Heterogeneous nuclear ribonucleoproteins (hnRNPs) participate in a variety of processes involving RNA, including transcription, splicing, processing, translation and turnover, and there are approximately 20 major members of the hnRNP family.1 High expression of some of these have been reported in several human malignant tumors and interest in the action of these proteins in malignancies has been growing.2, 3 HnRNP L is 68 kDa protein with 4 RNA recognition motifs (RRM). There have been several interesting reports demonstrating that cytoplasmic hnRNP L specifically interacts with VEGF mRNA in hypoxic cells in vivo, regulates VEGF mRNA stability4 and binds in a sequence-specific manner to a cis-acting RNA sequence element that enables intron-independent gene expression.5 The role of hnRNP L, however, still requires further study.
KW-2189 is a water-soluble derivative of antitumor antibiotic duocarmycin (DUM),6, 7, 8 and DUM and its derivatives have been reported to exert their anti-tumor activity through covalent binding to the DNA minor groove and inhibition of DNA synthesis. We identified previously a nuclear protein DARP (duocarmycin-DNA adduct recognizing protein) in human cervical carcinoma HeLa S3 cells.9 We purified the DARP from nuclear extract of HeLa S3 and its amino acid sequence was identical to hnRNP L. We investigated this, particularly in cancer cells.
MATERIAL AND METHODS
Cell cultures and reagents
Human small cell lung cancer cell lines SBC-3 and H69, human non-small cell lung cancer cell lines PC-14, and their respective cisplatin-resistant cell lines (SBC-3/CDDP, H69/CDDP,10 and PC-14/CDDP11) were maintained in RPMI 1640 (Sigma, St. Louis, MO) supplemented with 10% heat-inactivated FBS (Gibco BRL, Gaithersburg, MD). Murine fibroblast cell line NIH3T3 and sublines (including cDNA transfectants) were cultured in DMEM (Nissui Pharmaceutical Co. Ltd., Tokyo, Japan) supplemented with 10% FBS. KW2189 was provided by Kyowa Hakko Kogyo Co., Ltd. A monoclonal antibody specific for hnRNP L (4D11) was generously provided by Dr. G. Dreyfuss (University of Pennsylvania, Philadelphia).
Cells were washed twice with cold PBS and lysed in buffer (10 mM Tris-HCl pH 7.8, 1% Nonidet P-40, 0.15 M NaCl, 1 mM EDTA, 10 μg/ml aprotinin, 0.5 μg/ml leupeptin, 1 mM phenyl-methane-sulfonyl fluoride [PMSF], 1 tablet/50 ml ϕgrComplete™ and 10% glycerol) for 60 min on ice. The lysates were centrifuged at 8,000g for 20 min, and supernatants were obtained as total protein. Protein concentration was measured by bicinchoninic acid protein assay (Pierce, Rockford, IL).
SBC-3, PC-14 and H69 cells were lysed in buffer A containing 10 mM HEPES-KOH (pH 7.9), 10 mM KCl, 0.1 mM EDTA-NaOH (pH 8.0), 0.1 mM ethyleneglycol bis(2-aminoethyl ether) tetraacetic acid (EGTA), 1 mM dithiothreitol (DTT), 0.5 mM PMSF, 1 mM aprotinin, and leupeptin. Nonidet P-40 (final concentration 0.5%) was added after allowing to stand on ice for 15 min. The supernatant obtained by centrifugation at 7,000g for 30 sec after standing on ice for 5 min was collected as the cytoplasmic fraction. The pellet was resuspended with buffer A containing 0.25 M sucrose, and after buffer B′ (buffer A containing 0.6 M sucrose) was added, the solution was centrifuged at 5,000g for 1 min at 4°C. The nuclei, which were contained in the pellet, were sonicated in buffer C containing 20 mM HEPES-KOH (pH 7.9), 0.4 M NaCl, 1 mM EDTA-NaOH (pH 8.0), 1 mM EGTA, 1 mM DTT and 1 mM PMSF, and then rocked at 4°C for 30 min and centrifuged at 8,000g for 10 min. The supernatant was used as the nuclear fraction. The nuclear protein content was adjusted to 5 μg per well, and the same volume of cytoplasmic protein was applied to the next well. The cytoplasmic and nuclear fractions were subjected to SDS-PAGE and Western blotting with anti-hnRNP L antibody.
