Skip to main content
NCBI home page
Molecular Pharmacology logoLink to Molecular Pharmacology
. 2009 Oct;76(4):723–733. doi: 10.1124/mol.109.055772

Species-Specific Differences in Translational Regulation of Dihydrofolate Reductase

Yi-Ching Hsieh 1, Nancy E Skacel 1,1, Nitu Bansal 1, Kathleen W Scotto 1, Debabrata Banerjee 1, Joseph R Bertino 1, Emine Ercikan Abali 1,✉
PMCID: PMC2769051  PMID: 19570950

Abstract

We have observed that rodent cell lines (mouse, hamster) contain approximately 10 times the levels of dihydrofolate reductase as human cell lines, yet the sensitivity to methotrexate (ED50), the folate antagonist that targets this enzyme, is similar. Our previous studies showed that dihydrofolate reductase protein levels increased after methotrexate exposure, and we proposed that this increase was due to the relief of feedback inhibition of translation as a consequence of methotrexate binding to dihydrofolate reductase. In the current report, we show that unlike what was observed in human cells, dihydrofolate reductase (DHFR) levels do not increase in hamster cells after methotrexate exposure. We provide evidence to show that although there are differences in the putative mRNA structure between hamster and human mRNA in the dihydrofolate reductase binding region previously identified, “hamsterization” of this region in human dihydrofolate reductase mRNA did not change the level of the enzyme or its induction by methotrexate. Further experiments showed that human dihydrofolate reductase is a promiscuous enzyme and that it is the difference between the hamster and human dihydrofolate reductase protein, rather than the DHFR mRNA, that determines the response to methotrexate exposure. We also present evidence to suggest that the translational up-regulation of dihydrofolate reductase by methotrexate in tumor cells is an adaptive mechanism that decreases sensitivity to this drug.

Dihydrofolate reductase (DHFR, 5, 6, 7, 8-tetrahydrofolate: NADP+ oxidoreductase, EC 1.51.3) catalyzes the reduction of dihydrofolate to tetrahydrofolate using NADPH as a cofactor. Tetrahydrofolate and its one-carbon adducts are essential cofactors in the synthesis of thymidylate, purines, and some amino acids (Blakley, 1984). Inhibition of DHFR results in a depletion of the reduced folate pools, inhibition of DNA synthesis, and cell death. MTX, a tight binding inhibitor of DHFR, is used for the treatment of acute lymphocytic leukemia, non-Hodgkin's lymphoma, osteosarcoma, choriocarcinoma, breast, and head and neck cancer, as well as a variety of other nonmalignant diseases (Chu and Allegra, 1996; Bertino et al., 1997).

Within days after exposure to MTX, DHFR activity increases approximately 6-fold in blast cells of leukemic patients (Bertino et al., 1962). After this initial report, several studies of human cells in vivo and in vitro showed that DHFR protein levels rapidly increase, or are “induced,” in response to MTX treatment (Bertino et al., 1962, 1963; Hillcoat et al., 1967a; Hillcoat and Bertino, 1969). This rapid adaptive response differs from the increase in DHFR because of gene amplification in cells after long-term MTX selection (Alt et al., 1978). The rapid increase in DHFR protein levels seen upon exposure to MTX was unaffected by the transcriptional inhibitor actinomycin D but was blocked by the translational inhibitor cycloheximide (Hillcoat et al., 1967a,b) ruling out an increase in transcription as the mechanism. Because the increase in DHFR is associated with MTX bound to the protein, it was suggested that MTX protected DHFR from degradation (Bertino et al., 1970). Later studies showed that the half-life of DHFR protein was unaltered in the presence of MTX (Domin et al., 1982; Cowan et al., 1986), indicating that protection from degradation was not the explanation for the increase.

The first suggestion that translational regulation may play a role in DHFR induction came from studies showing that DHFR mRNA levels remain the same in the presence or absence of MTX (Bastow et al., 1984; Cowan et al., 1986). Based on studies demonstrating an interaction between DHFR protein and its cognate mRNA, a model was proposed to account for increases in DHFR protein levels in response to MTX exposure, namely that the binding of human DHFR protein to its cognate mRNA results in decreased translation and that the addition of MTX disrupts this autoregulation (Chu et al., 1993; Ercikan et al., 1993). Using an in vitro translation assay, our laboratory (Ercikan et al., 1993) and Chu et al. (1993) demonstrated that addition of exogenous DHFR protein in a rabbit reticulocyte system inhibited translation of DHFR mRNA. using both RNA gel-shift assays and UV cross-linking competition studies, the DHFR binding site was localized to a 100-base pair region in the coding region between nucleotides 380 and 480 (Ercikan-Abali et al., 1997). Computer modeling of mRNA folding in this region revealed two possible stem-loop structures suitable for DHFR binding between nucleotides 407 to 470. We postulated that DHFR binds to its own mRNA within this coding region, resulting in the repression of translation. The addition of inhibitor, substrate, or cofactor alters DHFR conformation, resulting in disruption of this RNA-protein complex and subsequent resumption of translation (Ercikan-Abali et al., 1997). Tai et al. (2004a) using gel-shift and nitrocellulose filter binding assays subsequently localized the binding site to an 82-nucleotide sequence corresponding to nucleotides 401 to 482, consistent with our initial findings.

In this report we show that DHFR induction by MTX differs between human and rodent cells. Although there is an increase in DHFR protein levels in human cell lines in response to MTX treatment, there is no change in DHFR protein levels in either mouse or hamster cells upon MTX treatment. We attribute this difference to differences in the protein structure of DHFR between rodents and human rather than differences in mRNA structure. It is noteworthy that we demonstrate that translational up-regulation of human DHFR decreases the sensitivity of cells to this drug. Finally, we propose that DHFR protein may be a moonlighting enzyme acting as an RNA binding protein.

Materials and Methods

Materials.

