Elevated FAM83A expression predicts poorer clincal outcome in lung adenocarcinoma

Abstract

BACKGROUND:

Family with sequence similarity 83 member A (FAM83A) can promote tumor cell proliferation and facilitate epidermal growth factor tyrosine kinase inhibitor resistance in some malignant tumors, but its role in lung cancer has not been directly explored.

OBJECTIVE:

We investigated FAM83A expression in lung adenocarcinoma (LUAD) and its significance in clinicopathologic characteristics and prognosis of the disease.

PATIENTS AND METHODS:

We analyzed the mRNA expression of FAM83A in LUAD and normal (or adjacent) lung tissues from Oncomine database firstly. Then, we detected FAM83A protein expression in five paired fresh LUAD and adjacent lung tissue specimens from patients in our hospital by Western blotting. In addtion, FAM83A expression in 86 paraffin-embedded archived LUAD samples was evaluated by Immunohistochemistry, and the correlations between FAM83A expression and clinicopathologic characteristics and prognosis of the patients were analyzed.

RESULTS:

Oncomine data analysis manifested that FAM83A mRNA expression was increased in LUAD. Western blotting revealed higher FAM83A expression in fresh LUAD tissues than in the adjacent lung tissues ( $P=$ 0.036). Immunohistochemistry analysis on 86 paraffin samples further demonstrated that the LUAD tissue had higher FAM83A expression than adjacent lung tissue ( $P<$ 0.001). The correlation analysis revealed that advanced stage tumors (stage III–IV) had higher FAM83A expression than early stage tumors (stage I–II) ( $P=$ 0.004). High FAM83A expression was significantly associated with lymphnode involvement and clinical staging ( $P=$ 0.008 and 0.008 respectively). Univariate and multivariate Cox regression analysis manifested that FAM83A expression was an independent predictive factor for poor survival. Kaplan-Meier survival curves showed that patients with higher FAM83A expression had shorter overall survival than those with lower FAM83A expressions ( $P=$ 0.002).

CONCLUSION:

FAM83A is upregulated in advanced LUAD and is related to unfavorible prognosis. FAM83A might be a novel diagnostic and prognositic biomarker for LUAD.

Keywords

Family with sequence similarity 83 member A (FAM83A)lung adenocarcinoma prognositic

1. Introduction

Lung cancer ranks the first in morbidity of malignant tumors and yields a high mortality worldwide [1]. In recent years, progress in identification of tumor driver genes have led to the development of many novel targeted therapies that can provide significant benefit to some patients, especially those with lung adenocarcinoma (LUAD). For example, epidermal growth factor-tyrosine kinase inhibitors (EGFR-TKIs) are highly effective in treating LUAD with mutated EGFRs [2, 3, 4]. These targeted drugs considerably extended the survival of some patients. However, upon the exposure to these targeted therapeutic drugs, cancer cells often develop resistance to these drugs through acquisition of mutations or other mechanisms in the targeted signaling pathways rendering the drugs ineffective. Therefore, clinical effects of these targeted drugs are not always satisfactory. Furthermore, most targeted drugs also inhibit normal cells and the dose-dependent toxicities limit their clinical application. Obviously, further strives in seeking novel and better therapeutic targets on LUAD remain urgently necessary.

Family with sequence similarity 83 member A (FAM83A) is one of the members of a novel protein family which consists of FAM83A-H. All FAM83 members contain a highly conserved 1699 (DUF1669) domain of unknown function located in the N terminus, while each FAM83 member has a distinct variable C terminus [5, 6, 7]. Early study demonstrated that breast cancer with EGFR mutations has poor response to EGFR-TKIs [8]. Lee and his colleagues first discovered that FAM83A was overexpressed and involved in the EGFR-TKIs resistance in breast cancer [5]. In the same study, the authors found that FAM83A was able to activate the signaling cascade downstream of EGFR, facilitate breast cell transformation, and confer EGFR-TKIs resistance. Subsequently, Liu and his colleagues, using a rapid nested PCR assay, detected the heightened levels of FAM83A in peripheral blood samples of breast cancer patients [9]. A recent study showed that HER2-positive breast cancer cells had overexpressed FAM83A, and the ablation of FAM83A inhibited HER2-positive breast cancer cell growth independently of trastuzumab sensitivity [10]. In addition, increased FAM83A was also found to be related to poor prognosis of the patients [11]. All these studies collectively suggested that FAM83A could be a potential biomarker for breast cancer prognosis.

