RPPA SPACE: an R package for normalization and quantitation of Reverse-Phase Protein Array data.

SUMMARY: Reverse-Phase Protein Array (RPPA) is a robust high-throughput, cost-effective platform for quantitatively measuring proteins in biological specimens. However, converting raw RPPA data into normalized, analysis-ready data remains a challenging task. Here, we present the RPPA SPACE (RPPA Superposition Analysis and Concentration Evaluation) R package, a substantially improved successor to SuperCurve, to meet that challenge. SuperCurve has been used to normalize over 170 000 samples to date. RPPA SPACE allows exclusion of poor-quality samples from the normalization process to improve the quality of the remaining samples. It also features a novel quality-control metric, 'noise', that estimates the level of random errors present in each RPPA slide. The noise metric can help to determine the quality and reliability of the data. In addition, RPPA SPACE has simpler input requirements and is more flexible than SuperCurve, it is much faster with greatly improved error reporting. AVAILABILITY AND IMPLEMENTATION: The standalone RPPA SPACE R package, tutorials and sample data are available via https://rppa.space/, CRAN (https://cran.r-project.org/web/packages/RPPASPACE/index.html) and GitHub (https://github.com/MD-Anderson-Bioinformatics/RPPASPACE). SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Analytical Platforms 3: Processing Samples via the RPPA Pipeline to Generate Large-Scale Data for Clinical Studies.

Reverse phase protein array (RPPA) is a functional proteomics technology amenable to moderately high throughputs of samples and antibodies. The University of Texas MD Anderson Cancer Center RPPA Core Facility has implemented various processes and techniques to maximize RPPA throughput; key among them are maximizing array configuration and relying on database management and automation. One major tool used by the RPPA Core is a semi-automated RPPA process management system referred to as the RPPA Pipeline. The RPPA Pipeline, developed with the aid of MD Avnderson's Department of Bioinformatics and Computational Biology and InSilico Solutions, has streamlined sample and antibody tracking as well as advanced quality control measures of various RPPA processes. This chapter covers RPPA Core processes associated with the RPPA Pipeline workflow from sample receipt to sample printing to slide staining and RPPA report generation that enables the RPPA Core to process at least 13,000 samples per year with approximately 450 individual RPPA-quality antibodies. Additionally, this chapter will cover results of large-scale clinical sample processing, including The Cancer Genome Atlas Project and The Cancer Proteome Atlas.

Development of a robust classifier for quality control of reverse-phase protein arrays.

MOTIVATION: High-throughput reverse-phase protein array (RPPA) technology allows for the parallel measurement of protein expression levels in approximately 1000 samples. However, the many steps required in the complex protocol (sample lysate preparation, slide printing, hybridization, washing and amplified detection) may create substantial variability in data quality. We are not aware of any other quality control algorithm that is tuned to the special characteristics of RPPAs. RESULTS: We have developed a novel classifier for quality control of RPPA experiments using a generalized linear model and logistic function. The outcome of the classifier, ranging from 0 to 1, is defined as the probability that a slide is of good quality. After training, we tested the classifier using two independent validation datasets. We conclude that the classifier can distinguish RPPA slides of good quality from those of poor quality sufficiently well such that normalization schemes, protein expression patterns and advanced biological analyses will not be drastically impacted by erroneous measurements or systematic variations. AVAILABILITY AND IMPLEMENTATION: The classifier, implemented in the "SuperCurve" R package, can be freely downloaded at http://bioinformatics.mdanderson.org/main/OOMPA:Overview or http://r-forge.r-project.org/projects/supercurve/. The data used to develop and validate the classifier are available at http://bioinformatics.mdanderson.org/MOAR.

Analysis of mutational and proteomic heterogeneity of gastric cancer suggests an effective pipeline to monitor post-treatment tumor burden using circulating tumor DNA.

