Registration Dossier

Diss Factsheets

Administrative data

Key value for chemical safety assessment

Effects on fertility

Additional information

The available data for this end point are an oral (diet) OECD 421 reproductive toxicity screening study conducted using the neutralized AMP-HCl (Hydrochloride salt of AMP) and a two-generation reproductive toxicity study conducted using 4,4-dimethyloxazolidine (which hydrolyzes in the stomach to AMP and formaldehyde). No data are available on 2-amino-2-methyl-1-propanol (AMP) as the free base.

OECD 421 reproductive/developmental screening study

In the OECD 421 one-generation reproductive toxicity screening study, groups of 12 male and female rats were administered diets containing 0, 100, 300, and 1000 mg/kg/day the hydrochloride salt of 2-amino-2-methyl-1-propanol (AMP-HCl) for two-weeks pre-breeding, through the breeding and gestation phases up to lactation day 4 (Carney et al., 2005). The dose levels were not corrected for AMP content; therefore, administered doses of 0, 100, 300, and 1000 mg/kg/day AMP-HCl represent 0, 72, 216, and 720 mg/kg/day of AMP. Administration of 1000 mg/kg/day of AMP-HCl resulted in no litters being produced (due to embryonic resorption), while the 300 mg/kg/day group induced approximately 70% litter loss (some complete litters lost, some partial, some no effects) by the same means. The 100 mg/kg/day group was the reproductive no-observed-effect level (NOEL) for this study. Where there was no, or partial litter loss, the surviving pups developed normally until the end of the study with no abnormalities identified. The study clearly identified a threshold for the effects at which, litter loss did not occur. With the exception of this reduction in litter size there were no other effects reported on reproductive or developmental parameters. The litter loss was considered likely to be due to early resorption of embryos after implantation. Taking into consideration the lack of any other reproductive effects, the effects in litter size were considered likely to be a developmental effect rather than a fertility effect. This was confirmed with subsequent studies on the mode of action.

Maternal liver toxicity was evident at all doses (including the lowest dose) and presented as fatty accumulation in hepatocytes. This was most severe in the mid and high dose groups. The relevance of these findings in the interpretation of the developmental toxicity caused by AMP is discussed below.

Two-generation reproductive toxicity study

Upon ingestion, 4,4-dimethyloxazolidine breaks down to AMP and formaldehyde with a molar ratio of 9:1 in the stomach (Saghiret al.,2007). As such, data generated on this substance is considered adequate for read across to predict AMP toxicity since oral dosing leads to systemic exposure to AMP (and formaldehyde to a lesser extent).

In the two-generation study, a reduction in litter size was observed in both generations at the highest dose tested (200 mg/kg/day). As in the OECD 421 study, the reduced litter size was considered to be due to early post implantation loss rather than reduction in viable embryos. The degree of postimplantation loss in the P1 generation was 18.7 versus 6.4% in controls and in the P2 generation was 16.3 versus 10.8% in controls. There were no other reproductive or developmental effects at this dose level however there was significant evidence of gastrointestinal irritation and liver toxicity (fatty accumulation and increased organ weight). There were no effects on reproductive performance or offspring growth/survival in the low and mid dose groups (20 or 60 mg/kg/day). The NOEL for reproductive effects was 200 mg/kg/day and the NOEL for developmental effects in this study was 60 mg/kg/day. The LOAEL for developmental effects was the highest dose of 200 mg/kg/day.

Data from repeated dose toxicity studies

In the available sub acute, sub chronic and chronic toxicity studies performed using AMP there have been no observed effects on any reproductive tissues (testis, epididymes, uterus, ovaries). Repeated dose toxicity is limited to effects in the liver.

Summary of toxicity to fertility

In the OECD 421, the 2-generation study and the available repeated dose toxicity studies there were no effects on reproductive organs or fertility parameters. The observed implantation loss is considered to be a developmental effect and as such further discussion of this effect is in the ‘Developmental toxicity’ section.

