Patent Application
20240200147
This invention provides a method of treating a patient having non-small cell lung cancer comprising administering to said patient an effective amount of a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, for treating non-small cell lung cancer in said patient or for producing an immune response and response to treatment selected from a group of 21 National Comprehensive Cancer Network (NCCN)-recommended drugs for treating non-small cell lung cancer in said patient. This invention provides a method of producing an immune response and response to treatment selected from a group of 21 NCCN-recommended drugs for treating non-small cell lung cancer in a patient having non-small cell lung cancer comprising administering to said patient a therapeutically effective amount of a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, for producing an immune response and response to treatment selected from a group of 21 NCCN-recommended drugs for treating non-small cell lung cancer in said patient. In addition, this invention provides a method of diagnosis and screening of lung cancer using a 7-gene assay.
There are currently no effective biomarkers for prognosis and optimal treatment selection to improve non-small cell lung cancer (NSCLC) survival outcomes.
A seven-gene prognostic panel is set forth in U.S. patent application Ser. No. 17/251,359, filed Dec. 11, 2020, and published as US Patent Application Publication Serial No. US 2021-0254173 A1 on Aug. 19, 2021, entitled Predictive 7-Gene Assay and Prognostic Protein Biomarker For Non-Small Cell Lung Cancer, is incorporated by reference herein, for predicting the risk of tumor recurrence and metastasis in non-small cell lung cancer as understood by those persons of ordinary skill in the art.
A seven-gene prognostic and predictive assay for Non-Small cell Lung Cancer is set forth in U.S. patent application Ser. No. 17/906,315, filed Sep. 14, 2022, and published as US Patent Application Publication Serial No. US 2023/0106465 A1 on Apr. 6, 2023, entitled 7-Gene Prognostic and Predictive Assay For Non-Small Cell Lung cancer in Formalin Fixed and Paraffin Embedded Samples, is incorporated by reference herein, for background methodology and ABCC4, CCL19, SLC39A8, CD27, FUT7, ZNF71 and DAG1 and including SEQ ID for gene and protein, as understood by those persons of ordinary skill in the art.
Artificial intelligence (AI) methodology for drug discovery is set forth in Patent Cooperation Treaty (PCT) Patent Application Serial No. PCT/US2022/075136 having an International Filing Date of Aug. 18, 2022 and published as WO 2023/023592 on Feb. 23, 2023, entitled Repurposing Drugs For Non-Small Cell Lung Cancer, is incorporated by reference herein, for background data and methodology as understood by those persons of ordinary skill in the art.
A method of treating a patient having non-small cell lung cancer comprises administering to said patient a therapeutically effective amount of a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, for treating non-small cell lung cancer in said patient. In certain embodiments of this method, the compound is PD-198306 or pharmaceutically acceptable salts thereof. In certain embodiments of this method, the compound is U-0126 or pharmaceutically acceptable salts thereof. In certain embodiments of this method, the compound is ZM-306416 or pharmaceutically acceptable salts thereof. In certain embodiments of this method, the compound is PQ-401 or pharmaceutically acceptable salts thereof.
Another embodiment of this invention provides a method of treating a patient having non-small cell lung cancer comprising administering to said patient a therapeutically effective amount of a composition comprising a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, and a therapeutically acceptable pharmaceutical carrier, for treating non-small cell lung cancer in said patient.
In another embodiment of this invention, a method of producing an immune response and response to treatment selected from a group of 21 NCCN-recommended drugs for treating non-small cell lung cancer in a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, for producing an immune response and response to treatment selected from a group of 21 NCCN-recommended drugs for treating non-small cell lung cancer in said patient. In certain embodiments of this method, the compound is PD-198306. In certain embodiments of this method, the compound is U-0126. In certain embodiments of this method, the compound is U-0126. In certain embodiments of this method, the compound is ZM-306416. In certain embodiments of this method, the compound is PQ-401.
Another embodiment of this invention provides a method for producing an immune response and response to treatment selected from a group of 21 NCCN-recommended drugs for treating non-small cell lung cancer in a patient having non-small cell lung cancer comprising administering to said patient a therapeutically effective amount of a composition comprising a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, and a therapeutically acceptable pharmaceutical carrier, for producing an immune response in said patient.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound that is an MEK1/2 inhibitor selected from the group consisting of PD-198306 and U-0126.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a composition comprising a compound that is a MEK1/2 inhibitor selected from the group consisting of PD-198306 and U-0126, and a therapeutically acceptable pharmaceutical carrier.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound that is a VEGFR inhibitor that is ZM-306416.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a composition comprising a compound that is a VEGFR inhibitor that is ZM-306416, and a therapeutically acceptable pharmaceutical carrier for treating a patient having non-small cell lung cancer.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound that is an IGF-1R inhibitor that is PQ-401.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a composition comprising a compound that is an IGF-1R inhibitor that is PQ-401 and a therapeutically acceptable pharmaceutical carrier for treating a patient having non-small cell lung cancer.
A method of diagnosis and screening of non-small cell lung cancer using protein expression levels of ABCC4, DAG1, and SLC39A8 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using protein expression levels of ABCC4 and DAG1 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using protein expression levels of ABCC4 and SLC39A8 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using protein expression levels of DAG1 and SLC39A8 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using protein expression levels of ABCC4 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using protein expression levels of DAG1 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using protein expression levels of SLC39A8 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using mRNA expression levels of ABCC4, CCL19, CD27, DAG1, FUT7, and ZNF71 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer using mRNA expression levels of the following gene combinations in patient samples.
A method of using expression of CDT1 and/or INCENP, or a combination thereof, to predict sensitivity to afatinib.
A method of using expression of IL4I1, LRP1, FAM156B, DCUN1D4, ESPN, PTPRJ, and/or NMNAT2, or a combination thereof, to predict resistance to afatinib.
A method of using expression of CSNK2A3, DYNC1H1, and/or PSME1, or a combination thereof, to predict sensitivity to alectinib.
A method of using expression of UCP3, CDC45, THSD7A, ANKRD52, EIF2A, FAM156B, KRT8, and/or CFAP44, or a combination thereof, to predict resistance to alectinib.
A method of using expression of MYOF, ABHD10, and/or TMTC4, or a combination thereof, to predict resistance to brigatinib.
A method of using expression of CSNK2A3, or a combination thereof, to predict sensitivity to cabozantinib.
A method of using expression of RNASEL, TMTC4, SMC2, ASB7, CFAP44, IL17RA, ZXDA, and/or THSD7A, or a combination thereof, to predict resistance to cabozantinib.
A method of using expression of ASB7 and/or RAB1A, or a combination thereof, to predict resistance to carboplatin.
A method of using expression of ZNF507, CSNK2A3, and/or WDR17, or a combination thereof, to predict sensitivity to cisplatin.
A method of using expression of ARHGAP12, CDC45, ESPN, EIF2A, IL17RA, MYOF, TMTC4, NMNAT2, PTPRJ, and/or KRT8, or a combination thereof, to predict resistance to cisplatin.
A method of using expression of RFC4, CSNK2A3, MCM2, MDN1, CDT1, and/or INCENP, or a combination thereof, to predict sensitivity to crizotinib.