Circulating tumor DNA (ctDNA) is released from tumor cells into blood in advanced cancer patients. Although gene mutations in individual tumors can be diverse and heterogenous, ctDNA has the potential to provide comprehensive biomarker information. Here, we performed multi-region sampling (three sites) per resected specimen from 10 gastric cancer patients followed by targeted sequencing and proteomic profiling using reverse-phase protein arrays. A total of 126 non-synonymous mutations were identified from 30 samples from 10 tumors. Of these, 16 (12.7%) were present in all three regions and were designated as founder mutations. Variant allele frequencies (VAFs) of founder mutations were significantly higher than those of non-founder mutations. Phylogenetic analysis also demonstrated a good concordance between founder and truncal mutations, defined as mutations shared by all simulated clones at the trunk of the tumor phylogenetic tree. These findings led us to prioritize founder mutations for quantitative ctDNA monitoring by digital PCR with individually-designed primer/probe sets. In preoperative plasma, the average ctDNA VAF of founder mutations was significantly higher than that of non-founder mutations (p = 0.039). Proteomic heterogeneity was present across the tumor regions both within and between patients independent of mutational status. Our results suggest that, in practice, mutations having high VAF identified without multi-regional sequencing may be immediately useful for quantitative ctDNA monitoring but do not provide sufficient information to predict the proteomic composition of tumors.

Suppression of Nuclear Factor-κB by Glucocorticoid Receptor Blocks Estrogen-Induced Apoptosis in Estrogen-Deprived Breast Cancer Cells.

Our clinically relevant finding is that glucocorticoids block estrogen (E2)-induced apoptosis in long-term E2-deprived (LTED) breast cancer cells. However, the mechanism remains unclear. Here, we demonstrated that E2 widely activated adipose inflammatory factors such as fatty acid desaturase 1 (FADS1), IL6, and TNFα in LTED breast cancer cells. Activation of glucocorticoid receptor (GR) by the synthetic glucocorticoid dexamethasone upregulated FADS1 and IL6, but downregulated TNFα expression. Furthermore, dexamethasone was synergistic or additive with E2 in upregulating FADS1 and IL6 expression, whereas it selectively and constantly suppressed TNFα expression induced by E2 in LTED breast cancer cells. Regarding regulation of endoplasmic reticulum stress, dexamethasone effectively blocked activation of protein kinase RNA-like endoplasmic reticulum kinase (PERK) by E2, but it had no inhibitory effects on inositol-requiring protein 1 alpha (IRE1α) expression increased by E2 Consistently, results from reverse-phase protein array (RPPA) analysis demonstrated that dexamethasone could not reverse IRE1α-mediated degradation of PI3K/Akt-associated signal pathways activated by E2 Unexpectedly, activated GR preferentially repressed nuclear factor-κB (NF-κB) DNA-binding activity and expression of NF-κB-dependent gene TNFα induced by E2, leading to the blockade of E2-induced apoptosis. Together, these data suggest that trans-suppression of NF-κB by GR in the nucleus is a fundamental mechanism thereby blocking E2-induced apoptosis in LTED breast cancer cells. This study provided an important rationale for restricting the clinical use of glucocorticoids, which will undermine the beneficial effects of E2-induced apoptosis in patients with aromatase inhibitor-resistant breast cancer.

GPNMB augments Wnt-1 mediated breast tumor initiation and growth by enhancing PI3K/AKT/mTOR pathway signaling and ß-catenin activity.