Short description of key information:
Studies on AMP
Standard Guideline OECD 421 Reproductive screening study (rat, oral, diet)
Standard Guideline OECD 414 Developmental toxicity study (rat, dermal)
Mode of action studies similar in protocol to OECD 421
2 Whole embryo culture studies

Studies on CS1135 (4,4-dimethyloxazolidine)
2-Generation reproductive toxicity study
OECD Guideline 414 Oral Developmental toxicity study (rabbit, gavage)
OECD Guideline 414 Oral Developmental toxicity study (rat, gavage)

Effects on developmental toxicity

Description of key information
Studies on AMP
Standard Guideline OECD 421 Reproductive screening study (rat, oral, diet)
Standard Guideline OECD 414 Developmental toxicity study (rat, dermal)
Mode of action studies similar in protocol to OECD 421
2 Whole embryo culture studies
Studies on CS1135 (4,4-dimethyloxazolidine)
2-Generation reproductive toxicity study
OECD Guideline 414 Oral Developmental toxicity study (rabbit, gavage)
OECD Guideline 414 Oral Developmental toxicity study (rat, gavage)
Additional information

For the assessment of developmental toxicity there are a number of studies available for AMP-HCl, AMP base and for the read across substance 4,4-dimethyloxazolidine.

In the standard guideline OECD 421 study (Carney et al.,2005) where rats were dosed via the diet with either 0, 100, 300 or 1000 mg. kg bw AMP-HCl salt, there was no evidence of developmental toxicity outside of the previously mentioned post-implantation loss in the mid and high dose groups. In the embryos that successfully implanted in the mid dose there were no developmental effects observed and no difference in post-partum mortality. In the litters where part of the litter had been resorbed, the remaining pups were slightly heavier than control pups and this was likely due to the smaller number of pups being supported by the dams.

In an OECD guideline 414 dermal developmental toxicity study (Carney and Thorsrud 2006) where doses of up to 300 mg/kg bw/day of the AMP base were applied to the skin of rats from gestation day 6 to 20 there was no evidence of any post-implantation loss or developmental toxicity in any dose group. Considering the estimated dermal penetration of approx 40% (Saghir et al.,2007), the potential systemic dose would have been between 100 and 150 mg/kg/bw. Therefore the NOEL from these data are reasonably consistent with that available from the OECD 421 and two generation reproduction studies described below but show that the NOEL for post-implantation loss may be slightly higher than defined by the previous studies (i.e., ~150 mg/kg/day).

As indicated previously, 4,4-dimethyloxazolidine is a reaction product of formaldehyde and AMP. This compound breaks down in the gastrointestinal tract releasing AMP and formaldehyde. This hydrolysis is almost immediate and results in systemic exposure to AMP predominantly (Saghir et al.,2007). Subsequently, data produced using 4,4-dimethyloxazolidine can be considered as relevant for the assessment of AMP.

There are two developmental toxicity studies available on 4,4-dimethyloxazolidine, one in the rat and one in the rabbit (Nemec 1989; Rasoulpour and Marshall 2008a). In the rat study (a probe study), animals were dosed with 0, 250, 500, 750, 1000 or 1500 mg/kg bw/day via gavage from days 6-15 of gestation. All animals at doses 750 mg/kg bw/day or higher died. At 250 and 500 mg/kg bw/day there were no developmental effects observed and maternal toxicity was limited to findings relating to the irritancy of the test material to the gastrointestinal tract. In the rabbit study, animals were dosed daily by gavage from gestation day 7 to 27 with doses ranging from 0 to 40 mg/kg bw/day. Aside from evidence of gastric irritation and a significant decrease in maternal bodyweight in the high dose group, there were no developmental effects at any dose level.

In the two-generation reproductive toxicity study available for 4,4-dimethyloxazolidine (Carney et al.,2008). Excluding the postimplantation loss observed at the highest dose (200 mg/kg bw/day), there was no other evidence of reproductive or developmental toxicity. Animals at all doses developed normally through the F1 generation and were capable of successfully producing a second generation. This confirms that the toxicity observed at implantation following exposure to AMP does not appear to effect the subsequent development of surviving embryos. In this study the high dose of 200 mg/kg bw/day 4,4-dimethyloxazolidine was equivalent to approximately 180 mg/kg bw/day of AMP. Assuming that the AMP released was associated with the observed increase in implantation loss, the level of postimplantation loss in both the P1 and P2 generation was far lower than that observed in the mid dose of the OECD 421 study (approx 18% versus 70%). The degree of postimplantation loss was also consistent between generations indicating that the P2 generation was no more susceptible to the effect than the P1 generation (i.e. there was no ‘trans-generational effect’). The study also demonstrates that the longer dosing period prior to gestation in this study design did not increase the severity of the effect.

