| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Veterinary Basic Sciences,3 Royal Veterinary College, North Mymms, Hatfield, Herts, AL9 7TA, United Kingdom
ARC Centre of Excellence in Biotechnology and Development,4 School of Environmental and Life Sciences, University of Newcastle, Callaghan, New South Wales 2308, Australia
ABSTRACT
In Westernized societies, average consumption of n-6 polyunsaturated fatty acids (PUFAs) far exceeds nutritional requirements. The ratio of n-6 to n-3 PUFAs is generally >10:1 whereas on a primitive human diet it was closer to 1:1. Diets fed to intensively farmed livestock have followed a similar trend. Both n-6 and n-3 PUFAs can influence reproductive processes through a variety of mechanisms. They provide the precursors for prostaglandin synthesis and can modulate the expression patterns of many key enzymes involved in both prostaglandin and steroid metabolism. They are essential components of all cell membranes. The proportions of different PUFAs in tissues of the reproductive tract reflect dietary consumption. PUFA supplements (particularly n-3 PUFAs in fish oil) are promoted for general health reasons. Fish oils may also benefit fertility in cattle and reduce the risk of preterm labor in women, but in both cases current evidence to support this is inconclusive. Gamma-linolenic acid containing oils can alter the types of prostaglandins produced by cells in vitro, but published data to support claims relating to effects on reproductive health are lacking. Spermatozoa require a high PUFA content to provide the plasma membrane with the fluidity essential at fertilization. However, this makes spermatozoa particularly vulnerable to attack by reactive oxygen species, and lifestyle factors promoting oxidative stress have clear associations with reduced fertility. Adequately powered trials that control for the ratios of different PUFAs consumed are required to determine the extent to which this aspect of our diets does influence our fertility.
female reproductive tract,, fertility,, fish oil,, gamma-linolenic acid,, parturition,, prostaglandins,, PUFAs,, sperm,, steroid hormones
Fats and oils together comprise more than 33% of the daily calorific intake of the American human diet [1]. These include both saturated fatty acids (SFA) and mono- and poly-unsaturated fatty acids (MUFA and PUFA, see Tables 1 and 2 for summary abbreviations used). PUFAs have more than one double bond present within the molecule and are further classified into three groups on the basis of their chemical structure: omega-3 (n-3), omega-6 (n-6) and omega-9 (n-9), where the first double bond is located 3, 6, or 9 carbons from the methyl end of the molecule. Animals cannot synthesize n-6 or n-3 fatty acids de novo as they lack the appropriate fatty acid desaturase enzymes. The n-6 PUFA linoleic acid (LA) and the n-3 PUFA 
-linolenic acid (ALA) therefore need to be provided in the diet as they are absolutely necessary for numerous processes, including growth, reproduction, vision, and brain development [2].
|
|
The main dietary n-6 PUFA is LA, which is abundant in vegetable oils such as corn, safflower, sunflower and rapeseed oils [2]. Most n-3 PUFAs are derived from ALA, found mainly in the chloroplasts of green vegetables and grass. These two essential fatty acids can be converted in the liver to longer chain PUFAs by desaturation and elongation enzyme systems common to both pathways (Fig. 1). Fatty acid desaturase 2 (FADS2) is rate limiting [2], so higher tissue concentrations of long chain PUFAs can be achieved by bypassing this step. For example, plant oils such as evening primrose and borage oils contain high concentrations of dihomo-
-linolenic acid (DGLA), whereas the n-3 PUFAs eicosapentaneoic acid (EPA) and docosahexaneoic acid (DHA) may be supplied directly from fish oils. In addition to general dietary consumption, many women of reproductive age take PUFA supplements for various health reasons such as rheumatoid arthritis [3] and menstrual dysfunction [4].
|
The type and quantity of dietary fat is a significant risk factor for cardiovascular disease and some cancers, with the result that health organizations generally recommend reducing the intake of saturated fatty acids in favor of more MUFAs and PUFAs. While diets high in PUFAs are supposedly healthier, most people consume quantities of n-6 PUFAs far in excess of those needed for normal physiological functioning, predominantly as LA [5]. It is thought that both man and livestock species such as cattle evolved on a diet with an n-6 to n-3 PUFA ratio of 1:1 [6, 7] but modern dietary trends have increased this ratio. It now ranges from 10:1 to 25:1 in westernized human populations [8]. Modern intensively farmed livestock have little, if any, access to fresh pasture. PUFA concentrations in preserved forage are low, and animal diets are often supplemented with fat derived from oilseeds rich in LA [6]. Therefore, in both human and animal diets there are grounds for reducing the n-6 intake and increasing the n-3 intake to promote better health [3, 6, 8].
The proportions of different PUFAs in cell membranes reflect the amounts consumed in the diet [5]. As outlined below, there is considerable evidence that dietary PUFA supplementation can influence biosynthetic pathways involved in both prostaglandin synthesis and steroidogenesis that have multiple roles in the regulation of reproductive function. Furthermore, the PUFA composition of the cell membranes of the sperm and oocyte is important during fertilization [9]. Surprisingly little is known about the overall effects of PUFAs on fertility, although both positive and negative actions are theoretically possible. This review will focus on four areas of reproduction where PUFAs are thought to make a difference: establishment of pregnancy, uterine activity, preterm labor, and male fertility.
