Highly outbred male and female Wistar rats of 70 - 72 days of age (respectively weighing 290 - 310 g and 170 - 185 g at the beginning of the experiment) were provided by our own animal facility at the University of Salamanca. Rats were housed and maintained on a 12:12 hour light: dark cycle (lights on at 8 am) in a room with controlled temperature.

For all experiments, the animals were allowed access to food and water ad libitum, and were maintained on a regular light-dark cycle (lights on: 07:00am - 19:00pm) with constant temperature (21˚C). The animals were handled and cared for according to the guidelines of the European Community’s Council Directive (2010/63/CE) and current Spanish legislation for the care and use of laboratory animals.

The acoustic startle reflex was measured in six identical startle-response cages, using the SR-LAB system (SDI, San Diego, CA, USA). Acoustic stimulus intensities and response sensitivities were calibrated (using an SR- LAB Startle Calibration System) so that they would be nearly identical in each of the six SR-LAB systems (maximum variability < 1% of stimulus range and <5% of response ranges). Each testing session consisted of an acclimatization period of ~5 min, followed by 64 trials presented pseudo-randomly, with a mean inter-trial interval of 30 s, as previously described [47] . Briefly, the sessions had four blocks of pulse and prepulse, with prepulse-to-pulse intervals of 50ms. Whole-body movements corresponding to startle responses were recorded and analyzed with the SR-LAB system, providing ASR latencies and amplitudes. The background noise of 65 dB SPL was generated throughout the entire session in order to avoid interference from external noise and to ensure equal experimental conditions. As depicted in the experimental design (Figure 1), four sessions of startle and sensory motor gating were performed in all animals. Before testing, the rats were habituated to the experimental conditions, especially regarding their introduction into the ASR device. All testing was carried out between 10:00 and 13:00 hours [48] .

In the first experiment, the effect of the stress was indexed by (Average ASR2/ Average ASR1) × 100) − 100), giving the percentage (%) of change in each individual. In the second experiment, long-term change (%) (LTC) in ASR amplitudes as a result of repeated testing were indexed by (Average ASR4/ Average ASR1) × 100) − 100, with body weight as a covariate. Thus, a positive and negative LTC respectively indicate the sensitization or habituation of the startle [11] . Immediately after each session, all animals were weighed and all females were subjected to estrous determination.

Vaginal smears were obtained by dipping a sterile swab (0.6 mm diameter, 0.025 in Fischer Scientific) in sterile saline, and then gently swabbing the vaginal lumen. The swabs were smeared onto labeled glass slides that were previously cleaned with 95% ethanol. The cells were fixed with 95% ethanol for 15 minutes and then air-dried before staining with haematoxylin-eosin. The vaginal smears were inspected and the phase of the estrous cycle was determined using an Olympus Microscope (×40), and following previous criteria [49] , a 4 to 5 day-cycle was considered (proestrus, estrous, metestrus and diestrus). Rats in proestrus with the highest estrogen levels (with high number of nucleated epithelial cells); by contrast, rats in estrus with the lowest estrogen levels (with anucleate cornfield cells) [24] [49] ; rats in metestrus with high number of white blood cells (WBC) and nucleated cornified cells; and rats in diestrus having some epithelial cells and still a predominance of WBC. Females were later distributed in order to have an equal number of animals in each phase in each group.

The day after the first startle/PPI session, animals of both sexes were placed in an isolated room, where they were distributed to either the restraint stress paradigm (N = 21 females and N = 20 males) or were left undisturbed (served as controls in the first experiment) (N = 19 females and N = 22 males). The rats exposed to restraint (RS) were placed daily in small transparent Plexiglas cylinders, three times a day for 45 minutes along 7 days. This device limited their body movements (the rats were not able to move forward or backward) and was positioned directly under a bright light. The length and diameter of the cylinder were based on body size, smaller-diameter cylinders being used in females than in males (70 vs. 80 mm), all with adjustable endplates.

