To our knowledge, this is the first report to show that exposure to PBs and their metabolites in human semen negatively correlated with sperm parameters. This study is also one of the few healthy population studies on the association between IE and sperm quality. We first proposed that PBs, as exogenous chemicals, were possibly contributing to an increase in IE (LPS) through antibacterial activity. Our results suggest a new pathway for the effects of PBs on male reproductive health; specifically, that PBs likely promote IETM by changing the composition of gut microbiota and intestinal permeability, thus affecting sperm quality.
PB levels in seminal plasma were associated with sperm quality
MeP, EtP, PrP, and BuP are the most commonly used representatives of PBs [10].They are easily transported into the circulation after absorption through the skin or dietary intake [47]. They are rapidly hydrolyzed by nonspecific esterases to PHBA, which is conjugated with sulfate, glucuronic acid, or glycine prior to being excreted, and is mainly eliminated through urine [21]. PHBA and its conjugates are the main PB metabolites and account for about 52.7–63.8% of all PB metabolites[12]. 3,4-DHB is a product of the hydroxylation of PHBA [48]. In this study, we detected the total state (the free state plus the binding state) of parent PBs, PHBA, and 3,4-DHB to evaluate the overall exposure level and the relationship with sperm parameters.
Characteristics of PBs in semen
In this study, the concentration of each parent PB and their metabolites in seminal plasma varied considerably (Table 2). The detection rate and level of MeP, EtP, and PrP were high (96–100%), while those of BuP were the lowest (56%), which was consistent with the previous findings of Buck Louis [30]. This may be because MeP, EtP, and PrP are the most widely used PBs, especially MeP and PrP. However, consistent with other exposure studies, we found that the concentration of PBs in semen was lower than in urine, which also differed from the concentration in blood [16, 17, 86] (Table S2), mainly because urine is the main route of excretion [21]. BuP was detected in seminal plasma at a higher rate than in urine[16, 17, 30, 49] (Table S2), which was consistent with the properties of long-chain esters, with decreased water solubility and increased fat solubility, compared to other short-chain PBs [50]. The concentration of PHBA was far higher than that of the parent PBs, and even higher than their sum (geometric mean (GM) = 12.451, range: 2.284–260.596 ng/mL, Table 2). The concentration of 3,4-DHB was high but only 38% was detected, and it was lower than PHBA. PHBA and 3,4-DHB are the metabolites of PBs, and are also derived from many other sources, such as other precursors of flavonoids and cinnamates, and natural sources (such as wine and berries) [21, 51]. Thus, both PHBA and 3,4-DHB are nonspecific exposure biomarkers. Currently, only four studies that reported PBs in semen were identified [16, 30, 31, 86], but the concentrations of PHBA and 3,4-DHB were not reported. Our study is the first to identify the exposure levels of parent PBs and their metabolites in semen of healthy men.
Relationship between PB exposure and sperm parameters
The results showed that the concentration of four PBs and two metabolites were negatively correlated (Fig. 1B) with total sperm count, sperm concentration, and sperm motility parameters. EtP and MeP had the strongest correlation with sperm motility parameters. EtP and BuP had the strongest correlation with sperm concentration and total sperm count. Therefore, EtP may have the most important influence on sperm quality. We also found that PHBA was significantly negatively correlated with sperm quantity parameters and motility parameters, and 3,4-DHB was only correlated with sperm concentration and total sperm count (Fig. 1B). To our knowledge, no other similar studies have been reported.
Rat studies have shown that PBs caused male reproductive toxicity through estrogenic or antiandrogenic activity [22, 23, 24, 52], including pathological changes in testis and spermatogenesis disorders. An in vitro experiment showed that BuP promoted the generation of human sperm reactive oxygen species, leading to decreased sperm motility and increased apoptosis [53]. At present, only one study explored the relationship between semen PBs and sperm parameters, but no relationship between the two was found [31]. In our study, for the first time, a negative correlation between PBs and metabolites in semen and sperm quality parameters was found in a healthy population, with EtP showing the strongest association with sperm quality parameters.
Semen PB level is more sensitive than urine for assessing sperm quality
It was previously reported that PBs in urine were negatively associated with human sperm concentration and motility rate [25, 27], and positively associated with abnormal morphology, sperm DNA damage [25], and sperm chromosome disomy [26]. The median concentrations of Mep, EtP, PrP, and BuP in urine in these reports were 3.48–15.6 µg/L (ng/mL) [25] and 0.03–6.55 ng/mL [27]; and concentrations of PrP and BuP in urine were 3.82 µg/L and 3.45 µg/L [26], respectively. However, in our study, Mep, EtP, PrP, and BuP median concentrations in semen were 0.018–0.505 ng/mL which was clearly one to two orders of magnitude lower than the concentrations of PBs in urine. This shows that exposure levels are differentially reflected in various biological media [16]. As there were no studies on the relationship between blood PBs levels and sperm quality, we studied the effect of PB exposure on sperm parameters as measured in semen, and found the concentrations lower in semen. Therefore, the levels of PB exposure as measured in semen are more sensitive for evaluating sperm quality.
