Historical exposure to chemicals reduces tolerance to novel chemical stress in Daphnia (waterflea)

Abstract Until the last few decades, anthropogenic chemicals used in most production processes have not been comprehensively assessed for their risk and impact on wildlife and humans. They are transported globally and usually end up in the environment as unintentional pollutants, causing long‐term adverse effects. Modern toxicology practices typically use acute toxicity tests of unrealistic concentrations of chemicals to determine their safe use, missing pathological effects arising from long‐term exposures to environmentally relevant concentrations. Here, we study the transgenerational effect of environmentally relevant concentrations of five chemicals on the priority list of international regulatory frameworks on the keystone species Daphnia magna. We expose Daphnia genotypes resurrected from the sedimentary archive of a lake with a known history of chemical pollution to the five chemicals to understand how historical exposure to chemicals influences adaptive responses to novel chemical stress. We measure within‐ and transgenerational plasticity in fitness‐linked life history traits following exposure of “experienced” and “naive” genotypes to novel chemical stress. As the revived Daphnia originate from the same genetic pool sampled at different times in the past, we are able to quantify the long‐term evolutionary impact of chemical pollution by studying genome‐wide diversity and identifying functional pathways affected by historical chemical stress. Our results suggest that historical exposure to chemical stress causes reduced genome‐wide diversity, leading to lower cross‐generational tolerance to novel chemical stress. Lower tolerance is underpinned by reduced gene diversity at detoxification, catabolism and endocrine genes in experienced genotypes. We show that these genes sit within pathways that are conserved and potential chemical targets in other species, including humans.

| 3099 ABDULLAHI et AL. exposure scenarios in situ (Kanno, 2016). These approaches also do not account for the cumulative toxicity that may arise from long-term exposures to sublethal concentrations of a chemical or chemical mixtures (Blair et al., 2015). Long-term exposure to environmentally relevant concentrations of anthropogenic chemicals has been shown to cause loss of genetic diversity, with consequences for adaptive potential to novel stress (Bijlsma & Loeschcke, 2012;Fasola et al., 2015;Ribeiro et al., 2012). Chemical stress has documented adverse effects on wildlife's ecological endpoints, including developmental defects and survival (Blahova et al., 2020;Jantzen et al., 2016), behaviour and metabolism (Xia et al., 2015), delayed growth and metamorphosis (Yoon et al., 2019), embryonic development (Balbi et al., 2018), fecundity and sexual maturity (Liu et al., 2017).
In an effort to understand the impact of long-term chemical stress on Daphnia fitness and susceptibility to novel chemical stress, here we study the transgenerational effects of five chemicals on the priority list of international regulatory frameworks because of their widespread presence in the environment and their potential adverse effects (Chang et al., 2016;Gosset et al., 2020;Pu et al., 2020). The chemicals used are a flame retardant (perfluorooctanesulfonic acid [PFOS]), a commonly used anti-inflammatory drug (diclofenac), the antibiotic trimethoprim, the herbicide atrazine and a heavy metal (arsenic). We use concentrations that are environmentally relevant based on literature research, opting for mid-to high environmental concentrations found either in high-income or developing countries (Graziano et al., 2006;Le Luu, 2019).
Daphnia plays a central role in freshwater foodwebs worldwide (Altshuler et al., 2011). It has a parthenogenetic life cycle, in which sexual and asexual reproduction alternate (Ebert, 2005). Sexual recombination results in early-stage embryos that arrest their development and enter dormancy, which can be exceptionally long (up to centuries; Kerfoot & Weider, 2004). Dormant embryos that are awakened (resurrected) from dormant stages are genetically distinct and can be propagated in the laboratory via clonal reproduction, allowing the rearing of populations of isogenic individuals (clones) from a single genotype . Daphnia magna strains were resurrected from the sedimentary archive of Lake Ring (Denmark), having a well-documented history of exposure to chemicals and other stressors Davidson et al., 2007). By exposing strains that have never been exposed to chemical stress (naïve) and strains that have been historically exposed to chemical stress (experienced) to the five chemicals mentioned above, we were able to study how historical exposure to safe doses of chemicals in the natural environment can influence susceptibility to novel chemical stress. We tested the hypothesis that "experienced genotypes" were less susceptible to novel stress by showing higher fitness than "naive genotypes" when exposed to novel chemical stress. We expected experienced genotypes to have higher detoxification abilities underpinned by enriched detoxification genes or pathways, suggesting evolution of tolerance to chemical stress.