MECHANISMS BY WHICH PUFAs CAN AFFECT REPRODUCTION
Twenty carbon PUFAs are the direct precursors of a large group of physiologically active compounds called eicosanoids which include prostaglandins (PGs), thromboxanes, leukotrienes and lipoxins [10]. The 1- and 2-series PGs are derived from the n-6 PUFAs DGLA and AA respectively, whereas the 3-series PGs are derived from EPA (the n-3 equivalent of AA) (Fig. 1) [10]. The synthesis of PGs in tissues throughout the body is a very tightly regulated process. Excess PUFAs are ‘stored' in most cells within phospholipids in the membranes in an esterified form. The initial step is the generation of the PUFA substrate within the cell (e.g., AA for 2-series PGs). AA is liberated by the action of a phospholipase A2 (PLA2). Many phospholipases have been identified, of which intracellular Group IV cytosolic A2 (GIVA) appears most important in controlling the availability of free AA for PG synthesis [11]. Activation of cytosolic PLA2 is dependent on release of stored calcium from the endoplasmic reticulum, leading in turn to activation of protein kinase C [12]. The free AA (or other PUFA) generated is then metabolized by prostaglandin endoperoxide synthase (PTGS) enzymes, of which PTGS1 and PTGS2 are most relevant to reproductive biology (Fig. 2). Although PTGS1 and PTGS2 fulfill the same function, they are encoded by different genes, regulated differentially, expressed in a cell specific manner, and found in different subcellular compartments [13].
|
The PTGS enzymes provide PGH to PG synthases in a spatially segregated manner [13]. Two forms of PGE synthase (PTGES) have been identified, of which the inducible form PTGES1, located in the microsomal fraction, is the enzyme responsible for PGE2 output in inflammatory situations [14]. There are multiple isoforms of PGF synthases belonging to the aldoketoreductase family [15, 16]. These will be referred to collectively as PGFS.
PGs act through cognate G-protein-coupled receptors (e.g., EP, FP) [17, 18]. To date, four subtypes of the prostaglandin E receptor (PTGER) have been cloned (PTGER1, PTGER2, PTGER3, PTGER4), with multiple isoforms of the PTGER3 receptor subtype [17]. Activation of PTGER receptors can result in an increase in [Ca2+]i (via PTGER1 receptors), phosphoinositol turnover (via PTGER3 receptors), cAMP generation (via PTGER2 and PTGER4 receptors), or inhibition of agonist-stimulated adenylate cylcase activity (via PTGER3 receptors) [18 and references therein]. While this summarizes the typical second messenger responses mediated via PTGER receptors, it must be emphasized that the coupling of a given receptor subtype can vary dramatically between tissues and between species. Due to the similarities in structure between different PG molecules, there is some cross-reactivity of different PGs between receptors. The 2-series PGs are generally considered to be more biologically active than the 1- and 3-series PGs. For example, PGF2 had a 10-fold greater potency compared to PGF3 in protecting against ethanol-induced injury to the gastric mucosa in rats [19]. In bovine luteal cells, PGF3 had a relatively high affinity for the PGF receptor (PTGFR) (39 nM versus 10 nM for PGF2), as compared to PGF1, whose affinity was considerably lower (153 nM versus 10 nM for PGF2) [20]. On the other hand, in Chinese hamster ovary cells expressing different PTGES receptor subtypes, PGE1 and PGE2 had very similar Ki values for all four receptor subtypes (e.g., about 
12 nM for the PTGER2 receptor) [21]. The binding characteristics of PGE1 and PGE2 to PTGER3 receptors in bovine luteal cells were also similar [22]. On the other hand, PGE1 acting through the PTGER2 or PTGER4 was more effective than PGE2 at inhibiting vascular smooth muscle cell proliferation [23]. It is therefore unwise to assume that the 2-series PGs are always more potent than their 1- and 3-series equivalents.
PUFAs affect PG production by acting as substrates for, and competitive inhibitors of, cyclooxygenation and by altering expression or cellular concentrations of the various relevant enzymes. The proportion of different PUFAs in the diet alters cell membrane phospholipid composition and this becomes quantitatively significant because the precursors of each group of PGs compete for the same enzyme systems for metabolism [24]. This in turn will have considerable effects on the types of PG synthesized—and hence, physiological responses [25]. For example, in our studies on PG synthesis in the ruminant uterus, LA added in vitro was generally inhibitory. Either in vivo supplementation of an LA enriched diet to cattle [26] or in vitro supplementation of LA to endometrial cells isolated from late pregnant ewes also caused a significant decrease in production of 2-series PGs [27]. In contrast, feeding a diet high in LA to late pregnant ewes increased endometrial and placental PG production [28–30]. This was associated with an approximate 50% increase in AA concentrations in blood and caruncular endometrium, suggesting that in this case, effective metabolism of LA to AA had occurred, thus providing more precursor. In support of this, supplementation of cultured endometrial cells with GLA or AA significantly increased PG synthesis [27].
