Lanthionine Synthetase Components C-like 2 Increases Cellular Sensitivity to Adriamycin by Decreasing the Expression of P-Glycoprotein through a Transcription-mediated Mechanism¹

Gene amplification is a mechanism for selectively increasing the copy number and expression of specific genes. Amplified MYCN (2p24) in neuroblastoma (1), CDK4³ (12q13–14) in soft tissue sarcomas (2), and HER2/NEU (17q11–12), as well as CYCD1 (11q13) in breast cancer (3, 4), are prime examples of genes whose increased expression promote cell proliferation via an amplification-mediated mechanism.

Results from studies directed at characterizing the size of amplification regions have consistently revealed that the amount of genomic DNA constituting an amplification repeat unit (amplicon) is invariably and considerably larger than that of the target gene itself (5, 6, 7). As a result, it is not surprising that tumor amplicons generally contain multiple genes. Genes that are amplified as a result of their physical proximity to a target sequence can be considered as coamplified bystanders, and a number of coamplified bystander genes has been identified, including DDX1 at 2p24 (8), TOPOIIα at 17q11–12 (9), and GLI at 12q13–14 (7).

Information has begun to emerge that suggests that the determination of amplicon bystander gene content may be of prognostic and/or therapeutic value to cancer patients, e.g., the coamplification of DDX1 with MYCN has been shown to identify more aggressive neuroblastoma (10). In breast cancer, the coamplification of TOPOIIα with ERBB-2 has been associated with a more favorable response to chemotherapy (11, 12). These results suggest that the occurrence of bystander gene coamplification has significant clinical ramifications. However, it is clear that our understanding of the molecular and biological basis that underlies these clinical observations is meager.

As with other coamplified bystanders, the amplification of LANCL2 results from its proximity to an amplification target, in this case, EGFR at chromosomal region 7p11.2 (13, 14). LANCL2 amplification occurs in approximately one-half of glioblastomas with EGFR amplification, or ∼20% of all such tumors, and is accompanied by elevated LANCL2 expression (13, 14). The consequences of elevated LANCL2 expression are unknown, because there has yet to be any functional analysis performed for this gene’s corresponding protein. The predicted amino acid sequence for LanCl-2 displays limited homology with bacterial LanC family proteins through its seven GXXG repeats, and this homology forms the basis for the name assigned to the human gene (15). In eukaryotes, the GXXG motif is associated with KH (2) domain proteins that bind single-stranded nucleic acids (16). Aside from this motif, however, the homology between LanCl-2 and any of the known KH proteins is negligible.

LANCL2 has also been designated as Testis Adriamycin Sensitivity Protein (GenBank accession no. AB035966). As suggested by this name, LANCL2 is most highly expressed in testis, although Northern analysis has revealed its expression in all normal tissues that have been surveyed (14, 15). The effect of its expression with respect to cellular drug sensitivity has yet to be investigated in detail. For this reason, in combination with its frequent amplification in glioblastoma, we have examined potential relationships between LanCl-2 expression and cellular sensitivity to Adriamycin, as well as between LanCl-2 and MDR1 expression because its cognate protein, P-gp, is known to increase cellular resistance to many chemotherapeutic agents, including Adriamycin (17). Our results show that increases in cellular LanCl-2 are accompanied by decreasing expression of P-gp, as well as by increasing cellular sensitivity to Adriamycin. In total, our data suggest that the identification of LANCL2 amplification and attendant overexpression may provide opportunity for targeted therapeutic intervention in the treatment of this cancer. In addition, these results provide further support for the importance of studying bystander gene coamplification in human tumors.

The human uterine sarcoma cell line, MES-SA (CRL-1976), and its multidrug-resistant derivative MES-SA/Dx5 (CRL-1977) were purchased from American Type Culture Collection. Cells were maintained in McCoy’s 5a medium supplemented with 10% FCS, 100 units/ml penicillin, and 100 μg/ml streptomycin. Media for maintaining transfectant cells additionally contained G418 at 800 μg/ml.

MES-SA/Dx5 cells were transfected with LANCL2 cDNA cloned into the pcDNA3.1-V5/His (Invitrogen) or empty vector using GenePORTER transfection reagent (Gene Therapy System) following the manufacturer’s protocol. LANCL2 transfectants were selected in the media described above supplemented with 800 μg/ml G418.

