M, Grabmeier-Pfistershammer K, et al. B cells sustain inflammation and predict response to immune checkpoint blockade in human melanoma. Nat Commun. 2019; 10:4186.
89. Lei X, Zhong Y, Huang L, Li S, Fu J, Zhang L, et al. Identification of a novel tumor angiogenesis inhibitor targeting Shh/Gli1 signaling pathway in non-small cell lung cancer. Cell Death Dis. 2020; 11:232.
90. Sun Z, Wang CY, Lawson DA, Kwek S, Velozo HG, Owyong M, et al. Single-cell RNA sequencing
reveals gene expression signatures of breast cancer-associated endothelial cells. Oncotarget. 2018; 9:10945–61.
91. Davidson S, Efremova M, Riedel A, Mahata B, Pramanik J, Huuhtanen J, et al. Single-cell RNA sequencing
reveals a dynamic stromal niche that supports tumor growth. Cell Rep. 2020; 31:107628.
92. Lee HW, Chung W, Lee HO, Jeong DE, Jo A, Lim JE, et al. Single-cell RNA sequencing
reveals the tumor microenvironment and facilitates strategic choices to circumvent treatment failure in a chemorefractory bladder cancer patient. Genome Med. 2020; 12:47.
93. Schouppe E, De Baetselier P, Van Ginderachter JA, Sarukhan A. Instruction of myeloid cells by the tumor microenvironment: open questions on the dynamics and plasticity of different tumor-associated myeloid cell populations. Oncoimmunology. 2012; 1:1135–45.
94. Gelse K, Pöschl E, Aigner T. Collagens–structure, function, and biosynthesis. Adv Drug Deliv Rev. 2003; 55:1531–46.
95. Sykes M. Immune tolerance: mechanisms and application in clinical transplantation. J Intern Med. 2007; 262:288–310.
96. Makkouk A, Weiner GJ. Cancer immunotherapy and breaking immune tolerance: new approaches to an old challenge. Cancer Res. 2015; 75:5–10.
97. Patel AJ, Richter A, Drayson MT, Middleton GW. The role of B lymphocytes in the immuno-biology of non-small-cell lung cancer. Cancer Immunol Immunother. 2020; 69:325–42.
98. Sandberg TP, Stuart MPME, Oosting J, Tollenaar RAEM, Sier CFM, Mesker WE. Increased expression of cancer-associated fibroblast markers at the invasive front and its association with tumor-stroma ratio in colorectal cancer. BMC Cancer. 2019; 19:284.
99. Alizadeh AA, Aranda V, Bardelli A, Blanpain C, Bock C, Borowski C, et al. Toward understanding and exploiting tumor heterogeneity. Nat Med. 2015; 21:846–53.
100. Ivashkiv LB. IFNγ: signalling, epigenetics and roles in immunity, metabolism, disease and cancer immunotherapy. Nat Rev Immunol. 2019; 18:545–58.
101. Pelosi G, Sonzogni A, Harari S, Albini A, Bresaola E, Marchiò C, et al. Classification of pulmonary neuroendocrine tumors: new insights. Transl Lung Cancer Res. 2017; 6:513–29.
102. Dang CV. MYC on the path to cancer. Cell. 2012; 149:22–35.
103. Nakamura S, Fukui T, Kawaguchi K, Fukumoto K, Hirakawa A, Yokoi K. Does ground glass opacity-dominant feature have a prognostic significance even in clinical T2aN0M0 lung adenocarcinoma? Lung Cancer. 2015; 89:38–42.
104. Shi X, Dong X, Young S, Chen AM, Liu X, Zheng Z, et al. The impact of angiogenesis inhibitors on survival of patients with small cell lung cancer. Cancer Med. 2019; 8:5930–8.
105. King JL, Zhang B, Li Y, Li KP, Ni JJ, Saavedra HI, et al. TTK promotes mesenchymal signaling via multiple mechanisms in triple negative breast cancer. Oncogenesis. 2018; 7:69.
106. Shi Q, Wang W, Jia Z, Chen P, Ma K, Zhou C. ISL1, a novel regulator of CCNB1, CCNB2 and c-MYC genes, promotes gastric cancer cell proliferation and tumor growth. Oncotarget. 2016; 7:36489–500.
107. Lim N, Townsend PA. Cdc6 as a novel target in cancer: oncogenic potential, senescence and subcellular localisation. Int J Cancer. 2020; 147:1528–34.
108. Shafer MER. Cross-species analysis of single-cell transcriptomic data. Front Cell Dev Biol. 2019; 7:175.
109. Liu FH, Cui YP, He YK, Shu RH. FBXO17 promotes malignant progression of hepatocellular carcinoma by activating wnt/β-catenin pathway. Eur Rev Med Pharmacol Sci. 2019; 23:8265–73.
110. Lundin A, Driscoll B. Lung cancer stem cells: progress and prospects. Cancer Lett. 2013; 338:89–93.
111. Denechaud PD, Fajas L, Giralt A. E2F1, a novel regulator of metabolism. Front Endocrinol (Lausanne). 2017; 8:311.
112. Cheng H, Shcherba M, Pendurti G, Liang Y, Piperdi B, Perez-Soler R. Targeting the PI3K/AKT/mTOR pathway: potential for lung cancer treatment. Lung Cancer Manag. 2014; 3:67–75.