Interpretation of findings

In the available standard guideline teratogenicity studies there is no evidence of developmental toxicity (oral or dermal dose routes). However the reproductive screening study and the 2-generation study available on AMP-HCl and 4,4-dimethyloxazolidine respectively identified that exposure to these substances for a period prior to implantation can lead to reduced litter size or complete litter loss. The pathology of the uteri in affected animals indicated that the implantation loss was occurring early on in the pregnancy. Post implantation loss is generally considered to be a developmental effect, however since this effect was not identified in the standard developmental toxicity studies additional studies were performed to characterize the effect and assess its relevance to humans.

AMP is not directly embryo toxic

To understand if AMP was capable of directly inhibiting embryo growth, two whole embryo culture assays (Rasoulpour and, 2008a; Rasoulpouret al.,2010) were conducted. These assays showed that embryos cultured in serum containing AMP (at concentrations up to three-fold higher than measured in rats given 300 mg/kg/day AMP-HCl; 3, 9, or 27 μg/mL of AMP-HCl) or the serum of rats exposed to 300 mg/kg/day AMP-HCl for at least three weeks showed no signs of embryotoxicity and developed normally. These data indicate that under the conditions of these studies, AMP was not directly toxic to the developing embryo, nor did AMP appear to affect the ability of maternal serum to support an embryo. This is consistent with the absence of developmental effects in surviving offspring in the reproductive toxicity screening study and the two-generation study.

Since AMP is not directly toxic to the developing embryo then it can be concluded a maternally mediated effect is most likely driving the implantation loss.

Dose response assessment

Considering the data from the OECD 421 and the 2-generation study, the dose response for the post implantation loss appears to be very steep with no effect at 72 mg/kg bw, 18% implantation loss at approximately 180 mg/kg and 70% implantation loss at 216 mg/kg bw/day. The steepness of the dose response suggests that exposure to AMP triggers a specific threshold event that is responsible for the effect leading to the implantation loss rather than a generalized systemic embryotoxicity. It is also clear that extending the dosing period prior to gestation does not produce implantation loss at lower dose levels. Therefore there appears to be a “window of susceptibility” during which the triggering event must occur.

In order to verify if such a window of susceptibility existed, studies were conducted to determine exactly when the embryo death was occurring and also to identify whether there was a critical exposure window necessary for AMP to induce embryo loss (Rasoulpour and Zablotny, 2009). These studies also assessed differences between dietary and gavage routes of administration to determine whether the toxicity was greater following a bolus dose compared to diet.

In a study using a similar protocol to the OECD 421, a group of rats was exposed to 300 mg/kg bw/day of AMP-HCl for 2 weeks prior to breeding, during breeding and then up to Gestation Day (GD) 14. At GD 14 the animals were sacrificed and assessed for the number of successful/viable implantations. In a second study the same dosing regime was used, however animals were sacrificed at GD8, 10, 12. These studies identified that postimplantation loss initiated shortly after implantation (between GD 6 and 8). Therefore the effects of AMP were being expressed close to the time of, or at implantation, rather than a further stage in development. This suggests a potential incompatibility between maternal tissues and the implanting embryos.

In a subsequent study, groups of 12 female rats were dosed daily with 300 mg/kg bw/day AMP-HCl for different periods prior to and during gestation. One group was dosed via the diet from 2 weeks prior to breeding, through to gestation day 8. A second group was dosed via gavage for the same period. Subsequent groups were dosed via gavage from GD 1-8, 6-8 or 9-11. Animals were sacrificed at the end of their respective dosing period and assessed for number of viable implantation sites.

Dosing AMP-HCl via the diet from 2 weeks prior to breeding through to GD8 produced approximately 70% implantation loss, dosing via gavage for the same period produced approximately 40% implantation loss, and dosing from GD 1-8 produced approximately 14% implantation loss. These results were considered significant compared to the historical control range (between 3 and approximately 10%). Dosing from GD 6-8 and 9-11 did not result in an increase in the percentage implantation loss compared to historical controls. From this data it is clear that in order to produce an increase in post implantation loss AMP had to be administered for at least 8 days prior to implantation; a short dosing period around the implantation period (e.g., GD 6-8) was insufficient to cause postimplantation loss. Dosing after GD 6 had no effect on embryo viability.

