Skip to main content

Table 2 Summary weight-of-evidence matrix for retinoid signalling perturbation by environmental contaminants

From: Highlighting the gaps in hazard and risk assessment of unregulated Endocrine Active Substances in surface waters: retinoids as a European case study

Associated condition Line of evidence, uncertainties, and limitations Biological relevance (B), strength (S) of the study, and correlation of data (Corr.) Reliability of the source and “test system” with respect to human health Conclusion: incl. identification of data gaps
Developmental effects (teratogenicity, malformations) Various (see Table 1); strongest associations are supported by in vivo loss-of-function and developmental toxicity/teratogenicity studies in, amongst others, mammals, fish and amphibians (also recently reviewed in [103] B: ++, S: ++ Very high  
Affective disorders RAS contributes to the development of: mood changes, aggressive behaviour, depression, and/or suicidal thoughts [177, 222, 223, 232, 234] B: +, S: + Very high (confirmed independently multiple times in human and rodent studies)  
  Mood changes, aggressive behaviour, depression, and/or suicidal thoughts [222] (Basis for decision to include it on the label is not provided.) B: -, S: ++, Corr.: ?/↑ Very high (official note for isotretinoin treatment); human Though a comment on the association of 13cRA with affective disorders is included on medication labelling, the causal link is not definitively established yet
  Depression and suicidal behaviour [223]
Systematic review; many of the sources are reported to be of limited use to the review, esp. with respect to suicidal thoughts
B: +. S: +, Corr.: 0/↓ High (systematic review); human A trend, but no statistically significant correlation between depression/suicidal thoughts and 13cRA treatment, but epidemiologic data are insufficient to draw associations esp. with respect to suicidal thoughts
  Depression [187, 234]
Very thorough and broad studies, covering longer-term (28 days) and chronic (> 100 days) exposure in adult/elderly rats; doses of 13cRA and atRA reflect serum levels in human during 13cRA treatment. No histopathological examination was conducted
B: +, S: ++, Corr.: 0 Moderate (original research); rat The data do not substantiate the hypothesis of 13cRA inducing depression. No effects were observed on behaviour, nor on monoamine levels in the brain. Observations differ from other studies perhaps due to interspecies variation
Autism, schizophrenia, ADHD, depression RA modulates the correct distribution of GABAergic interneurons in the prefrontal cortex via corticothalamic interaction [248]
Focus on CYP26B1 (co-)detection with parvalbumin-positive GABAergic inter-neurons in the prefrontal cortex in mice. The study concludes on the role of thalamic signalling (controlling CYP26B1 expression) during interneuron maturation, not on RA signalling itself
B: 0, S: +, Corr.: ↑ Weak/moderate (original research); mouse The study satisfactory links altered corticothalamic influence on CYP26B1 expression in the prefrontal cortex, together with subsequently altered RA signalling, however, the study only indirectly assessed RA signalling
Schizophrenia CYP26B1 mutation in humans is a risk factor for schizophrenia [240, 247]
Genome-wide association study (GWAS) including 36,989 cases of schizophrenia and 113,075 controls, and a multi-laboratory combined cohort of 153 patients vs. 153 controls examined for gene expression changes. While both studies identified CYP26B1 as a potential schizophrenia risk factor, other enzymes in the RAS cascade/pathway (e.g. RALDH, RARs, RXRs) were not identified
B: +, S: ++, Corr: 0/↑ High (genome-wide association study, large cohort); human Identification of 108 conservatively defined loci of genome-wide significance towards schizophrenia, including CYP26B1 (rank 74). While GWAS is a powerful tool to uncover rare mutations and genetic risk factors, the causal link needs to be confirmed in mechanistic studies
  RAS is disrupted in schizophrenia [497]
GWAS on Australian Schizophrenia Research Bank cohort (425 schizophrenia cases and 251 controls included in study). The polygenic risk score of 22 retinoid genes was significantly associated with the disorder. In addition, a rare variation in the gene encoding RARβ was associated with severe cognitive deficits
B: +, S: ++, Corr: ↑ High (genome-wide association study, large cohort); human The study strengthens the link between neurological disease, particularly schizophrenia with cognitive deficits, and disturbed (decreased) retinoid signalling
This study paves the way for preventive or therapeutic use of retinoids in schizophrenia, though responsiveness of cognitive symptoms needs to be investigated individually first
Learning disabilities Prolonged post-natal vitamin A deficiency decreases RAS and RAR expression in the brain [153]
