A4.4 Developmental and reproductive effects of glyphosate-based herbicide in amphibians and birds
Paganelli et al (2010) performed studies on neural crest development in three experimental systems:
(i) Xenopus laevis embryos, which were exposed from the 2-cell stage onwards to Roundup Classic (a Monsanto product containing 48% w/v of an unspecified glyphosate salt; no other constituents were identified) at 3000-, 4000- and 5000-fold dilutions in their incubation medium. The final concentrations of glyphosate were 717, 536 and 430 µM at the respective dilutions. Neurula stage embryos were fixed and examined at the by immunofluorescence following in situ hybridisation with antisense RNA probes. Retinoic acid (RA) activity was measured by chemiluminescence in neurula-stage embryos that had been injected with RAREZ reporter plasmid prior to Roundup exposure as described. For rescue experiments RAREZ-injected embryos were incubated with Roundup at 4000-fold dilution until the blastula stage, then exposed to the RA receptor antagonist Ro 41-5253 at 1.0 µM until assay of RA activity.
(ii) Two-cell Xenopus embryos were injected with 360 or 500 pg of glyphosate into one or both cells (producing intracellular concentrations of 8 – 12 µM) together with 10 ng of the visual marker Dextran Oregon Green. They were then incubated until sibling controls had reached the desired developmental stage, fixed, and examined visually or by immunofluorescence following in situ hybridisation as described above.
(iii) Fertilised chicken eggs were injected with 20 µL of 3500- or 4500-fold dilutions of Roundup Classic and incubated at 38 ̊C until fixation, in situ hybridisation and immunofluorescence examination as described for Xenopus embryos. Control embryos were treated similarly after injection of 20 µL of water.
Effects on neural crest markers, rhombomere formation and primary neuron differentiation: Compared with sibling controls, Roundup Classic at 5000-fold dilution impaired neural crest formation in 87% of Xenopus embryos (n = 30), seen as down-regulation of the neural crest marker slug and zinc finger transcription factor krox-20 in the r3 rhombomere. Neuron formation was suppressed, as evidenced by decreased numbers of primary motor, inter- and sensory neurons in 83% of treated embryos. Similar effects occurred in 70 – 80% of embryos injected unilaterally with 500 pg glyphosate. On their injected side these displayed abolition of slug expression, reduced krox-20 expression in r3 and r5, and decreased numbers of primary motor, inter- and sensory neurons. The study authors considered the Roundup-exposed and glyphosate injected embryos to be equivalent (although not identical) phenotypes. They did not present any results obtained at the 360 pg/cell dose or the 1/4000 or 1/3000 dilutions.
Effects on the development of the head and dorsal midline: In 85% of 1:5000rd Roundup-exposed neurula-stage Xenopus embryos, there was reduced expression of shh (a gene whose expression is responsible for resolving the brain and retina into two separate hemispheres) and pax6 (responsible for eye formation). After incubation was prolonged to the tailbud stage, ca 90% of treated embryos displayed a decrease in anterior shh expression with concomitant microphthalmia, microcephaly, shortening of the anterior-posterior (A-P) axis and delayed migration of neural crest cells into the eyes, genital ridges and pharyngeal arches. Bilateral injection of 360 pg glyphosate also reduced shh expression and induced microphthalmia and microcephaly in the majority of treated embryos. In older (tadpole stage) embryos, Roundup exposure caused microphthalmia and a generalised reduction of cranial cartilage structures; most unilaterally-injected embryos showed these effects on the treated side, while bilateral injection caused cyclopia in 3/8 embryos. The study authors did not provide any data obtained at the 500 pg/cell dose or the 1/4000 or 1/3000 dilutions.
Effects on retinoic acid signalling: A highly significant (p<0.0001) dose-dependent increase in RA signalling activity occurred in Roundup-exposed Xenopus embryos at 4000- and 3000-fold dilutions. The magnitude of the effect was intermediate between the activity seen after addition of exogenous RA at 0.50 and 5.0 µM. However, there was no apparent response to Roundup at 1:5000, which Paganelli et al attribute to a lack of sensitivity of the RAREZ reporter plasmid. Assuming a linear response of the luminescence system, the study authors estimated that the endogenous concentration of RA in Xenopus embryos is ca 0.2 µM. The RA receptor antagonist Ro 41-5253 blocked the signalling increase mediated by 1:4000 Roundup, and prevented 1:5000 Roundup from inhibiting shh activity and causing microcephaly. No data were presented on the influence of Ro on RA signalling or embryo phenotype at other dilutions.
Effects in chicken embryos: Roundup caused concentration-dependent reduction in pax6 expression and in the size of the optic vesicles, loss of the r3 and r5 domains and decrease in shh expression in midline cells, accompanied by microcephaly and loss of shh expression in the pre-chordal mesoderm.
