The expression of Bcl-2 family proteins differs between nonsmall cell lung carcinoma subtypes




Proteins of the Bcl-2 family play a key role in the control of apoptosis and carry out both proapoptotic and antiapoptotic functions. However, with the exception of Bcl-2 itself, little is known about the expression of these potentially critical proteins in nonsmall cell lung carcinoma.


Immunohistochemistry was used to study the expression of Bcl-2 and 6 other Bcl-2 family proteins in a pilot series of 41 archival nonsmall cell lung carcinoma specimens (19 adenocarcinomas and 22 squamous cell carcinomas).


Overexpression of the apoptosis inhibitors Bcl-2 and Bcl-XL was observed in 10 of 41 samples (24%) and in 11 of 41 samples (27%), respectively. Loss of expression of proapoptotic proteins was observed as follows: Bak, 24 of 41 samples (59%); Bad, 21 of 41 samples (51%); Bid, 20 of 41 samples (49%); Bax, 14 of 41 samples (34%); and Bim/Bod, 2 of 41 samples (5%). Statistically significant differences in expression between adenocarcinoma samples and squamous cell carcinoma samples were observed for Bcl-XL (overexpression in 11 of 19 adenocarcinomas [58%] vs. 0 of 22 squamous cell carcinomas [0%]; P < 0.001) and for Bad (loss of expression in 5 of 19 adenocarcinomas [26%] vs. 16 of 22 squamous cell carcinomas [73%]; P = 0.004).


Although this was only a pilot study, the results revealed significant differences in the expression of apoptosis-related proteins both between individual samples of nonsmall cell lung carcinoma and between the two main histologic subtypes. Such differences may play a role in the development of lung tumors; and, if it is found that these differences are of clinical importance, then it may be required to regard nonsmall cell lung carcinoma subtypes as separate entities rather than as one disease. Cancer 2005. © 2005 American Cancer Society.

There are two main histologic subtypes of nonsmall cell lung carcinoma (NSCLC): adenocarcinoma (AC) and squamous cell carcinoma (SCC). Compared with other solid tumors, such as colorectal carcinoma, relatively little is known about the genetic defects or pathways of events that lead to the development of NSCLC and whether specific events are involved in the pathogenesis of either histologic type.

The Bcl-2 family of proteins plays a critical role in the control of apoptosis and consists of both proapoptotic and antiapoptotic members. Because of their apoptotic role, these proteins are of great importance in tumor development; however, with the exception of Bcl-2 itself, little is known about the expression of these key proteins in human malignancies.

Both Bcl-2 and Bcl-XL are located mainly in the outer mitochondrial membrane. They exert an antiapoptotic function at least partly by preventing the release of cytochrome c into the cytoplasm. Bcl-XL also interacts with Apaf-1 and inhibits caspase 9 activation.1 Overexpression of antiapoptotic proteins would be expected to prolong cell survival and eventually allow neoplastic transformation. Paradoxically, Bcl-2 overexpression has been associated with a better prognosis in a recent meta-analysis of 21 studies that reported on Bcl-2 expression in NSCLC.2 The reasons for this are not understood fully, although there is a suggestion that Bcl-2 may play an inhibitory role in the hematogenous metastatic process.2

The Bax, Bak, Bad, Bid, and Bim/Bod proteins promote apoptosis. Bax and Bak oligomerize in the outer mitochondrial membrane and allow the release of cytochrome c, which leads to the activation of caspase 9.3 Bid is cleaved by caspase 8 to form truncated Bid (tBid), which translocates to the mitochondria and induces the Bax-dependent and Bak-dependent release of cytochrome c. It is believed that Bad and Bim/Bod act by binding to and inactivating antiapoptotic Bcl-2 family proteins.3

We used immunohistochemistry to assess the frequency of overexpression or loss of expression of 7 Bcl-2 family proteins (Bcl-2, Bcl-XL, Bax, Bak, Bad, Bid, and Bim/Bod) in a pilot series of resected NSCLC tumors. The objective of this study was to identify which proteins may be important in the development of NSCLC, especially those that may be specific to a particular histologic subtype. To our knowledge, this is the first report of the frequency of Bad, Bid, and Bim/Bod expression in primary NSCLC.