Affinities for precursors (substrates) decreases from the n-3 to the n-9 series PUFAs, thus when EPA is present in sufficient amounts, it can depress metabolism of n-6 PUFAs and hence the synthesis of the 2-series PGs. The preferred substrate for PTGS catalysis by either PTGS1 or PTGS2 is AA [31]. Although EPA is a substrate for both PTGS1 and PTGS2, its metabolism via PTGS1 is poor (about 10% that of AA), which means that its ability to generate 3-series PGs is also poor [31]. EPA is also an inhibitor of PTGS1 activity. This means that EPA not only inhibits 2-series PG generation (for example, in cultured bovine endometrial cells [32]), but does not induce, as expected, a concomitant increase in 3-series PG generation via PTGS1. Despite this, changes in the synthesis of 3-series PGs do occur following dietary supplementation with n-3 PUFAs in man [33]. As well as altering the proportions of 1-, 2-, and 3-series PGs produced, there is also evidence PUFAs can influence the balance between the production of PGF and PGE. We found that supplementing cultured ovine endometrial cells with the n-6 PUFAs LA, GLA, or AA all resulted in a significant shift to the production of more PGE2 relative to PGF2 [27]. Beef heifers fed a high fish oil diet had a higher mRNA expression of PTGES in the endometrium [34], suggesting a possible mechanism for this effect. On the other hand, dairy cows receiving chronic feeding of bypass calcium salts of fish oil (enriched in EPA and DHA) did not have altered PTGS2, PTGES, or PGFS mRNA expression in endometrial tissue collected at Day 17 of the cycle [35].
AA and its metabolites have long been implicated in steroidogenesis through direct effects on the steroidogenic machinery (e.g., steroid acute regulator [STAR] protein, cytochrome P450, family 11, sub family A, polypeptide 1 enzyme [CYP11A1]), or indirect effects via PGs. STAR plays a critical role in regulating steroid synthesis [36]. In MA-10 Leydig cells, inhibition of endogenous release of AA inhibited dibutyryl cAMP-induced steroid synthesis as well as STAR promoter activity, Star mRNA and STAR protein, whereas addition of exogenous AA reversed all these effects. Specific inhibition of PTGS2 was also associated with an increased Star expression and steroid output [37, 38]. This may be caused by a decrease in PGF2, as the latter inhibits STAR protein expression, consistent with its well-established role in inhibiting luteal progesterone synthesis [38]. In contrast, the age-dependent inhibition of testosterone production involves the suppression of STAR as a consequence of oxidative stress [39]. The cause of the latter is not known for certain, but is correlated with excessive AA flux through the PTGS2 pathway [40], possibly promoted by an age-dependent leakage of electrons from Leydig cell mitochondria [41].
PUFAs can also regulate adrenal steroidogenesis. Basal corticosterone synthesis was stimulated by LA [42] while ACTH-stimulated steroid synthesis was inhibited [43]. We have similarly shown that n-3 and n-6 PUFAs in vitro stimulated cortisol secretion by ovine adrenocortical cells [44]. Supplementation of ewes with a high n-3 PUFA diet (ALA) prior to tissue collection abolished this ability to respond to further PUFA additions in vitro. These changes in steroid synthesis and responsiveness were accompanied by changes in STAR protein and mRNA expression. The ability of PTGS metabolites of AA such as PGE2 to apparently mediate the stimulatory effects of ACTH on corticosterone secretion in the rat adrenal cortex [45] provides further evidence that PUFAs and eicosanoids modulate steroid synthesis in a variety of different steroidogenic tissues.
PUFAs and Transcription Factors
PUFAs may also alter the function of transcription factors controlling gene expression and can thus affect cellular concentrations of enzymes regulating both the PG and steroidogenic synthetic pathways. For example, upstream transcription factor proteins (USF-1 and USF-2) affect PTGS2 promoter activation [46]. Moreover, protein kinase A-dependent phosphorylation of USF enhances its ability to transactivate PTGS2 [47]. In luteinised granulosa cells and in bovine endometrial stromal cells, both n-6 and n-3 PUFAs activate protein kinase C, which in turn phosphorylates and activates phosphodiesterase [48, 49]. Therefore PUFAs should reduce USF-2 phosphorylation and PTGS2 transactivation through this effect. Given that PUFAs can activate mitogen-activated protein kinase 3 (MAPK3) [50], it is also possible that they could regulate PTGES2 and PTGS2 activation in a MAPK-dependent manner [51]. This raises the possibility that PUFAs could affect the activity of an enzyme such as PTGS2 through altered phosphorylation even when the absolute protein concentrations remain unchanged.
In relation to steroidogenesis, nuclear receptor subfamily 1, group H, member 3 (NR1H3) is a transcription factor known to respond to PUFAs [52] that affects expression of STAR, and therefore steroid synthesis in mouse adrenocortical cells [53]. NR1H3 is also present in the ovine ovary [54]. In addition to these factors, PUFAs affect another class of transcription factors, the peroxisome proliferator activated receptors (PPARs). PPARs are a family of nuclear receptors that are activated by endogenous ligands such as PUFAs [52]. A detailed discussion of PPARs and the consequences of possible interactions with PUFAs is beyond the scope of this review. It is, however, pertinent to draw the readers' attention to this class of molecules, as PPARs can affect many aspects of reproduction including gonadal steroid synthesis [55] and parturition [56].