These data lead to a conclusion that AMP itself is unlikely to be directly embryotoxic. It is more likely that AMP exposure for a sufficient period prior to implantation leads to physiological changes that reduce the capacity of the dam to support the implanting embryo. Taking into consideration the results of the two-generation study where the second generation had a similar level of postimplantation loss to the first generation it is also apparent that whilst a certain period of exposure prior to implantation is necessary (at least 8 days) increasing the exposure period prior to implantation from three weeks to thirteen weeks did not lead to an increase in postimplantation loss. Thus there is clearly a critical window during which a specific dose level must be administered in order to cause an effect. Dosing outside of this window does not produce any apparent effects on fertility or embryo survival.

In this study that examined if the timing of the dose was critical, it was also noted that dosing via gavage produced a less severe degree of postimplantation loss compared to the same dose and duration via the diet (approximately 40% postimplantation loss compared to 70%). The lower level of postimplantation loss following gavage dosing is unlikely to be just a random variation since in the mechanistic studies conducted to date, oral dosing of 300 mg/kg bw AMP-HCl via the diet has consistently resulted in a mean of 70% postimplantation loss. This reduced severity following gavage is unusual since it is more often the case that the bolus dose provided by gavage administration is more effective than dosing via the diet. The fact that the AMP data run counter to this trend indicates that Cmaxmay not be an important factor in the observed postimplantation loss, or that interactions in the gut when dosed via the diet play some role in the effect. For instance, AMP will exist in the small intestine in an ionized form (due to the pH in the intestine) therefore it is likely to be using some form of active absorption process since systemic bioavailability following an oral dose was approximately 100% (Saghiret al.,2007). Due to the similarity between choline, ethanolamine and AMP, it is probable that the active transport process for choline uptake in the gut is being used by AMP. This uses a sodium ion dependant carrier mechanism (Hegazy and Schwenk, 1984; Huerga and Popper, 1952) and appears to be saturable. Therefore when a bolus does of AMP is administered it is possible that the active transport mechanism becomes saturated and less AMP is absorbed.

Alternatively, it is possible that involvement in metabolic processes may also be responsible for the difference in effect following dietary exposure or gavage. Kinetic studies of AMP in rats (Saghiret al.,2007) have demonstrated that the majority (ca. 70%) of a single oral bolus dose is excreted in the urine, un-metabolized, within 24-48 hours. The excretion of the remainder of the dose (un-metabolized) takes up to 140 additional hours. This long beta-phase elimination kinetics is consistent with the knowledge that AMP may become incorporated into phospholipids within the liver, delaying excretion. Since the synthesis of phospholipids will be increased immediately following food intake, it is possible that dosing via the diet results in a greater incorporation of the AMP into phospholipids compared to dosing via gavage, which is done in the morning, several hours after feeding takes place. Dosing via the diet could also lead to additional interactions between AMP and dietary components such as choline or ethanolamine in the gut, reducing their uptake. Therefore a greater proportion of the dose would become incorporated into endogenous processes following dietary consumption compared to gavage. This indicates that non-dietary dosing is likely to be less effective at producing toxicity and this is important when considering the relevance of these effects to the non-oral exposure paradigm in humans.

Interactions with Choline and Postimplantation loss at levels that induce liver toxicity

Due to the structural similarity of AMP to the essential nutrient choline, it was hypothesized that AMP may be interacting with choline metabolism or uptake, perhaps leading to a depletion in maternal choline levels. In anin vitrostudy, AMP was found to inhibit the uptake of choline into CHO cells supporting this hypothesis (Stott and, 2006). Choline deficiency has been linked to fatty liver and increases in postimplantation loss since choline is a vital component of cell membranes, lipid transport and signaling molecules. Humans are less susceptible than rodents to choline deficiency due to greater capacity to synthesize it; therefore if the effects were due to choline deficiency then humans would likely be less sensitive.

Two studies were performed to assess the interaction with Choline and AMP and the significance for implantation loss. In the first study (Stottet al., 2006) the effect of AMP administration on the choline pools in the liver was assessed. Two groups of 6 female rats were fed either a control diet or a diet containing 300 mg/kg bw/day AMP-HCl for 2 weeks prior to breeding through to GD13, then sacrificed and examined for evidence of implantation loss. Samples of livers were either fixed for pathology or frozen and used to assess the levels ofcholine (Cho), betaine, glycerophosphocholine (GPCho), phosphocholine (PCho), phosphatidyl choline (PtdCho), and sphingomyelin (SM).