An INSTA-Blot human tissues membrane (Imgenex, San Diego, CA), which contains 10 μg per lane of different human tissue lysates, was soaked in 100% methanol and then washed with TBST. After blocking the membrane in 5% skim milk in TBST for 1 hr at room temperature, it was probed with anti-hnRNP L antibody diluted (1:500) in TBST with 1% skim milk for 1 hr at room temperature, washed 3 times in TBST, incubated with anti-mouse IgG horseradish peroxidase antibody diluted (1:5000) in TBST with 1% skim milk for 1 hr at room temperature, and then washed 3 times in TBST. The signal was visualized with ECL (Amersham Pharmacia Biotech UK Ltd., Buckinghamshire, England), and Hyperfilm-MP (Amersham) was exposed to it.
Purification of DARP and amino acid sequencing
Purification of DARP was conducted as described previously.9 DARP was detected by its ability to bind to DUMSA (one of DUMs)-DNA adduct in gel shift assays. Nuclear and cytoplasmic extracts from HeLa S3 cells (ATCC: American Type Culture Collection) were prepared according to previously published procedures. For identification of the DARP band, the aliquot of this material was subjected to DEAE-sephacel column again, and eluted with 0.5M stepwise procedure (0.1–0.5 M KCl) to give the small amount of purified DARP. Protein concentrations were estimated using Bio-Rad protein assay and the quality of the each fraction was checked by CBB (Coomassie brilliant blue) or silver staining of SDS-polyacrylamide gels. For analysis of amino acids sequence of DARP about 2 μg of affinity purified DARP was separated by 12% SDS-PAGE. The 60 kDa protein band was excised and digested with lysyl endopeptidase (WAKO, Japan) in 0.1 M Tris-HCl (pH 9.0), 4 M urea at 37°C for 16 hr. The resulting peptides were isolated by reversed phase HPLC on a RPC C2/C18 column (Amersham Pharmacia Biotech, Sweden). The amino acid sequence was determined by automated Edman degradation using a PPSQ-10 protein sequencer (Shimadzu, Japan).
Gel mobility shift assay
Labeled oligonucleotide (1 μg) was incubated with cell extract (final protein concentration, 20 μg/μl) at 30°C for 30 min in the presence of 2 μg of poly[dIdC]poly[dIdC] and 1 μg of BSA, except where stated, in a final volume of 15 μl of 0.1 M KCl HEDG. Where indicated, drug modified or unmodified calf thymus DNA was added to the reactions. Samples were electrophoresed in 6% polyacrylamide gel, dried and scanned.
Total RNA was prepared from HeLa cells with ISOGEN (Nippon Gene, Tokyo, Japan), and 14–784 and 636—1718 fragments of hnRNP L (2033 bp) were obtained by reverse transcription-polymerase chain reaction (RT-PCR). The PCR products were cloned in PCR II, a TA cloning plasmid vector, and then coupled at the Bcl I site. Subsequently, a fragment including hnRNP L was digested from the plasmid with Not I, and it was informed into the Not I site of the pRc/CMV vector. After confirming its sequence, this expression vector, pRc/CMV, containing cDNA of hnRNP L, was transfected into NIH3T3 cells with the Lipofectin reagent (Gibco BRL) according to the manufacturer's instructions. After 48 hr incubation, 1.5 mg/ml of G418 (Sigma) was added. Cells resistant to neomycin were selected, and isolated by limiting dilution methods.
Total RNAs were prepared from SBC-3, PC-14 and NIH3T3 cells, and the 10 stable transfectants described above with ISOGEN reagent. RNA (12 μg) was electrophoresed and transferred to a positively charged nylon membrane (Hybond-N+). The 1030 bp fragment of hnRNP L cDNA was labeled with [α32P]-dCTP by using the Rediprime II random primer labeling system (Amersham) and was used as a probe. The membrane was hybridized at 42°C overnight for blocking with sonicated salmon sperm DNA (Stratagene, La Jolla, CA) and hybridized at 42°C overnight with the labeled probe rotating. Washings were carried out in 2× SSC, 0.1% SDS, for 10 min at room temperature, 1× SSC, 0.1% SDS, for 1 hr at 42°C, and 0.2× SSC, 0.1% SDS, at 42°C for 1 hr. A BAS imaging plate (Fuji Photo Film Co. Ltd., Kanagawa, Japan) was exposed to the filter for 2 hr, and relative band intensities were measured with a BAS 2000 system (Fuji).
The effect of hnRNP L on cell sensitivity to KW2189 was estimated by the 3-(4,5-dimethylthiazol-2yl)-2,5-diphenoyltetrazoliumbromide (MTT) assay. NIH3T3, and stable transfectants of hnRNP L cDNA, Fw3and Fw9 cells were exposed to 0–50 nM KW2189 for 72 hr before measuring absorbance. The OD values at 562–630 nm were measured with a 96-well microtiter plate reader, EL340 (Bio-Tek, Winooski, VT).