Glycoprotein Nmb (GPNMB) is overexpressed in triple-negative and basal-like breast cancers and its expression is predictive of poor prognosis within this aggressive breast cancer subtype. GPNMB promotes breast cancer growth, invasion, and metastasis; however, its role in mammary tumor initiation remains unknown. To address this question, we overexpressed GPNMB in the mammary epithelium to generate MMTV/GPNMB transgenic mice and crossed these animals to the MMTV/Wnt-1 mouse model, which is known to recapitulate features of human basal breast cancers. We show that GPNMB alone does not display oncogenic properties; however, its expression dramatically accelerates tumor onset in MMTV/Wnt-1 mice. MMTV/Wnt-1 × MMTV/GPNMB bigenic mice also exhibit a significant increase in the growth rate of established primary tumors, which is attributable to increased proliferation and decreased apoptosis. To elucidate molecular mechanisms underpinning the tumor-promoting effects of GPNMB in this context, we interrogated activated pathways in tumors derived from the MMTV/Wnt-1 and MMTV/Wnt-1 × MMTV/GPNMB mice using RPPA analysis. These data revealed that MMTV/Wnt-1 × MMTV/GPNMB bigenic tumors exhibit a pro-growth signature characterized by elevated PI3K/AKT/mTOR signaling and increased ß-catenin activity. Furthermore, we extended these observations to an independent Wnt-1 expressing model of aggressive breast cancer, and confirmed that GPNMB enhances canonical Wnt pathway activation, as evidenced by increased ß-catenin transcriptional activity, in breast cancer cells and tumors co-expressing Wnt-1 and GPNMB. GPNMB-dependent engagement of ß-catenin occurred, in part, through AKT activation. Taken together, these data ascribe a novel, pro-growth role for GPNMB in Wnt-1 expressing basal breast cancers.

Validation of reverse phase protein array for practical screening of potential biomarkers in serum and plasma: accurate detection of CA19-9 levels in pancreatic cancer.

The current study analyzed reverse phase protein arrays (RPPA) as a means to experimentally validate biomarkers in blood samples. One microliter samples of sera (n = 71), and plasma (n = 78) were serially diluted and printed on NC-coated slides. CA19-9 levels from RPPA results were compared with identical patient samples as measured by ELISA. There was a strong correlation between RPPA and ELISA (r = 0.87) as determined by scatter plots. Sample reproducibility of CA19-9 levels was excellent (interslide correlation r = 0.88; intraslide correlation r = 0.83). The ability of RPPA to accurately distinguish CA19-9 levels between cancer and noncancer samples were determined using receiver operating characteristic curves and compared with ELISA. The AUC for RPPA and ELISA was comparable (0.87 and 0.86, respectively). When the mean CA19-9 levels of normal samples was used as a cutoff for RPPA and compared with the standard clinical ELISA cutoff, comparable specificities (71% for both) were observed. Notably, RPPA samples normalized to albumin showed increased sensitivity compared to ELISA (90% vs. 75%). As RPPA is a high-throughput method that shows results comparable to that of ELISA, we propose that RPPA is a viable technique for rapid experimental screening and validation of candidate biomarkers in blood samples.

Telomere 3' overhang-specific DNA oligonucleotides induce autophagy in malignant glioma cells.

Telomere 3' overhang-specific DNA oligonucleotides (T-oligos) induce cell death in cancer cells, presumably by mimicking telomere loop disruption. Therefore, T-oligos are considered an exciting new therapeutic strategy. The purpose of this study was to elucidate how T-oligos exert antitumor effects on human malignant glioma cells in vitro and in vivo. We demonstrated that T-oligos inhibited the proliferation of malignant glioma cells through induction of nonapoptotic cell death and mitochondria hyperpolarization, whereas normal astrocytes were resistant to T-oligos. Tumor cells treated with T-oligos developed features compatible with autophagy, with development of autophagic vacuoles and conversion of an autophagy-related protein, microtubule-associated protein 1 light chain 3 from type I (cytoplasmic form) to type II (membrane form of autophagic vacuoles). A reverse-phase protein microarray analysis and Western blotting revealed that treatment with T-oligos inhibited the mammalian target of the rapamycin (mTOR) and the signal transducer and activator of transcription 3 (STAT3). Moreover, pretreatment with T-oligos significantly prolonged the survival time of mice inoculated intracranially with malignant glioma cells compared with that of untreated mice and those treated with control oligonucleotides (P=0.0065 and P=0.043, respectively). These results indicate that T-oligos stimulate the induction of nonapoptotic autophagic also known as type II programmed cell death and are thus promising in the treatment of malignant glioma.

