The Reproductive Effects of the Cancer Chemotherapy Agent, Carmofur, on Daphnia magna are mediated by its metabolite, 5-Fluorouracil

Abstract

Carmofur is an antineoplastic agent that inhibits ceramidase, a key enzyme in the sphingolipid pathway. Previous research suggests carmofur represses reproductive maturity in Daphnia magna. The purpose of this experiment was to confirm carmofur’s effects on fecundity and reproductive maturity over two generations. A chronic toxicity test found reproductive maturity was delayed from 9 to 19 days by 0.80 μM carmofur with a 99.7% drop in reproduction, probably caused by delayed ovarian development. Second generation effects were even greater with 0% reproductive success at 0.40 μM. To our surprise, carmofur was not measured in the media by HPLC 24 hours after exposure. Previous research indicated that carmofur is unstable in water and hydrolyzed into 5-fluorouracil (5-FU). Therefore, the chronic toxicity study was repeated with 5-FU and similar effects on reproductive maturity were observed at similar concentrations despite very different acute toxicities (48 h carmofur LC50 = 1.93 μM; 5-FU LC50 = 207 μM). 5-FU delayed reproductive maturity from 9 to 21 days with a 71.12% drop in reproduction at 0.80 μM and greater effects in the 2^nd generation similar to carmofur. 5-FU was found stable in aquatic media and HPLC confirmed 5-FU was hydrolyzed from carmofur within 24 hours. In conclusion, carmofur and 5-FU reduce fecundity because they delay reproductive maturity and ovarian development in Daphnia magna. We conclude that the reproductive effects observed after carmofur treatment are primarily mediated by its breakdown product, 5-FU. This further underscores the importance of measuring chemical concentrations and evaluating chemical metabolism and decomposition when determining toxicity, especially of chemotherapeutic agents.

INTRODUCTION:

Daphnia magna are small planktonic Crustacea commonly used as a model species in ecotoxicology because of their role as primary consumers in aquatic food chains. Daphnia transfer energy rich nutrients from primary producers such as algae to other organisms higher in the trophic level (Colbourne et al. 2011; Stollewerk 2010). Consumption of primary producers that are rich with polyunsaturated fatty acids (PUFAs) are crucial for the tightly regulated processes of energy metabolism, growth and development (Ravet et al. 2012; Sengupta et al. 2017). Some key dietary PUFAs include eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA), linoleic acid, and arachidonic acid (AA) (von Elert 2002). Daphnia develop eggs in the brood chamber that are later released as neonates when the female molts. To do so the females need to obtain a threshold level of PUFAs (Ravet et al. 2012; Sengupta et al. 2016; Sperfeld and Wacker 2015). Of these, AA, DHA, and EPA are considered crucial in growth and reproduction within Daphnia and other organisms (Ginjupalli et al. 2015; Ravet et al. 2012; Taipale et al.2011). Therefore, disruption of PUFA transfer and metabolism by anthropogenic chemicals can have profound effects on reproduction and development, as well as other energy and endocrine-mediated processes such as cardiovascular disease, diabetes, and obesity (Ginjupalli et al. 2015;Heindel and Blumberg 2019; Heintz et al. 2019; Maradonna and Carnevali 2018; Sengupta et al. 2016).

Carmofur is an anti-neoplastic ceramidase inhibitor with some anti-viral properties that is also used as a pro-drug for oral 5-fluorouracel treatment (5-FU) (Jin et al. 2020; Realini et al. 2013; Straub 2010). In turn, carmofur perturbs normal PUFA metabolism and is associated with delayed development and reproduction in Daphnia magna (Sengupta et al. 2017). Antineoplastic agents have been found in wastewater and surface waters, of which more than 100 have been found with 5-FU, carmofur’s hydrolysis product, one of the top 2 and a known DNA damaging agent (Kümmerer et al. 2016).

Sphingolipids, including sphingomyelins and ceramides, are crucial lipids involved in cell signaling, growth and development, and may be involved in reproductive maturity such as ovarian or egg development (Cutler et al. 2014; Hernández-Coronado et al. 2015; Herr et al. 2003). Sphingomyelins are converted by sphingomyelinases to ceramide; ceramides are converted by ceramidases to sphingosine; that is metabolized by sphingosine kinase to sphingosine-1-phosphate (Cutler et al. 2014). All of these sphingolipids are involved in signaling processes and reproduction, especially sphingosine-1-phosphate (Jee et al. 2011; Kitatani et al. 2008; Phan et al. 2007).

