The effects of intravenous, glucose versus saline on ovarian follicles and their levels of some mediators of insulin signalling

Reproduction is subject to the influence of several factors related to the animal itself and to its environment. Among these, nutrition is one of the main factors affecting most aspects of the reproductive performance of the animal [1, 2]. Lindsay and his colleagues [1, 2] recognized that an increase in short-term food intake increased the lambing rate in ewes by stimulating ovulation rate. Three effects of nutrition on ovulation rate, the "acute", "dynamic" and "static" effects, have been described [2, 3] although there is uncertainty over the number of mechanisms involved [3]. The acute effect is an effect of diet associated with weight change, the static effect is associated with the absolute level of body weight while the dynamic effect is a short-term dietary effect not accompanied by a change in weight [reviewed in: 2, 3].

There have been numerous investigations of the relationship between diet and ovulation rate in farm animals and particularly sheep [2‐4] and of the physiological mechanism(s) responsible for this phenomenon [3, 5, 6]. The inclusion of energy-dense foods such as lupin grain [7‐13] or corn with soy meal [14] in the diet will increase both ovulation rate and the number of follicles. Other forms of nutritional manipulation for example, the infusion of leptin [15] or glucose [16‐18] have been investigated and shown to increase both the ovulation rate and the number of follicles. So clearly, there is a link between nutrition and folliculogenesis. This relationship could involve effects mediated by the intrafollicular glucose-insulin system [3, 5, 6] and other energy sensing mechanisms such as the AMPK system [6, 19]. Although there is an extensive literature (see references cited above) describing the effects of dietary energy on folliculogenesis and ovulation rate there are relatively few studies of the intrafollicular mechanisms that are affected by the energy content of the diet. The level of Aromatase P₄₅₀ was reduced in follicles of ewes whose diet was supplemented with lupin grain [20] or infused with glucose [18] and in the former study these effects were associated with alterations in the level of expression of the insulin receptor substrate (IRSs) proteins and in the later study, they were associated with alterations in the levels of Akt and AMPK; and Akt is a phosphorylation target in the insulin signalling pathway. In addition the presence of the insulin dependent glucose transporter (GLUT4 or the sugar transport facilitator [SLC2A-4]) has been confirmed in ovine [21], bovine [22] and rodent [23] ovarian follicles. Finally, insulin has been identified as a factor in several malfunctions of the ovary and its follicles in a number of ovarian pathologies [24]. These data suggest that insulin signalling to granulosa cells is implicated in the effects of dietary energy on folliculogenesis.

To improve the understanding of the interaction between folliculogenesis and the metabolic status of ewes we studied the effect of intravenous glucose on the insulin-signaling pathways in granulosa cells of follicles collected from ewes in either the luteal or follicular phases of the oestrous cycle. Although several potential metabolic sensors have been identified, including the IGF system, GH, and leptin [25, 26] we have focused our attention on the insulin-glucose and AMPK systems. A preliminary report of certain aspects of this study has been published [27].

In a ewe from a glucose-infused group the catheter failed and the infusion was interrupted. This animal was excluded from the experiment.

The results of this experiment show that in cyclic ewes during the breeding season, the intravenous infusion of glucose increased the total number of follicles present in their ovaries [17]. These findings agree with those obtained when using anoestrous ewes treated with eCG [18]. Furthermore the present experiment also shows that the effect of intravenous glucose on the follicle population of the ovary is present in both the luteal and follicular phases of the oestrous cycle. The concentrations of glucose achieved by the infusions peaked at 120 mg/dL or about 30% above the upper limit of the normal reference range for sheep (45-80 mg/dL). Additionally, this effect of glucose was also associated with a reduction in the circulating concentrations of oestradiol (Figure 3) but without any associated change in the concentrations of circulating FSH (Figure 3) and a reduction in the level of ovarian Aromatase P₄₅₀ in oestrogenic follicles (Figure 5). Furthermore, the effect of glucose was accompanied by an increase in the circulating concentration of insulin (Figure 2) and finally it was associated with changes in the insulin signalling and energy sensing mechanisms (Figures 6, 7 and 8) in granulosa cells.