To test these hypotheses, we quantified fitness responses of two naïve and two experienced genotypes across three clonal generations to the five chemicals in common garden experiments.
To link the observed changes in fitness to adaptive responses to chemicals, we quantified genome-wide nucleotide diversity in the four genotypes and identified enriched functional pathways diverging between experienced and naïve genotypes. To investigate the relevance of our findings to other species, we used a comparative functional analysis of protein domains to identify enriched pathways in Daphnia that were conserved in other species and that are potential chemical targets. This study thus enabled us to investigate how evolutionary responses to novel chemical stress is potentially influenced by historical exposure to chemicals. We also gained insights into how the interactions between historical and novel chemical stress may influence transgenerational adaptive responses.

| Study system and experimental design
Genotypes of Daphnia magna were previously resurrected (revived) from a biological archive of Lake Ring, a shallow mixed lake in Denmark (55°57′51.83″N, 9°35′46.87″E) with a well-documented history of anthropogenic impact (Cuenca-Cambronero, Marshall, et al., 2018;Davidson et al., 2007). In the early 1900s and until late 1940s the lake was semipristine. In the late 1950s, it experienced eutrophication due to sewage inflow from a nearby town, which was diverted in the late 1970s; from the 1980s until the late 1990s, the lake experienced an increase in biocide run-off due to agricultural land use intensification; the lake partially recovered from high nutrient levels in modern times but still received agricultural run-off from the 1999s onward. According to these records, the lake experienced no chemical exposure until the 1970s, and high chemical exposure from the 1975s onward ( Figure 1). Although the resident population of D. magna was probably exposed to multiple stressors over time, here we focus on tolerance to novel chemical stress and interpret fitness and functional responses to novel chemical stress in the context of historical chemical exposure. In our previous studies of the D. magna population from Lake Ring, we showed microevolutionary responses to recurrent chemical pollutants and that variance at fitness-linked life history traits was larger between than within temporal populations (Cuenca-Cambronero, Marshall, et al., 2018;Cuenca-Cambronero, Marshall, et al., 2018;. For the current study, a total of 360 exposures were completed. We selected a genotype from each of the four lake phases and exposed five clonal replicates per genotype across three clonal generations to six experimental conditions (five chemicals plus control; Figure 1). The genotypes are: LRII36_1 (<1950), LRV12_3 (1960( -1970( ), LRV8.5_3 (1980( -1990 and LRV0_1 (>1999). The former two genotypes are "naïve" and the latter two genotypes are "experienced" to chemical stress ( Figure 1). The concentrations of the five chemicals used are as follows, reflecting environmentally relevant concentrations in surface waters: PFOS (70 ng L −1 ), atrazine (0.2 mg L −1 ), trimethoprim (2 mg L −1 ), diclofenac (2 mg L −1 ) and arsenic (1,000 µg L −1 ). The concentrations of PFOS and its derivatives vary dramatically between geographical areas and proximity to contamination sources (<0.8 ng L −1 to >17 mg L −1 ; Sinclair et al., 2006;Huang et al., 2010) but it is typically found at concentrations of 50-100 ng L −1 in surface waters (Bai & Son, 2021). Nonsteroidal anti-inflammatory drugs such as diclofenac, acetaminophen and ibuprofen can be found at µg L −1 to g L −1 concentrations in seawater (Weigel et al., 2001) and surface waters (Fick et al., 2009), whereas their concentration is significantly lower in ground and drinking waters (ng L −1 ; Godfrey et al., 2007). A concentration between 1 and 2 mg L −1 is common in surface waters in high-income economies (Fick et al., 2009). Average levels of antibiotic drugs in surface water range between ng L −1 and µg L −1 (Kummerer, 2009), with the exception of effluent originating from chemical manufacturers where concentrations of antibiotic drugs can exceed 30 mg L −1 (Larsson et al., 2007) and fish farms where they range from 1 to 6 mg L −1 (Le & Munekage, 2004). Atrazine is one of the most widely used photosynthesis-inhibiting pre-emergent biocides worldwide (Prado et al., 2014), with more than 70 million kg produced yearly (Sass & Colangelo, 2006). Typically, streams and rivers receiving agricultural run-off display concentrations of 0.2 mg L −1 (Graziano et al., 2006).
