| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Embryo; |
School of Biomedical Sciences, University of Queensland, Brisbane 4072, Queensland, Australia
ABSTRACT
Concurrent with compaction, preimplantation mouse embryos switch from the high pyruvate consumption that prevailed during cleavage stages to glucose consumption against a constant background of pyruvate uptake. However, zygotes exposed to and subsequently deprived of glucose can form blastocysts by increasing pyruvate uptake. This metabolic switch requires cleavage-stage exposure to glucose and is one aspect of metabolic differentiation that normally occurs in vivo. Monocarboxylates, such as pyruvate and lactate, are transported across membranes via the SLC16 family of H+-monocarboxylate cotransporter (MCT) proteins. Thus, the increase in pyruvate uptake in embryos developing without glucose must involve changes in activity and localization of MCT. In mouse embryos, continued expression of Slc16a1 (MCT1) requires glucose supply. Messenger RNA for Slc17a7 (MCT2) and Slc16a3 (MCT4) has been detected in mouse preimplantation embryos; however, protein function, localization, and regulation of expression at the basis of these net pyruvate uptake changes remain unclear. The expression and localization of SLC16A7 and SLC16A3 have therefore been examined to clarify their respective roles in embryos derived from the reproductive tract and cultured under varied conditions. SLC16A3 appears localized to the plasma membrane until the morula stage and also maintains a nuclear distribution throughout preimplantation development. However, continued Slc16a3 mRNA expression is dependent on prior exposure to glucose. SLC16A7 localizes to apical cortical regions with punctate, vesicular expression throughout blastomeres, partially colocalizing in peroxisomes with peroxisomal catalase (CAT). In contrast to SLC16A3 and SLC16A1, SLC16A7 and CAT demonstrate upregulation in the absence of glucose. These striking differences between the two isoforms in expression localization and regulation suggest unique roles for each in monocarboxylate transport and pH regulation during preimplantation development, and implicate peroxisomal SLC16A7 as an important redox regulator in the absence of glucose.
conceptus, early development, environment, monocarboxylate transporters, nutrient
Early research provided fundamental tenets about nutrient utilization during preimplantation development, including that the monocarboxylate pyruvate, but not lactate, can support the first cleavage division of the mouse embryo [1] and that glucose cannot support development of the preimplantation mouse embryo until around the early eight-cell stage [2]. Measurement of the uptake and metabolism of nutrients revealed that the switch from pyruvate to glucose occurs around the time of compaction, from which point glucose becomes the predominant nutrient source for development to a blastocyst [3].
Subsequently, the role of glucose in preimplantation development has been more deeply explored. High glucose concentrations (1 mM) inhibited development of cleavage-stage embryos [4], yet, without glucose development, proceeded only to the morula stage [5, 6]. The timing of this apparent glucose requirement was narrowed to 56–92 h post hCG [5]. Metabolic studies revealed that two-cell mouse embryos cultured without glucose may still cavitate to form blastocysts by increasing their consumption of pyruvate [7]. Complimenting this was evidence that early exposure to glucose is required for morulae to express SLC2A3 (GLUT-3), a high-affinity facilitated glucose transporter determined to be essential for blastocyst formation [8–10]. This suggested that, while not required as a nutrient, glucose was indeed necessary for the changes in gene expression required for normal preimplantation development, perhaps as a signal of maternal nutritional status, as is the case for somatic tissues such as liver (for review see [11, 12]).
The molecular basis of this glucose-triggered metabolic differentiation facilitating increased pyruvate uptake is less clear, because pyruvate and lactate enter cells via diffusion facilitated by a family of monocarboxylate-proton cotransporters (SLC16, MCT). There are 14 MCT isoforms, each with unique substrate affinities and tissue distribution for close regulation of cellular monocarboxylate and pH homeostasis [13–15]. An MCT system that maintained the intracellular pH above electrochemical equilibrium, and the functional characteristics of which resembled those of Slc16A1, was detected in two-cell embryos [16]. Subsequently, mRNA for Slc16a1, Slc16a7, and Slc16a3 was detected throughout preimplantation development [17–19], but two different assays failed to demonstrate any gross kinetic changes in monocarboxylate transport activity in embryos cultured with or without glucose [17, 19]. However, neither technique could differentiate changes in activity location or function of individual MCT isoforms, which would be likely to change in such a metabolic switch. In contrast, immunocytochemical analysis demonstrated that early exposure to glucose was necessary for expression of Slc16a1 mRNA and protein beyond the morula stage [19].
