Alternative promoter usage and alternative splicing contribute to mRNA heterogeneity of mouse monocarboxylate transporter 2

Shelley X. L. Zhang, Tina R. Searcy, Yiman Wu, David Gozal, Yang Wang


Expression patterns of monocarboxylate transporter 2 (MCT2) display mRNA diversity in a tissue-specific fashion. We cloned and characterized multiple mct2 5′-cDNA ends from the mouse and determined the structural organization of the mct2 gene. We found that transcription of this gene was initiated from five independent genomic regions that spanned >80 kb on chromosome 10, resulting in five unique exon 1 variants (exons 1a, 1b, 1c, 1d, and 1e) that were then spliced to the common exon 2. Alternative splicing of four internal exons (exons AS1, AS2, AS3, and exon 3) greatly increased the complexity of mRNA diversity. While exon 1c was relatively commonly used for transcription initiation in various tissues, other exon 1 variants were used in a tissue-specific fashion, especially exons 1b and 1d that were used exclusively for testis-specific expression. Sequence analysis of 5′-flanking regions upstream of exons 1a, 1b, and 1c revealed the presence of numerous potential binding sites for ubiquitous transcription factors in all three regions and for transcription factors implicated in testis-specific or hypoxia-induced gene expression in the 1b region. Transient transfection assays demonstrated that each of the three regions contained a functional promoter and that the in vitro, cell type-specific activities of these promoters were consistent with the tissue-specific expression pattern of the mct2 gene in vivo. These results indicate that tissue-specific expression of the mct2 gene is controlled by multiple alternative promoters and that both alternative promoter usage and alternative splicing contribute to the remarkable mRNA diversity of the gene.

  • gene expression
  • transcriptional regulation
  • mRNA diversity
  • lactate metabolism
  • hypoxia

lactate movement across the cellular membrane (“intercellular lactate shuttle”) relies on a specific transport mechanism that is afforded by the proton-linked monocarboxylate transporters (MCTs) (22, 27). There is also evidence that these transporters may be responsible for lactate movement across the mitochondrial membrane (“intracellular lactate shuttle”), which facilitates lactate oxidation (7, 12, 33). MCTs catalyze the facilitated diffusion of lactate with a proton, a process that does not require energy input other than that provided by the concentration gradients of lactate and protons. Recent studies have implicated a critical role for MCTs in coordinating lactate metabolism under both physiological and pathological conditions. For example, in skeletal muscle cells the amount of MCT correlates with the capacity of lactate oxidation (35), while in cardiac myocytes MCT is operating close to its maximal capacity and may become rate limiting under hypoxic conditions (58). It remains largely unknown, however, how lactate transporters are regulated under conditions such as hypoxia/ischemia, in which enhanced lactate metabolism is desired (23, 33, 48, 52, 62). Interestingly, a recent study reported that at least one of the genes coding for lactate transporters was upregulated by hypoxia at the transcriptional level via a hypoxia-inducible factor 1 (HIF-1)-dependent mechanism (56).

The MCT gene family comprises multiple members, among which MCT1 (SLC16A1) and MCT2 (SLC16A7) have been relatively well characterized. These two transporters, although being similar in substrate selectivity and transport capability, have significantly different tissue/cell distribution, prompting the hypothesis that members of the MCT gene family perform distinct physiological functions (19, 29, 46). This hypothesis has been supported by recent studies showing functional relevance of neuron-specific expression of MCT2 and astroglial cell-specific expression of MCT1 in lactate traffic between these two cell types in the brain (11, 28, 39, 44). Furthermore, evidence suggests that MCT1 and MCT2 are differentially regulated by distinct molecular mechanisms. Notably, while the MCT1 gene produces a single mRNA transcript in various tissues under the control of a single promoter (15, 21), the MCT2 gene produces a variety of mRNA transcripts with different molecular weights whose relative abundance is tissue dependent (25, 29, 31, 40). However, the molecular mechanisms that underlie MCT2 mRNA diversity and the biological relevance of the various mRNA species in determining the overall expression of MCT2 are currently unknown.

We hypothesized that the MCT2 gene was a complex locus whose expression was controlled by alternative promoters and alternative splicing patterns. In the present study, we cloned and characterized multiple mct2 5′-cDNA ends from the mouse and determined the structural organization of the mouse mct2 gene. We demonstrate that the mouse mct2 gene is controlled by at least five alternative promoters that initiate transcription of this gene from distinct locations in a large genomic region and give rise to a variety of mRNA transcripts with distinct 5′-ends. Alternative splicing of multiple internal exons further increases the complexity of mRNA diversity. We further demonstrate that alternative promoter usage is responsible for the reported tissue-specific expression pattern of this important gene.


