Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Lung cancer is the leading cause of cancer death in both men and women, accounting for 30% of all cancer-related mortality in the United States (1). Even though some improvement has been made in clinical management of lung cancer, the current 5-yr survival rate is only about 13% (2). The diagnosis of lung cancer is usually too late and current therapeutic strategies are inadequate. Lung cancer, especially non-small cell lung cancer (NSCLC), remains refractory to all available systematic treatment options (3). Any major impact in the prevention and/or treatment of this disease is largely dependent upon a better understanding of the molecular aspects of the pathogenesis of lung cancer.

Proto-oncogenes and tumor suppressor genes are cellular genes that regulate other genes involved in the control of normal cellular growth and differentiation. These genes code for growth factors, growth factor receptors, kinases which phosphorylate diverse cellular substrates, and DNA binding proteins which regulate expression of numerous cellular genes. Oncogenesis may occur through activation of dominant growth promoting proto-oncogenes (such as Her-2/neu, ras, and myc) where mutation or overexpression of the gene product induces malignancy, and inactivation of growth inhibitory tumor suppressor genes (such as p53 and Rb) where the absence of expression or function of these genes results in malignancy, or both (4, 5). Defects in pathways controlling growth and differentiation can effect the development of malignancy including lung cancer.

One approach to identifying genes that may play key roles in cellular growth control is to focus on transcripts whose expression is low in nondividing cells but is rapidly increased when stimulated by mitogens. Immediate early genes are among the earliest downstream nuclear targets for ligand-receptor interactions. These genes are induced in the absence of de novo protein synthesis and were initially identified as promoters of cell proliferation because their expression is required for quiescent cells to enter the cell-cycle. The best characterized immediate early genes are c-jun and c-fos (6, 7); both genes encode transcription factors that bind to specific regulatory sequences and activate expression of responsive genes. c-jun has a leucine zipper motif, which forms a heterodimer with a variety of proteins including c-fos. The jun-fos protein complex, also known as AP-1, binds to a number of cellular promoters through a common element (5'-TGACTCA-3'). At present, it is thought that these genes have bifunctional activity because they appear to be able to both activate and repress transcription of responsive genes.

Phorbol myristate acetate (PMA), a tumor-promoting phorbol ester, is a potent mitogen that can activate gene expression of a number of genes, including immediate early genes like jun and fos. Recently, it has been reported that PMA can enhance the growth inhibitory activity of negative growth factors like transforming growth factor- (TGF-) in human prostate cancer cell lines (8). PMA has also been shown to increase the steady-state level of TGF-₁ mRNA in different cell types. In addition, it has also been shown that both PMA and TGF-₁ can inhibit mitogenesis through a similar signal transduction pathway involving protein kinase C (PKC) in both mesengial and vascular smooth muscle cells (9).

The differential mRNA display method, originally introduced by Liang and Pardee in 1992 (10), has been succesfully used to isolate genes differentially expressed between normal and cancer cells, and between cells cultured in the presence and absence of growth factors and chemical agents (11). In this method, only a small amount of RNA is required from two different populations of cells that is reverse-transcribed using anchored oligo-dT primers and random 5' oligonucleotide primers. The patterns of amplified cDNA fragments are displayed side-by-side on a polyacrylamide gel, individual bands that indicate differential expression are recovered, reamplified, subcloned, and sequenced. The result is typically a 50-500 bp cDNA sequence which may correspond to the 3' end of a cDNA of a known gene obtained from available DNA data bases.

In this study, differential mRNA display was used to identify PMA-regulated immediate early genes in NSCLC cells and normal human bronchial epithelial (NHBE) cells. One of these genes was identified to be early growth response gene-1 (Egr-1). Egr-1 mRNA expression was found to be regulated similarly by PMA in both NSCLC cells and NHBE cells reaching a maximum level by 1 h before decreasing; in contrast, it was shown that Egr-1 mRNA was differentially regulated by TGF-₁ in these different cell types. Our results suggest that Egr-1 may play different roles in response to mitogens in normal and malignant lung cells.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Cell Culture and RNA Isolation

