SEARCH

SEARCH BY CITATION

Keywords:

  • lung carcinomas;
  • mammalian target of rapamycin;
  • phosphorylation;
  • rapamycin;
  • epidermal growth factor receptor;
  • morphogenesis;
  • metastasis

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

BACKGROUND: Aberrant signaling cascades emanating from epidermal growth factor receptor (EGFR) are involved in the complex network of oncogenic signaling in lung carcinomas. One representative cascade is the phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin (mTOR) pathway. METHODS: The authors investigated the involvement of mTOR in the pathobiologic profiles of 150 specimens of lung carcinoma by immunohistochemistry and immunoblotting in correlation with the upstream and downstream proteins Akt and p70S6-kinase (S6K), respectively. RESULTS: Immunohistochemistry revealed Akt activation in 44% of tumors and mTOR expression in 68.7% of tumors, and the preponderance of activation was observed in adenocarcinoma (AC) (100%). Phosphorylated mTOR (p-mTOR) was observed in 53.3% of tumors and had the highest frequency in AC (89.7%). In AC, the frequency of p-mTOR staining was higher in the well differentiated subtype, in particular, in the acinar structure. However, little correlation was observed between the activation of mTOR and Akt, except in the 5 AC specimens that harbored an EGFR gene mutation, which exhibited constitutive activation of both Akt and mTOR. Conversely, in squamous cell carcinomas, mTOR activation was associated with a significantly higher frequency of lymph node metastasis. CONCLUSIONS: The results of this study suggested the dual functions of mTOR. First, mTOR may function not only in the proliferation of tumor cells as an effector molecule downstream of EGFR but also possibly in the morphogenesis of AC. Second, the activation of mTOR may play a key role in metastasis in squamous cell carcinoma. Overall, the current results demonstrated the potential for the application of rapamycin, an mTOR inhibitor, as an additional novel component of chemotherapy for a defined subset of patients with lung carcinoma. Cancer 2009. © 2008 American Cancer Society.

Despite the remarkable advances in oncology medicine and research, the prognosis for patients with lung cancer remains poor.1, 2 A great effort is being made to develop more effective modes of treatment by ‘molecularly targeted therapy.’3 In pursuit of this strategy, it has been revealed that many aberrant signaling cascades emanating from growth factor receptors are involved in oncogenic signaling; therefore, they and their effector molecules have been investigated as possible drug targets. This has led to the development of novel pharmacologic agents, including those that target epidermal growth factor receptor (EGFR).2, 3 However, although it has been demonstrated that therapies targeting EGFR are effective in a defined subset of patients,3 the results still have been heterogeneous because of the complexity of the signaling pathways involved in each case. Thus, more precise and comprehensive analyses are required to identify molecules that are critical to individual cases and to develop more effective drug combinations.

There are a growing number of reports describing amplification and mutation of EGFR in lung carcinomas. We and others have reported that EGFR overexpression is observed in 30% to 60% of nonsmall cell lung carcinomas (NSCLC), and up to 70% of those tumors exhibited amplification of EGFR.2, 4, 5 In addition, mutation of EGFR was observed in 25% to 40% of adenocarcinomas (AC), and these tumors transduce signals predominantly through Akt.2, 6, 7 These observations suggest that the phosphatidylinositol 3-kinase (PI3-K)/Akt pathway plays a critical role downstream of EGFR in a defined population of NSCLC, in particular, those that harbor EGFR mutations. Akt transduces signals to various molecules, including the tuberous sclerosis complex (TSC) protein.8, 9 Akt inhibits the activity of TSC, which leads to activation of the molecule farther downstream, mammalian target of rapamycin (mTOR), which is a 289-kD serine/threonine (Ser/Thr) protein kinase.8 The mTOR pathway is highly conserved,9, 10 and its activation positively regulates cell proliferation by promoting entry into G1-phase of the cell cycle through phosphorylation of the substrates p70S6 kinase (S6K) and 4E-BP1, which cooperate in translational initiation.10, 11 Because of these functions, mTOR has been regarded as an attractive target of anticancer agents.8, 10 The bacterial macrolide rapamycin, which was a well known immunosuppressant and also inhibits mTOR activity, has emerged as a validated, novel cancer therapeutic.10, 12 In vitro studies have established the potential of rapamycin to inhibit cellular transformation, and several kinds of tumors that exhibit the activation of PI3-K/Akt, including renal cell carcinoma, small cell lung carcinoma (SmCC), and others, are hypersensitive to rapamycin in vivo.8, 13 On the basis of these observations, rapamycin and its derivatives currently are being evaluated in clinical trials for solid tumors.9, 14, 15

