The former annotated human pseudogene dihydrofolate reductase ...

The former annotated human pseudogene dihydrofolate reductase-like 1 (DHFRL1) is expressed and functional Gráinne McEntee, Stefano Minguzzi, Kirsty O’Brien, Nadia Ben Larbi, Christine Loscher, Ciarán Ó’Fágáin, and Anne Parle-McDermott1 School of Biotechnology, Dublin City University, Glasnevin, Dublin 9, Ireland

Human dihydrofolate reductase (DHFR) was previously thought to be the only enzyme capable of the reduction of dihydrofolate to tetrahydrofolate; an essential reaction necessary to ensure a continuous supply of biologically active folate. DHFR has been studied extensively from a number of perspectives because of its role in health and disease. Although the presence of a number of intronless DHFR pseudogenes has been known since the 1980s, it was assumed that none of these were expressed or functional. We show that humans do have a second dihydrofolate reductase enzyme encoded by the former pseudogene DHFRP4, located on chromosome 3. We demonstrate that the DHFRP4, or dihydrofolate reductase-like 1 (DHFRL1), gene is expressed and shares some commonalities with DHFR. Recombinant DHFRL1 can complement a DHFR-negative phenotype in bacterial and mammalian cells but has a lower specific activity than DHFR. The K m for NADPH is similar for both enzymes but DHFRL1 has a higher K m for dihydrofolate when compared to DHFR. The need for a second reductase with lowered affinity for its substrate may fulfill a specific cellular requirement. The localization of DHFRL1 to the mitochondria, as demonstrated by confocal microscopy, indicates that mitochondrial dihydrofolate reductase activity may be optimal with a lowered affinity for dihydrofolate. We also found that DHFRL1 is capable of the same translational autoregulation as DHFR by binding to its own mRNA; with each enzyme also capable of replacing the other. The identification of DHFRL1 will have implications for previous research involving DHFR.

R

ecent knowledge of the dihydrofolate reductase (DHFR) gene family suggests one functional gene among several intronless pseudogenes (1, 2). The functional DHFR gene resides on chromosome 5 (1) and encodes an enzyme that catalyzes the reduction of dihydrofolate to the biologically active form, tetrahydrofolate. The DHFR gene/enzyme has been studied extensively in relation to health and disease given its crucial role in folate metabolism (3), use as an antifolate drug target (4), and as a commonly used reporter gene for molecular studies. Folate mediated one-carbon metabolism is a cellular pathway where the essential B vitamin folate acts as a cofactor for a variety of anabolic and catabolic reactions (5). This pathway is essential for the supply of cofactors for purine/pyrimidine synthesis, cellular methylation reactions, and the supply of formylated methionine for protein synthesis in the mitochondria. The DHFR enzyme forms part of folate metabolism, ensuring there is a supply of the biologically active form of folate, i.e., tetrahydrofolate. Up to now, DHFR was thought to be the only enzyme that could not only recycle folate metabolites back to tetrahydrofolate, but also reduce the synthetic form of folate, folic acid. This enzyme activity is significant given the widespread worldwide mandatory and voluntary folic acid fortification of foods that has occurred in recent years as a preventative measure against the occurrence of neural tube defects (6). Despite the importance of DHFR activity, recent work has demonstrated that human liver DHFR activity was quite variable between individuals and had limited www.pnas.org/cgi/doi/10.1073/pnas.1103605108

ability to reduce folic acid when compared to the rat version of the enzyme (7). Absence of DHFR activity leads to a rapid depletion of tetrahydrofolate and a consequent cessation in de novo DNA synthesis and cell proliferation. This effect has led to the development of a range of antifolate drugs that target DHFR (and other folate enzymes). Methotrexate is one such drug and has been in use in chemotherapy for more than 50 years. Cells, however, can become drug resistant through mutation or amplification of the DHFR gene (4). Normally, however DHFR expression is tightly controlled at the transcriptional, translational, and posttranslational level. Transcriptional control during the cell cycle is mediated by the transcription factors Sp1 and E2F (8, 9) plus a noncoding RNA that is transcribed from the minor promoter (10). Regulation of DHFR at the translational level involves the binding of the DHFR protein to its own mRNA (11). The initial response of cells to methotrexate exposure is to upregulate DHFR protein level. This upregulation is thought to be mediated at the translational level (12–14); likely due to a conformational change of the DHFR mRNA complex (11, 15). At the posttranslational level recent evidence suggests that DHFR is subject to both monoubiquitination and sumoylation (16, 17). These posttranslational modifications are thought to be important for its localization at specific phases of the cell cycle. DHFR has also been reported to be regulated posttranslationally by p14ARF by an unknown indirect mechanism that affects protein stability (18). It is clear that DHFR has been extensively investigated, but all the work to date has assumed that humans have just one expressed and functional DHFR. DHFR on chromosome 5 was thought to be the only human enzyme capable of carrying out the reduction of dihydrofolate to tetrahydrofolate as the four reported pseudogenes were regarded as nonfunctional (Table S1). The intronless nature of the four DHFR pseudogenes indicates that they arose through reintegration of an mRNA intermediate (2). Although there are other dihydrofolate reductase-like sequences in other species, this particular reintegration event may have been a primate-specific event as similar intronless pseudogenes of DHFR are not evident in nonprimate species (www.ensembl.org, blast.ncbi.nlm.nih.gov/Blast.cgi). The DHFRP1 pseudogene located on chromosome 18 is polymorphic in the human population which is indicative of its recent evolutionary origins (19). The open reading frame (ORF) of DHFRP1 is identical to the functional DHFR Author contributions: A.P.-M. designed research; G.M., S.M., K.O., and N.B.L. performed research; C.L. supervised confocal microscopy; C.O. supervised enzyme kinetics analysis; G.M., S.M., K.O., N.B.L., and A.P.-M. analyzed data; and G.M., S.M., and A.P.-M. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. 1

