Reduction, Plasma vitellogenin concentrations leads to Reduction, Cumulative fecundity and spawning

Biological plausibility for this KER is centered around vitellogenin as a hepatically derived protein in maternal circulation that is sequestered by vitellogenic oocytes during teleost vitellogenesis, where it accounts for the greatest proportion of oocyte growth and yields yolk protein derivatives needed for early embryonic development (Tyler & Lancaster, 1993). In teleosts, this uptake from the maternal circulation is selective and receptor-mediated; Tyler and Lancaster (1993) identified specific vitellogenin receptors in rainbow trout follicle membranes, providing a mechanistic basis for linking circulating vitellogenin availability to oocyte growth. Consistent with this mechanism, Li et al. (2011) modeled oocyte growth after recruitment into the growth phase as dependent on vitellogenin uptake, with uptake rate proportional to plasma vitellogenin concentration and oocyte volume. In that framework, reduced plasma vitellogenin would be expected to reduce vitellogenin uptake by growing oocytes, slow their growth, and delay or prevent attainment of the size threshold required for spawning (Li et al., 2011). This provides a mechanistic basis for expecting reduced plasma vitellogenin to decrease eggs per spawn, disrupt spawning dynamics, and reduce cumulative fecundity (Li et al., 2011; Watanabe et al., 2016). Additional teleost support comes from Kishida and Specker (1993), who quantified vitellogenin in plasma and characterized it in oocyte extract in tilapia. Watanabe et al. (2016) further concluded that oocyte growth and development are driven by absorption of vitellogenin from plasma, reinforcing the mechanistic linkage between reduced circulating vitellogenin and reduced reproductive output.

Direct empirical support for this KER comes primarily from short-term reproduction studies in adult female fathead minnows. In the 21-day fadrozole assay, mean fecundity declined from 20.5 eggs/female/day in controls to 8.9, 1.5, and 1.2 eggs/female/day at 2, 10, and 50 μg/L, respectively, while plasma vitellogenin also decreased in a concentration-dependent manner (Ankley et al., 2002). At 10 and 50 μg/L, reduced fecundity was almost solely because the fish stopped spawning within 2 days of exposure; at 2 μg/L, decreased fecundity reflected both a slight reduction in number of spawns and about a 50% decrease in eggs per spawn (Ankley et al., 2002). This study therefore provides direct evidence that reduced plasma vitellogenin was associated with both reduced cumulative fecundity and reduced spawning occurrence in reproductively active female fathead minnows (Ankley et al., 2002).

Cross-chemical evidence indicates that this relationship is not unique to fadrozole. Using data from 21-day reproduction studies with 17β-trenbolone, 17α-trenbolone, prochloraz, fenarimol, and fadrozole, Miller et al. (2007) found a highly significant linear relationship between fecundity and plasma vitellogenin concentration in female fathead minnows. This analysis shows that reduced plasma vitellogenin was consistently associated with reduced fecundity across chemicals with different upstream modes of action that converge on reduced plasma vitellogenin (Miller et al., 2007).

Additional quantitative support comes from oocyte growth dynamics model evaluation against empirical fathead minnow data. Watanabe et al. (2016) reported that the model robustly predicted the fecundity of control fish and the impact of five steroidogenesis inhibitors, fadrozole, propiconazole, prochloraz, fenarimol, and trilostane, on reproduction metrics using plasma vitellogenin as the key input variable. The same study found weaker prediction for ketoconazole, with average fecundity generally overpredicted, and attributed this discrepancy to compensatory changes in plasma vitellogenin during exposure that were not represented in the simulation (Watanabe et al., 2016). Together, these studies support a strong empirical association, and moderate quantitative understanding, between reduced plasma vitellogenin and reduced reproductive output in female fathead minnows.

Most direct quantitative evidence for this KER derives from fathead minnow assays. Extrapolation across fish taxa should be made cautiously because reproductive strategy can influence how reduced plasma vitellogenin is expressed in downstream reproductive metrics. Li et al. (2011) noted that oviparous fish include both total-spawning and batch-spawning species, and their oocyte growth dynamics model for fathead minnows used clutch size and spawning interval distributions specific to a batch-spawning teleost (Li et al., 2011). The same authors further noted that adaptation of the model to other batch- or total-spawning fishes would require modification of fish-specific parameters, including oocyte recruitment size, oocyte recruitment interval, and critical size for spawning (Li et al., 2011). Evidence is therefore strongest for batch-spawning teleosts, especially fathead minnow, and weaker for broader extrapolation to all fish taxa.

Second, not all chemicals produce identical quantitative relationships. Watanabe et al. (2016) found that the vitellogenin-based model robustly predicted fecundity for several steroidogenesis inhibitors but overpredicted mean fecundity for ketoconazole-exposed fish, likely because compensatory responses increased plasma vitellogenin during prolonged exposure and were not represented in the model. This suggests that time-varying endocrine compensation can weaken prediction of fecundity from a single terminal plasma vitellogenin measurement (Ankley et al., 2007; Watanabe et al., 2016).