Immunochemical cell staining
Human lung cancer cell lines, SBC-3, PC-9, PC-14 and H69 cells were prepared on slide glasses with cytospin (Shandon, Pittsburgh, PA). The cells were dried and then fixed in cold acetone for 2 min. All of the incubation steps were carried out at room temperature, and Step 2 and 3 were carried out in the dark. The steps included: 1) incubation with 10% horse serum for 30 min for blocking; 2) incubation with anti-human hnRNP L (1:500 diluted in PBS with 1.5% blocking serum) for 60 min; and 3) incubation with fluorescence anti-mouse IgG (1:500 diluted) for 45 min. Slides were washed with 3 changes of PBS between each step. After Step 2 each washing was carried out for 5 min. The slides were mounted with 90% glycerol in PBS and examined with a fluorescence microscope (Nikon, Tokyo, Japan), equipped with fluorescein isothiocyanate filter set B-2A (Nikon).
EGFP-hnRNP L deletion mutants
pRc/CMV containing the 14–1718 fragment of hnRNP L cDNA (2033bp) was constructed as described above. After digesting the plasmid with SacII and BamHI, and the resulting fragment was introduced into the SacII/BamHI site of the pEGFP-C3 vector (Clontech, Palo Alto, CA), with the Takara DNA ligation system. Construction of deletion plasmids was carried out as follows. EGFP-hnRNP L (Construct 2) was partially digested with StuI and self-ligated to generate Constructs 3 and 6. PEGFP-hnRNP L was digested with BglII, and after extracting the 570 bp and 1023 bp fragments with a QIAquick Gel Extraction Kit (Qiagen, Hilden, Germany), each fragment was inserted into the BglII site of the pEGFP-C2 and -C3 vectors to generate Constructs 4 and 7, respectively. The 384 bp fragment of hnRNP L extracted by digesting with StuI was inserted into the SmaI site of pEGFP-C3 vectors to generate Construct 5. PEGFP-hnRNP L was digested with KpnI, and it self-ligated to generate Construct 8. The 584 bp fragment digested with KpnI and BglII and extracted was inserted into the BglII site of pEGFP-C3 vectors to generate Construct 9, and the 626 bp fragment digested with AccI was inserted into the AccI site of pEGFP-C3 vectors to generate Construct 10 (Fig. 1).
A cover-glass was placed on the bottom of each well of a 6-well culture dish, and each well was seeded with 1.6 × 105 NIH3T3 cells and incubated for 48 hr at 37°C. After diluting 2.5 μg/well of plasmid DNA (pEGFP vectors containing deletion mutants of hnRNP L described above) in 1 ml/well of serum-free DMEM, 7.5 μl/well of 1 mM TransFast Reagent (Promega, Madison, WI) was added to the mixture. After allowing the mixture to stand for 15 min at room temperature, it was added to cells from which the growth medium was removed. The cells were then incubated for 1 hr at 37°C, and 1 ml/well of complete growth medium was added to them. At 24 hr after transfection, the cells were mounted on slides with aqueous mounting medium and examined under a fluorescence microscope (Nikon, B-2A filter, Tokyo, Japan).
Purification and sequence analysis of the DARP
Purification of the DARP was conducted as described previously.9 After affinity purification, 2 main proteins were detected in SDS-PAGE with silver staining. Further purification efforts with DEAE-sephacel column chromatography gave a single band of Mr ∼60,000 with the binding activity to the labeled duocarmycin-modified oligonucleotides (Fig. 2a). Coincubation of duocarmycin-treated calf thymus DNA with the labeled probe and purified DARP resulted in the retarded band in the gel mobility shift assay (Fig. 2b). Competition experiment in the presence of 30 and 300 ng of calf thymus DNA-DUMSA adduct demonstrated that 300 ng adduct reduced the intensity of the band in our previous study.9
The 60 kDa protein separated by SDS-PAGE was excised and digested with lysyl endopeptidase. The resulting peptides were eluted, separated by reversed phase HPLC, and sequenced. Three partial amino acid sequences were obtained, AAAGGGGGGGRYYGGG, DFSESRNNRFSTPEQAA and SDALETLGFLN, which were found to completely match parts of the predicted human heterogeneous nuclear ribonucleoprotein L. Gel mobility shift assay using anti-hnRNP L did not, however, show the supershift of the band induced by anti-hnRNP L (data not shown).