To whom correspondence should be addressed. E-mail: [email protected].

This article contains supporting information online at www.pnas.org/lookup/suppl/ doi:10.1073/pnas.1103605108/-/DCSupplemental.

PNAS ∣ September 13, 2011 ∣ vol. 108 ∣ no. 37 ∣ 15157–15162

BIOCHEMISTRY

Edited by Stephen J. Benkovic, Pennsylvania State University, University Park, PA, and approved July 20, 2011 (received for review March 8, 2011)

but despite this similarity, it does not appear to have a functional promoter and there is no evidence to suggest it is expressed. In this study, we provide evidence that in fact the pseudogene, formerly known as DHFRP4 (DHFRL1) on chromosome 3 is not only expressed, but that the translated protein (i) harbors enzyme activity, (ii) can complement a DHFR null phenotype, (iii) is likely to auto regulate itself and DHFR, and (iv) localizes to the mitochondria. Results Confirmation of Expression of DHFRL1 by Quantitative Reverse Transcribed PCR (RT-qPCR) and Sequencing. A large scale cDNA sequen-

cing project (20) was the first indication that the DHFRL1 (or DHFRP4) pseudogene was actually expressed. Annotation of the DHFRL1 mRNA entry (NM_176815) suggests that there are two transcripts produced by the DHFRL1 gene that differ in their 5′ untranslated (UTR) regions. Both transcripts would produce the same protein sequence. We designed a successful RT-qPCR assay to specifically amplify transcript variant 2 without a possibility of erroneous amplification of genomic DNA or other DHFR homologous sequences. The RT-qPCR assay was optimized and a number of human cell lines were screened for expression including SW480, SKBR3, L428, DG75, BT474, National Cancer Institute H1299, and Coriell lymphoblast cell lines. The DHFRL1 transcript was expressed in all the cell lines tested at either a similar or lower level (relative ratio ranged from 1.0 to 0.11) to the relatively abundant control transcript glucuronidase beta (Fig. S1). The highest level of expression was observed in cell line BT474. Direct sequencing of the purified PCR product confirmed that the DHFRL1 assay was specific and not amplifying the functional DHFR gene (Fig. S2). This data confirmed that the DHFRL1 gene was being transcribed. An examination of the DHFRL1 entry in the Unigene database (http://www.ncbi.nlm.nih.gov/UniGene/ ESTProfileViewer.cgi?uglist=Hs.718516) indicates that DHFRL1 is expressed in a variety of normal human tissues and developmental states.

Sequence Analysis of DHFRL1. Comparison of the amino acid sequences of DHFR with DHFRL1 shows that they are 92% identical (Fig. S3). The four motifs required for dihydrofolate reductase activity are conserved except for three amino acid residues. The most significant of these is a conserved tryptophan at position 24 (W24); DHFRL1 has an arginine (R) at this position. Previous site-directed mutagenesis experiments of the human enzyme showed that replacement of this tryptophan with phenylalanine (F) resulted in a 50% decrease in stability and a drop in efficiency of 48% under intracellular conditions (21). This data suggested that the W24 to arginine (R) change in DHFRL1 would result in an enzyme capable of dihydrofolate reductase activity but with altered catalytic characteristics versus wild type. The DHFRL1 sequence is preserved from primates to humans (www.ensembl.org) with the same R24 change preserved in the chimpanzee. This conservation may indicate that this amino acid change is significant for a functional role of DHFRL1 that may be distinct from DHFR, i.e., a type of subfunctionalization. The amino acid sequences necessary for DHFR mRNA binding (15) and sumoylation (17) are all conserved in DHFRL1 indicating that this new dihydrofolate reductase is subject to similar translational regulation and posttranslational modifications. Additional sequence analysis of DHFRL1 indicates that the translational signals required are present within the 5′ UTR of DHFRL1. A comparison of the DHFR and DHFRL1 5′ UTR sequences surrounding the initiating ATG encoding methionine shows that DHFRL1 differs at just one base: ctgtcAUGt (DHFRL1) versus ctgtcAUGg (DHFR). A reassessment of the translation initiation codons in vertebrates (22) found that approximately 50% of cytoplasmic translated transcripts do not contain a “g” at the +4 position. 15158 ∣