On the day after the second behavioral session (ASR/PPI 2) and two days after completion of the RS period, all animals of both sexes were subdivided into six subgroups according to the further treatment to be subjected (see Figure 1).

SERT-receiving rats were given Sertraline (Besitran^© Pfizer S.A. Madrid, SPAIN) intraperitoneally at 5 mg/kg/day once daily (i.p.) for 8 days. Sertraline was dissolved in 0.9% NaCl vehicle and were administrated in a dose of 1 ml/kg. The dose was based on previous reports [50] . Vehicle-injected animals (Control + Veh and RS + Veh) were given 0.9% saline solution in the same manner. These animals were used in order to control for the stressful effect of the intraperitoneal (i.p.) administration route. Also, two groups of non-injected (RS and Control) animals of both sexes were left undisturbed from the first to the last day of the i.p. procedure, and were used as a control of the experimental conditions.

On the day after the completion of the behavioural tests, all animals were exposed to immobilization stress (IMO), a procedural variation of restraint. The same cylinders that had been used previously were adjusted to achieve total restriction of movements, taking care not to hurt the animal [51] . Each rat was placed in the cylinder for 30 min, after which it was placed in a new cage and allowed to rest for 10 min. Then, the animal was anesthetized and trunk blood was taken by cardiac puncture [50] .

This procedure was carried out simultaneously on three animals from 10:00 h to 13:00 h. A group of undisturbed animals (No-IMO) of the same age was also used in order to achieve the effects of IMO stress. In order to reduce the inherent effects of differences in the estrous cycle, only females that were not in proestrus were used [52] .

For hematological analyses, trunk blood was collected to EDTA (K3)-containing tubes that were freshly proc- essed on an automatic cell counter (ADVIA 120 cytometer, Bayer, Leverkusen, Germany). The obtained hematological parameters were number of erythrocyte, hemoglobin concentration, mean corpuscular hemoglobin (MCH), hematocrit values, platelet number and mean platelet volume (MPV), and the number and type of leukocytes (WBC).

For biochemical analyses, blood samples were collected into heparinized tubes, which were then centrifuged at 10,000 × g for 20 min to obtain serum, which was drawn into Eppendorf tubes and used fresh on a SPOTCHEM^TM EZ device (QBC Europe). The levels of serum total protein (T-Pro), albumin and bilirubin, and the cytosolic enzymes glutamic-oxaloacetic transaminase (GOT), glutamic-pyruvic transaminase (GPT), and lactate dehydrogenase (LDH) were measured using the commercial kit (ARKRAY®) according to the manufacturer’s directions.

Statistical analyses were performed using IBM^® SPSS^® software, version 20 (IBM Crp. and SPSS Inc., Chicago, IL, USA, 2011). Differences between groups were analyzed by ANOVA (one, two and three way), followed by the Fisher-PLSD-test for post hoc comparison if appropriate, and ANOVA mixed (or “SPLIT-PLOT”) with the Bonferroni-test. Mean differences were subjected pairwise to Student’s t-test, using the Levene Test for equality of variances. Pearson’s coefficient was used to determine correlations. Differences were regarded as statistically significant when p ≤ 0.05.

This experiment aimed at obtaining preliminary data regarding the effects of one week of restraint stress on ASR and PPI, measuring the data one day after completion of the stress protocol.

This experiment had two main objectives: to examine the effects of one week of restraint stress 10 days after completion of the stress protocol, and to examine the effects of repeated injections (either vehicle or SERT) in previously stressed animals.

Given the large differences found in the behavioral parameters of each sex in response to each type of stressful condition and to SERT treatment, in this experiment our main goal was to extend the results by comparing the effects of each paradigm on later stress responsiveness.

Thus, one day after the last test ended the animals of both sexes subjected to each type of paradigm-restraint stress, i.p. injections with SERT or vehicle, or controls―were exposed to a new stressor, IMO. Blood samples were extracted and compared with those of a group of undisturbed animals (NO-IMO).