Post implantation loss was observed (67%) in the AMP-HCl treated rats in conjunction with evidence of fatty liver (hepatocyte vacuolation) and a decrease of between 20-25% in PCho. GPCho was elevated 2-3 fold compared to control dams indicating a conversion of PtdCho to Cho. All other metabolites were at similar levels to control. These data indicated that AMP treatment was capable of altering choline homeostasis in the pregnant dams. It was therefore considered possible that alterations in choline homeostasis could also be linked to the implantation loss. Subsequently a choline “rescue” experiment was performed to determine whether providing additional choline to AMP dosed animals had any effect on the degree of postimplantation loss (Rasoulpour and Ellis-Hutchings, 2009). Supplementing with choline (9000ppm in diet compared to 1800ppm) did not prevent postimplantation loss; however, it did ameliorate the degree of postimplantation loss compared to AMP treated animals (39% versus 66% loss). Thus whilst a simple choline deficiency was unlikely to be the key event leading to postimplantation loss, clearly AMP interactions with choline metabolism do play some role in the hepatic toxicity and may also be linked to the postimplantation loss. 

Relationship between hepatotoxicity and implantation loss

In both the OECD 421 and the two-generation study, AMP treated male and female rats had increased accumulation of lipid vacuoles in periportal hepatocytes at doses lower than those leading to postimplantation loss. There are a number of publications that have examined this effect of AMP on hepatocytes (Humeet al., 1965; Russellet al.,1965; Yueet al., 1970; Wells and Remy, 1961 and Akesson 1977) and it is hypothesized to be due to disruption of lipid transport out of hepatocytes subsequent to interference with phospholipid production, inhibition of ethanolamine and choline uptake into hepatocytes and inhibition of de-novo choline synthesis. The transport of lipids out of hepatocytes is dependent on the production of the phospholipids that package them, such as phosphatidyl choline and phosphatidyl ethanolamine. Interfering with the production of these subsequently inhibits the release of lipids from hepatocytes. Interference in phospholipid synthesis has been reported in other tissues such as swine coronary arteriesin vitro(Morin 1969), rabbit and human endometrial tissuesin vitro(Morin 1970), house fly larvae (Bridges and Ricketts, 1967), and murine fibroblastsin vitro(Schroeder 1980). The significance of these findings towards the reproductive toxicity of AMP is unclear since the studies used high doses of AMPin vitrothat would be very difficult to achieve systemicallyin vivoparticularly considering the irritancy of the AMP base. Thiseffect of AMP is the only other consistent effect observedin vitroand with repeated oral dosing of AMP to animals so it is plausible that it is related to the increase in postimplantation loss. This is supported to some degree by the research done into the importance of phospholipids in embryo implantation, particularly in relation to their spatial and temporal distribution during the implantation phase and their function as a source of fatty acids such as arachadonic acid (Burnumet al.,2009).

Whilst the available data do not allow a clear and direct association between the hepatotoxicity and reproductive toxicity to be made they do indicate a potential non-specific, physiological disturbance may be occurring in the tissues, and that the reproductive toxicity could be secondary to this. The available data also indicate that the liver is more sensitive to this effect than other tissues since effects on the liver were observed at doses lower than those producing effects on postimplantation.

AMP has a maternally mediated mode of action

In a final study (Rasoulpour et al., 2010), animals were treated with AMP-HCl at 300 mg/kg/day for two-weeks pre-breeding, through breeding and up to GD 6 (i.e., at the start of implantation). Extensive histopathology of the implantation sites and gene array data of the decidual swellings were generated. The purpose of this study was to examine the maternal/embryo interface at implantation in greater detail to identify whether effects in the uterus could be causing the implantation loss.