To further understand the molecular basis of the increased pyruvate uptake induced by glucose deprivation, the expression and localization during mouse preimplantation embryo development of the other two key MCTs. SLC16A7 and SLC16A3 were at the focus of this research, including any individual changes in location or protein concentration that may occur in response to the presence or absence of glucose. Mouse basigin (BSG) is a glycoprotein that participates in the cell surface orientation of SLC16A1 and SLC16A3 [20–24], and the mRNA of which is expressed throughout the pre- and peri-implantation period [18, 25, 26]. Therefore, its presence and its relationship to SLC16A3 in embryos were also investigated. SLC16A7 has been previously reported in peroxisomes [27] that have been described in oocytes [28]. Consequently, a peroxisomal location of SLC16A7 was also explored with catalase (CAT; a peroxisomal marker) colocalization.
The Animal Ethics and Experimentation Committees of the University of Queensland approved all experiments on mice. These committees are approved by the National Health and Medical Research Council of Australia.
The handling medium was M2 medium [29] modified to contain 0.33 mM sodium pyruvate and 4 g/L of BSA and lacking glucose. The culture medium used was KSOM [30]. The homogenization buffer consisted of 50 mM Tris-HCl, 150 mM NaCl, 10 mM NaF, 1 mM Na3VO4, 1% Triton X-100, one cOmplete protease inhibitor tablet per 50 ml buffer (Roche Diagnostics Australia P/L, NSW, Australia). Laemmli sample buffer [31] was used, and consisted of 10% SDS containing 1% glycerol, 0.125 mg/ml bromophenol blue, 0.125 mg/ml xylene cyanol, 100 mM dithiothreitol, and 125 mM Tris-HCl, pH 6.8. The Towbin transfer buffer [32] employed consisted of 25 mM Tris, 192 mM glycine, 20% methanol, pH 8.3.
Affinity-purified anti-SLC16A7 immunoglobulin (Ig) G was raised in goats against a synthetic peptide mapping to the carboxy terminus of mouse SLC16A7 (14926; Santa Cruz Biotechnology, Santa Cruz, CA). Anti-CAT antibody was prepared in rabbits against whole purified bovine liver CAT (Rockland Immunochemicals Inc., Gilbertsville, PA). Affinity-purified anti-SLC16A3 antiserum (courtesy N. Philp, Thomas Jefferson University, Philadelphia, PA) was raised in New Zealand white rabbits against mouse SLC16A3 [22]. Affinity-purified rat monoclonal anti-mouse BSG was raised against mouse clone RL73 (14–1471; eBioscience, San Diego, CA). Secondary antibodies used for immunofluorescence or Western blotting included: fluorescein isothiocyanate-conjugated rabbit anti-goat IgG (Merck Pty [Calbiochem], VIC, Australia); Texas Red-conjugated goat anti-rabbit IgG (Merck [Calbiochem]); Texas-Red conjugated goat anti-rat IgG (Sapphire Bioscience, NSW, Australia); horseradish peroxidase (HRP)-labeled donkey anti-rabbit IgG (Amersham Biosciences, NSW, Australia) or rabbit anti-goat IgG (Progen Biosciences, QLD, Australia).
Female 6 to 8-wk-old Quackenbush mice were superovulated by injection with 10 IU eCG (Folligon; Intervet P/L, NSW, Australia) and 10 IU hCG (Chorulon; Intervet) 48 h apart. Embryos from successfully mated females were collected at 24, 48, 72, and 96 h post-hCG into glucose-free M2 medium. Where necessary, cumulus cells were removed with 0.5 mg/ml hyaluronidase in glucose-free M2 followed by washing in fresh glucose-free M2.
For culture experiments, zygotes were collected into glucose-free M2 at 18 h post-hCG and cultured in KSOM microdroplets at a density of 1 embryo/µl under mineral oil, in a humidified atmosphere of 5% O2, 5% CO2, and 90% N2 at 37°C with a MINC incubator (Cook, QLD, Australia). KSOM contained 0.2 mM glucose or was modified to be glucose-free (KSOM-G). Cultured embryos were sampled at 48, 72, 90, and 96 h post-hCG.