Our data show that one-week of restraint stress was sufficient to affect growth in young rats, an effect that was more marked in males. It is assumed that the decrease in body weight is a good physiological marker of stress, and hence the intensity and duration of the stressor would be determining factors in the results. With more intense stressors, such as IMO for 14 days [53] or 8 hours restriction/day plus the use of a variable stress paradigm [3] , more spectacular results have been achieved. However, in the present work we failed to detect significant differences in weight gain in the restrained animals subjected later to the i.p. procedure or in those treated with SERT.

It is well known that the startle response is very sensitive to stress and anxiety both in humans and animals [11] [23] [52] [54] . In the present work, our results indicated there were no differences in ASR amplitude means in either sex as a function of stress paradigm or SERT treatment. However, we found that long-term habituation was disrupted as an effect of RS, in males.

In control animals of both sexes, startle amplitude decreased slightly each time it was examined (data not shown), indicating that a kind of habituation had occurred [11] . Previous works have confirmed the existence of acoustic startle habituation both between consecutive trials during a single session [11] [23] [36] [50] and between the ASR sessions [13] [55] . Nevertheless, ten days after the end of the restraint stress procedure, the RS males exhibited an increase in startle amplitude (when compared to the baseline). These results are partly in agreement with what has been reported in humans with psychopathological symptoms [19] [23] . These authors reported that their patients did not become habituated easily during the startle probe, increasing startle or maintaining it at the same levels in each exposure to the test.

When we analyzed PPI we observed a significant increase in PPI up to an age of 12 weeks in the animals of both sexes regardless of treatment, as we previously found in our laboratory, in males (unpublished data). At that moment, in the control animals PPI remained stable among sessions, as expected [36] [55] . According to the literature, stress affects PPI. PPI may be enhanced by fear or in response to emotional conditions [14] [55] [56] . Our data show that, alone, one week of restraint stress was not sufficient to induce changes in PPI. Also, a single i.p. injection with vehicle had no effects on either the startle reflex or PPI. However, when the animals were exposed to repeated injections with saline, a significant increase in PPI was observed in both sexes, an effect that was counteracted by SERT, although more efficiently in females. The PPI enhancement after daily injections could be interpreted in the sense that this injection procedure might have facilitated attention and vigilance after the animals had been exposed to an adverse condition [14] [55] [57] . In fact, other authors have reported that epeated injections [58] , or even blood sampling [52] [59] , are sufficient to increase corticosterone (CORT) levels (an indicator of stress response) and to increase the startle reflex (in the first presentation trials) [58] and PPI in rodents [52] . However, although it has been reported that changes in the environment are sufficient to enhance PPI in both humans and rodents [18] [55] [57] [60] ―in humans PPI may even be increased if the prepulse stimulus has an affective component [60] ―the route through which the drugs are supplied is not normally taken into account and its aversion value is not considered [13] [35] [37] [58] .

Furthermore, the importance of taking the time to respond to the startle stimulus as an adaptive ability has recently been reported [61] . This is meaningful, considering the protective function of the startle reflex; when animals are first challenged by unexpected loud noises, a faster startle response is expected because the environment is potentially unsecure. Then, after repeated and inconsequential encounters with the same stimulus sequences, the startle should slow down, due to habituation. With repeated exposure to the test sessions (from ASR1 to ASR4) we found that whereas restraint stress did not affect ASR or PPI latency in either sex, the procedure of repeatedly injecting the rats induced a significant decrease in latencies. These animals speeded up their startle response, in contrast to habituation. This suggests that they became more reactive and fearful [61] , an effect that SERT-pretreatment totally reversed in females but not in males.

In comparison with the amount of work conducted on the effects of stress and their modulation in startle and sensorimotor gating mechanisms in males, the respective literature addressing female responses is fairly limited. Here, we found that sex was a major factor affecting the results.

At the beginning of the experiment, the males exhibited higher startle amplitudes and PPI levels than the females, in agreement with the previous data [18] . In contrast to other authors [62] [63] , Lehmann’s group described the existence of consistent differences between male and female Wistar rats in sensorimotor gating mechanisms [18] . It is now well established that these differences also exist in humans [36] [64] .