Sample Selection

Forty-one samples of NSCLC (AC and SCC subtypes) were selected from the histopathology archive at the Hull Royal Infirmary (Hull, United Kingdom). All samples were taken from patients who underwent resection for operable (early-stage) NSCLC during 1999, and local ethical approval was granted for the study. Nineteen samples were classified histologically as AC (9 poorly differentiated AC and 10 moderately differentiated AC), and 22 samples were classified as SCC (3 poorly differentiated SCC and 19 moderately differentiated SCC). A representative block of formalin fixed, paraffin embedded tissue was selected for each sample. All blocks contained adjacent normal lung tissue to serve as an internal control.


Immunohistochemical assays were carried out essentially as described previously.4 Briefly, 4-μm sections of tissue were cut onto SuperFrost Plus slides and dried overnight at 37 °C. Sections were dewaxed in Histoclear (National Diagnostics, Hull, United Kingdom) and rehydrated in alcohol before blocking endogenous peroxidase by incubating in 400 mL methanol containing 8 mL hydrogen peroxide (30% volume/volume). Antigen retrieval was carried out by boiling slides in a pressure cooker for 3 minutes at 15 psi in 1500 mL distilled water with 15 mL Antigen Unmasking Solution (Vector Laboratories Ltd., Burlingame, CA). Nonspecific protein was blocked by incubation with 1 × casein (Vector Laboratories Ltd.) diluted in Tris buffered saline (TBS) for 10 minutes, and endogenous biotin and avidin were blocked by successive, 15-minute incubations with avidin and biotin, respectively (Avidin Biotin Blocking Kit; Vector Laboratories Ltd.). The primary antibody was diluted in 0.2 × casein in TBS and applied for 2 hours at room temperature. Antibody suppliers and dilutions are given in Table 1. The primary antibody was omitted from the negative control. Antibody detection was carried out using the Duet Kit (Dakocytomation Ltd., Ely, United Kingdom) according to the manufacturer's instructions. Finally, sections were counterstained with hematoxylin, dehydrated in alcohol, and mounted with Histomount (National Diagnostics).

Table 1. Antibody Suppliers and Dilutions Used
AntibodySupplier (catalog no.)Dilution
Bcl-2Serotec (MCA1550)1:50
Bcl-XL, clone 7D9Neomarkers (MS-1334-P0)1:75
Bax, clone 2C8Neomarkers (MS-1335-P0)1:100
BakNeomarkers (RB-1520-P0)1:50
Bim/BodNeomarkers (RB-1612-P0)1:100
Bid, clone 7BD Biosciences (611528)1:100
Bad, clone 48BD Biosciences (610391)1:50

Scoring of Sections and Statistical Analysis

Sections were scored as either positive or negative by two independent investigators. Discrepancies in the results were resolved by consensus after reassessment of the slide by both investigators. For Bcl-2 and Bcl-XL, a section was considered positive if > 10% of tumor cells were stained, because overexpression of these proteins inhibits apoptosis and, thus, is detrimental.5, 6 For Bax, Bak, Bad, Bid, and Bim/Bod, in which the loss of expression is detrimental, sections were scored as negative if expression was lost in > 50% of tumor cells.7 Significant differences in staining patterns between AC and SCC samples were evaluated by chi-square test using SPSS software (version 11; SPSS Inc., Chicago, IL).


Expression of Bcl-2 Family Proteins in NSCLC

Staining with all antibodies was observed in the cytoplasm of cells when present. Bcl-2 was not expressed in normal lung tissue, whereas weak staining was observed in some samples for Bcl-XL (Table 2). Normal lung tissue did express Bax, Bak, Bad, Bid, or Bim/Bod, with strong staining for Bid and Bim/Bod (Table 2), and was used as an internal positive control for these antibodies. Lymphocytes were used as an internal positive control for Bcl-2 and Bcl-XL. There was heterogeneity in the percentage of tumor cells that showed expression of each protein with the exception of Bim/Bod, which showed mainly homogeneous expression throughout tissue sections.