CMRL 1066 media, RPMI 1640, fetal bovine serum (FBS), dialyzed FBS, G418 sulfate (Geneticin), penicillin, streptomycin, and trypsin were purchased from Invitrogen (Carlsbad, CA). The pEGFPN3 vector was from Clontech (Mountain View, CA), and the pcDNA3 vector was from Invitrogen. The QuikChange Site-Directed Mutagenesis Kit was purchased from Stratagene (La Jolla, CA). Oligonucleotide primers were synthesized by Integrated DNA Technologies, Inc. (Coralville, IA) or Biosource International (Camarillo, CA). ECL reagents were supplied by GE Healthcare (Chalfont St. Giles, Buckinghamshire, UK). The antibody to green fluorescent protein was from Roche Molecular Biochemicals (Indianapolis, IN). The antibodies used as loading control, anti-β-tubulin antibody, and anti-GAPDH were obtained from Sigma (St. Louis, MO) and Millipore Bioscience Research Reagents (Temecula, CA). The rabbit polyclonal antibody to human DHFR was custom-produced by Research Genetics/Invitrogen, and the monoclonal antibody to DHFR was from BD Biosciences (San Diego, CA). Secondary antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Enzymes were from New England Biolabs (Ipswich, MA). Aprotinin, PMSF, and sodium orthovanadate were obtained from Sigma.

Tissue Culture and Transfections.

The Chinese hamster ovary cell line DG44, deleted for both alleles of DHFR, was a kind gift from Dr. L. Chasin (Columbia University, New York, NY). Parental DG44 cells were maintained in CMRL 1066 media supplemented with 10% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. Human (C85), hamster (CHO), and mouse (3T6) cells were grown in RPMI 1640 supplemented with 10% FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. All transfections were performed using DOTAP transfection reagent (Roche Molecular Biochemicals, Indianapolis, IN). Parental cells were seeded at a density of 2.5 × 104 cells/10-cm dish. When cells reached 10 to 15% confluence, they were transfected with 10 μg of plasmid DNA plus 70 μl of DOTAP reagent according to the manufacturer's protocol. Fresh media were supplied 24 h after transfection. Selection in 750 μg/ml G418 sulfate for plasmids containing the neomycin resistance gene began 48 h later. A parallel mock transfection, which contained everything but plasmid DNA, was performed concurrently with each transfection. Approximately 14 days later, individual single clones could be isolated by ring cylinders and expanded into stable resistant cell lines. At the same time pooled batches were harvested and expanded into stable lines. Neither mock-transfected (only DOTAP reagent) nor parental cells survived the G418 selection.

Preparation of Cell Lysates.

Cells were harvested by 2-min trypsinization and collected by centrifugation (800g) for 5 min. Cell pellets were washed twice in 1× PBS, and resuspended in ice-cold radioimmunoprecipitation assay (PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) containing 100 μg/ml PMSF, 30 μl/ml aprotinin (Sigma), and 1 mM sodium orthovanadate. After 30-min incubation on ice, cells were sonicated three times at 10-s bursts using a VibraCell Sonicator (30% output; Vibracell, Danbury, CT). After sonication, 100 μg/ml PMSF was added, and cells were incubated on ice for 30 min. The lysate was centrifuged at 15,000g for 20 min at 4°C. The supernatant was collected and used immediately or stored at −70°C.

Western Blotting.

Protein samples were resolved on 10% SDS-polyacrylamide gel electrophoresis gels. Transfer was performed by standard electroblotting protocols onto nitrocellulose membranes. Equal loading was determined by using Ponceau S and anti-β tubulin antibody. Immunoblots were incubated with anti-EGFP, anti-DHFR, anti-GAPDH, or anti-β-tubulin antibodies at room temperature for 1 h using buffer containing 0.1% Tween 20 and 5% nonfat dry milk powder. Blots were washed at room temperature and incubated for 1 h with goat anti-mouse IgG as per standard protocols. Chemiluminescence was used to visualize bands according to the manufacturer's instruction (ECL kit; GE Healthcare). Blots were stripped and reprobed according to protocol included with the ECL kit.

Site-Directed Mutagenesis.

The DHFR-EGFP fusion vector generated as described above served as the template in subsequent QuikChange site-directed mutagenesis reactions to create the majority of variants used in this study. Mutagenesis was performed according to the manufacturer's protocol included in the QuikChange kit (Stratagene). In brief, each 50-μl reaction contained 10 ng of double-stranded DNA template (pEGPN3-DHFR), 125 ng of sense mutagenic primer, 125 ng of antisense mutagenic primer, 1 μl of dNTP mix, and 5 μl of 10× manufacturer's reaction buffer, to which 1 μl of Pfu Turbo DNA polymerase (2.5 U/μl) was added. Thermocycling was performed in the GeneAmp 9600 (PerkinElmer/Cetus, Norwalk, CT) thermocycler according to the following parameters: T1, 95°C, 30 s, (1 cycle), followed by T1, 95°C, 30 s; T2, 55°C, 1 min; T3, 68°C, 12 min (16 cycles). After cycling, the reaction was cooled on ice. To digest the parental (nonmutated) template strain, 1 μl of Dpn I restriction enzyme (10 U/μl) was incubated with the reaction for 1 h at 37°C. One microliter of the digested reaction mix containing the mutated plasmid was transformed into XL-1 Blue Supercompetent cells according to standard protocols. Clones were isolated, and each variant was sequenced by University of Medicine and Dentistry of New Jersey–Robert Wood Johnson Medical School DNA Core Facility (Piscataway, NJ) to confirm the presence of the specific mutation. The sequences of the primers used for mutagenesis are shown in Table 1.

TABLE 1.

Primers used in the constructions of variants of human DHFR

Primer
DHFR
    Sense 5′-GGCGGCAAGCTTATGGTTGGTTCGCTAAACTGC-3′
    Antisense 5′-GCCGCCGGATCCATCATTCTTCTCATACTTC-3′
BR1
    Sense 5′-GGAAGCCATGAATCAGCCAGGCCATCTCAGACTCTTTGTGACAAGGATCATGC-3′
    Antisense 5′-GCATGATCCTTGTCACAAAGAGTCTGAGATGGCCTGGCTGATTCATGGCTTCC-3′
BR2
    Sense 5′-CCAGAAATTGATTTGGAGAAATATAAACTTCTCCCAGAATACCCAGGTG-3′
    Antisense 5′-CACCTGGGTATTCTGGGAGAAGTTTATATTTCTCCAAATCAATTTCTGG-3′
T419C
    Sense 5′-CTTAAACTATTTGTGACAAGGATCACGCAAGACTTTGAAAGTGAC-3′
    Antisense 5′-GTCACTTTCAAAGTCTTGCGTGATCCTTGTCACAAATAGTTTAAG-3′
Open in a new tab

Cytotoxicity Assay.