Expression of hnRNP L
Western blot analysis was carried out using a membrane containing normal human tissue lysates from different organs. A 68 kDa band of hnRNP L was detected in total protein extracts from brain and small intestine, but not in others, including normal lung (Fig. 3). The expression of hnRNP L protein, however, was detected in the human lung cancer cell lines (Fig. 4a). Northern blot analysis confirmed the expression of hnRNP L at the mRNA (∼2 kbp) level in these cells (Fig. 4b). In contradiction to our first result that hnRNP L was not detected in normal lung tissue, the expression of hnRNP L in malignant cells seemed to increase.
Effect of hnRNP L on drug sensitivity
To evaluate the function of hnRNP L, hnRNP L cDNA was transfected into NIH3T3 cells, and stable transfectant clones were characterized. The Fw3 and Fw9 clones showed higher expression of hnRNP L mRNA than other transfectants by 5.2-fold and 4.4-fold to control respectively detected by Northern blot analysis (Fig. 5a).
We measured the growth inhibitory effect of KW-2189 in the hnRNP L transfectant cells by MTT assay. The IC50 values for KW-2189 in the Fw3 and Fw9 clones were 3.5 nM and 4.3 nM, respectively, and the Fw3 and Fw9 cells were 13.4-fold and 10.9-fold, respectively, more sensitive to KW-2189 than the Mock transfectant C4 cells (IC50: 47 nM) (Fig. 5b). These results indicate that hnRNP L enhances cell sensitivity to the growth inhibitory effect of KW-2189 in vitro. We also examined the sensitivity of the transfectants to cisplatin and mitomycin C and no difference of the sensitivity was observed between the transfectants and the Mock cells (data not shown). The hnRNPs have been reported to regulate both nuclear and cytoplasmic events, as described above, and the intracellular localization of hnRNP L was examined in the next step to identify the site of action of hnRNP L in the sensitivity enhancement machinery.
Localization of hnRNP L protein in human lung cancer cell lines
We carried out immunofluorescence cell staining with anti-hnRNP L antibody to determine the subcellular localization of hnRNP L protein in human lung cancer cells (Fig. 6a). Based on the results, the localization of hnRNP L cells could be classified into two patterns: nuclear localization (N) and cytoplasmic localization (C) (Fig. 6b). As shown in Figure 6c, the cytoplasmic pattern was observed frequently in SBC-3 and H69 cells, whereas the nuclear pattern was common in PC-14 cells. To confirm this differential distribution, fractionated proteins from the nuclear and cytoplasmic fractions of these cells were immunoblotted with anti-hnRNP L antibody (Fig. 7). The results showed that hnRNP L was expressed equally in the nucleus and cytoplasm of the SBC-3 and H69 cells, whereas it was expressed predominantly in the nuclei of the PC-14 cells. These results are consistent with the immunocytological findings. In addition, the cisplatin-resistant sublines derived from these cells exhibited the same localization pattern as their parental cells. This indicates that the differential localization depends on the cell type.
Motifs required for the intracellular localization of hnRNP L
It has been reported that hnRNP L is localized in the nucleoplasm of HeLa cells, except the nucleoli,12 but the mechanism of its localization remains unknown. To identify the motifs responsible for the localization of hnRNP L, we constructed an hnRNP L deletion series fused to EGFP (Fig. 8a), transfected the constructs into NIH3T3 cells, and examined them under a fluorescence microscope. As shown in Figure 8b, EGFP protein itself was rather evenly distributed throughout the cell, the cytoplasm and the nucleus (Transfectant 1). Full-length hnRNP L was present in the nucleoplasm, except the nucleoli (Transfectant 2). Deletion mutants containing the N-terminal portion of RRM1 or of RRM3 (Transfectants 3, 6, 8 and 9) showed hnRNP L localization in the nucleus. Transfectants 4, 5 and 7, containing the N-terminal portion of RRM2 showed hnRNP L distributed through the cell, whether they also contained that portion of RRM1 and RRM3 or not. In Transfectant 10, which lacked the N-terminal region of all 3 RRMs, hnRNP L was distributed throughout the cell. These results indicate that the N-terminal portion of each RRM is required for determination of the intracellular localization of hnRNP L (Fig. 8a, arrows). We then searched the sequence of hnRNP L and found 2 sequences that were rich in alkaline amino acids (residues 25–31, 380–387) and a sequence that was rich in hydrophobic amino acids (residue 163–171, Fig. 9). The sequences rich in alkaline amino acids showed high homology with the NLS sequences of c-Jun and SV40 large T antigen,13, 14 and the sequence rich in hydrophobic amino acids showed high homology with the NES sequences of PKI15 and Rev16 respectively. The N-terminal portion of RRM1 and RRM3 contain the NLS-like sequences, residue 25–31 and residue 380–387, respectively, and the N-terminal portion of RRM2 contains the NES-like sequence, residue 163–171.