A method of using expression of LRP1 and/or IL17RA, or a combination thereof, to predict resistance to crizotinib.
A method of using expression of FBLN7, MAP2, CTNNB1, RNF128, CSNK2A3, WDR17, and/or ZNF507, or a combination thereof, to predict sensitivity to dabrafenib.
A method of using expression of FAM156B, SLC39A8, SMC2, KRT8, and/or NMNAT2, or a combination thereof, to predict resistance to dabrafenib.
A method of using expression of CDT1, or a combination thereof, to predict sensitivity to dacomitinib.
A method of using expression of MAP3K7, DCUN1D4, LRP1, THOC5, FAM156B, ANKRD52, and/or THSD7A, or a combination thereof, to predict resistance to dacomitinib.
A method of using expression of RFC4, ABCC5, FBLN7, and/or MAP2, or a combination thereof, to predict sensitivity to docetaxel.
A method of using expression of SLC39A8, IL17RA, KRT8, RAB1A, THSD7A, DCUN1D4, MYOF, and/or NMNAT2, or a combination thereof, to predict resistance to docetaxel.
A method of using expression of CSNK2A3, PSMB6, and/or TGFB2, or a combination thereof, to predict sensitivity to erlotinib.
A method of using expression of FAM156B, ASB7, TMTC4, and/or NMNAT2, or a combination thereof, to predict resistance to erlotinib.
A method of using expression of PSMB3, CDT1, DYNC1H1, and/or ZNF507, or a combination thereof, to predict sensitivity to etoposide.
A method of using expression of ESPN, NMNAT2, ZNF324B, and/or IL17RA, or a combination thereof, to predict resistance to etoposide.
A method of using expression of BRMS1L, CCNA2, CDT1, and/or INCENP, or a combination thereof, to predict sensitivity to gefitinib.
A method of using expression of IL4I1, ABHD10, FAM156B, MAP3K7, DCUN1D4, NMNAT2, PTPRJ, and/or KRT8, or a combination thereof, to predict resistance to gefitinib.
A method of using expression of MDGA1, MDN1, ZNF507, and/or WDR17, or a combination thereof, to predict sensitivity to gemcitabine.
A method of using expression of KRT8, ANKRD52, ASB7, RAB1A, FAM156B, IL17RA, DCUN1D4, IL4I1, LRP1, and/or NMNAT2, or a combination thereof, to predict resistance to gemcitabine.
A method of using expression of MDN1 and/or PSMB6, or a combination thereof, to predict sensitivity to lorlatinib.
A method of using expression of ANKRD52, ESPN, and/or RAB1A, or a combination thereof, to predict resistance to lorlatinib.
A method of using expression of ABCC5, MDGA1, TCF20, MAP2, and/or PSMB6, or a combination thereof, to predict sensitivity to osimertinib.
A method of using expression of ASB7, DCUN1D4, and/or THSD7A, or a combination thereof, to predict resistance to osimertinib.
A method of using expression of INCENP, CDT1, MCM2, and/or TCF20, or a combination thereof, to predict sensitivity to paclitaxel.
A method of using expression of RAB1A, THOC5, ASB7, IL17RA, NMNAT2, KRT8, LRP1, SLC39A8, THSD7A, and/or UCP3, or a combination thereof, to predict resistance to paclitaxel.
A method of using expression of MCM2 and/or CCNA2, or a combination thereof, to predict sensitivity to pemetrexed.
A method of using expression of EIF2A, IL17RA, MYOF, THSD7A, ZXDA, ARHGAP12, and/or LRP1, or a combination thereof, to predict resistance to pemetrexed.
A method of using expression of CTNNB1, or a combination thereof, to predict sensitivity to trametinib.
A method of using expression of RNASEL, IL17RA, ANKRD52, FAM156B, AP2S1, ARHGAP12, SRRM2, UCP3, ZXDA, TMTC4, DCUN1D4, NCOR1, and/or RAB1A, or a combination thereof, to predict resistance to trametinib.
A method of using expression of RNASE L and/or IL17RA , or a combination thereof, to predict resistance to vemurafenib.
A method of using expression of ABCC5, DYNC1H1, IRX1, and/or TCF20, or a combination thereof, to predict sensitivity to vinorelbine.
A method of using expression of RAB1A, SLC39A8, KRT8, DCUN1D4, PTPRJ, IL17RA, MYOF, and/or NMNAT2, or a combination thereof, to predict resistance to vinorelbine.
This invention provides a method of treating a patient having non-small cell lung cancer comprising administering to said patient a therapeutically effective amount of a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, for treating non-small cell lung cancer in said patient. In certain embodiments of this method, the compound is PD-198306 or pharmaceutically acceptable salts thereof. In certain embodiments of this method, the compound is U-0126 or pharmaceutically acceptable salts thereof. In certain embodiments of this method, the compound is ZM-306416 or pharmaceutically acceptable salts thereof. In certain embodiments of this method, the compound is PQ-401 or pharmaceutically acceptable salts thereof.
Another embodiment of this invention provides a method of treating a patient having non-small cell lung cancer comprising administering to said patient a therapeutically effective amount of a composition comprising a compound selected from the group consisting of PD-198306 or pharmaceutically acceptable salts thereof, U-0126 or pharmaceutically acceptable salts thereof, ZM-306416 or pharmaceutically acceptable salts thereof, and PQ-401 or pharmaceutically acceptable salts thereof, and a therapeutically acceptable pharmaceutical carrier, for treating non-small cell lung cancer in said patient.
In another embodiment of this invention, a method of producing an immune response in a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, for producing an immune response in said patient. In certain embodiments of this method, the compound is PD-198306. In certain embodiments of this method, the compound is U-0126. In certain embodiments of this method, the compound is U-0126. In certain embodiments of this method, the compound is ZM-306416. In certain embodiments of this method, the compound is PQ-401.
Another embodiment of this invention provides a method for producing an immune response in a patient having non-small cell lung cancer comprising administering to said patient a therapeutically effective amount of a composition comprising a compound selected from the group consisting of PD-198306, U-0126, ZM-306416, and PQ-401, and a therapeutically acceptable pharmaceutical carrier, for producing an immune response in said patient.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound that is a MEK1/2 inhibitor selected from the group consisting of PD-198306 and U-0126.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a composition comprising a compound that is a MEK1/2 inhibitor selected from the group consisting of PD-198306 and U-0126, and a therapeutically acceptable pharmaceutical carrier.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound that is a VEGFR inhibitor that is ZM-306416.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a composition comprising a compound that is a VEGFR inhibitor that is ZM-306416, and a therapeutically acceptable pharmaceutical carrier for treating a patient having non-small cell lung cancer.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a compound that is a IGF-1R inhibitor that is PQ-401.
A method of treating a patient having non-small cell lung cancer is provided comprising administering to said patient a therapeutically effective amount of a composition comprising a compound that is a IGF-1R inhibitor that is PQ-401 and a therapeutically acceptable pharmaceutical carrier for treating a patient having non-small cell lung cancer.