In Drosophila and Daphnia magna, the nuclear receptor HR96, which is related to mammalian Farnesoid-X-Reeptor (FXR) and Liver-X-Receptor (LXR) (Litoff et al. 2014), regulate responses to chemical stressors (Karimullina et al. 2012; King-Jones et al. 2006), cholesterol, and triacylglycerol uptake and utilization, including sphingomyelin metabolism (King-Jones et al. 2006; Schmidt et al. 2017; Sengupta et al. 2017; Sieber and Thummel 2012). Previous work showed that lipid and xenobiotic modulators of HR96 perturbed sphingomyelin concentrations as activators (linoleic acid) decreased sphingomyelin concentrations and inverse activators (DHA, triclosan) increased sphingomyelin concentrations (Sengupta et al. 2017). Subsequent experiments indicated that the sphingomyelinase inhibitor, GW4869, that blocks production of ceramides from sphingomyelin, did not perturb reproduction, but the ceramidase inhibitor, carmofur, that blocks production of sphingosine from ceramides, inhibited reproduction (Sengupta et al. 2017). However, the specific effects of carmofur on brood size and its muli-generational effects were not monitored. Therefore, the original purpose of this manuscript was to further elucidate the effects of carmofur on the reproductive development and fecundity of Daphnia magna for two generations.

However, Carmofur is not stable in water. It loses its hydrocarbon chain through hydrolysis and breaks down to 5-FU (Nakajima et al. 1981; Realini et al. 2013). Both carmofur and 5-FU are anti-cancer drugs, but work through different mechanisms (Realini et al. 2013). 5-FU works by inhibiting deoxythymidine monophosphate (dTMP) production that is essential for DNA replication leading to cytotoxicity (Zhang et al. 2008). Therefore, we ultimately need to assess the acute and chronic toxicity of both carmofur and 5-FU to estimate which is the ultimate reproductive toxicant in D. magna. Furthermore, recent research has indicated that cancer chemotherapy agents are potentially problematic in aquatic environments due to poor breakdown in wastewater treatment facilities (Kümmerer et al. 2016; Yadav et al. 2021). Our work demonstrates that carmofur most likely does not act as a sphingomyelin inhibitor in Daphnia and instead its actions on reproduction are initiated by 5-FU. Furthermore, it underscores the importance of understanding biotransformation and natural degradation and the need to confirm active concentrations of chemicals when assessing aquatic toxicity in test organisms.

MATERIALS AND METHODS

Daphnia magna Culture

A culture of D. magna has been maintained at Clemson University for about 25 years (Ginjupalli et al. 2015). Stock cultures of D. magna were maintained in moderately hard water at pH 8.2-8.4, 22°C with a 16:8 hour light cycle in an environmental chamber. Daphnia were fed daily with approximately 6×10⁶ cells of Raphidocelis subcapitata per daphnid and supplemented with 0.25 mg dry weight of blended TETRAFIN fish flakes.

Acute Toxicity of Carmofur and 5-FU

The acute toxicity of both carmofur and 5-FU was analyzed at 0, 0.5, 1.0, 2.0, 4.0, and 8.0 μM carmofur, and 0, 0.385, 0.9625, 1.925, and 3.85 mM 5-FU. Four daphnids placed in 40 mL of media were exposed to either carmofur or 5-FU (n = 4 beakers per concentration) and observed at 24 and 48 h for survival.

Quantification of Carmofur and 5-FU Media Concentrations

Carmofur or 5-FU containing media was collected at 0 and 24 hours (n = 3). Forty ml of media was divided in two subsamples and carmofur was extracted 2x with 20 ml of dichloromethane (DCM; Fisher Scientific, Atlanta, GA); 5-FU was extracted with ethyl acetate (Fisher Scientific, Atlanta, GA) after acidification with 1 μL of glacial acetic acid. The carmofur containing DCM fraction was evaporated under nitrogen and reconstituted in 1 ml of 56:44 methanol:water of which 30 μl was injected in the HPLC system; the 5-FU containing ethyl acetate was also evaporated under nitrogen and reconstituted in 1 mL of DMSO of which 20 μl was injected.