Dietary treatment of mice with either EPA or DHA suppressed mitogen induced T-lymphocyte proliferation by inhibiting interleukin-2 (IL-2) secretion [57]. In investigating a possible mechanism for this response, the authors showed that the 3-series PUFAs (but not AA) affected the kinetics of two important intracellular second messengers, diacyglycerol (the physiological activator of protein kinase C) and ceramide. In cultured ovine uterine epithelial cells treated with either GLA or AA, PG production was up-regulated, but the cells became unresponsive to oxytocin [27, 58]. Whilst it is possible that this reflected saturation of available PTGS2, an alternative explanation could be due to alterations in cell signaling pathways. Previous workers showed that the myometrial oxytocin receptor could be converted from a low to a high affinity state by the addition of cholesterol, an effect thought to be mediated via changes in membrane properties [59]. These experimental findings raise the possibility that PUFAs may also differentially affect cellular responses through changes in membrane fluidity, receptor binding characteristics or their downstream activation.
The Establishment of Pregnancy in Ruminants
The establishment of pregnancy in ruminants requires ovulation of a competent oocyte, insemination at the correct time and an appropriate pattern of estradiol and progesterone during the follicular and luteal phases of the estrous cycle [60, 61]. Furthermore, the embryo must develop adequately to prevent luteolysis by the production of sufficient interferon
to inhibit up-regulation of endometrial oxytocin receptors [62]. In dairy cattle there is extensive loss of embryos over this period, so that only around 40% of cows remain pregnant 28 days after insemination [63].
There is some evidence that all of these events can be influenced by dietary PUFAs. Oocytes of many species contain high levels of fatty acids (FAs) [64]. Granulosa cell and oocyte FAs were more saturated than those of plasma, suggesting a selective uptake mechanism in follicles [65]. FAs are used as an energy source during oocyte maturation and the extended period of embryo development before implantation. Bovine oocytes exposed to methyl palmoxirate to block FA oxidation showed reduced capacity to form blastocysts after fertilization [66]. The PUFA content of oocytes can affect maturation, cryopreservation and subsequent developmental competence. ALA has been implicated in oocyte growth and differentiation, in regulating meiotic arrest at the germinal vesicle stage, and in preventing germinal vesicle breakdown [67, 68]. Higher numbers of grade 1 oocytes were collected from n-3 PUFA supplemented ewes [69]. Differences in FA composition between grade 1 and grade 2/3 oocytes suggests that these may be important for oocyte competence as well as fertilization and developmental potential [68]. However, feeding PUFAs (vegetable or linseed oils) in comparison with MUFAs to lactating dairy cows failed to affect oocyte quality, assessed by embryo development after IVF [70].
In cattle, dietary supplementation with various long chain PUFAs (both n-3 and n-6) induced changes in several aspects of folliculogenesis, including both an increase in total follicular numbers and in the size of the dominant or pre-ovulatory follicle [70–74]. In contrast, Petit et al. [75–76] found little or no effect of linseed, fish oil, flax, and sunflower seed supplemented diets on follicular dynamics. Granulosa cells collected from follicles of n-6 PUFA supplemented cows showed increased steroid secretion in vitro [77] and follicular phase estradiol concentrations were higher in cows supplemented with ALA [78].
Dietary PUFAs have influenced several aspects of luteal function in some but not all studies. Cows on diets high in LA had decreased progesterone levels (50%) on Days 4 through 8 of the estrous cycle while cows on a high ALA diet had decreased levels of progesterone on Days 4 through 15 of the cycle [78]. This suggested that n-3 or n-6 PUFAs either directly or indirectly (via PGs) exerted differential effects on ovarian steroid synthesis. The finding that n-3 PUFAs in vitro inhibited bovine luteal cell progesterone secretion supported this contention [79]. Other workers, however, found no changes in luteal progesterone secretion following diets containing flaxseed (mainly n-3) or sunflower seed (n-6) [74], Menhaden fish meal [80], or fish oil [73].
PGs are thought to act as luteotrophic factors in the early luteal phase. PGE2 stimulated progesterone secretion by cyclic corpora lutea whereas in the corpora lutea of pregnancy, PGE1 was more potent than PGE2 in stimulating progesterone secretion [81]. Later in the cycle, PGF2 is the main luteolytic agent [12]. There is evidence that n-3 PUFAs (e.g., ALA and EPA) can inhibit PGF2 release by bovine endometrial cells [82]. In support of this, we have shown that a high n-3 diet (ALA) can delay luteal regression in cyclic ewes through delaying the onset of pulsatile release of PGF2a [83]. In contrast, a number of trials conducted on cattle that have measured the amount of the PGF metabolite 13,14-dihydro-15-keto-PGF2 (PGFM) released in response to an oxytocin challenge given in the late luteal phase have produced conflicting results. On Day 15 of the cycle, Mattos et al. [80] reported reduced PGFM secretion in cows fed fish meal, but Petit et al. [75] found a higher response in cows given a mixture of linseed and fish oil compared with linseed alone. Petit et al. [76] reported a trend to higher PGFM on Day 15 on a diet supplemented with sunflower seed compared with flaxseed or Megalac. Robinson et al. [78] found no effect of LA- or ALA-based diets on Days 15 and 16 of the cycle, but on Day 17 cows on the LA diet had a higher response to oxytocin. These results suggest that dietary changes involving PUFAs can influence the PGFM response, but no consistent pattern has emerged. This may in part be due to the fact that the amount of PGFM released in response to oxytocin increases as the time of natural luteolysis approaches, so treatment differences may be confounded by slight changes in cycle length between animals.