The histopathology of the implantation sites indicated the presence of vacuoles in the uterine cells immediately adjacent to the embryo in AMP treated dams, but not in controls. Staining with Oil Red O confirmed that these vacuoles did not contain lipids thus the nature of these vacuoles is unknown; however, since they appeared only in treated dams, they do indicate that AMP was producing a physiological effect within the uterine tissue. The severity of vacuolation varied across the uterus and the dams, with some implantation sites more heavily vacuolated than others. One issue with interpreting this finding is the lack of published data on what actually happens during an embryo resorption. For instance, this vacuolation could be a natural part of the resorption process, or alternatively it could be evidence of an adverse physiological effect that is causing a resorption to occur, nevertheless, the embryos in the implantation sites appeared completely normal indicating that the vacuoles precede resorption. In the repeated dose oral toxicity studies in dogs, rats and mice there have not been any reported effects observed in the uterine tissue indicative of a toxic effect. This lack of any observations in the uterine tissues in three species in the standard repeated dose protocols indicates that the uterus is not a specific target organ for toxicity, at least in the absence of pregnancy.

Gene expression arrays of the cells taken from decidual swellings from treated and control dams were analyzed to identify if there was any effect of AMP administration at the transcriptional level. The results of the analysis indicate that certain genes were less active in AMP treated rats compared to control. Specifically these genes (e.g., claudins and occludins) are associated with the formation of tight junctions. It is possible that this observation is associated with the postimplantation loss, particularly since these changes are occurring at the start of implantation rather than during or subsequent to the postimplantation loss. By influencing the formation of tight junctions, the decidual swelling surrounding the embryo could be made more ‘leaky’ and implantation sites may not be completely isolated from the maternal circulation. This, for example, could result in penetration of the maternal immune system into the implantation site leading to rejection and resorptions of the embryo.

When interpreting this data it is important to understand that this level of detailed gene array analysis has not been performed on other implantation sites where resorption of embryos is suspected. It is also not possible to state whether AMP directly or indirectly impacted gene transcription in the cells analyzed. If AMP was capable of directly interfering with the transcription of genes involved in tight junction formation it seems plausible that adverse effects in other tissues would be noticed, and this is not the case. Therefore it is not possible to concretely conclude that the lower gene activity observed is a consequence of AMP exposure or preparation for the resorption of an embryo; i.e. the changes in gene activity could be casually related rather than causally related to the observed effect. Further research would be needed to fully understand the relevance and importance of this information to the postimplantation loss observed.

Conclusions of mode of action experiments

Oral administration of AMP to rats prior to and during pregnancy is associated with an increase in postimplantation loss via a threshold, maternally mediated mechanism. The postimplantation loss occurs during or shortly after implantation (gestation day 6 to 8), and requires exposure to AMP for a period of at least 8 days prior to implantation via a dose route that provides significant systemic exposure (oral). There is a clear threshold for this effect and the dose response is steep, but very consistent (multiple studies using the same dose level have consistently produced the same degree of implantation loss). Considering the evidence of other toxicological effects at doses producing post-implantation loss (alteration in phospholipid synthesis in the liver leading to fat accumulation, gastrointestinal irritation) it is plausible that the post implantation loss could be secondary to a more general physiological effect in the dam rather than a specific embrotoxic effect. Taking this into consideration, AMP appears to indirectly affect the implanting embryo via a mechanism that involves maternal toxicity.

In a dermal developmental toxicity study there were no observed effects on any developmental parameter. The lack of postimplantation loss observed in that study is consistent with the lower systemic availability of AMP following dermal dosing, coupled with the shorter dosing period that covered gestation day 6 through to 20. However the highest dose that could be used in this study was limited by irritation to 300 mg/kg bw/day..Based on the applied dose, the maximum possible systemic dose from this study would have been approximately 100 to 150 mg/kg bw based on the absorption through the skin of between 30 to 40% AMP in the rat. Considering the significantly lower bioavailability via dermal dosing and potential need for oral exposure, even if this study had dosed for a longer period prior to gestation day 6, it is very unlikely post implantation loss would have occurred.

Thus this reproductive toxicity appears only relevant to those exposure routes where significant systemic availability is possible (oral). It is also important to note that dermal penetration data comparing human and rat skin demonstrated that penetration of AMP through human skin was less than half that observed in rat skin. As such exposure of humans dermally to AMP would be very unlikely to lead to sufficient systemic exposure to re-create the toxicity observed following oral dosing. Dermal exposure to AMP through industrial/professional/consumer uses would also be to the salt form (AMP is used as a neutralizer) and penetration of the salt through the skin will be lower than the base due to differences in lipophilicity, further reducing the potential for systemic exposure.