www.pnas.org/cgi/doi/10.1073/pnas.1103605108

Complementation of a DHFR-Negative Phenotype in a Bacterial System. Escherichia coli D3-157 is a streptomycin-resistant bacterial

strain that has been mutated so that it no longer has DHFR activity and requires the addition of thymidine to the medium for growth (23). To determine if our recombinant DHFRL1 has any enzyme activity, we transformed our DHFRL1 recombinant clone (in pCR2.1) into E. coli D3-157 strain and grew cultures in media both in the presence/absence of thymidine and/or isopropyl β-D-1-thiogalactopyranoside (IPTG) and ampicillin. The presence of ampicillin selects for those bacteria that have been transformed with pCR2.1 vector whereas IPTG induces expression of the recombinant protein. As a positive control the strain was also transformed with a DHFR recombinant clone (also in pCR2.1). Our other controls included untransformed cultures and cultures transformed with empty pCR2.1 vector. The results are shown in Fig. 1. As expected, both the original D3-157 strain and bacteria transformed with the empty pCR2.1 vector survived only in media where thymidine was present. They quickly died in the media without thymidine. Our positive control, D3-157 transformed with DHFR, also behaved as predicted and grew in media containing IPTG (plus ampicillin) but without thymidine. D3-157 cells transformed with DHFRL1 behaved similar to the positive control and grew in media both with and without thymidine. The conclusion drawn from this experiment is that recombinant DHFRL1 has sufficient DHFR enzyme activity to complement a DHFR-negative phenotype in a bacterial system. Complementation of a DHFR-Negative Phenotype in a Mammalian System. CHO DG44 cells are Chinese hamster ovary cell mutants

lacking in dihydrofolate reductase (24). Similar to the E. coli D3-157 bacterial cell line, these cells required the media to be supplemented with thymidine and hypoxanthine to grow. Having shown complementation of the phenotype in a bacterial system we wanted to replicate those results in a mammalian system. The CHO DG44 cells were transfected with a mammalian expression vector with either a DHFR or a DHFRL1 insert. The cells transfected with DHFR were used as a positive control, normal untransfected CHO DG44 cells were a negative control, and cells transfected with an empty vector acted as a quality control for the experiment. After transfection, cells were left in complete growth medium for 48 h. Transfected cells were then positively selected by exploiting the neomycin resistance gene on the expression vector by adding 500 μg∕mL G418. At this stage the untransfected CHO DG44 cells quickly died off. The cells were left in selective media for a further 14 d at which point they were switched to media containing G418 but without thymidine or hypoxanthine supplements. The cells remained in the complementation media for 12 d. Cells were counted using trypan blue on days 1, 5, and 12. Results are shown in Fig. 2. On day 1 both DHFR and DHFRL1 transfected cells had approximately 2 × 106 cells∕mL, with cells transfected with empty vector having slightly less at 1.4 × 106 cells∕mL. Cells transfected with empty vector did not survive and were completely dead by day 5. Cells transfected with either DHFR or DHFRL1 did show significant cell death by day 5, however, they had recovered sufficiently by day 12. The cell death in transfected cells is likely to be related to how the plasmid was incorporated into the genome. It is possible that for these cells the neomycin resistance gene could be active whereas the DHFR or DHFRL1 gene was silenced. If this type of gene activation was the case those cells would have survived in the selection media which had the required supplements but were unable to survive once the supplements were removed. At day 12, cells transfected with DHFR had recovered their numbers to what they were on day 1. Cells transfected with DHFRL1 were a lot slower to recover. This data may indicate that although DHFRL1 does have enzyme activity it is not as active as DHFR. This theory was tested by harvesting protein from transfected cells at day 12 in complementation media and carrying out an McEntee et al.

Fig. 1. Complementation of DHFR-negative phenotype in a bacterial system. E. coli D3-157 streptomycin-resistant cells were transformed with either DHFR or DHFRL1 and grown in media with and without supplements/antibiotics. Strep; streptomycin 100 μg∕mL; Amp, ampicillin 100 μg∕mL; IPTG ¼ 0.2 mM. Growth was measured at various time points until stationary phase was reached. (A) The original strain grew only in media containing thymidine and without ampicillin. (B) Cells transformed with empty vector only grew in the presence of thymidine with or without ampicillin and/or IPTG. (C) Cells transformed with DHFR grew as expected in media both with and without supplements, i.e., could grow in the absence of thymidine once DHFR expression was induced. (D) Cells transformed with DHFRL1 also complemented the phenotype similarly to recombinant DHFR.