Association between intestinal endotoxin and sperm quality in healthy men
Healthy people may have endotoxemia [35], which is the presence in the plasma of endotoxin (LPS) derived from the gut. LPS has been identified as a causal or complicating factor in multiple diseases [54], including impaired reproductive health [41, 42, 55].
The measurement of LPS in serum or plasma is the most direct way to quantify endotoxemia. LPS may enter the blood through transmucosal epithelial translocation [56]. It can be detected in serum or plasma under normal physiological conditions [56, 57]. Limulus amoebocyte lysate (LAL) is the gold standard for the detection of LPS in biological samples [58]. In this study, we examined LPS concentrations in the plasma of 315 healthy men by using LAL. The detection rate of LPS was 100%. LPS was generally at a low level, with a range of 0.021–0.195 EU/mL. The mean levels of LPS in the systemic circulation of healthy humans varied over a wide range, according to previous studies. The variation can be related to the method that was used to prepare plasma and serum samples prior to assay, differences among the subjects, or the reagent kit. For example, Nádházi et al. [56] detected the plasma endotoxin level of 116 healthy donors by LAL, which was less than 1 EU/mL (0.01–1.0 EU/mL) in most of the donors. Kallio et al. [59] reported that in a nutrition cohort study of 2452 people, the serum LPS of the control group was 54.2–63.2 pg/mL, equivalent to 0.135–0.157 EU/mL. Gnauck et al. [60] summarized 44 studies that examined endotoxin levels in healthy subjects and found that the overall mean level of systemic endotoxin in healthy subjects ranged from 0.15 to 61 EU/mL, with a median of 0.32 EU/mL. Our results of LPS were lower than the above studies and the range was within the results reported by Nádházi et al. [56] We speculated that this was because the subjects in our study were young and healthy men, and men with infectious diseases were excluded. In general, our results suggested that the LAL assay could be used to effectively detect circulating LPS and may be a more sensitive health indicator for a population. The data presented here is the first reported baseline data on LPS in Chinese healthy men.
The relationship between IE levels and reproductive health effects in the healthy population need to be evaluated. Previous studies have reported that animals treated with bacterial LPS showed testicular dysfunction at multiple levels [61], including steroidogenesis, spermatogenesis, and testicular cells apoptosis [62–64]. LPS in vitro disrupted steroidogenesis, induced apoptosis [65], disrupted the expression of the estrogen and androgen receptors [66], and induced oxidative stress of testicular cells [67, 68]. In addition, human and animal studies have shown that metabolic endotoxin associated with high-fat food or obesity could: (1) induce host immune responses, including local inflammation in the epididymis and testis, and produce a large number of pro-inflammatory cytokines (such as IL-1β, TNF-α, and IL-6) that further impaired Leydig cell function and testosterone production [69]; (2) cause germ cell apoptosis and morphological degeneration of testis; (3) upregulate the expression of nuclear factor-κB (NF-kB)/Jun N-terminal kinase (JNK) extracellular signals pathways and disrupt spermatogenesis [70]; and (4) reduce testicular mitochondrial function [42].
In this study, we found that LPS was significantly negatively correlated with sperm concentration and total sperm count, and that the correlation was more significant when LPS was higher than 0.104 EU/mL. Multiple linear regression showed that when LPS was higher than 0.104 EU/mL, the contribution rates of sperm concentration and total sperm count were higher (R2 = 0.295 and R2 = 0.236, respectively) (P < 0.01) (Table 5). In addition, the mean LPS concentration in group A2 and group A3 were 2.2- and 4.6-fold higher, respectively, than that of group A1, when LPS was significantly correlated with sperm parameters. In the report of Ding et al. [42], their analysis of 12 healthy and 48 infertile men revealed a strong negative correlation between the combined abundance of G− (Bacteroides and Prevotella) and sperm motility, as well as a strong positive association between the abundance of Bacteroides and the blood endotoxin concentration of patients. They also revealed that high-fat food induced microbiota dysbiosis in mice and that elevated endotoxin was associated with defects in spermatogenesis. Their study confirmed the harmful effects of IE on sperm quality. In our study, we analyzed the data of 315 healthy men and found that higher endotoxin (LPS) in plasma correlated with lower sperm quality, which supported the conclusion of Ding et al. [42] and provided additional significant evidence for population research.