Human lung cancer NSCLC cell lines (NCI-H157, -H226, -H727, -H838, -H1264, -H1299, and -H2170) were cultured in serum-supplemented medium at 37°C [RPMI-1640 medium (GIBCO/BRL, Grand Island, NY)] containing 10% heat-inactivated fetal bovine serum (GIBCO/BRL) (12, 13). NHBE cells obtained from Clonetics (San Diego, CA) were cultured in bronchial epithelial growth medium supplied with the cells. Then NSCLC cells were split weekly by using trypsin/EDTA (GIBCO/BRL, Rockville, MD). Routinely, the cells showed greater than 90% viability, were Mycoplasma free, and were used when they were in exponential growth phase. The NHBE cells were routinely used without splitting. For treatment of cells with PMA and TGF-₁, cells were washed with phosphate-buffered saline and incubated with RPMI-1640 medium containing 0.1% bovine serum albumin (BSA). PMA was obtained from Sigma (St. Louis, MO). Human TGF-₁ was obtained from R&D Systems (Minneapolis, MN). Total RNA was isolated from NSCLC and NHBE cells by the method of Chirgwin and associates (14), and was treated with RNase-free DNase (GIBCO/BRL).

Differential mRNA Display

Purified total cell RNA (0.2 µg) was reverse transcribed in a 20-ul reaction mixture with Superscript reverse transcriptase (GIBCO/BRL) and oligo-dT primer T₁₁C at 37°C for 60 min. For PCR amplification, 2 µl of reverse transcribed sample was mixed with 18 µl of mixture containing the oligo-dT primer T₁₁C and a random primer 5'-AAGCTTGATTGCC-3' (AP-1). Radioactive [³³P]--dATP was added in the PCR reaction and the reaction mixture was subjected to 40 cycles of PCR using parameters of 94°C for 30 s, 40°C for 2 min, and 72°C for 30 s according to the RNAimage kit (GenHunter Co., Nashville, TN). Radiolabeled PCR products were analyzed by electrophoresis on 6% polyacrylamide gels. The gels were dried and then exposed to Kodak XAR-5 film (Kodak Co., Rochester, NY).

Reamplification of cDNA Fragments

Differentially expressed cDNA fragments were identified and recovered from the gel and reamplified in a 40-cycle PCR using the same pair of primers described previously, T₁₁C and AP-1 (GenHunter Co.). Reamplified cDNA fragments were separated by electrophoresis on a low-melting agarose gel, extracted using the GENECLEAN II KIT (BIO 101, Vista, CA) and the purified cDNA fragments were used as templates for random priming and for TA cloning.

Northern Blot Analysis

Equal amounts of total RNA (10 µg) were electrophoresed on 1% agarose gels containing 0.66 M formaldehyde, transferred to "Nytran" filters (Scleicher and Schuell, Keene, NH), UV-cross-linked and baked for 3 h. Ethidium bromide (33 µg/ml) was included in both the gels and running buffers to visualize the positions of ribosomal RNAs by UV illumination after eletrophoresis. Blots were hybridized with [³²P]-labeled (3,000 Ci/mmol, Dupont, Boston, MA) random-primed cDNA probes at 65°C according to Church and Gilbert (15), and then exposed for various times at 70°C using an intensifying screen. Densitometry of autoradiograms was performed using a scanning laser densitometer.

cDNA Probes

Hybridization was performed using the following cDNA probes: 1.6 kb XbaI-HindIII fragment of human Egr-1, plasmid pOC68, obtained from V. Sukhatme (16, 17); 1.0 kb EcoRI fragment of human c-fos, obtained from T. Curran (18); 0.9 kb XbaI-HindIII fragment of rat TGF-₁, plasmid pRTGF-₁, obtained from S. W. Qian (19).

TA Cloning and Sequencing Analysis

Reamplified cDNA fragments were cloned into the plasmid vector pCRII using the TA cloning kit (Invitrogen, San Diego, CA) and purified by the QIAGEN plasmid kit (QIAGEN, Inc., Chatsworth, CA). The cDNAs were sequenced with T7 and SP6 primers using the dye terminator cycle sequencing core kit and an Applied Biosystems automated sequencer (Perkin Elmer, Foster City, CA). The nucleotide sequences obtained were compared with known sequences by searching the Genbank and EMBL databases with the BLAST (National Center for Biotechnology Information, Bethesda, MD).

