European Journal of Pharmacology 969 (2024) 176434 Available online 6 March 2024 0014-2999/© 2024 Elsevier B.V. All rights reserved. Sildenafil, alone and in combination with imipramine or escitalopram, display antidepressant-like effects in an adrenocorticotropic hormone-induced (ACTH) rodent model of treatment-resistant depression Juandré Lambertus Bernardus Saayman a, Brian Herbert Harvey a,b,c, Gregers Wegener d, Christiaan Beyers Brink a,* a Centre of Excellence for Pharmaceutical Sciences (Pharmacen™), Faculty of Health Sciences, North-West University, Private Bag X6001, Potchefstroom, 2520, South Africa b South African Medical Research Council Unit on Risk and Resilience on Mental Disorders, Department of Psychiatry and Neuroscience Institute, University of Cape Town, Rondebosch, 7700, South Africa c The Institute for Mental and Physical Health and Clinical Translation, School of Medicine, Deakin University, Geelong, Australia d Translational Neuropsychiatry Unit (TNU), Department of Clinical Medicine, Aarhus University, DK-8200 Aarhus N, Denmark A R T I C L E I N F O Keywords: Adrenocorticotropic hormone Major depressive disorder Phosphodiesterase type 5 inhibitor Sildenafil Sprague-dawley rat Treatment-resistant depression A B S T R A C T Background: Major depressive disorder (MDD) represents a challenge with high prevalence and limited effec tiveness of existing treatments, particularly in cases of treatment-resistant depression (TRD). Innovative strate gies and alternative drug targets are therefore necessary. Sildenafil, a selective phosphodiesterase type 5 (PDE5) inhibitor, is known to exert neuroplastic, anti-inflammatory, and antioxidant properties, and is a promising antidepressant drug candidate. Aim: To investigate whether sildenafil monotherapy or in combination with a known antidepressant, can elicit antidepressant-like effects in an adrenocorticotropic hormone (ACTH)-induced rodent model of TRD. Methods: ACTH-naïve and ACTH-treated male Sprague-Dawley (SD) rats received various sub-acute drug treat ments, followed by behavioural tests and biochemical analyses conversant with antidepressant actions. Results: Sub-chronic ACTH treatment induced significant depressive-like behaviour in rats, evidenced by increased immobility during the forced swim test (FST). Sub-acute sildenafil (10 mg/kg) (SIL-10) (but not SIL-3), and combinations of imipramine (15 mg/kg) (IMI-15) and sildenafil (3 mg/kg) (SIL-3) or escitalopram (15 mg/ kg) (ESC-15) and SIL-3, exhibited significant antidepressant-like effects. ACTH treatment significantly elevated hippocampal levels of brain-derived neurotrophic factor (BDNF), serotonin, norepinephrine, kynurenic acid (KYNUA), quinolinic acid (QUINA), and glutathione. The various mono- and combined treatments significantly reversed some of these changes, whereas IMI-15 + SIL-10 significantly increased glutathione disulfide levels. ESC-15 + SIL-3 significantly reduced plasma corticosterone levels. Conclusion: This study suggests that sildenafil shows promise as a treatment for TRD, either as a stand-alone therapy or in combination with a traditional antidepressant. The neurobiological mechanism underlying the antidepressant-like effects of the different sildenafil mono- and combination therapies reflects a multimodal action and cannot be explained in full by changes in the individually measured biomarker levels. 1. Introduction Major depressive disorder (MDD) is a serious mood disorder, char acterized by persistent depressive symptoms, often accompanied by anhedonia and other distressing symptoms (American Psychiatric Association, 2013; National Institute of Mental Health, 2023). Its impact on individuals and society is profound, given its chronic and recurrent nature (Reierson et al., 2011; Ten Have et al., 2018; Verhoeven et al., 2018). It ranks as a leading contributor to global disability (Preboth, 2000; Friedrich, 2017) and affects approximately 4.4% of the global * Corresponding author: Internal Box 16, Unit for Drug Research and Development, School of Pharmacy, Faculty of Health Sciences, North-West University, Private Bag X6001, Potchefstroom, 2520, South Africa. E-mail address: Tiaan.Brink@nwu.ac.za (C.B. Brink). Contents lists available at ScienceDirect European Journal of Pharmacology journal homepage: www.elsevier.com/locate/ejphar https://doi.org/10.1016/j.ejphar.2024.176434 Received 23 November 2023; Received in revised form 15 February 2024; Accepted 16 February 2024 mailto:Tiaan.Brink@nwu.ac.za www.sciencedirect.com/science/journal/00142999 https://www.elsevier.com/locate/ejphar https://doi.org/10.1016/j.ejphar.2024.176434 https://doi.org/10.1016/j.ejphar.2024.176434 https://doi.org/10.1016/j.ejphar.2024.176434 http://crossmark.crossref.org/dialog/?doi=10.1016/j.ejphar.2024.176434&domain=pdf European Journal of Pharmacology 969 (2024) 176434 2 population (World Health Organisation, 2017). Treatment-resistant depression (TRD) represents a particularly challenging and persistent form of MDD, characterized by severe, longer-lasting symptoms and a high recurrence rate (Rush et al., 2008; Cleveland, 2023), resulting in reduced quality of life, impaired social, cognitive, and occupational functioning, and an elevated risk of suicide (Hawton et al., 2013; Bergfeld et al., 2018; Lex et al., 2019; Gregory et al., 2020; Li, 2023). Pharmacological treatment of MDD typically involves selective se rotonin reuptake inhibitors (SSRIs) and other antidepressant classes (Malhi et al., 2013). For TRD, treatment strategies include optimization of pharmacotherapy, novel approaches such as ketamine or psilocybin, or augmentation with lithium or second-generation antipsychotics (Al-Harbi, 2012; Voineskos et al., 2020). However, limitations include side effects, delayed onset of action, and difficulties in managing cognitive deficits (Rosenzweig-Lipson et al., 2007; Carvalho et al., 2016). Notably, almost one-third of MDD patients do not achieve remission even after multiple trials of traditional antidepressants, who are classified as treatment-resistant (Rush et al., 2006; Zorumski et al., 2015; Carvalho et al., 2016). This underscores the need for novel and more effective treatment options. Existing antidepressants primarily target central monoaminergic neurotransmission, particularly serotonergic, noradrenergic, and dopa minergic systems (Malhi et al., 2013; Willner et al., 2013; Boku et al., 2018). However, the pathophysiology of MDD may involve complex and diverse mechanisms (Duarte-Silva et al., 2020), with inflammation and oxidative stress playing a central role in MDD, and contributing to pathological features of the condition such as neurodegeneration, cell death, reduced neurogenesis, and autoimmune responses (Bakunina et al., 2015; Brand et al., 2015). An imbalance between reactive oxygen and nitrogen species and antioxidants, such as glutathione, can lead to oxidative stress and inflammation (Bakunina et al., 2015). Additionally, glutamate-related pathology is intimately linked in TRD, e.g., its response to ketamine (Schwartz et al., 2016; Phillips et al., 2019). Indeed, glutamate’s association with the kynurenine pathway and sub sequent effects on excitotoxicity via quinolinic acid-kynurenine imbal ance, are causally linked to inflammation and changes in monoaminergic signalling, particularly serotonin (Niciu et al., 2014; Moller et al., 2015). Glutamatergic signalling is linked to the nitric oxide (NO)-cyclic guanosine monophosphate (cGMP) cascade in depression, especially relapse and treatment refractoriness (Harvey, 1996; Harvey et al., 2002, 2006b). NO-cGMP in turn emerges as a critical player in neuro modulation, neurotransmission, and redox-inflammatory signalling (Schulman, 1997; Calabrese et al., 2004; Straub et al., 2007; Banach et al., 2011). The cGMP/protein kinase G (cGMP/PKG) pathway, under the regulation of phosphodiesterase type 5 (PDE5), plays a pivotal role in modulating cyclic adenosine monophosphate (cAMP)-response element binding protein/brain-derived neurotrophic factor (CREB/BDNF) signalling and neuroplasticity (Feil et al., 2005; Dhir and Kulkarni, 2007). The antidepressant-like properties of selective PDE5 inhibitors, such as sildenafil, have been demonstrated in our labora tories (Brink et al., 2008; Liebenberg et al., 2010a, 2010b), supported by other pre-clinical studies (Baek et al., 2011; Matsushita et al., 2012; Tomaz et al., 2014; Wang et al., 2014; Socała et al., 2016). In this study, we explored the dose-dependent antidepressant-like effects of sildenafil, either alone or in combination with a known monoaminergic antide pressants, in an adrenocorticotropic hormone-induced (ACTH) rodent model of TRD (Kitamura et al., 2002, 2008; Walker et al., 2013; Pereira et al., 2019). We hypothesised that (1) low-dose sildenafil and high-dose sildenafil in combination with imipramine would induce antidepressant-like effects and (2) low-dose sildenafil would augment the antidepressant-like effects of traditional monoaminergic antide pressants, i.e., imipramine and escitalopram, in the ACTH model of TRD. 2. Materials and methods 2.1. Animals This study received approval from the Animal Care, Health, and Safety Research Ethics Committee (NWU-AnimCareREC; South African National Health Research Ethics Council reg. no.: AREC-130913-015) of the Faculty of Health Sciences, North-West University, South Africa (ethics approval no.: NWU-00598-19-A5). Male SD rats (n = 108) were bred, supplied, and housed at the An imal Centre (Vivarium) of the Pre-Clinical Drug Development Platform (PCDDP) of the South African Department of Science and Innovation (DSI) and North-West University (NWU), South Africa (South African Veterinary Council (SAVC) registered (reg. no.: FR15/13458) and As sociation for Assessment and Accreditation of Laboratory Animal Care (AAALAC) accredited (international file #1717)). All rats were group- housed (2–3 rats/cage) in 395 × 346 × 213 mm (w x d x h) individu ally ventilated polysulphone cages under conditions of constant tem perature (22 ± 1 ◦C), humidity (55 ± 10%) and positive air pressure. The Vivarium air was exchanged 16 to 18 times (fresh uncirculated air) per hour, and the air quality was controlled with high-efficiency par ticulate air (HEPA) filters. A 12-h light/dark cycle (lights on between 06:00 and 18:00) was maintained. Cages were cleaned, and bedding (chipped corncob) was replaced weekly. The rats had access to a poly vinyl chloride (PVC) tunnel shelter and nesting material (paper towel) as environmental enrichment. Food (standard rat chow) and tap water were provided ad libitum. The rats and all experimental facilities were protected from outside noise. 