The methods of this invention include the following compounds:
Compound ZM-306416 has the formula:
Compound PQ-401 has the formula:
Compound PD-198306 has the formula:
This invention provides seven-gene panel for diagnosis and prognosis of NSCLC using RNA sequencing and proteomic profiles of patient tumors. Within the seven-gene panel, ZNF71 expression combined with dendritic cell activities defines NSCLC patient subgroups (n=966) with distinct survival outcomes (p=0.04, Kaplan-Meier analysis). ZNF71 expression is significantly associated with the activities of natural killer cells (p=0.014) and natural killer T cells (p=0.003) in NSCLC patient tumors (n=1016) using Chi-squared tests. Overexpression of ZNF71 results in decreased expression of multiple components of the intracellular intrinsic and innate immune systems, including dsRNA and dsDNA sensors. Multi-omics networks of ZNF71 and the intracellular intrinsic and innate immune systems were computed as relevant to NSCLC tumorigenesis, proliferation, and survival using patient clinical information and in-vitro CRISPR-Cas9/RNAi screening data. From these networks, pan-sensitive and pan-resistant genes to 21 NCCN-recommended drugs for treating NSCLC were selected. Based on the gene associations with patient survival and in-vitro CRISPR-Cas9, RNAi, and drug screening data, MEK1/2 inhibitors PD-198306 and U-0126, VEGFR inhibitor ZM-306416, and IGF-1R inhibitor PQ-401 were discovered as targeted therapy that may also induce immune response for treating NSCLC.
As used herein, the term “patient” means members of the animal kingdom, including but not limited to, human beings.
As used herein, the term “effective amount” or “therapeutically effective amount” refers to that amount of any of the present compounds, salts thereof, and/or compositions required to bring about a desired effect in a patient. The desired effect will vary depending upon the illness or disease state being treated. For example, the desired effect may be reducing the tumor size, destroying cancerous cells, and/or preventing metastasis, any one of which may be the desired therapeutic response. On its most basic level, a therapeutically effective amount is that amount of a substance needed to inhibit mitosis of a cancerous cell. As used herein, “tumor” refers to an abnormal growth of cells or tissues of the malignant type, unless otherwise specifically indicated and does not include a benign type tissue. The “tumor” may be comprised of at least one cell and/or tissue. The term “inhibits or inhibiting” as used herein means reducing growth/replication. As used herein, the term “cancer” refers to any type of cancer, including for example but not limited to, non-small cell lung cancer, and the like.
The methods and novel compounds and pharmaceutically acceptable salts thereof of this invention provide for treatment of tumors, or other cancer cells, in cancer patients. The types of cancer can vary widely and in certain embodiments, the methods and novel compounds and pharmaceutically acceptable salts thereof of this invention are particularly useful for example, in treating patients with non-small cell lung cancer.
As used herein, the term “therapeutically effective pharmaceutical carrier” refers to any pharmaceutically acceptable carrier known in the art, absent compatibility problems with the novel compounds of the invention. Generally, therapeutically effective pharmaceutical carriers include for example but not limited to, physiologic saline and 5% dextrose in water.
As will be understood by one skilled in the art, a therapeutically effective amount of said compound can be administered by any means known in the art, including but not limited to, injection, parenterally, intravenously, intraperitoneally, orally or, where appropriate, topically.
It is well within the skill of one practicing in the art to determine what dosage, and the frequency of this dosage, which will constitute a therapeutically effective amount for each individual patient, depending on the severity or progression of cancer or cancer cells and/or the type of cancer. It is also within the skill of one practicing in the art to select the most appropriate method of administering the compounds based upon the needs of each patient.
Non-small cell lung cancer (NSCLC) has the second-highest cancer incidence rate and the highest cancer mortality rate for both men and women [1]. The major histological subtypes of NSCLC are lung adenocarcinoma (LUAD, 40% of NSCLC cases), squamous cell carcinoma (LUSC, 25-30%), and large cell carcinoma (LCC, 5-10%). Each subtype represents a distinct prognosis for patients and informs treatment options [2, 3]. According to the current NCCN standard of care [4], stage 1A NSCLC patients do not receive adjuvant therapy after surgery. Osimertinib is recommended for stage 1B patients with EGFR exon 19 deletion or L858R. Adjuvant therapy is recommended for patients in stage 1B with high-risk features, i.e., tumor size>4 cm, poor differentiation, vascular invasion, wedge resection, visceral pleural involvement, and unknown lymph node status, stage 2, and above. Patients in stages 3 and 4 receive additional radiotherapy [5]. Programmed Death 1 (PD-1) and its ligand PD-L1 compromise anti-tumor immunity while maintaining peripheral tolerance [6]. Anti-PD-L1 antibodies have revolutionized cancer immunotherapy, including NSCLC treatment [7]. NCCN guidelines [4] recently changed to establish adjuvant anti-PD-L1 immunotherapy (atezolizumab) after chemotherapy as the standard of care for stages 2/3A NSCLC patients with PD-L1>1%, following the FDA approval [8]. Yet, resectable NSCLC has a 5-year mortality rate of 40% in stage 1, 66% in stage 2, and 85% in stage 3A because of recurrence [9-13]. At present, there are no accurate prognostic tests that predict post-surgical recurrence/metastasis or inform the clinical benefits of adjuvant therapies, including chemotherapy and immunotherapy, in early-stage NSCLC patients—a significant unmet clinical need.
We recently discovered a seven-gene (ABCC4, CCL19, CD27, DAG1, FUT7, SLC39A8, and ZNF71) signature that accurately predicts the risk of recurrence/metastasis in retrospective analyses of 1,500 early-stage NSCLC patients for all histological subtypes, including clinical trials [14, 15]. Employing novel artificial intelligence (AI) methods and confirmed with qRT-PCR using frozen tumors (n=331) [14], our assay also predicts the clinical benefits of receiving adjuvant chemotherapy in both training and validation sets, including a clinical trial JBR.10. Results from our seven-gene panel were corroborated in The Cancer Genome Atlas (TCGA) cohort for risk stratification of early-stage NSCLC patients (n=923) [15]. Within this seven-gene panel, CD27 is a target for immune checkpoint inhibitors (ICIs) [16], and anti-CD27 mAb is being tested as an adjuvant immunotherapy in phase I/II clinical trials for multiple tumor types with promising results [17, 18]. Within the seven-gene panel, ZNF71 protein expression quantified with AQUA produced robust patient stratification in two separate NSCLC cohorts (n=191) in tissue microarrays [14]. We previously reported that the ZNF71 KRAB isoform was associated with epithelial-to-mesenchymal (EMT) transition and poor prognosis in NSCLC patients [19].