Justification for classification or non-classification

Summary of AMP induced toxicity

  • No data demonstrating developmental toxicity are available for AMP as the free base, rather all information comes from data generated with the hydrochloride salt of AMP or the4,4-dimethyloxazolidine. Pharmacokinetic data indicatethat upon ingestion, 4,4-dimethyloxazolidine is hydrolyzed to AMP (90% by weight) and formaldehyde in the stomach.Developmental studies conducted using AMP did not demonstrate evidence of developmental toxicity.
  • AMP HCl administration was associated with an increase in early post implantation loss. This effect has a well defined, and consistent dose response with a clear threshold, below which there is no developmental/reproductive toxicity even when the dosing period is increased significantly.
  • The reproductive toxicity of AMP HCl is only apparent in the presence of a more general toxicity that affected the physiological processes involved in phospholipid synthesis. It does not occur at doses where there are no effects observed in the liver indicative of disruption to phospholipid synthesis and lipid transport.
  • The effect is clearly maternally mediated and AMP is not directly toxic to the embryo or developing fetus after implantation. All surviving embryos develop normally through to sexual maturation and subsequent mating.
  • The post implantation loss has only been demonstrated in rats and there is no certainty that such an effect could occur in humans. AMP is excreted unchanged therefore differences in xenobiotic metabolism between humans and rats are therefore unlikely to make humans more vulnerable to such an effect.
  • The data generated to assess the reproductive toxicity of AMP have all used the oral route of exposure and a test material that has been neutralized to allow oral dosing. By altering the test material it was possible to dose far higher doses of AMP than would be possible if the AMP base itself was used instead. It is acknowledged that AMP acts as a pH neutralizer and as such dermal exposure to a lower pH formulation containing AMP is possible. However it is argued that whilst the data indicate a potential effect, the relevance of this effect to humans must be questioned since the oral exposure route is not considered a relevant route of exposure for this highly irritating substance or the neutralized form. Classification should take into account relevant routes of exposure to humans. In this case, irritancy and the lack of oral exposure scenarios reduces the relevance of orally generated data to humans. This does bring in an element of ‘risk’ into the traditional ‘hazard’ assessment; however it is important to label those hazards relevant to man rather than label for every technically possible hazard no matter whether they are relevant to human exposures or not.
  • Due to the irritancy of AMP it is not technically possible to dose sufficient AMP to produce the same effect via the dermal route. The pH will either limit the maximum dose that could be applied to the skin, or the fact that AMP is acting as a neutralizer and forms salts will significantly decrease its ability to penetrate the skin. The same would be true for inhalation, where the irritancy would also limit the exposure via a mist or aerosol exposure. Thus if a dermal or inhalation reproductive screening study were performed using AMP it is highly unlikely that any adverse reproductive effects would have been observed at the highest dose that could be tested. Therefore the relevance of this effect to humans is called into question considering the fact that AMP is used in applications where exposure will be limited to dermal or inhalation routes.


Conclusion on Classification

According to the guidance for classification for reproductive/developmental toxicity under the Dangerous substances Directive and the CLP regulation; where developmental toxicity is occurring only in the presence of maternal toxicity or by a specific maternally mediated-mechanism, where there are significant toxicokinetic differences that bring into question the relevance of the effects to man, or where the data were generated using a route of exposure not considered appropriate for humans, it can be argued that classification is not warranted. Considering the available evidence, AMP exposure via the oral route at doses greater than 100 mg/kg bw/day causes an increase in post implantation loss via a specific (yet not fully defined) maternally mediated mechanism. The relevance of this effect observed in rats to humans is unclear given the differences in absorption between humans and rats, the specific nature of the maternal effect, the lack of evidence from other species and differences between exposure routes in experimental studies versus human exposure. Therefore it is proposed not to classify this substance for reproductive toxicity at this time.




Akesson, B. (1977). Effects of analogues of ethanolamine and choline on Phospholipid metabolism in rat Hepatocytes.Biochem J.168, 401-408.