5′-Rapid amplification of cDNA ends.

Tissue samples from the brain, the skeletal muscle, the lung, and the testis, which are known to possess active lactate transporter activity, were harvested from normal male C57BL/6 mice. Total cellular RNA was extracted using the TRI reagent (Molecular Research Center). An improved 5′-rapid amplification of cDNA ends (RACE) protocol was performed using the RLM-RACE kit (Ambion) to isolate full-length 5′-cDNA ends. Briefly, total cellular RNA was treated with calf intestinal alkaline phosphatase to remove the 5′-phosphate from all 5′-truncated mRNA transcripts that lack the cap structure. The RNA samples were then treated with tobacco acid pyrophosphatase to remove the cap structure from transcripts with a complete 5′-end, leaving an intact 5′-monophosphate. An RNA adapter was ligated to those transcripts with the intact 5′-monophosphate, and the RNA samples were reverse transcribed using decamer random primers and Superscript III reverse transcriptase (Invitrogen). The resulting cDNA pool was used for the first round of PCR amplification using a sense primer compatible with the 5′-part of the adapter (5′-GCTGATGGCGATGAATGAACACTG-3′) and a gene-specific antisense primer localized in exon 5 of the mouse mct2 gene (GSP1, 5′-CAATGGAGATAAAGGACGCACAGA-3′) (29). The PCR product was used as templates for the second round of PCR using a nested sense primer compatible with the 3′-part of the adapter (5′-CGCGGATCCGAACACTGCGTTTGCTGGCTTTGATG-3′) and a nested gene-specific antisense primer localized in exon 4 of the mouse mct2 gene (GSP2, 5′-CCAACACACTACTGATGGGACCTC-3′) (29). Products from the second round of PCR were then subcloned into the pCR II vector (TA cloning kit, Invitrogen) and subject to DNA sequence analysis on both strands. The design of the RACE strategy was such that all mRNA diversities upstream of exon 4, where translation of mct2 was initiated would be identified (see Fig. 1A).

Fig. 1.

Mouse monocarboxylase transporter 2 (mct2) 5′-mRNA structure. A: various 5′-leader sequences are spliced to a common exon 2, which is then spliced, with or without the alternatively spliced exon 3, to exon 4 where translation initiates. Shaded area represents the coding region. The size of each exon is indicated. *31 represents the size of the noncoding region of exon 4. Bent arrows represent transcription start sites. Locations of the gene-specific antisense primers used in the 5′-rapid amplification of cDNA ends (RACE) are depicted as arrowheads GSP1 and GSP2. B: nucleotide sequences of the five exon 1 variants and three alternatively spliced exons of the mouse mct2 gene (GenBank accession numbers: DQ140161-140168).

Exon 1-specific RT-PCR and Southern blotting.

The first-strand cDNA was synthesized with total cellular RNA (5 μg) derived from various mouse organs using random primers and SuperScript III reverse transcriptase. A 532-bp fragment of mouse GAPDH was amplified to verify the quality of all cDNA samples (sense: 5′-AGCCTCGTCCCGTAGACAA-3′; antisense: 5′-CCTTCCACAATGCCAAAGTT-3′). The overall expression pattern of mct2 in these organs was assessed using a primer pair subtending the common region of the mct2 cDNA (sense in exon 4: 5′-ACACCACCTCCAGTCAGAT-3′; antisense in exon 6: 5′-AACCTTTTGTGCTGTTGAT-3′) (29). Five exon 1-specific sense primers (1a: 5′-CTCCGAGTCCCCAACAAGTG-3′; 1b: 5′-GGGCTGACTCACTCCGTTGT-3′; 1c: 5′-GCACCGGGTGTCAGAGTAC-3′; 1d: 5′-TTCTGACAGGCTCTCCATGA-3′; and 1e: 5′-ATCGCATTAAACATCTTCATTCA-3′) were designed based upon the results obtained from cDNA cloning and were paired with a common antisense primer (5′-GCACCGGGTGTCAGAGTAC-3′) localized in exon 4 for PCR amplification of cDNA samples. PCR amplifications were performed in a total volume of 50 μl for 30 cycles using Platinum Taq polymerase (Invitrogen).