Characterization of Recombinant DHFRL1 Protein. DHFR enzyme acts by reducing dihydrofolate into tetrahydrofolate in the presence of NADPH. To determine if DHFRL1 has similar enzyme activity we produced a purified recombinant DHFRL1 protein with a GST tag that we subsequently cleaved off. We then tested this recombinant protein for enzyme activity using a standard, and compared the results to a recombinant DHFR protein produced in the same manner. Initial results indicated that DHFRL1 protein did have enzyme activity. However the specific activity of DHFRL1 was roughly two-thirds (70%) that of DHFR (Table 1).

BIOCHEMISTRY

enzyme assay to test for activity (Fig. S4). As expected both DHFR and DHFRL1 transfected cells showed enzyme activity, as measured by a decrease in absorbance at 340 nm over a 10 min time period. DHFRL1 activity begins to level off after 5 min whereas DHFR still shows strong enzyme activity after 10 min. Measurement of specific activity showed that cells transfected with DHFR (0.5832 μmol∕ min ∕mg) had approximately five times higher activity than that of DHFRL1 (0.154697 μmol∕ min ∕mg). These results correlate with the specific activity measured using recombinant purified protein (see below).

Fig. 2. Complementation of DHFR-negative phenotype in a mammalian system. Cell counts of transfected CHO DG44 cells after switching cells to media without supplements. Cells were counted after 1, 5, and 12 d in complementation media. Cells transfected with either DHFR (A) or DHFRL1 (B) had some cell death on day 5; however, by day 12 both sets of cells had recovered and were growing well in the complementation media. The cells transfected with DHFR grew more quickly than those transfected with DHFRL1. (C) Cells transfected with the empty vector alone did not survive without supplements.

McEntee et al.

PNAS ∣ September 13, 2011 ∣

vol. 108 ∣

no. 37 ∣

15159

Table 1. Kinetic analysis of recombinant DHFRL1 versus DHFR Enzyme DHFR 2.5–300 μM‡ DHFRL1 2.5–700 μM‡

NADPH†

Dihydrofolic acid* K m μM 20.1–2.1¶ 209.3–18.5 ‡

V max μmol∕ min ∕mL 0.0132–0.0004 0.0226–0.0004

K m μM 3.6–0.4 3.4–0.7 §

V max μmol∕ min ∕mL 0.021–0.0004 0.038–0.001

Specific activity μmol∕ min ∕mg 6.1–0.3 4.3–0.3

*At constant 50 μM NADPH † At constant 60 μM dihydrofolic acid ‡ Dihydrofolic acid concentration ranges ¶ Standard error § NADPH concentration ranged from 2.5–150 μM

Having shown that recombinant DHFRL1 did have enzyme activity we then went on to characterize the enzyme by calculating K m values for both the cofactor NADPH and the substrate dihydrofolic acid (Table 1). The K m values of both enzymes for the cofactor NADPH were similar. For the substrate dihydrofolate, however, DHFRL1 displayed a K m value of 209.3 μM versus the DHFR value of 20.1 μM (Table 1). The altered values of DHFRL1 compared to DHFR may be driven by the W24R change in DHFRL1, but this suggestion requires further investigation. DHFRL1 has the Ability to Bind its own mRNA and that of DHFR.

Previous studies have established that the DHFR protein can act as an RNA binding protein and bind to its own mRNA to suppress translation (11). The amino acids essential for this process have also been identified in previous work (15, 25). Our sequence analysis of DHFRL1 shows that those essential amino acids are also present in the DHFRL1 protein (Fig. S3). Moreover, the 27-nt mRNA sequence of DHFR, which was shown to be strictly necessary for the binding with DHFR protein (25), differs in only one nucleotide from the DHFRL1 sequence. For these reasons, we hypothesized that DHFRL1 protein may bind to its own mRNA and also may have the ability to bind DHFR mRNA. Initially we tested the ability of DHFRL1 protein to bind to DHFR mRNA by electrophoretic mobility shift assay. We used DHFR binding to DHFR mRNA as a positive control. Results shown in Fig. 3A clearly show a mobility shift for both the positive control and for DHFRL1 protein. We expanded the experiment to include DHFRL1 mRNA and tested not only the ability of DHFRL1 protein to bind to its own mRNA but also explored the possibility that DHFR protein could bind to DHFRL1 mRNA (Fig. 3B). As expected, DHFRL1 protein did bind to DHFRL1 mRNA and a clear mobility shift can be seen for this sample. A clear shift can also be seen for the sample containing DHFRL1 mRNA and DHFR protein, indicating that DHFR protein can also bind to DHFRL1 mRNA as well as its own mRNA. We repeated the EMSA experiment but included purified GST protein as an additional negative control to ensure that the binding of DHFR and DHFRL1 was not being mediated by the GST tag (Fig. S5). No mobility shift was detected for DHFR nor DHFRL1 mRNA when only GST protein was added. From these experiments we conclude that recombinant DHFRL1 protein acts as an RNA binding protein not only for its own mRNA but also for DHFR mRNA. We have also shown that, in turn, DHFR protein can also bind to DHFRL1 mRNA in addition to its own mRNA. These results indicate that DHFRL1 can not only regulate its own translation but may also play a role in the regulation of DHFR protein translation. Subcellular Localization of DHFRL1. Folate enzymes are known to reside in the cytoplasm, nucleus, and mitochondria (26). DHFR has previously been reported to localize primarily in the cytoplasm but a small percentage has been shown to go to the nucleus at the synthesis phase of the cell cycle (17). We examined the subcellular localization of DHFRL1 in unsynchronized HEK293 cells that were transiently transfected with GFP-DHFRL1 and examined using immunofluoresence and confocal microscopy. 15160 ∣