In lung carcinomas, if the mTOR pathway is a predominant downstream effector of Akt even in the defined subset, then rapamycin could be a powerful therapeutic agent. However, the overexpression or activation of mTOR and/or its association with biologic behavior remain unclear. Therefore, we examined the expression, activation, and tissue distribution of mTOR protein in correlation with the upstream and downstream molecules Akt and S6K, respectively. Specifically, to evaluate the utility of mTOR as a molecular target, we investigated 1) whether mTOR is constitutively overexpressed and/or activated in lung carcinomas, 2) whether there is a correlation between the aberrations of EGFR and the activation of Akt/mTOR/S6K proteins, and 3) whether the level of mTOR protein or its activation state predicts the biologic profiles of tumors and/or patients' prognosis.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

Cases and Classification

For this study, 150 tumors, including 53 squamous cell carcinoma (SCC) specimens, 58 AC specimens, 9 large cell carcinoma (LCC) specimens, and 30 SmCC specimens, were obtained from surgeries or biopsies (29 of the SmCC specimens) that were performed at University of Yamanashi Hospital. Histologic diagnosis, differentiation, and stage were evaluated according to the World Health Organization classifications16, 17 and the International Union Against Cancer TNM classification18: The SCC specimens included 17 well differentiated (WD) subtypes, 29 moderately differentiated (MD) subtypes, and 7 poorly differentiated (PD) subtypes. The AC specimens included 29 WD subtypes, 21 MD subtypes, and 8 PD subtypes (Table 1). With regard to the lymph node status (N classification) of all NSCLC specimens, 58 tumors were classified as N0, 43 tumors were classified as N1, and 19 tumors were classified as N2. This project was approved by the institutional committee in each university, and informed consent was obtained from each patient.

Table 1. Results of Immunohistochemical Analysis of 150 Specimens
 No. of Specimens
 p-AKT ScoremTOR Scorep-mTOR Scorep-S6K Score
Histologic Type/Subtype012012012012
  • p-AKT indicates phosphorylated AKT; mTOR, mammalian target of rapamycin; p-mTOR, phosphorylated mTOR; p-S6K, phosphorylated S6K; BAC, broncioalveolar carcinoma; Mixed, adenocarcinoma of the mixed subtype; WD, well differentiated; MD, moderately differentiated; PD, poorly differentiated; Papillary, papillary adenocarcinoma; Solid, solid adenocarcinoma with mucin.

  • *

    Higher than that noted in the other histologic types (P < .0001).

Adenocarcinoma (n=58)            
 BAC (n=13)103002110310742
 Mixed            
  WD (n=14)12200410059635
  MD (n=21)137105161416687
  PD (n=7)151025430133
 Papillary (n=2)110011011110
 Solid (n=1)001010100101
 Total (n=58)3718301543*61636*211918
Squamous cell carcinoma (n=53)            
 WD (n=17)87264710431322
 MD (n=29)151131161218832441
 PD (n=7)421421430511
 Total (n=53)27206211220321564274
Large cell carcinoma (n=9)423423522621
Small cell carcinoma (n=30)16104226227212532
Total845016473568703545943125