The fact that males reach higher startle amplitudes than females is evident, since in rats the ASR is measured based on acceleration [18] . With regard to PPI, this physical law does not apply since the PPI is calculated as a percentage of reduction in ASR and not as an absolute value. Accordingly, body weight is not a determining factor. In fact, our data show that when the animals reached adulthood (at 12 weeks of age) the differences in startle habituation and PPI values between the sexes were no longer found between controls, but were found among the animals that experienced stressful paradigms. Apparently, females were somewhat more sensitive to the injection procedure and males were more sensitive to restraint. To the best of our knowledge, the differences in our data concerning startle and PPI modulation due to the type of the stressor in each sex have never been considered. Previously, some studies reported different results with physical vs. emotional stressors [14] [56] ; however, most of those studies were conducted in males.

The reported sex-differences could be due to changes in reproductive hormones, in particular estrogen (E2) [4] [62] . Nevertheless, when the differences at E2 levels throughout the estrous cycle were considered in females [49] , neither PPI nor the startle values were changed, as we did not find any correlation between E2 levels and PPI, or between E2 levels and startle at any time point. Moreover, recently, it has been reported [48] [65] , the estrous cycle had no effects on ASR amplitude and PPI when tested in random phases of the estrous cycle.

When we examined the physiological analyses, whereas no sex differences were found in the biochemical parameters or leukogram, some differences between the males and females were found in the haemostatic response to stress. In agreement with our results whereas under basal conditions, no differences between the sexes in chemical or hematological parameters are expected [66] , several authors have reported differences in the stress response [42] [67] - [69] .

According to the literature, the startle reflex and its modulations are sensitive to 5-HT levels and to the administration of SSRIs, such as SERT [34] [35] [37] . In the present study, SERT did not act in control animals; but in the animals exposed to stressful conditions SERT counteracted the effects of stress, although more efficiently in females. Results were only seen after 8 days of SERT administration; a single administration of SERT did not change any of the behavioral measures. This agrees with previous works, in the sense that antidepressants (such as SSRIs) may normalize or stabilize serotonin function and restore stress-induced behavioral changes [1] . Only repeated treatment with SSRIs affords the desired effects by blocking the harmful effects of stress [29] [30] [54] [70] [71] and no effects have been reported with acute treatment [35] . Nevertheless, in our study, the effects of SERT in modulating startle and PPI were different according to sex; SERT treatment was slightly more effective in female rats than in males. In fact, sex differences in 5-HT levels have been reported in many structures of the corticolimbic region known to regulate stress and emotion processing [72] [73] . Moreover, the spontaneous firing rate of dorsal raphe neurons (DRN) in males is more than 40% higher than in females [73] . Considering these sex differences, it could be suggested that the already increased DRN 5-HT activity in males could be related to the observed differences of SERT in reversing anxiety-related behaviors [74] . The serotonergic signaling can either facilitate or attenuate anxious states, depending on the site of action and the specific serotonin receptor subtype involved [37] . However, the present study, using the same administration route and treatment duration, highlights the differences between the sexes in the antidepressant response, with regard to anti-anxiety effects, i.e. modulating the startle reflex and PPI in response to stressors.

We found that physiological parameters, together with our behavioral observations, were useful for monitoring the stress responsiveness and the role of SERT protecting it. After the animals had been exposed to IMO stress, changes in the hematological and metabolic values were found in all the experimental groups in comparison with the undisturbed controls.

When animals are exposed to an acute stressor physiological changes in the animals’ body are expected in response to the strong increase in glucocorticoids (GCs) and catecholamines secreted by the adrenal glands [1] . Blood coagulation accelerates, eliciting hemoconcentration [2] ; the catabolic characteristics of GCs produces a rapid mobilization of amino acids and lipids [4] [75] and the immune response is faster, enhanced by catecholamines (increasing pro-inflammatory cytokine production) but soon suppressed if GCs remain at high levels [1] [5] [7] [76] . Together, these conditions have evident advantages for the animal’s short-term survival, which are of utmost importance when it must face a potential danger [7] [12] . Thus, and in agreement with previous studies, after being exposed to acute stress the young rats in our study exhibited a significant increase in hematocrit, hemoglobin and platelet numbers [2] [12] [77] . Surprisingly, in the animals previously exposed to daily injections, the stress-response was less evident.