PCR products were size-fractionated by agarose gel electrophoresis and transferred to GeneScreen Plus membranes (DuPont). Southern blots were hybridized with [γ-32P]ATP-labeled oligonucleotides. An oligonucleotide localized in exon 6 (5′-CCTTCCAGCCATAGTTGTT-3′) was used to hybridize with PCR products amplified from the common region to detect the overall expression pattern of mct2 in various organs. An oligonucleotide localized in exon 4 but internally to the common antisense primer used in the PCR (5′-AGGCTCTGATGGCATTTCTG-3′) was used to hybridize with PCR products amplified with exon 1-specific primers to detect tissue-specific expression patterns of each exon 1 variants. Radioactive signals were recorded and analyzed using a PhosphorImager (STORM 840 and ImageQuant 1.2 software, Molecular Dynamics).

Real-time RT-PCR.

Relative expression levels of exon 1-specific mct2 transcripts in several important organs were further examined with real-time RT-PCR using the Mx4000 Multiple Quantitative PCR System (Stratagene) and the Brilliant quantitative PCR core reagent (Stratagene). cDNA derived from 100 ng total RNA was used as template. New sets of primers were designed for this purpose: exon 1a, sense 5′-CAACAAGTGGCTGATGCG-3′, antisense 5′-GGTGGTGGGTAGAATATTTAAGC-3′; exon 1b, sense 5′-CTCCGTTGTGATACTTTAATCC-3′, antisense 5′-GTGGTGGGTAGAATATTTAAGC-3′; exon 1d, sense 5′-ACTTTAAGGAGATGCTGGAAG-3′, antisense 5′-TGGGTAGAATATTTAAGCCTAGC-3′; and exon 1e, sense 5′-TCGCATTAAACATCTTCATTCAC-3′, antisense 5′-GATGACACTTTAACAGCTTTGG-3′. Exon 1c was not included in this assessment because it was technically impossible to design a primer pair that would detect all variants of 1c-specific transcripts. A common TaqMan probe (Integrated DNA Technologies) was used for all PCR reactions. This probe was labeled at the 5′-end with a fluorescent reporter (6-carboxy-fluorescin) and at the 3′-end with a quencher dye (black hole quencher 1) and incorporated locked nucleic acids (LNAs) to increase the binding specificity: 5′-ctgGagGctGctCtacc-3′ (capitalized letters indicate LNAs). The PCR was performed for 40 cycles (30 s at 95°C and 30 s at 60°C) following denaturing for 10 min at 95°C. PCR products were continuously measured by the sequence detector and normalized by the endogenous reference β-actin (sense 5′-ATCTACGAGGGCTATGCTCTCC-3′, antisense 5′-GCTGTGGTGGTGAAGCTGTAG-3′, and TaqMan probe 5′-CCTGCGTCTGGACCTGGCTGGC-3′). Three replicates for each RNA sample were performed. Results are expressed as relative levels based on calibers derived from series dilutions.

Western blotting analysis.

Frozen tissue samples were pulverized and homogenized in sample buffer containing 50 mM Tris·HCl (pH 7.5), 5 mM EDTA, 10 mM EGTA, 10 mM benzamidine, 50 μg/ml phenylmethylsulfonyl fluoride, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 10 μg/ml pepstatin A, and 10% glycerol (vol/vol) (all chemicals from Sigma-Aldrich). The homogenate was then incubated on a rocking platform in the presence of 20 mM 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate [(CHAPS), Sigma-Aldrich] at 4°C for 4 h. Following centrifugation at 14,000 g for 30 min, the supernatant was collected as total cellular protein. Protein concentration was determined by a modified Lowry method (DC Protein Assay Kit, Bio-Rad). Protein samples (80 μg/lane) were then separated on 10% polyacrylamide/SDS gels and transferred by electroblotting onto nitrocellulose membrane, which was dyed with reversible Ponceau staining for verification of equal loading and transfer efficiency. After being blocked in 5% nonfat milk (Bio-Rad), membranes were incubated with an anti-MCT2 polyclonal antibody (Chemicon), followed by incubation with a horseradish peroxidase-conjugated secondary antibody (Cell Signaling Technology). Signal detection was facilitated with enhanced chemiluminescence (ECL kit, GE Lifesciences).

Characterization of genomic organization.

All newly identified cDNA sequences and the known 5′-cDNA sequence of the mouse mct2 were compared with the database of the Mouse Genome Project to confirm their authenticity and to determine their exonic-intronic boundaries, their positions in the mouse genome, and their relative positions to one another. Sequences in the 5′-flanking regions upstream of the exon 1 variants were then extracted from the genomic contig (GenBank accession number: NT_039502). The composition of these regions, including the presence of repetitive elements and potential binding sites for transcription factors, were analyzed using Genetics Computer Group sequence analysis software package and Eukaryotic Transcription Factor Database release 7.4.