2.2. Drug treatments During the sub-chronic treatment period, ACTH-(1–24) (ChinaPep tides Co., Ltd, Shanghai, China) was administered to the rats for 14 consecutive days at a fixed dosage of 0.1 mg/rat/day (Kitamura et al., 2008; Walker et al., 2013; Pereira et al., 2019). Then, rats received sub-acute administration of the following drugs alone and in different combinations: (1) imipramine hydrochloride (Sigma-Aldrich®, Schnelldorf, Germany) at 15 mg/kg (IMI-15) (Sales et al., 2011; Pereira et al., 2019), (2) escitalopram oxalate (Sigma-Aldrich®, Schnelldorf, Germany) at 15 mg/kg (ESC-15) (Mokoena et al., 2015) and (3) sil denafil citrate (Sigma-Aldrich®, Schnelldorf, Germany) at 3 mg/kg (SIL-3) or 10 mg/kg (SIL-10) (Liebenberg et al., 2010a). All rats were weighed daily during the sub-acute treatment period, with a dose adjustment calculation performed immediately before administration to enable accurate dosing in each rat. ACTH and drugs were all dissolved in phosphate-buffered saline (PBS) and subsequently administered via subcutaneous injection. The ACTH crystals were dissolved in a constant volume of PBS (0.1 ml), whereas the drug powders were dissolved in a maximum of 0.25 ml of PBS, using the smallest volume possible. Control rats received only vehicle control (VEH), i.e., PBS, as did the ACTH- naïve control rats. Previous studies from our laboratories informed the dosages selected for sildenafil and specific classes of traditional antidepressants chosen to constitute part of the combination therapies with sildenafil in this study. These studies found that the antidepressant-like properties of sildenafil display strict dose dependency, with only higher doses (≥10 mg/kg) exhibiting an interaction with the cholinergic system in rodents (Brink et al., 2008; Liebenberg et al., 2010a, 2010b). Hence, the antidepres sants imipramine (having anticholinergic properties) and escitalopram (not having anticholinergic properties) were used to investigate whether low- and/or high-dosage sildenafil can augment the antidepressant-like effects of an antidepressant with anticholinergic properties and one without these properties, thereby overcoming antidepressant treatment resistance in ACTH-treated SD rats. See Brink et al. (2008) for a detailed discussion of the role that the central cholinergic system plays in the antidepressant-like effects of sildenafil (Brink et al., 2008). J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 3 2.3. Experimental design All rat pups were weaned on postnatal day 21 (PND21) and randomly allocated to home cages (3 rats/cage). Rat pups were randomly divided into an ACTH-naïve and ACTH-treated group comprising 12 and 96 rats, respectively. Afterward, they were randomly divided into sub-acute treatment groups comprising 12 rats each, using the same procedure described above. All the rat pups, therefore, had an equal chance to be allocated to a specific home cage, the ACTH-naïve or ACTH-treated group, and any of the various sub-acute treatment groups, preventing possible cage effects. Importantly, the ACTH-naïve group comprised a single sub-acute treatment group, namely a VEH (negative control) [C1] group. However, the ACTH-treated group consisted of 8 different sub-acute treatment groups, including a VEH (negative con trol) [C2], IMI-15 (positive control) [D1], ESC-15 (positive control) [D2], SIL-3 [D3], SIL-10 [D4], IMI-15 + SIL-3 [D5], IMI-15 + SIL-10 [D6] and ESC-15 + SIL-3 [D7] group. Fig. 1 below illustrates the study layout. Starting on PND50, rats in the ACTH-naïve group daily received VEH (negative control) until PND63, whereas rats in the ACTH-treated group received ACTH. Sub-acute treatments followed immediately before the forced swim test (FST), as described below (Porsolt et al., 1977; Porsolt, 1979). • Dose 1: VEH and drugs corresponding to the various treatment groups were administered between 20:00 and 20:30 on PND63 (24 h before the forced swim test (FST) performed on the following day), • Dose 2: VEH and drugs corresponding to the various treatment groups were administered for a second time between 15:00 and 15:30 on PND64 (5 h before the FST performed on the same day), and • Dose 3: VEH and drugs corresponding to the various treatment groups were administered for a third and final time between 19:00 and 19:30 on PND64 (1 h before the FST performed on the same day). All the rats were habituated to (placed in) an open field test (OFT) arena for 10 min under red light (80 lux) on the evening of PND63, whereas the OFT test procedure was performed 24 h later, on the eve ning of PND64. Similarly, a 15-min pre-conditioning swim session was carried out for all the rats under red light (80 lux) on PND63 (directly after the OFT habituation session), followed 24 h later by the FST on PND64 (directly after the OFT test procedure). The OFT test procedure is usually performed in conjunction with the FST to eliminate possible drug-induced effects on the general locomotor activity of rats, viz., any changes in immobility observed during the FST can be ascribed to drug- induced changes in psychomotor and not general locomotor activity (Saayman et al., 2021). All rats were humanely euthanised on PND65 (at least 12 h following completion of the FST performed on the previous evening to eliminate the immediate effects that swimming stress may have on biochemical marker levels), whereafter plasma and brain (hippocampus) samples were collected and stored at − 80 ◦C for subse quent biochemical analyses. Finally, corticosterone levels were analysed in the plasma, whereas BDNF, serotonin, norepinephrine, KYNUA, QUINA, glutathione, and glutathione disulfide were analysed in the hippocampus. 2.4. Rodent model of antidepressant treatment resistance ACTH (0.1 mg/rat) was administered to rats daily between 17:00 and 19:00 through a single subcutaneous injection for 14 days, as described in previous studies (Kitamura et al., 2008; Walker et al., 2013; Pereira et al., 2019). This model is based on evidence suggesting that sub-chronic ACTH treatment negates the antidepressant-like action of monoaminergic antidepressants in novelty suppressed feeding (Antunes et al., 2015) and forced swim tests (FST) (Kitamura et al., 2002, 2008; Walker et al., 2013; Pereira et al., 2019). Therefore, the ACTH model of TRD possesses sound predictive validity, while its etiological and construct validity is founded on the underlying role of hyper cortisolaemia in the pathophysiology of the disorder (Hornig-Rohan et al., 1996; Markopoulou et al., 2009; Markopoulou, 2013; Caraci et al., 2018). 2.5. Statistical analyses Power analyses were performed to determine the minimum number Fig. 1. A schematic illustration of the study layout. ACTH: adrenocorticotropic hormone. BDNF: brain-derived neurotrophic factor. ESC-15: escitalopram oxalate (15 mg/kg). FST: forced swim test. IMI-15: imipramine hydrochloride (15 mg/kg). KYNUA: kynurenic acid. n: number of rats per treatment group. OFT: open field test. PND: postnatal day. QUINA: quinolinic acid. SIL-3: sildenafil citrate (3 mg/kg). SIL-10: sildenafil citrate (10 mg/kg). VEH: vehicle control. * A total of 3 VEH or drug doses were administered during this period (24, 5, and 1 h before the FST was performed). J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 4 of rats needed per treatment group to ensure statistically meaningful results. In cases where a one-way analysis of variance (ANOVA) was performed, power calculations were conducted considering the omnibus test (F statistic), and the standardised effect size was defined as medium to large (F = 0.25 and F = 0.40, respectively). Power calculations were conducted in all cases considering a type-I error rate of 5% (α = 0.05). The threshold for type II error was 20% (1-β ≥ 0.8). Power calculations were conducted using G*power version 3.1.9.2. GraphPad Prism® version 8 for Windows (GraphPad Prism® soft ware, Version 8.0, San Diego, California, United States of America) was used for all statistical analyses and graphical representations. Grubbs’ test was performed on all data sets to screen for outliers (with α = 0.05 accepted as significant). The Shapiro-Wilk’s and Levene’s tests were conducted to test for normality of distribution and homogeneity of variances (with p < 0.05 accepted as a violation of both assumptions), respectively. An unpaired Student’s t-test (data distributed normally) or Mann-Whitney U test (data not distributed normally) was used to determine the effects of sub-chronic ACTH treatment in rats (by comparing the ACTH-treated to the ACTH-naïve control group). Furthermore, to assess the impact of different sub-acute treatments in ACTH-treated rats, an Ordinary one-way ANOVA (multiple comparisons for data distributed normally) or Kruskal-Wallis one-way ANOVA (numerous comparisons for data not distributed normally) was per formed (by comparing the various sub-acute treatment groups to each other). Ordinary one-way ANOVAs were followed by a Dunnett’s post- hoc test for multiple comparisons. In contrast, Kruskal-Wallis one-way ANOVAs were followed by a Dunn’s test (comparing the various sub- acute drug treatment groups to the VEH treatment group in ACTH- treated rats). Data are presented as the mean ± standard error of the mean (SEM). Statistical significance was set at p ≤ 0.05 for all comparisons. Statistical analyses were followed by effect magnitude calculations, where appropriate. Effect magnitudes indicate strong trends and mini mise Type I (false positive) or Type II (false negative) errors, specifically in instances where the assumption of homogeneity of variances is violated. The unbiased Cohen’s d-value (d) was used to calculate the effect magnitude of interactions and intergroup differences (with a 95% confidence interval (CI) of the effect magnitude). The d was calculated to establish the practical (clinical) significance of effect magnitude. Only medium (d ≥ 0.6) and large (d ≥ 0.8) effect sizes were considered sig nificant. Effect magnitude indicators were calculated in Exploratory Software for CIs (Cohen, 1988). 2.6. Behavioural tests Prior studies in our laboratory have shown that preceding behav ioural tests do not influence the outcome of subsequent consecutive tests when behavioural tests are performed from least to most stressful (Mokoena et al., 2015). Thus, the OFT was performed (50 min after the third and final dose of VEH or drug treatment) immediately before the FST. As rats are nocturnal animals, behavioural tests were performed during the dark cycle (between 18:00 and 06:00). More specifically, to accommodate for the initial foraging and activity of rats, behavioural tests only commenced 1 h after the start of the dark cycle (at 19:00). 2.6.1. Open field test (OFT) The OFT test procedure is performed to evaluate the general loco motor activity of rodents and, hence, the ability to move around and negotiate their surroundings (Schoeman et al., 2017). Since more anxious rodents will tend to remain stationary in a corner or next to a wall of the OFT arena (Ramos et al., 1997; Prut and Belzung, 2003; Hiroi and Neumaier, 2006), increased anxiety-like behaviour in the OFT (neophobia, e.g., freezing) (Misslin and Cigrang, 1986) may influence general locomotor activity as assessed in the OFT. As a result, the habituation of rats in the OFT arena (24 h before performing the OFT test procedure) was performed to negate this possibility. The OFT apparatus consisted of four 100 × 100 × 45 cm (w x d x h) square arenas inscribed on the floor, with opaque black walls. A video camera situated directly above each OFT arena recorded locomotor behaviour. The test was performed on PND64 under red light (80 lux), as previously described (Schoeman et al., 2017; Saayman et al., 2021). In short, each rat was placed in the centre of an OFT arena and allowed 5 min for exploration. Following each test session, the rats were returned to their home cages, and whereafter the OFT arenas were wiped with a 10% ethanol solution to eliminate any olfactory cues during subsequent test sessions. The video recordings were scored using EthoVision XT 14 software (Noldus Information Technology BV, Wageningen, Netherlands), with the total distance moved (cm) used to measure general locomotor activity. 