Results of this study reinforce the concept that the developmental effects of OPs extend beyond the nervous system. Neonatal exposures to DZN and PRT altered the developmental trajectory of AC-mediated cell signaling in peripheral tissues, and to a greater extent in the liver than in the cerebellum. Although, in general, OP exposure elicited a net AC gain-of-function, the effects differed among the various tissues, effectively ruling out a global effect on expression of receptors, G-proteins, or AC and instead pointing to selectively greater effects on specific aspects of hepatic function. This conclusion was further reinforced by the fact that DZN and PRT exposures evoked different trajectories despite doses of each OP chosen to be toxicodynamically equivalent (Slotkin et al. 2006b
): The effects of DZN intensified with age, whereas those of PRT waned.
In our previous work with CPF, we found persistent, global increases in all measures of hepatic AC signaling, an effect restricted to males (Meyer et al. 2004
); this connotes global sensitization of the pathway, wherein up-regulation of AC activity itself (enhanced forskolin response) produces an augmented response to stimulants acting on Gs
-coupled receptors (isoproterenol, glucagon), as well as to those directly activating G-proteins (fluoride). Equally important, the CPF effects displayed a critical period of sensitivity, indicating that these are specifically developmental actions (Meyer et al. 2004
). As found in the present study, neonatal DZN exposure produced a similar global effect that appeared in adolescence and young adulthood at the higher dose but that also became significant at the lower dose by full adulthood; notably, although the effect emerged first in males, it eventually encompassed both sexes, thus differing in outcome from the sex-specific effects seen for CPF (Meyer et al. 2004
). We also found alterations in specific elements of the signaling cascade superimposed on global sensitization. If the up-regulation of AC itself were the only effect of DZN exposure, then all pathway stimulants should show the same degree of enhancement. Instead, at both PND60 and PND100, the response to fluoride was augmented to a significantly smaller extent than that for forskolin or either of the receptor stimulants. Fluoride differs from isoproterenol and glucagon in that it also activates Gi
, and consequently, our findings point to an increase in expression and/or function of Gi
after neonatal DZN exposure; again, this differs from the effects of CPF, which produces uniform enhancement of responses as expected from activation of AC itself (Meyer et al. 2004
). Effects on Gi
will also produce global alterations in AC signaling because they affect the response to any receptor acting through Gi
. Thus, all the effects we noted involved the signaling proteins downstream from the receptors, either AC itself or the G-proteins, rather than reflecting effects on the expression of the neurotransmitter receptors or their specific coupling to the control of AC. Indeed, none of the small effects seen for receptor binding could account for the robust augmentation of AC responses.
In contrast to DZN, neonatal PRT exposure produced a much larger initial sensitization of hepatic AC in adolescence, again involving global changes at the level of AC itself, but additional pathway effects were evident at the lower dose, reflecting smaller increments for glucagon and fluoride than for isoproterenol and forskolin. In turn, these effects imply homologous desensitization for glucagon and heterologous increases in Gi-mediated inhibition; the effect of PRT on the glucagon response was not shared by DZN, again pointing out specific differences related to each individual OP. Further, the effects of DZN intensified over time, whereas those of PRT waned. Thus, the two OPs differ completely in their effects on the developmental trajectory of hepatic AC signaling, with large effects in adolescence for PRT but not DZN, supplanted by the opposite pattern in adulthood. The fact that AC signaling in the liver was affected far more than in either the heart or cerebellum further demonstrates selective effects of early-life OP exposure. Finally, for the heart (but not the cerebellum), we found significant changes in receptor binding without apparent connection to the effects on AC signaling, again reinforcing the importance of sensitization downstream from the receptors as the primary site of regulatory disruption.
