Significant outcomes
-
Juvenile sub-chronic sildenafil treatment induced long-term antidepressant-like behavioural effects in young adult Flinders Sensitive Line but not Sprague-Dawley rats.
-
Juvenile sub-chronic sildenafil treatment did not have any long-term effect on the hippocampal brain-derived neurotrophic factor levels of either rat strain during young adulthood.
Limitations
-
Bio-behavioural analyses, immediately following juvenile sub-chronic sildenafil treatment, were not performed.
-
The FST alone was used to evaluate depressive-like behaviour. Other behavioural tests of depressive-like behaviour, such as the sucrose preference test, would be of value in prospective studies.
-
Monoaminergic levels were not measured, which would confirm the conclusions drawn from the behavioural data. Similarly, prospective studies may investigate the role of the nitric oxide/cyclic guanosine 3′,5′-monophosphate/protein kinase G signalling pathway as a novel antidepressant target.
Introduction
Major depressive disorder (MDD) is a serious mood disorder (NIMH, 2018) that globally affects an estimated 322 million people (WHO, 2017), making it the leading cause of global disability (Friedrich, Reference Friedrich2017). A recent study reported that 1.7% of children and 6.1% of adolescents suffer from MDD (Ghandour et al., Reference Ghandour, Sherman, Vladutiu, Ali, Lynch, Bitsko and Blumberg2019), highlighting the high prevalence in the paediatric population. Paediatric MDD results in social dysfunction and poor academic performance (Hazell & Mirzaie, Reference Hazell and Mirzaie2013), with most patients having recurrences during young adulthood (Melvin et al., Reference Melvin, Dudley, Gordon, Ford, Taffe and Tonge2013; Kovacs et al., Reference Kovacs, Obrosky and George2016), thereby accentuating the chronicity of MDD (Monroe & Harkness, Reference Monroe and Harkness2012).
Conventional antidepressants have major shortcomings, including a delayed onset of action, troublesome side-effect profiles and marked ineffectiveness in the treatment of treatment-resistant depression (Rosenzweig-Lipson et al., Reference Rosenzweig-Lipson, Beyer, Hughes, Khawaja, Rajarao, Malberg, Rahman, Ring and Schechter2007). Importantly, the United States’ Food and Drug Administration (FDA) has only approved fluoxetine for the treatment of childhood MDD and escitalopram as additional option in adolescent MDD (FDA, 2014). Still, antidepressants are issued with a ‘black-box’ warning of early increased risk of suicidal ideation and behaviour in paediatric patients (FDA, 2014). This is of note, considering adolescent MDD itself is a major risk factor for suicide (Zubrick et al., Reference Zubrick, Hafekost, Johnson, Sawyer, Patton and Lawrence2017) – the second leading cause of death in this age group (Mullen, Reference Mullen2018). Antidepressant-use in paediatrics has increased (Bachmann et al., Reference Bachmann, Aagaard, Burcu, Glaeske, Kalverdijk, Petersen, Schuiling-Veninga, Wijlaars, Zito and Hoffmann2016), highlighting the difficult position prescribers are placed in – weighing the severe, sometimes fatal consequences of untreated MDD against the uncertain long-term effects of approved therapy.
Neurodevelopment displays surprising similarities and alignment of age-related brain development between humans and other mammals (Eiland & Romeo, Reference Eiland and Romeo2013) (Fig. 1). Importantly, the serotonergic system in rats reaches maturity at the onset of the pre-pubertal period (viz. childhood in humans) on postnatal day 21 (PND21), whereas the noradrenergic system continues to develop throughout pre-pubertal development and only reaches maturity at the onset of puberty (viz. early adolescence in humans) on PND35 (Murrin et al., Reference Murrin, Sanders and Bylund2007; Eiland & Romeo, Reference Eiland and Romeo2013). These different maturity rates may explain why only serotonergic, and not adrenergic-targeting antidepressants are effective in paediatric MDD. Nevertheless, stimuli experienced by an individual during early-life development that affect these neurodevelopmental processes may either benefit or harm the functional integrity of the adult brain (Andersen, Reference Andersen2003; Gomes da Silva et al., Reference Gomes Da Silva, Unsain, Mascó, Toscano-Silva, De Amorim, Silva Araújo, Simoes, Da Graça Naffah-Mazzacoratti, Mortara and Scorza2012), suggesting early-life to be a unique ‘window of opportunity’ to induce long-term beneficial effects via novel antidepressant treatment targets.
A summarised comparison of rodent and human neurodevelopmental phases.
Adapted from (Steyn, Reference Steyn2018). PND: postnatal day.