Total RNA was extracted from embryos collected at 24, 48, 72, and 96 h post-hCG with phenol/chloroform and the RNeasy Mini Kit (Qiagen, VIC, Australia). RNA was also extracted from mouse liver and retina and 1 µg used for reverse transcription (RT). RNA was reverse transcribed and amplified by PCR [19] with specific primers for Slc16a7 and Slc16a3. The primer pairs derived from murine Slc16a7 and Slc16a3 cDNA sequences were as follows: Slc16a7 5' primer, 5'-ctgatgcggcttctctctct-3', and 3' primer, 5'-cgatctgactggaggtggtg-3'; Slc16a3 5' primer, 5'-ccatagttatcagccaccta-3', and 3' primer, 5'-ccttctacctgagaaggcag-3'. The 438-bp and 307-bp PCR products for Slc16a7 and Slc16a3, respectively, were resolved on 1% agarose gels containing 0.5 mg/L ethidium bromide. Genomic contamination of cDNA samples was tested by parallel PCR for mouse Actb (ACTB) with 5' primer 5'-cgtgggccgccctaggcacca-3', and 3' primer 5'-ttggccttagggttcagggggg-3'. This primer pair spans the first 87-bp intron of the mouse Actb gene and generates a predicted 243-bp fragment for the cDNA, and a 330-bp fragment if contaminating genomic DNA is present [33].
PCR products were extracted from gels (Gel Extraction Kit; Qiagen), ligated into a pGem-T Easy Vector (Promega, NSW, Australia), and amplified in transformed DH5
Escherichia coli cells. Purified plasmid clones (QIAprep Spin Miniprep Kit; Qiagen) were digested with EcoR1 (Roche Diagnostics GmbH, Pennzberg, Germany) and sequenced at the Australian Genome Research Facility (Brisbane, QLD, Australia) to confirm identity.
Embryos fixed in 2% paraformaldehyde in PBS (pH 7.4) underwent immunohistochemical analysis as previously described [8]. Briefly, embryos were incubated overnight in primary antibody (2–10 µg/ml) at 4°C and then exposed to secondary antisera (1:100) for 1 h at room temperature. Following extensive washing, samples were mounted in glycerol on concave glass slides for examination with either a Bio-Rad (Hercules, CA) MRC-2000 or MRC-1024 confocal laser scanning microscope mounted on a Zeiss (Oberkochen, Germany) Axioskop with a Zeiss Plan-APOCHROMAT 60x oil immersion objective. All experiments were completed at least in triplicate, with 10 or more samples of each embryo stage observed per experiment.
Mouse tissues were homogenized in ice-cold homogenization buffer, centrifuged at 14 000 x g for 15 min at 4°C, and supernatant was removed and placed with an equal volume of Laemmli sample buffer. For embryo samples, 80–100 embryos were placed in 5 µl homogenization buffer with an equal volume of Laemmli sample buffer. Tissue (10 µg protein/lane, as determined by bicinchoninic acid assay) and embryo samples were separated by polyacrylamide electrophoresis and transferred to 0.45 µm Immobilon-P PVDF membrane (Millipore Australia P/L, NSW, Australia) with Towbin transfer buffer. Membranes were incubated for 1 h at room temperature in blocking solution of 5% skim milk powder (for SLC16A7, CAT, and BSG) or 5% BSA (for SLC16A3) in PBS/0.1% Tween-20. Primary antisera (0.2 µg/ml) were applied overnight at 4°C, and membranes were exposed to secondary antibody conjugated to HRP for 1 h at room temperature. Immunoreactive bands were detected with Supersignal West Pico or West Femto Enhanced Chemiluminescence Detection Kits (Pierce Biotechnology Inc., Rockford, IL). Exposed films were developed in a Kodak M35 X-Omat Processor (Kodak Australasia P/L, Melbourne, Australia).
Pixel intensity of western blots was analyzed with Image-J Software (National Institutes of Health, Bethesda, MD), and a paired Student t-test with Welch's correction for unequal variance was performed on measured, standardized grey levels with Prism (GraphPad Software, Inc., San Diego, CA).