Staining of the nuclei demonstrated that DHFRL1 does not localize to the nucleus in these unsynchronized cells. However, intense fluorescence was detected in the mitochondria (Fig. 4) demonstrating that DHFRL1 localizes to this organelle. Discussion We have demonstrated that the former human pseudogene DHFRP4 or DHFRL1 is not only expressed but is functional. Previously, DHFR was thought to be the only enzyme capable of dihydrofolate reductase activity in humans. The data that we have presented refutes this long held assumption and the sequence similarity between DHFR and DHFRL1 indicates that much of the previous work on DHFR may well have been unable to distinguish between the two forms. DHFRL1 is capable of complementing DHFR knockout phenotypes in both bacterial and mammalian cells (Fig. 1, Fig. 2, and Fig. S4). However, our data also indicate that although both enzymes share commonalities they have distinct differences in their affinity for dihydrofolate. Gene duplication is a major contribution to diversity of function (27). The neofunctionalization model allows for the gain of a new function in one of the duplicates, which is thought to occur through an initial relaxation of selective constraints (28). Therefore, new genes can arise through duplications that will drive the evolution and adaptation of a species. The duplication event that gave rise to DHFRL1 may have happened after primates

Fig. 3. EMSA shows that DHFRL1 and DHFR can bind to their own and each other’s mRNA. All binding reactions were resolved on a 4% nondenaturing polyacrylamide gel, then transferred to a nylon membrane for detection using the LightShift® Chemiluminescent EMSA Kit (Thermo Scientific). Band shifts are indicated by the arrows. (A) EMSA involving DHFR mRNA. A clear band shift is only observed in the presence of recombinant DHFR or DHFRL1 (lanes 2 and 4). Lane order, 1: DHFR mRNA only; 2: DHFR mRNA + DHFR; 3: DHFR mRNA + DHFR + unlabeled DHFR mRNA; 4: DHFR mRNA + DHFRL1; 5: DHFR mRNA + DHFRL1 + unlabeled DHFRL1 mRNA. (B) EMSA involving DHFRL1 mRNA. A clear band shift is only observed in the presence of recombinant DHFR or DHFRL1 (lanes 7 and 9). Lane order, 6: DHFRL1 mRNA only; 7: DHFRL1 mRNA + DHFR; 8: DHFRL1 mRNA + DHFR + unlabeled DHFR mRNA; 9: DHFRL1 mRNA + DHFRL1; 10: DHFRL1 mRNA + DHFRL1 + unlabeled DHFRL1 mRNA.

McEntee et al.

diverged from their most recent common ancestor. An mRNA copy of DHFR reintegrated into the genome possibly before primates diverged from humans, giving rise to DHFRL1 plus three additional intronless pseudogenes (Table S1). A number of amino acid substitutions accumulated in DHFRL1, the most significant that we noted appears to be replacement of the conserved tryptophan in the catalytic site (Fig. S3). Under our conditions of measurement, the impact of the replacement of tryptophan with arginine (W24R) appears to result in a drop in the specific activity of DHFRL1 by nearly 30% compared to DHFR (Table 1). The K m for NADPH was not significantly different. DHFRL1, however, consistently showed a notably higher K m value than DHFR for dihydrofolate. This difference possibly underlies the apparently lower specific activity of DHFRL1, but further investigation of this point would require a complete kinetic analysis. Apart from enzyme activity, the amino acid conservation between DHFR and DHFRL1 also indicated that DHFRL1 may also be capable of binding its own mRNA in a similar fashion to DHFR; resulting in suppression of translation (11). Our EMSA analysis demonstrated that this proposal is in fact the case and that each enzyme can substitute for each other, i.e., DHFRL1 can bind DHFR mRNA and DHFR can bind DHFRL1 mRNA (Fig. 3). This binding is significant, particularly if the enzymes have different affinities for antifolate drugs such as methotrexate. The initial response of cells to methotrexate is to upregulate protein levels through disruption of the DHFR∶mRNA complex (29). Our finding that DHFRL1 can also prevent DHFR mRNA translation (and vice versa) indicates that this autoregulation mechanism and response to methotrexate needs to be reconsidered. We also examined the cellular localization of DHFRL1. The compartmentalization of folate metabolism has been previously established (26), with several mitochondrial-specific enzymes identified in recent years (30–33). Moreover, a small fraction of DHFR has been reported to localize to the nucleus during the synthesis phase of the cell cycle (17). In this context, we considered whether DHFRL1 can also localize to the mitoMcEntee et al.