Cytotoxicity to antifolates was determined by the 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt assay according to CellTiter 96 Aqueous One Solution protocol (Promega, Madison, WI). Four thousand cells were seeded in 96-well plates in RPMI 1640 media supplemented with 10% dialyzed FBS, 2 mM l-glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin. The wells in the first row of the plate were used as blank and contained only media but no cells. The wells in the second row of the plate were used as control and contained transfected DG44 cells in the media but no drug. After incubation of the cells with MTX for 96 h, 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, inner salt assay was performed. The absorbance at 490 nm was measured using a plate reader. The percentile cell survival was determined by Softmax Pro software, and sigmoidal dose-response curve fit of the graphs drawn by GraphPad Prism 4 software (GraphPad Software Inc., San Diego, CA) was used to determine EC50 values. All experimental points were set up in replicate eight wells, and all experiments were repeated seven times.

In Vitro Transcription and Translation Assay.

DHFR-EGFP constructs were generated with polymerase chain reaction using a primer at the 5′-end that the T7 promoter. After phenol chloroform extraction and precipitation of the polymerase chain reaction products using 90% ethanol, in vitro transcription was performed using T7 RNA polymerase according to the manufacturer's instruction (Promega). The integrity and the size of the RNA were verified using gel electrophoresis, and the concentration and the purity of RNA samples were determined by reading the absorbance at λ = 260 nm and taking the ratio of λ = 260 nm to λ = 280 nm, respectively.

The nuclease-treated rabbit reticulocyte lysate system (Promega) was used to translate the in vitro-transcribed RNA samples. A 7.5-pmol portion of each transcript, 23 μl of rabbit reticulocyte lysate, 40 U of rRNAsin inhibitor, 1 mM amino acid mixture that lacked methionine, 5 μl of [35S]methionine, and an aliquot of nuclease-free water to a final volume of 55 μl were added, and the reaction mixture was incubated for 30 min at 37°C. The cell-free translation products were run on a 12% SDS-polyacrylamide gel electrophoresis gel and autoradiographed for visualization.

Flow Cytometry.

Expression of EGFP after 1 μM MTX treatment for 48 h was determined using flow cytometry. After trypsinization of the DHFR-EGFP-transfected DG44 cells, the cells were washed and suspended in media. Cytometric analysis was performed using FACSCalibur (BD Biosciences) equipped with CellQuest software. EGFP fluorescence was detected using a 525/50 bandpass filter. The induction efficiency of variants of DHFR is expressed as a percentage (the number of cells expressing EGFP over the total number of cells).

Measurement of DHFR Activity.

The specific activity of DHFR from lysates of hamster and the variants of human DHFR-EGFP was determined as described by Ercikan-Abali et al. (1996).

Results

Rodent DHFR Levels Do Not Increase in the Presence of Antifolates.

DHFR protein levels are up-regulated in response to MTX in all human cell lines tested (Bertino et al., 1962; Bastow et al., 1984; Cowan et al., 1986; Chu et al., 1993; Ercikan-Abali et al., 1997). In contrast, there was no detectable increase in DHFR levels in hamster (CHO) and mouse (3T6) cell lines after exposure to 100 nM MTX for 24 or 48 h, whereas DHFR protein levels were increased in a human colon cancer cell line, C85 (Fig. 1a).

Fig. 1.

Fig. 1.

Open in a new tab

a, Western blot analysis of total cell lysates from human and rodent cell lines after MTX exposure. Human (C85), hamster (CHO), and mouse (3T6) cells were exposed to l μM MTX or media without MTX for 24 and 48 h. b, increase in EGFP fluorescence of various transfectants after 1 μM MTX treatment for 24 h. EGFP and wt human DHFR-EGFP transfectants are negative and positive control, respectively. c, human and hamster DHFR-EGFP transfectants of DG44 cells were exposed to either 1 or 10 μM MTX in the presence of hypoxanthine and thymidine to prevent cell death due to MTX treatment for 24 and 48 h. d, DHFR-EGFP transfectants were exposed to 1 μM MTX, trimetrexate (TMTX), raltitrexed (RTX), and pemetrexed (PTX), respectively. a to c, 50 μg of total cell lysate was loaded and probed with either an anti-DHFR antibody (a) or EGFP antibody (b and c). Equal loading was determined by either α-tubulin or GAPDH, respectively.

We considered several possible explanations for the difference in response of hamster versus human cells after exposure to MTX. Although DHFR levels are regulated at both the transcriptional and translational levels, we had shown previously that steady-state levels of DHFR mRNA are not altered by MTX in either hamster or human cells, suggesting that differential regulation occurs downstream of RNA synthesis/degradation, perhaps at the level of translation. Translation of DHFR mRNA is controlled by sequences within the coding region and the 3′-untranslated region (Mishra et al., 2007). To determine whether the differential response to MTX was regulated by differences in the coding region, chimeric constructs of either human or hamster DHFR coding region fused to EGFP and under the control of the constitutive CMV promoter were constructed. Using a functional rapid fluorescence assay, we tested the relative increase in EGFP fluorescence of the two DHFR-EGFP fusion proteins; EGFP alone was used as a negative control to set the baseline fluorescence. As shown in Fig. 1b, there was a specific increase (>2-fold) in the human DHFR-EGFP fusion protein levels in response to MTX treatment. However, when hamster DHFR-EGFP was transfected into the DG44 cell line, there was no detectable increase in fusion protein levels in response to MTX treatment (Fig. 1, b and c). Moreover, the three mutants of human DHFR (S118A, L22R, and E30A) that are not up-regulated by MTX and other antifolates (Skacel et al., 2004) were also not increased in response to MTX (Fig. 1b). We had reported previously that a mutant DHFR, with its sole cysteine residue changed to either serine or glycine, increased after treatment with MTX and trimetrexate (an antifolate with similar affinity toward DHFR as MTX), but not after treatment with raltitrexed (with thymidylate synthase as its major target (Ki = 1 nM), and DHFR as a secondary target (Ki = 92 nM) (Jackman et al., 1991) or pemetrexed [pemetrexed-polyglutamates potently inhibit both thymidylate synthase (Ki = 1.3 nM) and DHFR (Ki = 7.2 nM) and to a lesser extent, 5-phosphoribosylaminoimidazolecarboxamide transformylase (Ki = 65 nM)] (Skacel et al., 2005). We therefore determined whether DHFR from hamster cells would also show differences in up-regulation to various antifolates. Hamster DHFR-EGFP-transfected cells exposed to MTX, trimetrexate, raltitrexed, and pemetrexed did not show an increase in DHFR, similar to what had been observed for the DHFR mutants that lack negative feedback regulation of DHFR translation (Fig. 1d).