Immunohistochemistry

Paraffin sections were autoclaved and immunostained with the following primary antibodies, which are available commercially from Cell Signaling Technology (Beverly, Mass); phosphorylated Akt (p-AktSer473), 1:50 dilution; mTOR, 1:200 dilution; phosphorylated mTOR (p-mTORSer2448), 1:100 dilution; and phosphorylated S6K (p-S6KThr389), 1:150 dilution. Antibodies were observed by using the ChemMate Envision/peroxidase complex kit (DAKO Japan, Kyoto, Japan). The sensitivity and specificity of antibodies were validated previously by us and others on cell lines and on the tissue specimens by immunohistochemistry and immunoblotting.7, 19–21 Immunohistochemical (IHC) expression was evaluated by 2 of us (Y.D. and S.S.), and the intensity of reactivity was evaluated as either significant or not significant, with significant defined as definite staining with higher intensity than that observed in nonneoplastic cells. Furthermore, the IHC expression level was scored by using the following 3-tier system: negative expression (score 0), <10% of tumor cells with significant staining; low expression (score 1), ≥10% but <50% of tumor cells with significant staining; and high expression (score 2), <50% of tumor cells with significant staining.7, 22, 23 When scores were classified into 2 groups for statistical analysis, the groups with low and high expression were combined into a single positive expression group.

Fresh Surgical Tissues

In 27 NSCLC specimens (11 SCC, 15 AC, and 1 LCC) and in 1 stage I SmCC, fresh tumor tissues and adjacent non-neoplastic tissues were available. Among these, 23 tissues had been examined previously for EGFR protein expression and gene aberrations.7 The 5 newly obtained specimens (3 SCC, 1 AC, and 1 SmCC) were evaluated for their EGFR status by using identical methods.5, 7, 22

Immunoblotting Analysis

Immunoblotting analysis (IB) was performed as described previously with the antibodies described above and with anti-β-actin antibody (1:2500 dilution; Cytoskeleton, Denver, Colo).7, 22 Protein levels relative to the β-actin signal (using an arbitrary level of 10) were quantified by using an Image Gauge (Fujifilm, Tokyo, Japan), and the results were designated as the ‘expression value.’ Next, the ‘protein index’ in tumor tissue was calculated as follows: 1) The ‘expression value’ in tumor tissue was divided by that in paired normal tissue; and 2) when expression was barely detectable in non-neoplastic tissue, the tumor ‘expression value’ was used directly as the ‘protein index.’ Protein expression was interpreted as ‘up-regulated’ or ‘activated’ when 1) expression was observed only in tumor tissue and was higher than that in any non-neoplastic tissues, and 2) the ‘protein index’ was >1.5.7, 22, 24

Statistical Analysis

Agreement among observers in the interpretation of IHC specimens was evaluated by using the κ statistic, as described previously.5, 25 Differences in the rate of positive immunostaining between the 2 groups were analyzed by the Fisher test. Correlations between the IHC score and clinicopathologic factors were evaluated by using the Mann-Whitney U test or the Kruskal-Wallis test. Differences in the level of protein expression/activation were analyzed by using an unpaired comparison t test. Patient survival was analyzed by using the Kaplan-Meier method.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

Immunohistochemistry

The results of IHC are summarized in Table 1. Overall interobserver agreement was ‘nearly perfect’ (lowest κ [for p-S6K], 0.893; 95% confidence interval, 0.874-0.912).

Phosphorylated Akt

Positive p-Akt staining was observed in both the cytoplasm and the nucleus of tumor cells in 66 tumors (44%), including 26 of 53 SCC specimens (49.1%), 21 of 58 AC specimens (36.2%), 5 of 9 LCC specimens (55.6%), and 14 of 30 SmCC specimens (46.7%). The IHC expression level ranged from 13.5% (AC) to 56.2% (SCC) in ‘positive’ tumors. Nuclear staining was observed predominantly in the peripheral or solid areas of the tumor nodule (Fig. 1A).