ZNF71 is a member of a large family of KRAB zinc finger transcription factors, KRAB-ZNFs, which due to the presence of the KRAB domain function as transcriptional repressors. One of the main roles ascribed to KRAB-ZNFs is the repression of retrotransposon class repetitive elements (TEs) [20] that comprise up to 36% of the human genome [21]. Retro TEs are remnants of ancient invaded viruses, which could produce dsRNA molecules and RNA/DNA hybrids. Although the vast majority of TEs in the human genome are inactivated by mutations, a small number of full-length functional elements including long interspersed nuclear elements (LINEs) and human endogenous retroviruses (HERVs) are capable of retrotransposition, mimicking viral infection. While normally silenced, they could be reactivated in cancer or response to therapy [22, 23]. The majority of cellular dsRNA is the result of the TEs transcription [24]. Non-degraded dsRNAs are recognized by specific proteins of the intracellular innate immune system also called pattern recognition receptors (PRRs) such as MDA5/IFIH1 and others, ultimately leading to a Type I interferon production [25]. Acting as specific dsRNA sensors downstream or in parallel with PRRs, the OAS-RNase L and the PKR/EIF2AK2 pathways degrade endogenous and viral dsRNA and block global cellular translation, respectively [26, 27]. In addition, a growing number of host restriction factors of the intrinsic immune system can be engaged in anti-viral response, including SAMHD1, TRIM5a, MX, and IFITM proteins [28].
Most human tumors display chromosomal instability (CIN) phenotype and aneuploidy, which are often accompanied by the generation of micronuclei and the presence of cytosolic dsDNA. Cytosolic dsDNA activates the cGAS-STING signaling pathway [29]. STING facilitates activation of TBK1 kinase leading to its autophosphorylation on S172 [30] and subsequent phosphorylation of IRF3 transcription factor, which in turn activates interferon response genes.
In this study, we sought to (1) further evaluate the diagnostic and prognostic implications of the seven-gene panel using both RNA-sequencing and proteomic profiles in diverse NSCLC patient cohorts and examine the associated immune cell activities during NSCLC tumorigenesis and progression; (2) investigate the functional involvement of ZNF71 KRAB in innate immunity; (3) identify molecular networks mediated by ZNF71 relevant to innate immunity in NSCLC tumors and normal adjacent lung tissues using a novel AI technology based on Boolean implication networks; (4) discover pan-sensitive and pan-resistant genes to a panel of 21 NCCN recommended drugs for treating NSCLC from the above identified molecular association networks; and (5) explore therapeutic compounds as new or repositioning drugs for treating NSCLC for designed mechanisms of actions based on our analysis of patient tumor profiles and in-vitro CRISPR-Cas9/RNAi and drug screening data using Connectivity Map (CMap) [31, 32].
Our previous work [14] developed a prognostic and predictive seven-gene assay including ABCC4, CCL19, CD27, DAG1, FUT7, SLC39A8, and ZNF71 for early-stage NSCLC. In this study, we further explored the prognostic capacity of the seven-gene signature using proteomic profiles in a Chinese LUAD cohort from Xu et al [33] (n=103) and TCGA-NSCLC datasets (n=923, TCGA-LUAD and TCGA-LUSC combined) patient samples with sufficient survival information. Immune cell-type activities associated with different prognostic patient groups were investigated.
ABCC4, CCL19, CD27, DAG1, and SLC39A8 from the seven-gene assay were available in log10 transformed proteomics data of Xu's LUAD cohort [33]. A multivariate Cox model was built based on these five genes to calculate the coefficients for the risk score. A stepwise model selection that dropped the least significant variable in each iteration was used to reach an optimal model. The final risk-score equation was shown in
The Kaplan-Meier analysis results also showed significant stratifications for each of these five genes (ABCC4, CCL19, CD27, DAG1, and SLC39A8) in RNA-sequencing/proteomic profiling in Xu's LUAD [33] or TCGA-NSCLC patient cohorts. Patients with a higher expression of ABCC4 (cutoff=10.45) in TCGA-NSCLC RNA-sequencing data survived significantly longer than those with a lower expression of ABCC4 (
xCell scores were computed for each patient sample in TCGA-NSCLC and Xu's LUAD [33] cohorts with the corresponding RNA sequencing data. For each significant patient stratification in survival analysis, immune cell types with a significant difference in activities (two-sample t-tests; p<0.05) between high-risk and low-risk patient tumors were identified. The log2 ratio of xCell scores between high-risk vs. low-risk tumors was shown in
The log2 ratios of significantly different xCell scores between Xu's LUAD tumors and their paired non-cancerous adjacent tissues (NATs) [33] were also shown in
In
We examined the potential of using the seven-gene panel to separate NSCLC tumors from NATs. Within the seven-gene panel, there were six genes (ABCC4, CD27, DAG1, FUT7, CCL19, and ZNF71) available in the RNA sequencing data of Xu's LUAD cohort [33] (
To evaluate the accuracy of using the seven-gene panel in classifying tumors from NATs in Xu's LUAD cohort [33], we applied seven commonly used machine-learning classification algorithms. These algorithms included decision tree, k-nearest neighbor (KNN), logistic regression, Naïve Bayes, random forests, Support Vector Machine (SVM), and Radial Basis Function (RBF) network. Classification methods were performed in Weka with 10-fold cross-validation. In LUAD RNA sequencing data [33], six genes (ABCC4, CCL19, CD27, DAG1, FUT7, and ZNF71) with available mRNA expression data were used in classification. Random forests and RBF networks had the highest overall classification accuracy of 0.86. The random forests classification had a sensitivity of 0.882, a specificity of 0.837, a ROC of 0.927, and an odds ratio of 38.44 in the 10-fold cross-validation of tumors vs. NATs (n=51). The RBF network had a sensitivity of 0.824, a specificity of 0.898, a ROC of 0.896, and an odds ratio of 41.07 in the 10-fold cross-validation of tumors vs. NATs. Three genes (ABCC4, DAG1, and SLC39A8) with available protein expression data were used in the classification. SVM had the highest overall classification accuracy of 0.941, with a sensitivity of 0.867, a specificity of 0.978, a ROC of 0.922, and an odds ratio of 286 in classifying LUAD tumors from NATs (n=103). Overall, the 7-gene panel generated high accuracy in separating tumors from NATs using RNA-sequencing or proteomic profiles, indicating its diagnostic implications in NSCLC. Detailed information on the classification results was included in Table 4 and Table 5.
ZNF71 protein expression was not available in Xu's LUAD cohort [33]. In our previous study, ZNF71 protein expression quantified with AQUA was associated with good prognosis in NSCLC patients (n=191) in tissue microarrays [14]. Although ZNF71 overall mRNA expression was not associated with NSCLC patient survival outcomes, ZNF71 KRAB, the transcriptional repression isoform, was an independent poor prognostic factor in early-stage NSCLC [19]. Furthermore, ZNF71 KRAB was associated with EMT in NSCLC patient tumors (n=197) and epithelial cell lines (n=117). ZNF71 protein expression positively correlated with epithelial markers E-cadherin and Cytokeratin and negatively correlated with mesenchymal markers ZEB1 and Vimentin in Western blots [15], consistent with the association between ZNF71 protein expression and favorable patient prognosis [19]. Based on the structurally relevant function of KRAB-ZNFs, we hypothesized that ZNF71 could be involved in the suppression of endogenous transposable elements (TEs) expression, which is often activated in cancer and can trigger an innate immune response.