Lack of Up-Regulation of Rodent DHFR by Antifolates Is Due to the Failure of Hamster DHFR Protein to Bind to Its Cognate mRNA.

The current model for translational regulation of DHFR by its cognate mRNA is that DHFR protein interacts through its NADPH binding site within the coding region of its own mRNA (nucleotides 380–480). We proposed that DHFR exists in at least two conformers, one of which has higher affinity for NADPH, and the other for DHFR mRNA. Binding of MTX to the folate binding site of DHFR would result in a change in the conformation of DHFR protein and the release of DHFR mRNA from the DHFR-mRNA complex, hence overcoming translational detainment of the enzyme by its own mRNA (Skacel et al., 2005). The three mutants of human DHFR identified that have no changes within the putative mRNA binding region and yet are not up-regulated by MTX indicate that the lack of up-regulation by these mutants may be due to a change in their protein structure rather than a difference in their mRNAs. Because both the protein and mRNA sequence of hamster DHFR differ from human DHFR, we investigated the role of each in the divergent regulation of human and rodent DHFR. To accomplish this, we initially compared the cis-acting regulatory elements on hamster DHFR mRNA with those of human wt and non-RNA binding mutant DHFR mRNAs.

Our earlier work and studies by Tai et al. (2004b) demonstrated that the DHFR binding site was localized to a 100-base pair sequence in the coding region between nucleotides 380 to 480. Analysis of human DHFR mRNA in this region using the mfold program (Zuker and Stiegler, 1981) revealed two potential stem-loop structures that could form stable secondary structures suitable for DHFR protein binding (ΔG = −12.3 kcal/mol). The difference between hamster and human DHFR in the predicted stem-loop structure is only four nucleotides with 89% homology. Although the hamster DHFR RNA also formed two similar stem-loop structures, the stability of these two stem loops was lower than the stability of the human stem loops (ΔG = −6.1 versus −12.3 kcal/mol). Therefore, we hypothesized that the difference in RNA folding due to four nucleotide differences between human and hamster within this binding region may be responsible for the lack of translational up-regulation of hamster DHFR by antifolates (Fig. 2, a and b). To test this hypothesis, we “hamsterized” the human DHFR-EGFP fusion construct within the region that forms the putative two stem-loop structures by changing the four nucleotides in the human DHFR to the hamster nucleotides. The hamsterized human DHFR-EGFP was transfected into the DG44 cell line, and clones, as well as batches, were selected by resistance to G418, and stable cell lines were generated. Transfected cells were exposed to l μM MTX or media without MTX for 24 and 48 h, and the fusion protein levels were detected using an anti-EGFP antibody. As shown in Fig. 2, c and d, individual clones and the batch of transfected cells with hamsterized human DHFR were induced in response to MTX, indicating that a four-nucleotide difference between hamster and human DHFR mRNA could not explain the difference in response to MTX between the species.

Fig. 2.

Fig. 2.

Open in a new tab

a, comparison of the primary sequence of binding region (BR) of human DHFR to hamster DHFR. b, secondary structures of human and hamster DHFR within the binding region were obtained using the mfold program. c and d, hamsterized human DHFR-EGFP protein levels increase upon exposure to MTX. Batches (c) and four individual clones (d) from DG44 cells transfected with hamsterized DHFR-EGFP. Cells were exposed to 1 μM MTX or media without MTX for 24 and 48 h. Total cell lysate (50 μg) was loaded and probed with an anti-EGFP antibody. Equal loading was determined by Ponceau S staining (data not shown).

Earlier studies indicated that the putative DHFR mRNA binding site is localized to a larger stretch of nucleotides (100 base pairs) than just the nucleotides that formed the stem loop structures (Ercikan-Abali et al., 1997; Tai et al., 2004a). Therefore, we also examined the role of the nucleotides that are not part of the stem and loop structures but still lie within the l00-base pair binding region. Figure 3a shows the sequence of the entire putative binding site that includes nucleotides 380 to 480 within the coding region compared with the hamster DHFR sequence within the same region. There are four-nucleotide differences upstream of the stem-and-loop structures and a one-nucleotide difference downstream. We denoted the hamsterized part of the binding region as the core binding region and using this construct, we performed additional rounds of site-directed mutagenesis using the hamsterized human DHFR-EGFP as the template, generating three additional chimeric constructs. One construct, a “half-mutated” binding region (BR1) had the core four nucleotides and the four additional upstream nucleotides hamsterized. A second construct, “half-mutated” binding region (BR2) had the four core nucleotides hamsterized and the downstream nucleotide that was not conserved between human and hamster (G to C). The third construct, a fully mutated binding region (FMBR) DHFR-EGFP, had all nine nucleotides hamsterized within the binding region of human DHFR mRNA. After constructing these vectors, we transfected them into DG44 cells and derived stable clones and cell lines. Treatment of the cells with 1 μM MTX for 24 and 48 h resulted in induction of DHFR-EGFP levels in all cell lines (Fig. 3b). Thus, the difference in primary sequence within the putative DHFR binding regions of the human and hamster DHFRs did not account for the species difference in MTX-induced DHFR increases. Therefore we considered that the difference might lie within the DHFR protein itself, in that the human protein might have characteristics lacking in the hamster DHFR protein that allows interaction with its cognate mRNA.

Fig. 3.

Fig. 3.