Proteins from whole cell lysates were electrophoresed through 10 or 4–20% gradient gels and transferred to nitrocellulose or polyvinylidene difluoride membranes. Membranes were preincubated with 5% nonfat milk and 0.1% Tween 20 in PBS and subsequently incubated in the same solution with antibody. For detection of V5-tagged LanCl-2, anti-V5 antibody (Invitrogen) was used at 1:5000 dilution. P-gp was detected with C219 (Signet Laboratories) at a concentration of 0.5 μg/ml. Antimouse IgG secondary antibody (Amersham) was used at 1:4000 dilution. Enhanced chemiluminescence reagent (Amersham) was used for chemiluminescence detection.

Each cell type was plated in hexaplicate at a density of 8,000/well in a 96-well plate. After overnight incubation, Adriamycin was added to concentrations of 0; 10; 50; 500; 1,000; and 10,000 nm. Cells were grown in the presence of drug for 3 days, and 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide assay was performed as described by Gosland et al. (18). Sample absorbances were determined at 595 nm using a microplate reader (model UV-3550; Bio-Rad).

Cells were prepared for immunofluorescence confocal microscopy as described previously (19). Primary antibodies anti-V5 (1:1000 dilution) and MDR Ab-1 (4 μg/ml concentration; Oncogene Research) were used for detection of LanCl-2-V5 and P-gp, respectively. Alexa 488 (LanCl-2-V5) and Alexa 594 (P-gp) were used as secondary according to the manufacturer’s protocol (Molecular Probes). Confocal imaging was performed using a Zeiss LSM510 confocal microscope.

Cells were harvested in PBS and centrifuged at 500 × g for 3 min. Pelleted cells were homogenized in buffer H [0.25 m sucrose, 10 mm HEPES (pH 7.5), and protease inhibitor] and then centrifuged at 1,600 × g for 10 min to obtain a pellet and supernatant. The supernatant was centrifuged at 200,000 × g for 1 h to obtain the cytosolic fraction. The pellet was resuspended in buffer H and homogenized. The sucrose concentration of the homogenate was adjusted to 1.6 m and topped with buffer H. The plasma membrane fraction was obtained after centrifugation at 28,000 rpm for 70 min. The pellet was resuspended in 2.2 m sucrose, 5 mm HEPES, and 3 mm MgCl₂ and centrifuged at 41,000 rpm for 1 h to obtain the nuclear fraction.

Total cellular RNA was isolated using TRIzol reagent (Invitrogen). RT-PCR for cDNA synthesis was performed using the SuperScript First Strand Synthesis system according to the manufacturer’s specifications (Invitrogen). The primers used for the simultaneous amplification of MDR1 and GAPDH have been described previously (20). The PCR reaction profile was 1 cycle at 94°C for 2 min for initial denaturation and 25 cycles at 94°C for 30 s, 54°C for 30 s, and 72°C for 30 s with a final extension for 7 min at 72°C. Each reaction product (10 μl) was resolved in a 2% agarose gel and visualized under UV light after ethidium bromide staining. MDR1 and GAPDH band intensities were determined using Multianalyst software (Bio-Rad).

For MDR1 promoter activity analysis in LANCL2 clonal isolates, cells were seeded into six-well plates at a density of 2–5 × 10⁵ cells/well. After 24 h, LipofectAMINE reagent (Invitrogen) was used to cotransfect cells with 0.5 μg of MDR1 reporter construct (kindly provided by Dr. Kathleen Scotto, Fox Chase Cancer Center; Ref. 21) and 1 ng of Renilla luciferase vector pRL-TK (internal control for transfection efficiency). For promoter activity analysis of transient transfectants, cells were cotransfected with either LANCL2 in pcDNA3.1-V5/His or an expression vector and 0.5 μg of EGFR (−777 to −1; Ref. 22), MMP1 (−677 to −8; Ref. 23), or MDR1 promoter. At 30 h of post-transfection, luciferase activities were measured and normalized against corresponding protein concentrations and/or Renilla luciferase activities.