thumbnail image

Figure 1. Immunohistochemical staining for proteins downstream of epidermal growth factor receptor. (A) A case of adenocarcinoma (AC) demonstrates positive staining for phosphorylated Akt (p-Akt), mammalian target of rapamycin (mTOR), phosphorylated mTOR (p-mTOR), and phosphorylated S6K (p-S6K). p-Akt was observed in the nuclei of the solid regions. Although mTOR and p-mTOR were observed in the cytoplasm of tumor cells, the latter was localized within the acinar structure. p-S6K was observed in the nucleus and cytoplasm. The arrow on the p-S6K photomicrograph indicates a mitotic figure. (B) Cytoplasmic staining of mTOR. SCC indicates squamous cell carcinoma; LCC, large cell carcinoma; SmCC, small cell lung carcinoma. (C) There is intense p-mTOR staining in the acinar structure in an AC, focally positive staining in a nest of SCC, faint staining in an LCC, and rare positivity in SmCC (original magnification ×100 in Panels A and C, ×200 in Panel B).

Download figure to PowerPoint

Mammalian target of rapamycin

In normal tissues, cytoplasmic mTOR staining was observed in the germinal center of the lymphoid follicle and in bronchial epithelial cells. In tumor tissues, IHC revealed positive staining for mTOR in 103 of 150 tumors (68.7%) without any particular distribution pattern. This positive staining was observed, regardless of the degree of differentiation, in 58 of 58 AC specimens (100%), 32 of 53 SCC specimens (60.4%), and 5 of 9 LCC specimens (55.6%); but it was lower in 8 of 30 SmCC specimens (26.7%) (Fig. 1A,B). In addition, 74.1% of positive AC specimens had ‘high’ IHC scores with expression levels of up to 82%. In SCC specimens, expression levels ranged from 12.8% to 63.5%.

Phosphorylated mammalian target of rapamycin

Weakly positive staining in non-neoplastic cells was observed in the cytoplasm of bronchial epithelial cells, but not in the germinal center of the lymphoid follicles (Fig. 1A,C). In tumor tissues, p-mTOR was expressed in a particular manner and was observed in 52 of 58 AC specimens (89.7%), 4 of 9 LCC specimens (44.4%), 21 of 53 SCC specimens (39.6%), and 3 of 30 SmCC specimens (10%). In the AC specimens, 36 of 52 of p-mTOR-positive specimens had ‘high’ IHC scores of up to 86.4% expression levels. However, in the SCC specimens, the group with ‘high’ IHC scores included only 6 of 21 of positive tumors with expression levels of up to 70.3%. In the AC specimens, p-mTOR staining was observed predominantly in the acinar structure and less in the tumor cells with solid sheets (Fig. 1A,C). Furthermore, although p-mTOR staining was observed exclusively in the cytoplasm of tumor cells in the acinar structure, occasional positive staining in the solid part was localized in the nuclei.

Phosphorylated S6K

Staining for p-S6K was positive in 56 tumors (37.3%), including 37 of 58 AC specimens (63.8%), and 20 of those positive specimens (54.1%) had nuclear staining. Positive staining was observed in 20.8% of SCC specimens, 33.3% of LCC specimens, and 16.7% of SmCC specimens. IHC expression levels ranged from 11.5% (SmCC) to 72.9% (SCC). Carcinoma cells of all histologic subtypes exhibited p-S6K staining in the mitotic nuclei (Fig. 1A).

Immunoblotting Analysis

The protein levels detected by IB generally correlated well with the immunohistochemical results (Fig. 2, Table 2).

thumbnail image

Figure 2. Protein levels evaluated by immunoblotting analysis in representative tumors. The lysates were prepared from tissues with mixed epithelial and stromal cells. The intensity of the signals was expressed as the ratio relative to β-actin, which was designated as 10. Asterisks indicate the tumors that demonstrated lymph node metastasis. EGFR indicates epidermal growth factor receptor; IHC, immunohistochemistry; AC, adenocarcinoma; SCC, squamous cell carcinoma; LC, large cell carcinoma; SmCC, small cell lung carcinoma; N, non-neoplastic tissue; T, tumor tissue; p-Akt, phosphorylated Akt; mTOR, mammalian target of rapamycin; p-mTOR, phosphorylated mTOR; S6K, phosphorylated S6K.