Changes to the progesterone profile during the estrous cycle outlined above would be likely to affect early embryo development indirectly. Decreased progesterone concentrations in the early luteal phase in cattle reduce embryo survival, an effect thought to be mediated via changes in uterine secretions [84]. On the other hand, the ability of n-3 PUFA supplements to delay luteolysis has been proposed as a potential method of improving cattle fertility by giving the conceptus longer to develop before the onset of luteolysis [85]. A number of trials conducted over the past 20 years suggest that feeding fishmeal to cattle improves the conception rates [86–89]. However, results often did not achieve statistical significance and these data are far from conclusive at present. The two earlier studies [86, 87] failed to control for the level of protein in the feed, which was also increased, and in the study of Burke et al. [88] fertility was only improved in one of two dairies included in the trial, in which overall fertility was poor. Other studies have examined the effect of flaxseed on conception in cattle. Two earlier trials reported lower pregnancy losses in cows fed flaxseed [74, 90], but this effect was not repeated when larger numbers of cows were examined [91]. In order to detect a 10% difference in pregnancy rate with a power of 0.8 at P < 0.05, a minimum of 300 animals per group is required. These conditions have not been met in work conducted to date, so more large-scale, well-controlled trials are required to resolve this issue.
GLA and the Control of Uterine Activity in Women
Oils containing high concentrations of GLA (e.g., evening primrose and borage oil) have known anti-inflammatory properties, an effect thought to be mediated at least in part via increased production of PGE1 [23]. These products are promoted widely to the female population through health food shops and the Internet for a variety of reasons relating to reproduction. These include enhanced sperm transport for couples wishing to conceive, as GLA is said to aid sperm survival through improvements to the nature of cervical mucus [92]. Opinion is divided as to whether GLA containing oils increase uterine contractions, thus leading to induction of labor [93], or decrease them, resulting in relief of some symptoms of premenstrual tension [94].
LA can be metabolized into GLA by FADS2, but in practice only a small fraction of dietary LA is converted to GLA as this enzyme is highly rate-limiting (Fig. 1) [5, 23]. Previous work in various cell types indicates that supplemented GLA bypasses this step and is quickly elongated to DGLA [23, 95]. Due to the further rate-liming properties of fatty acid desaturase 1 (FADS1), only a small portion of DGLA can be converted into AA and the accumulated DGLA is converted into 1-series PGs by PTGS [23, 96]. Both Fads2 and Fads1 are highly expressed in liver, where expression is regulated by a variety of nutritional stimuli including composition of dietary fat [97–99]. Most PUFA metabolism is therefore likely to occur at this level following dietary consumption. The enzymatic activity of FADS2 is reportedly lower in other tissues, although significant expression is present elsewhere, for example in lung, heart, and skeletal muscle [97]. Fads1 expression likewise has a wide distribution including the placenta, but with a much lower abundance than Fads2 [98].
There is surprisingly little published data available to support the contention that GLA consumption modulates aspects of reproduction. Our studies based on ovine uterine epithelial cells confirmed that addition of GLA in vitro increased both the absolute and proportional PGF1 production and slightly enhanced PGF2 generation, whereas supplemental AA increased PGF2 generation as expected [58]. Both GLA and AA increased overall PGF output significantly but prevented the cells from responding to oxytocin. GLA treatment of isolated rat uterine strips similarly enhanced the output of PGE1 but did not influence the generation of PGE2 or PGF2 [100]. As the PUFAs in both experiments were added directly to the uterine tissues, these data suggest indirectly that elongase activity is present in uterus but that FADS1 expression is low or absent. There is, however, little known about how much PGF1 is produced by normal reproductive tissues with or without GLA supplementation, or what the biological effects of PGF1 are in the context of reproduction. PGF1 was unable to decrease luteal progesterone concentrations in bovine corpora lutea in vitro [101] and it did not cause luteolysis in pregnant rats, although there was evidence of intrauterine growth retardation [102]. This is consistent with its lower affinity for the PGF receptor in comparison with PGF2 [20]. Endometrial biopsies from women taking a 6-month dietary supplement of GLA had a reduced ability to synthesize PGF2 and PGE2 in vitro, but 1-series PGs were not measured [103]. Intravenous infusion of PGF1 to one woman induced labor, but it was estimated to be 10 times less potent that PGE1 in this respect [104]. An evening primrose oil supplement was also evaluated in relation to fertility in the blue fox. It tended to improve the conception rate but the abortion rate also increased, so there was no overall effect on the number of females producing litters [105]. Further studies to determine the relevance of 1-series PGs to reproductive processes are thus urgently required.