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

We have previously shown that PMA is able to induce expression of several genes in lung cancer cells including c-fos and TGF-₁ (20, 21). To examine the effect of PMA on expression of potentially inducible novel genes in NSCLC cells, exponentially growing subconfluent cultures of human lung squamous carcinoma cells (NCI-H157) were treated with PMA (1 µM) for different times from 1 h to 24 h in RPMI-1640 medium containing 0.1% BSA. The optimal amount of PMA added was based on prior studies (20). Total RNA was extracted from these cells and northern blot analysis was performed with a radiolabeled random-primed cDNA probe for c-fos. Within 1 h of addition of PMA, there was a transient 15-fold increase in the level of the 2.2 kb c-fos transcript which decreased by 3 h (Figure 1A). To examine expression of TGF-₁, the blot that was used to investigate expression of c-fos was dehybridized, exposed to film to ensure complete dehybridization, and then hybridized with a cDNA probe for TGF-₁. Within 6 h of addition of PMA, there was an 8-fold increase in the level of the 2.5 kb TGF-₁ transcript which persisted to 24 h (Figure 1B). To examine the regulation of c-fos and TGF-₁ by PMA in normal lung cells, sub-confluent cultures of NHBE cells were incubated with PMA as described for the NSCLC cells. Total RNA was extracted from these cells at the indicated times, and northern blot analysis was performed using a c-fos cDNA probe. Figure 1D shows that the level of expression of c-fos mRNA increased 12-fold by 1 h in NHBE cells following treatment with PMA. When this blot was rehybridized with a cDNA probe for TGF-₁, the level of TGF-₁ mRNA increased 2.5-fold by 3 h and then an additional 2- and 6-fold by 6 and 24 h, respectively, in these cells (Figure 1E). As a control, the gels were stained with ethidium bromide and photographed to ensure that approximately equal amounts of RNA had been applied to the gels (Figures 1C and 1F).

View larger version (60K): [in this window] [in a new window]   Figure 1.   Northern blot analysis of c-fos and TGF-1 mRNAs in NCI-H157 and NHBE cells after treatment with PMA. Total RNA (10 µg) was isolated from exponentially growing subconfluent (panels A and B) NCI-H157 cells and (panels D and E) NHBE cells cultured with 1 µM PMA for 0 h (lanes 1, 6), 1 h (lanes 2, 7), 3 h (lanes 3, 8), 6 h (lanes 4, 9) and 24 h (lanes 5, 10), separated by electrophoresis on a 1% agarose-formaldehyde gel, and transferred to a Nytran filter as described in MATERIALS AND METHODS. Hybridization was performed with (panels A and D) c-fos cDNA probe; (panels B and E) TGF-1 cDNA probe. The blots were exposed for 1 day. Panels C and F show the ethidium bromide staining pattern of the gels showing 18S and 28S rRNA. The blots shown in this and subsequent figures are representative of three separate experiments.

Differential mRNA display analysis was used to identify early response genes regulated by PMA in lung cancer cells. For this purpose, total RNAs were isolated from exponentially growing subconfluent cultures of NCI-H157 raised either in the absence of or presence of two concentrations of PMA. cDNAs were synthesized by reverse transcriptase and amplified using a one-base anchored oligo-dT primer (T₁₁C) as described in MATERIALS AND METHODS. Forty cycles of amplification were performed with an oligo-dT primer (5'-T₁₁C-3') and a random primer (5'-AAGCTTGATTGCC-3'). The amplified radiolabeled cDNA products were size fractionated on denaturing urea-polyacrylamide gels and differentially expressed cDNAs were identified by autoradiography. A representative differential display pattern observed with RNAs isolated from untreated NCI-H157 cells and cells treated with two concentrations of PMA is shown in Figure 2A. Several reproducible bands appeared to represent potential differentially regulated cDNA fragments (Figure 2A, arrows); while the majority of these bands showed a strong signal resulting after treatment with PMA, at least two bands showed a strong signal without treatment with PMA. Among the potential differential cDNAs indicated by arrows, one of approximately 100 bp indicated by the thick arrow had the strongest signal and the most dramatic change (thickness of the band) after PMA treatment. This cDNA fragment was excised from the gel and reamplified by another round of PCR using corresponding 5' and 3' primers.