To study the effects of LanCl-2 expression on drug sensitivity, we used the sarcoma cell line MES-SA and its multidrug-resistant derivative MES-SA/Dx5 that has elevated P-gp expression (24). Dx5 cells that were transiently transfected with V5-tagged LANCL2 showed a ∼2-fold decrease in P-gp expression (Fig. 1,A). Similar results were observed in long-term transfectant pools, as well as in clonal isolates of LANCL2 transfectants (Fig. 1 B). In fact, results from the clonal isolates suggested an inverse association between LanCl-2 and P-gp expression. No effect of LanCl-2 expression was evident with regard to cellular levels of β-actin or mitogen-activated protein kinase.

The observed decreases in P-gp expression suggest that the LANCL2 transfectants would show increased sensitivity to cytotoxic drugs that are exported by P-gp (17). To test this possibility, we subjected the clonal LANCL2 transfectants to various concentrations of Adriamycin. Dx5-L2, with the highest LanCl-2 and lowest P-gp expression, was >20-fold more sensitive to Adriamycin as compared with Dx5 cells transfected with vector alone (IC₅₀s of 7 versus 150 nm, respectively: Fig. 2). Dx5-L1 and -L3, which also show reduced P-gp, were 2–3-fold more sensitive (IC₅₀s of 55 nm) to Adriamycin than vector-transfected Dx5 cells.

To address the possibility of a direct interaction between LanCl-2 and P-gp proteins, we conducted coimmunoprecipitation assays. The results did not support a physical association between these proteins, irrespective of the order of use of antibodies for immunoprecipitation and Western blot detection (data not shown). This conclusion was corroborated by immunofluorescence analysis that did not support LanCl-2 and P-gp colocalization, despite each protein showing strong staining at the plasma membrane (Fig. 3,A). Our Western analysis of protein distributions in subcellular fractions indicated that LanCl-2 resides in each of the three cellular compartments: (a) plasma membrane; (b) cytosol; and (c) nucleus (Fig. 3,B). As expected, P-gp was exclusively localized to the plasma membrane in MES-SA/Dx5 and vector-transfected cells (Fig. 3, A and B). This was also the case for Dx5-L2, although the intensity of membranous P-gp staining in these cells was significantly reduced and therefore consistent with the results from the Western blot analysis of subcellular fractions.

To investigate the mechanism by which LanCl-2 reduces cellular P-gp, we used pulse-chase analysis to address potential changes in protein stability. These results, however, indicated no significant difference in P-gp half-life between vector versus LANCL2-transfected Dx5 cells (data not shown). Next, we used RT-PCR to determine whether decreased P-gp expression in LANCL2-transfected cells is associated with decreased cellular mRNA. This analysis revealed decreased MDR1 mRNA in each of the clonal isolates (Fig. 4,A) and to an extent that was consistent with the observed decreases in corresponding P-gp protein expression (Fig. 1 B).

To determine whether the observed decreases in cellular mRNA resulted from reduced MDR1 gene transcription, a luciferase reporter assay was used to compare relative levels of MDR1 promoter activity in MES-SA parental cells, the Dx5 derivative cell line, and in the three Dx5-LANCL2 clonal isolates. This analysis revealed that MDR1 promoter activity was highest in Dx5 cells and Dx5 vector transfectants, with each of these showing approximately seven times the level of MDR1 transcription relative to MES-SA parental cells (Fig. 4 B). Importantly, Dx5-L2, with the highest LanCl-2 expression, demonstrated MDR1 promoter activity comparable with that determined for MES-SA parental cells, indicating a significant level of MDR1 transcriptional suppression. Dx5-L1 and -L3 showed a ∼2-fold reduction in MDR1 promoter activity when compared with untransfected Dx5 cells.

To corroborate the repressive effect of LanCl-2 on MDR1 transcription in clonal isolates, we examined Dx5 cells that had been transiently transfected with the MDR1 promoter construct and increasing amounts of LANCL2 cDNA (Fig. 4,C). The results are consistent with MDR1 promoter activity being regulated in a LanCl-2 dose-dependent manner (Fig. 4,C). To test for specificity of MDR1 transcriptional suppression, we examined the effect of LanCl-2 expression on two other promoters: (a) EGFR; and (b) MMP1. The associated results show similar EGFR and MMP1 promoter activities, irrespective of exogenous LanCl-2 expression (Fig. 4 C).