Download figure to PowerPoint

Table 2. Overall Data of the Tissues Examined
  ImmunohistochemistryImmunoblotting Analysis   
Case No.Histologyp-AktmTORp-mTORp-S6Kp-AktmTORp-mTORp-S6KFISHMutationMetastasis
  • FISH indicates fluorescence in situ hybridization; p-AKT, phosphorylated AKT; mTOR, mammalian target of rapamycin; p-mTOR, phosphorylated mTOR; p-S6K, phosphorylated S6K; AC, adenocarcinoma; Mixed, adenocarcinoma of mixed subtype; −, negative; +, positive; SCC, squamous cell carcinoma; WD, well differentiated; MD, moderately differentiated; PD, poorly differentiated; Papillary, papillary adenocarcinoma; Solid, solid adenocarcinoma with mucin; LC, large cell carcinoma; SmCC, small cell lung carcinoma.

  • *

    In these tumors, the results were evaluated as either ‘up-regulated’ or ‘activated.’

1ACMixed++1.212.85*1.303.25*+ +
2SCCWD++1.243.36*3.50*0.82+ +
3SCCMD+1.081.401.081.83*+  
4SCCMD++1.153.11*1.003.25*+  
5ACMixed++++4.70*2.5*2.33*2.08*Del747−753+
6ACMixed+++3.69*2.07*1.88*0.81L858R+
7ACMixed+++2.36*2.67*2.33*1.27Del747−753insS+
8ACMixed+++3.50*1.88*2.08*0.67L858R 
9ACMixed++++3.33*1.87*2.33*2.00*L859R 
10SCCWD1.131.161.090.71  
11ACMixed+++1.142.07*2.50*2.62*  
12ACMixed+++1.332.46*2.15*2.43*  
13SCCPD+++2.58*3.36*3.75*0.73 +
14ACMixed++++1.60*2.29*2.17*3.07*  
15ACMixed++1.80*2.38*3.17*2.23*  
16ACMixed+++1.332.16*2.33*3.33*  
17ACMixed++1.70*2.64*1.383.00*  
18ACMixed+++1.333.40*2.70*2.89*  
19ACPapillary++1.382.53*2.00*1.13  
20ACSolid++0.893.00*0.782.10*  
21SCCMD++2.11*2.20*1.101.29  
22SCCMD+2.20*1.050.901.07  
23SCCMD+++2.73*1.95*1.60*2.33* +
24SCCMD++2.71*1.74*2.77*1.27  
25SCCPD+1.251.441.362.55*  
26SCCMD1.381.451.001.24  
27LC +++2.33*3.31*2.80*1.06  
28SmCC +1.62*2.29*1.201.00  
Phosphorylated Akt

Akt activation was noted in 15 of 28 tumors (53.6%), including 5 of 11 of SCC specimens, 8 of 15 of AC specimens, 1 LCC specimen, and 1 SmCC specimen. Tumor tissues and adjacent normal tissues revealed average p-Akt indices of 2.26 and 1.04, respectively.

Mammalian target of rapamycin

mTOR expression was observed as the 289-kD form in 28 tumor specimens and 26 normal tissue specimens with indices of up to 3.26 and 1.46, respectively. In tumor tissues, mTOR up-regulation was observed in 23 of 28 tumors (6 of 11 SCC specimens, 15 of 15 AC specimens, 1 LCC specimen, and 1 SmCC specimen).

Phosphorylated mammalian target of rapamycin

mTOR was activated in 17 of 28 tumors, including 4 of 11 of SCC specimens, 12 of 15 AC specimens, and 1 of 1 LCC specimen. The average p-mTOR indices in SCC and ACC were 1.74 and 2.08, respectively. In normal tissue specimens, p-mTOR was detectable in 20 of 28 tissues with indices of up to 1.3, which was far lower than the average index in tumor specimens (2.05).

Phosphorylated S6K

S6K was activated in 15 of 28 tumors, including 4 of 11 SCC specimens and 11 of 15 AC specimens. The average p-S6K indices in SCC and ACC were 1.55 and 2.20, respectively.