High performance liquid chromatography (HPLC) was performed to determine the concentration levels of carmofur or 5-FU in the media by comparing peak area to a standard curve from carmofur concentrations at 0, 0.075 μg/mL (0.303 μM), 0.15 μg/mL (0.606 μM), 0.3 μg/mL (1.21 μM), 0.6 μg/mL (2.43 μM), 1.25 μg/mL (4.85 μM), 2.5 μg/mL (9.7 μM) and 5 μg/mL (19.4 μM) to the samples, or 5-FU concentrations at at 0, 0.20, 4.0, and 100.0 μM to samples. Carmofur standards and samples (30 μL) were run on the HPLC (Waters, Corporation, Milford MA USA) for 10 minutes at a flow rate of 0.8mL/min of 56:44 methanol:water and peak volume quantified at a UV-vis of 254 nm. Carmofur eluted shortly after 4 minutes. 5-FU standards and samples (20 μL) were run on the HPLC for 10 minutes at a flow rate of 0.8 mL/min of 10:90 methanol:water with 0.1% acetic acid, and peak volume quantified.

First Generation Chronic Toxicity Test

Reproduction and developmental maturity of D. magna were tested via a chronic toxicity test using standard methods (Ginjupalli and Baldwin 2013; United States 1994) following exposure to carmofur at nominal concentrations of 0, 0.05, 0.10, 0.20, 0.40, 0.80 μM or 5-FU at 0, 0.20, 0.80, 4.0, 20.0, and 100.0 μM. Carmofur and 5-FU were dissolved in ethanol and final concentrations of solvent were 0.01% in all beakers including the controls. Neonates ≤ 24-h of age were selected at random and placed into their own 50 mL beaker containing 40 mL moderately hard media. The six different exposure groups (n = 13) were fed 3×10⁶ cells of R. subcapitata per daphnid (beaker) and 200 mg (50 μl) of blended TetraFin fish flakes until the daphnids were 7-days old; at that point the daphnids were fed 6×10⁶ cells of P. subcapitata and 50 μL of blended TetraFin fish flakes daily. Media was changed every other day at which point D. magna were monitored for survival, egg development, molt release, and fecundity until adults were 21-days old. The length of three daphnids per treatment group was determined from the head to the base of the tail (in mm) at 14-days old under a dissection microscope (American Optical Corporation, Buffalo, NY, USA) with a micrometer.

Second Generation Chronic Toxicity Test

Neonates from each concentration group were selected at random for a second generation test at day 21 of the first generation chronic toxicity test (n = 10). Neonates continued to be exposed at the same concentrations of the first chronic toxicity test. In addition, ten neonates were selected at random for a second generation test from concentrations at 0, 0.05, 0.1, 0.2 and 0.4 μM from the carmofur group and 0, 0.20, and 0.80 μM from the 5-FU exposed daphnids. Daphnids were then exposed to carmofur, 5-FU, or left unexposed to determine whether these chemical had effects regardless of continued exposure. As per the first generation, media was changed every other day at which point D. magna were monitored for survival, molting, egg development, age at first brood, and fecundity until 21-days old.

Statistical Analysis

Statistical analysis was performed by one-way ANOVA followed by Fishers LSD as the post-hoc test using GraphPad Prism 7 (GraphPad Software, San Diego, CA USA). LC50s and EC50s were determined by non-linear regression curve fit using GraphPad Prism 7. For LC50s, data was normalized to percent survival, log transformed, and LC50 values were determined by non-linear regression curve fit (sigmoidal dose-response curve), log(inhibitor) vs normalized response – variable slope, ordinary fit with GraphPad 7.0 (Schmidt et al. 2017; Sengupta et al. 2017). To compare chronic toxicity between carmofur and 5-FU exposed daphnids, EC50 values were deteremined for offspring / adult, offspring / brood, and number of broods. To determine EC50s, data was normalized to percent survival, log transformed, and LC50 values determined by non-linear regression curve fit (sigmoidal dose-response curve, 4PL), log(inhibitor) vs normalized response – variable slope, ordinary fit with GraphPad 7.0 (Schmidt et al. 2017; Sengupta et al. 2017).

The 24 and 48 h acute/chronic ratios (ACR) were determined by dividing the acute toxicity (LC50) by the chronic toxicity (EC50). Typically, to determine the ACR, the square root of the NOEC multipled to the LOEC is used to determine a maximum allowable toxicity concentration (MATC) and this is divided by the LC50. However, this method has much less precision than curve models that often show NOECs between the EC10 and EC30, and the traditional ACR calculation is not possible if the NOEC is 0 (Fox 2010; Landis and Chapman 2011; Rider and LeBlanc 2005). Therefore, the EC50 was used in our calculation as a basic standard in biology and often used along with Hillslopes in mixture assessments (Baldwin and Roling 2009; Rider and LeBlanc 2005).

