The present results of the acoustic startle reflex paradigm are in agreement with previous findings from other anxiety models [8–10], which indicated that SHR rats display less emotional reactivity than LEW rats. Moreover, SHR rats showed decreased prepulse inhibition when compared to LEW rats. The most important finding of the present study, however, was that the locus Ofil2 on rat chromosome 7  affected significantly the PPI levels in female rats without affecting their startle responses. Furthermore, neither Ofil1 nor Ofil 2 affected blood pressure thus confirming our previous studies [9, 11].
The startle reflex, a fast twitch of the body musculature triggered by a sudden and intense acoustic stimulus, is frequently used to investigate the neurobiology of anxiety and fear [16, 22]. The acoustic startle response is mediated by a relatively simple neuronal circuit located in the lower brainstem, but this primary pathway receives projections from higher structures that are known to control defensive behaviors [16, 22]. In line with these observations, in the present study, LEW rats (considered as more "anxious") showed increased acoustic startle responses in relation to SHR rats (less "anxious"). Other studies have reported that SHR rats showed reduced startle response to an auditory stimulus when compared to most other strains [25, 26]. It is noteworthy that whereas female LEW rats showed habituation in startle responses over acoustic trials the female SHRs did not habituate. Short-term habituation of startle is impaired in some neuropsychiatric disorders. For example, Braff et al.  reported that schizophrenic patients have extensive deficits in both PPI and acoustic startle habituation. Moreover, Grillon et al.  reported that startle magnitude was elevated in children with a parental history of anxiety disorder, whereas startle habituation and PPI were impaired in children with a parental history of alcoholism. Curiously, it has been reported that SHR rats show less anxiety-related behaviors [8–10] and higher alcohol intake than LEW rats [24, 29].
In the PPI test, the startle magnitude is reduced when the acoustic stimulus is preceded by a non-startling prepulse and thereby is considered as an operational measure of sensorimotor gating . In humans, impaired PPI has been associated with disorders of uncontrolled behavior, notably schizophrenia  and ADHD [18, 19]. In rodents, deficits in PPI can be induced with the administration of various pharmacological agents, such as apomorphine, D-amphetamine and phencyclidine, and normalized by antipsychotic drugs . We found that the SHR strain, considered as an animal model of ADHD [6, 7], showed significantly decreased levels of PPI when compared to the LEW strain. However, other studies [26, 30] reported that SHRs, compared to either Wistar-Kyoto or Sprague-Dawley rats, did not show any deficit in PPI. These findings, together with the present results, suggest that the LEW strain may show elevated levels of PPI in comparison not only with SHR but also with other strains. In any case, this behavioral contrast between LEW and SHR rats in PPI responses provides a useful tool to study the genetic mechanisms underlying sensorimotor gating. To our knowledge, no direct comparison between these two strains has been carried out previously. Since there is evidence that SHR rats may habituate more slowly to novel environments we cannot rule out this issue as a potential confounding factor in the behavioral differences observed between LEW and SHR rats.
It is puzzling that PPI-deficits can be induced by amphetamine in rodents and humans  while methylphenidate, another dopamine releaser drug, can ameliorate the symptoms of ADHD. Hawk et al.  recently reported that methylphenidate selectively increased PPI among boys with ADHD to a level comparable to that of controls. Very recently, Yamashita et al.  reported that a high dose of methylphenidate (60 mg/kg) significantly impaired PPI in wild-type mice but the same pharmacological treatment significantly reversed PPI deficits in dopamine transporter knockout mice. Thus, it should be of interest to test LEW and SHR rats treated with methylphenidate in the PPI test.
It is important to emphasize that SHR rats showed similar levels of startle reflex and PPI in response to either acoustic or tactile stimuli (airpuff) . These findings suggest that the differential startle responses and PPI between LEW and SHR rats observed in the present study are more likely to be related to their behavioral reactivity rather than to acoustic acuity. Current opinion generally assumes that startle amplitude and PPI are independent variables that are under different genetic control [16, 32]. Consistent with this idea, we found no statistical correlation between startle and PPI responses (data not shown), suggesting that the PPI differences are not likely produced by differences in startle magnitude.
