Antidepressants block cloned fish serotonin transporters at levels already measured in polluted water

Fish kept two serotonin transporter genes where mammals kept one, and Shinichi Miyagawa's group at Tokyo University of Science cloned both, plus DAT and NET, from medaka (Oryzias latipes) and ayu (Plecoglossus altivelis) ^[2]. The split ran the same way in both species: SERTa much more sensitive to the drugs, SERTb much less ^[4]. The sequence differences predicted which paralogue the drugs would hit hardest ^[5], so a screening lab handed a new fish genome would know which paralogue to clone first.

Duloxetine, fluoxetine, citalopram and paroxetine inhibited medaka SERTa somewhere between a few hundred and roughly 1,300 nanograms per litre ^[8]. A litre of water is a kilogram, which is a trillion nanograms, so the upper end of that range works out to about 1.3 parts per billion ^[17]. Wherever the tenfold gap with the human transporter holds, a guideline value read across from human potency would sit at least an order of magnitude above the fish number ^[18].

"By demonstrating that key molecular targets in fish can be more sensitive than their human equivalents, our work offers crucial insights into the potential risks of pharmaceutical exposure to aquatic wildlife," Miyagawa said ^[10]. The paper ran in Environmental Science & Technology on Aug. 24, co-authored by Masaru Ihara of Kochi University ^[1]. Miyagawa also said the work "provides a vital scientific basis for prioritizing specific pharmaceuticals in environmental monitoring programs and for deriving more protective, species-specific risk thresholds in water quality guidelines" ^[11].

The comparison with polluted water puts two numbers in the same units that are not the same quantity. One is the concentration in a dish around a cell expressing a fish protein; the other is what an analytical lab found in a waterway. The authors are explicit that in vivo studies at environmentally realistic exposures, and with pharmaceutical mixtures, are needed before these molecular effects can be tied to biological or ecological outcomes ^[15]. Behaviour in an exposed fish, meanwhile, does not identify the protein the drug acted on ^[12].

Medaka and ayu are only distantly related, and the team takes the shared pattern as a sign that the heightened sensitivity extends past one species ^[16]. Most mechanistic work of this kind has used a small number of fish, so whether the sensitivity is broadly shared is unsettled ^[14]. Two species is a real improvement on one, and I would want several more lineages in the panel before treating the pattern as general to fish.