2.6.2. Forced swim test (FST) The FST is routinely performed to evaluate antidepressant-like ac tivity for a broad spectrum of antidepressants (Borsini and Meli, 1988). The test is based on deficits in escape-directed behaviour when rats are subjected to an inescapable water cylinder and the reversal thereof by an antidepressant. Adopting an immobile posture, i.e., immobility, in the FST has been associated with a failure of perseverance in escape-directed behaviour (behavioural despair) and reflects depressive-like behaviour (Porsolt et al., 1977; Porsolt, 1979). A pre-conditioning swim on day 1 is required to sensitize the animals, which subsequently present with behavioural despair when reintro duced to the FST on day 2 of the test (Pereira et al., 2019; Saayman et al., 2021). The FST apparatus consisted of four 20 × 100 cm (d x h) cylindrical tanks spaced next to each other, each filled to a depth of 40 cm with ambient water (maintained at 25 ± 1 ◦C), and a video camera situated directly in front of the four FST cylinders. The FST was performed on PND64 under red light (80 lux), as previously described (Schoeman et al., 2017; Pereira et al., 2019). In short, each rat was placed into a water-filled FST cylinder and allowed to swim for 7 min while being video recorded, following which the rats were removed from the cyl inders, dried with a towel, and returned to their home cages. The water in the FST cylinders was replaced with clean water following every FST trial to avoid any influence of odour trails on subsequent FST trials (Abel and Bilitzke, 1990). The video recordings were scored using EthoVision XT 14 software (Noldus Information Technology BV, Wageningen, Netherlands), with total time spent immobile (sec) used to measure their depressive-like behaviour. However, the first and last minutes of the FST were not scored (leaving a total of 5 min that were scored) to maximise the accuracy of the results (Oberholzer et al., 2018). If necessary, the investigator was out of sight but close at hand to intervene should there be signs of untoward submersion. 2.7. Biochemical assays Specific biomarkers, including plasma corticosterone and hippo campal BDNF, serotonin, norepinephrine, KYNUA, QUINA, glutathione, and glutathione disulfide levels, were analysed, as described in sections 2.7.1.1., 2.7.2.1. and 2.7.2.2. 2.7.1. Blood collection and processing Following the euthanasia of rats by decapitation (without anaes thesia) on PND65, their trunk blood was collected in pre-chilled 4 ml BD Vacutainer® tubes containing 7.2 mg K2-EDTA (dipotassium ethyl enediaminetetraacetic acid). The blood-filled tubes were manually mixed before being placed on ice. The blood samples were subsequently centrifuged (at 14 000 rpm and 4 ◦C for 10 min), whereafter the plasma layer (top layer) was pipetted into 1.5 ml Eppendorf® microcentrifuge tubes and subsequently preserved at − 80 ◦C until the plasma cortico sterone assay (Möller et al., 2013). J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 5 2.7.1.1. Plasma corticosterone assay. Plasma samples were analysed for corticosterone using liquid chromatography-mass spectrometry (LC-MS) as described previously (Li et al., 2014; Yuan et al., 2015). 2.7.2. Brain tissue collection and preparation Immediately after euthanasia, the whole brain of individual rats was extracted and placed in ice-cold PBS. The hippocampi were quickly dissected on an ice-cooled dissection slab and placed into 1.5 ml Eppendorf® microcentrifuge tubes. The tubes were subsequently snap frozen in liquid nitrogen and preserved at − 80 ◦C until the hippocampal biomarker assays were performed (Harvey et al., 2006a; Brand and Harvey, 2017; Steyn et al., 2018). 2.7.2.1. Hippocampal brain-derived neurotrophic factor (BDNF) assay. Fig. 2. A graphical representation of the OFT and FST data. A Distance moved by rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. B Distance moved by rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. C Time spent immobile by rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. D Time spent immobile by rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. All group sizes are equal (n = 12). Data points represent the mean ± SEM. Statistical analyses are reported in the text with d vs the VEH + VEH treatment group in A, **p < 0.01, and d vs the ACTH + VEH treatment group in B, #p < 0.05, and d vs the VEH + VEH treatment group in C, and * p < 0.05, **p < 0.01 and d vs the ACTH + VEH treatment group in D. ACTH: adrenocorticotropic hormone. d: unbiased Cohen’s d-value. ESC-15: escitalopram oxalate (15 mg/kg). FST: forced swim test. IMI-15: imipramine hydrochloride (15 mg/kg). OFT: open field test. SIL-3: sildenafil citrate (3 mg/kg). SIL- 10: sildenafil citrate (10 mg/kg). VEH: vehicle control. J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 6 Hippocampal BDNF levels were measured using rat BDNF enzyme- linked immunosorbent assay (ELISA) kits (catalogue no.: E-EL-R1235, Elabscience Biotechnology Inc.) according to the manufacturer in structions (Steyn et al., 2018; Saayman et al., 2021). 2.7.2.2. Hippocampal serotonin, norepinephrine, kynurenic acid (KYNUA), quinolinic acid (QUINA), glutathione, and glutathione disulfide assay. According to previous methods, hippocampal tissue samples were assayed for serotonin, norepinephrine, KYNUA, QUINA, gluta thione, and glutathione disulfide using LC-MS (Moriarty et al., 2011; Squellerio et al., 2012; Fuertig et al., 2016; Wojnicz et al., 2016; Wang et al., 2019). 3. Results As illustrated by graphs A, C, E, and G (see sections 3.1.1. and 3.2.1. to 3.2.5.), the ACTH-treated group is compared to the ACTH-naïve control group, with both these groups comprising SD rats that received 3 doses of VEH during the sub-acute treatment period of this study. Graphs B, D, F, and H respectively depict the same ACTH-treated groups as in graphs A, C, E, and G, which serve as the control groups for the ACTH plus various sub-acute drug treatment groups. 3.1. Behavioural analysis Data were obtained from the behavioural tests performed on PND64 once all the VEH and drug treatments had been administered. 3.1.1. Distance moved and time spent immobile Fig. 2 below depicts the total distance moved in the OFT and time spent immobile in the FST by SD rats after they received sub-chronic (over 14 days) VEH or ACTH treatment, followed by sub-acute (over 24 h) VEH or drug treatment. Since FST data are corrected for possible drug-induced changes in the general locomotor activity of rats using analyses of covariance (ANCOVAs) in this study, changes observed in the time spent immobile in the FST can be directly ascribed to drug- induced changes in psychomotor (and not general locomotor) activity. An unpaired Student’s t-test revealed no statistically significant dif ference between the ACTH + VEH (ACTH-treated) and the VEH + VEH (ACTH-naïve) control group regarding distance moved in the OFT (p = 0.2874, d = 0.43) (see Fig. 2A). A Kruskal-Wallis one-way ANOVA test for multiple comparisons indicated statistically significant differences between the various sub-acute treatment groups receiving ACTH regarding distance moved in the OFT (p < 0.0001). Dunn’s post-hoc test for multiple comparisons showed in ACTH-treated rats that sub-acute IMI-15 + SIL-3 (p = 0.0013, d = 2.09) and IMI-15 + SIL-10 (p = 0.0065, d = 1.65) treatments reduce the distance moved in the OFT compared to VEH (see Fig. 2B). An unpaired Student’s t-test revealed a statistically significant dif ference between the ACTH + VEH and the VEH + VEH control group regarding time spent immobile in the FST. Sub-chronic ACTH treatment elevates the time spent immobile by SD rats in the FST compared to VEH (p = 0.0322, d = 0.92) (see Fig. 2C). An Ordinary one-way ANOVA test for multiple comparisons indicated, all in ACTH-treated rats, that sta tistically significant differences were evident between the various sub- acute treatment groups regarding time spent immobile in the FST (F(7, 88) = 2.983, p = 0.0074). Dunnett’s post-hoc test for multiple compar isons showed, all in ACTH-treated rats, that sub-acute SIL-10 (p = 0.0111, d = 1.30), IMI-15 + SIL-3 (p = 0.0016, d = 1.34) and ESC-15 + SIL-3 (p = 0.0108, d = 1.43) treatments significantly reduce the time spent immobile in the FST compared to VEH-treated rats (see Fig. 2D). 3.2. Rat biochemical marker levels Data were obtained from the various biochemical assays conducted on PND65 once all the VEH and drug treatments had been administered and behavioural tests performed. 3.2.1. Plasma corticosterone levels Fig. 3 below depicts the total plasma corticosterone levels of SD rats after receiving sub-chronic VEH or ACTH treatment, followed by sub- acute VEH or drug treatment. An unpaired Student’s t-test revealed no statistically significant dif ference in plasma corticosterone levels between the ACTH + VEH and the VEH + VEH control group (p = 0.1655) (see Fig. 3A). However, an effect magnitude calculation demonstrated a practical (clinical) signif icant difference between these groups regarding plasma corticosterone levels (d = 0.61; Fig. 3A). A Kruskal-Wallis one-way ANOVA test for multiple comparisons across ACTH-treated rats indicated statistically significant differences in plasma corticosterone levels between the various sub-acute treatment groups (p < 0.0001). Dunn’s post-hoc test for multiple comparisons showed, all in ACTH-treated rats, that sub- acute ESC-15 + SIL-3 (p = 0.0034, d = 1.28) treatment significantly reduces plasma corticosterone levels compared to VEH-treated rats (see Fig. 3B). 3.2.2. Hippocampal brain-derived neurotrophic factor (BDNF) levels Fig. 4 below depicts the total hippocampal BDNF levels of SD rats after receiving sub-chronic VEH or ACTH treatment, followed by sub- acute VEH or drug treatment. An unpaired Student’s t-test revealed a statistically significant dif ference between the ACTH + VEH and the VEH + VEH control group regarding hippocampal BDNF levels. Therefore, sub-chronic ACTH treatment significantly elevates hippocampal BDNF levels compared to VEH-treated SD rats (p = 0.0074, d = 1.16) (see Fig. 4A). An Ordinary one-way ANOVA test for multiple comparisons, all in ACTH-treated rats, indicated statistically significant differences between the various sub- acute treatment groups regarding hippocampal BDNF levels (F(7, 87) = 8.417, p < 0.0001). Dunnett’s post-hoc test for multiple comparisons showed, all in ACTH-treated rats, that sub-acute IMI-15 (p < 0.0001, d = 2.37), ESC-15 (p < 0.0001, d = 2.19), SIL-10 (p = 0.0336, d = 1.25), IMI- 15 + SIL-3 (p < 0.0001, d = 1.51), IMI-15 + SIL-10 (p = 0.0002, d = 1.54) and ESC-15 + SIL-3 (p < 0.0001, d = 1.74) treatments significantly reduce hippocampal BDNF levels compared to VEH-treated rats (see Fig. 4B). 3.2.3. Hippocampal serotonin and norepinephrine levels Fig. 5 below depicts the total hippocampal serotonin and norepi nephrine levels of SD rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH or drug treatment. An unpaired Student’s t-test revealed a statistically significant dif ference between the ACTH + VEH and the VEH + VEH control group regarding hippocampal serotonin levels. Therefore, sub-chronic ACTH treatment significantly elevates hippocampal serotonin levels compared to VEH-treated SD rats (p = 0.0174, d = 1.01) (see Fig. 5A). An Ordinary one-way ANOVA test for multiple comparisons, all in ACTH-treated rats, indicated statistically significant differences between the various sub- acute treatment groups regarding hippocampal serotonin levels (F(7, 85) = 3.521, p = 0.0023). Dunnett’s post-hoc test for multiple compar isons showed, all in ACTH-treated rats, that sub-acute IMI-15 (p = 0.0004, d = 1.40), ESC-15 (p = 0.0122, d = 1.30), SIL-3 (p = 0.0138, d = 1.25), SIL-10 (p = 0.0065, d = 1.22) and ESC-15 + SIL-3 (p = 0.0060, d = 1.26) treatments significantly reduce hippocampal serotonin levels compared to VEH-treated rats (see Fig. 5B). An unpaired Student’s t-test revealed a statistically significant dif ference between the ACTH + VEH and the VEH + VEH control group regarding hippocampal norepinephrine levels. Therefore, sub-chronic ACTH treatment significantly elevates hippocampal norepinephrine levels compared to VEH-treated SD rats (p = 0.0064, d = 1.22) (see Fig. 5C). An Ordinary one-way ANOVA test for multiple comparisons indicated, all in ACTH-treated rats, statistically significant differences J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 7 between the various sub-acute treatment groups regarding hippocampal norepinephrine levels (F(7, 87) = 9.317, p < 0.0001). Dunnett’s post-hoc test for multiple comparisons showed in ACTH-treated rats that sub- acute