The expression of FAM83A has also been found to be overexpressed in some other tumor types including brain tumors, pancreatic cancer, and ovarian cancer [11]. However, the significance of FAM83A in lung cancer has not yet been fully investigated. It has been reported that FAM83A mRNA was overexpressed in human lung cancer tissues and circulating tumor cells [12, 13]. Cipriano et al. further found that FAM83A mRNA was upregulated in lung cancer using an assay that detects mRNA of all FAM83 members (FAM83A, B, C, D, E, F, G, H) in several types of tumor tissues [14]. In addition, bioinformatic analysis of publically available data in GEO database also found FAM83A mRNA was overexpressed in lung cancer [11]. However, most of those studies explored FAM83A expressions merely at mRNA levels, not at protein level. It is well known that the level of mRNA expression does not necessarily correlate with the level of protein expression. Furthermore, the clinical significance of FAM83A overexpression in lung cancer is largely unknown. In this study, we focused on the expression of FAM83A protein in LUAD and its correlation with the clinicopathological characteristics of the patients.

2. Materials and methods

2.1 Bioinformatic analysis

We analyzed the FAM83A expression (mRNA) levels of LUAD tissues (44 cases) and normal lung tissues (65 cases) in Oncomine database. To exclude the impacts of individual differences, we compared the mRNA levels of FAM83A in 20 paired LUAD tissue and the matched adjacent lung tissue samples from the dataset.

2.2 Patient information and tissue samples

The patient enrollment criteria were: 1) Patient at stage I-III according to the TNM staging criteria established by American Joint Committee on Cancer (AJCC) [15]. 2) Patient who had complete clinical and pathological information including age, sex, pathological grading, pathological differentiation, TNM stage, and clinical stage etc. 3) Patient without distant metastasis. A total of 86 paraffin-embedded archived LUAD and adjacent non-tumor tissue samples from the patients in our hospital who met the inclusion criteria were used in this study. The clinicopathological characteristics of the patients were showed in Table 1. All the patients at stage II–IV accepted standard postoperative adjuvant chemotherapy. Five pairs of fresh LUAD samples and matched adjacent lung tissues were compared for their FAM83A protein expression using Western blotting. The protocol of this study was approved by the Institutional Research Ethics Committee of Anhui Provincial Hospital (No. 2019-P-017).

2.3 Western blotting

FAM83A protein expression was determined by Western blotting. Fragments of fresh tissues were ground in RIPA lysis buffer (P0013B, Beyotime, Shan-ghai, China) and protein concentration was measured with BCA protein concentration assay (P0010, Beyotime, Shanghai, China). Samples with equal amount of protein (20 $\mu$ g) were loaded on sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) (74255-250G, Sigma, USA) and transferred onto polyvinylidene difluoride (PVDF) membrane (IP VH00010, Millipore, MA, USA). After transferring, the PVDF membranes were blocked with 5% bovine serum albumin for two hours and incubated with anti-FAM83A rabbit polyclonal antibody (1:200, orb183 622, Biorbyt, Cambridge, UK) and anti-GAPDH (ab 8245, Abcam, Cambridge, UK) overnight at 4 ${}^{\circ}$ C, then incubated with horseradish peroxidase (HRP) conjugated goat anti-rabbit secondary antibodies (1:1000, DD23, Ascendbio, Guangzhou, China) and HRP conjugated goat anti-mouse secondary antibodies (1:1000, DD23, Ascendbio, Guangzhou, China) two hours at room temperature. The FAM83A expression was determined with chemiluminescence detection reagent (P0019, Beyotime, Shanghai, China) based on the operation instructions, GAPDH was used as a loading control.

2.4 Immunohistochemistry

We detected the protein expression of FAM83A in a tissue microarray by immunohistochemistry method. Immunohistochemistry analysis was performed on the paraffin-embedded LUAD specimens and adjacent lung tissue ones. Briefly, the tissue slices were incubated in dry oven at 63 ${}^{\circ}$ C for about one hour, deparaffinized and rehydrated on autostainer (LEICAST5010, LEICA, German) at room temperature. After that, the slides were washed with heat citric acid solution and immersed into staining dish with boiled citric acid solution to retrieve antigen. After retrieval, slides were bathed in peroxidase-blocking reagent for 15 minutes at room temperature, covered with rabbit anti-FAM83A (1:400; No: orb183622, Biorbyt, Britain) and incubated overnight at 4 ${}^{\circ}$ C in a humidified chamber. Then, slides were washed with PBS, incubated with the appropriate secondary antibody (Envision $+$ /HRP, Rabbit, Dako, Sweden) for 30 minutes at room temperature and incubated with DAB solution for 5 minutes and washed with tap water for 5 minutes. At last, slides were bathed for 40 seconds in hematoxylin solution and 2 seconds in 0.25% acid alcohol and dehydrated in the autostainer. Photomicrographs were read with Aperrio ImageScope v11.1.2.752.