Preclinical studies from our laboratories were the first to demonstrate that enhanced central nitric oxide/cyclic guanosine 3′,5′-monophosphate/protein kinase G (NO/cGMP/PKG) signalling, either through selective phosphodiesterase type 5 (PDE5) inhibition (Brink et al., Reference Brink, Clapton, Eagar and Harvey2008; Liebenberg et al., Reference Liebenberg, Harvey, Brand and Brink2010a) or intracerebroventricular infusion of a cGMP analogue (Liebenberg et al., Reference Liebenberg, Wegener, Harvey and Brink2010b), induces antidepressant-like effects, partly by activating the cyclic adenosine monophosphate response element-binding protein/brain-derived neurotrophic factor (CREB/BDNF) downstream signalling pathway (Wang et al., Reference Wang, Zhang, Lu, Lin, Pan, Zhao, Liu, Wang, Zhou and Zhang2014). Indeed, MDD is associated with reduced BDNF levels (a marker of neuroplasticity) that is restored with antidepressant treatment (Pittenger & Duman, Reference Pittenger and Duman2008; Lee & Kim, Reference Lee and Kim2010).
A disordered NO/cGMP/PKG signalling cascade has been reported in the Flinders Sensitive Line (FSL) rat (Wegener et al., Reference Wegener, Harvey, Bonefeld, Müller, Volke, Overstreet and Elfving2010), which is a widely described and validated genetic animal model of MDD (Overstreet et al., Reference Overstreet, Friedman, Mathé and Yadid2005). Regarding the hippocampal BDNF levels of FSL rats, findings between studies have been inconsistent. As such, a couple of studies have demonstrated that the hippocampal BDNF levels of FSL rats are comparable to those of controls (Angelucci et al., Reference Angelucci, Aloe, Vasquez and Mathé2000; Angelucci et al., Reference Angelucci, Aloe, Jiménez-Vasquez and Mathé2003), whereas another study indicated reduced hippocampal BDNF levels in FSL rats (Elfving et al., Reference Elfving, Plougmann, Müller, Mathé, Rosenberg and Wegener2010). Therefore, the FSL rat is a suitable animal model to investigate the long-term effects of juvenile sub-chronic sildenafil (SIL) treatment on depressive-like bio-behaviour during young adulthood. Conversely, Sprague-Dawley (SD) rats represent appropriate non-depressive-like comparisons (Magara et al., Reference Magara, Holst, Lundberg, Roman and Lindskog2015).
Taken together, PDE5 inhibitors (such as SIL) hold potential as novel candidates for the treatment of paediatric MDD, with potentially beneficial outcomes later in life. Specifically, the use of SIL in the treatment of paediatric pulmonary hypertension (Huddleston et al., Reference Huddleston, Knoderer, Morris and Ebenroth2009) makes its safety profile better understood, rendering its potential use in depressed children and adolescents especially plausible.
Aims of the study
To investigate the long-term antidepressant-like effects of pre-pubertal versus pubertal sub-chronic sildenafil treatment in young adult Sprague-Dawley versus Flinders Sensitive Line rats.
Materials and methods
Figure 2 illustrates the study layout where all pups (SD and FSL) were weaned on PND21 and randomly divided into pre-pubertal (treated from PND21-34) and pubertal (treated from PND35-48) intervention groups. Hereafter, rats were again randomly divided into saline (SAL) and SIL treatment cohorts. Following 14 days of treatment, rats were housed under standard laboratory conditions, for a ‘wash-out’ period, until PND60 (young adulthood), when behavioural tests were performed. Finally, rats were euthanized by decapitation on PND61 to measure hippocampal BDNF levels.
A schematic illustration of the study layout. BDNF: brain-derived neurotrophic factor. FSL: Flinders Sensitive Line. FST: forced swim test. OFT: open field test. PND: postnatal day. SAL: saline. SD: Sprague-Dawley. SIL: sildenafil.
Animals
Male SD (n = 48) and FSL (n = 48) rats were bred, supplied, and housed at the Vivarium of the Pre-Clinical Drug Development Platform (PCDDP) of the NWU, RSA [(SAVC reg. no.: FR15/13458), AAALAC accredited (international file #1717), GLP compliant (SANAS GLP compliance no.: G0019)]. All rats were group-housed (2–3 rats/cage) in polysulphone individually ventilated cages, under constant Vivarium conditions. Food and tap water were provided ad libitum.