Antitumor mechanisms of combined gastrin-releasing peptide receptor and epidermal growth factor receptor targeting in head and neck cancer.

Head and neck squamous cell carcinoma (HNSCC) is characterized by epidermal growth factor receptor (EGFR) overexpression, where EGFR levels correlate with survival. To date, EGFR targeting has shown limited antitumor effects in head and neck cancer when administrated as monotherapy. We previously identified a gastrin-releasing peptide/gastrin-releasing peptide receptor (GRP/GRPR) aurocrine regulatory pathway in HNSCC, where GRP stimulates Src-dependent cleavage of EGFR proligands with subsequent EGFR phosphorylation and mitogen-activated protein kinase (MAPK) activation. To determine whether GRPR targeting can enhance the antitumor efficacy of EGFR inhibition, we investigated the effects of a GRPR antagonist (PD176252) in conjunction with an EGFR tyrosine kinase inhibitor (erlotinib). Combined blockade of GRPR and EGFR pathways significantly inhibited HNSCC, but not immortalized mucosal epithelial cell, proliferation, invasion, and colony formation. In addition, the percentage of apoptotic cells increased upon combined inhibition. The enhanced antitumor efficacy was accompanied by increased expression of cleaved poly(ADP-ribose) polymerase (PARP) and decreased phospho-EGFR, phospho-MAPK, and proliferating cell nuclear antigen (PCNA). Using reverse-phase protein microarray (RPPA), we further detected decreased expression of phospho-c-Jun, phospho-p70S6K, and phospho-p38 with combined targeting. Cumulatively, these results suggest that GRPR targeting can enhance the antitumor effects of EGFR inhibitors in head and neck cancer.

Src activation regulates anoikis in human colon tumor cell lines.

Src is a non-receptor protein tyrosine kinase, the expression and activity of which is increased in >80% of human colon cancers with respect to normal colonic epithelium. Previous studies from this and other laboratories have demonstrated that Src activity contributes to tumorigenicity of established colon adenocarcinoma cell lines. Src participates in the regulation of many signal transduction pathways, among which are those leading to cellular survival. In this study, we addressed the potential role of Src activation to a specific aspect of tumor cell survival, resistance to detachment-induced apoptosis (anoikis). Using five colon tumor cell lines with different biologic properties and genetic alterations, we demonstrate that expression and activity of Src corresponds with resistance to anoikis. Enforced expression of activated Src in subclones of SW480 cells (of low intrinsic Src expression and activity) increases resistance to anoikis; whereas decreased Src expression in HT29 cells (of high Src expression and activity) by transfection with anti-sense Src expression vectors increases susceptibility to anoikis. In contrast, increasing or decreasing Src expression had no effect on susceptibility to staurosporine-induced apoptosis in attached cells. PD173955, a Src family-specific tyrosine kinase inhibitor, increases the susceptibility of HT29 cells to anoikis in a dose- and time-dependent manner. Increasing Src expression and activity led to increased phosphorylation of Akt, a mediator of cellular survival pathways, whereas decreasing Src activity led to decreased Akt phosphorylation. In colon tumor cells with high Src activity, the PI3 kinase inhibitor LY 294002 sensitized cells to anoikis. These results suggest that Src activation may contribute to colon tumor progression and metastasis in part by activating Akt-mediated survival pathways that decrease sensitivity of detached cells to anoikis.

KGF-induced motility of breast cancer cells is dependent on Grb2 and Erk1,2.

