Prognostic Significance of N-Glycolyl GM3 Ganglioside Expression in Non-Small Cell Lung Carcinoma Patients: New Evidences

The prognostic role of N-glycolyl GM3 ganglioside (NeuGcGM3) expression in non-small cell lung carcinoma (NSCLC) still remains controversial. In this study, the NeuGcGM3 expression was reevaluated using an increased number of NSCLC cases and the 14F7 Mab (a highly specific IgG1 raised against NeuGcGM3). An immunohistochemical score integrating the percentage of 14F7-positive cells and the intensity of reaction was applied to reassess the relationship between NeuGcGM3 expression, some clinicopathological features, and the overall survival (OS) of NSCLC patients. The double and the triple expression of NeuGcGM3 with the epidermal growth factor receptor (EGFR) and/or its ligand, the epidermal growth factor (EGF), were also evaluated. NeuGcGM3 expression correlates with both S-Phase fraction (p = 0.006) and proliferation index (p = 0.000). Additionally, NeuGcGM3 expression was associated with a poor OS of patients in both univariate (p = 0.020) and multivariate (p = 0.010) analysis. Moreover, the double and/or the triple positivity of tumors to NeuGcGM3, EGFR, and/or EGF permitted us to identify phenotypes of NSCLC with a more aggressive biological behavior. Our results are in agreement with the negative prognostic significance of NeuGcGM3 expression in NSCLC patients. However, standardization of techniques to determine the expression of NeuGcGM3 in NSCLC as well as the implementation of a universal scoring system is recommended.


Introduction
Lung cancer is one of the most frequent cancers in the world and usually has a very poor prognosis. There are two main forms of the disease but non-small cell lung cancer (NSCLC) represents about 80-85% of these tumors [1,2]. Consequently, numerous studies are currently focusing on the selection of newer biological and molecular prognostic factors as a potential complement of TNM (tumor, node, metastasis) staging system [3][4][5]. In line with this, unusual glycosylated gangliosides have been identified by immunohistochemistry (IHC) in NSCLC, also becoming attractive targets for immunotherapy [6,7].
The aberrant expression of N-glycolyl GM3 ganglioside (NeuGc 2-3Gal 1-4Glc -Cer) (NeuGcGM3) in NSCLC, by mean of 14F7 Mab reactivity, using frozen [8] and formalinfixed and paraffin-embedded (FFPE) [9,10] tissues, was previously demonstrated. This Mab is a highly specific IgG1 raised against NeuGcGM3 that does not cross-react with other NeuGc-containing gangliosides [11]. In addition, the expression of NeuGcGM3 was also evidenced by immunohistochemistry and TLC-immunostaining in FFPE samples with GMR8 Mab [7]. GMR8 Mab is an IgM that reacts with an epitope shared by more than one of the NeuGc-containing gangliosides [12]. In line with this, at least another NeuGccontaining ganglioside, GD1a (NeuGc), was recognized by GMR8 Mab in NSCLC tissues, although NeuGcGM3 was the predominant ganglioside identified in these samples [7].
Additionally, the relevance of NeuGcGM3 in cancer progression and its capability of modulating CD4 expression on T cells have been published [13,14]. Recently, the expression of NeuGcGM3, as shown as 14F7 Mab staining, has been associated with more aggressive disease in colon adenocarcinoma [15]. In NSCLC patients, the expression of NeuGcGM3 inhibits dendritic cells (DCs) differentiation, maturation, and migration leading to tumor-induced DCs suppression [9]. However, at present, the prognostic role of NeuGcGM3 expression in NSCLC still remains controversial [7,9,10].
For this reason, in this study the prognostic value of the 14F7 Mab reactivity was reevaluated using an increased number of Cuban NSCLC patients as well as an immunohistochemical score, which integrate the percentage of positive cells and the intensity of reaction. Additionally, the impact of the simultaneous expression of NeuGcGM3, epidermal growth factor receptor (EGFR), and/or epidermal growth factor (EGF) was assessed in the overall survival of these patients. A study of the antigen recognized by 14F7 Mab in FFPE tissues was also included.