Very similar findings have been reported for the effects of diets that increase the supply of energy. Feeding dietary supplements such as lupin grain [13, 17, 20, 35, 36], a mixture of soya meal and maize [14], or steamed corn flakes [37] or administering by gavage, an energy supplement in the form of a mixture of glycerol and propylene glycol [38, 39] to ewes all increased the number of follicles in their ovaries. In all these studies the nutritional treatment increased the concentrations of glucose and insulin in jugular venous plasma and these changes were associated with an increased number of follicles. Some of these studies also reported reduced concentrations of oestradiol in jugular plasma [20, 35], reduced levels of Aromatase P₄₅₀ in granulosa cell lysates [13, 20] and lower concentrations of oestradiol in follicular fluid from oestrogenic follicles [13, 20].

There is some controversy over the effects of either glucose or glucogenic diets on the circulating concentrations of oestradiol. We have reported decreased circulating concentrations of oestradiol following the infusion of glucose [18] or the feeding of a lupin grain supplement [20] and other authors have reported similar findings [35]. However other studies from our group found that feeding lupin grain [13] had no effect on oestradiol and other authors have also found no effect of glucogenic diets on the concentrations of oestradiol in jugular venous blood [14]. Using a different experimental model, these authors also reported that a supplement of lupin grain increased oestradiol concentrations and reduced those of FSH [36]. The main reason for these discrepancies is most probably because sheep have particularly low concentrations of oestradiol [40] and the oestradiol assays employed are being used at the very limits of their technical capability. The secretion rate of oestradiol was reduced by feeding a lupin grain supplement to ewes with ovarian auto-transplants [3] and in other experiments the concentration of oestradiol in the follicular fluid was reduced by feeding a supplement of lupin grain [13, 20] or a high energy diet [41] and in this experiment the infusion of glucose reduced the concentration of oestradiol in the follicular fluid of oestrogenic follicles (Figure 4). Furthermore these treatments were associated with reduced levels of Aromatase P₄₅₀ in granulosa cell lysates from oestrogenic follicles [13, 18, 20, 42]. So on balance we conclude tentatively, that it is very likely that both intravenous glucose and glucogenic diets can reduce the secretion of follicular oestradiol from oestrogenic follicles.

The effect of glucose on ovarian follicles appears to be dependent on the stage of development of the follicles. In the population of small follicles the effect of glucose was to stimulate their growth a conclusion that can be inferred from the increased number of small follicles in glucose-infused ewes (Table 2). Furthermore in small non-oestrogenic follicles this effect was not associated with any detectable change in the level of phosphorylated Akt, ERK1/2 and AMPK (Figures 6, 7 and 8) suggesting that this effect of glucose was not related to altered activity of insulin-signalling pathways in granulosa cells. This leads us to conclude that it is was possibly a direct effect of an increased supply of metabolic fuel (glucose) to the ovary. By contrast, in medium and large follicles glucose had little effect on the number of follicles (Table 2) but, in medium and large oestrogenic follicles it did have substantial effects on their physiological function through an inhibition of the level of aromatase P₄₅₀ (Figure 5) and a reduction in the concentration of oestradiol in follicular fluid (Figure 4). These actions were associated with changes in insulin signalling pathways and so these follicles are presumed to be responsive to both insulin and glucose and that the effect of glucose is insulin-mediated [43].

Extrapolating further, these data indicate that oestrogenic follicles are insulin-responsive [43] and that non-oestrogenic follicles are not. The population of non-oestrogenic follicles is a mixed population consisting of a sub-population of undifferentiated, small growing follicles and a sub-population of medium and large atretic follicles neither of which are insulin-responsive. Our findings suggest that glucose acts directly in growing follicles to stimulate growth leading to greater numbers of small follicles while in oestrogenic follicles which are physiologically functional but non-growing, glucose acts indirectly through an insulin-mediated mechanism to inhibit the synthesis and secretion of oestradiol [43].

The circulating concentration of IGF-I was not determined in this experiment. However, it was in our earlier experiment [18] where glucose increased the jugular venous concentrations of IGF-I. Despite the fact that IGF-I is a potent stimulator of both follicle growth [29] and oestradiol secretion [29, 44] we consider it unlikely that the effects of glucose in the follicle, in this experiment were mediated by IGF-I. IGF-I is a potent growth factor controlled by hepatic feedback systems that are independent of ovarian follicles [45]. Consequently there are intra-ovartian mechanisms that regulate the activity of IGF-I [46, 47] to protect the follicle from the potentially harmful effects of high concentrations of IGF-I. None-the-less it is possible that the insulin-independent effects of glucose on small follicles (Table 2) are an effect of IGF-I. Interestingly, glucose reduced circulating (Figure 3) and intrafollicular (Figure 4) concentrations of oestradiol suggesting that the intra-follicular bio-activity of IGF-I in oestrogenic follicles is suppressed.

