Transgenic mice lacking CREB and CREM in noradrenergic and serotonergic neurons respond differently to common antidepressants on tail suspension test

Evidence exists that chronic antidepressant therapy enhances CREB levels and activity. Nevertheless, the data are not conclusive, as previous analysis of transgenic mouse models has suggested that CREB inactivation in fact contributes to antidepressant-like behavior. The aim of this study was to evaluate the role of CREB in this context by exploiting novel transgenic mouse models, characterized by selective ablation of CREB restricted to noradrenergic (Creb1DBHCre/Crem−/−) or serotonergic (Creb1TPH2CreERT2/Crem−/−) neurons in a CREM-deficient background to avoid possible compensatory effects of CREM. Selective and functional ablation of CREB affected antidepressant-like behavior in a tail suspension test (TST) after antidepressant treatment. Contrary to single Creb1DBHCre mutants, Creb1DBHCre/Crem−/− mice did not respond to acute desipramine administration (20 mg/kg) on the TST. On the other hand, single Creb1TPH2CreERT2 mutants displayed reduced responses to fluoxetine (10 mg/kg) on the TST, while the effects in Creb1TPH2CreERT2/Crem−/− mice differed by gender. Our results provide further evidence for the important role of CREM as a compensatory factor. Additionally, the results indicate that new models based on the functional ablation of CREB in select neuronal populations may represent a valuable tool for investigating the role of CREB in the mechanism of antidepressant therapy.

(CREB) transcription factor 4 , one of transcription factors binding to cAMP-responsive elements (CRE). CREB appears to be involved in both the mechanisms of antidepressant action and in the disease itself 5 . Therefore, research has focused on the role of G-protein-coupled receptors and associated second messenger pathways, primarily involving the cyclic AMP (cAMP)-dependent pathway. Augmentation of cAMP leads to the activation of cAMP-dependent protein kinase A (PKA), and PKA activity is enhanced after chronic antidepressant treatment 6,7 . Alternative mechanisms implicated in the action of antidepressant drugs are related to Ca2±/ calmodulin-dependent protein kinase II (CaMKII) 8,9 . CREB has been demonstrated to mediate the transcriptional activation of genes in response to both cAMP and Ca 2+ influx signal transduction pathways 10,11 . For instance, according to the neurotrophic hypothesis of depression, levels of brain derived neurotrophic factor (BDNF, whose transcription is regulated by a CREB-dependent mechanism) are increased both by noradrenergic or/and serotonergic antidepressants, while exposure to stress is characterized by downregulation of BDNF expression in hippocampus 12 ; this downregulation can be prevented by antidepressant treatment 13,14 . However, it has to be mentioned, that this effect is not specific and although enhancement of BDNF expression seems to exert antidepressant-like effects in the hippocampus, its actions might be opposite in other brain regions i.e. nucleus accumbens, where chronic social defeat stress increases BDNF protein levels 15 . BDNF also regulates serotonin signaling, as its main receptor, TrkB, can be found on serotonergic neurons as well 16 .
The abovementioned monoamine systems exert mutual influence over each other. The serotonergic system may be inhibited by noradrenaline through action on α1and β-adrenergic receptors on serotonergic neurons of the raphe nuclei, while serotonergic projections can inhibit the activity of the locus coeruleus 6,17 .
Antidepressants can affect CREB in several different ways; however, the data are not conclusive. The generally accepted theory is that chronic antidepressant treatment enhances CREB levels and activity, thus implicating CREB as an important mechanism of antidepressant treatment 5 . In particular, chronic administration of desipramine and imipramine (common antidepressants whose action is based primarily on the inhibition of noradrenaline reuptake) increases expression of CREB mRNA and phospho-CREB (pCREB) in selected brain regions 18,19 . Similar effects have been observed after fluoxetine or citalopram (selective serotonin reuptake inhibitors, or SSRIs) treatment 18,20 . On the other hand, adverse effects on CREB expression have also been observed after desipramine or fluoxetine treatment 21,22 . Furthermore, treatment with venlafaxine, a dual monoamine reuptake inhibitor, was shown to reduce pCREB in frontal cortex with no change in the total CREB expression 23 .