Pathobiologic and Clinicopathologic Analyses

Specific correlations in staining patterns

On the basis of the known relations among the molecules that mediate intracellular signaling, we evaluated the staining patterns in IHC.

Akt and mammalian target of rapamycin/phosphorylated mammalian target of rapamycin

The expression and activation of mTOR was observed more frequently in the p-Akt-positive group than in the p-Akt-negative group: 74.2% (49 of 66 tumors) of p-Akt-positive tumors versus 64.3% (54 of 84 tumors) of p-Akt-negative tumors exhibited mTOR expression, and 56.1% (37 of 66 tumors) versus 51.2% (43 of 84 tumors) exhibited mTOR activation. However, the differences between the 2 groups were not statistically significant (P = .216 and P = .622, respectively). Furthermore, in the tumors that were positive for both p-Akt and p-mTOR, there was little overlap in the localization of positively stained cells in IHC, and coexpression was observed only occasionally on a cell-by-cell basis.

Mammalian target of rapamycin and phosphorylated mammalian target of rapamycin

p-mTOR-positive tumors also were positive for mTOR, and their staining patterns showed significant overlap: in AC specimens, the p-mTOR-positive acinar structure generally was contained within a broader mTOR-positive area (Fig. 1A-C). Furthermore, all of the tumors that exhibited positive staining for p-mTOR revealed expression of mTOR. Therefore, it is likely that mTOR overexpression is a prerequisite for activation.

Correlation Among Aberrations of EGFR and Phosphorylation of Downstream Molecules

The results obtained from the current study were combined with the results from our previously performed genetic analyses.7EGFR amplification was observed in 4 tumors (3 SCC and 1 AC) (Table 2). Among these, 3 tumors (2 SCC and 1 AC) exhibited up-regulation of mTOR, and only 1 of those tumors (1 SCC) exhibited mTOR activation in IB. However, this 1 tumor that exhibited mTOR activation did not show S6K activation.

Mutation of EGFR was observed in 5 tumors, none of which exhibited gene amplification. All of these 5 tumors exhibited markedly increased levels of p-Akt and p-mTOR (Table 2). However, activation of p-S6K was observed in only 2 tumors. Therefore, mutation of EGFR is associated with constitutive activation of the Akt/mTOR pathway, but not necessarily with the activation of p-S6K.

In the remaining 19 tumors that did not exhibit EGFR aberrations, mTOR expression was up-regulated in 15 tumors (78.9%) and activated in 11 tumors (57.9%). In those tumors, no specific correlation with the activation of Akt or S6K was observed.

Clinicopathologic Analysis

We statistically analyzed these results with clinicopathologic profiles. First, with respect to histologic subtypes and differentiation, the positive frequencies of mTOR (100%) and p-mTOR (89.7%) in AC specimens were significantly higher than in the other subtypes (48.9%, 30.4%, respectively; P < 0.0001)(Table 1). Moreover, in AC specimens, the frequency of positive staining of p-mTOR was correlated with the grade of histologic differentiation: WD and MD tumor subtypes had positive staining with greater frequency than the PD tumor subtype (98% vs 37.5%; P < .0001). In each tumor, p-mTOR positivity was present in the acinar structure, even in the tumor that was diagnosed as MD.

Second, with regard to lymph node status, a correlation between IHC expression of p-Akt and metastasis was observed for all NSCLC specimens (P = .0279). We observed that this correlation also was statistically significant when analyzed by IB (P = .0370). Among 7 tumors that had lymph node metastasis, the average p-Akt index was 2.64, but 20 tumors that did not exhibit metastasis had an index of 1.73 (Table 3). Thus, Akt phosphorylation was a possible predictive factor of metastasis. However, this correlation was not observed in each histologic subtype (P = .0929 in SCC; P = .0941 in AC). Furthermore, in SCC specimens, lymph node metastasis was observed more frequently in the p-mTOR-positive group (12 of 21 tumors) in IHC compared with the negative group (8 of 32 tumors), and the difference was statistically significant (P = .0235). This correlation also was observed when analyzed by IB: for 3 tumors that had metastasis, the average index was 2.95, compared with 8 tumors without metastasis that had an average index of 1.29 (P = .0114) (Table 3). This correlation between positive p-mTOR results and metastasis was not observed in the other histologic subtypes (P > .999 in AC; P = .1667 in LCC). Finally, positive IHC or protein levels evaluated by IB revealed no significant correlation with tumor classification survival rates (P > .05).