Preterm birth occurs in about 10% of all human pregnancies and is associated with 70% of neonatal deaths [106]. Spontaneous preterm labor is a worldwide problem that affects countries regardless of economic status. In most eutherian mammals, it is largely accepted that changes in the hypothalamo-pituitary-adrenal axis are responsible for initiating the onset of parturition [107]. This was originally demonstrated in sheep, where a rise in fetal cortisol initiates a chain of events that culminates in the birth of the offspring. The main action of cortisol in the context of parturition is to raise levels of 2-series PGs [106, 108]. Both fetal and maternal tissues, including amnion, chorion, and decidual endometrium, synthesize PGs in vitro [107]. The levels of PGs along with their synthetic enzymes (mainly PTGS2) increase either before or at the time of labor and inhibitors of PG synthesis prolong gestation [108].
PUFAs may thus be able to influence the timing of parturition through alterations to PG or adrenal steroid synthesis as previously discussed. A diet high in n-6 PUFAs is generally thought to be associated with an increased risk of preterm delivery. Plasma concentrations of AA increase throughout pregnancy, with the highest levels being observed during labor followed by a rapid decline postpartum [109]. The onset of normal labor is associated with increased levels of both LA and AA in blood [110]. Sheep fed a diet rich in n-6 PUFAs during late gestation had higher circulating concentrations of 2-series PGs (PGE2 and PGF2), produced more PGs in uterine and placental tissues, and appeared to be more susceptible to preterm parturition [28–30]. Similar observations have been reported in human subjects, where women who delivered prematurely were shown to possess higher n-6 PUFA concentrations in erythrocytes [111,112].
In women, low n-3 PUFA consumption during pregnancy is also associated with a higher risk of preterm labor and low birth weight [113]. Conversely, a diet high in n-3 PUFA is associated with an increase in gestational length and birth weight in rats [114–116] and human populations with high fish consumption such as that seen in Faroese women and Greenland Eskimos [117]. In rats, excessive n-3 PUFA consumption during pregnancy was associated with a prolonged and difficult labor [118, 119]. Various mechanisms have been suggested to explain these effects. They could be as a direct result of a shift from synthesis of 2-series to 3-series PGs and significant production of 3-series PGs by rat uterus has been confirmed using mass spectroscopy [120]. In late pregnant sheep, infusion of n-3 PUFAs was shown to inhibit the course of betamethasone-induced labor [121], an effect that was associated with a decrease in myometrial PTGS2 mRNA [122]. Based on ratios of FAs measured in serum phospholipids, Greenland Eskimos also have decreased FADS1 activity, leading to proportionately lower circulating concentrations of AA (see Fig. 1) [123].
These findings have led to the suggestion that dietary supplementation of pregnant women with n-3 PUFAs may be used to diminish the risk of premature delivery in high-risk subjects, as it represents a low-cost, minimally invasive strategy [124]. Some well-powered epidemiological studies have appeared to suggest that such interventions have been successful, with mean increases in gestation length of 4 and 6 days, respectively [125, 126]. Women with a previous preterm delivery had a significantly lower recurrence following EPA and DHA supplementation in comparison with a placebo group, reducing the risk from 33% to 21% [127]. A further trial that provided women with DHA-enriched eggs found fewer low-birth-weight and preterm babies and a larger placental size [128]. In contrast, another large randomized clinical trial with supplemental fish oil in Norwegian women had no effect on either birth weight or gestation length [129]. Furthermore, supplementation of seafood to an American population slowed fetal growth, decreased birth weight, and did not reduce the risk of preterm birth [130].
In their review of this topic, Allen and Harris [7] drew attention to the need for high doses of fish oil in order to obtain an effect and also highlighted the possible differences in action of DHA and EPA. These are hard to separate, as many preparations contain both PUFAs in varying proportions. The most comprehensive evaluation of the available data to date [131] concluded that results obtained so far are inconclusive and recommended that further interventionist studies be carried out both to assess the efficacy of n-3 PUFA supplementation in preventing prematurity and to understand the mechanistic nature of the PUFA action.
Long chain PUFAs have been detected in the spermatozoa of man [132, 133] and a variety of livestock species including both mammals (ram, bull, boar [134]) and birds (chickens, ducks, turkeys, [135]). In birds, the most abundant were the n-6 PUFAs AA (5%–9%) and docosatetranoic acid (22:4 n-6, 15%–21%). These were synthesized from LA, which was the most abundant PUFA in the diet (15%), but was present at a lower concentration in spermatozoa (2%–3%) [136]. Altering the PUFA sources in the diet resulted in concomitant changes in the n-6 and n-3 composition of sperm (e.g., boar [137], cockerel [138, 139]).