Materials and Methods Quantitative Reverse Transcribed PCR. Assay details and confirmation of DHFRL1 mRNA expression are detailed in SI Materials and Methods. Sequence Analysis. Alignment of the amino acid sequences for DHFR (P00374.2) and DHFRL1 (AAH63379.1) were carried out using CLUSTAL 2.0.08. Relevant catalytic motifs were identified using PRINTS (www.bioinf. manchester.ac.uk/dbbrowser/PRINTS/). Other relevant amino acid residues were identified from the literature (15, 17, 21, 38–40). The Ensembl genome browser (www.ensembl.org) and BLAST (blast.ncbi.nlm.gov/Blast.cgi) were used to examine DHFRL1 sequences in other species. PNAS ∣ September 13, 2011 ∣

vol. 108 ∣

no. 37 ∣

15161

BIOCHEMISTRY

Fig. 4. Localization of GFP-DHFRL1 in mitochondria by immunofluorescence microscopy. HEK293 cells were transiently transfected with GFP-DHFRL1 and visualized by confocal microscopy. The top left image shows GFP-DHFRL1 (green). The bottom left image shows mitochondria stained with MitoTracker CMTMRos (red). The top right image is the merged image; arrows show localization of GFP-DHFRL1 in the mitochondria. The bottom right image is the differential interference contrast (DIC) of the cells.

chondria or nucleus. We found no evidence that DHFRL1 localizes to the nucleus in unsynchronized cells, however, analysis of cells from the synthesis phase of the cell cycle would be required to definitely rule this localization out. We found clear evidence that DHFRL1 localizes to the mitochondria (Fig. 4). Although DHFRL1 does possess a potential sumoylation site, its localization to the mitochondria indicates that its primary role is to support mitochondrial DNA synthesis and replication. The relevance of DHFRL1 mRNA binding (discussed above) is not immediately apparent, given its mitochondrial localization. It is unlikely that the entire complement of DHFRL1-translated protein is present in the mitochondria at any one time and that a small percentage remains in the cytoplasm. This putative cytoplasmic percentage provides the opportunity for DHFRL1 to target DHFRL1 and DHFR mRNA. Moreover, the localization of DHFRL1 may also change at different phases of the cell cycle, in response to the extracellular environment or in a pathological situation. The fact that DHFR can also bind DHFR and DHFRL1 mRNA has obvious biological relevance given that DHFR principally resides in the cytoplasm. The mechanism of mitochondrial localization is unclear given the lack of any obvious sequence that would indicate that DHFRL1 is targeted to the mitochondria. Mitochondria targeting sequences (MTS) do not appear to share a consensus primary sequence rather they share similar overall characteristics. Most MTS lack acidic residues and are rich in the positively charged amino acids arginine and lysine and the hydroxylated amino acids serine and threonine (34). A secondary feature common among MTS is the formation of an amphiphilic α helices found on the surface of the protein (35). Hurt and Schatz found that amino acids 1–85 on mouse DHFR had the potential to be an MTS; however, it was inactive within the folded protein (36). These amino acids are highly conserved within DHFRL1 and, therefore, the localization of DHFRL1 in the mitochondria may be related to the folding of the enzyme following translation, possibly revealing the presence or absence of amino acids that facilitate its import (37). However, this suggestion requires further investigation. The identification and localization of DHFRL1 emphasizes the importance of the mitochondria in folate metabolism and will inform current research in this area. The conservation of DHFRL1 from primates to humans indicates that this second human dihydrofolate reductase enzyme has a specific role to play. The identification of DHFRL1 now means that DHFR regulation, function, and antifolate drug responses will have to be reassessed in the context of its paralogue. A differential response to antifolate drugs such as methotrexate may lead the way for more improved therapeutic treatments. Why do we need a second DHFR enzyme? The answer to this question is likely to relate to its subcellular localization and possibly its tissue specificity. A DHFR with reduced affinity for its substrate may be a specific requirement for one-carbon flux through the mitochondria. Its localization to the mitochondria adds to the recent list of folate enzymes that have also been identified in this organelle and highlights the importance of mitochondrial folate metabolism. It has been assumed that the DHFR pseudogenes are nonfunctional and thus irrelevant. Our data show that this assumption is incorrect and that the human DHFR enzyme is not alone.