Third, the evidence is stronger for reduced cumulative fecundity than for reduced spawning per se when considered across the full dataset. Ankley et al. (2002) directly documented spawning cessation at higher fadrozole concentrations and also distinguished effects on number of spawns and eggs per spawn at the lowest concentration. However, across many 21-day studies and their integrative analyses, fecundity is more consistently represented as total eggs over time, such as eggs/female/day or cumulative fecundity, than as separately quantified spawning frequency and eggs per spawn, particularly in group spawning designs where individual female contributions cannot always be resolved (Ankley et al., 2002; Watanabe et al., 2016). Consequently, the linkage to reduced spawning is well supported, but less uniformly quantified across the full dataset than the linkage to reduced cumulative fecundity.

Fourth, plasma vitellogenin is often measured at the end of a 21-day assay, which can support correlation and prediction but may not capture time-varying endocrine dynamics during exposure. This is particularly important when compensatory responses alter plasma vitellogenin over time or when vitellogenin protein persists after the exposure pattern has changed (Watanabe et al., 2016; Craft et al., 2004).

Response-response data for this KER are available from both direct empirical studies and model evaluation. Miller et al. (2007) reported a highly significant linear relationship between plasma vitellogenin concentration and fecundity in female fathead minnows across five 21-day reproduction studies, with the model fecundity = -0.042 + 0.95 × VTG, r = 0.94, and r2 = 0.88 (Miller et al., 2007). This provides the clearest quantitative description of how decreasing plasma vitellogenin is associated with decreasing cumulative fecundity.

Data from the fadrozole study show a similar graded relationship. Median plasma vitellogenin declined from 15.1 mg/mL in controls to 7.6, 1.0, and 0.0 mg/mL at 2, 10, and 50 μg/L, while fecundity declined from 20.5 to 8.9, 0.0, and 0.0 eggs/female/day, respectively (Miller et al., 2007). In the corresponding primary study, reduced fecundity at 2 μg/L reflected both fewer eggs per spawn and a slight reduction in number of spawns, whereas at 10 and 50 μg/L the fish stopped spawning within 2 days of exposure (Ankley et al., 2002). These data indicate that the downstream response can be expressed as graded reduction in cumulative fecundity at moderate upstream change and loss of spawning at more severe upstream change.

Watanabe et al. (2016) further evaluated this relationship using an oocyte growth dynamics model parameterized with plasma vitellogenin as the primary input variable. The model robustly predicted fecundity for control fish and for fish exposed to fadrozole, propiconazole, prochloraz, fenarimol, and trilostane, supporting the quantitative linkage between measured plasma vitellogenin and downstream reproductive output (Watanabe et al., 2016). Prediction was weaker for ketoconazole, indicating that time-varying compensation can modify the apparent response-response relationship when only terminal plasma vitellogenin values are used (Watanabe et al., 2016).

Available evidence indicates that downstream changes in cumulative fecundity and spawning can occur within days in strongly affected female fathead minnows under conditions associated with marked reduction in plasma vitellogenin.

In the fadrozole study, fish in the 10 and 50 μg/L treatments almost stopped spawning within 2 days of exposure, indicating that the downstream reproductive effect can emerge very quickly when vitellogenin is strongly suppressed (Ankley et al., 2002). At the lower concentration, the effect manifested over the assay as both fewer spawns and fewer eggs per spawn (Ankley et al., 2002).

The mechanistic time-scale is biologically consistent with the need for continuously available circulating vitellogenin to support growth of maturing oocyte cohorts. Li et al. and Watanabe et al. model oocyte growth as a continuous vitellogenin-dependent process that determines when oocytes reach spawning threshold size, which provides a basis for understanding why sustained vitellogenin reduction can alter both spawning interval and total egg output over a 21-day test (Li et al., 2011; Watanabe et al., 2016).

A remaining uncertainty is that plasma vitellogenin is often measured terminally rather than serially. Thus, empirical precision on lag time between vitellogenin decline and fecundity decline is limited in many assays, even though temporal concordance is supported overall (Craft et al., 2004; Watanabe et al., 2016).

No well-characterized feedforward or feedback loop has been identified at the level of vitellogenin uptake into oocytes for this KER. The principal feedback influencing this relationship appears to occur upstream in the endocrine axis, where compensatory responses to steroidogenesis inhibition can partially offset expected reductions in plasma vitellogenin and thereby alter the apparent strength of the linkage to fecundity (Ankley et al., 2007; Watanabe et al., 2016). Watanabe et al. (2016) specifically concluded that overprediction of fecundity for ketoconazole likely reflected compensatory changes in plasma vitellogenin that were not represented in the model