Table 3. Correlations Between Protein Expression/Activation Levels and Clinicopathologic Factors
  Experimental Results 
  Score: No. of Cases   
Clinicopathologic VariableTotal No.012RangeMeanP
  1. NSCLC indicates nonsmall cell lung carcinoma; p-Akt, phosphorylated Akt; IHC, immunohistochemistry; N, lymph node classification; p-mTOR, phosphorylated mammalian target of rapamycin; SCC, squamous cell carcinoma.

All NSCLC       
 p-Akt IHC score       
  N05839145  .0279
  N1/N26229267   
 p-Akt protein index       
  N020   0.89-3.501.73.0370
  N1/N27   1.21-4702.64 
 p-mTOR IHC score       
  N058261517  .0577
  N1/N262171827   
SCC       
 p-mTOR IHC score       
  N0332463  .0235
  N1/N220893   
 p-mTOR protein index       
  N08   1.00-2.771.29.0114
  N1/N23   1.60-3.752.95 

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

It has been noted recently that constitutive PI3-K/Akt or mTOR/S6K activation is involved critically in ovarian and lung carcinomas and in acute leukemia, respectively.12, 21, 26 These observations suggest that aberration of the EGFR-Akt-mTOR-S6K pathway may be critical to a wide spectrum of carcinomas. The results presented here reveal with some precision the role of this pathway in lung carcinomas and offer several different and novel observations.

First, we observed that mTOR protein was expressed widely in NSCLC (79.2%). Second, mTOR was activated more frequently in AC, particularly in tumor cells of the acinar structure. Furthermore, AC specimens that harbored mutation exhibited the activation not only of Akt, as described previously,7, 19 but also of mTOR. However, in contrast to these results obtained with tissue specimens, we did not observe significantly higher levels of p-mTOR in AC-derived cultured cells in our preliminary experiments (data not shown). Therefore, mTOR levels may be enhanced in tumor cells and activated at a higher level preferentially in AC only in vivo. These overall results suggest the novel idea that mTOR activation is involved not only in cell proliferation9, 10 but also in the maintenance (and possibly the morphogenesis) of AC. Although the mechanisms that operate to coordinate these diverse phenomena currently are unclear, mTOR is a candidate for a coordinator of these complex processes.

Third, the current study revealed an infrequent, simultaneous activation of both Akt and mTOR on tissue sections. A previous IHC study of NSCLC in which positive staining was described for p-Akt in 51%, and for p-mTOR in 74% of tumors, concluded that the latter was associated significantly with activation of Akt.21 However, those results are inconsistent with ours: We observed mTOR activation in both the p-Akt-positive group and the p-Akt-negative group without any significant difference. Moreover, our results suggest that frequent coactivation of Akt/mTOR detectable in individual cells is unlikely: Tumors that stained positive for both p-Akt and p-mTOR revealed little overlap in their positively stained areas within the tumor nodule. This is most likely because of the diversity of activation cascades that involve mTOR: eg, although Akt can activate mTOR, other kinases are also capable of phosphorylating mTOR.27 In addition, activation of mTOR by p-Akt may be executed in a transient manner by rapid dephosphorylation; thus, morphologic evidence of simultaneous activation of these kinases rarely is detected in tissue specimens.28 An exception was tumors that exhibited mutation of EGFR, in which the constitutive activation of both Akt and mTOR could be detected. Hence, mTOR is activated constitutively by mutated EGFR through the Akt pathway, but its activation was neither correlated with nor reciprocal to amplification of the EGFR.