These unsaturated fatty acids give the sperm plasma membrane the fluidity it needs to participate in the membrane fusion events associated with fertilization. However, these molecules are also vulnerable to attack by reactive oxygen species (ROS), initiating a lipid peroxidation cascade that can seriously compromise the functional integrity of these cells [132, 133, 140–142]. Such peroxidative damage would be expected to disrupt the fusogenicity of the sperm plasma membrane and the ability of this structure to support key membrane-bound enzymes such as ATPases. In addition, changes in membrane fluidity could theoretically impede the assembly and activation of signal transduction pathways critical to the fertilization process, although this has not to our knowledge been studied. The propagation of peroxidative damage is catalyzed by the presence of transition metals, such as iron or copper that can change their valency state by gaining or losing electrons. Measurements of the products of lipid peroxidation, such as malondialdehyde and 4-hydroxyalkenal, have been shown to correlate negatively with semen quality in both man and domestic animals [132, 143, 144]. Reductions in human semen quality as a consequence of smoking, infection, irradiation, varicocele, oligozoospermia, and toxicant/heavy metal exposure have all been linked with oxidative stress and the promotion of lipoperoxidation [145–148]. Similarly, in animal models the adverse reproductive effects of irradiation and xenobiotic exposure have been linked to the induction of peroxidative damage in the spermatozoa. For example, trichloroethylene, a widely used industrial solvent and common water contaminant, has been found to disrupt fertility in male rats as a consequence of oxidative damage to the spermatozoa [148].
Given the importance of lipid peroxidation in the generation of defective mammalian spermatozoa, there is an urgent need to determine the sources of oxidative stress that precipitate this condition. In this context, four major factors are currently acknowledged: 1) abstinence, 2) leukocytic infiltration, 3) antioxidant depletion, and 4) excess free radical generation by the spermatozoa themselves. Prolonged abstinence has a negative effect on the motility and functional competence of human spermatozoa, via mechanisms that involve oxidative stress and are particularly pronounced in spinal cord injury patients [149, 150]. Leukocytospermia (seminal leukocyte counts in excess of 1 x 106/ml) is usually a consequence of chronic infection and may be associated with oxidative damage to human spermatozoa depending on the site of infiltration, the number and type of leukocytes involved, the duration of exposure, and the state of leukocyte activation [9, 151, 152].
Extracellular antioxidants are extremely important for the protection of mammalian spermatozoa against oxidative stress because the cytoplasmic extrusion associated with sperm morphogenesis depletes these cells of their internal store of antioxidant enzymes. As a consequence, the fluids bathing the spermatozoa during their passage through the male reproductive tract are endowed with highly specialized secreted forms of antioxidant enzymes. The latter include members of the glutathione peroxidase and periredoxin families as well as superoxide dismutase and a host of small molecular mass free radical scavengers, including carnitine, tyrosine, uric acid, and vitamin C [153–157]. Since seminal plasma is recognized as one of the most powerful antioxidant fluids known to man, is not surprising that deficiencies in this protective milieu have been associated with oxidative stress and male infertility [158]. In animal models, surgical ablation of the male accessory glands creates a state of oxidative stress in the spermatozoa reflected in high levels of DNA damage and high rates of embryonic loss in mated females [159]. In the clinical situation, a variety of factors can deplete the seminal plasma of its antioxidant properties, including leukocytic infiltration and dietary/lifestyle factors, particularly smoking. Men who smoke heavily possess significantly depressed levels of antioxidants in their seminal plasma, exhibit clear signs of oxidative DNA damage in their spermatozoa, and produce children at increased risk of contracting childhood cancer [160–161]. In light of these data, antioxidant (vitamin E, glutathione, coenzyme Q10) therapy is being increasingly utilized as a treatment for male infertility [155]. Although many studies in this context are flawed because of the lack of appropriate patient selection, those that have been properly designed are returning promising results [162, 163]. Experiments on chickens have shown that feeding more PUFAs in the diet reduced the antioxidant status and quality of the semen (sperm concentration, semen volume). The importance of lipid peroxidation in this context was suggested by the ability of vitamin E, a chain breaking antioxidant, to reverse the negative impact of PUFA supplementation [138, 139].
The final source of oxidative stress is the spermatozoa themselves. Defective human spermatozoa generate ROS—the more intense this activity, the more impaired sperm function becomes [164]. One of the other features of defective human spermatozoa is that, according to some authors, they contain excessively high amounts of unsaturated fatty acid, particularly DHA and AA [165, 166]. A causative relationship between the retention of high levels of unsaturated fatty acid and ROS generation was indicated by a recent study indicating that exposure of human spermatozoa to the PUFAs LA, AA, and DHA triggered free radical generation, lipid peroxidation, and DNA damage in human spermatozoa [167]. Why some spermatozoa should contain more free unesterified unsaturated fatty acid than others is probably related to their state of developmental maturity, and specifically to the amount of residual cytoplasm retained by these cells. Immature, defective human spermatozoa generating high levels of ROS are characterized by the abnormal retention of cytoplasmic remnants and the presence of high levels of unsaturated fatty acid [164, 165, 167]. One of the predicted consequences of excess residual cytoplasm is not only increased unsaturated fatty acid content but also a corresponding abundance of cytoplasmic enzymes (particularly glucose-6-phosphate dehydrogenase) that will stimulate the generation of NADPH in the cytoplasm via the hexose monophosphate shunt [167]. The latter is then predicted to fuel the production of ROS by a putative membrane bound NADPH oxidase [168]. The lipid peroxidation induced in this manner should activate PLA2, thereby facilitating the release of yet more free unsaturated fatty acid from the phospholipid pool and further enhancing the generation of reactive oxygen metabolites. The resultant oxidative stress will generate peroxidative damage to the extent that the patients' antioxidant status and local transition metal availability will allow (Fig. 3). Under these circumstances, a combination of low antioxidant status and high ROS generation would be expected to seriously compromise the fertilizing potential of the spermatozoa, as is indeed the case [169].