RESULTS

First Generation Chronic Toxicity Test of Carmofur (Life Cycle assessment):

A chronic toxicity test was performed to determine the reproductive effects of the ceramidase inhibitor, carmofur on D. magna. The survival rate of first generation daphnids was high with the exception of the highest concentration (0.8 μM) where 10/13 daphnids (77%) survived during the exposures. Two of the daphnids died at day 12 and the third daphnid died on day 21 (Table 1). However, carmofur had significant effects on reproduction (Fig. 1). Carmofur reduced fecundity in a concentration-dependent manner with reproduction repressed 99.68% at the highest concentration tested (0.8 μM) (Table 1). Fecundity was repressed 31.46% and 85.2%, respectively at 0.2 and 0.4 μM. All female daphnids reproduced at 0 - 0.2 μM; however, only 77% and 23% reproduced, respectively at 0.4 and 0.8 μM (Fig. 2A). At 0.8 μM, only 4 neonates total were released from 3 reproductive adult female daphnids. This low fecundity precluded us from including 0.8 μM in the second generation.

Table 1:

Carmofur perturbs reproductive maturity and development in the first generation.

Concentration (μM)	Percent Survival	Offspring produced / Adult Daphnid	Number of Molts	Age at First Brood	Length
0	100%	95.62 ± 4.18	8.54 ± 0.29	9.00 ± 0.00	3.77 ±0.32
0.05	100%	94.54 ± 4.13	8.38 ± 0.24	9.00 ± 0.00	3.70 ± 0.12
0.1	100%	81.62 ± 2.89*	8.38 ± 0.31	9.46 ± 0.31	3.80 ± 0.06
0.2	92%	65.54 ± 6.21**	8.69 ± 0.41	11.08 ± 0.39*	3.70 ± 0.06
0.4	100%	14.15 ± 0.17**	8.92 ± 0.26	16.38 ± 1.11**	3.77 ± 0.15
0.8	77%	0.31 ± 0.17**	7.61 ± 0.38*	18.92 ± 1.09**	3.57 ± 0.12

Data provided as mean ± SEM. Statistical signifigance determed by ANOVA followed by Fisher’s LSD and compared to 0mM.

^(*)indicates p < 0.05; and

^(**)p < 0.0001 (n = 13).

Figure 1. Carmofur decreases fecundity in a concentration-dependent manner.

A chronic toxicity test was performed and neonates per daphnid were determined for 21-days. Statistical analysis was performed by one-way ANOVA with * p < 0.01; ** p < 0.001 (n = 13). (a) indicates all exposure groups are significantly different from the control (UT), (b) indicates all exposure groups 0.10 μM and greater are different from the control and (c) indicates exposure groups 0.20 μM and greater are significantly different from the control.

Figure 2. Carmofur reduced reproduction in a concentration-dependent manner in the first generation.

A chronic toxicity test was performed and (A) the total number of reproductive daphnids, (B) mean age of first showing ovaries, (C) mean number of broods per daphnid, (D) mean age at first brood, and (E) the mean number of offspring per brood was determined. Statistical analysis was performed by one-way ANOVA with * p < 0.05 , ** p < 0.0005; *** p < 0.0001 compared to the control (n = 13). Data are shown as mean ± SEM (n= 13).

The most sensitive parameters associated with reduced fecundity were age when first showing ovaries (Fig. 2B) and number of broods per adult female daphnid (Fig. 2C). Age at first brood (Fig. 2D) and offspring per brood (Fig. 2E) were not as sensitive of an indicator as fecundity. The reduced number of broods was probably due to a delay in reproductive maturity; delayed ovary production and age at first brood. On average, controls started releasing broods on day 9 (Table 1). This is consistent with previous research in our laboratory as we regularly observe the ovaries develop on about day 7 and brood release (reproduction) at day 9 (Sengupta et al. 2017). Higher concentrations of carmofur increased the length of time for reproductive maturity as the 0.2μM concentration took about 11 days on average to become reproductively mature and the 0.4μM concentration took 16 days (Table 1); nearly double the time of controls. This means that the carmofur had a strong enough effect on these two groups to take up to 7 more days to become reproductively mature compared to those not exposed to carmofur. Ovary production generally followed a similar trend of displaying ovaries at a later age for the higher concentrations (Fig. 2B). Overall, carmofur repressed fecundity in a concentration-dependent manner with brood number, associated with a delay in reproductive maturity, the most sensitive indicators.