View larger version (39K): [in this window] [in a new window]   Figure 2.   Differential mRNA display analysis of PMA-affected genes. Panel A: Total RNA was isolated from untreated () (lane 1) or PMA treated (+) (lane 2, 1 µM PMA; lane 3, 5 µM PMA) cultures of NCI-H157 cells and reverse transcribed to cDNA for further mRNA differential display amplification. The amplified cDNA fragments were separated by electrophoresis on 6% DNA sequencing gels and autoradiographed as described in MATERIALS AND METHODS. Thin arrows indicate the position of differentially displayed cDNA bands. The thick arrow indicates the position of the 100 bp PMA-induced cDNA fragment. DNA molecular weight markers are indicated at the right. Panel B: Northern blot analysis of PMA-induced mRNAs in NCI-H157 cells. Total RNA (10 µg) was isolated from exponentially growing subconfluent NCI-H157 NSCLC cells cultured for 1 h with: no addition (lane 1); 1 µM PMA (lane 2); 5 µM PMA (lane 3), and analyzed as before with (panel B) PMA-induced gene cDNA probe; (panel C) c-fos cDNA probe. The blots were exposed for 1 day. Panel D: The ethidium bromide staining pattern of the gel showing 18S and 28S rRNA.

To reconfirm the differential expression pattern of the 100 bp cDNA fragment, the excised reamplified cDNA fragment was used as a probe to hybridize to northern blots containing total RNA from NCI-H157 cells cultured in the presence or absence of PMA for 1 h. Northern blot analysis demonstrated that this cDNA fragment hybridized to a transcript of 3.4 kb in NCI-H157 cells after treatment with two concentrations of PMA; within 1 h of addition of PMA, there was at least a 30-fold increase in the level of this transcript (Figure 2B). Expression of c-fos mRNA was also examined in these cells cultured in the presence or absence of PMA. To examine expression of c-fos, the blot that was used to investigate expression of the 3.4 kb mRNA induced by treatment with PMA was dehybridized, exposed to film to ensure complete dehybridization, and then hybridized with a cDNA probe for c-fos. Hybridization with a cDNA probe for c-fos showed an increase in expression of c-fos mRNA of at least 60-fold after treatment with PMA in NCI-H157 cells (Figure 2C). As a control, the gel was stained with ethidium bromide and photographed to ensure that equal amounts of RNA had been applied to the gel (Figure 2D).

Following northern blot hybridization analysis, the PMA-induced cDNA fragment was subcloned into plasmid DNA and the insert from plasmid DNA was also shown to hybridize to the same size mRNA and exhibit differential expression in PMA-treated NCI-H157 cells (data not shown). It was determined that this plasmid DNA fragment has an insert of 88 bp and sequence analysis confirmed that it contained both 5' (AP-1) and 3' (T₁₁C) primer sequences used for amplification. After DNA sequence homology searches in Genbank database, the sequence was found to be identical to a portion of the 3'-untranslated region of the cDNA sequence of the human early growth response-1 gene (Egr-1) (Figure 3). One mismatch was observed in the random primer region.

View larger version (11K): [in this window] [in a new window]   Figure 3.   Nucleotide sequence of PMA-induced gene. Reamplified PMA-induced gene cDNA fragment was cloned into a TA cloning vector and sequenced using SP6 and T7 primers. The sequence was searched in GenBank data base. Homology of the PMA-induced gene with Egr-1 is indicated. The 5'-untranslated, coding and 3'-untranslated regions of human Egr-1 cDNA are shown.