Monoclonal Antibodies.
We used the 14F7, a murine IgG1 highly specific against the N-glycosylated variant of GM3 ganglioside [11], the ior egf/r3 (anti-human EGFR) [16], and the CB-EGF1 (anti-human EGF) [17]. 14F7 and ior egf/r3 Mabs were produced by the Center for Molecular Immunology (Havana, Cuba), while CB-EGF1 was kindly provided by the Center for Genetic Engineering and Biotechnology (Havana, Cuba). Additionally, the anti-human EGFR Mab (Dako EGFR pharmDx kit, K1494), the negative control included in this kit, the P3 Mab, an IgM that recognize NeuGcGM3, and other NeuGc-containing gangliosides as well as sulfated glycolipids [6] were employed.

Tissue Specimens.
A total of 90 formalin-fixed and paraffin-embedded (FFPE) tissues with diagnosis of NSCLC were received from the pathology department of both Hermanos Ameijeiras General Hospital and the National Institute of Oncology and Radiobiology, after approved consent by the institutional ethical committees. Specimens from patients with stages I-IIIA were surgically resected by means of wedge biopsy, lobectomy, or pneumonectomy before the standard radio-chemotherapeutic regimens, while samples from patients with stages IIIB-IV were obtained during the clinical autopsy.

Lipid Extraction.
Paraffin blocks were melted by heating in a 70 ∘ C oven for 4 h. The tissue samples were maintained in xylene for 1 h at 70 ∘ C and then at room temperature (3 × 1 h). The tissues were rehydrated in graded ethanol series (100%, 100%, 90%, and 70%) for 1 h each. Finally, they were washed in tap and distilled water for 30 and 15 minutes, respectively. For the lipid extraction, the method as it was previously described by Bligh and Dyer (1959) with minor modifications was used [18]. Briefly, one of two fragments of NSCLC samples (1-2 g) was mechanically homogenized. Then, 3.75 mL of a mixture chloroform/methanol (1/2) was added and vortexed for 30 minutes. Afterward, 1.25 mL of chloroform and 1.25 mL of distilled water were added and the preparation was mixed during 1 minute in each step. The samples were centrifuged, the upper phase was discarded, and finally the lower phase was collected with a Pasteur pipette. After evaporation, the lipid extract (lower phase) was redissolved in 20 L of chloroform/methanol (2/1).

Mass
Spectrometry. The gangliosides fractions extracted from FFPE tissues were analyzed by matrix-assisted laser desorption/ionization-time of flight (MALDI TOF) mass spectrometry. MALDI mass spectra were obtained using an AXIMA Performance MALDI TOF/TOF mass spectrometer (Shimadzu Biotech) with a 337 nm N 2 laser using delayed extraction, in positive reflectron mode. The glycolipid extracts were resuspended in 10 L of 50% methanol, from which 1 L was mixed with an equal volume of a saturated solution of 2,5-dihydroxybenzoic acid (DHB) (Sigma-Aldrich, 85707) in 50% acetonitrile (Sigma-Aldrich, 34851) on top of the MALDI target plate. Samples were allowed to dry off at room temperature and were recrystallized through the addition of 1 L of ethanol. Two hundred laser shots were accumulated for each mass spectrum, with a 50 Hz laser repetition rate. In this analysis, signals between m/z 1.000 and 2.000 were collected. The data was processed using MALDI-MS software from Axima Biotech Launch pad software pack.
The spectra were externally calibrated against Proteo-MassTM peptide MALDI-MS calibration kit (Sigma-Aldrich, MSCAL2-1KT). The peaks corresponding to NeuAcGM3 and NeuGcGM3 were confirmed by comparison with purified samples of both gangliosides.
2.6. Immunohistochemical Staining. Five-micrometer serial sections from each block were obtained in a micrometer (Leitz, 1512) and they were mounted on plus slides (Dako, S2024). All sections were attached to the slide by heating in a 60 ∘ C oven for 1 h. Afterward, the slides were dewaxed in xylene and rehydrated in graded ethanol series in the usual way. The samples were maintained in tap water until they were stained.
The immunolocalization of NeuGcGM3 ganglioside was performed as it was previously described in [10] with some modifications. Briefly, the slides were incubated with 14F7 Mab in a humid chamber for 1 h at room temperature followed by the labeled streptavidin biotin (LSAB) two steps' system (Dako, K0690) both for 30 minutes at room temperature. The enzymatic activity was visualized with 3,3diaminobenzidine (DAB) substrate chromogenic solution (Dako, K3465) and the tissues were counterstained with Mayer's Hematoxylin (Dako, S2020). Concerning the evaluation of both EGFR and EGF tissue antigens, the procedure as it was previously described in [19] was used.