There are approximately 20 major hnRNPs, and some of them have been reported to be highly expressed in cancer tissues. Sueoka et al.3 demonstrated elevated expression of hnRNP B1 mRNA in human lung cancer tissue, and hnRNP I and hnRNP K mRNA have been reported in malignant glioblastoma and breast cancer, respectively.2, 17 We demonstrated expression of hnRNP L in human lung cancer cell lines and high expression of hnRNP L is presumably present in lung cancer tissue.
We reported previously that a nuclear protein in human cancer cells binds to the DUM-DNA adduct. The protein, DARP, preferentially bound to the DNA damage induced by DNA-alkylating minor groove binders such as DUMs and CC-1065. Because the amino acid sequence of DARP was identical to hnRNP L, hnRNP L is a candidate protein that binds to the DNA damage induced by DUM. A water-soluble derivative of DUM, KW-2189, exhibits broad spectrum antitumor activity in a series of experimental tumor models and entered clinical trials. KW-2189 was designed as a prodrug to generate active species, DU86, in tumor cells and DARP bounds to the DNA induced by DU86 (unpublished results). Although KW-2189 alkylates DNA in vitro, only the DU86-DNA adduct was detected in the human cells treated with KW-2189.18, 19 The transfection study demonstrated that hnRNP L enhanced the cellular sensitivity to KW2189. As described previously, DARP did not recognize the DNA adducts of cisplatin and mitomycin C in vitro.18 We show that when we examined the transfectants for sensitivity to other DNA-damaging agents, i.e., the major groove binders mitomycin C and cisplatin (data not shown), ectopic hnRNP L expression had no affect on cell sensitivity to them. These results suggest that DARP could be hnRNP L and it acts specifically on DNA damage induced by the minor groove binder.
Other possible mechanisms of increased sensitivity to KW-2189 are: 1) that hnRNP L facilitates transportation of the drug to the nucleus, and 2) that hnRNP L increases the stability of the drug-DNA adduct in a sequence-specific manner.
We have described the difference in intracellular localization of hnRNP L in human lung cancer cell lines. Although there is a report claiming that hnRNP L localized in the nucleoplasm in HeLa cells transfected with hnRNP L,12 we showed that the intracellular localization of hnRNP L differs among human lung cancer cell lines.
There was a report that hnRNP A2 is located in the cytoplasm in post-mitotic phase.20 In this study, few mitotic cells were observed in the culture condition indicating that mitosis was not correlated with hnRNP L distribution. We speculate that in the case of hnRNP A2 a different mechanism might be involved in the intracellular localization of hnRNP L. Nevertheless, synchronization experiments must be examined.
SBC-3 and PC-14 cells grow faster than H69 cells. Even though cell growths of SBC-3 and PC-14 cells were equal in our culture condition, distribution of hnRNP L in these cells were different. This result indicate that the distribution depends on the cell type rather than difference of the cell growth.
To determine whether the localization of hnRNP L is altered by drug exposure, we examined the immunofluorescent staining of hnRNP L in lung cancer cells exposed to KW-2189 for 24 hr. An increased population of cells in which hnRNP L was localized in the nucleus was observed after exposure of a small cell lung cancer (SBC-3) cell line to KW-2189 (data not shown). Although this result was not observed in the rest two cell lines, it can support the hypothesis that hnRNP L helps drugs to transport into nuclear and involves in cell sensitivity mentioned above.
To test the hypothesis that the differences in intracellular localization in lung cancer cells are due to gene alterations, we compared the hnRNP L cDNA sequences in these cell lines. No mutations were detected in any of the lines (data not shown), suggesting that hnRNP L might be co-localized with other proteins. Interaction between hnRNPs has been reported and hnRNP L is known to have a binding domain for interaction with other hnRNPs (e.g., hnRNP I and hnRNP K),21 which are recognized to have NLS. Based on this evidence, the differences in localization of hnRNP L in these cell lines might be due to changes in the molecules that interact with hnRNP L, such as hnRNP I or K. In addition, the putative sites for regulation of localization signal in hnRNP L that we found (25–31, 380–387 and 163–171) would be involved in these interactions. Further studies should extend the potential use of hnRNP L as a factor to assess sensitivity to chemotherapy and candidate molecules for drug development. In addition, expression of hnRNP L needs to be investigated in tissue from lung cancer patients for therapeutic exploitation.
In summary, we have demonstrated the expression of hnRNP L with different intracellular localization in human lung cancer cell lines and that ectopic hnRNP L expression increases cellular sensitivity to a minor groove binder.
The authors are grateful to Dr. G. Dreyfuss for providing anti-hnRNP L antibody.