To determine whether Egr-1 mRNA is also induced by PMA in other squamous NSCLC cell lines, three additional squamous NSCLC cell lines including NCI-H226, NCI-H1264, and NCI-H2170 were cultured in the presence of PMA and analyzed by northern blot hybridization. Figure 4A shows that expression of Egr-1 mRNA increased approximately 20-, 40-, and 40-fold in NCI-H226, NCI-1264, and NCI-H2170, respectively, following treatment with PMA. To determine whether Egr-1 mRNA is induced by PMA in other NSCLC cell lines besides the squamous type, three other NSCLC cell lines representing additional types of NSCLC were examined including NCI-H727 (carcinoid), NCI-H838 (adenocarcinoma) and NCI-H1299 (large cell carcinoma). Exponentially growing subconfluent cultures of these NSCLC cells were treated with PMA for different times from 1 h to 24 h in RPMI-1640 medium containing 0.1% BSA. Total RNA was extracted from these cells at increasing times and northern blot analysis was performed with an Egr-1 cDNA probe. Within 1 h of addition of PMA, there was an increase in expression of Egr-1 mRNA by 60-, 100-, 30-, and 50-fold in NCI-H157, NCI-727, NCI-H838, and NCI-H1299, respectively (Figure 5A); this increase was transient and decreased by 3 h following treatment with PMA. A second increase in Egr-1 mRNA was observed 6 h following treatment with PMA in NCI-H157 and NCI-H727 before decreasing after 24 h to basal levels; this second increase was only about 30% of the level at 1 h in each cell line. No second increase was detected in NCI-H838 or NCI-H1299 following treatment with PMA.

View larger version (47K): [in this window] [in a new window]   Figure 4.   Northern blot analysis of EGR-1 mRNA in squamous NSCLC cells. Panel A: Total RNA (10 µg) was isolated from subconfluent exponentially growing NSCLC cells including NCI-H157, NCI-H226, NCI-H1264 and NCI-H2170 and cultured for 1 h with: no addition () (lanes 1, 3, 5 and 7); 1 µM PMA (+) (lanes 2, 4, 6 and 8). The blots were exposed for 1 day. Panel B: The ethidium bromide staining pattern of the gels showing 18S and 28S rRNA.

View larger version (41K): [in this window] [in a new window]   Figure 5.   Effect of PMA on expression of Egr-1 mRNA in NSCLC cell lines. Panel A: Total RNA (10 µg) isolated from subconfluent exponentially growing NSCLC cell lines NCI-H157, NCI-H727, NCI-H838, and NCI-H1299 cultured with 1 uM PMA for 0 h (lanes 1, 6, 11 and 16), 1 h (lanes 2, 7, 12 and 17), 3 h (lanes 3, 8, 13 and 18), 6 h (lanes 4, 9, 14 and 19) and 24 h (lanes 5, 10, 15 and 20) and analyzed as before with Egr-1 cDNA probe. The blot was exposed for 1 day. Panel B: The ethidium bromide staining pattern of the gel showing 18S and 28S rRNA.

In addition to investigating the effect of PMA on expression of Egr-1 mRNA in NSCLC cells, the effect of TGF-₁ on expression of this transcript was also examined in these cells. Expression of Egr-1 mRNA increased maximally at levels of 60-, 100-, 15-, and 50-fold in NCI-H157, NCI-727, NCI-H838, and NCI-H1299, respectively, by 1 h following treatment with TGF-₁ (Figure 6A). The Egr-1 mRNA expression decreased by 3 h in each cell line with a second increase occurring after 6 h in NCI-H727 which persisted for 24 h. A minor second increase in Egr-1 mRNA also occurred at 6 h in NCI-H157, and decreased to basal level by 24 h.

View larger version (41K): [in this window] [in a new window]   Figure 6.   Effect of TGF-1 on expression of Egr-1 mRNA in NSCLC cell lines. Panel A: Total RNA (10 µg) was isolated from subconfluent exponentially growing NSCLC cell lines cultured with 5 ng/ml TGF-1 for 0 h (lanes 1, 6, 11 and 16), 1 h (lanes 2, 7, 12 and 17), 3 h (lanes 3, 8, 13 and 18), 6 h (lanes 4, 9, 14 and 19) and 24 h (lanes 5, 10, 15 and 20) and analyzed as before with Egr-1 cDNA. The blot was exposed for 1 day. Panel B: The ethidium bromide staining pattern of the gel showing 18S and 28S rRNA.