MCT Expression in Fresh and Cultured Embryos
Slc16a3 mRNA is present in unfertilized oocytes and throughout all preimplantation stages in embryos developing in vivo or in vitro (Fig. 1J). There were no apparent differences in SLC16A3 immunolocalization in either freshly collected (Fig. 1, A–D) or cultured embryos (Fig. 1, F–H). Immunoreactivity for SLC16A3 is localized to the plasma membrane in zygotes and cleavage-stage embryos (Fig. 1, A, B, and F); however, this is not apparent in morulae and blastocysts (Fig. 1, C, D, G, and H). In all stages, SLC16A3 appears to have a distinct nuclear distribution. SLC16A3 antibody was confirmed to be monospecific on a Western blot, with an expected electrophoretic mobility of 42 kDa.
|
To determine if the lack of plasma membrane SLC16A3 arose from a defect in BSG location, BSG protein expression was examined. BSG was evident on plasma membranes and in nuclei in all stages (Fig. 2, A–E). While SLC16A3 was not apparent in plasma membranes in morulae and blastocysts, BSG expression persisted in this location. A Western blot with anti-BSG antibody in mouse retina and freshly collected embryos demonstrated multiple bands between 48 and 66 kDa, as expected (Fig. 2F). This is typical of the highly and variably N-glycosylated transmembrane protein.
|
Slc16a7 mRNA was present in unfertilized oocytes, and at all stages of preimplantation development in both freshly collected and cultured embryos (Fig. 3J). SLC16A7 protein was localized on or just beneath the plasma membrane in zygotes and two-cell embryos (Fig. 3, A, B, and F). In morulae and blastocysts, immunoreactivity was most intense in the apical region of outer blastomeres, but SLC16A7 lacked any membrane association in the inner blastomeres of morulae or in the inner cell mass of blastocysts (Fig. 3, C, D, G, and H). A distinct granularity was evident throughout the cytoplasm at all stages. There were no differences in SLC16A7 expression between in vivo- and in vitro-derived embryos. A Western blot for SLC16A7 antibody demonstrated a single band with an expected molecular weight of around 42 kDa (Fig. 3I).
|
The peroxisomal marker, CAT, was examined in embryos to see whether any cytoplasmic SLC16A7 granules costained with CAT. Staining was granular throughout the cytoplasm from zygote to blastocyst (blastocyst not shown in results; Fig. 4, A–C). In zygotes, CAT appeared brightest at the periphery of the cell (Fig. 4A), but this was not apparent in later stages. Colocalization of CAT with SLC16A7 demonstrated identical immunoreactivity for some vesicles of 0.5–2.0 µm diameter (Fig. 4, D–F, arrowheads), suggesting that some SLC16A7 protein may be located in peroxisomal compartments. A Western blot for anti-CAT antibody resulted in one band with an expected electrophoretic mobility of approximately 60 kDa (Fig. 4I), confirming identity and monospecificity.
|
Negative controls for immunofluorescence included normal rabbit serum (Figs. 1E, 4H, and 5H), normal goat serum (Figs. 3E, 4G, and 5D), or affinity-purified rat IgG (Fig. 2E) diluted to the same concentrations as relevant primary antibodies. Separate omission of primary and secondary antisera was also used successfully as negative controls (results not shown). All negative controls demonstrated no apparent specific immunofluorescence.
Effect of Glucose on MCT Expression and Location
Zygotes collected in glucose-free M2 were randomly assigned for culture in KSOM or KSOM lacking glucose. Some zygotes cultured without glucose received a "pulse" or 1 h exposure to 27 mM glucose between 32 and 35 h of culture (56–59 h post hCG). These were then washed several times to remove glucose and returned to culture in KSOM devoid of glucose. Late morulae (90 h post-hCG) from each group were selected and analyzed for SLC16A3 expression. SLC16A3 immunoreactivity did not appear to vary in intensity or cellular location, either in the presence or absence of glucose, nor after receiving a brief glucose pulse (Fig. 6, A–C). However, mRNA for Slc16a3 was not detected in morulae that had developed without glucose (Fig. 6D). A brief pulse of glucose around the four-cell stage appeared to enable Slc16a3 mRNA expression.
|
Development in the absence of glucose caused a dramatic increase in the intensity of SLC16A7 immunofluorescence at the apical surface of outer blastomeres in morulae, but also within the cytoplasm of blastomeres where granules appeared particularly dense compared to morulae cultured with glucose (compare Fig. 5, A and B). CAT granules also increased in intensity and density in morulae cultured without glucose (Fig. 5F) compared with embryos cultured with glucose (Fig. 5E). A brief pulse of glucose at the four-cell embryo prevented this upregulation for both proteins (Fig. 5, C and G). These observations were consistent across four experiments with 20 embryos observed in each treatment, suggesting a real increase in SLC16A7 and CAT expression in the absence of glucose.