Breast cancer metastasis is directly associated with breast cancer cell motility. Using a cell culture wounding model, we have demonstrated that keratinocyte growth factor (KGF) enhanced the motility of estrogen receptor-positive breast cancer cells. However, the mechanisms by which KGF enhanced motility of breast cancer cells are not known. In the present study, we report that KGF-induced motility requires intact tyrosine kinase signaling since genistein, a tyrosine kinase inhibitor, led to decreased motility of breast cancer cells mediated by KGF. Using cDNA microarrays, we previously found that KGF increased the expression of Grb2 mRNA by 2 3-fold. Since Grb2 plays an important role in tyrosine kinase signaling, we examined the involvement of Grb2 in KGF-induced motility. Down-regulation of Grb2 protein expression inhibited KGF-induced motility. Since Grb2 is known to regulate Erk1,2 and Akt kinase activities we determined whether these downstream proteins may be vital to KGF-induced motility. Inhibiting the activation of Erk1,2 by PD98059 suppressed KGF-induced motility whereas inhibiting the activation of Akt by wortmannin did not affect KGF-induced motility. In conclusion, these results indicate that KGF mediated signal transduction employs Grb2 to transduce the tyrosine kinase signals resulting in the activation of Erk1,2 and breast cancer cell motility.

HER2/neu uses Akt to suppress retinoic acid response element binding activity in MDA-MB-453 breast cancer cells.

We previously demonstrated that HER2/neu prevents all trans-retinoic acid (ATRA) from inducing growth inhibition in MDA-MB-453 breast cancer cells. For ATRA to induce growth inhibition, it needs to bind to retinoic acid receptors and modulate gene transcription via retinoic acid response elements (RAREs). We hypothesized that HER2/neu suppresses RARE binding activity to prevent ATRA from inducing growth arrest in breast cancer cells. Electrophoretic mobility shift assays showed that when HER2/neu was inhibited by the trastuzumab antibody, RARE binding activity increased, indicating that HER2/neu suppresses RARE binding. Since trastuzumab also decreased Akt activity, we determined whether Akt regulates RARE binding activity. Compared to parental MDA-MB-453 cells, MDA-MB-453 cells transfected with a dominant negative Akt mutant (MDA-MB-453/DN-Akt) had higher RARE binding activity. However, trastuzumab did not further increase RARE binding activity in MDA-MB-453/DN-Akt cells. These data indicate that HER2/neu predominantly uses Akt to suppress RARE binding activity, which may be one mechanism by which HER2/neu induces ATRA resistance in breast cancer cells.

The HER2/Grb2/Akt pathway regulates the DNA binding activity of AP-1 in breast cancer cells.

We showed that the HER2/Grb2/Akt pathway induces all-trans retinoic acid (ATRA) resistance in breast cancer cells by suppressing the DNA binding activity of retinoic acid receptors (RAR). AP-1 activation was shown to induce ATRA resistance. Here, we determined whether AP-1 binding activity is correlated with ATRA resistance in HER2-overexpressing cells. Inhibition of HER2/Grb2/Akt decreased AP-1 binding activity in HER2-transfected cells, but increased AP-1 activity in cells that are naturally HER2-overexpressing. Since HER2/Grb2/Akt inhibition sensitized both cell types to ATRA, our results indicate that, unlike RAR, AP-1 binding activity is not correlated with ATRA sensitivity in HER2-overexpressing breast cancer cells.

LKB1 is a central regulator of tumor initiation and pro-growth metabolism in ErbB2-mediated breast cancer.

