• Open Access

Prognostic impact of Raf-1 and p-Raf-1 expressions for poor survival rate in non-small cell lung cancer


To whom correspondence should be addressed.

E-mail: weimi003@yahoo.com


Overexpression of Raf-1 has commonly been observed in solid tumors including non-small cell lung cancer (NSCLC). The objective of this study was to investigate whether overexpression of Raf-1, phosphorylated-Raf-1 (p-Raf-1) or both correlates with poor survival rate in NSCLC patients and to explore associations between expression of these proteins and NSCLC cell fate both in vitro and in vivo. Expression of Raf-1 and p-Raf-1 were detected by immunohistochemistry in tumor specimens from 152 NSCLC patients and associations between their expression and the clinicopathological characteristics were assessed. Five-year median survival rate of patients were analyzed by Kaplan–Meier method, log-rank test and Cox regression. Cell fate was compared between normal tumor cells and those with Raf-1 silencing, in both the adenocarcinoma cell line A549 and xenografted mice that were infected with the A549 cell line. The incidence of overexpression of both Raf-1 and p-Raf-1 in NSCLC was much higher than normal control (< 0.05), and the survival rate of patients with positive expression of Raf-1, p-Raf-1 or both was found to be significantly lower than the negative group (< 0.05). Both univariate and multivariate analyses showed Raf-1 (P = 0.000, P = 0.010), p-Raf-1 (P = 0.004, P = 0.046), or both (P = 0.001, P = 0.016) was good prognostic markers for poor survival rate in NSCLC patients. Suppression of Raf-1 inhibited tumorigenesis by inducing apoptosis both in vitro and in vivo. These findings demonstrate that overexpression of Raf-1, p-Raf-1 or both could be considered as a new independent prognostic biomarker for poor survival rates for NSCLC patients.

The survival rate for patients with lung cancer remains unchanged in the last decade, despite major advances in cancer therapies.[1] Non-small cell lung cancer (NSCLC) accounts for 80% of all lung cancers and is a leading cause of cancer mortality worldwide.[1] Surgery is the main curative treatment for early stage NSCLC patients, but the prognosis for this disease is generally poor and the median survival for advanced cases treated with chemotherapy alone is about 10 months.[2] Therefore besides tumor pathological scores, practical prognostic biomarkers are needed for better screening and treatment modalities for this disease.

Raf was initially discovered just under three decades ago as the transforming principle shared by a murine sarcoma and an avian carcinoma virus.[3] As the linker between Ras (the most frequently mutated oncogene in human cancers) and the mitogen-activated protein kinase/extracellular signal-regulated kinase (MEK/ERK) module, Raf has been conclusively established as a major player in tumor development.[3] Raf encodes serine/threonine activity and for mammals the subtypes are: A-Raf, B-Raf and C-Raf (Raf-1),[4] and viral oncogene homolog v-Raf. In response to growth factor stimulation and Ras activation C-Raf (Raf-1) and B-Raf form heterodimers to yield higher MEK kinase activity for cell proliferation and tumor development.[5] Overexpression of Raf-1 has been reported in squamous cell carcinoma (SCC),[6] in lung adenocarcinoma (ADC),[7] and correlates with metastatic progression.[8]

The present study aimed to investigate the prognostic value of Raf-1, p-Raf-1 and clinicopathological markers in NSCLC patients. The associations between Raf-1 expression and tumor cell fate were further explored both in vitro and in vivo.

Materials and Methods

Patients and tissue collection

For preliminary experiments, Raf-1 expression was examined in both NSCLC specimens and normal samples nearby them from 20 patients (ADC 8, SCC 12) who did not receive neoadjuvant therapy. One hundred and seventy patients who were admitted to West China Hospital from 2004 to 2005 and who received complete resection for primary NSCLC without previous neoadjuvant therapy were subsequently recruited. Normal lung samples nearby the tumors were taken as controls. Patients underwent standard therapeutic procedure after surgical resection according to the Clinical Oncology Information Network guidelines for nonsurgical management of lung cancer.[9] A total of 152 patients with complete datasets after 5-year follow-up were finally enrolled. All the tissues were fixed in 10% formalin immediately and embedded in paraffin within 12–24 h of surgical resection. Tumors were staged according to the International Union Against Cancer's tumor-node-metastasis system, while their differentiation and histological types were assessed according to the World Health Organization's classification for NSCLC.[10, 11] The study design and procedure were approved by the institutional review board of West China Hospital, and all patients completed a written informed consent before enrollment.