The study authors suggest that the similarity between the phenotypes observed in Roundup-incubated and glyphosate-injected Xenopus embryos indicates that neural crest development is disrupted by the active constituent, rather than adjuvants present in the formulated product. Noting (a) similarities between the effects of Roundup and glyphosate with those of excess retinoic acid (RA) concentrations in Xenopus, mice and humans; (b) increased RA signalling levels in Xenopus embryos in response to Roundup; and (c) the effectiveness of the anti-retinoid Ro in preventing the developmental effects of Roundup in Xenopus, Paganelli et al hypothesise that glyphosate is a developmental toxin with a mode of action involving enhancement of RA signalling activity.
Given their belief that (d) glyphosate inhibits aromatase, a cytochrome P450 enzyme; and (e) retinoid activity is regulated by degradation of RA by CYP26, the study authors further hypothesise that glyphosate increases RA signalling by inhibiting the activity of CYP26 responsible for maintaining normal RA distribution by specific territorial degradation.
Williams et al (2012) have noted that in this study
The glyphosate solution was not pH-adjusted, and so the effects may be attributable to its acidic nature;
The injection route of exposure was inappropriate and irrelevant to risk assessment; and
The observations require further substantiation using appropriate methods before consideration in risk assessment.
Oliviera et al (2007): Adult drakes in breeding season (6/group) were gavaged with Roundup (360 g/L glyphosate, present as 480 g/L glyphosate isopropylamine salt; no other formulation constituents identified; Monsanto do Brasil Ltda, Sao Paulo, Brazil) in water at 5.0 or 100 mg/kg bw/d for 15 days. The study authors did not specify whether the dose levels applied to the active constituent, or the product. A control group received water only.
After the treatment period, the birds were anaesthetised and perfused intracardially with 2.5% gluteraldehyde. Fixed testes and epididymides (5/group) were then weighed, examined morphometrically and examined histochemically to investigate lysosomes and lipids within the epididymal region. Androgen receptor (AR) expression was studied by immunohistochemistry, with confirmation of antibody specificity by SDS-PAGE / Western blotting. Plasma testosterone and oestradiol concentrations were measured in three birds/group by RIA.
Body and organ weights: There was no treatment-related effect on bodyweight. Relative testicular weights were depressed by ca 13% at both doses, but the difference from control was not statistically significant. Data on absolute testis weight were not presented.
Hormones: Plasma testosterone levels were reduced by ca 90% at both doses (p<0.05). A significant (p<0.05) ca 30% decline in plasma oestradiol occurred at 5.0 mg/kg, but there was no such effect at 100 mg/kg.
Tissue histology: Within the testis, Roundup at 5.0 and 100 mg/kg respectively induced slight but statistically significant (p<0.05 vs control) reductions in the volumetric proportion of seminiferous tubule epithelium (by 4 and 5%) and interstitial tissue (by 12 and 10%), together with 20 and 22% increases in the lumen volume (p<0.05). Spermatogenesis appeared to be normal, however.
Within the epididymal region, there were dose-related trends towards reduced volumetric proportions of proximal efferent ductules and connecting duct, together with increases in the proportion of rete testis, distal efferent ductules and connective tissue. These features attained statistical significance (p<0.05) at 100 mg/kg but not the low dose.
In the proximal efferent ductules of treated birds, qualitative morphological alterations (increased epithelial lipid content and epithelial vacuolisation caused by increased numbers of lysosomes) were found, together with increases of 11 and 7% in epithelial height and 41 and 105% in lysosomal area at 5.0 and 100 mg/kg respectively (all p<0.05 vs control).
The morphology of the epididymal duct was also affected. Birds receiving 5 and 100 mg/kg, respectively, displayed significant (p<0.05) reductions of 28 and 49% in tubular diameter and increases of 23 and 34% in epithelial height. The epididymal ducts of treated birds presented collapsed and sometimes highly folded lumen, together with an increase in the basement membrane. By contrast, control birds presented wider and regular lumen and a slight basement membrane.
AR expression: At both doses, Roundup caused a major (but unquantified) decrease in AR expression within the Sertoli cell nuclei within the testis. However, the effect did not occur within the epididymal region. The specificity of the AR antibody used was confirmed.
Comment: This study is notable for the low numbers of birds used (especially for hormonal assay); the non-dose related depression of oestradiol concentration; and the lack of an experimental group treated with glyphosate alone, which prevented identification of the formulation constituent(s) causing the reported effects. The observed responses to treatment may have been associated with generalised physiological stress, rather than a specific effect on steroid hormone synthesis.