|
Confirmation and quantification of SLC16A7 and CAT upregulation in response to glucose was sought with Western immunoblotting. Groups of 100 embryos were cultured as described in the previous experiment, and 80 morulae per group selected for PAGE and Western blotting (see Materials and Methods). Immunoblots for SLC16A7 (Fig. 7A) or CAT (Fig. 7B) were then stripped and reprobed with ACTB (beta-actin) antibody as a loading control. The calculated grey level ratios SLC16A7:ACTB and CAT:ACTB for the two treatment groups were compared for three different experiments, with a paired Student t-test. Results demonstrate a significant increase of 30%–50% in both SLC16A7 and CAT in the absence of glucose (Fig. 7; P < 0.05 for each experiment shown).
|
This first report of protein localization for SLC16A7 and SLC16A3 during preimplantation development supports previous mRNA expression studies [17–19] and extends them to reveal novel compartmentalization of the proteins and regulation of expression by glucose exposure. Both SLC16A7 and SLC16A3 are expressed throughout this period, but display distinct localizations. Where SLC16A3 appears to be regulated by early exposure to glucose in a similar manner to SLC16A1 [19], SLC16A7 exhibits a paradoxical upregulation in the absence of glucose that is associated with increased peroxisomal activity.
SLC16A3 is a lower-affinity monocarboxylate transporter, with Michaelis constant values one order of magnitude higher than SLC16A1 and SLC16A7 [34]. It is highly expressed in white skeletal muscle fibers, astrocytes, white blood cells, and chondrocytes [35–38], and transports lactate with higher affinity than pyruvate [34]. This suggests that SLC16A3 may be important for lactate efflux from glycolytic cells. Its distribution in embryos is somewhat surprising then, with membrane localization disappearing as the embryo compacts and forms a blastocyst, concurrent with the timing of increased glycolysis in the embryo. However, while SLC16A3 favors lactate over pyruvate, its affinity for lactate is 10-times lower than that of SLC16A1, which can also mediate the exchange of one monocarboxylate for another [38]. As SLC16A3 is disappearing from the trophectodermal membrane in developing blastocysts, previous work has shown that SLC16A1 concentrates in the polar trophectoderm at this time [19]. It may be that SLC16A3 is suitable for efflux of a proton with lactate or pyruvate to maintain pH homeostasis in the cleavage-stage embryo, but, as metabolism moves to glucose reliance, SLC16A1 is better suited for transport of these products in either direction across the plasma membrane to regulate pH in a rapidly changing environment. It was concluded that SLC16A1 was responsible for much of the pH regulation during development [16, 17], but this conclusion, relying only on pharmacological and RT-PCR analysis, did not have the benefit of knowledge of MCT location.
The regulatory mechanisms for this relocation of SLC16A3 are less obvious. Two controlling factors were considered: BSG and glucose. BSG is a single transmembrane-spanning Ig-like glycoprotein involved in many physiological processes, including fertilization, implantation, neural function, inflammation, and matrix metalloproteinase induction for tissue remodeling/invasion [18, 26, 39, 40–42]. Its molecular weight, when treated with N-glycanase, is around 32 kDa, but, more commonly, it exists in variably N-glycosylated forms from around 43 to 66 kDa, and sometimes more than one form is evident in one tissue type [43]. Correct targeting and organization of SLC16A1 and SLC16A3 at the plasma membrane involves BSG, and this association occurs in its transmembrane and cytoplasmic domains [18, 22, 44]. In the embryo, Bsg mRNA is present throughout preimplantation development [18, 25]. Protein localization gave clear evidence that BSG exists on the plasma membrane throughout all stages of preimplantation development and, while there was some staining associated with nuclei, there was no evidence to suggest that BSG distribution alters during compaction to drive the changes that occur in SLC16A1 [19] and SLC16A3 localization. Two explanations present themselves: 1) the level of glycosylation can alter BSG function [45]; and 2) BSG may have heterogeneity in its N-terminal cytoplasmic domain presenting different isoforms [43], perhaps via exon 5 splicing variation [46]. The level of glycosylation did not affect the binding of MCT with BSG, so the first explanation is unlikely [44]. The second suggestion is more plausible, since SLC16A1 and SLC16A3 may be binding to BSG variants that were not differentially detected by the antiserum used in this investigation.