The immunohistochemistry scoring of FAM83A expression was graded as previously described [16]. The score of staining intensity (S1) was evaluated as follows: 1) no staining; 2) weak staining, showing light yellow; 3) moderate staining, showing yellow brown; 4) strong staining, showing brown. The score of staining cell proportion (S2) was graded as follows: 0) no staining cells; 1) $<$ 10% staining cells; 2) 10–35% staining cells; 3) 35–75% staining cells; 4) $>$ 75% staining cells. The staining index $(\textit{SI})=S1\times S2$ . The possible SI of the LUAD samples contain 2, 4, 6, 8, 10, 12 and 16. The median value – 8 – was determined as cut off value. Hence, if a specimen with a $\textit{SI}\geqslant$ 8, it means high expression, $\textit{SI}<$ 8, it means low expression.

2.5 Statistical analysis

SPSS version 13.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. The correlation between the clinicopathological characteristics of the patients and FAM83A expressions was analyzed using the Chi-squared test. Prognostic factor analyses were performed using univariate and multivariate Cox regression analyses. Survival analyses were performed using Kaplan-Meier and Log-rank test. A two-tailed $P$ value $<$ 0.05 was set as statistically significant.

3. Results

3.1 FAM83A was overexpressed in LUAD

Using Oncomine data, we compared the FAM83A mRNA expression in 44 LUAD samples and 65 normal lung tissues and found that FAM83A mRNA expression was elevated in LUAD ( $P<$ 0.001, Fig. 1a). Furthermore, the mRNA levels of FAM83A in 20 paired LUAD tissue samples and matched adjacent lung tissues in the dataset were compared and the results showed similarly that FAM83A transcription was up-regulated in LUAD tissues relative to the adjacent tissues ( $P<$ 0.001, Fig. 1b). In the five paired fresh LUAD tissue samples and the adjacent non-tumor lung tissue ones from patients (Fig. 1d), FAM83A protein levels were also specifically increased in LUAD tissues compare to that in the adjacent normal lung tissues ( $P=$ 0.036, Fig. 1c). Those results showed that FAM83A increased in LUAD both in mRNA and protein levels.

Figure 1.

FAM83A is overexpressed in LUAD. A. FAM83A expression level (mRNA) was transcriptionally upregulated in 44 LUAD tissues in comparison with that of 37 normal lung tissue in the Oncomine database. B. FAM83A mRNA expression in LUAD tissue samples was higher than that of matched adjacent normal tissues in the Oncomine database. C. The relative FAM83A expression of LUAD was higher than that of pared adjacent normal tissues in Western blotting assay. D. Western blotting of FAM83A expression in five paired LUAD (T) and adjacent normal tissues (ANT), GAPDH was used as the loading control.

Figure 2.

FAM83A overexpression in LUAD correlates with poor overall survival of the patients. A. The tissue microarray including adjacent lung and LUAD tissues, representative IHC images of FAM83A expression in adjacent normal tissues (ANT) and LUAD with low (LOW) and high (HIGH) FAM83A expression. The scale equals to 50 micron. B. The IHC staining index of FAM83A in LUAD tissues (Tumor) and adjacent normal lung tissues (Adjacent). C. The IHC staining index of FAM83A in LUAD tissues with Stage I-II and Stage III-IV. D. Kaplan-Meier survival curves of 86 LUAD patients grouped by low and high expression of FAM83A.