Open in a new tab

a, experimental design for fully mutated binding region (FBMR) and half-mutated binding regions (BR1 and BR2). Human DHFR-EGFP with FMBR and human DHFR-EGFP with a half-mutated binding regions demonstrate increases in fusion protein levels upon exposure to MTX. b, batches from DG44 cells transfected with an FMBR, BR1, and BR2 human were exposed to 1 μM MTX or media without MTX for 24 and 48 h. Total cell lysate (50 μg) was loaded and probed with an anti-EGFP antibody. Equal loading was determined by Ponceau S staining (data not shown).

We began to test this new hypothesis by analyzing a number of human DHFR mutants that alter the amino acid sequence. Tai and colleagues (2004b) showed that a mutation in nucleotide 419 (U to C) dramatically reduced the binding of wt human DHFR to its cognate mRNA. However it was not clear whether the loss of binding between DHFR protein with its cognate mRNA correlated with the loss of MTX-induced translational regulation. Therefore we created the same mutation and tested its ability to be induced by MTX. As shown in Fig. 4, DHFR 419 U→C, (DHFR M140T mutant) was still up-regulated by MTX. We then examined the human DHFR S118A mutant, which has characteristics similar to those of hamster DHFR; i.e., both are not up-regulated by MTX (Fig. 5a). DG44 cells that were already transfected with the mutant S118A-EGFP were transfected with a Flag-tagged wt DHFR protein expression vector. This allowed us to observe the changes in both human wt and mutant DHFR protein (Fig. 5b). Five individual cell lines established from stable transfectants containing both Flag-tagged DHFR and the mutant S118A-EGFP demonstrated increased fusion protein levels upon exposure to 1 μM MTX for 24 h. The levels of S118A-EGFP fusion protein increased in response to MTX comparable with Flag-tagged wt DHFR in the doubly transfected cell lines. Moreover, the basal level of S118A protein levels in doubly transfected cell lines were significantly reduced compared with the singly transfected cell line, indicating the restoration of negative feedback regulation of the human DHFR S118A mutant by expression of wt human DHFR in these cells. Hence, introduction of wt human DHFR into the S118A-transfected cells re-established the ability of this mutant to be induced by MTX (Fig. 5, a and b).

Fig. 4.

Fig. 4.

Open in a new tab

The only ATG within the binding region was mutated to ACG, which resulted in M140T. After exposure to 0.1 and 1 μM MTX for 24 and 48 h, human DHFR variant M140I resulted in increased levels of DHFR-EGFP fusion protein. Western blots were performed on total cell lysates from DG44 cells transfected with M140T. Cells were exposed to MTX or media without MTX in the presence of hypoxanthine and thymidine to prevent cell death due to MTX treatment. Total cell lysate (50 μg) was loaded and probed with an anti-EGFP antibody. Equal loading was determined by GAPDH antibody.

Fig. 5.

Fig. 5.

Open in a new tab

wt human DHFR restores the lack of feedback regulation in hamster and human DHFR S118A variant, both of which are not responsive to MTX-induced up-regulation. a, DG44 cells stably transfected with DHFR S118A-EGFP were exposed to 1 μM MTX for 24 h, demonstrating the lack of up-regulation of DHFR by MTX. DHFR S118A-EGFP levels were detected using antibody against EGFP. Equal loading was determined with GAPDH antibody. b, DG44 cells transfected with DHFR S118A-EGFP were transfected once more with Flag-tagged wt human DHFR. Although Flag-tagged wt human DHFR protein levels were detected with anti-Flag antibody, mutant DHFR S118A-EGFP levels were detected with anti-EGFP antibody. Five stable clones of doubly transfected cell lines were exposed to 1 μM MTX for 48 h, and DHFR protein was detected using Western blotting. Graph depicts the quantitative analysis of the data in b. The intensity of DHFR S118A-EGFP protein bands was normalized to the intensity of GAPDH. c, C85 cells transfected with hamster DHFR-EGFP were exposed to 0.l and 1 μM MTX for 48 h. Total cell lysate (50 μg) was loaded and probed with either DHFR antibody for endogenous levels of wt human DHFR and EGFP antibody for wt hamster-EGFP fusion protein. To control for equal loading, blots were stripped and reprobed with an α-tubulin antibody.

In a second experiment, we used a metastatic human colon cancer cell line (C85) in which the endogenous DHFR levels are up-regulated by MTX (Fig. 1a). These cells were transfected with either wt human or hamster DHFR-EGFP, and batches resistant to G418 were expanded. Polyclonal DHFR antibody was used to detect endogenous human DHFR, and anti-EGFP antibody was used to detect the fusion protein of wt hamster DHFR-EGFP. When these cells were exposed to 0.1 or 1 μM MTX for 48 h, there was a marked increase in hamster DHFR-EGFP protein levels in response to MTX treatment (Fig. 5c). Thus, in the presence of human DHFR protein, hamster DHFR protein was now induced by MTX, indicating that human DHFR protein is somewhat promiscuous (i.e., it also binds to DHFR hamster DHFR mRNA and inhibits its synthesis), similar to the results obtained with the human DHFR S118A variant that is not responsive to MTX-induced up-regulation (Fig. 5a). Therefore wt human DHFR protein by binding to both wt hamster DHFR and the human DHFR S118A variant mRNA regulated the translation of these mRNAs and DHFR levels were increased in the presence of MTX.

Translational Regulation of Human DHFR by MTX Is an Intrinsic Mechanism of Resistance.