|
Despite the potentially harmful effects of high PUFA concentrations on spermatozoa, experiments in farm livestock have been conducted to see if dietary supplementation with various PUFAs could improve fertility. In boars, enrichment of the semen diluent with DHA enriched egg yolk before freezing did not improve sperm quality following thawing. Addition of 3% fish oil to the daily boar ration increased the DHA content of the spermatozoa from 33% to 45% and increased the number of sperm in the ejaculate, but did not alter freezability [137]. Feeding boars shark oil improved sperm motility and velocity parameters [170], but another study reported no improvement in nonreturn rates to artificial insemination using semen derived from boars supplemented with cod liver oil [171].
There is overwhelming evidence that dietary PUFA consumption can alter both the type and amount of PGs produced by different tissues. Dogma predicts that DGLA will increase production of 1-series PGs and that n-3 PUFAs will both reduce overall PG output and alter the balance of production in favor of 3-series PG. While there are good grounds to support this view, hardly any studies related to the reproductive system have actually measured 1- or 3- series PG production, in large part due to limitations over appropriate methodology. Furthermore, there is also a lack of information on how organs such as the uterus will respond physiologically to different mixtures of the various PGs that may be generated. An additional complication is the growing evidence that PUFAs may also alter the ratio of PGE to PGF produced. While increased supply of AA at a tissue level generally increases the total output of 2-series PGs, this may also limit the ability of the tissue to respond to stimuli such as oxytocin. This effect could be due to limitations on PTGS availability, or to alterations in membrane properties, another topic that awaits study. In relation to steroidogenesis, most work to date has focused on AA. Again, the evidence is clear that AA can influence steroid output at a cellular level. However, both stimulatory and inhibitory effects have been reported in vitro and responses of animals fed different PUFA diets have also varied considerably. For example, n-3 PUFAs have decreased or had no effect on luteal progesterone output in different studies on cattle.
Reported effects of PUFAs on various aspects of female reproduction are equally unclear. Early enthusiasm for the possibility of using fish oil to improve fertility in cattle or to reduce the risk of preterm labor in women has not always been replicated in later studies. Some investigations have been underpowered. Others have lacked appropriate controls. Responses between individuals clearly vary according to the balance of PUFAs stored in particular tissues. This is influenced by the immediate life history of the animal and is very difficult to either regulate or measure accurately in human populations. While it is relatively easy to give more of a particular PUFA such as DGLA or DHA, it is harder to design appropriate control diets that are balanced for both protein and energy intake, as all other fats that may be used instead are likely to be bioactive in their own right. There are also real issues over the cost and palatability of some of the supplements, particularly in relation to animal diets.
With regard to male fertility, PUFAs are essential by virtue of their ability to confer upon the sperm plasma membrane the fluidity it needs to achieve fertilization. Moreover, they are critical components of a particular class of fucosylated glycosphingolipids that are essential for male fertility. They are, however, also a potential source of pernicious alkoxyl and peroxy radicals, should these cells experience oxidative stress as a consequence of the range of factors related to disease status or lifestyle previously discussed. PUFAs have similar dichotomous effects on other aspects of male reproduction such as Leydig cell steroidogenesis. On the one hand, AA can stimulate the production of testosterone via effects on STAR. On the other hand, the age-dependent inhibition of testosterone involves the suppression of STAR as a consequence of oxidative stress, possibly promoted by an age-dependent leakage of electrons from Leydig cell mitochondria.
In conclusion, PUFAs can affect many reproductive processes. As consumers, we are all bombarded with advertisements to alter our diets to promote healthy living. It is widely accepted that fertility of both human and farm animal populations in westernized societies is currently decreasing. However the jury is still out as regards appropriate advice over PUFA consumption. It appears that PUFAs are a two-edged sword—some are essential, but too many are potentially harmful. We remain largely ignorant as to the best balance to take at different points in our lives in order achieve optimum fertility.
ACKNOWLEDGMENTS
The authors gratefully acknowledge our many colleagues who have contributed both by assisting with the work and through helpful discussion.
FOOTNOTES
1Supported by BBSRC, Defra, the Wellcome Trust, and the Australian Research Council. 
Correspondence: 2D. Claire Wathes, Department of Veterinary Basic Sciences, Royal Veterinary College, Hawkshead Lane, North Mymms, Hatfield, Herts, AL9 7TA, UK. FAX 44 1707 666371; e-mail: dcwathes{at}rvc.ac.uk
Received: 1 February 2007.
First decision: 28 February 2007.
Accepted: 3 April 2007.
REFERENCES
receptor in the corpus luteum: recent information on the gene, messenger ribonucleic acid, and protein. Biol Reprod 2001; 64:1041–1047 stimulates cAMP phosphodiesterase via protein kinase C in cultured human granulosa cells. Mol Cell Endocrinol 1991; 2:207–214. Endocrinology 2002; 143:2571–2583