Metal contamination of groundwater is among the biggest health threats in low-and middle-income countries (Winkel et al., 2011).

| Common garden experiments
We previously resurrected several genotypes of D. magna from Lake Ring and maintained them in the laboratory for over a year as isoclonal lines in the following standard laboratory conditions: 16:8-hr light-dark photoperiod; 0.8 mg L −1 Chlorella vulgaris fed weekly; ambient temperature: 10°C (Cuenca-Cambronero, Marshall, et al., 2018).
Before the exposure experiment for this study, clonal lineages of the four genotypes (LRII36_1, LRV12_3, LRV8.5_3 and LRV0_1) were acclimatized for at least two generations to the following conditions to reduce interference from maternal effects: 16:8-hr light-dark photoperiod; 0.8 mg L −1 Chlorella vulgaris fed daily; ambient temperature: 20°C. After at least two generations in these conditions, 24-hr-old juveniles from the second or following broods were randomly isolated and assigned to experimental conditions ( Figure 1). Each clonal generation was established following the same criteria and starting from randomly selected 24-hr-old juveniles from the second or following broods. Where necessary, different broods from genotypes F I G U R E 1 Experimental design. Four genotypes of Daphnia magna, previously resurrected from dormant embryos, were used for transgenerational exposures to five chemicals: PFOS (70 ng L −1 ), diclofenac (2 mg L −1 ), trimethoprim (2 mg L −1 ), atrazine (0.2 mg L −1 ) and arsenic (1,000 µg L −1 ). The four genotypes were resurrected from different times in the past, and were either "naïve" (black -LRII36_1 [<1950] and blue, LRV12_3 [1960][1961][1962][1963][1964][1965][1966][1967][1968][1969][1970]) or experienced (green, LRV8.5_3 [1980][1981][1982][1983][1984][1985][1986][1987][1988][1989][1990] and red, LRV0_1 [>1999)] to chemicals. The clonal lines from the four genotypes were maintained in common garden conditions for at least two generations before the experiment to control for maternal effect. Five clonal replicates of each genotype, randomly selected from the control environment, were exposed to the five chemicals for three generations (G). Coloured squares represent genotypes at each generation (G). Some genotypes went extinct in G2 and G3 1960-1970 >1999 1980-1990 <1950 Control Environment -no chemicals The following fitness-linked life history traits were measured across three generations of four genotypes and five clonal replicates for exposed and nonexposed Daphnia (120 individuals across three clonal generations): age at maturity (first time parthenogenetic eggs are released in the brood pouch), size at maturity (distance from the head to the base of the tail spine), fecundity (sum of juveniles across two broods), interval between broods and mortality. Size at maturity was measured after the release of the first brood in the brood pouch using imagej software (https://imagej.nih.gov/ij/). The mortality rate per genotype was determined with the survival model fit using the psm function in the "rms" package in R version 3.6.0 (Harrell Jr, 2001). The day of mortality and mortality events were combined as a response variable while the term "genotype" was treated as a fixed effect. The mortality curves per generation were plotted with the survplot function from the rms package in R version 3.6.0 (Harrell Jr, 2001). Genotypes were fixed across all experimental conditions and clonal generations, enabling us to control for confounding factors, such as genetic changes occurring from one generation to the next and genetic variation among experimental exposures. Clonal replicates were nested within genotype in the statistical analyses as explained below. This design permits the analysis of within (WGP) and transgenerational plasticity (TGP), as well as the analysis of evolutionary differences among genotypes that originate from the same genetic pool.

| Fitness response to chemicals
We quantified genetic, WGP and TGP using an analysis of variance (ANOVA), and tested the effect of Generation (G), Genotype (Gen), Treatment (T) and their interaction terms on the five fitness-linked life history traits described above; a Wald chi-square test (Type III test) was used to generate an analysis of deviance table (Langsrud, 2003). Multivariate effects were calculated on the same terms using multivariate statistics (MANOVA). Both statistics were completed using the "car" package for R version 3.6.0 (Fox & Weisberg, 2019) after checking for normality assumptions by plotting model residual vs. fitted values (Q-Q plots) (Zuur et al., 2010). Clonal replicates were fit as a random effect nested within genotype. As the four genotypes used here belong to populations separated in time that originate from the same genetic pool, a significant genotype term indicates genetic evolution of the life history trait (Orsini et al., 2016). Differences in mean trait values between an individual treatment and its control within a generation are the expression of WGP. If genotype means varied by generation, we would have evidence of a G (generation) × Gen (genotype) interaction. We also measured the three-way interaction term, which measures how the treatment per genotype differed across generations (Gen × G × T).