SIL-3 (p = 0.0134, d = 1.01) and SIL-10 (p = 0.0417, d = 0.89) treatments significantly reduced hippocampal norepinephrine levels compared to VEH-treated rats (see Fig. 5D). Fig. 3. A graphical representation of the plasma corticosterone assay data. A Corticosterone levels of rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. B Corticosterone levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. All group sizes are equal (n = 12). Data points represent the mean ± SEM. Statistical analyses are reported in the text with d vs the VEH + VEH treatment group in A, and **p < 0.01 and d vs the ACTH + VEH treatment group in B. ACTH: adrenocorticotropic hormone. d: unbiased Cohen’s d-value. ESC-15: escitalopram oxalate (15 mg/ kg). IMI-15: imipramine hydrochloride (15 mg/kg). SIL-3: sildenafil citrate (3 mg/kg). SIL-10: sildenafil citrate (10 mg/kg). VEH: vehicle control. Fig. 4. A graphical representation of the hippocampal BDNF assay data. A BDNF levels of rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. B BDNF levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. All group sizes are equal (n = 12). Data points represent the mean ± SEM. Statistical analyses are reported in the text with ##p < 0.01 and d vs the VEH + VEH treatment group in A, and * p < 0.05, ***p < 0.001, ****p < 0.0001 and d vs the ACTH + VEH treatment group in B. ACTH: adrenocorticotropic hormone. BDNF: brain-derived neurotrophic factor. d: unbiased Cohen’s d-value. ESC-15: escitalopram oxalate (15 mg/kg). IMI-15: imipramine hydrochloride (15 mg/kg). SIL-3: sildenafil citrate (3 mg/kg). SIL- 10: sildenafil citrate (10 mg/kg). VEH: vehicle control. J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 8 3.2.4. Hippocampal kynurenic acid (KYNUA) and quinolinic acid (QUINA) levels Fig. 6 below depicts the total hippocampal KYNUA and QUINA levels of SD rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH or drug treatment. An unpaired Student’s t-test revealed a statistically significant dif ference between the ACTH + VEH and the VEH + VEH control group regarding hippocampal KYNUA levels. Therefore, sub-chronic ACTH treatment elevates hippocampal KYNUA levels compared to VEH- treated SD rats (p = 0.0041, d = 1.26) (see Fig. 6A). An Ordinary one- way ANOVA test for multiple comparisons, all in ACTH-treated rats, indicated statistically significant differences between the various sub- acute treatment groups regarding hippocampal KYNUA levels (F(7, 86) = 11.84, p < 0.0001). Dunnett’s post-hoc test for multiple comparisons Fig. 5. A graphical representation of the hippocampal serotonin and norepinephrine assay data. A Serotonin levels of rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. B Serotonin levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. C Norepinephrine levels of rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. D Norepinephrine levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. All group sizes are equal (n = 12). Data points represent the mean ± SEM. Statistical analyses are reported in the text with #p < 0.05 and d vs the VEH + VEH treatment group in A, *p < 0.05, **p < 0.01, ***p < 0.001, and d vs the ACTH + VEH treatment group in B, ##p < 0.01 and d vs the VEH + VEH treatment group in C, and * p < 0.05 and d vs the ACTH + VEH treatment group in D. ACTH: adrenocorticotropic hormone. d: unbiased Cohen’s d-value. ESC-15: escitalopram oxalate (15 mg/kg). IMI-15: imipramine hydrochloride (15 mg/kg). SIL-3: sildenafil citrate (3 mg/kg). SIL-10: sildenafil citrate (10 mg/kg). VEH: vehicle control. J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 9 showed, all in ACTH-treated rats, that sub-acute IMI-15 (p < 0.0001, d = 1.76), ESC-15 (p < 0.0001, d = 1.69), SIL-3 (p = 0.0457, d = 0.65), SIL- 10 (p < 0.0001, d = 2.05), IMI-15 + SIL-3 (p < 0.0001, d = 1.28), IMI-15 + SIL-10 (p < 0.0001, d = 1.25) and ESC-15 + SIL-3 (p < 0.0001, d = 1.60) treatments significantly reduce hippocampal KYNUA levels compared to VEH-treated rats (see Fig. 6B). An unpaired Student’s t-test revealed a statistically significant difference between the ACTH + VEH and the VEH + VEH control group regarding hippocampal QUINA levels. Therefore, sub-chronic ACTH treatment significantly elevates hippocampal QUINA levels compared to VEH-treated SD rats (p = 0.0052, d = 1.22) (see Fig. 6C). A Kruskal- Wallis one-way ANOVA test for multiple comparisons, all in ACTH- treated rats, indicated statistically significant differences between the various sub-acute treatment groups regarding hippocampal QUINA Fig. 6. A graphical representation of the hippocampal KYNUA and QUINA assay data. A KYNUA levels of rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. B KYNUA levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. C QUINA levels of rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. D QUINA levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. All group sizes are equal (n = 12). Data points represent the mean ± SEM. Statistical analyses are reported in the text with ##p < 0.01 and d vs the VEH + VEH treatment group in A, *p < 0.05, ****p < 0.0001, and d vs the ACTH + VEH treatment group in B, ##p < 0.01 and d vs the VEH + VEH treatment group in C, and **p < 0.01, ***p < 0.001 and d vs the ACTH + VEH treatment group in D. ACTH: adrenocorticotropic hormone. d: unbiased Cohen’s d-value. ESC-15: escitalopram oxalate (15 mg/kg). IMI-15: imipramine hydrochloride (15 mg/kg). KYNUA: kynurenic acid. QUINA: quinolinic acid. SIL-3: sildenafil citrate (3 mg/kg). SIL-10: sildenafil citrate (10 mg/kg). VEH: vehicle control. J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 10 levels (p = 0.0017). Dunn’s post-hoc test for multiple comparisons showed, all in ACTH-treated rats, that sub-acute IMI-15 (p = 0.0025, d = 1.49), SIL-10 (p = 0.0002, d = 1.61) and IMI-15 + SIL-3 (p = 0.0044, d = 1.36) treatments significantly reduce hippocampal QUINA levels compared to VEH-treated rats (see Fig. 6D). 3.2.5. Hippocampal glutathione and glutathione disulfide levels Fig. 7 below depicts the total hippocampal glutathione and glutathione disulfide levels of SD rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH or drug treatment. A Mann-Whitney U test revealed a statistically significant difference between the ACTH + VEH and the VEH + VEH control group regarding hippocampal glutathione levels. Therefore, sub-chronic ACTH treatment significantly elevates hippocampal glutathione levels compared to VEH- treated SD rats (p < 0.0001, d = 2.71) (see Fig. 7A). A Kruskal-Wallis one-way ANOVA test for multiple comparisons, all in ACTH-treated Fig. 7. A graphical representation of the hippocampal glutathione and glutathione disulfide assay data. A Glutathione levels of rats after receiving sub- chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. B Glutathione levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. C Glutathione disulfide levels of rats after receiving sub-chronic VEH or ACTH treatment, followed by sub-acute VEH treatment. D Glutathione disulfide levels of rats after receiving sub-chronic ACTH treatment, followed by sub-acute VEH or drug treatment. All group sizes are equal (n = 12). Data points represent the mean ± SEM. Statistical analyses are reported in the text with ####p < 0.0001 and d vs the VEH + VEH treatment group in A, *p < 0.05 and d vs the ACTH + VEH treatment group in B, and **p < 0.01 and d vs the ACTH + VEH treatment group in D. ACTH: adrenocorticotropic hormone. d: unbiased Cohen’s d-value. ESC-15: escitalopram oxalate (15 mg/kg). IMI-15: imipramine hydrochloride (15 mg/kg). SIL-3: sildenafil citrate (3 mg/kg). SIL-10: sildenafil citrate (10 mg/kg). VEH: vehicle control. J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 11 rats, indicated statistically significant differences between the various sub-acute treatment groups regarding hippocampal glutathione levels (p = 0.0016). Dunn’s post-hoc test for multiple comparisons showed, all in ACTH-treated rats, that sub-acute SIL-10 (p = 0.0125, d = 1.21) treatment significantly reduces hippocampal glutathione levels compared to VEH-treated rats (see Fig. 7B). A Mann-Whitney U test revealed no statistically significant differ ence between the ACTH + VEH and the VEH + VEH control group regarding hippocampal glutathione disulfide levels (p = 0.5899, d = 0.12) (see Fig. 7C). An Ordinary one-way ANOVA test for multiple comparisons, all in ACTH-treated rats, indicated statistically significant differences between the various sub-acute treatment groups regarding hippocampal glutathione disulfide levels (F(7, 88) = 6.091, p < 0.0001). Dunnett’s post-hoc test for multiple comparisons showed, all in ACTH- treated rats, that sub-acute IMI-15 + SIL-10 (p = 0.0015, d = 1.13) treatment significantly elevates hippocampal glutathione disulfide levels compared to VEH-treated rats (see Fig. 7D). 4. Discussion In this study, the ACTH-induced rodent model of TRD induced sig nificant antidepressant-like effects together with noteworthy changes in certain depression-related biomarkers. However, the latter were for the most part unresponsive to sub-acute treatments with IMI-15, ESC-15, SIL-3, and IMI-15 + SIL-10, but responsive to SIL-10, IMI-15 + SIL-3, and ESC-15 + SIL-3. Sub-chronic ACTH treatment had no significant impact on locomotor activity (Fig. 2A), consistent with earlier studies (Kitamura et al., 2002; Walker et al., 2013; Pereira et al., 2019). In ACTH-treated rats, sub-acute IMI-15 and ESC-15 did not significantly alter locomotor activity (Fig. 2B), but IMI-15 exhibited a tendency toward reduced activity (d = 1.02). Previous studies in ACTH-naïve rats reported variable effects for imipramine (Liebenberg et al., 2010b; Guan et al., 2014) and negligible effects for escitalopram (Mokoena et al., 2015). Furthermore, SIL-3 and SIL-10 did not significantly alter locomotor activity (Fig. 2B), consistent with prior research (Liebenberg et al., 2010a, 2010b; Saayman et al., 2021). However, IMI-15 + SIL-3 and IMI-15 + SIL-10 significantly reduced locomotor activity, suggesting that both doses of sildenafil augmented the effect of IMI-15 on locomotor activity. Following ANCOVA adjustments for locomotor activity, immobility in Fig. 2C and D reflects psychomotor activity, i.e., depressive-like behaviour. ACTH significantly increased depressive-like behaviour (time spent immobile) in the FST compared to VEH-treated animals (Fig. 2C), indicating a causal role for robust activation of the hypothalamic-pituitary-adrenal (HPA) axis, as seen in MDD (Pariante and Lightman, 2008; Baumeister et al., 2016) and TRD (Juruena et al., 2013; Markopoulou, 2013). Nevertheless, similar studies have failed to show exaggerated depressive-like behaviour following ACTH treatment (Walker et al., 2013; Pereira et al., 2019). As expected, IMI-15 and ESC-15 did not significantly alleviate depressive-like behaviour in ACTH-treated rats (Fig. 2D), confirming that the model exhibits treat ment resistance to at least two classes of monoaminergic active antide pressants, aligned with previous studies (Walker et al., 2013; Antunes et al., 2015; Pereira et al., 2019). SIL-10, but not SIL-3, significantly reduced depressive-like behaviour in ACTH-treated rats (Fig. 2D), not only reaffirming sildenafil’s previ ously demonstrated antidepressant properties (Brink et al., 2008; Wang et al., 2014; Saayman et al., 2021), but also highlighting for the first time its ability to overcome antidepressant resistance in a dose-dependent manner. Furthermore, that IMI-15 + SIL-3 and ESC-15 + SIL-3 signifi cantly reduced depressive-like behaviour, where either antidepressant as monotherapy was ineffective, demonstrates that low-dose sildenafil can augment the activity of traditional antidepressants. IMI-15 (having antimuscarinic properties) + SIL-10 treatment, unlike either drug alone, showed a notable trend toward reducing