Immunohistochemical Evaluation.
The tissue expression of the NeuGcGM3, EGFR, and EGF in both cell membrane and cytoplasm of epithelial malignant cells was evaluated for percentage of positive cells and intensity of reaction. The most representative regions of each section were selected and the percentage of tumor cells showing immunostaining in them was estimated (0-100%) using the 10x objective lens. The intensity of reaction of each sample was judged as negative (0), weak (1), moderate (2), and strong (3). The final results were considered according to two observers' agreement (Charles E. Rengifo and Rancés Blanco).

Generation of Integrated IHC Scores.
The percentage of positive cells and the intensity of reaction were multiplied for each specimen, resulting in a score ranging from 0 to 300. These scores ranges were grouped as follows: 0 (0), 1 (1 to 100), 2 (101-200), and 3 (>200) to generate an integrated score ( -score). Subsequently, the -scores were grouped in this manner: low expression (scores < 150) and high expression (scores ≥ 150) to create the final score ( -score).

DNA Content, S and G 2 /M Phases Measurements.
The flow cytometry methodology using nuclei isolated from formalin-fixed and paraffin-embedded tissues was used as it was previously described in [5]. The percentage of cells in S and G 2 /M phases of cell cycle was calculated for each sample based on DNA histograms. The S-Phase fraction (SPF) values were divided into two groups: a low-SPF group defined as having SPF less than 20% of proliferating cells and a high-SPF group with SPF greater than 20%. The proliferation index was calculated as the sum of cells in S and G 2 /M phases of cell cycle and was scored as it was previously described for SPF.

Statistical
Analysis. The relationships between 14F7Mab immunoreactivity and clinicopathological variables were analyzed using the Chi-square test. Correlations were assessed by Spearman ranks correlation coefficients. Survival distribution was estimated by the Kaplan-Meier method and compared with the log-rank test. For multivariate analysis, the Cox regression model was used to identify independent prognostic factors for overall survival (OS). The model included all variables and the backward stepwise elimination as the algorithm of selection was applied. OS was measured from the date of surgery to death for any cause or last follow-up (5 years) and was calculated for all patients. A value < 0.05 was considered statistically significant. Statistical analysis was carried out using SPSS (version 15.0; SPSS Inc., Chicago, IL). Table 1 shows a summary of patient characteristics and some pathological features. The gender ratio was close to 2 : 1 in favor of males, with a mean age of 57.4 ± 10.6 years. In general, the overall rate of NSCLC patients was 67.8% (61/90) and the median overall survival of this population at 5 years was 46.2 months (ranging from 0.7 to 67.1).