Table 2. Expression of Antiapoptotic and Proapoptotic Proteins in Normal Lung Tissue and Lymphocytes
ProteinNormal lung tissueLymphocytes
  1. −: No staining; +/−: weak staining: +: positive for staining; ++: strong staining.


The number of samples that demonstrated overexpression (Bcl-2 and Bcl-XL) and loss of expression (Bax, Bak, Bid, Bad, and Bim/Bod) are summarized in Table 3. Bcl-2 and Bcl-XL were overexpressed in 24% and 27% of samples, respectively. Loss of expression of Bim/Bod and Bax was observed in 5% and 34% of samples, respectively. Bak, Bad, and Bid showed more frequent loss of expression, with 59%, 51%, and 49% of samples scored as negative, respectively.

Table 3. Numbers and Percentages of Tumors Scored as Positive for Bcl-2 and Bcl-XL and as Negative for Bax, Bad, Bak, Bid, and Bim/Bod
ProteinNo. of tumors (%)P value (significance)a
All tumorsACSCC
  • AC: adenocarcinoma; SCC: squamous cell carcinoma.

  • a

    Chi-square analysis was used to test for differences in frequency of expression between AC and SCC samples.

  • b

    Significant result.

Antiapoptotic (% positive)    
 Bcl-210/41 (24)2/19 (11)8/22 (36)0.058
 Bcl-XL11/41 (27)11/19 (58)0/22 (0)< 0.001b
Proapoptotic (% negative)    
 Bax14/41 (34)8/19 (42)6/22 (27)0.252
 Bad21/41 (51)5/19 (26)16/22 (73)0.004b
 Bak24/41 (59)9/19 (47)15/22 (68)0.151
 Bid20/41 (49)11/19 (58)9/22 (41)0.220
 Bim/Bod2/41 (5)0/19 (0)2/22 (9)0.282

Significant Differences between Histologic Subtypes

Significant differences in the expression of and Bad were observed between AC samples and SCC samples. Bcl-XL was overexpressed more frequently in AC samples (P < 0.001), and loss of Bad expression was more frequent in SCC samples (P = 0.004) (Table 3). There was a trend toward more frequent overexpression of Bcl-2 in SCC samples; however, this trend did not reach significance (P = 0.058).


The current study was designed to investigate the frequency of overexpression and loss of expression of Bcl-2 family proteins in NSCLC samples using immunohistochemical analysis. With the exception of Bcl-2, there have been few reports to date describing the expression of these proteins in human malignancies. Seven proteins were investigated, including both antiapoptotic (Bcl-2 and Bcl-XL) and proapoptotic (Bax, Bad, Bak, Bid, Bim/Bod) Bcl-2 family proteins. Staining for all proteins was cytoplasmic, as expected.

Antiapoptotic Proteins

The overexpression of Bcl-2 in 10 of 41 samples (24%) is in close agreement with reports published to date. Results of 23%5 and 20%6 have been reported recently in series of 114 and 60 NSCLC samples, respectively. Lai et al.5 studied samples of Stage I–IIIa NSCLC, and Tomita et al.6 included only later-stage disease (Stage IIIa–IV). The similar results reported in those 2 studies suggest that Bcl-2 expression may be independent of stage in NSCLC. Other studies have reported overexpression of Bcl-2 in between 31% and 59% of samples.7–12

The results for expression of Bcl-2 in AC (2 of 19 samples; 11%) and SCC (8 of 22 samples; 36%) also are in close agreement with the results of Lai et al.,5 who reported expression in 15.5% of AC samples and 35% of SCC samples. Although it did not reach significance in the current study (P = 0.058) (Table 3), the trend toward increased overexpression of Bcl-2 in SCC samples compared with AC samples was in agreement with previous reports, which showed that expression of Bcl-2 was associated significantly with this histologic type of NSCLC.8, 10–11 Expression of Bcl-2 has been associated with a better prognosis in patients with NSCLC2 but is also known to increase resistance to chemotherapy; therefore, the clinical relevance of this finding remains to be elucidated.