depressive-like behaviour (d = 0.85) (Fig. 2D). Only higher doses of sildenafil (≥10 mg/kg) have been demonstrated before to require an antimuscarinic agent to reveal an antidepressant-like effect in a rodent depression model (Brink et al., 2008; Liebenberg et al., 2010a), which is not a finding replicated in the ACTH model of TRD. Despite inducing a depressive-like state (Fig. 2C), ACTH did not significantly raise plasma corticosterone levels (Fig. 3A). Others observed a transient increase in plasma corticosterone by ACTH, peak ing on day 4 and returning to normal levels afterward, suggesting HPA system down-regulation (Walker et al., 2013). Clinical data (Carroll et al., 2007) supports the idea that elevated ACTH secretion typically does not lead to long-term hypercortisolaemia (Walker et al., 2013). Nevertheless, ACTH caused a trend toward increased plasma cortico sterone levels, with a medium effect size (d = 0.61) (Fig. 3A), as reported before (Pereira et al., 2019). Our findings confirm that ACTH causes a sustained depressive-like state in rats resistant to antidepressant treat ment. HPA axis hyperactivity and impaired negative feedback regula tion are strongly linked to MDD and TRD (Barden, 2004; Boyle et al., 2005; Anacker et al., 2011; Baumeister et al., 2016), with chronic an tidepressant treatment downregulating the HPA axis (Satoh et al., 1985; Farahbakhsh and Radahmadi, 2022) as a critical drug target (Frost et al., 2003). IMI-15 and ESC-15 failed to significantly reduce plasma corticoste rone levels (Fig. 3B), possibly contributing to the treatment resistance. Similar results have been reported before for amitriptyline and imipra mine (Antunes et al., 2015), whereas Bano et al. (2010) reported con flicting results for antidepressants, e.g., that sertraline and moclobemide had no significant effect on corticosterone, while citalopram reduced corticosterone, and tianeptine increased them (Bano et al., 2010). Sub-acute SIL-3, SIL-10, IMI-15 + SIL-3, and IMI-15 + SIL-10 treatments also failed to significantly affect plasma corticosterone levels in ACTH-treated rats (Fig. 3B). Importantly, despite some evidence for drug-induced changes in corticosterone, i.e., ESC-15 + SIL-3 treatment reducing its levels, corticosterone was not elevated in ACTH-treated rats; a study limitation that complicates the interpretation of treatment ef fects on corticosterone, warranting further research. Consequently, no precise and reliable interpretation of treatment-related effects in ACTH-treated rats is possible. ACTH unexpectedly raised hippocampal BDNF levels (Fig. 4A), contrary to the typical association of reduced BDNF levels with MDD (Lee and Kim, 2010), TRD (Li et al., 2007), and HPA axis hyperactivity (Jacobsen and Mørk, 2006; Huang et al., 2011; Gong et al., 2016). That said, IMI-15 and ESC-15 reduced this elevation in hippocampal BDNF levels in ACTH-treated rats (Fig. 4B). While these findings oppose the negative neurotrophic hypothesis of MDD, which links BDNF reversal with antidepressant efficacy (Duman and Monteggia, 2006; Polyakova et al., 2015; Duman et al., 2016), there are considerations to keep in mind. Changes in BDNF may be related to co-presenting metabolic and redox factors, with BDNF potentially exerting counter-regulatory effects on glutathione oxidation or mediating redox effects itself, contributing to mood disorder development (Harvey et al., 2012; Brand et al., 2015). Yet sub-acute SIL-10, IMI-15 + SIL-3, IMI-15 + SIL-10, and ESC-15 + SIL-3 treatments all significantly reduced hippocampal BDNF levels in ACTH-treated rats, with sub-acute SIL-3 showing a trend toward reduction (d = 0.89) (Fig. 4B). The counter-regulatory role of BDNF may be dependent on co-morbid redox and inflammatory conditions (Harvey et al., 2012). Importantly, these states may reflect illness severity, and in turn response to antidepressants (Mokoena et al., 2010, 2015). Yet, it is unlikely that the antidepressant-like effects induced by sildenafil in the FST are solely mediated by a downregulation in BDNF signalling, considering the same reduction in the hippocampal levels of BDNF was observed following sub-acute IMI-15, ESC-15, and IMI-15 + SIL-10 treatments, which failed to reduce the immobility evident in ACTH-treated rats. ACTH significantly raised hippocampal serotonin and norepineph rine levels in rats (Fig. 5A and C, respectively), contrary to predictions of the biogenic amine hypothesis of MDD (Brand et al., 2015). Others have J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 12 also reported increased prefrontocortical and hippocampal serotonin and norepinephrine levels in rats after ACTH treatment (Li et al., 1990; Walker et al., 2013). Notably, IMI-15 and ESC-15 reduced elevated hippocampal serotonin levels in ACTH-treated rats while leaving norepinephrine levels unaffected (Fig. 5B and D). Interestingly, SIL-3 and SIL-10 significantly lowered norepinephrine levels (Fig. 5D), with ESC-15 + SIL-3 reducing only hippocampal serotonin levels (Fig. 5B) and showing a trend toward reducing norepinephrine levels (d = 0.67) (Fig. 5D). However, IMI-15 + SIL-3 and IMI-15 + SIL-10 did not significantly impact hippocampal serotonin and norepinephrine levels (Fig. 5B and D), with a trend toward reducing serotonin levels (d = 0.62 and d = 0.70, respectively). Unlike that observed in ACTH-treated SD rats, previous reports suggest that sildenafil partly exerts its antidepressant-like effects in Flinders Sensitive Line (FSL) rats, i.e., an animal model of MDD, by modulating the monoaminergic system, particularly enhancing serotonergic and/or noradrenergic neurotrans mission (Liebenberg et al., 2010a, 2010b). The reduction in time spent immobile in the FST induced by sildenafil can also not be fully explained by reduced hippocampal serotonergic and noradrenergic levels, since this was not consistently demonstrated across the various sildenafil mono- and combination therapies. Multimodal neurobiological mecha nisms appear to be involved. Based on the above findings, we conclude that the neurobiological profile of TRD may differ from that of MDD, which is not surprising since TRD resists monoaminergic active antidepressants. Thus, the ACTH- induced rodent model of TRD exhibits elevated hippocampal BDNF levels, whereas genetic rodent models of MDD, like the FSL rat, demonstrate the opposite (Elfving et al., 2010; Overstreet and Wegener, 2013). Both models, however, share increased hippocampal serotonin and norepinephrine levels, as observed here and in other studies (Zan gen et al., 1997, 1999). ACTH may, therefore, counteract the effects of monoaminergic antidepressants by blocking their primary mechanism of action, thereby failing to elevate central synaptic monoaminergic and subsequently BDNF levels. SIL-10, IMI-15 + SIL-3, and ESC-15 + SIL-3 may induce their antidepressant-like effects (in the FST) independently of the mono aminergic system and BDNF signalling pathway, explaining sildenafil’s effectiveness in a rodent model resistant to monoaminergic antidepres sants. In fact, the antidepressant-like action of PDE5 inhibitors may partially depend on PDE5 inhibition and subsequent cGMP signalling (Brink et al., 2008; Liebenberg et al., 2010a). cGMP modulates various monoamine transmitter systems (Harvey, 1996), along with roles in inflammation (Rapôso et al., 2014; Nguyen et al., 2022), oxidative stress (Masood et al., 2008; Curatola et al., 2011), and neuroplasticity (Gallo and Iadecola, 2011; Reierson et al., 2011), all implicated in MDD and antidepressant action (Brand et al., 2015). That said, the tryptophan-kynurenine pathway may serve as a con necting mechanism. Current MDD literature suggests a shift in trypto phan metabolism away from serotonin, playing a vital role in mood regulation (Marx et al., 2021; Hestad et al., 2022), specifically towards the kynurenine pathway (Brand et al., 2015) which is involved in neu roinflammation via glutamatergic actions (Moller et al., 2015). Elevated indoleamine-2,3-dioxygenase (IDO), as activated by pro-inflammatory cytokines, depletes tryptophan and its derived neurotransmitters like serotonin, while increasing downstream kynurenine metabolites, such as neuroprotective KYNUA and neurotoxic QUINA. The KYNUA:QUINA ratio determines a subtle balance between neuroprotection and neuro toxicity (Brown et al., 2021; Marx et al., 2021). In this study, ACTH significantly increased hippocampal KYNUA and QUINA levels (Fig. 6A and C), which may indicate maintenance of a neuroprotective balance (Moller et al., 2015). MDD is often associated with a shift in kynurenine metabolism towards QUINA, especially in cases of hippocampal shrinkage (Savitz et al., 2015; Marx et al., 2021; Öztürk et al., 2021; Hestad et al., 2022). Importantly, antidepressants can reverse or stabi lize this shift (Halaris et al., 2015; Sun et al., 2020; Ou et al., 2023). IMI-15 and ESC-15 significantly reduced elevated hippocampal KYNUA levels, whereas IMI-15 but not ESC-15 reduced QUINA (Fig. 6B and D). Contrary to some of our findings, monoaminergic antidepres sants usually reduce QUINA levels and increase KYNUA levels in MDD patients (Halaris et al., 2015; Kocki et al., 2018). Our findings in the ACTH model of TRD suggests a unique mechanism that might be specific to this model. SIL-3, SIL-10, IMI-15 + SIL-3, IMI-15 + SIL-10, and ESC-15 + SIL-3 significantly reduced hippocampal KYNUA levels (Fig. 6B), whereas SIL-10 and IMI-15 + SIL-3 significantly reduced hippocampal QUINA levels. Notably, SIL-3, IMI-15 + SIL-10, and ESC-15 + SIL-3 induced large effect size reductions in hippocampal QUINA levels (Fig. 6D). Sildenafil-induced elevations in the hippocam pal levels of KYNUA (neuroprotective) therefore do not appear to un derlie its antidepressant-like effects in ACTH-treated rats. Moreover, although reduced QUINA levels, and thus a reduction in its neurotoxic effects, may play a role in the mechanism of action of sildenafil, the failure of sildenafil mono- and combination therapies to consistently reduce QUINA levels in the ACTH model of TRD, while IMI-15 was able to reduce said levels as well as evoke antidepressant effects, suggest that sildenafil does not induce its antidepressant-like effects solely by reducing QUINA. That ACTH significantly raised hippocampal glutathione levels, without effect on glutathione disulfide (Fig. 7A and C), contrasts with oxidative stress and reduced antioxidant, i.e., glutathione, levels asso ciated with MDD (Gawryluk et al., 2011; Freed et al., 2017; Song et al., 2021). However, protective mechanisms may be at play, initially boosting state-dependent antioxidant processes to combat oxidative stress. IMI-15 and ESC-15 did not affect hippocampal glutathione and glutathione disulfide levels (Fig. 7B and D), whereas SIL-10, but not SIL-3, IMI-15 + SIL-3, IMI-15 + SIL-10, and ESC-15 + SIL-3 significantly reduced hippocampal glutathione levels (Fig. 7B). These findings sug gest that the antidepressant-like effects of sildenafil mono- and combi nation therapy cannot be ascribed exclusively to its antioxidant properties, i.e., ability to increase glutathione levels. In summary, the ACTH model has demonstrated treatment resistance to IMI-15 and ESC-15 while the antidepressant-like effects of sildenafil, whether used alone or as an augmentation strategy, appear to be dose- dependent. A higher/optimal dose of sildenafil (10 mg/kg) is needed as monotherapy to overcome antidepressant treatment resistance in this model, while a low dose (3 mg/kg) augments the effects of co- administered monoaminergic antidepressants (IMI-15 and ESC-15). Hippocampal BDNF, serotonin, norepinephrine, KYNUA, QUINA, and GSH levels were significantly elevated in ACTH-treated animals compared to controls, but plasma corticosterone remained unchanged in the model. It may be that the sustained effects on BDNF, serotonin, norepinephrine, KYNUA, QUINA, and glutathione levels were initially induced and sustained by elevated corticosterone early on in ACTH treatment. The effects of corticosterone eventually dissipated, repre senting an immediate early-gene-related event that switches on genes whose post-transcriptional effects are felt for a prolonged period even though corticosterone has long since been down-regulated. IMI-15 and ESC-15, as well as SIL-3 and SIL-10, and IMI-15 + SIL-3, IMI-15 + SIL- 10, and ESC-15 + SIL-3, reversed some of these changes. Concluding, sildenafil is therefore a typical example of a multimodal antidepressant, although further studies are needed to unravel the exact mechanisms of its promising antidepressant and augmentative effects. 