Our interest in LANCL2 stems from previous studies that revealed this gene as the most frequently coamplified bystander with EGFR (13, 14). The occurrence of LANCL2 amplification in only half of glioblastomas with amplified EGFR may be surprising given its proximity to EGFR (∼200 kb); however, our work as well as the work of others indicate that tumor amplicons often contain the amplification target gene with relatively little flanking genomic sequence (14, 25), and with respect to EGFR amplicons, this is clearly the case.

The cognate protein for LANCL2 has been suggested to sensitize cells to the cytotoxic effects of Adriamycin (GenBank accession no. AB035966), and here we have investigated this possibility by examining the effects of exogenous LANCL2 gene transfer into the well-characterized cell line pair MES-SA and its multidrug-resistant derivative MES-SA/Dx5 (24). Our results indicate that increasing the expression of LANCL2 does indeed sensitize cells to treatment with Adriamycin and that this effect is associated with decreased expression of P-gp. This relationship is most evident in the case of Dx5-L2 transfectant cells, which show the highest level of LanCl-2 protein expression with an associated 15-fold reduction in expression of P-gp, as well as a corresponding 20-fold increase in sensitivity to Adriamycin (Figs. 1,B and 2).

The effect of LanCl-2 on P-gp expression appears to be at the level of transcription, as indicated by decreased MDR1 promoter activity in LANCL2-transfected cells (Fig. 4, B and C) and supported by results from RT-PCR analysis of MDR1 mRNA levels in the LANCL2 transfectants (Fig. 4,A). Furthermore, the repressive effect of LanCl-2 on transcription shows some degree of specificity for the MDR1 gene promoter because the activities of two other promoters, EGFR and MMP1, were unaffected by LanCl-2 expression (Fig. 4,C). Importantly, we have shown that decreased MDR1 transcription is accompanied by a corresponding decrease in P-gp protein (Fig. 1, A and B).

Amino acid sequence analysis comparisons show that bacterial LanC protein and LanCl-2 share seven GXXG motifs (15). This motif is thought to be important for LanC function in antibiotic peptide processing (26). In eukaryotes, GXXG is a signature motif for KH domains (16). Structural analysis has proposed the GXXG loop as a DNA/RNA-binding surface (16). Although flanking regions of the GXXG motifs in LanCl-2 are quite different from the surrounding amino acid residues of GXXG motifs in known KH domain-containing proteins, the presence of this conserved sequence in LanCl-2 suggests that it may function as a single-stranded, nucleic acid-binding protein. Certainly, our data indicating LanCl-2-mediated MDR1 transcriptional suppression, as well as LanCl-2 nuclear localization, provide justification for investigating this possibility in detail. Interestingly, both immunofluorescence and subcellular fractionation analyses (Fig. 3, A and B) indicate that LanCl-2 also resides within the cytoplasmic compartment, with possible concentration of this protein along the inner surface of the plasma membrane. Whether these different cellular populations of LanCl-2 are distinguished through protein modification(s) has yet to be determined.

Bystander gene coamplification has been demonstrated for other loci in other types of cancer (7, 8, 9), and results from recent studies suggest that the determination of bystander gene coamplification may be of relevance to the prognosis and treatment of cancer patients (11, 12). In the present study, we have shown that the introduction of exogenous LANCL2, a bystander gene that is coamplified with EGFR in glioblastoma, causes increased cellular sensitivity to an anticancer drug. The implications of this investigation are therefore analogous to those showing that the coamplification of TOPOIIα with ERBB2 results in altered cellular sensitivity to topoisomerase II inhibitors (11, 12). A similar, albeit inverse, concept has been advanced for the frequent codeletion of the methylthioadenosine phosphorylase gene with CDKN2A, which can be exploited for selective chemotherapy with de novo purine synthesis inhibitors (27). In point of contrast, however, frequent homozygous deletions of genes other than CDKN2A have yet to be demonstrated in human tumors, whereas a substantial body of literature attests to the existence of multiple amplification targets in human cancer (7, 8, 9). Our results, combined with the results from studies published previously (10, 11, 12), support the importance for detailed characterization of tumor cell amplicons and for determining the functional consequences associated with the coamplification of individual bystander genes.