Rodent models have made substantial contributions to advancing our understanding of depression and the mechanism of antidepressant treatment. Several models for studying the role of CREB are based on transgenic rats overexpressing CREB 24,25 or mice with a constitutive deletion of the gene [26][27][28] . Results from these models are counterintuitive, as the majority of studies have demonstrated that CREB inactivation contributes to antidepressant-like behavior. However, it should be emphasized that these loss-of-function studies possess many caveats, possibly making interpretation of the data difficult and misleading. Namely, (i) the experimental mutations generally targeted CREB in several brain structures; and (ii) compensatory effects of related CREB heterodimerization gene products (i.e., cAMP response element modulator, CREM) were not taken into consideration. The last caveat seems to be of particular importance, as it was shown that CREB is not the only mediator of cAMP-dependent transcriptional regulation and other nuclear effectors of the cAMP-dependent signaling pathway can compensate lack of CREB function 29 . In particular, CREB spatiotemporal knockouts are usually not as deteriorated regarding their phenotype as expected, as other c-AMP driven transcriptional activators (CREM and ATF-1) can compensate for each other, and various CREB deficient mice overexpress CREM 29 . This strong interdependence between CREB and CREM was also confirmed in the study of Mantamadiotis et al., showing that the embryonic mutation evoking loss of CREB in neural and glial progenitors (Creb1 NesCre Crem−/− mice) was effective throughout the brains of these mutants only when both CREM alleles were also depleted 30 .
The aim of the current study was to evaluate the function of CREB in the mechanisms of antidepressant treatment by exploiting novel transgenic mouse models characterized by selective functional ablation of CREB restricted only to the noradrenergic or serotonergic neurons. Furthermore, considering the known compensatory effects of CREM 29,31 , both lines were maintained in a CREM-deficient (Crem−/−) background.
Male and female mutant mice were kept with their control (Cre-negative, CREM+/+) littermates of the same sex in self-ventilated cages under standard laboratory conditions (12 h light/dark cycle, food and water ad libitum). All mice were 3-4 months old (approx. 12-16 weeks). This study was carried out in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Behavioral protocols were approved by the Animal Ethical Committee at the Institute of Pharmacology, Polish Academy of Sciences (Permit Number: 1125, issued 11/24/2014).

Open field test (OFT).
The OFT was performed to assess spontaneous locomotor activity. Mice were tracked by video camera using automated video tracking software (EthoVision XT8, Noldus, Netherlands) for 60 min in 40 × 40 cm white square boxes; the total distance moved was scored in 10-min intervals.

Rotarod test (ROT).
The ROT was performed to assess motor coordination using an accelerated rotarod (Ugo Basile, Italy). The assessment was preceded by a training session 1 day before the experiment (5 min on the rotating rod, constant speed of 4 rpm). During the experiment the time spent on the accelerating rod (4-40 rpm in a 5-min period) was measured.

Tail suspension test (TST).
The TST was performed to evaluate depression-like and antidepressant behavior after drug treatment. We recorded the overall time that animals were immobile while suspended by the tail over a 6 min period. Scoring of immobility time was performed by means of automated video tracking software (EthoVision XT8, Noldus, Netherlands) as described previously 34 . Statistical analysis. Data were analyzed using Statistica 12 software (Statsoft, USA). All comparisons were performed using one-way analysis of variance followed by Fisher Least Significant Difference post-hoc test. Changes with p value lower than 0.05 were considered significant.
Data Availability Statement. The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Results
Ablation of CREB is cell type-specific in both studied transgenic lines. Immunofluorescent staining confirmed that the CREB protein was lost specifically in noradrenergic cells expressing tyrosine hydroxylase (TH) in the locus coeruleus (LC) of Creb1 DBHCre Crem−/− mice (Fig. 1a) as well as in serotonergic, tryptophan hydroxylase 2 (TPH2) positive cells in the dorsal raphe (DR) nucleus of Creb1 TPH2CreERT2 Crem−/− mice (Fig. 1b). The DBHCre and the TPH2CreERT2 specificity have been previously shown and extensively used for conditional gene targeting. Here, we have shown that in both Creb1 DBHCre Crem−/− and Creb1 TPH2CreERT2 Crem−/− lines, staining with anti-CREB antibodies provided no signal in the abovementioned brain structures of mutant animals. In other brain areas not targeted by the mutation (e.g., hippocampus), the CREB protein was preserved in both transgenic lines (Fig. 2a,b). These results confirmed that the mutations were highly specific in the central nervous system and restricted to noradrenergic and serotonergic neuronal cell populations.