5. Ethical standards All efforts were made to minimise animal suffering during this study. Ethical approval was given as described under materials and methods above. Housing conditions and all experiments and procedures per formed as part of this study complied with the institutional policies and other guidelines on the care and use of laboratory animals and national legislation. The guidelines of the South African National Standards: The Care and Use of Animals for Scientific Purposes (SANS 10386:2008) also informed how animals were kept and the way experiments and J.L. Bernardus Saayman et al. European Journal of Pharmacology 969 (2024) 176434 13 procedures were carried out, as well as the Ethics in Health Research: Principles, Processes and Structures guidelines of 2015. Smith et al. (2018) have recently developed a set of planning guidelines to improve the conduction, reporting, and appraisal of animal research, referred to as Planning Research and Experimental Procedures on Animals: Recommendations for Excellence (PREPARE) guidelines (Smith et al., 2018). As such, the PREPARE guidelines were carefully considered and implemented during the planning phase of this study to enhance the quality, reproducibility, and translatability of experimental results. All experimental data are reported according to the National Centre for the Replacement, Refinement, and Reduction of Animals in Research’s (NC3Rs) Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines to promote a transparent, reproducible, compre hensive, accurate, concise, logically ordered, and well-written manu script (Kilkenny et al., 2010). Financial support This research was made possible by funding from the Suid-Afrikaanse Akademie vir Wetenskap en Kuns (SAAWK), and the Jaschafoundation (grant 2022–0366). 6. Statement of interest GW reported having received research support/lecture/consultancy fees from H. Lundbeck A/S, Eli Lilly A/S, Takeda/Shire A/S, HB Pharma A/S, Arla Foods Amba., Janssen Pharma A/S and Mundipharma Inter national, Ltd. BHH and CBB declare that, except for income from the primary employer (NWU) and research funding to JLBS (SAAWK; Jascha foundation), these organisations did not have any vested interest in the study. BHH has participated in advisory boards and received honoraria from Adcock-Ingram, Servier and Lundbeck, and has received research funding from Servier, Lundbeck, and HG&H Pharma. No other financial support or compensation has been received from any individual or corporate entity over the past three years for research or professional services, and there are no personal financial holdings that could be perceived as constituting a potential conflict of interest. CRediT authorship contribution statement Juandré Lambertus Bernardus Saayman: Writing – original draft, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation. Brian Herbert Harvey: Writing – review & editing, Valida tion, Supervision, Methodology, Funding acquisition, Conceptualiza tion. Gregers Wegener: Writing – review & editing, Validation, Methodology, Funding acquisition, Data curation. Christiaan Beyers Brink: Writing – review & editing, Visualization, Validation, Supervi sion, Data curation, Conceptualization. Declaration of competing interest Gregers Wegener reported having received research support/lec ture/consultancy fees from H. Lundbeck A/S, Eli Lilly A/S, Takeda/Shire A/S, HB Pharma A/S, Arla Foods Amba., Janssen Pharma A/S and Mundipharma International, Ltd. Data availability Data will be made available on request. Acknowledgements The authors would like to thank Mr Cor Bester, Mr Kobus Venter, and Sr Irene Serage for overseeing the welfare of the animals, as well as Dr Francois P Viljoen and Mr Walter Dreyer for assisting with the biochemical analyses that were conducted during this study. References Abel, E.L., Bilitzke, P., 1990. A possible alarm substance in the forced swimming test. Physiol. Behav. 48 (2), 233–239. Al-Harbi, K.S., 2012. Treatment-resistant depression: therapeutic trends, challenges, and future directions. Patient Prefer. Adherence 6, 369. American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders: DSM-5, vol. 5. American psychiatric association Washington, DC. Anacker, C., Zunszain, P.A., Carvalho, L.A., Pariante, C.M., 2011. The glucocorticoid receptor: pivot of depression and of antidepressant treatment? Psychoneuroendocrinology 36 (3), 415–425. Antunes, M.S., Ruff, J.R., de Oliveira Espinosa, D., Piegas, M.B., de Brito, M.L.O., Rocha, K.A., de Gomes, M.G., Goes, A.T.R., Souza, L.C., Donato, F., 2015. Neuropeptide Y administration reverses tricyclic antidepressant treatment-resistant depression induced by ACTH in mice. Horm. Behav. 73, 56–63. Baek, S.-B., Bahn, G., Moon, S.-J., Lee, J., Kim, K.-H., Ko, I.-G., Kim, S.-E., Sung, Y.-H., Kim, B.-K., Kim, T.-S., 2011. The phosphodiesterase type-5 inhibitor, tadalafil, improves depressive symptoms, ameliorates memory impairment, as well as suppresses apoptosis and enhances cell proliferation in the hippocampus of maternal-separated rat pups. Neurosci. Lett. 488 (1), 26–30. Bakunina, N., Pariante, C.M., Zunszain, P.A., 2015. Immune mechanisms linked to depression via oxidative stress and neuroprogression. Immunology 144 (3), 365–373. Banach, M., Piskorska, B., J Czuczwar, S., K Borowicz, K., 2011. Nitric oxide, epileptic seizures, and action of antiepileptic drugs. CNS Neurol. Disord. - Drug Targets 10 (7), 808–819. Bano, S., Gitay, M., Ara, I., Badawy, A., 2010. Acute effects of serotonergic antidepressants on tryptophan metabolism and corticosterone levels in rats. Pak. J. Pharm. Sci. 23. Barden, N., 2004. Implication of the hypothalamic–pituitary–adrenal axis in the physiopathology of depression. J. Psychiatr. Neurosci. 29 (3), 185–193. Baumeister, D., Lightman, S.L., Pariante, C.M., 2016. The HPA axis in the pathogenesis and treatment of depressive disorders, Integrating clinical and molecular findings. Psychopathology Review 3 (1), 64–76. Bergfeld, I.O., Mantione, M., Figee, M., Schuurman, P.R., Lok, A., Denys, D., 2018. Treatment-resistant depression and suicidality. J. Affect. Disord. 235, 362–367. Boku, S., Nakagawa, S., Toda, H., Hishimoto, A., 2018. Neural basis of major depressive disorder: beyond monoamine hypothesis. Psychiatr. Clin. Neurosci. 72 (1), 3–12. Borsini, F., Meli, A., 1988. Is the forced swimming test a suitable model for revealing antidepressant activity? Psychopharmacology 94 (2), 147–160. Boyle, M.P., Brewer, J.A., Funatsu, M., Wozniak, D.F., Tsien, J.Z., Izumi, Y., Muglia, L.J., 2005. Acquired deficit of forebrain glucocorticoid receptor produces depression-like changes in adrenal axis regulation and behavior. Proc. Natl. Acad. Sci. USA 102 (2), 473–478. Brand, S., Moller, M., H Harvey, B., 2015. A review of biomarkers in mood and psychotic disorders: a dissection of clinical vs. preclinical correlates. Curr. Neuropharmacol. 13 (3), 324–368. Brand, S.J., Harvey, B.H., 2017. Exploring a post-traumatic stress disorder paradigm in Flinders sensitive line rats to model treatment-resistant depression I: bio-behavioural validation and response to imipramine. Acta Neuropsychiatr. 29 (4), 193–206. Brink, C., Clapton, J., Eagar, B., Harvey, B., 2008. Appearance of antidepressant-like effect by sildenafil in rats after central muscarinic receptor blockade: evidence from behavioural and neuro-receptor studies. J. Neural. Transm. 115, 117–125. Brown, S.J., Huang, X.-F., Newell, K.A., 2021. The kynurenine pathway in major depression: what we know and where to next. Neurosci. Biobehav. Rev. 127, 917–927. Calabrese, V., Boyd-Kimball, D., Scapagnini, G., Butterfield, D.A., 2004. Nitric oxide and cellular stress response in brain aging and neurodegenerative disorders, the role of vitagenes. In Vivo 18 (3), 245–268. Caraci, F., Calabrese, F., Molteni, R., Bartova, L., Dold, M., Leggio, G.M., Fabbri, C., Mendlewicz, J., Racagni, G., Kasper, S., 2018. International union of basic and clinical pharmacology CIV, the neurobiology of treatment-resistant depression, from antidepressant classifications to novel pharmacological targets. Pharmacol. Rev. 70 (3), 475–504. Carroll, B., Cassidy, F., Naftolowitz, D., Tatham, N., Wilson, W., Iranmanesh, A., Liu, P., Veldhuis, J.D., 2007. Pathophysiology of hypercortisolism in depression. Acta Psychiatr. Scand. 115, 90–103. Carvalho, A.F., Sharma, M.S., Brunoni, A.R., Vieta, E., Fava, G.A., 2016. The safety, tolerability and risks associated with the use of newer generation antidepressant drugs: a critical review of the literature. Psychother. Psychosom. 85 (5), 270–288. Cleveland Clinic, 2023. Treatment-Resistant Depression. Overview. What are the Symptoms of Treatment-Resistant Depression? https://my.clevelandclinic.org/h ealth/diseases/24991-treatment-resistant-depression.Dateofaccess, 2023-08-1.. Cohen, J., 1988. Statistical Power Analysis for the Behavioral Sciences, second ed. Lawrence Erlbaum Associates, Hillsdale. NJ, USA. Curatola, A.M., Xu, J., Hendricks-Munoz, K.D., 2011. Cyclic GMP protects endothelial progenitors from oxidative stress. Angiogenesis 14, 267–279. Dhir, A., Kulkarni, S., 2007. Effect of addition of yohimbine (alpha-2-receptor antagonist) to the antidepressant activity of fluoxetine or venlafaxine in the mouse forced swim test. Pharmacology 80 (4), 239–243. Duarte-Silva, E., Chaves Filho, A.J.M., Barichello, T., Quevedo, J., Macedo, D., Peixoto, C., 2020. Phosphodiesterase-5 inhibitors: shedding new light on the darkness of depression? J. Affect. Disord. 