Promoter-reporter assay.

DNA fragments containing ∼3 kb of the 5′-flanking region and a part of the exon 1 sequence were generated for exons 1a (−2706/+506, relative to the most 5′-transcription start site, similarly hereinafter), 1b (−3183/+112), and 1c (−3059/+181), respectively, by PCR using Pfx polymerase and mouse genomic DNA as template (1a, sense: 5′-GGCTCTCCTGCGTTTGTCTTCT-3′; antisense: 5′-ATCTGCTCTCCCTTCCGTCTCT-3′; 1b, sense: 5′-ATCCAGCCTCCGATTGTTCTAC-3′; antisense: 5′-GAGACATGGGACACAGGTTCTTC-3′; and 1c, sense: 5′-CCTTTATGCTGGAGGGAGAATG-3′; antisense: 5′-AGCACCGCCTTTACCTGACTCT-3′). The 1a and 1b fragments were directly inserted, whereas the 1c fragment was digested with FspI and the 3240-bp fragment was then inserted, into the BglII site, which had been made blunt-ended, in the poly-linker of the promoterless luciferase reporter vector pGL2/Basic (Promega) such that the expression of the firefly luciferase reporter was under the control of the inserted mouse genomic DNA fragments. Transient expression of these promoter-reporter constructs was assessed in LA-4 (mouse lung epithelium, ATCC) and Neuro-2a (mouse neuroblastoma, ATCC) cells. Cells were grown in six-well plates to 40–50% confluence and were transfected with 2 μg of the DNA constructs and 4 μl of GenePORTER 2 reagent (Gene Therapy Systems) per well. The pGL2/Control vector (Promega) whose reporter was under the control of the robust heterologous SV40 promoter was used as a positive control and the promoter-less pGL2/Basic vector as a negative control. A Renilla luciferase expression vector (phRL-TK, Promega) was cotransfected to control for transfection efficiency across wells. Forty-eight hours after transfection, cells were lysed and assayed for luciferase activity using the Dual-Luciferase Reporter Assay System (Promega) and a microplate reader (Victor3V, PerkinElmer). A total of six batches of transfection were performed, with each being conducted in duplicated wells.


Cloning of multiple exon 1 variants and their genomic organization.

A total of 68 5′-RACE clones derived from the four organs, including the brain, the skeletal muscle, the lung, and the testis, were analyzed. The 5′-ends of these clones were believed to represent true transcription start sites since the improved 5′-RACE protocol targeted only capped mRNA transcripts. A comparison with the overall structural organization of the mouse mct2 gene demonstrated that all clones contained a common 3′-sequence that was identical to the 5′-end of exon 4, but divergent 5′-sequences that occurred upstream of the first nucleotide of exon 2 (Fig. 1A). Mapping of these 5′-cDNA sequences with the Mouse Genome Database indicated that these divergent sequences were led by four unique exons designated as exons 1a, 1b, 1c, and 1d, respectively, based on the order in which they were identified (Fig. 1). Of note, the 1a sequence was similar to the 5′-end of the published mct2 cDNA sequence (29), but had a longer 5′-end. While exons 1a, 1b, and 1d were spliced directly to the common exon 2, exon 1c was spliced to exon 2 via complex combinations of three alternatively spliced internal exons, termed AS1, AS2, and AS3 based on their 5′ to 3′ order in the mouse genome (Figs. 1; 2, A and B). A search of mouse expressed sequence tags (ESTs) identified an additional exonic sequence derived from the mct2 gene (accession number AI315949). This unique sequence was originally cloned from the kidney and comprised, in the 5′ to 3′ order, a 5′-leader exon, AS2, AS3, exon 2, and exon 4 (Figs. 1; 2, A and B). The 5′-leader exon of this EST was deemed to be an exon 1 (1e) rather than an internal exon although it was localized between AS1 and AS2 (Fig. 2, A and B) because it was not alternatively spliced in any transcripts initiated from exon 1c and it was not preceded by a splicing acceptor sequence. These novel 5′-exonic sequences, including five leading exons and three internal exons, were distributed over a genomic region greater than 80 kb in mouse chromosome 10 (Fig. 2A). All exon-intron boundaries conformed to the GT/AG rule (36). Interestingly, as the EST (AI315949), the vast majority of the 5′-RACE clones lacked exon 3. Alternative splicing of this internal exon further increased the complexity of splicing patterns utilized by this gene (Fig. 2C). Since stop codons were evident in all three translation reading frames upstream of the initiating AUG in exon 4, all newly identified sequences represented mct2 mRNA diversity within the 5′-untranslated region (5′-UTR) and did not affect the structure of the encoded protein.