Next, we tested if ZNF71 expression and specific immune cell activities have any associations with NSCLC prognosis. The Kaplan-Meier analysis results showed that if we used the median value of ZNF71 mRNA expression as the cutoff to stratify TCGA-NSCLC, the low-expression group and high-expression group did not have a significant difference in the survival outcomes (
To investigate the potential function of ZNF71 in intracellular immune response, we overexpressed its KRAB and KRABless isoforms in lung adenocarcinoma A549 cells (
To explore gene interactions and pathways among ZNF71 and the intracellular intrinsic and innate immune systems, multi-omics association networks involving ZNF71 and genes examined in Western blot (
The computed direct gene associations between ZNF71 and IIIR genes (
Second, we also identified indirect gene association networks between ZNF71 and IIIR genes using RNA sequencing data from Xu et al [33]. ZNF71 had co-expression associations with IIIR genes (p<0.05, z tests) through some intermediate genes. To form a manageable size of networks, we filtered the intermediate genes with the following criteria: (1) the gene was differentially expressed between LUAD tumors vs. NATs (p<0.05, two-sample t-tests); (2) the gene was a proliferation gene that had a significant effect (dependency score<−0.5) on 50% or more NSCLC cell lines in RNAi or CRISPR-Cas9 screening data; (3) the gene was a prognostic gene that can significantly stratify the patient survival in RNA sequencing data of both TCGA-LUAD and Xu's LUAD [33] patient cohorts. The indirect gene association networks with intermediate genes that meet all three criteria were found when ZNF71 was up-regulated in LUAD tumors (
Third, although ZNF71 was not available in LUAD proteomics data from Xu et al [33], ZNF71 mRNA expression had indirect associations with the protein expression of some IIIR genes through intermediate genes. The indirect gene association networks of ZNF71 (mRNA expression)->intermediate genes (mRNA expression)->IIIR genes (protein expression) were found in LUAD tumors when ZNF71 was up-regulated (
We examined the pathways in ZNF71 co-expression networks with ToppGene. Genome-scale ZNF71 co-expression networks containing all the genes with a significant association (p<0.05, z tests) with ZNF71 were constructed with the Boolean implication networks using RNA sequencing data in LUAD tumors and NATs from Xu et al [33], respectively. For each disease state, the significantly (p<0.05) enriched pathways of the gene co-expression networks when ZNF71 was upregulated or downregulated were compared. There were 30 common pathways between the networks when ZNF71 was upregulated or downregulated in NATs, focusing on DNA repair, RTK signaling, and transcriptional regulation. There were no common pathways between the networks when ZNF71 was upregulated or downregulated in LUAD tumors. Of the 30 common pathways in NATs, generic transcription pathway (ToppGene ID: 1269650) and gene expression (ToppGene ID: 1269649) were present in the pathways when ZNF71 was upregulated in LAUD tumors; whereas membrane trafficking (ToppGene ID: 1269877) and vesicle-mediated transport (ToppGene ID: 1269876) showed up in the pathways associated with ZNF71 downregulated network in LUAD tumors. Once we focused the gene co-expression networks on ZNF71 and IIIR genes when ZNF71 was upregulated or downregulated, 21 common pathways relevant to immune response were found in LUAD tumors; two common pathways, gene expression and generic transcription pathway, were found in NATs. Detailed significantly (p<0.05) enriched pathways associated with ZNF71 genome-wide co-expression networks, indirect networks of ZNF71 and IIIR genes without any filtering criteria applied, and networks shown in
A total of 21 NCCN-recommended drugs for systemic or targeted therapy for treating NSCLC were available in the Cancer Cell Line Encyclopedia (CCLE) drug screening data. We sought to identify pan-sensitive and pan-resistant genes to these 21 drugs using CCLE RNA sequencing and proteomics profiles in human NSCLC cell lines. The following genes were included in the drug-sensitivity analysis: 1) the seven-gene panel, 2) selected epithelial genes (CDH1, EPCAM, ESRP1, ESRP2, DDR1, CTNNB1, CD24, CLDN7, KRT8, KRT19, and RAB25) and mesenchymal genes (ZEB1, VIM, and FN1), 3) IIIR genes (
To discover functional pathways and therapeutic targets to improve NSCLC treatment outcomes, pan-sensitive (n=23) and pan-resistant (n=31) genes in RNA sequencing data (Table 1) were used as CMap input candidate genes. The up-regulation of pan-sensitive genes and down-regulation of pan-resistant genes are expected to enhance NSCLC treatment response. Thus, the pan-sensitive genes and the pan-resistant genes were used as the initial up- and down-regulated gene lists in the CMap, respectively. The following steps were further applied to the gene lists to inhibit NSCLC proliferation, reverse EMT, and induce immune response: (1) excluding proliferation genes that had a significant effect (dependency score<−0.5) on at least 50% NSCLC cell lines in both CRISPR-Cas9 and RNAi from the up-regulated gene list; (2) excluding survival protective genes (p<0.05, HR<1; univariate cox-model in RNA sequencing data of Xu's LUAD cohort [33] and TCGA-NSCLC data) from the down-regulated gene list; (3) excluding the hazard genes (p<0.05, HR>1; univariate cox-model in RNA sequencing data of Xu's LUAD cohort [33] and TCGA-NSCLC data) from the up-regulated gene list; (4) excluding the mesenchymal genes from the up-regulated gene list; (5) excluding epithelial genes from the down-regulated gene list; (6) adding CD27, PD1 (PDCD1), and PDL1 (CD274) in the downregulated gene list.
With the final up- and downregulated genes (
To investigate if the identified compounds can effectively inhibit the growth of NSCLC cells, the average IC50 and EC50 values in the CCLE NSCLC cell lines (n=135) of the drugs available in PRISM were examined (Table 3). PD-0325901 (
To identify potential oncogenes and tumor suppressor genes among the seven-gene panel, EMT genes, the intracellular innate immune response (IIIR) genes, and all the intermediate genes in previously identified networks (
This study further validated the seven-gene panel in patient recurrence risk stratification using proteomic profiles in early-stage NSCLC patients. In addition, we showed that the seven-gene panel can accurately classify tumors from NATs on both RNA sequencing and proteomic platforms, suggesting its potential diagnostic implications for NSCLC. These findings warrant further clinical studies on liquid biopsies in the early detection of NSCLC. Blood-based assays for predicting NSCLC risk and metastasis are important for thoracic surgeons in clinical decision-making, given that the current accepted benign rate is 5% in surgery and 20-30% in biopsies [40]. Developing minimally invasive biomarker-based assays will reduce unnecessary surgeries and biopsies on patients who do not have lung cancer.