264, 138–149. J.L. Bernardus Saayman et al. http://refhub.elsevier.com/S0014-2999(24)00122-5/sref1 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref1 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref2 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref2 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref3 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref3 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref4 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref4 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref4 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref5 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref5 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref5 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref5 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref6 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref6 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref6 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref6 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref6 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref7 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref7 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref7 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref8 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref8 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref8 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref9 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref9 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref9 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref10 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref10 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref11 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref11 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref11 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref12 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref12 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref13 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref13 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref14 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref14 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref15 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref15 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref15 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref15 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref16 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref16 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref16 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref17 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref17 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref17 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref18 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref18 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref18 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref19 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref19 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref19 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref20 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref20 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref20 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref21 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref21 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref21 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref21 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref21 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref22 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref22 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref22 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref23 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref23 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref23 https://my.clevelandclinic.org/health/diseases/24991-treatment-resistant-depression.Dateofaccess https://my.clevelandclinic.org/health/diseases/24991-treatment-resistant-depression.Dateofaccess http://refhub.elsevier.com/S0014-2999(24)00122-5/sref25 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref25 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref26 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref26 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref27 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref27 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref27 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref28 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref28 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref28 European Journal of Pharmacology 969 (2024) 176434 14 Duman, R.S., Aghajanian, G.K., Sanacora, G., Krystal, J.H., 2016. Synaptic plasticity and depression: new insights from stress and rapid-acting antidepressants. Nat. Med. 22 (3), 238. Duman, R.S., Monteggia, L.M., 2006. A neurotrophic model for stress-related mood disorders. Biol. Psychiatr. 59 (12), 1116–1127. Elfving, B., Plougmann, P.H., Müller, H.K., Mathé, A.A., Rosenberg, R., Wegener, G., 2010. Inverse correlation of brain and blood BDNF levels in a genetic rat model of depression. Int. J. Neuropsychopharmacol. 13 (5), 563–572. Farahbakhsh, Z., Radahmadi, M., 2022. The protective effects of escitalopram on synaptic plasticity in the CA1 region of chronically stressed and non-stressed male rats. Int. J. Dev. Neurosci. 82 (8), 747–757. Feil, S., Zimmermann, P., Knorn, A., Brummer, S., Schlossmann, J., Hofmann, F., Feil, R., 2005. Distribution of cGMP-dependent protein kinase type I and its isoforms in the mouse brain and retina. Neuroscience 135 (3), 863–868. Freed, R.D., Hollenhorst, C.N., Weiduschat, N., Mao, X., Kang, G., Shungu, D.C., Gabbay, V., 2017. A pilot study of cortical glutathione in youth with depression. Psychiatr. Res. Neuroimaging 270, 54–60. Friedrich, M., 2017. Depression is the leading cause of disability around the world. JAMA 317 (15), 1517-1517. Frost, P., Bornstein, S., Ehrhart-Bornstein, M., O’Kirwan, F., Hutson, C., Heber, D., Go, V., Licinio, J., Wong, M.-L., 2003. The prototypic antidepressant drug, imipramine, but not Hypericum perforatum (St. John’s Wort), reduces HPA-axis function in the rat. Horm. Metab. Res. 35 (10), 602–606. Fuertig, R., Ceci, A., Camus, S.M., Bezard, E., Luippold, A.H., Hengerer, B., 2016. LC–MS/ MS-based quantification of kynurenine metabolites, tryptophan, monoamines and neopterin in plasma, cerebrospinal fluid and brain. Bioanalysis 8 (18), 1903–1917. Gallo, E.F., Iadecola, C., 2011. Neuronal nitric oxide contributes to neuroplasticity- associated protein expression through cGMP, protein kinase G, and extracellular signal-regulated kinase. J. Neurosci. 31 (19), 6947–6955. Gawryluk, J.W., Wang, J.-F., Andreazza, A.C., Shao, L., Young, L.T., 2011. Decreased levels of glutathione, the major brain antioxidant, in post-mortem prefrontal cortex from patients with psychiatric disorders. Int. J. Neuropsychopharmacol. 14 (1), 123–130. Gong, M.-j., Han, B., Wang, S.-m., Liang, S.-w., Zou, Z.-j., 2016. Icariin reverses corticosterone-induced depression-like behavior, decrease in hippocampal brain- derived neurotrophic factor (BDNF) and metabolic network disturbances revealed by NMR-based metabonomics in rats. J. Pharmaceut. Biomed. Anal. 123, 63–73. Gregory, E., Torres, I.J., Ge, R., Blumberger, D.M., Downar, J.H., Daskalakis, Z.J., Lam, R.W., Vila-Rodriguez, F., 2020. Predictors of cognitive impairment in treatment-resistant depression. J. Affect. Disord. 274, 593–601. Guan, X.-t., Shao, F., Xie, X., Chen, L., Wang, W., 2014. Effects of aspirin on immobile behavior and endocrine and immune changes in the forced swimming test: comparison to fluoxetine and imipramine. Pharmacol. Biochem. Behav. 124, 361–366. Halaris, A., Myint, A.-M., Savant, V., Meresh, E., Lim, E., Guillemin, G., Hoppensteadt, D., Fareed, J., Sinacore, J., 2015. Does escitalopram reduce neurotoxicity in major depression? J. Psychiatr. Res. 66, 118–126. Harvey, B.H., 1996. Affective disorders and nitric oxide: a role in pathways to relapse and refractoriness? Human Psychopharmacology. Clinical and Experimental 11 (4), 309–319. Harvey, B.H., Brand, L., Jeeva, Z., Stein, D.J., 2006a. Cortical/hippocampal monoamines, HPA-axis changes and aversive behavior following stress and restress in an animal model of post-traumatic stress disorder. Physiol. Behav. 87 (5), 881–890. Harvey, B.H., Hamer, M., Louw, R., Van Der Westhuizen, F.H., Malan, L., 2012. Metabolic and glutathione redox markers associated with brain-derived neurotrophic factor in depressed African men and women: evidence for counterregulation? Neuropsychobiology 67 (1), 33–40. Harvey, B.H., Jonker, L.P., Brand, L., Heenop, M., Stein, D.J., 2002. NMDA receptor involvement in imipramine withdrawal-associated effects on swim stress, GABA levels and NMDA receptor binding in rat hippocampus. Life Sci. 71 (1), 43–54. Harvey, B.H., Retief, R., Korff, A., Wegener, G., 2006b. Increased hippocampal nitric oxide synthase activity and stress responsiveness after imipramine discontinuation: role of 5HT 2A/C-receptors. Metab. Brain Dis. 21, 201–210. Hawton, K., i Comabella, C.C., Haw, C., Saunders, K., 2013. Risk factors for suicide in individuals with depression: a systematic review. J. Affect. Disord. 147 (1–3), 17–28. Hestad, K., Alexander, J., Rootwelt, H., Aaseth, J.O., 2022. The role of tryptophan dysmetabolism and quinolinic acid in depressive and neurodegenerative diseases. Biomolecules 12 (7), 998. Hiroi, R., Neumaier, J.F., 2006. Differential effects of ovarian steroids on anxiety versus fear as measured by open field test and fear-potentiated startle. Behav. Brain Res. 166 (1), 93–100. Hornig-Rohan, M., Wolkowitz, O.M., Amsterdam, J.D., 1996. Novel strategies for treatment-resistant depression. Psychiatr. Clin. 19 (2), 387–405. Huang, Z., Zhong, X.-M., Li, Z.-Y., Feng, C.-R., Pan, A.-J., Mao, Q.-Q., 2011. Curcumin reverses corticosterone-induced depressive-like behavior and decrease in brain BDNF levels in rats. Neurosci. Lett. 493 (3), 145–148. Jacobsen, J.P., Mørk, A., 2006. Chronic corticosterone decreases brain-derived neurotrophic factor (BDNF) mRNA and protein in the hippocampus, but not in the frontal cortex, of the rat. Brain Res. 1110 (1), 221–225. Juruena, M.F., Pariante, C.M., Papadopoulos, A.S., Poon, L., Lightman, S., Cleare, A.J., 2013. The role of mineralocorticoid receptor function in treatment-resistant depression. J. Psychopharmacol. 27 (12), 1169–1179. Kilkenny, C., Browne, W.J., Cuthill, I.C., Emerson, M., Altman, D.G., 2010. Improving bioscience research reporting: the ARRIVE guidelines for reporting animal research. PLoS Biol. 8 (6), e1000412. Kitamura, Y., Araki, H., Gomita, Y., 2002. Influence of ACTH on the effects of imipramine, desipramine and lithium on duration of immobility of rats in the forced swim test. Pharmacol. Biochem. Behav. 71 (1–2), 63–69. Kitamura, Y., Fujitani, Y., Kitagawa, K., Miyazaki, T., Sagara, H., Kawasaki, H., Shibata, K., Sendo, T., Gomita, Y., 2008. Effects of imipramine and bupropion on the duration of immobility of ACTH-treated rats in the forced swim test: involvement of the expression of 5-HT2A receptor mRNA. Biol. Pharm. Bull. 31 (2), 246–249. Kocki, T., Urbańska, E.M., Kocki, J., Kloc, R., Kocka, K., Olajossy, M., Owe-Larsson, B., 2018. Prolonged therapy with antidepressants increases hippocampal level of kynurenic acid and expression of Kat1 and Kat2 genes. Pharmacol. Rep. 70 (4), 737–745. Lee, B.-H., Kim, Y.-K., 2010. The roles of BDNF in the pathophysiology of major depression and in antidepressant treatment. Psychiatry investigation 7 (4), 231. Lex, H., Ginsburg, Y., Sitzmann, A.F., Grayhack, C., Maixner, D.F., Mickey, B.J., 2019. Quality of life across domains among individuals with treatment-resistant depression. J. Affect. Disord. 243, 401–407. Li, B., Suemaru, K., Kitamura, Y., Cui, R., Gomita, Y., Araki, H., 2007. Strategy to develop a new drug for treatment-resistant depression–role of electroconvulsive stimuli and BDNF. Yakugaku Zasshi: J. Pharm. Soc. Jpn. 127 (4), 735–742. Li, C.-T., 2023. Overview of treatment-resistant depression. Prog. Brain Res. 278, 1–23. Li, H., Liu, X., Poh, Y., Wu, L., Zhou, Q.G., Cai, B.C., 2014. Rapid determination of corticosterone in mouse plasma by ultra fast liquid chromatography-tandem mass spectrometry. Biomed. Chromatogr. 28 (12), 1860–1863. Li, X., Li, H., Zhao, B., 1990. Serotonin of hippocampus and hypothalamus taking part in the analgesic effect of adrenocorticotropic hormone in rats. Zhongguo yao li xue bao= Acta Pharmacologica Sinica 11 (1), 89–92. Liebenberg, N., Harvey, B.H., Brand, L., Brink, C.B., 2010a. Antidepressant-like properties of phosphodiesterase type 5 inhibitors and cholinergic dependency in a genetic rat model of depression. Behav. Pharmacol. 21 (5–6), 540–547. Liebenberg, N., Wegener, G., Harvey, B.H., Brink, C.B., 2010b. Investigating the role of protein kinase-G in the antidepressant-like response of sildenafil in combination with muscarinic acetylcholine receptor antagonism. Behav. Brain Res. 209 (1), 137–141. Malhi, G., Hitching, R., Berk, M., Boyce, P., Porter, R., Fritz, K., 2013. Pharmacological management of unipolar depression. Acta Psychiatr. Scand. 