BACKGROUND: Germline and somatic mutations in STK11, the gene encoding the serine/threonine kinase LKB1, are strongly associated with tumorigenesis. While loss of LKB1 expression has been linked to breast cancer, the mechanistic role of LKB1 in regulating breast cancer development, metastasis, and tumor metabolism has remained unclear. METHODS: We have generated and analyzed transgenic mice expressing ErbB2 in the mammary epithelium of LKB1 wild-type or LKB1-deficient mice. We have also utilized ErbB2-expressing breast cancer cells in which LKB1 levels have been reduced using shRNA approaches. These transgenic and xenograft models were characterized for the effects of LKB1 loss on tumor initiation, growth, metastasis and tumor cell metabolism. RESULTS: We demonstrate that loss of LKB1 promotes tumor initiation and induces a characteristic shift to aerobic glycolysis ('Warburg effect') in a model of ErbB2-mediated breast cancer. LKB1-deficient breast cancer cells display enhanced early tumor growth coupled with increased cell migratory and invasive properties in vitro. We show that ErbB2-positive tumors deficient for LKB1 display a pro-growth molecular and phenotypic signature characterized by elevated Akt/mTOR signaling, increased glycolytic metabolism, as well as increased bioenergetic markers both in vitro and in vivo. We also demonstrate that mTOR contributes to the metabolic reprogramming of LKB1-deficient breast cancer, and is required to drive glycolytic metabolism in these tumors; however, LKB1-deficient breast cancer cells display reduced metabolic flexibility and increased apoptosis in response to metabolic perturbations. CONCLUSIONS: Together, our data suggest that LKB1 functions as a tumor suppressor in breast cancer. Loss of LKB1 collaborates with activated ErbB2 signaling to drive breast tumorigenesis and pro-growth metabolism in the resulting tumors.

Small molecule ErbB inhibitors decrease proliferative signaling and promote apoptosis in philadelphia chromosome-positive acute lymphoblastic leukemia.

The presence of the Philadelphia chromosome in patients with acute lymphoblastic leukemia (Ph(+)ALL) is a negative prognostic indicator. Tyrosine kinase inhibitors (TKI) that target BCR/ABL, such as imatinib, have improved treatment of Ph(+)ALL and are generally incorporated into induction regimens. This approach has improved clinical responses, but molecular remissions are seen in less than 50% of patients leaving few treatment options in the event of relapse. Thus, identification of additional targets for therapeutic intervention has potential to improve outcomes for Ph+ALL. The human epidermal growth factor receptor 2 (ErbB2) is expressed in ~30% of B-ALLs, and numerous small molecule inhibitors are available to prevent its activation. We analyzed a cohort of 129 ALL patient samples using reverse phase protein array (RPPA) with ErbB2 and phospho-ErbB2 antibodies and found that activity of ErbB2 was elevated in 56% of Ph(+)ALL as compared to just 4.8% of Ph(-)ALL. In two human Ph+ALL cell lines, inhibition of ErbB kinase activity with canertinib resulted in a dose-dependent decrease in the phosphorylation of an ErbB kinase signaling target p70S6-kinase T389 (by 60% in Z119 and 39% in Z181 cells at 3 µM). Downstream, phosphorylation of S6-kinase was also diminished in both cell lines in a dose-dependent manner (by 91% in both cell lines at 3 µM). Canertinib treatment increased expression of the pro-apoptotic protein Bim by as much as 144% in Z119 cells and 49% in Z181 cells, and further produced caspase-3 activation and consequent apoptotic cell death. Both canertinib and the FDA-approved ErbB1/2-directed TKI lapatinib abrogated proliferation and increased sensitivity to BCR/ABL-directed TKIs at clinically relevant doses. Our results suggest that ErbB signaling is an additional molecular target in Ph(+)ALL and encourage the development of clinical strategies combining ErbB and BCR/ABL kinase inhibitors for this subset of ALL patients.

Tumor epidermal growth factor receptor and EGFR PY1068 are independent prognostic indicators for head and neck squamous cell carcinoma.