Fig. 2.

Structural organization and alternative splicing patterns of mouse mct2 5′ exons. A: exon/intron location on genomic DNA. Exonic sequences are depicted as vertical boxes. Five exon 1 variants and alternatively spliced exons span a genomic region >80 kb upstream of exon 2. B: splicing patterns of mouse mct2 5′ exons. Exons 1a, 1b, and 1d are spliced to the common exon 2; whereas exons 1c and 1e are spliced to exon 2 with various combinations of 3 alternatively spliced exons, AS1, AS2, and AS3. Those shown in the figure were the ones physically cloned using 5′-RACE (1e represents EST AI315949). Bent arrows depict transcription start sites. C: alternative splicing patterns involving the deletion/inclusion of exon 3 that is positioned between the common exons 2 and 4. Alternative splicing of this exon greatly increases the number of mct2 mRNA species.

Tissue-specific expression of alternative exon 1 variants.

An exon 1-specific RT-PCR approach was used to determine the expression patterns of each mouse mct2 exon 1 in normal tissues. Consistent with the trend noticed in 5′-RACE, the alternative exon 1 variants were expressed in a tissue-specific fashion when compared with PCR amplifications subtending the common region (exon 4–6, within the open reading frame) (Fig. 3). Notably, exons 1b and 1d were expressed almost exclusively in the testis, although very low levels of 1b were detected in a few other tissues. In contrast, exons 1a, 1c, and 1e were expressed in multiple tissues. In particular, exon 1c was ubiquitously expressed in all tissues with relatively uniform intensity, whereas exons 1a and 1e were also expressed in multiple tissues, but the expression intensity varied greatly among different tissues. For example, 1a was highly expressed in the lung, the liver, the kidney, and the bladder but was only marginally expressed in brain tissues, and 1e was mainly expressed in the kidney and the heart.

Fig. 3.

Tissue-specific expression of mct2 exon 1 variants and their splicing patterns. A: RT-PCR amplification was performed using varied exon 1-specific sense primers and a common antisense primer located in exon 4. An oligonucleotide probe located in exon 4 was used for Southern blot analysis that hybridized to PCR products both with (larger bands) and without (smaller bands) exon 3. B: the same blots hybridized with an oligonucleotide probe located in exon 3 that detected only PCR products with exon 3. C: PCR amplification of the coding region between exons 4 and 6 was used to assess the overall mct2 expression. Blots shown are representative results from 4 different experiments. Skeletal M., skeletal muscle; ORF, open reading frame.

In line with the finding made in 5′-RACE, exon 3 was found to be excluded from most mct2 transcripts under normal conditions, especially those initiated from exons 1a, 1b, and 1d (Fig. 3A). The identities of the high- and low-molecular-weight PCR amplicons were examined by cloning and sequencing and were confirmed to be those with or without exon 3, respectively. Additional confirmation was obtained by using an oligonucleotide probe located in exon 3 to reprobe the Southern blots (Fig. 3B). Meanwhile, mct2 mRNA transcripts initiated from exons 1c and 1e were accentuated by intricate alternative splicing of AS1, AS2, AS3, and exon 3 with a tissue-specific flavor (Fig. 3). For example, >10 different PCR products could be recognized from transcripts initiated from 1c, although only five such variants were physically cloned in 5′-RACE as shown in Fig. 2B. A similar pattern was observed with 1e-initiated transcripts. Given the extraordinary complexity of these 1c- and 1e-initiated transcripts, no additional attempt was made to clarify the molecular nature of these PCR products. Interestingly, sizes and relative densities of these PCR products differed among different tissues, suggesting the existence of tissue-specific and regulated alternative splicing mechanisms.

The relative expression levels of exon 1a-, 1b-, 1d-, and 1e-specific transcripts in several important organs, including the brain, the lung, the liver, the kidney, the heart, and the testis, were further assessed with real-time RT-PCR for better quantification (Fig. 4). While confirming the general trends observed in experiments using RT-PCR plus Southern blotting regarding the tissue-specific usage of alternative exon 1s and relative expression levels of each exon 1 variant in various tissues, real-time RT-PCR revealed additional findings. For example, it revealed that the expression level of 1a in the kidney was substantially higher than, rather than being similar to, those in the lung and the liver (Fig. 4A). It also detected marginal expression of 1d in both the brain and the lung (Fig. 4C), which was not seen with RT-PCR plus Southern blotting (Fig. 3A). Conversely, marginal expression of 1b in the brain and the kidney and 1e in the brain and the lung as seen with RT-PCR plus Southern blotting (Fig. 3A) was not detected with real-time RT-PCR (Fig. 4, C and D). The overall expression patterns of alternative exon 1 variants in various tissues assessed using both methods are summarized in Table 1.