The main effects of genotype, treatment and generation plus their interaction terms on individual life history traits were visualized through univariate reaction norms, which describe the pattern of phenotypic expression of each genotype across treatments and generations (Roff, 1997). We visualized the multivariate analysis results using phenotypic trajectory analysis (PTA) plots to describe the difference in multivariate reaction norms in terms of magnitude and direction of change (Adams & Collyer, 2009). In the PTA, reaction norms are described as multivariate vectors with varying magnitude (the amount of phenotypic change between environments) and direction (the covariation of phenotypic variables) projected onto principal components in a multivariate space (Collyer & Adams, 2007).
The R code provided by Adams and Collyer (2009) was used for the PTA. We visualized the principal mode of variation and covariation among traits (trade-offs) within and across generations through a principal components analysis (PCA) done with the "prcomp" function in R followed by visualization through the fviz_pca_biplot function in the factoextra package in R version 3.6.0 on log-transformed data (Kassambara & Mundt, 2020).

| Genomics and functional analysis
We used clonal replicates of the four genotypes used in the expo- The genome sequences were subjected to quality checking by mapping raw reads onto the newly assembled reference genome of D. magna obtained using a hybrid assembly of long and short readsthe detailed description of this reference genome will be presented elsewhere (NCBI: SUB9530054). Mapping and quality filtered single nucleotide polymorphism (SNP) variants were identified using the following steps: (i) read sequences base pair quality was assessed using fastqc (https://www.bioin forma tics.babra ham.ac.uk/proje cts/ fastq c/) and multiqc (https://multi qc.info/); (ii) trimmomatic version 0.33 was applied for adapter trimming and to remove low-quality sequences (Bolger et al., 2014). Paired-end reads with Q > 30 and read length >50 bp were retained and mapped against the reference genome of D. magna using the BWA-mem algorithm (Li & Durbin, 2010).
(iii) samtools were used for format conversion, sorting, indexing and merging of mapping files onto the reference genome following (Li et al., 2009); (iv) picard tools were used to mark and remove poly- We studied genome-wide and chromosomal-level alpha and beta SNP diversity following methods described in Ma et al. (2020). We quantified gene-level beta diversity between each pair of genotypes using noiseq with no replicates mode (Tarazona et al., 2015). We identified significant differences in SNP diversity per gene between each pair of genotypes using a probability approach (Tarazona et al., 2015). To understand the potential functional impact of the gene di- A significant three-way interaction term (Table 1; MANOVA, G × Gen × T) was observed, showing that treatment per genotype differed across generations in all chemicals, except for PFOS.
ANOVA revealed that these overall patterns were driven by: (i) mortality in the PFOS exposure; (ii) size at maturity, age at maturity, fecundity and mortality in the atrazine exposure; (iii) size and age at maturity in the diclofenac exposure; (iv) size at maturity, fecundity and mortality in the trimethoprim exposure; and (v) fecundity age at maturity in the arsenic exposure (Table 1; ANOVA, G × Gen × T).
Exposure to the chemicals induced a significant genetic response both within (Gen × T) and between (G × Gen) generations.
As the genotypes originate from the same genetic pool sampled at different times in the past, a significant difference in mean trait values among genotypes indicates evolutionary differences. A genotype-dependent response to treatment was observed in all chemicals, except for atrazine (Table 1; MANOVA, Gen × T). The individual fitness-linked life history traits contributing to this response differed among treatments: (i) size at maturity and mortality significantly varied between genotypes in the PFOS treatment; (ii) size at maturity varied by genotype in the diclofenac exposure; (iii) size at maturity and fecundity varied among genotypes in the trimethoprim exposure; and (iv) size and at maturity, and mortality varied among genotypes in the arsenic treatment (Table 1; ANOVA, Gen × T).
A significant genotype per generation effect (G × Gen) was observed across treatments, except for PFOS (Table 1;

TA B L E 1 (Continued)
and (v) all life history traits except mortality for arsenic (Table 1; ANOVA, G × Gen).