Male and female CREB-and CREB/CREM-deficient mice in serotonergic or noradrenergic cells show no impairment in basal behavioral phenotype.
Male Creb1 DBHCre Crem−/− and Creb1 TPH2CreERT2 Crem−/− mice did not exhibit any visible impairments, nor was their body weight affected in comparison to their control littermates (Fig. 3a,b). Moreover, their spontaneous locomotor activity on the open field test (OFT) did not differentiate them from control animals, although in the first 10 min of the test, Creb1 DBHCre Crem−/− mice were slightly less active (Fig. 3c,d). The ROT results demonstrated that the motor coordination of males in both transgenic mouse lines was unaffected by the mutation (Fig. 2e,f). Moreover, the TST did not reveal any differences in immobility times for single mutant Creb1 DBHCre Crem−/− and double mutant Creb1 TPH2CreERT2 Crem−/− mice, suggesting the lack of a depressive or antidepressant phenotype under basal conditions (Fig. 3g,h).
Similar to the mutant males, no differences were observed in the body weights of transgenic females of either the Creb1 DBHCre Crem−/− or Creb1 TPH2CreERT2 Crem−/− line (Fig. 4a,b). As measured by the OFT and ROT, spontaneous locomotor activity (Fig. 4c,d) and motor coordination of female mutant mice (Fig. 4e,f) remained at similar levels as detected in control littermates, irrespective of the loss of CREB in noradrenergic and serotonergic neurons. Additionally, mutant females, like males, were indistinguishable from control mice in their TST response under basal conditions (Fig. 4g,h).
Overall, behavioral screening of basal phenotypes did not reveal any differences between control, single mutant (either Creb1 DBHCre or Creb1 TPH2Cre ) and double mutant (Creb1 DBHCre Crem−/− or Creb1 TPH2CreERT2 Crem−/−)    administration in control mice (both males and females) evoked an antidepressant behavior on the TST; this response is shown by their shortened immobility time in comparison to saline treated animals (Fig. 5a,b, second bars from the left). However, while single Creb1 DBHCre mutants showed an antidepressant desipramine response of similar level to control animals (significant only in males), Creb1 DBHCre Crem−/− double mutants did not respond (Fig. 5a,b), thus presenting a drug-resistant phenotype. The immobility time scores in Creb1 DBHCre Crem−/− mice were similar to or greater than those of untreated control littermates; these scores were significantly different from those of single Creb1 DBHCre mutants. This phenomenon was observed in Creb1 DBHCre Crem−/− mice regardless of sex.

Antidepressant-like effects of fluoxetine in Creb1 TPH2CreERT2
Crem−/− double mutants on the TST are sex-dependent. Fluoxetine administration in control male and female mice resulted in significantly shorter immobility times on the TST in comparison to animals that received saline injections (Fig. 6a,b). However, in contrast to the effects observed in Creb1 DBHCre and Creb1 DBHCre Crem−/− mice after desipramine treatment, a single CREB ablation with conserved CREM (Creb1 TPH2CreERT2 mice) was sufficient to produce a drug-resistant phenotype after fluoxetine administration. The immobility time after fluoxetine injection did not change in these single mutants when compared to control littermates that did not receive the drug. Additionally, when the CREM protein was removed, animals reacted in a sex-dependent manner: Creb1 TPH2CreERT2 Crem−/− double mutant females did not react to fluoxetine treatment, while Creb1 TPH2CreERT2 Crem−/− double mutant males responded to fluoxetine similarly to control animals, exhibiting significantly decreased immobility times relative to saline-treated controls (Fig. 6a,b). These results suggest that cell-specific loss of CREB-dependent signaling could account for differential responses to anti-depressants.