Second Generation Chronic Toxicity Test of Carmofur:

To test the effects of maternal and continued exposure to carmofur, neonates from the first generation were selected at random and continued to be exposed at the same concentrations. This time a slightly greater effect was seen on survival. The only group with 100% survival (10/10) was the control group, while the other concentrations ranged form 1-4 deaths (Table 2). All but one death was due to an improper completion of the molting cycle as determined by the presence of an incomplete molt (Baldwin et al. 2001). Reproduction was also more severly effected in the second generation (Fig. 3). The controls and 0.05 μM concentration group averaged similar number of offspring between the first generation and second generation. The 0.1 μM group had a 37.13% decrease from the control in the second generation compared to only a 17.64% drop in the first generation (Table 2). Only 5 adult daphnids from the 0.2 μM reproduced and they only produced one brood each with low numbers of neonates. None of the daphnids in the highest concentration (0.4 μM) reproduced in the 2^nd generation (Fig 4A).

Table 2.

Carmofur decreased reproductive maturity and development in the second generation

Concentration (μM)	Percent Survival	Offspring produced / Adult Daphnid	Number of Molts	Age at First Brood	Length
0	100%	99.10 ± 2.24	6.80 ± 0.25	9.40 ± 0.48	4.13 ± 0.09
0.05	90%	95.30 ± 4.13	6.60 ± 0.40	10.80 ± 1.20	3.87 ± 0.03
0.1	80%	62.30 ± 9.39**	6.20 ± 0.66	12.30 ± 1.51	3.83 ± 0.06*
0.2	60%	1.60 ± 0.70**	5.50 ± 0.82	18.50 ± 1.19**	3.77 ± 0.03*
0.4	80%	0.00 ± 0.00**	5.70 ± 0.96	NA**	3.53 ± 0.17**

Data provided as mean ± SEM. Statistical significance determined by ANOVA followed by Fisher’s LSD and compared to 0 μM.

^(*)indicates p < 0.05;

^(**)indicates p < 0.001 (n = 10).

NA = not applicable

Figure 3. Total number of neonates produced decreases as carmofur concentration (μM) increases in the second generation.

A second generation chronic toxicity test was performed and the number of neonates produced per daphnid was recorded for 21-days. Statistical analysis was performed via one-way ANOVA. *p < 0.01 and **p < 0.001. (a) indicates all exposure groups are significantly different from the control and (b) indicates all exposure groups 0.10 μM and greater are different from the control.

Figure 4. There was an inverse relation between reproductive success and carmofur concentration in second generation chronic toxicity tests.

(A) The total number of reproductive daphnids, (B) mean age of first showing ovaries, (C) mean number of broods released per daphnid, (D) mean age at first reproduction, and (E) mean number of offspring per brood were determined. Statistical analysis was performed by one-way ANOVA with * p < 0.05 , ** p < 0.0005; *** p < 0.0001 compared to the control. * indicates p <0.05; ** indicates p < 0.0001 (n = 10). Data are shown as mean ± SEM (n = 10).

The age at which the adult daphnids develop ovaries, produce broods, and the number of broods produced were relatively similar to the first generation with slightly more sensitivity to carmofur as to number of broods in the 2^nd generation (Figure 4BCD). The number of offspring per brood was also more sensitive in the 2^nd generation at the 0.10 mM concentration when compared to the 1 first generation (Fig. 4E). Overall, the data was similar between the 1^st and 2^nd generations with slightly greater sensitivity in the second generation.

Measured Carmofur and 5-FU Concentrations:

Carmofur was measured in the exposure media directly after adding aliquots of the stock solution (t = 0) and again after 24 h. While the measured concentrations at t = 0 were very close to the nominal concentrations (80-130%), after 24 h no carmofur was detected at any of the concentrations (Table 3). When 5-FU was measured in the exposure media, both at t = 0 and t = 24 the actual concentrations were very close to the nominal concentrations (86-138%) (Table 4; Suppl fig 1–2 show HPLC histograms). It was confirmed that this breakdown was not a result of microbial activity, as the breakdown over 24 h was also observed when carmofur was incubated with sterile water (Table 3). Clearly the carmofur is unstable in aqueous solutions. Based on the literature, the carmofur breaks down through spontaneous hydrolysis into 5-FU (21, 22). Therefore, we also quantified 5-FU concentrations after the carmofur exposures and found that about 76 – 91% of the original carmofur was detected as 5-FU (Table 5), confirming that carmofur is spontaneous hydrolyzed to 5-FU.

Table 3:

Nominal and measured concentrations of carmofur after carmofur exposure.