Plasticity was pervasive within and across generations.
Significant plastic responses to treatment (T) within generations, indicative of WGP, were observed in PFOS, atrazine and diclofenac (Table 1; Figure S1. An overview of mortality across treatments and generations is presented in Figure S2. Significant TGP was observed in all exposures (Table 1, G × T).
These plastic responses had a significant effect on: (i) size at maturity, age at maturity and fecundity in PFOS exposures; (ii) size at maturity in atrazine exposure; (iii) size at maturity, age at maturity, fecundity and mortality in diclofenac exposures; (iv) age at maturity in trimethoprim exposures; and (v) age at maturity, size at maturity and mortality in arsenic exposures (Table 1;

| Detoxification pathways and genome-wide diversity in naïve and experienced genotypes
The genome of the four genotypes of D. magna was assembled (NCBI: SUB9530054) and the raw depth of coverage was: 84× for LRV0_1; 42× for LRV8.5_3; 49× for LRV12_3; and 45× for LRII36_1.
The genome-wide SNP-alpha diversity was comparable among the genotypes LRII36_1, LRV12_3 and LRV8.5_3, and lower in the most recent genotype (LRV0_1) (Figure 3a; Table S2). The genome-wide alpha diversity patterns were reflected equally across chromosomes ( Figure 3b). Beta diversity was significant in all pairwise comparisons but comparatively higher in pairwise comparisons including the LRV0_1 genotype (Table S2; p =.05).
The number of divergent genes between LRV0_1 and the other genotypes ranged between 1,093 (3.3% of the total number of Daphnia genes) and 1,317 (4%). Conversely, the number of significantly divergent genes in the pairwise comparisons involving the other genotypes ranged between 514 (1.6% of the total number of Daphnia genes) and 697 (2.1%) (Figure 3c; Table S3). The mean gene diversity at the divergent genes was significantly lower in the experienced than in the naïve genotypes (t test; p =.03).
To understand the potential functional impact of the diver-  (Table S4). Genes involved in endocrine processes (GPCR), epigenetic regulation (the Histone deacetylase family and the Elongator complex protein 1) and neuronal activities (learning and memory) were also enriched (Table S4) (Table S4).
We used the manually curated "Reactome" database to assess how many of the functional pathways diverging between naïve and experienced genotypes in our study were conserved across other animals, based on conservation of the protein sequence (Table   S4) (1960( -1970( ) LRV8.5_3 (1980( -1990 (Table S4). Fat and carbohydrate metabolism pathways were differentially enriched between "naive" and "experienced" genotypes (Table S5).

| Naïve genotypes show higher fitness in response to novel chemical stress
We hypothesized that experienced genotypes had an evolutionary advantage over naïve genotypes when exposed to novel chemical stress. We expected the experienced genotypes to always have higher overall fitness underpinned by enrichment at detoxification genes or pathways.
Our results show significant fitness differences among geno- in 60% of the chemicals tested between the second and the third generation. Where mortality did not occur (e.g., trimethoprim and arsenic), the experienced genotypes showed larger fitness changes than the naïve genotypes as the generations progressed. These fitness changes occurred between the second and the third generation of exposure (PFOS, atrazine and diclofenac), indicating cumulative toxicity effects. A complex interplay between genetic adaptation, WGP and TGP has been previously observed in population-level studies of the Daphnia population from Lake Ring, suggesting that the genotypes used in this study are representative of the Daphnia subpopulations from Lake Ring (Cuenca-Cambronero, Marshall, et al., 2018;Cuenca-Cambronero et al., 2021;Toyota et al., 2019).
A population-level analysis will be required to confirm the observed patterns in our study.
Highly plastic traits tend to show strong maternal effect variance and little to no genetic variance, because they are more strongly influenced by the environment, including parental environment, and because additive genetic variance may be masked by high environmental variation (Donelan et al., 2020). In our study, the average fitness of the genotypes increased in the second generation, but it F I G U R E 3 Genomic diversity. Alpha diversity measured at (a) genome-wide and (b) chromosome-level in the four genotypes used in this study; (c) number of significantly divergent genes between each pair of genotypes. The genotypes are colour coded as in Figure 1: LRII36_1 (<1950; black), LRV12_3 (1960LRV12_3 ( -1970blue), LRV8.5_3 (1975blue), LRV8.5_3 ( -1985green) and LRV0_1 (> 1999; red).