In this invention, we further extended our previous findings on prognostic implication of the seven-gene signature for NSCLC patient survival using proteomics/RNA-Seq data of Xu's LUAD (n=103) cohort and RNA-Seq data of TCGA-NSCLC patient cohort (n=923). From the available expression data, we found that ABCC4 was positively associated with patient survival in both Xu's LUAD protein and TCGA-NSCLC RNA-Seq data. DAG1 was positively associated with patient survival at both mRNA and protein levels, while CCL19 and SLC39A8 showed similar associations at the protein level in Xu's LUAD dataset. At the same time, CD27 mRNA expression was associated with worse patient outcomes. Noteworthy, the combined protein expression score of ABCC4, DAG1, and SLC39A8 efficiently stratified patient outcome (p=0.0013, HR: 8.378 [1.774, 39.57];
Expression of ZNF71 was only available at the mRNA level in the TCGA dataset, and we found only a trend for its negative association with patient survival (
To identify potential new and repositioning medication candidates, we developed mechanisms of action to improve therapy response, extend patient survival, decrease proliferation, and reverse EMT. Therapeutic targets were identified from pan-sensitive genes and pan-resistant genes from the following list: 1) the seven-gene panel, 2) selected epithelial genes and mesenchymal genes, 3) IIIR genes (
Several MEK1/2 inhibitors were selected as potential targeted therapy also with an inhibitory effect on ICIs for treating NSCLC. Selumetinib is an FDA-approved drug to treat neurofibromatosis type 1 with symptomatic, inoperable plexiform neurofibromas. It is also a designated orphan drug for treating thyroid cancer. Selumetinib is being investigated as a secondary therapy for treating late-stage, metastatic, Kras-mutant NSCLC in several trials [41]. Compared with various combination therapies, chemo-, and immune therapy, selumetinib does not have superior efficacy but does have a better safety profile in treating NSCLC [42]. AS-703026 (pimasertib) has a clinical activity of phosphorylated extracellular signal-regulated kinase (pERK) inhibition in peripheral blood mononuclear cells in patients with locally advanced/metastatic melanoma, particularly BRAF- and NRAS-mutated tumors at clinically relevant doses in a phase I study [43]. In a phase I trial of patients with solid tumors, pimasertib inhibited pERK and the recommended phase II dose (RP2D) was defined as 60 mg bid [44]. PD-198306, an orally active inhibitor of MEK1/2, acts as a potent mitochondrial protonophore and uncouples oxidative phosphorylation [45]. PD-198306 is being studied in rabies virus infection [46], neuropsychiatric disorders [47, 48], breast cancer [49], and osteoarthritis [50] research. Resulting from direct inhibition of MEK1 and MEK2, U-0126 inhibits endogenous promoters containing activator protein 1 (AP-1) response elements but does not affect genes lacking an AP-1 response element in their promoters [51]. U-0126 inhibits anchorage-independent growth of Ki-ras-transformed rat fibroblasts by concurrently suppressing both ERK and mammalian target of rapamycin (mTOR)-p70(S6K) pathways, and sensitizes human breast cancer MDA-MB-231 and HBC-4 cells to anoikis [52].
Among other selected compounds with an in-vitro inhibitory effect on ICIs, ZM-306416 is a VEGFR antagonist and a potent inhibitor of EGFR function [53]. As an inhibitor of placental growth factor (PGF) receptor FLT1, ZM-306416 impaired trophoblast proliferation and migration in fetal growth restriction [54]. PQ-401 is an IGF-IR inhibitor that induces apoptosis and inhibits in-vitro viability, proliferation, and mobility of U87MG glioma cells and in-vivo glioma tumor growth in a mouse xenograft model [55]. In a separate study, PQ-401 inhibits osteosarcoma cell proliferation, migration, and colony formation in U2OS and 143B lines [56]. Overall, the above analysis identified several compounds, including PD-198306, U-0126, ZM-306416, and PQ-401, as potential targeted therapy that may also induce immune response for treating NSCLC, which was not known before.
Dasatinib was reported as a potential repositioning drug for treating NSCLC in our previous publication [15]. PD-0325901 was used to treat refractory NSCLC patients but did not meet its primary endpoint in an open-label, phase II study [57]. Combinations of indirubin and arylidene derivatives showed antimetastatic effects on human NSCLC A549 and NCI-H460 cells [58]. Saracatinib, an orally available inhibitor of src kinases, improved progression-free survival in a subset of patients with advanced, platinum-pretreated NSCLC in a phase II clinical trial [59]. BMS-754807 alone reduced cell survival and wound closure and enhanced apoptosis in human NSCLC A549 and NCI-H358 cells, particularly in NSCLC cells expressing high levels of IGF-IR [60]. In addition, BMS-754807 enhanced cisplatin and carboplatin in A549 cells [60]. A combination of trametinib and bosutinib can synergistically suppress the growth of NSCLC by inhibiting both the mitogen-activated protein kinase (MAPK) and proto-oncogene tyrosine-protein kinase (SRC) pathways, suggesting the potential for treating NSCLC, especially in the treatment of erlotinib-resistant NSCLC [61]. Maintenance therapy of adding linsitinib to erlotinib did not improve PFS or OS in non-progressing NSCLC patients in a phase II randomized trial [62]. In a separate phase II study, adding linsitinib to erlotinib resulted in worse patient outcomes compared with erlotinib alone, suggesting that biomarkers are needed to select responding patients [63]. Morphine, a μ-opioid receptor (MOR) agonist, promoted the growth of NSCLC H460 cells both in-vitro and in-vivo, a higher morphine dosage shortens the survival time of patients with lung cancer [64]. Treatment with the Src inhibitor protein phosphatase 1 (PP-1) and the MOR antagonist methylnaltrexone (MNTX) decreased the phosphorylation induced by morphine. Furthermore, the antiapoptotic impact of morphine on NSCLC cells was reversed by MNTX, PP-1, and the PI3K/AKT inhibitor deguelin. Lapatinib (EGFR and HER2 tyrosine kinase inhibitor (TKI)), gefitinib (EGFR TKI), ZD4054 or BQ-123 (ETAR antagonist), GM6001 (matrix metalloprotease inhibitor), PP-2 (Src inhibitor) or Tiron (superoxide scavenger) all inhibited the increase in EGFR and HER2 transactivation induced by the addition of ET-1 to NSCLC cells [65]. These results indicate that our AI pipeline is capable to select relevant compounds for further clinical studies.
Among the identified potential NSCLC oncogenes, a reduction in the germline copy number of EIF4G3 is linked to breast cancer susceptibility in the Japanese population [66]. IkappaB kinase (IKK) promotes tumorigenesis via inhibiting forkhead FOXO3a which can be reversed by FOXO3a [67]. Conditional suppression of IKBKB inhibits melanoma tumor development in mice, and IKBKB-mediated NFkB activity is required in mutant Hras-initiated tumorigenesis [68]. IKBKB also promotes osteosarcoma cancer progression [69]. Minichromosome maintenance proteins (MCMs) are essential in DNA replication, genomic stability, and cell proliferation [70, 71]. MCMs, in particular, MCM2 and MCM4, are potential biomarkers to identify high-risk NSCLC patients [72]. MCM2 was significantly overexpressed in almost all human cancers/subtypes in TCGA and was associated with tumor mutation burden, tumor stage, immune therapy response, immune infiltration, and patient poor prognosis [73]. RFC4 is frequently overexpressed in colorectal cancer (CRC), and RFC4 overexpression is associated with tumor progression and shorter patient survival, possibly due to RFC4-mediated cell cycle arrest and the regulation of CRC cell proliferation [74]. RFC4, along with other genes and microRNAs, might promote osteosarcoma initiation and development [75]. GAPVD1, a cytoplasmic trafficking factor, is involved in the regulation of the mammalian circadian clock [76]. Mutations in GAPVD1 and other genes, estrogen- and growth factor-dependent regulation are involved in both transcriptional and post-transcriptional dysregulation of syndecan-4 in breast cancer [77]. Together, our identified NSCLC oncogene genes are supported by the literature.