Expression and Purification of Recombinant DHFR and DHFRL1. The expression vectors were constructed similar to the approach described by Wang et al. (41) and are described in SI Materials and Methods. The induced GST-DHFRL1 or -DHFR fusion proteins were purified using glutathione agarose [Invitrogen catalog (cat.) no. G2879] and further purified as described in SI Materials and Methods. Enzyme Activity Assay and Determination of K m Values. Enzyme activity and K m values were tested using a Dihydrofolate Assay Kit (Sigma cat. no. CS03040-1KT) as described in SI Materials and Methods. Complementation of DHFR-Negative Phenotype in a Bacterial System. Recombinant clones for DHFRL1 and DHFR were constructed as detailed in SI Materials and Methods. Recombinant plasmid DNA was transformed into a DHFR-negative E. coli cell line D3-157 (American Type Culture Collection cat. no. 47050). Cultures were then grown for 72 h in media containing streptomycin (100 μg∕mL) and various combinations of Ampicillin (100 μg∕mL), IPTG (0.2 mM), and Thymidine (50 μg∕mL). Samples were taken at 0, 2, 4, 6, 22, 30, 48, and 72 h and growth was measured in a spectrophotometer at 600 nm.

EMSA. EMSA analysis was carried using a Pierce ®RNA 3′ End Biotinylation Kit (Thermo Scientific cat. no. 20160) plus purified recombinant DHFRL1 or DHFR. Details are provided in SI Materials and Methods. Localization of DHFRL1 Protein. The Invitrogen DHFRL1 ORF clone (cat. no. IOH26763) was inserted into pcDNA6.2/N-EmGFP-DEST (Invitrogen, cat. no. V356-20) expression vector using the Gateway Cloning System. Plasmid DNA was isolated using a Qiagen Mini Prep Kit (cat. no. 12123). HEK293 cells, grown on cover slips (1 × 105 cells∕mL), were transfected with 5 μg of plasmid DNA using Lipofectamine 2000 reagent (Invitrogen, cat. no. 11668500). Six hours after transfection, the transfection medium was removed and replaced with complete growth medium. The cells were retransfected again 24 h after the initial transfection using the same conditions. A further 48 h later, the cells were incubated for 20 min at 37 °C with MitoTraker CMTMROS (Molecular Probes Invitrogen, cat. no. M7512) at 200 nM. The cells were then fixed in paraformaldehyde on ice for 30 min. Following rinsing 3 × 5 min in PBS baths, the cover slips were mounted on slides with antifade medium (Dako). Slide preparations were observed using a Zeiss Axio Observer. Z1 equipped with a Zeiss 710 and ConfoCor3 laser scanning confocal head (Carl Zeiss, Inc.). Images were analyzed using Zen 2008 software.

Complementation of DHFR-Negative Phenotype in a Mammalian System. Mammalian expression vectors of DHFRL1 and DHFR were constructed and transfected as detailed in SI Materials and Methods.

ACKNOWLEDGMENTS. This work was funded by the Health Research Board of Ireland, HRB/2009/54.

1. Maurer BJ, Barker PE, Masters JN, Ruddle FH, Attardi G (1984) Human dihydrofolate reductase gene is located in chromosome 5 and is unlinked to the related pseudogenes. Proc Natl Acad Sci USA 81:1484–1488. 2. Anagnou NP, et al. (1984) Chromosomal organization of the human dihydrofolate reductase genes: Dispersion, selective amplification, and a novel form of polymorphism. Proc Natl Acad Sci USA 81:5170–5174. 3. Litwack G, ed. (2008) Folic Acid and Folates. Vitamins and Hormones (Elsevier, New York), 79. 4. Bertino JR (2009) Cancer research: From folate antagonism to molecular targets. Best Pract Res Clin Haematol 22:577–582. 5. Bailey LN, ed. (1995) Folate in Health and Disease (Marcel Dekker, Inc., New York), pp 1–22. 6. Czeizel AE, Dudas I (1992) Prevention of the first occurrence of neural-tube defects by periconceptional vitamin supplementation. N Engl J Med 327:1832–1835. 7. Bailey SW, Ayling JE (2009) The extremely slow and variable activity of dihydrofolate reductase in human liver and its implications for high folic acid intake. Proc Natl Acad Sci USA 106:15424–15429. 8. Jensen DE, Black AR, Swick AG, Azizkhan JC (1997) Distinct roles for Sp1 and E2F sites in the growth/cell cycle regulation of the DHFR promoter. J Cell Biochem 67:24–31. 9. Johnson LF, Fuhrman CL, Wiedemann LM (1978) Regulation of dihydrofolate reductase gene expression in mouse fibroblasts during the transition from the resting to growing state. J Cell Physiol 97:397–406. 10. Martianov I, Ramadass A, Barros AS, Chow N, Akoulitchev A (2007) Repression of the human dihydrofolate reductase gene by a non-coding interfering transcript. Nature 445:666–670. 11. Ercikan-Abali EA, et al. (1997) Dihydrofolate reductase protein inhibits its own translation by binding to dihydrofolate reductase mRNA sequences within the coding region. Biochemistry 36:12317–12322. 12. Hillcoat BL, Swett V, Bertino JR (1967) Increase of dihydrofolate reductase activity in cultured mammalian cells after exposure to methotrexate. Proc Natl Acad Sci USA 58:1632–1637. 13. Cowan KH, et al. (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. 14. 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. 15. Skacel N, et al. (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. 16. Maguire M, et al. (2008) MDM2 regulates dihydrofolate reductase activity through monoubiquitination. Cancer Res 68:3232–3242. 17. Anderson DD, Woeller CF, Stover PJ (2007) Small ubiquitin-like modifier-1 (SUMO-1) modification of thymidylate synthase and dihydrofolate reductase. Clin Chem Lab Med 45:1760–1763. 18. Magro PG, Russo AJ, Li W-W, Banerjee D, Bertino JR (2004) p14ARF expression increases dihydrofolate reductase degradation and paradoxically results in resistance to folate antagonists in cells with nonfunctional p53. Cancer Res 64:4338–4345. 19. Anagnou NP, Antonarakis SE, O’Brien SJ, Modi WS, Nienhius AW (1988) Chromosomal localization and racial distribution of the polymorphic human dihydrofolate reductase pseudogene (DHFRP1). Am J Hum Genet 42:345–352.