PURPOSE: To assess the prognostic value of epidermal growth factor receptor (EGFR) molecular characteristics of head and neck squamous cell carcinoma (HNSCC). PATIENTS AND METHODS: HNSCC tumors from patients prospectively enrolled in either an Early Detection Research Network (EDRN) study and treated with surgery without an EGFR-targeted agent (N = 154) or enrolled in a chemoradiation trial involving the EGFR-targeted antibody cetuximab (N = 39) were evaluated for EGFR gene amplification by FISH and EGFR protein by immunohistochemical staining. Fresh-frozen tumors (EDRN) were also evaluated for EGFR protein and site-specific phosphorylation at Y992 and Y1068 using reverse-phase protein array (n = 67). Tumor (n = 50) EGFR and EGFRvIII mRNA levels were quantified using real-time PCR. RESULTS: EGFR expression by immunohistochemistry (IHC) was significantly higher in the EDRN tumors with EGFR gene amplification (P < 0.001), and a similar trend was noted in the cetuximab-treated cohort. In the EDRN and cetuximab-treated cohorts elevated EGFR by IHC was associated with reduced survival (P = 0.019 and P = 0.06, respectively). Elevated expression of total EGFR and EGFR PY1068 were independently significantly associated with reduced progression-free survival in the EDRN cohort [HR = 2.75; 95% confidence interval (CI) = 1.26-6.00 and HR = 3.29; 95% CI = 1.34-8.14, respectively]. CONCLUSIONS: In two independent HNSCC cohorts treated with or without cetuximab, tumor EGFR levels were indicative of survival. Tumor EGFR PY1068 levels provided prognostic information independent of total EGFR.

Targeting the epidermal growth factor receptor in epithelial ovarian cancer: current knowledge and future challenges.

The epidermal growth factor receptor is overexpressed in up to 60% of ovarian epithelial malignancies. EGFR regulates complex cellular events due to the large number of ligands, dimerization partners, and diverse signaling pathways engaged. In ovarian cancer, EGFR activation is associated with increased malignant tumor phenotype and poorer patient outcome. However, unlike some other EGFR-positive solid tumors, treatment of ovarian tumors with anti-EGFR agents has induced minimal response. While the amount of information regarding EGFR-mediated signaling is considerable, current data provides little insight for the lack of efficacy of anti-EGFR agents in ovarian cancer. More comprehensive, systematic, and well-defined approaches are needed to dissect the roles that EGFR plays in the complex signaling processes in ovarian cancer as well as to identify biomarkers that can accurately predict sensitivity toward EGFR-targeted therapeutic agents. This new knowledge could facilitate the development of rational combinatorial therapies to sensitize tumor cells toward EGFR-targeted therapies.

Proteomic profiling identifies pathways dysregulated in non-small cell lung cancer and an inverse association of AMPK and adhesion pathways with recurrence.

INTRODUCTION: The identification of key pathways dysregulated in non-small cell lung cancer (NSCLC) is an important step toward understanding lung pathogenesis and developing new therapeutic approaches. METHODS: Toward this goal, reverse-phase protein lysate arrays (RPPA) were used to compare signaling pathways between NSCLC tumors and paired normal lung tissue from 46 patients and assess their association with clinical outcome. RESULTS: After RPPA quantification of 63 proteins and phosphoproteins, tissue pairs were randomized to a training set (n = 25 pairs) and test set (n = 21 pairs). In the training set, 15 protein markers were differentially expressed between tumors and normal lung (p ≤ 0.01), including markers in the PI3K/AKT and p38 MAPK signaling pathways (e.g., p70S6K, S6, p38, and phospho-p38), as well as caveolin-1 and ß-catenin. A four-protein signature (p70S6K, cyclin B1, pSrc(Y527), and caveolin-1) independent of histology classified specimens as tumor versus normal with a predicted accuracy of 83%, sensitivity of 67%, and specificity of 100%. The signature was validated in the test set, correctly classifying all normal tissues and 14 of 21 tumor tissues. RPPA results were confirmed by immunohistochemistry for caveolin-1 and p70S6K. In tumors from patients with resected NSCLC, expression of proteins in the energy-sensing AMPK pathway (pLKB1, AMPK, p-Acetyl-CoA, pTSC2), adhesion, EGFR, and Rb signaling pathways was inversely associated with NSCLC recurrence. CONCLUSIONS: These data provide evidence for dysregulation of several pathways including those involving energy sensing and adhesion that are potentially associated with NSCLC pathogenesis and disease recurrence.