In addition to investigating the effect of PMA and TGF-₁ on expression of Egr-1 mRNA in NSCLC cells, the effect of these agents on expression of Egr-1 mRNA was also examined in NHBE cells. Figure 7A shows that Egr-1 mRNA was induced maximally (20-fold) at 1 h after PMA treatment before decreasing to 50% of this level by 3 h and to basal level by 24 h. Following treatment with TGF-₁, expression of Egr-1 mRNA increased 2-fold by 1 h and reached a maximal level only after 6 h where expression was 10-fold higher than at the initial time (Figure 7B); expression of Egr-1 mRNA decreased to basal levels by 24 h following treatment with TGF-₁ in NHBE cells.

View larger version (37K): [in this window] [in a new window]   Figure 7.   Effect of PMA and TGF-1 on expression of Egr-1 mRNA in NHBE cells. Total RNA (10 µg) was isolated from subconfluent NHBE cells cultured with panel A: 1 µM PMA; panel B: 5 ng/ml TGF-1 for 0 h (lane 1); 1 h (lane 2); 3 h (lane 3); 6 h (lane 4); 24 h (lane 5). The blots were exposed for 2 days. The ethidium bromide staining pattern showing 18S and 28S rRNA is shown below each blot.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Regulation of cellular growth and proliferation is dependent on a number of gene families including growth factors, growth factor receptors, proto-oncogenes, and transcription factor genes. It has been suggested that some transcription factors play a role in the regulation of normal lung physiology. In the present study, differential mRNA display was used to identify Egr-1 as a gene regulated by PMA in NSCLC cells. In addition, we have shown that expression of Egr-1 mRNA is also regulated by PMA in NHBE cells. The biochemical processes that precede immediate early gene induction after mitogenic stimulation include signal transduction through protein kinase activation. PKC is the cellular receptor for phorbol esters like PMA that can initially activate some PKC isoenzymes and then downregulate them (22, 23). Since PMA has been shown to elicit both growth promotion and growth inhibition in different cells and both higher and lower levels of PKC have been found in cancer cells as compared to normal cells, it has been suggested that PKC may participate in negative feedback of growth. Although the expression of Egr-1 has been most commonly observed after treatment with mitogenic stimuli, a consistent requirement for progression through the cell cycle has not been observed, leaving the role of Egr-1 in control of cell proliferation unclear. Immediate early response genes were initially identified as promoters of cell proliferation because their expression was thought to be required for quiescent cells to enter the cell cycle (24). However, because some of these genes like fos have been shown to be inducible throughout almost all stages of the cell cycle, activation of some of these genes may not be limited to the mitogenic response. Our study shows that Egr-1 mRNA is rapidly induced by PMA in NSCLC and NHBE cells. While maximal expression of Egr-1 mRNA was detected within 1 h following treatment with PMA, a second increase in expression of this mRNA, although smaller than the first, was detected by 6 h in some of these cells. This secondary increase in Egr-1 mRNA may indicate the induction of other genes which also act to regulate Egr-1. Also, while PMA was shown to rapidly regulate Egr-1 mRNA in NSCLC cells, it remains to be determined whether the Egr-1 protein is also rapidly translated and regulated in these different lung cell types.