Vitamin A deficiency simulated age-related decline of RAS in mice (C57BL6 Jico). Unlike in age-related decline in function, RA treatment (150 µg/kg) could not rescue the effects to control levels
The results need to be taken with caution; the trends observed in treated and control mice are very similar
B: +, S: 0, Corr.: ↑ Moderate (original research); mouse Vitamin A deprivation during adolescence is associated with a sustained decrease in learning/memory function and, which cannot be rescued by RA treatment
These results need further investigation and independent reproduction
  Decreased retinoid signaling impairs locomotion [226]
Open field behaviour of morphologically and histologically normal mutant mice (RXRβ–RXRγ double null or single mutant and WT) was observed. Double null mutants showed impaired locomotion. Additionally, double mutants (but not single-mutants) showed a decreased striatal expression of dopamine receptors, which are involved in voluntary movement
Though the authors are addressing mechanistic endpoints (retinoid and dopamine receptors), the mechanistic investigation is insufficient
B: +, S: +, Corr.: ↑ Moderate/high (original research); mouse This study is amongst the first to link retinoid signalling and cognitive function mechanistically.
Follow-up studies based on these initial findings have been conducted, though the link is still not understood entirely
  RARα modulates synaptic plasticity and affects hippocampal learning [174]
In mice (homozygous RARα floxed mice, C57BL/6 background. 10 days environmental enhancement in young adults before assessment at PND 60-70), deletion of RARα in hippocampal circuits blocks homeostatic synaptic plasticity and enhances long-term potentiation via/together with mTOR. The Morris water maze was used to assess memory/learning capacity (RARα KO mice performed better than WT) and for reversal learning/memory modification testing (RARα KO mice performed worse than WT)
The results are the first to look into stable (i.e. not chronically suppressed) synaptic plasticity, which makes comparisons to other studies difficult
B: ++, S: + , Corr.: ↑ High (original research); mouse Hippocampal RARα deletion enhances spatial learning but decreases learning flexibility. In this study, RARα alters AMPA receptor expression indirectly by an intermediate pathway rather than having direct transcriptional/translations effects
The conflicting observation between initial learning and memory modification, and its implications for cognitive performance has to be investigated further
This publication may contribute to teasing out the reasons for apparently conflicting findings regarding retinoid signalling in cognitive functions
Dementia/neurodegenerative diseases Modulation of RAS contributes to Parkinson’s disease (PD) by modulating the dopaminergic system. [498]
No statistically significant correlation between vitamin A/carotene blood levels and PD was identified in the pooled dataset; significant associations detected only in case-studies. Out of 362 potentially relevant papers, the sample size included in the meta-analysis was low (n = 8 studies)
B: ++, S: -, Corr.: 0 Very high (23-y. epidemiology-based meta-analysis) Epidemiological data on the association of PD with vitamin A/carotenoid blood levels are insufficient to draw conclusions; recommended reporting schemes for epidemiological data need to be followed
  Age-related memory deficit in mice (C57BL6 Jico) due to decreased RAS associated with decreased RAR expression [153, 203]
Observations are clearly linked to RAS and RAR activity due to the experimental design
B: +, S: ++, Corr.: ↑ Moderate (original research); mouse Age-related, but not postnatal vitamin A deficiency-induced, decline in RAS and learning/memory could be restored to normal adult levels by subcutaneous injection of 150 µg RA/kg bw.
These results need independent reproduction
  1. AMPA: α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid, KO: knock out, mTORC1: mammalian target of rapamycin, PD: Parkinson’s disease, PND: post-natal day, RA: retinoic acid, RAR: retinoic acid receptor, RAS: Retinoic acid signalling/retinoid signalling, RXR: retinoid X receptor, WT: wild type. +: relevant/strong/reliable line of evidence, 0 neutral, - not relevant/strong/reliable line of evidence. “Biological relevance” is interpreted as: “the working hypothesis to be tested is scientifically sound”, “strength” is interpreted as: “the tools to investigate the hypothesis are fit-for purpose”. Arrows indicate if the results of the study support the hypothesis (positive correlation, ↑), contradict the hypothesis (negative correlation, ↓), are neutral (no correlation/effect, 0), a trend was observed but no statistical significance (0/↓; 0/↑), or are still under development (?)