Cells transfected with the three human DHFR mutants that lacked translational feedback regulation had a higher cellular baseline level of DHFR than cells transfected with wt DHFR (Skacel et al., 2004). Because hamster DHFR baseline levels are also high, we hypothesized that the reason for the increased level of DHFR in hamster cells compared with human cells was a difference in translation efficiency. Therefore, we compared the translational efficiency of hamster and human mutant DHFR mRNAs. When equal concentrations of in vitro-transcribed DHFRs (human wt, S118A, L22R, and E30A, and hamster DHFR transcripts) were prepared and translated using a rabbit reticulocyte assay, the human mutant and hamster wt DHFR transcripts were translated to a greater extent than wt human DHFR mRNA (Fig. 6a). The half-lives of the mutant and the wt human and hamster DHFRs were compared. After treating the transfected DG44 cells with cycloheximide to inhibit protein synthesis, cell lysates were collected at different time points, and DHFR levels were analyzed using Western blotting. There was no significant difference in half-lives among the mutants of human DHFR that are not induced by MTX, wt human, and hamster DHFR (data not shown). Therefore, the higher basal levels of the variants of human DHFR and hamster DHFR may be attributed to the fact that these proteins do not bind to their own mRNA to inhibit translation. Next, we determined the basal levels and specific activity of DHFR in DG44 cells transfected with either wt human DHFR-EGFP or wt hamster DHFR-EGFP. The basal levels were established by titrating wt hamster DHFR levels and by comparing 30 μg of total lysate of wt human DHFR with wt hamster DHFR (Fig. 6b). The basal level of wt hamster DHFR was 10-fold higher than that of wt human DHFR transfectants. In addition, DHFR protein levels were comparable with the specific activity obtained for human and hamster DHFR (Fig. 6c). The kinetic properties of human and hamster DHFR are similar (the Km values of dihydrofolate for human and hamster DHFR are 0.02–0.37 and 0.42 μM, respectively, and the Km values for NADPH for human and hamster are 0.16–1.44 and 0.64 μM, respectively) (Schweitzer et al., 1989; Lewis et al., 1995; Ercikan-Abali et al., 1996; Wu et al., 1997). In addition, there was no statistical difference (p = 0.26) in the ED50 values for MTX between DG44 transfectants of human and hamster DHFR when one-tailed paired t test was used to analyze the data using GraphPad Prism 4 (Fig. 6d). The observation that hamster DHFR is expressed at ∼10 times the level of human DHFR, but human DHFR is induced ∼10-fold by MTX, indicates that human cells respond to MTX by translationally up-regulating DHFR levels and that this mechanism is an intrinsic mechanism of resistance to antifolates. Hence, the up-regulation of DHFR protein in the presence of MTX leads to decreased sensitivity to this drug.

Fig. 6.

Fig. 6.

Open in a new tab

a, synthesis of wt human, hamster, and variants of human DHFR-EGFP using an in vitro reticulocyte translation assay. b, DHFR-EGFP protein levels of hamster is higher than the wt human DHFR-EGFP. Serial dilutions of hamster DHFR-EGFP lysates were compared with 30 μg of wt human DHFR-EGFP transfected cells. Western blots were probed with an anti-EGFP antibody. Loading was controlled using GAPDH antibody. c, DHFR activity of hamster and human DHFR variants that are not up-regulated by MTX are higher than the wt DHFR-EGFP. d, cytotoxicity of MTX to DG44 cells transfected with wt human and hamster DHFR-EGFP.

Discussion

Earlier, we observed that DHFR in rodent (mouse and hamster) cells did not increase after MTX treatment compared with DHFR levels in human, monkey, and dog cells (Skacel et al., 2004). Herein, we demonstrate that this is due to a critical difference in the translational regulation of human and rodent DHFR. Based on the aforementioned model, we considered two explanations for the lack of MTX-induced translational regulation of hamster DHFR: 1) a difference between the cis-acting regulatory elements on hamster versus human DHFR mRNA; or 2) a difference in the trans-acting regulatory domain of hamster DHFR versus human DHFR. To test the first possibility, we compared the human DHFR mRNA cis element with that of the hamster. Target elements for RNA binding proteins are generally found in loops, bulges, or interior loops, which offer conformations for recognition and binding (Frankel, 2000). Modeling studies using Zuker's RNA mfold program identified two putative stem-and-loop structures within this region for both species. However, these structures had significant species-specific differences in their predicted stability, with the human structures approximately twice as stable as the comparable structures in rodent DHFR mRNA structures. The bulges within these structures were also dissimilar. However, “hamsterizing” the human protein by replacing the human sequence with the respective hamster nucleotides had no affect on the ability of MTX to up-regulated the hamsterized DHFR, strongly suggesting that the observed difference in DHFR mRNA structure was not the explanation for the lack of up-regulation of hamster DHFR.

Next we addressed the second possibility that the divergent translational regulation between hamster and human DHFR is due to the differences in the trans-acting regulatory (i.e., protein) domains between human and hamster DHFR. Previously, we identified a variant of human DHFR (S118A) residing in the NADPH binding pocket was similar to wt human DHFR in its enzymatic properties and yet was not up-regulated by MTX (Skacel et al., 2005). We hypothesized that this human DHFR mutant might mimic the hamster DHFR, perhaps in a shared inability to bind to DHFR mRNA, thereby affecting translational control. If this were the case, the wild-type human DHFR protein should be able to compensate for the hamster and S118A defects, bind to their respective mRNAs, and regulate their translation. Using cotransfection experiments, we indeed showed that both hamster and S118A DHFR protein levels were elevated after MTX administration in the presence of wt human DHFR, strongly implicating differences in DHFR protein/conformation in the lack of a translational autoregulatory mechanism for the mutant and hamster DHFR sequence. By extrapolation, human DHFR protein is an RNA-binding protein, whereas hamster is not. These findings also may explain the increased basal levels of hamster DHFR protein and variants of human DHFR that are not up-regulated by MTX (Skacel et al., 2005). There is growing evidence that many enzymes have additional regulatory functions that are not related to their enzymatic activities. These enzymes have been referred to as “moonlighting” enzymes (Jeffery, 1999). Thus, the ability of DHFR to regulate its translation would indicate that DHFR may also be a moonlighting enzyme.

The involvement of the DHFR NADPH binding site in autoregulation of DHFR through RNA binding is notable but not unprecedented. A number of dehydrogenases have been shown to interact with RNA through their NAD/NADPH binding sites (Hentze, 1994; Nagy et al., 2000). One of the most noteworthy, glyceraldehyde-3-phosphate dehydrogenase, binds tRNA, adenylate uridylate-rich sequences at the 3′-untranslated region of mRNA, hammerhead ribozyme, and viral cis-acting regulatory elements (Evguenieva-Hackenberg et al., 2002). The possibility that human DHFR may bind to other mRNA species is currently being explored.