F I G U R E 2 Phenotypic trajectory analysis. PTA on the four genotypes of Daphnia magna used in transgenerational exposure to five chemical classes (PFOS [70 ng L −1 ], diclofenac [2 mg L −1 ], trimethoprim [2 mg L −1 ], atrazine [0.2 mg L −1 ] and arsenic [1,000 µg L −1 ]), resulting from multivariate response of five fitness-linked life history traits. Open circles represent the control (nonexposed clonal replicates) and full circles represent the exposed clonal replicates. Genotype centroids are connected by reaction norms (solid lines), showing phenotypic change in direction and length. Differences among genotypes in terms of magnitude (M) and direction (θ) of plastic response are all significant. The statistics supporting the PTA are given in Table S1. Genotypes are colour-coded as in Figure 1 declined again in the third generation, indicating transient positive maternal effects. Transient positive maternal effects have been observed in transgenerational studies of D. magna exposed to gamma radiation (Parisot et al., 2015). Conversely, a persistent positive maternal effect has been observed in transgenerational studies on photoperiod length (Toyota et al., 2019) and endocrine disruptors (Clubbs & Brooks, 2007;Tanaka & Nakanishi, 2002). A positive maternal effect is experienced when the offspring environment perfectly matches the maternal one. In human-altered environments, such as the one linked to chemical run-off, it may be more difficult for the parental generation to detect and correctly identify novel environmental conditions that lack historical context or that increase environmental variability. In these conditions the parents may fail to respond because they lack appropriate cue-response systems (Burgess et al., 31. 2014

| Higher fitness in naïve genotypes is underpinned by higher diversity in detoxification genes
Our findings reject the hypothesis that "experienced genotypes" have an evolutionary advantage in the presence of novel chemical stress. In our seminal study on the Daphnia population from Lake Ring, the experienced genotypes showed a comparatively higher fitness when exposed to the same chemicals recorded in the historical environment, even if this was dependent on the severity of the stress (Cuenca-Cambronero, Marshall, et al., 2018). The limitation of our study is in the small number of clones used, which may not be representative of the local population genetic diversity, therefore providing a qualitative rather than quantitative support to our hypothesis testing. However, if the patterns observed here are validated at the population level, the results of our study and of past studies on the Lake Ring Daphnia population suggest that the acquired tolerance to chemical stress may be evolutionarily advantageous to recurring but not novel chemical stress. It is noteworthy that one of the two naïve genotypes showed the smallest trade-offs and an overall highest fitness (LRII 36_1) when exposed to novel chemical stress.
This may be explained by this genotype being resurrected from a semipristine environment whereas the naïve genotype LRV12.3 was historically exposed to other stressors (e.g., eutrophication). It has been shown that multiple stressors linked to anthropogenic activities can influence how organisms adapt and evolve, with evolutionary mechanisms underpinning multiple stress responses being often synergistic (Cuenca-Cambronero et al., 2021;Jackson et al., 2016;Orr et al., 2021).
Lower tolerance to novel chemical stress was associated with reduced genome-wide genetic diversity in the modern genotype (LRV_1). These patterns suggest that genetic erosion occurred as result of multidecadal exposure to chemical stress (Diez-Del-Molino et al., 2018). The theory that genetic erosion occurred in Lake Ring has to be validated through population-level analysis of genomewide variation. However, the significant decline in genetic diversity in the most modern genotype (LRV0_1), which was exposed to chemical pollution for 10-15 years/sexual generations (starting from the 1980s and accounting for a sexual generation per year in the Daphnia population), agrees with experimental studies on genetic erosion, showing a decline in genetic diversity following >12 generations of exposure to chemicals (e.g., Nowak et al., 2009). Our findings on pervasive plasticity both within and between generations align with the hypothesis of genetic erosion. Previous studies have shown that plasticity can be maintained in the face of genetic erosion, but it comes with fitness costs (Luquet et al., 2011) and with reduced tolerance to environmental stress (Bijlsma & Loeschcke, 2012).
Genes significantly divergent between naïve and experienced genotypes showed lower diversity in experienced genotypes. These These pathways are potential targets in other species, including humans.