Drug treatment
Rats were either treated with SAL or SIL citrate (3 mg/kg/day; purchased from Sigma Aldrich) (Liebenberg et al., Reference Liebenberg, Harvey, Brand and Brink2010a) dissolved in SAL, via daily subcutaneous (sc) injection, from PND21 (pre-pubertal intervention groups) or PND35 (pubertal intervention groups) for 14 days (Liebenberg et al., Reference Liebenberg, Harvey, Brand and Brink2010a; Schoeman et al., Reference Schoeman, Steyn, Harvey and Brink2017).
Behavioural tests
Behavioural tests were performed during the dark cycle, with the forced swim test (FST) performed 1 h after the open field test (OFT). Behaviour was recorded with a video camera, situated above (OFT) and in front of the test arenas (FST).
Open field test
The OFT was performed as previously described for our laboratories (Schoeman et al., Reference Schoeman, Steyn, Harvey and Brink2017), with each rat placed in the centre of the arena and allowed to explore for 5 min. Total distance moved was analysed with Ethovision XT 14 software (Noldus Information Technology BV, Wageningen, NLD).
Forced swim test
The FST consisted of four Perspex® cylindrical tanks, each filled to a depth of 30 cm with ambient water. The FST was performed as previously described for our laboratories (Schoeman et al., Reference Schoeman, Steyn, Harvey and Brink2017), with rats placed into the water-filled tanks and allowed to swim for 7 min. Behaviour was manually scored by an investigator blind to the treatment groups using a manual continuous timer software (FST Scoreboard 2.0 software; Academic Support Services: Information Technology in Education, NWU, RSA), and scoring only the mid-5 min of each trial (Badenhorst et al., Reference Badenhorst, Brand, Harvey, Ellis and Brink2017; Schoeman et al., Reference Schoeman, Steyn, Harvey and Brink2017). Scored behaviour included immobility, struggling and swimming (Cryan et al., Reference Cryan, Valentino and Lucki2005).
Because the FSL rat presents with inherent increased immobility in the FST, it does not require a pre-conditioning swim trial 24 h before the testing swim trial (Overstreet et al., Reference Overstreet, Friedman, Mathé and Yadid2005). Consequently, neither SD nor FSL rats had a pre-conditioning swim trial to directly compare the behaviour of the two rat strains.
Neurochemical analysis
After euthanasia by decapitation on PND61, the whole brain was immediately extracted and placed in ice-cold SAL. The hippocampi were dissected out promptly on an ice-cooled dissection slab and individually placed into Eppendorf® tubes, which were immediately placed into liquid nitrogen before being stored at −80 °C (Harvey et al., Reference Harvey, Brand, Jeeva and Stein2006).
Hippocampal BDNF level analysis
Analysis of hippocampal BDNF levels was performed using rat BDNF enzyme-linked immunosorbent assay (ELISA) kits (catalog no.: E-EL-R1235) that were purchased from Elabscience Biotechnology Incorporated and according to the instructions of the manufacturer. Importantly, each sample contained the same wet weight of tissue, homogenised in phosphate-buffered saline (PBS) in a 1:9 ratio (hippocampal tissue (g):PBS (ml) = 1:9). Therefore, the differences between the samples in terms of BDNF protein yield of the extraction process were negligible and the measurement of total protein levels was not required. Results were obtained from a standard curve plotted and expressed as pg/g wet weight of the hippocampus.
Statistical analysis
The minimum number of rats needed for statistically significant results were used in this study, as estimated by an evidence-based estimation (Liebenberg et al., Reference Liebenberg, Harvey, Brand and Brink2010a). A 5% confidence limit for error was taken as statistically significant (p ≤ 0.05). A three-way analysis of variance (ANOVA) was performed on all data sets, followed by analyses of two-way interactions, simple main effects, and Bonferroni pairwise comparisons. The unbiased Cohen’s d (d unb ) was also calculated and reported ± 95% confidence interval (CI) of the effect magnitude (Cumming, Reference Cumming2013). IBM® SPSS® Statistics (version 25.0. Armonk, NY: IBM Corp), together with Laerd Statistics® (https://statistics.laerd.com) were used for statistical analyses, whereas figures were created in GraphPad Prism® (version 7.0, San Diego California, USA). Effect magnitude indicators were calculated in Exploratory Software for CIs (Cumming, Reference Cumming2013).
Results
Data are presented according to significant two-way interactions, as identified by ANOVA analyses. Accordingly, Figure 3 reflects only the two independent variables that were identified to have significant effects on inter-group differences. Still, a complete description of the statistical analyses is described in the text.
