Background
The pre-ovulatory LH surge induces massive changes within the follicle, including morphological, physiological and molecular alterations [
1,
2]. The reorganization of the follicle is accompanied by a tremendous shift of the gene expression profile particularly in cells of the granulosa layer [
3,
4]. The aromatase-encoding gene
CYP19A1, and the gonadotropin receptors,
FSHR and
LHCGR, are significantly down-regulated immediately after the pre-ovulatory LH surge in vivo [
5,
6]. In contrast,
RGS2, encoding the regulator of G protein signaling 2, and the inflammatory-related genes
VNN2, encoding the vascular non-inflammatory molecule 2, and
PTX3, encoding pentraxin 3, were highly up-regulated [
4,
7,
8]. It was hypothesized that hypoxic conditions occur during folliculogenesis and support ovulation [
9,
10]. However, it is still unclear whether hypoxia is an essential signal during the folliculo-luteal transition [
11]. As for granulosa cells (GC) it was proposed that pericellular hypoxia present under high cell density conditions may induce differentiation [
12]. Likewise, we showed that high cell density induces specific changes of the gene expression and steroid hormone profiles in cultured GC mimicking an early post-LH status of GC differentiation [
13]. Furthermore, applying a whole-genome approach we could demonstrate in GC cultured at high plating density, that genes connected to hypoxia are affected [
14]. In particular
LDHA transcripts encoding L-lactate dehydrogenase were remarkably up-regulated. It was demonstrated also by others that
LDHA expression is regulated by hypoxia [
15]. In addition, a binding site for HIF1/2α could be identified in the
LDHA promotor explaining the hypoxia-related expression [
16,
17]. On the other hand Lee et al. [
18] proposed a HIF-independent mechanism of lactate accumulation under hypoxic conditions. Hence, L-lactate might play a role during the folliculo-luteal transition. Reports from different species demonstrated higher L-lactate concentrations within the follicular fluid than in the respective serum, ranging from 6 mM in human up to 27 mM in rats [
19‐
21]. Generated even during adequate oxygen provision L-lactate might represent an important regulator of metabolism [
22]. A study in L6 cells revealed that L-lactate may be involved in the delivery of oxidative and gluconeogenic substrates thus leading to the cell-cell and intracellular lactate shuttle hypothesis [
23,
24]. In this context L-lactate is also described as a metabolic signal. L-lactate affects its own metabolism by stimulating the expression of the lactate transporter MCT1 in rat muscle cells [
25]. In mouse granulosa cells MCTs were identified regulating the transport of L-lactate within the female reproductive tract [
26]. Moreover, in neuronal cells L-lactate affects the expression of genes linked to neuronal plasticity during establishment of the long-term memory [
27,
28]. In this study we therefore tested the hypothesis that L-lactate is a signaling molecule in the bovine follicle. To that end, we analyzed the effects of L-lactate in a serum-free estradiol (E2)-producing GC culture model [
13,
29,
30] on specific morphological, physiological and molecular parameters.
Discussion
Initial experiments with different lactate concentrations revealed significant effects on
RGS2 expression with 7.5 mM and on
CYP19A1 or
FSHR expression with 30 mM.
LHCGR was significantly affected only with 45 mM. However, this might be due to the generally observed greater variance of
LHCGR, of which the very low abundance levels were close to the detection limit. Nonetheless, based on these data, an L-lactate concentration of 30 mM was used as an effective concentration throughout the following experiments. Interestingly, the non-metabolized enantiomer D-lactate elicited more prominent effects on gene expression than L-lactate. An effect of D-lactate was also described by Latham et al. [
35] and Wagner et al. [
37], analyzing HDAC activity. However, it must be considered that the concentrations of D-lactate in mammalian species has been shown to be three orders of magnitude lower than those of L-lactate [
38], thus raising the question whether D-lactate plays any important role after all. But nonetheless, our data on increased effects of the non-metabolized D-lactate support our hypothesis that L-lactate can act as a signaling molecule apart from being metabolized. Moreover, it is reasonable that the enhanced effects of D-lactate are due to the non-metabolized status in mammalian cells. At this point we cannot exclude the possibility that a part of L-lactate is metabolized, which in turn leads to a reduction of L-lactate concentration and hence results in minor effects than the non-metabolized D-lactate. This leads to the suggestion that chirality might not be the primary parameter for signaling efficiency. This hypothesis has to be further taken into account in subsequent studies.
Interestingly, our results of different glucose concentration in GC media revealed no alterations in the gene expression. This indicates that an increased energy supply and thus higher metabolic activity alone is not sufficient to induce differentiation of bovine GCs in culture, which was shown in a previous work in sheep GC [
39].
In follicular fluids of rats L-lactate concentrations of 27 mM had been detected at the time of the expected LH surge, when the serum concentration of L-lactate remained considerably low at 5 mM [
21]. Studies in humans showed concentrations of 6.2 mM L-lactate at the time of oocyte recovery from women undergoing IVF [
19]. In the bovine different values have been reported ranging from 5 mM up to 14.4 mM depending on the size or stage of the follicle [
20,
40,
41]. Until now, effects of L-lactate on the cell physiology and gene expression have been only studied in rat L6 cells, human mesenchymal stem cells, human HCT116 cells or mouse neurons, where effective concentrations ranged from 10 to 30 mM lactate [
27,
34,
35,
42]. To our knowledge the present study is the first one that aims at clarifying the effects of L-lactate on GC differentiation in any species.