From a clinical standpoint, the demonstrated ability of MTX to up-regulate human wt DHFR could result in increased resistance of human tumor cells to MTX treatment. Higher DHFR levels are correlated with increased resistance to MTX; therefore, it would be reasonable to expect that the DG44 cells transfected with hamster DHFR would be more resistant to MTX than the DG44 cells transfected with human DHFR. However, despite the divergent basal levels (∼10-fold) of DHFR, IC50 values for MTX in both cell lines were statistically similar because of translational up-regulation of human DHFR, but not hamster DHFR, in the presence of MTX. Therefore, in contrast to the genetic change (gene amplification) that results in resistance after multiple exposures to MTX, translational up-regulation of human DHFR by antifolates is a rapid adaptive mechanism of resistance. Although this mechanism may protect normal renewal tissues from the toxicity of MTX, it may also decrease the sensitivity of tumor cells to therapy. This adaptive mechanism of human DHFR was also shown with pyrimethamine, an antifolate that is used in the treatment of malaria. The malaria parasite, Plasmodium falciparum, has a bifunctional DHFR-thymidylate synthase enzyme synthesized from the same mRNA transcript. This transcript binds to the bifunctional enzyme, but administration of antifolates for the treatment of malaria does not release DHFR-TS mRNA from binding to its protein Therefore, although the parasite is killed by pyrimethamine, the ability of the human host to up-regulate DHFR allows the host to overcome drug toxicity by increasing human DHFR protein levels (Zhang and Rathod, 2002).

We conclude that divergent translational regulation of human and hamster DHFR is due to the differences in their protein properties rather than differences in their mRNAs (Fig. 7). Not only does human DHFR bind to its cognate mRNA, it also binds to hamster DHFR mRNA. Translational regulation of human DHFR allows cells to rapidly respond to cytotoxic effects of MTX; hence, this is an adaptive mechanism. We are currently testing novel antifolates that do not share MTX's ability to up-regulate DHFR levels and thus may be more effective anticancer agents.

Fig. 7.

Fig. 7.

Open in a new tab

Model for divergent translational regulation of rodent and human DHFR. wt human DHFR mRNA translation is feedback-regulated through the interaction of DHFR protein with its cognate mRNA. The cis-acting regulatory elements on human DHFR mRNA were localized within the coding region, and the trans-acting regulatory domain was suggested to be within the NADPH binding site of DHFR. wt human DHFR protein has at least two conformations, one of which is bound preferentially to NADPH, and the other is bound to DHFR mRNA. These two conformers are in equilibrium and can interconvert. Binding of MTX to the binary complex of DHFR-NADPH shifts the equilibrium toward the NADPH-bound conformer, releasing DHFR mRNA to be translated. However, three variants of human DHFR protein (S118A, L22R, and E30A) and hamster are unable to bind their own cognate mRNA, which leads to increased synthesis of these proteins. Therefore, although binding of MTX to the wt human DHFR-mRNA complex leads to resumption of DHFR synthesis, hamster and three variants of human DHFR proteins are not induced by MTX.

This work was supported by the National Institutes of Health National Cancer Institute [Grant CA08010]; the U.S. Public Health Service; and the Department of Medicine Award from the Robert Wood Johnson Medical School–University of Medicine and Dentistry of New Jersey.

Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org.

ABBREVIATIONS:
DHFR
dihydrofolate reductase
MTX
methotrexate
EGFP
enhanced green fluorescent protein
CHO
Chinese hamster ovary
PMSF
phenyl methyl sulfonyl fluoride
PBS
phosphate-buffered saline
FBS
fetal bovine serum
ECL
enhanced chemiluminescence
GAPDH
glyceraldehyde-3-phosphate dehydrogenase
wt
wild type
FMB
fully mutated binding region
DOTAP
N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate.