127, 6–23. Markopoulou, K., 2013. HPA axis Dysfunction in Treatment Resistant Affective Disorders. King’s College London (University of London. Markopoulou, K., Papadopoulos, A., Juruena, M.F., Poon, L., Pariante, C.M., Cleare, A.J., 2009. The ratio of cortisol/DHEA in treatment resistant depression. Psychoneuroendocrinology 34 (1), 19–26. Marx, W., McGuinness, A.J., Rocks, T., Ruusunen, A., Cleminson, J., Walker, A.J., Gomes-da-Costa, S., Lane, M., Sanches, M., Diaz, A.P., 2021. The kynurenine pathway in major depressive disorder, bipolar disorder, and schizophrenia: a meta- analysis of 101 studies. Mol. Psychiatr. 26 (8), 4158–4178. Masood, A., Nadeem, A., Mustafa, S.J., O’Donnell, J.M., 2008. Reversal of oxidative stress-induced anxiety by inhibition of phosphodiesterase-2 in mice. J. Pharmacol. Exp. Therapeut. 326 (2), 369–379. Matsushita, H., Matsuzaki, M., Han, X.-J., Nishiki, T.-I., Ohmori, I., Michiue, H., Matsui, H., Tomizawa, K., 2012. Antidepressant-like effect of sildenafil through oxytocin-dependent cyclic AMP response element-binding protein phosphorylation. Neuroscience 200, 13–18. Misslin, R., Cigrang, M., 1986. Does neophobia necessarily imply fear or anxiety? Behav. Process. 12 (1), 45–50. Mokoena, M.L., Harvey, B.H., Oliver, D.W., Brink, C.B., 2010. Ozone modulates the effects of imipramine on immobility in the forced swim test, and nonspecific parameters of hippocampal oxidative stress in the rat. Metab. Brain Dis. 25, 125–133. Mokoena, M.L., Harvey, B.H., Viljoen, F., Ellis, S.M., Brink, C.B., 2015. Ozone exposure of Flinders Sensitive Line rats is a rodent translational model of neurobiological oxidative stress with relevance for depression and antidepressant response. Psychopharmacology 232, 2921–2938. Möller, M., Du Preez, J.L., Viljoen, F.P., Berk, M., Emsley, R., Harvey, B.H., 2013. Social isolation rearing induces mitochondrial, immunological, neurochemical and behavioural deficits in rats, and is reversed by clozapine or N-acetyl cysteine. Brain Behav. Immun. 30, 156–167. Moller, M., Swanepoel, T., Harvey, B., 2015. Neurodevelopmental animal models reveal the convergent role of neurotransmitter systems, inflammation, and oxidative stress as biomarkers of schizophrenia: implications for novel drug development. ACS Chem. Neurosci. 6 (7), 987–1016. Moriarty, M., Lee, A., O’Connell, B., Kelleher, A., Keeley, H., Furey, A., 2011. Development of an LC-MS/MS method for the analysis of serotonin and related compounds in urine and the identification of a potential biomarker for attention deficit hyperactivity/hyperkinetic disorder. Anal. Bioanal. Chem. 401, 2481–2493. National Institute of Mental Health, 2023. Depression. What is depression? https://www. nimh.nih.gov/health/topics/depression. (Accessed 23 September 2023). Nguyen, T.H., Axell, A., Turek, I., Wright, B., Meehan-Andrews, T., Irving, H.R., 2022. Modulation of inflammatory cytokine production in human monocytes by cGMP and IRAK3. Int. J. Mol. Sci. 23 (5), 2552. Niciu, M.J., Ionescu, D.F., Richards, E.M., Zarate, C.A., 2014. Glutamate and its receptors in the pathophysiology and treatment of major depressive disorder. J. Neural. Transm. 121, 907–924. Oberholzer, I., Möller, M., Holland, B., Dean, O.M., Berk, M., Harvey, B.H., 2018. Garcinia mangostana Linn displays antidepressant-like and pro-cognitive effects in a genetic animal model of depression: a bio-behavioral study in the flinders sensitive line rat. Metab. Brain Dis. 33, 467–480. J.L. Bernardus Saayman et al. http://refhub.elsevier.com/S0014-2999(24)00122-5/sref29 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref29 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref29 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref30 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref30 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref31 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref31 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref31 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref32 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref32 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref32 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref33 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref33 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref33 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref34 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref34 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref34 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref35 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref35 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref36 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref36 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref36 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref36 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref37 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref37 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref37 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref38 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref38 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref38 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref39 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref39 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref39 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref39 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref40 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref40 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref40 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref40 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref41 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref41 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref41 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref42 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref42 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref42 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref42 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref43 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref43 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref43 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref44 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref44 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref44 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref45 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref45 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref45 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref45 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref46 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref46 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref46 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref46 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref47 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref47 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref47 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref48 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref48 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref48 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref49 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref49 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref50 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref50 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref50 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref51 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref51 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref51 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref52 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref52 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref53 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref53 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref53 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref54 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref54 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref54 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref55 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref55 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref55 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref56 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref56 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref56 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref57 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref57 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref57 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref58 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref58 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref58 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref58 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref59 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref59 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref59 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref59 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref60 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref60 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref61 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref61 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref61 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref62 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref62 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref62 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref63 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref64 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref64 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref64 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref65 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref65 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref65 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref66 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref66 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref66 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref67 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref67 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref67 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref68 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref68 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref69 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref69 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref70 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref70 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref70 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref71 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref71 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref71 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref71 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref72 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref72 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref72 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref73 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref73 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref73 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref73 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref74 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref74 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref75 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref75 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref75 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref75 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref76 http://refhub.elsevier.com/S0014-2999(24)00122-5/sref76 http://refhub.elsevier.com/S0