Discussion
The current study was based on transgenic lines lacking CREB in two important neurotransmitter systems, noradrenergic and serotonergic neurons, both of which play a crucial role in the modulation of antidepressant drug action. The objective of this study was to re-evaluate the role of CREB in antidepressant drug action by considering the known compensatory effects of CREM. This factor has been often ignored in previous investigations of the role of CREB in depression and antidepressant treatment carried out in transgenic animal models. To avoid the compensatory effects of CREM, both lines were maintained in a CREM deficient (Crem−/−) background.
The specificity of the targeted mutation in single mutant Creb1 DBHCre line has been previously validated in the central nervous system 35 . The current study confirmed this specificity in the double mutant line Creb1 DBHCre / Crem−/− as well as in the newly created Creb1 TPH2CreERT2 /Crem−/− double mutant mice. In both cases, the selectivity of the mutation was confirmed by immunofluorescent staining. CREB expression was selectively and completely lost in regions expressing DBH and TPH2: the locus coeruleus (LC; Creb1 DBHCre /Crem−/− mice;   Fig. 1b). Other brain structures remained intact (e.g., hippocampus; Fig. 2a,b).
We quantified the basic phenotype of both transgenic lines in selected behavioral tests. When compared to control littermates, single CREB mutants (Creb1 DBHCre , Creb1 TPHCreERT2 ) and double mutants (Creb1 DBHCre / Crem−/−, Creb1 TPH2CreERT2 /Crem−/−) displayed no obvious alterations in daily cage behavior, weight gain (Figs 3a,b and 4a,b), spontaneous locomotor activity (Figs 3c,d and 4c,d), or motor coordination (Figs 3e,f and 4e,f), regardless of genotype and sex. Since the main purpose of this study was to investigate the influence of select antidepressants on the tail suspension test (TST), basal depressive behavior was evaluated beforehand to ensure that results were not biased by reactiveness on the part of non-treated animals on the TST (Figs 3g,h and 4g,h). Male and female animals were investigated separately because sex differences frequently occur in studies of depression-like and antidepressant behavior 36 , an issue often neglected by researchers. TST in basal conditions did not reveal any differences between mutant and control animals in both studied transgenic lines. Only difference noted in basal phenotype of studied animals was a diminished locomotor activity of Creb1 DBHCre single, and Creb1 DBHCre Crem−/− double male mutants in the first 10 min interval of OFT. This might suggest increased anxiety behavior of male mutant mice, but even assuming such phenotype, it does not reflect the results obtained in TST at basal conditions (Fig. 3g). Overall, the only noticeable difference in basal phenotype between mutants and controls was the infertility of CREM-deficient male mice, a well-known issue due to the crucial role of CREM in spermatogenesis 37 .
Abovementioned basal phenotype results indicated that both transgenic models could be evaluated on the TST following antidepressant treatment without risk of confounds due to prior behavioral impairments. We selected a single-dose paradigm using the most common, representative antidepressants: a potent noradrenaline reuptake inhibitor (desipramine) and selective serotonin reuptake inhibotor (SSRI; fluoxetine). Doses (20 mg/kg, i.p. and 10 mg/kg, i.p., respectively) were based on our previous experience and published literature 38,39 .
In the transgenic line targeting the noradrenergic system, neither male nor female Creb1 DBHCre /Crem−/− mice responded to acute desipramine treatment on the TST. This result differed from findings in single Creb1 DBHCre mutants and was consistent with the initial hypothesis that CREM-dependent compensation for CREB function plays a role in antidepressant treatment. Apparently a single mutation was insufficient for disordering the mechanism of drug action, and the effects of desipramine on the TST were abolished only after concomitant CREM removal. This finding is also in line with other results highlighting the pivotal role of CREB in the action of antidepressant drugs targeting the noradrenergic system, in particular desipramine. Specifically, it has been proposed that the therapeutically relevant action of this drug may be related to attenuation of CREB-mediated gene transcription 22 . Recent studies have shown that desipramine improves depression-like behavior on the TST by upregulating p-CREB in the hippocampus 40 . Thus, the functional loss of CREB in Creb1 DBHCre /Crem−/-mice potentially interferes with these proposed mechanisms.