Nominal Conc (μM)	Measured Conc (μM) 0 h	Measured Conc (μM) 24 h	Measured Conc (μM) 24 h (sterile water)
0	n.d.	n.d.	n.d.
0.05	0.0574 ± 0.0078	n.d.	n.p.
0.1	0.1345 ± 0.0136	n.d.	n.p.
0.2	0.1638 ± 0.0318	n.d.	n.d.
0.4	0.3389 ± 0.0448	n.d.	n.p.
0.8	0.7083 ± 0.342	n.d.	n.d.

n.d. = not detected

n.p. = not performed

Data shown as mean ± SEM

Table 4:

Nominal and measured concentrations of 5-fluorouracil after 5-fluorouracil exposure.

Nominal Conc (μM)	Measured Conc (μM) 0 h	Measured Conc (μM) 24 h
0	n.d.	n.d.
0.2	0.183 ± 0.023	0.172 ± 0.0167
4	4.003 ± 0.026	3.728 ± 0.727
100	139.352 ± 4.829	137.815 ± 4.41

n.d. = not detected

Data shown as mean ± SEM

Table 5:

Nominal and measured concentrations of 5-fluorouracil after carmofur exposure.

Nominal Conc (μM)	Measured Conc (μM) 0 h	Measured Conc (μM) 24 h
0	n.d.	n.d.
0.2	n.d.	0.1523 ± 0.0050
0.8	n.d.	0.7290 ± 0.0531

n.d. = not detected

Data shown as mean ± SEM

Comparing Acute Toxicity Between Carmofur and 5-FU

Carmofur’s and 5-FU’s acute toxicity were remarkably different (Fig. 5). The 24 and 48 h LC50s of carmofur were 2.43 and 1.93 μM, respectively, while the 24 and 48 h LC50s of 5-FU were 26790 and 207 μM, respectively. The difference between the two were stark with much greater acute toxicity following carmofur exposure. Furthermore, carmofur’s toxicity was relatively stable over time, potentially a product of it’s breakdown as it’s acute toxicity is probably caused by the parent compound, which is not available after 24 hours (Table 3). 5-FU’s acute toxicity increased 100X over the last 24 hours of exposure, potentially a product of its stability and its effects on DNA (Realini et al. 2013; Zhang et al. 2008).

Figure 5: Acute Toxicity of Carmofur and 5-FU.

An acute toxicity test was conducted on Daphnia magna comparing carmofur and 5-FU over 24 and 48 hours of exposure. LC50’s were determined using GraphPad Prism 7.0 as described in the Materials and Methods.

First Generation Chronic Toxicity Test of 5-FU

A chronic toxicity test was performed on D. magna to compare the reproductive effects of the DNA-replication inhibitor, 5-FU with the reproductive effects of carmofur. The survival rate of first generation daphnids was low at 0 μM (10/12 daphnids; 83% survival; however one of these daphnids died because of improper handling during the exposure. The 20 μM concentration had 92% survival (11/12) and this daphnid died on day 21 (Table 6).

Table 6:

5-FU perturbs reproductive maturity and development in the first generation.

Concentration (μM)	Percent Survival	Offspring produced / Adult Daphnid	Number of Molts	Age at First Brood	Length (mm)
0	83%	67.8 ± 2.49	6.1 ± 0.1	9.3 ± 0.21	3.99 ± 0.03
0.2	100%	58 ± 3.17*	5.5 ± 0.15	12.08 ± 0.08**	4.09 ± 0.02
0.8	100%	19.58 ± 2.68**	6.18 ± 0.37	18.75 ± 0.33**	3.85 ± 0.03*
4	100%	0.9167 ± 0.43**	5.67 ± 0.22	21.67 ± 0.14**	3.70 ± 0.04**
20	92%	0.0 ± 0.0**	5.55 ± 0.28	NA**	3.59 ± 0.03**
100	83%	0.0 ± 0.0**	6.3 ± 0.21	NA**	3.21 ± 0.08**

Data provided as mean ± SEM. Statistical signifigance was determined by ANOVA followed by Fisher’s LSD and compared to 0 μM control.

^*indicates p < 0.05;

^**p < 0.0001 (n = 12).

NA = Not applicable

Exposure to 5-FU reduced fecundity in a concentration-dependent manner with reproduction repressed 100% at 20 and 100 μM (Fig 6). At concentrations between 0 - 0.8 μM all daphnids produced neonates; however, as the concentration increased the number of neonates produced decreased. The 4 μM concentration exposed daphnids only produced 11 neonates from 4 different daphnids on the 21^st day. At 0.8 μM, most of the reproduction occurred during the last brood as the first brood was often released around day 18 (Fig. 7a).