PB exposure was associated with intestinal endotoxemia
There are various reasons for the increase of IE in blood and further promotion of IETM. Although excessive intake of common nutritional factors has been shown to induce IETM [33, 71, 72], we chose to focus on exploring the effects of endotoxemia induced by exogenous chemicals on reproductive health.
PBs are widely used antimicrobial preservatives [73]. Our previous study found that PBs, especially BuP, could increase the proportion of G−/G + in the intestine of rats and the levels of LPS in blood (Supporting Information Table S3; Figures S1–S9). In this study we tested PB exposure levels as measured in seminal plasma from 315 healthy sperm donors, and first analyzed the relationship between PB levels and LPS concentrations. The results showed that the levels of EtP and BuP in seminal plasma were positively correlated with the concentration of LPS in plasma (P < 0.05, Fig. 1E), and the higher the EtP level was, the stronger the correlation with LPS (r = 0.480, P < 0.01, Table 6). When EtP was higher than the upper tertile of 0.880 ng/mL, the regression coefficient for LPS was β = 1.139, and R2 = 0.293 (P < 0.05, Table 6). Similarly, the concentration of BuP was associated with the concentration of LPS. We analyzed the possible reasons. First, PBs have the ability to change the diversity of gut microbiota according to their antibacterial activity. A rat study provided initial evidence that postnatal low-dose exposure to MeP was capable of modifying the gut microbiota in adolescent rats [43]. G+ (such as Staphylococcus aureus) are more sensitive to PBs with longer alkyl chains than G− (such as Escherichia coli) [44, 74], due to the significant differences in lipid composition of plasma membranes between G + and G − bacteria [75]. When PBs inhibited G+, the abundance of G − increased, and more LPS was released. Second, PBs may increase intestinal permeability. BuP had direct toxicity on human colorectal adenocarcinoma (Caco-2 cells) [45], and our previous animal experiment showed that BuP decreased the expression of the intestinal tight junction protein ZO-1 (Supporting Information Figures S8 and S9), and was speculated to contribute to increased intestinal permeability, which possibly promoted more LPS to leak into circulation.
The results also showed the total concentrations of PHBA and 3,4-DHB had no significant correlation with LPS. Although the growth and metabolism of many microorganisms were inhibited by PHBA [76], there were no specific difference in antibacterial ability against G + and G− [77]. Similarly, 3,4-DHB possessed antibacterial properties towards both G + and G− [78], so it could not influence the composition of gut microbiota.
There have been examples of other antibacterial agents that altered gut microbiota and affected endotoxin levels. It was reported that oral administration of vancomycin, a widely used G + antibiotic, resulted in dysbiosis of gut microbiota, and the absolute number of G + decreased [79, 80], while the proportion of G − increased to compensate [81, 82]. At the same time, fasting plasma LPS [83] drastically increased. Antimicrobial agent triclosan greatly disturbed the homeostasis of gut microbiota, resulting in the overproduction of LPS, and significantly increased intestinal permeability, thus promoting LPS translocation [84].
Based on a literature review, we hypothesized for the first time in a population study that IE (LPS) was associated with exposure levels of parent PBs and was detectable in semen. The above results showed that the LPS level in semen was negatively correlated with sperm concentration and total sperm count. Therefore, it is speculated that EtP and BuP affect sperm quality possibly by promoting an LPS increase through their antibacterial activity.
Limitations
This study has some limitations. First, dietary patterns and nutritious factors have potential effects on blood LPS levels [85]. To reduce the complexity of the research, dietary factors were not investigated. Previous studies showed exposure to exogenous chemicals could change the composition of gut microbiota and increase the level of LPS in blood [38]. Based on the evidence, this research did not analyze gut microbiota. Second, previous studies have shown that IE can pass through the blood-testosterone barrier and cause sperm damage, but population studies on the effects of LPS in blood on human reproduction are still limited, so our first step was to measure LPS in the blood. In the future, LPS should be measured in seminal plasma, and the difference in LPS levels between seminal plasma and blood as well as their toxicity characteristics should be explored so as to provide direct evidence for the effects of IE on sperm parameters. Finally, due to the limits of cross-sectional design, our results cannot conclude causal relationships between IE and sperm quality, between PB exposure and IE, or the definite effects of PB exposure on sperm quality. These need to be further researched in combination with animal or in vitro experiments to confirm the causal relationships.