FST data of SD and FSL rats on PND60, following sub-chronic pre-pubertal or pubertal treatment with either SAL or SIL. Time spent immobile (A) and struggling (B), regardless of treatment period, and time spent swimming (C), regardless of rat strain, according to ANOVA-identified independent variables. All group sizes are equal (n = 24). Data points represent the mean ± 95% confidence interval of the mean. Statistical analyses are reported in the text with ** p ≤ 0.01, **** p ≤ 0.0005 versus SAL + SD in A and B and SAL + Pre-pubertal in C; ^^ p ≤ 0.01, ^^^ p ≤ 0.001, ^^^^ p ≤ 0.0005 versus indicated test group. FSL: Flinders Sensitive Line. SAL: saline. SD: Sprague-Dawley. SIL: sildenafil.
Forced swim test
Although no significant three-way interactions for immobility (F 1,88 = 1.31, p = 0.26) or time spent struggling (F 1,88 = 0.01, p = 0.94) in the FST existed, significant rat strain*drug interactions were identified (F 1,88 = 6.80, p = 0.01) (Fig. 3(A)) and (F 1,88 = 19.40, p ≤ 0.0005) (Fig. 3(B)). Compared to SAL + FSL animals, SAL + SD (d unb = 1.4 [0.8; 2.1]) and SIL + FSL (d unb = 1.4 [0.8; 2.0]) groups spent less time immobile in the FST, regardless of treatment period (p ≤ 0.0005). For struggling, rat strain differences were only identified between SAL-treated animals (p ≤ 0.0005), with SD spending more time struggling compared to FSL rats, regardless of treatment period (d unb = 2.3 [1.6; 3.1]). Further, SIL treatment, regardless of treatment period, decreased struggling behaviour in SD rats (p = 0.007; d unb = 0.9 [0.3; 1.5]) and increased it in FSL rats (p = 0.001; d unb = 0.8 [0.2; 1.4]). Again, no significant three-way interaction (F 1,88 = 1.79, p = 0.18) existed for swimming behaviour (Fig. 3(C)); however, a significant treatment period*drug interaction was identified (F 1,88 = 4.74, p = 0.03). Pre-pubertal SIL treatment, regardless of rat strain, increased swimming behaviour relative to SAL-treated age-matched controls (p ≤ 0.0005, d unb = 1.3 [0.7; 1.9]) and compared to pubertal SIL-treated animals (p = 0.02, d unb = 0.8 [0.3; 1.4]).
Open field test and hippocampal brain-derived neurotrophic factor levels
There were no significant three-way (F 1.88 = 2.08, p = 0.15 and F 1,88 = 0.01, p = 0.94) or two-way interactions for distance moved in the OFT or hippocampal BDNF levels (Table 1).
Total distance moved in the OFT and hippocampal BDNF levels of young adult SD and FSL rats, following sub-chronic pre-pubertal or pubertal treatment with either SAL or SIL
BDNF: brain-derived neurotrophic factor. FSL: flinders sensitive line. OFT, open field test. SAL: saline. SD: Sprague-Dawley. SIL: sildenafil.
Distance moved (cm) on PND60 and hippocampal BDNF levels (pg/g wet weight of the hippocampus) on PND61. All groups are equal (n = 12). Data points represent the mean ± standard error of the mean.
Discussion
On PND60 (representing young adulthood), SAL-treated FSL rats, regardless of treatment period, displayed increased depressive-like behaviour (Fig. 3(A)) and reduced struggling behaviour (putatively associated with impaired noradrenergic neurotransmission) in the FST (Fig. 3(B)) compared to SAL-treated SD rats, thereby confirming the depressive-like phenotype of the FSL rat (Overstreet et al., Reference Overstreet, Friedman, Mathé and Yadid2005). Importantly, because general locomotor activity was comparable across all treatment groups (Table 1), FST group differences could be ascribed to treatment-induced changes or inter-strain variation in psychomotor (and not locomotor) activity.
Perhaps the most important finding is that early-life sub-chronic SIL treatment induced antidepressant-like effects in young adult FSL rats, regardless of treatment period (Fig. 3(A)). This is promising, as neither escitalopram (Steyn et al., Reference Steyn, Harvey and Brink2018) nor venlafaxine (Steyn et al., Reference Steyn, Harvey and Brink2020) were able to induce any robust long-term antidepressant-like bio-behavioural effects in FSL rats in previous studies. However, the antidepressant-like effects of early-life SIL treatment was not observed in young adult SD rats, suggesting that a genetic predisposition that makes individuals more susceptible to the development of MDD (i.e. FSL rats) may play an important role in the long-term antidepressant-like effects of SIL.