Figure 6. 5-Fluorouracil reduces the fecundity of Daphnia magna in a concentration-dependent manner.

A chronic toxicity test was performed and neonates per daphnid were determined for 21-days. Statistical analysis was performed by one-way ANOVA with * p < 0.001; ** p < 0.0001 (n = 12). (a) indicates all exposure groups are significantly different from the control, (b) indicates all exposure groups different from the control with a p < .0001 significance, except for 0.2 μM which is different from the control with a p < 0.001 significance.

Figure 7. 5-Fluorouracil reduced reproduction in Daphnia magna in a concentration-dependent manner in the first generation.

A chronic toxicity test was performed and (A) the total number of reproductive daphnids, (B) the mean age of Daphnia magna first showing eggs, (C) the mean number of broods released per Daphnia (D) the mean age at first brood, and (E) the mean number of offspring per brood were determined. Statistical Analysis was performed by one-way ANOVA and ** indicates a p-value < 0.0001 (n = 12).

The age at which reproductive maturity was achieved increased in a concentration-dependent manner as measured by the appearance of ovaries (Fig 7B), broods per daphnid (Fig 7C), and age of first brood release (Fig 7D). As with carmofur, 5-FU reduced the number of broods, most likely due to a delay in reproductive maturity. The most sensitive indicators of 5-FU toxicity were the same as those for carmofur; appearance of ovaries, broods per daphnid, and age of first brood release with offspring / brood the least sensitive variable (Fig 7E).

Second Generation Chronic Toxicity Test of 5-FU

A second-generation chronic toxicity test was conducted using random neonates from the first-generation study. All groups had a 100% survival. Reproduction was heavily affected in the second generation as the number of offspring per Daphnia dropped 61% at 0.2 μM and 99% at 0.8 μM compared to the control (Fig. 8)(Table 7).

Figure 8: 5-Fluorouracil reduced fecundity of Daphnia magna in a concentration-dependent manner in the second generation.

A second generation chronic toxicity test was performed and neonates per daphnid were detennined for 21-days. Statistical Analysis was performed by one-way ANOVA with * indicating p < .05, ** indicating p < .01, and *** indicating p < 0.0001 between the control and the treated groups.

Table 7.

5-FU decreased reproductive maturity and development in the second generation.

Concentration (μM)	Percent Survival	Offspring produced / Adult Daphnid	Number of Molts	Age at First Brood	Length (mm)
0	100%	87.5 ± 3.39	7.7 ± 0.21	7.0 ± 0.0	3.77 ± 0.06
0.2	100%	33.9 ± 3.35**	8.0 ± 0.0	16.0 ± 0.26**	3.63 ± 0.02
0.8	100%	0.9 ± 0.9**	7.9 ± 0.41	21.7 ± 0.3**	3.37 ± 0.08**

Statistical Analysis was performed by one-way ANOVA;

^(*)indicates a significance of p < .001 and

^(**)indicates a significance of p < 0.0001 (n = 10)

The cause of reduced fecundity was once again associated with delayed reproductive maturity and poor reproduction potential (Fig. 9A) associated with delayed egg/ovary production (Fig. 9B), lower brood numbers (Fig. 9C), and delayed brood release (Fig. 9D). Number of offspring per brood was not as sensitive a marker (Fig. 9E).

Figure 9. 5-Fluorouracil reduced reproduction of Daphnia magna in the second generation.

A second generation chronic toxicity test was performed and (A) the total number of reproductive daphnids, (B) the mean age of Daphnia magna first showing eggs, (C) the mean number of broods released per Daphnia, (D) the mean age at first brood release, and (E) the mean number of offspring per brood were determined. Statistical Analysis was performed by one-way ANOVA; * p < .001; ** p < 0.0001 (n = 10).

Second generation toxicity was also investigated in daphnids in which after first generation exposure to carmofur and 5-FU, the second generation was not exposed. This experiment was performed to determine if first generation exposure caused developmental toxicity that perturbed the reproductive system. Few significant differences were observed and none of biological consequence (Suppl Fig 3–7).

Not only are the Lowest Observable Effect Concentrations (LOEC) and specific reproductive paramaters (delayed ovary production, broods / daphnid) effected in a similar manner and at similar concentrations between carmofur and 5-FU, but so are their EC50s for brood number, offspring/adult and offspring/brood. This despite the tests not being performed at the same time. The EC50s for brood number were 0.323 and 0.510 μM for carmofur and 5-FU, respectively; for offspring/adult were 0.268 and 0.503 μM, and for offspring/brood were 0.336 and 0.864 μM, further indicating the chronic effects of carmofur were elicited by 5-FU. Data and curves are shown in Suppl Fig. 8. The 24 and 48 h acute/chronic ratio (ACR; acute LC50/chronic EC50) of carmofur are 9.06 and 7.20 when offspring/adult is used for chronic toxicity. The 24 and 48 h ACRs for 5-FU are 53,260 and 411.5. 5-FU shows much greater chronic toxicity relative to its acute toxicity.