Table 1

Clinicopathological characteristics and FAM83A expression of the patients

Characteristics	Number of cases (%)
Ages (y)
$\geqq$ 60	34 (39.5)
$<$ 60	52 (60.5)
Gender
Male	47 (54.7)
Female	39 (45.3)
Pathological grading
I	3 (3.5)
II	56 (65.1)
III	27 (31.4)
T classification
T1	16 (18.6)
T2	49 (57.0)
T3	16 (18.6)
T4	5 (5.8)
N classification
N0	36 (41.9)
N1	26 (30.2)
N2	19 (22.1)
N3	5 (5.8)
M classification
M0	81 (94.2)
M1	5 (5.8)
Clinical stage
I	25 (29.1)
II	24 (27.9)
III	32 (37.2)
IV	5 (5.8)
Vital states (at follow-up)
Alive	20 (23.3)
Dead	66 (76.7)
Expression of FAM83A
Low expression	37 (43.0)
High expression	49 (57.0)

Table 2

Correlation between clinicopathological characteristics and FAM83A expressions

Characteristics	FAM83A		$P$ value
	Low expression	High expression
Age (y)
$\geqq$ 60	13	21
$<$ 60	24	28	0.468
Gender
Male	19	28
Female	18	21	0.593
Clinical stage
I	17	8
II	11	13
III	8	24
IV	1	4	0.008
T classification
T1	8	8
T2	22	27
T3	6	10
T4	1	4	0.644
N classification
N0	23	13
N1	8	18
N2	4	15
N3	2	3	0.008
M classification
M0	36	45
M1	1	4	0.284

Table 3

Univariate and multivariate cox-regression analysis of diverse prognostic factors in patients with lung adenocarcinoma

	Univariate analysis			Multivariate analysis
	$P$	Hazard ratio	95% confidence interval	$P$	Harzard ratio	95% confidence interval
FAM83A	0.028	1.765	1.065–2.926	0.025	1.899	1.085–3.323

3.2 Upregulated FAM83A expression was associated with advanced clinicopathological features

The clinicopathological features of the patients used in this study were summarized in Table 1. We then examined FAM83A protein in LUAD tissue sections by immunohistochemistry staining. The results showed that FAM83A was mainly located in cytoplasm (Fig. 2A). The staining index of immunohistochemistry for FAM83A was significantly higher in LUAD tissues than that in the adjacent normal tissues ( $P<$ 0.001, Fig. 2B), which indicated that FAM83A was overexpressed in LUAD. Subgroup analysis by clinical stage showed FAM83A was expressed lower in tumors at early stages (stage I-II) and higher in advanced stages (stage III-IV) ( $P=$ 0.004, Fig. 2C). The correlation analysis revealed a strong correlation between FAM83A expression and clinical stages ( $P=$ 0.008) and lymph node stage classification ( $P=$ 0.008). Nevertheless, FAM83A expression had no association with age ( $P=$ 0.468), gender ( $P=$ 0.593), tumor classification ( $P=$ 0.644), and metastasis classifications ( $P=$ 0.284) (Table 2). All of those results suggested that high expression of FAM83A was related to advanced clinicopathological features.

3.3 FAM83A overexpression predicts poorer survival and prognosis

Univariate Cox regression analysis showed that high FAM83A expression increased the risk of death in comparison to low FAM83A expression (Hazard Ratio: 1.765, 95% Confidence Interval: 1.065–2.926, $P=$ 0.028). Multivariate Cox regression analysis demonstrated that FAM83A expression was a independent predictive factor for poorer survival of the patients (Hazard Ratio: 1.899, 95% Confidence Interval: 1.085–3.323, $P=$ 0.025) (Table 3). Kaplan Meier survival analysis manifested that patients with high FAM83A expression ( $n=$ 37) had poorer overall survivals than those with low FAM83A expression ( $n=$ 49), as showed in Fig. 2D, $P=$ 0.024. Patients with elevated FAM83A expression had higher mortality than those with lower expressions. Survival rate at the end of 1, 3, and 5 years for patients with high and low FAM83A expression were also different (0.857 vs 0.919, 0.653 vs 0.757, 0.204 vs 0.378 respectively). Those results demonstrated that FAM83A overexpression correlates with poorer prognosis and survival.

4. Discussion

FAM83A is a novel signaling protein that found to be overexpressed in some tumors and confer chemo-resistance to breast cancer [11]. Previous study indicated FAM83A expression was elevated in lung cancer at mRNA levels. In this study, we further demonstrated that FAM83A was also overexpressed in LUAD at protein in addition to mRNA levels, and the heightened FAM83A expression correlated with advanced clinicopathological features and poor prognosis.