In addition to regulation by PMA, it was also demonstrated in the present study that Egr-1 mRNA is rapidly induced by TGF-₁ in NSCLC cells as well. We have shown in a recent study expression of the mRNAs and proteins for the different TGF- ligands and receptors in NSCLC cells (25). Egr-1 is a member of a zinc-finger gene family consisting of at least four transcription factors (Egr-1, -2, -3, and -4) that preferentially binds GC-rich regulatory elements (GCEs) with the sequence 5'-GCGT/GGGGCG-3'or 5'-TCCT/ACCTCCTCC-3' (26). The promoter of the human TGF-₁ gene has been shown to contain multiple GC-rich sequences (27). The expression of TGF-₁ is regulated in part by these GC-rich sequences (27, 28). Wilms tumor-1 protein (WT1) is a strong repressor of the transcriptional activity of the human TGF-₁ promoter whose effect is counteracted in a dose-dependent manner by coexpression of Egr-1 (29). Recently, it has been shown that expression of Egr-1 in human HT-1080 fibrosarcoma cells causes increased secretion of biologically active TGF-₁ in direct proportion to the amount of Egr-1 expressed (30); addition of exogenous TGF-₁ to Egr-1-expressing HT1080 cells inhibits their growth and addition of monoclonal anti-TGF-₁ antibodies to these cells completely reverses the growth inhibitory effects of Egr-1. The more persistent increase in Egr-1 mRNA in response to TGF-₁ in NHBE cells compared to NSCLC cells may act to feedback and increase the level of TGF-₁ and thus inhibit the proliferation of these cells. It remains to be determined whether the persistent increase in Egr-1 mRNA in NHBE cells also occurs at the protein level to enable this to occur.

The effects of PMA and TGF-₁ have been examined previously in normal and neoplastic human bronchial epithelial cells. Differential effects of PMA and TGF-₁ have been demonstrated in these different human bronchial epithelial cells. For example, while as little as 3 nM PMA was shown to induce NHBE cells to undergo terminal differentiation and completely inhibit their proliferation, the growth of several NSCLC cell lines was not significantly inhibited at this concentration NSCLC cells and continued to proliferate in as much as 100 nM PMA (31). While NHBE cells and several NSCLC cell lines possess comparable numbers of TGF- receptors with similar affinities, TGF-₁ has been shown to inhibit DNA synthesis, proliferation and differentiation of NHBE cells, but not to inhibit these activities in NSCLC cells (32). Lung cancer cells apparently acquire differentiation-resistent phenotypes at a step of the differentiation-induction pathway other than at the ligand binding of the surface receptors for the differentiation inducers. We have demonstrated a transient increase in Egr-1 mRNA in NSCLC cells following exposure to either PMA or TGF-₁. While the effect of PMA on Egr-1 mRNA is also transient in NHBE cells, the effect of TGF-₁ on Egr-1 mRNA is more prolonged in these cells. Cell density has been shown to govern the ability of TGF-₁ to control differentiation in NHBE cells. For example, it has been demonstrated that whereas irreversible inhibition of DNA synthesis occurs in sparse cell density cultures of NHBE cells within 24 h after exposure to TGF-₁, only a transient depression in DNA synthesis is seen in high density cultures (33). Recently, we have demonstrated the ability of TGF-₁ to inhibit the proliferation of some NSCLC cells in soft agarose (25). It is possible that cell density has a role in controlling the ability of TGF-₁ to inhibit proliferation of lung cancer cells just as it does in NHBE cells. It is also possible that cell density may have an effect on the regulation of Egr-1 in these cells. Additional studies will be needed to determine whether this is the case.

Since Egr-1 is a positive regulator of TGF-₁, it may be important that Egr-1 is low or undetectable in most human tumors that have been investigated. For example, in a series of 101 paired NSCLC carcinomas and normal specimens, it was observed that the level of Egr-1 is less in the carcinomas than in adjacent normal tissue in more than 72% of the specimens (34). Our data show that TGF-₁ stimulates expression of Egr-1 mRNA rapidly in NSCLC cells. It may be possible that Egr-1 may have growth inhibitory properties in NSCLC cells by stimulating secretion of TGF-₁, which in turn might act as an autocrine or paracrine agent to inhibit cell proliferation as in HT-1080 cells. However, because expression of Egr-1 mRNA is only transient in NSCLC cells, this inhibitory effect might be predicted to be transient as well. In contrast, TGF-₁ stimulates expression of Egr-1 mRNA in a more persistent manner in NHBE cells so that the growth inhibitory activities of Egr-1 might be predicted to be more prolonged than in NSCLC cells. Future studies will focus on whether this is the case to better understand the mechanism of growth control by Egr-1 in normal and malignant lung cells.