The expected effect of reduced secretion of oestradiol (Figure 3) would be an increase in circulating FSH. How then can we explain the absence of the expected compensatory increase in FSH? Published data on the effects of glucogenic diets on the plasma concentrations of FSH presents a mixed picture. Some authors have reported increased concentrations of FSH and others have reported no change. It has been suggested that the reasons for these differences are technical and associated with the assays employed to measure ovine FSH [45] or that they reflect the dynamic nature of the negative relationship between FSH and the follicle [3]. However, it is also possible that the variation simply indicates the presence of uncontrolled and unidentified physiological differences among animals leading to unconscious bias in randomisation procedures. In an attempt to control this variability, Vinoles and her colleagues [36] used a first wave model to synchronise follicles waves and thus theoretically, reduce between ewe variability in the follicle population. Their results showed that supplementation with a glucogenic diet increased concentrations of oestradiol and reduced those of FSH in jugular plasma. However, the control of FSH by negative feedback from the ovary has two hormonal components namely oestradiol and inhibin and until the effects of glucogenic diets on the secretion of ovarian inhibin have been described a complete understanding of the effects of glucogenic diets on FSH will not be possible. Regrettably, at present there in no satisfactory technique available for measuring the concentration of inhibin in jugular venous plasma from sheep.

It is worth noting that the population of small growing follicles is an important source of inhibin [48]. Because glucogenic diets increase their number then theoretically, these diets should also increase the total secretion of follicular inhibin and the concentration of inhibin in the peripheral circulation. We can speculate that the resulting increase in inhibin negative feedback on FSH may thus compensate for any reduced oestradiol feedback and may explain why in some studies oestradiol decreased but FSH did not increase. We suggest that variability of the FSH responses to glucogenic treatments may reflect subtle changes in the balance of the effects of glucogenic diets on small growing (i.e. inhibin secreting) and large oestrogenic (i.e. oestradiol secreting) follicles.

The follicle has a functional insulin-glucose system [5] and the level of activity in the pathway can be estimated from the ratio of phosphorylated to non-phosphorylated forms of the various protein kinases in the pathway. Some of the kinases activated by insulin, although not exclusively, including Akt, ERK and AMPK have been detected in ovarian follicles from ewes [18, 49‐52] and in this experiment the phosphorylation ratios of all three of these were altered by the infusion of glucose. The Irish group [49‐51] have shown that both Akt and ERK are implicated in the mechanism of follicle selection and this study confirms their findings [18] showing that intravenous glucose decreased the level of phosphorylation of Akt and AMPK and furthermore, shows for the first time that glucose reduced the level of phosphorylation of ERK. How do these effects relate to the mechanism through which a glucogenic diet can increase ovulation rate? Ovulation rate in sheep is ultimately determined by the number of gonadotrophin-dependent, ovulatory follicles present at the time of the LH surge. The reduction in the secretion of oestradiol by the ovulatory follicle suggests that the effect of glucose is to impair FSH-stimulated synthesis of oestradiol, the reduced levels of Aromatase P₄₅₀ are consistent with such an action, and reduce its inhibitory influence (or dominance) on subordinate follicles in the cohort thus allowing for the emergence of additional ovulatory follicles. Potential mechanisms probably involve “cross-talk” [53] between the insulin signalling and FSH signalling pathways in granulosa and theca cells of the ovulatory follicle [5]. However, the details of how and where these signalling pathways interact are topics for future research.

These results show that an intravenous glucose that maintains the concentration of blood glucose slightly above the upper end of the normal reference range increased the number of small follicles and altered the function of follicular granulosa cells in large oestrogenic follicles but not in small follicles or non-oestrogenic follicles of any size. The glucose reduced the levels of Aromatase P₄₅₀ in large oestrogenic follicles and this was reflected in a reduced blood concentration of oestradiol. These reduced levels of Aromatase P₄₅₀ were associated with reduced levels of phosphorylated Akt, ERK and AMPK in large follicles. However, glucose did not affect the phosphorylation state of the energy sensor, AMPK and mediators of insulin action (Akt, ERK) in small follicles. These data suggest that the effect of glucose in small follicles is a direct action of glucose that increases the number of small follicles while the effect of glucose in large, oestrogenic follicles is an indirect insulin-mediated action that inhibits their capacity to secrete oestradiol.