While the presence or absence of glucose did not affect SLC16A3 distribution before or after compaction, the expression of Slc16a3 mRNA was indeed affected. In the absence of glucose, Slc16a3 mRNA was present at 24 and 48 h post-hCG, but could not be detected at 90 h post-hCG. A 1-h pulse of glucose around the four-cell stage prevented this downregulation of Slc16a3 mRNA. This effect of glucose mirrors that previously reported for Slc16a1 expression under the same culture conditions [19]. Poor correlation with SLC16A3 expression at 90 h post-hCG may arise from a slower turnover of protein compared with mRNA, as loss of SLC16A1 protein in the absence of glucose was not evident until about 100 h post-hCG [19]. Similarly, there was no correlation between Slc16a3 mRNA and SLC16A3 in rat muscle [47], and, in mouse germ cells, Slc16a7 mRNA was transcribed 6 days before SLC16A7 could be detected, suggesting that posttranscriptional control of MCT protein is not unusual [48]. Nevertheless, this indicates that glucose is a positive regulator of SLC16A3 expression in the same way as it is for SLC16A1. Early availability of oviductal glucose is therefore important for correct expression of SLC16A3 and SLC16A1 and the metabolic differentiation during preimplantation development that enables the large increase in glucose metabolism.
Despite the loss of membrane localization of SLC16A3 after compaction, the brightest intensity of immunolocalized SLC16A3 at this time was nuclear or closely associated with the nucleus. While this may represent only a small amount of protein, because of the volume of the nucleus relative to the cytoplasm and plasma membranes, there is some supporting evidence for such a nuclear localization of MCT in the nuclear membrane of white blood cells [49]. In addition, SLC16A1 staining of aster-like structures near or associated with astrocyte nuclei suggested a role in the maintenance of the pH gradients that exist between cellular compartments [50]. This explanation is consistent with proton shuttle function of MCT, but would require further study to confirm the exact ultrastructural location. This discovery of a nutrient transporter in the nucleus is not isolated. SLC2A1 (GLUT-1) was detected in the nucleus of mouse preimplantation embryos [51], correlating with the demonstration of glycogen and glycogen synthase in nuclear fractions of rat liver and Xenopus oocytes [52]. Indeed, lactate dehydrogenase (LDH) has been located in the nuclei of a variety of tissue types [53–57]. This nuclear LDH has single-stranded DNA-binding properties, and can be controlled by NADH levels and tyrosine phosphorylation [53, 54, 57]. Nuclear LDH may play a role in controlling DNA replication and transcriptional events, and provides a valid reason for the existence of a lactate exporter, such as SLC16A3 in the nucleus. Abnormalities in pH have been associated with preimplantation developmental arrest [58] and pH alterations in subcellular compartments in zygotes (including pronuclei and nucleolus precursor bodies) have been demonstrated to mark transitions through mitosis [59]. It seems plausible that nuclear MCT may regulate nuclear pH; however, this suggestion requires further investigation.
SLC16A7 has the highest affinity for pyruvate and lactate of all the functionally characterized MCT [60, 61]. SLC16A1 and SLC16A7 do not colocalize, suggesting a unique function for each isoform [62–64]. SLC16A7 appears to be inserted into the plasma membrane of preimplantation embryos. This location in cleavage-stage embryos indicates that SLC16A7 may initially be important for nutrient (pyruvate) import. In later stages, SLC16A7 appears to be polarized to the apical surfaces of blastomeres in morulae and the apical trophectoderm of blastocysts, which predominantly utilize glucose via glycolysis. This location of SLC16A7 may facilitate more rapid exchange of monocarboxylates and protons for pH regulation. SLC16A1 on inner-cell mass cells [19] may act to ferry lactate to the trophectoderm for rapid exchange at the embryo/maternal interface via SLC16A7.
The large cytoplasmic pool of SLC16A7 that appears to exist within tiny granules in every blastomere was intriguing. SLC16A7 has previously been reported to exist in peroxisomes [27], where it may act as a lactate shuttle, importing pyruvate to allow peroxisomal LDH to restores NAD+ for continued β-oxidation, and exporting lactate as the resultant by-product. Peroxisomes have been observed in rat gametes [28], so the possibility that some of these SLC16A7-positive granules were peroxisomes was therefore investigated.