20. Baross A, et al. (2004) Systematic recovery and analysis of full-ORF human cDNA clones. Genome Res 14:2083–2092. 21. Beard WA, et al. (1991) Role of the conserved active site residue tryptophan-24 of human dihydrofolate reductase as revealed by mutagenesis. Biochemistry 30:1432–1440. 22. Peri S, Pandey A (2001) A reassessment of the translation initiation codon in vertebrates. Trends Genet 17:685–687. 23. Singer S, Ferone R, Walton L, Elwell L (1985) Isolation of a dihydrofolate reductasedeficient mutant of Escherichia coli. J Bacteriol 164:470–472. 24. Urlaub G, Chasin LA (1980) Isolation of Chinese hamster cell mutants deficient in dihydrofolate reductase activity. Proc Natl Acad Sci USA 77:4216–4220. 25. Tai N, Ding Y, Schmitz JC, Chu E (2002) Identification of critical amino acid residues on human dihydrofolate reductase protein that mediate RNA recognition. Nucleic Acids Res 30:4481–4488. 26. Appling DR (1991) Compartmentation of folate-mediated one-carbon metabolism in eukaryotes. FASEB J 5:2645–2651. 27. Ohno S (1970) Evolution by Gene Duplication (Springer, New York), pp 23–30. 28. Hughes T, Liberles DA (2007) The pattern of evolution of smaller-scale gene duplicates in mammalian genomes is more consistent with neofunctionalization than subfunctionalization. J Mol Evol 65:574–588. 29. Ercikan E, et al. (1993) Translational regulation of the synthesis of dihydrofolate reductase. Adv Exp Med Biol 338:537–540. 30. Peri KG, Belanger C, Mackenzie RE (1989) Nucleotide sequence of the human NAD-dependent methylene tetrahydrofolate dehydrogenase-cyclohydrolase. Nucleic Acids Res 17:8853. 31. Christensen KE, Patel H, Kuzmanov U, Mejia NR, MacKenzie RE (2005) Disruption of the Mthfd1 gene reveals a monofunctional 10-formyltetrahydrofolate synthetase in mammalian mitochondria. J Biol Chem 280:7597–7602. 32. Prassannan P, Pike S, Peng K, Shane B, Appling DR (2003) Human mitochondrial C1-tetrahydrofolate syntase. J Biol Chem 278:43178–43187. 33. Stover PJ, et al. (1997) Molecular cloning, characterization, and regulation of the human mitochondrial serine hydroxymethyltransferase gene. J Biol Chem 272: 1842–1848. 34. Hurt EC, van Loon APGM (1986) How proteins find mitochondria and intramitochondrial compartments. Trends Biochem Sci 11:204–207. 35. Omura Tsuneo (1998) Mitochondria-targeting sequence, a multi-role sorting sequence recognized at all steps of protein import into mitochondria. J Biochem 123:1010–1016. 36. Hurt EC, Schatz G (1987) A cytosolic protein contains a cryptic mitochondrial targeting signal. Nature 325:499–503. 37. Yogev O, Pines O (2011) Dual targeting of mitochondrial proteins: Mechanism, regulation, and function. Biochim Biophys Acta 1808:1012–1020. 38. Jackson RC, Hart LI, Harrap KR (1976) Intrinsic resistance to methotrexate of cultured mammalian cells in relation to the inhibition kinetics of their dihydrofolate reductases. Cancer Res 36:1991–1997. 39. Simonsen CC, Levinson AD (1983) Isolation and expression of an altered mouse dihydrofolate reductase cDNA. Proc Natl Acad Sci USA 80:2495–2499. 40. Thillet J, Absil J, Stone SR, Pictet R (1988) Site-directed mutagenesis of mouse dihydrofolate reductase. J Biol Chem 263:12500–12508. 41. Wang Y, Buruenn JA, Queener SF, Cody V (2001) Isolation of rat dihydrofolate reductase gene and characterization of recombinant enzyme. Antimicrob Agents Chemother 45:2517–2523.

15162 ∣


McEntee et al.