References

  1. Alt FW, Kellems RE, Bertino JR, Schimke RT. (1978) Selective multiplication of dihydrofolate reductase genes in methotrexate-resistant variants of cultured murine cells. J Biol Chem 253:1357–1370 [PubMed] [Google Scholar]
  2. Bastow KF, Prabhu R, Cheng YC. (1984) The intracellular content of dihydrofolate reductase: possibilities for control and implications for chemotherapy. Adv Enzyme Regul 22:15–26 [DOI] [PubMed] [Google Scholar]
  3. Bertino JR, Cashmore AR, Hillcoat BL. (1970) “Induction” of dihydrofolate reductase: purification and properties of the “induced” human erythrocyte and leukocyte enzyme and normal bone marrow enzyme. Cancer Res 30:2372–2378 [PubMed] [Google Scholar]
  4. Bertino JR, Donohue DR, Gabrio BW, Silber R, Alenty A, Meyer M, Huennekens FM. (1962) Increased level of dihydrofolic reductase in leucocytes of patients treated with amethopterin. Nature 193:140–142 [DOI] [PubMed] [Google Scholar]
  5. Bertino JR, Kamen BA, Romanini A. (1997) Folate antagonists, in Cancer Medicine, 4th ed ( Holland JF, Bast RC, Jr, Morton DL, Frei E, 3rd, Kufe DW, Weichselbaum RR. eds) pp 907–922, Williams and Wilkins, Baltimore: [Google Scholar]
  6. Bertino JR, Silber R, Freeman M, Alenty A, Albrecht M, Gabrio BW, Huennekens FM. (1963) Studies on normal and leukemic leukocytes. IV. Tetrahydrofolate-dependent enzyme systems and dihydrofolic reductase. J Clin Invest 42:1899–1907 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Blakley RL. (1984) Dihydrofolate reductase, in Folates and Pterins (Blakley RL, Benkovic SJ. eds) pp 191–253, Wiley, New York: [Google Scholar]
  8. Chu E, Takimoto CH, Voeller D, Grem JL, Allegra CJ. (1993) Specific binding of human dihydrofolate reductase protein to dihydrofolate reductase messenger RNA in vitro. Biochemistry 32:4756–4760 [DOI] [PubMed] [Google Scholar]
  9. Chu EA, Allegra CJ. (1996) Antifolates, in Cancer Chemotherapy and Biotherapy: Principles and Practice (Chabner BA, Longo DL. eds) pp 109–148, Lippincott-Raven, Philadelphia: [Google Scholar]
  10. Cowan KH, Goldsmith ME, Ricciardone MD, Levine R, Rubalcaba E, Jolivet J. (1986) Regulation of dihydrofolate reductase in human breast cancer cells and in mutant hamster cells transfected with a human dihydrofolate reductase minigene. Mol Pharmacol 30:69–76 [PubMed] [Google Scholar]
  11. Domin BA, Grill SP, Bastow KF, Cheng YC. (1982) Effect of methotrexate on dihydrofolate reductase activity in methotrexate-resistant human KB cells. Mol Pharmacol 21:478–482 [PubMed] [Google Scholar]
  12. Ercikan-Abali EA, Banerjee D, Waltham MC, Skacel N, Scotto KW, Bertino JR. (1997) Dihydrofolate reductase protein inhibits its own translation by binding to dihydrofolate reductase mRNA sequences within the coding region. Biochemistry 36:12317–12322 [DOI] [PubMed] [Google Scholar]
  13. Ercikan-Abali EA, Waltham MC, Dicker AP, Schweitzer BI, Gritsman H, Banerjee D, Bertino JR. (1996) Variants of human dihydrofolate reductase with substitutions at leucine-22: effect on catalytic and inhibitor binding properties. Mol Pharmacol 49:430–437 [PubMed] [Google Scholar]
  14. Ercikan E, Banerjee D, Waltham M, Schnieders B, Scotto KW, Bertino JR. (1993) Translational regulation of the synthesis of dihydrofolate reductase. Adv Exp Med Biol 338:537–540 [DOI] [PubMed] [Google Scholar]
  15. Evguenieva-Hackenberg E, Schiltz E, Klug G. (2002) Dehydrogenases from all three domains of life cleave RNA. J Biol Chem 277:46145–46150 [DOI] [PubMed] [Google Scholar]
  16. Frankel AD. (2000) Fitting peptides into the RNA world. Curr Opin Struct Biol 3:332-340 [DOI] [PubMed] [Google Scholar]
  17. Hentze MW. (1994) Enzymes as RNA-binding proteins: a role for (di)nucleotide-binding domains? Trends Biochem Sci 19:101–103 [DOI] [PubMed] [Google Scholar]
  18. Hillcoat BL, Bertino JR. (1969) Vriation of dihydrofolate reductase activity in mammalian cells in vitro. Exp Cell Res 56:304–308 [DOI] [PubMed] [Google Scholar]
  19. Hillcoat BL, Perkins JP, Bertino JR. (1967a) Dihydrofolate reductase from the L1210 R murine lymphoma. Further studies on the binding of substrates and inhibitors to the enzyme. J Biol Chem 242:4777–4781 [PubMed] [Google Scholar]
  20. Hillcoat BL, Swett V, Bertino JR. (1967b) Increase of dihydrofolate reductase activity in cultured mammalian cells after exposure to methotrexate. Proc Natl Acad Sci U S A 58:1632–1637 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jackman AL, Taylor GA, Gibson W, Kimbell R, Brown M, Calvert AH, Judson IR, Hughes LR. (1991) ICI D1694, a quinazoline antifolate thymidylate synthase inhibitor that is a potent inhibitor of L1210 tumor cell growth in vitro and in vivo: a new agent for clinical study. Cancer Res 51:5579-5586 [PubMed] [Google Scholar]
  22. Jeffery CJ. (1999) Moonlighting proteins. Trends Biochem Sci 24:8–11 [DOI] [PubMed] [Google Scholar]
  23. Lewis WS, Cody V, Galitsky N, Luft JR, Pangborn W, Chunduru SK, Spencer HT, Appleman JR, Blakley RL. (1995) Methotrexate-resistant variants of human dihydrofolate reductase with substitutions of leucine 22. Kinetics, crystallography, and potential as selectable markers. J Biol Chem 270:5057–5064 [DOI] [PubMed] [Google Scholar]
  24. Mishra PJ, Humeniuk R, Mishra PJ, Longo-Sorbello GS, Banerjee D, Bertino JR. (2007) A miR-24 microRNA binding-site polymorphism in dihydrofolate reductase gene leads to methotrexate resistance. Proc Natl Acad Sci U S A 104:13513–13518 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Nagy E, Henics T, Eckert M, Miseta A, Lightowlers RN, Kellermayer M. (2000) Identification of the NAD(+)-binding fold of glyceraldehyde-3-phosphate dehydrogenase as a novel RNA-binding domain. Biochem Biophys Res Commun 275:253–260 [DOI] [PubMed] [Google Scholar]
  26. Schweitzer BI, Srimatkandada S, Gritsman H, Sheridan R, Venkataraghavan R, Bertino JR. (1989) Probing the role of two hydrophobic active site residues in the human dihydrofolate reductase by site-directed mutagenesis. J Biol Chem 264:20786–20795 [PubMed] [Google Scholar]
  27. Skacel N, Ercikan-Abali EA, Banerjee B, Bertino JR. (2004) Human dihydrofolate reductase (DHFR) protein can regulate translation of hamster DHFR by virtue of its promiscuity. Proc Am Assoc Cancer Res 51:982 [Google Scholar]
  28. Skacel N, Menon LG, Mishra PJ, Peters R, Banerjee D, Bertino JR, Abali EE. (2005) Identification of amino acids required for the functional up-regulation of human dihydrofolate reductase protein in response to antifolate Treatment. J Biol Chem 280:22721–22731 [DOI] [PubMed] [Google Scholar]
  29. Tai N, Schmitz JC, Chen TM, Chu E. (2004a) Characterization of a cis-acting regulatory element in the protein-coding region of human dihydrofolate reductase mRNA. Biochem J 378:999–1006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Tai N, Schmitz JC, Liu J, Lin X, Bailly M, Chen TM, Chu E. (2004b) Translational autoregulation of thymidylate synthase and dihydrofolate reductase. Front Biosci 9:2521–2526 [DOI] [PubMed] [Google Scholar]
  31. Wu JW, Wang ZX, Zhou JM. (1997) Inactivation kinetics of dihydrofolate reductase from Chinese hamster during urea denaturation. Biochem J 324:395–401 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Zhang K, Rathod PK. (2002) Divergent regulation of dihydrofolate reductase between malaria parasite and human host. Science 296:545–547 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Zuker M, Stiegler P. (1981) Optimal computer folding of large RNA sequences using thermodynamics and auxiliary information. Nucleic Acids Res 9:133–148 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Molecular Pharmacology are provided here courtesy of American Society for Pharmacology and Experimental Therapeutics

RESOURCES