Among the identified potential tumor suppressor genes, DAG1 was co-deleted with the Von Hippel Lindau (VHL) tumor suppressor gene in clear cell renal cell carcinoma [78]. In glioblastomas, DAG1 correlated with tumor grade, and the patient group with higher expression of DAG1 survived shorter than the patient group with lower expression of DAG1 [79]. These results suggest that DAG1 may have different functions in tumor initiation and progression in different cancer types. SLC39A8 is responsible for pasting zinc to the cytoplasm when zinc is depleted for maintaining many critical biological processes. SLC39A8 suppresses the progression of clear cell renal cell carcinoma [80]. TMEM173/STING1 was expressed higher in normal samples in lung adenocarcinoma, lung squamous carcinoma, prostate adenocarcinoma, uterine corpus endometrial carcinoma, but was expressed higher in tumor tissues in colorectal carcinoma, kidney renal clear cell carcinoma, stomach adenocarcinoma, and thyroid adenocarcinoma [81]. MTA1-induced inhibition of TJP1 protein co-localized in the cytoplasm and membrane of NSCLC cells leads to weakened cell junctions and changes in adhesion, migration, and invasion capabilities of cells, putatively promoting the invasion and metastasis of NSCLC [82]. β-catenin/CTNNB1 is an intracellular scaffold protein. Aberrant CTNNB1 signaling is one of the fundamental processes in many human cancers [83, 84]. Both gain-of-function and loss-of-function CTNNB1 mutations are found in multiple human cancer types [85]. SMC3 haploinsufficiency accelerates lymphomagenesis in mice with constitutive BCL6 expression and is considered a putative tumor suppressor for germinal center B cells [86]. RDX knockdown increased the intracellular SN-38 concentration, indicating enhanced anti-tumor activity, in human clear cell renal cell carcinoma Caki-1 cells [87]. To date, the literature supports the tumor suppressor functions of DAG1, SLC39A8, TMEM173/STING1, TJP1, CTNNB1, and SMC3, but not RDX.
This study obtained NSCLC patient sample data from public resources including Xu et al [33] and The Cancer Genome Atlas (TCGA Research Network: https://www.cancer.gov/tcga, accessed on 28 Apr. 2021). The primary lung adenocarcinoma (LUAD) cohort collected samples from 103 randomly selected treatment-naïve Chinese patients between 2010 and 2016 from Xu et al [33]. Proteomics data of 103 paired LUAD tumors and non-cancerous adjacent tissues (NATs), and RNA sequencing data of 51 paired LUAD tumors and 49 NATs were used in this study. RNA sequencing data of TCGA NSCLC patient cohort, i.e. TCGA-LUAD (n=515) and TCGA-LUSC (n=501), with patient clinical information were downloaded from an openly accessible entry LinkedOmics (http://www.linkedomics.org, accessed on 28 Apr. 2021) [88].
4.2. xCell
The xCell (https://xcell.ucsf.edu/, accessed on 12 Jul. 2022) tool [89] was used to predict the levels for 64 immune and stroma cell types based on gene expression data. The xCell scores for patient samples were calculated using single-sample gene set enrichment analysis (ssGSEA) to analyze the immune microenvironment. Low xCell scores indicated the cell type had similar levels across all samples; whereas high xCell scores indicated the cell type had different levels across all samples
The Weka software (Version 3.8.6) [90] was utilized to conduct machine learning classifier approaches to differentiate between tumors and NATs with selected genes in Xu's LUAD RNA sequencing and proteomics data [33]. Commonly used machine learning classification methods, including decision tree, k-nearest neighbors (KNN), logistic regression, naïve Bayes, random forests, support vector machine (SVM), and radial basis function (RBF) network, were used. Ten-fold cross-validation was applied in each session.
A549 cells (kind gift of Dr. Ivan Martinez, WVU) were grown in DMEM (Corning, cat. #15-018-CV) supplemented with 10% FBS (HyClone, UT, USA), 2 mM L-glutamine (Corning, cat. #25-005-CI), and 1×Antibiotic Antimycotic Solution (Corning, cat. #30-004-CI). All cells were maintained at 37 C in a 5% CO2 incubator.
ZNF71 KRABless isoform was PCR-amplified from cDNA obtained from MDA231 cells using PfuUltraII DNA polymerase using primers #5ZNF71 ex1-Mun: 5′-AGAGCAATT-GATGGCTGCTCAGCTGC-3′ and #3ZNF71-X: 5′-AGACTCGAGTCAGGTGTGAATCCG-CAG-3′ and cloned into Dox-inducible pLUT lentiviral vector using EcoRI and XhoI cloning sites. To generate the KRAB isoform, a 397 bp long KRAB containing 5′-terminal fragment was synthesized at GenScript and cloned into pLUT-ZNF71-KRABless using NheI and EcoRI sites. The cloned cDNAs were sequenced to verify the absence of mutations.
Lentiviral particles were packaged in HEK293T cells (RRID:CVCL_0063) following calcium phosphate cotransfection of constructed pLUT-ZNF71 vectors, psPAX2 (Addgene, 12260), and pCMV-VSV-G (Addgene, 8454) as previously described [91]. pLUT vector expressing turbo red fluorescent protein (RFP) was used as control. After two rounds of infection, transduced A549 cells were selected in puromycin (1 ug/mL) containing media for at least 5 days Ectopic ZNF71 expression was induced by culturing cells in media containing 0.5 ug/mL doxycycline for 7 days.
Whole-cell lysates were prepared in nonreducing Laemmli buffer as described in [91]. Protein concentration was quantified by Pierce BCA Protein Assay (ThermoFisher, cat #23225). Lysates with an equal amount of total protein were separated on 4% to 12% Bis-Tris NuPAGE Novex gels and transferred to a polyvinylidene difluoride (PVDF) membrane. Protein bands were detected using standard chemiluminescence techniques using GE Healthcare Amersham Imager 680.
The following antibodies were used in Western blotting: ZNF71 (GeneTex, cat. #GTX116553), from Cell Signaling Biotechnology: STING (cat. #13647), TBK1 pSer172 (cat. #5483), TBK1 (cat. #3504), OAS1 (cat. #14498), RNase L (cat. #27281), TRIM5α (cat. #14326), SAMHD1 pThr592 (cat. #89930), and GAPDH (Millipore, MAB374).
The Boolean implication network algorithm [92-94] was used in this study to generate gene associate networks. Details of the algorithm were included in our previous publications [15, 95]. Thresholds of scope and precision for filtering the implication rules were computed based on a one-tailed z-score of 1.64 (95% level of significance).