113. Tang YA, Tsai YT, Lin RK, Hsu HS, Chen CY, Wang YC. Deregulation of p53 and RB transcriptinal control leads to overexpression of DNA methyltransferases in lung cancer. J Cancer Res Pract. 2014; 1:14–27.
114. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011; 144:646–74.
115. Brinton LT, Brentnall TA, Smith JA, Kelly KA. Metastatic biomarker discovery through proteomics. Cancer Genomics Proteomics. 2012; 9:345–55.
116. Pandey R, Zhou M, Islam S, Chen B, Barker NK, Langlais P, et al. Carcinoembryonic antigen cell adhesion molecule 6 (CEACAM6) in pancreatic ductal adenocarcinoma (PDA): an integrative analysis of a novel therapeutic target. Sci Rep. 2019; 9:18347.
117. Kurotani R, Kumaki N, Naizhen X, Ward JM, Linnoila RI, Kimura S. Secretoglobin 3A2/uteroglobin-related protein 1 is a novel marker for pulmonary carcinoma in mice and humans. Lung Cancer. 2011; 71:42–8.
118. Yuan X, Larsson C, Xu D. Mechanisms underlying the activation of TERT transcription and telomerase activity in human cancer: old actors and new players. Oncogene. 2019; 38:6172–83.
119. Owusu BY, Galemmo R, Janetka J, Klampfer L. Hepatocyte growth factor, a key tumor-promoting factor in the tumor microenvironment. Cancers (Basel). 2017; 9:35.
120. Zhang Y, Feng S, Nie K, Li Y, Gao Y, Gan R, et al. TREM2 modulates microglia phenotypes in the neuroinflammation of Parkinson’s disease. Biochem Biophys Res Commun. 2018; 499:797–802.
121. Lim YJ, Lee J, Choi JA, Cho SN, Son SH, Kwon SJ, et al. M1 macrophage dependent-p53 regulates the intracellular survival of mycobacteria. Apoptosis. 2020; 25:42–55.
122. Kratochvill F, Neale G, Haverkamp JM, Van de Velde LA, Smith AM, Kawauchi D, et al. TNF counterbalances the emergence of M2 tumor macrophages. Cell Rep. 2015; 12:1902–14.
123. Zargar A, Chang S, Kothari A, Snijders AM, Mao JH, Wang J, et al. Overcoming the challenges of cancer drug resistance through bacterial-mediated therapy. Chronic Dis Transl Med. 2019; 5:258–66.
124. Ni L, Lu J. Interferon gamma in cancer immunotherapy. Cancer Med. 2018; 7:4509–16.
125. Hamarsheh S, Osswald L, Saller BS, Unger S, De Feo D, Vinnakota JM, et al. Oncogenic KrasG12D causes myeloproliferation via NLRP3 inflammasome activation. Nat Commun. 2020; 11:1659.
126. Venkateswaran N, Conacci-Sorrell M. Kynurenine: an oncometabolite in colon cancer. Cell Stress. 2020; 4:24–6.
127. Kumari S, Malla R. New insight on the role of plasminogen receptor in cancer progression. Cancer Growth Metastasis. 2015; 8:35–42.
128. Wang Q, Guldner IH, Golomb SM, Sun L, Harris JA, Lu X, et al. Single-cell profiling guided combinatorial immunotherapy for fast-evolving CDK4/6 inhibitor-resistant HER2-positive breast cancer. Nat Commun. 2019; 10:3817.
129. Brave SR, Odedra R, James NH, Smith NR, Marshall GB, Acheson KL, et al. Vandetanib inhibits both VEGFR-2 and EGFR signalling at clinically relevant drug levels in preclinical models of human cancer. Int J Oncol. 2011; 39:271–8.
130. Rosenzweig SA. Acquired resistance to drugs targeting tyrosine kinases. Adv Cancer Res. 2018; 138:71–98.
131. Cipponi A, Goode DL, Bedo J, McCabe MJ, Pajic M, Croucher DR, et al. MTOR signaling orchestrates stress-induced mutagenesis, facilitating adaptive evolution in cancer. Science. 2020; 368:1127–31.
132. Patwa A, Shah A. Anatomy and physiology of respiratory system relevant to anaesthesia. Indian J Anaesth. 2015; 59:533–41.
133. Cheng YH, Chen YC, Lin E, Brien R, Jung S, Chen YT, et al. Hydro-Seq enables contamination-free high-throughput single-cell RNA-sequencing for circulating tumor cells. Nat Commun. 2019; 10:2163.
134. Nguyen QH, Pervolarakis N, Nee K, Kessenbrock K. Experimental considerations for single-cell RNA sequencing
approaches. Front Cell Dev Biol. 2018; 6:108.
135. Choi YH, Kim JK. Dissecting cellular heterogeneity using single-cell RNA sequencing
. Mol Cells. 2019; 42:189–99.