Antibodies and immunohistochemistry

All paraffin-embedded tissues were sliced to a thickness of 4 μm for immunohistochemistry studies. Primary antibodies were: Rabbit anti-Raf-1 (Signalway Antibody, MD, USA), Rabbit anti-Raf-1 (phospho-ser259) (Signalway Antibody, College Park, MD, USA), Rabbit anti-phospho P70S6kinase (Cell Signaling, Shanghai, China), Mouse anti-GAPDH Ki67 (Thermo Fisher Scientific, Rockford, IL, USA). Secondary antibodies were Goat anti-Rabbit IgG (Dako, Shanghai, China) and Goat anti-Mouse IgG (Dako). Evision method according to kit instructions, and antigen retrieval was done by heating (Tris)-ethylenediaminetetraacetic acid retrieval solution (pH 8.0) at 95°C for 45 min in a water bath. Immunohistochemistry staining was undertaken as has been previously described in the literature.[12]

Immunohistochemistry scoring

All the stained slides were assessed by two independent pathologists without knowledge of the patients' clinical data according to a dual rate semi-quantitative method.[12] Ten randomly selected fields for each slide were scored for area and intensity of positively stained (brown) cytoplasm and/or cell membrane under light microscopy. Scores for positively stained area were: no staining, 0; <20% cells stained, 1; 20–50% cells stained, 2; and >50% cells stained, 3. Intensity scores for positively stained area were: no appreciable staining, 0; barely detectable staining, 1; readily appreciable brown staining, 2; and dark brown staining, 3. The total score was calculated by multiplying the area with the intensity score, producing a number from 0 to 9. For statistical analysis, scores of 2–9 were defined as positive expression (overexpression) while scores of 0–2 were negative (low-expression).

Raf-1 gene silencing and tumorigenesis in xenografted mice

Raf-1-shRNA and shRNA-Neg (with normal Raf-1 expression) were constructed in lentivirus vector (LV) using RNA interference target sequence (GAGACATGAAATCCAACAATA) of Raf-1 gene according to dsRNAs design principles and manuals (BD Biosciences, Shanghai, China). Lentivirus vector shRNA-Neg (NC group) and LV Raf-1-shRNA (KD group) were then transfected into the human lung ADC cell line A549 obtained from American Type Culture Collection (ATCC). Expression of Raf-1 and its downstream molecule p-P70S6 kinase were tested by flurogenic quantitative polymerase chain reaction and Western blot using GAPDH as an internal control.

BALB/c nu/nu mice of 6–8 weeks old were supplied by the Laboratory Animal Center of Sichuan University, and were housed in laminar flow cabinets under specific pathogen-free conditions. The lung ADC model was established by subcutaneous injection of the A549 cell line (5 × 106 each) with either LV shRNA-Neg (NC group, = 6) or LV Raf-1-shRNA (KD group, = 6) into the back of the mice. Control mice (CON group, = 6) received only injections of A549 cell line. Tumor lump was observed 7 days after the vaccination and tumor volume was calculated every 3 days, according to a previously described formula.[13] The mice were killed 4 weeks after tumor implantation, and the tumor weight was measured. Expression of Raf-1 and the cellular proliferation marker Ki-67 was examined by immunohistochemistry. The TUNEL method was used to determine apoptosis in xenograft tumor tissue. Ki-67 and apoptotic cells were expressed as a percentage of the total number of tumor cells. All data analyses were conducted by Image Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, USA).

Statistical analysis

The association of clinical variables with protein expression status was determined by Pearson's χ2 test. The Kaplan–Meier method was used to estimate univariate survival. The log-rank test and univariate Cox regression analysis were used to compare survival distributions between positive and negative staining groups. Independent prognostic factors of survival were identified with a multivariate Cox regression analysis. The statistical significance of difference between the groups was determined with the one-way anova followed by Fisher's protected least significant difference (PLSD) post hoc test. The Scheffe method was used to compare differences between the two groups. Data analyses and summarizations were conducted using SPSS 17.0 for Windows (SPSS, Chicago, IL, USA).