Xu's LUAD cohort [33] was used to construct the gene association networks. Each gene's expression was divided into three categories: under-expressed (−1), normal (0), and over-expressed (1). The categorization was based on the distribution of the selected housekeeping genes (B2M, ESD, FLOT2, GAPDH, GRB2, HPRT1, HSP90AB1, LDHA, NONO, POLR2A, PPP1CA, RHOA, SDCBP, and TFRC) [14, 96-98]. The percentage of under-expressed and over-expressed samples for all the housekeeping genes was fixed to be 30% in each dataset. Standard deviations were calculated for the normal range based on the housekeeping genes and applied to the rest genes. The numbers of standard deviation used for each dataset were: 0.68 for RNA sequencing data of LUAD tumors; 0.75 for RNA sequencing data of NATs; 0.88 for proteomics and RNA sequencing data of LUAD tumors; 0.95 for proteomics data and RNA sequencing data of NATs. The networks were visualized with Cytoscape (version 3.9.1) [99].
Genome-scale CRISPR-Cas9 knockdown data from project Achilles [100, 101] were obtained from the DepMap (release 21Q4) [102]. CRISPR-Cas9 dependency scores of human NSCLC cell lines (n=94) from the DepMap portal (https://depmap.org/portal/download/all/, accessed on 12 Sep. 2022) were used in this study. The CERES method was used to normalize gene-level dependency scores. The median of the normalized dependency scores of common essential genes was −1.0, and that of non-essential genes was 0 in each cell line. In this study, a dependency score less than −0.5 indicated that the gene had a significant effect on the cell line in CRISPR-Cas9 knockout.
Genome-wide RNA interference (RNAi) knockdown screening data [103] in project Achilles was analyzed in this study. RNAi dependency scores were processed with the DEMETER2 v6 algorithm [103] to distinguish between on- and off-target effects. The dependency scores of human NSCLC cell lines (n=92) were obtained from the DepMap portal (https://depmap.org/portal/download/all/, accessed on 12 Sep. 2022). The median of the normalized dependency scores of the positive control gene set was −1, and that of the negative control gene set was 0. The gene that had a dependency score less than −0.5 was considered as having a significant effect on the cell line in RNAi knockdown.
The ToppFun web tool (https://toppgene.cchmc.org/enrichment.jsp, accessed on 28 Jun. 2022) from the ToppGene suite [104] is an online resource for gene list enrichment analysis. We used the ToppGene application with default parameters (FDR correction, p-value cutoff=0.05, gene limits 1≤n≤2000) to perform pathway enrichment analysis.
RNA sequencing data of 135 human NSCLC cell lines were obtained from the Cancer Cell Line Encyclopedia (CCLE) in DepMap release 22Q2 (https://depmap.org/portal/download/all/, accessed on 12 Sep. 2022). Proteomics data of 64 human NSCLC cell lines were obtained from Nusinow et al [105].
The drug sensitivity data of human NSCLC cell lines were taken from multiple resources. The secondary PRISM repurposing dataset [35] was obtained from the DepMap release 19Q4. It has 1,448 compounds screened in 499 human cell lines, 84 of which were NSCLC cell lines and were used in this study. The Genomics of Drug Sensitivity in Cancer (GDSC) datasets (GDSC1 and GDSC2 [36-38]) were downloaded (https://www.cancerrxgene.org/downloads/bulk_download, accessed on 12 Sep. 2022). A total of 64 NSCLC cell lines from GDSC1 and 58 NSCLC cell lines from GDSC2 were included in this study. Details of the categorization of drug sensitivity in each cell line were included in our previous publications [15, 95].
The connectivity map (CMap) online tool (https://clue.io/, accessed on 12 Sep. 2022) [31, 32] was used to explore functional pathways and potential repositioning of drugs with selected gene expression signatures. Raw connectivity scores higher than 0.9 and a p-value<0.05 were considered significant.
Statistical analysis was performed using R software (version 4.1.3) with RStudio (version 2022.02.3 Build 492) [106]. Two sample t-tests (two-tailed) were performed for the comparisons of two groups of continuous variables. The independence of categorical variables was evaluated with χ2 tests. Principle component analysis (PCA) was used to generate the separation of NATs and LUAD tumors with selected genes. Visualization was carried out in R and Cytoscape. Kaplan-Meier method was performed to conduct survival analysis and create survival curves. The difference between the survival rates of patient groups was evaluated using log-rank tests. Univariate and multivariate Cox regression analyses were performed to evaluate the prognostic capacity of the studied factors. R packages survival, survminer, and ggplot2 were used in survival analysis. Pearson's correlation coefficients were used to determine the association between the two sample groups. All hypothesis tests were two-sided, and test results with a p-value<0.05 were considered statistically significant.
This invention extended a seven-gene panel for NSCLC prognosis using proteomic profiles. The results also showed that the seven-gene panel can accurately classify NSCLC tumors from NATs on both RNA-sequencing and proteomic platforms, suggesting its diagnostic implications for the early detection of lung cancer. Gene expression of ZNF71, a marker within the seven-gene panel, when combined with dendritic cell activities, can further stratify NSCLC into different prognostic groups. ZNF71 expression is associated with the activities of NK cells and NKT cells. Overexpression of ZNF71 results in decreased expression of multiple components of the intracellular intrinsic and innate immune systems, including dsRNA and dsDNA sensors, confirming a hypothesis that ZNF71 suppresses the transcription of genomic transposable elements. Multi-omics networks of ZNF71 and the intracellular innate immune response genes were revealed in NSCLC using a computational Boolean implication network algorithm. From these constructed networks, pan-sensitive and pan-resistant genes to 21 NCCN-recommended drugs for treating NSCLC were selected. We designed mechanisms of action to enhance treatment response, prolong patient survival, inhibit proliferation, and reverse EMT to screen candidates for new and repositioning drugs. PD-198306, U-0126, ZM-306416, and PQ-401 were identified as potential targeted therapy that may also induce immune response for treating NSCLC, which was not known before. Our future research will identify direct targets of ZNF71 for the development of novel therapeutic strategies to improve NSCLC survival outcomes.
A method of diagnosis and screening of non-small cell lung cancer is provided comprising using protein expression levels of one elected from the group consisting of ABCC4, DAG1, and SLC39A8 in patient samples, of ABCC4 and DAG1 in patient samples, of ABCC4 and SLC39A8 in patient samples, of DAG1 and SLC39A8 in patient samples, of ABCC4 in patient samples, of DAG1 in patient samples, and of SLC39A8 in patient samples.
A method of diagnosis and screening of non-small cell lung cancer is provided comprising using mRNA expression levels of ABCC4, CCL19, CD27, DAG1, FUT7, and ZNF71 in patient samples.
It will be appreciated by those skilled in the art that changes could be made to the embodiments described above without departing from the broad inventive concept thereof. It is understood, therefore, that this invention is not limited to the particular embodiments disclosed, but it is intended to cover modifications that are within the spirit and scope of the invention, as defined by the appended claims.
This utility non-provisional patent application claims the benefit of priority to U.S. Provisional Patent Application Ser. No. 63/384,696, filed Nov. 22, 2022. The entire contents of U.S. Provisional Patent Application Ser. No. 63/384,696 are incorporated by reference into this utility non-provisional patent application as if fully rewritten herein.
This invention was made with government support under Grant Nos. R01 LM009500, R56 LM009500, P20 RR016440 and GM130174 awarded by the National Institute of Health. The government has certain rights in the invention.
| Number | Date | Country | |
|---|---|---|---|
| 63384696 | Nov 2022 | US |