FAM83A overexpression was firstly discovered in breast cancer. It may promote proliferation, invasion, and EGFR-TKI resistance of breast cancer cell [5]. FAM83A overexpression was also found later in some other tumors including pancreatic cancer, and overexpressed FAM83A may promote cancer stem cells-like traits and chemo-resistance [16]. However, there have been only a few studies on FAM83A expression in lung cancer. In present study, by analyzing the data from Oncomine database, we confirmed that FAM83A was overexpressed in LUAD at mRNA levels. To exclude the impact of individual differences, we compared twenty paired data and confirmed the earlier findings by others [12, 14].

It is well known that the expression of mRNA and protein are not always in agreement [17], meaning an elevated mRNA expression is not necessarily accompany with an increased protein expression. To ascertain the expression of FAM83A protein in LUAD, we detected FAM83A levels in fresh LUAD tissue samples using Western blotting. Our results demonstrated that FAM83A protein was indeed overexpressed in LUAD as suspected. Therefore, we are the first to report FAM83A protein is overexpressed in LUAD in addition to mRNA overexpression.

Previous studies have demonstrated that FAM83A may be important in tumor cell proliferation, invasion, metastasis and drug resistance [5, 18]. It has been suggested that FAM83A achieves this by serving as the downstream of EGFR, and significantly activate the downstream signaling pathway, such as PI3K/AKT and RAS/MAPK [18]. In the study of breast cancer, FAM83A expression was found to be invertly associate with patient’s survival. Given that EGFR signaling cascade is essential to the proliferation and survival of LUAD [19, 20], it is conceivable to hypothesize that FAM83A may also play an important role in LUAD and affect the clinical course of the disease as demonstrated in breast cancer. In this study, the results of immunohistochemistry analysis performed on the paraffin-embedded LUAD specimens and adjacent lung tissues showed FAM83A is indeed overexpressed in LUAD tissues. Furthermore, patients with higher FAM83A expression were associated with early lymph node metastasis and advanced tumor stages, poorer survivalship as presented by M-H survival analysis. Log-rank test showed a significant difference between patients with high and low FAM83A expressions ( $P=$ 0.024). We demonstrated remarkable correlations between high FAM83A expression and advanced clinicopathological features and poor prognosis.

In conclusion, our findings demonstrated aberrant overexpression of FAM83A in LUAD and it predicts more advanced clinicopathological features and poor prognosis. The results of our study suggested a possible oncogenic role of FAM83A in LUAD. It may also potentially serve as a biomarker for LUAD, which requires additional investigation and testing.

Footnotes

Acknowledgments

We sincerely thank Dr. Wu Xiaosheng (Mayo Clinic, Minnesota, US) for his help in language editing this research paper.

References

Parkin

D.M.

, Global cancer statistics in the year 2000, The Lancet Oncolog 2 (2001), 533–543.

Hanna

Johnson

Temin

Baker

Brahmer

Ellis

P.M.

Giaccone

Hesketh

P.J.

Jaiyesimi

Leighl

N.B.

Riely

G.J.

Schiller

J.H.

Schneider

B.J.

Smith

T.J.

Tashbar

Biermann

W.A.

and Masters

, Systemic therapy for stage IV non-Small-Cell lung cancer: american society of clinical oncology clinical practice guideline update, Journal of clinical oncology: Official journal of the American Society of Clinical Oncolog 35 (2017), 3484–3515.

Lindeman

N.I.

Cagle

P.T.

Aisner

D.L.

Arcila

M.E.

Beasley

M.B.

Bernicker

E.H.

Colasacco

Dacic

Hirsch

F.R.

Kerr

Kwiatkowski

D.J.

Ladanyi

Nowak

J.A.

Sholl

Temple-Smolkin

Solomon

Souter

L.H.

Thunnissen

Tsao

M.S.

Ventura

C.B.

Wynes

M.W.

and Yatabe

, Updated molecular testing guideline for the selection of lung cancer patients for treatment with targeted tyrosine kinase inhibitors: Guideline from the college of american pathologists, The International Association for the Study of Lung Cance 20 (2018), 129–159.

Copur

M.S.

Crockett

Gauchan

Ramaekers

and Mleczko

, Molecular testing guideline for the selection of patients with lung cancer for targeted therapy, Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncolog (2018), JCO2018788240.

Lee

S.Y.

Meier

Furuta

Lenburg

M.E.

Kenny

P.A.

and Bissell

M.J.

, FAM83A confers EGFR-TKI resistance in breast cancer cells and in mice, The Journal of Clinical Investigatio 122 (2012), 3211–3220.