Discussion:

Our hypothesis was that carmofur reduced fecundity by inhibiting ceramidase activity and delaying reproductive maturity based on previous results in our laboratory (Realini et al. 2013; Sengupta et al. 2017). Carmofur did delay reproductive maturity by delaying ovary development, brood release, and ultimately the number of broods produced. However, carmofur was not detected in the media at 24 h. Carmofur readily hydrolyzes to 5-FU (Nakajima et al. 1981;Realini et al. 2013), and is especially susceptible to base or water-based hydrolysis (Uekama and Hirayama 2011). Carmofur is often complexed to cyclodextrins wrapped in organic acids to reduce the hydrolysis to 5-FU (Kikuchi et al. 1987; Uekama and Hirayama 2011). Therefore, it is highly unlikely that the toxic effects observed due to carmofur are due to its effects as a ceramidase inhibitor (Realini et al. 2013) or pertubations in sphingolipids levels (Sengupta et al. 2017). This research highlights the importance of understanding biotransformation of the chemical in our studies and measuring the chemical when possible.

Interestingly, 5-FU has identical chronic toxicity effects to carmofur at nearly identical concentrations despite widely different acute toxicities. 5-FU is readily detectible in the media following exposure. Therefore, our data indicates that carmofur (and 5-FU) delay reproduction through 5-FU’s toxic effects. 5-FU inhibits DNA synthesis and therefore is especially active in rapidly dividing cells. In turn, it appears to have very powerful effects on the rapidly developing ovaries and its eggs, causing reproductive delay. Its lack of acute toxicity is telling and indicates that other tissues are not nearly as sensitive. Although it may not be acutely toxic, 5-FU is rather chronically toxic. It’s 48 h ACR is approximately 411.5 using the EC50 as a measure of chronic toxicity. In contrast, carmofur has an ACR of 7.20 because of its increased acute toxicity. Using the LOEC would provide higher ACRs for both compounds. This makes 5-FU one of the most specifically chronically (reproductively) toxic compounds that can be found in the environment (May et al. 2016). More importantly, basic acute or short-term tests would not find 5-FU toxic and would miss its environmental effects due to its effects on fecundity.

The second generations experienced stronger effects than the first generation following exposure to either carmofur or 5-FU. For example, the higher concentrations used in the first generation significantly repressed fecundity, but in the second generation these concentrations either ablated fecundity or caused death associated with the molting cycle, a time of growth. Thus, exposure to 5-FU based chemotherapeutic agents, 5-FU and carmofur, has significant effects on reproduction that worsen with continued exposure over multiple generations.

As the life expectancy in the US and other developing countries increases, we are seeing a rise in diseases such as cancer (Rabii et al. 2014). As such, anti-neoplastics are increasingly found in our wastewater and eventually our surface waters, and therefore anti-neoplastics can exhibit potentially toxic effects on aquatic biota (Mello et al. 2020). However, along with killing rapidly proliferating cells, they have the potential to cause harmful reproductive effects based on the 1^st and 2^nd generation chronic studies. As with many toxic chemicals, carmofur and 5-FU can end up in wastewater and eventually water sources, and can exhibit toxic effects not only on us but also on the environment. The concentrations of 5-FU that induced reproductive toxicity in Daphnia magna are about 12.3 – 360X less than estimated in wastewater and 8277X less than measured in surface waters (Straub 2010). Therefore, the risk of 5-FU and other anti-neoplastics agents in the environment appears low (Kümmerer et al. 2016); however, the difficulty in eliminating them from wastewater coupled with their lipophilicity, increased use, and additive toxicity of 100s of compounds may ultimately increase their risk to the environment along with other pharmaceuticals (Yadav et al. 2021). In conclusion, the toxic effects of carmofur on reproductive maturity are mediated by its metabolite 5-FU. Our work also underscores the importance of measuring chemical concentrations and examining chemical stability and breakdown. Last, the chemotheraputic drugs, carmofur and 5-FU and potentially other drugs that effect DNA replication, are potential causes of concern for the environment because of their effects on reproduction and egg development.

Data Availability

All data is found in the manuscript. Raw data is available on request