NeuGcGM3 Detection by Immunohistochemical and Biochemical Methods.
Firstly, the specificity of the 14F7 Mab immunoreaction was confirmed with sialidase (Clostridium perfringens) treatment of NSCLC sections. The enzymatic digestion of tissues did not alter the level of EGFR expression, while the reactivity of P3 Mab was only reduced with this treatment (Figure 1). The staining with 14F7 Mab was significantly removed after sialidase digestion of tissue sections. No reaction was evidenced in NSCLC samples incubated with the negative control. TLC-immunostaining of gangliosides fraction extracted from 3 different histological subtypes of NSCLC (epidermoid carcinoma, adenocarcinoma, and large cell carcinoma) that were positive for NeuGcGM3 in the immunohistochemistry study was performed. Both NeuAcGM3 and NeuGcGM3 gangliosides were evidenced in all NSCLC samples (Figure 2(a)). However, only NeuGcGM3 was recognized by 14F7 Mab, while no reaction was evidenced with NeuAcGM3 ( Figure 2(b)). The expression of these gangliosides after formalin fixation and routine histological procedures was also confirmed by mass spectrometry (Figure 3). NeuAcGM3 was the major ganglioside detected in NSCLC samples using this technology, although a little amount of NeuGcGM3 was also evidenced.   index of cell proliferation ( = 0.006 and = 0.000; Fisher exact test, resp.) ( Table 4). When cases were analyzed independently, according to the intensity of reaction or the percentage of positive cells (data not shown), no significant associations with clinicopathological features were obtained. Nevertheless, higher levels of 14F7 Mab positive cells correlated with both S-Phase fraction and index of cell proliferation ( = 0.028 and = 0.000; Fisher exact test, resp.), while for the intensity of reaction only a significant association ( = 0.037; Fisher exact test) was obtained with the proliferation index.

Survival Analysis.
The results of univariate and multivariate survival analyses at 5 years are summarized in Table 5. Univariate analysis showed that lymph node metastasis ( = 0.000), other sites metastasis ( = 0.000), stage ( = 0.000), disease recurrence ( = 0.000), and the expression of NeuGcGM3 ( = 0.020) were significant prognostic factors for OS. Patients with high level of NeuGcGM3 expression had significantly impaired OS compared to those with low level by mean of the -score (58.5% versus 81.1%; = 0.001; Fisher's exact test). Five-year survival probabilities in NSCLC patients with higher levels of NeuGcGM3 expression were 3.4 times lower than those with reduced levels of this ganglioside. Kaplan-Meier curves according to 14F7 reactivity are represented in Figure 4(a). Among the studied variables, gender ( = 0.047), tumor size ( = 0.003), other sites metastasis ( = 0.029), disease recurrence ( = 0.001), and NeuGcGM3 expression ( = 0.010) were independent prognostic factors on multivariate analysis. No association was obtained when the intensity of reaction or the percentages of positive cells were analyzed ( = 0.134 and = 0.113; log-rank test, resp.).    (Figure 4(b)). But no association with the OS of patients was evidenced when NeuGcGM3 + /EGF + and NeuGcGM3 + / EGF − phenotypes of NSCLC were compared (Figure 4(c)).