The clear implication is that neonatal OP exposure is likely to affect hepatic responses to a greater extent than those in the heart or in the central nervous system, and it is therefore critical to examine how the cellular changes seen here might then contribute to alterations in tissue function. In the liver, where βARs and glucagon receptors are linked through AC to enhanced gluconeogenesis and lipolysis, global sensitization of AC signaling leads to corresponding metabolic abnormalities. In our earlier work with CPF, we established the presence of hyperlipidemia, but serum glucose levels were maintained within normal limits (Slotkin et al. 2005
); however, glucose homeostasis was maintained only by compensatory hypersecretion of insulin, thus producing a metabolic profile akin to prediabetes. Critical to the proposed mechanistic connections, the sex selectivity of the metabolic effects (males) exactly matched that for the sensitization of hepatic AC signaling (Meyer et al. 2004
). More recently, we also found evidence of prediabetes after neonatal PRT exposure (Lassiter et al. 2008
). In this case, there was no corresponding increase in serum insulin, and as a result, the animals displayed a frank prediabetic profile, characterized by hyperglycemia and impaired glucose and lipid utilization (Lassiter et al. 2008
). It is thus important to note that the effects of PRT on hepatic AC signaling seen here were restricted to adolescence and, unlike those of CPF, did not persist into adulthood. Accordingly, for PRT, either the AC changes are unrelated to the metabolic disorders, or the effects in adolescence may be sufficient to reprogram metabolism so that defects emerge later, despite the subsequent normalization of signaling parameters. Future work will need to dissect the temporal emergence of prediabetes after neonatal PRT exposure in order to distinguish these two possibilities. However, the present data point toward the latter interpretation, because PRT-exposed animals show a switch from enhanced to suppressed weight gain coinciding with the time course for the disappearance of the effects on AC signaling (Lassiter et al. 2008
). Although detailed metabolic studies have not been done for DZN, based on the results for hepatic AC signaling, we would expect to see greater metabolic defects than for PRT, consistent with the greater weight loss seen here and in earlier studies (Lassiter et al. 2008
; Roegge et al. 2008
); we further predict that, unlike CPF, DZN will target metabolic function in both males and females because the cellular effects were not sex selective. Finally, human studies suggest a connection of diabetes to long-term OP exposure (Morgan et al. 1980
; Saldana et al. 2007
) and a link between gain-of-function AC gene polymorphisms and diabetes susceptibility (Nordman et al. 2008
). Our results thus provide a mechanistic underpinning for these population studies. Nevertheless, it is clear that actions on the hepatic AC pathway are likely to be just the tip of the iceberg for the metabolic abnormalities wrought by early-life OP exposure. In our earlier work, we showed how additional stresses imposed by elevated dietary fat intake can reveal defects that likely originate in adipocyte dysfunction rather than in hepatic metabolism (Lassiter et al. 2008
), and it would be worthwhile to pursue these additional contributions to prediabetes.
There are similar implications for the significant, albeit lesser, effects of DZN and PRT exposure on AC signaling in the heart and cerebellum. Transgenic animals that produce AC hyperstimulation through overexpression of βARs or Gs
show development of cardiomyopathies and abnormal heart rate regulation (Gao et al. 2003
; Iwase et al. 1996
). Alterations in m2
AChR expression, which we detected for PRT, also have corresponding effects on cardiac function and the response to autonomic input (LaCroix et al. 2008
). Importantly, the fact that DZN reduced heart weight significantly and to a greater extent than body weight indicates the need to pursue potential consequences for cardiac function, which have been much less studied than hepatic function. Indeed, the prediabetic changes seen after neonatal OP exposure are themselves likely to contribute to further cardiovascular morbidities. In the cerebellum, we found a nonmonotonic effect of DZN, with enhanced AC signaling at the low dose that disappeared or was reversed at the high dose; this likely represents the positive neurotrophic effect of acetylcholine produced by a small degree of cholinesterase inhibition, offsetting the direct effects of low exposure (Timofeeva et al. 2008a
). In contrast, PRT produced up-regulation with a typical, monotonic dose–effect relationship. Most important, the disparate dose–effect patterns of effects on AC signaling for both DZN and PRT correspond to the differences in behavioral findings between the two agents (Timofeeva et al. 2008a
Our results support the view that developmental exposure to OPs targets the trajectory of AC signaling in peripheral tissues, thus extending their actions outside the nervous system, with the consequences that a
) the effects on signaling occur with nonsymptomatic exposures, and b
) effects differ among OPs even at toxicodynamically equivalent exposures as assessed at the level of brain cholinesterase inhibition (Slotkin et al. 2006b
). Further, the liver appears to be especially sensitive to persistent disruption of AC signaling, involving sensitization of the entire pathway via induction of AC activity. In turn, this provides a likely mechanism for the metabolic consequences of neonatal OP exposure identified in earlier studies, indicative of a prediabetic state (Lassiter et al. 2010
; Slotkin et al. 2005
). Indeed, global sensitization is likely to have widespread consequences in general because it affects all humoral signals that operate through cAMP, rather than involving any single input. Future studies will need to address how the tissue-selective changes in cell signaling come about, that is, whether they involve a direct impact of the OPs on expression or function of the signaling proteins, whether they are downstream events secondary to effects on differentiation of hepatocytes and myocytes, or whether they originate in upstream effects on hormonal or neuronal input. In any case, our findings extend the Barker (2003)
hypothesis, which originally related prenatal growth restriction to subsequent development of cardiovascular disease and diabetes, to include otherwise nonsymptomatic chemical exposures that may produce similar outcomes without the precondition of fetal/neonatal growth restriction. Our findings point out the need to explore the possibility that developmental exposure to common chemical contaminants contributes to the explosive worldwide increase in diabetes and obesity.