Fourth, with regard to subcellular localization, p-mTOR was observed predominantly in the cytoplasm in cells of the acinar structure in AC specimens, but it also was observed occasionally in the nucleus in other specimens. Indeed, mTOR is a nuclear-cytoplasmic shuttling protein, and this shuttling is required for the activation of its target proteins, S6K and 4E-BP1.11, 29 It has been reported that mTOR and S6K shuttle between the cytoplasm and the nucleus and that S6K is activated by p-mTOR in the nucleus and relocalizes to the cytoplasm after messenger RNA translation is initiated.30 During this process, the activation and cytoplasmic relocalization of p-mTOR corresponds with movement of the actin arc; thus, mTOR activation is involved in the organization of actin cytoskeleton and, consequently, cell migration.31, 32 It has been established that this coordinated signaling of mTOR/S6K leading to cytoskeleton reorganization plays a critical role in tissue repair through cell-cell attachment. By this function, mTOR may coordinate the morphogenesis of the acinar structure, in which remarkable mTOR activation was observed.

Finally, we observed a correlation between high p-mTOR expression and lymph node metastasis in SCC. Therefore, in this subset of carcinomas, signaling pathways mediated by p-mTOR may promote cell motility, migration, and, subsequently, metastasis, although it is unclear why this phenomenon was not observed in AC. The correlation between Akt activation and lymph node metastasis was observed when NSCLC specimens were viewed as a whole, as we reported previously.7 However, because this correlation was not observed within each histologic subtype, and because the activation of Akt and mTOR was not correlated exactly, the metastatic capability enhanced by p-Akt and p-mTOR may be regulated by different mechanisms.

From the clinical aspect, despite their potent activity in model systems, inhibitors of mTOR clinically exhibit more modest antitumor activity.33 The underlying mechanism is that constitutive activation of the Akt/mTOR pathway induces upstream feedback inhibition of signaling through the EGFR and that inhibition of mTOR abrogates this feedback inhibition and promotes Akt activation. However, mTOR still may be a possible target of chemotherapy for a defined subset: Although mTOR activation often was observed in AC, its localization was observed almost exclusively in the acinar structure. Therefore, we speculate that mTOR is a critical factor involved in morphogenesis but is not an inducer of differentiation. Moreover, it may be possible to potentiate anti-EGFR therapy in lung carcinoma by using rapamycin.12 Treatment with EGFR-tyrosine kinase inhibitors (TKI) alone reportedly suppressed the activity but increase the levels of EGFR expression and subsequently activated its downstream signaling mediators in NSCLC cells.34 Therefore, combined treatments involving EGFR-TKI and rapamycin could abrogate TKI-induced downstream activation and rapamycin-induced upstream activation by each other; indeed, a clinical investigation of this strategy is underway.14 Alternatively, the combined inhibition of activated mTOR together with another upstream molecule that, otherwise, is down-regulated by activated mTOR, may be of greater effect than either treatment alone. In this sense, combined treatments with rapamycin and Akt inhibitors, such as lactoquinomycin and KP372-1, could be novel regimens.33, 35, 36 The current results could provide a rationale for tailoring combination therapies for lung cancer, especially AC, using inhibitors of EGFR signaling and rapamycin.2

In conclusion, the current study on lung carcinomas provides a detailed description of the proteins downstream of EGFR, in particular, those in the mTOR pathway, and the results have several clinicopathologic implications. mTOR exhibited dual roles: First, the mTOR protein frequently is activated and constitutes part of the characteristic profile of AC. Second, activated mTOR may be involved in metastasis in SCC. Together, these results suggest that mTOR plays multiple roles in promoting cancers and, thus, indicate that the mTOR inhibitor rapamycin may be a useful as a molecularly targeted therapy in combination with inhibitors of EGFR signaling in patients with lung carcinoma.

Conflict of Interest Disclosures

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References

Supported by the Japanese Ministry of Education, Sports, Science, and Culture grant C-20590351 (Y.O.) and grant C-19590342 (A.O.)

References

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Conflict of Interest Disclosures
  7. References