Interestingly, we did not observe analogous effects in the second line targeting the serotonergic system after fluoxetine application. Regardless of sex, single-mutant Creb1 TPH2Cre mice displayed resistance to fluoxetine treatment on the TST. Moreover, this effect was sustained only in case of Creb1 TPH2CreERT2 /Crem−/− female double mutants; in contrast, Creb1 TPH2CreERT2 /Crem−/− male double mutants responded to fluoxetine treatment in the same way as control mice. One of possible explanation that mires interpretation of obtained results might be the fact, that due to technical limitations and mice availability, in our experiments we compare constitutive (Creb1 TPH2CreERT2 /Crem−/−) vs inducible (Creb1 TPH2CreERT2 /Crem−/−) line. Therefore, before introducing drug factor, we thoroughly analyzed the basic behavior of both transgenic lines and did not find any discerning effects of introduced mutation in neither of them. Nevertheless, one cannot exclude that compensatory effects may emerge when triggered by external stimuli i.e. exposure to the drug. We speculate that in case of Creb1 TPH2CreERT2 mice, CREM-dependent brain plasticity was insufficient to compensate for the single CREB ablation especially because the mutation took place in adult mice which eliminates the possible developmental compensatory mechanisms present in constitutive knock-out animals. This assumption is supported by prior findings that CREM overexpression in transgenic models is sometimes insufficient to prevent the CREB-deletion phenotype, particularly in the context of drug addiction 31 . Moreover, the precise mechanisms of interaction between CREB, CREM and ATF factors (including the extent and kinetics of mCREB dimerization) are not well understood 37,41 . On the other hand, the response to fluoxetine treatment observed in Creb1 TPH2CreERT2 /Crem−/− male mice could be interfered by different mechanism of action of drugs acting via the serotonergic system, particularly fluoxetine, whose effectiveness is additionally determined by various environmental factors 42 . This finding is compatible with observations of differential regulation of CREB expression by desipramine and fluoxetine and with the regional specificity of their effects 43 .
In relation to the gender-dependent response to fluoxetine observed in serotonergic-specific mutants, interestingly, clinical studies have reported sex differences among prevalence of depression, particularly among perimenopausal women 44 . Differential responsiveness to antidepressants, including SSRIs, has been reported between men and women in clinics, although this finding remains controversial 45 . Furthermore, sex differences in the behavioral responsiveness of transgenic mouse models targeting the serotonergic system have been described in other studies as well 34,46,47 .
Finally, it has to be clearly stated that all previous studies regarding the role of CREB in depression and antidepressant drug action have been carried out in multiple brain regions, while our mutation is focused on the origin sites of noradrenergic and serotonergic neurons in the central nervous system: in particular, the locus coeruleus (LC) and dorsal raphe (DR) nuclei. It remains uncertain whether and how CREB can influence other brain structures receiving inputs from noradrenergic and/or serotonergic projections. These structures are traditionally regarded as the most important areas in depression pathophysiology and the mechanism of antidepressants (i.e., prefrontal cortex, hippocampus, nucleus accumbens). Clarifying these influences will be our goal in future research. Due to technical limitations (i.e., the demands of obtaining proper groups of controls, single mutants, and double mutant littermates for behavioral experiments, which forced us to split the analysis of basal and drug-induced behavior in mutant mice), we were unable to include a separate group of single-mutation CREM-deficient mice (Crem−/−). Crem−/− male mutants are known to be sterile 48 , but overall this mutation does not appear to be phenotypically meaningful, being not associated with any profound impairment 49 . Namely, no significant difference were observed in daily cage behavior, spontaneous and amphetamine-induced locomotor activity, conditioned suppression of motility (reactiveness to stress) 49 . Crem−/− mice were characterized as slightly hyperactive only when analyzed in the dark phase, and by diminished anxiety behavior as revealed by elevated plus maze (EPM) 49 . However, this phenotype (even assuming exertion on obtained results) should in fact promote antidepressant behavior of double mutants analyzed in our experiments i.e. decreased immobility in TST at basal conditions, and such behavior was certainly not observed.
These initial observations, in particular those regarding the effects of fluoxetine, require confirmation in other behavioral paradigms. Nevertheless, these results provide overall confirmation of the crucial role of CREB in response to antidepressant treatment and clearly highlight CREM as an important compensatory factor, despite the different regulation observed in the serotonergic line. Additionally, these newly-created models based on functional ablation of CREB in select neuronal populations may represent a unique, valuable tool for investigating the role of CREB in the mechanism of antidepressants.