Bartel

C.A.

Parameswaran

Cipriano

and Jackson

M.W.

, FAM83 proteins: Fostering new interactions to drive oncogenic signaling and therapeutic resistance, Oncotarge 7 (2016), 52597–52612.

Fulcher

L.J.

Bozatzi

Tachie-Menson

K.Z.L.

Cummins

T.D.

Bufton

J.C.

Pinkas

D.M.

Dunbar

Shrestha

Wood

N.T.

Weidlich

Macartney

T.J.

Varghese

Gourlay

Campbell

D.G.

Dingwell

K.S.

Smith

J.C.

Bullock

A.N.

and Sapkota

G.P.

, The DUF1669 domain of FAM83 family proteins anchor casein kinase 1 isoforms, Sci Signa 11 (2018).

Mueller

K.L.

Hunter

L.A.

Ethier

S.P.

and Boerner

J.L.

, Met and c-Src cooperate to compensate for loss of epidermal growth factor receptor kinase activity in breast cancer cells, Cancer Researc 68 (2008), 3314–3322.

Liu

Cheng

Xiao

Gao

Wang

and Song

, A rapid nested polymerase chain reaction method to detect circulating cancer cells in breast cancer patients using multiple marker genes, Oncology Letter 7 (2014), 2192–2198.

10.

Bartel

C.A.

and Jackson

M.W.

, HER2-positive breast cancer cells expressing elevated FAM83A are sensitive to FAM83A loss, PloS One 12 (2017), e0176778.

11.

Snijders

A.M.

Lee

S.Y.

Hang

Hao

Bissell

M.J.

and Mao

J.H.

, FAM83 family oncogenes are broadly involved in human cancers: An integrative multi-omics approach, Molecular Oncolog 11 (2017), 167–179.

12.

Dong

Yin

Zhang

Pang

Zhang

and Chen

, BJ-TSA-9, a novel human tumor-specific gene, has potential as a biomarker of lung cancer, Neoplasi 7 (2005), 1073–1080.

13.

Liu

Liao

G.Q.

Zhu

Liu

P.H.

Y.M.

Song

X.M.

Q.W.

Gao

Zhang

Chen

W.F.

and Yin

Y.H.

, Detection of circulating cancer cells in lung cancer patients with a panel of marker genes, Biochemical and Biophysical Research Communication 372 (2008), 756–760.

14.

Cipriano

Miskimen

K.L.

Bryson

B.L.

Foy

C.R.

Bartel

C.A.

and Jackson

M.W.

, Conserved oncogenic behavior of the FAM83 family regulates MAPK signaling in human cancer, Molecular Cancer Research: MC 12 (2014), 1156–1165.

15.

Carey

L.A.

Metzger

Dees

E.C.

Collichio

Sartor

C.I.

Ollila

D.W.

Klauber-Demore

Halle

Sawyer

Moore

D.T.

and Graham

M.L.

, American joint committee on cancer tumor-node-metastasis stage after neoadjuvant chemotherapy and breast cancer outcome, Journal of the National Cancer Institut 97 (2005), 1137–1142.

16.

Chen

Huang

Liu

Liang

Zhang

and Jin

, FAM83A is amplified and promotes cancer stem cell-like traits and chemoresistance in pancreatic cancer, Oncogenesi 300 (2017).

17.

Zhao

Akbani

Liu

Ling

Vellano

C.P.

Roebuck

Eterovic

A.K.

Byers

L.A.

Davies

M.A.

Deng

Gopal

Y.N.

Chen

Von Euw

E.M.

Slamon

Conklin

Heymach

J.V.

Gazdar

A.F.

Minna

J.D.

Myers

J.N.

Mills

G.B.

and Liang

, Characterization of human cancer cell lines by reverse-phase protein arrays, Cancer Cell 31 (2017), 225–239.

18.

Grant

, FAM83A and FAM83B: candidate oncogenes and TKI resistance mediators, The Journal of Clinical Investigatio 122 (2012), 3048–3051.

19.

Avraham

and Yarden

, Feedback regulation of EGFR signalling: decision making by early and delayed loops, Nature Reviews Molecular Cell Biolog 12 (2011), 104–117.

20.

Kolch

and Pitt

, Functional proteomics to dissect tyrosine kinase signalling pathways in cancer, Nature Reviews Cance 10 (2010), 618–629.