Discussion
Despite recent advances in the diagnosis and treatment, lung cancer still remains the leading cause of death from cancer worldwide and one of the 10 leading causes of death from all causes [21]. Usually patients with NSCLC have a poor prognosis because most of them present with advanced or metastatic disease at the time of diagnosis [22]. Even after early diagnosis and surgical resection with curative intent, in stages I-III NSCLC patients, recurrent disease or distant metastasis results in a 5-year survival rate of <50% [23][24][25]. It has been estimated that only 10-15% of NSCLC patients will ultimately be cured [22]. In this sense, a number of independent prognostic factors for survival in NSCLC patients have been evaluated, including disease stage, performance status, gender, age, histology, haemoglobin level, some primary tumor characteristics such as tumor size or local extension, and mediastinal neoplastic infiltration [22].
In this study, we showed the prognostic significance of gender, tumor size, the occurrence of lymph node and other sites metastasis, stage, and disease recurrence in NSCLC patients. However, only the existence of other sites metastasis and the disease recurrence were identified as prognostic factor in both univariate and multivariate analyses. In previous reports, TNM (tumor, node, metastasis) staging system [26], disease stage, and performance status at time of diagnosis have been considered among most noticeable prognostic factors in these patients [22]. Nevertheless, the pathology-based TNM stage classification has been considered to provide imprecise information about the survival rates [27]. Since outcomes can be different even among patients with the same disease stage, numerous studies are currently focusing on the evaluation of other biological and molecular prognostic factors as a potential complement of TNM staging system [24,25,28].
GM3 is the major glycosphingolipid expressed in both tumor and nontumor tissues from human lung cancer patients [29]. In non-small cell lung cancer (NSCLC), an increased expression of GM3 and GM3 synthase (sialyltransferase-I or SAT-I) mRNA has been detected. However, the cDNA sequences of the cytidine monophospho-N-acetylneuraminic acid hydroxylase (CMAH) suggest an inactivating mutation, similar to that in normal tissues [7]. In this sense, these results suggest the existence of an alternative pathway for NeuGc-containing gangliosides, different from the normal pathway that is mediated by CMAH enzyme activity [30]. In line with this, Yin et al. reported a hypoxic culture induced NeuGc-containing ganglioside expression on human cells as a possible candidate for the alternative pathway because a hypoxic culture markedly induced mRNA for the sialic acid transporter, sialin [31].
In this study, the expression of NeuGcGM3 in NSCLC samples after the routine histopathological procedures was confirmed using both immunohistochemistry after sialidase digestion and TLC-immunostaining with 14F7 Mab. However, GM3 was the major ganglioside detected in this kind of samples, similar to previous report [7]. The tissue expression of NeuGcGM3 was confirmed by mean of mass spectrometry of lipid fraction extracted from FFPE samples. NeuGcGM3 is not completely extracted from NSCLC tissues after formaldehyde fixation and the subsequent histological procedures [8]. In a previous work, our group published the NeuGcGM3 ganglioside biochemical measure in breast cancer using frozen tissues and mass spectrometry [32]. However, to our knowledge, this is the first report of the NeuGcGM3 expression in FFPE tissues using mass spectrometry.
Previously, Van Cruijsen et al. [9] and Blanco et al. [10] reported that NeuGcGM3 expression is associated with a favorable survival of NSCLC patients. However, in recent studies Hayashi et al. [7] obtained opposite data, considering the reduced number of cases used by Blanco et al. responsible for these differences ( = 28). Nevertheless, Van Cruijsen et al. used 165 patients in their research [9], which is a larger sample size compared with the rest of studies [7,10]. According to our experience, other factors such as the immunohistochemical analysis (membrane and/or cytoplasmic staining), the IHC score, the specificity of Mabs, and the differences in the histological type proportion could be also responsible for these controversial results.
Here, the reactivity of 14F7 Mab in both cell membrane and cytoplasm of malignant cells was analyzed, as it was previously described [10]. However, in previous reports the expression of NeuGcGM3 was only considered in the cytoplasm [9] or in plasmatic membranes [7] of malignant cells, excluding part of Mabs staining. Moreover, the expression of NeuGcGM3 was scored ( -score) by mean of the integration of both percentage of 14F7-positive cells and the intensity of reaction. It is known that any scoring decision could directly be influenced by bias because the analysis of IHC assays is visually judged [33]. Nevertheless, the introduction of the score systems in the clinical practice had permitted us to reduce variability, particularly for markers that aim to select patients for specific treatments [34].