Fig. 4.

Real-time RT-PCR assessment of relative expression levels of mct2 exon 1 variants in various tissues. A: exon 1a is expressed in multiple tissues at different levels. It is highly expressed in the kidney. B: exon 1b is exclusively expressed in the testis. C: exon 1d is almost exclusively expressed in the testis with marginal expression in the brain and the lung. D: exon 1e is selectively expressed in the kidney, the heart, and the testis. Data are means ± SE; n = 4 for each group. Levels <1% of the highest level in each PCR reaction were deemed as undetected.

View this table:
Table 1.

Expression patterns of alternative exon 1s in various tissues

Interestingly, mct2 mRNA expression (Fig. 3C) was not always correlated quantitatively to its protein expression (Fig. 5) in all tissues. For instance, although mct2 mRNA was expressed in the lung at a level similar to those in the brain and in the testis, its protein was expressed in the former at a much lower level. On the other hand, while mct2 mRNA was barely detectable in the skeletal muscle, its protein was expressed in this tissue at a level similar to those in the liver and the heart.

Fig. 5.

Western blotting analysis of MCT2 protein expression in various tissues. Shown is a representative blot from 4 independent experiments. The size and the general distribution of the indicated band are consistent with those of MCT2 reported in earlier studies using noncommercial MCT2 antibodies (18, 25). The ≈42-kDa MCT2 protein is expressed in multiple tissues at different levels.

Cell type-specific usage of alternative promoters.

Transient transfection of promoter-reporter constructs containing ∼3 kb of 1a, 1b, or 1c 5′-flanking genomic regions demonstrated that both the 1a and the 1c regions contained functional promoters (Fig. 6). Furthermore, transcriptional activities of these two promoters in this in vitro assay generally displayed cell type-specific patterns that were consistent with their in vivo tissue-specific usage as reflected by the expression patterns of their respective exon 1 variants (Fig. 3). For example, the 1a promoter was highly active in LA-4 cells (mouse lung epithelium) as it was in the lung, while it was only marginally active in Neuro-2A cells (mouse neuroblastoma) as it was in the brain (Fig. 6). Interestingly, the 1c promoter was highly active in Neuro-2A cells (Fig. 6B) as it was in the brain, but its activity in LA-4 cells compared with that in Neuro-2A cells (Fig. 6A) seemed to be lower than its activity in the lung compared with that in the brain. This minor discrepancy in 1c promoter activities in vivo vs. in vitro may reflect the possibility that the 3-kb genomic region upstream of exon 1c lacks certain cis-elements that are important for maximum promoter activity in LA-4 cells. Alternatively, it may reflect the difference between the activity of a promoter assessed in a tissue and that assessed in a specific cell type. Although significant 1b promoter activity was not detected in these two cell types as expected based upon the unique testis-specific expression pattern of this promoter in vivo, preliminary studies suggested that a 5′-deletion 1b construct (−1254/+112) was moderately active in LA-4 cells (data not shown). Thus, the ∼3-kb genomic regions upstream of exons 1a, 1b, and 1c, respectively, represent functional, alternative promoters of the mouse mct2 gene, which are likely involved in the tissue-/cell type-specific expression of the gene.

Fig. 6.

Cell type-specific promoter activities of selected mct2 promoters assessed with promoter-reporter assay. A: reporter activities in LA-4 cells (mouse lung epithelium) transiently transfected with promoter-reporter constructs. The 1a promoter is the most active in these cells as it is in the lung in vivo. B: reporter activities in Neuro-2A cells (mouse neuroblastoma) where the 1c promoter is the most active as it is in brain tissues in vivo. The promoterless pGL2/Basic vector was used as a negative control. Data are means ± SE; n = 6 for each group; *P < 0.001 vs. pGL2/Basic; #P < 0.01 vs. pGL2/Basic.