Although there are contradictory results in humans in regard to physiological habituation to stress [12] [78] , in agreement with our data, it has been reported that subjects displaying negative affect showed consistently reduced coagulation activation in response to acute psychosocial stress [79] , and that this was the exact opposite of what they expected in their first hypothesis. Also, in rodents, it was reported that rats exposed to an early type of stress (social deprivation) exhibited failure in fear potentiation [14] and, specifically, the same authors stated that previously injected rats exhibited signs of learned helplessness. As those rats had lost, somewhat, the ability to attend to ecologically important sensory signals. Moreover, when animals are subjected to inescapable stress conditions, the ascendant activity of the serotonergic system may become deregulated [80] , the involvement of the serotoninergic system in regulating the autonomous functions having been suggested [40] [41] . The present study shows that SERT administration attenuates the changes induced by the i.p., affecting the hemostatic response to stress.

Moreover, our data also point to a significant increase in cytoplasmic enzymes, (GOT), and lactate dehydrogenase (LDH) in response to IMO stress in the animals of both sexes, as expected [75] . Even so, it almost doubled the intensity in previously stressed animals as compared to controls, indicating potential tissue-damage in previously stressed animals [81] . Rats exposed to either restraint stress or vehicle injections plus IMO had twice the absolute values of LDH and GOT enzymes in comparison with animals only subjected to IMO. Importantly, SERT administration prevented the prior stress-induced increase in enzyme levels, showing that the drug plays an important role in the preservation of cellular integrity. Nevertheless, recent reports have associated the use of SERT with acute liver failure [82] [83] . Despite intensive investigation using SERT in both animals and humans, the hepatotoxicity of Sertraline remains to be fully elucidated. The toxicity of higher doses, collateral factors, and the method used to study it (in vitro) are major factors that can affect the effects of SERT [83] . Moreover when the impact of SERT on post-myocardial infarction was studied [84] [85] , it was observed that besides being safe, this antidepressant drug was efficacious in patients with unstable ischemic heart disease, decreasing the incidence of severe adverse cardiovascular events [85] .

In the present study, stress induced a dramatic decrease in the total number of leukocytes in rats of both sexes. The significant decrease in the number of leukocytes and lymphocytes after stress exposure has already been reported in several species [86] [87] and should be followed by a slow and complete recovery once it ends [76] . In fact, we found no changes in the animals exposed to IMO when compared to the undisturbed animals (after the recovery period had ended). By contrast, the animals that were exposed to stressful conditions prior to IMO exhibited a persistent fall in the leukocyte count. Even at 10 days after the stress had ended, the animals exhibited a decrease in the total WBC concentration. This prolonged effect is consistent with recent findings reported from our laboratory, where differences were found in prenatally stressed offspring when they were studied in adulthood [50] . We also reported that by administering SERT for two months during adolescence it was possible to reverse the deleterious effects of prenatal stress on the leukocyte count. In the present study, pretreatment with SERT i.p. for 8 days was not sufficient to normalize stress-induced immunosuppression. The inhibition of the reuptake and synthesis of 5-HT does not occur in an immediate way [29] [71] [88] [89] . One week of administration of the drug is sufficient to maintain plasma drug levels steady [90] , but normally the antidepressant effects can only be reached after a few weeks of treatment [46] . It could be surmised that this period of SERT treatment would be sufficient to lead to an increase in the basal levels of 5-HT in the brain [71] [88] [91] , but insufficient to induce the adaptive changes in monoaminergic neurotransmission or in neurotrophic factors, and as consequence, unable to modulate the immune reactivity and possibly the central actions of the cytokines [9] .