Raf-1 overexpression of in NSCLC and normal control groups

For the preliminary 20 NSCLC patients, the integrated optical density of Raf-1/GAPDH (IODRaf-1/GAPDH) shown by immunohistochemistry in ADC (48.34 ± 20.12) and SCC (55.43 ± 23.88) groups were significantly higher than normal controls (5.98 ± 3.93). In line with this, the IODRaf-1/GAPDH shown by Western blot in ADC (0.82 ± 0.11) and SCC (0.91 ± 0.23) groups were also significantly higher than normal controls (0.16 ± 0.03).

Figure 1 shows the positive expression of Raf-1 and p-Raf-1, which were both detected in the cytoplasm and/or cell membrane. Of the 152 specimens included in each group, there were 83 (54.6%) specimens with Raf-1 overexpression in the NSCLC group and 21 (13.8%) in the normal control group. The Raf-1 overexpression rate was significantly different between the two groups (= 0.000). There were 104 (68.4%) specimens with p-Raf-1 overexpression in the NSCLS group and 18 (11.8%) in the normal control group. The overexpression rate of both proteins was also significantly different between the two groups (= 0.000).

Figure 1.

Expression of Raf-1 and p-Raf-1 in non-small cell lung cancer (NSCLC) by immunohistochemistry. Representative images for immunohistochemistry scores of 0–2, 3–5, and 6–9, respectively. Negative expression of Raf-1 (a) and positive expression of Raf-1 (b, c). Negative expression of p-Raf-1 (d) and positive expression of p-Raf-1 (e, f). Original magnification, ×400.

The relationships between Raf-1, p-Raf-1 or both protein expression profiles and clinical variables are summarized in Table 1. The results showed that the overexpression of p-Raf-1 was associated with T stage (= 0.049). Overexpression of both Raf-1 and p-Raf-1 was also associated with T stage (= 0.024). There were no associations between Raf-1 overexpression and gender, age, histological subtypes, differentiation, tumor size, lymph node invasion, or distant metastasis (all P > 0.05).

Table 1. Expression of Raf-1 and p-Raf-1 in non-small cell lung cancer (NSCLC) patients
 Patients (n)Raf-1p-Raf-1Raf-1 and p-Raf-1
  1. ADC, adenocarcinoma; CI, confidence interval; HR, hazard ratio; p-Raf-1, phosphylated-Raf-1; SCC, squamous cell carcinoma.

Age (years)
Lymph node

Univariate anaylsis

The Kaplan–Meier method was used to evaluate the correlations between expression of Raf-1 proteins and patients' 5-year median survival rate (Fig. 2). Log-rank test was used to assess the associations between clinicopathological factor and patient outcomes. The associations of Raf-1 overexpression and clinical variables with survival rate were estimated by the univariate analysis (Table 2). Overexpression of Raf-1, p-Raf-1, or Raf-1 and p-Raf-1 showed significantly less survival rate (all  0.001) and shorter median survival time (all  0.001) compared with those with negative results, respectively. Histological subtypes, differentiation, tumor size, lymph node invasion and distant metastasis were all also shown to associate with poor prognosis of NSCLS patients (all P < 0.05).

Figure 2.

Correlations between Raf-1/p-Raf-1 expression and survival rate in non-small cell lung cancer (NSCLC). Kaplan–Meier curve of protein expression comparing positive (green line) with negative (red line). Raf-1 expression and survival (a), p-Raf-1 expression and survival (b) and both Raf-1 and p-Raf-1 expression and survival (c).

Table 2. Cox model estimates of the prognostic significance of each parameter for 5-year survival rates
VariablesMedian survival time (month)UnivariateMultivariate
HR95% CIP-valueHR95% CIP-value
  1. ADC, adenocarcinoma; CI, confidence interval; HR, hazard ratio; p-Raf-1, phosphylated-Raf-1; SCC, squamous cell carcinoma.