Sequence inspection of the proximal regions (−1200 bp) of the three promoters revealed that only 1b contained a TATA-like element (CATAAAWG) in the core promoter region near the transcription start site (−23) (Fig. 7). However, numerous potential binding sites for ubiquitous transcription factors were evident in all three promoters (Fig. 7). For example, the 1a promoter comprised two CAAT boxes (−180 and −789), a strong Sp1 site (−382), an NF-Y site (−618), and two AP-1 sites (−892 and −1003). In the 1b promoter, a CAAT box (−157) and an Sp1 site (−373) were closely localized with the TATA-like element. Interestingly, this region contained three copies of the steroidogenic factor 1 (SF-1) binding site (252, −364, and −1132) that was thought to be important to testis-specific gene expression (4, 32). And finally, the 1c promoter possessed two strong Sp1 sites (33 and 151), two insulin response elements (IRE, 35 and 244), one of which overlapped with the proximal Sp1 site, an upstream CAAT box (−832), a GATA site (−271), and three AP-1 sites (188, −332, and −564).

Fig. 7.

Potential binding sites for selected transcription factors in the proximal regions (−1200 to +200 bp, relative to transcription start sites of the 3 exons, respectively) of alternative promoters 1a, 1b, and 1c of the mouse mct2 gene. Exonic sequences are depicted as filled horizontal boxes. A tetranucleotide repeat (TGGA)6 is shown in the 1c region as a shaded horizontal box. Positions, but not sizes, for all elements are drawn to scale.


The current study shows that the mouse mct2 gene is a highly complex locus whose expression in diverse tissues is likely initiated by multiple alternative promoters. This notion is supported by the following lines of evidence: 1) transcription of the mct2 gene starts from five distinct exon 1 variants that are distributed over a large genomic region; 2) the usage of these exon 1 variants displays a tissue-specific fashion; 3) the 5′-genomic regions flanking these exon 1 variants contain arrays of cis-acting regulatory elements constituting common promoters; 4) three of the 5′-flanking regions have been tested in vitro, each of which has been shown to contain a functional promoter; and 5) the in vitro, cell type-specific activities of the promoters are consistent with the tissue-specific expression pattern of the gene in vivo. Interestingly, the human and the rat homologs appear to utilize the same transcriptional strategy. For example, both genes are transcribed from multiple first exons, as deduced from cDNA sequences deposited in the GenBank database (accession numbers AF049608 and AF058056 for hMCT2; X97445, U62316, and BC081701 for rMCT2). A recent study further demonstrated the presence of two distinct promoters in the human MCT2 gene (56). The fact that this intricate transcriptional strategy is conserved across several species suggests that it may serve important biological functions.

Once regarded as “unconventional,” alternative promoter usage is now recognized as a common mechanism in the transcriptional regulation of mammalian genes (3, 30, 55, 64). A recent analysis of 21,000 full-length mouse cDNA clones has revealed that 9% of the mouse genes contain alternative first exons (61), whereas a subsequent analysis of 67,000 human transcripts has shown that ∼18% of human genes may use alternative promoter (30). Alternative promoters create genomic diversity and flexibility and, in turn, provide a framework for a given gene to effectively respond to diverse physiological and pathological signals, including cell type-, developmental stage-, and environmental stress-specific signals. Many important genes whose diverse and complex function under various conditions requires accurate and exquisite transcriptional regulation are equipped with multiple promoters. For instance, the human dystrophin gene uses five independent promoters to direct its transcription in a cell-specific and developmentally controlled fashion (1), while at least 10 promoters are responsible for tissue-specific expression of the human neuronal NO synthase in the brain, the skeletal muscle, the testis, and other organs (59, 60). Other examples include the genes for c-Myc (6, 17), platelet-derived growth factor β-receptor (57), aromatase cytochrome P450 (24), and glucocorticoid receptor (34). In the case of the mouse mct2 gene, it is apparent that the use of alternative promoters is at least important for tissue-specific expression of the gene, as shown in the current study. While it remains to be determined whether such mechanism is also involved in the expression of this gene in response to development (5, 9, 49), differentiation (31), metabolic stress (45), trauma (47), hormones (9, 20), and hypoxia (63), the interesting composition of the 1b promoter appears to suggest that this particular region might be involved in the induction of mct2 by the latter two types of signals. Indeed, this genomic region not only contains multiple copies of the SF-1 binding site that is thought to be important to testis-specific gene expression (4, 32) but also features clusters of potential binding sites for transcription factors implicated in hypoxia-induced gene expression (Fig. 7), including two HIF-1 sites (53, 54), four STAT sites (26), three NF-κB sites (50, 51), and three C/EBP/p300 sites (2, 16). In this regard, a recent study demonstrated that both the mouse and the human MCT4 genes were upregulated by hypoxia through HIF-1α-dependent transcriptional activation (56). It is possible that HIF-1α plays a central role in regulating the expression of lactate transporters and, in turn, lactate metabolism under hypoxic conditions. However, it was noted that the human MCT1 promoter and the two human MCT2 promoters tested in the same study were not responsive to hypoxia (56).