Short-term treatment of cultured bovine GC at the end of the culture period did not lead to considerable alterations. Neither hormone production nor gene expression profiles were significantly affected. Interestingly, this observation is in line with a conclusion by others that short-term exposure to L-lactate does not affect gene expression, but may already have some effects on the cell metabolism [
23]. Furthermore, it was suggested that effects on gene expression potentially require long-term exposure to L-lactate [
23]. This is clearly supported by our data showing that only long-term treatment could induce significant changes of physiological and molecular parameters in cultured GC, although the typical GC-like morphology and formation of cell clusters [
13,
30,
32] was not affected by L-lactate. Estradiol production clearly decreased whereas progesterone production was nearly unaffected compared to the controls. Estradiol down-regulation can be easily attributed to the remarkable down-regulation of the aromatase coding gene
CYP19A1. Besides
CYP19A1 we also analyzed additional marker genes of an early LH-dependent differentiation, which have been shown to be highly regulated following the pre-ovulatory LH-surge in vivo [
4,
5,
7]. In particular,
FSHR and
LHCGR were found down-regulated and
RGS2,
VNN2 and
PTX3 clearly up-regulated upon L-lactate treatment. These data together with the alterations of the hormone production suggest that L-lactate can induce a stage of differentiation in cultured GC that is similar to the early post-LH stage in vivo. Previous studies identified increased L-lactate concentrations after hCG stimulation in human GC culture or in the antral fluid of follicles in macaques thus supporting our conclusion [
43,
44]. L-lactate is well known as a product of anaerobic respiration, but may also play a role as regulatory molecule even during adequate oxygen provision [
22,
24]. Interestingly, it was described in several studies that L-lactate can act as a signaling molecule in fibroblasts [
45], macrophages [
46] and even neuronal cells [
27,
28]. Yang and colleagues showed that in neurons L-lactate affected the expression of genes involved in the establishment of long-term memory [
27]. In another recent study L-lactate exhibited neuroprotective properties against excitotoxicity by coordinating specific cellular pathways [
47]. The results of our study indicate for the first time a specific role of L-lactate as a signaling molecule in bovine GC.
As a next approach we tested if L-lactate uptake is an essential prerequisite to induce the observed effects. The inhibition of L-lactate uptake was accomplished with UK5099, a potent MCT inhibitor of L-lactate transport [
27]. This inhibitor is characterized to target the group of MCT 1–4 that are responsible for lactate as well as pyruvate transport. Accordingly, the transport of L-lactate into the cells should be completely inhibited by UK5099. Our results showed that the estradiol concentrations as well as the expression of selected marker genes nearly regained control levels after inhibiting the MCTs. This clearly suggests that the observed effects are vitally dependent upon L-lactate uptake into GC. It has been shown in neurons, that transport of L-lactate into the cells and K
ATP channels [
48] or NMDA receptor activation [
27] is necessary to promote the observed changes. In contrast to our data demonstrating strong effects of the non-metabolized D-lactate, the conversion of L-lactate to pyruvate is a crucial prerequisite in both cases. Whereas the rise of ATP production is essential for K
ATP channel actions, the increase in the NADH/NAD ratio leads to the modulation of NMDA receptor activity [
28]. Here we can only speculate on mechanisms occurring in bovine GC. But in human ovaries functional ATP-sensitive potassium channels could be identified and moreover it was shown that blocking of these ion channels negatively influences progesterone production [
49]. In ovarian tissue and granulosa cells expression of specific subunits of the NMDA receptor, the subunits 1 and 2B, had been identified [
50] thus supporting the view that this receptor might be involved in these cells as well. Expression of NMDAR subunits, especially of
GRIN2D and
GRINA has been found in vivo in GC (see database GranulosaIMAGE [
51]) as well as in vitro (own GC culture microarray dataset [
14]). Whether these receptor subunits can actually form functional receptors to be modulated by L-lactate in bovine GC has to be investigated in further studies. However, our data clearly show that L-lactate uptake is mandatory for most of the observed effects thus excluding the possibility that a G-protein coupled receptor for lactate, coded by
HCAR1, might be necessary to carry the L-lactate signal across the cell membrane [
52]. In addition,
HCAR1 transcripts were not detectable in our GC culture system (data not shown) thus indicating that
HCAR1 is not expressed in cultured GC. This is further supported by in vivo data demonstrating that
HCAR1 expression could not be detected in GC [
51]. The mode of action of L-lactate within bovine GC, however, is still an open question and has to be further elucidated.
At present, we cannot completely exclude the possibility that endogenous L-lactate production by the cultured cells might contribute to the observed effects. However, in an initial study without any lactate treatment we measured lactate concentrations in the media of approximately 7 mM (not shown), which is far from the effective concentrations used during the presented study. Therefore, it is reasonable to assume that endogenously produced lactate did not significantly interfere with our experimental settings.
The present results demonstrate that L-lactate has positive effects on the expression of
LDHA, the enzyme regulating the L-lactate-pyruvate-ratio as well as on the expression of its transporters,
SLC16A1 and
SLC16A7. This is in accordance with earlier observations by others that L-lactate directly stimulates the expression of the MCT1 protein (encoded by
SLC16A1) in L6 cells [
34]. Another study showed that endurance training, leading to a higher L-lactate production, increased MCT1 expression in rats and humans [
23]. This suggests that also in bovine GC L-lactate can stimulate its own metabolism and transport, thus being part of a positive feedback loop. Supportively, in skeletal muscle cells it was shown that MCT1 expression and L-lactate uptake is dependent on the activation of PKA and PKC signaling [
53]. This is an interesting hypothesis, having in mind that the PKA signaling is one pathway involved in inducing luteinization of GC [
54].