Histology (ADC/SCC/others)43.1/55.4/56.40.510.27–0.960.0360.200.06–0.730.015
Differentiation (poor/well)64.0/–3.620.0132.121.17–3.850.014
Tumor (T1/T2/T3/T4)74.1/54.0/34.7/34.31.841.32–2.560.0000.360.20–0.650.001
Lymph node (N0/N1/N2)57.6/46.8/27.71.781.24–2.550.0020.670.09–5.120.665
Metastasis (M0/M1)65.7/34.31.741.18–2.570.0051.761.17–2.640.007
Raf-1 (−/+)58.1/–3.820.0041.681.01–2.800.046
p-Raf-1 (−/+)62.8/44.52.881.56–5.180.0012.141.16–3.970.016
Raf-1 and p-Raf-1 (−/+)63.5/41.13.631.77–7.430.0003.711.37–10.020.010

Multivariate analysis

Based on the outcomes above, subsequent multivariate analysis was conducted by incorporating the overexpressions of Raf-1, p-Raf-1, and Raf-1/p-Raf-1, histology type, differentiation, tumor size, lymph node invasion, and TNM stage as covariates. The results from the Cox's regression model suggested that all the elements above were independent prognosis factors for NSCLC development (log-rank, all < 0.05) except lymph node invasion.

Raf-1 expression and tumor cell fate in vitro and in vivo

The Raf-1 mRNA expression in the KD (0.10 ± 0.01) group was significantly down regulated compared with the CON (1.00 ± 0.09) and NC (0.88 ± 0.14) groups. The IODRaf-1/GAPDH of Raf-1 protein expression in the KD (0.20 ± 0.03) group was also significantly downregulated compared with the CON (1.56 ± 0.35) and NC (1.46 ± 0.42) groups. Moreover, we have also confirmed that the downstream molecule of Raf-1, p-P70S6 kinase, in the KD (0.36 ± 0.23) group were significantly downregulated compared to the CON (1.19 ± 0.41) and NC (1.36 ± 0.45).

The immunohistochemical results of tumor tissue isolated from xenografted mice demonstrated that RNA interference signicantly decreased expression of Raf-1 (Fig. 3–a–c). The tumor growth and cell fate for each group are shown in Table 3. Tumor weight of the KD group was significantly lighter than both NC and CON groups (P < 0.05), while there was no difference between the NC group and the CON group (> 0.05). The TUNEL staining demonstrated that the apoptotic rate for tumor cells in the KD group (27.2%) was significantly higher than the CON (3.9%) and NC (1.8%) groups (Table 3 and Fig. 3–d–f). Cell proliferation as expressed by Ki-67 was significantly lower in the KD group compared to both of the other two groups (both < 0.05), while there was no difference between the NC and CON groups (Table 3 and Fig. 3–g–i).

Figure 3.

The expression of Raf-1, Ki-67 and terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL). Expression of Raf-1 in control (CON) (a), lentivirus vector chRNA-Neg (NC) (b) and LV Raf-1-shRNA (KD) (c) groups. TUNEL staining in the CON (d), NC (e) and KD (f) groups. Expression of Ki-67 CON (g), NC (h) and KD (i) groups. The arrows show the positive stained cells. Original magnification, ×400.

Table 3. The results of tumor growth, TUNEL analysis and Ki-67 expression
GroupsTumor weightTUNEL positive rate (%)Ki-67 positive rate (%)
  1. *< 0.01 versus control (CON); **< 0.01 versus NC group (Lentivirus vector shRNA-Neg). KD group, LV Raf-1-shRNA; TUNEL, terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling.

CON (= 6)110.50 ± 28.941.8 ± 0.673.2 ± 16.3
NC (= 6)140.67 ± 61.843.9 ± 1.269.9 ± 13.2
KD (= 6)58.83 ± 21.79*,**27.2 ± 9.5*,**26.4 ± 4.4*,**


Non-small cell lung cancer is a highly heterogeneous disease, the prognosis of which is difficult to determine by clinical variables and molecular/gene markers.[14] Therefore, it is vital to understand the biological behaviors of this tumor to provide useful predictors for clinical outcomes and for guiding treatment. Currently, besides TNM staging, clinical staging and differentiation, other factors such as angiogenic markers,[15] genetic markers,[16, 17] surviving,[18] Bad expression,[19] insulin receptor,[20] E-cadherin,[21] cell cycle deregulation,[22] and pERK1/2 and pAkt-1 have been reported to be independent prognostic markers for NSCLC patients.