CAT, a typical peroxisomal marker [65], was brighter at the periphery of oocytes, as previously reported [28], and stippled throughout the cytoplasm of every blastomere. There is evidence of fatty acid oxidation in preimplantation embryos [66–70] that may occur in these peroxisomes. Furthermore, the expression of peroxisome proliferators and peroxisome-associated enzymes has been demonstrated in embryonic cells [28, 71, 72], so embryos almost certainly contain peroxisomes. In addition, colocalization of CAT and SLC16A7 suggested that a large proportion of these granules, ranging in size from around 0.5 to 2 µm, contain both CAT and SLC16A7 (Fig. 4, arrowheads). This implies that embryos utilize peroxisomal metabolism, perhaps for long and very long chain fatty acids, and that this metabolism may include activity of SLC16A7 as a lactate shuttle for redox regulation within the preimplantation embryo, as described above.
In this study, culture from the zygote stage to 90 h post-hCG in the absence of glucose caused an upregulation in expression of SLC16A7 and CAT by about 50% and 30%, respectively. Embryos deprived of glucose from the zygote stage demonstrate a continual decline in glucose and pyruvate utilization until they eventually degenerate [7, 73]. Moreover, functional differences in MCT transport in the absence of glucose are not apparent [17, 19], although both transport assays lacked the sensitivity to detect changes in individual MCT activity. Poor correlation of the expression data with net utilization of pyruvate supports the concept that SLC16A7 behaves as a lactate shuttle and redox regulator in the face of metabolic stress, as proposed for liver [27]. As the greatest part of the increased CAT in the absence of glucose was found in these peroxisomes, along with much of the elevated SLC16A7, SLC16A7 may function in this stressed state to shuttle lactate and pyruvate across the peroxisomal membrane as substrates for peroxisomal and cytoplasmic LDH. It has previously been reported that early mouse embryos incubated with radiolabeled pyruvate produced some labeled lactate [74]. LDH has been demonstrated in cleavage-stage mouse (as LDHB4) changing to LDHA4 in late blastocysts [75], and its activity found in human embryos [76] and in peroxisomes as LDHA3B and LDHA4 [77], and around 40% of pyruvate consumption in eight-cell human embryos can be accounted for by lactate production [78]. This lactate-pyruvate circuit via SLC16A7 between cytoplasm and peroxisomes would generate a finely balanced NAD+:NADH ratio in both compartments, without changes in net production or consumption of either monocarboxylate. Alternatively, it is possible that SLC16A7 is acting to shuttle cytosolic lactate into the peroxisome for reoxidation to pyruvate, generating cytosolic NAD+ and peroxisomal H2O2. Clearly, these hypotheses require further investigation.
The localization and distribution of SLC16A7 and SLC16A3 under different physiological conditions provide insights into their regulation and function during preimplantation development. SLC16A3 requires early glucose exposure for continued expression in morulae and blastocysts. It is located on the plasma membrane, ideally placed for lactate export during the cleavage stages, but also resides in nuclei, perhaps for pH homeostasis during DNA repair, replication, and transcription. SLC16A7 is present cortically and on the plasma membrane of early embryos and trophectoderm-fated blastomeres, most likely for rapid pyruvate-lactate exchange and pH regulation. Cytoplasmic SLC16A7 also appears to colocalize with CAT in peroxisome-like granules, consistent with previous immunolocalization in rat oocytes [28]. SLC16A7 in this location may serve to shuttle lactate/pyruvate between cytoplasm and peroxisomes to maintain redox status. SLC16A7 and CAT appear to be similarly elevated in the absence of glucose, and this could reflect an augmented redox response to conditions of metabolic stress. In light of this evidence, further investigation of MCT function will help to refine their roles and regulation in the preimplantation embryo.
ACKNOWLEDGMENTS
We thank Dr. Nancy Philp (Thomas Jefferson University, Philadelphia, PA) for her generous gift of antibodies. Special thanks also to Emmy Hung for her technical advice and support throughout.
FOOTNOTES
1Supported by National Health and Medical Research Council of Australia project grant 210194 to M.P. and P.L.K. and National Institute of Child Health and Human Development National Cooperative Program on Female Health and Egg Quality cooperative agreement U01 HD044644. S.J. was supported by a University of Queensland Graduate School Joint Research Scholarship. 
Correspondence: 2FAX: 617 3365 1766; e-mail: p.kaye{at}uq.edu.au
Received: 10 December 2007.
First decision: 14 January 2008.
Accepted: 11 March 2008.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