Bridges, R.G. and Ricketts, J. (1967). The incorporation,in vivo,of aminoalcohols into the phospholipids of the larvae of the Housefly,Musca domestica.J. Insect Physiol.13,835-850

Burnum, K.E., Cornett, D.S., Puolitaival, S.M., Milne, S.B., Myers, D.S., Tranguch, S., Brown, H.A, Dey, S.K and Caprioli, R.M. (2009). Spatial and temporal alterations of phopholipids determined by mass spectrometry during mouse implantation.J. Lip. Res.50, 2290-2298

Carney, E. W., Quast, J. F., and Zablotny, C. L. (2005). 2-Methyl-2-Aminopropanol Hydrochloride: A Reproductive / Developmental Toxicity Screening Test in CD Rats. Study number: DR-0038-3186-001.

Carney, E. W. and Thorsrud, B. (2006). AMP: Dermal Developmental Toxicity Study in CRL: CD(SD) Rats. Study number: DR-0399-3940-001.

Carney EW, Zablotny CL, Andrus AK,RasoulpourRJ, Quast JF, Yano BL. (2008). CS-1135: Two Generation Oral Gavage Reproductive Toxicity Study in CRL: CD(SD) Rats. Study number: DR-0365-7690-070.

Hegazy E. and Schwenk M. (1984). Choline uptake by isolated enterocytes of guinea pig.J. Nutr.114, 2217-2220,

de la Huerga, J. and Popper, H. (1952). Factors influencing choline absorption in the intestinal tract.J. Clin. Invest.31(6), 598–603.

Hume, J.W., Mulford, D.J. and Russell, P.J. (1965). Effect of 3-aminopropanol on choline deficiency in rats.Proc. Soc. Exp. Biol. Med.118(2), 430-433

Morin, R.J. (1969).In Vitroinhibition by metabolic antagonists of Phosphate-32P into the major phospholipids of swine coronary and pulmonary arteries.. Atheroscler. Res.10(3), 283-289

Morin, R.J. (1970). Inhibitionin vitroof [32P]-phosphate into rabbit and human endometrial phospholipids. Reprod. Fertil.23(3), 457-462

Nemec, M. D. (1989). A Range-Finding Teratology Study in Rats with AMINE CS-1135. Study number: DR-0365-7690-029.

Rasoulpour, R.J. and, V.A. (2008a); CS-1135: ORAL GAVAGE DEVELOPMENTAL TOXICITY STUDY IN NEWWHITE RABBIT. Study Number: DR-0365-7690-084

Rasoulpour R.J. and Marshall, V.A. (2008b). 2-AMINO-2-METHYL-1-PROPANOL HYDROCHLORIDE (AMP-HCL): WHOLE EMBRYO CULTURE IN CRL: CD(SD) RATS. Study number:DR-0038-3186-006.


Rasoulpour R.J. and Zablotny, C.L. (2009). 2-AMINO-2-METHYL-1-PROPANOL HYDROCHLORIDE (AMP-HCL): A DURATION OF EXPOSURE DIET AND GAVAGE STUDY IN Crl: CD(SD) RATS. Study number: DR-0038-3186-007


Russell, P.J., Green, S.T. and Mulford, D.J. (1965). The effect of 2-amino-2-methylpropanol on rat-liver mitochondria.Biochem. Biophys. Acta.98(3), 445-453

Saghir,,, A.J., McClymont, E.L. and Staley, J.L, (2007); CS-1135: PHARMACOKINETICS AND METABOLISM IN FISCHER 344 RATS.Study number:DR-0365-7690-066

Schroeder, F. (1980). Regulation of aminophospholipid asymmetry in murine fibroplast plasma membranes by choline and ethanolamine analogues.Biochim. Biophys. Acta.599(1), 254-270

Stott, W. T. and, H. L. (2006). 2-Methyl-2-aminopropanol (AMP): Effects Upon Choline Uptake in Chinese Hamster Ovary Cells. Study number: DR-0038-3186-002.

Stott, W.T., Zablotney, C.L. and Thomas, J. (2006). Potential of AMP-HCl to alter hepatic choline and phospholipid synthesis in pregnant CRL:CD(SD) rats when administered by the dietary route. DR-0038-3186-004   

Wells, I.C. and Remy, C.N. (1961). Inhibition ofde novocholine synthesis by 2-amino-2-methylpropanol.Arch. Biochem. Biophys.95, 389-399

Yue, K.T., Mulford, D.J. and Russell, P.J. (1970). Metabolism of 2-amino-2-methylpropanol n young rats.Arch. Biochem. Biophys.136(1), 47-53

Additional information