To our knowledge, there is only one previous study of the frequency of expression of Bcl-XL in NSCLC using immunohistochemistry.12 The authors of that study reported expression in > 30% of tumor cells in 89% of samples, which is in contrast to the expression in 27% of tumor cells obtained in the current study. The reasons for this discrepancy are not apparent, because similar specimens (paraffin embedded tissue) were used in each study, and the higher threshold for positivity (30% rather than 10% used in our study), if anything, would be expected to yield a lower result. A possible reason is the use of polyclonal immune serum in contrast to the monoclonal antibody used here; however, further studies will be needed to resolve this issue. Groeger et al.12 did not report the percentage of positive AC and SCC samples; therefore, it is not possible to confirm the more frequent expression of Bcl-XL in AC samples observed in the current study, which will require investigation in a larger series of samples to verify the result. It is interesting to note that, whereas Bcl-2 was overexpressed more frequently in SCC samples, Bcl-XL (also an antiapoptotic protein) was overexpressed more frequently in AC samples (P < 0.001). Therefore, it is possible that the inhibition of apoptosis (leading to cell survival and neoplastic transformation) is due to overexpression of different antiapoptotic proteins in each histologic type of NSCLC. If this is the case, then these proteins may be useful as diagnostic markers and would be potential specific therapeutic targets.

Proapoptotic Proteins

Loss of Bax expression was observed in 14 of 41 samples (34%). This is in line with the reported underexpression of Bax in 31% of samples11 and is in general agreement with reports by Groeger et al.12 and Caputi et al.,13 which demonstrated the loss of expression of Bax in > 40% of tumor cells in 23% and 27% of samples, respectively. Those reports of relatively infrequent loss of Bax expression suggest that such loss is unlikely to be a major event in the pathogenesis of NSCLC.

We found that the expression of Bak was lost in 24 of 41 samples (59%). This is in general agreement with Borner et al.,7 who showed loss of expression in > 50% of tumor cells in 42% of NSCLC samples. To our knowledge, this is the only previous report of the frequency of expression of Bak in NSCLC. Loss of this proapoptotic protein in approximately 50–59% of NSCLC samples suggests that it may play a role in the development of NSCLC.

Loss of expression of Bad, Bid, and Bim/Bod was observed in 51% (21 of 41 samples), 49% (20 of 41 samples), and 5% (2 samples 41), respectively. To our knowledge, this is the first immunohistochemical study of the frequency of expression of these proteins in primary NSCLC samples. The infrequent loss of Bim/Bod expression implies that this is unlikely to be important in the pathogenesis of NSCLC. In contrast, loss of expression of Bad and Bid in approximately 50% of samples may suggest a role for these proteins in facilitating the development of NSCLC.

A significant difference in the frequency of loss of expression of Bad was observed between AC and SCC, with loss more frequent in SCC samples (P = 0.004). Because there are no previous studies of expression of this protein in NSCLC, this result cannot be confirmed, and further study of a larger number of samples will be needed to verify the result. The more frequent loss of expression of Bad in SCC suggests that this is more likely to be a significant factor in the development of this form of NSCLC rather than in AC, which retained expression in 74% of samples.

Although this was only a pilot study, the current results showed that the Bcl-2 family proteins Bcl-2, Bcl-XL, Bad, Bak, and Bid may be important in the pathogenesis of NSCLC, with Bad (and Bcl-2) more important in SCC and with Bcl-XL more important in AC. These results, especially the differences in expression between AC samples and SCC samples, will require verification in a larger series of matched samples (according to disease stage and degree of differentiation) to confirm the findings and to investigate the basis of the difference. Markers that are useful in distinguishing SCC from AC subtypes of NSCLC also may be useful in separating AC from mesothelioma and may aid in the histologic diagnosis of these diseases. Apoptosis proteins have fundamental cellular roles, and aberrant expression may be related to patient prognosis and response to therapy. These avenues remain to be explored.