The fact that all five mct2 promoters are active in the testis appears to contradict the concept of “alternative promoters.” It is possible that in this case multiple promoters serve as a safeguard mechanism that secures proper expression of a vital gene. Recent studies have implicated MCT2 in spermatogenesis and other important testicular functions (9, 10, 20, 25). A sustained lactate transporter activity might be vital to these functions, such that additional security measures may have developed through evolution to preserve the expression of this transporter. Indeed, two of the five mct2 promoters are shown to be designated specifically for this organ. It would be a fascinating topic as to how these promoters are coordinated to achieve the level of expression of mct2 proper to the vital organ of the testis. Alternatively, the five promoters might direct mct2 expression in different types of cells within the testis. The biological significance of these promoters within this important organ will require additional investigation.

In addition to alternative promoter usage, alternative splicing of pre-mRNA transcripts, especially those derived from exons 1c and 1e, is a major contributor to mct2 mRNA diversity (Fig. 3). The alternative splicing patterns of these transcripts are so complex that it is difficult, if not impossible, to determine the molecular identity for each one of them. However, alternative inclusion of exon 1a, 1b, 1d, or 1e in 1c transcripts has been ruled out since Southern blotting analysis of the 1c-specific PCR products using probes located in these exons did not detect any bands (data not shown). It is conceivable that the various 1c transcripts are products of alternative splicing of internal exons AS1, AS2, AS3, and possibly exon 3 and that the 1e transcripts are derived from alternative splicing of exons AS2, AS3, and possibly exon 3. Of note, while exon 3 was predominantly excluded among mct2 transcripts from C57BL/6 mice as shown in the current study, it was included in a cDNA cloned from 129 mice (accession number AF058054) (29). It is unclear whether the latter reflects a true strain-based preference in the usage of this exon or just a random occurrence in a single cDNA sequence.

Although all mct2 mRNA diversities, resulting from both alternative promoter usage and alternative splicing, are restricted to the 5′-UTR and hence are not expected to affect MCT2 protein structure, their potential significance cannot be ignored. The 5′-UTR of eukaryotic mRNA has long been known to play crucial roles in posttranscriptional regulation of gene expression through the modulation of RNA transport (8), translational efficiency (37, 60), and RNA stability (13). Several recent studies have compared the effect of 5′-UTR variants from a single gene on translation and have revealed significant differences in translational efficiency among naturally occurring mRNA variants from the neuronal NO synthase (38, 60), the endothelial argininosuccinate synthase (41), and the RARbeta2 (42). Considering the substantial differences among the mct2 mRNA variants in their primary composition and possibly in their secondary structure, these transcripts are likely translated at different efficiencies. While it needs to be experimentally confirmed, data presented in the current study showing discrepancies between mRNA and protein levels in some organs are supportive of this notion. Furthermore, recent studies demonstrated that MCT2 genes indeed were regulated at the translational level (14, 25, 43). It is thus reasonable to assume at this time that the remarkable mRNA diversity of the mouse mct2 gene is more than a mere consequence of the need to initiate transcription from different genomic locations and in fact may represent an additional regulatory mechanism that insures appropriate expression of this important gene under various conditions in diverse tissues.

In summary, we demonstrate that transcription of the mct2 gene is initiated from five distinct variants of exon 1, likely under the control of five alternative promoters, and that a large number of mature mRNA are produced from these pre-mRNA through alternative splicing of multiple internal exons in the 5′-UTR. Both alternative promoter usage and alternative splicing may contribute to the tissue-specific expression pattern of mct2 and may also be involved in development-specific and environmental stress-induced expression of this important gene.


This study was supported in part by National Heart, Lung, and Blood Institute Grants R01HL-074369 and P50HL-060296, American Heart Association Grant 0455418B, Defense Advanced Research Projects Agency Grant N66001-02-C-8056, and the Center for Genetics and Molecular Medicine, University of Louisville.


Nucleotide sequences of the 5 exon 1 variants and 3 alternatively spliced exons of the mouse mct2 gene identified in this study have been deposited in GenBank (accession numbers: DQ140161-140168).


  • Address for reprint requests and other correspondence: Y. Wang, Dept. of Pediatrics, Univ. of Louisville, 570 S. Preston St., Ste. 211, Louisville, KY 40202 (e-mail:{at}

    Article published online before print. See web site for date of publication (


View Abstract