Interestingly, early-life sub-chronic SIL treatment, irrespective of treatment period, appears to elevate noradrenergic neurotransmission (mechanism to be confirmed) in young adult FSL rats (increased struggling behaviour in the FST) and reduce it in young adult SD rats (healthy controls) (Fig. 3(B)). Therefore, it cannot be ruled out that sub-chronic SIL treatment in paediatric patients may have potentially detrimental effects in normal healthy individuals later in life in terms of hampering coping responses and therefore requires further investigation. Still, these findings suggest that having a genetic susceptibility to develop MDD (or the lack thereof) dictates the type of effects that early-life sub-chronic SIL treatment might have on noradrenergic neurotransmission during young adulthood and warrants further investigation.
Our data further suggest that putative enhancement of serotonergic-associated behaviour (mechanism to be confirmed) later in life by early-life sub-chronic SIL treatment is age-dependent, since increased swimming behaviour in the FST was only observed in pre-pubertal- and not pubertal-treated animals, irrespective of rat strain (Fig. 3(C)). Genetic susceptibility does not appear to play a role in the long-term serotonergic-associated effects of SIL, since both adult SD and FSL rats were similarly affected in this study, which is in line with the current approval of only serotonergic-targeting antidepressants in paediatric MDD patients.
Increases in BDNF levels have been associated with antidepressant action (Pittenger & Duman, Reference Pittenger and Duman2008; Lee & Kim, Reference Lee and Kim2010). In this study, early-life sub-chronic SIL treatment had no long-term or lasting effect on the hippocampal BDNF levels of either adult SD or FSL rats (Table 1). This may be due to the transient enhancement of neuroplasticity observed immediately following early-life antidepressant treatment. As such, our group demonstrated that pre-pubertal sub-chronic escitalopram treatment increased hippocampal BDNF levels on PND35 in FSL rats (immediately following treatment), however on PND60 (after a ‘wash-out’ period), these levels had returned to baseline (Steyn et al., Reference Steyn, Harvey and Brink2018). Prospective studies should further investigate the effect of PDE5 inhibition on neuroplasticity.
Although the biochemical mechanisms underlying the antidepressant-like effects of SIL cannot be elucidated solely based on the present findings, the pro-adrenergic effects of SIL, as delineated from the SIL-induced changes in struggling behaviour in the FST, are an intriguing possibility. In this regard, whereas it has been demonstrated that the antidepressant activity of SIL is associated with a novel mechanism of action, namely the modulation of the NO/cGMP/PKG and the cyclic adenosine monophosphate/protein kinase A (cAMP/PKA) cell signalling pathways (Liebenberg et al., Reference Liebenberg, Wegener, Harvey and Brink2010b, Reference Liebenberg, Müller, Fischer, Harvey, Brink, Elfving and Wegener2011), we see here that SIL may induce its long-term antidepressant-like effects in FSL rats, at least in part, by ultimately normalising noradrenergic signalling. As a working hypothesis, when translating the current results to humans, it may be that individuals with a genetic susceptibility to develop MDD could benefit from early-life sub-chronic SIL treatment, putatively by modulating neurodevelopment, which results in a significantly reduced risk of developing MDD during young adulthood. Taken together, the PDE5 inhibitors (such as SIL) have potential as novel antidepressant strategies in the treatment of paediatric MDD, with beneficial behavioural outcomes later in life.
Acknowledgements
Mr. Cor Bester, Mrs. Antoinette Fick, and Mr. Kobus Venter for overseeing the welfare of the animals.
Author contributions
JLBS conducted behavioural tests, assisted with neurochemical and statistical analyses, data interpretation, wrote the first draft of the manuscript, and prepared it for submission. SFS supervised the study, assisted with statistical analyses, and data interpretation, edited and finalised the manuscript for submission. CBB: designed, planned, and funded the study, assisted with data interpretation, edited and finalized the manuscript for submission.
Financial support
This study was funded by Professor Brink (NRF – grant no. 103371IFR160118156926).
Conflict of interest
No conflict of interest to declare.
Animal welfare
The study was approved by the Animal Care, Health and Safety Research Ethics Committee (NHREC reg. no.: AREC-130913-015) of the North-West University (NWU-AnimCareREC ethics approval no.: NWU-00277-17-A5). All experiments adhered to the guidelines of the South African National Standards: The Care and Use of Animals for Scientific Purposes (SANS 10 386 : 2008). Animals were maintained, and all procedures performed in accordance with the code of ethics in research, training, and testing of drugs in South Africa and complied with national legislation.
Ethical standards
The authors assert that all procedures contributing to this work comply with the code of ethics of the South African National Standards and institutional guides on the care and use of laboratory animals.
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/neu.2021.4
Open access