Raf is considered as a unique gene of cancer and various studies have e stablished its important role in Ras/Raf/MEK/ERK classic signaling pathway.[3] Raf-1 and B-Raf have been proved to contribute to the prognostic role in a variety of malignant tumors. Indeed, high expression of Raf-1 has commonly been observed in solid tumors, including renal carcinoma, hepatocellular carcinoma and NSCLC.[3] Increased expression of Raf-1 was also observed in cell lines and most malignant cells, suggesting that overexpression of Raf-1 protein may be related to the malignant transformation of cells.[23] Overexpression of Raf-1 in transgenic mice makes them prone to develop lung cancer. Based on these findings, we postulated that Raf-1 might have reasonable prognostic role for NSCLC patients.

Immunohistochemistry and Western blot analyses from 20 NSCLC patients suggest that Raf-1 protein was significantly upregulated in the ADC and SCC compared to the normal controls. But there was no significant difference of Raf-1 expression between ADC and SCC (P > 0.05). Subsequently, results from 152 NSCLC patients suggest that the incidence of overexpression of Raf-1, p-Raf-1, or both was significantly higher in the NSCLC specimens compared to their normal controls. There were no significant differences in expressions of Raf-1, p-Raf-1, or both in terms of gender, age, histological subtypes, differentiation, lymph node invasion, and distant metastasis. However, overexpression of p-Raf-1 (P = 0.049), Raf-1 and p-Raf-1 (P = 0.024) were shown to relate to tumor size. These findings clearly demonstrate that in NSCLC specimens, Raf-1, p-Raf-1, or both was upregulated and the latter two were associated with tumor size. Subsequently, the prognostic value of Raf-1 and p-Raf-1 for poor survival rate of NSCLC patients was investigated and compared with clinical variables. Univariate analysis revealed that the median survival time was associated with Raf-1, p-Raf-1, Raf-1 and p-Raf-1, histology subtypes, differentiation, tumor size, lymph node invasion and distant metastasis. Except lymph node invasion, all these factors were confirmed in the multivariate analysis to be independent prognostic factors in NSCLC (log-rank, < 0.05). The median 5-year survival time of positive Raf-1, p-Raf-1, and Raf-1 and p-Raf-1 group was significantly shorter than the negative group (all < 0.05). These findings suggest that Raf-1 and p-Raf-1 had similar predictive value for NSCLC patients compared to TNM staging system and tumor differentiation.

The disadvantage for dividing NSCLC patients into risk groups based on clinical and pathological description is the variation in interpreting these qualitative variables. Therefore, researchers have been focusing on potential prognostic biomarkers extensively, especially on NSCLC related protein expression profiles. Compared to genetic study, which is expensive and not widely available, the protein expression studies by immunohistochemistry are routinely carried out in many hospitals, and they are rapid, reproducible and costless. As our data have shown that the overexpression of Raf-1, p-Raf-1, or both correlated well with poor survival rate of NSCLC patients, these proteins could be used as predictors for clinical outcomes. However, to draw such a conclusion, precautions mush be made for the limitations of this study. First, the sample size for this study was not large enough and patients were all from a local region, whether geography distribution could affect Raf-1 protein expression in NSCLC patients remains unknown. Second, as it was a retrospective study and the lack of biosamples of lung tumors, the expression of these proteins was not further confirmed by precise quantitative methods but was only tested by immunohistochemistry, which is at best a semi-quantitative method. Third, this study included various stages of patients, which might introduce bias into the multivariate analysis.

Raf-1 participates in regulating the MEK cascade, which is the main target of Ras protein,[24] thus the signal from the extracellular receptor/Ras complex is transferred to the nucleus where it regulates a variety of physiological and pathological processes such as cell proliferation, differentiation, apoptosis, survival and malignant transformation of cell.[25] To understand how Raf-1 expression affects tumor cell fate, a gene silencing technique was adopted to knock down Raf-1. In human lung ADC cell line A549, the KD group had significantly less expression of Raf-1 compared with NC and CON groups. Our in vitro experiments also showed that Raf-1 could induce the proliferation of tumor cells, promote invasive ability, and arrest the cell cycle at the stage of G0/G1 (data not shown). Results from our in vivo experiments were consistent with our in vitro observations, indicating that downregulation of Raf-1 expression attenuates proliferation and induces apoptosis, as the weight and volume of the tumor tissue isolated from nude mice was significantly lower in the KD group compared to the other two groups.

Disclosure Statement

The authors have no conflict of interest.