Regarding the influence of Ofil1 and Ofil2, we have found previously  that they had a female-specific effect on central locomotion in the open field. The effect of Ofil1 was inverted as compared to the parental strains (i.e. LEW alleles promoted more instead of less central locomotion in the open field), whereas Ofil2 acted in the expected direction (i.e. LEW alleles reducing the trait). In a subsequent study, Mormède et al.  confirmed the role of these loci on behavioral responses by producing two rat lines with extreme genotypes for Ofil1 and Ofil2. It was found that the Low line (corresponding to S4/L7 rats in the present study) was less active than the High line (corresponding to the present L4/S7 rats) in the center of the open-field. This inhibition was not attributable to a classical "anxiety" factor as measured in the elevated plus-maze, in which the open-arm behaviors were not different between the lines. The High line also showed a deficit in PPI responses, suggesting that Ofil1 and Ofil2, which had been previously described as being related to anxiety, were indeed involved in sensorimotor gating mechanisms . This study, however, did not allow us to sort apart the respective influences of these two loci because the effects of Ofil1 and Ofil2 were combined in only two rat lines. Recently, we have found that Ofil1, but not Ofil2, affected central locomotion in the open field in females . Herein, it was found that females carrying two LEW alleles at Ofil2 showed increased PPI levels when compared to females carrying two SHR alleles at this locus. Therefore, it is possible to conclude that the PPI effects observed in the study by Mormède et al.  were likely produced by Ofil2 and not Ofil1. These findings further suggest that central locomotion in the open field and PPI responses are genetically dissociated in this specific model. It is important to note that the startle reflex was not affected by Ofil2. Furthermore, the PPI genetic effects reported herein are consistent with the profile of the parental strains, i.e. SHR alleles reducing the trait and LEW alleles increasing it. Finally, an interaction between Ofil1 and Ofil2 (i.e. epistatic effect) was also found for the acoustic startle reflex. The L4/S7 and S4/L7 lines displayed increased startle response in relation to L4/L7 and S4/S7 lines. These findings demonstrate the complexity of genetic influences on behavioral traits, probably involving gene-gene, gene-gender and gene-environment interactions.
Swerdlow et al.  reported that Sprague-Dawley rats displayed increased disruption of PPI responses by apomorphine, a dopaminergic agonist, when compared to Long Evans rats. Moreover, strain differences were observed in the efficacy of dopamine D2-like receptor-G-protein in specific brain areas. Correlational analysis revealed that in the striatum, cingulum, and cortex, greater dopamine-stimulated G-protein binding predicted less sensitivity to the PPI-disruptive effects of apomorphine . As pointed by the authors, genes that regulate these differences may contribute to identification of heritable differences in patients with specific neuropsychiatric disorders. To date, only one study identified QTLs for PPI in rats. Palmer and collaborators  found a significant QTL on chromosome 2 and another suggestive QTL on chromosome 18. Furthermore, Joober et al.  reported several provisional QTLs specific for PPI on mouse chromosome 3, 5, 7 and 16. Very recently, Petryshen et al.  reported two QTLs on mouse chromosome 16. Therefore, the present study provides additional information by reporting a sex-specific QTL for PPI on rat chromosome 7. This region contains the Syn3 gene, which encodes Synapsin III (see http://www.ncbi.nlm.nih.gov/mapview/map_search.cgi?taxid=10116). Synapsins have been implicated in synaptogenesis and in the modulation of neurotransmitter release . Since the human synapsin III gene is a potential locus for schizophrenia , we considered Syn3 as a candidate gene for Ofil2.
In conclusion, the present study confirmed the importance of sex and genetic factors on behavioral responses. The LEW and SHR strains, in addition to being useful in the study of anxiety-related behaviors, also seem appropriate for studying mechanisms of sensorimotor gating. Moreover, we identified a female-specific QTL on chromosome 7 that modulates PPI in rats, a neuro-behavioral trait found in ADHD (and in other psychiatric disorders). Further dissection of this locus should give information about sex and molecular mechanisms influencing disorders of uncontrolled behaviors and thereby open up new perspectives for neuropsychiatric therapies in humans.