Interestingly, NSCLC patients displaying higher levels of NeuGcGM3 expression had a significantly poorer overall survival than those with lower levels of this ganglioside.
It is the first report describing the relation between the 14F7 Mab reactivity and a more aggressive biological behavior of NSCLC. Comparable results were recently obtained in colon adenocarcinoma by mean of a similar integrated IHC score [15]. Notably, when only the intensity of reaction or the percentage of 14F7-positive cells was analyzed, in general, no significant associations with clinicopathological features were noted [15], resembling the present research. In this sense, our data permit us to suggest the use of an IHC score integrating both parameters, although further studies are needed to validate these findings.
On the other hand, increasing levels of NeuGcGM3 expression correlated with higher S-Phase fraction and proliferation index measured by flow cytometry. It is known that the upregulation of EGFR expression by EGF is considered a mechanism that promotes the development and progression of lung tumors [35,36]. Additionally, GM3 diminishes the cancer cell proliferation inhibiting of EGFR tyrosine kinase [37], modulating the expression of cell cycle regulation proteins [38][39][40]. However, NeuGcGM3 is not able to inhibit the EGFR tyrosine kinase as compared with GM3 [7]. Interestingly, the triple expression of NeuGcGM3 with EGFR and EGF was detected in the 20.0% of studied patients. In line with this, the aberrant expression of NeuGcGM3 in NSCLC probably causes a disruption in the mechanism of EGFR tyrosine kinase inhibition by GM3, permitting an uncontrolled EGFR system activation mediated by the ligand EGF.
In addition, the simultaneous expression of EGFR and its ligands in tumor and adjacent lung tissues was associated with lower overall and relapse-free survival in NSCLC patients [41]. Here, the dual expression of EGFR and EGF was evidenced in 58.7% of NSCLC samples, while the double expression of NeuGcGM3 and EGFR or EGF was detected in 65.0% and 77.5% of these patients, respectively. Outstandingly, patients with EGFR + /EGF + , NeuGcGM3 + / EGFR + , and NeuGcGM3 + /EGFR + /EGF + phenotypes of NSCLC displayed poor overall survival rates as compared with those with EGFR + /EGF − , NeuGcGM3 + /EGFR − , and NeuGcGM3 − /EGFR + /EGF + , respectively. The dual expression of NeuGcGM3 and EGFR in a variety of human malignant tumors, including NSCLC, was evidenced [42]. However, to our knowledge, it is the first report regarding the double and the triple expressions of NeuGcGM3 with EGFR and/or EGF and its relationship with the overall survival of NSCLC patients.
Finally, differences in predominant histological type might also lead to differences in results. In the first two studies using the 14F7 Mab [9,10] the proportion of squamous subtype and adenocarcinoma was major in favor of squamous carcinoma, while in the third report an opposite relation was obtained [7], resembling the present work. Lung adenocarcinoma had been reported to be the highest levels of SAT-I mRNA expression as compared with epidermoid carcinoma and the rest of tumors [29]. Interestingly, in this study, a statistically significant difference was obtained when the reactivity of 14F7 Mab was compared in function of the histological subtype of tumors. But no differences in NeuGcGM3 expression were evidenced between epidermoid carcinoma and adenocarcinoma (data not shown). In this sense, experiments using increasing number of these tumors of the lung should be designed.

Conclusions
In summary, in this paper the expression of NeuGcGM3 in NSCLC is reported for the first time using mass spectrometry and TLC-immunostaining with 14F7 Mab. Interestingly, high levels of NeuGcGM3 expression, by means of an IHC integrated score ( -score), correlate with the proliferation index and with a poor overall survival of NSCLC patients. These results support the use of a universal scoring system to assess the prognostic value of NeuGcGM3 expression in NSCLC patients. Moreover, some NSCLC phenotypes with poor overall survival rates based on the simultaneous expression of NeuGcGM3 with EGFR and/or EGF were identified. Our data permit us to consider the immunohistochemical detection of these tumor-associated antigens as a useful complement of established prognostic factors in NSCLC. Moreover, our findings support the potential use of combined passive and active immunotherapy in tumors overexpressing NeuGcGM3, EGFR, and/or EGF.